Forthcoming Current Anthropology Wenner-Gren Symposium Curren t Supplementary Issues (in order of appearance)

     

e Biological Anthropology of Modern Human Populations: World Anthropolog y Current Histories, National Styles, and International Networks. Mary Susan Lindee and Ricardo Ventura Santos, eds. Human Biology and the Origins of Homo. Susan Antón and Leslie C. Aiello, Anthropology eds.

THE WENNER-GREN SYMPOSIUM SERIES Previously Published Supplementary Issues October THE ORIGINS OF AGRICULTURE: NEW DATA, Engaged Anthropology: Diversity and Dilemmas. Setha M. Low and Sally NEW IDEAS 2

Engle Merry, eds. V 01 1 GUEST EDITORS: Working Memory: Beyond Language and Symbolism. omas Wynn and T. DOUGLAS PRICE AND OFER BAR-YOSEF Frederick L. Coolidge, eds. The Origins of Agriculture 52 olum e Corporate Lives: New Perspectives on the Social Life of the Corporate Form. Climatic Fluctuations and Early Farming in West and East Asia Damani Partridge, Marina Walker, and Rebecca Hardin, eds. Neolithization Processes in the Levant Becoming Farmers The Origins of Agriculture in the Near East The Neolithic Southwest Asian Founder Crops

Supplement The Early Process of Mammal Domestication in the Near East The Beginnings of Agriculture in China New Archaeobotanic Data for Origins of Agriculture in China The Transition from Foraging to Farming in Prehistoric Korea Advances in Understanding Early Agriculture in Japan 4 Finding Plant Domestication in the Indian Subcontinent Holocene Population Prehistory in the Pacific Region Current Anthropology is sponsored by e Early Agriculture and Plant Domestication in Island Southeast Asia Wenner-Gren Foundation for Anthropological Domestication Processes and Morphological Change Research, a foundation endowed for scientific, Westward Expansion of Farming from Anatolia to the Balkans

educational, and charitable purposes. e Page s Foundation, however, is not to be understood as The Spread of Agriculture from Central Europe to the Atlantic endorsing, by virtue of its financial support, any of Origins of Plant Cultivation in the New World Tropics

the statements made, or views expressed, herein. S161–S512 Cultural Context of Plant Domestication in Eastern North America Genetics and Domestication Agricultural Demographic Transition and the Agriculture Inventions

0011-3204(201110)52:5+4;1-1 Sponso r e d b y the W enner-Gren Founda tion f o r Anth r opologic a l Rese a r c h

THE UNIVERSIT Y O F CHICAGO PRESS Wenner-Gren Symposium Series Editor: Leslie Aiello Wenner-Gren Symposium Series Managing Editor: Victoria Malkin Current Anthropology Editor: Mark Aldenderfer Current Anthropology Managing Editor: Lisa McKamy Book Reviews Editor: Holley Moyes Corresponding Editors: Claudia Briones (IIDyPCa-Universidad Nacional de Rı´o Negro, Argentina; [email protected]), Anne de Sales (Centre National de la Recherche Scientifique, France; [email protected]), Michalis Kontopodis (Humboldt Universita¨t zu Berlin, Germany; [email protected]), Jose´ Luis Lanata (Universidad Nacional de Rı´o Negro San Carlos de Bariloche, Argentina; [email protected]), David Palmer (Hong Kong University, China; [email protected]), Zhang Yinong (Shanghai University, China; [email protected])

Please send all editorial correspondence to Reasons of practicality or law make it necessary or desirable Mark Aldenderfer to circulate Current Anthropology without charge in certain School of Social Sciences, Humanities, and Arts portions of the world; it is hoped, however, that recipients of University of California, Merced this journal without charge will individually or collectively in 5200 North Lake Road various groups apply funds or time and energy to the world Merced, CA 95343, U.S.A. good of humankind through the human sciences. Information (fax: 209-228-4007; e-mail: [email protected]) concerning applicable countries is available on request.

Individual subscription rates for 2012: $68 print ϩ elec- ᭧ 2011 by The Wenner-Gren Foundation for Anthropological tronic, $40 print-only, $39 e-only. Institutional print ϩ elec- Research. All rights reserved. Current Anthropology (issn tronic and e-only subscriptions are available through JSTOR’s 0011-3204) is published bimonthly in February, April, June, Current Scholarship Program and include unlimited online August, October, and December by The University of Chicago access; rates are tiered according to an institution’s type and Press, 1427 East 60th Street, Chicago, IL 60637-2954. research output: $287 to $575 (print ϩ electronic), $243 to Periodicals postage paid at Chicago, IL, and at additional $4869 (e-only). Institutional print-only is $288. For additional mailing offices. Postmaster: Send address changes to rates, including single copy rates and print-only or electronic- Current Anthropology, P.O. Box 37005, Chicago, IL 60637. only subscriptions, please visit www.journals.uchicago.edu/ CA. Additional taxes and/or postage for non-U.S. subscrip- tions may apply. Free or deeply discounted access is available to readers in most developing nations through the Chicago Emerging Nations Initiative (www.journals.uchicago.edu/ ceni/).

Please direct subscription inquiries, back-issue requests, and address changes to the University of Chicago Press, Jour- nals Division, P.O. Box 37005, Chicago, IL 60637. Telephone: (773) 753-3347 or toll-free in the United States and Canada (877) 705-1878. Fax: (773) 753-0811 or toll-free (877) 705- 1879. E-mail: [email protected] Current Anthropology Volume 52 Supplement 4 October 2011

The Origins of Agriculture: New Data, New Ideas

Leslie C. Aiello The Origins of Agriculture: New Data, New Ideas S161 T. Douglas Price and Ofer Bar-Yosef The Origins of Agriculture: New Data, New Ideas: An Introduction to Supplement 4 S163 Ofer Bar-Yosef Climatic Fluctuations and Early Farming in West and East Asia S175 A. Nigel Goring-Morris and Anna Belfer-Cohen Neolithization Processes in the Levant: The Outer Envelope S195 Anna Belfer-Cohen and A. Nigel Goring-Morris Becoming Farmers: The Inside Story S209 Melinda A. Zeder The Origins of Agriculture in the Near East S221 Ehud Weiss and Daniel Zohary The Neolithic Southwest Asian Founder Crops: Their Biology and Archaeobotany S237 Jean-Denis Vigne, Isabelle Carre`re, Franc¸ois Briois, and Jean Guilaine The Early Process of Mammal Domestication in the Near East: New Evidence from the Pre-Neolithic and Pre- Pottery Neolithic in Cyprus S255 David Joel Cohen The Beginnings of Agriculture in China: A Multiregional View S273 Zhijun Zhao New Archaeobotanic Data for the Study of the Origins of Agriculture in China S295 Gyoung-Ah Lee The Transition from Foraging to Farming in Prehistoric Korea S307 Gary W. Crawford Advances in Understanding Early Agriculture in Japan S331 http://www.journals.uchicago.edu/CA Dorian Q Fuller Finding Plant Domestication in the Indian Subcontinent S347 Peter Bellwood Holocene Population History in the Pacific Region as a Model for Worldwide Food Producer Dispersals S363 Tim Denham Early Agriculture and Plant Domestication in New Guinea and Island Southeast Asia S379 Fiona Marshall and Lior Weissbrod Domestication Processes and Morphological Change: Through the Lens of the Donkey and African Pastoralism S397 Mehmet O¨ zdog˘an Archaeological Evidence on the Westward Expansion of Farming Communities from Eastern Anatolia to the Aegean and the Balkans S415 Peter Rowley-Conwy Westward Ho! The Spread of Agriculturalism from Central Europe to the Atlantic S431 Dolores R. Piperno The Origins of Plant Cultivation and Domestication in the New World Tropics: Patterns, Process, and New Developments S453 Bruce D. Smith The Cultural Context of Plant Domestication in Eastern North America S471 Greger Larson Genetics and Domestication: Important Questions for New Answers S485 Jean-Pierre Bocquet-Appel The Agricultural Demographic Transition During and After the Agriculture Inventions S497 Erratum S511 Current Anthropology Volume 52, Supplement 4, October 2011 S161

The Origins of Agriculture: New Data, New Ideas Wenner-Gren Symposium Supplement 4

by Leslie C. Aiello

The Origins of Agriculture: New Data, New Ideas resulted from terest in the origins of agriculture and domestication. One of a Wenner-Gren-sponsored symposium held at the Hacienda the earliest meetings organized by the Foundation in July 1960 Temozon, Merida, Yucatan, Mexico, March 6–13, 2009 (fig. led to the seminal publication Courses toward Urban Life: 1). The symposium was organized by T. Douglas Price (Uni- Archaeological Considerations of Some Cultural Alternates versity of Wisconsin–Madison and the University of Aber- (Braidwood and Willey 1962). Other influential meetings in- deen) and Ofer Bar-Yosef (Harvard University). cluded the Origins of African Plant Domestication (Harlan, De The major aim of the symposium was to better understand Wet, and Stemler 1972) and Where the Wild Things Are Now the origins of agriculture in light of new fieldwork, new sites, (Mullin and Cassidy 2007), which invited anthropologists new analytical techniques, and more radiocarbon dates. The from all subfields to rethink the concept of domestication in global nature of agricultural origins was a key theme, and a anthropology. Information on these meetings and others can major focus of the discussions was on East Asia as well as be found on our Web site at http://wennergren.org/history. lesser-known regions such as Papua New Guinea, Africa, and Most recently, agricultural origins were explored in a special eastern North America, alongside more traditional areas such issue of Current Anthropology titled Rethinking the Origins of as the Near East and Mesoamerica. The papers presented in Agriculture introduced by Mark Cohen (Cohen 2009). The this supplementary issue are designed to provide the latest current supplementary issue continues the discussions and information on the antiquity of agriculture covering at least debates explored in this earlier contribution but is perhaps 10 different centers of domestication. more data rich and geographically diverse. Together these two The organizers, Price and Bar-Yosef, note in their intro- CA issues provide an excellent contemporary overview of the duction that emerging data point to an unexpected synchron- state of research in this exciting area of inquiry. icity in the timing of the first domesticates around the end The Wenner-Gren Foundation is always looking for in- of the Pleistocene. They also note that, contrary to earlier novative new directions in the field for future Foundation- thought, the environments in which agriculture originated sponsored and organized symposia and eventual CA publi- were not marginal and that agricultural experimentation took cation. We encourage anthropologists to contact us with their place in areas of concentrations of populations and resources. ideas for future meetings. Information about the Wenner- Each major area may also have included multiple loci for Gren Foundation and the Symposium program can be found domestication. These were major areas of agreement in a on the Foundation’s Web site (http://wennergren.org/ meeting that was characterized by lively debate over the va- programs/international-symposia). riety of hypotheses proposed for agricultural origins and whether global or more area-specific explanations were most appropriate. As in any good meeting, there were more ques- References Cited tions than answers, but this is the sign of a dynamic field. The degree of collegiality and collaboration among the diverse Braidwood, Robert John, and Gordon Randolph Willey, eds. 1962. Courses toward urban life: archaeological considerations of some cul- symposium participants and the speed at which new data are tural alternates. Viking Fund Publications in Anthropology, no. 32 accumulating are good signs that our understanding of this (Wenner-Gren Foundation for Anthropological Research). Chi- important period in human adaptation will continue to evolve cago: Aldine. rapidly. Cohen, Mark Nathan. 2009. Introduction: rethinking the origins of The Wenner-Gren Foundation has had a long-standing in- agriculture. Current Anthropology 50:591–595. Harlan, Jack R., Jan M. J. De Wet, and Ann B. L. Stemler, eds. 1976. Origins of African plant domestication. World Anthropology Series. The Hague: Mouton. Leslie C. Aiello is President of the Wenner-Gren Foundation for Mullin, Molly, and Rebecca Cassidy, eds. 2007. Where the wild things Anthropological Research (470 Park Avenue South, 8th Floor North, are now. Wenner-Gren International Symposium Series. Oxford: New York, New York 10016, U.S.A.). Berg.

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0001$10.00. DOI: 10.1086/660154 Figure 1. Participants in the symposium “The Origins of Agriculture: New Data, New Ideas.” Front row from left: Laurie Obbink (Wenner- Gren staff), Carolyn Freiwald (monitor), Leslie Aiello, Fiona Marshall, Ofer Bar-Yosef, Gyoung-Ah Lee, Ehud Weiss, Anna Belfer-Cohen. Middle row from left: Tim Denham, Peter Bellwood, Melinda A. Zeder, Dolores R. Piperno, Greger Larson, Richard H. Meadow, Jean-Denis Vigne, Meh- met O¨ zdog˘an, Peter Rowley-Conwy. Back row from left: David Joel Co- hen, Zhao Zhijun (“Jimmy”), Dorian Q Fuller, Bruce D. Smith, Gary W. Crawford, T. Douglas Price, Jean-Pierre Bocquet-Appel, A. Nigel Goring- Morris. A color version of this photo appears in the online edition of Current Anthropology. Current Anthropology Volume 52, Supplement 4, October 2011 S163

The Origins of Agriculture: New Data, New Ideas An Introduction to Supplement 4

by T. Douglas Price and Ofer Bar-Yosef

This introduction to the symposium and to this issue of Current Anthropology attempts to provide some sense of the topic, the meeting itself, the participants, and some of the initial results. Our symposium brought together a diverse international group of archaeological scientists to consider a topic of common interest and substantial anthropological import—the origins of agriculture. The group included individuals working in most of the places where farming began. This issue is organized by chronology and geography. Our goal was to consider the most recent data and ideas from these different regions in order to examine larger questions of congruity and disparity among the groups of first farmers. There is much new information from a number of important areas, particularly Asia. Following a review of the history of investigation of agricultural origins, this introduction summarizes the results of the conference. There are at least 10 different places around the world where agriculture was independently developed, and the antiquity of domestication is being pushed back in time with new discoveries. Our symposium has emphasized the importance of a multidis- ciplinary approach to such large questions in order to assemble as much information as possible. We anticipate that the results and consequences of this symposium will have long-term ripple effects in anthropology and archaeology.

In the middle of March 2009, we were part of a group of 22 The Symposium and the Participants individuals that included archaeologists, archaeobotanists, ar- chaeozoologists, a geneticist, and a physical anthropologist We, the coorganizers of this symposium, first met in 1968 at that took a 6-day trip to a lovely hacienda near Merida in the Paleolithic site of ‘Ubeidiya in the rift valley of Israel. It the Yucatan of Mexico. We were well cared for and well fed, was about 120ЊF in the shade. Doug was a graduate student but it was an intense and demanding journey and at the same visiting from another excavation. Ofer was digging in the hot time one of those rare opportunities for like-minded indi- sun and seemed to enjoy it a lot. We next spent a week viduals (we use that term loosely) to get together and explore together at a School for American Research Seminar in Santa a subject of shared interest, even fascination. We spent those Fe in 1993. We had a small, fine meeting on the earliest days trying to better understand the origins of agriculture. farming. We have been talking ever since, and a few years ago The people were passionate, the ideas powerful, the infor- we decided that there was so much new information on ag- mation thought provoking, and it is small meetings like this riculture that another meeting was needed, and we ap- that are the ones that we remember, that leave an imprint, proached the Wenner-Gren Foundation. and that generate messages and ideas that form and transform Wenner-Gren has organized more than 140 symposia, so our views of the past. they have a pretty good feel for what works and for how to run a successful academic meeting. However, one of the com- plexities of small, intensive workshop meetings is that each T. Douglas Price is Weinstein Professor of European Archaeology, participant arrives with his or her own set of perspectives, Department of Anthropology, University of Wisconsin–Madison experiences, and biases that provide insight but that can also (5240 W. H. Sewell Social Science Building, 1180 Observatory Drive, act as blinders. The insight is of course important, but the Madison, Wisconsin 53706, U.S.A. [[email protected]]). Ofer Bar- Yosef is MacCurdy Professor of Prehistoric Archaeology, Department blinders can impede broader synthesis. We observe only the of Anthropology, Harvard University (Cambridge, Massachusetts world we know. For example, Doug does archaeology in 02138, U.S.A.). This paper was submitted 13 XI 09, accepted 1 III northern Europe, and from that perspective he witnesses only 11, and electronically published 19 VIII 11. the arrival of farming, not its origins. But the larger questions

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0002$10.00. DOI: 10.1086/659964 S164 Current Anthropology Volume 52, Supplement 4, October 2011 are often the same. Why do hunters become farmers? Why terests are concerned with ancient DNA and the evidence for is agriculture so dynamic in changing human adaptation and domestication. Gyoung-Ah Lee is an assistant professor at the behavior? University of Oregon. Trained as an archaeobotanist, her in- Our symposium brought a number of biases: there was an terests focus on the origins of agriculture and its impact in Old World bias, an East Asia bias, a plant bias, a male bias, East Asia, including Korea, China, and Japan. and an age bias; someone even suggested an anticamel bias. Fiona Marshall is a professor at Washington University in Specific individual knowledge and bias were critical for our St. Louis, where she directs research in African prehistory and discussions. At the same time, most of the participants did zooarchaeology. Her interests are in domestication, pastor- manage to put aside their blinders and take a wider point of alism, climate change, and mobility. Mehmet O¨ zdog˘anisan view. archaeologist and chair of the prehistory department at Is- The biases at our symposium can be documented in part tanbul University. His research for the past 2 decades has by a brief glimpse at the background of the participants. There focused on the emergence of early food-producing sedentary were 22 scientists invited to the symposium—nine archae- communities and the spread of agriculture from Anatolia to ologists, six archaeobotanists, five archaeozoologists, a phys- Europe. ical anthropologist/demographer, and an archaeological ge- Dolores R. Piperno is Senior Scientist and Curator of Ar- neticist. These individuals came from eight countries and chaeobotany and South American Archaeology at the Na- work in many major areas of agricultural origins and/or tional Museum of Natural History in Washington, DC, and spread. A brief biography of each participant (in alphabetical the Smithsonian Tropical Research Institute, Balboa, Panama. order) may provide some sense of the perspectives at the Her interests are in domestication and tropical archaeology symposium. Coauthors who did not attend the meeting are not included. and paleoecology. T. Douglas Price is Weinstein Professor of Ofer Bar-Yosef is an archaeologist and MacCurdy Professor European Archaeology at the University of Wisconsin–Mad- of Prehistoric Archaeology at Harvard University. Among ison and 6th Century Chair in Archaeological Science at the many other interests, he has investigated the origins of ag- University of Aberdeen. His major interests involve the tran- riculture in both Southwest Asia and China. Anna Belfer- sition from hunters to farmers in northern Europe and ar- Cohen is professor of archaeology at the Hebrew University chaeological chemistry. He is director of the Laboratory for in Jerusalem with interests in the transition to farming in the Archaeological Chemistry in Madison. Peter Rowley-Conwy Near East. Peter Bellwood is an Australian archaeologist and is a professor at Durham University, United Kingdom, and professor at Australian National University. His interests focus an archaeozoologist with interests in the spread of pigs and on Southeast Asia and the Pacific, and he is a proponent of agriculture across Europe. Bruce D. Smith is Curator of North the connectedness of culture, linguistics, and human biology. American Archaeology and Senior Research Scientist in Ar- Jean-Pierre Bocquet-Appel is an anthropological demogra- chaeobiology at the National Museum of Natural History in pher and research director at Centre National de la Recherche Washington, DC. He is currently interested in integrating bi- Scientifique in Paris. His current interest concerns the agri- ological and anthropological approaches to documenting culture demographic transition worldwide. plant and animal domestication. David Joel Cohen is an archaeologist and adjunct assistant Jean-Denis Vigne is an archaeozoologist with interests in professor at Boston University, where he helped establish the mammal domestication in Eurasia. He is employed by Centre International Center for East Asian Archaeology and Cultural National de la Recherche Scientifique, Paris, and is in charge History. His current research focuses on the origins of agri- of the laboratory of archaeozoology and archaeobotany at the culture in East Asia. Gary W. Crawford is an archaeobotanist French National Museum of Natural History. Ehud Weiss is and professor at the University of Toronto Mississauga, Can- a senior lecturer and the director of the archaeobotanical ada, and is interested in the origins and intensification of laboratory at Bar-Ilan University and the Weisman Institute agriculture in China, Japan, Korea, and Eastern North Amer- of Science in Israel. He is interested in the beginning of plant ica. Tim Denham is a research fellow at Monash University domestication in the Near East through the prism of archaeo- in Victoria, Australia. Over the past decade his research has focused on the emergence of agriculture in Papua New botany. Melinda A. Zeder is Senior Research Scientist and Guinea. Curator of Old World Archaeology and Zooarchaeology at Dorian Q Fuller is a lecturer in archaeobotany at the In- the National Museum of Natural History in Washington, DC. stitute of Archaeology, University College London. His re- Her interests concern the domestication of animals in general search focuses on the origins of agriculture in South and East and the social and environmental implication of early agri- Asia. A. Nigel Goring-Morris is an archaeologist at Hebrew culture in the ancient Near East more specifically. Zhijun University in Jerusalem. His interests are in the beginnings (Jimmy) Zhao is a staff scientist at the Institute of Archae- of complexity in the Near East and Neolithization processes. ology, Chinese Academy of Social Sciences, Beijing. His re- Greger Larson is a research fellow at Durham University in search is focused on the origins of agriculture in China and the United Kingdom. As an archaeological geneticist, his in- East Asia using archaeobotanical evidence. Price and Bar-Yosef The Origins of Agriculture S165 Definitions and Dates Dating Also for purposes of consistency, the authors in this issue have been asked to report radiocarbon dates in terms of cal- SD, unless 1ע Definitions ibrated years before present (BP cal) with otherwise noted. Authors were also asked to note the cali- To enhance consistency in our discussions we have tried to bration program that was used. define some terms of common usage. There was not universal agreement on the meaning of these words, of course, and the individual authors use these definitions or not as they find appropriate in the following articles. Our purpose here is not The Organization of This Issue to set the final definitions of important terminology but rather simply to try to use these words consistently in the papers Because of the number of participants and the size and format presented here. A number of authors have suggested defini- of this publication, we are limited to brief contributions and a few illustrations each. For these reasons as well, our intro- tions for these terms over the years (e.g., Harris 1996, 2007), duction will be brief. All of the papers have been rewritten but there are still no widely accepted meanings. In the papers following the symposium to incorporate the ideas and input in this issue of Current Anthropology, the following definitions from our discussions. These revisions clearly reflect the impact will be used unless noted by the individual authors. of the discussions on the participants and the stimulus that Mobility and sedentism. These are relative terms that de- was provided by the symposium. scribe a range from completely mobile to completely sed- The papers in this issue have been organized largely by entary. Sedentism is difficult to measure in the archaeology chronology and geography to facilitate access for the reader. of the last hunters and first farmers, and this definition at- The earliest evidence for the origins of agriculture comes from tempts to recognize that. It was suggested that the presence Southwest Asia. Bar-Yosef (2011) writes about the role of of commensals, such as house mice, and the seasonal distri- climate change and the congruence in the chronology of ag- bution of plant foods within the same site may indicate an ricultural origins in the Near East and China with the Younger annual long-term occupation. Dryas cold episode at the end of the Pleistocene. Goring- Management. Manipulation and some degree of control Morris and Belfer-Cohen (2011; see also Belfer-Cohen and of wild species (plants or animals) without cultivation or Goring-Morris 2011) have provided two contributions de- morphological changes. In the 1970s this kind of treatment tailing the Neolithization process in the Near East as seen was referred to as “cultural control.” within the core area and from the larger region. One of the Cultivation. Intentional preparation of the soil (Oxford fascinating things about their contribution about the “Neo- English Dictionary) for planting wild or domesticated plants. lithic revolution” is that is has little to do with subsistence. Identified in many cases by arable weeds in cereal caches Animal and plant management as well as domestication in (Bogaard et al. 1999; Harris 2007). The term is often used to the Near East are the focus of papers by Zeder (2011), Weiss indicate cultivation of wild plants before domestication. and Zohary (2011), and Vigne et al. (2011). Zeder provides Domestication. Morphological or genetic changes in plant a thorough discussion of current evidence for domestication. and animal species. Archaeozoologists also use criteria such Weiss and Zohary consider the evidence for the major founder as age profiles, milking, and osteological pathologies. species of plants. Vigne and colleagues provide a look at the Farming. Utilization of domestic plants and/or animals remarkable evidence from Cyprus where Pre-Pottery Neo- lithic people brought plants and animals, essentially re-cre- for food as well as other resources. ating an Early Neolithic ecosystem on that rather barren Med- Agriculture. Farming and/or herding predominate the ac- iterranean island. tivities of a particular community and determine the main The next set of papers deals with East Asia, largely from diet, although hunting and gathering may continue. an archaeobotanic perspective. Cohen (2011) offers an over- There were also a number of terms that were not used view of the beginnings of agriculture in China in the context often at the symposium. Words such as “horticulture,” “ar- of interregional interaction. Zhao (2011) details the abundant boriculture,” “herding,” “pastoralism,” “husbandry,” “stor- new botanical data that have appeared in the past 10 years. age,” “agropastoralism,” and “intensity of food production” Lee (2011) elaborates on agricultural origins in Korea. Craw- may also be important and useful to scholars of early agri- ford (2011) provides an overview of recent advances in our culture in both the Old and New Worlds. Bruce Smith and understanding of early cultivation in East Asia with a focus others, for example, found utility in the concept of “low-level on Japan. food production” (Smith 2001). In an ideal world, all of these South Asia and Africa provide a very different statement terms would have fixed meanings that could be used easily on the beginnings of farming, and questions continue about in discussions of the first farmers, but for the present such the appearance of native domesticates. Fuller (2011) tracks consensus does not exist. this issue, finding plant domestication in the Indian subcon- S166 Current Anthropology Volume 52, Supplement 4, October 2011 tinent. Animal domestication in Africa is the subject of the sidering the early theories first. Explanations of why domes- contribution by Marshall and Weissbrod (2011), with a focus tication occurred include the oasis hypothesis, the natural- on the donkey in particular and African pastoralism in gen- habitat hypothesis, the population-pressure hypothesis, the eral. edge hypothesis, the social hypothesis, and more. A consid- In the Pacific, two studies are presented. Bellwood (2011) eration of these ideas also reveals much about the nature of provides an overview of Holocene population history of the archaeology and archaeologists. region as a model for worldwide food-producer dispersals. During the first half of the twentieth century, the best in- Denham (2011), on the other hand, offers details of the evi- formation on early farming villages came from riverine areas dence for domestication and early agriculture in New Guinea or oases in Northeast Africa and Southwest Asia—along the and Island Southeast Asia. Nile River in Egypt and at Jericho in the Jordan Valley, for The spread of agriculture subsequent to its origins is the example. Early views on the origins of agriculture focused on subject of papers dealing with the European evidence. O¨ z- climate change. At that time, the end of the Pleistocene was dog˘an (2011) provides an in-depth look at the archaeological assumed to have been a period of increasing warmth and evidence for the westward movement of early farmers from dryness in the earth’s climate. Scholars reasoned that because Anatolia to the Aegean and Balkans of Europe. The new evi- the ice ages were cold and wet, they should have ended with dence from European Turkey and Bulgaria is rewriting our higher temperatures and less precipitation. Given that view understanding of this spread. Rowley-Conwy (2011) contin- of past climate, logic suggested that areas such as Southwest ues the story of westward movement from Central Europe to Asia, a dry region to begin with, would have witnessed sig- the Atlantic. The picture of the expansion of early farming nificant aridity at the end of the Pleistocene when vegetation across Europe is now one of repeated episodes of very rapid grew only around limited water sources. The oasis hypothesis spread followed by long periods of stability. Bursts of regional suggested a circumstance in which plants, animals, and hu- expansion are not in contrast to the overarching model of mans would have clustered in constrained zones near water. continuous millennial-scale demic diffusion but provide a V. Gordon Childe, one proponent of this idea, argued that more detailed picture that brings the information closer to the only successful solution to the competition for food in the social history of particular populations. these situations would be for humans to domesticate and Turning to the New World, Piperno (2011) examines the control the animals and the plants. In this sense, domesti- evidence for origins of plant cultivation and domestication cation emerged as a symbiotic relationship for the purpose in the tropics of Central and South America. Abundant new of human survival. evidence from starch grains and phytoliths as well as mac- During the 1940s and 1950s, however, new evidence sug- robotanical remains provides exciting new information and gested that there had been no dramatic climatic changes in pushes back the dates for early domestication. Smith (2011) Southwest Asia at the close of the Pleistocene—no crisis dur- continues his documentation of an early center for domes- ing which life would concentrate at oases. The new infor- tication in Eastern North America with a consideration of the mation forced a reconsideration of the origins of agriculture. cultural context of this process, the evidence coming from The late Robert Braidwood pointed out in his natural-habitat archaeology. hypothesis that the earliest domesticates therefore should ap- Two papers were topical rather than areal and covered im- pear where their wild ancestors lived. That area, the “hilly portant aspects of the study of agricultural origins. Larson flanks” of the Fertile Crescent in Southwest Asia, should be (2011) provides a refreshingly candid view of the role of ge- the focus of investigations. Braidwood and a large team of netics in the study of plant and animal domestication. There researchers excavated at the site of Jarmo in northern Iraq is great promise in this method, but serious problems remain and elsewhere in the Fertile Crescent and found evidence of to be solved. Jean-Pierre Bocquet-Appel (2011), proponent early agriculture, supporting his hypothesis that domestica- of the concept of the Neolithic demographic transition, de- tion did indeed begin in the natural habitat. scribes the background and development of this model of Braidwood did not offer a specific reason as to why do- population growth across the transition to agriculture (Boc- mestication occurred other than to point out that technology quet-Appel and Bar-Yosef 2008). and culture were ready by the end of the Pleistocene and that humans were familiar with the species that were to be do- mesticated. At that time, archaeologists and others generally The State of Play considered that farming was a highly desirable and welcome invention providing security and leisure time for prehistoric In order to place the results of our symposium and some of peoples. Once human societies had recognized the possibilities the new ideas and information that emerged in perspective, of domestication, they should have immediately started farm- it may be useful to briefly review the development of ideas ing following the perspectives of the time. about the origins of agriculture and some of the explanations Lewis Binford (1968) challenged those ideas in the 1960s that have been proffered. We can best understand ideas about and focused on population. Binford argued that farming was the origins of agriculture from a historical perspective, con- backbreaking, time consuming, and labor intensive. Citing Price and Bar-Yosef The Origins of Agriculture S167 studies of ethnographically known hunter-gatherers, he in the ability of certain individuals to generate a surplus of pointed out that they spend only a few hours a day obtaining food and to transform that surplus into more valued items, food; the rest of their time is for visiting, talking, gambling, such as rare and valued materials and objects. From this per- and the general pleasures of life. Even in very marginal areas, spective, agriculture was the means by which social inequality such as the Kalahari Desert of South Africa, food collecting emerged and egalitarian societies became hierarchical. Such is a successful adaptation, and people rarely starve. Binford a view is intriguing if difficult to document. It is very much argued, therefore, that human groups would not become a chicken-egg question, like the issue of population pressure, farmers unless they had no other choice, that the origins of of which came first. There is subtle evidence for social in- agriculture was not a fortuitous discovery but a last resort. equality in the Early Neolithic of the Near East (Price and Binford made his point by positing an equilibrium between Bar-Yosef 2010), but such reorganization of social relations people and food, a balance that could be upset by either a and wealth accumulation may be a consequence of agricul- decline in available food or an increase in the number of tural production rather than a cause. people. Because climatic and environmental changes appeared There are a number of other intriguing theories about why to be minimal in Southwest Asia, Binford thought it must human societies began to cultivate the earth along with some have been increased population size that upset the balance. not so enlightening ideas. Geographer Carl Sauer (1952) sug- Population pressure was thus introduced as a causal agent for gested that agriculture began in the hilly tropics of Southeast the origins of agriculture: more people required more food. Asia, where sedentary groups with knowledge of the rich plant The best solution to the problem was domestication, which life of the forest might have domesticated plants for poisons provided a higher yield of food per acre of land. At the same and fibers. Botanist David Rindos (1984) has argued that time, however, agricultural intensification required more la- domestication was a process of interaction between humans bor to extract the food. and plants evolving together into a more beneficial symbiotic Binford’s concern with population was elaborated and ex- relationship. tended as a global explanation by Mark Cohen (1977, 2009). In a fascinating volume titled Naissance des divinite´s, naiss- Cohen argued for an inherent tendency for growth in human ance de l’agriculture: la re´volution des symboles au Ne´olithique population, a pattern responsible for the initial spread of the (The Birth of the Gods and the Origins of Agriculture), French human species out of Africa, the colonization of Asia and archaeologist Jacques Cauvin (1994) argued that the impor- Europe, and eventually colonization of the Americas as well. tant changes associated with the “Neolithic revolution” were According to Cohen, after about 10,000 BC, all the habitable more cultural than economic. He meant that the transition areas of the planet were occupied, and population continued to farming involved concepts and ideas as much as or more to grow. At that time, there was an increase in the use of less than food production. Specifically, he suggested that agricul- desirable resources in many areas. Land snails, shellfish, birds, ture was preceded by the emergence of new cosmologies, and many new plant species were added to the human diet religious practices, and symbolic behaviors. This transfor- around the end of the Pleistocene. Cohen argued that the mation of hunter-gatherers that allowed them to view their only way for a very successful but rapidly increasing species habitat in a different way also promoted the more active to cope with declining resources was for them to begin to exploitation of that environment according to the views of cultivate the land and domesticate its inhabitants rather than Cauvin. simply to collect the wild produce. In recent years, the school of evolutionary ecology in ar- Domestication for Cohen was a solution to problems of chaeology has provided its take on the origins of agriculture overpopulation on a global scale. There is, however, very little in a series of papers and volumes (e.g., Bleed and Matsui evidence for population pressure in the record of agricultural 2010; Gremillion and Piperno 2009; Winterhalder and Ken- origins. Population is a notoriously difficult parameter to nett 2006, 2009). Evolutionary ecology developed out of an grasp in prehistory. In the very few cases where some infor- earlier perspective known as cultural ecology, which focused mation is available, such as in the Levant as presented in this on the dynamic relationship between human society and its issue by Belfer-Cohen and Goring-Morris (2011), there may environment using culture as the primary mechanism of ad- even be some population decline shortly before the first ap- aptation. Culture is a giant concept and hard to work with, pearance of domesticates. As Bocquet-Appel (2011) points so evolutionary ecologists have emphasized human abilities out in this issue and elsewhere, most of the evidence for to reason and to optimize their behavior. In this view, natural growing population comes after the origins of agriculture, not selection is thought to operate on the behavior of individuals. before. Evolutionary ecologists assume that natural selection designed Others, arguing that the transition to farming and food organisms to adapt to local conditions in fitness-enhancing storage and surplus cannot be understood simply in terms of or optimizing ways. environment and population, have developed social hypoth- One of the major tenets of this perspective involves a con- eses to explain the origins of agriculture. Barbara Bender cept known as optimal foraging theory—borrowed from bi- (1975) and Brian Hayden (1992, 1995), among others, have ology—to explain the food-getting behavior of humans, es- suggested that the origins of food production may lie more pecially hunter-gatherers. Optimal foraging theory argues that S168 Current Anthropology Volume 52, Supplement 4, October 2011 the most efficient foraging strategies produce the greatest re- single accepted theory for the origins of agriculture—rather, turn in energy relative to time and effort expended. Optimal there is a series of ideas and suggestions that do not quite foraging assumes that humans make rational decisions based resolve the question. At the same time, of course, the evidence on economic efficiency. Evolutionary ecologists examine the we have is scanty and limited. A great deal more research and archaeological and ethnographic record looking for things discussion needs to be done. That is why we convened this such as “optimization goals,” “currencies,” and “constraints” symposium. and applying ecological and mathematical models to explain human behavior. Ideas about the origins of agriculture have sometimes been categorized as either push or pull models. Hunter-gatherers Some Observations are either pushed, or forced, to become farmers or they are The focal point of the symposium was, of course, the origins pulled, drawn by the benefits of a new lifestyle. Population- of agriculture. We hoped to bring together new data and new pressure models, for example, force human societies to find ideas to push our understanding of this remarkable phenom- new ways to feed growing numbers of members. Social hy- enon further along. The origins of agriculture is one of the potheses usually involve pull, in which members of society are drawn into relationships of inequality in order to benefit most important developments in our past. Virtually every- from new arrangements that reduce risk or increase wealth. thing we as humans know and do today stems from this The perspective of evolutionary ecology involves push remarkable transition. Detailed study of this issue—the pre- models. Hunter-gatherers operate on the premise of efficiency sentation of evidence and the evaluation of potential an- to acquire sufficient food to eat. Foods are ranked by net swers—has significance for students of the past, for anthro- energy value, and lower-ranked subsistence resources (such pology as a whole, and for a wide range of related areas of as seeds) are added only as higher-ranked foods become un- scholarly interest. available. Such push perspectives appear to assume that There are both practical and theoretical implications of the hunter-gatherers are surviving among limited resources in study of agricultural transitions. The documentation of when difficult environments. One of the most important and en- and where farming began provides a powerful statement re- lightening realizations in recent years is the fact that agri- garding the global nature of this event. Investigation of the cultural origins take place in relatively abundant environ- shift from hunting and gathering to farming invokes virtually ments, not in places where little food is available (Price and all aspects of anthropological perspectives on human behavior Gebauer 1995). It is in such situations of sufficient subsistence and cultural change. The transition to agriculture is a com- resources, where risk is limited, that experiments leading to mon human experience that has effected us all in terms of the origins of agriculture took place. Higher-ranked food re- rapid population growth and aggregation and social inequal- sources do not appear to have declined or become unavailable ity. Is it “the worst mistake in the history of the human race” in such contexts, bringing into question the utility of such (Diamond 1987) or an inevitable step in the evolution of optimal foraging models. human society? More recently, detailed information on climate change has A multitude of developments concerned with the origins come from a most unlikely place, the glaciers of Greenland. and spread of agriculture have taken place in recent years. Deep corings of the ice sheets there have provided a layered New fieldwork and new sites in new and old places, more record of changes in temperature and other aspects of climate radiocarbon dates, and new methods have documented earlier for the past 100,000 years and more. One of the very inter- transitions to agriculture in parts of Asia, the south Pacific, esting results of this research is the documentation of a 33% and the Americas. Studies of microscopic plants remains, es- pecially starch grains and phytoliths, have revolutionized increase in atmospheric CO2 at the end of the Pleistocene identification of plant exploitation before the emergence of (Sage 1995). Higher levels of CO2 would foster the expansion of temperate species such as grasses, which include many of cultivation as well as the appearance of domesticated plants. the ancestors of the major domesticated species. The full im- Advances in the genetics of domestication, such as utilizing plications of such changes in the atmosphere are not yet clear ancient DNA to examine the relationships among prehistoric but may play a role in the transition from hunting to farming. domestics, are beginning to resolve standing questions about

The evolutionary ecologists have picked up on CO2 and argue where and when. It is time to assemble this new information, that climatic amelioration that followed in the Early Holocene to sift and winnow, and to summarize our current under- and the accompanying rise in CO2 made the origins of ag- standing of the origins and spread of agriculture. riculture “compulsory” (Bettinger, Richerson, and Boyd Our symposium on the state of the art in the study of the 2009). origins of agriculture was intended to provide a baseline for The simple fact is that we do not yet have a good grasp continuing and future work. We wanted to learn new facts, on the causes for the origins of agriculture. The how and the examine a wide range of variables, and use our knowledge to why of the Neolithic transition remain among the more in- evaluate current explanations and to explore new ideas for triguing questions in human prehistory. There is as yet no understanding what took place at the origins of agriculture. Price and Bar-Yosef The Origins of Agriculture S169

Above all we wanted to think in new directions about this (Piperno 2011). New work is beginning to provide data from large, complex, and obstinate issue. critical regions in Mesoamerica, one of the least understood There is too little collaboration and interaction between regions of early agricultural origins (Piperno et al. 2009; Ran- the varied disciplines investigating this question (e.g., genetics, ere et al. 2009). Archaeobotany is moving forward rapidly botany, zoology, archaeology, linguistics, demography). An with a variety of techniques for recording information related enormous amount of research is going on today in a rather to domestication (Allaby, Fuller, and Brown 2008; Fuller uninformed context. This symposium was intended to inte- 2011). Genetic studies of modern and ancient DNA in do- grate that context and provide shared perspectives on the mesticated plants and animals are also providing remarkable question of agricultural origins. This work is going on around data on species distribution and evolution (e.g., Dobney and the world, and one of our primary goals was to bring together Larson 2006). Genetic markers for domestication are starting the leading scholars concerned with this question from diverse to be identified. At the same time, a note of caution regarding places and origins. Such gatherings are extremely rare and genetic studies, especially age estimation based on mutation often very fruitful. rate, permeated the symposium and was reiterated by our We were a volatile mix of scholars, from many times and resident archaeogeneticist (Ho and Larson 2006). places. At the end of our time together, we did not determine One of the most interesting phenomena we noted was not why agriculture originated. We did not even agree on whether pattern but variation. In the one or two places where data its causes were global or local. Archaeozoologists decried the on the transition are relatively rich, there appears to be a difficulties involved in identifying domestication and seek period of chaos, a “zone of variability” at the origins of ag- complex forms of evidence for the process (Zeder 2006a, riculture (Weiss, Kislev, and Hartmann 2006). There seems 2006b); archaeobotanists seek changes in morphology and to be a period for the auditioning of many possible new genetic makeup as indicators of changes in plants (e.g., Smith options in human adaptation. This is the beginning of a new 2006; Zohary and Hopf 2000). These differences mean distinct way of life. views on the question of agricultural origins from the two Three recent discoveries from the earliest Neolithic in the subdisciplines. Some might conclude that our symposium was Near East highlight this zone of variability, change our un- not successful. derstanding of this period in human prehistory, and raise In fact, we believe this symposium was a great success. enormous new questions. The colonization of the Mediter- Perhaps most importantly there was a strong sense of collab- ranean island of Cyprus by Late Pre-Pottery Neolithic A and oration at the meeting. There was much new in terms of both Pre-Pottery Neolithic B people carrying domestic plants and information and ideas. There was a major emphasis on the domestic and wild animals by boat is an extraordinary story origins of agriculture in East Asia. Lesser-known regions such (Guilaine et al. 1998, 2000; Peltenburg and Wasse 2004; Pel- as Papua New Guinea, Africa, and eastern North America tenburg et al. 2000; Simmons 2007; Vigne et al. 2000). Ex- were included in our discussions. Lots of new data were pre- cavations at Go¨bekli Tepe in southern Turkey have revealed sented from East and West Asia, Africa, and Central and South a series of remarkable shrines or centers associated with large America. We were able to put together a table of the latest stone architecture and art from the same time period (Peters information on the antiquity of agriculture in various parts and Schmidt 2004; Schmidt 2001, 2003, 2006; Schmidt and of the world and recognized that there are at least 10 different Hauptman 2003). The roughly contemporary burial ground places with claims as original centers of domestication (fig. of Kafar HaHoresh in Israel documents enormous new var- 1). Information on estimated dates BP cal for domestication iation in the treatment of the dead and indications of emerg- in these areas is provided in table 1 (see also fig. 1). ing social inequality at this time (Goring-Morris 2005; Gor- The antiquity of domestication has been pushed deeper ing-Morris et al. 1998). into the past in many areas. Today, an eerie synchronicity in A number of potentially important variables involved in the timing of the first domesticates around the end of the the shift from foraging to farming were discussed at the sym- Pleistocene is emerging. Another commonality among the posium. These include sedentism, storage, population density, cradles of agriculture is the rich environments in which farm- population pressure, resource abundance, resource availabil- ing originates. Experiments in domestication do not take place ity, niche construction, processing and harvesting technolo- in marginal areas but amid concentrations of population and gies, climate and environmental changes, ownership of pro- resources across the globe. It also appears that in each area duce and resource localities, potential domesticates, where several different species are involved in the transition competition, inequality, risk reduction, nutritional require- to agriculture, there are multiple centers of domestication ments, choice, chance, and a receptive social/cultural context. within the region. A number of groups appear to be manip- ulating their natural world. Remarkable new studies are documenting this evidence. Some Conclusions Microscopic studies of starch grains in South America have identified a number of early crops, and more specific infor- Farming is a way of obtaining food that involves the culti- mation on their origin and distribution is becoming available vation of plants and the controlled herding of animals. But Figure 1. Major centers of domestication and dates for earliest plants and animals. Illustration by Marcia Bakry. A color version of this figure is available in the online edition of Current Anthropology. Price and Bar-Yosef The Origins of Agriculture S171

Table 1. Approximate dates for the appear- the process of domestication and to distinguish biology from ance of domesticated species in various parts culture in the transition from hunting to farming. The criteria of the world for identifying domestication differ significantly for plants and animals. Plants rather quickly exhibit distinct morphological Date of appearance changes; animals are much slower to show such develop- Place and species (cal BP) ments. Archaeozoology has introduced a number of impor- Southwest Asia: tant new concepts regarding the process. Three classes of do- Plants 11,500 mesticates can be identified: (1) commensals, adapted to a Animals 10,500 China: human niche (e.g., dogs, cats, guinea pigs); (2) prey animals Millet 10,000 sought for food (e.g., cows, sheep, pig, goats); and (3) targeted Rice 17000 animals for draft and nonfood resources (e.g., horse, camel, Mexico: donkey). Changes related to domestication may be more spe- Corn 9000 cies specific in animals. These differences mean that non- South America: Plants 10,000 morphological criteria, such as changes in age profiles of Animals 6000 herds, may be required for fauna. New Guinea: Increasing site size is one of the primary archaeological Plants 17000 indicators of changes in human subsistence and organization South Asia: associated with the agricultural revolution. This increase is a Plants 5000 Animals 8000 reflection of population growth as well as new forms of set- Africa: tlement and organization. One of the more striking devel- Plants 5000 opments associated with the arrival of farming is the increas- Animals 9000 ing visibility of a human presence in the archaeological record. Eastern North America: Hunter-gatherers rarely leave visible traces, few bumps in the Plants 5000 landscape. Shell middens, some ditches, and other features remain today, but even the most complex hunter-gatherer adaptations did not modify the landscape to a large extent or leave many traces that are visibly on the surface of the earth the beginnings of new subsistence systems were much more today. than cultivation, herding, or the ensuing domestication of The phantom of causality floated at the edge of our delib- various species. This revolution entailed major long-term erations at Temozon, always there but not often addressed. changes in the structure and organization of the societies that This transition from hunting to farming poses one of the adopted this new way of life as well as a totally new rela- most intriguing questions about the human past and one of tionship with the environment. Humans truly began to har- the most difficult to answer: why did hunters become farmers? ness the earth. While hunter-gatherers live off the land in an Causality is such a thorny issue. There is no common playing extensive fashion, generally exploiting a diversity of resources field or shared rules of engagement. The evidence from dif- over a broad area, farmers utilize the landscape intensively ferent parts of the world varies in both quantity and quality. and create a milieu that suits their needs. That discordance is disconcerting. Variation generates many The symposium at Temozon and the papers proffered in perspectives and divergence. Consensus on fundamentals is this issue attempt to describe some of the latest evidence and lacking. to comprehend these extraordinary developments. Here, we Are there general causes? The almost simultaneous devel- hope to provide some sense of the results of the symposium opment of agriculture in so many different places is not simple and our own perception of the state of knowledge concerning coincidence. Should we invoke climate, environment, pop- the origins of agriculture. We learned a great deal. Some of ulation, subsistence intensification, brain capacity, religion, the major threads woven throughout the symposium included inequality, entrepreneurs? Are there specific conditions? Are the importance of integrating the subdisciplines, the aban- there immediate and local causes distinct from global ones? donment of dichotomies for the study of process and trans- Are the origins of agriculture the results of a “perfect storm” formation, the chaos of transitions (a period of “auditioning,” of factors that forced or encouraged human societies to do- as Bruce Smith termed it), the importance of attention to mesticate plants and animals? detail in the study of landscapes and species as well as ar- The division between generalists and particularists was clear chaeological sites, the punctuated nature of agricultural at Temozon. Particularists wanted to look at each individual spread, the many major gaps in our knowledge, and the ne- case of agricultural origins as unique; generalists sought a cessity for critically reevaluating existing information. There more global explanation that would encompass all of the areas was also lots of terminology, much new data, some innovative where early agriculture appeared. Some in the group insist ideas about causality, and many remaining questions. that causality is a local or regional phenomenon that varies It is important to separate the origins of agriculture from across time and space. Zeder and Smith (2009) have argued S172 Current Anthropology Volume 52, Supplement 4, October 2011 at some length that global views, “one-size-fits-all” approaches important factors in the transition from the perspective of to the explanation of agricultural emergence and dispersal, the authors presented here include, in order of suggested are not feasible. Bruce Smith enlivened the symposium one importance, available protodomesticates, human sedentism, morning by arguing that “causality is in the eye of the be- higher population density, resource abundance, geographic holder”—that generalization was not possible because of the and/or social constraints, processing and harvesting technol- complexity of individual contexts. ogy, storage, and wealth accumulation. A change from com- Others believe that the simultaneity of the origins of ag- munity to household levels of economic organization ob- riculture in time argues for general or global causes. It is served in several areas may have accompanied the transition completely remarkable that the process of domesticating to agriculture, including a shift from communal sharing to plants and animals appears to have taken place separately and familial or individual accumulation. Economic intensification independently in a number of areas at about the same time. and competition were frequent companions of the Neolithic Given the long prehistory of our species, why should the revolution society (Price and Bar-Yosef 2010). Wealth accu- transition to agriculture happen within such a brief period, mulation and status differentiation appear at the individual, a few thousand years in a span of more than 6 million years household, or lineage level. of human existence? An important and dramatic shift in the We were, however, unable to agree on the primacy of causal trajectory of human adaptation would seem to demand gen- factors or on the major issue of general versus specific ex- eral explanation. But such answers are hard to reach. planation. The frustrations of explanation leave us asking Anna Belfer-Cohen brightened the symposium room one questions—many important questions were raised at the sym- morning with a quote from Nigel Barley (1989, p. 205): “An- posium. Beyond the specific details of local sequences, the thropology largely neglects the individual to deal in gener- nuances of plant and animal domestication, and concerns alizations. Generalizations always tell a little lie in the service with the meaning of the evidence, certain larger questions of greater truth.” Conclusions, of course, involve generali- arose again and again. What determines where the first farm- zations. It is important to recognize some of the limitations ers appeared? What makes centers of origin special places? and constraints on such broad speculation. In spite of ex- Why do humans domesticate plants and animals? Why are traordinary advances in a variety of fields, many detailed at some plants and animals selected for manipulation and not the symposium, we really know very little about the origins others? Is the domestication process determined solely by the of agriculture. There is some information from Southwest biology of the selected species? What can we say about timing? Asia, a little data from East Asia and North America, and How long does it take to domesticate plants and animals? next to nothing from the rest of the world. We are still in the How does the timing of this process interlace with develop- early stages of the process of identifying and understanding ments such as sedentism, population growth, and social in- this transition from hunting to farming. equality? Why was agriculture such a successful adaptation? Moreover, as we learn more about a specific region, it be- How does agriculture spread quickly to areas with different comes clearer how complex the past was, exhibiting much cultures, climates, and environments? Can the spread of ag- more variability than we have admitted or realized. The Near riculture tell us about its origins? What spreads, people or East is the prime example. There is more information from things? How do we best explain the origins of agriculture? this region about the origins of agriculture than anywhere Questions provide the right subject to end our introduction else in the world: more sites, more excavations, more analyses, and enter the assembled essays that follow. Questions and more publications, and so on. Yet new discoveries in the past curiosity, of course, are inherent in the pursuit of knowledge; 20 years have completely altered our understanding of this unanswered questions drive continuing research. Archaeology area and revealed levels of complexity not even imagined 2 is concerned with questions about the past. The origins of decades ago. agriculture is one of the most important and most obstinate. A number of important general factors in the origins of Yet we have no doubt that while it will be a long and arduous agriculture were recognized at the symposium. These factors journey, our search for answers will have a successful end. can be categorized as exogenous, or natural (e.g., climate/ Our goal for the symposium was to develop and explore environment, population growth), and endogenous, or cul- a rich and productive dialog among scholars from diverse tural (e.g., social change, religion). Theories on the transition branches of archaeology and related disciplines focused on to agriculture have most often focused on external factors the beginnings of farming. Through the course of the sym- such as climatic change or inherent growth in population as posium there was a growing respect and a leveling of bound- problems solved by the cultivation of plants and animals. aries between the subdisciplines. There was lots of news— Exogenous factors are generally natural forces over which much information and inspiration. We believe that the par- human groups have little control; endogenous factors reflect ticipants in this symposium returned home with renewed internal change within society and decisions that humans optimism about the state of research—both data and ideas— make. on the origins of agriculture. It is our hope that enthusiasm A series of significant variables involved in the shift from will be conveyed through the continuing studies of the par- foraging to farming were discussed at Temozon. The most ticipants and will be passed to their colleagues and students. Price and Bar-Yosef The Origins of Agriculture S173

In this way, our symposium will have a large impact on the Cohen, David Joel. 2011. The beginnings of agriculture in China: a archaeological community, and we can help to push future multiregional view. Current Anthropology 52(suppl. 4):S273–S293. Cohen, Mark. 1977. The food crisis in prehistory: overpopulation and research along a well-lit path. the origins of agriculture. New Haven, CT: Yale University Press. ———. 2009. Introduction: rethinking the origins of agriculture. Current Anthropology 50:591–595. Crawford, Gary W. 2011. Advances in understanding early agriculture Acknowledgments in Japan. Current Anthropology 52(suppl. 4):S331–S345. Denham, Tim. 2011. Early agriculture and plant domestication in Special thanks to Leslie Aiello, whose presence we enjoyed at New Guinea and Island Southeast Asia. Current Anthropology the symposium and whose constant interest and occasional 52(suppl. 4):S379–S395. participation were much appreciated. Leslie has also played Diamond, Jared. 1987. The worst mistake in the history of the human a key role in the final editing of this issue of Current An- race. Discover,May. thropology. Kudos and enormous thanks to Laurie Obbink of Dobney, Keith, and Greger Larson. 2006. Genetics and animal do- mestication: new windows on an elusive process. Journal of Zoology the Wenner-Gren Foundation, who is responsible for the lo- 269:261–271. gistics and function of these meetings and has been doing a Fuller, Dorian Q. 2011. Finding plant domestication in the Indian magnificent job for many years. Our symposium ran re- Subcontinent. Current Anthropology 52(suppl. 4):S347–S362. markably smoothly and enjoyably. Carolyn Freiwald of the Goring-Morris, A. Nigel, and Anna Belfer-Cohen. 2011. Neolithi- University of Wisconsin–Madison attended the symposium zation processes in the Levant: the outer envelope. Current An- thropology 52(suppl. 4):S195–S208. as an invited graduate student, and her presence was appre- Goring-Morris, N. 2005. Life, death and the emergence of differential ciated by all. Thanks also to the staff of the Hacienda Temozon status in the Near Eastern Neolithic: evidence from Kfar Ha- who went beyond the call of duty to ensure a pleasant stay Horesh, Lower Galilee, Israel. In Archaeological perspectives on the for all of us with wonderful food, flower-strewn rooms, and transmission and transformation of culture in the Eastern Mediter- much individual attention. David Meiggs has done a masterful ranean. J. Clarke, ed. Pp. 89–105. Oxford: Council for British Research in the Levant/Oxbow. job of editing the manuscripts for publication. Goring-Morris, N., R. Burns, A. Davidzon, V. Eshed, Y. Goren, I. Hershkivitz, S. Kangas, and J. Kelecevic. 1998. The 1997 season References Cited of excavations at the mortuary site of Kfar Hahoresh, Galilee, Israel. Neo-Lithics 98:1–4. Allaby, R. G., D. Q Fuller, and T. A. Brown. 2008. The genetic ex- Gremillion, K. J., and D. R. Piperno. 2009. Human behavioral ecol- pectations of a protracted model for the origins of domesticated ogy, phenotypic (developmental) plasticity, and agricultural ori- crops. Proceedings of the National Academy of Sciences of the USA gins: insights from the emerging evolutionary synthesis. Current 105:13982–13986. Anthropology 50:615–619. Barley, Nigel. 1989. Not a hazardous sport. London: Penguin. Guilaine, J., F. Briois, J. Coularou, P. Deveze, S. Philibert, J.-D. Vigne, Bar-Yosef, Ofer. 2011. Climatic fluctuations and early farming in West and I. Carre`re. 1998. La site neolithique preceramique de Shill- and East Asia. Current Anthropology 52(suppl. 4):S175–S193. ourokambos (Parekklisha, Chypre). Bulletin de Correspondance Belfer-Cohen, Anna, and A. Nigel Goring-Morris. 2011. Becoming Hellenique 122:603–610. farmers: the inside story. Current Anthropology 52(suppl. 4):S209– Guilaine, J., F. Briois, J.-D. Vigne, and I. Carre`re. 2000. Decouverte S220. d’un neolithique preceramique ancien chypriote (fin 9e debut 8e Bellwood, Peter. 2011. Holocene population history in the Pacific millenaires cal. BC), apparente au PPNB ancien/moyen du Levant region as a model for worldwide food producer dispersals. Current nord. Earth and Planetary Sciences 300:75–82. Anthropology 52(suppl. 4):S363–S378. Harris, David R. 1996. Domesticatory relationships of people, plants, Bender, Barbara. 1975. Farming in prehistory: from hunter-gatherer to and animals. In Redefining nature: ecology, culture, and domesti- food-producer. New York: St. Martin’s. cation. R. Ellen and K. Fukui, eds. Pp. 437–463. Oxford: Berg. Bettinger, R., P. Richerson, and R. Boyd. 2009. Constraints on the ———. 2007. Agriculture, cultivation and domestication: exploring development of agriculture. Current Anthropology 50:627–631. the conceptual framework of early food production. In Rethinking Binford, Lewis R. 1968. Post-Pleistocene adaptations. In New per- agriculture: archaeological and ethnoarchaeological perspectives.T. spectives in archaeology. L. R. Binford and S. R. Binford, eds. Pp. Denham, J. Iriarte, and L. Vrydaghs, eds. Pp. 16–35. Walnut Creek, 313–341. Chicago: Aldine. CA: Left Coast. Bleed, P., and A. Matsui. 2010. Why didn’t agriculture develop in Hayden, B. 1992. Models of domestication. In Transitions to agri- Japan? a consideration of Jomon ecological style, niche construc- culture in prehistory. A. B. Gebauer and T. D. Price, eds. Pp. 11– tion, and the origins of domestication. Journal of Archaeological 19. Madison, WI: Prehistory. Method and Theory 17:356–370. ———. 1995. A new overview of domestication. In Last hunters, first Bogaard, A., C. Palmer, G. Jones, and M. Charles. 1999. A FIBS farmers: new perspectives on the transition to agriculture. T. D. Price approach to the use of weed ecology for the archaeobotanical and A. B. Gebauer, eds. Pp. 273–300. Santa Fe, NM: School of recognition of crop rotation regimes. Journal of Archaeological Sci- American Research. ence 26:1211–1224. Ho, S. Y. W., and G. Larson. 2006. Molecular clocks: when times are Bocquet-Appel, J.-P., and O. Bar-Yosef, eds. 2008. The Neolithic de- a-changin’. Trends in Genetics 22(2):79–83. mographic transition and its consequences. Dordrecht: Springer. Larson, Greger. 2011. Genetics and domestication: important ques- Bocquet-Appel, Jean-Pierre. 2011. The agricultural demographic tions for new answers. Current Anthropology 52(suppl. 4):S485– transition during and after the agriculture inventions. Current An- S495. thropology 52(suppl. 4):S497–S510. Lee, Gyoung-Ah. 2011. The transition from foraging to farming in Cauvin, Jacques. 1994. Naissance des divinite´s, naissance de Prehistoric Korea. Current Anthropology 52(suppl. 4):S307–S329. l’agriculture: la re´volution des symboles au Ne´olithique. Paris: CNRS. Marshall, Fiona, and Lior Weissbrod. 2011. Domestication processes S174 Current Anthropology Volume 52, Supplement 4, October 2011

and morphological change: through the lens of the donkey and Schmidt, K., and H. Hauptmann. 2003. Go¨bekli Tepe et Nevali Cori. African pastoralism. Current Anthropology 52(suppl. 4):S397–S413. Dossiers d’Arche´ologie 281:60–67. O¨ zdog˘an, Mehmet. 2011. Archaeological evidence on the westward Simmons, Alan H. 2007. The Neolithic revolution in the Near East: expansion of farming communities from eastern Anatolia to the transforming the human landscape. Tucson: University of Arizona Aegean and the Balkans. Current Anthropology 52(suppl. 4):S415– Press. S430. Smith, B. D. 2001. Low level food production. Journal of Archaeo- Peltenburg, E., and A. Wasse, eds. 2004. Neolithic revolution: new logical Research 9:1–43. perspectives on Southwest Asia in light of recent discoveries on Cyprus. ———. 2006. Documenting domesticated plants in the archaeolog- Oxford: Oxford University Press. ical record. In Documenting domestication: new genetic and ar- Peltenburg, Edgar, Sue Colledge, Paul Croft, Adam Jackson, Carole chaeological paradigms. M. A. Zeder, D. G. Bradley, E. Emshwiller, McCartney, and Mary Anne Murray. 2000. Agro-pastoralist col- and B. D. Smith, eds. Pp. 15–24. Berkeley: University of California onization of Cyprus in the 10th millennium BP: initial assessments. Press. Antiquity 74:844–853. Smith, Bruce D. 2011. The cultural context of plant domestication Peters, Joris, and Klaus Schmidt. 2004. Animals in the symbolic world in eastern North America. Current Anthropology 52(suppl. 4): of Pre-Pottery Neolithic Go¨bekli Tepe, south-eastern Turkey: a S471–S484. preliminary assessment. Anthropozoologica 39:179–218. Vigne, J.-D., I. Carre`re, J.-F. Saliege, A. Person, H. Bocherens, J. Piperno, D. R., A. J. Ranere, I. Holst, J. Iriarte, and R. Dickau. 2009. Guilaine, and F. Briois. 2000. Predomestic cattle, sheep, goat and Starch grain and phytolith evidence for early ninth millennium pig during the late 9th and the 8th millennium cal. BC on Cyprus: B.P. maize from the Central Balsas River valley, Mexico. Proceedings preliminary results of Shillourokambos (Parekklisha, Limassol). In of the National Academy of Sciences of the USA 106:5019–5024. Archaeozoology of the Near East IV. H. Buitenhaus, M. Mashkour, Piperno, Dolores R. 2011. The origins of plant cultivation and do- and F. Poplin, eds. Pp. 83–106. Groningen: Archaeological Re- mestication in the New World tropics: patterns, process, and new search and Consultancy. developments. Current Anthropology 52(suppl. 4):S453–S470. Vigne, Jean-Denis, Isabelle Carre`re, Franc¸ois Briois, and Jean Gui- Price, T. Douglas, and Ofer Bar-Yosef. 2010. Traces of inequality at laine. 2011. The early process of mammal domestication in the the origins of agriculture in the ancient Near East. In Pathways to Near East: new evidence from the Pre-Neolithic and Pre-Pottery power: new perspectives on the origins of social inequality. T. Douglas Neolithic in Cyprus. Current Anthropology 52(suppl. 4):S255–S271. Price and Gary M. Feinman, eds. Pp. 147–168. New York: Springer. Weiss, Ehud, Mordechai E. Kislev, and Anat Hartmann. 2006. Au- Price, T. Douglas, and Anne B. Gebauer, eds. 1995. Last hunters, first tonomous cultivation before domestication. Science 312:1608– farmers: new perspectives on the prehistoric transition to agriculture. 1610. Santa Fe, NM: School for American Research. Weiss, Ehud, and Daniel Zohary. 2011. The Neolithic Southwest Ranere, A. J., D. R. Piperno, I. Holst , R. Dickau, and J. Iriarte. 2009. Asian founder crops: their biology and archaeobotany. Current The cultural and chronological context of Early Holocene maize Anthropology 52(suppl. 4):S237–S254. and squash domestication in the Central Balsas River valley, Mex- Winterhalder, Bruce, and Douglas J. Kennett, eds. 2006. Behavioral ico. Proceedings of the National Academy of Sciences of the USA ecology and the transition to agriculture. Berkeley: University of 106:5014–5018. California Press. Rindos, D. 1984. The origins of agriculture: an evolutionary perspective. ———. 2009. Four neglected concepts with a role to play in ex- Orlando, FL: Academic Press. plaining the origins of agriculture. Current Anthropology 50:645– Rowley-Conwy, Peter. 2011. Westward ho! the spread of agriculture 648. from Central Europe to the Atlantic. Current Anthropology Zeder, M. A. 2006a. Archaeological approaches to documenting an- 52(suppl. 4):S431–S451. imal domestication. In Documenting domestication: new genetic and

Sage, R. F. 1995. Was low atmospheric CO2 during the Pleistocene archaeological paradigms. M. A. Zeder, D. Bradley, E. Emshwiller, a limiting factor for the origin of agriculture? Global Change Biology and B. D. Smith, eds. Pp. 171–180. Berkeley: University of Cali- 1:93–106. fornia Press. Sauer, Carl O. 1952. Agricultural origins and dispersals. Washington, ———. 2006b. Central questions in the domestication of plants and DC: American Geographical Society. animals. Evolutionary Anthropology 15:105–117. Schmidt, K. 2001. Go¨bekli Tepe, southeastern Turkey: a preliminary Zeder, Melinda A. 2011. The origins of agriculture in the Near East. report on the 1995–1999 excavations. Pale´orient 26:45–54. Current Anthropology 52(suppl. 4):S221–S235. ———. 2003. “Kraniche am See”: Bilder und Zeichen vom fru¨h- Zeder, Melinda A., and Bruce D. Smith. 2009. A conversation on neolithischen Go¨bekli Tepe (Su¨dosttu¨rkei). In Der Turmbau zu agricultural origins: talking past each other in a crowded room. Babel: Ursprung und Vielfalt von Sprache und Schrift: eine Aus- Current Anthropology 50:681–691. stellung des Kunsthistorischen Museums Wien fu¨r die Europa¨ische Zhao, Zhijun. 2011. New archaeobotanic data for the study of the Kulturhauptstadt Graz 2003, vol. 3A. Wilfried Seipel, ed. Pp. 23– origins of agriculture in China. Current Anthropology 52(suppl. 4): 29. Milan: Skira. S295–S306. ———. 2006. Sie bauten den ersten Tempel: das ra¨tselhafte Heiligtum Zohary, D., and M. Hopf. 2000. Domestication of plants in the Old der Steinzeitja¨ger: die archa¨ologische Entdeckung am Go¨bekli Tepe. World: the origin and spread of cultivated plants in West Asia, Europe, Munich: Beck. and the Nile Valley. 3rd edition. Oxford: Oxford University Press. Current Anthropology Volume 52, Supplement 4, October 2011 S175

Climatic Fluctuations and Early Farming in West and East Asia

by Ofer Bar-Yosef

This paper presents a Levantine model for the origins of cultivation of various wild plants as motivated by the vagaries of the climatic fluctuation of the Younger Dryas within the context of the mosaic ecology of the region that affected communities that were already sedentary or semisedentary. In addition to holding to their territories, these communities found ways to intensify their food pro- curement strategy by adopting intentional growth of previously known annuals, such as a variety of cereals. The Levantine sequence, where Terminal Pleistocene and Early Holocene Neolithic archae- ology is well known, is employed as a model for speculating on the origins of millet cultivation in northern China, where both the archaeological data and the dates are yet insufficient to document the evolution of socioeconomic changes that resulted in the establishment of an agricultural system.

Opening Remarks: Observations cieties in core areas and their reactions to abrupt climatic and Statements changes, require some brief statements and clarifications. Although history was not recorded during the onset of the Neolithic Revolution, the foundations laid by early farmers The Neolithic Revolution, or the Agricultural Revolution, was unexpectedly led in due course to the invention of writing a major “point of no return” in human evolution. After 2.6 systems. Indeed, in an evolutionary sense it was a set of rapid million years of hunting and gathering, low population den- socioeconomic changes. We should therefore refer to our data sities, and a series of dispersals to the edges of Eurasia and sets from the first half of the Holocene as the history of beyond to the Americas and Australia, human societies de- “people without names,” and like all historical documents veloped a new economic system that changed the course of they are amenable to different interpretations. prehistory. Instead of survival based on foraging with partial Investigations into the reasons why people became farmers or full-time residential and logistical mobility, the appearance are not new. Since the nineteenth century, scholars have and disappearance of sedentary and semisedentary commu- searched for “where,” “when,” “how,” and “why” cultivation nities, and episodes of genetic “bottlenecks,” the first farming became an attractive survival strategy. Botanists such as Al- villages in a limited number of regions led unintentionally to fonse de Candolle, Nikolai Vavilov, and Jack Harlan explored a revolutionary social upheaval. Although a few Late Pleis- the natural habitats where the wild progenitors of domesti- tocene hunter-gatherers developed low-level food production, cated plants were found, assuming that these were the “cen- the people who started cultivating the wild species of the ters” of plant domestications. The desire to find primordial locations is still pertinent, as recent investigations in the Le- “winning” plants—those that feed the world of today (wheat, vant (e.g., Zohary and Hopf 2000) or the efforts to find the barley, rice, and corn)—were in retrospect responsible for the location of millet progenitors in East Asia (e.g., Lu et al. 2009) rapidly increasing world population that led to the Industrial demonstrate. Revolution. Historically, these natural habitats or their margins were Before delving into the particular cases of West and East thought by archaeologists to have been the “core areas” where Asia (fig. 1), where I believe the earliest manifestations of this cultivation and domestication occurred (e.g., Binford 1968; socioeconomic revolution were triggered by the impact of the Braidwood 1952; Flannery 1973). What seems to be a feasible Younger Dryas (YD), several issues, such as small-scale so- research target in the Levant because of its relatively small dimensions and the large number of different schools of ar- Ofer Bar-Yosef is George Grant MacCurdy and Janet G. B. MacCurdy chaeology involved in fieldwork is more complex in East Asia, Professor of Prehistoric Archaeology in the Department of where investigations began almost half a century later and Anthropology, Harvard University (Cambridge, Massachusetts where the core area is at least two to four times as large. 02138, U.S.A. [[email protected]]). This paper was submitted Hence, using the primacy of the Levant, I discuss the role of 13 XI 09, accepted 17 II 11, and electronically published 24 VIII 11. “core areas” and refer to places into which agricultural systems

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0003$10.00. DOI: 10.1086/659784 S176 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 1. Regions discussed in the text: the Levant and China. were transmitted to as “secondary core areas.” For example, In this paper, I suggest that at the two ends of the Asian the Levant was the “core area” where the cultivation of wild continent, a climatic change was the main trigger for the onset cereals, also known as “predomestication cultivation” or “low- of cultivation (i.e., growing the progenitors of wheat, barley, level food production” (e.g., Smith 2001) began and where rye, millet, and other grains) in a context of “low level food these plants became the cultigens after more than 1,000 years production” (Smith 2001). This either evolved into major of farming. These domesticated crops were introduced into food production or failed (e.g., Weiss, Kislev, and Hartmann Europe and the Nile Valley, both “secondary centers.” 2006; Weiss and Zohary 2011). Hence, before delving into the The archaeological evidence for the initiation of cultivation, climatic, social, and economic evidence acquired from ar- the corralling of wild herd animals, and the ensuing trans- chaeological investigations, we need to examine briefly the mission of knowledge, techniques, plants, and animals, as well past human experience in the face of abrupt climatic change. as voluntary and/or forced movements of humans among More than once during the Pleistocene, humans faced en- village-based societies, is characteristic of interregional con- vironmental changes that improved, worsened, or did not nections. Short- and long-distance interactions did not always affect their basic ecological conditions. While we sometimes occur under stable environmental conditions. Paleoclimatic tend to see these climatic changes as the causes for the ex- proxies from the Terminal Pleistocene and Early Holocene tinction of particular populations or eventual dispersals, the record numerous fluctuations. Even minor shifts in local con- overall picture is too coarse-grained. History, however, has ditions, given the available technology and social structure of taught us different lessons. Slow or abrupt shifts of environ- human groups, could make the difference between successful mental conditions that resulted in social and economic dis- biological survival and failure. asters have been recorded (e.g., Bell 1971; Hassan 2002; Shen Bar-Yosef Early Farming in the Levant and China S177 et al. 2007; Weiss and Bradley 2001). The impacts of droughts, or increasing the distance of forays by task groups if neigh- harsh winters, flooding by rivers or the sea, failing harvests, boring territories are less affected, thereby facing foes or famines, wars, spreads of disease, disturbances of the social friends (in the best situation, kin relations); (b) to stay put order, and the demise of peoples are reasonably well docu- and defend the reliable natural resources within their im- mented. Some societies were able to cope with or to minimize mediate territory; or (c) to actively intensify the yield of avail- disastrous effects, while others collapsed (Rolett 2008). The able plant resources through technological improvements, challenge for archaeologists is to find how humans handled new inventions, part-time cultivation, the tending of fruit a situation and either succeeded or failed in coping with the trees, and even the corralling of herd animals as pets and impacts of natural calamities (Rosen 2007). When biological food. The sum of all these traits, or only a few, provide the survival is critical for individuals or a group of people—be “cultural filter” through which environmental changes are it a band, a macroband, or a tribe (using an approximate mitigated by the affected society or negotiated with its neigh- scale of population size; see Binford 2006; Hawkes and Paine bors, whether they belong to the same culture (and speak the 2007; Marcus 2008; Roscoe 2008, 2009)—they will use a num- same language or the same dialect) or to a different society. ber of strategies within their knowledge, depending on the Many archaeologists are reluctant to accept the notion that resilience of their social structure and cultural concepts, to climatic changes expressed in environmental degradation, re- insure their physical existence in the world. flected in climatic proxies, and seemingly of the same date of The capacity to minimize risk, for example, can be in- an observed cultural “break” in the archaeological sequence creased by actively intervening in the natural environment were the cause for turnover or crisis. They are right. Chro- when K-selected (e.g., large-bodied mammals) and r-selected nological correlation is not causation. Employing approxi- (e.g., small-bodied mammals such as hare and tortoise, as mate chronological correlations as the basis for proposed in- well as fruits and seeds) resources decrease or when required terpretations is merely a hypothesis to be verified or falsified. foray distances increase. This can be accomplished by inten- This paper relies on that approach but advocates, even if it sifying food-acquisition techniques such as tending particular is not yet completely feasible, that sound interpretations plants, using fire to enhance the growth of wild stands of should be grounded on precisely dated paleoclimatic infor- herbaceous vegetation, digging simple irrigation canals, and mation obtained from archaeological deposits and their con- in some cases being involved in “low-level food production” tents (e.g., plants and animal bones). In addition, it should (e.g., Binford 2006; Denham et al. 2003; Kelly 1995; Lewis be stressed that proposals by nonarchaeologists that abrupt 1972; Lourandos 1997; Roscoe 2009; Smith 2001). climatic changes resulted in a socioeconomic shift should in- The ability to handle the complexity of the social organi- corporate an anthropologically oriented explanation for re- zation through the articulation of its members and the spatial sponding to the “why” question. However, this is rarely done. distribution of the mating system under stressful circum- Shying away from the impacts of climatic changes has stances is another strategy. The increased social cohesion by marked the archaeological literature of the last decades. Pref- increasing group size (agglomeration) or reducing (fissioning) erence has been given to explanatory models that involved in face of natural disasters includes the perception of home social changes emanating from intra- and intersociety political territory and its ranges, incorporation of sacred localities, and pressures, physical conflicts and wars, rapid population the degree of preparedness of the group to pay for costly growth, and fissioning of settlements attributed to “scalar signaling of their ownership (e.g., Roscoe 2006). stress” (Johnson 1982) or the “Irritation Coefficient” (Rap- The level of preparedness of those who inhabit ecologically paport 1968; cited in Bandy 2004), although New Guinean marginal zones—for example, those with increased frequency ethnography offers an alternative model (Roscoe 2009). In- of droughts or prolonged cold winters—when facing natural deed, the main question is how did foragers and farmers disasters is tested when and if they are ready to change or minimize the risk to survival (Rosen 2007)? successfully move out (Kawecki 2008). Much depends on the In studying human responses to environmental disasters, survival skills, ability to mitigate relations with neighbors, and an important source of information comes from recent in- socioeconomic information transmitted through the “living ternational relief programs, which are forced to identify the memory” of the social unit (e.g., Minc and Smith 1989; Rosen specific interactions between populations affected by calam- 2007). ities such as earthquakes, droughts, or floods in their im- The ability of the population to sustain its biological sur- mediate environments and within given political regimes. vival for future generations within a large geographic region Some reactions are common across the human spectrum, is tested in times of stress. The intervariability of the social while others express the individual nature of the impaired organization, alliances with neighboring groups, and the his- human group. In each case, anthropologists discover that the tory of conflicts often play an important role in adaptation outcome resulted from previously held beliefs, social and po- to new conditions, although recovering all this information litical organization, and historical interactions with neigh- from the archaeological record is difficult (e.g., Marcus 2008). boring societies (e.g., Glantz et al. 1998). As a result, when disasters strike, the options for all foragers Currently, information concerning changing past climates are (a) to increase mobility by moving to another territory relies on speleothems, terrestrial and marine pollen cores, and S178 Current Anthropology Volume 52, Supplement 4, October 2011 supporting evidence from long-distance correlations with the determine their 18O content (see, e.g., Kolodny, Stein, and Greenland ice core (Greenland Ice Sheet Project 2 [GISP2]) Machlus 2005). Subregional ecological variability is fully ex- for the following reasons. (1) Speleothems are found in many pressed in the Levant because precipitation, the determinant caves (whether occupied by humans or not) located in dif- factor, decreases dramatically from west (the sea) to east (the ferent types of environments (forest, parkland, steppe, arid). desert) and from north (the foothills of the Taurus and moun- Mass spectrometers produce dates with calendar ages, and the tain areas) to south (the Negev and Sinai). Thus, a paleo- basic isotopic information (d18O and d13C) traces the sources ecological mosaic of habitats should be taken into account and the amounts of local precipitation (e.g., Lachniet 2009). when sites are discussed. (2) Lake and marsh pollen cores provide information con- The speleothems, despite disagreements concerning the cerning vegetational fluctuations in their drained basins but precise conversion of their fluctuating d18O and d13C contents do not always produce precise radiocarbon dates. Correlations to the amount of annual rainfall, reflect past centennial and with marine pollen cores may refute or substantiate the ter- millennial fluctuations (e.g., Bar-Matthews and Ayalon 2003; restrial sequences. (3) Ice cores provide the most detailed Frumkin, Ford, and Schwarcz 2000; Vaks et al. 2003). Spe- information concerning climatic changes, and studies of prox- leothem data from Soreq Cave, on the western flanks of the ies from other regions compare their results to GISP2. How- central hilly ridge, and Ma’aleh Efriam Cave, on the eastern ever, in regions farther from the Arctic, the role of local con- flanks (Bar-Matthews et al. 1999; Vaks et al. 2003; fig. 2), ditions increases and time correlations become more tentative compare well with marine and lake cores from Greece and and less secure, although the major trends of climatic fluc- the Aegean Sea (e.g., Rohling et al. 2002). tuations remain the same. (Examples of each are given below.) The Late Glacial Maximum (LGM; ca. 24,000–18,000 cal Taking the above comments into account, I try to demonstrate below that the conditions imposed by the YD triggered cul- tivation in two subregions in West and East Asia (fig. 1).

The Levantine Paleoclimatic Sequence and the Onset of Intentional Cultivation The Levant is a region located on the edge of the eastern Mediterranean basin, geographically bordered by the Taurus Mountains on the north, the Mediterranean Sea on the west, the Syro-Arabian desert on the east, and the Sinai Desert on the south (Cauvin 2000; also see Belfer-Cohen and Goring- Morris 2011; Goring-Morris and Belfer-Cohen 2011; Vigne et al. 2011; Weiss and Zohary 2011; Zeder 2011). Its optimal habitats for exploitation are within the Mediterranean and Irano-Turanian (steppic) vegetational belts, which stretch in parallel from north to south, have variable topography, and are watered from west to east in decreasing amounts of annual precipitation. Every model of optimal foraging should take these conditions into account. Under favorable climatic con- ditions, the Levantine flora and fauna expand mainly to the north, along the foothills of the Taurus and the Zagros arc and beyond the Euphrates, the Balikh, the Khabur, and the Tigris rivers (an area called “Upper Mesopotamia” that bears historic connotations irrelevant to the prehistoric northern Levant). The paleoclimatic information is derived from cave spe- leothems, Mediterranean marine pollen, and lake cores. The current interpretation of the terrestrial pollen cores is based on the correlations of vegetational zones within the marine cores (Rossignol-Strick 1995; van Zeist, Baruch, and Bottema 2009). Latitudinal and longitudinal shifts in several atmo- spheric systems dictate changing climatic patterns in this re- gion (e.g., Enzel et al. 2008). Most of the storm tracks that Figure 2. Paleoclimatic changes based on Soreq Cave speleothem transport the winter rain to the region originate in the Atlantic (after information from M. Bar-Matthews and A. Ayalon, with Ocean and cross the Mediterranean along different paths that permission). Bar-Yosef Early Farming in the Levant and China S179

Figure 3. Chronological chart of Levantine late prehistory.

BP), as elsewhere, was a cold and dry period in the Levant, (Goring-Morris 1995; Goring-Morris and Belfer-Cohen 2011; and its conditions are reflected in decreasing precipitation fig. 3). More or less at the same time, the Mushabian and over the drainage basin of Lake Lisan, which began shrinking Ramonian entities were competing/coexisting with the Lev- (Bartov et al. 2002). While this period witnessed a discernible antine foragers (Goring-Morris and Belfer-Cohen 2011). One reduction of global human populations in the higher latitudes, interpretation suggests their origins in Northeast Africa, while it affected the eastern Mediterranean less. another proposal has them originating from local Levantine A rapid post-LGM rainfall increase is recorded in speleo- groups. It is conceivable that additional groups of hunter- thems, marine pollen cores, and lake pollen cores in the Hula gatherers were attracted by the improving ecological condi- and the Ghab valleys (van Zeist, Baruch, and Bottema 2009). tions and penetrated the eastern Levantine marginal areas These sources demonstrate that the return of wetter condi- through the Syro-Arabian desert and/or the Taurus foothills. tions from ca. 16,500 to 14,700/14,500 cal BP facilitated a By the end of these several millennia, there were groups of pan-Levantine distribution of foragers, bearers of the mi- mobile foragers everywhere in the Levant. Most intriguing is crolithic Geometric Kebaran industry in every ecological hab- the question of whether a short, cold climatic episode (known itat from the northern Levant through the Sinai Peninsula in Europe as Dryas I) caused a temporary retraction of the S180 Current Anthropology Volume 52, Supplement 4, October 2011 steppic belt and triggered a relative “demographic pressure.” ditions in the Levant probably began somewhat later, ca. It is at that point in time that certain groups established the 12,600/12,500 cal BP (e.g., Mayewski et al. 2004; Rohling et Early Natufian hamlets, although none are as yet known in al. 2002; Sima, Paul, and Schultz 2004). Thus, the duration the northern Levant (e.g., Bar-Yosef 2002; Bar-Yosef and Bel- of the YD in the Levant is still unresolved, but it could have fer-Cohen 1989; Henry 1989). This initial formation of hu- been shorter than that indicated by the ice cores and as long man agglomerations, departing from the old lifeway of resi- as that in western Europe or eastern China (Liu et al. 2008), dential mobility by building pit houses and burying the dead that is, about 1,000 years. on site (Belfer-Cohen 1995) and by forming sedentary or The Ghab pollen core in the Orontes River valley dem- semisedentary permanent camps, is indicated by the presence onstrates a clear decline of arboreal pollen during the YD, of commensals such as house mice, rats, and sparrows (Tcher- with a major recovery of the oak-pistachio forests during the nov 1991). Accommodating a few families or even subclans Early Holocene (van Zeist, Baruch, and Bottema 2009; van within a settlement marked territorial ownership, expressed Zeist and Bottema 1991; Yasuda 2002). The reduction in ar- in intrasite cemetery areas, enabling the group to achieve a boreal pollen is less well marked in pollen cores from Ana- sense of security and defense either by force or by symbolic tolian lakes located in the steppic areas or from the coast of acts (see Roscoe 2009 and references therein). Mount Carmel (Kadosh et al. 2004; Lev-Yadun and Weinstein- The success of the Natufian as a well-organized society of Evron 2005). Localities close to the Mediterranean Sea were foragers could be related to the wetter and warmer climatic generally forested, and even a reduction of ca. 30% of annual conditions of the Bølling-Allerød (ca. 14,500–13,000/12,800 precipitation had a minimal ecological impact. cal BP). The establishment of the Early Natufian sites is what The proxy data from marine cores across the eastern Med- we once referred to as a “point of no return” (Bar-Yosef and iterranean—from the Adriatic Sea, the Aegean, and Cyprus— Belfer-Cohen 1989; Belfer-Cohen and Bar-Yosef 2000). Their support the overall picture described above. The temperature small villages and hamlets were constructed from a series of cline from the Atlantic Ocean through the Mediterranean, brush huts built over circular stone foundations, sometimes shown by the analysis of planktonic foraminifera (Kuhlemann with wooden poles. Larger structures could have served for et al. 2008), demonstrates the general time correlations of special uses such as rituals and were the forerunners of the climatic fluctuations between the two water bodies. Com- Pre-Pottery Neolithic A (PPNA)–Pre-Pottery Neolithic B parisons between the oxygen and stable carbon isotopes from (PPNB) “kiva”-type subterranean buildings (e.g., house 131 cave speleothems and those from the marine cores indicate in Eynan [Valla 1989]; Stordeur and Abbe`s 2002; Stordeur that the same sequence of climatic changes occurred in the and Iba´n˜ez 2008). The on-site burials (often of different so- Levant (Bar-Matthews and Ayalon 2003). Hence, in spite of cially unexplained styles; Belfer-Cohen 1995); the rich lithic, probable attenuation due to local conditions, proxies from bone, and horn-core objects, including incised items; the nu- neighboring regions within the Northern Hemisphere can be merous marine shells and stone beads for body decoration; employed (e.g., Enzel et al. 2008; Willcox, Buxo´, and Herveux animal figurines; incised slabs; and much more are among 2009). Missing from the proposed paleoclimatic interpreta- the markers of this culture. Our knowledge of Natufian econ- tions are the simulated impacts of reduced precipitation on omy is limited to hunted, trapped, and gathered mammals, the Mediterranean forests, open parklands, and steppic en- birds, and reptiles, with evidence of fishing in some sites. It vironments. The overall trend was marked by diminishing is, however, obvious from sickle blades that bear the special yields of wild plant seeds and annual fluctuations in the pro- sheen resulting from harvesting cereals and the mortars in duction of acorns and pistachio nuts, which were intensively which cereals were processed, in addition to cup holes and collected or harvested by Terminal Pleistocene foragers, as rare grinding slabs, that a considerable amount of plant food well as changes in the spatial distribution of game animals. was consumed. Given the wealth of literature concerning the Thus, the impact in a land “full of people” was that those Natufian culture and the variable social interpretations, only who occupied the better areas probably stayed put and in- a few selected references are mentioned here in addition to tensified their food acquisition. Others increased their mo- those above (e.g., Bar-Yosef 1998, 2002; Bar-Yosef and Valla bility. 1991 and references therein; Belfer-Cohen and Bar-Yosef 2000; The Late and Final Natufian sites in the southern Levant Belfer-Cohen and Goring-Morris 2011; Belfer-Cohen and (ca. 13,000/12,800–11,700/11,500) produced poorer remains Hovers 2005; Byrd 2005; Cauvin 2000; Edwards 2007; Goring- than the Early Natufian. It probably reflects the return in Morris and Belfer-Cohen 2011; Grosman, Munro, and Belfer- many of the exploited habitats to a mobility greater than their Cohen 2008; Henry 1989; Munro 2004; Price and Bar-Yosef ancestors’. Flimsy dwelling structures characterize these sites, 2010; Valla 1995, 1999; Valla et al. 2007; Weinstein-Evron and the dead were rarely buried with adornments, but rich 2009). lithic and bone-tool assemblages were produced (Valla et al. Currently, the radiocarbon dates for the cultural transition 2007). The Late and Final Natufians increased consumption from the Early to the Late Natufian indicate that it occurred of low-ranked resources such as bone grease, juvenile gazelles, before the climatic crisis of the YD, which started ca. 13,000/ and fast-moving small game such as hare (Munro 2004; Stiner, 12,800 cal BP in the northern latitudes. The worsening con- Munro, and Surovell 2000). Two mounds, Abu Hureyra Bar-Yosef Early Farming in the Levant and China S181 (Moore, Hillman, and Legge 2000) and Tell Mureybet (Iba´n˜ez PPNA Communities and Early Farming 2008), provided the only Late Natufian archaeobotanical as- semblages to be discussed below. In sum, in the face of the The PPNA communities (ca. 11,700/11,500–10,700/10,500 cal difficulties of the YD, the solutions triggered for risk mini- BP) are considered the direct descendants of the Natufians, mization by Late Epi-Paleolithic societies were diverse and although we lack evidence of their contemporaries who in- included the following: habited southeast Turkey because of a paucity of research, 1. Increased mobility and additional specialized adaptations and they invested more energy and materials than their fore- to the mosaic ecology of the Levant. In the Negev and the bears in building houses. Circular and oval stone foundations northern Sinai, these led to the emergence of the unique continued to be the standard shape of the domestic unit, but Harifian culture and the invention of a typical arrowhead, the quarrying of clay and hand-molding of planoconvex bricks Harif Point (Goring-Morris 1991; see Belfer-Cohen and Gor- for the walls, as well as the mounting of flat roofs that required ing-Morris 2011; Goring-Morris and Belfer-Cohen 2011). supporting posts, represent an increased investment in cre- 2. Increased sedentism for security and defense from other ating the human space (Stordeur and Willcox 2009; Watkins noncultivator groups of foragers. This is demonstrated in the 2006). Private and public storage facilities were erected (Kuijt and Finlayson 2009). The villages grew up to 2.5 ha in size, establishment of the village of Hallan C¸emi Tepesi (11,900– with populations of at least 150–300 people practicing a mixed 10,500 cal BP) on the banks of a tributary of the Tigris River economy of cultivating different suites of plants, according (Rosenberg and Redding 2000). No cereals were found at this to their local ecology, and fig trees in the Jordan Valley (Kislev, site (Savard, Nesbitt, and Jones 2006), indicating that during Hartmann, and Bar-Yosef 2006). Hunting the common game the YD the distribution of einkorn and barley did not extend in the area and gathering wild plants provided a major part farther east from its main western habitat. Barley (Hordeum of the diet (Willcox, Fornite, and Herveux 2008). cf. spontaneum) makes its first major appearance in this region Interpretations of the archaeobotanical data indicate that farther east (Demirko¨y and Qermez Dereh) only during the initiation of intentional cultivation varied. In each subregion, PPNA (after 11,700/11,500 cal BP; Savard, Nesbitt, and Jones a different set of wild plants were cultivated and were either 2006). successful or total failures (Weiss, Kislev, and Hartmann 3. Intensified hunting and gathering and part-time culti- 2006). The first experiments in cultivation could have begun vation. This is evident in the presence of arable weeds, re- during the Early Natufian, but a step forward was made during flecting increased sedentism, and it emerges in Tel Qaramel the Late Natufian (Hillman 2000; Hillman et al. 2001). Most west of the Euphrates valley, Mureybet, and Jerf el-Ahmar authorities agree that during the closing centuries of the YD (Willcox, Buxo´, and Herveux 2009; Willcox, Fornite, and Her- or the first centuries of the Holocene, bearers of the earliest veux 2008). Hence, wild cereals were available only along the PPNA tool kits, defined as the Khiamian culture in the north- western wing of the Fertile Crescent, as predicted by the con- ern Levant, were the first farmers, because their carbonized ditions of the YD and the plant remains from the Late Na- plant remains contain the weeds that grow in tilled fields in tufian at Abu Hureyra and Mureybet (Hillman et al. 2001). addition to the cereals (Colledge 2001; Kislev, Hartmann, and Exploitation of small seeds was already known from the days Bar-Yosef 2006; Willcox, Buxo´, and Herveux 2009; Willcox, of Ohalo II (ca. 23,000–21,000 cal BP) and probably from Fornite, and Herveux 2008). The suite of plants grown by the earlier times as well. The decision to include the cultivation first cultivators included rye (Secale cereale), einkorn (Triticum of cereals in the economy of these foragers seems to have boeoticum), emmer wheat (Triticum dicoccoides), barley (Hor- started in the northern Levant, probably before the end of deum spontaneum), and oats (Avena sterilis). Several grass the YD (ca. 11,700/11,500 cal BP), as in Tel Qaramel (Ma- species, such as Aegilops and Stipa, may represent wild weeds zurowski, Michczynska, and Padzur 2009 and references that grew in cultivated fields or the results of gathering. Pulses therein; Willcox, Buxo´, and Herveux 2009), and the idea such as lentils (Lens culinaris), peas (Pisum sativum), grass spread rapidly southward. It has been suggested that the first peas (Lathyrus), bitter vetch (Vicia ervilia), and common appearance of green beads among Late Natufian body dec- vetch (Vicia sativa) are well recorded, while chickpeas (Cicer orations marked the onset of beliefs related to the practice of arietinum) and faba beans (Vicia faba) first appeared during cultivation (Bar-Yosef Mayer and Porat 2008 and references the PPNB (Lev-Yadun, Gopher, and Abbo 2002). Currently, therein). Indeed, botanical evidence indicates that within a the prevailing view is that systematic cultivation was carried few centuries, the climatic conditions improved and farming out by several PPNA villages, including Tel Qaramel, where became successful because there were stable and sufficient all the 14C dates were produced by one laboratory (Mazu- amounts of winter rain (e.g., Willcox, Buxo´, and Herveux rowski, Michczynska, and Padzur 2009). 2009; Willcox, Fornite, and Herveux 2008). The ensuing mil- According to several archaeobotanists, it took some 1,000 lennia of the Holocene enjoyed better climatic conditions, in or 2,000 years of systematic cultivation of wild cereals (Fuller spite of rapid climatic changes that had variable impacts on 2007; Kislev 1989, 1997; Tanno and Willcox 2006) before a human communities as population increased and social struc- major portion of the plants acquired the mutation of non- ture became more complex (Weninger et al. 2009). shattering ears and increased their grain size. This means that S182 Current Anthropology Volume 52, Supplement 4, October 2011 without the continuous activities of sowing and harvesting, origins of millet cultivation is focused in a large area of about the domesticated plants would not have taken over in the 500,000 km2, incorporating the middle and lower Yellow River fields. This is a process whose meaning is not fully understood basin (Cohen 2011; Zhao 2004). The number of sites where by some archaeologists, for whom the terms “agriculture” and plant remains have been carefully recovered and reported is “cultivation” are interchangeable (e.g., Hodder 2007). An ad- still small, and the cultural relationships among the different ditional marker of intentional cultivation is the presence of subregions is debated among archaeologists (Cohen 2011). typical wild weeds that grow annually in cultivated and har- Hence, we must first consider the overall geographic features vested fields (Willcox, Buxo´, and Herveux 2009). Therefore, of China and the current climate and then proceed to sum- the domesticated cereals that characterized the agricultural marize the proxies for past climatic fluctuations before delving economy of PPNB villages were the result of a prolonged into the particular information concerning their impact on period of cultivation. One should refer to the historical use the local hunter-gatherers. of terms such as “cultivation,” “domestication,” “agriculture,” The physiography of China (ca. 9.6 million km2) is com- and others as clearly presented by Harris (2007). If the def- monly subdivided into three topographic landforms, each inition of “cultivation” incorporates the entire set of activi- with its own regional variability. These are defined by elevation ties—such as tillage, sowing, irrigation, harvesting, and stor- above sea level (a.s.l.): (1) the Tibetan Plateau, some 4,000– age of seeds for consumption and next year’s planting—then 5,000 m a.s.l.; (2) the central mountain plateau area, ca. 1,000– regardless of the genetically determined morphological traits 2,000 m a.s.l., incorporating Inner Mongolia, the Loess Pla- of the plants, early cultivators were simply farmers. One may teau, the Sichuan Basin, and the Yunnan-Guizhou Plateau; argue whether this is a fully “agricultural” subsistence system and (3) the plains and seacoast, generally below 200 m a.s.l. or an indication of “low-level food production” or that the and crossed by numerous copious rivers. This region is strewn definition should be retained for societies where husbandry with hilly areas, mostly south of the Yangtze River, that can of animals was part and parcel of annual subsistence activities reach ca. 500 m a.s.l. (Zhao 1994). (Vigne et al. 2009 and references therein). But farmers who The climate of China is characterized by the tropical and grow tubers that are not fully domesticated are classified as subtropical Pacific and Indian Ocean summer monsoons. The “agriculturalists” and/or “horticulturalists,” because it is the arrival of the monsoon marks the onset of the rainy season, practice and not the state of “domestication” of the plants starting in the south and advancing northward from early that counts when the economic system is categorized. March to June–July (fig. 4). Later, the rains retreat to the Finally, all PPNA villages in the Levant show the same south and may last from late August through September and crowded clustering that a millennium later became the hall- October. During the winter, the entire landmass is dominated mark of several PPNB sites. Calibrated radiocarbon chro- by Siberian-Mongolian high-pressure systems that often pro- nology—mostly derived from short-lived, site-by-site sam- duce strong winds. But the winter monsoon carries some ples—indicates that almost every village, including those moisture from the Pacific into eastern China, and the north- situated next to a copious spring, like Jericho, or along the west enjoys the westerlies that bring some precipitation from river banks, as in the Euphrates valley, was abandoned within western Eurasia. Winter temperatures are close to or below a few centuries. 0ЊC in the north, while summer temperatures may rise to above 30ЊC, particularly in the south, and higher in the west- Terminal Pleistocene and Early Holocene ern deserts. Topographic variability within each of the sche- Climatic Fluctuations in China matically averaged levels results in a mosaic distribution of precipitation and temperature and thus of flora and fauna The basic assumptions for discussing the origins of cultivation (Zhao 1994). in China are the same as for the Levant, namely, that Late The Last Glacial in China was characterized by significant Pleistocene–Early Holocene climatic fluctuations in North and frequent oscillations well recorded in a suite of proxies China played a similar role as in the Levant, triggering the such as the Himalaya ice cores, loess sediments, pollen cores, transition to cultivation of wild millets for the intensification marine cores, and cave speleothems (e.g., Cosford et al. 2008; of a staple food. In this argument, I follow the footsteps of Lin et al. 2006; Yu, Luo, and Sun 2008; Yu et al. 2000; fig. others who have already suggested, either partially or fully, 4). Unfortunately, the distribution of caves with studied spe- the relationship between the impact of the cold and dry YD leothems is uneven, and most are located in southeastern and conditions on the survival strategies of mobile foragers and central China and in Tibet, where karstic landscapes prevail the primacy of millet cultivation (e.g., Barton et al. 2009; (fig. 4). Hence, most of the paleoclimatic information for the Bettinger, Barton, and Morgan 2010; Bettinger et al. 2007; Lu western and northern regions are drawn from lake and marine 1999, 2006; Shelach 2000). The emergence of rice cultivation, pollen cores, loess sequences, and deep-sea cores in the East which followed during the Early Holocene (Cohen 2011; Zhao China Sea (e.g., Wen et al. 2010; Yi et al. 2003). The different 2010, 2011), is briefly discussed below in relation to climatic data sets reflect the impacts of the Pacific and Indian Ocean fluctuations and resources in South China. monsoons and show some differences between the strengths It is important to stress that the ongoing search for the of the two systems as well as the impact of the westerlies. Figure 4. Location of Chinese caves with studied speleothems mentioned in the text. S184 Current Anthropology Volume 52, Supplement 4, October 2011

Among the eastern sites, Hulu Cave, near Nanjing (Wang Bølling-Allerød increased the potential for hunter-gatherers et al. 2001), produced a long paleoclimatic curve that best to expand their populations into previously arid or semiarid fits the GISP2 ice core. Other caves include Dongge (Dykoski habitats in western China. Thus, the abrupt change to the et al. 2005), Heshang (Hu et al. 2005), Qingtian (Liu et al. YD (e.g., Liu et al. 2008), from around 12,800/12,500 cal BP 2008), Sanbao (Wang et al. 2008), Songjia (Zhou et al. 2008), until 11,700/11,600 cal BP, was a major calamity. This natural and Timta in Tibet (Sinha et al. 2005). It should be stressed crisis provides early testimony to the problems that North that all speleothem sequences demonstrate similar trends but China has faced through history from fluctuations in the that not all correlate well chronologically. In addition, the monsoonal system (Shen et al. 2007). transition from one climatic stage to another (e.g., from the Allerød to the YD) took 1 or 2 centuries longer than the The Role of the YD in North China comparable transition recorded by the Greenland ice cores (Liu et al. 2008). However, detailed discussion of these issues Understanding the impact of the rapid climatic change of the is beyond the scope of this paper. YD on social systems is first appreciated from historical rec- LGM conditions in North China were cold and dry, and ords. The absence or paucity of summer rains is not an iso- except for a few protected habitats, most of these steppic- lated phenomenon. This can be seen in every historical review desertic environments were desolate landscapes (Yu et al. that documents droughts; droughts begin in the north, as a 2000). By ca. 16,000 cal BP, climatic amelioration was wit- recent review of major droughts during the past five centuries nessed in slowly rising temperatures, increasing rainfall, and in China indicates (Shen et al. 2007). The effects were dra- a moderate return of forest habitats to the loessic areas, as matic, because anthropogenic activities had already altered judged from the evidence for the earliest Holocene (e.g., Cai the local environments. Exceptionally severe droughts oc- et al. 2010; Ren and Beug 2002). As the monsoon system curred in AD 1586–1589 (when Taihu Lake, the third-largest became stronger, it penetrated farther north, particularly dur- freshwater lake in China, dried up), in AD 1638–1641, and ing the Bølling-Allerød (ca. 14,500–ca. 13,000/12,800 cal BP), in AD 1965–1966. More frequent droughts were recorded in and facilitated the spread of foragers within this region. tree rings in the Tien Shan area (Li et al. 2006). All these Several researchers have reported information concerning events were caused by a weak summer monsoon, together the environmental conditions that facilitated the growth of with the westward and northward movement of the western human populations, producers of the microblade (micro- Pacific subtropical high. lithic) industries, before the YD (e.g., Bettinger, Barton, and We therefore expect a sudden increase in dryness across Morgan 2010; Bettinger et al. 2007; Chen 2007; Madsen et North China to have caused the same reactions as observed al. 1998; Wu¨nnemann et al. 2007). Most sites from this period in the Levant. Unfortunately, the archaeological literature is are small and ephemeral and reflect varying degrees of mo- not sufficiently detailed, and accelerator mass spectrometry bility. Series of such occupations with microblade industries (AMS) dates for localities where humans stayed during the have been sampled and studied (e.g., Chen 1984, 2007; Cohen YD are rarely available. On the other hand, we have more 2003; Madsen et al. 1998). Xiachuan is one of the important information concerning Early Holocene conditions, including clusters of sites where a few radiocarbon dates indicate the reconstructed vegetation maps based on pollen records. These presence of at least two major occupations rich in microblades may help us speculate about what happened during the pre- (ca. 25,000 and ca. 15,000 cal BP) and a large number of vious period (e.g., Ren and Beug 2002; Wen et al. 2010). grinding slabs (Lu 1999). Lu (2002) reports siliceous sheen Hunter-gatherers retreated to more favorable habitats, in- on several flakes that resemble her experimental pieces em- cluding river valleys, as in Shizitan, and probably established ployed in harvesting foxtail grass panicles. Another multilayer semisedentary communities and increasingly intensified ex- cluster of sites excavated at Shizitan, where occupational ho- ploitation of resources arising from their reduced mobility, rizons were interspersed with loess accumulations several me- causing “population pressure” and increased competition for ters thick (Shizitan Archaeological Team 2002, 2010), is dated resources. Hence, during the YD and in particular during the to ca. 20,000–ca. 9,000 cal BP. On the whole, microblade first two millennia of the Holocene (11,500–9500 cal BP), we industries occur at several hundred sites across North China, note the appearance of larger sites as agglomerations of fam- southern Siberia, Korea, and Japan (Chen 2007; Kajiwara ilies and possibly subclans reflecting the need for security and 2008; Kuzmin, Keates, and Shen 2007). Larger tools were often territorial defense, in view of real or imaginary enemies (Ros- made from local raw material, such as quartz or quartzite. coe 2009). Earlier ephemeral occupations in sites such as Grinding slabs and rubbing stones are a common component Donghulin, Nanzhuangtou, and Zhuannian date to the YD in these sites, indicating small-seed grass processing. In ad- and/or Early Holocene and are of variable sizes (Cohen 2011). dition to plant resources, this vast region, dissected by the These are sites of foragers who successfully survived in the large Yellow River valley and numerous smaller ones, was region (fig. 5). frequented by several species of deer, equids, wild boar, and Nanzhuangtou (Hebei) contained some rare microblades, a few carnivores. no pottery, and a rich bone and antler assemblage, including It seems that the penetration of the westerlies during the the remains of deer, dog, pig, wolf, chicken, softshell turtle, Figure 5. Partial map of China with all the early farming sites and the locations of several of the main caves with speleothems. The list of sites is from Cohen (2011). 1, Yuchanyan; 2, Chengtoushan; 3, Pengtoushan; 4, Bashidang; 5, Xianrendong and Diaotonghuan; 6, Shangshan; 7, Kua- huqiao; 8, Xiaohuangshan; 9, Hemudu and Tianluoshan; 10, Dadiwan; 11, Shizitan; 12, Xiachuan; 13, Jiahu; 14, Peiligang; 15, Cishan; 16, Yue- zhuang; 17, Xiaojingshan; 18, Houli; 19, Nanzhuangtou; 20, Yujiagou; 21, Zhuannian; 22, Xinglongwa; 23, Jiahu. S186 Current Anthropology Volume 52, Supplement 4, October 2011 and shellfish (Cohen 2003; Underhill 1997). Only the recovery (Hu, Ambrose, and Wang 2006), propose that only about of plant remains by flotation will clarify whether the late 1,000 years later did millet become a predominant component foragers in the north were only collectors or were also part- in daily consumption. In Xinglongwa-type sites (ca. 8100– time cultivators, growing broomcorn or foxtail millet or even 7200 cal BP), d13C values in human bones that mark the both. consumption of millet (a C4 plant) reflect the presence of At the site of Donghulin, dated to ca. 10,500–9600 cal BP, both species (broomcorn and foxtail), probably indicating the the excavations and additional field research exposed a pit level of agricultural development (Barton et al. 2009). Inter- house, numerous stone artifacts (including microblades), pot- estingly, flotation samples from a Houli culture site—Yue- tery, grinding stones, faunal remains, and three burials, one zhuang (Jinan, Shandong), with one AMS date of 7900 cal of a woman decorated with 68 sea shells (Cohen 2011; Hao BP—demonstrate the presence of 40 broomcorn seeds and et al. 2001; Zhao et al. 2006). one foxtail millet seed along with 26 rice seeds, indicating an In spite of a relative paucity of excavations that have pro- unexpectedly early arrival of the latter plant in the Yellow vided reliable assemblages of plant remains and radiocarbon River area (Crawford, Chen, and Wang 2006). dates (some of which may represent the use of wood for Animal domestication in North China is an issue raised by building), the next phase is represented by the cultures or the several authors in spite of the paucity of detailed zooarchae- cultural groups named Huoli, Cishan, and Peiligang (fig. 5; ological studies (Flad, Yuan, and Li 2007; Yuan and Flad 2002; further details in Cohen 2011). The bearers of these different Yuan, Flad, and Luo 2008). Given the relative scarcity of fish groups (identified by their pottery types) emerged as culti- in the Yellow River (when compared with the Yangtze River) vators of millet within the middle and lower Yellow River and the abundance of nondomesticated species such as deer basin (Crawford 2006, 2009; Lu et al. 2009; Zhao 2004, 2011). and carnivores (including wolves), the best candidate was the It seems that they started as dryland farmers of broomcorn wild boar. There is little doubt that pigs were the first animal and foxtail millets (Zhao 2004, 2010), and they are possibly to be adopted by farmers, who continued to hunt. The pro- incorporated in the primary “core area” where agriculture (in cess, possibly similar to the one in the Levant, began with terms of the set of activities as defined above) was established. “cultural control” of individuals attracted to the garbage The first farming communities are characterized as 1–2 ha in dumps of villages such as Cishan, at least by 8000 cal BP. By size with semisubterranean rounded houses; a large number 6000 cal BP, pig meat was 60% of consumed mammal tissues of storage pits (some containing abundant millet grains); gar- (Yuan, Flad, and Luo 2008). bage pits; distinct cemetery areas; abundant pottery, stone The two domesticated varieties of millet, P. miliaceum and adzes, axes, and spades; and four-legged grinding stones best Setaria italica, were identified in Xinglongwa from about known from Cishan. The architectural change to rectangular 8200/8100 cal BP as well as at Dadiwan (7800–7300 cal BP) houses marks a second phase within the developing sedentary in the Laoguantai area, which is located farther west and in communities. a higher altitude. The geographic location of both and their A new biomolecular study of plant remains from Cishan rectangular houses mark the later phase of the Early Neolithic suggests that broomcorn millet (Panicum miliaceum) was first (by contrast with the rounded ones that characterized the cultivated/domesticated by ca. 10,300–8700 cal BP (Chang earlier phase) and indicate that they are situated within a 1986; Cohen 2011; Crawford 2009; Lu et al. 2009), although “secondary core area.” Archaeological observations in Inner the earliest radiocarbon dates in this study are older than the Mongolia have already led Shelach (2000) to suggest that lowermost reported layer of the village. Northward (such as to the Xinglongwa culture in Inner Mongolia) and southward millet cultivation must have started earlier than at the Xing- to the Peiligang area, dispersals of early farming probably longwa site. In addition, the well-ordered rectangular-square occurred during the second millennium of the Holocene (ca. houses oriented in the same direction, often attached or very 10,500–9500 cal BP; Cohen 2011). If this scenario is supported close to one other and surrounded by a trench, hint at a social by new evidence, we may suggest that cultivation of wild hierarchy (represented by a central house and a burial with varieties of millet possibly started during the last centuries of jade earrings) that indicates further changes within farming the YD and probably during the first millennium of the Ho- societies. The location of the site on top of a low hill indicates locene and that millet became domesticated some 1,500–2,000 that the trench was probably not to prevent water from flood- years later. If such a scenario is supported by additional evi- ing the site but rather to physically and/or symbolically deter dence, we may conclude that it is an interesting coincidence real or imaginary enemies. Warfare among agricultural tribes that the impact of the YD on populations in both North China is a well-known phenomenon (e.g., Keeley 1996; Roscoe and the northern Levant led to the onset of wild-plant cul- 2009). If long-distance similarities are meaningful, then the tivation (see also Shelach 2000). arrangement of the houses at Xinglongwa resembles sites such Isotopic analysis of human bones from the Xiaojingshan as As¸ıklı and C¸atalho¨yu¨k in Anatolia that belong to the second site (ca. 8000 cal BP) suggests that millet made up only 25% phase of the Neolithic Revolution in the Levant, and culti- of the diet of both males and females (Hu et al. 2008). Hu vation had already been practiced in this region for some et al. (2008), supported by the isotope analysis from Jiahu 2,000 years. Bar-Yosef Early Farming in the Levant and China S187 The Role of Paleoclimate in South China A somewhat similar picture emerges from the reports on Xianrendong and Diaotonghuan (MacNeish et al. 1998). The South China stretches from south of the Qinling Mountains caves were abandoned by the end of the Bølling-Allerød and and the Huai River to the south and southeast coast, and the early YD (i.e., 13,700–12,300 cal BP), and thus direct although it enjoyed somewhat better climatic conditions than cultural connection with the early villages of the Middle Yang- the North, several fluctuations are clearly recorded in cave tze basin is unknown. Early Holocene conditions were im- speleothems and in the South China Sea. Not surprisingly, proved, with more stable monsoonal systems that allowed these are correlated with Timta Cave in Tibet. However, the foragers to carry on their gathering and hunting activities for origins of rice cultivation and domestication, currently de- several millennia (Cohen 2011; Zhao 2011). However, the bated among scholars (e.g., Fuller, Harvey, and Qin 2007; Liu impetus for the onset of cultivation of wild rice is unclear, et al. 2007; Zhao 2011), are sought in three geographic basins and we should regard as among the potential triggers the social south of the Yangtze River, namely, the Lake Dongting area connections through the river network with the north, where (Hunan), the Lake Poyang area (Jiangxi), and the lower Yang- millet was already grown. An additional option is the local tze River. It should also be remembered that the sea rise during “demographic pressure,” which can hardly be imagined in an the post-LGM reduced the coastal belt by at least 250 km. area rich in plant and animal resources, and some social mech- Hence, the region we briefly examine is about 400,000– anism related to competition with other foragers. In spite of 500,000 km2. the huge areas discussed, long-distance connections in the The Late Upper Paleolithic sites in South China (ca. 23,000/ Chinese landmass were enormously facilitated by river trans- 20,000–11,500 cal BP) preserved the old tradition of cobble port. Simple craft could be made from a bunch of bamboo tools such as choppers; cores and flakes; small cup holes on tied together, and the early making of canoes is evidenced in cobbles; perforated cobbles; and bone, antler, and shell tools. Kuahuqiao. This region produced the earliest evidence for pottery making, Rice exploitation predates the available evidence of phy- which dates to ca. 18,000–17,000 cal BP in Yuchanyan Cave toliths and carbonized plant remains obtained in Xianrendong (Boaretto et al. 2009) and probably to an earlier time in Cave (Jiangxi: Zhao 1998), Bashidang (Hunan, ca. 8150–7600 Xianrendong and Diaotonghuan (MacNeish et al. 1998); the cal BP: Zhang 2002; H. Gu, personal communication, 2008), pottery may have been used to make special liquids, to cook Kuahuqiao (ca. 7900–7300 cal BP) in the Lower Yangtze basin bones for grease extraction, or for storage and undoubtedly (Zheng, Sun, and Chen 2007; Zong et al. 2007), and Tian- had special social meaning (Pearson 2005). Rice phytoliths luoshan (by 6600/6400 cal BP; Fuller et al. 2009). Although found in these Terminal Pleistocene cave deposits are now the plant evidence is missing, quite possibly the subsistence considered evidence of gathering or of first experiments in system of Pengtoushan (Hunan, ca. 9300–8300 cal BP), an cultivation (Zhao 2010). early village in the Dongting area, was partially based on rice Open-air sites of Late Pleistocene foragers that may rep- gathering and perhaps cultivation. By ca. 8000–7000 cal BP, resent early rice exploitation within the basin of the Yangtze at least half of the rice recorded in Kuahuqiao was already of River and its small tributary valleys are rare and mostly buried the domesticated variety (Zheng, Sun, and Chen 2007). The under the rapid Holocene alluviation. Therefore, caves are presence of rice in Jiahu and Yuezhuang in the Yellow River regarded as the main sources of information. The most cited basin may indicate that there were long-distance interactions are Yuchanyan Cave (Hunan), Xianrendong and Diaotong- by ca. 8000 cal BP (Zhang and Wang 1998). While the overall huan (Jiangxi), and Miaoyan (Guangxi). impression may be of the simultaneous emergence of two The recently studied deposits of Yuchanyan Cave (Boaretto farming systems, I believe that detailed scrutiny of the avail- et al. 2009; Prendergast, Yuan, and Bar-Yosef 2009; Yuan 2002) able calibrated radiocarbon dates may still raise the interpre- represent many events of building fires with wood during the tation that the middle and lower Yangtze River basin could early Bølling-Allerød period. Rice phytoliths identified in the have been a “secondary core area” influenced by the Yellow first round of research (Zhang 2002) probably reflect gath- River “primary core area” (Zhang and Hung 2008; Zhao 2010, ering in the natural wetlands during fall. Most revealing are 2011). the animal bones, frequently of several deer species, with a few macaque, hare, small carnivores, and large rodents, par- Concluding Remarks ticularly bamboo rat (Prendergast, Yuan, and Bar-Yosef 2009; Yuan 2002). Identified birds such as heron, tern, crane, goose, When viewed from the perspective of a longue dure´e, subsis- and others winter in the area, while the ducks were all wetland tence strategies adopted by foragers during the Terminal Pleis- taxa. Fish included carp and catfish. In sum, it seems that the tocene in western and eastern Asia have much in common. young age of the deer and the presence of wintering wetland In a land “full of people,” the winning option was to stay put birds (whose breeding grounds are in North China, Mongolia, and intensify the exploitation of plant resources—this meant or Siberia) indicate that Yuchanyan Cave was ephemerally starting to cultivate in suitable ecological niches. The strategy occupied by a small group mainly during early fall and winter worked best within the natural habitats of the cereals in the and possibly early spring. Levant and North China. None of the early farmers aban- S188 Current Anthropology Volume 52, Supplement 4, October 2011 doned the gathering of wild plants, hunting, trapping, fishing, as the decision to settle down. Other options were available, or, particularly in China, collecting land snails, freshwater and the final choice was made within the social arena. The mollusks, and water plants. The evidence from the Japanese viable option to move to other people’s territories in China archipelago indicates that the mixed strategy of low-level food could have taken place in a vast region were waterways were production with broad-spectrum exploitation of the sur- the prehistoric highways. Facing intergroup conflicts and min- rounding natural environment also characterized the Jomon imizing mobility was a Levantine solution that would be fa- people (Crawford 2008). We may label these foragers as “in- vored where walking was the common means of crossing the cipient farmers” or “affluent foragers” who practiced culti- landscape. vation, and we should be fully aware of their entire gamut of Thus, the processes in both China and the Levant were subsistence resources. In the Levant and North and South reasonably similar, and sedentism was the common group China, “incipient cultivation” resulted in the domestication strategy. The building of domestic dwellings followed the same of the harvested species and the stable, steady provisioning pattern, starting with round pit houses and shifting gradually of staple foods under favorable climatic conditions; this led to square and rectangular ground plans. Materials varied. In to the rapid increase of local populations and the development China, wood and bricks became the standard building ma- of full-fledged farming and herding economies. terials, while in the Levant, undressed and dressed stones Over the first four millennia of the Holocene (ca. 11,700/ played a major role, joined by bricks. Earlier small-scale farm- 11,500–8200 cal BP), the process of annual cultivation in the ing was additionally supplemented by gathering and hunting, Levant ended in the domestication of several species of cereals a strategy that lasted longer in South China than in either as well as a suite of other plants (e.g., legumes, flax). Corralling North China or the Levant. While rapid climatic change of selected animals (goat, sheep, cattle, and pig) caused their served as a trigger during the closing centuries of the YD, domestication, and together with the plants, these species such changes continue to punctuate the Holocene sequences provided the foundations of the agropastoral societies of later of both regions, a subject beyond the scope of this paper (e.g., periods. The rapid population growth in Southwest Asia re- Berger and Guilaine 2009; Chen et al. 2008; Weninger et al. sulted is what is known as the “Neolithic demographic tran- 2009). sition” (e.g., Bocquet-Appel 2011; Bocquet-Appel and Bar- Yosef 2008 and references therein). The same phenomenon is observed across other regions during the Holocene, and it makes clear that farming was a winning economic strategy Acknowledgments and that its consequences were disastrous to hunting-and- This paper is based on several of my previous writings from gathering groups. the past decade concerning the impact of climatic changes in In light of the Late Pleistocene paleoclimatic and archae- Levantine prehistory. I have added to the current version ological information from North and South China, it seems information gathered recently from Chinese Quaternary stud- that the middle and lower Yellow River basin was prone to ies. I am grateful to Anna Belfer-Cohen, Nigel Goring-Morris, droughts much more frequently than South China, and given and Leore Grosman (Institute of Archaeology, Hebrew Uni- the reconstructed demography of mobile hunter-gatherers in versity) for numerous discussions in the past. Thanks to M. this region, we should expect that the establishment of millet Bar-Matthews and A. Ayalon (Geological Survey of Israel, cultivation preceded the earliest rice cultivation by a millen- Jerusalem) for the information in figure 2. I thank David nium or two. There is clear evidence for the attenuated impact Cohen (Boston University) for discussions concerning Chi- of the YD in South China on the local vegetation. However, nese archaeological issues. Thanks to David Meiggs (Univer- Holocene conditions ranged from subtropical to tropical, with sity of Wisconsin) for his copyediting skills. I am, however, high frequencies of rainfall brought by the monsoons, and solely responsible for any shortcomings of this paper. therefore the impact of the YD is not easily detectable. In both regions, there was probably a long time between the cultivation of the wild progenitors and the establishment of References Cited domesticated, nonshattering varieties as the dominant plants Bandy, Matthew S. 2004. Fissioning, scalar stress, and social evolution in the fields (e.g., Crawford, Chen, and Wang 2006; Fuller, in early village societies. American Anthropologist 106(2):322–333. Harvey, and Qin 2007; Fuller et al. 2009; Lee et al. 2007; Liu Bar-Matthews, Miryam, and Avner Ayalon. 2003. Climatic conditions in the eastern Mediterranean during the Last Glacial (60–10 ky) et al. 2007; Lu 1999, 2006; Zhao 2011; Zheng, Sun, and Chen and their relations to the Upper Palaeolithic in the Levant as 2007). inferred from oxygen and carbon isotope systematics of cave de- Attributing the incipient cereal and millet cultivation to the posits. In More than meets the eye: studies on Upper Palaeolithic impacts of the YD is theoretically couched in several behav- diversity in the Near East. A. Nigel Goring-Morris and Anna Belfer- ioral options that hunter-gatherers had when trying to min- Cohen, eds. Pp. 13–18. Oxford: Oxbow. Bar-Matthews, Miryam, Avner Ayalon, Aaron Kaufman, and Gerald imize risks to their survival and create economic buffer con- J. Wasserburg. 1999. The Eastern Mediterranean region paleocli- ditions. The decision to start cultivation as a planned mate as a reflection of regional events: Soreq Cave, Israel. Earth food-acquisition strategy had its own consequences, as much and Planetary Science Letters 166(1–2):85–95. Bar-Yosef Early Farming in the Levant and China S189

Barton, Loukas, Seth D. Newsome, Fa-Hu Chen, Hui Wang, Thomas Bocquet-Appel, Jean-Pierre. 2011. The agricultural demographic P. Guilderson, and Robert L. Bettinger. 2009. Agricultural origins transition during and after the agriculture inventions. Current An- and the isotopic identity of domestication in northern China. Pro- thropology 52(suppl. 4):S497–S510. ceedings of the National Academy of Sciences of the USA 106(14): Bocquet-Appel, Jean-Pierre, and Ofer Bar-Yosef, eds. 2008. The Neo- 5523–5528. lithic demographic transition and its consequences. Dordrecht: Bartov, Yuval, Mordechai Stein, Yehouda Enzel, Amotz Agnon, and Springer. Ze’ev Reches. 2002. Lake levels and sequence stratigraphy of Lake Braidwood, Robert J. 1952. The Near East and the foundations for Lisan, the Late Pleistocene precursor of the Dead Sea. Quaternary civilization. Condon Lectures. Eugene: Oregon State System of Research 57(1):9–21. Higher Education. Bar-Yosef, Ofer. 1998. The Natufian culture in the Levant: threshold Byrd, Brian F. 2005. Reassessing the emergence of village life in the to the origins of agriculture. Evolutionary Anthropology 6(5):159– Near East. Journal of Archaeological Research 13(3):231–290. 177. Cai, Yanjun, Liangcheng Tan, Hai Cheng, Zhisheng An, R. Lawrence ———. 2002. The Natufian culture and the Early Neolithic: social Edwards, Megan J. Kelly, Xinggong Kong, and Xianfeng Wang. and economic trends in southwestern Asia. In Examining the farm- 2010. The variation of summer monsoon precipitation in central ing/language dispersal hypothesis. Peter Bellwood and Colin Ren- China since the last deglaciation. Earth and Planetary Science Letters frew, eds. Pp. 113–126. McDonald Institute Monographs. Cam- 291(1–4):21–31. bridge: McDonald Institute for Archaeological Research. Cauvin, Jacques. 2000. The birth of the gods and the origins of agri- Bar-Yosef, Ofer, and Anna Belfer-Cohen. 1989. The origins of sed- culture. Trevor Watkins, trans. Cambridge: Cambridge University entism and farming communities in the Levant. Journal of World Press. Prehistory 3(4):447–498. Chang, Kwang-Chih. 1986. The archaeology of ancient China.New Bar-Yosef, Ofer, and Franc¸ois Valla, eds. 1991. The Natufian culture Haven, CT: Yale University Press. in the Levant. Archaeological Series 1. Ann Arbor, MI: International Chen, Chun. 1984. The Microlithic in China. Journal of Anthropo- Monographs in Prehistory. logical Archaeology 3(2):79–115. Bar-Yosef Mayer, Daniella E., and Naomi Porat. 2008. Green stone ———. 2007. Techno-typological comparison of microblade cores beads at the dawn of agriculture. Proceedings of the National Acad- from East Asia and North America. In Origin and spread of mi- emy of Sciences of the USA 105(25):8548–8551. croblade technology in northern Asia and North America. Yaraslov Belfer-Cohen, Anna. 1995. Rethinking social stratification in the Na- V. Kuzmin, Susan G. Keates, and Chen Shen, eds. Pp. 7–38. Bur- tufian culture: the evidence from burials. In The archaeology of naby, British Columbia: Archaeology Press, Simon Fraser Univer- death in the ancient Near East. Stuart Campbell and Anthony sity. Green, eds. Pp. 9–16. Oxbow Monograph 51. Oxford: Oxbow. Chen, Zhongyuan, Yongqiang Zong, Zhanghua Wang, Hui Wang, Belfer-Cohen, Anna, and Ofer Bar-Yosef. 2000. Early sedentism in and Jing Chen. 2008. Migration patterns of Neolithic settlements the Near East: a bumpy ride to village life. In Life in Neolithic on the abandoned Yellow and Yangtze River deltas of China. Qua- farming communities: social organization, identity, and differentia- ternary Research 70(2):301–314. tion. Ian Kuijt, ed. Pp. 19–37. New York: Plenum. Cohen, David Joel. 2003. Microblades, pottery, and the nature and Belfer-Cohen, Anna, and Erella Hovers. 2005. The ground stone chronology of the Palaeolithic-Neolithic transition in China. Re- assemblages of the Natufian and Neolithic societies in the Levant: view of Archaeology 24(2):21–36. a brief review. Journal of the Israel Prehistoric Society 35:299–308. ———. 2011. The beginnings of agriculture in China: a multiregional Belfer-Cohen, Anna, and A. Nigel Goring-Morris. 2011. Becoming view. Current Anthropology 52(suppl. 4):S273–S293. farmers: the inside story. Current Anthropology 52(suppl. 4):S209– Colledge, Sue. 2001. Plant exploitation on Epipaleolithic and Early S220. Neolithic sites in the Levant. BAR International Series 986. Oxford: Bell, Barbara. 1971. The Dark Ages in ancient history. I. The first Hedges. Dark Age in Egypt. American Journal of Archaeology 75(1):1–26. Cosford, Jason Cosford, Hairuo Qing, Bruce Eglington, Dave Mattey, Berger, Jean-Franc¸ois, and Jean Guilaine. 2009. The 8200 cal BP Daoxiang Yuan, Meiliang Zhang, and Hai Cheng. 2008. East Asian abrupt environmental change and the Neolithic transition: a Med- iterranean perspective. Quaternary International 200(1):31–49. monsoon variability since the Mid-Holocene recorded in a high- Bettinger, Robert L., Loukas Barton, and Christopher Morgan. 2010. resolution, absolute-dated aragonite speleothem from eastern The origins of food production in North China: a different kind China. Earth and Planetary Science Letters 275(3–4):296–307. of agricultural revolution. Evolutionary Anthropology 19(1):9–21. Crawford, Gary W. 2006. East Asian plant domestication. In Ar- Bettinger, Robert L., Loukas Barton, Peter J. Richerson, Robert Boyd, chaeology of Asia. Miriam T. Stark, ed. Pp. 77–95. Blackwell Studies Wang Hui, and Choi Won. 2007. The transition to agriculture in in Global Archaeology. Malden, MA: Blackwell. northwestern China. In Late Quaternary climate change and human ———. 2008. The Jomon in early agriculture discourse: issues arising adaptation in arid China. David B. Madsen, Fa-Hu Chen, and Xing from Matsui, Kanehara and Pearson. World Archaeology 40(4):445– Gao, eds. Pp. 83–101. Developments in Quaternary Science 9. 465. Amsterdam: Elsevier. ———. 2009. Agricultural origins in North China pushed back to Binford, Lewis R. 1968. Post-Pleistocene adaptations. In New per- Pleistocene-Holocene boundary. Proceedings of the National Acad- spectives in archaeology. Sally R. Binford and Lewis R. Binford, eds. emy of Sciences of the USA 106(18):7271–7272. Pp. 313–341. Chicago: Aldine. Crawford, Gary W., Xuexiang Chen, and J. H. Wang. 2006. Houli ———. 2006. Bands as characteristic of “mobile hunter-gatherers” culture rice from the Yuezhuang site, Jinan. East Asia Archaeology may exist only in the history of anthropology. In Archaeology and 3:247–251. [In Chinese.] ethnoarchaeology of mobility. Fre´de´ric Sellet, Russell Greaves, and Denham, Timothy P., Simon G. Haberle, Carol Lentfer, Richard Full- Pei-Lin Yu, eds. Pp. 3–22. Gainesville: University Press of Florida. agar, Judith Field, Michael Therin, Nicholas Porch, and Barbara Boaretto, Elisabetta, Xiaohong Wu, Jiarong Yuan, Ofer Bar-Yosef, Winsborough. 2003. Origins of agriculture at Kuk Swamp in the Vikki Chu, Yan Pan, Kexin Liu, et al. 2009. Radiocarbon dated highlands of New Guinea. Science 301(5630):189–193. early pottery at Yuchanyan Cave, Hunan Province, China. Pro- Dykoski, Carolyn A., R. Lawrence Edwards, Hai Cheng, Daoxian ceedings of the National Academy of Sciences of the USA 106(24): Yuan, Yanjun Cai, Meiliang Zhang, Yushi Lin, Jiaming Qing, Zhi- 9595–9600. sheng An, and Justin Revenaugh. 2005. A high-resolution absolute S190 Current Anthropology Volume 52, Supplement 4, October 2011

dated Holocene and deglacial Asian monsoon record from Dongge human life history. Santa Fe, NM: School of American Research Cave, China. Earth and Planetary Science Letters 233(1–2):71–86. Press. Edwards, Philip C. 2007. A 14 000 year-old hunter-gatherer’s toolkit. Henry, Donald O. 1989. From foraging to agriculture: the Levant at Antiquity 81(314):865–876. the end of the Ice Age. Philadelphia: University of Pennsylvania Enzel, Yehouda, Rivka Amit, Uri Dayan, Onn Crouvi, Ron Kahana, Press. Baruch Ziv, and David Sharon. 2008. The climatic and physio- Hillman, Gordon C. 2000. Abu Hureyra I: the Epipaleolithic. In graphic controls of the eastern Mediterranean over the Late Pleis- Village on the Euphrates: from foraging to farming at Abu Hureyra. tocene climates in the southern Levant and its neighboring deserts. Andrew M. T. Moore, Gordon C. Hillman, and Anthony J. Legge, Global and Planetary Change 60(3–4):165–192. eds. Pp. 327–398. Oxford: Oxford University Press. Flad, Rowan K., Jing Yuan, and Shuicheng Li. 2007. Zooarchaeolog- Hillman, Gordon C., Robert Hedges, Andrew Moore, Susan Colledge, ical evidence for animal domestication in Northwest China. In and Paul Pettitt. 2001. New evidence of Lateglacial cereal culti- Late Quaternary climate change and human adaptation in arid vation at Abu Hureyra on the Euphrates. Holocene 11(4):383–393. China. David B. Madsen, Fa-Hu Chen, and Xing Gao, eds. Pp. Hodder, Ian. 2007. C¸atalho¨yu¨k in the context of the Middle Eastern 167–204. Developments in Quaternary Science 9. Amsterdam: Neolithic. Annual Review of Anthropology 36:105–120. Elsevier. Hu, Chaoyong, Junhua Huang, Nianqiao Fang, Shucheng Xie, Gid- Flannery, Kent V. 1973. The origins of agriculture. Annual Review of eon M. Henderson, and Yanjun Cai. 2005. Absorbed silica in sta- Anthropology 2:271–310. lagmite carbonate and its relationship to past rainfall. Geochimica Frumkin, Amos, Derek C. Ford, and Henry P. Schwarcz. 2000. Pa- et Cosmochimica Acta 69(9):2285–2292. leoclimate and vegetation of the last glacial cycles in Jerusalem Hu, Yaowu, Stanley H. Ambrose, and Changsui Wang. 2006. Stable from a speleothem record. Global Biogeochemical Cycles 14(3):863– isotopic analysis of human bones from Jiahu site, Henan, China: 870. implications for the transition to agriculture. Journal of Archaeo- Fuller, Dorian Q. 2007. Contrasting patterns in crop domestication logical Science 33(9):1319–1330. and domestication rates: recent archaeobotanical insights from the Hu, Yaowu, Shougong Wang, Fengshi Luan, Changsui Wang, and Old World. Annals of Botany 100(5):903–924. Michael P. Richards. 2008. Stable isotope analysis of humans from Fuller, Dorian Q, Emma Harvey, and Ling Qin. 2007. Presumed Xiaojingshan site: implications for understanding the origin of domestication? evidence for wild rice cultivation and domestica- millet agriculture in China. Journal of Archaeological Science 35(11): tion in the fifth millennium BC of the Lower Yangtze region. 2960–2965. Antiquity 81(312):316–331. Iba´n˜ez, Juan Jose´, ed. 2008. Le site ne´olithique de Tell Mureybet (Syrie Fuller, Dorian Q, Ling Qin, Yunfei Zheng, Zhijun Zhao, Xugao Chen, du Nord): en hommage a`Jacques Cauvin. BAR International Series Leo Aoi Hosoya, and Guo-Ping Sun. 2009. The domestication 1843. Oxford: Archaeopress. process and domestication rate in rice: spikelet bases from the Johnson, Gregory A. 1982. Organizational structure and scalar stress. lower Yangtze. Science 323(5921):1607–1610. In Theory and explanation in archaeology. A. Colin Renfrew, Mi- Glantz, Michael H., David G. Streets, Thomas R. Stewart, N. Bhatti, chael J. Rowlands, and Barbara A. Segraves, eds. Pp. 389–421. Christopher M. Moore, and C. H. Rosa. 1998. Exploring the concept London: Academic Press. of climate surprises: a review of the literature on the concept of surprise Kadosh, Dafna, Dorit Sivan, Haim Kutiel, and Mina Weinstein- and how it is related to climate change. ANL/DIS/TM-46. Argonne, Evron. 2004. A Late Quaternary paleoenvironmental sequence IL: Decision and Information Sciences Division, Argonne National from Dor, Carmel coastal plain, Israel. Palynology 28:143–157. Laboratory, U. S. Department of Energy. Kajiwara, Hiroshi. 2008. Microlithization in Eurasia: a brief review Goring-Morris, A. Nigel. 1991. The Harifian of the southern Levant. on the microblade revolution technology and its significance as In The Natufian culture in the Levant. O. Bar-Yosef and Franc¸ois behavioural threshold of modern humans. Bulletin of Tohoku Fu- R. Valla, eds. Pp. 173–216. Archaeological Series 1. Ann Arbor, kushi University 32:207–234. MI: International Monographs in Prehistory. Kawecki, Tadeusz J. 2008. Adaptation to marginal habitats. Annual ———. 1995. Complex hunter-gatherers at the end of the Palaeo- Review of Ecology, Evolution, and Systematics 39:321–342. lithic (20,000–10,000 B.P.). In The archaeology of society in the Holy Keeley, Lawrence H. 1996. War before civilization. New York: Oxford Land. Thomas E. Levy, ed. Pp. 141–168. London: Leicester Uni- University Press. versity Press. Kelly, Robert L. 1995. The foraging spectrum: diversity in hunter- Goring-Morris, A. Nigel, and Anna Belfer-Cohen. 2011. Neolithi- gatherer lifeways. Washington, DC: Smithsonian Institution Press. zation processes in the Levant: the outer envelope. Current An- Kislev, Mordechai E. 1989. Pre-domesticated cereals in the Pre-Pot- thropology 52(suppl. 4):S195–S208. tery Neolithic A period. In People and culture in change. Israel Grosman, Liore, Natalie D. Munro, and Anna Belfer-Cohen. 2008. Hershkovitz, ed. Pp. 147–152. BAR International Series 508. Ox- A 12,000-year-old Shaman burial from the southern Levant (Is- ford: British Archaeological Reports. rael). Proceedings of the National Academy of Sciences of the USA ———. 1997. Early agriculture and paleoecology of Netiv Hagdud. 105(46):17665–17669. In An Early Neolithic village in the Jordan Valley. Part I: the ar- Hao, Shou-Gang, Xue-Ping Ma, Si-Xun Yuan, and John Southon. chaeology of Netiv Hagdud. Ofer Bar-Yosef and Avi Gopher, eds. 2001. The Donghulin woman from western Beijing: 14C age and Pp. 209–236. Cambridge, MA: Peabody Museum of Archaeology associated compound shell necklace. Antiquity 75(289):517–522. and Ethnology, Harvard University. Harris, David R. 2007. Agriculture, cultivation, and domestication: Kislev, Mordechai E., Anat Hartmann, and Ofer Bar-Yosef. 2006. exploring the conceptual framework of early food production. In Early domesticated fig in the Jordan Valley. Science 312(5778): Rethinking agriculture: archaeological and ethnoarchaeological per- 1372–1374. spectives. Timothy Denham, Jose Iriarte, and Luc Vrydaghs, eds. Kolodny, Yehoshua, Mordechai Stein, and Malka Machlus. 2005. Sea- Pp. 16–35. Walnut Creek, CA: Left Coast. rain-lake relation in the Last Glacial East Mediterranean revealed Hassan, Fekri A. 2002. Droughts, food and culture: ecological change by d18O-d13C in Lake Lisan aragonites. Geochimica et Cosmochimica and food security in Africa’s later prehistory. New York: Kluwer Acta 69(16):4045–4060. Academic/Plenum. Kuhlemann, Joachim, Eelco J. Rohling, Ingrid Krumrei, Peter Kubik, Hawkes, Kristen, and Richard R. Paine, eds. 2007. The evolution of Susan Ivy-Ochs, and Michal Kucera. 2008. Regional synthesis of Bar-Yosef Early Farming in the Levant and China S191

Mediterranean atmospheric circulation during the Last Glacial Marcus, Joyce. 2008. The archaeological evidence for social evolution. Maximum. Science 321(5894):1338–1340. Annual Review of Anthropology 37:251–266. Kuijt, Ian, and Bill Finlayson. 2009. Evidence for food storage and Mayewski, Paul A. Mayewski, Eelco J. Rohling, J. Curt Stager, Wi- predomestication granaries 11,000 years ago in the Jordan Valley. bjo¨rn Karle´n, Kirk A. Maasch, L. David Meeker, Eric A. Meyerson, Proceedings of the National Academy of Sciences of the USA 106(27): et al. 2004. Holocene climate variability. Quaternary Research 10966–10970. 62(3):243–255. Kuzmin, Yaroslav V., Susan G. Keates, and Chen Shen, eds. 2007. Mazurowski, Ryszard F., Danuta J. Michczyn´ska, Anna Pazdur, and Origin and spread of microblade technology in Northeastern Asia Natalia Piotrowska. 2009. Chronology of the early Pre-Pottery and North America. Burnaby, British Columbia: Archaeology Press, Neolithic settlement Tell Qaramel, northern Syria, in the light of Simon Fraser University. radiocarbon dating. Radiocarbon 51(2):771–781. Lachniet, Matthew. S. 2009. Climatic and environmental controls on Minc, Leah D., and Kevin P. Smith. 1989. The spirit of survival: speleothem oxygen-isotope values. Quaternary Science Reviews cultural response to resource variability in North Alaska. In Bad 28(5–6):412–432. year economics: cultural responses to risk and uncertainty.PaulHal- Lee, Gyoung-Ah, Gary W. Crawford, Li Liu, and Xingcan Chen. 2007. stead and John O’Shea, eds. Pp. 8–39. Cambridge: Cambridge Plants and people from the Early Neolithic to Shang periods in University Press. North China. Proceedings of the National Academy of Sciences of Moore, Andrew M. T., Gordon C. Hillman, and Anthony J. Legge, the USA 104(3):1087–1092. eds. 2000. Village on the Euphrates: from foraging to farming at Abu Lev-Yadun, Simcha, Avi Gopher, and Shahal Abbo. 2002. The cradle Hureyra. Oxford: Oxford University Press. of agriculture. Science 288(5471):1602–1603. Munro, Natalie D. 2004. Zooarchaeological measures of hunting Lev-Yadun, Simcha, and Mina Weinstein-Evron. 2005. Modeling the pressure and occupation intensity in the Natufian. Current An- influence of wood use by the Natufians of El-wad in the forest of thropology 45(suppl. 4):S5–S34. Mount Carmel. Journal of the Israel Prehistoric Society 35:285–298. Pearson, Richard. 2005. The social context of early pottery in the Lewis, Henry T. 1972. The role of fire in the domestication of plants Lingnan region of South China. Antiquity 79(306):819–828. and animals in Southwest Asia: a hypothesis. Man, n.s., 7(2):195– Prendergast, Mary. E., Jiarong Yuan, and Ofer Bar-Yosef. 2009. Re- 222. source intensification in the Late Upper Paleolithic: a view from Li, Jinbao, Xiaohua Gou, Edward R. Cook, and Fahu Chen. 2006. southern China. Journal of Archaeological Science 36(4):1027–1037. Tree-ring based drought reconstruction for the central Tien Shan Price, T. Douglas, and Ofer Bar-Yosef. 2010. Traces of inequality at area in Northwest China. Geophysical Research Letters 33:L07715, the origins of agriculture in the ancient Near East. In Pathways to doi:10.1029/2006GL025803. power: new perspectives on the emergence of social inequality.T. Lin, Da-Cheng, Chiung-Hui Liu, Tien-Hsi Fang, Cheng-Han Tsai, Douglas Price and Gary M. Feinman, eds. Pp. 147–168. New York: Masafumi Murayama, and Min-Te Chen. 2006. Millennial-scale Springer. changes in terrestrial sediment input and Holocene surface hy- Rappaport, Roy A. 1968. Pigs for the ancestors: ritual in the ecology drography in the northern South China Sea. Palaeogeography, Pa- of a New Guinea people. New Haven, CT: Yale University Press. laeoclimatology, Palaeoecology 236(1–2):56–73. Ren, Guoyu, and Hans-Ju¨rgen Beug. 2002. Mapping Holocene pollen Liu, Dianbing, Yongjin Wang, Hai Cheng, R. Lawrence Edwards, data and vegetation of China. Quaternary Science Reviews 21(12– Xinggong Kong, Xianfeng Wang, Jiangying Wu, and Shitao Chen. 13):1395–1422. 2008. A detailed comparison of Asian monsoon intensity and Rohling, Eelco J., James Casford, Ramadan Abu-Zied, Stephen Cooke, Domenico Mercone, John Thomson, Ian Croudace, et al. Greenland temperature during the Allerød and Younger Dryas 2002. Rapid Holocene climate changes in the eastern Mediterra- events. Earth and Planetary Science Letters 272(3–4):691–697. nean. In Droughts, food and culture: ecological change and food Liu, Li, Gyoung-Ah Lee, Leping Jiang, and Juzhong Zhang. 2007. security in Africa’s later prehistory. Fekri A. Hassan, ed. Pp. 35–46. Evidence for the early beginning (c. 9000 cal. BP) of rice domes- New York: Kluwer Academic/Plenum. tication in China: a response. Holocene 17(8):1059–1068. Rolett, Barry V. 2008. Avoiding collapse: pre-European sustainability Lourandos, Harry. 1997. Continent of hunter-gatherers: new perspec- on Pacific Islands. Quaternary International 184:4–10. tives in Australian prehistory. Cambridge: Cambridge University Roscoe, Paul B. 2006. Fish, game and the foundations of complexity Press. in forager society: the evidence from New-Guinea. Cross-Cultural Lu, Houyuan, Jianping Zhang, Kam-biu Liu, Naiqin Wu, Yumei Li, Research 40(1):29–46. Kunshu Zhou, Maolin Ye, et al. 2009. Earliest domestication of ———. 2008. Settlement fortification in village and “tribal” society: common millet (Panicum miliaceum) in East Asia extended to evidence from contact-era New Guinea. Journal of Anthropological 10,000 years ago. Proceedings of the National Academy of Sciences Archaeology 27(4):507–519. of the USA 106(18):7367–7372. ———. 2009. Social signaling and the organization of small-scale Lu, Tracey Lie-Dan. 1999. The transition from foraging to farming and society: the case of contact-era New Guinea. Journal of Archaeo- the origin of agriculture in China. BAR International Series 774. logical Method and Theory 16(2):69–116. Oxford: Archaeopress. Rosen, Arlene M. 2007. Civilizing climate: social responses to climate ———. 2002. A green foxtail (Setaria viridis) cultivation experiment change in the ancient Near East. London: Altamira. in the Middle Yellow River Valley and some related issues. Asian Rosenberg, Michael, and Richard W. Redding. 2000. Hallan C¸emi Perspectives 41(1):1–14. and early village organization in eastern Anatolia. In Life in Neo- ———. 2006. The occurrence of cereal cultivation in China. Asian lithic farming communities: social organization, identity, and differ- Perspectives 45(2):129–158. entiation. Ian Kuijt, ed. Pp. 39–61. New York: Plenum. MacNeish, Richard S., Geoffrey Cunnar, Zhijun Zhao, and Jane Libby. Rossignol-Strick, Martine. 1995. Sea-land correlation of pollen rec- 1998. Sino-American Jiangxi (PRC) Origin of Rice Project. Andover, ords in the eastern Mediterranean for the glacial-interglacial tran- MA: Andover Foundation for Archaeological Research. sition: biostratigraphy versus radiometric time-scale. Quaternary Madsen, David B., Jingzen Li, Robert G. Elston, Cheng Xu, Robert Science Reviews 14(9):893–915. L. Bettinger, Kan Geng, P. Jeff Brantingham, and Kan Zhong. 1998. Savard, Manon, Mark Nesbitt, and Martin K. Jones. 2006. The role The loess/paleosol record and the nature of the Younger Dryas of wild grasses in subsistence and sedentism: new evidence from climate in Central China. Geoarchaeology 13(8):847–869. the northern Fertile Crescent. World Archaeology 38(2):179–196. S192 Current Anthropology Volume 52, Supplement 4, October 2011

Shelach, Gideon. 2000. The earliest Neolithic cultures of northeast palaeoecology of the Hula area, northeastern Israel. In A timeless China: recent discoveries and new perspectives on the beginning vale: archaeological and related essays on the Jordan Valley in honour of agriculture. Journal of World Prehistory 14(4):363–413. of Gerrit van der Kooij on the occasion of his sixty-fifth birthday. Shen, Caiming, Wei-Chyung Wang, Zhixin Hao, and Wei Gong. 2007. Eva Kaptijn and Lucas P. Petit, eds. Pp. 29–64. Archaeological Exceptional drought events over eastern China during the last five Studies Leiden University 19. Leiden: Leiden University Press. centuries. Climatic Change 85(3–4):453–471. van Zeist, Willem, and Sytze Bottema. 1991. Late Quaternary vege- Shizitan Archaeological Team. 2002. The excavation of the Paleolithic tation of the Near East. Beihefte zum Tu¨binger Atlas des Vorderen site, Shizitan (loc. S14), Jixian County, Shanxi Province. Archae- Orients, Series A (Naturwissenschaft), no. 18. Wiesbaden: Reichert. ology (Tiayuan) 4:1–28. [In Chinese.] Vigne, Jean-Denis, Isabelle Carre`re, Franc¸ois Briois, and Jean Gui- ———. 2010. Report on the excavations of Shizitan site 9, Jixian laine. 2011. The early process of mammal domestication in the County, Shanxi. Kao-Gu 10(2010):7–17. [In Chinese.] Near East: new evidence from the Pre-Neolithic and Pre-Pottery Sima, Adriana, Andre´Paul, and Michael Schulz. 2004. The Younger Neolithic in Cyprus. Current Anthropology 52(suppl. 4):S255–S271. Dryas: an intrinsic feature of the Late Pleistocene climate change Vigne, Jean-Denis, Antoine Zazzo, Jean-Franc¸ois Salie`ge, Franc¸ois at millennial timescales. Earth and Planetary Science Letters 222(3– Poplin, Jean Guilaine, and Alan Simmons. 2009. Pre-Neolithic wild 4):741–750. boar management and introduction to Cyprus more than 11,400 Sinha, Ashish, Kevin G. Cannariato, Lowell D. Stott, Hong-Chun Li, years ago. Proceedings of the National Academy of Sciences of the Chen-Feng You, Hai Cheng, R. Lawrence Edwards, and Indra B. USA 106(38):16135–16138. Singh. 2005. Variability of southwest Indian summer monsoon Wang, Yongjin, Hai Cheng, R. Lawrence Edwards, Zhisheng An, precipitation during the Bølling-Allerød. Geology 33(10):813–816. Jiangyin Wu, Chuan-Chou Shen, and Jeffrey A. Dorale. 2001. A Smith, Bruce D. 2001. Low level food production. Journal of Ar- high-resolution absolute-dated Late Pleistocene Monsoon record chaeological Research 9(1):1–43. from Hulu Cave, China. Science 294(5550):2345–2348. Stiner, Mary C., Natalie D. Munro, and Todd A. Surovell. 2000. The Wang, Yongjin, Hai Cheng, R. Lawrence Edwards, Xinggong Kong, tortoise and the hare: small game use, the broad-spectrum revo- Xiaohua Shao, Shitao Chen, Jiangyin Wu, Xiouyang Jiang, Xian- lution, and Paleolithic demography. Current Anthropology 41(1): feng Wang, and Zhisheng An. 2008. Millennial- and orbital-scale 39–73. changes in the East Asian monsoon over the past 224,000 years. Stordeur, Danielle, and Fre´de´ric Abbe`s. 2002. Du PPNA au PPNB: Nature 451(7182):1090–1093. mise en lumie`re d’une phase de transition a` Jerf el Ahmar (Syrie). Watkins, Trevor. 2006. Architecture and the symbolic construction Bulletin de la Socie´te´ Pre´historique Franc¸aise 99(3):563–595. of new worlds. In Domesticating space: construction, community, Stordeur, Danielle, and Juan Jose´ Iba´n˜ez. 2008. Stratigraphie et re´- and cosmology in the Late Prehistoric Near East. Edward B. Banning partition des architectures a` Mureybet. In Le site ne´olithique de and Michael Chazan, eds. Pp. 16–24. Studies in early Near Eastern Tell Mureybet (Syrie du Nord): en hommage a`Jacques Cauvin. Juan production, subsistence, and economy 12. Berlin: Ex Oriente. Jose´Iba´n˜ez, ed. Pp. 33–94. BAR International Series 1843. Oxford: Weinstein-Evron, Mina. 2009. Archaeology in the archives: unveiling Archaeopress. the Natufian culture of Mount Carmel. American School of Pre- Stordeur, Danielle, and George Willcox. 2009. Indices de culture et historic Research Monograph Series. Boston: Brill. d’utilisation des ce´re´ales a` Jerf el Ahmar. In De Me´diterrane´e et Weiss, Ehud, Mordechai E. Kislev, and Anat Hartmann. 2006. Au- d’ailleurs. . .: me´langes offerts a`Jean Guilaine. Pp. 693–710. Tou- tonomous cultivation before domestication. Science 312(5780): louse: Archives d’E´ cologie Pre´historique. 1608–1610. Tanno, Ken-ichi, and George Willcox. 2006. How fast was wild wheat Weiss, Ehud, and Daniel Zohary. 2011. The Neolithic Southwest domesticated? Science 311(5769):1886. Asian founder crops: their biology and archaeobotany. Current Tchernov, Eitan. 1991. Biological evidence for human sedentism in Anthropology 52(suppl. 4):S237–S254. Southwest Asia during the Natufian. In The Natufian culture in Weiss, Harvey, and Raymond S. Bradley. 2001. What drives societal the Levant. Ofer Bar-Yosef and Franc¸ois R. Valla, eds. Pp. 315– collapse? Science 291(5504):609–610. 340. Archaeological Series 1. Ann Arbor, MI: International Mono- Wen, Ruilin, Jule Xiao, Zhigang Chang, Dayou Zhai, Qinghai Xu, graphs in Prehistory. Yuecong Li, and Shigeru Itoh. 2010. Holocene precipitation and Underhill, Anne P. 1997. Current issues in Chinese Neolithic ar- temperature variations in the East Asian monsoonal margin from chaeology. Journal of World Prehistory 11(2):103–160. pollen data from Hulun Lake in northeastern Inner Mongolia. Vaks, Anton, Miryam Bar-Matthews, Avner Ayalon, Bettina Schil- Boreas 39(2):262–272. man, Mabs Gilmour, Chris J. Hawkesworth, Amos Frumkin, Weninger, Bernhard, Lee Clare, Eelco Rohling, Ofer Bar-Yosef, Utz Aaron Kaufman, and Alan Matthews. 2003. Paleoclimate recon- Bo¨hner, Mihael Budja, Manfred Bundschu, et al. 2009. The impact struction based on the timing of speleotherm growth and oxygen of rapid climate change on prehistoric societies during the Ho- and carbon isotope composition in a cave located in the rain locene in the Eastern Mediterranean. Documenta Praehistorica 36: shadow in Israel. Quaternary Research 59(2):182–193. 551–583. Valla, Franc¸ois R. 1989. Aspects de sol de L’abri 131 de Mallaha Willcox, George, Ramon Buxo´, and Linda Herveux. 2009. Late Pleis- (Eynan). Pale´orient 14(2):283–296. tocene and Early Holocene climate and the beginnings of culti- ———. 1995. The first settled societies: Natufian (12,500–10,200 vation in northern Syria. Holocene 19(1):151–158. BP). In The archaeology of society in the Holy Land. Thomas E. Willcox, George, Sandra Fornite, and Linda Herveux. 2008. Early Levy, ed. Pp. 169–189. London: Leicester University Press. Holocene cultivation before domestication in northern Syria. Veg- ———. 1999. The Natufian: a coherent thought? In Dorothy Garrod etation History and Archaeobotany 17(3):313–325. and the progress of the Palaeolithic. William Davies and Ruth Wu¨nnemann, Bernd, Kai Hartmann, Manon Janssen, and Zhang Charles, eds. Pp. 224–241. Oxford: Oxbow. Hucai Zhang. 2007. Response of Chinese desert lakes to climate Valla, Franc¸ois R., Hamoudi Khalaily, He´le`ne Valladas, E´ velyne Kalt- instability during the past 45,000 years. In Late Quaternary climate necker, Fanny Bocquentin, Teresa Cabellos, Daniella A. Bar-Yosef change and human adaptation in arid China. David B. Madsen, Mayer, et al. 2007. Les fouilles de Ain Mallaha (Eynan) de 2003 Fa-Hu Chen, and Xing Gao, eds. Pp. 11–24. Developments in a` 2005: quatrie`me rapport pre´liminaire. Journal of the Israel Pre- Quaternary Science 9. Amsterdam: Elsevier. historic Society 37:135–383. Yasuda, Yoshinori. 2002. The second east side story: origin of agri- van Zeist, Willem, Uri Baruch, and Sytze Bottema. 2009. Holocene culture in West Asia. In The origins of pottery and agriculture. Bar-Yosef Early Farming in the Levant and China S193

Yoshinori Yasuda, ed. Pp. 15–38. Yangtze River Civilization Pro- River Civilization Programme, International Research Center for gramme, International Research Center for Japanese Studies. New Japanese Studies. New Delhi: Roli/Lustre. Delhi: Roli/Lustre. Zhao, C., Tao Wang, Xiaohong Wu, Mingli Liu, Xuemei Yuan, Jin- Yi, Sangheon, Yoshiki Saito, Hideaki Oshima, Yongqing Zhou, and cheng Yu, and Jing Guo. 2006. Donghulin, Zhuannian, Nan- Helong Wei. 2003. Holocene environmental history from pollen zhuangtou, and Yujiagou sites. In Prehistoric archaeology of South assemblages in Huangbe (Yellow River) delta, China: climatic China and Southeast Asia: collected papers of an international ac- change and human impact. Quaternary Science Reviews 22(5–7): ademic workshop: celebrating 30 years for the excavations of Zeng- 609–628. piyan Cave. Institute of Archaeology of the Chinese Academy of Yu, C., Y. Luo, and X. Sun. 2008. A high resolution pollen record Social Science, ed. Pp. 116–127. Beijing: Wen wu chu ban she. from Ha’ni Lake, Jilin, Northeast China showing climate changes Zhao, Songqiao. 1994. Geography of China: environment, resources, between 13.1 k cal BP and 4.5 k cal BP. Disiji Yanjiu 28:929–938. population and development. New York: Wiley. [In Chinese.] Zhao, Zhijun. 1998. The Middle Yangtze region in China is one place Yu, Ge, Xudong Chen, Jian Ni, Rachid Cheddadi, Joe¨l Guiot, H. Han, where rice was domesticated: phytolith evidence from the Diao- Sandy P. Harrison, et al. 2000. Paleovegetation of China: a pollen tonghuan Cave, northern Jiangxi. Antiquity 72(278):885–897. data-based synthesis for the Mid-Holocene and Last Glacial Max- ———. 2004. New evidence of the origin of dry farming in Northeast imum. Journal of Biogeography 27(3):635–664. China. Wenwubo 10:89–96. [In Chinese.] Yuan, Jiarong. 2002. Rice and pottery 10,000 yrs. BP at Yuchanyan, ———. 2010. New data and new issues for the study of origins of Dao County, Hunan Province. In The origins of pottery and ag- rice agriculture in China. Archaeological and Anthropological Sci- riculture. Yoshinori Yasuda, ed. Pp. 157–166. Yangtze River Civi- ences 2(2):99–105. ———. 2011. New archaeobotanic data for the study of the origins lization Programme, International Research Center for Japanese of agriculture in China. Current Anthropology 52(suppl. 4):S295– Studies. New Delhi: Roli/Lustre. S306. Yuan, Jing, and Rowan K. Flad. 2002. Pig domestication in ancient Zheng, Yun-Fei, Guo-Ping Sun, and Xu-Gao Chen. 2007. Charac- China. Antiquity 76(293):724–732. teristics of the short rachillae of rice from archaeological sites Yuan, Jing, Rowan Flad, and Yunbing Luo. 2008. Meat-acquisition dating to 7000 years ago. Chinese Science Bulletin 52(12):1654– patterns in the Neolithic Yangzi River Palley, China. Antiquity 1660. 82(316):351–366. Zhou, Houyun, Jianxin Zhao, Yuexing Feng, Michael K. Gagan, Zeder, Melinda A. 2011. The origins of agriculture in the Near East. Guoqing Zhou, and Jun Yan. 2008. Distinct climate change syn- Current Anthropology 52(suppl. 4):S221–S235. chronous with Heinrich event one, recorded by stable oxygen and Zhang, Chi, and Hsiao-Chun Hung. 2008. The Neolithic of southern carbon isotopic compositions in stalagmites from China. Quater- China: origin, development, and dispersal. Asian Perspectives 47(2): nary Research 69(2):306–315. 299–329. Zohary, Daniel, and Maria Hopf. 2000. Domestication of plants in the Zhang, Juzhong, and Xiangkun Wang. 1998. Notes on the recent Old World: the origins and spread of cultivated plants in West Asia, discovery of ancient cultivated rice in Jiahu, Henan Province: a Europe, and the Nile Valley. 3rd edition. Oxford: Oxford University new theory concerning the origin of Oryza japonica. Antiquity Press. 72(278):897–901. Zong, Yongqiong, Zhongyuan Chen, James B. Innes, Chun Chen, Zhang, Wenxu. 2002. The bi-peak-tubercle of rice, the character of Zhanghua Wang, and Hui Wang. 2007. Fire and flood management ancient rice and the origin of cultivated rice. In The origins of of coastal swamp enabled first rice paddy cultivation in East China. pottery and agriculture. Yoshinori Yasuda, ed. Pp. 205–216. Yangtze Nature 449(7161):459–463.

Current Anthropology Volume 52, Supplement 4, October 2011 S195

Neolithization Processes in the Levant The Outer Envelope

by A. Nigel Goring-Morris and Anna Belfer-Cohen

The Near East is one of those unique places where the transition(s) from hunter-gatherers to farmers occurred locally, so it is possible to observe the whole sequence of these processes within the region as a whole. We discuss the archaeological evidence pertaining to those transformations within the Levant, presenting the particularistic local changes in settlement patterns and the character of the different communities juxtaposed with the landscapes and environmental background. The asyn- chronous developments clearly reflect the mosaic nature of the Levant in terms of specific local environmental conditions that influenced the scope and pace of Neolithization processes.

Introduction Bar-Yosef and Belfer-Cohen 1989; Bellwood 2005; Binford 1968; Boserup 1981; Cauvin 2000; Flannery 1969; Grosman 2004; Hayden 1990; Redman 1978; Rindos 1984; Verhoeven The Levant is a distinct and limited area of the Near East 2004; Watkins 2005). Yet many models suffer from a disregard characterized by its unique geographic location and the pres- of the basics, namely, that we are dealing with real people in ence of a mosaic of different phytogeographic zones. There real places and in real time. It is important to stress that are few if any regions in the world where precipitation and developments appear to have been directional only in ret- vegetation zones vary so markedly over such small distances rospect. The processes that took place were multifaceted, with (Goring-Morris, Hovers, and Belfer-Cohen 2009 and refer- various options available at the time; some of the choices, ences therein). It was against this unusual backdrop that the momentous changes from small groups of mobile foragers to ultimately, were significant to future developments, but others large settled agricultural communities first occurred. These were “sideshows” or culs-de-sac in the evolutionary sense. conditions need to be taken into account when considering Accordingly, within the archaeological record, we may stum- Neolithization processes as a whole. Thus, in order to present ble on evidence for both categories. as coherent a picture as possible, we here discuss and expand As we move through the chronological sequence, increas- on the external issues pertinent to such developments. Belfer- ingly detailed information is available; of course, the question Cohen and Goring-Morris (2011) explore the internal facets arises as to whether this evidence was present earlier and of these phenomena. simply hidden by the foggy curtain of time. We are able to The region is also one in which relatively intensive field identify and isolate more archaeological entities—using tra- and laboratory research has been conducted, albeit still with ditional archaeological criteria—during the earlier Epipaleo- significant “blank areas.” It is fascinating to observe the lithic than previously during the Upper Paleolithic. Yet even changes in the scientific approaches to Neolithization. These if the archaeological cultures differed in specific settlement comprise early simplistic paradigms tracing the inevitable uni- patterns and cultural markers, the general nature of each such directional “progress” (in a Marxist sense) from hunting-gath- entity was quite similar. This similarity most likely relates to ering to agricultural practices by means of a specific trigger the obvious fact that all these groups were mobile hunter- such as climate change (e.g., Childe 1934). At the other end gatherers exploiting their environments by similar means and of the spectrum are complex multifactor approaches em- in similar modes (Bar-Yosef 1981; Byrd 1990; Goring-Morris bracing demographic and social dynamics as well as quan- 1995; Henry 1989). Resources during the Late Glacial Max- titative and simulation studies (to name but a few, see, e.g., imum (LGM) varied from one region to another, reflecting local patchiness in availability and sometimes seasonality. Dia- chronic changes did occur, but, in general, the overall picture A. Nigel Goring-Morris and Anna Belfer-Cohen are Professors in remained broadly stable. Within this frame of reference, grad- the Department of Prehistory, Institute of Archaeology, Hebrew University of Jerusalem (Jerusalem 91905, Israel [goring@ ual demographic increase during the Early and climatically mscc.huji.ac.il]). This paper was submitted 13 XI 09, accepted 12 ameliorated Middle Epipaleolithic would have led to locally XII 10, and electronically published 12 VII 11. increased packing in some areas following previous long-lived

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0004$10.00. DOI: 10.1086/658860 S196 Current Anthropology Volume 52, Supplement 4, October 2011

Table 1. Chronology of cultural entities in the southern Levant based on available radiometric date ranges

Sociocultural entities Dates BP Cal Time stratigraphic units Mediterranean zone Steppe and desert zone 24,000–21,750 Early Epipaleolithic Nebekian 24,200–19,250 Masraqan (Late Ahmarian) Masraqan (Late Ahmarian) 21,250–17,575 Kebaran Kebaran 20,200–18,900 Nizzanan Nizzanan 18,000–16,250 Middle Epipaleolithic Geometric Kebaran Geometric Kebaran 17,000–15,500 Classic Mushabian 16,850–14,400 Early Ramonian 14,900–13,700 Late Epipaleolithic Early Natufian Terminal Ramonian Early Natufian 13,500–12,750 Late Natufian Late Natufian 12,500–11,750 Final Natufian Harifian 12,175–11,800 Early Neolithic PPNA Khiamian 11,625–11,000 Sultanian (Final Harifian) 10,950–10,300 Early Neolithic PPNB Early PPNB Early PPNB 10,150–9725 Middle PPNB Middle PPNB 9400–8900 Late PPNB Late PPNB 9050–8450 Final PPNB (PPNC) Final PPNB (PPNC) 8400–7700 Late Neolithic PNA Yarmukian Tuwailan 7750–7450 Jericho IX 7500–6500 Late Neolithic PNB Wadi Raba Qatifian Timnian Note. There are slight discrepancies between this table and table 1 in Belfer-Cohen and Goring-Morris (2011). PNA p Pottery Neolithic A; PNB p Pottery Neolithic B; PPNA p Pre-Pottery Neolithic A; PPNB p Pre-Pottery Neolithic B; PPNC p Pre-Pottery Neolithic C.

Upper Paleolithic patterns (Goring-Morris, Hovers, and Bel- and incipient processes of Neolithization in the Near East fer-Cohen 2009). should be traced all the way back to the local Late Upper While, superficially, the observed changes during the earlier Paleolithic/Early Epipaleolithic. This corresponds to a chro- Epipaleolithic appear to have been incremental, gradually nological span of some 15,000 years (after calibration) until speeding up so that the pace of change eventually became the end of the Neolithic, that is, the equivalent of some 500– logarithmic, common sense suggests that changes may more 600 generations (table 1). likely have been quite random, isolated, and independent. Obviously, we are limited by the very nature of the archae- The Geographic Setting ological record. The changes would have involved techno- logical innovations, differential social interactions, and re- It is vital to define the geographic boundaries in which these sponses to extraneous factors. Ultimately, it is the specific processes occurred. The “Near (or Middle) East” is an am- articulations of these various factors, internal (intra- and in- biguous term that covers Southwest Asia between the Med- tergroup) and/or external (climate, carrying capacities, etc.) iterranean and Iran. The region (Western Asia) encompasses that are particularly relevant in terms of the nature and tempo Anatolia, the Levant, Cyprus, Mesopotamia, and Transcau- of cultural change. This is one of the pitfalls of simulation casia. Here our primary focus of study is the Levant, its most studies, because they are based on consideration of long distinctive feature being its ecological diversity. In addition, stretches of time and averaging out the data available. Actually, where pertinent, we relate briefly to adjacent regions (e.g., there are indications that at least at certain points in time, Cyprus and central Anatolia). The Levant is a small region the course of events is more apt to be described in the mode enclosed between the Taurus and Zagros mountains to the of “punctuated equilibrium,” with periods of stasis inter- north, the Mediterranean coastline to the west, the Sinai Pen- spersed by bursts of activity (and see below for specific ex- insula to the south, and the Syro-Arabian desert to the east, amples). This pattern is reported from various parts of the ca. 1,000 km north to south by up to 400 km east to west world in the articles in this issue. (fig. 1). The topography of the Levant is characterized by a Another caveat concerns the very nature of the archaeo- north-south longitudinal series of alternating elevated and logical data as well as research paradigms that influence in- low-lying regions: the coastal plain and western piedmont; terpretations. The “rule of thumb” is that what is observed the central hill range reaching up to 2,000 m a.s.l.; the Dead archaeologically more often than not reflects the end of the Sea Rift lying below sea level; and the Trans-Jordanian/Syrian process (from invention to innovation to its wide acceptance), plateau (the central-south Levant), which rises steeply to el- “solid” enough to be observed. evations between 800 and 2,000 m a.s.l., followed by a gradual Considering all of the above, it transpires that the origins descent eastward into Saudi Arabia. Today there are relatively Goring-Morris and Belfer-Cohen Neolithization Processes in the Levant S197

Figure 1. Phytogeographic regions of the Near East. The line between Beirut and Damascus differentiates between southern and northern Le- vant. Note that some relate the Upper Tigris as part of Mesopotamia while others include the Middle Euphrates as southeast Anatolia. few perennial rivers or streams in the region, the most notable regional levels. This is especially valid in comparisons between being the Tigris and the Euphrates in the north and the Oron- the southern and northern Levant. For example, Terminal tes and the Jordan farther south along the rift valley. Almost Pleistocene/Early Holocene geomorphological changes appear all other drainages are seasonal and ephemeral. Changes in to have acted differently in the north and the south, with evapotranspiration rates were major factors in the presence more extensive aggradation in the north that may have caused and extent of Late Quaternary inland water bodies. Springs the burial of sites there (e.g., O¨ zdog˘an 1997). are common in the Mediterranean zone but are widely dis- In this article, we use the term “southern Levant” to dif- persed in more arid areas. Obviously, the specifics would have ferentiate between the area south of an east-west line from changed at various times, depending on the particular climatic the coast across to the Damascus Basin, as opposed to the conditions. It is important to stress that within the Levant, “northern Levant,” stretching from that line north to the as indeed elsewhere, global climatic changes would have dif- Taurus and the western end of the Zagros Mountains, in- ferentially affected environments at the micro- and macro- cluding the Middle Euphrates and the Upper Tigris region S198 Current Anthropology Volume 52, Supplement 4, October 2011

(sometimes called the “Golden Triangle”). Physical and cul- dividuals for minimal sustainable mating networks should be tural interactions are documented between the south and the emphasized (e.g., Johnson 1982; Kosse 1994; Wobst 1976). north during the relevant periods; developments were not Site numbers, site sizes, and their relative durations through always synchronous, and centers of innovation appear to have most of the Upper Paleolithic indicate that demographic in- shifted from the south to the north (but see Watkins 2008, creases were slow and incremental. Relative population den- concerning the so-called primacy of the southern Levant). sities began to grow starting with the Early Epipaleolithic; this Whereas in the past this geographic shift of the hub of Neo- rise in density was initially somewhat artificial, inflated by lithization appeared to reflect more the history of research, local environmental deterioration with the onset of the LGM it is less so the case today, notwithstanding ongoing lacunae that brought about declining carrying capacities in more mar- in field investigations in certain areas. ginal areas. Accordingly, populations congregated in refugia, so relative packing was greater within smaller areas while other The Paleoenvironmental Background areas were, to all intents and purposes, abandoned. In regions such as the eastern Trans-Jordanian steppes, groups could In order to begin to comprehend developments in human congregate seasonally in winter/spring because of the presence adaptations, it is vital to know the physical circumstances of of large herds of the seasonally migratory subspecies of gazelle the landscapes that populations occupied. This is especially Gazella subgutturosa subgutturosa (Martin 1994, 2000; Mu- applicable for areas such as the Levant, which comprises a hesen and Wada 1995 and references therein). Later, during particularistic mosaic in terms of environment and ecology. the Middle Epipaleolithic, significant climatic amelioration Such conditions would dictate fine-tuned adaptation by local relaxed previous physical and ecological constraints, providing populations to specific and frequently seasonal niches. Studies more and different stimuli to both local and overall popu- have demonstrated that climatic changes during the Terminal lation increases (see Grosman 2004 for additional references; Pleistocene and Early Holocene fluctuated markedly, and rap- fig. 2). idly and were often of an intensity hardly encountered during It is important to stress that, indeed, general trends varied much of the Quaternary. We have in mind the time span from from one geographical region to another throughout the Le- the LGM through the Bo¨lling/Allerød, the Younger Dryas, and vant, as reflected by the numbers of sites documented. We then to the Preboreal and the end of the Early Holocene also need to take into account the sizes and the duration of Optimum (e.g., Bar-Matthews and Ayalon 2003; Byrd 2005; sites in attempting to reconstruct actual demographic in- Enzel et al. 2008; Robinson et al. 2006; Wick, Lemcke, and creases. There are likely to have been major jumps in relative Sturm 2003 and references therein). Particularly relevant for population densities within communities that we believe cor- human populations would have been short-term annual or relate with a combination of increasing sedentism (likely to decadal climatic events, which are frequently lost in paleo- induce shorter birth spacing) on the one hand and techno- climatic reconstructions that tend to average out data over logical innovations (enabling more efficient exploitation of longer periods (Grosman and Belfer-Cohen 2002). These cli- the environment) on the other. Clearly, these tendencies are matic shifts influenced vegetation and faunal distributions and apparent within the southern Levant from the Late Epipaleo- densities throughout the region. The most marked influences, lithic Natufian onward. for better or worse, would have been at the interfaces/ecotones With regard to developments in the northern Levant and between the more mesic Mediterranean regions and the semi- adjacent areas, we remain in the dark for much of the Epi- arid peripheries. In terms of human adaptations, these paleolithic (until the very end of that period). The few known changes would have resulted in “push-pull” or “concertina” sites are located in or at the margins of the mountainous effects, alternately bringing about retreats into refugia or en- zones, perhaps reflecting the geomorphological burial of sites abling dispersals into newly opened-up areas. Reconstructions in this region (and see above). Most of these sites were ex- of specific human adaptations and subsistence should take cavated 50 or more years ago, when stratigraphic control and those dynamics into consideration. Additionally, late Qua- excavation techniques were much less rigorous than today. ternary physical changes to the landscape were more marked Whether the apparent absence of pre-Pre-Pottery Neolithic in some areas (e.g., lower sea level at the height of the LGM A sites in more open environments reflects reality or the and its subsequent rise) as well as the extent of lakes—the relative lack of systematic archaeological exploration and pe- Dead Sea Basin and inland shallow lakes farther to the east destrian survey remains unresolved. (e.g., the Azraq, Damascus, el-Kowm, and el-Jafr basins). Setting the Stage: The Natufian Epipaleolithic Settlement Patterns The Late Epipaleolithic Natufian complex lasted for some In addressing the issues of the Early and Middle Epipaleolithic 3,000 years (ca. 100–125 generations), with the Negev/Sinai (ca. 23–15,000 cal BP), consideration of the scale of social local variant of the Final Natufian, the Harifian (Goring-Mor- networks is crucial. Here the “magic” numbers of ca. 25 in- ris 1991), lasting a mere 750 years (!25–30 generations; Stutz dividuals for many mobile-band societies and 250–500 in- 2004). It is crucial to differentiate between the various chro- Goring-Morris and Belfer-Cohen Neolithization Processes in the Levant S199

Figure 2. Changing site distribution densities in various areas of the Near East. Note that the differing longevity of each period is taken into account by presenting the relative number of sites per 1,000 years calibrated for each period (table 1). EEPI p Early Epipaleolithic; EPPNB p Early Pre- Pottery Neolithic B; FPPNB p Final Pre-Pottery Neolithic B (Pre-Pottery Neolithic C); LEPI p Late Epipaleolithic; LN1 p Early Late (Pottery) Neolithic; MEPI p Middle Epipaleolithic; MPPNB p Middle Pre-Pottery Neolithic B; PPNA p Pre-Pottery Neolithic A; PPNB p Pre-Pottery Neolithic B; UP p Upper Paleolithic. nological phases within the Natufian while also acknowledg- and still others on the margins were even more mobile. In- ing the high degree of variability in specific Natufian adap- deed, the Natufian is not just about the sedentary logistic tations at the regional level (Bar-Yosef 2002; Goring-Morris collectors of the Mediterranean zone versus their (“poorer”) and Belfer-Cohen 1997; Valla 1999). In general, during the foraging cousins in the periphery, because there are also in- Late Epipaleolithic, a greater degree of differentiation between termediate situations and interactions. Accordingly, the geo- archaeological entities/cultures can be identified, as some graphic scale of individual territorial ranges would have varied groups became increasingly sedentary, incorporating tech- significantly, and a greater degree of packing of populations nological developments on various levels (including those would have occurred within those areas settled by more sed- connected with food procurement and processing). This led entary communities. to the beginnings of the dichotomy between “the desert and Natufian subsistence modes were based primarily on the the sown”; thus, human groups with different economic intensification of hunting and gathering and associated pro- modes of existence interacted differently with their environ- cessing techniques. This is reflected in the exploitation of a ments. particularly wide spectrum of faunal species coupled with the When discussing the subsistence mode of the Natufian, we intense targeting of their preferred prey, the gazelle (see have to acknowledge the ranges of specific Natufian adap- Munro 2004 and references therein). Although botanical re- tations. While some groups were more or less sedentary in mains are rarely preserved, the large numbers of groundstone favorable ecological settings (e.g., on the shores of lakes or utensils, especially mortars and pounding tools, clearly reflect marshes), others likely practiced seasonal residential mobility, more intensive plant-food preparation than previously (Bel- S200 Current Anthropology Volume 52, Supplement 4, October 2011 fer-Cohen and Hovers 2005 and references therein). There is appear to result from external factors, especially the delete- clear use of sickles for harvesting (whether wild cereals and/ rious effects of the rapid onset of the YoungerDryas (Grosman or reeds) during the Natufian, which has been suggested to and Belfer-Cohen 2002). Indeed, this is reflected most ob- imply more efficient methods of exploitation of smaller areas viously in more marginal regions, such as the Negev and Sinai, (Hillman and Davies 1992). However, the automatic associ- with the finely tuned and short-lived Harifian entity. At the ation of composite tools—blades inserted in bone handles— end of the day, conditions in these peripheral settings dete- with cereal harvesting requires caution, because few if any of riorated beyond a critical threshold, and continued occupa- the inserts actually display evidence for sickle gloss ( Edwards tion of the area simply became untenable. There may have 1991; Garrod and Bate 1937). It appears that while the Na- been no choice but to abandon the region and retreat to join tufian does exhibit intensification of plant collection, there is communities elsewhere—whether in adjacent areas, in which no overt evidence for intentional plant cultivation. The re- the Final Natufians themselves were experiencing precarious liability of claims for Late Natufian domestication of cereals conditions, or even farther afield. at Abu Hureyra remains dubious, given the taphonomy of In the northern Levant, broadly contemporary with the that site (Hillman 2000; Weiss and Zohary 2011). Late Natufian, a series of occupations have been described Without delving into the specifics of social and mating around the edges of the piedmont zone. This “Round-House networks operating during the Natufian here (see Belfer- Horizon” continues in time through the equivalent of the Cohen and Goring-Morris 2011), suffice it to note that in southern Pre-Pottery Neolithic A (PPNA; Peasnall 2000). Ad- some areas community sizes increased significantly, although aptations in the north were based primarily on hunting and the spacing between individual hamlets may have remained gathering of nuts and fruits. There have also been claims for constant. The impact of the growth in community sizes and some degree of management of wild boar (Redding and Ro- their observed territoriality is reflected in disparities between senberg 1998; Vigne et al. 2011). While there have been claims various components of the profane and symbolic material for the initial occupation of Cyprus during the Late Epipa- culture in the Natufian “core area” and more peripheral areas. leolithic, the available evidence for pre-Pre-Pottery Neolithic Indeed, the quantum rise in the number of burials within occupation of the island to date remains equivocal and open “core area” Natufian sites, in areas clearly designated as burial to debate (Ammerman et al. 2006, 2007; Simmons 2001, 2007 grounds, undoubtedly reflects the increasing individual sense and references therein). of “belonging” to specific localities and communities. Com- bined with this is the development of geographically wide- Archaic Villages of the PPNA ranging exchange networks for both mundane (e.g., ground- stone utensils; Weinstein-Evron, Kaufman, and Bird-David The PPNA is a short-lived phenomenon with regard to both 2001) and exotic (e.g., marine mollusks, greenstone, and other what preceded it and what succeeded it—a mere 1,000 ca- minerals; Bar-Yosef Mayer 2005) items, exchange that was lendrical years (or ∼40 generations). However, the scale of previously at best sporadic (Goring-Morris 1989). change for the PPNA in the southern Levant is of a lesser There are also indications for technological developments order of magnitude compared with the northern Levant, in other realms, including the beginnings of what can be where there was a rapid increase in population and in the termed “cottage industries.” This includes circumstantial evi- accompanying elements of social organization. The arid mar- dence from the increased abundance and scope of bone tool gins throughout the Levant were all but deserted during the assemblages for basketry and matting (e.g., Campana 1991; course of the PPNA and were recolonized only during the Le Dosseur 2008; Stordeur 1991; see also the isolated find, a Pre-Pottery Neolithic B (PPNB). wood and flax “comb” from Wadi Murabba’at; Schick et al. The southern PPNA displays numerous elements of con- 1995). As for pyrotechnology, or “playing with fire,” in ad- tinuity from the Natufian. It likely represents the amalga- dition to increased lime-plaster production, we also witness mation of Final Natufian populations in locally favorable set- technological developments with regard to fired clay and tings. This includes the Harifian, which represents Final ochre (Belfer-Cohen 1988; Perrot and Ladiray 1988; Valla Natufian refugees from the arid margins retreating into the 1999; Zackheim 1997). Mediterranean zone because of the cumulative effects of the It is important to stress the documented diachronic changes dire events associated with the Younger Dryas (Goring-Morris in Natufian lifeways with explicit evidence for greater com- 1991). During the PPNA, conditions improved but appear to plexity during the early rather than the later phase. The Early have been marked by torrential episodes (e.g., at Netiv Hag- Natufian displays a high degree of complexity and spatial dud and perhaps also at Jericho; Bar-Yosef 1986; Bar-Yosef, organization, with large well-built structures that were surely Goring-Morris, and Gopher 2010). occupied by units larger than nuclear families (Goring-Morris Major PPNA sites in the southern Levant display a pro- and Belfer-Cohen 2003; Valla 2008). Subsequently, there ap- pensity for lowland settings in a north-south alignment, es- pears to have been a return to domicile by smaller social units pecially along the edges of the rift valley (and also along the in smaller structures. However, the bulk of the evidence for eastern edge of the coastal plain). Smaller seasonal exploita- increasing mobility later on, during the Final Natufian, would tion sites are found in a gradient up the neighboring slopes Goring-Morris and Belfer-Cohen Neolithization Processes in the Levant S201 to the west and to the east. A few sites, such as Jericho and from the area between the Middle Euphrates and the coast Netiv Hagdud, are of a different order of magnitude, each (Abbe`s 2008; Mazurowski and Jamous 2001). encompassing areas of 1.6–2.0 ha, which may have housed The northern intrasite pattern is one of dispersed small up to a couple of hundred inhabitants. The scale of individ- household units together with associated semisubterranean ually spaced oval semisubterranean residential units is indic- communal structures at sites such as Jerf el-Ahmar and Tell ative of nuclear family residences and reflected by the presence ‘Abr 3 (Stordeur 2007; Stordeur et al. 2001). The economy of grinding and pounding installations within each domestic was presumably based on floodplain cultivation, but there is unit (Rosenberg 2008; Wright 2000). Well-constructed exter- little if any solid evidence for domesticates until the end of nal silos are found, although storage was probably also in the PPNA (Willcox 2005). Similarly, there is no unequivocal baskets suspended from beams (e.g., Gilgal and Netiv Hagdud; confirmation for animal domestication (for an overview see Bar-Yosef and Gopher 1997; Bar-Yosef, Goring-Morris, and Zeder 2009, 2011). Hunting often appears to have focused Gopher 2010). At Dhra, a raised silo is documented (Kuijt on herd species such as Gazella subgutturosa subgutturosa and and Finlayson 2009), and at Jericho, what have been inter- Equus hemionus, although in some sites wild boar Sus scrofa preted as larger “plastered” silos around the tower may in- is also common and may have been managed and/or culled dicate some form of community storage (Kenyon 1981). (e.g., Vigne 2008; Vigne et al. 2011). Communal structures are few, with the exception of the These developments would have necessitated changes in enigmatic tower and wall at Jericho. Numerous interpreta- the scale of mating networks, which can be tied in with the tions have been proposed for their functions, ranging from foundation of various ritual-cum-aggregation sites in water- the mundane (defenses against potential enemies or flooding) shed locations (e.g., the PPNA–B hilltop ritual site of Go¨bekli to the cultic (a ritual precinct and/or astronomical device; Tepe; Schmidt 2005, 2007 and references therein). Clearly, the Barkai and Liran 2008; Bar-Yosef 1986; Crowfoot-Payne 1976; nature of the relationships between the various sites, as well Naveh 2003; Ronen and Adler 2001). as between the northern and southern Levant and other areas, Were these farming communities? While they probably underwent transformation. Exchange networks were wide- practiced small-scale cultivation on the alluvial fans on which spread, as exemplified by the systematic acquisition and use of Cappadocian and Bingo¨l obsidian, although no PPNA sites the settlements were located, the plant remains recovered were are presently documented near those sources (Delerue 2007). not always of species that were soon to be domesticated locally Here, it is interesting to note the similarity of the “southern” (e.g., oats Avena sterilis at Gilgal and Netiv Hagdud; Kislev Harifian projectile points to the “northern” Nemrik projectile 1997; Weiss, Kislev, and Hartmann 2006). There is some de- points; are we facing the beginnings of what will be the hall- bate as to whether populations of medium-sized ungulates mark of “convergence” (i.e., the development of a PPNA had already been depleted during the Natufian (Cope 1991 koine—a regional interaction sphere sharing focal cultural and Davis 1983, 1987 vs. Bar-Oz 2004 and Sapir-Hen et al. characteristics—preceding the better-known PPNB one) and/ 2009), but there certainly appears to have been an increasing or actual population movements? While the issue of the sys- focus on smaller species, including birds (Horwitz et al. 2010; tematic colonization of Cyprus during the Late Epipaleolithic Tchernov 1994). The presence of silos may provide indirect is open to some debate, it was clearly well under way during evidence for surplus, which could have contributed to ex- the PPNA (McCartney et al. 2007). Various lines of material panding exchange networks of foodstuffs. In addition, some evidence indicate that the origin of the colonists was probably sites appear to have served as nodes for procurement and from the north Levantine littoral to the Cilician plain. This exchange networks of valued materials (e.g., obsidian, mal- likely reflects major demographic increases in that region. achite, salt, bitumen, seashells, etc.), which may have influ- The demise of the PPNA in the southern Levant displays enced their location and size. This may explain the otherwise little evidence for in situ continuity; most sites were aban- anomalous location of the site of Wadi Faynan 16 in a mar- doned, whether from overexploitation, declining yields, cli- ginal ecological setting at the edge of the rift valley (Finlayson mate shifts, changing geomorphology, shrinking water and Mithen 2007) yet adjacent to abundant sources of mal- sources, or other causes. One should also add the possibility achite. of contagious diseases causing havoc at the local level, because Coevally, in the northern Levant there was a veritable pop- the geographical scale of local interactions was still quite lim- ulation explosion. This trend can already be detected along ited. By contrast, in the north there is unequivocal evidence the Upper Tigris (the “Round-House Horizon”) toward the for direct continuity from the PPNA to the PPNB, which end of the local equivalent of the Late Natufian before the commonly (but not always) took place on-site (e.g., Mur- onset of the Younger Dryas (Peasnall 2000). Farther west, a yebet; Iba´n˜ez 2008). linear riparian settlement pattern of communities developed at intervals of 20–25 km along the Euphrates and its tribu- PPNB Village Society taries. The settlement at Jerf el-Ahmar was initially very small but rapidly increased in size (D. Stordeur, personal com- With the emergence of PPNB village societies, the center of munication). Recently, PPNA sites have also been reported innovation had already shifted to the northern Levant. In the S202 Current Anthropology Volume 52, Supplement 4, October 2011 southern Levant, the shift from PPNA to PPNB was quite way of life here was one of mobile foraging supplanted during abrupt, with the Early PPNB representing but a short bridging the end of the period by pastoral nomadism, which seemingly episode. So it is only with the Middle PPNB that we find the originated in the Syrian Desert and then diffused southward emergence of the full-scale PPNB koine that developed and later westward (Bar-Yosef and Khazanov 1992; Betts 1998; through the Late and Final PPNB. The koine stretched across Goring-Morris 1993). the entire Levant and beyond, even incorporating central An- Specialized sites were located in boundary or neutral geo- atolia and Cyprus (Bar-Yosef and Belfer-Cohen 1989; Cauvin graphic settings, serving most probably as cultic centers, ag- 2000; Esin and Harmankaya 1999; Hodder 2007; O¨ zdog˘an gregation sites, and/or mortuary sites, as, for example, Nahal 2011; Simmons 2007). It is important to note that this in- Hemar cave in the Judean Desert and Kfar HaHoresh in the teraction sphere encompassed a wide spectrum of specific and Nazareth Hills of lower Galilee (Bar-Yosef and Alon 1988; varied adaptations. These included large-scale permanent vil- Goring-Morris 2005). In the north, settlement patterns con- lages based on farming and herding or hunting and farming tinued to be largely linear, with a focus on the Euphrates and and fishing, as well as mobile foraging groups and, toward its tributaries (e.g., the Balikh) and sites usually located at the end of the period, pastoral societies (fig. 3). Because of intervals of 20–25 km from one another (about a day’s walk the nature of local geographic zoning, these adaptations were away). The individual sites expanded significantly in size in all in close proximity to one another. The PPNB coincides comparison with those of the PPNA. The demise of the PPNB with the Early Holocene Climatic Optimum (Byrd 2005), a koine may reflect a combination of factors, including a rapid time of plenty as conditions improved from one year to the climate deterioration (the so-called 8.2-kyr event), the effects next. of some 2,000 years of farming without manuring (Bogaard The southern Levant settlements were initially founded in 2005; Bogaard and Isaakidou 2010), local overexploitation of a narrow corridor along a north-south axis on the western nonsustainable nearby resources (Rollefson and Ko¨hler-Roll- edge of the Trans-Jordanian Plateau (and to some extent the efson 1989; but see Campbell 2010), outbreaks of contagious Jordan Valley) from Aswad to southern Edom along the later diseases, and increased tensions between neighboring com- “King’s Highway” adjacent to major wadi systems. These sites munities (Goring-Morris and Belfer-Cohen 2010). subsequently expanded to become the Late and Final PPNB “megasites,” reaching up to 30 acres in extent (Bienert, Gebel, Discussion and Neef 2004 and references therein). The megasites were accompanied by complementary perpendicular settlement Although most of the above represents a synopsis of hard- patterns along an east-west axis with much smaller sites. The core data, there are points of debate regarding their inter- north-south axis formed the basis for the main long-range pretation and/or incorporation within the evolving matrix of exchange (“down-the-line”) networks, the “megasites” serv- the Levantine Neolithization process. Pertinent among them ing as nodes for subsidiary distribution to the east and west are the following. (Goring-Morris, Hovers, and Belfer-Cohen 2009). Issues of continuity. Contrary to what was believed during Still, there is little direct evidence as to the density of pack- earlier days of research, it transpires that there is no linear ing of architectural units within PPNB sites at any given point development from the Epipaleolithic to the end of the PPNB in time. This is obviously a crucial issue concerning esti- (e.g., Mellaart 1975; Perrot 1968). Each major period wit- mations of village populations, which vary greatly. But, con- nesses a breakdown, being separated from the next by a short- servatively, communities at the top end of the scale numbered term yet major rupture, or “bottleneck,” followed by realign- at least several hundred individuals (Campbell 2010; Kuijt ment of the new system (i.e., the Early PPNA, the Early PPNB, 2000; Simmons 2007). and the Early Pottery Neolithic). However, it is important to Economy displays considerable variability within the Med- stress that cultural complexes build on the past, sharing many iterranean zone. Along the King’s Highway and down in the common traits with their predecessors. They accordingly rep- rift valley, communities subsisted primarily on domestic ce- resent generational links within an extended family rather reals. But farther west, in the Carmel and in the Galilee, the than a simple grandfather-son-grandson chain. emphasis was on lentils (e.g., Garfinkel, Kislev, and Zohary People or ideas in motion. For many years there was debate 1988). So, too, there are differences in the animals exploited. as to whether it was people or ideas that moved and dispersed At Middle PPNB ‘Ain Ghazal, a full 50% of the fauna was from one area to the next. Currently, it is abundantly clear still hunted, a situation that changed dramatically by the Late that both require consideration. Unequivocal examples of the PPNB, with 75% domesticated goats and sheep (Driesch and former are the directed colonization of Cyprus and probably Wodtke 1997; Wasse 1997). But west of the rift—in the Gal- also of central Anatolia (e.g., Guilaine and Le Brun 2003; ilee, Carmel, and Samaria—hunting continued to predomi- O¨ zdog˘an 2011; Peltenberg and Wasse 2004). The latter in- nate throughout (Horwitz et al. 2000; Sapir-Hen et al. 2009). cludes the diffusion of the bidirectional (naviform) chipped- With climatic amelioration, the desert margins of eastern stone technology from the Central Euphrates southward (Bar- Trans-Jordan as well as in the Negev and Sinai either were zilai 2009; and see also Perle`s 2005). The situation with regard recolonized or relict populations increased significantly. The to the diffusion of animal and plant domesticates within the Goring-Morris and Belfer-Cohen Neolithization Processes in the Levant S203

Figure 3. Pre-Pottery Neolithic B koine in the Near East, showing the wide range of economic adaptations throughout the region.

Levant remains uncertain and debated (e.g., Bar-Yosef 1998; ward, diverse types of mating networks are likely to have Horwitz et al. 2000; Nesbitt 2002; Weiss and Zohary 2011; operated. Variability is especially relevant as individual com- and see the never-domesticated animal species introduced to munities became more sedentary and increased in size. Nu- Cyprus in Vigne et al. 2011). Indeed, the apparent retardation merical thresholds were crossed, necessitating realignments in in the diffusion of ideas across northern Sinai to the Nile social intercourse within and between communities because Delta at a time of amelioration is an enigmatic example (Gor- previous social mechanisms were unable to cope with novel ing-Morris 1993; Smith 1989). situations (and see Belfer-Cohen and Goring-Morris 2011). Population increase and demographic transition. Population This makes it all the more difficult to predict or state when increase and the Neolithic demographic transition (Bocquet- and where, precisely, the ADT occurred; should we start Appel and Bar-Yosef 2008) or, more recently, the agricultural counting from the Natufian, the PPNA, or the PPNB? demographic transition (ADT; Bocquet-Appel 2011) in the Differences between developments in the northern and south- Near East are crucial in terms of the independence of the ern Levant. Was the southern Levant during the PPNB an scale of mating networks. From the Early Epipaleolithic on- independent center of innovation? Do developments in this S204 Current Anthropology Volume 52, Supplement 4, October 2011 region simply reflect diffusion of novelties from the Euphrates taken for granted that present-day distributions of the wild region? Or are we witnessing combinations of both? Certainly, precursors broadly correlate with Terminal Pleistocene/Early various facets of the material culture derive from the north, Holocene patterns such that advances in DNA sourcing would such as bidirectional naviform lithic technology. On the other enable the straightforward pinpointing of domestication lo- hand, the evidence regarding the origins of domesticates is calities (e.g., Heun et al. 1997; Kilian et al. 2006, 2007; Naderi more equivocal. et al. 2007). Others have argued for a single hearth of do- The when, where, how, and why of plant and animal do- mestication for the whole package of founder crops (Lev- mestication. Archaeological evidence clearly demonstrates that Yadun, Gopher, and Abbo 2000). Notwithstanding attempts there were more than two variants of economic existence; it to reconstruct late Quaternary distributions (Hillman 2000), was not just about farmers versus foragers. Fully fledged set- current distributions of wild plants and animals surely have tled agriculturalists did coexist with hunter-gatherers during been affected by almost 10 millennia of agriculture and pas- the Pre-Pottery Neolithic, but other groups, for a while at toralism. Furthermore, if domestication of plants and animals least, filled intermediate roles as forager-farmers so that there was a gradual process, then there are likely to have been was a wide spectrum of lifeways (with the geographic distances considerable genetic interactions between the “predomesti- involved being very small). Indeed, it is only toward the end cated” but manipulated/cultivated species and wild popula- of the Late Neolithic that subsistence throughout the region tions. These interactions probably continued, affecting the became based mostly on domesticated plants and animals genetic fingerprints of both the wild progenitors and the early (with the continuing dichotomy of agriculturists and pastor- domesticated forms. alists, as between “the desert and the sown”). This complexity If we take into consideration all of the above, the concept should be acknowledged when we try to interpret the ar- of Neolithization involved much more than plant and animal chaeological record, just as in the case of the rise of a new domestication, for Neolithization processes also involved the species. Even when we recognize the new species that will “domestication” of fire (pyrotechnological developments eventually replace its parent species, both the parent species leading eventually to pottery production) and water (man- and the offspring species coexist for some time, a simile to agement in the form of wells and irrigation). Additionally, be aware of while deciphering the archaeological data at hand. and of paramount significance, is social “domestication” with Plant and animal domestication processes. For how long was new means of molding community identity and interaction, cultivation and animal herding practiced before their recog- whose very essence changed; these range from bonding nizable signatures in DNA, morphological changes, or age through kinship, exchange networks, craft specialization, composition? What was the duration of plant and animal feasting, and so on, to rivalry, political boundaries, and intra- domestication processes? Domestication can be accomplished and intercommunity confrontational violence. Ultimately, the quite rapidly, but it seems more likely that initial Neolithic “Neolithic revolution,” in the Near East at least, was a long- domestication in the Levant was a prolonged and undirected term, incremental, and undirected process marked by signif- process. Not all those species that were intensively harvested icant threshold events, the outcome of which was by no means and/or cultivated during the Early Neolithic were then do- certain. mesticated (e.g., Avena sterilis and Secale cereale; Weiss, Kislev, It is important to stress the novel nature of innovations and Hartmann 2006; Willcox 2005). And perhaps the same associated with Neolithization processes. This is especially rel- observations are valid with regard to the culling of wild an- evant when considering ethnographic analogies. Observed imals (and see the situation in Cyprus with Dama mesopo- phenomena in the archaeological record mostly reflect oc- tamica; Vigne et al. 2011; Zeder 2011). currences that had “made it in the first cut” if not later on Prey stress, overhunting, and climatic changes with regard to in the sequence. But, surely, there is also evidence of phe- animal domestication. Contrary to previous assertions that nomena and innovations that ultimately did not pass the trials initial animal domestication focused on stocking the larder of time. We thus face amalgamations of both material culture for meat, recently it has been claimed that the process was evidence of what was ultimately successful and what fell by directed first at milk products (Vigne and Helmer 2007). the wayside, as we discuss in Belfer-Cohen and Goring-Morris Should we be looking at the central areas or the margins (and (2011). see Binford 1968, 2001; Flannery 1969)? Last but not least, we should bear in mind that domestication processes, whether of plants or animals, were intricately tied in with the non- Acknowledgments material realms of human existence. Those promoted and encouraged the domestication venue, paced its tempo, or even We are grateful to the organizers (Doug Price and Ofer Bar- arrested its progress. In short, one can be quite sure that the Yosef) and fellow participants of the Temozon Wenner-Gren social realm controlled to a certain extent both the domes- workshop (Merida, Yucatan) in March 2009 for a most stim- tication processes themselves and their tempo and extent (and ulating and intensive exchange of data and opinions con- see Belfer-Cohen and Goring-Morris 2011). cerning agricultural beginnings around the world. Thanks are Primary habitats and hearths of domestication. It has been due to the Wenner-Gren Foundation for the superb organi- Goring-Morris and Belfer-Cohen Neolithization Processes in the Levant S205 zation of the event and especially to Leslie Aiello and Laurie as reflected through lithic studies: the bidirectional blade indus- Obbink for their personal commitments to the success of the tries. PhD thesis, Hebrew University of Jerusalem. Belfer-Cohen, A. 1988. The Natufian settlement at Hayonim cave: a meeting. We also acknowledge the support of the Israel Sci- hunter-gatherer band on the threshold of agriculture. PhD thesis, ence Foundation funded by the Israel Academy of Sciences Hebrew University of Jerusalem. (A. N. Goring-Morris: grants 840/01, 558/04, and 755/07; A. Belfer-Cohen, A., and E. Hovers. 2005. The groundstone assemblages Belfer-Cohen: grants 855/03 and 202/05). of the Natufian and Neolithic societies in the Levant: a brief review. Journal of the Israel Prehistoric Society 35:299–308. Belfer-Cohen, A., and A. N. Goring-Morris. 2011. Becoming farmers: References Cited the inside story. Current Anthropology 52(suppl. 4):S209–S220. Bellwood, P. 2005. First farmers: the origins of agricultural societies. Abbe`s, F. 2008. Wadi Tumbaq 1: a Khiamian occupation in the Bal’as Malden, MA: Blackwell. Mountains. Neo-Lithics 1(8):3–9. Betts, A., ed. 1998. The Harra and the Hamad: excavations and surveys Ammerman, A., P. Flourentzos, R. Gabrielli, C. McCartney, J. Noller, in eastern Jordan, vol. 1. Sheffield Archaeological Monographs 9. D. Peloso, and D. Sorabji. 2007. More on the new early sites on Sheffield: Sheffield Academic Press. Cyprus. Report of the Department of Antiquities, Cyprus 2007:1– Bienert, H. D., H.-G. K. Gebel, and R. Neef, eds. 2004. Central 26. settlements in Neolithic Jordan. Berlin: Ex Oriente. Ammerman, A., P. Flourentzos, C. McCartney, J. Noller, and D. Binford, L. R. 1968. Post-Pleistocene adaptations. In New perspectives Sorabji. 2006. Two new early sites in Cyprus. Report of the De- in archeology. S. Binford and L. R. Binford, eds. Pp. 313–341. partment of Antiquities, Cyprus 2006:1–22. Chicago: Aldine. Barkai, R., and R. Liran. 2008. Midsummer sunset at Neolithic Jer- ———. 2001. Constructing frames of reference: an analytical method icho. Time and Mind: Journal of Archaeology, Consciousness, and for archaeological theory building using ethnographic and environ- Culture 1:273–284. mental data sets. Berkeley: University of California Press. Bar-Matthews, M., and A. Ayalon. 2003. Climatic conditions in the Bocquet-Appel, J.-P. 2011. The agricultural demographic transition eastern Mediterranean during the Last Glacial (60–10 ky BP) and during and after the agriculture inventions. Current Anthropology their relations to the Upper Palaeolithic in the Levant: oxygen and 52(suppl. 4):S497–S510. carbon isotope systematics of cave deposits. In More than meets Bocquet-Appel, J.-P., and O. Bar-Yosef, eds. 2008. The Neolithic de- the eye: studies on Upper Palaeolithic diversity in the Near East.A. mographic transition and its consequences. Dordrecht: Springer. N. Goring-Morris and A. Belfer-Cohen, eds. Pp. 13–18. Oxford: Bogaard, A. 2005. “Garden agriculture” and the nature of early farm- Oxbow. ing in Europe and the Near East. World Archaeology 37:177–196. Bar-Oz, G. 2004. Epipalaeolithic subsistence strategies in the Levant: a Bogaard, A., and V. Isaakidou. 2010. From megasites to farmsteads: zooarchaeological perspective. American School of Prehistoric Re- community size, ideology and the nature of early farming land- search Monograph Series. Boston: Brill. scapes in Western Asia and Europe. In Landscapes in transition: Bar-Yosef, O. 1981. The Epi-palaeolithic complexes in the southern understanding hunter-gatherer and farming landscapes in the Early Levant. In Pre´histoire du Levant: chronologie et l’organisation de Holocene of Europe and the Levant. B. Finlayson and G. Warren, l’espace depuis les origines jusqu’au VIe`me mille`naire. J. Cauvin and eds. Pp. 192–207. Levant Supplementary Series, vol. 8. Oxford: P. Sanlaville, eds. Pp. 389–408. Lyon: Maison de L’Orient. Oxbow. ———. 1986. The walls of Jericho: an alternative explanation. Cur- Boserup, E. 1981. Population and technological change: a study of long- rent Anthropology 27:157–162. term trends. Chicago: University of Chicago Press. ———, ed. 1998. The transition to agriculture in the Old World. Byrd, B. F. 1990. Late Pleistocene assemblage diversity in the Azraq Thematic issue, Review of Archaeology 19. Basin. Pale´orient 14:257–265. ———. 2002. Natufian: a complex society of foragers. In Beyond ———. 2005. Reassessing the emergence of village life in the Near foraging and collecting: evolutionary change in hunter-gatherer set- East. Journal of Archaeological Research 13:231–290. tlement systems. B. Fitzhugh and J. Habu, eds. Pp. 91–149. New Campana, D. V. 1991. Bone implements from Hayonim cave: some York: Kluwer Academic/Plenum. relevant issues. In The Natufian culture in the Levant. O. Bar-Yosef Bar-Yosef, O., and D. Alon. 1988. Nahal Hemar cave. ‘Atiqot English and F. R. Valla, eds. Pp. 459–466. Ann Arbor, MI: International Series, vol. 18. Jerusalem: Department of Antiquities and Muse- Monographs in Prehistory. ums. Campbell, D. 2010. Modelling the agricultural impacts of the earliest Bar-Yosef, O., and A. Belfer-Cohen. 1989. The origins of sedentism large villages at the Pre-Pottery Neolithic–Pottery Neolithic tran- and farming communities in the Levant. Journal of World Prehis- sition. In Landscapes in transition: understanding hunter-gatherer tory 40:447–498. and farming landscapes in the Early Holocene of Europe and the Bar-Yosef, O., and A. Gopher, eds. 1997. An Early Neolithic village Levant. B. Finlayson and G. Warren, eds. Pp. 173–183. Levant in the Jordan Valley. I. The archaeology of Netiv Hagdud. American Supplementary Series, vol. 8. Oxford: Oxbow. School of Prehistoric Research Bulletin 43. Cambridge, MA: Pea- Cauvin, J. 2000. The birth of the gods and the origins of agriculture. body Museum of Archaeology and Ethnology, Harvard University. Cambridge: Cambridge University Press. Bar-Yosef, O., A. N. Goring-Morris, and A. Gopher, eds. 2010. Gilgal: Childe, V. G. 1934. New light on the most ancient Near East.New excavations at Early Neolithic sites in the lower Jordan Valley: the York: Kegan Paul, Trench, Trubner. excavations of Tamar Noy. Oakville, CT: ASPR & David Brown/ Cope, C. 1991. Gazelle hunting strategies in the southern Levant. In Oxbow. The Natufian culture in the Levant. O. Bar-Yosef and F. R. Valla, Bar-Yosef, O., and A. Khazanov, eds. 1992. Pastoralism in the Levant. eds. Pp. 341–358. Ann Arbor, MI: International Monographs in Monographs in World Archaeology, vol. 10. Madison, WI: Pre- Prehistory. history. Crowfoot-Payne, J. 1976. The terminology of the Aceramic Neolithic Bar-Yosef Mayer, D. E. 2005. The exploitation of shells as beads in period in the Levant. In Deuxie`me colloque sur la terminologie de the Palaeolithic and Neolithic of the Levant. Pale´orient 31:176– la pre´histoire du Proche-Orient. R. Wendorf, ed. Pp. 131–137. Nice: 185. UISPP. Barzilai, O. 2009. Social complexity in the southern Levantine PPNB Davis, S. J. M. 1983. The age profiles of gazelles predated by ancient S206 Current Anthropology Volume 52, Supplement 4, October 2011

man in Israel: possible evidence for a shift from seasonality to Goring-Morris, A. N., E. Hovers, and A. Belfer-Cohen. 2009. The sedentism in the Natufian. Pale´orient 9:55–62. dynamics of Pleistocene settlement patterns and human adapta- ———. 1987. The archaeology of bones. London: Batsford. tions in the Levant: an overview. In Transitions in prehistory: papers Delerue, S. 2007. L’obsidienne dans le processus de neolithisation du in honor of Ofer Bar-Yosef. J. J. Shea and D. Lieberman, eds. Pp. Proche-Orient (12000–6500 av. J.-C. cal.). PhD dissertation, Uni- 187–254. Oakville, CT: Brown. versite´ Bordeaux 3. Grosman, L. 2004. Computer simulation of variables in models of Driesch, A. von den, and U. Wodtke. 1997. The fauna of ‘Ain Ghazal, the origins of agriculture in the Levant. PhD dissertation, Hebrew a major PPN and Early PN settlement in central Jordan. In The University of Jerusalem. prehistory of Jordan, II. H.-G. K. Gebel and G. O. Rollefson, eds. Grosman, L., and A. Belfer-Cohen. 2002. Zooming onto the “Younger Pp. 511–556. Berlin: Ex Oriente. Dryas.” In The dawn of farming in the Near East.R.T.J.Cappers Edwards, P. C. 1991. Wadi Hammeh 27: an Early Natufian site at and S. Bottema, eds. Pp. 49–54. Berlin: Ex Oriente. Pella, Jordan. In The Natufian culture in the Levant. O. Bar-Yosef Guilaine, J., and A. Le Brun, eds. 2003. Le ne´olithique de Chypre. and F. R. Valla, eds. Pp. 123–148. Ann Arbor, MI: International Bulletin de correspondence helle´nique, supple´ment 43. Athens: Monographs in Prehistory. E´ cole franc¸aise d’Athe`nes. Enzel, Y., R. Amit, U. Dayan, O. Crouvi, R. Kahana, B. Ziv, and D. Hayden, B. 1990. Nimrods, piscators, pluckers, and planters: the Sharon. 2008. The climatic and physiographic controls of the east- emergence of food production. Journal of Anthropological Archae- ern Mediterranean over the Late Pleistocene climates in the south- ology 9:31–69. ern Levant and its neighboring deserts. Global and Planetary Henry, D. O. 1989. From foraging to agriculture: the Levant at the Change 60:165–192. end of the Ice Age. Philadelphia: University of Pennsylvania Press. Esin, U., and S. Harmankaya. 1999. Asikli. In Neolithic in Turkey: Heun, M., R. Schafer-Pregl, D. Klawan, R. Castagna, M. Accerbi, B. the cradle of civilization: new discoveries.M.O¨ zdog˘an and N. Bas- Borghi, and F. Salamini. 1997. Site of einkorn wheat domestication gelen, eds. Pp. 115–131. Istanbul: Arkeologie ve Sanat Yayinlari. identified by DNA fingerprinting. Science 278:1312–1314. Finlayson, B., and S. Mithen, eds. 2007. The early prehistory of Wadi Hillman, G. C. 2000. Overview: the plant-based components of sub- Faynan, southern Jordan. Oxford: Oxbow. sistence in Abu Hureyra 1 and 2. In Village on the Euphrates: from Flannery, K. V. 1969. Origins and ecological effects of early domes- foraging to farming at Abu Hureyra. A. M. T. Moore, G. C. Hillman, tication in Iran and the Near East. In The domestication and ex- and A. J. Legge, eds. Pp. 416–422. Oxford: Oxford University Press. ploitation of plants and animals. P. J. Ucko and G. W. Dimbleby, Hillman, G. C., and M. S. Davies. 1992. Domestication rates in wild eds. Pp. 23–53. London: Duckworth. wheats and barley under primitive cultivation. In Pre´histoire de Garfinkel, Y., M. E. Kislev, and D. Zohary. 1988. Lentil in the Pre- l’agriculture: nouvelles approches experimentales et ethnographiques. Pottery Neolithic B Yiftah’el: additional evidence of its early do- P. C. Anderson-Gerfaud, ed. Pp. 113–158. Paris: CNRS. mestication. Israel Journal of Botany 37:49–51. Hodder, I. 2007. C¸atalho¨yo¨k in the context of the Middle Eastern Garrod, D. A. E., and D. M. A. Bate. 1937. Excavations at the Wadi Neolithic. Annual Review of Anthropology 36:105–120. Mughara,vol.1ofThe Stone Age of Mount Carmel. Oxford: Clar- Horwitz, L. K., T. Simmons, O. Lernau, and E. Tchernov. 2010. Fauna endon. from the sites of Gilgal I, II and III. In Gilgal: excavations at Early Goring-Morris, A. N. 1989. Socio-cultural aspects of marine mollusc Neolithic sites in the Lower Jordan Valley: the excavations of Tamar use in the Terminal Pleistocene of the Negev and Sinai regions of Noy. O. Bar-Yosef, A. N. Goring-Morris, and A. Gopher, eds. Pp. the southern Levant. In Proceedings of the shell bead conference: 263–296. Oakville, CT: Brown. selected papers. C. F. Hayes III, L. Ceci, and C. C. Bodner, eds. Pp. Horwitz, L. K., E. Tchernov, P. Ducos, C. Becker, A. von den Driesch, 175–188. Rochester Museum and Science Center Research Rec- L. Martin, and A. N. Garrard. 2000. Animal domestication in the ords, no. 20. Rochester, NY: Rochester Museum and Science Cen- southern Levant. Pale´orient 25(2):63–80. ter. Iba´n˜ez, J. J., ed. 2008. Le site ne´olithique de Tell Mureybet (Syrie du ———. 1991. The Harifian of the southern Levant. In The Natufian Nord): en hommage a`Jacques Cauvin. BAR International Series culture in the Levant. O. Bar-Yosef and F. R. Valla, eds. Pp. 173– 1843. Oxford: Archaeopress. 234. Ann Arbor, MI: International Monographs in Prehistory. Johnson, G. A. 1982. Organizational structure and scalar stress. In ———. 1993. From foraging to herding in the Negev and Sinai: the Theory and explanation in archaeology. C. A. Renfrew, M. J. Row- Early to Late Neolithic transition. Pale´orient 19:63–87. lands, and B. A. Segraves, eds. Pp. 389–421. London: Academic ———. 1995. Complex hunter-gatherers at the end of the Paleolithic Press. (20,000–10,000 BP). In The archaeology of society in the Holy Land. Kenyon, K. M. 1981. The architecture and stratigraphy of the Tell, vol. T. E. Levy, ed. Pp. 141–168. London: Leicester University Press. 3ofExcavations at Jericho. London: British School of Archaeology ———. 2005. Life, death and the emergence of differential status in in Jerusalem. the Near Eastern Neolithic: evidence from Kfar HaHoresh, lower Kilian, B., H. O¨ zkan, J. Kohl, A. von Haeseler, F. Barale, O. Deusch, Galilee, Israel. In Archaeological perspectives on the transmission and A. Brandolini, C. Yucel, W. Martin, and F. Salamini. 2006. Hap- transformation of culture in the eastern Mediterranean. J. Clark, ed. lotype structure at seven barley genes: relevance to gene pool bot- Pp. 89–105. Oxford: Oxbow. tlenecks, phylogeny of ear type and site of barley domestication. Goring-Morris, A. N., and A. Belfer-Cohen. 1997. The articulation Molecular Genetics and Genomics 276:230–241. of cultural processes and Late Quaternary environmental changes Kilian, B., H. O¨ zkan, A. Walther, J. Kohl, T. Dagan, F. Salamini, and in Cisjordan. Pale´orient 23:71–93. W. W. Martin. 2007. Molecular diversity at 18 loci in 321 wild ———. 2003. Structures and dwellings in the Upper and Epi- and 92 domesticate lines reveal no reduction of nucleotide diversity Palaeolithic (ca. 42–10 K BP) Levant: profane and symbolic uses. during Triticum monococcum (einkorn) domestication: implica- In Perceived landscapes and built environments. O. Soffer, S. Vasil’ev, tions for the origin of agriculture. Molecular Biology and Evolution and J. Kozlowski, eds. Pp. 65–81. BAR International Series 1122. 24:2657–2668. Oxford: Archaeopress. Kislev, M. E. 1997. Early agriculture and paleoecology of Netiv Hag- ———. 2010. “Great Expectations”; or, the inevitable collapse of the dud. In An Early Neolithic village in the Jordan Valley. I. The ar- Early Neolithic in the Near East. In Early village society in global chaeology of Netiv Hagdud. O. Bar-Yosef, and A. Gopher, eds. Pp. perspective. M. S. Bandy, and J. Fox, eds. Pp. 62–77. Tucson: Uni- 209–236. Cambridge, MA: Peabody Museum of Archaeology and versity of Arizona Press. Ethnology, Harvard University. Goring-Morris and Belfer-Cohen Neolithization Processes in the Levant S207

Kosse, K. 1994. The evolution of large, complex groups: a hypothesis. (Guinea) model for pig domestication in the Middle East. MASCA Journal of Anthropological Archaeology 13:35–50. Research Papers in Science and Archaeology 15:65–76. Kuijt, I. 2000. People and space in early agricultural villages: exploring Redman, C. 1978. The rise of civilization. San Francisco: W. H. Free- daily lives, community size, and architecture in the Late Pre-Pot- man. tery Neolithic. Journal of Anthropological Archaeology 19:75–102. Rindos, D. 1984. The origins of agriculture. Orlando, FL: Academic Kuijt, I., and B. Finlayson. 2009. Evidence for food storage and pre- Press. domestication granaries 11,000 years ago in the Jordan Valley. Robinson, S. A., S. Black, B. W. Sellwood, and P. J. Valdes. 2006. A Proceedings of the National Academy of Sciences of the USA 106: review of palaeoclimates and palaeoenvironments in the Levant 10966–10970. and eastern Mediterranean from 25,000 to 5000 years BP: setting Le Dosseur, G. 2008. La place de l’industrie osseuse dans la neolith- the environmental background for the evolution of human civ- isation au Levant sud. Pale´orient 34:59–89. ilisation. Quaternary Science Reviews 25:1517–1541. Lev-Yadun, S., A. Gopher, and S. Abbo. 2000. The cradle of agri- Rollefson, G. O., and I. Ko¨hler-Rollefson. 1989. The collapse of Early culture. Science 288(5741):1602–1603. Neolithic settlements in the southern Levant. In People and culture Martin, L. A. 1994. Hunting and herding in a semi-arid region: an in change. I. Hershkovitz, ed. Pp. 73–89. BAR International Series archaeological and ethological study of the faunal remains from 508. Oxford: Archaeopress. the Epipalaeolithic and Neolithic of eastern Jordan. PhD thesis, Ronen, A., and D. Adler. 2001. The walls of Jericho were magical. University of Sheffield. Archaeology, Ethnology and Anthropology of Eurasia 2:97–103. ———. 2000. Gazelle (Gazella spp.) behavioural ecology: predicting Rosenberg, D. 2008. Serving meals making a home: the PPNA lime- animal behaviour for prehistoric environments in Southwest Asia. stone vessel industry of the southern Levant and its importance Journal of the Zoological Society of London 250:13–30. to the Neolithic revolution. Pale´orient 34(1):23–32. Mazurowski, R., and B. Jamous. 2001. Tell Qaramel: excavations Sapir-Hen, L., G. Bar-Oz, H. Khalaily, and T. Dayan. 2009. Gazelle 2000. Polish Archaeology in the Mediterranean 12:327–341. exploitation in the Early Neolithic site of Motza, Israel: the last of McCartney, C., S. W. Manning, D. Sewell, and S. T. Stewart. 2007. the gazelle hunters in the southern Levant. Journal of Archaeological The EENC 2006 field season: excavations at Agia Varvara- Science 36:1538–1546. Asprokremmos and survey of the local Early Holocene landscape. Schick, T., E. Werker, C. Shimony, and G. Bonani. 1995. A 10,000 Report of the Department of Antiquities, Cyprus 2007:27–44. year old comb from Wadi Murabba’at in the Judean Desert. ‘Atiqot Mellaart, J. 1975. The Neolithic of the Near East. London: Thames & 27:199–206. Hudson. Schmidt, K. 2005. “Ritual centers” and the Neolithisation of Upper Muheisen, M., and H. Wada. 1995. An analysis of the microliths at Mesopotamia. Neo-Lithics 2(5):13–21. Kharaneh IV, phase D, square A20/37. Pale´orient 21:75–95. ———. 2007. Go¨bekli Tepe. In Vor 12.000 Jahren in Anatolien: die Munro, N. D. 2004. Zooarchaeological measures of hunting pressure altesten Monumente der Menschheit. C. Lichter, ed. Pp. 74–78. and occupation intensity in the Natufian: implications for agri- Karlsruhe: Badisches Landesmuseum cultural origins. Current Anthropology 45(suppl.):S5–S33. Simmons, A. H. 2001. The first humans and last pygmy hippopotami Naderi, S., H. R. Rezaei, P. Taberlet, S. Zundel, S. A. Rafat, H. R. of Cyprus. In The earliest prehistory of Cyprus: from first colonization Nagash, M. A. El-Barody, O. Ertugrul, and F. Pompanon. 2007. to exploitation. S. Swiny, ed. Pp. 1–18. CAARI Monograph Series Large-scale mitochondrial DNA analysis of the domestic goat re- 2. ASOR Archaeological Reports 5. Boston: American Schools of veals six haplogroups with high diversity. PLoS ONE 10:1–23. Oriental Research. ———. 2007. The Neolithic revolution in the Near East: transforming Naveh, D. 2003. PPNA Jericho: a socio-political perspective. Cam- the human landscape. Tucson: University of Arizona Press. bridge Archaeological Journal 13:83–96. Smith, A. 1989. The Near Eastern connection: Early to Mid-Holocene Nesbitt, M. 2002. When and where did domesticated cereals first relations between North Africa and the Levant. In Late prehistory occur in Southwest Asia? In The dawn of farming in the Near East. of the Nile Basin and the Sahara. L. Krzyzaniak and M. Kobusiewcz, R. T. J. Cappers and S. Bottema, eds. Pp. 113–132. Berlin: Ex eds. Pp. 69–78. Studies in African Archaeology, vol. 2. Poznan: Oriente. Poznan Archaeological Museum. O¨ zdog˘an, M. 1997. Anatolia from the Last Glacial Maximum to the Stordeur, D. 1991. Le Natoufien et son evolution a travers les artifacts Holocene Climatic Optimum: cultural formations and the impact en os. In The Natufian culture in the Levant. O. Bar-Yosef and F. of the environmental setting. Pale´orient 23:25–38. Valla, eds. Pp. 467–482. Ann Arbor, MI: International Monographs ———. 2011. Archaeological evidence on the westward expansion in Prehistory. of farming communities from eastern Anatolia to the Aegean and ———. 2007. Les baˆtiments collectifs des premiers Ne´olithiques de the Balkans. Current Anthropology 52(suppl. 4):S415–S430. l’Euphrate: creation, standardisation et me`moire des formes ar- Peasnall, B. L. 2000. The Round House Horizon along the Taurus- chitecturales. Subartu 17:19–31. Zagros Arc: a synthesis of recent excavations of Late Epipaleolithic Stordeur, D., M. Brenet, G. Der Aprahamian, and J.-C. Roux. 2001. and Early Aceramic sites in southeastern Anatolia and northeastern Les baˆtiments communautaires de Jerf el-Ahmar et Mureybet Ho- Iraq. PhD dissertation, University of Pennsylvania, Philadelphia. rizon PPNA (Syrie). Pale´orient 26:29–44. Peltenburg, E., and A. Wasse, eds. 2004. Neolithic revolution: new Stutz, A. J. 2004. The Natufian in real time? radiocarbon date cali- perspectives on Southwest Asia in light of recent discoveries on Cyprus. bration as a tool for understanding Natufian societies and their Oxford: Oxbow. long-term prehistoric context. In The end of the hunter-gatherer Perle`s, C. 2005. From the Near East to Greece: let’s reverse the focus: societies in the Near East. C. Delage, ed. Pp. 13–37. BAR Inter- cultural elements that didn’t transfer. In How did farming reach national Series 1320. Oxford: Archaeopress. Europe? C. Lichter, ed. Pp. 275–290. Istanbul: Ege Yayinlari. Tchernov, E. 1994. An Early Neolithic village in the Jordan Valley. II. Perrot, J. 1968. La pre´histoire palestinie`nne. In Supplement au dic- The fauna of Netiv Hagdud. American Schools of Prehistoric Re- tionnaire de la Bible, vol. 8. Pp. 286–446. Paris: Letougey & Ane. search Bulletin, no. 44. Cambridge: Peabody Museum of Archae- Perrot, J., and D. Ladiray. 1988. Les hommes de Mallaha (Eynan), ology and Ethnology, Harvard University. Israel.Me´moires et travaux du Centre de Recherche Franc¸aise de Valla, F. R. 1999. The Natufian: a coherent thought? In Dorothy Je´rusalem, no. 7. Paris: Association Pale´orient. Garrod and the progress of the Palaeolithic. W. Davies and R. Redding, R. W., and M. Rosenberg. 1998. Ancestral pigs: a New Charles, eds. Pp. 224–241. Oxford: Oxbow. S208 Current Anthropology Volume 52, Supplement 4, October 2011

———. 2008. L’homme et l’habitat. l’invention de la maison durant stones: basalt implements as evidence for trade/exchange in the la pre´histoire. Paris: CNRS. Levantine Epipalaeolithic. Journal of the Israel Prehistoric Society— Verhoeven, M. 2004. Beyond boundaries: nature culture and a holistic Mitekufat Haeven 31:25–42. approach to domestication in the Levant. Journal of World Pre- Weiss, E., M. E. Kislev, and A. Hartmann. 2006. Autonomous cul- history 18:179–282. tivation before domestication. Science 312:1608–1610. Vigne, J.-D. 2008. Zooarchaeological aspects of the Neolithic diet Weiss, E., and D. Zohary. 2011. The Neolithic Southwest Asian foun- transition in the Near East and Europe, and their putative rela- der crops: their biology and archaeobotany. Current Anthropology tionships with the Neolithic demographic transition. In The Neo- 52(suppl. 4):S237–S254. lithic demographic transition and its consequences.J.P.Bocquet- Wick, L., G. Lemcke, and M. Sturm. 2003. Evidence of Late Glacial Appel and O. Bar-Yosef, eds. Pp. 179–205: New York: Springer. and Holocene climatic change in eastern Anatolia: high-resolution Vigne, J.-D., I. Carre`re, F. Briois, and J. Guilaine. 2011. The early pollen, charcoal, isotopic and geochemical records from the lam- process of mammal domestication in the Near East: new evidence inated sediments of Lake Van, Turkey. Holocene 13:665–675. from the Pre-Neolithic and Pre-Pottery Neolithic in Cyprus. Cur- Willcox, G. 2005. The distribution, natural habitats and availability rent Anthropology 52(suppl. 4):S255–S271. of wild cereals in relation to their domestication in the Near East: Vigne, J.-D., and D. Helmer. 2007. Was milk a “secondary product” multiple events, multiple centres. Vegetation History and Archaeo- in the Old World Neolithisation process? its role in the domes- botany 14:534–541. tication of cattle, sheep and goats. Anthropozoologica 43:9–40. Wobst, H. M. 1976. Locational relationships in Palaeolithic society. Wasse, A. 1997. Preliminary results of an analysis of sheep and goat Journal of Human Evolution 5:49–58. bones from ‘Ain Ghazal, Jordan. In The prehistory of Jordan II: Wright, K. I. 2000. The social origins of cooking and dining in early perspectives from 1997. H.-G. K. Gebel, Z. Kafafi, and G. O. Roll- villages of western Asia. Proceedings of the Prehistoric Society 66: efson, eds. Pp. 575–592. Berlin: Ex Oriente. 89–121. Watkins, T. 2005. The Neolithic revolution and the emergence of Zackheim, O. 1997. Provenance of ochre from the Natufian assem- humanity: a cognitive approach to the first comprehensive world- blages of el-Wad, Eynan and Hayonim. MA thesis, University of view. In Archaeological perspectives on the transmission and trans- Haifa. formation of culture in the eastern Mediterranean. J. Clark, ed. Pp. Zeder, M. A. 2009. The Neolithic macro-(r)evolution: macroevolu- 84–88. Oxford: Oxbow. tionary theory and the study of culture change. Journal of Ar- ———. 2008. Supra-regional networks in the Neolithic of Southwest chaeological Research 17:1–63. Asia. Journal of World Prehistory 21:139–171. ———. 2011. The origins of agriculture in the Near East. Current Weinstein-Evron, M., D. Kaufman, and N. Bird-David. 2001. Rolling Anthropology 52(suppl. 4):S221–S235. Current Anthropology Volume 52, Supplement 4, October 2011 S209

Becoming Farmers: The Inside Story

by Anna Belfer-Cohen and A. Nigel Goring-Morris

Neolithization processes in the Levant differed from those in Europe. A major population growth was already occurring in the former at the onset of the Late Glacial Maximum. Population growth was not linear but rather reflected local circumstances, both external and internal. In addition to changing environmental conditions, the social implications of growth in community sizes within specific areas should be taken into account. The solutions and mechanisms that people devised during the transition to agriculture in order to counter the stresses stemming from those developments pertain to the tempo and scope of the changes as well as to endemic traditions.

Introduction lations as living fossils. Facing the problem of translating ma- terial data into behavioral patterns, one should also remember that human conduct is frequently unpredictable, and hence This article focuses on the internal social issues associated the application of common sense and optimized models with the processes of “Neolithization” in the Near East, with sometimes can be counterproductive. Nevertheless, some ba- particular emphasis on the Levant. It is intended to be read sic tenets of human behavior directly reflect the constraints in conjunction with our other article in this volume, “Neo- of our inherent neurological wiring as Homo sapiens sapiens lithization Processes in the Levant: The Outer Envelope” (e.g., Dunbar 1996; Johnson 1983; Kosse 1994; Wobst 1974). (Goring-Morris and Belfer-Cohen 2011). With the rapid accumulation of archaeological data in the It is vital to emphasize the marked local geographic vari- Near East, it seems that we have to view Neolithization from ability within the general region of the Levant, a point fre- the perspective of the longue dure´e (Braudel 1993). These quently overlooked in sweeping treatments of Neolithization processes had already begun during the Early Epipaleolithic phenomena. Change was not homogenous throughout the (ca. 23,000 years ago) and were heralded through seemingly region, with regard to either intensity or pace. Developments minor but in retrospect significant changes in human behav- were affected by circumstances pertaining to the shift from ior. Thus, we begin our overview almost 15,000 years before mobile to increasingly sedentary lifeways, the scales of com- the emergence of the fully fledged Neolithic “package” of munity sizes together with associated mating and social net- village farming communities. works, the nature of interactions within and between com- munities, and worldviews evolving in terms of ideology and ritual. Setting the Stage It is important to stress the “first-time” nature of these Epipaleolithic groups in the Levant were mobile-band soci- processes all through the sequence of human prehistory, a eties, just as in the preceding Upper Paleolithic (table 1), but case in point in the Near East. Moreover, the processes en- the scale of individual territories in some areas shrank as a tailed the “domestication” of multiple elements; Neolithiza- result of increased packing, leading to relatively higher pop- tion was not just about subsistence. We can also observe the ulation densities at local as opposed to pan-Levantine levels domestication of landscape, fire, water, social institutions, and (e.g., Bar-Yosef 2001). Here, the magic numbers of ca. 25 even “the gods” (e.g., Cauvin 2000). individuals for small-scale mobile-band societies with mating Ethnographic analogies are at best ambiguous even without networks encompassing some 250–500 individuals are rele- the problematics of considering recent and subrecent popu- vant (e.g., Wobst 1974). Virtually no data are available at this time for the northern Levant (see Goring-Morris and Belfer- Anna Belfer-Cohen and A. Nigel Goring-Morris are Professors in Cohen 2011), and accordingly, our discourse focuses on the the Department of Prehistory, Institute of Archaeology, Hebrew south. University of Jerusalem (Jerusalem 91905, Israel [belferac@ Shifts and realignments throughout the course of the Early mscc.huji.ac.il]). This paper was submitted 13 XI 09, accepted 12 and Middle Epipaleolithic are documented, demonstrating XII 10, and electronically published 27 VII 11. that the nature of diachronic and synchronic social interac-

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0005$10.00. DOI: 10.1086/658861 S210 Current Anthropology Volume 52, Supplement 4, October 2011

Table 1. Duration of the Epipaleolithic and Neolithic enti- quantities of ochre and especially lime plaster (probably for ties in the Levant hafting) as well as increased variability in hearth types in Mushabian sites provide evidence for increasing concerns with Approximate range pyrotechnology, presaging future developments (e.g., Bar- Period (cal BP) Duration (yr) Yosef and Goring-Morris 1977; Goring-Morris 1988; Kingery, Early Epipaleolithic 24,000–18,000 6,000 Vandiver, and Pickett 1988). Middle Epipaleolithic 18,000–15,000 2,900 Late Epipaleolithic 15,000–11,600 3,500 PPNAa 11,600–10,500 1,100 The Natufian Early PPNB 10,500–10,100 400 Middle PPNB 10,100–9500 600 With the beginnings of sedentism during the Late Epipaleo- Late PPNB 9500–8750 750 lithic Natufian in the Mediterranean “core area” at least (Gor- Final PPNB 8750–8400 350 ing-Morris and Belfer-Cohen 2011; fig. 1), there is clear evi- Early Pottery Neolithic 8400–7600 800 dence for significant increases in community sizes, with Note. Discrepancies between this table and table 1 of Goring-Morris and hamlets of up to 75–100 individuals (?), and perhaps no less Belfer-Cohen (2011) derive from uncertainties of 14C dating. PPNA p Pre-Pottery Neolithic A; PPNB p Pre-Pottery Neolithic B. important, for those communities lasting over extended pe- 14 a The “Round-House Horizon” in the Upper Tigris region (in two riods of time. Calibration of C dates demonstrates that the phases) is estimated at ca. 12,750–11,750–10,750 BP cal. duration of the Natufian was considerably longer than pre- viously assumed, 13,500 calendrical years as opposed to 2,000 tions across the region as a whole would have been quite years before (see, e.g., Bar-Yosef 1983; Stutz 2004). variable (fig. 1). Localized packing of groups west of the rift Social mechanisms were necessarily of great importance to valley is reflected in the different stylistic elements displayed and had a great impact on enabling the Natufians in some by the chipped-stone assemblages, indicating that specific Ke- areas to stay together in larger sedentary groups for the first baran group encounters commonly occurred on a one-to-one time in the archaeological record. The increased community basis. The excellent organic preservation at Early Epipaleo- sizes apparently crossed a critical threshold in terms of social lithic Ohalo II (ca. 22,000–24,000 cal BP, OxCal 4) on the interactions. The changes indeed represent a radical “exper- shores of Lake Lisan represents a rare window on palimpsest iment” of long duration; thus, the Natufian phenomenon short-term occupations in a refugium-type setting (Weiss et should not be viewed as simply intermediate between Pale- al. 2008). In other not-so-distant areas east of the rift valley, olithic and Neolithic lifeways. It is important to consider the networks differed in geographic scale, as portrayed by large- impact of the Natufian novelty in terms of the changes in scale Early Epipaleolithic aggregation megasites (e.g., Khar- social organization that occurred along its sequence. These aneh IV and Jilat 6; Garrard 1998; Muheisen and Wada 1995), clearly relate to a dramatic increase in the range and scope where the intensity and scale of social interaction would have of the material culture remains (e.g., Bar-Yosef 2002). Overall, been on a quite different level (i.e., multiple bands converging it is hardly surprising that from the very beginning of the on a seasonal basis; see Goring-Morris and Belfer-Cohen Early Natufian there are indications of scalar stress (Johnson 2011). 1982), which is accompanied by a dramatic increase in artistic/ The later emergence during the Middle Epipaleolithic of a symbolic activities (e.g., Belfer-Cohen 1988a; Belfer-Cohen distinct regional entity in the more arid margins of the Negev and Bar-Yosef 2000; Boyd 1995; Edwards 2009; Valla 1989). and Sinai—the Mushabian, synchronous with the Geometric The degree of Natufian sedentism varies from one area to Kebaran—may indicate the creation of local boundaries be- another. This is reflected by the combination of increased tween specific mating networks (Goring-Morris 1995). Evi- idiosyncrasy at the individual community level and regional dence for symbolic activities during the Early and Middle variability (e.g., Belfer-Cohen 1988a; Stordeur 1991; Wright Epipaleolithic tends to be sporadic, but when found it is quite 1978). Social and mating networks would also likely have been sophisticated (e.g., the engraved plaque from Urkan e-Rubb diverse. Some Natufian groups may have continued to be II; Hovers 1990). It remains unclear to what extent the tiny organized along the lines of small-scale bands, the basic unit Mediterranean marine mollusk assemblages recovered in being the nuclear family. But in more favorable settings, Early south Sinai sites were directly procured or derived from small- Natufians probably adopted quite different patterns of dom- scale exchange (Bar-Yosef and Killebrew 1984). icile, whether by extended family, moieties, or the like, as The relative paucity of burials throughout the Earlier Epi- reflected by the dramatic shifts in the size of architectural paleolithic indicates different means of relating to specific remains (e.g., at Wadi Hammeh 27 and el-Wad; Edwards 1991; localities and the wider landscape than in the later periods. Goring-Morris 1996; Goring-Morris and Belfer-Cohen 2008 Still, some changes may be detected during the course of the and references therein). Another notable feature is the internal Middle Epipaleolithic (e.g., in the Geometric Kebaran sites of and external organization of space, with areas or whole sites Neve David, ‘Uyyun el-Hammam, and Wadi Mataha stone designated for special activities (e.g., burial grounds and cem- bowls sometimes accompanied burials; Kaufman and Ronen etery sites such as Nahal Oren, Hilazon Tachtit, and Raqefet; 1987; Maher 2005; Stock et al. 2005). The presence of small Grosman 2003; Grosman and Munro 2007; Lengyel and Boc- Belfer-Cohen and Goring-Morris Becoming Farmers S211

Figure 1. Reconstructed ranges of some Early Epipaleolithic groups in the southern Levant based on the technotypological and stylistic attributes of lithic assemblages. Note different scales of ranges east and west of rift valley. quentin 2005; Stekelis and Yizraely 1963). One can refer also wide range of symbolic behaviors, including changes in at- to the symbolic spatial arrangements within structures (e.g., titudes toward the dead and accompanying funerary practices. the monoliths at Wadi Hammeh 27 and Rosh Zin; Edwards The overall numbers of Natufian burials within sites increase 2009; Goring-Morris and Belfer-Cohen 2003, 2008) or the dramatically. They now represent location markers within the various pebble arrangements, including an anthropomorphic landscape, whether as burial grounds adjacent to living quar- one at Eynan (Boyd 1995; Perrot 1966; Valla 1989). Indeed, ters, as foundation deposits (?), or as separate dedicated cem- there is an exponential increase in elements pertaining to a eteries (e.g., Perrot and Ladiray 1988). Indeed, it appears that S212 Current Anthropology Volume 52, Supplement 4, October 2011 we can detect the beginnings of a “dialogue with the dead” Belfer-Cohen, Schepartz, and Arensburg 1991]). Although the by members of the living community (and see Belfer-Cohen burials are seemingly orderly and in well-defined spaces, there 1995; Stekelis and Yisraely 1963). is great variability in the position of the interred individuals Another indication of communal activities above the level as well as in who was buried in any specific grave (whether of individual nuclear families concerns the sizes and distri- single burial, multiple burials, burials of mixed sexes, or buri- butions of groundstone utensils. The presence of huge mor- als of young and old). Besides the individually decorated buri- tars as well as off-site bedrock mortar arrays provides another als of the Early Natufian (making up ca. 10% of the total indication of communal activities, perhaps simply the scaling burials), there are also unique burials of the kind exposed at up of earlier egalitarian band principles (Belfer-Cohen and Late Natufian Hilazon Tachtit: the “shaman” burial (Gros- Hovers 2005). The range and quantities of raw materials and man, Munro, and Belfer-Cohen 2008) or the “gazelle-horned” types of jewelry and other symbolic items expand considerably individuals in Grave 10 at Eynan (Perrot and Ladiray 1988). to include not only marine mollusks but also colored min- The joint human and dog burials observed at both Eynan erals, clay, animal bones and teeth, and so forth (e.g., Bar- and Hayonim Terrace represent another unique mortuary Yosef Mayer 2005; Bar-Yosef Mayer and Porat 2008; Belfer- practice presaging later developments (Davis and Valla 1978; Cohen 1991; Marechal 1991; Weinstein-Evron and Ilani Tchernov and Valla 1997). 1994). While some items were produced locally on site, others How does all of the above tie in with what we know about derive from considerable distances and obviously involved the nature of the Natufian entity? Most probably the Natufian regional exchange networks, but usually these were internal, reflects aspects of emerging social identity in the face of shar- within the Natufian world itself (e.g., basalt groundstone tools; ing space with distant and second-tier kin for extensive pe- Weinstein-Evron, Kaufman, and Bird-David 2001); the spo- riods. At the same time, the increased packing of sites as well radic appearance of obsidian toward the end of the period is as other indications of territoriality (and see Goring-Morris an exception (Cauvin 1991; Delerue 2007; Valla 1999). Artistic and Belfer-Cohen 2011) brought about latent competition manifestations include rare figurines of all kinds and raw both within and between sites. Given postulated community materials (e.g., a female ochre figurine from Hayonim Terrace; sizes, some mating networks incorporated more than the in- Valla 1999). dividual base camp. Thus, there must have been at least two Another interesting phenomenon is the appearance of “dec- points of public social reference for each individual besides orated” items that are rarely observed in earlier contexts: bone her/his private status regarding gender, age, or matrimony. It tools and groundstone utensils. Why do people decorate seems that Natufian lifeways, at least in the “core area,” seemingly mundane items? Are they used on special occa- brought forth community identity (on the level of the indi- sions? Are they produced as goods to be exchanged or as vidual base camp), a concept that hardly existed previously; poorer-quality imitations as grave goods (e.g., the bone tools this may explain the individual styles seen in various domains deposited in grave XVII at Hayonim Cave), to be removed in any given base camp. We do not know how the individual from circulation? Do the patterns denote signatures of own- defined her/himself within their community or within the ership (Edwards 2007)? Concurrently, there is also a plethora Natufian as a whole because currently we lack the means to of incised stone and bone plaques with abstract schematic isolate which attributes pertain to the personal, the com- patterns reminiscent of those observed on items from con- munity, or the broader mating-network identity. Undoubtedly temporary Paleolithic horizons in Europe (e.g., Bar-Yosef and there was a need for more markers of identity within these Belfer-Cohen 1999; Belfer-Cohen and Bar-Yosef 2009; Svo- more socially complex contexts. Another factor to consider boda 1997). relates to the issue of possible local surpluses, whether as- Jewelry and symbolic/artistic motifs differ in intensity and sociated with subsistence (primarily but not only vegetal, on type from one base campsite and region to another, most a seasonal basis) or of an artisan-related or symbolic nature. probably denoting territories and group identities and affil- Such surpluses would have required regulation—whether at iations as well as personal status (e.g., Belfer-Cohen 1991; the family, community, or some intermediate level—desig- Garrod 1936–1937; Stordeur 1981). Such items are found all nating new social agendas (i.e., those individuals/families/ through the Natufian sequence even though they largely dis- communities with more personal goods versus those with appear from Late Natufian grave contexts. less). Many Natufian burials reflect considerable investment of In this new world full of novelty and tension, the question time and effort. While there is significant variability in mor- arises as to whether we can detect a rise in personal or in- tuary practices, some general patterns can be discerned, sev- tercommunity violence? There is indeed sporadic evidence for eral of which remain ambiguous. For example, there are more violence during the Natufian (Bocquentin and Bar-Yosef primary individual burials in the Early Natufian as opposed 2004), but it should be emphasized that signs of trauma can to more secondary burials in the Later Natufian, when post- also reflect mundane, everyday accidents. Here, as an aside, mortem skull removal is documented, a precursor of sub- we note that although we are dealing with scientific endeavors sequent Neolithic developments (Belfer-Cohen 1988b). There and hard-core data, one cannot escape the feeling of changing are also instances of family burials (e.g., Hayonim Grave VII; research fashions. The mode changed from observing pre- Belfer-Cohen and Goring-Morris Becoming Farmers S213 historic populations as “brutish and violent” in the early twen- are found (Hershman and Belfer-Cohen 2010) together with tieth century to considering them in the later 1960s/early “decorated” stone plaques in abstract designs (Bar-Yosef and 1970s as “flower children,” not aware that they were on the Gopher 1997; Edwards 2007). At the same time, there may verge of becoming us (i.e., violent, greedy, impatient, etc.). have been a greater dichotomy between private and public Of late, the pendulum appears to have swung back, and now domains as reflected in communal edifices, such as the unique research focuses on potential indications for violence that tie tower, walls, and silos of PPNA Jericho with a concentration in with the changing ways of life (e.g., Bocquentin 2003; and of burials nearby (Kenyon 1957). Do these represent en- further afield, Solecki, Solecki, and Agelarakis 2004). deavors to strengthen social cohesion? It seems plausible that the larger PPNA communities correspond to aggregations of Pre-Pottery Neolithic A residual local Natufian populations together with refugees from the periphery contracting back to refugia following the The Pre-Pottery Neolithic A (PPNA) is a short-lived phe- debilitating effects of the Younger Dryas. nomenon both with regard to what preceded as well as what The hierarchy of PPNA site sizes, in addition to reflecting succeeded it—a mere 1,000 years (and see Goring-Morris and different subsistence modes, undoubtedly had significant so- Belfer-Cohen 2011). However, the scale of change for the cial implications at the local and regional levels. This would PPNA of the southern Levant is of a different order of mag- have included differentiated relationships between nuclear nitude in comparison with that of the northern Levant, where and extended families and larger kinship lines at community there was apparently a radical increase in population size and and intercommunity scales. in the accompanying elements of social organization. In this In the north, too, domicile was apparently also organized context, the systematic colonization of Cyprus already from by nuclear families clustered into extended family com- the PPNA (if not earlier—a matter of some debate) represents pounds, each associated with a communal ritual semisubter- an important contribution unequivocally demonstrating the ranean structure as at Jerf el-Ahmar, Tell ‘Abr 3, and Mureybet “budding off” of groups on the mainland to found settlements (Stordeur 2007). It is interesting to note that these “socio- in new territories (Guilaine and Le Brun 2003; McCartney et cultic” structures were originally interpreted as typical resi- al. 2007; Peltenburg et al. 2001). Undoubtedly, similar pro- dential ones (e.g., Mureybet: Stordeur et al. 2001; and Qermez cesses occurred also around the Mediterranean core area on Dere: Watkins 1990). the mainland itself (e.g., Abbe`s 2008). However, in the north there is an additional ritual phe- In the south, certain sites became nodes of regional or even nomenon that emerged during the course of the PPNA and panregional interactions as reflected in the exchange of ob- locally continued through the earlier part of the PPNB in the sidian and its relative abundance in various sites. Such an form of massive supraregional cult sites (e.g., Go¨bekli Tepe example is Jericho, which may have functioned as a center and possibly Karahan; C¸elik 2000; Hauptmann and Schmidt for the distribution of various minerals—obsidian, malachite, 2007; Schmidt 2005, 2006). These huge monumental sites are and salt, among others. Surpluses would have enabled the found in prominent watershed locations away from the linear development of a protomarket—forces of supply and de- riparian habitation settlements in the low-lying plains. They mand—and would have encouraged increasing social inter- would have required the mobilization and organization of action. Nevertheless, social organization continued to be large groups (on a quite different scale from the tower of based on the concept of families, and a certain tendency to- Jericho) because their construction would have involved long- ward privatization at the individual family level is observed term investment to quarry and move massive pillars weighing (e.g., groundstone tool furniture within each structure; Belfer- tens of tons. The abundance of rich and intricate artistic Cohen and Hovers 2005; Rosenberg 2008). We suggest that manifestations portrays the intense investment of time and individual domiciles would likely have been spatially orga- effort in the ritual sphere. The overall spiritual domain of the nized along kinship lines, that is, as extended families within northern PPNA far surpasses that of the south in terms of the wider community or as wards or the like (although the both scale and in the motifs depicted. In particular, the ico- available data remain ambivalent on this issue). nography focuses on wild and frightening bestiary often hav- In general the southern PPNA articulates more comfortably ing emphatically gendered phallocentric connotations (Hod- with the Natufian than with the following Pre-Pottery Neo- der and Meskell, forthcoming). What remains elusive is the lithic B (PPNB). An illustration of this is the nature of PPNA sociocultural background to such an explosion. It is of interest burials, which often continue in the manner of the Late Na- to note that the larger structures/units are the oldest, similar tufian. Could it be that the lack of burial ornamentation and to phenomena observed within the Natufian sequence (Gor- grave goods reflects an improvement through time (observed ing-Morris 1996). Furthermore, at Go¨bekli as elsewhere in already in the Late Natufian) of social mechanisms that re- the north, there is evidence for repeated intentional infilling placed the simple “beads and feathers” system of social no- and burial of structures before the construction of new ones menclature? However, for another approach invoking the pro- (O¨ zdog˘an 2006). motion of egalitarian principles, see work by Kuijt (1996). Perhaps such sites fulfilled another function in that they Quantities of figurines made of soft stone and burned clay portrayed the wealth and the merchandise of individual com- S214 Current Anthropology Volume 52, Supplement 4, October 2011 munities; they would thus become active centers of social icantly increased the complexity of social and other interac- (competitive?) interaction sanctioned by “cultic” activities, tions at the intra- and intercommunity levels. Undoubtedly, with the separate units in these ritual megasites representing the individual now faced considerable competition, both individual groups within the regional framework. In this they within and between communities. On the other hand, the may parallel the individual ritual features in the habitation individuality as reflected in the stone masks and plastered sites, which are associated with clearly defined compounds of skulls with modeled faces can be interpreted as personalized extended families (and see above). Currently we are unable representatives of the community rather than actual privati- to decipher the association between particular social units and zation (other interpretations vary considerably, from trophies their specific symbolic vocabulary, but there is clearly evidence to gods; see discussions in Kuijt 2008). Apparently, as com- for the existence of such an association (e.g., some complexly munities burgeoned, not all members could take an active incised designs at Jerf el-Ahmar approximate pictograms; part in all and every aspect of their existence, whether eco- Stordeur 2004). nomic, social, and/or cultic. In consequence, the community It is quite obvious that as social circles increased in scale as a whole was represented by “individuals,” perhaps person- and scope, there was a need for more markers/symbols of ifying groups’ ancestors, who thus epitomized the group itself. identity at all levels of social interaction. Accordingly, we wit- There is some evidence within settlements of the desig- ness a phenomenon reminiscent of the concept of amphyc- nation of separate areas for communal and cultic purposes tyonies—leagues of neighboring communities associated with (e.g., Beidha, ‘Ain Ghazal, and Atlit Yam; Byrd 1994; Galili sacred locales (Belfer-Cohen and Goring-Morris 2002:144; 2004; Rollefson 2000). These installations display some con- Schmidt 2005)—known to exist in protohistoric and later tinuity from the Natufian as reflected by monoliths and paved periods in the Eastern Mediterranean. Indeed, these Pre-Pot- areas (Goring-Morris and Belfer-Cohen 2003). Further ele- tery Neolithic (PPN) phenomena could find parallels in the ments of continuity may also be reflected by the cult-cum- “Chaco phenomenon” of Great Houses and associated kivas cemetery site of Kfar HaHoresh, which is located in a prom- in the American Southwest (e.g., Cordell 1994; Kantner 2004; inent yet secluded neutral place (i.e., extraterritorial relative and see articles in American Antiquity 66 [2001]). Both cor- to nearby low-lying village sites; Goring-Morris 2005). The relates appear to provide avenues for further investigation in glimpses we gain of the rituals performed indicate that we analyzing the scope and scale of the north Levantine PPN are facing primarily intensifications and embellishments of reality. past customs more than elements of a “new order” (Belfer- This complexity reflects the additional commitments of Cohen and Goring-Morris 2002, 2005). society—whether at the level of the individual, the nuclear By contrast, in the northern Levantine PPNB, there is a and extended family, or the community—in order to maintain greater degree of continuity from the local PPNA that is re- and consolidate regional cohesion at the supracommunal flected in many elements as well as further innovations. The level. In the south, these nodes may often have been the actual colonization of central Anatolia as well as the continued col- settlements themselves, located in strategic settings. The dif- onization of Cyprus most probably represents the need for ferences between the southern and northern Levant are clearly new territories as overall population densities increased within perceptible; while in the south most of the ritual behaviors core areas on the mainland. are more or less on a private level, in the north there are in From the beginning of the PPNB, impressive “Houses of addition large-scale interactions pertaining to the ritual do- the Dead” make their appearance within habitation sites (e.g., main, indicating that we are facing a multitiered system (i.e., at Djade´and at C¸ayo¨nu¨; Coqueugniot 2006; O¨ zdog˘an 1999); adding yet more rings to the social “onion”). whether these represent charnel houses of ancestors or of enemies remains debatable (e.g., Testart 2008). Concurrently, Pre-Pottery Neolithic B PPNA traditions of designated cult structures continued (e.g., at Nevali C¸ori and at C¸ayo¨nu¨; Hauptmann 1999; O¨ zdog˘an The PPNB of the south emerges after a significant break with 1999; for detailed discussion of some of these issues see the PPNA, albeit with various elements of continuity of earlier Croucher 2006; Verhoeven 2002a, 2002b, 2002c). Interestingly, traditions. By contrast, the transformation in the northern the large-scale cult sites such as Go¨bekli Tepe ceased to func- Levant is relatively seamless. tion sometime during the course of the PPNB; this may reflect The primary trends in the south regarding social organi- a degree of social disintegration and a retreat back to the zation include the later emergence, during the Late PPNB, of individual community. the “megasite” village phenomenon (e.g., ‘Ain Ghazal, es- The rise of surpluses within the PPNB koine also un- Sifiya, and Basta; Gebel 2004; Goring-Morris, Hovers, and doubtedly influenced the social and ritual fabric. The whole Belfer-Cohen 2009; Mahasneh 2004; Rollefson 2001). notion of organized exchange networks affected and brought Whether one adopts a maximalist or minimalist approach in about nonkin social segmentation. In this context, one should terms of estimating population sizes, these likely double or consider the emergence of incipient craft specialization in- triple the size of the largest communities inferred for the volving artisans and associated merchants who would have PPNA (Kuijt 2000; Simmons 2007). This would have signif- operated not just systematically within but also between com- Belfer-Cohen and Goring-Morris Becoming Farmers S215 munities (e.g., Barzilai 2009; Quintero 1998 with regard to in the southern Levant during the shift from the Natufian to lithics). Indeed, one may even speculate about the emergence PPNA by way of the short-lived “Khiamian” entity. of protoguilds, because we witness high levels of artisan com- There are inherent uncertainties regarding the mode and petence executed in particular raw materials and dispersed tempo of changes in subsistence that are assumed to go hand over large geographic areas. It is not clear as to whether the in hand with social structures and systems and that accord- products of such know-how were circulated by wandering ingly influence our perceptions. Moreover, it seems to us that individual specialists or whether “secrets of the trade” were if during the course of Neolithization, changes occurred spo- disseminated among a network of a chosen few who acted in radically and took a long time to gel and take root and that unison. Whatever the case, all of the above would have fos- their courses of evolution differed from one region to the tered further levels of communication between communities. next, we need to readjust the ways in which we review and The end of the PPNB world likely relates to a wide array evaluate the social realm. of interconnected factors, whether climatic changes, demo- Undoubtedly, human groups modified their attitudes to graphic pressures including diseases, and declining yields (Co- the environment in accordance with the role it played within hen and Crane-Kramer 2007; Goring-Morris and Belfer- the Neolithization process. Starting at least by the Natufian, Cohen 2010; Rollefson and Ko¨hler-Rollefson 1989). This was the first orderly burial grounds functioned as location markers more marked in the southern Levant than the northern Le- in the landscape. These burial customs persisted through the vant. Yet in terms of economy, the PPNB represents a point following Neolithic sequence, indicating the continuity and of no return. Human subsistence in the Levant continued persistence of ancient traditions. Accordingly, Neolithic thereafter to be based in one form or another on the suite groups looked back and took comfort in relying on the past of domesticates (plants and animals) initiated during the and its institutions rather than inventing a complete package PPNB (as well as the addition of pastoralism). Breakdowns of “new order” regulations. Changes took time and were ini- in social interactions—brought about, among other things, tially of marginal impact (e.g., skull removal and skull mod- by the rise in contagious diseases and interpersonal violence ification). Still, it is of interest to note that there was a general (as perhaps reflected in more multiple burials in the Late Pre- consistent trajectory throughout the Levant in terms of the Pottery Neolithic B [LPPNB])—heralded the demise of the persistence of local traditions. PPNB koine. Vagaries in the history of research have led to Debates have arisen regarding possible population move- an emphasis on the presence of a major break between the ments as opposed to the diffusion of ideas and local conti- PPNB and Early Pottery Neolithic, but to the contrary, recent nuity. For example, PPNB skull plastering originated as a research indicates numerous elements of continuity, albeit southern Levantine phenomenon—the foundations of which with local trajectories rather than the previous pan-Levantine lie in Natufian skull removal—but persisted into the Late ¨ scope of the PPNB (e.g., Garfinkel and Miller 2002). What Neolithic in the north (Ozbek 2009). It is of interest to note is lost is the unique character of the PPNB as reflected in its that some ritual phenomena, though rare, persist through the social interactions and networks and its material culture com- entire sequence, thousands of years apart, as exemplified by ponents related to the sacred and ritual, in short, the attributes the pebble depiction at Natufian Eynan, which recalls the considered as the hallmarks of what is traditionally referred human bone arrangement at PPNB Kfar HaHoresh and also to as a particular culture. the stone drawings at Late Neolithic Khasm et-Tarif and Biqat Uvda (Avner 2002; Goring-Morris 2005; Perrot 1966). Changes observed in institutionalized behavior were gen- Discussion erally of a minor magnitude; just as decorated burials in the Early Natufian did not exceed ca. 10%, so too PPNB plastered The role of climate change (and especially rapid climate skulls do not exceed 10% of total burials. There were changes, change) in the Neolithization process is certainly a factor to from an emphasis on primary to multiple secondary burials be considered at various specific points in time. Climate did during the course of the Natufian, a situation mirrored toward change at a global scale, but the impact and effects of the the end of the PPNB. Perhaps these similarities reflect stress, changes differed in various and varied environmental settings with differences stemming from specific circumstances. What with diverse technological and social milieus, bringing about do the Early Natufian decorated burials represent and why different, particularistic responses of the human populations did they disappear? Most of the otherwise unique burials were exposed to those climatic changes. Overall, the climatic in the Late Natufian and thenceforth were not decorated. Still, changes caused internal “bottleneck” situations in the passage animal themes were part and parcel of burial rituals from the from the Natufian to PPNA, from PPNA to PPNB, and at Natufian through the PPNB, whether the Late Natufian Hil- the end of the PPNB, all of which require more detailed azon “shaman” burial (Grosman, Munro, and Belfer-Cohen investigations. Sometimes these bottlenecks may have acted 2008), the burial with a gazelle-horn headdress at Eynan (Per- as tipping points or threshold events in the sense of providing rot and Ladiray 1988), turtle carapaces at el-Wad, Hayonim catalysts for realignments of social systems. As an example, Terrace, and Eynan, and so forth (Garrod and Bate 1937; one may cite the likely presence of a refugium at low elevations Valla 1999). By the PPNB they include complete or partial S216 Current Anthropology Volume 52, Supplement 4, October 2011 animal skeletons and pars pro toto (e.g., fox mandibles; Gor- that burgeoned as part of the growing complexity of the Early ing-Morris and Horwitz 2007; Horwitz Kolska and Goring- Neolithic Levantine koine. Morris 2004). These coassociations occur but sporadically. Do Was violence an integral element to the observed social they represent professional personal achievements or some processes of Neolithization? Interpersonal violence is without manner of status differentiation? What do animal bones as doubt a human constant. And certainly at “Proto-Neolithic” grave goods denote? Are they talismans, totemic affiliation Zawi Chemi Shanidar there is evidence for extensive head signs, and/or remnants of ritual feasting (e.g., Twiss 2008)? traumas that could correlate with violence on an intercom- Furthermore, the concept of the “House of the Dead” in the munity scale (Solecki, Solecki, and Agelarakis 2004). During north—present in Djade´, C¸ ayo¨nu, Aswad, Tell Hallula, and the Natufian, the evidence for intercommunity violence is C¸atalho¨yu¨k (Du¨ring 2009; Guerrero et al. 2009; Stordeur and more equivocal, but later the presence of multiple burials in Khawam 2009 and see references above)—could represent the LPPNB could potentially be interpreted (in addition to lineages or other social associations (and see Kuijt 2008 for the possibility of poor health) as yet another mechanism for a discussion of societal consolidation and the creation of com- structured social intercourse. Indeed, it has even been sug- munal memory). gested that slavery may have been a feature of PPNB lifeways Scalar stress began to be significant from the Natufian on- (Bar-Yosef and Bar-Yosef Mayer 2002). ward, the common explanation for the exponential rise in We believe the above represent significant issues that merit artistic activities and representations and accompanying sty- further discussion with respect to Neolithization processes in listic variability (Belfer-Cohen and Bar-Yosef 2000 and ref- the Near East and that ultimately were not just about the erences therein). Without delving into a detailed discussion “origins of agriculture” per se. Thus, we encounter societies of the social correlates of stylistic variation (e.g., Conkey and guided by combinations of both new and old principles that Hastorf 1990; Jochim 1987; Stark 1998; Wiessner 1989; Wobst had undergone profound modifications. Human societies are, 1977, 1999 and references therein), there was a growing need by definition, conservative institutions that cling to the sym- for additional markers of status/affiliation/kinship. For ex- bols that define them. It was the radical transformations in ample, competition for potential partners would have in- interpersonal and especially intercommunity relations that herald the real “revolution.” creased both within the community as well as between com- munities. From the beginning of the Natufian through the entire Neolithization sequence, the thickness of the different social layers varied; there were circles of family, kinsmen, the Acknowledgments village, nearby communities, exchange partners, and the list grew to encompass the whole PPNB koine. We are grateful to the organizers (Doug Price and Ofer Bar- There were now in the PPNB levels of segmentation that Yosef) and fellow participants of the Temozon Wenner-Gren were not necessarily always based on kinship (and see, e.g., workshop “The Origins of Agriculture: New Data, New Ideas” Du¨ring 2006; Hodder 2005; Hodder and Cessford 2004). (Merida, Yucatan) in March 2009 for a most stimulating and intensive exchange of data and opinions concerning agricul- Whereas in the Earlier Epipaleolithic, contacts between ex- tural beginnings around the world. Thanks are due to the ternal groups were more sporadic, by the PPNB the whole Wenner-Gren Foundation for the superb organization of the system involved much more intense interactions, especially event and especially to Leslie Aiello and Laurie Obbink for with regard to the “outer rings”—“strangers” now became their personal commitments to the success of the meeting. permanent fixtures of the everyday world, and people had to We also acknowledge the support of the Israel Science Foun- devise new codes of behavior toward new strata of social dation funded by the Israel Academy of Sciences (A. Belfer- interaction. Cohen: grants 855/03 and 202/05; A. Nigel Goring-Morris: Additionally, in many areas sedentary communities existed grants 840/01, 558/04, and 755/07). virtually cheek by jowl with mobile foraging communities, and this differentiation by economic mode of existence called for yet another societal differentiation (e.g., David, Sterner, References Cited and Gavua 1988; Kent 2002 for an ethnoarchaeological per- Abbe`s, F. 2008. Wadi Tumbaq 1: a Khiamian occupation in the Bal’as spective). If, as Schmidt (2006) claims, the isolated hilltop Mountains. Neo-Lithics 1(8):3–9. site of Go¨bekli Tepe is characterized only by cultic (rather Avner, U. 2002. Studies in the material and spiritual culture of the Negev populations, during the 6th–3rd millennia B.C. PhD dis- than domestic) structures, then it is possible to hypothesize sertation, Hebrew University of Jerusalem. that this was a sacred site serving as an aggregation locale for Bar-Yosef, O. 1983. The Natufian of the southern Levant. In The hilly neighboring competitive groups within a “neutral” setting flanks and beyond. P. E. L. Smith and P. Mortensen, eds. Pp. 11– much in the manner of the later classical amphictyony (Belfer- 42. Chicago: University of Chicago Press. ———. 2001. The world around Cyprus: from Epi-Paleolithic for- Cohen and Goring-Morris 2002). Such a situation could have agers to the collapse of the PPNB civilization. In The earliest pre- served to mitigate and facilitate intercommunity relations re- history of Cyprus: from colonization to exploitation. S. Swiny, ed. garding the extended mating, exchange, and other networks Pp. 129–164. Boston: American Schools of Oriental Research. Belfer-Cohen and Goring-Morris Becoming Farmers S217

———. 2002. Natufian: a complex society of foragers. In Beyond Bocquentin, F. 2003. Pratiques fune´raires, parametres biologiques et foraging and collecting: evolutionary change in hunter-gatherer set- identites culturelles au Natoufien: une analyse arche´o-anthropol- tlement systems. B. Fitzhugh and J. Habu, eds. Pp. 91–149. New ogique. PhD thesis, Universite´ Bordeaux 1. York: Kluwer Academic/Plenum. Bocquentin, F., and O. Bar-Yosef. 2004. Early Natufian remains: evi- Bar-Yosef, O., and D. E. Bar-Yosef Mayer. 2002. Early Neolithic tribes dence for physical conflict from Mt. Carmel, Israel. Journal of in the Levant. In The archaeology of tribal societies. W. A. Parkinson, Human Evolution 47:19–23. ed. Pp. 340–371. Ann Arbor, MI: International Monographs in Boyd, B. 1995. House and hearths, pits and burials: Natufian mor- Prehistory. tuary practices at Mallaha (Eynan), Upper Jordan Valley. In The Bar-Yosef, O., and A. Belfer-Cohen. 1999. Encoding information: archaeology of death in the ancient Near East. S. Campbell and A. unique Natufian objects from Hayonim Cave, western Galilee, Green, eds. Pp. 17–23. Oxford: Oxbow. Israel. Antiquity 73:402–410. Braudel, F. 1993. A history of civilizations. New York: Penguin. Bar-Yosef, O., and A. Gopher, eds. 1997. An Early Neolithic village Byrd, B. F. 1994. Public and private, domestic and corporate: the in the Jordan Valley. Part I: the archaeology of Netiv Hagdud.Amer- emergence of the southwest Asian village. American Antiquity 59: ican School of Prehistoric Research Bulletin 43. Cambridge, MA: 639–666. Peabody Museum of Archaeology and Ethnology, Harvard Uni- Cauvin, J. 2000. The birth of the gods and the origins of agriculture. versity. Cambridge: Cambridge University Press. Bar-Yosef, O., and A. Nigel Goring-Morris. 1977. Geometric Kebaran Cauvin, M.-C. 1991. L’obsidie`nne au Levant pre´historique: prove- A occurrences. In Prehistoric investigations in Gebel Maghara, nance et fonction. Cahiers de l’Euphrate 5–6:163–190. northern Sinai, Qedem 7. O. Bar-Yosef and J. L. Phillips, eds. Pp. C¸elik, B. 2000. A new Early Neolithic settlement: Karahan Tepe. Neo- 115–148. Monographs of the Institute of Archaeology. Jerusalem: Lithics 2(3):6–8. Institute of Archaeology, Hebrew University. Cohen, M. N., and G. Crane-Kramer, eds. 2007. Ancient health. Bar-Yosef, O., and A. Killebrew. 1984. Wadi Sayakh: a Geometric Gainesville: University Press of Florida. Kebaran site in southern Sinai. Pale´orient 10:95–101. Conkey, M. W., and C. A. Hastorf, eds. 1990. The uses of style in Bar-Yosef Mayer, D. E. 2005. The exploitation of shells as beads in archaeology. Cambridge: Cambridge University Press. the Palaeolithic and Neolithic of the Levant. Pale´orient 31:176– Coqueugniot, E. 2006. Diversified funerary practices: the case of 185. Dja’de (Syria, Euphrates Valley, 9th millennium cal. BC, late PPNA Bar-Yosef Mayer, D. E., and N. Porat. 2008. Green stone beads at and EPPNB). In Abstract of 5th International Congress on the Ar- the dawn of agriculture. Proceedings of the National Academy of chaeology of the Ancient Near East workshop Houses for the Living Sciences of the USA 105:8548–8551. and a Place for the Dead.J.M.Co´rdoba, M. Molist, M. Pe´rez, I. Barzilai, O. 2009. Social complexity in the southern Levantine PPNB Rubio, and S. Martı´nez, eds. Madrid: Universidas Autonoma de as reflected through lithic studies: the bidirectional blade indus- Madrid. tries. PhD thesis, Hebrew University of Jerusalem. Cordell, L. S. 1994. Ancient Pueblo peoples: exploring the ancient world. Belfer-Cohen, A. 1988a. The appearance of symbolic expression in Montreal: St. Remy. the Upper Pleistocene of the Levant as compared to western Eu- Croucher, K. 2006. Death, display and performance: a discussion of rope. In L’homme de Neanderthal: la pense. M. Otte, ed. Pp. 25– the mortuary remains at C¸ayo¨nu Tepesi. In The archaeology of cult 29. Lie`ge: University of Lie`ge. and death. M. Georgiadis and C. Gallou, eds. Pp. 11–44. Budapest: ———. 1988b. The Natufian graveyard in Hayonim Cave. Pale´orient Archaeolingua Alapı´tva´ny. 14:297–308. David, N., J. Sterner, and K. Gavua. 1988. Why pots are decorated. ———. 1991. Art items from layer B, Hayonim Cave: a case study Current Anthropology 29:365–389. of art in a Natufian context. In The Natufian culture in the Levant. Davis, S. J. M., and F. R. Valla. 1978. Evidence for domestication of O. Bar-Yosef and F. R. Valla, eds. Pp. 569–588. Ann Arbor, MI: the dog 12,000 years ago in the Natufian of Israel. Nature 276: International Monographs in Prehistory. 608–610. ———. 1995. Rethinking social stratification in the Natufian culture: Delerue, S. 2007. L’obsidienne dans le processus de Neolithisation the evidence from burials. In The archaeology of death in the ancient du Proche-Orient (12000–6500 av. J.-C. cal.). PhD dissertation, Near East. S. Campbell and A. Green, eds. Pp. 9–16. Edinburgh: Universite´ Bordeaux 3. Oxbow. Dunbar, R. 1996. Grooming, gossip, and the evolution of language. Belfer-Cohen, A., and O. Bar-Yosef. 2000. Early sedentism in the Cambridge, MA: Harvard University Press. Near East: a bumpy ride to village life. In Life in Neolithic farming Du¨ring, B. S. 2006. Constructing communities: clustered neighbourhood communities: social organization, identity, and differentiation.I. settlements of the central Anatolian Neolithic ca. 8500–5500 cal. BC. Kuijt, ed. Pp. 19–38. New York: Kluwer Academic/Plenum. Leiden: Nederlands Instituut voor het Nabije Oosten. ———. 2009. First things first: abstract and figurative artistic ex- ———. 2009. Sub-floor burials at C¸atalho¨yu¨k: exploring relations pressions in the Levant. In In memory of A. Marshack. P. Bahn, between the dead, houses, and the living. In Proceedings of the 5th ed. Pp. 25–37. Cambridge, MA: Harvard University Press. International Congress on the Archaeology of the Ancient Near East. Belfer-Cohen, A., and A. Nigel Goring-Morris. 2002. Recent devel- J. M. Co´rdoba, M. Molist, M. Pe´rez, I. Rubio, and S. Martı´nez, opments in Near Eastern Neolithic research. Pale´orient 28:143– eds. Pp. 603–620. Madrid: Centro Superior de Estudios sobre el 148. Oriente Pro´ximo y Egipto. ———. 2005. Which way to look? conceptual frameworks for un- Edwards, P. C. 1991. Wadi Hammeh 27: an Early Natufian site at derstanding Neolithic processes. In Dialogue on the Early Neolithic Pella, Jordan. In The Natufian culture in the Levant. O. Bar-Yosef origin of ritual centres. Special issue, Neo-Lithics 2(5):22–24. and F. R. Valla, eds. Pp. 123–148. Ann Arbor, MI: International Belfer-Cohen, A., and E. Hovers. 2005. The groundstone assemblages Monographs in Prehistory. of the Natufian and Neolithic societies in the Levant: a brief review. ———. 2007. The context and production of incised Neolithic Journal of the Israel Prehistoric Society 35:299–308. stones. Levant 39:27–33. Belfer-Cohen, A., L. Schepartz, and B. Arensburg. 1991. New bio- ———. 2009. The symbolic dimensions of material culture at Wadi logical data for the Natufian populations in Israel. In The Natufian Hammeh 27. In Proceedings of the 5th International Congress on culture in the Levant. O. Bar-Yosef and F. R. Valla, eds. Pp. 411– the Archaeology of the Ancient Near East.J.M.Co´rdoba, M. Molist, 424. Ann Arbor, MI: International Monographs in Prehistory. M. Pe´rez, I. Rubio, and S. Martı´nez, eds. Pp. 507–520. Madrid, S218 Current Anthropology Volume 52, Supplement 4, October 2011

Spain: Centro Superior de Estudios sobre el Oriente Pro´ximo y Guerrero, E., M. Molist, I. Kuijt, and J. Anfruns. 2009. Seated mem- Egipto. ory: new insights into Near Eastern Neolithic mortuary variability Galili, E. 2004. Submerged settlements of the ninth to seventh Mil- from Tell Halula, Syria. Current Anthropology 50:379–391. lennia BP off the Carmel coast. PhD thesis, University of Tel Aviv. Guilaine, J., and A. Le Brun, eds. 2003. Le Ne´olithique de Chypre. Garfinkel, Y., and M. A. Miller, eds. 2002. Sha’ar Hagolan 1: Neolithic Bulletin de correspondence helle´nique, supple´ment 43. Athens: art in context. Oxford: Oxbow. E´ cole Franc¸aise d’Athe`nes. Garrard, A. N. 1998. Environment and cultural adaptations in the Hauptmann, H. 1999. The Urfa region. In Neolithic in Turkey: the Azraq Basin: 24,000–7,000 BP. In The prehistoric archaeology of cradle of civilization: new discoveries.M.O¨ zdogan and N. Basgelen, Jordan. D. O. Henry, ed. Pp. 139–148. BAR International Series eds. Pp. 65–86. Istanbul: Arkeologie ve Sanat Yayinlari. 705. Oxford: BAR. Hauptmann, H., and K. Schmidt. 2007. Anatolien vor 12 000 Jahren: Garrod, D. 1936–1937. Notes on some decorated skeletons from the die Skulpturen des Fru¨hneolithikums. In Anatolien vor 12 000 Mesolithic of Palestine. Annual of the British School in Athens 37: Jahren: die a¨ltesten Monumente der Menschheit. C. Lichter, ed. Pp. 123–127. 67–82. Karlsruhe: Badisches Landesmuseum. Garrod, D. A. E., and D. M. A. Bate. 1937. Excavations at the Wadi Hershman, D., and A. Belfer-Cohen. 2010. “It’s magic!”: artistic and Mughara,vol.1ofThe Stone Age of Mount Carmel. Oxford: Clar- symbolic material manifestations from the Gilgal sites. In Gilgal: endon. Early Neolithic occupations in the lower Jordan Valley: the excava- Gebel, H.-G. K. 2004. There was no centre: the polycentric evolution tions of Tamar Noy. O. Bar-Yosef, A. Nigel Goring-Morris, and A. of the Near Eastern Neolithic. Neo-Lithics 1(4):28–32. Gopher, eds. Pp. 185–216. Oakville, CT: Brown. Goring-Morris, A. Nigel. 1988. Trends in the spatial organization of Hodder, I. 2005. Socialization and feasting at C¸atalho¨yu¨k. American Terminal Pleistocene hunter-gatherer occupations as viewed from Antiquity 70:189–191. the Negev and Sinai. Pale´orient 14:231–243. Hodder, I., and C. Cessford. 2004. Daily practice and social memory ———. 1995. Complex hunter-gatherers at the end of the Paleolithic at C¸atalho¨yu¨k. American Antiquity 69:17–40. (20,000–10,000 BP). In The archaeology of society in the Holy Land. Hodder, I., and L. Meskell. 2011. A “curious and sometimes a trifle T. E. Levy, ed. Pp. 141–168. London: Leicester University Press. macabre artistry”: some aspects of the symbolism of the Neolithic ———. 1996. The Early Natufian occupation at el-Wad, Mt. Carmel, in Anatolia. Current Anthropology 52(2):235–263. reconsidered. In Nature et culture. M. Otte, ed. Pp. 417–428. Lie`ge: Horwitz Kolska, L., and A. Nigel Goring-Morris. 2004. Animals and Universite´ de Lie`ge. ritual during the Levantine PPNB: a case study from the site of ———. 2005. Life, death and the emergence of differential status in Kfar Hahoresh, Israel. Anthropozoologica 39:165–178. the Near Eastern Neolithic: evidence from Kfar HaHoresh, Lower Hovers, E. 1990. Art in the Levantine Epi-Palaeolithic: an engraved Galilee, Israel. In Archaeological perspectives on the transmission and pebble from a Kebaran site in the lower Jordan Valley. Current transformation of culture in the Eastern Mediterranean. J. Clark, ed. Anthropology 31:317–322. Pp. 89–105. Oxford: Oxbow. Jochim, M. 1987. Late Pleistocene refugia in Europe. In The Pleis- Goring-Morris, A. Nigel, and A. Belfer-Cohen. 2003. Structures and tocene Old World: regional perspectives. O. Soffer, ed. Pp. 317–333. dwellings in the Upper and Epi-Palaeolithic (ca. 42–10 K BP) New York: Plenum. Levant: profane and symbolic uses. In Perceived landscapes and Johnson, G. A. 1982. Organizational structure and scalar stress. In built environments. O. Soffer, S. Vasil’ev, and J. Kozlowski, eds. Pp. Theory and explanation in archaeology. C. A. Renfrew, M. J. Row- 65–81. BAR International Series 1122. Oxford: BAR. lands, and B. A. Segraves, eds. Pp. 389–421. London: Academic ———. 2008. A roof over one’s head: developments in Near Eastern Press. residential architecture across the Epipalaeolithic-Neolithic tran- ———. 1983. Decision-making organization and pastoral nomad sition. In The Neolithic demographic transition and its consequences. camp size. Human Ecology 11:175–199. J.-P. Bocquet-Appel and O. Bar-Yosef, eds. Pp. 239–286. Dordrecht: Kantner, J. 2004. Ancient Puebloan Southwest. Cambridge: Cambridge Springer. University Press. ———. 2010. “Great Expectations”; or, the inevitable collapse of the Kaufman, D., and A. Ronen. 1987. La Se´pulture Ke´barienne Ge´o- Early Neolithic in the Near East. In Early village society in global perspective. M. S. Bandy, and J. Fox, eds. Pp. 62–77. Tucson: Uni- metrique de Neve-David, Haifa, Israel. L’anthropologie 91:335–342. versity of Arizona Press. Kent, S., ed. 2002. Ethnicity, hunter-gatherers, and the “other”: asso- ———. 2011. Neolithization processes in the Levant: the outer en- ciation or assimilation in Africa. Washington, DC: Smithsonian velope. Current Anthropology 52(suppl. 4):S195–S208. Institution. Goring-Morris, A. Nigel, and L. K. Horwitz. 2007. Funerals and feasts Kenyon, K. M. 1957. Digging up Jericho. London: Benn. in the Near Eastern Pre-Pottery Neolithic B. Antiquity 81:902–919. Kingery, D. W., P. B. Vandiver, and M. Pickett. 1988. The beginnings Goring-Morris, A. Nigel, E. Hovers, and A. Belfer-Cohen. 2009. The of pyrotechnology. II. Production and use of lime and gypsum dynamics of Pleistocene settlement patterns and human adapta- plaster in the Pre-Pottery Neolithic Near East. Journal of Field tions in the Levant: an overview. In Transitions in prehistory: papers Archaeology 15:219–244. in honor of Ofer Bar-Yosef. J. J. Shea and D. Lieberman, eds. Pp. Kosse, K. 1994. The evolution of large, complex groups: a hypothesis. 187–254. Oakville, CT: Brown. Journal of Anthropological Archaeology 13:35–50. Grosman, L. 2003. Preserving cultural traditions in a period of in- Kuijt, I. 1996. Negotiating equality through ritual: a consideration stability: the Late Natufian of the hilly Mediterranean zone. Current of Late Natufian and Pre-Pottery Neolithic A period mortuary Anthropology 44:571–580. practices. Journal of Anthropological Archaeology 15:313–336. Grosman, L., and N. Munro. 2007. The sacred and the mundane: ———. 2000. People and space in early agricultural villages: ex- domestic activities at a Late Natufian burial site in the Levant. ploring daily lives, community size, and architecture in the Late Before Farming: The Archaeology and Anthropology of Hunter-Gath- Pre-Pottery Neolithic. Journal of Anthropological Archaeology 19: erers 4:1–14. 75–102. Grosman, L., N. D. Munro, and A. Belfer-Cohen. 2008. A 12,000- ———. 2008. The regeneration of life: Neolithic structures of sym- year-old Shaman burial from the southern Levant (Israel). Pro- bolic remembering and forgetting. Current Anthropology 49:171– ceedings of the National Academy of Sciences of the USA 105:17665– 197. 17669. Lengyel, G., and F. Bocquentin. 2005. Burials of Raqefet Cave in the Belfer-Cohen and Goring-Morris Becoming Farmers S219

context of the Late Natufian. Journal of the Israel Prehistoric Society Neolithic cemetery in Shanidar Cave. College Station: Texas A&M 35:271–284. University Press. Mahasneh, H. M. 2004. Spatial and functional features of Area B in Stark, M. T., ed. 1998. The archaeology of social boundaries. Wash- Neolithic es-Sifiya, Jordan. In Central settlements in Neolithic Jor- ington, DC: Smithsonian Institution. dan. H. Bienert, H. G. K. Gebel, and R. Neef, eds. Pp. 45–63. Stekelis, M., and T. Yisraely. 1963. Excavations at Nahal Oren: pre- Berlin: Ex Oriente. liminary report. Israel Exploration Journal 13:1–12. Maher, L. 2005. Recent excavations at the Middle Epipalaeolithic Stock, J. T., S. K. Pfeiffer, M. Chazan, and J. Janetski. 2005. F-81 encampment of U` yun al-Hammam, northern Jordan. Annual of skeleton from Wadi Mataha, Jordan, and its bearing on human the Department of Antiquities of Jordan 49:101–114. variability in the Epipaleolithic of the Levant. American Journal of Marechal, C. 1991. Elements de Parure de la Fin du Natoufien: Mall- Physical Anthropology 128:453–465. aha Niveau I, Jayroud 1, Jayroud 3, Jayroud 9, Abu Hureyra et Stordeur, D. 1981. La contribution de l’industrie de l’os a la de´li- Mureybet IA. In The Natufian culture in the Levant. O. Bar-Yosef mination des aires culturelles: l’example du Natoufien. In Pre´his- and F. R. Valla, eds. Pp. 589–612. Ann Arbor, MI: International toire du Levant: chronologie et l’organisation de l’espace de´puis les Monographs in Prehistory. origines jusqu’au VIe mille`naire. J. Cauvin and P. Sanlaville, eds. McCartney, C., S. W. Manning, D. Sewell, and S. T. Stewart. 2007. Pp. 433–437. Paris: CNRS. The EENC 2006 field season: excavations at Agia Varvara- ———. 1991. Le Natoufien et son evolution a travers les artifacts Asprokremmos and survey of the local Early Holocene landscape. en os. In The Natufian culture in the Levant. O. Bar-Yosef and F. Report of the Department of Antiquities, Cyprus 2007:27–44. Valla, eds. Pp. 467–482. Ann Arbor, MI: International Monographs Muheisen, M., and H. Wada. 1995. An analysis of the microliths at in Prehistory. Kharaneh IV, phase D, square A20/37. Pale´orient 21:75–95. ———. 2004. Symboles et imaginaire des premie`res cultures ne´o- O¨ zbek, M. 2009. Remodeled human skulls in Ko¨sk Ho¨yu¨k (Neolithic lithiques du Proche-Orient (haute et moyenne valle´e de age, Anatolia): a new appraisal in view of recent discoveries. Journal l’Euphrate). In Arts et symboles du Ne´olithique a` la protohistoire. of Archaeological Science 36:379–386. J. Guilaine, ed. Pp. 15–36. Paris: Errance. O¨ zdog˘an, A. 1999. C¸ayo¨nu. In Neolithic in Turkey: the cradle of civ- ———. 2007. Les baˆtiments collectifs des premiers Ne´olithiques de ilization: new discoveries.M.O¨ zdogan and N. Basgelen, eds. Pp. l’Euphrate: creation, standardisation et me`moire des formes ar- 35–63. Istanbul: Arkeologie ve Sanat Yayinlari. chitecturales. Subartu 17:19–31. ¨ Ozdog˘an, M. 2006. “Dead houses”: the burial of houses. In Abstract Stordeur, D., M. Brenet, G. Der Aprahamian, and J.-C. Roux. 2001. of 5th International Congress on the Archaeology of the Ancient Near Les baˆtiments communautaires de Jerf el-Ahmar et Mureybet Ho- East workshop Houses for the Living and a Place for the Dead.J. rizon PPNA (Syrie). Pale´orient 26:29–44. M. Co´rdoba, M. Molist, M. Pe´rez, I. Rubio, and S. Martı´nez, eds. Stordeur, D., and R. Khawam. 2009. Une place pour les morts dans Madrid: Universidas Autonoma de Madrid. les maisons de Tell Aswad (Syrie) (Horizon PPNB ancien et PPNB Peltenburg, E., P. Croft, A. Jackson, C. McCartney, and M. Murray. moyen). In Proceedings of the 5th International Congress on the 2001. Well-established colonialists: Mylouthkia 1 and the Cypro Archaeology of the Ancient Near East.J.M.Co´rdoba, M. Molist, Pre-Pottery Neolithic B. In The earliest prehistory of Cyprus: from M. Pe´rez, I. Rubio, and S. Martı´nez, eds. Pp. 561–590. Madrid, colonization to exploitation. S. Swiny, ed. Pp. 61–93. CAARI Mono- Spain: Centro Superior de Estudios sobre el Oriente Pro´ximo y graph Series 2. Boston: CAARI. Egipto. Perrot, J. 1966. Le Gisement Natoufien de Mallaha (Eynan), Israe¨l. Stutz, A. J. 2004. The Natufian in real time? radiocarbon date cali- L’anthropologie 70:437–483. bration as a tool for understanding Natufian societies and their Perrot, J., and D. Ladiray. 1988. Les hommes de Mallaha (Eynan), long-term prehistoric context. In The end of the hunter-gatherer Israel.Me´moires et travaux du Centre de Recherche Franc¸aise de societies in the Near East. C. Delage, ed. Pp. 13–37. BAR Inter- Je´rusalem, no. 7. Paris: Association Pale´orient. national Series 1320. Oxford: BAR. Quintero, L. A. 1998. Evolution of Lithic economies in the Levantine Svoboda, J. 1997. Symbolisme gravettien en Moravie: espace, temps Neolithic: development and demise of naviform core technology. et formes. Bulletin de la Socie´te´ Pre´historique de l’Arie`ge 52:87–103. PhD dissertation, University of California, Riverside. Rollefson, G. O. 2000. Ritual and social structure at Neolithic ‘Ain Tchernov, E., and F. R. Valla. 1997. Two new dogs, and other Natufian Ghazal. In Life in Neolithic farming communities: social organiza- dogs, from the southern Levant. Journal of Archaeological Science tion, identity, and differentiation. I. Kuijt, ed. Pp. 163–190. New 24:65–95. York: Kluwer Academic/Plenum. Testart, A. 2008. Des cranes et des vautours ou la guerre oublie. ———. 2001. The Neolithic period. In The archaeology of Jordan. Pale´orient 34:33–58. B. MacDonald, R. Adams, and P. Bienkowski, eds. Pp. 67–105. Twiss, K. C. 2008. Transformations in an early agricultural society: Sheffield: Sheffield Academic Press. feasting in the southern Levantine Pre-Pottery Neolithic. Journal Rollefson, G. O., and I. Ko¨hler-Rollefson. 1989. The collapse of Early of Anthropological Archaeology 27:418–442. Neolithic settlements in the southern Levant. In People and culture Valla, F. R. 1989. Aspects de sol de L’abri 131 de Mallaha (Eynan). in change. I. Hershkovitz, ed. Pp. 73–89. BAR International Series In Pre´histoire du Levant 2. O. Aurenche, M.-C. Cauvin, and P. 508. Oxford: BAR. Sanlaville, eds. Pp. 283–296. Paris: CNRS. Rosenberg, D. 2008. Serving meals making a home: the PPNA lime- ———. 1999. The Natufian: a coherent thought? In Dorothy Garrod stone vessel industry of the southern Levant and its importance and the progress of the Palaeolithic. W. Davies and R. Charles, eds. to the Neolithic revolution. Pale´orient 34(1):23–32. Pp. 224–241. Oxford: Oxbow. Schmidt, K. 2005. “Ritual centers” and the Neolithisation of Upper Verhoeven, M. 2002a. Ritual and ideology in the Pre-Pottery Neo- Mesopotamia. Neo-Lithics 2(5):13–21. lithic B of the Levant and southeast Anatolia. Cambridge Archae- ———. 2006. Sie Bauten die ersten Tempel: das Ratselhafte Heiligtum ological Journal 12:233–258. der Steinzeitjager. Berlin: Beck. ———. 2002b. Ritual and its investigation in prehistory. In Magic Simmons, A. H. 2007. The Neolithic revolution in the Near East: practices and ritual in the Near Eastern Neolithic. H.-G. K. Gebel, transforming the human landscape. Tucson: University of Arizona B. Dahl Hermansen, and C. Hoffman Jensen, eds. Pp. 5–42. Berlin: Press. Ex Oriente. Solecki, R. S., R. L. Solecki, and A. P. Agelarakis. 2004. The proto- ———. 2002c. Transformations of society: the changing role of ritual S220 Current Anthropology Volume 52, Supplement 4, October 2011

and symbolism in the PPNB and the PN in the Levant, Syria and and society. In The meaning of things. I. Hodder, ed. Pp. 57–61. south-east Anatolia. Pale´orient 28:5–14. London: Harper-Collins Academic. Watkins, T. 1990. The origins of house and home? World Archaeology Wobst, H. M. 1974. Boundary conditions for Palaeolithic social sys- 21:336–347. tems: a simulation approach. American Antiquity 39:147–178. Weinstein-Evron, M., and S. Ilani. 1994. Provenance of ochre in the ———. 1977. Stylistic behavior and information exchange. In For Natufian layers of el-Wad Cave, Mount Carmel, Israel. Journal of the director: research in honor of J. B. Griffin. C. E. Cleland, ed. Archaeological Science 21:461–467. Pp. 317–342: Ann Arbor: University of Michigan, Museum of Weinstein-Evron, M., D. Kaufman, and N. Bird-David. 2001. Rolling Anthropology. stones: basalt implements as evidence for trade/exchange in the ———. 1999. Style in archaeology or archaeologists in style. In Ma- Levantine Epipalaeolithic. Journal of the Israel Prehistoric Society— Mitekufat Haeven 31:25–42. terial meanings. E. S. Chilton, ed. Pp. 118–132. Salt Lake City: Weiss, E., M. E. Kislev, O. Simchoni, D. Nadel, and H. Tschauner. University of Utah Press. 2008. Plant-food preparation area on an Upper Paleolithic brush Wright, G. A. 1978. Social differentiation in the Early Natufian. In hut floor at Ohalo II, Israel. Journal of Archaeological Science 35: Social archaeology: beyond subsistence and dating.C.L.Redman, 2400–2414. M. J. Berman, E. V. Curint, J. W. T. Langhorne, N. M. Versaggi, Wiessner, P. 1989. Style and changing relations between the individual and J. C. Wanser, eds. Pp. 201–233. New York: Academic Press. Current Anthropology Volume 52, Supplement 4, October 2011 S221

The Origins of Agriculture in the Near East

by Melinda A. Zeder

The emerging picture of plant and animal domestication and agricultural origins in the Near East is dramatically different from that drawn 16 years ago in a landmark article by Bar-Yosef and Meadow. While in 1995 there appeared to have been at least a 1,500-year gap between plant and animal domestication, it now seems that both occurred at roughly the same time, with initial management of morphologically wild future plant and animal domesticates reaching back to at least 11,500 cal BP, if not earlier. A focus on the southern Levant as the core area for crop domestication and diffusion has been replaced by a more pluralistic view that sees domestication of various crops and livestock occurring, sometimes multiple times in the same species, across the entire region. Morphological change can no longer be held to be a leading-edge indicator of domestication. Instead, it appears that a long period of increasingly intensive human management preceded the manifestation of archaeologically detectable morphological change in managed crops and livestock. Agriculture in the Near East arose in the context of broad-based systematic human efforts at modifying local environ- ments and biotic communities to encourage plant and animal resources of economic interest. This process took place across the entire Fertile Crescent during a period of dramatic post-Pleistocene climate and environmental change with considerable regional variation in the scope and intensity of these activities as well as in the range of resources being manipulated.

Introduction understanding of Near Eastern plant and animal domestication and agricultural emergence in the years between the Santa Fe and Me´rida conferences. Eighteen years ago, a week-long seminar was held in Santa Fe, New Mexico, that, much like the Wenner-Gren Me´rida con- ference featured in this special issue of Current Anthropology, Near Eastern Agricultural Origins: 1995 focused on the context, timing, and possible causes of the emer- gence of agriculture in different world areas. Sponsored by the While comprehensive in its geographic scope, Bar-Yosef and School of American Research, this seminar resulted in the pub- Meadow (1995) had a special emphasis on the Levant, es- lication of an influential edited volume, Last Hunters, First Far- pecially on the southern Levant (figs. 1, 2). Decades of survey mers: New Perspectives on the Prehistoric Transition to Agriculture and excavation, especially in the parts of the Levant that fell (Price and Gebauer 1995), a comprehensive global overview of within the borders of modern Israel, had yielded a remarkably agricultural origins. The contribution by Ofer Bar-Yosef and detailed and well-controlled archaeological record of the tran- Richard Meadow, in particular, provided a richly detailed ac- sition from foraging to farming in this part of the Near East. count of the transition from foraging to farming in the Near Similar coverage had not yet been accomplished in other parts East (Bar-Yosef and Meadow 1995). The scope and breadth of of the Fertile Crescent. When the Bar-Yosef and Meadow the Bar-Yosef and Meadow article likely explains why it has article was published, documenting domestication in plants been the most authoritative and most widely cited synthesis of and animals required the detection of morphological modi- Near Eastern agricultural origins. This work, then, serves as an fications caused by domestication. In cereals, the marker of ideal benchmark against which to measure advances in our choice was the development of a tough rachis, a change in the plant’s dispersal mechanism thought to arise when hu- mans sowed harvested cereal grains. In pulses, the primary Melinda A. Zeder is a Senior Research Scientist in the Program for domestication marker was an increase in seed size, a response Human Ecology and Archaeobiology, National Museum of Natural History, Smithsonian Institution, Department of Anthropology to seedbed pressures that allowed sown seeds to germinate (MRC 112, National Museum of Natural History, Smithsonian more quickly and shade out competing seedlings. In animals, Institution, Washington, DC 20013-7012, U.S.A. [[email protected]]). archaeozoologists relied primarily on the demonstration of This paper was submitted 13 XI 09, accepted 3 I 11, and electronically overall body-size reduction, held to be a rapid response to published 23 VIII 11. herd management.

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0006$10.00. DOI: 10.1086/659307 S222 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 1. Distribution of main Late Epipaleolithic and Neolithic sites in the Near East. 1, Ohalo II; 2, Ein Gev IV; 3, Neve David; 4, Kharaheh IV; 5, Beidha; 6, Hayonim; 7, Wadi al-Hammeh 27; 8, Ain Mallaha; 9, Jericho; 10, Iraq ed Dubb; 11, Hatoula; 12, Dhra; 13, Netiv Hagdud; 14, Gigal I; 15, Aswad; 16, Ghoraife; 17, Wadi el-Jilat 7; 18, Yiftah’el; 19,Ain Ghazal; 20, Basta; 21, Ramad; 22, Khirbet Hammam; 23, Abu Hureyra; 24, Mureybit; 25, Dja’de; 26, Jerf el Ahmar; 27, Kosak Shamali; 28, Halula; 29, Qaramel; 30, Tel el-Kerkh; 31, Ras Shamra; 32, Bouqras; 33, Hallan C¸emi; 34, Demirko¨y; 35,Ko¨rtik; 36,Go¨bekli Tepe; 37, Nevali C¸ori; 38, C¸ayo¨nu¨; 39, Cafer Ho¨yu¨k; 40, Grittle; 41, Palegawra; 42, Shanidar cave; 43, Zawi Chemi Shanidar; 44, Qermez Dere; 45, Nemrik; 46, M’lefaat; 47, Asiab; 48, Ganj Dareh; 49, Ali Kosh; 50, Jarmo; 51, Guran; 52, Sarab; 53, Pinarbassi A; 54,As¸ikli Ho¨yu¨k; 55, Suberde; 56, Can Hasan III; 57, C¸atal Ho¨yu¨k; 58, Erbaba; 59, Aetokremnos; 60, Mylouthikia; 61, Shil- lourokambos.

Based on these criteria, crop domestication was thought to domesticated in other parts of the Fertile Crescent, the south- have originated in the southern Levant during the Pre-Pottery ern Levant was thought to be the home of initial cultivation Neolithic A (PPNA) period, around 11,500–11,000 cal BP (fig. from which domesticates and domestic technology spread 2). Animal domestication seemed to have been a delayed de- quickly into the rest of the Fertile Crescent through an “un- velopment, with different livestock species brought under do- even series of movements affecting different areas at different mestication in different parts of the region (from the Levant times” (Bar-Yosef and Meadow 1995:41). The coalescence of to the Zagros), beginning with goats sometime during the disparate elements of this subsistence system into an agri- Middle Pre-Pottery Neolithic B (PPNB), around 10,000 cal cultural economy was thought to have occurred over a 2,000- BP, followed by sheep, with cattle and pigs domesticated later year period, from about 10,000 to 8000 cal BP, during which still. While livestock and some crop plants may have been Zeder Origins of Agriculture in the Near East S223

Figure 2. Time line of Near Eastern sites, Levantine chronology, and climatic conditions compiled using information from Aurenche et al. (2001); Bar-Yosef and Meadow (1995); Byrd (2005); Kuijt and Goring- Morris (2002); Nesbitt (2002); and Willcox (2005). PPNA, PPNBp Pre- Pottery Neolithic A and B, respectively. time it gradually became the dominant subsistence economy thoroughly explored in 1995. A number of new archaeo- throughout the region. biological approaches to documenting domestication have been developed that are providing powerful new insights into the initial phases of domestication in both plants and animals. Also contributing to the emerging picture of Near Eastern Near Eastern Agricultural Origins: 2010 agricultural origins are genetic analyses that have identified In the 16 years since publication of Bar-Yosef and Meadow the progenitors of Near Eastern domestic crops and livestock (1995), there has been an exponential increase in information species and defined the likely geographic regions of their do- on this transition not only from the southern Levant but also mestication. More widespread use of small-sample accelerator from other parts of the Fertile Crescent that had not been as mass spectrometry (AMS) radiocarbon dating has made it S224 Current Anthropology Volume 52, Supplement 4, October 2011 possible to directly and precisely date the remains of domestic assemblage before it can be considered domestic (Tanno and plants and animals, greatly enhancing the temporal control Willcox 2006a; Weiss, Kislev, and Hartmann 2006). This of our understanding of this transition. The result is a vastly means that the barley recovered from these early sites (where changed picture of the origins of agriculture in the Near East. tough-rachis varieties constitute about 4% of the total barley recovered) more likely represent intensively collected and pos- sibly cultivated morphologically wild cereals (Kislev 1997; New Archaeological Insights into Plant Domestication Weiss, Kislev, and Hartmann 2006). The application of AMS Cereals. When Bar-Yosef and Meadow (1995) was written, dating to carbonized material recovered from new excavations the presence of a few large domestic grains of einkorn (Tri- at Tell Aswad has moved up the dates of the more securely ticum monococcum cf. monococcum) and rye (Secale cf. cereale) identified domesticated emmer and barley from this site. from Epipaleolithic levels at Abu Hureyra I had raised the Originally thought to date to the PPNA (ca. 11,500 cal BP), possibility that initial cereal domestication occurred in the the levels that yielded these domestic cereals are now known northern Levant during the Younger Dryas climatic downturn to date to the end of the Early and beginning of the Middle (Hillman, Colledge, and Harris 1989). Subsequent AMS dat- PPNB (ca. 10,300–10,000 cal BP; Stordeur 2003; Willcox ing of these grains found that, as suspected, most were in- 2005). trusive from Middle PPNB levels. However, three grains of Nesbitt’s comprehensive evaluation of the evidence for the domestic-morphotype rye were found to date to between appearance of domesticated cereals in the Near East concludes 13,000 and 12,000 cal BP, and on the basis of this early date, that the evidence for morphologically altered cereal domes- Hillman argued that these grains represented the earliest mor- ticates before about 10,500 cal BP is either too poorly doc- phologically altered domestic cereals in the Near East (Hill- umented or too poorly dated to be accepted as marking the man 2000; but see Nesbitt 2002:118–119). Hillman acknowl- initial threshold of cereal domestication (Nesbitt 2002). The edged that grains consistent in size with domestic varieties earliest securely identified and dated domestic emmer (Tri- are known to occur in low numbers in wild cereals but argued ticum turgidum ssp. dicoccum) and einkorn (T. monococcum that the probability of finding these rare mutant forms within ssp. monococcum) grains and chaff, according to Nesbitt, come archaeological assemblages of collected wild rye was essentially from sites in the Upper Euphrates valley (Nevali C¸ori, Cafer zero (Hillman 2000:382). If rye was domesticated at this early Ho¨yu¨k, and possibly C¸ayo¨nu¨) that date to the Early PPNB, date, however, it does not seem to have made much of a mark at about 10,500–10,200 cal BP. Nesbitt contends that securely on Near Eastern subsistence economies. With the onset of identified and dated domestic barley is not seen until the warmer and wetter climatic conditions in the Early Holocene, Middle PPNB, when it is found throughout the Fertile Cres- the utilization of this cool-climate cereal first declined and cent and Anatolian Plateau. then ceased altogether in the Middle Euphrates (Willcox, For- Additional evidence for the late or at least delayed ap- nite, and Herveux 2008). Domesticated rye is not seen again pearance of morphologically domestic cereals in the Near East for another 2,000 years, when it is found in low numbers in is provided by Tanno and Willcox (2006a), who document central Anatolia at Can Hasan III (ca. 9400 cal BP; Hillman the gradual increase in the proportion of tough-rachis do- 1978). Never a prominent component of Near Eastern cereal mestic morphotypes among wheat and barley recovered from crops, modern domestic rye traces its heritage to European sites in the Middle and Upper Euphrates valley. Domestic wild rye (Weiss, Kislev, and Hartmann 2006). morphotypes constitute only 10% of the single-grained ein- Arguments advanced in 1995 for the appearance of mor- korn recovered from Nevali C¸ori (ca. 10,200 cal BP), barely phologically altered domestic barley and emmer during the meeting the threshold for demonstrating the presence of do- PPNA in the southern Levant have been largely overturned mestic cereals. Only 35% of the barley recovered from some- in the intervening years (Weiss and Zohary 2011). When Bar- what later levels at Aswad (10,200–9500 cal BP) and a little Yosef and Meadow (1995) went to press, the handful of tough- over 50% of the barley recovered from Ramad (9500–8500 rachis barley Hordeum vulgare ssp. distictum found among cal BP) are nonshattering varieties. Even as late as 7500 cal the thousands of brittle-rachis wild barley grains H. vulgare BP, domestic morphotypes constitute only around 60% of the ssp. spontaneum recovered at Gigal and Netiv Hagdud in the two-grained einkorn recovered from Kosak Shamali, a variety southern Levant seemed likely candidates for the earliest do- that Willcox postulates represents a second domestication of mesticated cereal crop. Although this evidence was questioned diploid wheats (Willcox 2005:537). at the time by Kislev (1989, 1992), who maintained that the low proportion of tough-rachis barley in the Netiv Hagdud Pulses. Although substantial quantities of lentils had been assemblage was consistent with the representation of this mor- recovered from PPNA sites in both the southern and the photype in wild stands, others seemed more comfortable with northern Levant by 1995, the absence of clear morphological the attribution of these cereal remains as domestic (i.e., Bar- markers of domestication (i.e., larger seed sizes) precluded Yosef and Meadow 1995:66–67; Hillman and Davies 1992; Bar-Yosef and Meadow from drawing any conclusions about Zohary 1992). There is now a more general consensus that their domestic status. However, Weiss, Kislev, and Hartmann tough-rachis grains must constitute at least 10% of a cereal (2006; also Weiss and Zohary 2011) and Tanno and Willcox Zeder Origins of Agriculture in the Near East S225

(2006b; also Willcox, Buxo´, and Herveux 2009; Willcox, For- represent this infertile variety. Domestication of this shrubby nite, and Herveux 2008) have subsequently concluded that pioneer plant, they maintain, could be accomplished by re- the hundreds of lentils found in storage bins at Netiv Hagdud planting cut branches of trees that naturally produce these and Jerf el Ahmar are unlikely to represent wild, unmanaged sweeter fruits. Such an activity underscores the degree to plants. Wild lentils, they argue, are not a common component which people were likely modifying local environments and of Near Eastern plant communities, and the yield of seeds biotic communities as well as their willingness to invest in per plant, at about 10–20, is very low. Moreover, wild lentils nurturing resources, such as slowly maturing trees, with de- have an exceptionally high rate of seed dormancy; only about layed rewards. 10% of wild lentil seeds germinate after sowing. Thus, the quantity of lentils recovered from these of PPNA sites suggests Plant management. There is, in fact, increasing evidence that lentils were likely being transplanted from wild patches, that humans were actively modifying local ecosystems and aggregated in new environments, and tended by humans. manipulating biotic communities to increase the availability Weiss, Kislev, and Hartmann (2006) also argue that these early of certain economically important plant resources for hun- lentils had undergone a lowering in the rate of seed dormancy dreds of years before the manifestation of morphological in- and an increase in the number of seeds per plant, initial steps dicators of plant domestication (Weiss, Kislev, and Hartmann toward domestication that would not be archaeologically de- 2006; Willcox, Buxo´, and Herveux 2009; Willcox, Fornite, tectable. and Herveux 2008). First, the presence of distinctive com- Similarly, Tanno and Willcox (2006b) maintain that the plexes of weedy species characteristic of fields under human chickpeas (Cicer sp.) recovered from Tel el-Kerkh (ca. 10,200 cultivation suggests that humans were actively tilling and cal BP) in northwestern Syria represent an early stage in the tending wild stands of einkorn and rye at both Abu Hureyra cultivation of this well-known Near Eastern crop plant. While and nearby Mureybit during the Late Epipaleolithic (ca. these are not definitively domestic morphotypes, the high 13,000–12,000 cal BP; Colledge 1998, 2002; Hillman 2000: degree of morphological variability of the chickpeas from this 378). Increases in this weed complex at Qaramel (ca. 11,500 site, together with the rarity and sparseness of wild chickpea cal BP) and Jerf el Ahmar (ca. 11,000 cal BP) signals an stands (which do not grow in the region today), is again intensification of plant cultivation in the Middle Euphrates suggestive of intentional transplanting and cultivation. A sim- during the ensuing PPNA period (Willcox, Fornite, and Her- ilar case is made for the faba beans (Vicia faba) recovered veux 2008). Willcox, Fornite, and Herveux (2008; also Will- from this site (Tanno and Willcox 2006b). Although not as cox, Buxo´, and Herveux 2009) also interpret the increase in large as modern faba beans, they are very similar to the faba the quantity of wild einkorn in Early Holocene assemblages beans recovered in large numbers from the Late PPNB (ca. from the Middle Euphrates sites as additional evidence of 8800 cal BP) at Yiftah’el (Garfinkel, Kislev, and Zohary 1988), human management of this plant. Wild einkorn T. monococ- which are almost certainly cultivated varieties. In fact, the cum ssp. baeoticum is not well adapted to the chalky soils of large-seeded modern variety of faba bean is not seen in the the Middle Euphrates, and it would not have responded well Near East until about AD 1000 (Tanno and Willcox 2006b). to the rising temperatures of the Early Holocene. Today the region is too hot and arid for wild einkorn, which can be Figs. Recently, Kislev, Hartmann, and Bar-Yosef (2006a) found only on basalt lava flows 100 km north of Jerf el Ahmar. have argued that the earliest morphologically altered plant The dramatic increase in the representation of wild einkorn domesticate in the Near East was neither a cereal nor a pulse in Middle Euphrates assemblages over the course of the PPNA but a tree crop. The presence of parthenocarpic figs at the to Early PPNB could happen, these authors argue, only if PPNA site of Gigal in the southern Levant (ca. 11,400–11,200 people were actively tending plants transplanted from pre- cal BP) has been interpreted as a clear indication of human ferred habitats, altering local microhabitats, removing com- selection for this mutant infertile fig variety that remains on petition, and artificially diverting water to tended plants (Will- the tree longer and develops sweeter, softer fruit. Other re- cox, Fornite, and Herveux 2008:321). A subtle increase in the searchers have noted, however, that parthenocarpy is known thickness and breadth of barley and einkorn grains from these among wild female fig trees (Denham 2007; Lev-Yadun et al. sites without a corresponding increase in grain length is in- 2006) and therefore, as with the presence of tough-rachis terpreted as a plastic response to cultivation (Willcox 2004). varieties or larger cereal grains in low quantities, their oc- The progressive decrease in indigenous plants of the Euphrates currence in an archaeobotanical assemblage cannot be con- floodplain and the concurrent adoption of and increase in sidered definitive proof of domestication. Kislev, Hartmann, morphologically wild representatives of founder crops such and Bar-Yosef (2006b) have responded that if, as their critics as barley, emmer, lentils, chickpeas, and faba beans have also contend, these figs represent the selective harvest of mutant been used to argue that humans were modifying local plant figs from wild fig trees, at least some seeded varieties would communities and managing morphologically wild but culti- be expected to have been collected along with these rare, vated cereals and pulses (Willcox, Buxo´, and Herveux 2009; naturally occurring parthenocarpic figs. Instead, all of the nine Willcox, Fornite, and Herveux 2008). In addition to the quan- carbonized fruits and 313 single druplets recovered from Gigal tities of lentils recovered from PPNA sites such as Netiv Hag- S226 Current Anthropology Volume 52, Supplement 4, October 2011 dud and Gigal, the large number of morphologically wild harvesting cereals before they were fully ripe or even gleaning barley and wild oats (Avena sterilis) recovered from these sites shattered heads of grain from the ground might have led to (e.g., 260,000 grains of wild barley and 120,000 of wild oats the retention of the brittle rachis in cultivated cereals (Hart- from a single granary at Gigal) suggests that people in the mann, Kislev, and Weiss 2006; Lev-Yadun, Gopher, and Abbo southern Levant were also cultivating plants of economic in- 2006; Tanno and Willcox 2006a; Willcox and Tanno 2006). terest (Weiss, Kislev, and Hartmann 2006). The appearance of morphological change in these founder A study of plant assemblages from the northern Fertile crops is, then, most likely an artifact of a change in man- Crescent by Savard, Nesbitt, and Jones (2006) demonstrates agement or harvesting practices of cultivated crops and not that people in the more eastern parts of the Fertile Crescent a leading-edge indicator of plants being brought under human were also intensively utilizing a wide variety of plant resources, control. with considerable regional variation in the plant species ex- ploited. Late Epipaleolithic residents of Hallan C¸emi, for ex- New Archaeological Insights into Animal Domestication ample, utilized a diverse range of plant species with a special focus on valley-bottom plants such as sea club-rush (Bolbo- Caprines. The utility of morphological markers as leading- schoenus maritimus) as well as dock/knotgrass, large-seeded edge indicators of livestock domestication is even more prob- legumes, and, to a lesser extent, almonds (Amygdalus sp.) and lematic. This is especially true of body-size reduction, the pistachio (Pistacia sp.). A similar assemblage was found at primary marker used to document animal domestication for Demirko¨y, a nearby site occupied shortly after Hallan C¸ emi, the past 30 years. Recent analysis of modern and archaeo- where a number of as yet unidentified small-seeded grasses, logical skeletal assemblages from the Zagros region has shown small-seeded barleys (Hordeum murinum), and some wild that sex and, to a lesser extent, temperature are the most barley (H. vulgare cf. spontaneum) were also recovered. The important factors affecting body size in both sheep (Ovis aries) plant assemblages from roughly contemporary sites in steppic and goats (Capra hircus). Domestic status, on the other hand, environments of northern Iraq (Qermez Dere and M’lefaat) has no effect on the size of female caprines and only a limited are dominated by large-seeded legumes, followed by small- effect on males, manifested as a decrease in the degree of seeded grasses, with small-seeded legumes and wild cereals sexual dimorphism (Zeder 2001, 2005). This work has also (barleys and einkorn/rye) also represented. shown that apparent evidence of domestication-induced The antiquity of this broad-spectrum plant-exploitation body-size reduction in Near Eastern archaeological assem- strategy stretches back at least to the Late Glacial Maximum blages is not, as had been assumed, the result of a morpho- (ca. 23,000 cal BP), as evidenced by the remarkably well- logical response to human management. Instead, the apparent preserved plant assemblage recovered from the waterlogged shift toward smaller animals is an artifact of the different Levantine site of Ohalo II, which contained a diverse array culling strategies employed by hunters, whose interest in max- of large- and small-seeded grasses and legumes (Piperno et imizing the return of the hunt often results in an archaeo- al. 2004; Weiss et al. 2004). There is some indication that the logical assemblage dominated by large prime-age males (Sti- intensive exploitation of this complex of small- and large- ner 1990), and herders, who seek to maximize the long-term seeded cereals, legumes, and other locally available plant re- growth of a herd by culling young males and delaying the sources may reach as far back as the Middle Paleolithic (Albert slaughter of females until they have passed peak reproductive et al. 2003; Lev, Kislev, and Bar-Yosef 2005). It is still an open years (Redding 1981). Because of various taphonomic factors question when, over the course of this long period of in- and methodological practices, the herder’s harvest strategy creasingly intense utilization of plant resources, humans began produces an archaeological assemblage dominated by smaller to actively modify local ecosystems and biotic communities adult females (Zeder 2001, 2008). Comparing assemblages of to encourage the availability of economically important hunted prey animals primarily made up of large adult males plants. But it is clear that by at least 11,500 years ago, humans with those of harvested managed animals dominated by had brought a number of plant species under cultivation and smaller females led to the erroneous conclusion that domes- that except for the manifestation of certain morphological tication-induced body-size reduction had taken place. traits seen in later-domesticated varieties, these plants might The consistent size difference between the skeletal elements arguably be considered domesticated crops. of male and female caprines, however, makes it possible to The delayed expression of domestication-induced mor- compute sex-specific harvest profiles for sheep and goats that phological changes in managed plants (at 10,500–10,000 cal are capable of distinguishing the herding harvest signature BP in cereals and later still in pulses) may be attributable to from the hunter’s prey strategy. In the central Zagros, the the frequent importation of new wild plants when cultivated herding signature of young-male harvest and delayed female crops failed (Tanno and Willcox 2006a). It is also possible slaughter is first detected within the highland natural habitat that early harvesting practices may not have encouraged the of wild goats among the goat remains from the site of Ganj morphological changes to cereal dispersal mechanisms once Dareh, directly dated to 9900 cal BP (Zeder 1999, 2005). The thought to be a first-line marker of cereal domestication. same signal was also detected in the goats from the site of Beating ripened grain heads into baskets, for example, or Ali Kosh, located outside the natural habitat of wild goats on Zeder Origins of Agriculture in the Near East S227 the lowland piedmont of southwestern Iran and first occupied at about 10,200 cal BP (Peters, von den Driesch, and Helmer at about 9500 cal BP. Progressive changes in the size and shape 2005:111). Helmer’s (2008) recent reconsideration of the fau- of goat horns has been noted over the 1,000-year occupation nal remains from Cafer Ho¨yu¨k, (ca. 10,300–9500 cal BP), of this site (Hole, Flannery, and Neely 1969). These changes which focuses on sex ratios and harvest profiles, leads him were a direct response to human management that arose when to conclude that sheep (and likely goats) at this site, though humans assumed control over breeding and eliminated the morphologically wild, were not hunted animals, as he had selective pressure for large horns used in mate competition. originally thought (Helmer 1991). Instead, he maintains that The unequivocal signatures of goat management documented these were “agriomorphic” herded animals, a new term he in the central Zagros are not, however, the earliest evidence coins for “domestic animals which are morphologically close of caprine management in the Near East. As with plants, it to wild ones” (Helmer 2008:169). now seems that the leading edge of animal management At As¸ikli Ho¨yu¨k in central Anatolia (ca. 10,200–9500 cal stretches back at least 1,000 years before the manifestation of BP), Buitenhuis (1997; also Vigne, Buitenhuis, and Davis archaeologically detectable morphological change in managed 1999) has detected demographic evidence for management of animals. morphologically wild caprines, predominately sheep. Ar- Perkins (1964) interpreted the younger age profile of the buckle’s (2008) analysis of faunal remains from Suberde, a sheep from the site of Zawi Chemi Shanidar in the north- site roughly contemporaneous with the latest occupation western Zagros as evidence of sheep domestication in the Late phases at As¸ikli Ho¨yu¨k and the initial occupation of C¸ atal Epipaleolithic (ca. 12,000–11,500 cal BP). A new analysis of Ho¨yu¨k (ca. 9500–8900 cal BP), questions the original inter- this assemblage finds a prey profile focused on 2–3-year-old pretation of this site as a “hunters’ village” (Perkins and Daly male sheep that is, as Perkins noted, a departure from the 1968). Demographic patterns detected among the caprines at prime-adult-male strategy detected for goats in Mousterian this site (again mostly sheep) are instead argued to represent and Upper Paleolithic levels at nearby Shanidar cave (Zeder an early and perhaps transitional form of management of 2008). But this demographic profile is also not consistent with morphologically unaltered animals. Management of morpho- the herd-management signature of young-male and delayed logically domesticated sheep and goats is found in central female harvest detected for goats at Ganj Dareh and Ali Kosh. Anatolia by 9500 cal BP in basal occupation levels of C¸atal A similar focus on 2–3-year-old males has been reported at Ho¨yu¨k (Russell and Martin 2005). the roughly contemporary site of Hallan C¸emi, 300 km to Moving farther south, the first appearance of goats in the the northwest of Zawi Chemi and part of the same Taurus/ assemblage from Abu Hureyra (ca. 9600 cal BP) is accom- Zagros “round-house tradition” (Redding 2005; Rosenberg et panied by demographic data that suggest culling strategies al. 1998). Redding interprets this demographic pattern as a similar to those detected at Ganj Dareh (Legge 1996; Legge prime-male hunting strategy practiced under conditions of and Rowley-Conwy 2000). Goats dominate the assemblage intensive pressure on local wild herds. The eradication of local from the site after about 9300 cal BP, reversing a many-mil- males by sedentary hunters, he argues, created a vacuum that lennia emphasis on hunted gazelle. A similar replacement of attracted younger males with less-established home territories a once-dominant focus on gazelles by one on goats is seen from outside regions. This “male sink” effectively assured a first in the Jordan Valley during the Middle PPNB (10,000– continuous supply of preferred prey while preserving a local 9200 cal BP), with an emphasis on gazelle still evident on the population of females and young. Although this strategy does Mediterranean coastal plain until the Final PPNB/Pre-Pottery not entail the same degree of intentional control over herd Neolithic C (ca. 8500 cal BP; Horwitz 2003; Horwitz et al. demographics found in managed herds, it certainly signals an 1999; Sapir-Hen et al. 2009). Horwitz interprets demographic attempt at increasing the availability of prey by setting a prec- patterns observed in morphologically wild goats from Middle edent for the slaughter of young males and the preservation PPNB sites in the Jordan Valley (composed of immature males of female breeding stock characteristic of herd management. and adult females) as evidence of an ongoing process of in- The demographic profile of the sheep remains from Ko¨rtik dependent domestication (Horwitz 1993, 2003). Other re- Tepe, a somewhat later (ca. 10,900 cal BP) site located 50 km searchers have argued that these managed goats were intro- to the south of Hallan C¸emi, has also been interpreted as a duced from outside the region (Bar-Yosef 2000; Peters et al. transitional strategy between game management and herd 1999). Managed sheep were a late arrival in the Levant, ap- management (Arbuckle and O¨ zkaya 2006). pearing sometime around 9200 cal BP (Horwitz and Ducos The transition from hunting to herding appears to have 1998). The introduction of managed sheep was also delayed been complete by about 10,500 cal BP at Nevali C¸ori, where, in the eastern arm of the Fertile Crescent, where a shift toward using lower-resolution demographic profiling methods, Peters adult-female-dominated assemblages appears in both high- and collaborators have detected changes in the age and sizes land and lowland sites in the Zagros at about 9000 cal BP of caprines consistent with the harvest of herded caprines (Zeder 2008). (Peters, von den Driesch, and Helmer 2005; Peters et al. 1999). Sheep seem to have been the initial early focus of herd man- Cattle. The outlines of cattle (Bos taurus) domestication in agement here, with managed goats introduced from elsewhere the Near East are still sketchy. Although cattle remains from S228 Current Anthropology Volume 52, Supplement 4, October 2011

Early and Middle PPNB (11,000–10,000 cal BP) sites in the through time at the site and data on age and sex as indicative upper and Middle Euphrates valley fall within the size range of a developing association between humans and wild boar of wild aurochs (Bos primigenius), Helmer finds evidence for (Redding 2005; Rosenberg et al. 1998; Rosenberg and Redding a reduction in the degree of sexual dimorphism at several 2000). The larger data set from nearby C¸ayo¨nu¨ clearly shows sites (especially at Halula and Dja’de, but less so at Cafer gradual change in multiple indexes (tooth size, age structure, Ho¨yu¨k and Aswad) that he links to an ongoing process of and biometry) thought to reflect a gradual process in which domestication (Helmer 2008; Helmer and Gourichon 2008; pigs moved from a wild to a commensal to a fully domestic Helmer et al. 2005). Cattle from contemporary sites in the status (Ervynck et al. 2001). Helmer’s (2008) recent reeval- same region (i.e., Mureybit III, Jerf el Ahmar, and Go¨bekli) uation of the Cafer Ho¨yu¨k fauna leads him to conclude that are still highly sexually dimorphic and are thus seen as rep- on the basis of demographic patterns, domestic pigs were resenting wild, hunted cattle. Domestic cattle spread slowly present at the site by 10,300 cal BP. out of this heartland of initial domestication, reaching the As with sheep and cattle, pigs seem to have spread slowly southernmost reaches of the Levant only during the Late down the western and eastern arms of the Fertile Crescent. PPNB (9500–9000) at the earliest (Horwitz et al. 1999) and Domestic pigs are thought to have been present in Middle the southern Zagros around 8500 cal BP (Hole, Flannery, and PPNB levels at Aswad (Helmer and Gourichon 2008), but Neely 1969:303). they did not reach the southernmost end of the Levantine In the 1960s, Perkins argued for a center of cattle domes- corridor until about 9000–8500 cal BP (Horwitz et al. 1999). tication in central Anatolia on the basis of an initial study of Domestic pigs have been identified at Jarmo in the north- remains from C¸atal Ho¨yu¨k (Perkins 1969). The analysis of a western Zagros by 9000 cal BP (Flannery 1983), but they did large sample of remains from the renewed excavations at the not reach lowland southwestern Iran until 6000 cal BP (Hole, site, however, does not support this conclusion (Russell, Mar- Flannery, and Neely 1969). Swine are not evident in central tin, and Buitenhuis 2005). C¸atal Ho¨yu¨k cattle show no evi- Anatolia until sometime after about 8500 cal BP (Martin, dence of body-size reduction, as had been claimed by Perkins, Russell, and Carruthers 2002). and are dominated by older male animals in earlier levels (ca. 9400–8500 cal BP). Final occupation levels at the site (ca. New Genetic Insights into Plant and Animal Domestication 8500–8300 cal BP) see a shift toward a female-dominated profile, although the continued emphasis on animals older Plants. Heun et al.’s (1997) study of domestic einkorn, one than 4 years of age raises doubts about the domestic status of the first multilocus genetic analyses of Near Eastern founder of these animals. crops, concluded that this crop plant had a monophyletic origin. This work traced the ancestry of all modern domestic Pigs. In pigs (Sus scrofa), a reduction in the size of molars, einkorn to a single wild population growing on Karacadag˘ especially in the length of the M3, is thought to be an early volcano in southeastern Turkey, only a few kilometers from marker of domestication (Flannery 1983). Changes in molar archaeological sites that have yielded the earliest evidence of lengths in pigs have been linked to a neotonization of skull single-grained einkorn domestication. A monophyletic origin morphology, which itself is believed to be an artifact of the of this domestic cereal is consistent with a highly localized selection for reduced aggression in animals undergoing do- and relatively rapid domestication process (Brown et al. 2008). mestication. A similar morphological response is also seen in A subsequent study by Kilian et al. (2007), however, contends dogs, where juvenilization of skull morphology is thought to that the wild race named by Heun et al. (1997) as the ancestor result in tooth crowding and size reduction. Like pigs, dogs of domestic einkorn is instead a closely related sister group. are animals thought to have entered into domestication This more distant relationship and the high level of genetic through a commensal route initiated when less wary individ- diversity evident in domestic einkorn, they maintain, argues uals approached human habitations to scavenge for food against a monophyletic origin of this early cereal crop. Their (Zeder, forthcoming). It is hard to know, then, whether the study does not find support for a polyphyletic model in which initial changes in skull, jaw, and tooth morphology seen in multiple geographically and genetically distinct races were these animals reflect the initiation of a true domestic part- separately brought under domestication (see Jones 2004). In- nership with humans or simply an adaptation to anthropo- stead, they propose a “dispersed specific model” in which genic environments required of commensal animals. The large multiple local populations of the originally more widely dis- litter sizes of wild pigs and the demographic partitioning of tributed sister race of wild einkorn were taken under culti- wild boar herds may result in prey profiles that mimic what vation and eventually domesticated multiple times by com- might be expected with management, complicating the ap- munities across a broad area. This model is more in line with plication of demographic modeling to distinguish between archaeological evidence indicating that multiple communities hunting and herding of pigs. from southeastern Turkey to the Middle Euphrates were in- Redding has reported that pigs at Hallan C¸emi show some volved in a protracted process of cultivation of both local and evidence of tooth-size reduction (Redding and Rosenberg imported wild progenitors of later crop plants (Willcox 2005). 1998). He also interprets an increase in the numbers of pigs Heun, Haldorsen, and Vollan (2008) have since defended their Zeder Origins of Agriculture in the Near East S229 initial determination of a monophyletic origin of single- domestication into the southern Levant (Weiss, Kislev, and grained einkorn in southeastern Turkey. They also suggest, Hartmann 2006). A separate southern Levantine domestica- however, that the Middle Euphrates may have been the site tion of a variety of lentils no longer represented among mod- of the domestication of a now extinct two-grained variety of ern domestic lentils cannot, however, be ruled out. wheat from Triticum urartu, a more arid-adapted wheat that commonly contains two-grained spikelets. Animals. While there are multiple genetically distinguish- Similarly, earlier genetic analyses of domestic emmer had able lineages in all major livestock species (Bradley 2006), the concluded that this crop plant had a monophyletic origin degree to which these different lineages represent spatially and with the closest living wild population found in the same temporally discrete “domestication events” in which different Karacadag˘region identified as the home of einkorn domes- populations were brought under domestication independently tication (O¨ zkan et al. 2002). Subsequent studies now point of one another is not entirely clear (Dobney and Larson 2006). to at least two separate domestications of emmer (Brown et In domestic goats, for example, there now appear to be as al. 2008). The geographic distance and degree of cultural in- many as six highly divergent maternal lineages. These include dependence between these events are unclear. In addition to the three lineages originally identified by Luikart et al. (2001, a major domestication event at Karacadag˘, O¨ zkan et al. (2005) 2006)—a dominant A lineage and smaller B and C lineages— now think that there may have been another secondary do- and three additional lineages (D, F, and G) identified in the mestication of a population near the Kartal Mountains 300 past few years (Chen et al. 2005; Joshi et al. 2004; Naderi et km to the west of Karacadag˘. They find no evidence that al. 2007; Sultana, Mannen, and Tsuji 2003). The divergence populations from the southern Levant were involved in em- of these lineages has considerable time depth (ca. 100,000– mer domestication, although there is some indication that 500,000 years), suggesting that each represents a different seg- populations in Iraq and Iran may have contributed to the ment of a larger wild goat population brought under do- gene pool of domesticated emmer (O¨ zkan et al. 2005:1057). mestication (Naderi et al. 2007). A genetic analysis of a large Luo et al. (2007) agree that emmer was most likely first do- sample of modern wild bezoar (Capra aegargrus) goats from mesticated in southeastern Turkey, but they also propose that Iran, Iraq, and Turkey finds all six major domestic maternal there was subsequent hybridization and introgression into lineages represented in present-day bezoar populations (Na- domesticated emmer from wild emmer in the southern Le- deri et al. 2008), patterning that, as Naderi et al. (2008) argue, vant. A less likely scenario for the results of their analysis is traces its roots to the Early Holocene. The weak phylogeo- that a population of wild emmer was independently domes- graphic structure of the domestic lineages among these be- ticated in the southern Levant and later absorbed into the zoars is seen as an artifact of human translocation of animals gene pool of domesticated emmer from southeastern Turkey. during the initial phases of the domestication process, before Initial indications of a single domestication of barley in the the morphological modifications that separate wild from do- Jordan Valley (Badr et al. 2000) have also recently been revised mestic goats arose. Evidence of rapid population growth to accommodate a second domestication of this crop. Morell among bezoars belonging to the C lineage resembles that and Clegg’s (2007) study of wild and traditional races of found in domestic goats in Iran and is not seen among bezoars cultivated barley from the Levant to western China has found that do not belong to domestic lineages. This pattern is taken evidence of the domestication of a variety of barley ancestral as a further sign of human-mitigated demographic control to landraces grown in Central and East Asia. Thought to have and protection of bezoars before complete isolation of man- occurred in the Zagros, this second, geographically quite dis- aged animals from wild ones. Bezoars belonging to the now- tinct domestication corresponds well to archaeological evi- dominant A domestic matriline are concentrated in eastern dence that finds domesticated barley in Zagros sites at about Turkey, conjectured to be the most likely region of initial 10,000 years ago (van Zeist et al. 1984). management of A-lineage goats. While C-lineage bezoars are There is also a concordance between archaeological and most frequently found in southern and central Iran, the C- phylogeographic evidence for pulse domestication. The wild lineage bezoars most closely related to C-lineage domestic chickpea population genetically closest to domestic chickpeas goats were found in southeastern Anatolia. These bezoars are was found growing at the far western end of the current thought to be the descendents of animals translocated from distribution of this plant in southern Turkey (Sudupak, Ak- southern Iran that, along with A-lineage goats, were the first kaya, and Kence 2004), the closest wild samples to Tel el- herded goats to leave the homeland of initial management, Kerkh, which, as noted above, yielded early evidence for cul- animals whose remains are found at sites such as Nevali C¸ ori tivation of this important Near Eastern pulse crop. Genetic at 10,200 cal BP. evidence also points to the initial domestication of lentils This study indicates, then, that all six modern maternal somewhere in southeastern Turkey or northern Syria (Ladi- lineages of domestic goats were brought under initial human zinsky 1989), where there is early evidence for the initial management in a region that stretches from the eastern Taurus chickpea cultivation. The appearance of quantities of culti- to the southern Zagros and Iranian Plateau. Although this vated lentils at contemporary sites such as Gigal suggests that process apparently involved individual communities taking this pulse crop spread quickly out of the homeland of initial local populations of wild goats under control, the geographic S230 Current Anthropology Volume 52, Supplement 4, October 2011 proximity of these populations and the evident human-mit- that the southern Levant was the core area for initial do- igated movement of animals across this region suggests that mestication, and a case could be made that all subsequent these activities were part of a more broad-based, culturally crop and livestock domestication in other parts of the Fertile connected set of economic strategies. Persistence of the ge- Crescent followed on the precedent of the crops, domestic netic signature of these activities among modern bezoars adds technology, and the Neolithic way of life introduced from this support to archaeological indications of a long period of active core region. Since then, the spotlight has shifted to the central and intentional human management of animals before the Fertile Crescent, especially the upper reaches of the Tigris and manifestation of archaeologically detectable morphological Euphrates rivers, which appears to be the homeland of the change in managed animals. initial domestication of a number of founder crops (einkorn, Sheep also show a polyphyletic signature, with three do- emmer, pulses) and three, if not four, livestock species (sheep, mestic lineages all thought to be derived from different pop- pigs, cattle, and possibly goats). By the late 1990s, a compelling ulations of wild mouflon Ovis orientalis, an animal with a case could be made that this region was a “cradle of agri- current range that extends from Anatolia to southeastern Iran culture” in a true Vavlovian sense (Lev-Yadun, Gopher, and (Bruford and Townsend 2006; Hiendleder et al. 2002; Pedrosa Abbo 2000). Genetic and archaeobiological evidence gener- et al. 2005). As in goats, one lineage dominates and is pro- ated since then paints a much less focused, more diffuse pic- posed to have originated among sheep whose descendents are ture of agricultural origins. The emergence of agriculture in now found in eastern Turkey and western Iran. Another lin- the Near East now seems to have been a pluralistic process eage is thought to have arisen from a more eastern population with initial domestication of various crops and livestock oc- of O. orientalis (Hiendleder et al. 2002), while a third is also curring, sometimes multiple times in the same species, across thought to have Turkish roots (Pedrosa et al. 2005). the entire region. Three and perhaps four of the five maternal lineages of In 1995, morphological change in both plants and animals domestic taurine cattle were also likely domesticated in the marked the threshold between wild and domestic. We now Fertile Crescent (Bradley and Magee 2006). One lineage (T3) know that morphological change may have occurred quite is the primary variety that spread throughout Europe (Troy late in the domestication process and can no longer be con- et al. 2001). The T1 lineage, the dominant lineage in North sidered a leading-edge indicator of domestication. In cereals, Africa, has been argued to represent an independent African the transition from brittle to tough rachises may actually have domestication of local wild cattle (Bradley and Magee 2006: been the result of changes in harvest timing and technology 324). Subsequent analysis, however, suggests that like the T3 that took place well after people began sowing harvested seed variety, the T1 lineage also likely arose in the Near East and stock. In pulses, seed-size change lagged behind changes in subsequently spread to North Africa through trade and hu- seed dormancy and plant yield that cannot be detected in the man migration (Achilli et al. 2008). archaeological record. In animals, the impact of human man- While Near Eastern wild boar haplotypes are not repre- agement on body size is now known to have been restricted sented among modern domestic swine (Larson et al. 2005), to a decrease in the degree of sexual dimorphism; alterations ancient-DNA analysis points to at least four different lineages in skull morphology may have resulted from a developing of Near Eastern domestic pigs, two of which were found commensal relationship rather than a two-way domestic part- among Neolithic assemblages in western Europe (Larson et nership; and changes in horn size and form may, like changes al. 2007). Near Eastern matrilines of domestic pigs are re- in rachis morphology, have reflected a change in management placed by domestic swine with maternal origins from Euro- practice rather than the initiation of animal management. In pean wild boar by about 6000 cal BP. fact, in both plants and animals, archaeologically detectable morphological indicators of domestication may have occurred A New Picture of Agricultural Origins in only once managed plants and animals were isolated from the Near East free-living populations and the opportunity for introgression or restocking managed populations with wild ones was elim- The emerging picture of plant and animal domestication and inated. While some may prefer not to call a plant or an animal agricultural origins in the Near East is dramatically different a domesticate until this separation has occurred, concentrat- from that drawn 16 years ago in the landmark Bar-Yosef and ing solely on this late stage of the process will not help us Meadow (1995) article. In 1995, there appeared to have been understand how it began. at least a 1,500-year gap between initial crop domestication When Bar-Yosef and Meadow published their synthesis ar- (ca. 11,400 cal BP) and livestock domestication (ca. 10,000 ticle, it seemed to have taken up to 2,000 years after initial cal BP). It now seems that plant and animal domestication domestication of both plants and animals for fully developed occurred at roughly the same time, with signs of initial man- agricultural economies to coalesce across the entire Fertile agement of morphologically wild future plant and animal Crescent. With the removal of morphological change as a domesticates reaching back to at least 11,500 cal BP, if not leading-edge indicator of domestication, this process seems earlier. to have taken longer still. Stable and highly sustainable sub- At the time this influential article was published, it appeared sistence economies based on a mix of free-living, managed, Zeder Origins of Agriculture in the Near East S231 and fully domesticated resources now seem to have persisted This backward look at the mainland from Cyprus provides for 4,000 years or more before the crystallization of agricul- us with new insight into the initial context of plant and animal tural economies based primarily on domestic crops and live- domestication in the Near East. It suggests that domestication stock in the Near East. and agriculture arose in the context of broad-based systematic The exciting recent discoveries on Cyprus provide a special human efforts at modifying local environments and biotic perspective on this long period of low-level food production communities to encourage plant and animal resources of eco- (sensu Smith 2001) in the mainland Fertile Crescent. This nomic interest, a practice that has been characterized as hu- work has produced solid evidence for the early importation man niche construction or ecosystem engineering (Smith of morphologically domesticated cereals (einkorn, emmer, 2007a, 2007b, 2011). The data emerging over the past 15 years and barley) and morphologically wild but managed animals clearly indicate that active human engagement in ecological (goats and cattle) to Cyprus by 10,500 cal BP, with managed niche construction was taking place across the entire Fertile pigs possibly brought to the island even earlier (Vigne, Car- Crescent during a period of dramatic post-Pleistocene climate re`re, and Guilaine 2003; Vigne et al. 2011; Willcox 2003). This and environmental change (Bar-Yosef 2011), with consider- means that at the same time that the earliest morphologically able regional variation in the scope and intensity of these domesticated einkorn and emmer is found in the Upper Eu- activities as well as in the range of resources being manipu- phrates valley (and even earlier than there is solid evidence lated. In certain instances, this context of human niche con- for morphologically altered domestic barley) and when we struction gave rise to coevolutionary relationships between see the first indications of animal management in the main- humans and certain species that eventually resulted in a full- land Fertile Crescent, people were loading these managed fledged domestic partnership, as it did with einkorn, emmer, plants and animals into boats and carrying them, along with and pulses in southeastern Anatolia, with barley in both the the knowledge of how to successfully care for them, to an southern Levant and the Zagros, and with the four major island 160 km off the Levantine coast. The importation and livestock species in different parts of the broad territory be- successful exploitation of these nascent domesticates on Cy- tween Anatolia to southern Iran. In others instances, the re- prus, where none of these plants and animals occur naturally, lationship never developed beyond the first tentative steps suggests that human control over these budding domesticates toward domestication. Failure to move beyond these initial was more established than is apparent on the mainland, where stages of the domestication process may have been due to the likely continued utilization of free-living populations of either behavioral or biological barriers on the part of the plant these species makes it hard to determine the degree of human or animal species, as perhaps in the case of gazelle in the investment in plant and animal management. southern Levant that are behaviorally unsuited to domesti- What is perhaps even more interesting about the Cyprus cation or with rye that could not survive conditions of in- data is that people also imported fallow deer and foxes to the creasing heat and aridity in the northern Levant. Or it might island, as well as other elements of the mainland biotic com- simply be due to a lack of follow-through by humans, as may munity that do not appear to have been subjected to the same have happened with wild oats that were likely cultivated in degree of human control. People who colonized Cyprus in the southern Levant during the PPNA but were not domes- the eleventh millennium did not selectively choose to import ticated until much later. only those plants and animals with which they had a devel- This broad middle ground between wild and domestic, oping domestic partnership. Instead, they seem to have trans- foraging and farming, hunting and herding makes it hard to ported from the mainland their entire ecological niche, made draw clean lines of demarcation between any of these states. up of a wide range of economically important species ex- Perhaps this is the greatest change in our understanding of ploited with a diverse array of more and less intensive strat- agricultural origins since 1995. The finer-resolution picture egies. It is unlikely that these early pioneers drew strict clas- we are now able to draw of this process in the Near East (and, sificatory boundaries between resources collected from as seen in the other contributions to this volume, in other free-living populations, resources that required a higher de- world areas) not only makes it impossible to identify any gree of encouragement and protection from competition or threshold moments when wild became domestic or hunting predation, and resources that had begun to show physiolog- and gathering became agriculture but also shows that drawing ical, behavioral, or morphological responses to human man- such distinctions actually impedes rather than improves our agement. They simply took with them the world that they understanding of this process. Instead of continuing to try to knew. The apparent relaxation of management strategies over pigeonhole these concepts into tidy definitional categories, a time that Vigne, Carre`re, and Guilaine (2003) have proposed more productive approach would be to embrace the ambi- for goats (perhaps because of the lack of major predators and guity of this middle ground and continue to develop tools a reduced threat from human poaching on this sparsely in- that allow us to watch unfolding developments within this habited island) underscores the fluidity of exploitation strat- neither-nor territory. egies and the blurring of distinctions between degrees of en- In the Near East, this means looking more closely at the gagement in the management of important resources that relationships between humans and plants and animals, es- existed at that time. pecially within the natural habitats of future domesticates. S232 Current Anthropology Volume 52, Supplement 4, October 2011

Continuing to develop archaeobiological and archaeological Publicaties 32. Groningen: Centre for Archaeological Research and tools for examining these evolving relationships will be key, Consultancy. ———. 2011. Climatic fluctuations and early farming in West and as will the development of new genetic approaches, including, East Asia. Current Anthropology 52(suppl. 4):S175–S193. one hopes, the analysis of ancient DNA of plant and animal Bar-Yosef, Ofer, and Richard H. Meadow. 1995. The origins of ag- remains from Near Eastern archaeological contexts. In 2011, riculture in the Near East. In Last hunters, first farmers: new per- we are clearly on the cusp of a new understanding of agri- spectives on the prehistoric transition to agriculture. T. Douglas Price cultural origins in the Near East and elsewhere. One can only and Anne-Birgitte Gebauer, eds. Pp. 39–94. Santa Fe, NM: School of American Research Press. imagine what the picture will look like in 2025. Bradley, Daniel G. 2006. Documenting domestication: reading animal genetic texts. In Documenting domestication: new genetic and ar- chaeological paradigms. Melinda A. Zeder, Daniel G. Bradley, Eve Emshwiller, and Bruce D. Smith, eds. Pp. 273–279. Berkeley: Uni- Acknowledgments versity of California Press. I am very grateful to Ofer Bar-Yosef and Doug Price for Bradley, Daniel G., and David A. Magee. 2006. Genetics and origins of domestic cattle. In Documenting domestication: new genetic and including me among the participants of the Me´rida confer- archaeological paradigms. Melinda A. Zeder, Daniel G. Bradley, Eve ence and to the Wenner-Gren Foundation for sponsoring this Emshwiller, and Bruce D. Smith, eds. Pp. 317–328. Berkeley: Uni- wonderful meeting. Both the venue and the unique format versity of California Press. of this meeting made it the single most enjoyable and pro- Brown, Terence A., Martin K. Jones, Wayne Powell, and Robin G. ductive one I have ever attended. The opportunity to interact Allaby. 2008. The complex origins of domesticated crops in the Fertile Crescent. Trends in Ecology & Evolution 24(2):103–109. with the varied group of researchers brought together in Me´- Bruford, Michael W., and Saffron J. Townsend. 2006. Mitochondrial rida, both in formal sessions and in less-formal evening meet- DNA diversity in modern sheep: implications for domestication. ings on the veranda of the Hacienda Temozon, had a trans- In Documenting domestication: new genetic and archaeological par- formative impact on my understanding of agricultural adigms. Melinda A. Zeder, Daniel G. Bradley, Eve Emshwiller, and origins—as a general process and as it played out in the many Bruce D. Smith, eds. Pp. 306–316. Berkeley: University of Cali- fornia Press. world areas where domestication arose and spread. This paper Buitenhuis, Hijlke. 1997. As¸ikli Ho¨yu¨k: a “protodomestication” site. could not have been written without this intensive and most Anthropozoologica 25–26:655–662. rewarding meeting. The manuscript also benefited from the Byrd, Brian F. 2005. Reassessing the emergence of village life in the comments of two anonymous reviewers and the gentle (and Near East. Journal of Archaeological Research 13(3):231–290. sometimes not so gentle) suggestions and proddings of the Chen, Shan-Yuan, Yan-Hua Su, Shi-Fang Wu, Tao Sha, and Ya-Ping Zhang. 2005. Mitochondrial diversity and phylogeographic struc- conference organizers/volume editors. ture of Chinese domestic goats. Molecular Phylogenetics and Evo- lution 37(3):804–814. References Cited Colledge, Susan M. 1998. Identifying pre-domestication cultivation using multivariate analysis. In The origins of agriculture and plant Achilli, Alessandro, Anna Olivieri, Marco Pellecchia, Cristina Uboldi, domestication. Ardeshir B. Damania, Jan Valkoun, George Willcox, Licia Colli, Nadja Al-Zahery, Matteo Accetturo, et al. 2008. Mi- and Calvin O. Qualset, eds. Pp. 121–131. Aleppo, Syria: ICARDA. tochrondrial genomes of extinct aurochs survive in domestic cattle. ———. 2002. Identifying pre-domestication cultivation using mul- Current Biology 18(4):R157–R158. tivariate analysis: presenting the case for quantification. In The Albert, Rosa, Ofer Bar-Yosef, Liliane Meignen, and Steve Weiner. dawn of farming in the Near East. Rene´ T. J. Cappers and Sytze 2003. Quantitative phytolith study of hearths from the Natufian Bottema, eds. Pp. 141–152. Studies in Near Eastern Production, and Middle Paleolithic levels of Hayonim Cave (Galilee, Israel). Subsistence, and Environment, no. 6. Berlin: Ex Oriente. Journal of Archaeological Sciences 30(4):461–480. Denham, Tim. 2007. Debate: early fig-domestication, or gathering Arbuckle, Benjamin S. 2008. Revisiting Neolithic caprine exploitation of wild parthenocarpic figs? Antiquity 81(312):457–461. at Suberde, Turkey. Journal of Field Archaeology 33(2):219–236. Dobney, Keith, and Greger Larson. 2006. Genetics and animal do- Arbuckle, Benjamin S., and Vecihi O¨ zkaya. 2006. Animal exploitation mestication: new windows on an elusive process. Journal of Zoology at Ko¨rtik Tepe: an early Aceramic Neolithic site in southeastern 269(2):261–271. Turkey. Pale´orient 32(2):113–136. Ervynck, Anton, Keith Dobney, Hitomi Hongo, and Richard Aurenche, Olivier, Philippe Galet, Emmanuelle Re´gangnon-Caroline, Meadow. 2001. Born free? new evidence for the status of pigs from and Jacques E´ vin. 2001. Proto-Neolithic and Neolithic cultures in C¸ayo¨nu¨ Tepesi, eastern Anatolia. Pale´orient 27(2):47–73. the Middle East: the birth of agriculture, livestock raising and Flannery, Kent V. 1983. Early pig domestication in the Fertile Cres- ceramics: a calibrated 14C chronology 12,500–5500 cal BC. Radi- cent: a retrospective look. In The hilly flanks: essays on the prehistory ocarbon 43(3):1191–1202. of Southwest Asia. T. Cuyler Young, Philip E. L. Smith, and Peder Badr, Abdelfattah, K. Mu¨ller, Ralf Scha¨fer-Pregl, Haddad El Rabey, Mortensen, eds. Pp. 163–188. Studies in Ancient Oriental Civili- Sieglinde Effgen, H. H. Ibrahim, Carlo Pozzi, Wolfgang Rohde, zation, no. 36. Chicago: Oriental Institute, University of Chicago. and Francesco Salamini. 2000. On the origin and domestication Garfinkel, Yosef, Mordechai E. Kislev, and Daniel Zohary. 1988. Lentil history of barley (Hordeum vulgare). Molecular Biology and Evo- in the Pre-Pottery Neolithic B Yiftah’el: additional evidence of its lution 17(4):499–510. early domestication. Israel Journal of Botany 37:49–51. Bar-Yosef, Ofer. 2000. The context of animal domestication in South- Hartmann, Anat, Mordechai E. Kislev, and Ehud Weiss. 2006. How western Asia. In Archaeozoology of the Near East IV: proceedings of and when was wild wheat domesticated? Science 313(5785):296. the 4th International Symposium on the Archaeozoology of South- Helmer, Daniel. 1991. Les changements des stra´tegies de chasse dans western Asia and Adjacent Areas. Marjan Mashkour, Alice M. le Ne´olithique pre´ce´ramique de Cafer Ho¨yu¨k est (Turquie). Cahiers Choyke, Hijlke Buitenhuis, and Franc¸ois Poplin, eds. Pp. 185–195. de l’Euphrate 5–6:131–138. Zeder Origins of Agriculture in the Near East S233

———. 2008. Re´vision de la faune de Cafer Ho¨yu¨k (Malatya, Tur- ———. 2003. Temporal and spatial variation in Neolithic caprine quie): apports des me´thodes de l’analyse des me´langes et de exploitation strategies: a case study of fauna from the site of l’analyse de Kernel a` la mise en e´vidence de la domestication. In Yiftah’el (Israel). Pale´orient 29(1):19–58. Archaeozoology of the Near East VIII: proceedings of the 8th Inter- Horwitz, Liora K., and Pierre Ducos. 1998. An investigation into the national Symposium on the Archaeozoology of Southwestern Asia origins of domestic sheep in the southern Levant. In Archaeo- and Adjacent Areas. Emannuelle Vila, Lionel Gourichon, Alice M. zoology of the Near East III: proceedings of the 3rd International Choyke, and Hijlke Buitenhuis, eds. Pp. 169–195. Travaux de la Symposium on the Archaeozoology of Southwestern Asia and Ad- Maison de l’Orient 49. Lyon: Maison de l’Orient et de la Me´di- jacent Areas. Hijlke Buitenhuis, La´szlo´Bartosiewicz, and Alice M. terrane´e. Choyke, eds. Pp. 80–95. Groningen: Centre for Archaeological Helmer, Daniel, and Lionel Gourichon. 2008. Premie`res donne´es sur Research and Consultancy. les modalite´s de subsistence a`Tell Aswad (Syrie, PPNB moyen et Horwitz, Liora K., Eitan Tchernov, Pierre Ducos, Cornelia Becker, re´cent, Ne´olithique ce´ramique ancien): fouilles 2001–2005. In Ar- Angela von den Driesch, Louise Martin, and Andrew Garrard. chaeozoology of the Near East VIII: proceedings of the 8th Inter- 1999. Animal domestication in the southern Levant. Pale´orient national Symposium on the Archaeozoology of Southwestern Asia 25(2):63–80. and Adjacent Areas. Emannuelle Vila, Lionel Gourichon, Alice M. Jones, Martin K. 2004. Between Fertile Crescents: minor grain crops Choyke, and Hijlke Buitenhuis, eds. Pp. 119–152. Travaux de la and agricultural origins. In Traces of ancestry: studies in honor of Maison de l’Orient 49. Lyon: Maison de l’Orient et de la Me´di- Colin Renfrew. M. K. Jones, ed. Pp. 127–135. Cambridge: Mac- terrane´e. donald Institute for Archaeological Research. Helmer, Daniel, Lionel Gourichon, Herve´Monchot, Joris Peters, and Joshi, Manjunath B., Pramod K. Rout, Ajoy K. Mandal, Chris Tyler- Maria San˜a Segui. 2005. Identifying early domestic cattle from Smith, Lalji Singh, and Kumarasamy Thangaraj. 2004. Phylo- Pre-Pottery Neolithic sites on the Middle Euphrates using sexual geography and origin of Indian domestic goats. Molecular Biology dimorphism. In The first steps of animal domestication: new ar- and Evolution 21(3):454–462. chaeobiological approaches. Jean-Denis Vigne, Joris Peters, and Kilian, Benjamin., Hakan O¨ zkan, Alexander Walther, Jochen Kohl, Daniel Helmer, eds. Pp. 86–95. Oxford: Oxbow. Tal Dagan, Francesco Salamini, and William Martin. 2007. Mo- Heun, Manfred, Sylvi Haldorsen, and Kari Vollan. 2008. Reassessing lecular diversity at 18 loci in 321 wild and 92 domesticate lines domestication events in the Near East: einkorn and Triticum ur- reveal no reduction of nucleotide diversity during Triticum mon- artu. Genome 51(6):444–451. ococcum (einkorn) domestication: implications for the origin of Heun, Manfred, Ralf Scha¨fer-Pregl, Dieter Klawan, Renato Castagna, agriculture. Molecular Biology and Evolution 24(12):2657–2688. Monica Accerbi, Basilio Borghi, and Francesco Salamini. 1997. Site Kislev, Mordechai E. 1989. Pre-domesticated cereals in the Pre-Pot- of einkorn wheat domestication identified by DNA fingerprinting. tery Neolithic A period. In People and culture in change. Israel Science 278(5341):1312–1314. Hershkovitz, ed. Pp. 147–152. BAR International Series 508. Ox- Hiendleder, Stefan, Bernhard Kaupe, Rudolf Wassmuth, and Axel ford: British Archaeological Reports. Janke. 2002. Molecular analysis of wild and domestic sheep ques- ———. 1992. Agriculture in the Near East in the seventh millennium tions current nomenclature and provides evidence for domesti- B.C. In Pre´histoire de l’agriculture: nouvelle approches expe´rimentales cation of two different subspecies. Proceedings of the Royal Society et ethnographiques. Patricia C. Anderson, ed. Pp. 87–93. Paris: B: Biological Sciences 269(1494):893–904. Centre National de la Recherche Scientifique. Hillman, Gordon C. 1978. On the origins of domestic rye Secale ———. 1997. Early agriculture and paleoecology of Netiv Hagdud. cereale: the finds from Aceramic Can Hasan III in Turkey. Anatolian In An Early Neolithic village in the Jordan Valley. Ofer Bar-Yosef Studies 28:157–174. and Avi Gopher, eds. Pp. 209–236. Cambridge, MA: Peabody Mu- ———. 2000. Abu Hureyra I: the Epipaleolithic. In Village on the seum of Archaeology and Ethnology, Harvard University. Euphrates: from foraging to farming at Abu Hureyra.AndrewM. Kislev, Mordechai E., Anat Hartmann, and Ofer Bar-Yosef. 2006a. T. Moore, Gordon C. Hillman, and Anthony J. Legge, eds. Pp. Early domesticated fig in the Jordan Valley. Science 312(5778): 327–398. Oxford: Oxford University Press. 1372–1374. Hillman, Gordon C., Susan M. Colledge, and David R. Harris. 1989. ———. 2006b. Response to comment on “Early domesticated fig in Plant-food economy during the Epipaleolithic period at Tell Abu the Jordan Valley.” Science 314(5806):1683b. Hureyra, Syria: dietary diversity, seasonality, and modes of ex- Kuijt, Ian, and Nigel Goring-Morris. 2002. Foraging, farming, and ploitation. In Foraging and farming: the evolution of plant exploi- social complexity in the Pre-Pottery Neolithic of the southern tation. David R. Harris and Gordon C. Hillman, eds. Pp. 240–268. Levant: a review and synthesis. Journal of World Prehistory 16(4): London: Unwin Hyman. 361–440. Hillman, Gordon C., and M. Stuart Davies. 1992. Domestication Ladizinsky, Gideon. 1989. Origin and domestication of the Southwest rates in wild wheats and barley under primitive cultivation: pre- Asian grain legumes. In Foraging and farming: the evolution of plant liminary results and archaeological implications of field measure- exploitation. David R. Harris and Gordon C. Hillman, eds. Pp. ments of selection coefficient. In Pre´histoire de l’agriculture: nou- 374–389. London: Unwin Hyman. velle approches expe´rimentales et ethnographiques. Patricia C. Larson, Greger, Umberto Albarella, Keith Dobney, Peter Rowly- Anderson, ed. Pp. 113–158. Paris: Centre National de la Recherche Conwy, Jo¨ry Schibler, Anne Tresset, Jean-Denis Vigne, et al. 2007. Scientifique. Ancient DNA, pig domestication, and the spread of the Neolithic Hole, Frank, Kent V. Flannery, and James A. Neely. 1969. Prehistory into Europe. Proceedings of the National Academy of Sciences of the and human ecology on the Deh Luran plain. Memoirs of the Mu- USA 104(39):15276–15281. seum of Anthropology, no. 1. Ann Arbor: University of Michigan Larson, Greger, Keith Dobney, Umberto Albarella, Meiying Fang, Press. Elizabeth Matisoo-Smith, Judith Robins, Stewart Lowden, et al. Horwitz, Liora K. 1993. The development of ovicaprine domesti- 2005. Worldwide phylogeography of wild boar reveals multiple cation during the PPNB of the southern Levant. In Archaeozoology centers of pig domestication. Science 307(5715):1618–1621. of the Near East: proceedings of the 1st International Symposium on Legge, Anthony J. 1996. The beginning of caprine domestication in the Archaeozoology of Southwestern Asia and Adjacent Areas. Hijlke Southwest Asia. In The origins and spread of agriculture and pas- Buitenhuis and Anneke T. Clason, eds. Pp. 27–36. Leiden: Uni- toralism in Eurasia. David R. Harris, ed. Pp. 238–263. Washington, versal Book Service. DC: Smithsonian Institution. S234 Current Anthropology Volume 52, Supplement 4, October 2011

Legge, Anthony J., and Peter Rowley-Conwy. 2000. The exploitation Pedrosa, Susana, Metehan Uzun, Juan-Jose´Arranz, Beatriz Guitie´r- of animals. In Village on the Euphrates: from foraging to farming rez-Gil, Fermı´n San Primitivo, and Yolanda Bayo´n. 2005. Evidence at Abu Hureyra. Andrew M. T. Moore, Gordon C. Hillman, and of three maternal lineages in Near Eastern sheep supporting mul- Anthony J. Legge, eds. Pp. 423–474. Oxford: Oxford University tiple domestication events. Proceedings of the Royal Society B: Bi- Press. ological Sciences 272(1577):2211–2217. Lev, Efraim, Mordechai Kislev, and Ofer Bar-Yosef. 2005. Mousterian Perkins, Dexter Jr. 1964. Prehistoric fauna from Shanidar, Iraq. Sci- vegetal food in Kebara Cave, Mt. Carmel. Journal of Archaeological ence 144(3626):1565–1566. Science 32(3):475–484. ———. 1969. The fauna from C¸atal Hu¨yu¨k: evidence for early cattle Lev-Yadun, Simcha, Avi Gopher, and Shahal Abbo. 2000. The cradle domestication in Anatolia. Science 164(3876):177–179. of agriculture. Science 288(5471):1602–1603. Perkins, Dexter Jr., and Patricia Daly. 1968. A hunters village in ———. 2006. How and when was wild wheat domesticated? Science Neolithic Turkey. Scientific American 219(5):96–106. 313(5785):296. Peters, Joris, Daniel Helmer, Angela von den Driesch, and Maria Lev-Yadun, Simcha, Gidi Ne’eman, Shahal Abbo, and Moshe A. San˜a Segui. 1999. Animal husbandry in the northern Levant. Pa- Flaishman. 2006. Comment on “Early domesticated fig in the Jor- le´orient 25(2):27–48. dan Valley.” Science 314(5806):1683a. Peters Joris, Angela von den Driesch, and Daniel Helmer. 2005. The Luikart, Gordon, Helena Ferna´ndez, Marjan Mashkour, Philip R. upper Euphrates-Tigris basin: cradle of agro-pastoralism? In The England, and Pierre Taberlet. 2006. Origins and diffusion of do- first steps of animal domestication: new archaeobiological approaches. mestic goats inferred from DNA markers: example analyses of Jean-Denis Vigne, Joris Peters, and Daniel Helmer, eds. Pp. 96– mtDNA, Y-chromosome and microsatellites. In Documenting do- 124. Oxford: Oxbow. mestication: new genetic and archaeological paradigms. Melinda A. Piperno, Dolores R., Ehud Weiss, Irene Holst, and Dani Nadel. 2004. Zeder, Daniel G. Bradley, Eve Emshwiller, and Bruce D. Smith, Processing of wild cereal grains in the Upper Paleolithic revealed eds. Pp. 294–305. Berkeley: University of California Press. by starch grain analysis. Nature 430(7000):670–673. Luikart, Gordon, Ludovic Gielly, Laurent Excoffier, Jean-Denis Vigne, Price, T. Douglas, and Anne Birgitte Gebauer, eds. 1995. Last hunters, Jean Bouvet, and Pierre Taberlet. 2001. Multiple maternal origins first farmers: new perspectives on the prehistoric transition to agri- and weak phylogeographic structure in domestic goats. Proceedings culture. Santa Fe, NM: School of American Research Press. of the National Academy of Sciences of the USA 98(10):5927–5930. Redding, Richard W. 1981. Decision making in subsistence herding Luo, Ming-.Cheng, Zu-Li Yang, Frank M. You, Taihachi Kawahara, of sheep and goats in the Middle East. PhD dissertation, University J. Giles Waines, and Jan Dvorˇa´k. 2007. The structure of wild and of Michigan, Ann Arbor. domesticated emmer wheat populations, gene flow between them, ———. 2005. Breaking the mold: a consideration of variation in the and the site of emmer domestication. Theoretical and Applied Ge- evolution of animal domestication. In The first steps of animal netics 114(6):947–959. domestication: new archaeobiological approaches. Jean-Denis Vigne, Martin, Louise, Nerissa Russell, and Denise Carruthers. 2002. Animal Joris Peters, and Daniel Helmer, eds. Pp. 41–48. Oxford: Oxbow. remains from the central Anatolian Neolithic. In The Neolithic of Redding, Richard W., and Michael Rosenberg. 1998. Ancestral pigs: central Anatolia: internal developments and external relations during a New (Guinea) model for pig domestication in the Near East. In the 9–6th millennia cal. BC. Fre´de´ric Ge´rard and Laurens Thissen, Ancestors for the pigs: pigs in prehistory. Sarah M. Nelson, ed. Pp. eds. Pp. 193–206. Istanbul: Ege Yayınları. 65–76. MASCA Research Papers in Science and Archaeology 15. Morrell, Peter L., and Michael T. Clegg. 2007. Genetic evidence for Philadelphia: University of Pennsylvania Museum. a second domestication of barley (Hordeum vulgare) east of the Rosenberg, Michael, R. Mark Nesbitt, Richard W. Redding, and Brian Fertile Crescent. Proceedings of the National Academy of Sciences L. Peasnall. 1998. Hallan C¸emi, pig husbandry, and post-Pleisto- of the USA 104(9):3289–3294. cene adaptations along the Taurus-Zagros arc (Turkey). Pale´orient Naderi, Saeid, Hamid-Reza Rezaei, Franc¸ois Pompanon, Michael G. 24(1):25–41. B. Blum, Riccardo Negrini, Hamid-Reza Naghash, O¨ zge Balkız, et Rosenberg, Michael, and Richard W. Redding. 2000. Hallan C¸emi al. 2008. The goat domestication process inferred from large-scale and early village organization in eastern Anatolia. In Life in Ne- mitochondrial DNA analysis of wild and domestic individuals. olithic farming communities: social organization, identity, and dif- Proceedings of the National Academy of Sciences of the USA 105(46): ferentiation. Ian Kuijt, ed. Pp. 39–61. New York: Plenum. 17659–17664. Russell, Nerissa, and Louise Martin. 2005. The C¸atalho¨yu¨k mammal Naderi, Saeid, Hamid-Reza Rezaei, Pierre Taberlet, Ste´phanie Zundel, remains. In Inhabiting C¸atalho¨yu¨k: reports from the 1995–1999 sea- Seyed-Abbas Rafat, Hamid-Reza Nagash, Mohamed A. A. El-Bar- sons. Ian Hodder, ed. Pp. 33–98. McDonald Institute Monographs. ody, Okun Ertugrul, and Franc¸ois Pompanon. 2007. Large-scale Cambridge: McDonald Institute for Archaeological Research. mitochondrial DNA analysis of the domestic goat reveals six hap- Russell, Nerissa, Louise Martin, and Hijlke Buitenhuis. 2005. Cattle logroups with high diversity. PLoS One 2(10):e1012, doi:10.1371/ domestication at C¸atalho¨yu¨k revisited. Current Anthropology journal.pone.0001012. 46(suppl. 5):S101–S108. Nesbitt, Mark. 2002. When and where did domesticated cereals first Sapir-Hen, Lidar, Guy Bar-Oz, Hamoudy Khalaily, and Tamar Dayan. occur in Southwest Asia? In The dawn of farming in the Near East. 2009. Gazelle exploitation in the Early Neolithic site of Motza, Rene´ T. J. Cappers and Sytze Bottema, eds. Pp. 113–132. Studies Israel: the last of the gazelle hunters in the southern Levant. Journal in Early Near Eastern Productions, Subsistence, and Environment of Archaeological Science 36(7):1538–1546. 6. Berlin: Ex Oriente. Savard, Manon, Mark Nesbitt, and Martin K. Jones. 2006. The role O¨ zkan, Hakan, Andrea Brandolini, Carlo Pozzi, Sieglinde Effgen, J. of wild grasses in subsistence and sedentism: new evidence from Wunder, and Francesco Salamini. 2005. A reconsideration of the the northern Fertile Crescent. World Archaeology 38(2):179–196. domestication geography of tetraploid wheats. Theoretical and Ap- Smith, Bruce D. 2001. Low level food production. Journal of Ar- plied Genetics 110(6):1052–1060. chaeological Research 9(1):1–43. O¨ zkan, Hakan, Andrea Brandolini, Ralf Scha¨fer-Pregl, and Francesco ———. 2007a. Niche construction and the behavioral context of Salamini. 2002. AFLP analysis of a collection of tetraploid wheats plant and animal domestication. Evolutionary Anthropology 16(5): indicates the origin of emmer and hard wheat domestication in 188–199. southeast Turkey. Molecular Biology and Evolution 19(10):1797– ———. 2007b. The ultimate ecosystem engineers. Science 315(5820): 1801. 1797–1798. Zeder Origins of Agriculture in the Near East S235

———. 2011. Niche construction and the management of wild plant Weiss, Ehud, and Daniel Zohary. 2011. The Neolithic Southwest and animal resources by small-scale pre-industrial societies. Phil- Asian founder crops: their biology and archaeobotany. Current osophical Transactions of the Royal Society B: Biological Sciences Anthropology 52(suppl. 4):S237–S254. 366(1566):836–848. Willcox, George. 2003. The origins of Cypriot farming. In Le Ne´o- Stiner, Mary C. 1990. The use of mortality patterns in archaeological lithique de Chypre: actes du colloque international organise´par le studies of hominid predatory adaptations. Journal of Anthropo- De´partement des antiquite´s de Chypre et l’E´cole franc¸aise d’Athe`nes, logical Archaeology 9(4):305–351. Nicosie 17–19 mai 2001. Jean Guilaine and Alain Le Brun, eds. Pp. Stordeur, Danielle. 2003. Tell Aswad 2001 et 2002. NeoLithics 1(1): 231–238. Bulletin de Correspondance Helle´nique, supple´ment 43. 7–15. Athens: E´ cole franc¸aise d’Athe`nes. Sudupak, Mehmet A., Mahinur S. Akkaya, and Aykut Kence. 2004. ———. 2004. Measuring grain size and identifying Near Eastern Genetic relationships among perennial and annual Cicer species cereal domestication: evidence from the Euphrates valley. Journal growing in Turkey assessed by AFLP fingerprinting. Theoretical of Archaeological Science 31(2):145–150. and Applied Genetics 108(5):937–944. ———. 2005. The distribution, natural habitats, and availability of Sultana, Saira, Hideyuki Mannen, and Soichi Tsuji. 2003. Mitochon- wild cereals in relation to their domestication in the Near East: drial DNA diversity of Pakistani goats. Animal Genetics 34(6):417– multiple events, multiple centres. Vegetation History and Archaeo- 421. botany 14(4):534–541. Tanno, Ken-ichi, and George Willcox. 2006a. How fast was wild Willcox, George, Ramon Buxo´, and Linda Herveux. 2009. Late Pleis- wheat domesticated? Science 311(5769):1886. tocene and Early Holocene climate and the beginnings of culti- ———. 2006b. The origins of cultivation of Cicer arietinum L. and vation in northern Syria. Holocene 19(1):151–158. Vicia faba L.: early finds from Tell el-Kerkh, north-west Syria, late Willcox, George, Sandra Fornite, and Linda Herveux. 2008. Early 10th millennium B.P. Vegetation History and Archaeobotany 15(3): Holocene cultivation before domestication in northern Syria. Veg- 197–204. etation History and Archaeobotany 17(3):313–325. Troy, Christopher S., David E. MacHugh, Jillian F. Bailey, David A. Willcox, George, and Ken-ichi Tanno. 2006. Response to “How and Magee, Ronan T. Loftus, Patrick Cunningham, Andrew T. Cham- when was wild wheat domesticated?” Science 313(5785):296–297. berlain, Bryan C. Sykes, and Daniel G. Bradley. 2001. Genetic Zeder, Melinda A. 1999. Animal domestication in the Zagros: a review evidence for Near-Eastern origins of European cattle. Nature of past and current research. Pale´orient 25(2):11–25. 410(6832):1088–1091. ———. 2001. A metrical analysis of a collection of modern goats van Zeist, Willem, Philip E. L. Smith, Rita M. Palfenier-Vegter, M. (Capra hircus aegargus and Capra hircus hircus) from Iran and Suwijn, and W. A. Casparie. 1984. An archaeobotanical study of Iraq: implications for the study of caprine domestication. Journal Ganj Dareh Tepe, Iran. Palaeohistoria 26:201–224. Vigne, Jean-Denis, Hijlke Buitenhuis, and Simon Davis. 1999. Les of Archaeological Science 28(1):61–79. premiers pas de la domestication animal a`l’Ouest de l’Euphrate: ———. 2005. New perspectives on livestock domestication in the Chypre et L’Anatolie centrale. Pale´orient 25(2):49–62. Fertile Crescent as viewed from the Zagros Mountains. In The first Vigne, Jean-Denis, Isabelle Carre`re, and Jean Guilaine. 2003. Unsta- steps of animal domestication: new archaeobiological approaches. ble status of early domestic ungulates in the Near East: the example Jean-Denis Vigne, Joris Peters, and Daniel Helmer, eds. Pp. 125– of Shillourokambos (Cyprus, IX–VIIIth millennia cal. B.C.). In Le 147. Oxford: Oxbow. Ne´olithique de Chypre: actes du colloque international organise´par ———. 2008. Animal domestication in the Zagros: an update and le De´partement des antiquite´s de Chypre et l’E´cole franc¸aise directions for future research. In Archaeozoology of the Near East d’Athe`nes, Nicosie 17–19 mai 2001. Jean Guilaine and Alain Le VIII: proceedings of the 8th International Symposium on the Ar- Brun, eds. Pp. 239–251. Bulletin de Correspondance Helle´nique, chaeozoology of Southwestern Asia and Adjacent Areas. Emannuelle supple´ment 43. Athens: E´ cole franc¸aise d’Athe`nes. Vila, Lionel Gourichon, Alice M. Choyke, and Hijlke Buitenhuis, Vigne, Jean-Denis, Isabelle Carre`re, Franc¸ois Briois, and Jean Gui- eds. Pp. 243–278. Travaux de la Maison de l’Orient 49. Lyon: laine. 2011. The early process of mammal domestication in the Maison de l’Orient et de la Me´diterrane´e. Near East: new evidence from the Pre-Neolithic and Pre-Pottery ———. forthcoming. Pathways to animal domestication. In Biodi- Neolithic in Cyprus. Current Anthropology 52(suppl. 4):S255–S271. versity in agriculture: domestication, evolution and sustainability. Weiss, Ehud, Mordechai E. Kislev, and Anat Hartmann. 2006. Au- Paul Gepts, Thomas R. Famula, Robert L. Bettinger, Stephen B. tonomous cultivation before domestication. Science 312(5780): Brush, Ardeshir B. Damania, Patrick E. McGuire, and Calvin O. 1608–1610. Qualset, eds. Cambridge: Cambridge University Press. Weiss, Ehud, Wilma Wetterstrom, Dani Nadel, and Ofer Bar-Yosef. Zohary, Daniel. 1992. Domestication of the Neolithic Near Eastern 2004. The broad spectrum revolution revisited: evidence from crop assemblage. In Pre´histoire de l’agriculture: nouvelle approches plant remains. Proceedings of the National Academy of Sciences of expe´rimentales et ethnographiques. Patricia C. Anderson, ed. Pp. the USA 101(26):9551–9555. 81–86. Paris: Centre National de la Recherche Scientifique.

Current Anthropology Volume 52, Supplement 4, October 2011 S237

The Neolithic Southwest Asian Founder Crops Their Biology and Archaeobotany

by Ehud Weiss and Daniel Zohary

This article reviews the available information on the founder grain crops (einkorn wheat, emmer wheat, barley, lentil, pea, chickpea, and flax) that started agriculture in Southwest Asia during the Pre-Pottery Neolithic period, some 11,000–10,000 years ago. It provides a critical assessment for recognizing domestication traits by focusing on two fields of study: biology and archaeobotany. The data in these fields have increased considerably during the past decade, and new research techniques have added much to our knowledge of progenitor plants and their domesticated derivatives. This article presents the current and accumulated knowledge regarding each plant and illustrates the new picture that emerged on the origin of agriculture.

Introduction (einkorn wheat Triticum monococcum, emmer wheat Triticum turgidum subsp. dicoccum, and barley Hordeum vulgare), four pulses (lentil Lens culinaris, pea Pisum sativum, chickpea Cicer The occasion of the Wenner-Gren conference “The Origins arietinum, and bitter vetch Vicia ervilia), and a single oil and of Agriculture: New Data, New Ideas” raised our awareness fiber crop (flax Linum usitatissimum). of the importance of critical, high-quality data. We suspected From many aspects these eight plants belong to the same that the newly obtained data sets from well-dated archaeo- group—the grain crops. All of them are annuals, self-polli- logical contexts may generate new hypotheses concerning the nated, diploid (except emmer wheat), native to the Fertile first steps of the agricultural revolution in Southwest Asia. Crescent belt, and interfertile within each crop and between For this purpose we concentrated on two domains: the the crop and its wild progenitor. Both the agronomic com- newly obtained rich archaeobotanical assemblages and the pensation and the dietary complementation between plants molecular analysis of present-day accessions. Some of these in this group have been appreciated since the early days of new data are inconclusive for pinpointing the shift from wild agriculture up until now. to domesticated plants because of bad preservation of the The aim of this article is to review the available information archaeological samples and problematic field or lab proce- on these founder crops that started agriculture in the Levant dures. Following the fruitful discussions during the confer- during the Pre-Pottery Neolithic (PPN) period some 11,000– ence, we critically reassessed the available archaeobotanical 10,000 years ago (table 2). We shall review the current knowl- data in an attempt to “separate the grain from the chaff.” In edge of these plants by focusing on two fields of study: (i) the following pages we present the data sets, both old and biology and (ii) archaeobotany. Biological data derived from new, that we consider to be reliable enough to indicate the the research on living plants can indicate what the wild pro- first appearances of domesticated forms in the Fertile Cres- genitors of the domesticated plants may have been and what cent. the selection pressures for domesticated types were. Archaeo- Eight plants are considered to be the domesticated founder botanical information identifies the plants used by hunting crops in the Levant (i.e., the western “arc” of the “Fertile and gathering groups, which were the plants first domesti- Crescent”; table 1). This assemblage includes three cereals cated, and when and where these processes took place. Large amounts of critical new data, botanical and archae- Ehud Weiss is Senior Lecturer at the Institute of Archaeology, Martin ological, have been gathered in the Fertile Crescent belt in (Szusz) Department of Land of Israel Studies and Archaeology, Bar- the past 40 years (fig. 1). These findings established the Levant Ilan University (Ramat-Gan, 52900 Israel [[email protected]]) as the critical area for understanding early domestication of and Visiting Scientist, Kimmel Center for Archaeological Sciences, Weizmann Institute of Science (Rehovot, 76100 Israel). Daniel both plants and animals (and see additionally Zeder 2011). Zohary is Professor in the Department of Evolution, Systematics, (Note that we use the terms “Southwest Asia” and “Near East and Ecology, Hebrew University Jerusalem (91904 Israel). This paper arc” interchangeably in this article; we also use the term “Le- was submitted 13 XI 09, accepted 1 XII 10, and electronically vant” to refer to the western horn of the Fertile Crescent in published 13 V 11. the Near East.)

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0007$10.00. DOI: 10.1086/658367 S238 Current Anthropology Volume 52, Supplement 4, October 2011

Table 1. Eight founder grain (three grass [cereal], four legume [pulse], and one oil and fiber) crops and their wild progenitors that started Neolithic agriculture in the Levant

Domesticated crop Wild progenitor Common name Scientific name Common name Scientific name Einkorn wheat Triticum monococcum subsp. monococcum Wild einkorn T. monococcum subsp. baeoticum Emmer wheat Triticum turgidum subsp. dicoccum Wild emmer T. turgidum subsp. dicoccoides Barley Hordeum vulgare subsp. distichum Wild barley H. vulgare subsp. spontaneum Lentil Lens culinaris Wild lentil Lens orientalis Pea Pisum sativum Wild pea Pisum humile Chickpea Cicer arietinum Wild chickpea Cicer reticulatum Bitter vetch Vicia ervilia Wild bitter vetch Vicia ervilia Flax Linum usitatissimum Wild flax Linum bienne

In this article we will adopt a critical assessment intended earliest definite domesticates as well as molecular and other to recognize domestication traits in Levantine archaeobotany biological data. (Zohary, Hopf, and Weiss 2011). In some way this assessment From the eight founder crops, this article does not deal is partially intended to follow and update Mark Nesbitt’s with bitter vetch Vicia ervilia. According to its large quantities (2002) article. The basics of this assessment are twofold. in Neolithic contexts, this pulse might have been taken into 1. The most reliable trait for the identification of domes- domestication in Anatolia or the Levant (i.e., in the general tication is ear shattering in the cereals, pod indehiscence in area in which it still grows wild today). However, because the pulses, and capsule indehiscence in flax. In wild plants, currently there are no reliable diagnostic traits to morpho- ears and pods/capsules disarticulate at maturation and shatter logically discriminate between its wild and domesticated the seed-dispersal devices or the seeds; in contrast, ears and forms in archaeological remains, the early Turkish finds could pods/capsules stay intact in the domestic plant. This trait is be either. the best effective diagnostic indication for recognizing grain- The source of most dates is Radiocarbon CONTEXT da- crop domestication (i.e., the shift from wild plants to do- tabase (Bo¨hner and Schyle 2002–2006). Table 4 lists the un- mesticated plants), which is controlled mostly by a single calibrated and calibrated radiocarbon dates. In an attempt to major gene locus or two such loci (table 3). simplify understanding of the domestication process, all dates 2. The second diagnostic morphological trait is seed size. in the text are representative dates only and were rounded to In the wild progenitors the seeds are relatively small, while the nearest 50 years (please refer to table 4 for range of dates). in domesticated plants they are frequently larger. This trait Periods mentioned in this article, such as Pre-Pottery Neo- takes longer to develop, it is variable within the plant com- lithic A (PPNA) or Early Pre-Pottery Neolithic B (EPPNB), munity, it is apparently a later development under domesti- do not necessarily imply cultural similarities but are used for cation, and it is not as diagnostic as the first trait. This trait the convenience of comparing among sites that belong to the is controlled by various genes and other factors. same approximate time period. Additional markers for domesticated traits (e.g., Fuller 2007) are relevant mostly for differentiating between wild and domesticated traits in living plants. This is apparently much less so in the archaeobotanical assemblage. Table 2. Cultural-historical sequence for the During the past decade or so, molecular studies became Southern Levantine Pre-Pottery Neolithic periods central in research toward understanding the beginning of Calibrated 14C agriculture. Whether the mode of domestication was mono- Period and entity/phase years BP phyletic or polyphyletic, these studies went in two lines Late Epipaleolithic, Final Natufian 12,500–11,700 (Brown et al. 2009). In the first half of this decade, molecular PPNA 11,600–10,500 studies were regarded as supporting monophyletic origin of PPNB: a single localized event or at least very rare events (e.g., Badr Early 10,500–10,100 et al. 2000; Heun et al. 1997; O¨ zkan et al. 2002). In the second Middle 10,100–9500 half, however, such studies were interpreted as supporting Late 9500–8750 Final/PPNC 8750–8400 polyphyletic domestication of multiple events in more than Early Pottery Neolithic, Yarmukian 8400–7600 one location, such as of einkorn wheat (Kilian et al. 2007) Sources. Goring-Morris and Belfer-Cohen 2011; Kuijt and and barley (e.g., Molina-Cano et al. 2005; Morrell and Clegg Goring-Morris 2002. 2007). An attempt is made to evaluate the available knowl- Note. PPNA p Pre-Pottery Neolithic A; PPNB p Pre- edge, crop by crop, for the above-mentioned signs of the Pottery Neolithic B; PPNC p Pre-Pottery Neolithic C. Weiss and Zohary Founder Crops S239

Figure 1. Distribution of Epipaleolithic and Pre-Pottery Neolithic sites mentioned in the text.

Cereals In domesticated forms, the kernels tend to be wider compared with the wild forms. Wild einkorn is widely distributed over Western Asia and Einkorn Wheat Triticum monococcum penetrates also into the southern Balkans (Harlan and Zohary 1966; Zohary, Harlan, and Vardi 1969). Its distribution center Biology. Einkorn is a small plant, rarely more than 70 cm lies in the Fertile Crescent (i.e., northern Syria, southern Tur- high, with a relatively low yield, but it survives on poor soils key, northern Iraq, and adjacent Iran, as well as some parts where other wheat types usually fail. It is a relatively uniform of western Anatolia; fig. 2). In these regions, wild einkorn is diploid (2np2xp14 chromosomes) wheat with characteristic massively distributed as a component of oak park-forests and hulled grains and delicate ears and spikelets. Most domesti- steppelike formations. cated einkorn varieties produce one grain per spikelet, hence Heun and colleagues (Heun, Haldorsen, and Vollan 2008; its name, but cultivars with two grains exist as well (Harlan Heun et al. 1997) analyzed domesticated and wild einkorn 1981; Schiemann 1948). from the Fertile Crescent and beyond for amplified fragment The wild ancestry of domesticated einkorn wheat is well length polymorphism (AFLP) DNA analysis. According to established. Domesticated Triticum monococcum is closely re- their research, wild einkorn wheat was domesticated in the lated to a group of wild and weedy wheat forms spread over Karacadag˘ range in southeast Turkey. It also supports the Southwest Asia and adjacent territories and traditionally re- assumption of its monophyletic origin. These findings were ferred to as “wild einkorn” or Triticum boeoticum. The most recently supported by Kilian et al. (2007), who found, how- distinguishing trait between wild einkorn and domesticated ever, that several domesticate lines arose at the beginning of einkorn is the mode of seed dispersal. Wild forms have brittle einkorn cultivation (i.e., that einkorn went through a number ears, and the individual spikelets disarticulate at maturity to of independent domestication events). These lines are all part disperse the seed. In domesticated einkorn the mature ear of the einkorn gene pool that grows wild in southeastern remains intact and breaks into individual spikelets only on Turkey. pressure (threshing). The shape of the grain is another di- agnostic character indicating domestication (van Zeist 1976). Archaeobotany. The earliest definite domesticated einkorn S240 Current Anthropology Volume 52, Supplement 4, October 2011

Table 3. Recessive mutations that changed wild-type trait into domestic-type trait in the Near East founder crops

No. recessive Crop Wild-type trait Domestication trait mutations Source Einkorn wheat Shattering ears Nonshattering ears 1 Love and Craig 1924 Emmer wheat Shattering ears Nonshattering ears 2 Sharma and Waines 1980 Barley Shattering ears Nonshattering ears 2 Takahashi 1955; Zohary 1960:41 Lentil Dehiscent pod Dehiscent pod 1 Ladizinsky 1979 Pea Dehiscent pod Indehiscent pod 1 Waines 1975 Chickpea Dehiscent pod Indehiscent pod 1 Kazan et al. 1993 Bitter vetch Dehiscent pod Indehiscent pod Unknown Flax Dehiscent capsule Indehiscent capsule 1 Diederichsen and Hammer 1995; Gill and Yermanos 1967 Note. These mutations are responsible for the shift from wild-type seed dispersal to human-dependent crops. There is no conclusive information yet regarding the situation in bitter vetch. wheat appears in two EPPNB sites, namely ca. 10,600–9900 review, see Feldman, Lupton, and Miller 1995; Zohary 1969) cal BP C¸ayo¨nu¨ (van Zeist and de Roller 1991–1992, 2003) that the domesticated tetraploid turgidum wheats (both hulled and Cafer Ho¨yu¨k (de Moulins 1997), southern Turkey (fig. dicoccum forms and free-threshing durum-type varieties) are 1). In these sites, situated within the present range of distri- closely related to wild wheat native to Southwest Asia and bution of wild einkorn (fig. 2), numerous charred spikelet traditionally called Triticum dicoccoides (wild emmer wheat). forks were found showing rough breakage scars. This is annual, predominantly self-pollinated wheat with char- From these localities, einkorn spreads farther south to Mid- acteristic large and brittle ears and big elongated grains that dle Pre-Pottery Neolithic B (MPPNB), ca. 10,200–9550 cal show a striking similarity to some domesticated emmer and BP, Tell Aswad near Damascus (van Zeist and Bakker-Heeres durum cultivars. 1982 [1985]) and Jericho (Hopf 1983), where its remains, In the tetraploid turgidum wheats, the most conspicuous however, are relatively few. Generally, this wheat appears to diagnostic difference between wild and tame is the seed- be less frequent than the emmer wheat and barley. dispersal mechanism (Zohary 1969; Zohary and Brick 1961). Wild dicoccoides wheats have brittle ears that shatter on ma- Emmer and Durum-Type Wheats Triticum turgidum turity into individual spikelets. Each spikelet operates as an arrowlike device disseminating the seed by inserting them into Biology. Triticum turgidum is a varied aggregate of domes- the ground. The “wild-type” rachis disarticulation and the ticated and wild tetraploid (2np4xp28 chromosomes) spikelet morphology reflect specialization in seed dissemi- wheats. According to their response to threshing, the domestic turgidum wheats fall into two groups of cultivars that are nation that ensures the survival of the wild forms under wild recognizable in archaeological remains. conditions. In the human-made system of reaping, threshing, 1. Hulled nonshattering emmer wheat, Triticum turgidum and sowing, this major adaptation broke down, and selection subsp. dicoccum (traditionally called Triticum dicoccum), in resulted in the evolution of human-dependant nonbrittle which the products of threshing are the individual spikelets. types. Significantly, the shift from a brittle spike (in wild The grains remain invested by the glumes and pales. In do- dicoccoides wheats) to a nonbrittle spike (“hulled type” in mestic emmer, as in einkorn, threshing results in breaking the domesticated dicoccum wheats) is governed by a single re- rachis of the ear in its weakest points below each spikelet. cessive mutation in a major gene (table 3). In addition, wild Emmer represents the primitive situation in domesticated tur- and domesticated forms differ from one another in kernel gidum wheats. morphology (van Zeist 1976). In domesticated dicoccum and 2. The more advanced free-threshing emmer, which evolved durum forms, the grain tends to be wider and thicker as well under domestication from hulled emmer. This is a group of as rounder in cross section compared with the wild dicoccoides intraspecific taxa; its common representative today is durum counterpart. wheat, T. turgidum conv. durum. The glumes in all these do- Wild emmer, T. turgidum subsp. dicoccoides,ismorere- mesticated tetraploid types are relatively thin. Threshing stricted in its distribution and more confined in its ecology breaks the glumes at their bases and frees the naked grains. than wild einkorn. Its range covers Israel, Jordan, south- The rachis is usually uniformly tough. Threshing breaks it western Syria, Lebanon, southeastern Turkey, northern Iraq, into irregular fragments. and western Iran (fig. 3). Hulled emmer (T. turgidum subsp. dicoccum) was the prin- At first, molecular studies placed a probable single origin cipal wheat of Old World agriculture in the Neolithic and of emmer wheat in Turkey. O¨ zkan and colleagues (O¨ zkan et early Bronze Age. al. 2002; Salamini et al. 2002) examined 204 AFLP loci of 99 Genetic and morphological evidence clearly indicates (for wild emmer lines from all over the Fertile Crescent belt and Weiss and Zohary Founder Crops S241

Table 4. Representative radiocarbon dates

Site Lab no. 14C dates BP Cal BP (1j) 13C Dated material Period Bone/ovicaprine, burnt PPNB . . . 9790–10,180 100 ע Abu Hureyra OxA-881 8870 Charcoal PPNB . . . 8710–9093 100 ע Abu Hureyra BM-1724R 8020 Organic material/carbon PPN . . . 9440–9600 90 ע Ali Kosh B-122721 8540 Organic material/carbon PPN . . . 8770–9000 50 ע Ali Kosh B-108256 8000 Charcoal/ash Chalcolithic . . . 8770–9090 78 ע Arpachiyah P-585 8064 Charcoal Chalcolithic . . . 7680–7830 60 ע Arpachiyah BM-1531 6930 Charcoal PPNB . . . 10,200–10,480 103 ע Beidha P-1380 9128 Pistacia, nuts PPNB . . . 9440–9630 100 ע Beidha P-1379 8546 Charcoal EPPNB . . . 10,650–11,200 190 ע Cafer Ho¨yu¨k Ly-4436 9560 Charcoal EPPNB . . . 9920–10,220 80 ע Cafer Ho¨yu¨k Ly-4437 8950 Charcoal/juniper Early Chalcolithic . . . 8580–8700 36 ע Can Hasan I AA-41170 7853 Charcoal Early Chalcolithic . . . 7020–7270 78 ע Can Hasan I P-793 6254 Charcoal Aceramic Neolithic . . . 9490–9610 65 ע Can Hasan III HU-11 8584 Charcoal Aceramic Neolithic . . . 8580–8780 70 ע Can Hasan III HU-9 7874 Charcoal EPPNB . . . 10,420–10,590 55 ע Cayo¨nu¨ GrN-6243 9320 Charcoal EPPNB . . . 9930–10,240 80 ע Cayo¨nu¨ GrN-6244 8980 Charcoal Neolithic/Samarra . . . 7560–7930 182 ע Choga Mami BM-483 6846 Charcoal PPNA . . . 11,230–11,470 70 ע Gilgal Pta-4588 9920 Charcoal PPNA . . . 11,080–11,230 70 ע Gilgal Pta-4585 9710 Charcoal PPNB . . . 9910–10,230 103 ע Jericho P-382 8956 Charcoal PPNB . . . 9480–9550 64 ע Jericho BM-1320 8539 ND Cypro-EPPNB . . . 10,410–10,650 60 ע Kissonerga-Mylouthkia OxA-7460 9315 ND Cypro-EPPNB . . . 10,200–10,380 70 ע Kissonerga-Mylouthkia AA-33129 9110 Charcoal PPNA . . . 10,580–10,880 122 ע Mureybit P-1224 9492 Charcoal PPNA . . . 12,390–12,800 140 ע Mureybit Lv-607 10,590 Charcoal PPNA . . . 11,090–11,320 90 ע Netiv Hagdud Pta-4557 9780 Charcoal PPNA . . . 11,060–11,200 70 ע Netiv Hagdud Pta-4556 9660 Ϫ20.60 Charcoal Masraqan 24,600–25,500 290 ע Ohalo II RT-1624 20,840 Ϫ22.70 Charcoal Masraqan 20,350–20,950 200 ע Ohalo II RT-1297 17,500 ND Cypro-EPPNB . . . 10,400–10,660 80 ע Parekklisha-Shillourokambos Ly-290 9310 ND Cypro-EPPNB . . . 9530–9740 80 ע Parekklisha-Shillourokambos Ly-930 8670 Charcoal PPNB . . . 9030–9270 50 ע Ramad GrN-4426 8210 Charcoal PPNB . . . 8980–9130 50 ע Ramad GrN-4821 8090 Rizokarpaso-Cape-Andreas- ND Cypro-PPNB . . . 8410–8720 125 ע Kastros MC-805 7775 Rizokarpaso-Cape-Andreas- ND Cypro-PPNB . . . 8160–8390 120 ע Kastros MC-807 7450 Charcoal Obeid . . . 8050–8420 160 ע Tepe Sabz I-1501 7460 Charcoal Obeid . . . 5990–6400 160 ע Tepe Sabz I-1500 5410 Horsebean seeds MPPNB . . . 10,190–10,390 80 ע Yiftahel KN 3571 9100 Horsebean seeds MPPNB . . . 9430–9740 130 ע Yiftahel RT 736A 8570 Note. PPN p Pre-Pottery Neolithic; PPNA p Pre-Pottery Neolithic A; PPNB p Pre-Pottery Neolithic B; EPPNB p Early Pre-Pottery Neolithic B; MPPNB p Middle Pre-Pottery Neolithic B; ND p unknown. several dozen different cultivars. They concluded that the 9900 cal BP, C¸ayo¨nu¨ (van Zeist and de Roller 1991–1992, most likely place of origin for emmer wheat is the Karacadag˘ 2003). Here, hundreds of spikelet forks were discovered, all Mountains in southeastern Turkey. There is, however, still a showing rough breakage scars characteristic of domestic em- debate regarding the validity of these results and the number mer. Numerous spikelet forks, with similar telltale rough scars, of possible domestication events as attested by recent molec- are also available from contemporary contexts at Cafer Ho¨yu¨k ular studies (e.g., Brown et al. 2009; Feldman and Kislev (de Moulins 1997). These finds indicate that at EPPNB, em- 2007). As Feldman and Kislev (2007) pointed out recently, mer domestication must have been well under way in the ¨ the data of Mori et al. (2003), Ozkan et al. (2002, 2005), and Fertile Crescent. We wish to have more data before being fully Luo et al. (2007) may indicate a diphyletic or polyphyletic convinced about the domestication status of the EPPNB Cyp- origin of domesticated emmer, possibly in the Karacadag˘ area, riot find from Kissonerga-Mylouthkia and Shillourokambos as well as in the southern Levant. (Murray 2003; Willcox 2000). Archaeobotany. Hulled emmer wheat. The earliest fully con- To date, the earliest conspicuous grain-size increase is re- vincing sign to date of domesticated emmer comes from the ported from EPPNB Tell Aswad, near Damascus, Syria (van numerous spikelet forks retrieved from EPPNB, ca. 10,600– Zeist and Bakker-Heeres 1982 [1985]). Here dicoccum-like S242 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 2. Distribution of wild einkorn wheat Triticum monococcum subsp. baeoticum (pTriticum baeoticum). The area in which wild einkorn is massively spread is shaded. Dots represent additional sites outside the main area harboring mainly weedy forms (Zohary, Hopf, and Weiss 2011). plump kernels start to appear in the lowest habitation level, cereals, and it prevails quantitatively over domesticated barley Ia (ca. 10,500–10,200 cal BP; Willcox 2005), and also in and domesticated einkorn. MPPNB phases II (ca. 10,200–9550 cal BP). In these phases, Free-threshing emmer wheats. Free-threshing wheat forms, however, no rachis segments have been retrieved. Significantly, identifiable by their rachis fragments, make their appearance no wild dicoccoides-like narrow kernels were retrieved from in Southwest Asia soon after the firm establishment of emmer Tell Aswad. As van Zeist and Bakker-Heeres emphasize, the wheat cultivation (for review, see Maier 1996). They are al- continuous presence of morphologically discernible dicoccum ready present among the plant remains of Late Pre-Pottery kernels, the total absence (from the very start) of dicoccoides- Neolithic B (LPPNB), ca. 9600–8600 cal BP, Can Hasan III, like material, and the extreme dryness (less than 200 mm Turkey (Hillman 1972), and of MPPNB, ca. 10,200–9550 cal annual rainfall) suggest that emmer wheat was introduced BP, Tell Aswad (van Zeist and Bakker-Heeres 1982 [1985]) into the Damascus Basin as a domesticated cereal not later and LPPNB, ca. 9450–8600 cal BP, Tell Bouqras (van Zeist than the second half of the eleventh millennium BP. From and Waterbolk-van Rooijen 1985), Syria. They also occur in ca. 10,100–8700 cal BP onward, charred grains that morpho- the MPPNB/LPPNB, ca. 9300–9000 cal BP, C¸ atalho¨yu¨k, Tur- logically conform with dicoccum appear also at Tell Abu key (Fairbairn, Near, and Martinoli 2005; Fairbairn et al. 2002; Hureyra, northeast Syria, again with no rachis segments (Hill- Helbaek 1964), and Tell Ramad, Syria (van Zeist and Bakker- man 1975, 2000; Hillman, Colledge, and Harris 1989), and Heeres 1982 [1985]). in contemporary PPNB Can Hasan III and C¸ atalho¨yu¨k East, Konya plain, Turkey (Fairbairn, Near, and Martinoli 2005; Barley Hordeum vulgare Fairbairn et al. 2002, 2007; Helbaek 1964, 1969; Hillman 1972, 1978), Ali Kosh, Deh Luran Plain, Khuzistan (Helbaek 1969), Biology. Domesticated barley Hordeum vulgare subsp. vul- and Jericho, Israel (Hopf 1983). From the very beginnings of gare is one of the main cereals of the belt of Mediterranean agriculture in Southwest Asia, emmer is one of the principal agriculture and a founder crop of Old World Neolithic food Weiss and Zohary Founder Crops S243

Figure 3. Distribution of the wild tetraploid emmer wheat Triticum tur- gidum subsp. dicoccoides (pTriticum dicoccoides) and Timopheev’s wheat Triticum timopheevi subsp. araraticum (pTriticum araraticum; Zohary, Hopf, and Weiss 2011). production. All over this area barley is a common companion lateral spikelets are reduced and are grainless and awnless. of wheat, but in comparison with the latter, it is regarded as Each ear thus contains only two rows of fertile spikelets. an inferior staple. Yet barley withstands drier and warmer 2. Six-rowed forms, traditionally called Hordeum hexasti- environments, poorer soils, and some salinity. Because of these chum, in which the three spikelets in each triplet bear seed qualities, it has been the principal grain produced in nu- and usually all are awned. Ears in these varieties therefore merous areas and an important element of the human diet. have six rows of fertile spikelets. Barley is also the main cereal used for beer fermentation in Wild barley is spread over the East Mediterranean Basin the Old World. The crop was and still is a most important and the West Asiatic countries (fig. 4), penetrating as far as feed supplement for domestic animals. Turkmenia, Afghanistan, Ladakh, and Tibet. Wild barley oc- Barley is a diploid (2np2xp14 chromosomes) and pre- cupies both primary and segetal human-made habitats. Its dominantly self-pollinated crop. All cultivars have nonbrittle distribution center lies in the Fertile Crescent. In this area, ears, a sharp contrast with wild barleys, in which ears are wild barley is continuously and massively distributed. It con- always brittle. Nonbrittleness in domesticated barley is gov- stitutes an important annual component of open herbaceous erned by a recessive mutation in either one of two tightly formations, and it is particularly common in the summer- linked “brittle” genes (table 3). dry deciduous oak park-forest belt east, north, and west of Barley ears have a unique structure. They contain triplets the Syrian Desert and the Euphrates Basin and on the slopes of spikelets arranged alternately on the rachis. According to facing the Jordan Rift Valley. From here, it spills over the drier the morphology of the spikelets, domestic barley can be di- steppes and semideserts. vided into two principal types. The origins of barley are still not fully understood. Early 1. Two-rowed forms, traditionally called Hordeum dis- crossing experiments and chloroplast DNA (cpDNA) typing tichum, in which only the median spikelet in each triplet is have suggested that barley is of one, two, or at most a very fertile and usually armed with a prominent awn. The two few major domestication events (Zohary 1999). Later, Badr S244 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 4. Distribution of wild barley Hordeum vulgare subsp. spontaneum (pHordeum spontaneum). The area in which wild barley is massively spread is shaded. Dots represent additional sites mainly of weedy forms. Wild barley extends eastward beyond the boundaries of this map as far as Tibet (Zohary, Hopf, and Weiss 2011). et al. (2000) examined 400 AFLP loci in wild and domesticated replaced by nonbrittle broad-seeded barley forms. Hulled bar- lines and found that barley is probably of a monophyletic ley has been found in Cyprus from the EPPNB, ca. 10,650– origin in the Israel-Jordan area. However, in recent studies 9550 cal BP (Murray 2003; Willcox 2000). that included sequencing of seven genetic loci, Morell and Domesticated barley continued to be a principal grain crop Clegg (2007; as well as other molecular studies by Molina- in Southwest Asia throughout the Neolithic period. Its re- Cano et al. [2005], [Saisho and Puruggana 2007], and Wang, mains have been recovered side by side with wheats in most Yu, and Ding [2009]) suggested two origins: one within the Neolithic sites. Shortly afterward, we are faced with more Fertile Crescent and a second farther east, possibly at the advanced forms (i.e., six-rowed hulled as well as naked cul- eastern edge of the Iranian Plateau. Apparently, European and tivars of barley). North African barley is largely connected to the Fertile Cres- In conclusion, the archaeological finds indicate that barley cent, while much of Asian barley is connected to the eastern is a founder crop of the Levantine Neolithic agriculture and center. a close companion of emmer and einkorn wheats. The ar- chaeological remains make it also possible to trace the main Archaeobotany. Unmistakable remains of nonbrittle barley developments of barley under domestication: first the fixation (i.e., forms that could survive only under domestication) of nonbrittle mutations and subsequently the emergence of came from phase II (ca. 10,200–9550 cal BP) in Tell Aswad six-rowed hulled and naked types. (van Zeist and Bakker-Heeres 1982 [1985]) and from ca. 9450 to 9300 cal BP Jarmo, Iraq. In the latter site (Braidwood 1960; Helbaek 1959a, 1960, 1966), Helbaek was the first to show Pulses two-rowed barley remains still closely resembling wild spon- taneum but also displaying a nonbrittle rachis. Similar finds Lentil Lens culinaris were reported and verified by Hopf (1983) in PPN Jericho. Indicative clues come from ca. 9600–8750 cal BP Ali Kosh Biology. Lentil ranks among the oldest and the most- (Helbaek 1969), where the brittle spontaneum-like material appreciated grain legumes of the Old World (Smartt 1990; characterized the lower layers, and in the upper strata it was Zohary 1995). In Mediterranean agriculture it is a character- Weiss and Zohary Founder Crops S245

Figure 5. Distribution of wild lentil Lens culinaris subsp. orientalis (pLens orientalis; Zohary, Hopf, and Weiss 2011). Wild lentil extends eastward beyond the boundaries of this map into north Afghanistan and adjacent Central Asia. istic companion of wheat and barley. Compared with the in seed size. The pod’s indehiscence is governed by a single cereals, yields are relatively low, but lentil stands out as one mutation (table 3), the nondehiscent condition being recessive of the most nutritious and tasty pulses. The protein content to the dehiscent one. is about 25%, and lentil constitutes an important meat sub- The wild progenitor of the domesticated plant, subsp. cu- stitute in peasant communities. linaris, shows close morphological, cytogenetic, and molecular The domesticated crop Lens culinaris (psyn. L. culinaris affinities to wild subsp. orientalis, native to the East Medi- subsp. culinaris; Lens esculenta) manifests a wide range of terranean Basin and Southwest and Central Asia (fig. 5; Fer- morphological variation in both its vegetative parts and its guson et al. 1998, 2000; Ladizinsky 1993; van Oss, Aron, and reproductive parts. Like many other annual grain crops, lentil Ladizinsky 1997; Zohary and Hopf 1973). In fact, subsp. ori- is predominantly self-pollinated, diploid (2np2xp14), and entalis looks like a miniaturized subsp. culinaris but bears pods interfertile, or largely so. that burst open immediately after maturation. Conventionally, lentil cultivars are grouped in two inter- Lens orientalis grows primarily on shallow stony soils and grading clusters of seed sizes: (a) small-seeded lentils (subsp. in gravelly hillsides in open or steppelike habitats. It also microsperma), with small pods and small 3–6-mm seeds, and enters disturbed localities such as stony patches or stone heaps (b) large-seeded lentils (subsp. macrosperma), with larger pods bordering orchards and cereal cultivation. In most parts of and with seeds attaining 6–9 mm in diameter. Macrosperma its distribution, L. orientalis is rather inconspicuous or even forms are to be regarded as more advanced; they start to rare. It usually forms small scattered colonies. However, on appear rather late in archaeological sequences, only in the stony slopes of Mount Hermon, the Anti-Lebanon, the oak third millennium BP. The occurrence of pods, or their frag- park-forest belt of southern Turkey, and the western escarp- ments, is extremely rare. ments of the Zagros range, L. orientalis is occasionally locally As we described in the introduction to this article, the first common at 1,200–1,600 m altitude (Ladizinsky and Abbo sign of domestication is the retention of the seed in the pod 1993). Frequently it grows side by side with bitter vetch (Vicia (pod’s indehiscence) and the second is the gradual increase ervilia). S246 Current Anthropology Volume 52, Supplement 4, October 2011

As argued by Zohary (1999), the rich chromosomal poly- over, it is doubtful whether comparative morphology will pro- morphism found in the wild progenitor compared with the vide us with such clues in the future. Yet once Neolithic chromosomal uniformity in the cultivars suggests that this agriculture is soundly established, cultivation of lentil is part pulse crop was taken into domestication only once or very of it. The available archaeological information on early re- few times. mains of lentil comes from the Levant, the very territory over Ladizinsky (1999) examined cpDNA restriction patterns. which wild L. orientalis is distributed. Using those results and additional information from cross- ability and chromosomal architecture led him to conclude Pea Pisum sativum that Lens originated from the territories around southern Tur- key and north Syria. The same areas were also mentioned in Biology. The pea ranks among the oldest grain legumes of later recombinant DNA studies conducted by Sonnante and the Old World. From its early beginnings, this crop has been coworkers (Sonnante, Galasso, and Pignone 2003; Sonnante, a close companion of wheat and barley domestication (Zohary Hammer, and Pignone 2009). and Hopf 1973). Pea Pisum sativum is well adapted to both warm Mediterranean-type and cool temperate conditions. In Archaeobotany. Lentils seem to be closely associated with peasant communities in Southwest Asia, the Mediterranean the start of wheat and barley domestication in the Levant. Basin, temperate Europe, Ethiopia, and northwestern India, Very possibly this legume was introduced into domestication it constitutes an important source of protein for human con- in this region together with emmer, einkorn, and barley; that sumption. The protein content of the seed is about 22%. is, it should be regarded as a founder crop of Old World Today, pea ranks among the world’s most important pulses Neolithic agriculture (Zohary and Hopf 1973). (Davies 1995; Smartt 1990). At the end of the tenth and in the ninth millennia BP, Pea is a diploid (2np2xp14 chromosomes) and predom- charred seeds of lentil appear in most of the PPNB early inantly self-pollinated crop. As a consequence of the self- farming villages in Southwest Asia. The seeds are still similar pollination system, variation in pea is molded in numerous in size to those of wild forms (2.5–3.0 mm in diameter) and true breeding lines. Domesticated pea shows a wide range of usually do not occur in quantities. Yet they are always asso- morphological variation. As we already described, the first ciated with domesticated wheat and barley. Among the richest sign of domestication is the retention of the seed in the pod sites are ca. 10,200–9550 cal BP Tell Aswad (van Zeist and (pod’s indehiscence), and the second is the gradual increase Bakker-Heeres 1982 [1985]), ca. 10,200–8700 cal BP Tell Abu in seed size, from 3–4 to 6–8 mm in diameter. The pod’s Hureyra (Hillman 1975, 2000, 2001; Hillman, Colledge, and indehiscence is governed by a single mutation (table 3), the Harris 1989), ca. 10,250–9500 cal BP Jericho (Hopf 1983), nondehiscent condition being recessive to the dehiscent one. ca. 10,600–9900 cal BP C¸ayo¨nu¨ (van Zeist 1972; van Zeist A third character here is the reduction of the relatively thick and de Roller 1991–1992, 2003), and ca. 9600–8800 cal BP texture and rough surface of the seed coat, resulting in the Ali Kosh, Iran (Helbaek 1969). breakdown of the germination inhibition of wild peas. A large hoard of carbonized lentils was recovered from the The crop complex of P. sativum contains the variable col- MPPNB, ca. 10,400–9450 cal BP, Yiftah’el, north Israel. The lection of pea cultivars and its closely related wild races. The size of the hoard (ca. 1,400,000 seeds) and its contamination wild forms of P. sativum fall into two main morphological by the fruits of the weed Galium tricornutum, a characteristic types (fig. 6): (i) Pisum sativum subsp. elatius, a tall “maquis weed in lentil cultivation, indicate that there and then lentil type” that thrives as a sporadic climber in maquis formations was already domesticated (Garfinkel, Kislev, and Zohary in the relatively mesic parts of the Mediterranean region and 1988). Large amounts of lentil seeds were discovered also in as a weed, and (ii) Pisum sativum var. pumilio (known for- somewhat later phases of the Neolithic settlements in South- mally as Pisum humile), a shorter, more xeric “steppe type” west Asia: in ca. 9450–9300 cal BP Jarmo, Iraq (Braidwood geographically restricted to southwest Asia. It occurs in the 1960; Helbaek 1959a, 1960, 1966); in ca. 9250–9000 cal BP deciduous oak park-forest belt and in open steppelike her- Tell Ramad, Syria (van Zeist and Bakker-Heeres 1982 [1985]); baceous vegetation formations characteristic of the Fertile in ca. 8200–7800 cal BP ceramic Hacilar (Helbaek 1970); and Crescent (i.e., in the same zone that harbors the wild pro- in ca. 8350–7750 cal BP Tepe Sabz, Deh Luran Valley, Iran genitors of wheat, barley, lentil, and flax). From such primary (Helbaek 1970). The Tepe Sabz lentils had already attained habitats humile peas spill over to secondary habitats and oc- 4.2 mm in diameter. This is an obvious development under casionally infest cereal cultivation. domestication. Lentils are repeatedly encountered in the early The available evidence from the living plants implicates the European and Southwest Asian Pottery Neolithic sites situated wild humile peas as the progenitor stock for pea domestica- far outside the distribution area of L. orientalis, suggesting tion. Humile peas show closer morphological similarities than that these lentils were already domesticated. elatius peas to the domesticated aggregate and grow in steppe- In summary, archaeological remains do not provide us yet like habitats (i.e., under open conditions similar to the cul- with direct diagnostic traits (such as indehiscent pod remains) tivated field). Within humile peas, the Turkish and the south for a determination of the start of lentil domestication. More- Levant forms having chromosomes identical to those present Weiss and Zohary Founder Crops S247

Figure 6. Distribution of the two main wild races of pea Pisum sativum: (i) “steppe-type” humile forms and (ii) “maquis-type” elatius forms (Zo- hary, Hopf, and Weiss 2011). In the West Mediterranean Basin, wild elatius peas extend beyond the boundaries of the map and reach as far as Spain. in the cultivars should be regarded as the primary ancestral later Neolithic phases in the Levant. Large quantities of car- stock. This is also supported by cpDNA comparisons. Yet it bonized seed accompany the domesticated wheats and barley is very likely that the two subspecies contributed some genes in MPPNB/LPPNB C¸atalho¨yu¨k (Fairbairn, Near, and Mar- to the cultivated ensemble through occasional secondary hy- tinoli 2005; Fairbairn et al. 2002; Helbaek 1964) and Final bridization (Ben Ze’ev and Zohary 1973; Palmer, Jorgensen, PPNB/PPNC Erbaba (van Zeist and Buitenhuis 1983). and Thompson 1985). In contrast to those of the wheats and barley, the earliest Nasiri and coworkers (Nasiri, Haghnazari, and Saba 2009) archaeological remains of pea do not provide us with simple have examined SSR (microsatellite) markers of wild peas and traits for a foolproof recognition of domestication (Zohary cultivars. They have found that these markers effectively dif- and Hopf 1973). In peas under domestication there is a gen- ferentiate between the cultivars and the wild accessions. In eral trend toward an increase in the size of the seed and the addition, Pisum sativum subsp. fulvum was found to be the length of the hilum, but such changes occurred gradually in closest relative of the cultivars, but it should be noted that the course of domestication. In early finds there is a consid- subsp. humile was not tested in this study. erable overlapping in the dimensions of wild and domesti- Archaeobotany. Remains of peas are present in many of the cated forms. Perhaps the most reliable indication of domes- PPNB farming villages that developed in the Fertile Crescent tication in peas is provided by the surface of the seed coat. arc from 10,500 cal BP years onward. Some of the earliest Wild peas are characterized by a rough or granular surface, finds were retrieved from ca. 10,200–9550 cal BP Tell Aswad while domesticated varieties have smooth seed coats. How- in south Syria (van Zeist and Bakker-Heeres 1982 [1985]), ever, seed coats survive only very rarely, and if they do not, ca. 10,600–9900 cal BP C¸ayo¨nu¨ in southeast Turkey (van Zeist it is impossible to ascertain whether the material retrieved and de Roller 1991–1992, 2003), ca. 10,250–9500 cal BP PPNB represents wild or domesticated forms. The lower levels (ca. Jericho (Hopf 1983), and ‘Ain Ghazal, Jordan (Rollefson et 10,600–9900 cal BP) of C¸ayo¨nu¨ (van Zeist and de Roller 1991– al. 1985). Much richer remains were available from somewhat 1992) retained some fragments of rough-surface wild seed S248 Current Anthropology Volume 52, Supplement 4, October 2011 coats. Wild-type seed coats occur even much later in Pottery remains are the only material recovered to date in archaeo- Neolithic Hacilar (Helbaek 1970). Significantly, the remains logical excavations. from MPPNB/LPPNB C¸atalho¨yu¨k (Helbaek 1964) and The domesticated chickpea C. arietinum is a member of a Bouqras (van Zeist and Waterbolk-van Rooijen 1985) and leguminous genus comprising some 40 species centered in those from LPPNB C¸ayo¨nu¨ and Can Hasan I (Renfrew 1968) Central and Western Asia (Cole, Maxted, and van der Maesen include a single seed with smooth seed coat characteristic of 1998; van der Maesen 1972). The domesticated pulse shows domesticated varieties. This smoothness suggests that do- close morphological affinities and an almost identical seed mestication of peas in the Levant is as old, or almost as old, protein profile to two wild species of chickpea: Cicer echi- as the domestication of wheat and barley. nospermum and Cicer reticulatum. These two wild chickpeas Although it is currently less definite, the archaeological evi- are diploid self-pollinated annuals known only from south- dence establishes pea as one of the founder crops of the Le- eastern Turkey. The available morphological and cytogenetic vantine Neolithic agriculture. Since this early start, pea seems evidence implicates C. reticulatum as the wild ancestor of the to be a consistent and common element of food production domesticated plant (Ladizinsky and Adler 1976), and it is and a common companion of wheats and barley. The evidence therefore referred to as C. arietinum subsp. reticulatum. Be- from the living plants complements the archaeological finds. cause the distribution of the wild progenitor is restricted to The wild humile forms, with chromosomes and cpDNA iden- southeast Turkey (fig. 7), the area of origin of the domesti- tical to those prevailing in the cultivated crop, should be cated crop can be outlined there. regarded as the closest wild relatives from which this pulse crop evolved. Archaeobotany. Like lentil and pea, chickpea seems to be closely associated with the start of food production in the Chickpea Cicer arietinum Levant, but unlike them it is much rarer in Neolithic contexts. Biology. Chickpea is a valued grain legume of the traditional Earlier chickpea seeds were found in EPPNB, ca. 10,600– agriculture in the Mediterranean Basin and Western Asia as 10,250 cal BP, Tell el-Kerkh, northwest Syria (Tanno and Will- well as India and Ethiopia. It is a member of the grain en- cox 2006). A few of these seeds are somewhat larger and semble found in Levantine Neolithic and Bronze Age remains. plumper than the wild progenitor, and the authors argue Chickpea is adapted to a subtropical or Mediterranean-type (Tanno and Willcox 2006:200) that this might indicate that climate; it grows almost exclusively in the postrainy season they are an “intermediate stage” between wild and domesti- on moisture stored in the soil. Like lentil and pea, chickpea cated chickpeas. Some charred chickpea seeds were recovered (with a seed protein content of some 20%) constitutes an from EPPNB, ca. 10,600–9900 cal BP, C¸ayo¨nu¨ (van Zeist 1972; important meat substitute in peasant communities. van Zeist and de Roller 1991–1992). A few more were found Domesticated chickpea Cicer arietinum, like all other eight in the MPPNB level of Tell Abu Hureyra, northern Syria founder crops, is a predominantly self-pollinated annual crop (Hillman 1975, 2000), and As¸ıklı Ho¨yu¨k (van Zeist and de with pods containing one to two seeds. The cultivars show a Roller 1995) in Turkey. The seeds from these sites correspond wide range of variation in size, color, and shape of the seed in size to those of C. reticulatum. Because these sites are and in the size and form of leaves and flowers. All domes- situated within (or close to) the very restricted geographic ticated varieties are diploid (2np2xp16 chromosomes) and distribution area of the wild progenitor, it is difficult to be interfertile. Chickpea landraces are grouped into two inter- sure whether they represent wild or domestic plants. The seeds connected clusters (Smartt 1990; Smithson, Thompson, and retrieved from MPPNB Jericho (Hopf 1983) and ‘Ain Ghazal Summerfield 1985). Large-seeded varieties (known as “Ka- (Rollefson et al. 1985) and those from LPPNB Ramad near buli” type) with relatively smooth, rounded, light-colored seed coats and pale cream flowers predominate in the Mediter- Damascus (van Zeist and Bakker-Heeres 1982 [1985]) very ranean countries and Southwest Asia. Varieties producing probably represent domesticated forms. The latter sites lie far small wrinkled seeds (“Desi” type) with dark-colored seed away from the territory of the wild progenitor, and the spec- coats and usually purple flowers prevail in the eastern and imens from Jericho seem to have smooth seed coats. southern parts of the distribution area of the crop (i.e., in The evidence from the living plants and the plant remains India, Afghanistan, and Ethiopia). discovered in archaeological excavations indicate that C. As in most other seed legumes, the conspicuous features arietinum belongs to the early Neolithic grain-crop assemblage of evolution under domestication in chickpea are the reten- of the Levant. Archaeological data are still limited and less tion of seeds in the pods (pod’s indehiscence) and the gradual definite, but in this legume, the delimitation of the place of increase in seed size, from 3.5 to 6.0 mm and more. The pod’s origin is relatively simple: the wild progenitor of the domes- indehiscence is governed by a single mutation (table 3). An- ticated chickpea is endemic to the central part of Fertile Cres- other change under domestication is the development of a cent. Here, very likely, this pulse was first brought into do- smooth seed coat and the reduction of its thickness. Seed mestication. Weiss and Zohary Founder Crops S249

Figure 7. Distribution of wild chickpea Cicer arietinum subsp. reticulatum (pCicer reticulatum; Zohary, Hopf, and Weiss 2011).

Fiber and Oil Plants Eurasia (Durrant 1976). In antiquity, flax fibers were the prin- cipal vegetable fiber used for weaving textiles in Europe and Western Asia. The seed contains about 40% oil, and in peasant Flax Linum usitatissimum communities linseed was used as a source for edible oil and Biology. Flax Linum usitatissimum is an annual crop with high-grade lighting oil. characteristic slender strong stems and rounded capsules that The fibers for spinning are obtained from the tall stems, in domesticated forms do not dehisce but retain the oval which are harvested before the maturation of the seed. Tra- compressed shining seed. The crop is diploid (2np2xp30 ditionally, they were first dried and then immersed (wetted) chromosomes) and predominantly self-pollinated. Conse- in water to allow the microbial decomposition (retting) of quently, variation has been molded in the form of numerous the pectin connecting the fibers with other cells and tissues true breeding lines and aggregates of landraces. Two special- of the stem. After retting, the stems were dried, and the fibers izations are apparent: (i) oil varieties, which are relatively (averaging 4 cm in length) were separated by pounding short (30–70 cm) and branched and usually bear large seeds, (breaking) and combing. and (ii) fiber varieties, which are taller and sparsely branched Domesticated flax L. usitatissimum is most closely related and usually produce small seeds. Transitional forms, culti- to wild Linum bienne (syn. Linum angustifolium). These two vated for both oil and fiber, occur as well. flaxes have the same chromosome number (see above), in- Flax was a principal oil and fiber source in the Old World tercross readily, and are fully interfertile (Gill and Yermanos and probably the earliest domesticated plant used for textiles. 1967). Linum bienne—with its characteristic strong branches, Until recently, flax was extensively cultivated in vast areas of blue flowers, and dehiscent capsules—is widely distributed S250 Current Anthropology Volume 52, Supplement 4, October 2011 over West Europe, the Mediterranean Basin, North Africa, 9900 cal BP Tell Mureybit (van Zeist and Casparie 1968). Southwest Asia, Iran, and Caucasia (fig. 8). Some wild forms Soon after, seeds of flax were found in many of the PPNB are annual, and others are biennial or perennial; all are pre- farming villages that appeared in the Fertile Crescent from dominantly self-pollinated. Linum bienne grows mainly in wet 10,500 cal BP onward (fig. 1). Some of the earlier finds come places such as moist grassy areas, springs, seepage areas on from ca. 10,600–9900 cal BP C¸ayo¨nu¨, Turkey (van Zeist 1972; rocky slopes, moist clay soils, and marshy lands. On the basis van Zeist and de Roller 1991–1992, 2003); ca. 10,200–9550 of its close morphological and cytogenetic affinities to the cal BP Tell Aswad, near Damascus, Syria (van Zeist and domesticated crop, L. bienne is identified as the wild progen- Bakker-Heeres 1982 [1985]); Ali Kosh, Iran (Helbaek 1969); itor of L. usitatissimum (Diederichsen and Hammer 1995). Jericho, Israel (Hopf 1983); and ‘Ain Ghazal, Jordan (Rol- The main changes under domestication are the shift to non- lefson et al. 1985). The seeds are still small, similar in size to splitting capsules and the increase of seed size (like in many those of wild bienne forms, yet they are almost always asso- grain crops) as well as the selection for higher oil yield or ciated with domesticated wheats and barley. longer stems with a high amount of long fibers. Fragments of a capsule from ca. 10,250–9500 cal BP Phylogenetic evidence from modern accessions (Allaby et MPPNB Jericho (Hopf 1983) is probably the earliest indi- al. 2005; Fu and Allaby 2010) gives indications that (i) there cation we have today of domesticated flax. Another indication was a single domestication event for flax and (ii) the oil- of early flax domestication comes from linseed remains re- producing variety was domesticated, suggesting that domes- covered from LPPNB, ca. 9250–9000 cal BP, levels of Tell tication selected for the larger oil-rich seeds rather than its Ramad, Syria. The calculated size of the seed from this site, fibers. corrected for charring shrinkage, ranges from 3.2 to 4.1 mm Archaeobotany. Flax was apparently used by humans already in length. This is already within the size class of the L. usi- before its domestication. Recently, twisted and dyed flax fibers tatissimum seed, the lower limit of which lies at 3.0 mm. It were reported in Upper Paleolithic Dzudzuana Cave, Georgia is therefore an attractive indication for flax domestication (ca. 30,000 years old; Bergfjord et al. 2010; Kvavadze et al. under rain-dependent conditions before ca. 8600 cal BP (van 2009, 2010). The oldest wild linseed remains retrieved from Zeist and Bakker-Heeres 1975). Such domesticated linseeds archaeological sites in Southwest Asia come from ca. 10,900– were found in Pottery Neolithic Nahal Zehora, Mount Car-

Figure 8. Distribution of wild flax Linum usitatissimum subsp. bienne (pLinum bienne; Zohary, Hopf, and Weiss 2011). Weiss and Zohary Founder Crops S251 mel, Israel, ca. 8150–7850 cal BP (Kislev and Hartmann, The earliest definite domestic forms of barley and lentil forthcoming). In the Mesopotamian Basin, linseed sizes in ca. first appear during the MPPNB. In this period, domesticated 8400–6000 cal BP Tell Sabz, Iran (Helbaek 1969), and in barley is present in Aswad (van Zeist and Bakker-Heeres 1982 Halafian Arpachiya (Helbaek 1959b) are even bigger (4.7–4.8 [1985]) and later on (still in the MPPNB) in Jarmo (Helbaek mm long). As argued by Helbaek, these large seeds indicate 1969). Domesticated lentil appears in MPPNB Yiftah’el (Gar- advanced domestication and demonstrate that flax was part finkel, Kislev, and Zohary 1988). of the irrigated grain agriculture system that evolved in this The earliest domesticated flax was retrieved from MPPNB region. In Choga Mami, Iraq, Helbaek (1972) reports a rare Jericho (Hopf 1983) and later in LPPNB Tell Ramad (van find in the earlier, ca. 7950–7550 cal BP, stratum, which be- Zeist and Bakker-Heeres 1982 [1985]) and Nahal Hemar, Is- came frequent during the following one, dated to the second rael (Schick 1988). half of the eighth millennium BP. Linseed also appears in The evidence regarding the remaining three founder several later Southwest Asian Neolithic and Bronze Age sites. crops—chickpea, pea, and bitter vetch—is as yet inconclusive. Remains of flax textiles also make an early appearance. The The geographical boundaries of the first group of domesti- best examples come from the drier parts of Southwest Asia, cated plants therefore focus on a rather restricted part of the where because of low humidity, woven material survived with- Fertile Crescent (i.e., from southeastern Turkey in the north out carbonization. Pieces of exquisitely woven linen were dis- to Israel in the south). covered among PPNB (beginning of the ninth millennium We see now (and see Zeder 2011 for details and references) BP) remains in Nahal Hemar cave near the south tip of the the first wave of animal domestication (all herbivores: goat, Dead Sea, Israel (Schick 1988). A single piece of linen was sheep, cow, and pig) more or less at the same time as early found in Neolithic (eighth millennium BP) Fayum, Egypt plant domestication. It is possible, therefore, that the remains (Caton-Thompson and Gardner 1934). of food processing for human food—stalks and inflores- The available archaeological evidence clearly suggests that cences—were used for animal feed of the emerging Neolithic flax belongs to the first group of grain crops that started culture. agriculture in the Levant during the MPPNB. The gradual At present, no agreement prevails among researchers and increase in seed size and the use of linen indicate that flax between research disciplines regarding the mode of origin of domestication was very probably already practiced in the the “first wave” of the Southwest Asian grain crops (einkorn PPNB Levant as attested first by MPPNB capsules and then wheat, emmer wheat, barley, lentil, pea, chickpea, bitter vetch, by LPPNB seeds and linen textile. The evidence from the living and flax). Was it of monophyletic origin (derived from the plants shows that wild bienne forms are widespread in the arc same ancestral taxon representing a single event of domes- and fully supports a Fertile Crescent domestication. tication) or of polyphyletic derivation (resulting from several ancestral taxa representing several independent events)? Conclusions The accumulated morphological, anatomical, and cytoge- netic information seems to support monophyletic speciation This article deals with current biological and archaeobotanical as the plausible mode of origin (Zohary 1996, 1999). Im- information from Southwest Asia (fig. 1) in an attempt to portant in domestication is the interaction between annaulity, identify the earliest finds of domesticated grain crops. If we reproductive biology, and self-pollination and selection; self- adopt the criteria suggested in the beginning of this article, pollination is significant because it causes the plant com- we can now separate the appearance of the eight founder munity to be built of almost only pure inbred lines. The crops into several distinguishable groups and dates. combination of selfing and selection of inbred lines further From the group of eight Southwest Asian founder crops, supports this notion because it causes the selection toward two crops seem to appear earlier as cultivated wild plants. domestic types to be efficient and fast. These “pioneer crops” are (i) wild barley, found in PPNA As far as the molecular research goes, however, there are Gilgal, and (ii) wild lentil, found in PPNA Jerf el-Ahmar and currently two views based on analytical as well as methodo- in PPNA Netiv Hagdud (Weiss, Kislev, and Hartmann 2006). logical perspectives. Some (e.g., Badr et al. 2000; Heun et al. Some lines of evidence suggest that cultivation of wild ein- 1997; O¨ zkan et al. 2002) support a monophyletic origin, while korn, wild emmer, wild barley, wild rye, and wild lentil might others (e.g., Kilian et al. 2007; Molina-Cano et al. 2005; Mor- have been practiced at PPNA Jerf el Ahmar and Dja’de, north- rell and Clegg 2007) support a polyphyletic one. It seems that ern Syria (Willcox, Fornite, and Herveux 2008). this major research question will get more attention in the From the group of founder crops, the earliest definite do- near future. No doubt the research avenue of molecular anal- mesticated plants are einkorn wheat and emmer wheat from ysis of domesticated crops and their origin is a fast-growing two EPPNB sites in Turkey, Cafer Ho¨yu¨k (de Moulins 1997) field that has not yet reached its climax. and C¸ayo¨nu¨ (van Zeist 1972; van Zeist and de Roller 1991– Although it is beyond the scope of this article, it is im- 1992, 2003). Unfortunately, study of the third possible site in portant to mention the relatively new data available from this cluster, Nevali C¸ori (Pasternak 1998), was published only Cyprus (e.g., Peltenburg et al. 2000, 2001; Vigne 2011; Vigne as a preliminary publication and therefore remains uncertain. et al. 2000; Willcox 2000). It is apparent that already during S252 Current Anthropology Volume 52, Supplement 4, October 2011 the EPPNB, human pioneers crossed the sea and settled the plant seed storage at Neolithic C¸atalho¨yu¨k East, Turkey. Vegetation island, bringing with them the Neolithic culture. These finds History and Archaeobotany 16:467–479. Fairbairn, A., J. Near, and D. Martinoli. 2005. Macrobotanical in- hint that the beginning of agriculture on the mainland, as vestigations of the north, south and KOPAL areas at C¸atalho¨yu¨k. attested in this article, was transferred almost immediately In Inhabiting C¸atalho¨yu¨k: reports from the 1995–1999 seasons.I. (on an archaeological timescale) outside the core area. Cyprus, Hodder, ed. Pp. 137–201. Cambridge/Ankara: McDonald Institute as an isolated phenomenon, teaches us the pace of the bio- for Archaeological Research/British Institute of Archaeology at An- logical and cultural shifts from wild to domesticated plants kara. Feldman, M., and M. E. Kislev. 2007. Domestication of emmer wheat and from hunter-gatherers to farmers. and evolution of free-threshing tetraploid wheat. Israel Journal of Plant Sciences 55:207–221. Feldman, M., F. G. H. Lupton, and T. E. Miller. 1995. Wheats. In Evolution of crop plants. 2nd edition. J. Smartt and N. W. Sim- Acknowledgments monds, eds. Pp. 184–192. Harlow: Longman Scientific & Technical. Ferguson, M. E., B. V. For-Lloyd, L. D. Robertson, N. Maxted, and This article is partially based on materials from D. Zohary, H. J. Nenbury. 1998. Mapping the geographical distribution of M. Hopf, and E. Weiss, Domestication of Plants in the Old genetic variation in the genus Lens for the enhanced conservation World, 4th edition, forthcoming from Oxford University of plant genetic diversity. Molecular Ecology 7:1743–1755. Ferguson, M. E., N. Maxted, M. van Slageren, and L. D. Robertson. Press. 2000. A re-assessment of the taxonomy of Lens Mill. (Leguminosae, Papilionoideae, Vicieae). Botanical Journal of the Linnean Society References Cited 133:41–59. Fu, Y.-B., and R. Allaby. 2010. Phylogenetic network of Linum species Allaby, R. G., G. W. Peterson, D. A. Merriwether, and Y.-B. Fu. 2005. as revealed by non-coding chloroplast DNA sequences. Genetic Evidence of the domestication history of flax (Linum usitatissimum Resources and Crop Evolution 57:667–677. L.) from genetic diversity of the sad2 locus. Theoretical and Applied Fuller, D. Q. 2007. Contrasting patterns in crop domestication and Genetics 112:58–65. domestication rates: recent archaeobotanical insights from the Old Badr, A., K. Mu¨ller, R. Scha¨fer-Pregl, H. E. L. Rabey, S. Effgen, H. World. Annals of Botany 100:903–924. H. Ibrahim, C. Pozzi, W. Rohde, and F. Salamini. 2000. On the Garfinkel, Y., M. Kislev, and D. Zohary. 1988. Lentil in the Pre-Pottery origin and domestication history of barley (Hordeum vulgare). Neolithic B Yiftah’el: additional evidence of its early domestication. Molecular Biology and Evolution 17:499–510. Israel Journal of Botany 37:49–51. Ben Ze’ev, N., and D. Zohary. 1973. Species relationships in the genus Gill, K. S., and D. M. Yermanos. 1967. Cytogenetic studies on the Pisum L. Israel Journal of Botany 22:73–91. genus Linum. I. Hybrids among the taxa with 15 as the haploid Bergfjord, C., S. Karg, A. Rast-Eicher, M. L. Nosch, U. Mannering, chromosome number. Crop Science 7:27–31. R. G. Allaby, B. M. Murphy, and B. Holst. 2010. Comment on Goring-Morris, A. N., and A. Belfer Cohen. 2011. Neolithization “30,000-year-old wild flax fibers.” Science 328:1634-b. processes in the Levant: the outer envelope. Current Anthropology Bo¨hner, U., and D. Schyle. 2002–2006. Radiocarbon CONTEXT da- 52(suppl. 4):S195–S208. tabase. http://context-database.uni-koeln.de/index.php. Harlan, J. R. 1981. The early history of wheat: earliest traces to the Braidwood, R. J. 1960. The agricultural revolution. Scientific American sack of Rome. In Wheat science: today and tomorrow. L. T. Evans 203:130–148. and W. J. Peacock, eds. Pp. 1–19. Cambridge: Cambridge Uni- Brown, T. A., M. K. Jones, W. Powell, and R. G. Allaby. 2009. The versity Press. complex origins of domesticated crops in the Fertile Crescent. Harlan, J. R., and D. Zohary. 1966. Distribution of wild wheats and Trends in Ecology & Evolution 24:103–109. barley. Science 153:1074–1080. Caton-Thompson, G., and E. W. Gardner. 1934. The desert Fayum. Helbaek, H. 1959a. Domestication of food plants in the Old World. London: Royal Anthropological Institute of Great Britain and Ire- Science 130:365–372. land. ———. 1959b. Notes on the evolution and history of Linum. Kuml Cole, S., N. Maxted, and L. J. G. van der Maesen. 1998. Identification 103–129. aids for Cicer (Leguminosae, Cicereae) taxa. Edinburgh Journal of ———. 1960. The paleobotany of the Near East and Europe. In Botany 55:243–265. Prehistoric investigations in Iraqi Kurdistan. R. J. Braidwood and Davies, D. R. 1995. Peas. In Evolution of crop plants. 2nd edition. J. B. Howe, eds. Pp. 99–118. Studies in Ancient Oriental Civilizations, Smartt and N. W. Simmonds, eds. Pp. 294–296. Harlow: Longman no. 31. Chicago: University of Chicago Press. Scientific & Technical. ———. 1964. First impressions of the C¸atal Hu¨yu¨k plant husbandry. de Moulins, D. 1997. Agricultural changes at Euphrates and steppe Anatolian Studies 14:121–123. sites in the mid-8th to 6th millennium BC. BAR International Series ———. 1966. Commentary on the phylogenesis of Triticum and 683. Oxford: Archaeopress. Hordeum. Economic Botany 20:350–660. Diederichsen, A., and K. Hammer. 1995. Variation of cultivated flax ———. 1969. Plant collecting, dry-farming, and irrigation agricul- (Linum usitatissimum L. subsp. usitatissimum) and its wild pro- ture in prehistoric Deh Luran. In Prehistory and human ecology of genitor pale flax (subsp. angustifolium (Huds.) Thell.). Genetic Re- the Deh Luran Plain: an early village sequence from Khuzistan, Iran. sources and Crop Evolution 42:263–272. F. Hole, K. V. Flannery, and J. A. Neely, eds. Pp. 383–426. Memoirs Durrant, A. 1976. Flax and linseed. In Evolution of crop plants.N. of the Museum of Anthropology, no. 1. Ann Arbor: University of W. Simmonds, ed. Pp. 190–193. London: Longman. Michigan. Fairbairn, A., E. Asouti, J. Near, and D. Martinoli. 2002. Macro- ———. 1970. The plant husbandry of Hacilar. In Excavations at botanical evidence for plant use at Neolithic C¸atalho¨yu¨k, south- Hacilar, vol. 1. J. Mellaart, ed. Pp. 189–244. Edinburgh: Edinburgh central Anatolia, Turkey. Vegetation History and Archaeobotany 11: University Press. 41–54. ———. 1972. Samarran irrigation agriculture at Choga Mami in Fairbairn, A., D. Martinoli, A. Butler, and G. Hillman. 2007. Wild Iraq. Iraq 34:35–48. Weiss and Zohary Founder Crops S253

Heun, M., S. Haldorsen, and K. Vollan. 2008. Reassessing domesti- Love, H. H., and W. T. Craig. 1924. The genetic relation between cation events in the Near East: Einkorn and Triticum urartu. Ge- Triticum dicoccum dicoccoides and a similar morphological type nome 51(6):444–451. produced synthetically. Journal of Agriculture Research 28:515–519. Heun, M., R. Scha¨fer-Pregl, D. Klawan, R. Castagna, M. Acerbi, B. Luo, M.-C., Z.-L. Yang, F. M. You, T. Kawahara, J. G. Waines, and Borghi, and F. Salamini. 1997. Site of einkorn wheat domestication J. Dvorˇa´k. 2007. The structure of wild and domesticated emmer identified by DNA fingerprinting. Science 278:1312–1314. wheat populations, gene flow between them, and the site of emmer Hillman, G. C. 1972. Plant remains. In D. H. French, G. C. Hillman, domestication. Theoretical and Applied Genetics 114:947–959. S. Payne, and R. J. Payne. Report on excavations at Can Hasan Maier, U. 1996. Morphological studies of free-threshing wheat ears III 1969–1970. In Papers in economic prehistory. E. S. Higgs, ed. from a Neolithic site in southwest Germany, and the history of Pp. 182–188. Cambridge: Cambridge University Press. naked wheats. Vegetation History and Archaeobotany 5:39–55. ———. 1975. The plant remains from Tell Abu Hureyra: a prelim- Molina-Cano, J. L., J. R. Russell, M. A. Moralejo, J. L. Escacena, G. inary report. Proceedings of the Prehistoric Society 41:70–73. Arias, and W. Powell. 2005. Chloroplast DNA microsatellite anal- ———. 1978. On the origins of domestic rye Secale cereale: the finds ysis supports a polyphyletic origin for barley. Theoretical and Ap- from aceramic Can Hasan III in Turkey. Anatolian Studies 28:157– plied Genetics 110:613–619. 174. Mori, N., T. Ishii, T. Ishido, S. Hirosawa, H. Watatani, T. Kawahara, ———. 2000. Abu Hureyra 1: the Epipalaeolithic. In Village on the M. Nesbitt, et al. 2003. Origins of domesticated emmer and com- Euphrates: from foraging to farming at Abu Hureyra.A.M.T. mon wheat inferred from chloroplast DNA fingerprinting. Pro- Moore, G. C. Hillman, and A. J. Legge, eds. Pp. 327–399. Oxford: ceedings of the 10th International Wheat Genetics Symposium, Paes- Oxford University Press. tum, Italy. Pp. 25–28. Rome: Instituto Sperimentale per la ———. 2001. Archaeology, Percival, and the problems of identifying Cerealicolutra. wheat remains. In Wheat taxonomy: the legacy of John Percival.P. Morrell, P. L., and M. T. Clegg. 2007. Genetic evidence for a second D. S. Caligari and P. E. Brandham, eds. Pp. 27–36. London: Lin- domestication of barley (Hordeum vulgare) east of the Fertile Cres- nean Society. cent. Proceedings of the National Academy of Sciences of the USA Hillman, G. C., S. M. Colledge, and D. R. Harris. 1989. Plant-food 104:3289–3294. economy during the Epipalaeolithic period at Tell Abu Hureyra, Murray, M. A. 2003. Archaeobotanical remains from the Aceramic Syria: dietary diversity, seasonality, and modes of exploitation. In Neolithic wells (ca. 9000 BP). In The colonisation and settlement Foraging and farming: the evolution of plant exploitation.D.R. of Cyprus: investigations at Kissonerga-Mylouthkia 1977–1996.E. Harris and G. H. Hillman, eds. Pp. 240–268. London: Unwin Peltenburg, ed. Pp. 59–72. Lemba Archaeological Project Cyprus, Hyman. vol. 3.1. Studies in Mediterranean Archaeology, vol. 70:4. Sa¨ve- Hopf, M. 1983. Jericho plant remains. In Excavations at Jericho, vol. dalen, Sweden: A˚ stro¨ms. 5. K. M. Kenyon and T. A. Holland, eds. Pp. 576–621. London: Nasiri, J., A. Haghnazari, and J. Saba. 2009. Genetic diversity among British School of Archaeology in Jerusalem. varieties and wild species accessions of pea (Pisum sativum L.) Kazan, K., F. J. Muehlabuer, N. W. Weeden, and G. Ladizinsky. 1993. based on SSR markers. African Journal of Biotechnology 8:3405– Inheritance and linkage relationships of morphological and iso- 3417. zyme loci in chickpea (Cicer arietinum L.). Theoretical and Applied Nesbitt, M. 2002. When and where did domesticated cereals first Genetics 86:417–426. occur in Southwest Asia? In The dawn of farming in the Near East. Kilian, B., H. Ozkan, A. Walther, J. Kohl, T. Dagan, F. Salamini, and R. T. J. Cappers and S. Bottema, eds. Pp. 113–132. Studies in Early W. Martin. 2007. Molecular diversity at 18 loci in 321 wild and Near Eastern Production, Subsistence, and Environment, no. 6. 92 domesticate lines reveal no reduction of nucleotide diversity Berlin: Ex Oriente. during Triticum monococcum (Einkorn) domestication: implica- O¨ zkan, H., A. Brandolini, C. Pozzi, S. Effgen, J. Wunder, and F. tions for the origin of agriculture. Molecular Biology and Evolution Salamini. 2005. A reconsideration of the domestication geography 24:2657–2668. of tetraploid wheats. Theoretical and Applied Genetics 110:1052– Kislev, M. E., and A. Hartmann. Forthcoming. Food crops from 1060. Nahal Zehora II. In Village communities of the Pottery Neolithic O¨ zkan, H., A. Brandolini, R. Scha¨fer-Pregl, and F. Salamini. 2002. period in the Menashe Hills, Israel: archaeological investigations at AFLP analysis of a collection of tetraploid wheats indicates the the sites of Nahal Zehora. A. Gopher, ed. Sonia and Marco Nadler origin of emmer and hard wheat domestication in southeast Tur- Institute of Archaeology Monograph Series. Tel Aviv: Tel Aviv Uni- key. Molecular Biology and Evolution 19:1797–1801. versity. Palmer, J. D., R. A. Jorgensen, and W. F. Thompson. 1985. Chlo- Kuijt, I., and N. Goring-Morris. 2002. Foraging, farming, and social roplast DNA variation and evolution of Pisum: patterns of change complexity in the Pre-Pottery Neolithic of the southern Levant: a and phylogenetic analysis. Genetics 109:195–213. review and synthesis. Journal of World Prehistory 16:361–440. Pasternak, R. 1998. Investigations of botanical remains from Nevali Kvavadze, E., O. Bar-Yosef, A. Belfer-Cohen, E. Boaretto, N. Jakeli, C¸ori PPNB, Turkey: a short interim report. In The origins of ag- Z. Matskevich, and T. Meshvelani. 2009. 30,000-year-old wild flax riculture and crop domestication: proceedings of the Harlan Sym- fibers. Science 325:1359. posium 10–14 May 1997, Aleppo, Syria. A. B. Damania, J. Valkoun, ———. 2010. Response to comment on “30,000-year-old wild flax G. Willcox, and C. O. Qualset, eds. Pp. 170–176. Aleppo, Syria: fibers.” Science 328:1634-c. ICARDA. Ladizinsky, G. 1979. The genetics of several morphological traits in Peltenburg, E., S. Colledge, P. Croft, A. Jackson, C. McCartney, and the lentil. Journal of Heredity 70:135–137. M. A. Murray. 2000. Agro-pastoralist colonization of Cyprus in ———. 1993. Wild lentils. Critical Reviews in Plant Sciences 12:169– the 10th millennium BP: initial assessments. Antiquity 74:844–853. 184. ———. 2001. Neolithic dispersal from the Levantine corridor: a ———. 1999. Identification of the lentil’s wild genetic stock. Genetic Mediterranean perspective. Levant 33:35–64. Resources and Crop Evolution 46:115–118. Renfrew, J. M. 1968. A note on the Neolithic grain from Can Hasan. Ladizinsky, G., and S. Abbo. 1993. Wild lentils of Central Asia. FAO/ Anatolian Studies 18:55–56. IBPGR Plant Genetic Resources Newsletter 93:5–8. Rollefson, G. O., A. H. Simmons, M. L. Donaldson, W. Gillespie, Z. Ladizinsky, G., and A. Adler. 1976. The origin of chickpea Cicer Kafafi, I. U. Ko¨hler-Rollefson, E. McAdam, S. L. Ralston, and M. arietinum L. Euphytica 25:211–217. K. Tubb. 1985. Excavation at the Pre-Pottery Neolithic B village S254 Current Anthropology Volume 52, Supplement 4, October 2011

of ‘Ain Ghazal (Jordan), 1983. Mitteilungen der Deutschen Orient- van Zeist, W., and W. Waterbolk-van Rooijen. 1985. The palaeo- Gesellschaft zu Berlin 117:69–116. botany of Tell Bouqras, eastern Syria. Pale´orient 11:131–147. Saisho, D., and M. D. Puruggana. 2007. Molecular phylogeography Vigne, J.-D., I. Carre`re, F. Briois, and J. Guilaine. 2011. The early of domesticated barley traces expansion of agriculture in the Old process of mammal domestication in the Near East: new evidence World. Genetics 177:1765–1776. from the Pre-Neolithic and Pre-Pottery Neolithic in Cyprus. Cur- Salamini, F., H. O¨ zkan, A. Brandolini, R. Scha¨fer-Pregl, and W. Mar- rent Anthropology 52(suppl. 4):S255–S271. tin. 2002. Genetics and geography of wild cereal domestication in Vigne, J.-D., I. Carre`re, J.-F. Salie´ge, A. Person, H. Bocherens, J. the Near East. Nature Reviews Genetics 3:429–441. Guilaine, and F. Briois. 2000. Predomestic cattle, sheep, goat and Schick, T. 1988. Nahal Hemar cave: cordage, basketry and fabrics. pig during the late 9th and the 8th millennium cal BC on Cyprus: Atiqot 38:31–43. preliminary results of Shillourokambos (Perkklisha, Limasol). In Schiemann, E. 1948. Weizen, Roggen, Gerste: Systematik, Geschichte, Archaeozoology of the Near East IV: proceedings of the 4th Inter- und Verwendung. Jena: Fischer. national Symposium on the Archaeozoology of Southwestern Asia Sharma, H. G., and J. G. Waines. 1980. Inheritance of tough rachis and Adjacent Areas. H. Buitenhuis, M. Mashkour, and F. Poplin, in crosses of Triticum monococcum and T. boeoticum. Journal of eds. Pp. 52–75. Groningen: Centre for Archeological Research and Heredity 71:214–216. Consultancy, Groningen Institute for Archaeology, Rijksuniversi- Smartt, J. 1990. Grain legumes: evolution and genetic resources. Cam- teit. bridge: Cambridge University Press. Waines, J. G. 1975. The biosystematics and domestication of peas Smithson, J. B., J. A. Thompson, and R. G. Summerfield. 1985. (Pisum L.). Bulletin of the Torrey Botanical Club 102(6):385–395. Chickpea (Cicer arietinum L.). In Grain legume crops.R.J.Sum- Wang, A., Z. Yu, and Y. Ding. 2009. Genetic diversity analysis of wild merfield and E. H. Roberts, eds. Pp. 312–390. London: Collins. close relatives of barley from Tibet and the Middle East by ISSR Sonnante, G., I. Galasso, and D. Pignone. 2003. ITS sequence analysis and phylogenetic inference in the genus Lens Mill. Annals of Botany and SSR markers. Comptes Rendus Biologies 332:393–403. 91:49–54. Weiss, E., M. E. Kislev, and A. Hartmann. 2006. Autonomous cul- Sonnante, G., K. Hammer, and D. Pignone. 2009. From the cradle tivation before domestication. Science 312:1608–1610. of agriculture a handful of lentils: history of domestication. Rendi- Willcox, G. 2000. Pre´sence des ce´re´ales dans le Ne´olithique pre´- conti Lincei 20:21–37. ce´ramique de Shillourokambos a`Chypre: re´sultats de la campagne Takahashi, R. 1955. The origin and evolution of cultivated barley. 1999. Pale´orient 26:129–135. Advances in Genetics 7:227–266. ———. 2005. The distribution, natural habitats and availability of Tanno, K., and G. Willcox. 2006. The origins of cultivation of Cicer wild cereals in relation to their domestication in the Near East: arientinum L. and Vicia faba L.: early finds from Tell el-Kerkh, multiple events, multiple centres. Vegetation History and Archaeo- northwest Syria, late 10th millennium BP. Vegetation History and botany 14:534–541. Archaeobotany 15(3):197–204. Willcox, G., S. Fornite, and L. Herveux. 2008. Early Holocene cul- van der Maesen, L. J. G. 1972. Cicer L., a monograph on the genus, tivation before domestication in northern Syria. Vegetation History with special reference to the chickpea (Cicer arietinum L.), its ecology and Archaeobotany 17:313–325. and cultivation. Wageningen: Veenman & Zonen. Zeder, M. A. 2011. The origins of agriculture in the Near East. Current van Oss, H., Y. Aron, and G. Ladizinsky. 1997. Chloroplast DNA Anthropology 52(suppl. 4):S221–S235. variation and evolution in the genus Lens Mill. Theoretical and Zohary, D. 1960. Studies on the origin of cultivated barley. Bulletin Applied Genetics 94:452–457. of the Research Council of Israel 9D:21–42. van Zeist, W. 1972. Palaeobotanical results in the 1970 season at ———. 1969. The progenitors of wheat and barley in relation to C¸ayo¨nu¨, Turkey. Helinium 12:3–19. domestication and agriculture dispersal in the Old World. In The ———. 1976. On macroscopic traces of food plants in Southwestern domestication and exploitation of plants and animals.P.J.Ucko Asia (with some references to pollen data). Philosophical Trans- and G. W. Dimbleby, eds. Pp. 47–66. London: Duckworth. actions of the Royal Society B 275:27–41. ———. 1995. Lentil. In Evolution of crop plants. 2nd edition. J. Smartt van Zeist, W., and J. A. H. Bakker-Heeres. 1975. Evidence of linseed and N. W. Simmonds, eds. Pp. 271–274. Harlow: Longman Sci- cultivation before 6000 BC. Journal of Archaeological Science 2: entific & Technical. 215–219. ———. 1996. The mode of domestication of the founder crops in ———. 1982 (1985). Archaeological studies in the Levant. 1. Neo- Southwest Asian agriculture. In The origins and spread of agriculture lithic sites in the Damascus Basin: Aswad, Ghoraife´, Ramad. Pa- and pastoralism in Eurasia. D. R. Harris, ed. Pp. 142–158. London: laeohistoria 24:165–256. University College London Press. van Zeist, W., and H. Buitenhuis. 1983. A palaeobotanical study of ———. 1999. Monophyletic vs. polyphyletic origin of the crops on Neolithic Erbaba, Turkey. Anatolica 10:47–89. which agriculture was founded in the Near East. Genetic Resources van Zeist, W., and W. A. Casparie. 1968. Wild einkorn wheat and barley from Tell Mureybit in northern Syria. Acta Botanica Neer- and Crop Evolution 46:133–142. landica 17:44–53. Zohary, D., and Z. Brick. 1961. Triticum dicoccoides in Israel: notes van Zeist, W., and G. J. de Roller. 1991–1992. The plant husbandry on its distribution, ecology and natural hybridization. Wheat In- of aceramic C¸ayo¨nu¨, S.E. Turkey. Palaeohistoria 33/34:65–96. formation Service 13:6–8. ———. 1995. Plant remains from Asikli Ho¨yu¨k, a Pre-Pottery Neo- Zohary, D., J. R. Harlan, and A. Vardi. 1969. The wild diploid pro- lithic site in central Anatolia. Vegetation History and Archaeobotany genitors of wheat and their breeding value. Euphytica 18:58–65. 4:179–185. Zohary, D., and M. Hopf. 1973. Domestication of pulses in the Old ———. 2003. The C¸ayo¨nu¨ archaeobotanical record. In Reports on World. Science 182:887–894. archaeobotanical studies in the Old World. W. van Zeist, ed. Pp. Zohary, D., M. Hopf, and E. Weiss. 2011. Domestication of plants in 143–166. Groningen: Groningen Institute of Archaeology, Uni- the Old World. 4th edition. Oxford: Oxford University Press. Forth- versity of Groningen. coming. Current Anthropology Volume 52, Supplement 4, October 2011 S255

The Early Process of Mammal Domestication in the Near East New Evidence from the Pre-Neolithic and Pre-Pottery Neolithic in Cyprus

by Jean-Denis Vigne, Isabelle Carre`re, Franc¸ois Briois, and Jean Guilaine

Recent archaeological investigations on Cyprus have unveiled unsuspected Late Glacial and Early Holocene (twelfth–tenth millennia cal BP) pieces of the island’s human history. Based on a review of the archaeological data and of the final results of the archaeozoological analyses of Sector 1 of the prepottery site at Shillourokambos, this paper examines how Cyprus improves our understanding of the process of mammal domestication in the Near East. Early introduction of controlled wild animals and then of early domestic lineages provides information about the modalities of the do- mestication process on the mainland. This information emphasizes the importance of technical skills, of local opportunities and adaptations, and of long-distance and increasing exchanges in the larger Near East area. Cyprus was a recipient of wild or domestic taxa from the continent through recurrent introductions, but it was fully part of the wider area of incipient farming, as seen in local innovations such as the intensive hunting/control of wild deer and boar or local domestication of wild/feral goats. The transition to farming during the tenth millennium appears to follow an unstable and oppor- tunistic Early and Middle Pre-Pottery Neolithic B phase of low-level food production based on rapidly changing combinations of hunting, control, and breeding.

Introduction recent field projects and archaeozoological studies have led to a reconsideration of the general scenarios (cf., e.g., Bar- Yosef and Meadow 1995; Ducos 1968; Helmer 1992; Legge During recent decades, many new data have been accumulated 1996; Zeder 2005). by archaeology, archaeozoology, and genetics on the begin- Since the 1990s, Cyprus has provided startling new evidence nings of mammal domestication in several regions of the from archaeological discoveries dating to the late eleventh and world. This includes the Near East, where the study of early the tenth millennia cal BP (Guilaine and Le Brun 2003; Pel- domestication began very early and has provided more evi- tenburg and Wasse 2004),1 contemporaneous with the end of dence and models than many other regions (e.g., Davis 2005; the Early Pre-Pottery Neolithic B (EPPNB) and the Middle Dobney and Larson 2006; Redding 2005; Zeder 2006, 2011; Pre-Pottery Neolithic B (MPPNB) and Late Pre-Pottery Neo- Zeder et al. 2006). However, even in the Near East, our knowl- lithic B (LPPNB) of the mainland. Because islands are clearly edge remains poor, as demonstrated by the fact that many defined territories, because the Mediterranean islands were massively colonized beginning only in the Neolithic (Cherry Jean-Denis Vigne is a senior researcher at the Centre National de 1990), because their endemic native fauna lacked any of the la Recherche Scientifique (CNRS) and Director of Unite´ Mixte de wild ancestors of domesticates (Vigne 1999), and because Cy- Recherche 7209, “Arche´ozoologie, Arche´obotanique: Socie´te´s, prus is the only large island in the Near East, it provides Pratiques et Environnements,” Muse´um National d’Histoire unique information. As Cyprus was necessarily dependent on Naturelle–CNRS (75005 Paris, France [[email protected]]). Isabelle Carre`re and Franc¸ois Briois are researchers at Unite´ Mixte de influences from the continent, at least at the beginning of the Recherche 5608, “Travaux et Recherches Arche´ologiques sur les Neolithic, it may be thought of as an observation post re- Cultures, les Espaces et les Socie´te´s” (TRACES), Ecole des Hautes cording what occurred on the nearby mainland. Etudes en Sciences Sociales, Universite´ Le Mirail (31000 Toulouse, In this paper, we present a synthesis from the Cypriot Pre- France). Jean Guilaine is Professor Emeritus at the Colle`ge de France (75005 Paris, France). This paper was submitted 13 XI 09, accepted 1. All the calibrated dates are given cal BP with 1j, using Calib Rev 28 XII 10, and electronically published 26 VIII 11. 5.0 (Reimer et al. 2004).

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0008$10.00. DOI: 10.1086/659306 S256 Current Anthropology Volume 52, Supplement 4, October 2011

Pottery Neolithic (PPN) perspective on early mammal do- Pleistocene terrestrial mammalian fauna was reduced to only mestication in the Near East. This is based mainly on the four or five endemic species (Boekschoten and Sondaar 1972; recent results of the final analyses of the first sector of Pa- Simmons 1999): mouse (Mus cypriacus; Cucchi et al. 2006), rekklisha, Shillourokambos (Limassol district), the largest genet (Genetta plesictoides), dwarf elephant (Elephas cypriotes), known Cypriot PPN site (10,400–9000 BP; Guilaine 2003; dwarf hippopotamus (Phanourios minutus), and perhaps a Guilaine and Briois 2007; Guilaine, Briois, and Vigne 2011). shrew. Our approach to animal domestication lies within both the The upper layer of the rock shelter at Akrotiri Aetokremnos technological (Mauss 1947 [1967]) and the structuralist con- provides the earliest known evidence of visits to the island ceptual frameworks (Le´vi-Strauss 1958): technoeconomic and by humans, dated to 12,776–12,461 cal BP (nine wood char- symbolic uses of animals by human societies are part of their coal dates; Simmons 1999; Simmons and Mandel 2007), cor- technical, social, and symbolic systems and thus a character- responding to the Late Natufian in the northern Levant. Sim- istic part of their cultural systems (Vigne 1998). However, as mons (1988, 1999) has proposed that these human groups part of the ecosystem (another structural entity), the rela- hunted elephants and hippos to extinction. This is, however, tionships between animals and humans also involve ecological questioned by numerous authors (e.g., Ammerman and approaches and should be considered within the structural Noller 2005; Binford 2000; Bunimovitz and Barkai 1996; Da- framework of the “anthroposystem,” that is, a metasystem vis 2003; Olsen 1999; Vigne 1999; Wasse 2007), who argue that groups together the cultural and ecological systems and that the two native large mammals disappeared before that their interactions and dynamics through time (Muxart et al. time for climatic (Bromage et al. 2002) or anthropic reasons 2003; Pascal, Lorvelec, and Vigne 2006; Vigne 2011). (or both; Wasse 2007). The diet of the occupants of Aeto- From this viewpoint, in addition to the different kinds of kremnos shelter seems to have consisted mainly of fish, shell- ecological affinities of animals to the human-made or human- fish, and birds (Simmons 1999). Ammerman et al. (2006, modified ecosystems, including commensalisms, domestica- 2008) recently found two other sites that may be more or less tion appears as a process of intensification of animal-human contemporaneous with Aetokremnos, but no radiometric data relationships (Pascal, Lorvelec, and Vigne 2006) boosted by have corroborated their interpretations until now. human intentionality.2 Consequently, we consider that control In addition to thousands of hippo bones, 18 suid bones in the wild and control/protection/use of commensals are were found at Aetokremnos, mainly in the upper layer (Sim- initial stages of the domestication process, although they in- mons 1999). They have recently been directly dated to 11,746– BP [AA79923; degraded bone 69 ע volve animal populations that are still wild from a biological 11,396 cal BP (10,045 13 viewpoint. Such situations can be detected mostly as an in- collagen]; d C, Ϫ25.1‰; Vigne et al. 2009). These suids were crease in the frequency of the targeted species, as changes in 9%–20% smaller than the Near Eastern Late Glacial and Ho- age and sex proportions in the archaeological record (Zeder locene wild boars and were also significantly smaller than 2009, 2011), and/or as transport. We restrict the term “do- those of the PPN and Early Pottery Neolithic of the Near mestic mammals” to animal populations that show pheno- East. The most probable scenario (already partly suggested typic modifications due to human breeding/herding. As some by Wasse 2007) is that wild boars were artificially introduced of these cannot be archaeologically detected, sometimes the to Cyprus sometime at the end of the Late Glacial (Natufian domestic status cannot be determined. In the same way, we on the mainland), rapidly decreased in size because of the restrict the term “farming” to technoeconomic systems based isolation effect, and then were hunted by people who fre- mostly on animal (and plant) breeding, which appears to have quented Aetokremnos during the second quarter of the started in the middle of the tenth millennium (transition twelfth millennium (Late Khiamian on the mainland; fig. 1). between MPPNB and LPPNB) in large villages of the Near At least three recent discoveries indicate that whether per- East (Vigne 2008, 2011). We distinguish farming from “low manently or not, humans continued to frequent the island level food production” as defined by Smith (2001). in the twelfth millennium cal BP and during the first half of the eleventh. At Ayios Tychonas Throumbovounos, Franc¸ois Briois and collaborators (Briois, Petit-Aupert, and Pe´choux The Early Neolithization of Cyprus: 2005; Guilaine and Briois 2007) found an eroded site with A Brief Updated Review large flint series similar to Mureybet II (Khiamian of the Middle Euphrates valley, late Dryas). Unfortunately, there The island of Cyprus emerged from the bottom of the Med- were no associated faunal remains (F. Briois and J.-D. Vigne iterranean Sea during the Miocene and has never been con- “Throumbovounos (Ayios Tychoˆnas, Chypre), rapport de nected to any continent (Held 1989). As a result, the Upper sondage,” unpublished report for the Department of Antiq- uities, Cyprus, 2003). At Agia Varvara Asprokremnos, painted 2. The conceptual frame summarized in this paragraph has been stone vessels, pendants, a shaft-straightener, and unidirec- strengthened, clarified, and here and there modified by discussions during the Me´rida meeting “The Beginnings of Agriculture,” mainly among the tional flint de´bitage provide strong parallels to materials dated archaeozoology working group: Fiona Marshall, Richard Meadow, Peter to the late Pre-Pottery Neolithic A [PPNA]/transitional Rowley-Conwy, and Melinda Zeder. EPPNB in the northern Levant (McCartney 2010; McCartney Figure 1. Evolution of large mammals on Cyprus during Neolithization, with particular focus on the data from Shillourokambos. The chrono- S258 Current Anthropology Volume 52, Supplement 4, October 2011 et al. 2007, 2008, 2010). The associated fauna is dominated Khirokitia culture, contemporaneous with the pre-Halaf/ by suids (Sus scrofa). A series of six radiocarbon dates ranging Halaf phases of the northern Levant (fig. 1). -BP spans a likely cal 49 ע BP and 9432 49 ע between9525 endar age of 10,846–10,675 cal BP (Manning et al. 2010). The last site, Ayios Tychonas-Klimonas, is also characterized by a The Main Archaeological and PPN (EPPNB-type) industry (Guilaine and Briois 2007) as- Archaeozoological Characteristics sociated with suids and dog bones (Vigne et al., forthcoming). of Shillourokambos The subsistence of the occupants of the site should have been Shillourokambos is an open-air site located on the southern based on hunting of only one game species, the small au- coastal plain of Cyprus, near Limassol (Guilaine 2003). From tochthonous wild boar, with the help of dogs. Only one wood 1992 to 2004 it was excavated over more than 5,000 m2 under charcoal (Prunus sp.) radiocarbon date indicates the first the responsibility of Jean Guilaine. Because it is located on quarter of the eleventh millennium cal BP (11,070–10,741 cal the top of a low hill near the confluence of two small rivers, ,.BP [AA88551]; d13C, Ϫ27.3‰; Vigne et al 53 ע BP, 9544 it has been strongly eroded. No building is preserved in the forthcoming). upper part of the site (Sector 1, ca. 3,000 m2), which is mainly The lithic de´bitage of these three sites and the radiocarbon composed of large sedimentary layers and numerous pits, dates of two of them clearly postdate those of Aetokremnos narrow ditches, post holes, and wells. The bases of several and predate the earliest phase (A) at Shillourokambos (Gui- buildings have, however, been preserved in the less eroded laine 2003; Guilaine and Briois 2007; Guilaine et al. 2000) lower part of the site (Sector 3, ca. 2,500 m2). and Kissonerga-Mylouthkia (Peltenburg 2003; Peltenburg et Four main phases of occupation of the site have been rec- al. 2000, 2001a, 2001b), both of which are radiocarbon dated ognized (Early A, Early B, Middle, and Late), spanning the (fig. 1). The Early A phase of Shillourokambos, where suids period from ca. 10,400–10,300 cal BP to the end of the tenth are still overwhelmingly dominant (Vigne, Carre`re, and Gui- millennium (fig. 1; Guilaine 2003; Guilaine et al. 2000). laine 2003), provided 12 dates (Guilaine, Briois, and Vigne Thirty-seven radiocarbon dates from Sector 1 suggest that at 2011) ranging from 10,696 to 10,297 cal BP (mean: 10,380 least this part of the site had been abandoned for several cal BP). Well 116 at Mylouthkia is dated to 10,740–10,290 decades or centuries between the Early phases A and B, at BP [AA33128]; the end of the eleventh millennium (Guilaine, Briois, and 70 ע BP [OxA7460]; 9235 60 ע cal BP (9315 BP [1133129]; Peltenburg 2003). At Shillourokam- Vigne 2011). However, it was occupied continuously during 70 ע 9110 bos, the blade technology of this late eleventh-millennium the tenth millennium. phase is very well documented and shows clear relationships Preliminary analyses of the animal remains of Sector 1 with Early MPPNB in the Levant (Briois 2003). (Vigne et al. 2000) showed that dog (Canis familiaris), fox From the beginning of the tenth millennium, prepottery (Vulpes vulpes), pig (Sus scrofa), the Mesopotamian fallow deer Cypriot sites became more numerous (Guilaine and Briois (Dama mesopotamica), goat (Capra aegagrus/hircus), sheep 2007; Guilaine and Le Brun 2003; Peltenburg and Wasse 2004; (Ovis aries), and cattle (Bos primigenius/taurus) were already Wasse 2007). They show a progressive shifting of all the ma- present before or at the very beginning of the tenth millen- terial culture toward a local Cypriot model, which culminated nium. Cat (Felis silvestris lybica) was thought to have been during the ninth and eighth millennia BP in the Aceramic introduced during the middle phases and partly controlled

cultural frame is that proposed by Hours et al. (1994). Radio- carbon dates are given in cal BP and calibrated at 1j (Calib Rev 5.0; Reimer et al. 2004). For each chronological phase of Shil- lourokambos, a box on the right-hand side contains the following information: the total number of identified specimens (NISP), the NISP of large mammals, the material dated (Ch p charcoal; Cl p bone collagen), and a reference (1, Simmons 1999 [only the nine charcoal dates]; 2, Vigne et al. 2009; 3, Vigne et al., forthcoming; 4, Manning et al. 2010; 5, Peltenburg 2003; Pel- tenburg et al. 2001b; 6, Guilaine 2003; Guilaine, Briois, and Vigne 2011; 7, Le Brun and Daune-Le Brun 2003 [Khirokitia, earliest level, G]). The Middle A1 and A2 phases of Shillouro- kambos are grouped together. Black arrows indicate the earliest evidence for the arrival of new species of fauna on Cyprus. Gray arrows indicate the probable introduction of new lineages of domestic or commensal animals to Shillourokambos and Khi- rokitia. Vigneetal. Mammal Domestication in the Near East S259

Table 1. Final archaeozoological analyses of the large mammals from Sector 1 at Shillourokambos

Morphometrics Culling profiles Animal species NISP No. measurements No. teeth MNI NISP epiphyseal Vulpes vulpes 56 214 Canis familiaris 11 77 Felis silvestris lybica 311 Sus scrofa 1,001 2,917 472 113 1,124 Dama dama mesopotamica 1,361 4,017 785 173 2,374 Capra aegagrus/hircus 394 1,284 Ovis aries 378 1,318 Capra/Ovis 219 515 Total Caprini 991 3,117 321 108 754 Bos taurus 124 336 26 Total 3,547 10,689 1,578 394 4,278 Note. Number of measured specimens, measurements, and teeth, the minimum number of individuals (MNI) used for the culling profiles, and the number of the bones used for epiphyseal profiles. NISP p number of identified specimens.

(“tamed”; Vigne et al. 2004). In the first preliminary study, The first important result of this final analysis is that the all the ungulate species except the fallow deer were considered proportions of the different mammal species vary significantly to have been introduced in the form of already-domesticated throughout the successive occupations of the site to such an lineages, mainly because of their small size in comparison extent and in such a well-structured way that it was possible with their wild counterparts on the mainland. This interpre- independently to find again the four phases based on the tation has been questioned by Horwitz, Tchernov, and Hongo lithics only through faunal characteristics. Correspondence (2004), who suggested, though without any direct analytical analysis of the faunal spectra even allows refinement of the data, that they were introduced as wild animals for stocking chronology and identification of several subphases: Early A1 the island with meat sources. and A2 (the second being more or less the initial stage of the Further investigation of the Early phase B at Shillouro- Early phase B), Early C, and Middle A1, A2, and B phases kambos suggested a more complex situation (Vigne, Carre`re, (the last more probably being the first stage of the late phase). and Guilaine 2003; Vigne and Guilaine 2004): suids, sheep, This new chronological framework enables a more precise and cattle were small in size and might therefore be considered approach to the succession of events on this site between as resulting from early domestication; but morphological 10,400 and 9000 cal BP. analyses and the slaughtering profiles suggest that (almost?) The final archaeozoological results presented in this paper all the goats and approximately half of the suids were wild/ are based on 32,500 animal remains, of which 9,225 (the num- feral and were obtained by hunting, while sheep and probably ber of identified specimens [NISP]) are taxonomically deter- cattle were bred. These observations underlined the unstable mined (8,864 vertebrates and 361 invertebrates; Guilaine, Bri- status of the domestic ungulates during several centuries after ois, and Vigne 2011). On the basis of cut marks (Vigne 2006), the earliest domestication and suggested that the human so- fragmentation, and skeletal frequencies, reconstruction of the cieties on the mainland could also have released domestic (or butchery and cooking processes and estimation of the distance even a mixture of wild and transported domesticated) un- between the kill sites and the village for each species of ungulate gulates into the wild and redomesticated them many times were possible. Nearly 11,000 bone and tooth measurements during the 15 centuries of the Pre-Pottery Neolithic B (table 1) provided statistics and enabled multivariate analyses. (PPNB), as well as later. For deer, goat, cattle, and to a much lesser extent sheep, mixture At that time, the question of what happened on Cyprus analyses (Monchot and Le´chelle 2002) provided reliable esti- before the beginning of the tenth millennium (i.e., during the mates of the sex ratio of adults and the decrease in sexual Early phase A) could not, however, be answered, because this dimorphism due to domestication (see Helmer 2008; Helmer earliest phase had provided less than 100 specimens. The last et al. 2005). By this technique, the evolution of the size of the season of excavation at Shillourokambos in 2004 produced different parts of the skeleton of the dimorphic ungulates could many more animal bones and much more archaeological in- also be studied for each sex, providing reliable approaches to formation for this phase (Guilaine et al. 2008). The final study morphological variations through time. The tooth and epi- of all the animal bones of Sector 1 could be carried out physal mortality profiles were based on large series of teeth and (Guilaine, Briois, and Vigne 2011). Jean-Denis Vigne and Is- long bones (table 1) and processed according to Vigne and abelle Carre`re were in charge of the archaeozoological analyses Helmer (2007). We did not use statistical comparisons (see and were active participants in the excavation. Consequently, Marom and Bar-Oz 2009), which are not adapted to these they were in a good position to select the most reliable faunal questions (Vigne 2000), but rather multivariate analyses (e.g., assemblages. Helmer, Gourichon, and Vila 2007). S260 Current Anthropology Volume 52, Supplement 4, October 2011

These final analyses led to modification and clarification can result only from humans (Vigne 1999), these species of several aspects of the preliminary interpretations. Figure 1 would have been introduced between the end of the Late presents the frequencies of species according to the phases. Glacial and ca. 10,400 cal BP. During the Early phase A1, suids, which are the same size as All the large true Mediterranean islands have undergone a those at Aetokremnos, are dominant (90%). Cats, goats, and similar turnover of their mammal fauna. That on Cyprus is cattle are clearly present in small numbers. The presence of the earliest, beginning during the Late Glacial, because do- dog is probable, that of fox questionable. The statistical anal- mestication in the Eastern Mediterranean was early (Vigne yses indicate that the absence of fallow deer and sheep is very 1999). unlikely to be due to the size of the sample (NISP p 282 ). It is confirmed by the absence of both species in well 116 at Pre-Neolithic Wild Boar Management More than Mylouthkia, which is contemporaneous with Early phase A 11,400 Years Ago of Shillourokambos (Peltenburg et al. 2001b). Deer appeared as the dominant species at the very beginning of Early phase The presence of small wild boar dated to 11,746–11,396 cal B. Sheep probably arrived slightly later. BP at Aetokremnos, the dominance of suids at Asprokremnos, Early phase B is characterized by high frequencies of cattle Klimonas, and Shillourokambos Early A1, and the morpho- (8%–12% of the total NISP of the phase), which decrease to logical similarities of suids at Aetokremnos and Shillouro- less than 1% in the middle and late phases. The Middle A kambos are convergent evidence that small island suids were phases are characterized by an increase in suids and foxes. living on Cyprus and exploited by humans during the twelfth– Sheep and goats become dominant in the Middle B and late eleventh millennia. This would explain why people frequented phases. (or inhabited) Cyprus during this period. The introduction of wild boar during the Late Natufian or Transport and Control: The Beginnings Early Khiamian strengthens the hypothesis of control of wild boar in the Final Late Glacial, as proposed by Redding and of Mammal Domestication before Rosenberg (1998), Rosenberg and Redding (1998), and Rosen- 10,300 Cal BP berg et al. (1998; see also Redding 2005; Starkovitch and Stiner 2009; Wasse 2007) on the basis of the Hallan C¸emi site in the upper Tigris basin. The evidence from Cyprus makes a strong Faunal Turnover on the Island contribution to the picture of a long span of increasingly in- Although Mus cypriacus is attested to at Mylouthkia and is tensive and skilled control of wild boars. This took place over present today on the island (Cucchi et al. 2006), animal bones a very large area, at least in eastern Anatolia (and rapidly ex- at Asprokremnos, Klimonas, Mylouthkia, and Shillourokambos panded to Cyprus), but it would have extended to western confirm that the Pleistocene endemic hippo fauna did not sur- Anatolia and the Aegean, as suggested by evidence of small vive the Late Glacial–Holocene transition. According to Bu- Mesolithic wild boar on the islands of Youra and Kythnos nimovitz and Barkai (1996) and Ammerman and Noller (2005), during the eleventh–tenth millennia (Trantalidou 2008). Here this fauna would have been extinct before the human occu- and there, it developed into true pig breeding in the middle of pation of layer 2 at Aetokremnos in the first centuries of the the eleventh millennium, as observed in the Upper Euphrates twelfth millennium (see also Wasse 2007). The introduction of valley at Nevalı C¸ori and Gu¨rcu¨tepe (Peters, von den Driesch, wild boar before the occupation of Aetokremnos renders it and Helmer 2005) and at Cafer Ho¨yu¨k (Helmer 2008). At the possible that people frequented the island before 12,000 BP and same time, control in the wild would have continued during then caused the extinction of the hippos. a large part of the tenth millennium, as suggested at C¸ayo¨nu¨ The endemic Cyprus genet (Genetta plesictoides) is present (Ervynck et al. 2001) and in the Aegean. in the Aetokremnos deposits without any direct dating (Sim- mons 1999). It is absent from the small-mammal fauna of Transport of Wild or Early Domestic Goats? Mylouthkia. It is unlikely that it became extinct at the same time as hippos because its presumed main prey, M. cypriacus, In our present state of knowledge, the earliest evidence of did survive. The introduction of dogs (before 11,000 cal BP) goats on Cyprus is about 40 bone specimens at Shillouro- or cats (mid-eleventh millennium at the latest) may have kambos Early A. The final analyses for Shillourokambos con- caused the extinction of the genet sometime in the twelfth– firmed that the morphology of the horn cores of the early eleventh millennia. phases was similar to that of the wild bezoar goat. But separate Suids, cats, goats, cattle, and dogs were living on Cyprus morphometric analyses of males and females has shown that before or starting from the earliest phase at Shillourokambos contrary to our preliminary conclusions (Vigne, Carre`re, and (fig. 1). In addition, data from Mylouthkia provide evidence Guilaine 2003; Vigne et al. 2000), these animals were signif- for the presence of the house mouse (Mus musculus domes- icantly smaller than the early domestic goat, as measured by ticus; Cucchi et al. 2002). Because they have no Cypriot Pleis- Helmer (2008) at Cafer Ho¨yu¨k (Early MPPNB). Peters, von tocene ancestor and because such a high rate of immigration den Driesch, and Helmer (2005), Helmer (2008), and Hongo Vigneetal. Mammal Domestication in the Near East S261 et al. (2009) presented convincing evidence of domestic goat for cat at Shillourokambos is a complete fifth right metacarpus in the Upper Euphrates and Tigris valleys in the middle of found in the bottom of the earliest stratigraphic unit, dated the eleventh millennium (late EPPNB), and Helmer and to 10,322–10,288 cal BP (Guilaine, Briois, and Vigne 2011). Gourichon (2008) did the same for the beginning of the Although there are only two cat remains in Sector 1 of Shil- MPPNB of the Damascus plain (Tell Aswad), ca. 10,300– lourokambos (fig. 1), the species is frequently present on tenth 10,200 BP. It is thus likely that the earliest goats of Cyprus to eighth millennium Cypriot sites (review in Vigne and Gui- were introduced as already modified domesticates. laine 2004). The house mouse and the cat were introduced However, on the basis of convincing theoretical inferences, to Cyprus in the middle of the eleventh millennium at the Redding (2005) argued for wild-goat control before the PPNB latest, and the probably unintentional transport of the former (see also Hole 1996). Genetic investigations of the Near and possibly led to the introduction of its predator. Middle Eastern modern bezoar goats (Capra aegagrus) have Auffray, Tchernov, and Nevo (1988) showed that house shown that the wild lineages that mothered modern domestic mice were already commensal in the Natufian layers at Hay- goats in the Zagros were subjected to a drastic demographic onim, but nothing was known about early domestication of increase ca. 10,000 BP, whereas the other lineages showed no cats in the Near East. The presence of both house mice and similar genetic signature (Naderi et al. 2008). They also sug- cats in Cyprus during the second half of the eleventh millen- gest that wild goats might have been transferred out of the nium not only confirms that the house mouse spread as a original area of the species, though still in mountain areas. pest to all the territories of the PPN complex (Cucchi and Like that of wild boar, the domestication of goats seems to Vigne 2006), including islands, but is also evidence that cat have begun as control in the wild in a very large area stretching domestication had already begun. It also supports the idea, from the central Iranian plateau to southeastern Anatolia. Like suggested by Malek (1993) in relation to Egypt, that the most the pig, morphologically domestic goat would have appeared likely reason for cat domestication was use of this predator here and there, as seen in the data from Nevalı C¸ori, ca. 10,500 against the commensal rodents and their consumption of ag- cal BP (Peters, von den Driesch, and Helmer 2005), and later ricultural products. Cat domestication would have followed (10,000–9500 cal BP) in the Zagros (Zeder 2005, 2011; Zeder two steps. First, “commensalization” would have occurred, and Hesse 2000). that is, cats being attracted to the villages by high concen- The final analyses at Shilloroukambos confirmed that (all?) trations of mice, whose presence was due to an increase in goats were hunted not only during the Early phase B (Vigne, stocked foodstuffs. Second, the villagers would merely have Carre`re, and Guilaine 2003) but also during the Early phase provided protection or simply impunity to some individuals C and the beginning of the middle phases. The bone sample for the commensalism to evolve toward domestication (Pas- for the Early phase A is too small for characterizing the mode cal, Lorvelec, and Vigne 2006). This would have been all the of exploitation, but the low frequency in comparison with more true if the cat had a symbolic significance for those the middle phases (fig. 1) makes hunting more likely. It is Neolithic societies, as suggested by the numerous represen- therefore difficult to determine whether the introduction of tations of felids in the PPN (Helmer, Gourichon, and Stordeur goats to Cyprus in the course of the eleventh millennium 2004), including in Shillourokambos Early A (Guilaine et al. resulted from the transport of controlled bezoar goats (as 1999). suggested by Horwitz, Tchernov, and Hongo 2004), which Should we expect future evidence for domestic cats in the would have decreased in size because of insularity, or was a Natufian on the mainland? This is not certain, because the consequence of a rapid spread of early domestic goats. The proposal of Auffray, Tchernov, and Nevo (1988) has to be ecology of the bezoar goat (which excludes its presence in the reevaluated with modern morphometric techniques and be- coastal plains, from which the visitors to Cyprus necessarily cause, although it began during the Natufian, food storage came), its introduction much later than the wild boar, and does not seem to have developed on a large scale before the the absence of any evidence of goat control in the nearby PPNA–EPPNB (Kuijt 2008; Kuijt and Finlayson 2009). In mainland areas, however, make the latter scenario more likely. addition, the abundance of foxes in the PPNA at sites such In any case, Cyprus provides strong evidence that the process as Jericho (Clutton-Brock 1979) and their introduction to of goat domestication in the Near East was well advanced in Cyprus at the beginning of the tenth millennium (fig. 1) the middle of the eleventh millennium. indicate that other species of small carnivores could also have played the same role as cats throughout the PPN. As domestic dogs are in evidence on the mainland since From Commensalism to Domestication: Mice, Cats, and the Natufian (Helmer 2008; Tchernov and Valla 1997) and Other Carnivores on Cyprus as early as the beginning of the eleventh millen- Cucchi et al. (2002) demonstrated that the house mouse (M. nium (Vigne et al., forthcoming), it would not be surprising musculus domesticus), absent from Aetokremnos (Simmons to find them on Cyprus during the twelfth millennium. They 1999), was well represented in well 116 of Mylouthkia, dated probably played a role in autochthonous-wild-boar hunting to 10,740–10,290 cal BP, together with the less commensal or control at Klimonas during the first half of the eleventh endemic Cypriot mouse, M. cypriacus. The earliest evidence millennium. At Shillourokambos, the scarceness of their re- S262 Current Anthropology Volume 52, Supplement 4, October 2011 mains and the total absence of any gnawing mark on the The Mesopotamian Fallow Deer other mammal bones suggest, however, that they lived away The final archaeozoological analyses provide evidence that from the village, at least from the beginning in the Early phase deer were never domesticated at Shillourokambos. In all B (Vigne and Guilaine 2004). phases, age classes were slaughtered in proportion to their natural abundance without any distinction between sexes. The Introduction of Early Domestic Cattle during the EPPNB morphological distance between males and females remained constant, and the body size not only did not decrease but Although absent at Mylouthkia, cattle were present during slightly and constantly increased from the tenth to the eighth the Early phase A1 of Shillourokambos, represented by a com- millennium (Khirokitia), perhaps as a result of constant hunt- plete right pelvis found at a depth of 4 m in the Early A1 ing pressure. fills of well 431. Cattle were introduced to Cyprus ca. 10,300 The late introduction of the Mesopotamian fallow deer to cal BP or shortly before (fig. 1), at a time when domestic Cyprus (fig. 1) probably illustrates the persistence, several cattle had begun to appear on the mainland (Helmer et al. centuries after the beginning of mammal domestication, of 2005; Hongo et al. 2009). wild-ungulate control similar to that which we have shown Cattle bones are, however, too rare in the earliest phase at for wild boars during the twelfth–eleventh millennia. Even if Shillourokambos to provide any information on their mor- the economic importance of the fallow deer in Cyprus appears phology and status. Their size does not seem to differ sig- to be unique in all the PPN and therefore specifically Cypriot, nificantly from that of Early phase B cattle bones. The latter it does suggest that somewhere in the Near East, human are significantly smaller than those of the aurochs of the Na- groups were already controlling fallow deer either as a do- tufian and PPNA layers at Mureybet (Gourichon and Helmer mestication experiment or as a common subsistence practice. 2008). They are very similar to those of Dja’de (L. Gourichon Who were these pioneers of deer control? The Cypriot and D. Helmer, unpublished data), which are mainly com- fallow deer were noticeably smaller than their northern-Le- posed of early domestic bovids (Helmer et al. 2005). Estimated vant PPNA counterparts at Mureybet (according to Gouri- through mixture analyses of log size index, the sexual di- chon and Helmer 2008). The highest density of archaeological morphism of the bovid from the early phases of Shillouro- evidence for the Mesopotamian fallow deer is located in the kambos appears also to be significantly less than that of the southern Levant (Davis 2003). The PPNB complex of cultures aurochs of Mureybet I–IIIA and similar to that of Dja’de. In reached the southern Levant only at the end of the eleventh addition, the presence on the site of all the skeletal elements millennium BP (Khalaily, Marder, and Barzilai 2007), several centuries later than in the northern Levant. Mammal do- of the limb extremities and all the vertebrae suggests that the mestication was then in a more experimental phase in the animals were killed very near or in the village. The epiphysal south than in the north (Conolly et al. 2011). All this may age profile and sex ratio of the early phases indicate a well- suggest that the fallow deer was introduced from the southern managed breeding strategy focused on the exploitation of the Levant. meat from the young adult males. The early Cyprus cattle bred during the Early phases at Shillourokambos thus probably resulted from the introduc- Sheep and Goats: Complementary Stories tion of already-domestic lineages such as those at Dja’de or The large number of bone remains (table 1) of sheep and at less distant Tell Aswad (Damascus), where Helmer and goats and morphometric, age, and sex-ratio analyses (Gui- Gourichon (2008) found evidence that cattle were used for laine, Briois, and Vigne 2011; Vigne, Carre`re, and Guilaine, carrying heavy loads as early as the beginning of the MPPNB. forthcoming) enable reconstruction of the codevelopment of sheep and goat exploitation at Shillourokambos (fig. 2). Goats were hunted throughout the early phases. Beginning in the Tenth Millennium: Cyprus Reflects middle phases, the villagers of Shillourokambos began to ex- Continental Complexity and ploit them more intensively, strictly culling the young males, Intensive Exchanges although no additional morphological modification appeared at that time. However, they continued to hunt wild/feral goats. Numerous archaeozoological data from the Early phase B to Size decrease and morphological modifications of the horn the Late phase of Shillourokambos are summarized in figure cores become visible only in the late phases (i.e., 2–4 centuries 2 (after Guilaine, Briois, and Vigne 2011). This provides an later). The modification of the culling profiles through time exceptional case study for the evolution of the status of un- indicates that at that time, the exploitation of goats had shifted gulate species and of subsistence strategy through the tenth toward their milk (Vigne, Carre`re, and Guilaine, forthcom- millennium, a time of emergence and consolidation of farm- ing). Even though this concerns goat lineages that may have ing on the mainland (Bar-Yosef and Meadow 1995; Vigne been domesticated long before on the mainland and then 2008, 2011). became feral on the island, this is the first evidence for a Vigneetal. Mammal Domestication in the Near East S263

Figure 2. Summary of the technicoeconomic and sociosymbolic char- acteristics of ungulates through the different chronological phases of Shil- lourokambos, after Guilaine, Briois, and Vigne (2011). Shading indicates the importance of hunting; darker shading indicates greater importance. domestication process on a Mediterranean island. It resulted (Vigne, Carre`re, and Guilaine 2003) similar to their contem- in a “new” lineage of domestic goats ca. 9400–9000 BP, that poraneous counterparts of the MPPNB at Tell Aswad. As soon is, more than a millennium later than in sites of the northern as Early phase B, they were herded in a sophisticated way for Levant such as Nevalı C¸ori, C¸ayo¨nu¨, Tell Halula, or Tell Aswad meat and milk production (“milk B,” as defined by Vigne (Helmer and Gourichon 2008; Hongo et al. 2009; Peters, von and Helmer 2007). Just after an episode of rapid size decrease den Driesch, and Helmer 2005; San˜a Segui 1999). and stress increase (malnutrition marks on the horn cores) Sheep were introduced up to 5 centuries later than goats that we interpreted as a breeding failure or collapse (Middle (fig. 1), in the form of small, horn-modified domestic animals A1), new larger-sized lineages abruptly appeared, possibly in- S264 Current Anthropology Volume 52, Supplement 4, October 2011 troduced from the central or southern Levant. During the attested to several centuries earlier (Hongo et al. 2009; Peters, late phases, herding concentrated on meat production and von den Driesch and, Helmer 2005). possibly hair production. A second rapid morphological The cattle of the Early C, Middle, and Late phases of Shil- change perhaps suggests another new lineage. Davis (1994) lourokambos were much larger than those of the Early phases has demonstrated an event of very rapid and significant size A and B. They are approximately the same size as the domestic change between the early and late levels at Khirokitia and cattle of the Middle and Late PPNB layers of Halula (see San˜a interprets it as the possible introduction of a new lineage. Segui 1999) and therefore significantly larger than the cattle There is no evidence for feralization or sheep hunting at Shil- of Dja’de (L. Gourichon and D. Helmer, unpublished data). lourokambos. The exploitation of sheep on this site was These data suggest the introduction of a new lineage of do- strictly limited to breeding periodically strengthened by the mestic cattle between Early phases B and C. introduction of new lineages that probably came from the The tooth morphometry of the modern Cypriot house mainland. mouse M. musculus domesticus does not differ from that of Basically, for the villagers of early Shillourokambos, goats the mainland mouse and has not varied since the PPN (Cucchi were game animals, while sheep were the most “domestic” of et al. 2002). Taking into account that genetically isolated mice all the ungulates. This probably implies quite different sym- diverge very rapidly from their ancestor (Pergams and Ashley bolic values for each of them. Archaeozoology tends to mix 2001), this morphological stability is evidence of a constant these two species only because their bones are difficult to and intensive genetic flow from the mainland to Cyprus from discriminate from one another. The evidence from Shillouro- the Neolithic up to today. Mice would have been successfully kambos demonstrates once again that they are really very introduced into Cyprus at least several times per year (Vigne different animals from bioecological, technoeconomic, and and Cucchi 2005). symbolic points of view (e.g., Balasse and Ambrose 2005). The arrows in figure 1 summarize the introductions of new Because of these differences, the two species were able to play species (black) and new domestic (gray) lineages to Cyprus complementary roles in the technoeconomic system. If the and more specifically to Shillourokambos during the twelfth– rapid Middle A1 size decrease of the sheep at Shillourokambos eighth millennia. Of course, introductions are possibly un- derevaluated because lineages that did not morphologically actually resulted from a herding failure, we could consider differ from the autochthonous ones could not be detected. that the domestication of goats in this village, which began On the other hand, the appearance of new lineages at Shil- at approximately that time, was developed to offset it. The lourokambos did not necessarily result from introductions complementarities between the two species are again illus- from the mainland, as some of them may have resulted only trated by the end of the story, when the goats take over milk from transfers from other Cypriot sites. In any case, figure 1 production from the sheep, which are then used for meat and clearly shows evidence of intensive movements of animals hair production. This scenario provides a good illustration of either from the mainland or from other Cypriot sites through- both a process of domestication (goat) and the early tech- out this period, a phenomenon that has until now been largely noeconomic complementarities of sheep and goat during the unrecognized in the PPN of the Near East. As evidenced by tenth millennium. several papers in this volume, mobility is observed in nu- merous regions of the world at the time of the Neolithization. Introduction of Mammals from the Mainland In addition to the fallow deer and several waves of domestic Shillourokambos: A Case Study for sheep, we also found evidence for the introduction of new Incipient Farming in the Near East lineages of domestic pig and cattle in the course of the tenth millennium (figs. 1, 2; Guilaine, Briois, and Vigne 2011). Be- tween Early phases A and B of Shillourokambos, the size Development of Meat Production: From Low Level variability of suids abruptly tends toward smaller values and to Farming becomes organized into two peaks. Geometric-morphomet- We estimated meat production on the basis of bone weights rics analyses of the outline of the molars also suggest the (Vigne 1992). For the species that were exploited through presence of two distinct types of shape. During the Middle both hunting and breeding (suids, goats), we drew from the and Late phases, the two different-sized peaks tend to fuse frequencies of age classes of the slaughtering profiles both because of interbreeding between the two lineages, but the minimal and maximal estimates of production. The final re- two tooth morphotypes persist. This suggests that another sults are therefore given with a large range of uncertainty, but lineage of pig was introduced shortly before 10,000 BP and they show interesting tendencies (fig. 3). These estimates have then mixed with the local lineage that had been introduced two other weaknesses: (1) although attested to for sheep and beginning with the twelfth millennium. As the former was goats at Shillourokambos (Vigne and Helmer 2007), dairying still smaller than the latter, we concluded that it was composed is not taken into account; and (2) the percentages of bone of domestic types coming from the mainland, where they are weights do not represent an actual value of meat weight (i.e., Vigneetal. Mammal Domestication in the Near East S265

Evolution of the Technical System: Instabilities and Opportunistic Strategies Marine resources played only a small role in the animal supply of the Shillourokambos villagers. Fishing was represented only by some large groupers (Epinephelus sp.; Desse and Desse- Berset 2003), and most of the seashells collected were battered, worn, or naturally pierced—that is, already dead on the beach—for ornaments and pendants (Serrand, Vigne, and Guilaine 2005). Most of the protein and lipid supply came from large terrestrial mammals. Botanical data are scarce (Guilaine, Briois, and Vigne 2011; Willcox 2003). During the Early phase A1, the meat supply came mainly from hunting the small autochthonous Cypriot wild boar. The kill-off profile is abnormally rich for the very old individuals, indicating that the suid population was subjected to low pre- dation pressures. As humans were indeed the only large pred- Figure 3. Evolution of meat production through the chronolog- ator on the island, this suggests hunting of abundant and ical phases of Shillourokambos. Bone weight percentages are not easy-to-catch populations or control, similar to that which exact estimates of meat procurement. Only the relative variations entailed the introduction of wild boar to the island approx- between the different phases can be taken into consideration in this diagram. This is why ordinates are not graduated. EPPNB p imately two millennia earlier. Accessibility to wild/feral goats Early Pre-Pottery Neolithic B; MPPNB p Middle Pre-Pottery was probably lower because of their natural mountain be- Neolithic B; LPPNB p Late Pre-Pottery Neolithic B. havior. Movement of humans between coastal and mountain game cannot be excluded. The apparently extreme speciali- zation of the meat supply of the Early phase A appears, there- 50% does not mean that hunting provided 50% of the meat). fore, as an opportunistic adaptation to both the island faunal We can take into consideration only the relative value for each resources and the coastal-plain environments of the site. In phase with reference to the others. The least unreliable ref- addition, cattle were actually bred on the site, and although erence is the Late phase, where we can reasonably consider tenuous, the evidence suggests that people had already begun that breeding produced all the meat for all mammal species to domesticate some suids. Thus, these (perhaps mobile) hu- except deer, for which the rate of production was 60%, 70%, man groups appear as mostly opportunistic hunters profiting or 76%, depending on the mode of estimation (weight of from a favorable island situation and practicing hunting/con- meat and offal, according to Vigne [1992], NISP, or bone trol and breeding as well. Any attempt to classify them as weight, respectively). strictly hunters or farmers would fail. Figure 3 indicates that meat production increased slowly This image seems to agree with the data from the mainland and irregularly but constantly over time, especially starting in (Goring-Morris and Belfer-Cohen 2011; Zeder 2011). There the Middle phases. It always remained below 60%–75% and is no indication of animal breeding in the southern Levant was probably less than 50% during the Early phases. However, for the middle of the eleventh millennium, and data collected it was probably significant as early as the earliest phases, then for southeastern Anatolia and the northern Levant indicate represented by cattle and an unclear proportion of the suids that early domestic goats, sheep, pigs, and cattle actually first (fig. 2). Even if milk production is added, this development appeared at that time (Helmer et al. 2005; Peters, von den implies a long stage of low-level food production (see Smith Driesch, and Helmer 2005) but that well-controlled hunting 2001) with important variations that suggest frequent mod- of hemiones, gazelles, wild boar, aurochs, and wild sheep and ifications of the technoeconomic strategies and then a slow goats still provided the major part of the meat supply (Vigne transition to farming. The former took place during the main- and Helmer 2007). Only the absence of sheep at Shillouro- land MPPNB, when numerous sites on the mainland also kambos Early A (and at Mylouthkia; Peltenburg et al. 2001b) show a dominance of hunting/control in their meat pro- is surprising, and it indicates that the processes of domesti- curement (Conolly et al. 2011; Hongo et al. 2009; Vigne and cation of sheep and goats in the Near East resulted from two Helmer 2007). The latter closely corresponds to the end of very different processes, as already shown by Zeder (2005) the MPPNB and the LPPNB. This evolution fits the macro- for the beginning of their domestication in Iran. regional historical tendency rather well (see Vigne 2008). Shil- Early phase B differs highly from the previous one in terms lourokambos can be taken as an example of the numerous of animal exploitation (fig. 2). The introduction of domestic local scenarios that led to farming in the Near East (Bar-Yosef pigs and sheep, immediately bred with sophisticated tech- and Meadow 1995) and in numerous regions, as evidenced niques, indicate that Cyprus then benefited from some of the during the Me´rida conference. late-MPPNB technical innovations of the mainland. As the S266 Current Anthropology Volume 52, Supplement 4, October 2011 local goats were quite abundant, the absence of goat intro- supported by the increase in technical efforts that appear duction and herding reveals a selection among the potential during the Early phase C. From the end of the Early phase mainland inputs that depended on local availability. Meat B onward, the archaeozoological data indeed suggest an in- procurement also became more eclectic (with reference to the tensification of henceforth well-controlled pig breeding fo- low taxonomic diversity) than during the Early phase A: all cused on the production of fat and meat by winter slaugh- the species present—suids, goats, and cattle, but also newly tering of first- and second-year pigs. For the first time since introduced fallow deer and sheep—were exploited, and they the beginning of the tenth millennium, the NISP of pigs, contributed almost in the same proportion, complementing which mainly result from breeding, exceeds that of deer. The the existing subsistence system. In parallel with the Middle presence of cattle clearly began to decline, perhaps only be- PPNB evolution on the mainland, animal husbandry in- cause cattle breeding is more difficult and less immediately creased in diversity and probably played an increasing buf- profitable than pig breeding. fering role in the seasonal subsistence, with a higher propor- Until Middle phase A1, sheep herding continued at a high tion of cattle; with the beginning of seasonal pig breeding, level. Then there was a strong and rapid decrease in the size which very quickly mixed autochthonous and new alloch- of the sheep, together with increasing frequency of environ- thonous lineages; and in general with sophisticated sheep mental stress marks on the horn cores and an extreme slen- herding for milk and meat. The latter practice, in which lambs derness of the long bones. At the same time, pigs also de- are kept alive but shared among the ewes for part of the time creased in size, and there was a significant increase in the rate (Vigne and Helmer 2007), implies a high technical level for of linear enamel hypoplasia, which indicates an intensification separate herding of young and adults, particular attention to of environmental stress. New innovations and important lambs, and processing of milk and its derivatives. For the first changes in animal control seem to have emerged very rapidly half of the tenth millennium, this has been attested to on the to compensate for this probable breeding failure. These were mainland only for goats at Cafer Ho¨yu¨k, Tell Halula, and Tell based both on imports from mainland “rear bases” and on Aswad (Gourichon and Helmer 2008; Helmer 2008; Vigne local improvements of the technical system. As mentioned and Helmer 2007). above, a new, larger domestic sheep lineage arrived from the However, hunting remained the main source of meat dur- central or southern Levant as well as a new cattle lineage ing this Early phase B. Almost immediately after its intro- (which might be at the origin of the slight increase of cattle duction, the Mesopotamian fallow deer became the major frequencies that we observed at that time). But the restoration source of meat through a very efficient strategy of exploita- of the system was also based on local resources, not through tion, which was probably collective hunting by beating of hunting, as the deer and then pig populations declined, but mixed herds of males, females, and young (Vigne 2000), quick through the new domestication of the local wild goats (see butchering of carcasses on the kill site, and selection of the above). Here again, these Cypriot populations appear to have most profitable joints. Wild pigs and goats were also subject been involved in a macroregional system of exchanges and to intensive hunting based on similar practices. Cooking tech- progress, but they were also able to provide local responses niques indicate this intensive exploitation (bone breakage for to the ongoing consolidation of the new Neolithic way of life. the consumption of marrow) with relatively standardized and During the Middle phase A2, there followed a complete refined practices (different cutting patterns for suids and ru- remodeling of the utilization of animal resources: besides deer minants, skilled cutting, roasting for some joints and boiling (and secondarily goat) hunting, which persisted, seasonal pig for others). Evidence of seasonal exploitation suggests that breeding, which provided fat, is dominant; from then on, deer were hunted in fall and winter, suids were slaughtered domestic goats progressively played the role of milk produc- in winter, and spring and early summer were devoted to lambs ers, as they usually did on the mainland (Vigne and Helmer and milk exploitation, together with plant cultivation. 2007). Complementarily, the restored sheep herding tends to However, in that period, and again more clearly during the be devoted to meat production, then also to hair exploitation, Early C and Middle phases, clear evidence of overexploitation a tendency that is also known from some pre-Halaf conti- of game appears mainly in the form of a demographic erosion nental sites, such as Tell Sotto and El Kowm 2 (Helmer, Gou- of deer populations. This reminds us that island resources richon, and Vila 2007). The increase of meat production at were not as diversified and abundant as those on the main- that time is demonstrated by a 25-fold increase in the density land, where MPPNB villages still practiced diversified hunting of animal food refuse (in relation to the volume of excavated of wild equids, gazelles, and aurochs. In this context, the earth) in comparison with the early phases. This development development of skilled breeding of cattle, pigs, and sheep in appears to be a parallel to the intensification of animal breed- Cyprus appears not only as a counterpoint to the ongoing ing on the mainland at the transition between the MPPNB evolution on the mainland but also as an answer to the local and LPPNB, which corresponds to the emergence of farming necessity of dealing with both the limitation of island re- (Vigne 2008). sources and the probably increasing population of these pi- On the other hand, this emphasizes the relatively low level oneer human communities. of animal products in the diet of the early phases at Shil- This vision of the dietary history of Shillourokambos is lourokambos. The decrease in grinding stones at the transition Vigneetal. Mammal Domestication in the Near East S267 between the Early and Middle phases (Perrin 2011) suggests in each of the different regions or even in each site, different that plant cultivation, which is well documented at Mylouth- scenarios and rates of intensification of exploitation existed, kia (Peltenburg et al. 2000, 2001a), would have played a more from control of wild/feral populations to sophisticated breed- important role in the diet of the early phases at Shillouro- ing that entailed the morphological changes that characterize kambos than suggested by the relatively poor set of botanical domestic lineages; (4) the beginnings of domestication, that remains (Willcox 2003). The strong development of animal is, control in a wide area extending from the Iranian Plateau husbandry in the middle of the tenth millennium may be to the Mediterranean shores (though ancillary to the conti- connected to a drastic change from a mostly vegetarian diet nent through the introduction of wild or domestic taxa, Cy- to a more carnivorous one. It is also probably connected to prus was part of this area, with special Cypriot innovations a population increase in the village, as seen in Sector 3 of the such as intensive hunting of fallow deer and local domesti- site, the study of which is still ongoing. cation of the wild/feral goat); (5) intensive and long-distance In summary, the history of meat supply at Shillourokambos exchanges of raw materials, including not only obsidian (Bri- comes within the scope of the PPNB Near Eastern Neolithic ois, Gratuze, and Guilaine 1997) but also living animals across transition. This is seen in the exploitation of local animals by large continental and marine areas (such interaction explains hunting and control in local domestication as well as by breed- the relative homogeneity of the PPN cultural sphere); and (6) ing, which led to the slow but constant intensification of food social organization and technical skills in the Near Eastern production relative to predation, and in genetic inputs from Final Late Glacial and Early Holocene societies that were more the mainland in the form of new lineages of domestic pigs, sophisticated than generally believed: in addition to rapid and sheep, and perhaps cattle. In comparison with the earliest high-level improvements in lithic (Astruc, Binder, and Briois known Anatolian sites with domestic animals, Shillouro- 2007) and construction (Schmidt 2003; Stordeur et al. 2000) kambos differs in its slightly later manifestation of local do- technologies, these groups were able to regularly and more mestication processes (pigs and goats), which are possibly and more intensively navigate to Cyprus (probably by sailing; adaptations to the local conditions, because of a low and Vigne 2009) and to control animal populations using diver- insular biological diversity. The emergence of Cypriot animal sified and sophisticated techniques (for meat, milk, hair, and production appears to be a local version of that seen on the draft products). Illustrating a part of the socioeconomic and mainland, driven by resource shortage and human demo- historical complexity involved in the beginning of the Neo- graphic growth (Bocquet-Appel 2011). This interpretation of lithic in the Near East, Cyprus also emphasizes the multifac- the subsistence strategy of Shillourokambos implies a knowl- torial causes of the transition, including, of course, climatic edge of seafaring (Broodbank 2006; Vigne and Cucchi 2005), and environmental conditions (Bar-Yosef 2011) but also the but this was not sufficient for sustainable development of dominant influence of demographic (Bocquet-Appel 2011), island populations; it also required a high level of control of technical, and social threshold effects. the sources of subsistence that could be attained only through the Neolithic way of life and strong connections to continental “rear bases.” Acknowledgments Conclusions We thank the Wenner-Gren Foundation and the organizers of the Me´rida conference, who invited J.-D. Vigne. We also Though still poorly documented for the thirteenth–eleventh thank all the archaeozoologists who provided us with often millennia, the prepottery history of the relationships between unpublished data: Paul Croft, Simon Davis, Sheelagh Frame, mammals and humans on Cyprus provides substantial new Lionel Gourichon, and Daniel Helmer. Elizabeth Willcox and information about the beginnings of domestication in the Doug Price greatly improved our English-language writing. Near East. In that context, the recent Cypriot data document We are grateful to all the scientists and excavators who con- the following: (1) a transition to farming during the tenth tributed to the study of this site. The French School at Athens, millennium, after an unstable and opportunistic late-EPPBN the Cyprus Department of Antiquities, and the French Min- and MPPNB phase of low-level food production based on istry of Foreign Affairs supported our stays for fieldwork and rapidly changing combinations of hunting, control, and analysis in Cyprus. breeding; (2) initiation of this process during the Final Late Glacial, evidenced in Cyprus by control and long-distance transport (and thus acclimatization) of wild boars before the References Cited mid-twelfth millennium (Natufian-Khiamian) followed by Ammerman, Albert J., Pavlos Flourentzos, R. Smadar Gabrieli, the introduction of domestic dogs, commensal mice, and wild Thomas Higham, Carole McCartney, and Tim Turnbull. 2008. or early domestic goats, cats, and cattle during the eleventh Third report on early sites on Cyprus. Report of the Department of Antiquities, Cyprus 2008:1–32. millennium (late PPNA, EPPNB) and the subsequent addition Ammerman, Albert J., Pavlos Flourentzos, Carole McCartney, Jay S. of wild fallow deer, foxes, and domestic sheep at the beginning Noller, and Daniel Sorabji. 2006. Two new early sites on Cyprus. of the tenth millennium (MPPNB); (3) for each of the species Report of the Department of Antiquities, Cyprus 2007:1–22. S268 Current Anthropology Volume 52, Supplement 4, October 2011

Ammerman, Albert J., and Jay S. Noller. 2005. New light on Aeto- Cucchi, Thomas, Annie Orth, Jean-Christophe Auffray, Sabrina Re- kremnos. World Archaeology 37(4):533–543. naud, Laurent Fabre, Josette Catalan, Eleftherios Hadjisterkotis, Astruc, Laurence, Didier Binder, and Franc¸ois Briois, eds. 2007. Sys- Franc¸ois Bonhomme, and Jean-Denis Vigne. 2006. A new endemic te`mes techniques et communaute´s du Ne´olithique pre´ce´ramique au species of the subgenus Mus (Rodentia, Mammalia) on the Island Proche-Orient. Antibes, France: APDCA. of Cyprus. Zootaxa 1241:1–36. Auffray, Jean-Christophe, Eitan Tchernov, and Eviatar Nevo. 1988. Cucchi, Thomas, and Jean-Denis Vigne. 2006. Origin and diffusion Origine du commensalisme de la souris domestique (Mus musculus of the house mouse in the Mediterranean. Human Evolution 21(2): domesticus) vis-a`-vis de l’homme. Comptes Rendus de l’Acade´mie 95–106. des Sciences de Paris se´r. III: Sciences de la Vie 307(9):517–522. Cucchi, Thomas, Jean-Denis Vigne, Jean-Christophe Auffray, Paul Balasse, Muriel, and Stanley H. Ambrose. 2005. Distinguishing sheep Croft, and Edgar Peltenburg. 2002. Introduction involontaire de and goats using dental morphology and stable carbon isotopes in la souris domestique (Mus musculus domesticus)a` Chypre de`s le

C4 grassland environments. Journal of Archaeological Science 32(5): Ne´olithique pre´ce´ramique ancien (fin IXe et VIIIe mille´naires av. 691–702. J.-C.). Comptes Rendus de l’Acade´mie des Sciences, Pale´vol 1(4): Bar-Yosef, Ofer. 2011. Climatic fluctuations and early farming in West 235–241. and East Asia. Current Anthropology 52(suppl. 4):S175–S193. Davis, Simon J. M. 1994. Even more bones from Khirokitia: the Bar-Yosef, Ofer, and Richard H. Meadow. 1995. The origin of ag- 1988–1991 excavations. In Fouilles re´centes a` Khirokitia (Chypre) riculture in the Near East. In Last hunters, first farmers. T. Douglas 1988–1991. Alain Le Brun, ed. Pp. 305–334. E´ ditions Recherches Price and Anne Birgitte Gebauer, eds. Pp. 39–94. Santa Fe, NM: sur les Civilisations. Paris: ADPF. School of American Research Press. ———. 2003. The zooarchaeology of Khirokitia (Neolithic Cyprus), Binford, Lewis. 2000. Review of Faunal extinctions in an island society: including a view from the mainland. In Le Ne´olithique de Chypre: pygmy hippopotamus hunters of Cyprus, by Alan H. Simmons. actes du colloque international organise´par le De´partement des An- American Antiquity 65(4):771. tiquite´s de Chypre et l’E´cole Franc¸aise d’Athe`nes, Nicosie 17–19 mai Bocquet-Appel, Jean-Pierre. 2011. The agricultural demographic 2001. Jean Guilaine and Alain Le Brun, eds. Pp. 253–278. Bulletin transition during and after the agriculture inventions. Current An- de Correspondance Helle´nique, supple´ment 43. Athens: E´ cole Fran- thropology 52(suppl. 4):S497–S510. c¸aise d’Athe`nes. Boekschoten, G. J., and Paul Y. Sondaar. 1972. On the fossil mam- ———. 2005. Why domesticate food animals? some zoo-archaeo- malia of Cyprus. Proceedings of the Kononklijke Nederlandse Aka- logical evidence from the Levant. Journal of Archaeological Science demie van Wetenschappen B 75(4):306–308. 32(9):1408–1416. Briois, Franc¸ois 2003. Nature et e´volution des industries lithiques de Desse, Jean, and Nathalie Desse-Berset. 2003. Les premiers peˆcheurs Shillourokambos. In Le Ne´olithique de Chypre: actes du colloque de Chypre. In Le Ne´olithique de Chypre: actes du colloque inter- international organise´par le De´partement des Antiquite´s de Chypre national organise´par le De´partement des Antiquite´s de Chypre et et l’E´cole Franc¸aise d’Athe`nes, Nicosie 17–19 mai 2001. Jean Guilaine l’E´cole Franc¸aise d’Athe`nes, Nicosie 17–19 mai 2001. Jean Guilaine and Alain Le Brun, eds. Pp. 121–133. Bulletin de Correspondance and Alain Le Brun, eds. Pp. 279–291. Bulletin de Correspondance Helle´nique, supple´ment 43. Athens: E´ cole Franc¸aise d’Athe`nes. Helle´nique, supple´ment 43. Athens: E´ cole Franc¸aise d’Athe`nes. Briois, Franc¸ois, Bernard Gratuze, and Jean Guilaine. 1997. Obsi- Dobney, Keith, and Greger Larson. 2006. Genetics and animal do- diennes du site ne´olithique pre´ce´ramique de Shillourokambos mestication: new windows on an elusive process. Journal of Zoology (Chypre). Pale´orient 23(1):95–112. 269(2):261–271. Briois, Franc¸ois, Catherine Petit-Aupert, and Pierre-Yves Pe´choux. Ducos, Pierre. 1968. L’origine des animaux domestique en Palestine. 2005. Histoire des campagnes d’Amathonte. I. L’occupation du sol Bordeaux: Institut de Pre´histoire de l’Universite´. au Ne´olithique.E´ tudes Chypriotes 16. Athens: E´ cole Franc¸aise Ervynck, Anton, Keith Dobney, Hitomi Hongo, and Richard d’Athe`nes. Meadow. 2001. Born free? new evidence for the status of pigs from Bromage, Timothy G., Wendy Dirks, Hediye Erdjument-Bromage, C¸ayo¨nu¨ Tepesi, eastern Anatolia. Pale´orient 27(2):47–73. Mathias Huck, Ottmar Kulmer, Rasime O¨ ner, Oliver Sandrock, Goring-Morris, A. Nigel, and Anna Belfer-Cohen. 2011. Neolithi- and Friedemann Schrenk. 2002. A life history and climate change zation processes in the Levant: the outer envelope. Current An- solution to the evolution and extinction of insular dwarves: a thropology 52(suppl. 4):S195–S208. Cypriot experience. In World islands in prehistory: international Gourichon, Lionel, and Daniel Helmer. 2008. Etude arche´ozoolo- insular investigations. William Waldren and Josep Ensenyat, eds. gique de Mureybet. In Le site ne´olithique de Tell Mureybet (Syrie Pp. 420–427. British Archaeological Reports, International Series du Nord): en hommage a`Jacques Cauvin. Juan Jose´Iban˜ez, ed. Pp. 1095. Oxford: Archaeopress. 115–227. British Archaeological Reports International Series 1843. Broodbank, Cyprian. 2006. The origins and early development of Oxford: Archaeopress. Mediterranean maritime activity. Journal of Mediterranean Ar- Guilaine, Jean. 2003. Parekklisha-Shillourokambos. Pe´riodisation et chaeology 19(2):199–230. ame´nagements domestiques. In Le Ne´olithique de Chypre: actes du Bunimovitz, Shlomo, and Ran Barkai. 1996. Ancient bones and mod- colloque international organise´par le De´partement des Antiquite´s de ern myths: ninth millennium BC hippopotamus hunters at Ak- Chypre et l’E´cole Franc¸aise d’Athe`nes, Nicosie 17–19 mai 2001.Jean rotiri Aetokremnos, Cyprus. Journal of Mediterranean Archaeology Guilaine and Alain Le Brun, eds. Pp. 3–14. Bulletin de Corres- 9(1):85–96. pondance Helle´nique, supple´ment 43. Athens: E´ cole Franc¸aise Cherry, John F. 1990. The first colonization of the Mediterranean d’Athe`nes. islands: a review of recent research. Journal of Mediterranean Ar- Guilaine, Jean, and Franc¸ois Briois. 2007. Shillourokambos and the chaeology 3(2):145–221. Neolithization of Cyprus: some reflections. Eurasian Prehistory Clutton-Brock, Juliet. 1979. The mammalian remains from the Jer- 4(1–2):159–175. icho Tell. Proceedings of the Prehistoric Society 45:135–157. Guilaine, Jean, Franc¸ois Briois, and Jean-Denis Vigne, eds. 2011. Conolly, James, Sue Colledge, Keith Dobney, Jean-Denis Vigne, Joris Shillourokambos: un e´tablissement ne´olithique pre´-ce´ramique a`Chy- Peters, Barbara Stopp, Katie Manning, and Stephen Shennan. 2011. pre: les fouilles du secteur 1. Paris: Errance–E´ cole Franc¸aise Meta-analysis of zooarchaeological data from SW Asia and SE d’Athe`nes. Europe provides insight into the origins and spread of animal Guilaine, Jean, Franc¸ois Briois, Jean-Denis Vigne, and Isabelle Car- husbandry. Journal of Archaeological Science 38(3):538–545. re`re. 2000. De´couverte d’un Ne´olithique pre´ce´ramique ancien chy- Vigneetal. Mammal Domestication in the Near East S269

priote (fin 9e de´but 8e mille´naires cal. BC), apparente´ au PPNB from the mainland. In Neolithic revolution: new perspectives on ancien/moyen du Levant nord. Comptes Rendus de l’Acade´mie des Southwest Asia in light of recent discoveries on Cyprus. Edgar Pel- Sciences, se´r. IIA: Sciences de la Terre et des Plane`tes 330(1):75–82. tenburg and Alexander Wasse, eds. Pp. 35–48. Levant Supple- Guilaine, Jean, Franc¸ois Briois, Jean-Denis Vigne, Thomas Perrin, mentary Series 1. Oxford: Oxbow. Yann Be´liez, Isabelle Carre`re, Patrice Ge´rard, and Maxime Re´- Hours, Francis, Olivier Aurenche, Jacques Cauvin, Marie-Claire Cau- micourt. 2008. Shillourokambos (Paerekklisha): l’e´tablissement vin, Lorraine Copeland, and Paul Sanlaville. 1994. Atlas des sites ne´olithique pre´ce´ramique: e´tudes et rapports 2.1. Bulletin de Cor- du Proche-Orient (ASPO): 1400–5700 BP. Travaux de la Maison respondance Helle´nique 128–129(2):1006–1021. de l’Orient 24. Lyon: Maison de l’Orient Me´diterrane´en. Guilaine, Jean, Philippe Deve`ze, Jacques Coularou, and Franc¸ois Bri- Khalaily, Hamudi, Ofer Marder, and Omry Barzilai. 2007. An early ois. 1999. Teˆte sculpte´e en pierre dans le Ne´olithique pre´ce´ramique Pre-Pottery Neolithic B blade cache from Motza, west of Jerusalem, de Shillourokambos (Parekklisha. Chypre). Report of the Depart- Israel. In Syste`mes techniques et communaute´s du Ne´olithique pre´- ment of Antiquities, Cyprus 1999:1–12. ce´ramique au Proche-Orient. Laurence Astruc, Didier Binder, and Guilaine, Jean, and Alain Le Brun, eds. 2003. Le Ne´olithique de Chy- Franc¸ois Briois, eds. Pp. 269–276. Antibes, France: APDCA. pre: actes du colloque international organise´par le De´partement des Kuijt, Ian. 2008. Demography and storage systems during the south- Antiquite´s de Chypre et l’E´cole Franc¸aise d’Athe`nes, Nicosie 17–19 ern Levantine Neolithic demographic transition. In The Neolithic mai 2001. Bulletin de Correspondance Helle´nique, supple´ment 43. demographic transition and its consequences. Jean-Pierre Bocquet- Athens: E´ cole Franc¸aise d’Athe`nes. Appel and Ofer Bar-Yosef, eds. Pp. 287–313. New York: Springer. Held, Steven O. 1989. Colonization cycles on Cyprus. 1. The bio- Kuijt, Ian, and Bill Finlayson. 2009. Evidence for food storage and geographic and paleontological foundations of early prehistoric predomestication granaries 11,000 years ago in the Jordan Valley. settlement. Report of the Department of Antiquities, Cyprus 1989: Proceedings of the National Academy of Sciences of the USA 106(27): 7–28. 10966–10970. Helmer, Daniel. 1992. La domestication des animaux par l’homme Le Brun, Alain, and Odile Daune-Le Brun. 2003. Deux aspects du pre´historique. Paris: Masson. Ne´olithique pre´-ce´ramique re´cent de Chypre: Khirokitia et Cap ———. 2008. Re´vision de la faune de Cafer Ho¨yu¨k (Malatya, Tur- Andreas-Kastros. In Le Ne´olithique de Chypre: actes du colloque quie): apports des me´thodes de l’analyse des me´langes et de international organise´par le De´partement des Antiquite´s de Chypre l’analyse de Kernel a` la mise en e´vidence de la domestication. In et l’E´cole Franc¸aise d’Athe`nes, Nicosie 17–19 mai 2001. Jean Guilaine Archaeozoology of the Near East VIII: proceedings of the 8th Inter- and Alain Le Brun, eds. Pp. 45–59. Bulletin de Correspondance national Symposium on the Archaeozoology of Southwestern Asia Helle´nique, supple´ment 43. Athens: E´ cole Franc¸aise d’Athe`nes. and Adjacent Areas. Emannuelle Vila, Lionel Gourichon, Alice M. Legge, Anthony J. 1996. The beginning of caprine domestication in Choyke, and Hijlke Buitenhuis, eds. Pp. 169–195. Travaux de la Southwest Asia. In The origins and spread of agriculture and pas- Maison de l’Orient 49. Lyon: Maison de l’Orient et de la Me´di- toralism in Eurasia. David R. Harris, ed. Pp. 238–262. Washington, terrane´e. DC: Smithsonian Institution. Helmer, Daniel, and Lionel Gourichon. 2008. Premie`res donne´es sur Le´vi-Strauss, Claude. 1958. Anthropologie structurale. Agora 7. Paris: les modalite´s de subsistance dans les niveaux re´cents (PPNB moyen Plon. a`Ne´olithique a` Poterie) de Tell Aswad en Damasce`ne (Syrie): Malek, Jaromir. 1993. The cat in ancient Egypt. London: British Mu- fouilles 2001–2005. In Archaeozoology of the Near East VIII: pro- seum. ceedings of the 8th International Symposium on the Archaeozoology Manning, Sturt W., Carole McCartney, Bernd Kromer, and Sarah T. of Southwestern Asia and Adjacent Areas. Emannuelle Vila, Lionel Stewart. 2010. The earlier Neolithic in Cyprus: recognition and Gourichon, Alice M. Choyke, and Hijlke Buitenhuis, eds. Pp. 120– dating of a Pre-Pottery Neolithic A occupation. Antiquity 84(325): 151. Travaux de la Maison de l’Orient 49. Lyon: Maison de l’Orient 693–706. et de la Me´diterrane´e. Marom, Nimrod, and Guy Bar-Oz. 2009. Culling profiles: the in- Helmer, Daniel, Lionel Gourichon, Herve´Monchot, Joris Peters, and determinacy of archaeozoological data to survivorship curve mod- Maria San˜a Segui. 2005. Identifying early domestic cattle from elling of sheep and goat herd maintenance strategies. Journal of Pre-Pottery Neolithic sites on the Middle Euphrates using sexual Archaeological Science 36(5):1184–1187. dimorphism. In The first steps of animal domestication: new ar- Mauss, Marcel. 1947 (1967). Manuel d’ethnographie. Petite Biblio- chaeobiological approaches. Jean-Denis Vigne, Joris Peters, and the`que Payot. Lausanne, Switzerland: Payot. Daniel Helmer, eds. Pp. 86–95. London: Oxbow. McCartney, Carole. 2010. Outside the corridor? the Neolithisation Helmer, Daniel, Lionel Gourichon, and Danielle Stordeur. 2004. A` of Cyprus. In The development of pre-state communities in the l’aube de la domestication animale: imaginaire et symbolisme an- ancient Near East: studies in honour of Edgar Petenburg. Diane imal dans les premie`res socie´te´s ne´olithiques du nord du Proche- Bolger and Louise Maguire, eds. Pp. 185–196. British Association Orient. Anthropozoologica 39(1):143–163. for Near Eastern Archaeology Publication Series 2. Oxford: Oxbow. Helmer, Daniel, Lionel Gourichon, and Emmanuelle Vila. 2007. The McCartney, Carole, Sturt W. Manning, Sandra Rosendahl, and Sarah development of the exploitation of products from Capra and Ovis T. Stewart. 2008. Elaborating Early Neolithic Cyprus (EENC): pre- (meat, milk and fleece) from the PPNB to the Early Bronze in the liminary report on the 2007 field season: excavations and regional northern Near East (8700 to 2000 BC cal.). Anthropozoologica field survey at Agia Varvara-Asprokremnos. Report of the Depart- 42(2):41–69. ment of Antiquities, Cyprus 2008:67–86. Hole, Frank. 1996. The context of caprine domestication in the Zag- McCartney, Carole, Sturt W. Manning, David Sewell, and Sarah T. ros region. In The origins and spread of agriculture and pastoralism Stewart. 2007. The EENC 2006 field season: excavations at Agia in Eurasia. David R. Harris, ed. Pp. 263–281. Washington, DC: Varvara-Asprokremnos and survey of the local Early Holocene Smithsonian Institution. landscape. Report of the Department of Antiquities, Cyprus 2007: Hongo, Hitomi, Jessica Pearson, Banu O¨ ksu¨z, and Gu¨lc¸in I˙lgezdi. 27–44. 2009. The process of ungulate domestication at C¸ayo¨nu¨, south- ———. 2010. Reconsidering Early Holocene Cyprus within the east- eastern Turkey: a multidisciplinary approach focusing on Bos sp. ern Mediterranean landscape. In Landscapes in transition. Bill Fin- and Cervus elaphus. Anthropozoologica 44(1):63–73. layson and Graeme Warren, eds. Pp. 133–146. Levant Supple- Horwitz, Liora K., Eitan Tchernov, and Hitomi Hongo. 2004. The mentary Series 8. Oxford: Oxbow. domestic status of the Early Neolithic fauna of Cyprus: a view Monchot, Herve´, and Jacques Le´chelle. 2002. Statistical nonpara- S270 Current Anthropology Volume 52, Supplement 4, October 2011

metrics methods for the study of fossil populations. Paleobiology Sarah M. Nelson, ed. Pp. 55–64. MASCA Research Papers in Sci- 28(1):55–69. ence and Archaeology 15. Philadelphia: University of Pennsylvania Muxart, Tatiana, Franck-Dominique Vivien, Bruno Villalba, and Museum. Joe¨lle Burnouf, eds. 2003. Des milieux et des hommes: fragments San˜a Segui, Maria. 1999. Arqueologı´a de la domesticacı´on animal: la d’histoires croise´es. Paris: Elsevier. gestio´n de los recursos animales en Tell Halula (Valle del E´ufrates, Naderi, Saeid, Hamid-Reza Rezaei, Franc¸ois Pompanon, Michael G. Siria) de 8.800 al 7.000 BP. Treballs d’Arqueologia del Pro`xim B. Blum, Riccardo Negrini, Hamid-Reza Naghash, O¨ zge Balkız, et Orient 1. Barcelona: Universitat Auto`noma de Barcelona. al. 2008. The goat domestication process inferred from large-scale Schmidt, Klaus. 2003. “Kraniche am See”: Bilder und Zeichen vom mitochondrial DNA analysis of wild and domestic individuals. fru¨hneolithischen Go¨bekli Tepe (Su¨dosttu¨rkei). In Der Turmbau Proceedings of the National Academy of Sciences of the USA 105(46): zu Babel: Ursprung und Vielfalt von Sprache und Schrift. Wilfried 17659–17664. Seipel, ed. Pp. 23–29. Vienna: Kunsthistorischen Museums. Olsen, Sandra L. 1999. Investigation of the Phanourios bones for Serrand, Nathalie, Jean-Denis Vigne, and Jean Guilaine. 2005. Early evidence of cultural modification. In Faunal extinction in an island Preceramic Neolithic marine shells from Shillourokambos, Cyprus society: pygmy hippopotamus hunters of Cyprus. Alan H. Simmons, (late 9th–8th mill. cal BC): a mainly ornamental set with simi- ed. Pp. 230–238. New York: Kluwer Academic/Plenum. larities to mainlands PPNB. In Archaeomalacology: molluscs in for- Pascal, Michel, Olivier Lorvelec, and Jean-Denis Vigne. 2006. Inva- mer environments of human behaviour. Daniella Bar-Yosef Mayer, sions biologiques et extinctions: 11,000 ans d’histoire des verte´bre´s en ed. Pp. 122–129. Oxford: Oxbow. France. Paris: Belin. Simmons, Alan H. 1988. Extinct pygmy hippopotamus and early Peltenburg, Edgar. 2003. The colonisation and settlement of Cyprus: man in Cyprus. Nature 333(6173):554–557. investigations at Kissonerga-Mylouthkia (1976–1996). Studies in ———, ed. 1999. Faunal extinction in an island society: pygmy hip- Mediterranean Archaeology. Sa¨vedalen, Sweden: Astro¨ms. popotamus hunters of Cyprus. New York: Kluwer Academic/Ple- Peltenburg, Edgar, Sue Colledge, Paul Croft, Adam Jackson, Carole num. McCartney, and Mary Anne Murray. 2000. Agro-pastoralist col- Simmons, Alan H., and Rolfe Mandel. 2007. Not such a new light: onization of Cyprus in the 10th millennium BP: initial assessments. a response to Ammerman and Noller. World Archaeology 39(4): Antiquity 74(286):844–853. 475–482. ———. 2001a. Neolithic dispersals from the Levantine corridor: a Smith, Bruce D. 2001. Low level food production. Journal of Ar- Mediterranean perspective. Levant 33:35–64. chaeological Research 9(1):1–43. ———. 2001b. Well-established colonists: Mylouthkia 1 and the Starkovich, Britt M., and Mary C. Stiner. 2009. Hallan C¸emi Tepesi: cypro-pottery Neolithic B. In The earliest prehistory of Cyprus: from high-ranked game exploitation alongside intensive seed processing colonization to exploitation. Stuart Swiny, ed. Pp. 61–93. Boston: at the Epipaleolithic-Neolithic transition in southeastern Turkey. American Schools of Oriental Research. Anthropozoologica 44(1):41–61. Peltenburg, Edgar, and Alexander Wasse, eds. 2004. Neolithic Revo- Stordeur, Danielle, Michel Brenet, Ge´rard Der Aprahamian, and Jean- lution: new perspectives on Southwest Asia in light of recent discov- Claude Roux. 2000. Les baˆtiments communautaires de Jerf el Ah- eries on Cyprus. Levant Supplementary Series 1. Oxford: Oxbow. mar et Mureybet, Horizon PPNA, Syrie. Pale´orient 26(1):29–44. Pergams, Oliver R. W., and Mary V. Ashley. 2001. Microevolution Tchernov, Eitan, and Franc¸ois R. Valla. 1997. Two new dogs and in island rodents. Genetica 112–113:245–256. other Natufian dogs from the southern Levant. Journal of Ar- Perrin, Thomas. 2011. Les donne´es du macro-outillage. In Shillouro- chaeological Science 24(1):65–95. kambos: un e´tablissement ne´olithique pre´-ce´ramique a` Chypre: les Trantalidou, Katerina. 2008. Glimpses of Aegean Island communities fouilles du secteur 1. Jean Guilaine, Franc¸ois Briois, and Jean-Denis during the Mesolithic and Neolithic Periods: the zooarchaeological Vigne, eds. Paris: Errance–E´ cole Franc¸aise d’Athe`nes. point of view. In A colloquium on the prehistory of the Cyclades. Peters Joris, Angela von den Driesch, and Daniel Helmer. 2005. The Neil Brodie, Jennifer Doole, Giorgos Gavalas, and Colin Renfrew, upper Euphrates-Tigris basin: cradle of agro-pastoralism? In The eds. Pp. 19–27. Cambridge: McDonald Institute. first steps of animal domestication: new archaeobiological approaches. Vigne, Jean-Denis. 1992. The meat and offal weight (MOW) method Jean-Denis Vigne, Joris Peters, and Daniel Helmer, eds. Pp. 96– and the relative proportion of ovicaprines in some ancient meat 124. Oxford: Oxbow. diets of the north-western Mediterranean. Rivista di Studi Liguri Redding, Richard W. 2005. Breaking the mold: a consideration of A 57(2):21–47. variation in the evolution of animal domestication. In The first ———. 1998. Facie`s culturels et sous-syste`me technique de steps of animal domestication: new archaeobiological approaches. l’acquisition des ressources animales: application au Ne´olithique Jean-Denis Vigne, Joris Peters, and Daniel Helmer, eds. Pp. 41– ancien me´diterrane´en. In Production et identite´ culturelle: actualite´ 48. Oxford: Oxbow. de la recherche. Andre´D’Anna and Didier Binder, eds. Pp. 27–45. Redding, Richard W., and Michael Rosenberg. 1998. Ancestral pigs: Antibes, France: APDCA. a New (Guinea) model for pig domestication in the Middle East. ———. 1999. The large “true” Mediterranean as a model for the In Ancestors for the pigs: pigs in prehistory. Sarah M. Nelson, ed. Holocene human impact on the European vertebrate fauna? recent Pp. 65–76. MASCA Research Papers in Science and Archaeology data and new reflections. In The Holocene history of the European 15. Philadelphia: University of Pennsylvania Museum. vertebrate fauna: modern aspects of research. Norbert Benecke, ed. Reimer, Paula J., Mike G. L. Baillie, E´ douard Bard, Alex Bayliss, J. Pp. 295–322. Archa¨ologie in Eurasien 6. Berlin: Deutsches Ar- Warren Beck, Chanda J. H. Bertrand, Paul G. Blackwell, et al. cha¨ologisches Institut. 2004. Intcal04 terrestrial radiocarbon age calibration, 0–26 cal kyr ———. 2000. Outils pour restituer les strate´gies de chasse au cerf BP. Radiocarbon 46(3):1029–1058. en Europe au Me´solithique et au Ne´olithique: analyses graphiques, Rosenberg, Michael, R. Mark Nesbitt, Richard W. Redding, and Brian statistiques et multivarie´es de courbes d’aˆges d’abattage. Ibex Jour- L. Peasnall. 1998. Hallan C¸emi, pig husbandry, and post-Pleisto- nal of Mountain Ecology 5/Anthropozoologica 31:57–67. cene adaptations along the Taurus-Zagros arc (Turkey). Pale´orient ———. 2006. De´coupe du cerf (Cervus elaphus)auMe´solithique 24(1):25–41. moyen, a` Noyen-sur-Seine (Seine-et-Marne): analyses trace´olo- Rosenberg, Michael, and Richard Redding. 1998. Early pig husbandry gique et expe´rimentale. Revue de Pale´obiologie vol. spe´cial 10:69– in Southwestern Asia and its implications for modeling the origins 82. of food production. In Ancestors for the pigs: pigs in prehistory. ———. 2008. Zooarchaeological aspects of the Neolithic diet tran- Vigneetal. Mammal Domestication in the Near East S271

sition in the Near East and Europe, and their putative relationships Vigne, Jean-Denis, and Jean Guilaine. 2004. Les premiers animaux with the Neolithic demographic transition. In The Neolithic de- de compagnie 8500 ans avant notre e`re?...oucomment j’ai mographic transition and its consequences. Jean-Pierre Bocquet Ap- mange´ mon chat, mon chien et mon renard. Anthropozoologica pel and Ofer Bar-Yosef, eds. Pp. 179–205. New York: Springer. 39(1):249–273. ———. 2009. Introductions et re´introduction de mammife`res a`Chy- Vigne, Jean-Denis, Jean Guilaine, Karyne Debue, Laurent Haye, and pre aux IXe et VIIIe mille´naires av. J.-C. (Ne´olithique pre´ce´ra- Patrice Ge´rard. 2004. Early taming of the cat in Cyprus. Science mique): indices indirects de l’usage de la voile au Ne´olithique? In 304(5668):259. De Me´diterrane´e et d’ailleurs. . .: me´langes offerts a` Jean Guilaine. Vigne, Jean-Denis, and Daniel Helmer. 2007. Was milk a “secondary Pp. 807–820. Toulouse, France: Archives d’E´ cologie Pre´historique. product” in the Old World Neolithisation process? its role in the ———. 2011. The origins of animal domestication and husbandry: domestication of cattle, sheep and goats. Anthropozoologica 42(2): a major change in the history of humanity and the biosphere. 9–40. Comptes Rendus de l’Acade´mie des Sciences, Biologie. 334(3):171– Vigne, Jean-Denis, Antoine Zazzo, Jean-Franc¸ois Salie`ge, Franc¸ois 181. Poplin, Jean Guilaine, and Alan Simmons. 2009. Pre-Neolithic wild Vigne, Jean-Denis, Franc¸ois Briois, Antoine Zazzo, Isabelle Carre`re, boar management and introduction to Cyprus more than 11,400 Julie Daujat, and Jean Guilaine. Forthcoming. Preliminary data years ago. Proceedings of the National Academy of Sciences of the on a new Early Pre-Pottery Neolithic site on Cyprus (Ayios Ty- USA 106(38):16131–16138. chonas, Klimonas, ca. 9000 cal. BC). Pale´orient. Wasse, Alexander. 2007. Climate, economy and change: Cyprus and Vigne, Jean-Denis, Isabelle Carre`re, and Jean Guilaine. 2003. Unsta- the Levant during the Late Pleistocene to mid-Holocene. In On ble status of early domestic ungulates in the Near East: the example the margins of Southwest Asia: Cyprus during the 6th to 4th millennia of Shillourokambos (Cyprus. IX–VIIIth millennia cal. B.C.). In Le BC. Joanne Clarke, ed. Pp. 44–63. Oxford: Oxbow. Ne´olithique de Chypre: actes du colloque international organise´par Willcox, George. 2003. The origins of Cypriot farming. In Le Ne´o- le De´partement des Antiquite´s de Chypre et l’E´cole Franc¸aise lithique de Chypre: actes du colloque international organise´par le d’Athe`nes, Nicosie 17–19 mai 2001. Jean Guilaine and Alain Le De´partement des Antiquite´s de Chypre et l’E´cole Franc¸aise d’Athe`nes, Brun, eds. Pp. 239–251. Bulletin de Correspondance Helle´nique, Nicosie 17–19 mai 2001. Jean Guilaine and Alain Le Brun, eds. Pp. supple´ment 43. Athens: E´ cole Franc¸aise d’Athe`nes. 231–238. Bulletin de Correspondance Helle´nique, supple´ment 43. ———. Forthcoming. Evolution of the use of sheep and goat during Athens: E´ cole Franc¸aise d’Athe`nes. the Pre-Pottery Neolithic at Shillourokambos (8400–7000 cal. BC). Zeder, Melinda A. 2005. A view from the Zagros: new perspectives In Archaeozoology of the Near East IX: Proceedings of the 9th In- on livestock domestication in the Fertile Crescent. In The first steps ternational Symposium on the Archaeozoology of Southwestern Asia of animal domestication: new archaeobiological approaches. Jean- and Adjacent Areas. Marjan Mashkour and Mark Beech, eds. Ox- Denis Vigne, Joris Peters, and Daniel Helmer, eds. Pp. 125–146. ford: Oxbow. Oxford: Oxbow. Vigne, Jean-Denis, Isabelle Carre`re, Jean-Franc¸ois Salie`ge, Alain Per- ———. 2006. Central questions in the domestication of plants and son, Herve´ Bocherens, Jean Guilaine, and Franc¸ois Briois. 2000. animals. Evolutionary Anthropology 15(3):105–117. Predomestic cattle, sheep, goat and pig during the late 9th and ———. 2009. The Neolithic macro-(r)evolution: macroevolutionary the 8th millennium cal. BC on Cyprus: preliminary results of theory and the study of culture change. Journal of Archaeological Shillourokambos (Perkklisha, Limassol). In Archaeozoology of the Research 17(1):1–63. Near East IV: proceedings of the 4th International Symposium on ———. 2011. The origins of agriculture in the Near East. Current the Archaeozoology of Southwestern Asia and Adjacent Areas. Marjan Anthropology 52(suppl. 4):S221–S235. Mashkour, Alice M. Choyke, Hijlke Buitenhuis, and Franc¸ois Pop- Zeder, Melinda A., Eve Emshwiller, Bruce D. Smith, and Daniel G. lin, eds. Pp. 52–75. Publicaties 32. Groningen: Centre for Ar- Bradley. 2006. Documenting domestication: the intersection of ge- chaeological Research and Consultancy. netics and archaeology. Trends in Genetics 22(3):139–155. Vigne, Jean-Denis, and Thomas Cucchi. 2005. Premie`res navigations Zeder, Melinda A., and Brian Hesse. 2000. The initial domestication au Proche-Orient: les informations indirectes de Chypre. Pale´orient of goats (Capra hircus) in the Zagros mountains 10.000 years ago. 31(1):186–194. Science 287(5461):2254–2257.

Current Anthropology Volume 52, Supplement 4, October 2011 S273

The Beginnings of Agriculture in China A Multiregional View

by David Joel Cohen

CAϩ Online-Only Material: References Cited PDF with Chinese and Japanese Characters Included

By 9000 cal BP, the first sedentary villages, marking the Early Neolithic, are present in Northeast China, North China, and the Middle and Lower Yangtze regions, but plant and animal domesticates do not make a substantial contribution to subsistence until after several more millennia, when domesticated millet or rice agricultural production is finally in place. While archaeobotanical ap- proaches have been the primary focus of recent studies, this paper looks specifically at the cultural developments in these four regions leading up to the emergence of agriculture in each. It is hy- pothesized that agriculture does not emerge independently in each of these regions but rather in interrelated steps through variable forms of interaction and information and social exchange within and between these regions. Interaction is currently evidenced through shared forms of material culture and by parallel and contemporaneous cultural developments.

Introduction systems that we witness the gradual and interrelated devel- opment of millet (dry) and rice (wet) agriculture: there are not multiple independent “inventions” of agriculture. Our understanding of the origins of agriculture in China is The processes of Neolithization are rooted in the Late Pa- still of low resolution, particularly during a critical time period leolithic (Cohen 2003). Because of a gap in archaeological between 12,500 and 9000 cal BP when hunter-gatherers in four distinct but interconnected geographical regions— coverage for the last few millennia of the Pleistocene, available Northeast China, the North China plains, and the Middle and data begin with the earliest-known villages already established Lower Yangtze River regions—establish the first sedentary vil- from ca. 9000 cal BP. These typically contain clusters of small lages. Although reliance on the cereals rice and millet was round houses and storage pits, abundant and increasingly once thought to be a major part of this fundamental behav- diversified pottery assemblages, edge-polished axes or adzes, ioral change, in recent years it has become clear that this is bone tools, ritual features, and discreet cemetery areas, but not the case. The advent of farming and the domestication subsistence still does not rely on domesticated plants. It is of cereals was a slow process occurring over 4 millennia in a such sites that we here call “Early Neolithic.” What follows number of small steps, region to region, after sedentism and is an overview of the archaeology of the Early Neolithic in other social and ideological changes of “Neolithization” had these four regions. begun. Recent debate has focused on the timing and pace of the Most previous work for China has focused on the archaeo- domestication of cereals—rice or millet—found in the Early botany of rice or millet or has looked at the archaeology only Neolithic and Middle Neolithic villages and how cereals are of single regions. The view here is purposefully broader, look- incorporated into changing modes of subsistence over a ing at parallel contemporaneous developments across both 3,000-year period (Crawford 2011; Fuller and Qin 2009; Fuller northern and southern China, as it is within this broad con- et al. 2009, 2010 and references therein; Liu, Lee, and Jiang text of continuous but variable contact between Neolithic 2007; Zhao 2011). We still do not know where cereal do- mestication first occurred. South China, where rice produc- tion begins, was assumed to have the earliest domestication David Joel Cohen is Adjunct Assistant Professor, International and agriculture, but new data have given rise to a counter- Center for East Asian Archaeology and Cultural History, Department of Archaeology, Boston University (650 Beacon Street, 5th Floor, argument for northern millet farming as having primacy or Boston, Massachusetts 02215, U.S.A. [[email protected]]). This paper independent invention (Barton et al. 2009; Bar-Yosef 2011; was submitted 13 XI 09, accepted 1 III 11, and electronically Bettinger et al. 2007; Lu 1999, 2006; Shelach 2000). The pau- published 17 IX 11. city of securely dated sites leaves this an open question.

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0009$10.00. DOI: 10.1086/659965 S274 Current Anthropology Volume 52, Supplement 4, October 2011

Cross-regional interactions leading to farming’s establish- climatic fluctuation in “Neolithization” require renewed at- ment and spread were facilitated by the interlacing river sys- tention (Bar-Yosef 2011). tems of eastern China. By 8900 cal BP, people in sedentary The north-south division of Early Neolithic cultures mir- villages in northern and southern China lived in fundamen- rors that of their Upper Paleolithic predecessors (Cohen tally different ways than their Late Pleistocene ancestors, and 2003). Terminal Pleistocene northern China features small the socioeconomic and ideological changes of the Neolithic ephemeral sites with lithics reflecting a broad range of tech- were well in place but still without agriculture (e.g., Cohen nological strategies, including microblade production, pol- 1998; Crawford 2006; Lu 1999, 2005; Shelach 2000; Underhill ished bone tools, grinding stones and chipped stone axes and and Habu 2006; Yasuda 2002 and references therein; Zhang adzes, and body ornamentation, perhaps reflecting specialized and Hung 2008). The organizational and ideological changes adaptations to the more extreme and unstable environments of the Early Neolithic arise first and out of synchronization of the region during the late Terminal Pleistocene (Elston and with the slower pathways leading from increasing exploitation Brantingham 2002). Microblade sites in northern China such of wild plants to their management and cultivation, to the as Xiachuan (27,500–18,200 cal BP) and Shizitan fixing of morphological traits and domestication, and then to (20,300–13,800 cal BP; Shanxi; fig. 1) are at the southernmost the spread of the domesticates (e.g., Fuller and Qin 2009; margin of an extensive Northern Asian microblade sphere Fuller et al. 2009; Jones and Liu 2009). (An 2000; Kuzmin, Keates, and Shen 2007; Lu 1999; Xia et al. 2002; Zhang 2002a). Upper Paleolithic South China features stability in its typ- The Geographical and Environmental ical cobble-tool industries (cores, flakes, choppers), which Setting and Upper Paleolithic Precursors persist into the Early Neolithic. This has been attributed to greater environmental stability in this warmer, wetter, more The vast landmass of China consists of three major topo- resource-abundant region. It is in hunter-gatherer contexts in graphic “steps” (Ren, Yang, and Bao 1985). While Late Pa- South China that pottery vessels—the earliest in the world— leolithic sites are found in many regions, it is in the river first appear, such as at Yuchanyan (Hunan) by 18,300 drainages of the lowest “step” (sea level to 1,000 m) that cal BP (Boaretto et al. 2009; Gu and Yuan 2006; Prendergast, hunter-gatherers first became sedentary and later domesti- Yuan, and Bar-Yosef2009; Yuan 2002) and Xianrendong cated plants and animals. Between 10,000 and 8900 cal BP, and Diaotonghuan (Jiangxi; 21,000–14,500 cal BP; four Early Neolithic cultural clusters are recognized. Along MacNeish et al. 1998; Peng and Zhou 2006; Sun and Zhan the Middle and Lower Yellow River, on the North China and 2004; Zhang 2002b; fig. 2). While it is not clear whether Central plains, we find the large village and cemetery sites of sedentary villages and plant domestication emerge first in the the Peiligang culture and closely related neighboring south or north, if environmental pressure is a causal factor cultures. These feature the intensive exploitation of millet for either, the resource-rich Yangtze River “land of plenty” (and some rice). The interdigitating tributaries of the Huai would require different modeling than the north. River system tie the Peiligang cultures to two clusters in the In Chinese archaeology, the presence of pottery has been Yangtze River system: the Pengtoushan culture of the taken as a marker for the Neolithic, but it is now apparent Middle Yangtze region and Shangshan and related sites that these are seasonal sites of mobile hunter-gatherers with in the Lower Yangtze. These feature early sedentary villages a small amount of coarse low-fired pottery. By 16,000–14,000 and the exploitation of probably wild rice. A fourth cluster, cal BP, pottery is distributed as widely as South China, Japan, in Northeast China along the Xiliaohe River in Inner and the Russian Far East (Cohen 1998, 2003; Kuzmin 2006; Mongolia, features large hillside settlements of the Xing- Wu and Zhao 2003). We hypothesize that pottery making longwa culture and possibly domesticated millet.1 spreads through these regions via social networks and is not Through the Terminal Pleistocene, Younger Dryas (ca. reinvented in each. Pearson (2005) sees the pottery as invented 12,800–11,600 cal BP), and onset of the Holocene (11,500 cal for feasting that reaffirmed group bonds and collective iden- BP), marked variation in annual monsoon cycles and their tities (but see Hayden 2009). Even if so, pottery takes on new intensities causes a shifting mosaic of localized environmental roles in the Early Neolithic as it diversifies in form and func- changes in China (Cohen 1998, 2002; Morrill, Overpeck, and tions. Cole 2003; Yasuda et al. 2004; Yi and Saito 2004; Yi et al. In North China, pottery appears at larger open-air sites 2003; Yuan et al. 2004). These would have been more severe featuring new adaptations by foragers. At Yujiagou (He- and abrupt in the north and would have more greatly affected bei; TL 11,600–11,100 BP), on the cold arid grasslands north- people at the margins of major phytogeographic zones (e.g., west of the Taihang Mountains, pottery appears in a typical Hong and Ricklefs 2001; Yang and Ding 2008). As there is microblade assemblage (Guo and Li 2000, 2002). Pottery sites some correlation with cultural changes, the localized roles of on the eastern warmer and wetter side of the Taihang range stand in ready contrast and can be seen as apparent precursors 1. Below, the lowercase terms “the north” and “northern China” are to the Early Neolithic. The lakeside Nanzhuangtou site used to refer jointly to the North China and Northeast China regions. (Hebei; 12,500–10,500 cal BP), dated to the Younger Dryas Cohen Beginnings of Agriculture in China S275

Figure 1. Map of Early Neolithic cultures and sites in South China (Mid- dle and Lower Yangtze regions), North China, and Northeast China men- tioned in the text. 1, Yuchanyan; 2, Chengtoushan; 3, Pengtoushan; 4, Bashidang; 5, Xianrendong and Diaotonghuan; 6, Shangshan; 7, Kuahu- qiao; 8, Xiaohuangshan; 9, Hemudu and Tianluoshan; 10, Dadiwan; 11, Shizitan; 12, Xiachuan; 13, Jiahu; 14, Peiligang; 15, Cishan; 16, Yuezhuang; 17, Xiaojingshan; 18, Houli; 19, Nanzhuangtou; 20, Yujiagou; 21, Zhuan- nian; 22, Xinglonggou. and Early Holocene, features flake tools (no microblades), cooking over fire and a new dependence on pottery in food polished axes, bone and antler tools, and querns and rollers, processing (Zhu 2002:98). Decorative elaboration may indi- indicating seed processing. Chiseled wooden rods give evi- cate increased social signaling using ceramics. dence of woodworking. Nanzhuangtou has activity areas for Two other nearby sites, both with microblades, also show fauna processing and cooking and two natural ditches filled important transitional elements. Zhuannian (Beijing; with midden deposits; these become typical features of Neo- 11,300–10,300 cal BP) features a larger site size (ca. 0.5 ha), lithic sites. The earliest domesticated dog is identified at Nan- querns and rollers, small stone axes, and fragments of stone zhuangtou (Yuan 2010). Small amounts of pig and chicken— vessels (mortars?) as well as pottery (Zhao et al. 2006). Dong- later Neolithic domesticates—are present, but deer are the hulin (Beijing; 11,000–9300 cal BP) has small pit buri- primary focus. Smoke blackening on pottery may indicate als. One (M2) contained a stone axe placed at the head of a Figure 2. Dates for Early Pottery (Late Paleolithic and Pottery Pre- Neolithic), Early Neolithic, and Middle Neolithic cultures and sites in South China, North China, and Northeast China mentioned in the text. Sources: A, Zhao et al. 2006; B, Cui and Zhou 2008; C, Zhejiang Shang- shan 2007:16; D, Zhang 2007; E, Gansu Dadiwan 2006 and Liu 2006b; F, Cai 2000; G, Beijing Donghulin 2006; H, Qu et al., forthcoming; I, MacNeish et al. 1998; J, Shelach 2000, 2006; K, Yasuda et al. 2004; L, Boaretto et al. 2009; M, Institute of Archaeology 1991; N, Zhejiang Kuahuqiao 2004:226–227; O, Henan Jiahu 1999:515; P, Yin 2006; Q, Zhejiang Tianluoshan 2007; R, Hunan Pengtoushan 2006; S, Guo and Li 2002; T, Guo jia wen wu ju 2006 and Wang 2006; U, Lu et al. 2009. Cohen Beginnings of Agriculture in China S277

Figure 2. (Continued)

flexed (perhaps bundled) individual with perforated shells exploitation included millet or whether or not these are sed- placed about the rest of the body. Ten pit hearths, neatly lined entary sites. with cobbles, show greater pyrotechnological controls (Beijing Donghulin 2006:5). The Early Neolithic Nanzhuangtou, Zhuannian, and Donghulin show impor- tant transitional steps toward the Early Neolithic. We find The appearance of Early Neolithic settlements sometime after new technologies such as grinding stones, pottery, and mi- 10,100 cal BP marks increased sedentism, population agglom- croblade knives as part of an adoption of broad-spectrum eration, and departures from Late Paleolithic behavioral pat- subsistence strategies (Lu 2006), but it is still not clear if plant terns. Many sites are enclosed by ditches and have distinct Figure 2. (Continued) Cohen Beginnings of Agriculture in China S279

Figure 2. (Continued) areas for houses, for domestic activities, for storage facilities, well as through stable isotope studies (e.g., Barton et al. 2009; and for the earliest planned multigenerational cemeteries, Hu, Ambrose, and Wang 2006; Hu et al. 2008; Pechenkina sometimes with hundreds of burials. Early houses are typically et al. 2005), a more complex picture is emerging of the degree small, round, and semisubterranean. Later they shift to a rect- of human manipulation of plants and their environment. angular plan. Houses are frequently found in clusters of sev- Early Neolithic populations continued to base their subsis- eral tens, and with a few exceptions, most houses in a village tence on hunting-gathering-fishing but with intensified ex- are similar in size. Burials are uniform in size, and grave goods ploitation of the plants later cultivated in agricultural systems. tend to be few. There is little evidence for functional differ- Pig domestication occurs perhaps at 9000 cal BP (Larson et entiation between households or social status differences be- al. 2010; Yuan 2010). Other early domesticates are dog and tween individuals. Within regions, there can be differences in chicken, with cattle, water buffalo, and sheep added later, but site size. wild animals continue to be a major meat source throughout A key issue, discussed in detail elsewhere (Crawford 2009, the Neolithic, although regional patterns of reliance on do- 2011; Fuller and Qin 2008, 2009; Fuller et al. 2009; Jones and mesticated animals versus wild animals vary (An 2004; Yuan Liu 2009; Liu 2008; Liu, Lee, and Jiang 2007; Liu et al. 2007; 2003; Yuan, Flad, and Luo 2008). Lu et al. 2009; Pan 2008; Zhao 2011), is the nature of plant and animal exploitation and whether domestication occurs North China: Peiligang and Related in this period. Since the 1970s, abundant cereal remains ex- Cultures cavated from storage pits and also found in pottery tempering were assumed to be from domesticates. In recent years, as The Early Neolithic cultures of North China date from at least archaeobotanical remains become systematically recovered, as ca. 9000 cal BP (or possibly 10,000 cal BP for Cishan) through S280 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 2. (Continued)

7400 cal BP. These cultures share many similarities in settle- iwan culture (Gansu and Shaanxi) in the western Middle ment structure, houses, burials, pottery forms and decoration, Yellow River. tools, grinding stones, and subsistence practices, and so we cluster them into one group here. The individual archaeo- The Peiligang Culture, Jiahu, and North-South Connections logical cultures within this cluster include the Peiligang (Henan) and Cishan (Hebei) cultures in the lowland plains Peiligang culture sites range in size from under 1 to 20 ha. of the eastern Middle Yellow River, the Houli culture (Shan- Excavated sites (all in Henan) include the typesite Peiligang dong) on the Lower Yellow River, and, dating later, the Dad- , Shawoli , Egou , Shigu , and Shuiquan Cohen Beginnings of Agriculture in China S281

(see Li 2003), but the best understood is Jiahu (5.5 ha, are capable of four- and seven-tone scales like ours today 9000–8000 cal BP; Henan Jiahu 1999:515–519). The Peiligang (Zhang, Xiao, and Lee 2004). component at the largest settlement, Tanghu , is said to Jiahu is rich in animal and plant remains. Subsistence was cover 20 ha (Zhang and Xin 2008). Both Tanghu and Jiahu based primarily on hunting-gathering-fishing. Domesticated feature a ditch enclosing part of the settlement, with Jiahu’s dogs are present at Jiahu, as are pigs (Yuan 2010). Jiahu stor- being the oldest. As discussed by Crawford (2011) for Jo¯mon age pits contain many plant remains, including rice. Although agriculture in Japan, human activities around these early vil- cereals have been the focus of discussions of subsistence sys- lage sites such as ditch construction and clearing would have tems, other plant foods probably played a larger role in Early a significant impact on plant populations and need further Neolithic subsistence and need more consideration. At Jiahu, consideration. these include acorns (Quercus), trapa, and wild soybean. Peiligang houses are semisubterranean and round in plan, Jiahu’s small-sized rice remains are at the center of the debate ca. 2.5 m in diameter. Round houses are a common feature on plant domestication (Crawford 2011; Fuller and Qin 2008; of the earliest sites across most regions. Interiors usually fea- Fuller, Qin, and Harvey 2008a, 2008b; Fuller et al. 2009; Liu, ture a hearth and a narrow sloped entryway or steps. At Jiahu Lee, and Jiang 2007; Zhao 2011; Zhao and Zhang 2009). (Henan Jiahu 1999), 42 of 45 structures are round and semi- Because cereal exploitation at other Peiligang and related cul- subterranean; most are a single room. At Tanghu, more than ture sites shares a primary focus on millet, it is extremely 60 semisubterranean round houses, some with double rooms, interesting that rice is abundant but millet is absent at Jiahu have been excavated along with 204 pits (Zhang and Xin even though Jiahu maintains many defining cultural traits of 2008). Peiligang-related sites can have hundreds of these “ash- the Peiligang. The exploitation of rice rather than millet had pits” for refuse disposal, storage, and ritual use (such as dog no impact on material culture at Jiahu, differentiating it from burials) found in both domestic and cemetery areas (Li 2003: other Peiligang sites, and so it was other factors—not millet 40–46). consumption—that may have maintained cultural boundaries Peiligang sites feature separate cemeteries within which can during this period. be several spatial clusters of graves, and two or three cemetery The Peiligang sites demonstrate evolutionary leaps in sed- entism and site structure, social organization and population areas can be located at one site. Contemporaneous cemeteries dynamics, technologies, and symbolic activities on first en- may serve separate social groups. At Jiahu, the early period trance into the Early Neolithic. Jiahu also supports the hy- has two cemeteries, and the middle and late periods have six pothesis that interregional contacts through social networks (Yan 2006:112). Thus, we have evidence for new forms of drive shared parallel development. For example, Jiahu mixes social organization emerging with Neolithization, with indi- Peiligang material culture related to subsistence—including cations of both population agglomeration and social differ- grinding stones, sickles, and pottery—with foods from South entiation (such as clan formation) over time. Jiahu burials China cultures such as lotus, trapa, and rice (Zhao 2011). show some differentiation by sex and by grave-good amounts, Jiahu’s location ties it into the Huai River communication but this differentiation in burial wealth occurs in a still egal- channels that connect northward into the Peiligang–Yellow itarian society with a continuum from poorest to richest, River system and southward into the Pengtoushan–Yangtze without clustering, and with random spatial distribution of River system. Shared pottery vessel forms with the Peng- the richest graves (Smith and Lee 2008). Unique to Jiahu are toushan culture—such as the double-eared, round-bellied, its numerous group primary burials and its many secondary high-neck, small-mouth hu jugs—demonstrate this (fig. 3), burials, both group and individual. as does the appearance of ding tripod vessels and three-leg Peiligang assemblages contain large numbers of adzes and cooking stands in both regions (Wu 1997). It is unlikely that the prototypical tongue-shaped spades, denticulate sickles, all of these identical forms developed independently. Also, and four-footed large stone querns with rollers. Bone tools Pengtoushan and Jiahu (unlike other Peiligang sites) have include elaborate barbed spears, points, and needles. The wide many secondary burials (and these burials in both areas con- variety of pottery vessel types include mostly guan jars, hu tain the double-ear hu vessels; Henan Jiahu 1999:533). They tall-neck small-mouth jars, and bo small bowls; second most each also have the earliest examples of settlement ditches. common are ding tripod vessels and wan bowls, and there are also pan platters, dou pedestal serving stands, and vessel lids (Li 2003:29–30; fig. 3). Cishan: Broomcorn Millet at 10,000 cal BP? Jiahu provides evidence for special ritual practices such as The Cishan culture is centered in Hebei, north of the dog pits and sets of tortoise shell rattles placed in some burials. Peiligang culture region (see Liu, Hunt, and Jones 2009). Symbolic systems become more greatly manifested in pottery Excavated sites include the 8-ha Cishan typesite (Hebei Cishan decoration as well as in individual “characters” found in- 1981) and Beifudi (Duan Beifudi 2007). Cishan’s ex- scribed on tortoise shells (but not related to later writing as cavations in the 1970s revealed 474 storage pits containing some suggest; Li et al. 2003). Jiahu inhabitants also had a an estimated 50,000ϩ kg of ashy remains of cereals, originally well-developed cognitive scheme of music: crane bone flutes identified as domesticated foxtail millet Setaria italica (Zhao Figure 3. Comparison of artifact types found in the Early Neolithic North China Peiligang and Cishan cultures and the South China Pengtoushan culture. 1, 14, and 24 are footed grinding stones of the Peiligang culture also found in Cishan later phases. 6 and 37 are cooking stands found in both North and South China. 19, 21, 32, 33, 43, and 44 are double-eared small-mouth hu jugs common at Pengtoushan and Peiligang sites. 7, 8, 17, 18, 27, and 28 are ding tripods. Objects are not to the same scale. 1–13 after Hebei Cishan 1981; 14, 15, 18, 19, 21, and 22 after Peiligang A 1978; 16, 17, and 20 after Peiligang B 1982; 23–34 after Henan Jiahu 1999; 36–44 after Hunan Pengtoushan 2006. Cohen Beginnings of Agriculture in China S283

2005:91; 2011). Recent phytolith and biomolecular analyses fu cauldrons are the common cooking vessels. There are leg of new samples from exposed sections at the site instead iden- stands, as in South China and Peiligang (Tong 2006). Stable tify these as domesticated broomcorn (common) millet Pan- isotope studies of human bone from Xiaojingshan show C4 icum miliaceum, with new radiocarbon dates ranging from plants (millet) contributing 25% of dietary protein at 8000 10,300 to 8700 cal BP, which would mean Cishan could have cal BP; this increases, becoming the most significant protein the earliest domesticated cereals and dryland farming in China source a millennium later (Hu et al. 2008). Flotation at the (Crawford 2009; Lu et al. 2009). But these radiocarbon dates, Yuezhuang site (ca. 7900 cal BP) recovered 40 broomcorn taken on samples from exposed surfaces and not from ar- and one foxtail seed but also 26 rice grains, showing Oryza’s chaeological excavation, are more than a millennium earlier arrival in the Lower Yellow River by this time (Crawford, than those from the earlier excavations and may be problem- Chen, and Wang 2006). Nuts, a significant component of atic (Zhao 2011). Further verification is called for through Early Neolithic diets in South China, have been identified at systematic stratigraphic sampling and micromorphological Houli sites but in smaller quantities (Crawford 2011). study to assure secure contexts and associations for radio- carbon samples. Foxtail millet appears after broomcorn, at Dadiwan and the Highlands 8700 cal BP, and only in small quantities (which is the pattern at other sites, as well). Domesticated pig is present at Cishan After the Peiligang-related cultures developed millet produc- as well (Yuan 2010). Cishan’s occupation ends at 7500 cal BP. tion in the lower elevations, they probably expanded into Cishan’s material culture shows close connections with the higher elevations and new ecological zones. The appearance Peiligang regional phase. For example, grinding querns in- of the Dadiwan (or Laoguantai ) culture may crease greatly through time. In its second phase, 52 four- result from the spread of farming populations into new footed querns are found—a unique form coming from Pei- regions, a pattern that was to continue and accelerate through ligang. Evidence for participation in wider-ranging the Neolithic. The Dadiwan distribution area is centered on interactions also is found. Vessel stands, ding tripods, and the Wei and Upper Hanshui Rivers and the loess plateau at some ring-foot vessels are also present, and these vessel forms ca. 1,500-m elevation, a higher and more arid region than are distributed widely, even into South China. Flat-based ce- the other Peiligang-related cultures. Sites include Dadiwan ramic jars and bowls and straight-sided wide-mouth cups (yu) (Gansu), Lower Beishouling , Laoguantai, Yuanjunmiao may indicate ties to the northeast (below; fig. 3). Beifudi is , and Beiliu (Bai and Zhang 2001:15). At Dadiwan, notable for its ritual features, ceramic masks (earliest found 240 houses, 325 pits, 35 kilns, and 65 burials have been ex- in China), and jades (Duan Beifudi 2007). Early jades are also cavated at 10 localities. Dadiwan dates later than other Pei- found in the northeast and at Dadiwan to the west. ligang-related regions. The earliest occupation, ca. 7900–7200 cal BP, features round semisubterranean houses, pits, and burials. Ceramics share common forms with Peiligang, in- The Houli Culture cluding tripod vessels and hu small-mouth jars, but one dif- Thirteen Houli sites are found at the margins of the ference is the presence of painted pottery—bowls with a red central hills and outer floodplains of the Shandong peninsula painted band at the rim. At 6500 cal BP, the Dadiwan settle- skirting the northern edge of the highlands (Tong 2006:176). ment expands from the second to the third river terrace. At Houli sites are larger than Peiligang sites, but their chronology this stage, a ditch is dug enclosing its 156 now rectangular is poorly understood and may date later. Hu et al. (2008) semisubterranean houses positioned around a central plaza, date the Xiaojingshan site to 8200–7500 cal BP. The and burial shifts to a separate cemetery zone. Ceramic forms Houli typesite is 15 ha. The Xiaojingshan site features a three- increase, and fine wares and complex painted designs appear. sided ditch 1,130 m in total length. The earliest houses and Jade chisels also are found (Gansu Dadiwan 2006). two cemeteries are found within its bounds. Later houses and Bettinger et al. (2007, 2010) include Dadiwan in a scenario a cemetery are added outside the ditch. There are also early for an independent center of millet agricultural origins created pottery kilns and many ash pits. Although houses are semi- by “low-level food producer” forager groups from the desert subterranean, they are square with rounded corners, 35–50ϩ margins north of the site. Intensively specializing on broom- m2 in size, with one to three hearths in the center and fired corn millet, they were forced to migrate to the unpopulated earthen walls. The rectangular shape may indicate a later date. plains of the loess plateau when in the Early Holocene the Many have grinding querns installed in the floors near the region became more arid. In semipermanent settlements such hearths. Unlike Peiligang, the querns are footless and used as Dadiwan, they domesticated broomcorn millet, which is on both sides, and they are found in domestic contexts, while more tolerant to cold and drought than foxtail millet (Bet- Peiligang querns primarily are found in burials. Many houses tinger et al. 2007:91–93). In Barton et al.’s (2009) stable iso- also have large pottery cauldrons, sometimes half buried at tope study, they define this stage as a low-intensity domestic the interior wall. relationship between millet, humans, and their hunting dogs. Houli ceramics feature more than 12 forms, but unlike The proposal that Dadiwan is an independent center because Peiligang, there are few ding tripods. Instead, round-bottom of its location 700 km from Peiligang, higher elevation, dif- S284 Current Anthropology Volume 52, Supplement 4, October 2011 ferent environment, and focus on broomcorn is hardly ac- baogou period, mortality profiles, but not morphology, of ceptable. Dadiwan, established at 7900 cal BP, appears more pigs suggest domestication (Shelach 2000:283). In Xinglongwa than a millennium after the early phase at Jiahu or Cishan. sites, early sacrificial usage of pigs and deer is found, too. Dadiwan is not a closed cultural system, and this view does Such animal sacrifice becomes a mainstay of later Neolithic not consider its evidence for interaction with other clusters and Bronze Age religious activities. In grave M118 at Xing- in North and Northeast China, including broomcorn millet longwa, a 50-year-old male was interred with one male and exploitation (found earlier at Xinglongwa and Cishan sites; one female pig. In house F5 at Xinglonggou, 12 pig crania Lu et al. 2009), settlement layout, house forms, kiln tech- and 3 deer crania—most with a hole drilled through the fore- nology, adzes, some burial practices, jades, and shared ceramic head—were placed into the house floor (Liu 2006b:9). vessel forms such as tripods and hu vessels. These make it Xinglongwa lithics are more diverse than other regions, difficult to account for Dadiwan as an independent center. retaining Late Paleolithic flake tools and microblades and Furthermore, the late adoption of foxtail millet at Dadiwan blades while adding polished tools. Both microblades and at 5900 cal BP (Barton 2009) mirrors foxtail millet’s later blades are also found at Beifudi in the Cishan culture to the growth in usage in the lower elevations as well. Xinglongwa south. Because true blades are rare in the Chinese Late Pa- produces jades, and the Beifudi Cishan site has jades. It is leolithic and Early Neolithic, this commonality may be sig- unlikely that all such features would have been reinvented nificant. Tools include flaked and sometimes fully polished independently at Dadiwan. Moreover, pottery production has or edge-polished rectangular stone “spades” (some with not been found in the loess plateau or northern desert margins waists) and polished axes/adzes (suggesting forest clearing) earlier than Dadiwan, and thus Dadiwan’s earliest but ad- and awls. The percentage of polished tools increases into the vanced pottery would most likely be the product of com- Zhaobaogou period (Liu 2006b:8–9). Bone-handle composite munication with or migration from farming cultures to the tools are inset with microblades. Use-wear analysis of stone east. Therefore, Dadiwan can be seen as a secondary center “plows,” which are up to 30 cm long, suggests they were used representing an early spread of farming populations into new to work soil. Stone querns and rollers are found in installa- environmental zones. Interaction is also seen in the opposite tions near the walls inside Xinglongwa houses and may have direction: when Dadiwan potters added painting to their rep- been used for grinding grains. Querns are relatively large, up ertoire, it quickly worked its way into the northeast and down to 40 cm across and 10–30 cm thick. These are sometimes the Yellow River into the Yangshao -related cultures. found associated with large ceramic basins installed in the house floors (Shelach 2000:384), and this is a domestic pattern The Early Neolithic in Northeast China: shared with North China Houli culture sites. The Xinglongwa Culture Xinglongwa pottery is sand tempered, coarse, and low fired. Vessel forms are few and mostly jars and bowls. Flat-based The occupations of the few dated settlements in Northeast vessels predominate, with straight vessel walls angling outward China’s Early Neolithic Xinglongwa culture begin 5 to form wide vessel mouths. Vessel surfaces can be covered centuries later than the earliest in the Yellow River region. with rows of impressed or incised patterns, most notably the These settlements, however, parallel what is seen in contem- Z-shaped motif and net patterns. Zhaobaogou culture pottery poraneous settlements in the Yellow River region. Xinglongwa becomes elaborated with more complex decorative motifs, sites, in eastern Inner Mongolia, include Xinglongwa, Baiyin- including geometric and net forms over the vessel surface. changhan , Chahai , Nantaizi , Xinglonggou During the Zhaobaogou period the first fine clay pottery ap- , and Beichengzi (Liu 2006a). Radiocarbon dates pears in the region. Decorative motifs combine intertwined from Xinglongwa sites range from ca. 8300 to 7400 cal BP. elements from deer, birds, boar, and other animals (see Liu The cultural sequence continues with the Fuhe culture 2000). The intricate designs, possibly produced by individuals and then the Zhaobaogou culture ca. 7100–6700 cal with specialized skills, are early examples of iconographic or BP (Liu 2006b; Shelach 2006). artistic motifs that blossom in the later Neolithic (Shelach Subsistence in Xinglongwa may include domesticated mil- 2000:389–393). lets (Zhao 2011), but wild plants and animals make up the During the Xinglongwa period, early examples of carved majority of the diet. Flotation at Xinglonggou recovered many jades are found in house floors and burials (Shelach 2000: carbonized seeds, most of which are wild grasses but that also 394), establishing the roots of the ritual jade production that include much broomcorn millet and a small amount of foxtail becomes a hallmark of this region by the Middle Neolithic millet (Liu 2006b:9; Zhao 2005:91–92)—a pattern that is sim- Hongshan culture. Jades found in the Dadiwan culture ilar to North China sites. These millets, dating ca. 7650 cal (above) date slightly later. BP, are morphologically domesticated (Zhao 2011). No domesticated animals are reported for Xinglongwa sites, Xinglongwa Settlements although domesticated pigs are already found in South and North China by this time period, but the hunting of deer and Unlike other regions, Xinglongwa sites are located primarily wild boar is an important part of the economy. By the Zhao- on hillsides, suggesting unique considerations for their place- Cohen Beginnings of Agriculture in China S285 ment. Larger sites range 2–10 ha, some with shallow ditches rice agriculture’s demands for available level ground for man- around them. Houses are numerous, closely spaced, and ar- aging water flow (Yan 2006:113). ranged in rows, with one to three clusters at each site. At Houses at Pengtoushan are semisubterranean and com- Beichengzi, there are 214 houses in 11 rows within the outer paratively small, with irregular round shapes and hearths of ditches, while Xinglongwa has 94 houses in 11 rows (Liu piled fired earth and clay. A few are rectangular and built at 2006b:4–5; Wu 2002:109; Yan 2006:112). Public spaces and ground level on a low base of clay and sand. Among Bashi- ceremonial installations are also found. Some sites have a dang’s 24 houses there are fewer semisubterranean and more small plaza at the center containing the largest structure at surface structures, including pile dwellings. Two houses built the site. At Chahai, a 120-m2 structure is in the plaza and has on earthen platforms 30–50 cm high have a large central post an associated 19.7-m-long “dragon-shaped” stone piling with with animal bones deposited near it, perhaps ritually. A ritual burials and a sacrificial pit (Liu 2006b:4–5). function may also hold for most of the 80 excavated pits at Xinglongwa houses are semisubterranean and rectangular Bashidang. They are carefully dug, 30–50 cm in diameter, with with central hearths, and most range in size from 50 to 80 straight walls and a flat base. Fired earth, pottery sherds, m2. A common and unique Xinglongwa practice is burial— polished black stone ornamental rods, a few polished stone individual male or female or child group internments—under objects, or cobbles are cached inside many. These are similar house floors (Liu 2006a:11). Pits possibly used for storage are to the sacrificial pits found in the Middle Neolithic “altar” features at the nearby Chengtoushan settlement (Yin also found dispersed among the houses as well as inside. These 2006). contain potsherds and animal bones, while one at Nantaizi Funerary treatment in Pengtoushan is simpler than in had more than 200 stone tools (Shelach 2000:401). The Xing- North China. The 100 graves at Bashidang and 21 at Peng- longwa rectangular houses date later than the round semi- toushan are small shallow pits of rectangular, rounded, or subterranean houses from the early phases in North China. irregular shape. Most are secondary, but some are primary Without formal reports being available for Xinglongwa sites, and flexed. Grave goods include only one to four objects, we still know little about their internal organization, their usually a fu cauldron or a high-neck small-mouth jar or a economic activities, their community integration, and the stone tool (Bai and Zhang 2001:23). Bashidang burials are structure of these early sedentary communities (for the sub- placed at the perimeter of the residential area but are not sequent Zhaobaogou period, see Shelach 2006). clustered together. Elaborate vertical pit graves with greater numbers of grave goods in larger distinct cemetery areas do The Early Neolithic in South China: not appear until the Middle Neolithic Daxi (6400–5500 Pengtoushan and Shangshan (Middle cal BP) and Qujialing (5500–4500 cal BP) cultures, such as at Chengtoushan (Chen 2006:210). Yangtze Region) Archaeobotanical remains at Bashidang include 67 species of plants. Rice at Bashidang, probably a secondary food re- The Pengtoushan culture represents the Early source, has been described as a cultivated form that shows Neolithic of the Middle Yangtze region. Sites are distributed signs of human selective pressure (Hunan Pengtoushan 2006: in the Liyang Plain of the Lishui River in northern Hunan 519; Zhang and Pei 1997), but Fuller, Harvey, and Qin (2007: Province. Five sites have been excavated: Pengtoushan, Ba- 323), noting markedly thin rice at the site, interpret it as shidang , Fenshanbao , Huangjiayuan ,and immature and morphologically wild. Other plants include Tujiatai (Hunan Pengtoushan 2006; Yin 2006). The wild soybeans, trapa, plum (Prunus mumu), peach (Prunus Pengtoushan culture sites have radiocarbon dates ca. 10,000– persica), and wild grape. Fauna includes six species of fish, 8400 cal BP but probably continued for another several cen- birds, Sus sp., Cervus sp., water deer, muntjac, and wild water turies. Sites are probably permanent villages with subsistence buffalo (Bubalus sp.; Hunan Pengtoushan 2006:513). still based on hunting-fishing-gathering but also with ex- Remarkably, although the Pengtoushan culture settlements ploitation and perhaps manipulation of wild rice. Soundings are most likely permanent sedentary villages with a new and at Bashidang and other sites show that an earlier but poorly increasingly intensive pattern of exploitation of rice and other understood village phase exists before the Pengtoushan cul- plants that would also become domesticates, its lithic assem- ture called the “Lower Bashidang” culture. blage still retains regional Paleolithic characteristics. Peng- Positioned on a slight rise above the current fields and water toushan still features mostly cobble tools, small flaked-flint channels of the plain, the Pengtoushan site is ca. 2 ha in size. tools (e.g., scrapers), and a few axes and adzes possibly used The later Bashidang site is ca. 3 ha in size and better preserved for woodworking (Hunan Pengtoushan 2006:587); this is not (Hunan Pengtoushan 2006). Bashidang is one of the largest like later specialized assemblages associated with rice agri- Early Neolithic sites here, with others typically under 1 ha culture (He 1995). (Yan 2006:113). Through time, South China settlements re- Pengtoushan pottery features uneven forming of vessels, main smaller than northern ones, perhaps because of the but vessel walls are thinner than Pleistocene pottery, mea- watery topography of the Yangtze basin or later because of suring 0.6–0.8 cm. Round-bottom vessel forms predominate, S286 Current Anthropology Volume 52, Supplement 4, October 2011 and there is much coarse cord marking (He 1995). Wares BP (Zhejiang Shangshan 2007), while Xiaohuangshan’s esti- have fine sand or organic tempering, including rice husks and mated age is ca. 10,000/9000–8000/7000 cal BP (Guo jia wen leaves (Pei 2002). Double-eared hu jars make up 12%–15% wu ju 2006; Wang 2006). At Shangshan, associations and dates of the assemblages and vessel stands 4%–11% (Hunan Peng- of its earliest features are still preliminary, and some may toushan 2006:586, 622–628). As mentioned above, both are belong to the later Kuahuqiao and Hemudu culture phases. also found at Jiahu and in other Peiligang-related cultures, Both Shangshan and Xiaohuangshan have been interpreted perhaps reflecting contact between these regions (fig. 3). as permanent settlements with a hunter-gatherer economy Bashidang, like Jiahu, is the earliest settlement found in the and early exploitation, perhaps cultivation, of rice (Guo jia Yangtze region to feature a surrounding ditch, and it dates to wen wu ju 2006; Wang 2006; Zhao 2011). While the size and the same time period, ca. 8800 cal BP. With sides of 170–200 layout of Shangshan are not reported, within the 5-ha area m in length, the ditch was ca. 4 m wide, with sloping edges of Xiaohuangshan are found three clusters of houses; it is also to a depth of 0.5–2 m. It may have in part made use of a the largest Early Neolithic site in the south. Houses at both diverted river channel, allowing drainage from the settlement sites are rectangular and up to 14 m long, with some built into the river (Pei 2006 [2004]:280). Bashidang’s ditch build- on U-shaped foundation trenches and others, perhaps later ing establishes a pattern of intensive landscape modification in date, as pile dwellings (Jiang and Liu 2006). Like Peng- that through the Neolithic becomes increasingly used to tie toushan, Xiaohuangshan burials are located about the edges together settlements and surrounding fields by water channels. of the settlement but are not clustered. Grave goods include As discussed by Crawford (2011), such engineering of site a few pottery vessels, usually small flat-bottom basins with environments could also produce opportunities for the pre- flared rims, ring-foot jars, or high-neck small-mouth jars. ferred growth of certain plant species such as wild rice. Water There are numerous pits within both settlements of various control, too, is central to rice agriculture, and as the water- shapes but typically shallow—some with one or two pottery management systems of these channels were developed fol- vessels or lithics—that may have been for burials or other lowing the Pengtoushan culture period, they may have been ritual purposes. At Xiaohuangshan, there are many 1-m- critical in the domestication process for rice (see Fuller and diameter pits with straight walls and flat bottoms. The ex- Qin 2009). Following Bashidang, Middle Yangtze settlements cavators believe they were used for stem tuber and nut storage increase in size, and the ditches become moats permanently (acorns are a common resource at later sites). filled with water. At 7400 cal BP, Chengtoushan’s moat is 15 Shangshan sites have large amounts of pottery. At Xiao- m wide surrounding a circular 5-ha area. At 6850 cal BP, the huangshan, 1,000 pottery vessels could be reconstructed. The Chengtoushan site expands again. A new embankment wall early-phase pottery can be grouped with Shangshan’s, where 10 m wide by 2 m high and a 2.2-m-deep moat are con- slab-molded or coil-made flat-bottom basins predominate structed. Moats such as this connected the settlements to (ca. 85%). As with Pengtoushan pottery, rice chaff, stalks, and extensive waterway networks, natural and man-made, leading leaves can be found in the sherd fabric—this tempering pro- to fields and neighboring settlements. Oars and other wooden vides the rice evidence that is used in recent discussions of boat parts are found. By the Late Neolithic, Chengtoushan the state of rice domestication at the site (e.g., Liu, Lee, and reaches 8 ha in size (Pei 2006 [2004]:281–283). As the Cheng- Jiang 2007). However, the flat-bottom vessel forms at Shang- toushan settlement is expanding, paddy-field rice agriculture shan sites are different from Early Neolithic sites in other is developing at the site as well. The earliest paddy-field sys- regions and from later cultures in the Lower Yangtze, such tems, with long rows of raised bunding demarcating fields of as the Kuahuqiao and Hemudu, that instead feature round- 2.7 m by 20ϩ m, belong to the Tangjiagang culture, bottom cauldrons and vessel stands (as does the Pengtoushan with radiocarbon dates ca. 6300–6000 cal. BP (Fuller and Qin culture). No tripod vessels are found, either. Later-phase 2009:97; Hunan Chengtoushan 2007:164–167; Zhao 2010). Shangshan pottery has more vessel forms and greater deco- By this period, the mode of subsistence can be called agri- rative elaboration, with everted-rim bowls, double-belly ped- cultural, here defined as having domesticated plants being estal dishes, cordmarked egg-shaped cauldrons, and platters grown in human-prepared fields. with a tall open-work ring foot (a new technique) all found. This formal differentiation continues into the later cultures in the region. The Lower Yangtze Early Neolithic Shangshan and Xiaohuangshan, like Pengtoushan, retain the South China Late Paleolithic lithic assemblage, featuring mainly flakes with some retouch and cobble tools, stone balls, Shangshan and Xiaohuangshan “digging stick weights,” and shallow “cupholes”: all are The Shangshan (Mao et al. 2008; Zhejiang Shangshan formed by percussion. Only a few adzes and chisels and a few 2007) and Xiaohuangshan (Guo jia wen wu ju 2006) grinding stones (up to 30–50 cm long) are found (Ren 2005; sites (Zhejiang) represent the earliest Early Neolithic culture Sheng, Zheng, and Jiang 2006; Zhejiang Shangshan 2007). found in the Lower Yangtze River region. A few radiocarbon Early Neolithic behavioral patterns are established in both determinations date the Shangshan site to 10,000–8500 cal the Pengtoushan and the Shangshan cultures, including ag- Cohen Beginnings of Agriculture in China S287 glomerated permanent settlements, house construction, heavy Qin (2007) and Fuller, Qin, and Harvey (2008a) observe that use of underground storage facilities, active human manip- Kuahuqiao rice, as well as later rice at Hemudu, has significant ulation of the site environment, more highly developed pot- proportions of immature harvested spikelets (as does Bashi- tery production and functional differentiation of forms, in- dang rice). A harvesting pattern involving such high numbers cipient use of axes and grinding stones, and increasingly of immature spikelets along with the split numbers between intensive exploitation of the cereal rice (and other select wild and japonica implies that the rice plants at Kuahuqiao plants, including future domesticates and acorns). In both were under cultivation. Furthermore, the low relative amounts cultures, these early village residents maintain a primarily of rice present at the site as a percentage of all plants would hunting-fishing-gathering subsistence system. Although rice, also indicate that this cultivated rice is still a supplementary which is most likely wild but possibly managed, has entered resource behind collected wild nuts and acorns (Fuller et al. the diet, it appears to play a secondary role in nutritional 2009). Other plant remains at Kuahuqiao include trapa (water intake. Yet these earliest villages may already have crossed a chestnut), foxnuts, peach, and apricot. certain threshold and were on a pathway toward rice culti- Microcharcoal and pollen studies at Kuahuqiao show hu- vation, domestication, and agriculture. We still lack data for mans actively manipulating the coastal swamp environment understanding seasonality and subsistence patterns at these around the site by clearing the wetland scrub through fire. sites as well as models for how the Early Neolithic behavioral This could have been part of managing natural stands of rice. patterns are established. When the South China hunter-gath- Kuahuqiao residents also faced regular flooding by slightly erer annual subsistence cycle came to incorporate wild rice, brackish water, and the site may have been abandoned because even in relatively small amounts, did this effect mobility pat- of Holocene marine transgression. If people were planting terns or allow year-round sedentism in these early villages? rice or managing and protecting wild stands, the rice could And how did the construction of villages affect the landscape have needed bunding and other forms of water management and wild plant populations? (Zong et al. 2007), or inhabitants may have carefully selected areas with advantageously fluctuating water levels to promote rice plant growth and grain production (Fuller and Qin 2009). Kuahuqiao While there is possibly stand management of rice and other The Kuahuqiao site (Zhejiang), situated at the modern plants at Kuahuqiao, the amount of human dependence on mouth of the Qiantang River, occupied 7900–7000 cal BP rice may still be limited. Fuller et al. (2009) suggest that do- (Zhejiang Kuahuqiao 2004:225–227), shows the next stage in mesticated rice (indicated by a higher percentage of fixed the development of the Early Neolithic in the Lower Yangtze nonshattering phenotype rice present and increased repre- region. The 3-ha settlement is heavily disturbed, but seven sentation of rice among all plant remains) appears after Ku- partial structures with new construction and woodworking ahuqiao, during the Hemudu culture period, between techniques were excavated, including an ovular platform 1.6 6900 and 6600 cal BP, as seen at the Tianluoshan site m high and 10 m wide as well as rectangular foundations up (Zhejiang; Zhao 2010; Zhejiang Tianluoshan 2007; Zheng et to5.7 # 4.7 m (Zhejiang Kuahuqiao 2004:26–38). As in other al. 2009).2 During this time period, in addition to an increas- settlements, numerous pits surround the structures, and some ing percentage of nonshattering (domesticated) rice spikelets, have their wooden covers preserved, indicating their use in rice consumption increases significantly as well, with rice re- storage. No burials are reported. mains jumping from 8% to 24% of total plant remains. Zheng Kuahuqiao sat near coastal swamp marsh, and the inhab- et al. (2009) identify land clearing and tilling for rice fields itants exploited a wide range of resources from the diverse at Tianluoshan, with many associated weeds as well. rich environments surrounding the site. Fauna include sika Preserved organic remains at Kuahuqiao show a high level deer, wild water buffalo, serow (Capricornis sumatraensis), of development in woodworking, basketry, and other prac- alligator, turtle, waterfowl, various fish species, and dolphin. tices. A 5.6-m long-(partial) dug-out pine canoe and wooden Deer, however, dominate the assemblage, as is typical in Neo- lithic cultures in the Yangtze and elsewhere (Yuan, Flad, and 2. Hemudu (Zhejiang Hemudu 2003)—the Hemudu culture type- Luo 2008). The inhabitants also were probably managing site—and Tianluoshan are both waterlogged sites with excellent preser- plants and animals. Domesticated dog is present, and the vation. As a Middle Neolithic culture, the Hemudu culture features a small numbers of pigs present show decreased mandible more highly diversified pottery assemblage with a rich iconography of plant and animal decorative elements; lacquerware vessels; bone and stone length, indicating domestication (Yuan 2010). tools associated with farming such as scapula spades; ivory objects, or- The rice remains from Kuahuqiao are central in the current namental stone, and jade objects; large, raised-pile dwellings; and wood- debate on rice domestication (e.g., Liu et al. 2007; Pan 2008). framed wells. Hemudu was excavated in the early 1970s, but the current Kuahuqiao rice grains and phytoliths have been interpreted Tianluoshan excavations now incorporate systematic sampling, water- as being in a “primitive stage of domestication,” with 41.7% screening, and flotation. As Zhao (2011) notes, preliminary results show that although rice agriculture is taking place, hunting-gathering is still of the spikelet bases identified as japonica type (domesticated) making a major contribution to subsistence. One focus at Tianluoshan and the remaining 58.3% as wild (Zhejiang Kuahuqiao 2004: is acorns and nuts, as we have seen indications for at the earlier sites as 375–376; Zheng, Sun, and Chen 2007). Fuller, Harvey, and well. S288 Current Anthropology Volume 52, Supplement 4, October 2011 oars give direct evidence of the importance of water transport Shixia site, Guangdong), over water to Taiwan and be- in the Yangtze region. Bow and arrow are used, with a 121- yond in the Austronesian dispersal, and northward into the cm-long mulberry bow preserved. Mortise and tenon joinery Middle and Lower Yellow River basin and possibly Korea, all are in use to fashion timbers and to make ladders. There is by 5000 cal BP (see Bellwood 2006, 2011; Fuller and Qin also woven matting. 2009:101 and references therein; Lee 2011; Ruddiman et al. Pottery and lithics at Kuahuqiao show significant advance- 2008; Zhang and Hung 2010). ments over Shangshan’s. Iron-containing slips are used to produce red, gray, or black surfaces, and red and white pat- Conclusions terns and abstract motifs are painted over the slip. In the Middle Yangtze, intricate geometric surface patterns occur Early Neolithic villages emerge shortly following the Younger during the same period in the Tangjiagang culture (7400– Dryas period of climatic stress, and the role of climate change 7200 cal BP), and the Daxi culture (6400–5500 cal BP) has needs to be revisited (Bar-Yosef 2011). Unfortunately, we are similar painted pottery. The number of vessel forms increases missing critical data for the period of the establishment of over Shangshan, and the pottery is more finely made, with these villages, and future work should focus on this. The role carefully formed rims, angled shoulders, and ring feet. There of pottery also needs further consideration. While pottery was is also a considerable increase in serving vessels, such as ped- already in use for over 8 millennia, once sedentary villages estal serving stands (dou). The new decorative motifs and appear, it immediately takes on prominent new roles in food vessel forms represent not only technological improvements preparation, serving, storage, ritual, and social signaling. As but also new roles for pottery in domestic activities (pro- outlined above, we have only recently come to understand cessing, cooking, serving, and storage), ritual, and social sig- that these first sedentary villages are not synonymous with naling. With Kuahuqiao, the majority of stone tools are now the presence of full-scale farming. Village settlements emerge polished, including adzes, axes, chisels, arrow points, rollers, in China before cultivated plants make a significant contri- and grinding stones. There are also jade ornaments. bution to diet and several millennia before agriculture systems The Kuahuqiao cultural inventory signals the transition to with domesticated plants and prepared fields are in place. the “Middle Neolithic” represented by the subsequent He- Instead, many of the earliest Early Neolithic villages still rely mudu culture in the Lower Yangtze region. During the Middle primarily on hunting-fishing-gathering. Adaptations involv- Neolithic, as seen at Tianluoshan after 6900 cal BP (Fuller et ing the exploitation of wild resources were able to trigger and al. 2009), people are farming rice (i.e., they are growing do- support sedentism, population growth and agglomeration, mesticated plants in prepared fields). This early rice farming and the other cultural changes that mark the Early Neolithic, occurs within a broad-spectrum subsistence system based on including pottery elaboration, house construction, ditch waterside and aquatic plants, as Hemudu culture sites are building, and new ritual activities (indicated by caches, plat- found primarily in wetland catchments (Nakamura 2010; Qin, forms, cemeteries, etc.) that all can be related to new ideo- Fuller, and Hai 2010). logical conceptions of social grouping and territoriality. In The subsequent development of more highly productive China, the Neolithic did not require farming at its start. rice farming by means of the paddy-field system marks a During the Early Neolithic across the four cultural areas major turning point in Neolithic societies. Soon after the discussed above, we see the sharing of specific material culture appearance of paddy fields, we find hierarchical settlement traits and parallel developments on the road to agriculture. patterns, higher population densities, increasing specialization Information exchange is hinted at typologically in shared pot- of labor, increasingly greater internal social status differenti- tery, lithics, house forms, spatial organization of settlements, ation, and greater accumulation of wealth and disposal of burial practices, subsistence practices, and ritual activities. I wealth in ritual (see Zhang 2003; Zhang and Hung 2008). posit here that Early Neolithic cultures emerged and devel- The earliest paddy fields in the Lower Yangtze, found east of oped—and the farming of millet and of rice was invented Lake Taihu in the late Majiabang culture (6300–6000 and spread—within wide-ranging webs of information ex- cal BP) sites of Caoxieshan and Chuodun ,are change and broad social exchange networks whose roots are contemporaneous with the earliest paddies in the Middle in the interactions of Late Paleolithic hunter-gatherer societies Yangtze at Chengtoushan. The Majiabang paddy fields consist (Cohen 2003). None of these regions was geographically or of small dug-out units connected by small water channels, culturally isolated from others. Because of this, it is unlikely with water reservoir pits from which stored water could be that there could have been multiple independent inventions fed into the system. This system of small paddies allows tight of major classes of material culture such as the shared pottery control of the water level in all units, which is essential to vessel forms. Indeed, the long, complex developmental path- raising the productivity of the rice plants (Fuller and Qin ways in each of these regions for millet and for rice farming, 2009). With the advent of paddy fields and the concomitant from cultivation through domestication and agriculture, demographic and organizational changes in societies that fol- would have occurred with information exchange from other low, rice farming “gains legs” and disperses quickly and regions—regions that contemporaneously were going through widely. Rice farming reaches to southernmost China (the parallel processes. In other words, the advent of agriculture Cohen Beginnings of Agriculture in China S289 did not result from independent isolated processes in each Early Neolithic period in the Yellow River valley). Hua xia kao gu region. 2001(2):14–28. Barton, Loukas, Seth D. Newsome, Fa-Hu Chen, Hui Wang, Thomas To accept this scenario, when looking at the origins of both P. Guilderson, and Robert L. Bettinger. 2009. Agricultural origins millet and rice agriculture, we still need to consider how or and the isotopic identity of domestication in northern China. Pro- even whether knowledge of a millet-centered dryland or a ceedings of the National Academy of Sciences of the USA 106(14): rice-centered wetland system could drive the development of 5523–5528. the other. Could the particular steps in the development of Bar-Yosef, Ofer. 2011. Climatic fluctuations and early farming in West and East Asia. Current Anthropology 52(suppl. 4):S175–S193. agriculture centered on one species “jump” across to the other Beijing da xue kao gu wen bo xue yuan, Beijing da xue kao gu xue species, when rice and millet have biological requirements yan jiu zhong xin, and Beijing shi wen wu yan jiu suo (Beijing that are so different? This is a question for future consider- Donghulin). 2006. Beijing shi Mentougou qu Donghulin shi qian ation. To answer it, we need a better appreciation of the many yi zhi (The prehistoric site of Donghulin in Mentougou, Beijing). steps involved. At present, we do not have enough archaeo- Kao gu 2006(7):3–8, pls. 1–2. Bellwood, Peter. 2006. Asian farming diasporas? agriculture, lan- logical and paleoethnobotanical data to establish the basic guages, and genes in China and Southeast Asia. In Archaeology of steps in the development of agriculture within each region or Asia. Mariam T. Stark, ed. Pp. 96–118. Malden, MA: Blackwell. the chronological control to properly order them across ———. 2011. Holocene population history in the Pacific region as regions. One possibility is that agriculture, the end product a model for worldwide food producer dispersals. Current Anthro- of a 3-millennia process in each region, may be an admixture pology 52(suppl. 4):S363–S378. Bettinger, Robert L., Loukas Barton, Christopher Morgan, Fahu of small steps coming from different regional sources over Chen, Hui Wang, Thomas P. Guilderson, Duxue Ji, and Dongju this lengthy time period: a new development in one region Zhang. 2010. The transition to agriculture at Dadiwan, People’s could be related to a previous development in another region Republic of China. Current Anthropology 51(5):703–714. and onward across regions. Such a possibility needs consid- Bettinger, Robert L., Loukas Barton, Peter J. Richerson, Robert Boyd, eration because we already have good evidence for the bound- Hui Wang, and Won Choi. 2007. The transition to agriculture in northwestern China. In Late Quaternary climate change and human aries of a material culture system—the Peiligang culture— adaptation in arid China. D. B. Madsen, F. Chen, and X. Gao, eds. extending across the boundary between where human groups Pp. 83–101. Developments in Quaternary Science 9. Amsterdam: were exploiting rice on one side and millet on the other, as Elsevier. at Jiahu, a site whose material culture classifies well within Boaretto, Elisabetta, Xiaohong Wu, Jiarong Yuan, Ofer Bar-Yosef, the millet-exploiting Peiligang cultural sphere but that has no Vikki Chu, Yan Pan, Kexin Liu, et al. 2009. Radiocarbon dating of charcoal and bone collagen associated with early pottery at millet, only rice. The stage is set for better understanding the Yuchanyan Cave, Hunan Province, China. Proceedings of the Na- lengthy and likely interrelated processes leading to the de- tional Academy of Sciences of the USA 106(24):9595–9600. velopment of millet- and rice-based agricultural systems in Cai, Yizhi. 2000. Guangdong zao qi nong ye fa zhan kui shi (A look Northeast China, North China, and the Middle and Lower at the development of early agriculture in Guangdong). Nong ye Yangtze regions. An interregional perspective can guide the kao gu 2000(1):59–64. Chen, Yuanbei. 2006. Huanan di qu shi qian mu zang tan xi (As- gathering of new data to help answer the many questions we certaining about prehistoric graves in southern China). In Huanan still have. ji Dong Nan Ya di qu shi qian kao gu—Ji nian Zengpiyan yi zhi fa jue 30 zhou nian guo ji xue shu yan tao hui lun wen ji. Zhongguo ke xue yuan kao gu yan jiu suo, ed. Pp. 207–222. Beijing: Wen Chinese and Japanese Characters wu chu ban she. Cohen, David J. 1998. The origins of domesticated cereals and the In the “References Cited” below, Chinese and Japanese titles Pleistocene-Holocene transition in East Asia. Review of Archaeology and author names are romanized. The full reference list with 19(2):22–29. ———. 2002. New perspectives on the transition to agriculture in Chinese and Japanese characters included is available as a China. In The origins of pottery and agriculture. Y. Yasuda, ed. Pp. PDF in the online edition. 217–227. New Delhi: Roli. ———. 2003. Microblades, pottery, and the nature and chronology of the Palaeolithic-Neolithic transition in China. Review of Ar- chaeology 24(2):21–36. References Cited Crawford, Gary W. 2006. East Asian plant domestication. In Ar- chaeology of Asia. M. T. Stark, ed. Pp. 77–95. Malden, MA: Black- An, Jiayuan. 2004. Zhongguo xin shi qi shi dai de jia zhu (Domestic well. pigs in Neolithic China). In Tao li cheng xi ji: qing zhu An Zhimin ———. 2009. Agricultural origins in North China pushed back to xian sheng ba shi shou chen. Deng Cong and Chen Xingcan, eds. the Pleistocene-Holocene boundary. Proceedings of the National Pp. 67–71. Xianggang zhong wen da xue Zhongguo wen hua yan Academy of Sciences of the USA 106(18):7271–7272. jiu suo Zhongguo kao gu yi shu yan jiu zhong xin zhuan kan 12. ———. 2011. Advances in understanding early agriculture in Japan. Hong Kong: Zhongguo kao gu yi shu yan jiu zhong xin. Current Anthropology 52(suppl. 4):S331–S345. An, Zhimin. 2000. Zhongguo xi shi qi fa xian yi bai nian 100 (100 Crawford, Gary W., Chen Xuexiang, and Wang Jianhua. 2006. Shan- years since the discovery of microliths in China). Kao gu 2000(5): dong Ji’nan Changqing qu Yuezhuang yi zhi fa xian Houli wen 45–56. hua shi qi de tan hua dao (Discovery of Houli culture period Bai, Yunxiang, and Zhang Jianfeng. 2001. Huang he liu yu qian qi carbonized rice at the Yuezhuang site, Changqing district, Ji’nan, xin shi qi shi dai mu zang de yan jiu (Research on graves of the Shandong). Dong fang kao gu 3:247–251. S290 Current Anthropology Volume 52, Supplement 4, October 2011

Cui, Jianxin, and Zhou Shangzhe. 2008. Zhongguo xin shi qi shi dai bunka. Wasano Kikuo, ed. Pp. 138–143. Saga: Saga University Ag- wen hua yan bian dong li fen xi (An analysis of the evolutionary riculture Department. forces of Neolithic period cultures in China). Zhongguo li shi di li Hebei sheng wen wu guan li chu and Handan shi wen wu bao guan lun cong 23(2):33–45. suo (Hebei Cishan). 1981. Hebei Wuan Cishan yi zhi (The Cishan Duan, Hongzhen, and Hebei sheng wen wu yan jiu suo (Duan Bei- site, Wuan, Hebei). Kao gu xue bao 1981(3):303–338. fudi). 2007. Beifudi Yishui liu yu shi qian yi zhi (The Beifudi pre- Henan sheng wen wu kao gu yan jiu suo (Henan Jiahu). 1999. historic site in the Yishui River valley [Hebei]). Beijing: Wen wu Wuyang Jiahu (Jiahu site, Wuyang [Henan]). Beijing: Ke xue chu chu ban she. ban she. Elston, Robert G., and P. Jeffrey Brantingham. 2002. Microlithic tech- Hong, Qian, and Robert E. Ricklefs. 2001. Diversity of temperate nology in northern Asia: a risk-minimizing strategy of the Late plants in East Asia. Nature 413:129–130. Paleolithic and Early Holocene. In Thinking small: global perspec- Hu, Yaowu, Stanley Ambrose, and Changsui Wang. 2006. Stable iso- tives on microlithization. Robert G. Elston and Steven L. Kuhn, topic analysis of human bones from Jiahu site, Henan, China: eds. Pp. 103–116. Archaeological Papers of the American Anthro- implications for the transition to agriculture. Journal of Archaeo- pological Association 12. Washington, DC: American Anthropo- logical Science 33:1319–1330. logical Association. Hu, Yaowu, Shougong Wang, Fengshi Luan, Changsui Wang, and Fuller, Dorian Q, Emma Harvey, and Ling Qin. 2007. Presumed Michael P. Richards. 2008. Stable isotope analysis of humans from domestication? evidence for wild rice cultivation and domestica- Xiaojingshan site: implications for understanding the origin of tion in the fifth millennium BC of the Lower Yangtze region. millet agriculture in China. Journal of Archaeological Science 35: Antiquity 81(312):316–331. 2960–2965. Fuller, Dorian Q, and Ling Qin. 2008. Immature rice and its ar- Hunan sheng wen wu kao gu yan jiu suo (Hunan Pengtoushan). chaeobotanical recognition: a reply to Pan. Antiquity 82(316): on- 2006. Pengtoushan yu Bashidang (Pengtoushan and Bashidang). line project gallery, http://www.antiquity.ac.uk/projgall/fuller316/. Beijing: Ke xue chu ban she. ———. 2009. Water management and labour in the origins and Hunan sheng wen wu kao gu yan jiu suo and He Jiejun (Hunan dispersal of Asian rice. World Archaeology 41(1):88–111. Chengtoushan). 2007. Li Xian Chengtoushan: xin shi qi shi dai yi Fuller, Dorian Q, Ling Qin, and Emma Harvey. 2008a. A critical zhi fa jue bao gao (The Chengtoushan site, Lixian County [Hunan]: assessment of early agriculture in East Asia, with emphasis on excavation report for the Neolithic site). Beijing: Wen wu chu ban Lower Yangtze rice domestication. Pradghara (Journal of the Uttar she. Pradesh State Archaeology Department) 18:17–52. Institute of Archaeology, CASS. 1991. Radiocarbon dates in Chinese ———. 2008b. Rice archaeobotany revisited: comments on Liu archaeology 1965–1991. Beijing: Wen wu chu ban she. et al. (2007). Antiquity 82(315): online project gallery, http:// Jiang, Leping, and Li Liu. 2006. New evidence for the origins of antiquity.ac.uk/projgall/fuller1/index.html. sedentism and rice domestication in the Lower YangziRiver, China. Fuller, Dorian Q, Ling Qin, Yunfei Zheng, Zhijun Zhao, Xugao Chen, Antiquity 80(308):355–361. Leo Aoi Hosoya, and Guoping Sun. 2009. The domestication pro- Jones, Martin K., and Xinyi Liu. 2009. Archaeology: origins of ag- cess and domestication rate in rice: spikelet bases from the Lower riculture in East Asia. Science 324(5928):730–731. Yangtze. Science 323:1607–1610. Kaifeng di qu wen guan hui and Xinzheng xian wen guan hui (Pei- Fuller, Dorian Q, Yo-ichiro Sato, Cristina Castillo, Ling Qin, Alison ligang A). 1978. Henan Xinzheng Peiligang xin shi qi shi dai yi R. Weisskopf, Eleanor J. Kingwell-Banham, Jixiang Song, Sung- zhi (The Peiligang Neolithic site in Xinzheng, Henan). Kao gu Mo Ahn, and Jacob van Etten. 2010. Consilience of genetics and 1978(2):73–79. archaeobotany in the entangled history of rice. Archaeological and Kuzmin, Yaroslav V. 2006. Chronology of the earliest pottery in East Anthropological Sciences 2(2):115–131. Asia: progress and pitfalls. Antiquity 80(308):362–371. Gansu sheng wen wu kao gu yan jiu suo (Gansu Dadiwan). 2006. Kuzmin, Yaroslav V., Susan G. Keates, and Chen Shen, eds. 2007. Qin’an Dadiwan: xin shi qi shi dai yi zhi fa jue bao gao (Dadiwan Origin and spread of microblade technology in Northern Asia and site, Qin’an: excavation report for the Neolithic site). Beijing: Wen North America. Burnaby, British Columbia: Archaeology, Simon wu chu ban she. Fraser University. Gu, Haibin, and Yuan Jiarong. 2006. Yuchanyan yi zhi wei ti luo lei Larson, Greger, Ranan Liu, Xingbo Zhan, Jing Yuan, Dorian Fuller, dui tan tao gu ren lei huo dong te zheng de zhi shi yi yi (The Loukas Barton, Keith Dobney, et al. 2010. Patterns of East Asian indicative significance of microscopic snails from the Yuchanyan pig domestication, migration, and turnover revealed by modern site on investigating the characteristics of ancient human activity). and ancient DNA. Proceedings of the National Academy of Sciences Di si ji yan jiu 26(4):563–570. of the USA 107(17):7686–7691. Guo jia wen wu ju. 2006. Zhejiang Shengzhou Xiaohuangshan yi zhi Lee, Gyoung-Ah. 2011. The transition from foraging to farming in fa jue (The excavation of the Xiaohuangshan site, Shengzhou, prehistoric Korea. Current Anthropology 52(suppl. 4):S307–S329. Zhejiang). In 2005 Zhongguo zhong yao kao gu fa xian. Guo jia Li, Xueqin, Garman Harbottle, Juzhang Zhang, and Changsui Wang. wen wu ju, eds. Pp. 9–13. Beijing: Wen wu chu ban she. 2003. The earliest writing? sign use in the seventh millennium BC Guo, Ruihai, and Li Jun. 2000. Cong Nanzhuangtou yi zhi kan Hu- at Jiahu, Henan Province, China. Antiquity 77:31–44. abei di qu nong ye he tao qi de qi yuan (Exploring the origins of Li, Youmou. 2003. Peiligang wen hua (The Peiligang culture). Beijing: agriculture and pottery in North China through the Nanzhuangtou Wen wu chu ban she. site). In Daozuo taoqi he dushi de qiyuan. Yan Wenming and Yasuda Liu, Guoxiang. 2000. Guan yu Zhaobaogou wen hua de ji ge wen ti Yoshinori, eds. Pp. 51–63. Beijing: Wenwu chu ban she. (Concerning a few questions about the Zhaobaogou culture). Bei ———. 2002. The Nanzhuangtou and Hutouliang sites: exploring fang wen wu 2000(2):8–17. the beginnings of agriculture and pottery in North China. In The ———. 2006a. Xinglonggou yi zhi di i di dian fa jue hui gu yu si origins of pottery and agriculture. Yoshinori Yasuda, ed. Pp. 193– kao (A review and considerations of the excavations of locality 204. New Delhi: Roli. one of the Xinglonggou site). Neimenggu wen wu kao gu 2006(2): Hayden, Brian. 2009. The proof is in the pudding: feasting and the 8–15, 30. origins of domestication. Current Anthropology 50(5):597–601. ———. 2006b. Xinglongwa wen hua yu Fuhe wen hua bi jiao yan He, Jiejun. 1995. Ho¯to¯zan iseki no inasaku ibutsu (Rice remains at jiu (Comparative research on the Xinglongwa culture and Fuhe Pengtoushan). In Higashi Ajia no inasaku kigen to kodai inasaku culture). Bei fang wen wu 2000(2):1–10. Cohen Beginnings of Agriculture in China S291

Liu, Li. 2008. When an issue becomes ethical in academic debate: Prendergast, Mary E., Jiarong Yuan, and Ofer Bar-Yosef. 2009. Re- response to Fuller et al. Antiquity 82(316), online project gallery, source intensification in the Late Upper Paleolithic: a view from http://antiquity.ac.uk/projgall/liu2/index.html. southern China. Journal of Archaeological Science 36:1027–1037. Liu, Li, Gyoung-Ah Lee, and Leping Jiang. 2007. Evidence for the Qin, Ling, Dorian Q Fuller, and Zhang Hai. 2010. Zao qi nong ye early beginning (c. 9000 cal. BP) of rice domestication in China: ju luo de ye sheng shi wu zi yuan yu yan jiu: yi Changjiang xia a response. Holocene 17(8):1059–1068. you he Zhongyuan di qu wei li (Modeling wild food resource Liu, Li, Gyoung-Ah Lee, Leping Jiang, and Juzhong Zhang. 2007. catchments among early agricultural settlements: case studies from The earliest rice domestication in China. Antiquity 81(313): online the Lower Yangzi and Central Plains). Quaternary Sciences 30(2): project gallery, http://antiquity.ac.uk/projGall/liu1/index.html. 245Ϫ261. Liu, Xinyi, Harriet V. Hunt, and Martin K. Jones. 2009. River valleys Qu, Tongli, Ofer Bar-Yosef, Youping Wang, and Xiaohong Wu. Forth- and foothills: changing archaeological perceptions of North coming. The geography, chronology, and techno-typology of the China’s earliest farms. Antiquity 83(319):82–95. Chinese Upper Paleolithic. Journal of Archaeological Research. Lu, Houyuan, Jianping Zhang, Kam-biu Liu, Naiqin Wu, Kunshu Ren, Mei’e, Renzhang Yang, and Haosheng Bao. 1985. An outline of Zhou, Maolin Ye, Tianyu Zhang, et al. 2009. Earliest domestication China’s physical geography. Beijing: Foreign Languages. of common millet (Panicum miliaceum) in East Asia extended to Ren, Shinan. 2005. Shi qian nong ye kao gu de xin jin zhan (New 10,000 years ago. Proceedings of the National Academy of Sciences progress in prehistoric agricultural archaeology). In Xin shi ji de of the USA 106(18):7367–7372. Zhongguo kao gu xue: Wang Zhongshu xian sheng ba shi hua dan Lu, Tracey Lie Dan. 1999. The transition from foraging to farming and ji nian lun wen ji. Zhongguo she hui ke xue yuan kao gu yan jiu the origin of agriculture in China. BAR International Series, vol. suo, ed. Pp. 74–85. Beijing: Ke xue chu ban she. 774. Oxford: British Archaeological Reports. Ruddiman, William F., Zhengtang Guo, Xin Zhou, Hanbin Wu, and ———. 2005. The origin and dispersal of agriculture and human Yanyan Yu. 2008. Early rice farming and anomalous methane diaspora in East Asia. In The peopling of East Asia: putting together trends. Quaternary Science Reviews 27(13–14):1291–1295. archaeology, linguistics and genetics. R. M. Blench, L. Sagart, and Shelach, Gideon. 2000. The earliest Neolithic cultures of Northeast A. Sanchez-Mazas, eds. Pp. 51–62. London: Curzon. China: recent discoveries and new perspectives on the beginning ———. 2006. The occurrence of cereal cultivation in China. Asian of agriculture. Journal of World Prehistory 14(4):363–413. Perspectives 45(2):129–158. ———. 2006. Economic adaptation, community structure, and shar- MacNeish, Richard S., Geoffrey Cunnar, Zhijun Zhao, and Jane G. ing strategies of households at early sedentary communities in Libby. 1998. Re-revised second annual report of the Sino-American Northeast China. Journal of Anthropological Archaeology 25(3):318– Jiangxi (PRC) Origin of Rice Project (SAJOR). Andover, MA: An- 345. dover Foundation for Archaeological Research. Sheng, Danping, Zheng Yunfei, and Jiang Leping. 2006. Zhejiang Mao, Longjiang, Duowen Mo, Leping Jiang, Yaofeng Jia, Xiaoyan Pujiang xian Shangshan xin shi qi shi dai zao qi yi zhi: changjiang Liu, Minglin Li, Kunshu Zhou, and Chenxi Shi. 2008. Environ- xia you wan nian qian dao zuo yi cun de zui xin fa xian (The mental change since mid-Pleistocene recorded in Shangshan Shangshan Early Neolithic site in Pujiang County, Zhejiang: newest achaeological [sic] site of Zhejiang. Journal of Geographical Sciences discoveries of 10,000 year old rice remains from the lower Chang- 18(2):247–256. jiang River). Nong ye kao gu 2006(1):30–32, cover plate. Morrill, Carrie, Jonathan T. Overpeck, and Julia E. Cole. 2003. A Smith, Barbara Li, and Yun Kuen Lee. 2008. Mortuary treatment, synthesis of abrupt changes in the Asian summer monsoon since pathology, and social relations of the Jiahu community. Asian Per- the last deglaciation. Holocene 13(4):465–476. spectives 47(2):242–298. Nakamura, Shin-ichi. 2010. The origin of rice cultivation in the Sun, Jiahua, and Kaixun Zhan. 2004. Xue bai dao mi cong zhe li Lower Yangtze region, China. Archaeological and Anthropological zhong qi: Wannian Xianrendong he Diaotonghuan yi zhi (Snow Sciences 2(2):107–113. white grains of rice started being grown here: the Xianrendong Pan, Yan. 2008. Immature wild rice harvesting at Kuahuqiao, China? and Diaotonghuan sites in Wannian). In Shou chan xia de wen Antiquity 82(316): online project gallery, http://www ming: Jiangxi zhong da kao gu fa xian. Sun Jiahua and Zhan Kaixun, .antiquity.ac.uk/projgall/pan316/. eds. Pp. 1–43. Nanchang: Jiangxi ren min chu ban she. Pearson, Richard. 2005. The social context of early pottery in the Tong, Peihua. 2006. Shandong Houli wen hua ju luo xing tai chu Lingnan region of south China. Antiquity 79(306):819–828. tan (Preliminary explorations of the settlement patterns of the Pechenkina, Ekaterina A., Stanley H. Ambrose, Xiaolin Ma, and Rob- Houli culture in Shandong). In Er shi yi shi ji de Zhongguo kao ert A. Benfer Jr. 2005. Reconstructing northern Chinese Neolithic gu xue: Qing zhu Tong Zhuchen xian sheng ba shi wu hua dan xue subsistence practices by isotopic analysis. Journal of Archaeological shu wen ji. Zhongguo ke xue yuan kao gu yan jiu suo, ed. Pp. Science 32:1176–1189. 173–192. Beijing: Wen wu chu ban she. Pei, Anping. 2002. Rice paddy agriculture and pottery from the mid- Underhill, Anne P., and Junko Habu. 2006. Early communities in dle reaches of the Yangtze River. In The origins of pottery and East Asia: economic and sociopolitical organization at the local agriculture. Yoshinori Yasuda, ed. Pp. 167–181. New Delhi: Roli. and regional levels. In Archaeology of Asia. M. T. Stark, ed. Pp. ———. 2006 (2004). Zhongguo shi qian de ju luo wei gou (Pre- 121–148. Malden, MA: Blackwell. historic settlement enclosure ditches in China). In Nong ye wen Wang, Xin. 2006. Changjiang xia you yuan shi wen ming xin yuan hua she hui: shi qian kao gu wen ji. Pei Anping, ed. Pp. 280–295. tou: Zhejiang Shengzhou Xiaohuangshan xin shi qi shi dai zao qi Beijing: Ke xue chu ban she. yi cun de kao gu xue yan tao (A new starting point for the primitive Peng, Shifan, and Zhou Guangming. 2006. Jiangxi Wannian Xian- civilization in the Lower Changjiang River: archaeological delib- rendong yu Diaotonghuan yi zhi: jiu shi qi shi dai xiang xin shi erations on the Early Neolithic remains at Xiaohuangshan, Sheng- qi shi dai guo du mo shi de ge an yan jiu (The Xianrendong and zhou, Zhejiang). Wen bo 2006(4):72–77. Diaotonghuan sites in Wannian, Jiangzi: a case study on modeling Wu, Xiaohong, and Chaohong Zhao. 2003. Chronology of the tran- the Paleolithic to Neolithic transition). In Hua nan ji Dong Nan sition from Palaeolithic to Neolithic in China. Review of Archae- Ya di qu shi qian kao gu: ji nian Zengpiyan yi zhi fa jue 30 zhou ology 24(2):15–20. nian guo ji xue shu yan tao hui lun wen ji 30. Zhongguo ke xue Wu, Yaoli. 1997. Xin shi qi shi dai zao qi wen hua tao san zu qi chu yuan kao gu yan jiu suo, ed. Pp. 102–115. Beijing: Wen wu chu lun (A preliminary discussion of tripod ceramic vessels of Early ban she. Neolithic cultures). Kao gu 1997(3):58–67. S292 Current Anthropology Volume 52, Supplement 4, October 2011

———. 2002. Zhongguo shi qian ju luo xing tai de yan bian (The (Research on prehistoric settlements of the Middle and Lower evolution of settlement patterns in prehistoric China). In 21 shi Yangzi River regions). Beijing: Wen wu chu ban she. ji Zhongguo kao gu xue yu shi ji kao gu xue (Twenty-first century Zhang, Chi, and Hsiao-chun Hung. 2008. The Neolithic of southern Chinese archaeology and world archaeology). Zhongguo she hui China: origin, development, and dispersal. Asian Perspectives 47(2): ke xue yuan kao gu yan jiu suo, ed. Pp. 89–105. Beijing: Zhongguo 299–329. she hui ke xue chu ban she. ———. 2010. The emergence of agriculture in southern China. An- Xia, Zhengkai, Ge Chen, Gongwang Zheng, Junqing Han, and Fuyou tiquity 84(323):11–25. Chen. 2002. Climate background of the evolution from Paleolithic Zhang, Juzhong, Xinghua Xiao, and Yun Kuen Lee. 2004. The early to Neolithic cultural transition during the last deglaciation in the development of music: analysis of the Jiahu bone flutes. Antiquity middle reaches of the Yellow River. China Science Bulletin (English 78(302):769–778. edition) 2002(1):71–75. Zhang, Songlin, and Xin Yingjun. 2008. Henan Xinzheng Tanghu Yan, Wenming. 2006. Zhongguo shi qian ju luo de kao gu yan jiu xin shi qi shi dai yi zhi (The Tanghu Neolithic site in Xinzheng, (Archaeological research on prehistoric settlements in China). In Henan). In 2007 Zhongguo zhong yao kao gu fa xian (Major ar- Er shi yi shi ji de Zhongguo kao gu xue: Qing zhu Tong Zhuchen chaeological discoveries in China, 2007). Guo jia wen wu ju, ed. xian sheng ba shi wu hua dan xue shu wen ji. Zhongguo ke xue Pp. 6–9. Beiing: Wen wu chu ban she. yuan kao gu yan jiu suo, ed. Pp. 107–120. Beijing: Wen wu chu Zhang, Wenxu, and Pei Anping. 1997. Li xian Mengxi Bashidang chu ban she. tu dao gu de yan jiu (Rice remains unearthed from Bashidang, Yang, Shiling, and Zhongli Ding. 2008. Advance-retreat history of Mengxi, Lixian County). Wen wu 1997(1):36–41. the East-Asian summer monsoon rainfall belt over Northern China Zhang, Zhiheng. 2007. Zhongguo xin shi qi shi dai tao yan qi de during the last two glacial-interglacial cycles. Earth and Planetary yan hua gui lu¨ (Patterns of evolution of ceramic cooking vessels Science Letters 274(3–4):499–510. in Neolithic China). Zhong yuan wen wu 2007(4):44–47. Yasuda, Yoshinori, ed. 2002. The origins of pottery and agriculture. Zhao, Chaohong, Wang Tao, Wu Xiaohong, Liu Mingli, Yuan Xue- New Delhi: Roli. mei, Yu Jincheng, and Guo Jing. 2006. Donghulin, Zhuannian, Yasuda, Yoshinori, Toshiyuki Fujiki, Hiroo Nasu, Megumi Kato, Nanzhuangtou yu Yujiagou (Donghulin, Zhuannian, Nanzhuang- Yoshimune Morita, Yuichi Mori, Masaaki Kanehara, Shuichi To- tou, and Yujiagou sites). In Hua Nan ji Dong Nan Ya di qu shi yama, Azusa Yano, and Mitsuru Okuno. 2004. Environmental ar- qian kao gu: Ji nian Zengpiyan yi zhi fa jue 30 zhou nian guo ji chaeology at the Chengtoushan site, Hunan Province, China, and xue shu yan tao hui lun wen ji. Zhongguo ke xue yuan kao gu yan implications for environmental change and the rise and fall of the jiu suo, ed. Pp. 116–127. Beijing: Wen wu chu ban she. Yangtze River civilization. Quaternary International 123–125:149– Zhao, Zhijun. 2005. You guan Zhongguo nong ye qi yuan de xin zi 158. liao he xin si kao (New materials and new ideas on the origins of Yi, Sangheon, and Yoshiki Saito. 2004. Latest Pleistocene climate agriculture in China). In Xin shi ji de Zhongguo kao gu xue: Wang variation of the East Asian monsoon from pollen records of two Zhongshu xian sheng ba shi hua dan ji nian lun wen ji. Zhongguo East China regions. Quaternary International 121(1):75–87. she hui ke xue yuan kao gu yan jiu suo, ed. Pp. 86–101. Beijing: Yi, Sangheon, Yoshiki Saito, Quanhong Zhao, and Pinxian Wang. Ke xue chu ban she. 2003. Vegetation and climate changes in the Changjiang (Yangtze ———. 2010. New data and new issues for the study of origin of River) Delta, China, during the past 13,000 years inferred from rice agriculture in China. Archaeological and Anthropological Sci- pollen records. Quaternary Science Reviews 22:1501–1519. ences 2(2):99–105. ———. 2011. New archaeobotanic data for the study of the origins Yin, Jianshun. 2006. Pengtoushan wen hua de bian nian ji qi ju luo of agriculture in China. Current Anthropology 52(suppl. 4):S295– te zheng (Chronology of the Pengtoushan culture and the char- S306. acteristics of its settlements). In Hua Nan ji Dong Nan Ya di qu Zhao, Zhijun, and Zhang Juzhong. 2009. Jiahu yi zhi 2001 nian du shi qian kao gu: ji nian Zengpiyan yi zhi fa jue 30 zhou nian guo fu xuan jie guo fen xi bao gao 2001 (Report on the analysis of ji xue shu yan tao hui lun wen ji. Zhongguo ke xue yuan kao gu results from the 2001 flotation work at the Jiahu site). Kao gu yan jiu suo, ed. Pp. 444–465. Beijing: Wen wu chu ban she. 2009(8):86–95. Yuan, Daoxian, Hai Cheng, R. Lawrence Edwards, Carolyn A. Dy- Zhejiang sheng wen wu kao gu yan jiu suo (Zhejiang Hemudu). koski, Megan J. Kelly, Meiliang Zhang, Jiaming Qing, et al. 2004. 2003. Hemudu: xin shi qi shi dai yi zhi kao gu fa jue bao gao Timing, duration, and transitions of the Last Interglacial Asian (Hemudu: excavation report for the Neolithic site). Beijing: Wen Monsoon. Science 304(5670):575–578. wu chu ban she. Yuan, Jiarong. 2002. Rice and pottery 10,000 yrs. BP at Yuchanyan, Zhejiang sheng wen wu kao gu yan jiu suo and Pujiang bo wu guan Dao County, Hunan Province. In The origins of pottery and ag- (Zhejiang Shangshan). 2007. Zhejiang Pujiang xian Shangshan yi riculture. Yoshinori Yasuda, ed. Pp. 157–166. New Delhi: Roli. zhi fa jue jian bao (Short report on the excavation of the Shangshan Yuan, Jing. 2003. The problem of the origin of domestic animals in site, Pujiang County, Zhejiang). Kao gu 2007(9):7–18, pls. 1–3. Neolithic China. Chinese Archaeology 3:154–156. Zhejiang sheng wen wu kao gu yan jiu suo and Xiaoshan bo wu ———. 2010. Zhongguo gu dai jia yang dong wu de dong wu kao guan (Zhejiang Kuahuqiao). 2004. Kuahuqiao (Kuahuqiao site). gu xue yan jiu (The zooarchaeology of domesticated animals in Beijing: Wen wu chu ban she. ancient China). Quaternary Sciences 30(2):298–306. Zhejiang sheng wen wu kao gu yan jiu suo, Yuyao shi wen wu bao Yuan, Jing, Rowan Flad, and Yunbing Luo. 2008. Meat-acquisition hu guan li suo, and Hemudu yi zhi bo wu guan (Zhejiang Tian- patterns in the Neolithic Yangzi River valley, China. Antiquity luoshan). 2007. Zhejiang Yuyao Tianluoshan xin shi qi shi dai yi 82(316):351–366. zhi 2004 nian fa jue jian bao 2004 (Short report on the 2004 Zhang, Chi. 2002a. The discovery of early pottery in China. Docu- excavations at the Tianluoshan Neolithic site, Yuyao, Zhejiang). menta Praehistorica 29:29–35. Wen wu 2007(11):4–24, 73. ———. 2002b. Early pottery and rice phytolith remains from Xian- Zheng, Yunfei, Guoping Sun, and Xugao Chen. 2007. Characteristics rendong and Diaotonghuan sites, Wannian, Jiangxi Province. In of the short rachillae of rice from archaeological sites dating to The origins of pottery and agriculture. Yoshinori Yasuda, ed. Pp. 7000 years ago. China Science Bulletin (English edition) 52(12): 185–191. New Delhi: Roli. 1654–1660. ———. 2003. Changjiang zhong xia you di qu shi qian ju luo yan jiu Zheng, Yunfei, Guoping Sun, Ling Qin, Chunhai Li, Xiaohong Wu, Cohen Beginnings of Agriculture in China S293

and Xugao Chen. 2009. Rice fields and modes of rice cultivation of the Chinese Neolithic, with discussion of the signs of the start between 5000 and 2500 BC in East China. Journal of Archaeological of the Chinese Neolithic). In 21 shi ji Zhongguo kao gu xue yu shi Science 36:2609–2616. ji kao gu xue (Twenty-first century Chinese archaeology and world Zhongguo she hui ke xue yuan kao gu yan jiu suo (Peiligang B). archaeology). Zhongguo she hui ke xue yuan kao gu yan jiu suo, 1982. 1979 nian Peiligang yi zhi fa jue jian gao (Short report on ed. Pp. 89–105. Beijing: Zhongguo she hui ke xue chu ban she. the 1979 excavations at Peiligang). Kao gu 1982(4):337–341. Zong, Y., Z. Chen, J. B. Innes, C. Chen, Z. Wang, and H. Wang. Zhu, Naicheng. 2002. Zhongguo xin shi qi shi dai ji zhong zhu yao 2007. Fire and flood management of coastal swamp enabled te zheng de qi yuan jian lun Zhongguo xin shi qi shi dai kai shi first rice paddy cultivation in East China. Nature 449(7161): de biao zhi (The origins of several kinds of major characteristics 459–462.

Current Anthropology Volume 52, Supplement 4, October 2011 S295

New Archaeobotanic Data for the Study of the Origins of Agriculture in China

by Zhijun Zhao

In the past 10 years, flotation techniques have been introduced and implemented in Chinese ar- chaeology. As a result, a tremendous quantity of plant remains have been recovered from archaeo- logical sites located all over China. These plant remains include crops that might have been do- mesticated in China—such as rice, foxtail millet, broomcorn millet, and soybean—as well as crops that were introduced into China from other parts of world—such as wheat and barley. The new archaeobotanic data provide direct archaeological evidence for the study of the origins and devel- opment of agriculture in China. This paper attempts a synthesis of these new archaeobotanic data while presenting some new ideas about the origins and development of ancient agriculture in China, including the rice agriculture tradition that originated around the middle and lower Yangtze River areas; the dry-land agriculture tradition, with millets as major crops, centered in North China; and the ancient tropical agriculture tradition located in the tropical parts of China, where the major crops seem to be roots and tubers, such as taro.

Introduction pathway of early Chinese agricultural development in a pre- historic context. There remain many unsolved questions in the study of the origins of agriculture in China that are partly China is one of the major centers for the origins of agriculture. due to the lack of reliable archaeological evidence. In the past, intensive study of early Chinese agriculture made Ancient agriculture is generally considered to be composed it clear that there were primarily two independent subcenters of two major elements: crop farming and animal husbandry, of origin within China (An 1989; Cohen 2011; Crawford 1992; which relied on crops or by-products of farming for fodder. Ho 1977; Jones and Liu 2009; Yan 1992; Zhao 2005c): one in Thus, plant remains, especially crop remains, are the most the middle and lower Yangtze River areas, where rice agri- important evidence in the study of origins of agriculture. culture was first developed, with rice (Oryza sativa) the major However, plant remains are more often than not poorly pre- cultivar; and the other in North China along the Yellow River, served, fairly small in size, and consequently difficult to iden- where the origin of dry-land agriculture is centered, with tify with the naked eye in the field. Thus, without special foxtail millet (Setaria italica) and broomcorn millet (Panicum attention to their recovery, plant remains from the past may miliaceum) the most representative crops. However, the origin be easily underrepresented in archaeological excavations. of domesticated rice and the beginning of rice farming in Well-preserved plant remains more frequently survive at China remain uncertain and highly controversial, while the archaeological sites where fire was used by the residents. study of the origin of dry-land agriculture has lagged even Charred plant remains are chemically stable and relatively farther behind. Apart from these two major areas central to resistant to soil and water erosion, and they can be preserved the origin of agriculture, a third important area that deserves for a long time in the sediments of archaeological sites. thorough investigation is the southernmost part of China, Charred plant remains are usually lighter in specific gravity centered along the Zhujiang River, where the major crops of than soil particles and water. This makes it possible to separate early farming seem to have been roots and tubers, probably the plant remains from the soil matrix, as the charred remains including taro (Colocasia sp.; Zhao 2006). This investigation will separate from sediment and float on the surface of water. should be of great importance because it may reveal a third Such is the widely used and recognized flotation technique, which has been commonly accepted as an effective way of Zhijun Zhao is Professor at the Institute of Archaeology, Chinese extracting plant remains from soil samples. Academy of Social Sciences (27 Wangfujing Street, Beijing, 100710, In recent years, flotation work has been carried out at more People’s Republic of China [[email protected]]). This than 80 archaeological sites throughout China. About 7,000 paper was submitted 13 XI 09, accepted 3 I 11, and electronically soil samples have been processed, and a tremendous quantity published 5 VIII 11. of charred plant remains have been recovered (Zhao 2005a).

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0010$10.00. DOI: 10.1086/659308 S296 Current Anthropology Volume 52, Supplement 4, October 2011

The charred plant material, especially the crop remains, ob- 8200–7000 cal BP (Zhejiang Institute of Archaeology and tained from the sites by flotation provide important new data Xiaoshan Museum 2004:228). for the study of the origins of agriculture. Identifiable do- Flotation work was conducted during the excavation in the mesticated taxa include rice, foxtail millet, broomcorn millet, Shangshan site in 2005. A diagnostic sampling strategy was wheat (Triticum aestivum), barley (Hordeum vulgare), soybean employed by which soil samples were collected from each (Glycine max), adzuki bean (Vigna angularis), oats (Avena unambiguous archaeological feature found in every layer. A sativa), buckwheat (Fagopyrum esculentum), and so on. froth-flotation device with a 0.2-mm mesh screen was used. The plant remains recovered by flotation in China, con- A total of 459 soil samples were collected and processed. Yet sidered together with plant remains from previous investi- only a very small number of charred plant seeds were obtained gations, allow us to gain a number of new insights into the from the samples. The most important result of the flotation origins of agriculture in China. This paper presents a brief work was the rice remains. A total of 33 charred rice grains outline of this new scenario based mainly on quantitative (fig. 2) and 7 charred rice spikelet bases were found in the analysis of the crop data derived from flotation samples col- soil samples. However, most of these rice remains were re- lected at several major archaeological sites. covered from the Kuahuqiao culture layers and only a few from the Shangshan culture layers. The Origin of Rice Agriculture in the Apart from flotation, other methods were also employed Yangtze River Areas for obtaining rice remains. For example, heaps of rice husks were found in burnt soil blocks from the early-period layers Before we go into the details of the new flotation evidence of the site (Jiang and Li 2006, fig. 6). More significant evidence of rice agriculture, it is necessary to clearly distinguish two of rice used by humans was found in the temper of pottery. terms for the study of the origin of rice agriculture: “do- A large number of pottery sherds were excavated from the mestication” and “cultivation.” Benefiting from the discussion Shangshan site. It is possible to identify with the naked eye during the Wenner-Gren Temozon conference, I believe that plant remains such as stems and leaves at the edge of these there is no inevitable relationship between rice domestication low-fired ceramic materials (Jiang and Li 2006, fig. 6). It is and the beginning of rice cultivation. fairly interesting that in the two ash pits of the Shangshan Rice domestication was primarily an evolutionary process culture period we examined, more than 80% of the pottery influenced by human activities that were by no means a de- sherds were tempered with plant remains and that some rice liberate effort aimed at changing the genetic or biological husks were also observed in the temper. All this evidence characteristics of plants but were simply intended to increase indicates that the Shangshan people had a close relationship the yield of wild rice. People may have been unaware of the with rice and that rice, whether domesticated or wild, might characteristics of a new species when it first appeared, and have been one of the foods consumed by the Shangshan res- the plants were therefore not viewed as “domestic” at all. It idents. is possible for domesticated rice to occur in a typical hunting- Some sort of rice cultivation activity might have taken place gathering adaptation, and the occurrence of domesticated rice at the Shangshan site during the Shangshan period whether might not immediately result in rice farming, which was char- or not such rice remains belong to the domesticated species. acterized by intentional production and utilization of do- Therefore, a complete study of the Shangshan site may pos- mesticated rice for food. sibly be used as a good case for the study of the beginning Also, the cultivation of rice might not be directly related to the domestication of rice. Cultivation is a series of inten- of rice cultivation in China. Future studies should take into tional human activities—such as planting, protecting, and account all factors, including the environmental setting of the harvesting—for increasing the yield of plants that could be site, the settlement pattern, pottery types, stone tools, and so morphologically and genetically either wild or domesticated. forth. Therefore, it is possible for early rice cultivation to be prac- As discussed at the Wenner-Gren Temozon conference, ag- ticed without domesticated rice. riculture can be defined as a subsistence economy character- The flotation work at the Shangshan site has provided some ized by farming, although hunting and gathering continue. clues for the study of early rice cultivation in China. The Therefore, the earliest archaeological site to date that clearly Shangshan site is an early Neolithic settlement located in the exhibits the characteristics of rice agriculture in China is the center of Zhejiang Province, which geologically belongs to the Jiahu site, which is located in the upper Huai River region lower Yangtze River region (fig. 1). The cultural assemblage in the center of China (fig. 1) and is dated to a period between at the site can be clearly divided into two periods. The early 9000 and 7800 cal BP by 19 radiocarbon dates (Henan In- period is named the Shangshan culture and is dated to around stitute of Archaeology 1999:515). The Jiahu site covers an area 10,000 cal BP by a number of radiocarbon dates, some of of 50,000 m2, within which the living areas, manufacturing which have been published (Jiang and Li 2006:356). The later areas, and cemeteries were organized in an orderly fashion. period belongs to the Kuahuqiao culture, which was distrib- The clear layout of the site documents a permanent village uted around Hangzhou Bay and is radiocarbon-dated to (Henan Institute of Archaeology 1999). The excavations in Zhao Origins of Agriculture in China S297

Figure 1. Locations of the sites with early rice remains in the Yangtze River areas: 1, Xianrendong/Diaotonghuan; 2, Shangshan; 3, Kuahuqiao; 4, Hemudu/Tianluoshan; 5, Liangzhu; 6, Jiahu; 7, Chengtoushan. the 1980s unearthed many charred grains that have been iden- the majority of food resources while domesticated rice might tified as domesticated rice (Zhang and Wang 1998). have played only a minor role in the subsistence of the Jiahu Flotation work at the Jiahu site began in 2001. A diagnostic people. sampling strategy was employed. A simple bucket-flotation A number of animal bones were unearthed during the ex- device was used, and plant remains were caught in a 0.2-mm cavation, and about 20 species were identified, among which mesh screen. A total of 125 soil samples were collected and dog is the only clearly domesticated animal. The identification processed. A large number of charred plant remains have been of domestic pig from Jiahu has been controversial (Yuan counted and identified, including about 4,100 seeds of 16 taxa 2001). A recent study suggests that the Jiahu pig was fully (Zhao and Zhang 2009). They can be divided into two groups: domesticated but was not significant for the meat supply (Luo edible plants and weedy grasses. The former embraces tubers and Zhang 2008). Nevertheless, zooarchaeological studies sug- (e.g., lotus roots Nelumbo nucifera), nuts (e.g., acorn and gest that provisioning of animal feed was not an important trapa), grains (e.g., rice and wild soybean Glycine soja), and part of Jiahu subsistence practices. so on. The latter includes such taxa as Digitaria and Echino- The abundance of fish bones and shells at the site is based chloa, among others. not only on the absolute count but also on the fact that almost More than 400 charred rice grains were found in the flo- all trash pits contained fish bones. This is a strong indication tation samples (fig. 2). From analysis of the shape and size that fishing played a very important role in the subsistence of the grains and the morphological characteristics of the economy. Considering that water plants such as lotus and embryos, and considering the fact that the Jiahu site is located trapa are also abundant in the plant assemblage, aquatic spe- far from the traditionally known distribution area of wild rice, cies, both animal and plant, were a dominant food resource we believe that the rice remains recovered at the Jiahu site for the Jiahu residents. are domesticated. Various quantitative measurements—such We might conclude from the above analyses that the overall as absolute counts, total weight, and ubiquity—were em- subsistence economy of the Jiahu people mainly relied on ployed to compare the plant remains recovered from the Jiahu fishing, hunting, and gathering and that agricultural products site (Zhao and Zhang 2009). These data make it clear that produced by rice farming and animal husbandry were only wild plants derived from gathering practices accounted for supplements to their diet. Therefore, the Jiahu site is a good Figure 2. Charred crops recovered by flotation: A, Shangshan rice; B, Jiahu rice; C, Tianluoshan rice; D, Xinglonggou broomcorn millets; E, Yuhuazhai broomcorn millets; F, Yuhuazhai foxtail millets. A color ver- sion of this figure is available in the online edition of Current Anthro- pology. Zhao Origins of Agriculture in China S299 example for the study of an early stage in the transition to the Hemudu culture, as seen at both the Tianluoshan site and rice agriculture, and it thus shows that the transition from the Hemudu site. hunting and gathering to rice farming was a gradual evolu- Thus, the above sites show that the transition to rice ag- tionary process rather than a revolutionary replacement. riculture from hunting and gathering was not a clear-cut rev- The discovery of the Hemudu site in the 1970s resounded olutionary change but a slow evolutionary process. During throughout the world. The site is located on the south bank this long-term process, hunting and gathering gradually of Hangzhou Bay, about 20 km from the present sea coast waned while rice agriculture slowly achieved a dominant po- (fig. 1). Cultural assemblages have been divided into four sition and finally became the major subsistence practice. On phases dated to a period between 7000 and 5800 cal BP by the basis of preliminary analysis of the flotation materials from 27 radiocarbon dates (Zhejiang Institute of Archaeology 2003: the Tianluoshan site, the subsistence pattern of the Hemudu 411). Because of the waterlogged environment of the site, culture seems to be still in the transition to rice agriculture. organic materials have been well preserved. Large numbers When was a hunting-and-gathering scenario completely re- of plant remains have been recovered and identified, including placed by rice-agriculture subsistence in the Yangtze River rice, trapa, acorns, and other edible plants (Liu 2006). The areas of China? We have no answer yet. More flotation work most important result of the excavation was the discovery of should be done in the future, especially at those sites im- large numbers of rice grains. Some scholars accordingly sug- portant in the key stages of early agriculture. This is the only gested that the Hemudu people might have had a very mature way that reliable evidence can be obtained. agricultural economy based on rice farming (Zhou 2002). Yet However, on the basis of previous archaeological discov- no quantitative analysis was carried out at that time, and it eries and studies, we believe that the overall subsistence econ- is therefore impossible to make comparisons between differ- omy of the Liangzhu culture, dated to a period between 5200 ent plant taxa. For this reason, the importance of rice culti- and 4300 cal BP in the lower Yangtze River region, already vation in the subsistence economy of the Hemudu society relied heavily on rice agriculture. Rice remains were occa- remained a question. However, the recent excavation of the sionally found at some archaeological sites of the Liangzhu Tianluoshan site, only 7 km from the Hemudu site, provides period in the past, but they cannot be used for a quantitative an opportunity to reexamine the unsolved questions about analysis because of the lack of systemic sampling strategies. the Hemudu site. In recent years, the flotation technique has been introduced, The Hemudu and Tianluoshan sites are similar in size, and new data has begun to emerge, but the results are not environmental context, and the nature of the cultural deposits. yet published. Nevertheless, some indirect evidence suggests Unlike at Hemudu, a systematic sampling strategy (i.e., a the establishment of full rice agriculture in the Liangzhu pe- composite sampling strategy) was applied during the exca- riod. vation of the Tianluoshan site. Both water screening and flo- For example, a large walled town of the Liangzhu culture tation techniques were employed to recover plant remains. A was unearthed near Hangzhou City recently (fig. 1), and it total of 222 soil samples were collected and processed, and a measures about 1,900 m in length and about 1,700 m in width, tremendous quantity of plant remains have been found by a total area of 3.23 km2. The foundation of the wall is about flotation as well as by water screening. 50 m in width and was consolidated by a layer of large pebbles The identification and counting of plant remains are still (Liu 2007). It is difficult to imagine that this vast construction in process because of the enormous quantities recovered by could be undertaken without the food support of fully de- flotation. Nevertheless, abundant rice remains, including veloped rice agriculture. charred rice grains and rice spikelet bases, have been found Also, archaeological surveys indicate that the Liangzhu cul- in the treated samples (fig. 2). A detailed study of rice spikelet ture sites sharply increased in number in the lower Yangtze bases shows that the Tianluoshan rice consisted of a high River region, especially in the Hangzhou Bay area, where the proportion of the shattering wild type, which suggests that distribution of Liangzhu sites is even denser than that of the process of rice domestication had not yet culminated in modern villages. This clearly indicates a significant increase the Hemudu period, sometime about 6500 cal BP (Fuller et in human population, which is likely to have been correlated al. 2009). with a rapid development of rice agriculture. In other words, Besides rice, plant remains recovered and identified from as late as about 5,000 years ago, rice agriculture had become the processed samples of the Tianluoshan site include acorns, the dominant food source in the lower Yangtze River region. Trapa sp., Diospyros sp., Gordon euryale, Ziziphus jujuba, Chi- The development of rice agriculture in the middle Yangtze nese gooseberry seeds, bottle-gourd seeds, and seeds of var- River region may have followed a pattern different from that ious weedy grasses. The preliminary results of quantitative in the lower Yangtze River region: full rice agriculture may analysis suggest that rice farming, though important, was only have been established earlier. This hypothesis can be partly part of more general subsistence activities of the Tianluoshan supported by studies of the Chengtoushan site, located in residents. The products of hunting and gathering, especially Lixian County of Hunan Province (fig. 1) in the middle Yang- acorns, also made great contributions to the diet. Therefore, tze River region (Hunan Institute of Archaeology 1999). Water the importance of rice farming should not be overstated for sieving was carried out during the excavation, and rice re- S300 Current Anthropology Volume 52, Supplement 4, October 2011 mains were recovered. Also, a possible rice field was identified middle Yellow River region (fig. 3) and is radiocarbon-dated BP (charcoal), and 100 ע BP (charcoal), 7141 100 ע by survey as well as by the phytolith technique, The site, dated to 6856 BP (charcoal). (These dates were corrected from 105 ע to ca. 6000 cal BP by 14 radiocarbon dates (Hunan Institute 7024 of Archaeology and International Research Center of Japan a 5,730- to a 5,568-year half-life by dividing by 1.03.) The Culture 2007:186), is one of the earliest walled towns in China calibrated age is ca. 7600–8100 cal BP (Institute of Archae- and was possibly constructed and used as a central settlement ology CASS 1991:21). A large number of millet remains, iden- in response to rapid population growth. Therefore, it is likely tified as foxtail millet, were recovered during the excavations that the subsistence of the Chengtoushan society already relied in the 1970s (Tong 1984). However, the taxonomic identifi- on rice agriculture. cation of the remains was problematic because all the so- called millet remains had already decayed to ash when they The Origin of Dry-Land Agriculture in were discovered (CPAM, Hebei Province, and Handan Relics Preservation Station 1981). The identification had to rely on North China the spodogrammic technique, which is the same as the pres- In the past, studies on the origin of dry-land agriculture in ent-day phytolith technique. The work was based on a direct North China have focused on the middle Yellow River region comparison between the morphological characteristics of phy- of North China because this region is believed to be the toliths extracted from modern foxtail millet and the ash re- motherland of ancient Chinese civilizations, where ancient mains recovered from the Cishan site, and a conclusion was millet farming was centered, and abundant prehistoric cul- drawn when similarities were identified (Huang 1982). How- tural remains have been recovered there (Lu 2002). However, ever, it is noteworthy that using phytoliths to identify plant new archaeological data suggest that the whole Yellow River species is not straightforward because of the redundancy of area in North China had the potential to have been the place phytolith production; that is, similar phytolith types may be of origin of dry-land agriculture. produced in widely divergent plant groups (Piperno 1988). Millet remains have been reported in several early Neolithic Recently, a new study was carried out on the Cishan millet sites distributed along the Yellow River. Those from the Cishan by a research team, and the results indicate that the millet site draw the most attention. The Cishan site is located in the remains consist of both foxtail and broomcorn millet, but the

Figure 3. Locations of the sites with early millet remains in North China: 1, Cishan; 2, Xinglonggou; 3, Shawoli; 4, Peiligang; 5, Yuezhuang; 6, Dadiwan; 7, Yuhuazhai. Zhao Origins of Agriculture in China S301 latter is more abundant (Lu et al. 2009a). The identification localities that were fully investigated and excavated from 2001 was done with the same technique (i.e., phytolith analysis), to 2003 (Liu 2004). Locality 1 is a large village settlement site but this time it was based on a systemic comparative study radiocarbon-dated to 8000–7500 cal BP. More than 150 house with modern millets (Lu et al. 2009b), which makes the iden- foundations were recovered from the site. The artifacts in- tification much more precise and believable. clude pottery, stone tools, and bone and jade artifacts. Large- New radiocarbon dates also have been obtained from Ci- scale flotation work with a composite sampling strategy was shan, and some of them are much older than the original carried out during the excavations. A froth-flotation device dates (Lu et al. 2009a). However, these new dates are highly with a 0.2-mm mesh screen was used. More than 1,200 soil suspect because of the ambiguous context of the materials samples were collected and processed. Abundant charred used for the dating. For example, the term “newly excavated plant remains were recovered. The most important result of storage pits” in the report is misleading, because no excavation the flotation work was the recovery of domesticated millets has been conducted at the Cishan site since the 1990s because identified morphologically as broomcorn and foxtail millet. of the stringent policy of the Chinese National Administration The broomcorn millet is more abundant, with about 1,400 of Cultural Relics for its protection. Therefore, the samples charred grains found. Only about 60 grains of foxtail millet likely had been collected from pits naturally exposed by ero- were recovered. Other plant remains recovered by flotation sion on the cliff. This is suggested from a photo shown in include Astragalus, Chenopodium, Vitis, and so forth. Nut the report (Lu et al. 2009a, fig. 1C). The mention of materials fragments of acorns were also identified. of “millet grain crop” in the report is also confusing because Morphologically different from the wild grasses, the millet the Cishan millet remains were already decayed to ashes mixed grains recovered here exhibit the distinctive characteristics of with soil when they were unearthed, according to the 1970s domesticated species, being more spherical in shape and larger and 1980s reports and the materials now exhibited in the in size (fig. 2). As crops, the millet grains are generally not Cishan Museum. Note that it would not be necessary to use as compacted inside as is usually seen in other wild grass the complicated phytolith technique for the identification if seeds, and therefore they are more readily expanded in the actual millet grains had really been found, because it is quite embryo portion when burned, exhibiting a popcornlike ap- straightforward and easy to identify and distinguish foxtail pearance. Such morphological characteristics have been found and broomcorn millet simply on the basis of morphological in the millet remains recovered from the Xinglonggou site. characteristics of the grains (Liu and Kong 2004). Therefore, The millet remains are directly dated by accelerator mass the material used for dating likely consisted of the bulk sam- spectrometry to ca. 7670–7610 cal BP in three independent ples rather than actual millet grains. It is interesting that one radiocarbon laboratories (Z. Zhao, L. Zhou, G. W. Crawford, of the new dates came from sample CS-BWG, which had G. Liu, Y. Li, K. Liu, X. Wu, and T. Nishimoto, unpublished actually been collected during the 1970s excavation and stored manuscript). Therefore, the Xinglonggou millet remains re- in the Cishan Museum. Its date is 7671–7596 cal BP (Lu et covered by flotation, with their precise dating and clear iden- al. 2009a; their fig. 3), which perfectly matches the original tification, are some of the earliest domesticated millets yet dates from the site. found in China. An archaeological chronology relies not only on chrono- Domesticated millets account for nearly 50% of the charred metric dating techniques but also on the classification and seeds recovered at the Xinglonggou site. However, the ubiq- seriation of the cultural assemblages. A number of archaeo- uity of millet remains was as low as only 5% in the soil samples logical sites located in North China have been classified in processed. A number of pig skulls were unearthed, and most the same cultural phase as the Cishan site (i.e., the early of them were identified as wild, while a few exceptions show Neolithic period, 9000–7000 cal BP) on the basis of com- some characteristics of domesticated pig. Combining the re- parison of their cultural assemblages, including settlement sults of studies of the plant remains, animal bones, and stone- patterns, burial traditions, pottery types, stone tools, and so tool use and wear, the subsistence of the Xinglonggou site forth (Zhao 2008). Millet remains were reported from some still seems to have been dominated by hunting and gathering, of these sites, for example, the Shawoli site (Wang 1984) and although millet farming was already practiced and the pig was the Peiligang site (Institute of Archaeology CASS 1984) in the probably in the process of domestication. Therefore, the Xing- middle Yellow River region, the Yuezhuang site (Crawford, longgou site should be regarded as in an early stage of the Chen, and Wang 2006) in the lower Yellow River region, the transition from hunting and gathering to agriculture; that is, Dadiwan site (Gansu Museum 1982) in northwestern China, the Xinglonggou residents still relied on hunting and gath- and the Xinglonggou site (Zhao 2005b) in northeastern China ering for food, and agricultural products produced by millet (fig. 3). Among these, the most important new data for millets farming and animal husbandry only supplemented their diet. come from the systematic flotation work at the Xinglonggou On the basis of flotation results from the Xinglonggou site site. and new data from other sites, we have found several signif- The Xinglonggou site is located in the upper Liao River icant characteristics of the early stage of the transition to millet region in northeastern China, along the eastern edge of the agriculture in North China. First, morphologically, millets at Mongolian Grassland. The site consists of three independent this stage were already fully domesticated species. In other S302 Current Anthropology Volume 52, Supplement 4, October 2011 words, the emergence of millet domestication should be dated sica, Papaver, Vitis, and, of course, crops that included foxtail as earlier than the early Neolithic period. Second, millet grains millet, broomcorn millet, and rice. recovered by flotation from these early Neolithic sites include The plant remains recovered from the Yuhuazhai site are both broomcorn and foxtail millet, but the former is always obviously dominated by the millets (fig. 2), which account more abundant than the latter in this time period. Third, for nearly 90% of the total number of unearthed plant seeds. besides the middle Yellow River region, where ancient Chinese Only four charred rice grains were found. In absolute num- civilization and early dry-land agriculture were centered, new bers, foxtail millet (about 36,000 grains) is much more abun- data on early millet remains are widely found farther north, dant than broomcorn millet (about 14,000 grains), while in along the Great Wall in North China, an economic area that terms of ubiquity, the percentage of broomcorn millet (67%) combines farming and herding today. is higher than that of foxtail millet (63%). The results of The middle Neolithic period (7000–5000 cal BP) is a very flotation at the Yuhuazhai site clearly indicate that millet farm- important stage of cultural development in North China. This ing was dominant in the food production of the Yangshao period is represented by the well-known Yangshao culture people. In other words, the overall subsistence of the Yangshao distributed all over North China. Thousands of archaeological culture, including the Yuhuazhai, Banpo, and other related sites of the Yangshao culture have been found in this broad sites, mainly relied on millet farming and animal husbandry. area, especially along several important branches of the Yellow Hunting and gathering might still play a role in economic River, such as the Wei, Fen, Yi, and Luo rivers. Among these practices but only a secondary one in subsistence. Therefore, sites, the best known is the Banpo site in Xi’an City. the transition from hunting and gathering to agriculture The Banpo site is a large village settlement enclosed by a seems to have been faster in North China than in the Yangtze moat inside of which house foundations, kilns, and ash pits River areas. As early as the Banpo phase of the Yangshao were found arranged in an orderly fashion. Outside the set- culture (i.e., 7000–6000 cal BP), full dry-land agriculture char- tlement, a complete lineage cemetery was discovered. Exca- acterized by millet farming was already established in North vations during the 1950s found large quantities of pottery and China. stone tools, among which stone hoes, spades, knives, and pestles and mortars are most likely to have been used as The Origin of Ancient Tropical agricultural tools in relation to farming practices such as har- Agriculture in South China vesting and grinding (Institute of Archaeology CASS and Banpo Museum 1963). Pigs and dogs are both identified as South China here refers to the region south of the Nanling domesticated. Furthermore, a few charred millet remains were Mountains, particularly in the Zhujiang (Pearl) River areas. found inside a pottery jar. All this evidence suggests that The study of the origin of agriculture in these areas has mainly agriculture was a compulsory element in the subsistence prac- focused on the Zengpiyan site in the last century. tice of the Yangshao culture. Unfortunately, flotation work Zengpiyan is a cave site located in Guilin City, Guangxi was not practiced at the time of excavation, and we therefore Province (fig. 4). Cultural deposits at the Zengpiyan site can have no reliable data for quantitative analysis regarding sub- be divided into five phases, with the first phase dated to sistence in the Yangshao culture. 12,000–11,000 cal BP and the last to 8000–7000 cal BP. Since However, the recent excavation of the Yuhuazhai site, only its excavation in the 1970s, the site has attracted much aca- a couple of kilometers from the Banpo site, provided an op- demic attention throughout the world. Because the pig bones portunity to do more-detailed analysis about subsistence in found there have been identified as possibly from domesti- the Yangshao culture. The Yuhuazhai site is also located in cated pigs, the site was believed by some scholars to be the Xi’an City (fig. 3) and was excavated in 2002. It is almost a earliest archaeological site with rice farming, and therefore it copy of the Banpo site. Both sites are large village settlements has been cited as evidence for a hypothesis stating that South enclosed by moats and dated to 7000–6000 cal BP. The cul- China was the center for the origin of rice agriculture. How- tural remains are identical in these two sites, including features ever, no plant remains were ever reported from the site to such as house foundations, kilns, pits, and artifacts such as support this idea. Therefore, the question of the nature of pottery and stone tools. subsistence at the Zengpiyan site has been vigorously debated. During the 2002 excavation season, we carried out a flo- In order to solve this problem, the Zengpiyan site was tation project at the Yuhuazhai site. A diagnostic sampling reexcavated in 2001 with a multidisciplinary approach, in- strategy was employed, and a froth-flotation device with a cluding archaeobotanic methods such as flotation and phy- 0.2-mm mesh screen was used. A total of 106 soil samples tolith analysis (Institute of Archaeology CASS et al. 2003). were collected and processed. Very abundant plant remains The new excavation was limited to the reserved areas between were recovered, including about 55,000 seeds. On the basis the excavated squares of the 1970s. Therefore, a complete of taxonomic identification, the seeds belong to the families sampling strategy was used for collecting flotation soil sam- of Poaceae, Leguminosae, Cruciferae, Chenopodiaceae, Po- ples: all cultural sediments from the excavation were floated lygonaceae, and Compositae, among others. Also, some eco- to recover plant remains. More than 8,000 L of soil samples nomically important plant remains were found, such as Bras- was processed by a froth-flotation device with a 0.2-mm mesh Zhao Origins of Agriculture in China S303

Figure 4. Locations of the sites related to the ancient tropical agriculture in South China: 1, Zengpiyan; 2, Xiaojing; 3, Dingsishan; 4, Guye. screen. The charred plant remains obtained included wood, another place. This was further corroborated by the excava- seeds, nuts, roots, and tubers. The seeds were identified as tions at the Dingsishan site. belonging to more than 16 plant species, but no rice was The Dingsishan site is located in YongningCounty, Guangxi found. Soil samples were also collected from each layer for Province (fig. 4). The site is a well-preserved shell midden, phytolith analysis, but no phytolith type of the Oryza genus, and the cultural deposits can be divided into four phases. such as the double-peaked glume type or rice leaf types, was Phase I (early Neolithic) was dated to 10,000 cal BP. The observed in the samples. Therefore, both flotation results and major part of the cultural deposits of the site is in phases II phytolith analysis clearly indicate that the Zengpiyan site had and III (8000–7000 cal BP). Phase IV was dated to 6000 cal nothing to do with the origin of rice agriculture. BP. No systematic flotation work was conducted during the Information about local subsistence after 7000 cal BP can excavations in 1997, but soil samples were collected for phy- be obtained from the Xiaojing site. The Xiaojing site is located tolith study at that time. in Ziyuan County, Guangxi Province, about 100 km north of The results of phytolith analysis of the Dingsishan site in- Zengpiyan (fig. 4). Cultural deposits of the Xiaojing site can dicate that no rice phytoliths are present from phase I through be classified into three phases based on cultural remains and phase III but that large numbers of rice phytoliths are found radiocarbon dates. Phase I was dated to 6500–6000 cal BP, phase II to 6000–4000 cal BP, and phase III to 4000–3000 cal in phase IV (Zhao, Lu, and Fu 2005). Regarding the pro- BP (Guangxi Archaeological Team and Ziyuan Administration duction of phytoliths, the Poaceae is the most productive of Culture Relics 2004). Water sieving was applied to recover family in the plant kingdom, while Oryza is one of the most plant remains during the excavation at the Xiaojing site. A productive genera in the grass family. Rice phytoliths are sig- large number of charred rice grains, numbering in the tens nificant not only for their abundance but also because the of thousands, were recovered from phases II and III, but not large size of the phytoliths enables a more definitive mor- a single rice grain was found in phase I. This stark contrast phological identification. From our experience, once rice has suggests that a rapid change to rice farming occurred at the appeared in an archaeological site, it is relatively easy to dis- Xiaojing site around 6,000 years ago. This rapid change can cover rice phytoliths in the cultural deposits. For these rea- be interpreted only as the introduction of rice farming from sons, the surprising abundance of rice phytoliths in phase IV S304 Current Anthropology Volume 52, Supplement 4, October 2011 indicates that rice was possibly introduced into this area about and 6500 cal BP was a transitional period from hunting and 6,000 years ago. gathering to rice agriculture. This transition features a rather According to the above archaeobotanic studies, South slow evolutionary process. The early stage of the transition is China would have had nothing to do with the origin of rice characterized by a mixed subsistence economy in which peo- agriculture. Rice agriculture was introduced into this area, ple still relied on hunting and gathering for their major food probably from the middle Yangtze River region, later than sources but began to practice rice farming and animal hus- 6,000 years ago, and then it rapidly became the major sub- bandry as a supplement. (3) During the time from 6500 to sistence resource in South China. Therefore, the question be- 4500 cal BP, rice agriculture–based subsistence was gradually comes, what kind of subsistence pattern was present in South established; that is, people began to rely on rice agriculture China before 6000 BP? more and more as their major food source. Full rice agri- Abundant charred remains of roots and tubers recovered culture was first established in the middle YangtzeRiver region by flotation from the Zengpiyan site have attracted our at- about 6400–5300 BP, a period known as the Daxi culture, tention. Most plants with edible roots or tubers feature veg- and then in the lower Yangtze River region about 5200–4300 etative propagation, and they should be easier to domesticate BP, during the Liangzhu culture period. than seed-bearing plants. As early as the 1950s, Carl Sauer The history of the origin of dry-land agriculture in North (1952) suggested that vegetative propagation should precede China can also be divided into three periods. (1) Cultivation seed agriculture. He believed that the first domesticated plants of millet and millet domestication might have begun as early in agricultural history were not seed plants such as wheat, as 10,000 years ago, but this initial stage of dry-land agri- rice, or corn but plants that lived in Southeast Asia, had edible culture is still vague. It is urgent to carry out thorough in- tubers and roots, and reproduced asexually. Note that South vestigations and get more solid and reliable archaeobotanic China occupies the same climatic and environmental setting data in future work. (2) From 9000 to 7000 cal BP, there was as, as well as sharing a similar Neolithic cultural tradition a transitional period from hunting and gathering to dry-land with, the northern part of Southeast Asia. Therefore, the flo- agriculture, evidenced by new data from several early Neo- tation results of the Zengpiyan site remind us to reconsider lithic sites. The subsistence pattern in this time period relied Sauer’s model of the origins of agriculture. At least we know on hunting and gathering, with millet farming as a supple- that in the tropical area of Asia, there should be a type of ment. (3) During the time from 7000 to 6000 cal BP, millet subsistence pattern primarily based on vegetative propagation farming–based subsistence (i.e., full dry-land agriculture) was before rice agriculture was introduced. established in North China, represented by the Yangshao cul- ture. However, it seems that the dry-land agriculture in North Conclusions China developed faster than the rice agriculture in the Yangtze River areas. The data obtained by flotation indicate that ag- From studying new archaeobotanic data obtained by flotation riculture accounted for a higher proportion of the subsistence over the past 10 years or so, we now have greater insight into at the Xinglonggou site in the north than at the contempo- the origins of agriculture in China. These can be summarized raneous Jiahu site in the south at ca. 8000 cal BP. Analysis of as follows. the plant remains recovered from the Yuhuazhai site also Three independent agricultures have centers of origin revealed that dry-land agriculture was already established in within China: (1) dry-land agriculture, represented by the North China by 7000–6000 cal BP, earlier than the time when major crops of foxtail and broomcorn millet, with a center rice agriculture was fully adopted in the middle and lower located in a wide area of loess along the Yellow River between Yangtze River areas. the southern border of the Mongolian steppe to the north Compared with the above two subcenters, the origins and and the Huai River to the south; (2) rice agriculture, with a early development of ancient agriculture based on vegetative center located in an area of wetland along the middle and propagation in South China are not yet clear. Further inves- lower Yangtze River between the Huai River to the north and tigations are needed to deepen our understanding of this pro- the Nanling Mountains to the south; and (3) ancient tropical cess. agriculture, characterized by crops that propagate vegetatively, From the discussion during the Wenner-Gren Temozon often roots and tubers and with a center located in the area conference, it is clear that, from a worldwide perspective, there along the Zhujiang River south of the Nanling Mountains. are many similarities between China and the other major The history of the origin of rice agriculture can be divided centers for agricultural origins regarding the questions of into three periods. (1) About 10,000 cal BP, rice cultivation when and how agriculture emerged in these centers. For ex- began in the Yangtze River areas. Although we need more ample, new studies from the Near East suggest that the be- research to clarify whether the rice cultivated at this period ginning of agriculture was indicated by the cultivation of wild belongs to a domesticated species in morphological and ge- plants such as wild barley and wild lentil. The archaeological netic terms, we believe that the purpose of the beginning of evidence for the early cultivation has been dated to the Pre- rice cultivation was only to increase the production of wild Pottery Neolithic A sometime around 11,000 cal BP. The ear- rice, not for plant domestication. (2) The time between 9000 liest domesticated plants, such as einkorn wheat and emmer Zhao Origins of Agriculture in China S305 wheat, were found from sites of the early Pre-Pottery Neolithic Relics. 2004. Excavation of the Neolithic site of Xiajing at Ziyuan B, dated to about 10,000 cal BP (Weiss and Zohary 2011). County, Guangxi. Kaogu [Archaeology] 2004(3):7–30. [In Chinese.] Henan Institute of Archaeology. 1999. Wuyang Jiahu. Beijing: Science. Therefore, the origins of agriculture in both the Near East [In Chinese.] and China seem to follow the same pattern: they both began Ho, Ping-Ti. 1977. The indigenous origin of Chinese agriculture. In with plant cultivation. However, there is a slight difference in Origins of agriculture. C. A. Reed, ed. Pp. 413–484. The Hague: the timing between these two centers: the beginning of cul- Mouton. tivation, as well as the occurrence of domestication, seems to Huang, Qixu. 1982. Spodogram study and its application in archae- ology. Kaogu [Archaeology] 1982(4):418–420. [In Chinese.] be 1,000 years or so later in China than in the Near East. Hunan Institute of Archaeology. 1999. Brief report of 1997–1998 This might be due to inadequate archaeobotanic data in China excavation seasons at Chengtoushang site in Lixian. Wenwu [Cul- at the moment, as flotation has been applied in the Near East tural Relics] 6:4–17. [In Chinese.] for about 50 years but has been practiced in China for only Hunan Institute of Archaeology and International Research Center 10 years. of Japan Culture. 2007. Chengtoushan in Lixian: Sino-Japan co- Studies in other major centers, such as the Near East, also operative research on environmental archaeology in the Liyang Plain. Beijing: Cultural Relics. [In Chinese.] suggest that morphological change in crops may occur late Institute of Archaeology CASS (Chinese Academy of Social Sciences). in the domestication process not only for legumes but also 1984. Excavation at Peiligang site in 1979. Kaogu Xuebao [Acta for cereals (Zeder 2011). This pattern is also supported by Archaeologica Sinica] 1:23–52. [In Chinese.] new data from China, as can be seen in the study of rice ———. 1991. Radiocarbon dates in Chinese archaeology (1965–1991). spikelet bases from the Tianluoshan site (Fuller et al. 2009). Beijing: Cultural Relics. [In Chinese.] Institute of Archaeology CASS, and Banpo Museum. 1963. The Ne- This fact provides a challenge for the identification of do- olithic village at Banpo, Xian. Beijing: Cultural Relics. [In Chinese.] mesticated plants recovered from archaeological sites because Institute of Archaeology CASS, Guangxi Archaeological Team, Zeng- there are no generally recognized criteria to draw a line within piyan Museum, and Archaeological Team of Guilin City. 2003. this long, uninterrupted process in order to distinguish wild Zengpiyan: a prehistoric site in Gulin. Beijing: Cultural Relics. [In and domestic species. Also, it raises some new issues for the Chinese.] Jiang, Leping, and Liu Li. 2006. New evidence for the origins of study of the origins of agriculture. For example, how long sedentism and rice domestication in the lower Yangzi River, China. did it take to complete the process of plant domestication? Antiquity 80(308):355–361. Are there different rates in the domestication processes be- Jones, Martin K., and Xinyi Liu. 2009. Origins of agriculture in East tween different crops or between different areas? What are Asia. Science 324(5928):730–731. the factors that can affect the rate of the process of plant Liu, Bin. 2007. A walled town of 5000 BP found in the Liangzhu domestication? All of these issues need to be considered in site. Wenwubao [News of Chinese Relics], December 5. [In Chinese.] Liu, Changjiang, and Zhaochen Kong. 2004. The morphological com- our future research on the origins of agriculture in China as parison of grains between foxtail and broomcorn millet and its well as in the world. application for identification in archaeology remains. Kaogu [Ar- chaeology] 2004(3):76–83. [In Chinese.] Liu, Guoxiang. 2004. Collective articles of northeast archaeology in China. Pp. 58–86. Beijing: Science. [In Chinese.] References Cited Liu, Jun. 2006. Hemudu culture. Beijing: Cultural Relics. [In Chinese.] Lu, Houyuan, Jianping Zhang, Kam-biu Liu, Naiqin Wu, Yumei Li, An, Zhimin. 1989. Prehistoric agriculture in China. In Foraging and Kunshu Zhou, Maolin Ye, et al. 2009a. Earliest domestication of farming. D. R. Harris and G. C. Hillman, eds. Pp. 643–649. Lon- don: Unwin Hyman. common millet (Panicum miliaceum) in East Asia extended to Cohen, David Joel. 2011. The beginnings of agriculture in China: a 10,000 years ago. Proceedings of the National Academy of Sciences multiregional view. Current Anthropology 52(suppl. 4):S273–S293. of the USA 106(18):7367–7372. CPAM (Committee for Preservation of Artifact Management), Hebei Lu, Houyuan, Jianping Zhang, Naiqin Wu, Kam-biu Liu, Deke Xu, Province, and Handan Relics Preservation Station. 1981. The Ci- and Quan Li. 2009b. Phytoliths analysis for the discrimination of shan site in Wu’an, Hebei Province. Kaogu Xuebao [Acta Archaeo- foxtail millet (Setaria italica) and common millet (Panicum mil- lica Sinica] 3:303–338. [In Chinese.] iaceum). PLoS ONE 4(2):e4448, doi:10.1371/journal.pone.0004448. Crawford, Gary W. 1992. Prehistoric plant domestication in East Asia. Lu, Tracey Lie Dan. 2002. A green foxtail (Setaria viridis) cultivation In The origins of agriculture: an international perspective. C. Wesley experiment in the middle Yellow River valley and some related Cowan and Patty Jo Watson, eds. Pp. 7–38. Washington, DC: issues. Asian Perspectives 41(1):1–14. Smithsonian Institution. Luo, Yunbing, and Juzhong Zhang. 2008. Restudy of the pig bones Crawford, Gary W., Xuexiang Chen, and J. H. Wang. 2006. Houli from the Jiahu site in Wuyang County, Henan. Kaogu [Archaeology] culture rice from the Yuezhuang site, Jinan. Orient Archaeology 3: 2008(1):90–96. [In Chinese.] 247–251. [In Chinese.] Piperno, Dolores R. 1988. Phytolith analysis: an archaeological and Fuller, Dorian Q, Ling Qin, Yunfei Zheng, Zhijun Zhao, Xugao Chen, geological perspective. San Diego, CA: Academic Press. Leo Aoi Hosoya, and Guo-Ping Sun. 2009. The domestication Sauer, Carl O. 1952. Agricultural origins and dispersal. Cambridge, process and domestication rate in rice: spikelet bases from the MA: MIT Press. lower Yangtze. Science 323(5921):1607–1610. Tong, Weihua. 1984. Original agricultural relics of Cishan sites and Gansu Museum. 1982. Excavation report of the first phase of Dadi- related problems. Agricultural Archaeology 1:194–207. [In Chinese.] wan culture at the Dadiwan site, Qinan. Kaogu Yu Wenwu [Ar- Wang, Jihua. 1984. Charred foxtail millets found in the Shawoli site, chaeology and Cultural Relics] 2:1–4. [In Chinese.] Xinzheng. Agricultural Archaeology 2:194–207. [In Chinese.] Guangxi Archaeological Team and Ziyuan Administration of Cultural Weiss, Ehud, and Daniel Zohary. 2011. The Neolithic Southwest S306 Current Anthropology Volume 52, Supplement 4, October 2011

Asian founder crops: their biology and archaeobotany. Current ———. 2005b. Discussion of the Xinglonggou site flotation results Anthropology 52(suppl. 4):S237–S254. and the origin of dry farming in northern China. Antiquities of Yan, Wenming. 1992. Origins of agriculture and animal husbandry Eastern Asia 2005A:188–199. [In Chinese.] in China. In Pacific Northeast Asia in prehistory: hunter-fisher-gath- ———. 2005c. New archaeobotanic data for the study on the origins erers, farmers, and sociopolitical elites. C. Melvin Aikens and Song of agriculture and civilization in China. Management and Review N. Rhee, eds. Pp. 113–123. Pullman: Washington State University of Social Sciences 2:82–91. [In Chinese.] Press. ———. 2006. New study on the ancient agriculture in South China. Yuan, Jing. 2001. Issues on the origins of domesticated animals in In Prehistoric archaeology of South China and Southeast Asia. In- Chinese Neolithic times. Wenwu [Cultural Relics] 5:51–58. [In Chi- stitute of Archaeology CASS, ed. Pp. 145–156. Beijing: Cultural nese.] Relics. [In Chinese.] Zeder, Melinda A. 2011. The origins of agriculture in the Near East. Zhao, Zhijun, Tracey Lie Dan Lu, and Xianguo Fu. 2005. Phytolith Current Anthropology 52(suppl. 4):S221–S235. study for the Dingsishan site in Yongning County, Guangxi. Kaogu Zhang, Juzhong, and Xiangkun Wang. 1998. Notes on the recent [Archaeology] 2005(11):76–84. [In Chinese.] discovery of ancient cultivated rice at Jiahu, Henan Province: a Zhao, Zhijun, and Juzhong Zhang. 2009. The report of flotation work new theory concerning the origin of Oryza japonica in China. at the Jiahu site. Kaogu [Archaeology] 2009(8):84–93. [In Chinese.] Antiquity 72(278):897–901. Zhejiang Institute of Archaeology. 2003. Hemudu. Beijing: Cultural Zhao, Hui. 2008. Study on the cultural sequence of the Neolithic Relics. [In Chinese.] period in China. In Review of Chinese archaeology in the last century. Zhejiang Institute of Archaeology and Xiaoshan Museum. 2004. Kua- Yan Wenming, ed. Pp. 19–34. Beijing: Science. [In Chinese.] huqiao. Beijing: Cultural Relics. [In Chinese.] Zhao, Zhijun. 2005a. Archaeobotany and its recent advances in Zhou, Xinhua. 2002. Rice farming tribe. Hangzhou: Zhejiang Wenyi. China. Kaogu [Archaeology] 2005(7):42–49. [In Chinese.] [In Chinese.] Current Anthropology Volume 52, Supplement 4, October 2011 S307

The Transition from Foraging to Farming in Prehistoric Korea

by Gyoung-Ah Lee

As a secondary setting of agricultural origins, prehistoric Korea may offer insights into the social interactions involved in crop acquisition and the human modification of local landscapes to accom- modate a new agrarian way of life. Recent data may point to Korea as one of several areas where the local domestication of two crops, azuki and soybean, may have taken place. This paper explores various economic adaptations and transitions to resource production in several ecological and social settings, including the central west and southeast coasts, islands on the west coast, and the floodplains in central and southeastern Korea during the Chulmun period (7500–3400 BP). The paper reviews two popular explanations of the transition to food production, environmental impulse and migration, in the context of Korean archaeology and beyond.

Introduction in northeastern China and the maritime region of the Russian Far East. Two types of millet—foxtail (Setaria italica ssp. ital- ica) and broomcorn (Panicum miliaceum)—were the first Prehistoric Korea1 is situated in a context of agricultural tran- crops to appear in Korea. Shortly after their initial domes- sition in which introduced crops became important in sub- tication in the Huanghe Basin (Crawford 2009; Lu et al. 2009), sistence and indigenous plants may have undergone domes- millets seem to have spread to the northeast, as suggested by tication. The origins of agriculture in primary regions have the findings at the Xinglonggou (8000–7500 BP) and Xinle received more scholarly attention than that in secondary sites in Liaoning Province (Cohen 2011; Zhao 2011). Millets regions, where domesticated species spread later, but it was had dispersed farther east to the Zaisanovka Neolithic culture the “spread of farming” that had the major impact on human in the Primorie Province of maritime Russia by 6000 BP society (Price 2000b:316). For example, case studies in Europe (Kang 2009). Some materials at Zaisanovka also indicate in- have produced in-depth understanding of how the spread of teractions with the Xinle culture and the Chulmun2 culture agriculture shaped tremendous changes in social organiza- in northeastern Korea. Crops were probably one of the shared tions, technologies, and ideologies (e.g., Ammerman and Biagi items in this interaction sphere. 2003; Hodder 1990; Price 2000a; Rowley-Conwy 2004). This Considering its proximity to early farming communities in supplement to Current Anthropology is also dedicated to il- northeastern China and Primorie, northeastern Korea may luminating the agricultural dispersal worldwide, including the have begun millet farming earlier than current data from Pacific region (Bellwood 2011), Japan (Crawford 2011), At- southern Korea suggest.3 In southern Korea, the earliest ap- lantic Europe (Rowley-Conwy 2011), and the Aegean and pearance of domesticated foxtail and broomcorn millet is Balkans (O¨ zdog˘an 2011). These papers not only cover the dated to the Middle Chulmun period, around 5500 cal BP new regional data but also provide insights into the issues of (table 1). population changes, social interactions, and ecological views Not long after the initial adoption of millet cultivation, the on the transition. As a part of these efforts, this paper attempts to explore the mechanisms of agricultural spreads and the 1. “Korea” is used to indicate the geography (the Korean Peninsula) social consequences with a regional example. rather than the term for political and ethnic distinction in this paper. The initial appearance of domesticated crops in Korea can 2. “Chulmun” and “Neolithic” are the terms that refer to the same be understood in terms of interactions with Neolithic entities time span, from the onset of pottery production and before the beginning of metallurgy. According to the new romanization of Korean issued in 2000, “Chulmun” now reads “Jeulmun.” This paper follows the new system except for commonly cited terms in previous publications (e.g., Gyoung-Ah Lee is Assistant Professor in the Department of Chulmun, Tongsamdong). Anthropology, University of Oregon (308 Condon Hall, 1321 Kincaid 3. The separation of North and South Korea has hampered our un- Street, Eugene, Oregon 97403-1218, U.S.A. [[email protected]]). derstanding of early agriculture in Korea. Neither systematic archaeo- This paper was submitted 13 XI 09, accepted 9 XII 10, and botanical study nor direct dates on domesticated crops are available in electronically published 28 VI 11. North Korea.

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0011$10.00. DOI: 10.1086/658488 Table 1. Accelerator mass spectrometry (AMS) dates

Age (BP) Period, province, site Material Lab ID Provenience Uncalibrated Calibrated Early Chulmun: Seoul: 100 ע 5760 70 ע Amsadonga Charred wood AERIK-189 House 75-2 5000 260 ע 5390 200 ע Charred wood AERIK House 75-4 4730 220 ע 5690 200 ע Charred wood AERIK House 75-5 4950 260 ע 5270 200 ע Charred wood AERIK House 75-5 4610 120 ע 6300 110 ע Charred wood KAERI-188 House 75-10 5510 150 ע 6940 105 ע Charred wood N-2336 House 6050 130 ע 7130 110 ע Charred wood N-2337 House 6230 Gyeonggi: 160 ע 5870 140 ע Misaria Charred wood KSU-497 Layer VI 4950 90 ע 7120 60 ע Ilsana Wood Beta46227 Layer I-Ga, 6210 420–430 cm a.s.l. 70 ע 6430 60 ע Peat Beta48387 Peat layer 1-Da 5650 90 ע 6080 60 ע Peat Beta48386 Peat layer 1-Da, 5290 420–425 cm a.s.l. 90 ע 5910 60 ע Wood Beta46226 Daehwari layer I 5170 Gangwon: 250 ע 6950 210 ע Osannia Charred wood KSU-619 House A- V-2 5900 230 ע 6580 210 ע Charred wood KSU-620 Layer V3 5570 90 ע 7040 50 ע Charred wood KSU-616 Layer V7 5950 230 ע 6580 210 ע Charred wood KSU-615 Layer V3 5740 South Gyeongsang: 50 ע 7580 50 ע Bibongrib Wooden plank Beta219086 Layer 45 6710 40 ע 7640 50 ע Wooden plank SNU06306 Layer 45 6800 50 ע 7540 60 ע Wooden plank SNU06208 Layer 45 6670 50 ע 7400 50 ע Juglans shell Beta219089 Layer 41, shell layer 5 6490 40 ע 7470 50 ע Juglans shell SNU06204 Layer 41, shell layer 5 6550 60 ע 7330 60 ע Charred wood SNU06210 Layer 39, shell layer 4 6390 60 ע 6810 40 ע Charred wood Beta219088 Layer 34, shell layer 3 5970 80 ע 7170 60 ע Charred wood SNU06203 Layer 34, shell layer 3 6270 80 ע 6810 60 ע Juglans shell SNU06209 Layer 31, shell layer 2 5970 170 ע 6190 150 ע Charred wood SNU06A002 Layer 26, pit 17 5420 70 ע 6110 40 ע Charred wood SNU05343 Layer 25-2, shell layer 1 5330 50 ע 5650 50 ע Juglans shell SNU06206 Layers 19–21, pit 9 4900 100 ע 5190 40 ע Charred wood SNU05345 Layers 19–21, pit 11 4530 80 ע 5430 50 ע Charred wood SNU05346 Layers 19–21, pit 12 4680 100 ע 5160 50 ע Quercus shell Beta219090 Layer 19, pit 1 4500 50 ע 4920 40 ע Quercus shell SNU05344 Layer 19, pit 1 4340 70 ע 5400 60 ע Charred wood SNU06201 Layer 19, pit 1 4650 140 ע 5070 50 ע Wooden plank SNU06205 Layer 19, pit 2 4420 110 ע 7390 120 ע Sejukric Incrustation SNU00385 Pit B2, layer III-3a 6480 270 ע 7110 250 ע Incrustation SNU00386 Pit B4, layer III-2b 6260 50 ע 7200 40 ע Incrustation SNU00387 Pits A8–A9, layer III-6 6260 50 ע 7260 40 ע Incrustation SNU00388 Pit A 6330 140 ע 6770 110 ע Incrustation SNU00390 Pit C 5930 40 ע 7220 40 ע Incrustation SNU00393 Pit B2, layer III-1 6280 50 ע 7200 40 ע Incrustation SNU00394 Pit B2, layer III-1 6260 120 ע 7010 80 ע Incrustation SNU00395 Pits A2–A3, layer III-2b 6110 100 ע 7330 110 ע Incrustation SNU00397 Pit B2, layer III-3c 6420 80 ע 6510 60 ע Incrustation SNU00398 Pit B2, layer III-3c 5700 180 ע 7300 180 ע Sanggnodaedoa Charred wood Layer V 6430 150 ע 5700 140 ע Tongsamdonga Charred wood GX-0378 Level E (Mokdo phase) 5890 140 ע 5720 130 ע Charred wood GX-0379 Level C (Busan phase) 4950 200 ע 5620 160 ע Charred wood N-1213 Level C 4880 60 ע 7340 50 ע Tongsamdongd Bone SNU00092 D-X-IX 6400 80 ע 6600 70 ע Bone SNU00090 F-V-3-4-2 5800 60 ע 6380 70 ע Bone SNU00091 F-V-4-1 5580 60 ע 5400 50 ע Bone SNU00089 F-V-2-3-3 4650 210 ע 6890 160 ע Yeondaedoa Charred wood Layer III 6010 200 ע 6970 160 ע Charred wood Layer III 6090

S308 Table 1 (Continued) Age (BP) Period, province, site Material Lab ID Provenience Uncalibrated Calibrated South Jeolla: 50 ע 6260 60 ע Gadoe Charred wood Beta92374 SW trench 2 clay layer 5460 100 ע 4700 60 ע Charred wood Beta92373 S5W5 clay layer 4160 190 ע 6210 170 ע Dolsan Songdoa Charred wood NUTA1334 House floor layer III-C 5280 190 ע 6200 170 ע Charred wood NUTA1335 Layer IV, pit S5-E1 5430 50 ע 5540 40 ע Seonyudoa Charred wood AERIK 4810 Middle Chulmun: Seoul: 70 ע 4950 50 ע Amsandonga Charred wood House excavated in 1988 4360 Gyeonggi: 50 ע 3780 40 ע Nambukdongf Charred wood SNU455 Outdoor hearth 3 3500 90 ע 4270 50 ע Charred wood SNU456 Outdoor hearth 9 3840 130 ע 5100 50 ע Charred wood SNU457 Outdoor hearth 32-1 4440 130 ע 5110 50 ע Charred wood SNU458 Outdoor hearth 32 4450 60 ע 5400 40 ע Charred wood SNU459 Outdoor hearth 50 4650 100 ע 5450 60 ע Charred wood SNU460 Hearth next to 50 4700 110 ע 5340 60 ע Charred wood SNU461 Hearth 52 4620 100 ע 5450 40 ע Sammokdof Charred wood SNU03255 Floor fill, house 1 4700 50 ע 5530 40 ע Charred wood SNU03256 Floor fill, house 1 4780 100 ע 5500 80 ע Charred wood SNU03258 Floor fill, house 2 4800 100 ע 5430 80 ע Charred wood SNU03257 Floor fill, house 2 4670 70 ע 5370 40 ע Charred wood SNU03259 Floor fill, house 3 4610 60 ע 4930 50 ע Charred wood SNU03260 Floor fill, house 4 4340 110 ע 4910 80 ע Charred wood SNU03261 Floor fill, house 4 4310 100 ע 5190 50 ע Charred wood SNU03262 Floor fill, house 5 4540 100 ע 5170 50 ע Charred wood SNU03263 Floor fill, house 5 4510 120 ע 5130 60 ע Charred wood SNU03264 Floor fill, house 8 4480 110 ע 5140 50 ע Charred wood SNU05200 Floor fill, house 8 4480 120 ע 5130 40 ע Charred wood SNU05202 Floor fill, house 9 4460 60 ע 5380 40 ע Charred wood SNU05201 Floor fill, house 9 4620 100 ע 5460 60 ע Charred wood SNU05203 Floor fill, house 9 4740 90 ע 5480 60 ע Charred wood SNU05204 Floor fill, house 11 4770 100 ע 5470 40 ע Neunggok Foxtail millet Beta252973 Floor fill, house 41 4740 North Chungcheong: 140 ע 4620 80 ע Ilsana Peat Beta48384 Seongjeori (zone 1) 4070 100 ע 4950 80 ע Charred wood Beta45536 Gawaji (zone 2) 4330 150 ע 5060 70 ע Peat Beta46237 Juyeopni 4410 80 ע 4910 70 ע Peat Beta52101 Juyeopni 4310 110 ע 4740 80 ע Peat Beta46236 Juyeopni 4220 40 ע 4480 25 ע Charred wood Juyeopni 4010 60 ע 4340 30 ע Gahyeonnia Peat KSU 3890 140 ע 5050 60 ע Daechonrig Charred wood SNU263 Hearth in house 2 4400 160 ע 4780 110 ע Charred wood SNU369 Pit in pit 2 4240 160 ע 5270 70 ע Charred wood SNU367 Floor fill, house 2 4590 90 ע 5160 40 ע Charred wood SNU368 Pit in house 2 4490 South Jeolla: 100 ע 4700 60 ע Gadoe Charred wood Beta92371 S4-W4 shell layer 6 4060 130 ע 4600 50 ע Charred wood Beta92372 SW tr. 2 shell layer 6 4830 South Gyeongsang: 170 ע 5260 100 ע Tongsamdongh Foxtail millet TO8783 Floor fill, house 1 4590 170 ע 4540 100 ע Tongsamdonga Charred wood AERIK-23 Layer III 4020 160 ע 4460 100 ע Charred wood AERIK-24 Layer III 3980 150 ע 4370 100 ע Charred wood AERIK-25 Layer III 3930 140 ע 4300 100 ע Charred wood AERIK-26 Layer III 3880 160 ע 5060 90 ע Charred wood AERIK-27 Layer III 4400 130 ע 4690 100 ע Charred wood AERIK-22 Layer II 4170 190 ע 4620 150 ע Charred wood Layer Mokdo 4110 170 ע 5270 100 ע Tongsamdongd Bone SNU00094 H-X-VIII 4600 60 ע 4940 40 ע Bone SNU00087 I-V-3-5-2 4360 50 ע 4890 40 ע Bone SNU00088 I-V-3-10-1 4300 80 ע 4740 40 ע Bone SNU00093 H-X-VIII 4200

S309 Table 1 (Continued) Age (BP) Period, province, site Material Lab ID Provenience Uncalibrated Calibrated 40 ע 4910 25 ע Pyeonggeodongi Azuki bean KCCAMS60748 Grid 20, pit 3C 4350 70 ע 4730 25 ע Azuki bean KCCAMS60749 Grid 21, pit 51 4175 80 ע 4740 40 ע Soybean SNU252971 Gird, 20 pit 3-A 4200 50 ע 4920 40 ע Broomcorn millet SNU252972 Gird 21, pit 50 4340 80 ע 4710 30 ע Sangchon Bh Acorn shell SNU01377 Elongate pit 6-1 4150 70 ע 3730 40 ע Bibongrib Charred wood Beta219087 Layer 17, hearth 2 3450 90 ע 3850 60 ע Charred wood SNU06202 Layer 17, hearth 2 3560 60 ע 3920 50 ע Charred wood SNU05347 Layer 17, hearth 2 3600 80 ע 3830 60 ע Charred wood SNU06207 Layer 17, hearth 4 3540 80 ע 2930 60 ע Charred wood SNU05348 Layer 14, burnt area 1 2810 Late Chulmun: Seoul: 320 ע 3740 250 ע Amsadonga Charred wood QL House dug in 1967 3430 Gyeonggi: 60 ע 3320 60 ע Sidoa Charred wood AERIK-13 Cairn 3100 90 ע 3240 60 ע Charred wood AERIK-11 Area III 3040 90 ע 3240 60 ע Charred wood AERIK-14 Cairn 3040 80 ע 3000 60 ע Charred wood AERIK-12 Cairn 2870 South Jeolla: 130 ע 4600 60 ע Gadoe Charred wood Beta92371 S4W4, shell layer 6 4060 60 ע 5560 50 ע Charred wood Beta92372 SW trench, shell layer 6 4830 South Gyeongsang: 60 ע 4510 40 ע Sangchon Bh Juglans shell SNU01376 Floor fill, house 22 4030 100 ע 4690 60 ע Acorn shell SNU01303 Pit 5-1 4150 200 ע 4560 140 ע Foxtail millet TO8608 Outdoor hearth 1 4060 170 ע 4560 100 ע Oun 1h Foxtail millet TO8607 Outdoor hearth 6 4030 260 ע 3690 220 ע Tongsamdonga Charred wood GX-0492 Layer B, Yongdo phase 3400 150 ע 3670 120 ע Charred wood GX-0493 Layer C, Dudo phase 3400 90 ע 3710 60 ע Sangnodaedoa Charred wood Layer 2 3430 160 ע 4600 150 ע Bonggaeria Walnut NUTA1034 III phase, house 9 4060 Early Mumun: South Gyeongsang: 370 ע 3970 280 ע Oun 1h Rice TO8605 Floor fill, house104 3610 90 ע 2980 60 ע Rice SNU125 Floor fill, house104 2850 80 ע 2960 60 ע Foxtail millet SNU126 Floor fill, house104 2830 130 ע 2950 100 ע Foxtail millet TO8637 Floor fill, house104 2800 Middle Mumon: South Gyeongsang: 70 ע 2900 60 ע Okbangh Soybean TO8611 Pit in house 658 2790 120 ע 2580 70 ע Daundongh Adzuki bean TO8965 Floor fill, house 7 2510 80 ע 2590 25 ע Soybean KCCAMS60750 Floor fill, house 7 2485 180 ע 4660 135 ע Daldongj Shell TH1003 Floodplain 4630 160 ע 2500 125 ע Shell TH1004 Floodplain 2910 160 ע 2700 110 ע Shell TH1005 Floodplain 3090 Sources. References for individual sites are in footnotes. Dates for sites not marked by source footnotes are for samples analyzed by the author. This paper uses conventions for laboratory code abbreviations provided in the journal Radiocarbon. The Carbon Dating Laboratory of the Tohoku University (TH) is not listed in Radiocarbon. Note. All dates except for those at Daldong are calibrated with the calibration curve CalPal 2007_HULU. Dates at Daldong are calibrated .(years (Stuiver and Braziunas 1993:156 45 ע with a DR correction from Enewetak Atoll in the northwestern part of Pacific:DR p 140 Lab IDs in boldface indicate AMS dates. a.s.l. p above sea level. a Kang, Choo, and Na (1993). b Gimhae National Museum (2008). c Dongguk University Museum (2006). d Yoon, Im, and Oh (2004). e Park et al. (2001). f Seoul National University Museum (2006). g Hannam University Museum (2003). h G.-A. Lee (2003). i G.-A. Lee et al. (2011). j Hwang and Yoon (2002). Lee Transition to Farming in Korea S311

Chulmun culture used legumes of genera Vigna and Glycine, poses. As demonstrated in many of the papers in this volume which may have paved the path to the domestication of azuki (e.g., Crawford 2011; Fuller 2011; Piperno 2011; Smith 2011; (Vigna angularis ssp. angularis) and soybean (Glycine max ssp. Weiss and Zohary 2011), detailed archaeobotanical study can max)4 at least around 5000 cal BP (table 1). Recently, possible contribute to distinguishing different types and intensities of independent domestication of soybean has also been docu- plant use. mented in the Jomon context in Japan, dating to 5300–4400 Second, I address the problems of the current discourse on cal BP (Crawford 2011; Obata 2008). These studies emphasize how and why the foraging system shifted to agriculture. The the importance of research on the “evolutionary history of popular models—emphasizing either resource-population individual crops and regional crop associations” (Harris 1990: imbalance due to decreasing marine productivity (J.-j. Lee 15) beyond the well-known primary origins of agriculture. 2001, 2006; Norton 2000, 2007) or social conflicts (J. Kim Definition of the term “agriculture” often hinders under- 2003, 2006)—are the two sides of one coin in terms of a standing of the Chulmun economy. As in the Jomon case shared view of agriculture as a stress response to human pop- discussed by Crawford (2008), a simplistic dichotomy of ulation increases by either internal growth or external influx. hunter-gathers versus rice farmers does not illustrate the For example, it is often suggested that a population imbalance Chulmun way of life. This tendency leads to the presumption resulting from decreasing marine productivity pushed the that agriculture became a primary economic activity only after emergence and intensification of agriculture. Agriculture nat- the Middle Mumun period (2800–2400 BP),5 when irrigated urally became a reluctant final resolution for survival when rice farming was firmly established (e.g., B. Kim 2006; J. Kim certain conditions were met, including a sea level decline, 2008; Norton 2007). A simple dichotomy has further limited marine-resource reduction, and population increase beyond our understanding of the factors involved in the transition the carrying capacity of the accessible lands. What these mod- from the Chulmun to the Mumun, often described as an els do is simply insert all the necessary conditions into a magic abrupt shift from hunting and gathering to agricultural so- formula that itself takes care of all the explanation or necessity cieties. for explanations. One of the achievements of the Temozo´n symposium (Price A few problematic assumptions are inherent in these ap- and Bar-Yosef 2011) was to shape a comprehensive perspective proaches. It is assumed that a new intensive farming technique on agriculture (or food/resource production in a broader would have been adopted as a response to immediate crises. sense) as a range of dynamic human involvements with target Because other options would have been available to humans, species. Documenting agriculture becomes the task of ex- population growth and climatic change cannot be justified as ploring the biology of the target species, the surrounding immediate direct causes of a shift to agriculture without ex- environments, the societies involved, and the technologies plaining why other options were not explored (Price and available. Domestication of species is no longer the sole cri- Gebauer 1995:4). Without identifying specifics in food terion for determining the presence of agriculture. Well before stresses, external-stress models tend to generalize the stress as the fully domesticated species came to be, many prehistoric a fatal one. The food-provisioning problem can be neither societies tempered their landscapes, creating engineered defined objectively nor perceived homogenously across cul- niches, as seen in the Jomon (Crawford 2011) and in eastern ture; it depends on the types and dimensions of stresses, so- North America (Smith 2011). Taking this holistic view on ciopolitical organizations, existing subsistence strategies, and agriculture, this paper adds a refreshing perspective on the population densities (Minnis 1985:5–6). Chulmun and Mumun subsistence by focusing on two as- pects. Previous Research on the First, the paper reviews plant manipulation in the Chulmun Chulmun Subsistence culture (7500–3400 BP) in central and southern regions. The archaeological sites examined are located in different land- The Chulmun period is often subdivided into the three scapes, including the shoreline (Sejukri, Tongsamdong), is- phases: Early (7500–5500 BP), Middle (5500–4000 BP), and lands (Nambukdong, Sammokdo), alluvial fans and plains Late (4000–3400 BP; fig. 2). The Middle Chulmun is distin- (Bibongri, Pyeonggedong, Oun 1, Sangchon B), and uplands guished from the early phase as the material culture in the (Anganggol, Neunggok; fig. 1). Reference is also made to the southeastern region became more incorporated in the central Early Mumun (3400–2800 BP) period for comparative pur- western regions. The majority of the Chulmun sites are shell middens along the coast or on islands (Nelson 1993). Such 4. Throughout this paper, “soybean” refers to G. max without spec- a preference for the coasts and islands leads to the impression ification of a subspecies because of the ambiguity in the archaeological that the Chulmun culture was exclusively marine oriented. record. I do not wish to bias the terminology to favor either wild (G. From the Early Chulmun, however, year-round village sites max ssp. soja) or domesticated soybean. were present along the main tributaries (e.g., Amsadong and 5. “Mumun” (3400–2000 BP) refers to the period after the Chulmun and is marked by the beginning of Mumun-style pottery and metallurgy. Missari along the Han River; Nelson 1993). Since the 1990s, The Late Mumun phase (2400–2000 BP) is often called the Early Iron more Chulmun settlements have been found on floodplains Age. and hillslopes, which the Mumun people occupied later. Some Figure 1. Map showing the Chulmun sites mentioned in the text. 1, Amsadong, Misari; 2, Neunggok; 3, Nambukdong, Sammokdo; 4, An- ganggol; 5, Daejukri; 6, Konamri; 7, Nukdo; 8, Pyeonggeodong, Nam River (Oun 1, Sangchon B); 9, Yondaedo; 10, Bibongri; 11, Tongsamdong; 12, Sejukri, Daldong; 13, Imdang. Lee Transition to Farming in Korea S313 of the sites in this study—Pyeonggeodong, Nam River, domesticated cattle, came from several Late Mumun sites dat- Neunggok, and Anganggol—are examples of these settle- ing as early as 2300 BP (J.-j. Lee 2009). Cattle foot prints on ments. the rice paddy fields at the Pyeonggeodong site indicate that The Chulmun subsistence in Korea is commonly regarded cattle were well integrated into rice farming as draft animals as a “broad-spectrum” system that combined fishing, hunting, by the Three Kingdom period (1700–1400 BP; Gyeongnam and gathering (Ahn 1997; Barnes 1999; Choe and Bale 2002; Development Institute 2007). The onset of the domesticated Crawford and Lee 2003; Nelson 1992; Pearson 1977). A pop- pig in Korea is ambiguous. Some pig remains at the Chulmun ular view in Korean archaeology is that millet farming spread sites resemble domesticated pigs, but clear evidence comes diachronically from northeastern China to northwestern Ko- from the much later Unified Shilla period (1400–1100 BP; J.- rea in the Middle Chulmun and then to the south by the Late j. Lee 2009). Overall, domesticated animals may have played Chulmun. Crop cultivation, however, remained an insignif- a small role in subsistence throughout the prehistoric period icant strategy, and a foraging economy remained dominant in Korea. throughout the Chulmun (J. Kim 2003, 2006). It is difficult to assess how much domesticated resources Before systematic archaeobotanical research began in the contributed to human diet, because dietary reconstruction late 1990s, at least 16 Chulmun sites yielded plant remains based on human remains has been rare in Korea. Recent (G.-A. Lee 2003). None of these sites has been subjected to attempts indicate dietary patterns mainly based on plant re- systematic recovery, and taxonomical identification of plant sources from the Mumun (table 3). The importance of C4 remains from these sites is often questionable.6 These data plants in the Early-Middle Mumun at Konamri is contrasted suggest that acorn (Quercus sp.) was the most important plant to the dominance of C3 plants in the Late Mumun and his- resource throughout the Chulmun period. torical periods, as seen in the Konamri, Neukdo, and Imdang Although rice cultivation is conventionally assumed to have sites (An 2006b; Choy and Richards 2009; Shin and Lee 2009). begun during the Early Mumun, some researchers have pro- The results of these cases, although limited, are correlated posed the existence of domesticated rice in the Chulmun or with an increase of large cereals (rice, wheat, barley) from the even much earlier in the Late Pleistocene on the basis of on Late Mumun onward (G.-A. Lee 2003). rice impressions (Son 1987), phytoliths (Kwak, Fujihara, and On the basis of one human remain, Choy and Richards Yanaki 1995), and uncharred grains (Im 1997; Lee and Park (2010) suggest that the Chulmun people at Tongsamdong 1997; Lee and Woo 2002). All the proposed data, however, were largely dependent on marine protein resources.7 Other are inconclusive because the context of these finds is circum- analyses, however, reflect a more diverse, complex picture. stantial, and identification of rice phytoliths in this case is Recent trace-element analysis at the Yondaedo shell midden controversial (Ahn 2010; Crawford and Lee 2003). suggests a plant-based diet with a sizable supplementary por- Recently, foxtail millet, barley, rice, wheat, hemp, and le- tion of shellfish and fish during the Early Chulmun (An 2009: gumes were reported from a pit house at the Middle Chulmun 20). Two Chulmun shell middens in similar environments Daechonri site in North Chungcheong Province (Hannam and cultural settings on the southwestern coast show differ-

University Museum 2003). Because of an incorrect identifi- ences in diet: a mixed consumption of C3 plants and terrestrial cation and a lack of direct accelerator mass spectrometry and marine proteins at Konamri and a plant-dominant diet (AMS) dates on these crop remains, the question of the pres- at Daejukri (An 2006a, 2006b; table 3). ence of multiple crops at Daechonri remains unanswered (G.- Again, limited numbers of cases cannot represent the di- A. Lee 2003). etary pattern of the entire Chulmun population. An intriguing Domestication of animals is one of the least-studied areas theme in these studies is that Chulmun diets cannot be sum- in Korean archaeology. Most reports on domesticated animals marized as one homogenous pattern. Tongsamdong and Yon- are controversial because the criteria for domestication and daedo (also Konamri and Daejukri) accentuate a possible dif- the context of discovery are unclear (J.-j. Lee 2009). The ferentiation of resource uses even in close proximity in similar earliest domesticated animals found in the archaeological con- environmental and cultural settings. text are dogs, dating as early as the Middle Chulmun, from the Bibongri, Tongsamdong, and Konamri sites (J.-j. Lee 2009; Case Studies table 2). Recently, cattle were reported from the Chulmun context at Bibongri (Kaneko 2008:317), but it is not clear A decade of effort in recovering plant remains through flo- whether the specimen represents a domesticated variety. The tation has increased the database and yielded promising re- earliest confirmed evidence of domesticated horse, as well as sults that help us to understand the transition to agriculture

6. In most cases, identification was made without microscopic aids, 7. Although Choy and Richards (2010:9) realized the limitation caused and the resolution of seed images in the reports is not high enough to by the small sample size in their analysis, their conclusion, interpreting identify their taxa. Finds at a few sites in North Korea have been re- a single sample (one human metacarpal of unknown gender and age) as currently cited as evidence of millet farming in the Middle Chulmun, representative of the Tongsamdong people in the Middle Chulmun pe- although the initial site report cast doubt on the identification of plant riod, is a far reach by any standard. The analyzed sample size of human remains from these sites (Do and Hwang 1961:52). remains is 1, not “a small number of individuals” (emphasis added). Figure 2. Chulmun chronology. This graph illustrates 36 dates from the 12 Early Chulmun sites, 57 dates from the 12 Middle Chulmun sites, and 10 dates from the seven Late Chulmun sites. Dates with standard errors of more than 100 years are not illustrated here. The bars of the date ranges are arranged in reverse order of the Chulmun samples listed in Lee Transition to Farming in Korea S315 in Korea. The following case studies provide good examples and Hokkaido, northern Japan (Crawford 1983). An anthro- because the archaeobotanical materials were collected and an- pogenic community is often evident before the emergence of alyzed systematically. food production in archaeological records worldwide. The Bibongri shell midden (35Њ24
N, 128Њ37
E) is located Early Chulmun (ca. 7500–5500 BP) on a foothill in a steep valley along a subsidiary channel of the Nakdong River. Tools made of perishable materials are Because of waterlogged conditions, two Early Chulmun shell exceptionally well preserved, including dugout canoe planks, middens, Sejukri and Bibongri, yielded well-preserved organic wooden posts, and fiber mats (Gimhae National Museum Њ
Њ
remains (table 2). Sejukri (35 27 N, 129 21 E) is located in 2008). Faunal diversity here is high, consisting of terrestrial the northeast corner of a small bay along the east coast of (e.g., wild boar, deer), marine (e.g., shark, rays), and riverine Korea (fig. 1). Remains of both terrestrial and marine mam- (e.g., carp) species. Moreover, cranial bones, femurs, and hu- mals are abundant, along with fish and shellfish. meri of domesticated dogs were found in the Middle Chul- I identified at lest 16 taxa of seeds, along with nut remains mun context (Kaneko 2008:311–312). One maxillary second (G.-A. Lee 2003). A few pits are filled with remains of several molar of cattle was also recovered, and it is reported to be Quercus species (table 2). The AMS dates of charred acorn similar to those of wild cattle Bos taurus that are distributed encrustations on vessels are associated with the Early Chul- in Japan (Kaneko 2008:317). mun (table 1). A high proportion (80%) of the weeds in seed I found 19 taxa of flora at Bibongri (G.-A. Lee 2008b). counts were found in several pits and a hearth, but no crop Nutmeat and shells of Quercus and Juglans are dominant. remains were found. The absence of domesticates does not Possible dietary sources other than nuts are several taxa of necessarily mean that plant resources at Sejukri were less im- edible fruits and wild green onion (Alium sp.) that were en- portant. Besides nut remains, edible seeds/fruits of annual crusted on pottery. Possibly useful resources for medicinal herbaceous plants and shrubs are abundant. Early Chulmun purposes are spicebush, dogwood, and Japanese snowbell (ta- inhabitants may have familiarized themselves with these ble 1). Like Sejukri, Bibongri indicates use of anthropogenic plants by observing or even making use of them. In particular, flora from both disturbed upland and lowland zones. A few chenopod, abundant in most samples at Sejukri, may have charred grains of foxtail millet were found at Bibongri, but had economic value to the inhabitants as a source of greens it is not clear whether they belong to the early or the middle or as grains. Chenopodium album, a common Asian chenopod, phase. occurs frequently in the archaeological record in Asia where Bibongri and Sejukri data suggest the possibility of early millet farming was practiced (Crawford 2006; Weber 1998). management of wild/weedy species that had economic value, A relevant ethnographic example is found among the Central including edible nuts, annual herbaceous greens and seeds, Mountain aborigines in Taiwan, who mix foxtail millet with and fruit-bearing shrubs. Wild green onion at Sejukri is the C. album and sow them together as crops (Fogg 1983:100). only evidence for tuber use. Nuts were stored intensively. This annual weed is still used as greens in Korea (Pemberton Further research is required to know whether intentional and Lee 1996:62). planting of wild plants took place in the Early Chulmun pe- Like millet, chenopod may have been valued for its wide riod. adaptability to short growing seasons and deficient soils as well as for its large, compact heads with many small seeds Middle and Late Chulmun (ca. 5500–4000 BP) that were easy to collect and process (Weber 1998:267). The abundance of chenopod seeds suggests that it may have been The distinction between the Middle and Late Chulmun is a protected garden plant, if not intentionally propagated, dur- mainly based on ceramic typology. Most sites that I have ing the Early Chulmun. In addition to chenopod, the presence examined were occupied in both phases (table 1). Three sites of knotweed, panic grass, and grass of Hordeae (Elymus or are on islands along the southeast (Tongsamdong: 35Њ05
N, Agropyron sp.) document other annuals at Sejukri that thrive 37Њ04
E) and central west coasts (Nambukdong: 37Њ27
N, in disturbed areas. They seem to take advantage of human- 126Њ24
E; Sammokdo: 37Њ29
N, 126Њ28
E), while five sites oc- induced disturbance that often leads to generating open can- cur in inland settings; three are situated in the alluvial flats opies and an increase in soil nutrients by waste. These plants, along the Nam River (Oun 1: 35Њ14
N, 127Њ58
E; Sangchon often called “anthropogenic flora,” were important in sub- B: 35Њ13
N, 127Њ58
E; Pyeonggeodong: 35Њ13
N, 128Њ07
E), and sistence from the Initial Jomon (8000 BP) onward in Tohoku two are located on the hillslopes of the central west (Neung-

table 1, with the oldest Early Chulmun date at the bottom and the youngest Late Chulmun date at the top. The black and white parts of each bar indicate 1- and 2-j ranges of the calibrated BP dates, respectively. Gray boxes indicate the Middle Chulmun (left) and the Early Mumun (right) periods that are based on both chronometric data and typology. See sources listed in table 1. Table 2. Taxa identified in the Chulmun sites

Period Early Middle–Late Category, family, common name Scientific name Bbr Sjr Ng Nbd Smd Ang Tsd Nam Pgd Cultigens: Poaceae: Broomcorn millet Panicum miliaceum XXXXX Foxtail millet Setaria cf. italica italica X? X X X X X Wheat Triticum aestivum Rice Oryza sativa X? Fabaceae: Azuki bean Vigna angularis X Soybean Glycine max X Weeds: Aliaceae: Wild onion Alium X X Asteraceae: Dandelion? Taraxacum XX Cannabaceae: Hop Humulus X Caryophyllaceae: Snow in summer Cerastium X Chenopodiaceae: Chenopod Chenopodium XX X X X X Brassicaceae: Wild mustard Brassica X Labiatae: Honey weed Leonurus sibiricus X Beefsteak plant Perilla frutescens XX Lauraceae: Spicebush Lindera X Fabaceae: Wild bean cf. Lespedeza/Melilotus XXX Poaceae: Barnyard grass? Echinochloa cf. crusgalli X Green foxtail Setaria cf. italica viridis XX XXX Panic grass Panicum XX Panicoid grass Paniceae X X X X X Wheatgrass/wild rye Agropyron/Elymus X Polygonaceae: Knotweed Polygonum XXXX Dock Rumex X Solanaceae: Nightshade Solanum cf. nigrum X Rubiaceae: Bedstraw Galium XX Cyperaceae: Bulrush Scirpus juncoides/ X Scirpus fluviatilis Alismataceae: Arrowwood Sagittaria latifolia X Actinidiaceae: Wild kiwi Actinidia XX Rosaceae: Plum Prunus X X Rosaceae: Bramble Rubus XX X Other X X X Vitaceae: Wild grape Vitis XX X X Ramanaceae: Chinese jujube Ziziphus jujuba X Lee Transition to Farming in Korea S317

Table 2 (Continued) Period Early Middle–Late Category, family, common name Scientific name Bbr Sjr Ng Nbd Smd Ang Tsd Nam Pgd Solanaceae: Ground cherry Physalis X Styracaceae: Japanese snowbell Styrax japonicus XX Nuts: Fagaceae: Oak Quercus aliena, Quer- XX X cus serrata, Quercus variata Pinaceae: Pine Pinus X Juglandaceae: Manchurian walnut Juglans mandshurica XX Cornaceae: Dogwood Cornus XX Domesticated animals: Canidae: Dog Canis familiaris XX Bovidae: Cattle Bos cf. taurus XX Note. Bbr p Bibongri; Sjr p Sejukri; Ng p Neunggok; Nbd p Nambukdong; Smd p Sammokdo; Ang p Anganggol; Tsd p Tongsamdong; Namp Nam River; Pgd p Pyeonggeodong. The Oun 1 and Sangchon B sites are collectively called the Nam River sites because they share the same cultural traits in close proximity. Faunal remains at Bibongri were recovered from the Middle Chulmun context. gok: 36Њ22
N, 126Њ48
E) and southwest regions (Anganggol: sites yield domesticated foxtail and broomcorn millets (fig. 36Њ47
N, 127Њ06
E). 3). In addition, beefsteak plant from house floor fills at The Tongsamdong site is located on an islet in Busan Neunggok may also represent a domesticated variety. At Harbor on the southeast coast (fig. 1). A shallow embayment Neunggok, acorn remains of several oak species are abundant, adjacent to the site probably provided rich marine resources. similar to those from Bibongri and Sejukri. Chenopod and The numbers of terrestrial and marine mammal bones and panicoid weeds are dominant at Anganggol and Neunggok, various fish and shellfish remains are enormous. House floor similar to the evidence at the Tongsamdong site (G.-A. Lee fill at Tongsamdong yielded two domesticated taxa—foxtail 2008a). and broomcorn millet—and several herbaceous annuals, in- The Oun 1, Sangchon B, and Pyenggedong sites are located cluding bedstraw, chenopod, knotweed, green foxtail, and di- within a 100-km reach of the Nam River channel (fig. 1), verse panicoid weeds (table 1). Domesticated taxa made up sharing similarity in surrounding landscapes and cultural re- more than half of all seeds. As in the Sejukri samples, che- mains. A total of 400 L of soil were collected from outdoor nopod was the most abundant weed and is almost as common hearths and pits from Oun 1 and Sangchon B (G.-A. Lee as foxtail millet. 2003). Domesticated species include foxtail and broomcorn Two island sites along the central west coast, Nambukdong millets, which are dominant in the samples. Two legume seeds and Sammokdo, show different patterns of plant use. Sam- at the Oun 1 site seem to belong to the genus Vigna, which mokdo was a dwelling site, while Nambukdong was probably is the genus of domesticated azuki bean (table 1). Weed taxa a seasonal task camp for gathering shellfish (J.-j. Lee 2006). are similar to those found at other Middle-Late Chulmun Both sites show a similarity in material culture. At Sammokdo, samples, but the presence of chenopod is negligible. Instead, one charred seed each of chenopod and wild legume was panicoids account for 80% of all the weedy seeds. Panicoid- recovered, along with possible tuber fragments from 75 L of dominant weed composition continued until the Mumun pe- floor fills (G.-A. Lee 2009). Plant remains were also rare at riod. Panicoid weeds probably flourished in gardens and fields Nambukdong; a total of 184 L of hearth fill revealed one and were included with harvested crops and left in activity dogwood seed (G.-A. Lee 2006). or dumping areas after threshing. Nuts, both acorn and Man- Neunggok and Anganggol are dwelling sites on a gentle churian walnut, are present at Sangchon B. Although one of hillslope dating to the Middle–Late Chulmun (fig. 1). Both the AMS dates for charred rice grains at Oun 1 raises a ques- S318 Current Anthropology Volume 52, Supplement 4, October 2011

Table 3. Human dietary reconstruction based on isotope and trace-element analyses

Period, site, province Sample Results Implication Late Mumun: Nukdo, South p 15 ע a 13 p Ϫ Gyeongsang 48 human remains: skull, d C 18.3‰ 0.4‰, d N 11.2‰ DietbasedonC3 plants and terrestrial for adult (n p 15); d13C p protein with a supplementary marine 0.7‰ ע limb, rib, femur, vertebra protein ע d15N p 12.5‰ ,0.7‰ ע Ϫ18.7‰ 1.1‰ for juveniles (n p 33) Middle Chulmun: Tongsamdong, South Gyeongsangb 1 human metacarpal d13C p Ϫ14.8‰; d15N p 18.1‰ Diet based on marine protein Late Chulmun: Konamri, South Chungcheongc 1 human femur Sr p 309 ppm; Ba p 73 ppm; Mixed diet of plants, low terrestrial Zn p 54.2 ppm protein Daejukri, South Chungcheongc 1 human femoral diaphysis Sr p 435 ppm; Ba p 280 ppm; Diet based on plants Zn p 66.3 ppm Konamri, South d 13 p Ϫ 15 p Chungcheong 1 human bone d C 17.83‰; d N 9.12‰ Diet based on C3 plants with a supple- mentary animal protein Mumun: Konamri, South d 13 p Ϫ 15 p Chungcheong 1 human bone d C 12.2‰; d N 10.1‰ Diet based on C4 plants Early Chulmun: Yondaedo, South Gyeongsange 2 human male and 1 female Ba/Sr p 0.408, 0.416, 0.442; Diet based on plants with supplemen- Zn p 203, 195, 280 ppm tary fish and shellfish United Shilla: Imdang, North ע f 13 p Ϫ Gyeongsang 2 human males and 1 fe- d C 17.5‰ 0.7‰; DietbasedonC3 plants 1.3‰ ע male; possibly 3 human d15N p 10.3‰ males, and 3 females; 8 unidentified a Choy and Richards (2009). b Choy and Richards (2010). c An (2006b). d An (2006a). e An (2009). f Shin and Lee (2009). tion concerning the earlier use of rice (table 1), rice use in veals a large quantity of legumes of two genera, Vigna and the Chulmun is not confirmed yet (Crawford and Lee 2003; Glycine. The pumule-hypocotyl (embryonic leaves and stem) G.-A. Lee 2003).8 shape and size relative to the cotyledon appear to distinguish My ongoing analysis finds at least 16 taxa of charred seeds East Asian azuki (Vigna angularis) from the South Asian mung and nut remains at the Pyeonggeodong site (table 1). Foxtail bean (Vigna radiata; Yoshizaki 1992). The morphology of and broomcorn millet, as well as acorn, are abundant here Vigna cotyledons resembles that of azuki rather than that of 9 (fig. 4). Unlike any other Chulmun site, Pyeonggeodong re- mung bean (fig. 5). Anthropogenic vegetation is well documented at Pyeong- 8. Rice grains from an Early Mumun context are dated to the Late Chulmun, although four other AMS dates from the same context all fall geodong, including panicoid grass, chenopod, knotweed, into the Early Mumun (table 1). Postdepositional disturbance may have chickweed, honey grass, nightshade, wild legume, and plum occurred, considering the proximity of the Chulmun and Mumun fea- (table 2). The AMS dates for broomcorn millet, azuki, and tures at Oun 1 (G.-A. Lee 2003). Although Norton (2007:141) quotes soybean (4950–4660 cal BP) here fit the Middle Chulmun the earliest accepted evidence of rice from Oun 1 in the Late Chulmun from Crawford and Lee (2003), the original lines read that “no rice has chronology (table 1). Soybean seeds at Pyeonggeodong (fig. been recovered from unequivocal Chulmun contexts at Oun 1,” and that 6) are about half the size of seeds of the domesticated soybean “rice as well as millet was grown during the Late Chulmun, but the from the Mumun component at the Nam River and Daun- evidence needs strengthening” (92–94). A recent publication by Ahn dong sites (Crawford and Lee 2003; fig. 7). Pyeonggeodong (2010:91) also misread Crawford and Lee’s interpretation on this find. 9. Legumes of genus Vigna were identified at Oun 1, but the quantity soybean specimens fall into the size ranges of Late Longshan is small. (ca. 4500–4000 BP) soybeans in China (Lee et al. 2007). Soy- Lee Transition to Farming in Korea S319

Figure 3. Foxtail millet from the Chulmun Neunggok site. Scale bar p 0.5 mm. beans from the Middle Jomon sites in Kyushu (ca. 5300–4400 insula is probably a part of the large area in which azuki and cal BP) were interpreted as belonging to a domesticated va- soybean were domesticated. riety on the basis of their morphology and size (Obata 2008). Size, however, should not be the sole distinguishing trait of Chulmun Resource Production domestication in earlier archaeological contexts (Crawford et Recent archaeobotanical study confirms Chulmun as a al. 2005). Small legumes could be from purposefully managed resource-producing economy by the middle phase. Several populations that resulted in initial domestication even though sites confirm foxtail and broomcorn millet cultivation they still retained the morphology of wild species. Even if they throughout the Middle–Late Chulmun, beginning by 5500 cal were wild, the dense concentration of soybeans at Pyeong- BP. Recent discovery of azuki and soybean at Pyeonggeodong geodong indicates their important economic value to the set- and beefsteak plant at Neunggok and Angangol indicates the tlers. The seed morphology and size of azuki beans at Pyeong- earlier appearance of upland cultivation of multiple crops in geodong match those of the domesticated specimens found Korea by the Middle Chulmun.10 As supported by three AMS at the Mumun sites as well as those found in the Middle and Late Jomon sites (Obata 2008). The Pyeonggeodong data sup- 10. The seeds found here resemble those of the modern domesticated port phylogenetic studies that propose multiple origins of beefsteak plant (Perilla frutescens), although their full domestication status soybean (Abe et al. 2003; Xu et al. 2002). The Korean Pen- has yet to be confirmed. S320 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 4. Broomcorn millet from the Chulmun Pyeonggeodong site. Scale bar p 0.5 mm. dates on soybean and azuki at Pyeonggeodong (table 1), use carefully considering any possible risk embedded in such a of both legumes around 5000 cal BP is clear, regardless of rapid shift. Could foragers adjust promptly to new sets of their domestication status. They were probably managed an- domesticates that required cultivation techniques and affected thropogenic plants, and the domestication process was cer- scheduling rhythms? I view the Chulmun economy from a tainly under way even if they were not fully domesticated. different perspective. Upland dry farming of millet and le- Millet farming in the Chulmun and Jomon is often con- gumes per se was a well-adapted sustainable strategy for a sidered a transitory phenomenon that was a prelude to in- prolonged time during the Chulmun. tensive (rice) agriculture in the Mumun and Yayoi (Crawford Weedy plants at Chulmun sites tell another interesting and Takamiya 1990:891). The gap between foraging and in- story. The diversity of taxa and morphology of weeds are tensive agricultural systems, exemplified by Chulmun and Jo- comparable with those found in Yangshao-Longshan sites mon food production, is often seen as a temporary, unstable (5500–4000 BP) in the Yiluo region of China (Lee et al. 2007) stage or a developmental transition that was both radical and and Jomon sites in northeastern Japan (Crawford 2011). Most rapid. In other words, this unidirectional view assumes that weedy remains represent millet-tribe annuals and other arable a foraging economy shifted to agriculture rapidly without species common in upland cultivation. The composition may Lee Transition to Farming in Korea S321

Figure 5. Vigna cotyledon from the Chulmun Pyeonggeodong site. Arrow indicates plumule hypocotyl. Scale bar p 1mm. represent a viable “garden” for Chulmun people as sources out Korea where diverse terrestrial resources of both plants of both domesticated and useful wild plants. The Chulmun and game were exploited. One of the common biases in Chul- may be another example of the earlier efforts of humans to mun subsistence studies is selective use of evidence for certain actively engineer their niche for food and other vital resources, edible resources. The visibility of sizable remains gives further as documented in Jomon sites and in Late Archaic eastern weight to the importance of shellfish and edible nuts, whereas North America (Smith 2001). mostly smaller, invisible seed/fruit and tuber remains are sim- Confining the Chulmun economy to a narrowly defined ply overlooked unless systematic recovery efforts are applied. concept of either specialized hunting and gathering or broad- Fortunately, a recent increase in archaeobotanical and faunal- spectrum subsistence does not seem to be the right direction. archaeological efforts illustrates the Chulmun as a complex The large number of Chulmun shell midden sites is often economy beyond a specialized subsistence based on shellfish used as evidence of a specialized economy focusing on marine and nuts. For instance, several coastal and offshore sites in resources. A substantial number of residential shell midden central western Korea demonstrate a variety of faunal remains, sites (e.g., Tongsamdong, Sejukri, Bibongri), however, reveal including 36 species of terrestrial mammals, birds, fish, and richness of both marine and terrestrial fauna as well as plant shellfish (E.-y. Kim 2007; J.-j. Lee 2006; Shin 2007). Dominant resources (J.-j. Lee 2001). Moreover, more sites in inland set- species are different across locations, but a complete picture tings (e.g., floodplains, hillslopes) have been found through- does not support the idea of a specialized economy based S322 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 6. Glycine specimen from the Chulmun Pyeonggeodong site. The ventral view of this specimen shows the hypocotyl radicle (embryonic stem and root, left arrow) and the hilum (right arrow). Scale bar p 1 mm. mostly on shellfish and fish. Isotope and trace-element anal- Lee (2001:317) regards the use of crops as a risk-reduction yses, although limited, show different dietary patterns, de- strategy against sea level decline on the east and south coats pending on the location (table 3). Overall, the Chulmun peo- and as population–marine resource imbalance on the west ple seem to have diversified the subsistence strategies into coast from 4000 BP onward. procuring seasonally available niche-specific marine and ter- Norton (2000) emphasizes population increase as a key restrial resources, possibly managing wild/weed plants, and condition for the adoption of rice farming along the west tending domesticates (millet, legumes, beefsteak plant, and coast. Based on the overrepresentation of cranial bones of dog), from at least the middle phase. large-sized fish at the Konamri shell midden in South Jeolla Province, Norton suggests the differential processing of large Assumptions Revealed fish as evidence of residential stability (fig. 1). Residential stability increased human population throughout the Chul- Most hypotheses about the origins of agriculture in Korea mun, causing a decrease in fish sizes and favorable taxa and offer a direct cause-and-effect explanation, whether the cause subsequently pushing the Chulmun hunter-gatherers to adopt is population increase or climate change. This paper aims to rice farming. test the plausibility of these arguments against available ar- On the other hand, J. Kim (2003, 2006) suggests the com- chaeobotanical and paleoenvironmental data, radiocarbon bination of environmental deterioration and subsequent dates, and cultural conditions. In this attempt, it is crucial to southward migration as a major cause of the agricultural tran- distinguish two transitions in agriculture in Korea: the emer- sition in Korea. Because of a cooling climate around 4000– gence of resource production and the shift to intensive farm- 3000 BP, the farmers in the Jilin-Duman regions along the ing of both wet and dry crops. current border with China migrated to the south, which was Quantitative analyses of marine resources have been used better suited for farming.11 Kim assumes constrained logistical to examine subsistence changes in the Chulmun-to-Mumun transition (e.g., J.-j. Lee 2001, 2006; Norton 2000, 2007). J.-j. 11. Duman is spelled “Tumen” in Chinese. Lee Transition to Farming in Korea S323

Figure 7. Glycine specimen from the Mumun Daundong site. The ventral view of this specimen shows the hypocotyl radicle (left arrow) and the hilum (right arrow). Scale bar p 1 mm. mobility for indigenous hunters-gatherers when immigrant and Tongsamdong in the south indicate that their inhabitants agriculturalists blocked the way to resource patches. The in- managed complicated scheduling of diverse marine and ter- accessibility of foraging territories prompted the conversion restrial resources. A change to a mixed forest probably di- of hunter-gatherers to farmers. versified resources rather than impoverishing the choices for the Middle Chulmun people. In most cases of primary origins, The Beginning of Resource Production as a Stress Response agriculture appeared in complex foraging economies in afflu- ent environments where risk was affordable (Price and Ge- As reviewed above, the Middle Chulmun data indicate that bauer 1995:8). Chulmun does not seem to be an exception. cultivation of millet and legume was practiced by 5500 cal Resource stress due to environmental degradation probably BP and around 5000 cal BP, respectively, before the proposed was not a direct cause of the appearance of millet and legumes. period of sea level decline and subsequent resource reduction around 3000 BP (table 2). It is not clear whether the ap- Furthermore, evidence of resource stress due to high popu- pearance of crops in Korea was concurrent with the climatic lation density in the Chulmun period is lacking (Lim 2009). amelioration. A recent palynological study suggests that a Current research on paleoclimate in Korea is too preliminary cooling episode between 6000 and 4500 cal BP caused a switch to address any environmental impact on emerging food pro- from oak-dominated forests to mixed conifer-deciduous duction conclusively. broad-leaved forests in the central west coast (Jun, Yi, and Another problem with previous research is the assumption Lee 2010). Deprivation of resources because of a cooling con- of incompatibility between agricultural and foraging strate- dition, however, is not evident in the Middle Chulmun rec- gies. For example, some claim that Jomon inhabitants in ords (5500–4000 BP). For example, my ongoing research northeastern Japan maintained their foraging economy be- shows an abundance of acorn at Neunggok in the central cause of the conflicts in resource scheduling that farming western region. Faunal and plant data from Bibongri, Sejukri, might have brought (e.g., Akazawa 1986; Matsui and Kane- S324 Current Anthropology Volume 52, Supplement 4, October 2011 hara 2006). This notion of scheduling conflicts is based on energy and time and is storable for an extended period, that limited ethnographic data but not on the substantial archae- resource can be highly desirable to settlers. Certainly millet ological data of early Jomon farming (Crawford 2008). and legumes fit the profile. In brief, the resource production A misunderstanding of Chulmun scheduling practices is in the Chulmun is comparable with one argued for the Ar- similar to the case of Jomon. Was the complex scheduling of chaic period of eastern North America (Smith 2011). As in exploiting marine/terrestrial resources seriously compromised eastern North America, the Chulmun does not provide much by the demands of agriculture? It depends on what kind of support for general models that incorporate environmental cultivation and crops were adopted. Ethnographic data from downturn, external environmental stress, population growth, Korea indicate that oysters and other common shellfish are landscape packing, constricted resource zones, and carrying- mostly exploited between late fall and early spring (J.-j. Lee capacity imbalance in explaining the initial domestication (or 2001:300). Shellfish harvesting is rarely pursued in summer adoption) process. because spawning makes shell meat less palatable and more susceptible to pest infection. In contrast, June and July are Sea-Level Changes as the Cause of Rice Farming the best sowing months for the autumn varieties of millet, azuki, and soybean, the early crops that appeared in the Mid- As reviewed above, the population-resource-imbalance model dle Chulmun, and these crops can be harvested by early fall, depends on an assumption of an abrupt sea level decrease at just before fishing and acorn-harvesting season (Jo 1995). the transitional period around 4000–3000 BP. Information on Moreover, none of these crops require demanding soil con- sea level fluctuations around the Korean Peninsula is, how- ditions. Millet, in particular, is a low-maintenance, quick- ever, inconclusive. Before the hypotheses of cultural changes growing, drought-resistant crop. Even nomads often culti- related to the oscillating sea level can be scrutinized, a brief vated broomcorn millet because of its ability to mature summary of different estimations of sea level changes around quickly, sometimes within 6 weeks (Purseglove 1972:199). It the peninsula during the Holocene is necessary. is, therefore, unlikely that coastal inhabitants were prevented According to a popularly quoted hypothesis on sea level from millet (and legume) cultivation because of the sched- fluctuations, the sea level rose almost to the present level uling conflicts or the unsuitable landscape along the coast. around 6000 BP (Early Chulmun), decreased slightly around Why did foragers invest their efforts in cultivation when 4000 BP (Late Chulmun), increased to a maximum at 3200 no immediate external stress existed during the Middle Chul- BP, and decreased again between 3000 and 2300 BP (Early– mun, supposedly during the time of a climatic optimum? Late Mumun), with a fluctuation gradient of 1–2 m (Hwang Sedentary conditions may give a hint to the answer. Sedentism and Yoon 1999, 2002). On the contrary, Kim et al. (1999) existed in many areas where agriculture originated and spread and Jang and Park (2002) suggest a gradual sea level rise on (Price and Gebauer 1995:8). the west coast until the mid-Holocene and stability thereafter. Evidence of winter occupation and abundance of pottery Models of sea level fluctuation in Japan are often cited in for storage can be found in most Chulmun sites (Nelson 1992: Korean archaeology. Several studies in Japan present a max- 180). House structures and faunal seasonality at Tongsamdong imum sea level around 6000 BP and a decrease afterward, but indicate that inhabitants may have lived there year-round on the timing and magnitude of oscillations are estimated dif- a permanent basis (J.-j. Lee 2001). A prerequisite for seden- ferently according to the locations and methods applied. For tism is a sufficiently reliable food source to sustain a group example, Toizumi (1999) explains the changing frequency of without having to move around (Higham 1995:151). Complex Jomon shell middens in Tokyo Bay as a result of fluctuations foragers sometimes use storage to guard against seasonal or in shellfish availability corresponding to an abrupt sea level annual fluctuations of resources. Food-storing foragers must fall (3500–2000 BP) below the current sea level. In contrast, be able to exploit diverse species so that fluctuations in the Sato’s (2008) reconstruction in the Seto Inland Sea does not availability of any one of them are not catastrophic (Rowley- suggest any downfall below the current level during the Late Conwy, comment on Testart 1982:533). Adoption of domes- Holocene. ticates may have provided an opportunity to expand storable Unlike the common assumption of the synchronicity be- food and thus to increase a stability in food supply. Testart tween the maximum sea level and shoreline transgression, (1982:535) even suggests that once foragers are practicing Shimoyama and Nishida (1999) reveal that the highest sea intensive storage and living a sedentary life, they can, without level and the Jomon Transgression were not concurrent.12 Di- any immediate major changes in their way of life, adopt a atom and marine clay deposition from the Saga Plain in new crop. I suggest that this depends on the required level northern Kyushu suggests that the transgression had already of investment in crops adopted. As the Bibongri, Tongsam- turned to a regression at 7000 BP, at least 1,000 years earlier dong, and Sejukri data indicate, Chulmun inhabitants along than the maximum sea level span. Considering the proximity the coast managed a complex seasonal foraging schedule with labor-intensive techniques and a storing strategy in an affluent 12. The forward movement of shorelines from 15,000 BP is commonly environment (Kaneko and Nakayama 1994; Sample 1974). If called the “Jomon Transgression” because the transgression became re- a new resource can be acquired with minimum expense of markable in the Jomon period (10,000–2500 BP). Lee Transition to Farming in Korea S325 of Kyushu to Korea, this study also provides a cautionary tale the Huanghe Delta plain at the mouth of the Bohai Sea show for Chulmun studies. Shoreline retreats and marine-resource an abrupt cooling period (4500–2700 cal BP) at the end of fluctuations may not have occurred concurrently with the sea the Holocene hypsithermal climatic conditions from 9800 to level decline. The local topography of coasts and channel 4500 cal BP (Yi et al. 2003:624). However, the effects of the systems probably affected shorelines differently. Direct cooling trend on the landscape and agricultural productivity triangulation of the sea level fall, shoreline regression, and in the Jilin-Duman regions have not been studied in detail. marine-resource decrease as explanations for the onset of ag- It is premature to conclude that migration resulted from a riculture should be reconsidered. reduction in agricultural productivity because of the cooling Another archaeological concern regarding sea level history trend. is a simplistic correlation between uncalibrated dates on ma- Let us review the claim that the transition from the Late rine samples and archaeological events without a proper cal- Chulmun to the Early Mumun in central western Korea was ibration based on marine reservoir effects. One of the ex- one of the most rapid transitions from foraging to agricultural amples is the two dates for cored shell fragments from economies in the world (J. Kim 2003, 2006). The rapidity of Daldong, Ulsan, along the east coast (table 1). Simply adding the transition should be put to the test in light of the chrono- error ranges seems to confirm that the sea level decreased metric dates. Radiocarbon dates have often been used to ex- during the Chulmun-to-Mumun transition, around 3200 BP plain cultural changes without a distinction between the ter- uncal BP) and 3000 BP (TH-1004: restrial and marine calibrations. If only the dates of terrestrial 110 ע TH-1005: 3090) uncal BP; Hwang and Yoon 2002). After being samples are considered, most Late Chulmun dates overlap 125 ע 2910 calibrated,13 however, the two dates correspond not to the with either the Middle Chulmun or the Early Mumun (fig. beginning of the Mumun (and the emergence of agriculture) 2) and do not support Kim’s claim of little overlap of dates but to the Middle Mumun phase, 2860–2340 cal BP. In other between the Late Chulmun and Early Mumun sites (J. Kim words, a sea level decrease was coincident with neither the 2003:304). In contrast, a recent study by Kim and Bae (2010) Early Mumun nor the transition to intensive agriculture. In suggests a prolonged transition from the Neolithic (Chulmun) conclusion, the hypothesis that sea level decline facilitated the to the Bronze Age (Mumun). Based on limited numbers of emergence of agriculture between 3000 and 2300 BP is un- dates from Late Chulmun sites, it is inconclusive whether and supported. Accordingly, external-stress models rarely have the how long the Chulmun may have overlapped with the Mu- strength to explain the transition as clearly and simply as they mun or whether the boundary between the two is different may seem to. from the one suggested. Simplistic explanations of a rapid An assumption implicit in this correlation is that marine transition, therefore, should be reconsidered. resources were the most important food source. Terrestrial The mechanism of the agricultural transition is explained taxa, particularly plants, were considered a minor resource as the conflicts over land use between farmers (newcomers) for foragers in coastal regions. Accordingly, a decline in ma- and foragers (indigenes) and a subsequent reduction in the rine resources was considered devastating to subsistence, logistical mobility of foragers (J. Kim 2003, 2006). J. Kim which must have led to the adoption of new strategies. Al- (2003) claims that no cultivated millet was found in the cen- though the Jomon culture in northeastern Japan is commonly tral west “despite recent effort to find cultigens by archaeo- 14 regarded as marine oriented, wild and domesticated plants botanists” (298). In reality, the irregular presence of do- played a substantial role (D’Andrea 1999). The data from mesticates in Korea reflects a research paucity rather than a settlements on the shoreline (Bibongri, Sejukri, and Tong- real picture, at least at the inland sites. Sizable Chulmun set- samdong) prove that plants, either crops or nuts, were a cru- tlements have rarely been examined for evidence of plant use. cial part of Chulmun subsistence. Moreover, as seen at Neung- Consequently, these Chulmun people were regarded as for- gok and Pyeonggeodong, domesticated crops and wild plants agers emphasizing riverine food sources and nuts. Whenever were important food sources for the Chulmun inhabitants in rigorous archaeobotanical work is applied, however, the Mid- the inland and riparian habitat, where both freshwater and dle–Late Chulmun sites in the central west also reveal evidence terrestrial resources were plentiful. of crop cultivation. For example, the Neunggok and An- ganggol sites (G.-A. Lee 2008a) confirm that millet cultivation was well established at least as early as in the other regions Migration as a Force for Intensive Agriculture from the Middle Chulmun (table 1). Contrary to the view Like the stress models, some of the social models also em- that any findings of domesticated taxa in the Chulmun are phasize a single factor—migration—as a strong impulse for rare and irrelevant exceptions in the overall Chulmun econ- the agricultural shift (J. Kim 2003, 2006). Populations moved omy (J. Kim 2006:170), it is more likely that Chulmun set- to Korea across the Duman River when the climate began to tlement sites would yield evidence of farming and its impor- deteriorate in northeastern China. Palynological records from tance if efforts were made to recover plant remains. The

13. Because there is no correction data on marine reservoirs around 14. No systematic archaeobotanical study, however, had been con- Korea, I used DR from the North Pacific for calibration (see table 1). ducted or reported in the Chulmun sites in the central west before 2003. S326 Current Anthropology Volume 52, Supplement 4, October 2011

Chulmun people were not pure foragers but resource pro- has yet to be confirmed.15 Unfortunately, systematic archaeo- ducers who acknowledged the importance of crops. botanical data are lacking for the critical transitional period According to J. Kim (2003:311), the Late Chulmun foragers from the Chulmun to the Mumun between 4000 and 3400 cal were forced to adopt agriculture because they could not afford BP. Because of a lack of data from several hundred years during the increasing mobility costs of resource exploitation when the transition, the question of the rate of changes to intensive the colonist Early Mumun farmers blocked the traffic. This agriculture cannot be resolved. hypothesis is based on the general fact that Korea is predom- Recently, some researchers have cast doubt on the abrupt inantly mountainous terrain and that the routes of movement replacement of the Chulmun by the Mumun. Shin (2007:11) are limited in number. Kim does not provide any geographical argues that cultural continuity in pottery making and site se- information on the distribution of resource patches and the lection indicates a gradual transition. The longevity of the Chul- location of the high-traffic nodes. This hypothesis, therefore, mun culture (over 4,000 years) may have been based on de- remains highly conjectural. veloping economic, technological, and social complexities. We must review carefully how cultural continuity and external in- fluence were integrated into the Chulmun-Mumun transition Discussion and Conclusion rather than simplify the transition as a sudden entire demise of the Chulmun and a wholesale replacement by the Mumun. Focus on only one causal factor is not a proper approach to What most studies in Korea overlook is the distinction understanding the complicated issue of food production. In- between the first adoption of crops and the later establishment stead, the transition to agriculture must be seen as a complex of intensive agriculture. We should consider these as two sep- process resulting from several forces operating simultaneously arate contexts of crop adoption during the Middle Chulmun on different chronological and spatial scales of resolution and the Chulmun-Mumun transition, respectively. The Early (Zvelebil 1986:167). Foragers adopted farming selectively to Chulmun people had already begun to manage useful wild fit local needs, which varied from region to region. There is plants, including small-seeded annual grass, fruit-bearing no single ultimate cause that would fit all situations. As re- shrubs, and nuts. Evidence from the Early Chulmun Bibongri viewed above, models heavily based on sea level changes and site indicates that tuberlike wild onion was used early on. resource-population imbalance do not provide a plausible ex- Because of a lack of preservation, it remains unanswered planation in the Chulmun-Mumun case. whether other starchy tubers were utilized, as in the Early Europe, a well-studied secondary region of agriculture, can Neolithic in the Yangtze Basin of South China (Cohen 2011; give a comparative perspective to prehistoric Korea. As Rowley- Zhao 2011) and the Initial and Early Jomon (Crawford 2011). Conwy (2011) summarizes succinctly, the spread of farming in From at least 5500 cal BP, Chulmun culture developed specific Europe is explained best by rapid but lurching migrations of subsistence solutions to floodplains, hilly regions, and coastal Neolithic farmers with a package of southwest-Asian domes- areas that involved distinctive combinations of wild (e.g., ticates. The argument is supported by multiple lines of evidence, acorn, Manchurian walnut), possibly managed (e.g., cheno- including a gap in radiocarbon dates between Mesolithic and pod, panicoid grass), and domesticated (e.g., foxtail and Neolithic sites, absence of evidence for local domestication, and broomcorn millets, possibly soybean, azuki, and beefsteak plant) species. Changes in plant use seem to have gone differences in ancient DNA between the Mesolithic and Neo- through similar phases in the Chulmun and Jomon periods. lithic populations. Detailed debate over rapid large-scale mi- Early Chulmun subsistence resembles that of the Initial Jomon gration in the Chulmun-Mumun transition is beyond the scope (9500–7000 BP), which is characterized by low densities of of this paper, but key evidence for the debate is lacking, unlike herbaceous annual plants while shrubs and nuts predominate the case in Europe. Although small in number, radiocarbon (Crawford 2011). The increasing importance of annual her- dates do not well establish the disjunction between the Late baceous species and shrubs is well documented in the Middle– Chulmun and the Early Mumun. Ancient human genes are Late Chulmun, similar to the case of the Early to Late Jomon simply not available in numbers sufficient to document mi- periods (7000–3000 BP). Chulmun and Jomon are good ex- gration. Most of all, in Korea, adoption of several domesticated amples of attempts at ecological engineering well before in- animals and plants did not occur all together, unlike the Eu- tensive agriculture was established. ropean case. Clearly, millet, azuki, and soybean were available The clearest archaeological evidence for intensive agricul- as early as 5500–5000 cal BP, while wheat, barley, and rice ture in Korea is found in the Early Mumun by 3400 cal BP. arrived later. However, the dates for the initial adoption of the Mumun people devoted multiple seasons, almost year-round, three large cereals are controversial. Were they incorporated to tending various domesticated plants, as evidenced by a into millet farming individually over a number of generations series of new crops that required labor-intensive care and the during the transitional period? Or were all these crops incor- construction of massive raised dry fields and irrigated rice porated as a package at once? Some Middle–Late Chulmun sites raise the question of the early availability of rice and wheat, 15. The argument on the beginning of rice farming in Korea is well as seen in the case of the Nam River sites, but such a claim summarized in Ahn (2010). See also footnote 6. Lee Transition to Farming in Korea S327 paddy fields (G.-A. Lee 2003). An overly simplified focus on Akazawa, T. 1986. Hunter-gatherer adaptations and the transition to rice farming often blurs the real complexity of Mumun ag- food production in Japan. In Hunters in transition: Mesolithic so- cieties of temperate Eurasia and their transition to farming. M. Zve- riculture. For example, rice farming has been viewed as a lebil, ed. Pp. 151–165. Cambridge: Cambridge University Press. cure-all remedy that solved the marine-resource reduction Ammerman, A. J., and P. Biagi, eds. 2003. The widening harvest: the (Norton 2007). Also, according to B. Kim (2006), the Early Neolithic transition in Europe: looking back, looking forward. Boston: Mumun agricultural economy, based heavily on rice culti- Archaeological Institute of America. vation, spread suddenly and swiftly into the Late Chulmun An, D.-I. 2006a. Dietary reconstruction at the Neolithic Konam-ri and Daejuk-ri sites by trace element analysis. Journal of the Korean foraging contexts with little evidence of a transitional period. Ancient Historical Society 53:23–41. [In Korean.] Rice farmers seem to be portrayed as a highly able group that ———. 2006b. Dietary reconstruction by stable isotopic analysis: the could have eradicated the Chulmun foragers. According to Konam-ri shell midden in Korea. Journal of the Korean Ancient the available archaeobotanical evidence, a heavy reliance on Historical Society 54:5–20. [In Korean.] rice probably occurred much later than the Early Mumun ———. 2009. Dietary reconstruction through trace element (Ba, Sr, Zn) analysis of human bones from the Neolithic Yondae-do site. period (G.-A. Lee 2003). Newcomers, if there were any, prob- Journal of the Korean Ancient Historical Society 66:5–26. [In Korean.] ably needed time to adjust to local environmental conditions, Barnes, G. L. 1999. The rise of civilization in East Asia: the archaeology particularly for rice, which required irrigation techniques. Al- of China, Korea and Japan. London: Thames & Hudson. ready available crops from the Middle Chulmun, such as mil- Bellwood, P. 2011. Holocene population history in the Pacific region let and legumes, would have been the first choice. Detailed as a model for worldwide food producer dispersals. Current An- thropology 52(suppl. 4):S363–S378. archaeobotanical study shows that millet and legumes were Choe, C. P., and M. T. Bale. 2002. Current perspectives on settlement, the most dominant and frequently used crops throughout the subsistence, and cultivation in prehistoric Korea. Arctic Anthro- Mumun period (Crawford and Lee 2003). Rice alone seems pology 39(1–2):96–121. to have played no more than a minor role in mitigating re- Choy, K., and M. P. Richards. 2009. Stable isotope evidence of human source stress, if there was any stress, and it did not serve as diet at the Nukdo shell midden site, South Korea. Journal of Ar- chaeological Sciences 36(7):1312–1318. a prime impetus for sociocultural changes. ———. 2010. Isotopic evidence for diet in the Middle Chulmun Contrary to the common view of millet cultivation in the period: a case study from the Tongsamdong shell midden, Korea. Chulmun as a transient, minor strategy, resource production Anthropological and Archaeological Sciences 2(1):1–10. was in fact a foundation for the long-lasting adaptation of Cohen, D. J. 2011. The beginnings of agriculture in China: a mul- the Chulmun culture. Chulmun subsistence represents a rich tiregional view. Current Anthropology 52(suppl. 4):S273–S293. Crawford, G. W. 1983. Paleoethnobotany of the Kameda Peninsula and diverse array of comparatively stable and successful so- Jomon. Anthropological Papers, no. 73. Ann Arbor: University of ciopolitical and economic solutions that should be recognized Michigan Press. and studied in their own right. ———. 2006. East Asian plant domestication. In Archaeology of Asia. M. T. Stark, ed. Pp. 77–95. Malden, MA: Blackwell. ———. 2008. The Jomon in early agriculture discourse: issues arising from Matsui, Kanehara, and Pearson. World Archaeology 40(4): Acknowledgments 445–465. ———. 2009. Agricultural origins in North China pushed back to I thank Leslie Aiello, Ofer Bar-Yosef, Laurie Obbink, and T. the Pleistocene-Holocene boundary. Proceedings of the National Douglas Price for organizing and leading the Wenner-Gren Academy of Sciences of the USA 106(18):7271–7272. conference successfully. I am grateful to Martin Bale, Gary ———. 2011. Advances in understanding early agriculture in Japan. Crawford, Li Liu, and Rory Walsh for critically reading the Current Anthropology 52(suppl. 4):S331–S345. manuscript. Many helped me collect the data: Jaeho Ahn, Crawford, G. W., and G.-A. Lee. 2003. Agricultural origins in Korea. Antiquity 77(295):87–95. Hyunjeong Cho, Seongsu Hong, Chunyoung Kim, Miyoung Crawford, G. W., and H. Takamiya. 1990. The origins and impli- Kim, Mungjin Kim, Minjeong Ko, Donghun Lee, Sangkil Lee, cations of late prehistoric plant husbandry in northern Japan. Sinya Shoda, Sieun Yang, Hopil Yoon, and more. My gratitude Antiquity 64(245):889–911. is to them. This research is partially funded by a Luce/Amer- Crawford, G. W., A. Underhill, Z. Zhao, G.-A. Lee, G. Feinman, L. ican Council of Learned Society East/Southeast Asian Ar- Nicholas, F. Luan, H. Yu, H. Fang, and F. Cai. 2005. Late Neolithic plant remains from northern China: preliminary results from chaeology and Early History Fellowship. Liangchengzhen, Shandong. Current Anthropology 46(2):309–317. D’Andrea, C. 1999. The dispersal of domesticated plants into north- References Cited eastern Japan. In The prehistory of food: appetites for change.C. Gosden and J. Hather, eds. Pp. 166–183. London: Routledge. Abe, J., D. Xu, Y. Suzuki, A. Kanazawa, and Y. Shimamoto. 2003. Do, Y. H. and K.-T. Hwang. 1961. Excavation report of the Jitapri site. Soybean germplasm pools in Asia revealed by nuclear SSRs. The- Pyeongyang, Democratic People’s Republic of Korea: Scientific oretical and Applied Genetics 106(3):445–453. Press. [In Korean.] Ahn, S.-M. 1997. The Neolithic subsistence system. In Korean history: Dongguk University Museum. 2006. Sejukri site. Gyeongju, Republic the Paleolithic and Neolithic periods, vol. 2. Y. H. Chong, ed. Pp. of Korea: Dongguk University Museum Press. [In Korean.] 491–504. Seoul: Korean History Editing Board. [In Korean.] Fogg, W. H. 1983. Swidden cultivation of foxtail millet by Taiwan ———. 2010. The emergence of rice agriculture in Korea: archaeo- aborigines: a cultural analogue of the domestication of Setaria botanical perspectives. Archaeological and Anthropological Sciences italica in China. In The origins of Chinese civilization. D. N. Keight- 2(2):89–98. ley, ed. Pp. 95–115. Berkeley: University of California Press. S328 Current Anthropology Volume 52, Supplement 4, October 2011

Fuller, D. Q. 2011. Finding plant domestication in the Indian sub- navia and central-western Korea. Journal of Archaeological Method continent. Current Anthropology 52(suppl. 4):S347–S362. and Theory 10(3):277–321. Gimhae National Museum, ed. 2008. Bibong-ri site. Gimhae, Republic ———. 2006. Resource patch sharing among foragers: lack of ter- of Korea: Gimhae National Museum. [In Korean.] ritoriality or strategic choice? In Beyond affluent foragers: rethinking Gyeongnam Development Institute. 2007. Pyeonggeodong site, Jinju: hunter-gather complexity. C. Grier, J. Kim, and J. Uchiyama, eds. a briefing for the excavation evaluation committee. Jinju, Republic Pp. 168–191. Oxford: Oxbow. of Korea: Gyeongnam Development Institute. [In Korean.] ———. 2008. Socioeconomic implications of storage facilities of the Hannam University Museum. 2003. Daechon-ri Neolithic site, Ok- Songgukri period. Journal of the Korean Archaeological Society 67: cheon. Daejeon, Republic of Korea: Hannam University Museum 4–39. [In Korean.] Press. [In Korean.] Kim, J. C., and C. J. Bae. 2010. Radiocarbon dates documenting the Harris, D. R. 1990. Vavilov’s concept of centres of origin of cultivated Neolithic–Bronze Age transition in Korea. Radiocarbon 52(2–3): plants: its genesis and its influence on the study of agricultural 483–492. origins. Biological Journal of the Linnean Society 39(1):7–16. Kim, Y. H., H. J. Lee, S. S. Chun, S. J. Han, and S. K. Chough. 1999. Higham, C. 1995. The transition to rice cultivation in Southeast Asia. Holocene transgressive stratigraphy of a macrotidal flat in the In Last hunters, first farmers: new perspectives on the prehistoric southeastern Yellow Sea: Gomso Bay, Korea. Journal of Sedimentary transition to agriculture. T. D. Price and A. B. Gebauer, eds. Pp. Research 69(2):328–337. 127–156. Santa Fe, NM: School of American Research Press. Kwak, J.-C., H. Fujihara, and J. Yanaki. 1995. Rice plant-opals ex- Hodder, I. 1990. The domestication of Europe: structure and contin- tracted from Neolithic pottery. Journal of the Korean Archaeological gency in Neolithic societies. Cambridge: Cambridge University Society 32:149–162. [In Korean.] Press. Lee, G.-A. 2003. Changes in subsistence patterns from the Chulmun Hwang, S.-I., and S.-O. Yoon. 1999. Influence of Holocene environ- to Mumun periods: archaeobotanical investigation. PhD disser- mental change on human activities in the Daegu Basin during tation, University of Toronto. prehistoric and ancient times. Journal of the Korean Archaeological ———. 2006. Plant remains recovered from the Nambukdong site. Society 41:1–36. [In Korean.] In Excavation report at Nambuk-Dong and Eulwang-Dong I site.S. ———. 2002. The environmental change and human activities at B. Yi, S. Lim, S. Yang, and E. Hong, eds. Pp. 231–242. Seoul: Seoul Sejuk, Hwangseong-Dong, Ulsan City, southeastern coast of Korea National University Museum. [In Korean.] during the middle Holocene. Journal of the Korean Archaeological ———. 2008a. Archaeobotanical study of the Asan New Town and Society 48:35–57. [In Korean.] three neighboring sites. In Jangjae-ri Anganggol Site (I),Asan. Im, H.-J. 1997. Neolithic chronology. In Korean history: the Paleolithic Chungcheong Research Institute of Cultural Heritage, ed. Pp. 341– and Neolithic periods, vol. 2. Y. H. Chong, ed. Pp. 305–315. Seoul: 345. Gongju, Republic of Korea: Chungcheong Research Institute Korean History Editing Board. [In Korean.] of Cultural Heritage. [In Korean.] Jang, J., and Y. A. Park. 2002. Sea-level changes around the Korean ———. 2008b. Plant remains from the Bibongri site. In Bibong-ri Peninsula. In Quaternary environment in Korea: environment and site. Gimhae National Museum, ed. Pp. 391–398. Gimhae, Re- humans. Y. A. Park and W. S. Gong, eds. Pp. 125–143. Seoul: Seoul public of Korea: Gimhae National Museum. [In Korean.] National University Press. [In Korean.] ———. 2009. Plant remains recovered from the Sammokdo site. In Jo, J.-Y. 1995. Dryfield farming. 4th edition. Seoul: Hyangmunsa. [In Excavation report of the Sammok III site, Incheon. S. Yi, S. Lim, S. Korean.] Yang, E. Hong, and Y. Hue, eds. Pp. 165–170. Seoul: Seoul National Jun, C. P., S. Yi, and S. J. Lee. 2010. Palynological implication of University Museum. [In Korean.] Holocene vegetation and environment in Pyeongtaek wetland, Ko- Lee, G.-A., G. W. Crawford, L. Liu, and X. Chen. 2007. Plants and rea. Quaternary International 227(1):68–74. people from the Early Neolithic to Shang periods in North China. Kaneko, H. 2008. Faunal remains from the Bibongri site. In Bibong- Proceedings of the National Academy of Sciences of the USA 104(3): ri site, Gimhae National Museum, ed. Pp. 305–390. Changnyeong, 1087–1092. Republic of Korea: Gimhae National Museum. [In Korean.] Lee, G.-A., H. Yun, M. Ko, and C. Kim. 2011. Plant use in the Nam Kaneko, H., and K. Nakayama. 1994. Tongsamdong shellmidden and River valley during the Neolithic period. Journal of the Yeongnam animal remains: focusing on the archaeological data from L. L. Archaeological Society 56:5-42. [In Korean.] Sample’s excavation. Journal of the Korean Archaeological Society Lee, J.-j. 2001. From shellfish gathering to agriculture in prehistoric 31:297–339. [In Korean.] Korea: the Chulmun to Mumun transition. PhD dissertation, Uni- Kang, I.-U. 2009. The appearance and spread of agriculture in the versity of Wisconsin, Madison. Zaisanovka Neolithic Culture, Primorie, Russia. In New approaches ———. 2006. Analysis of the shellfish from the Youngyudo and to prehistoric agriculture. S.-M. Ahn and J.-j. Lee, eds. Pp. 372– Ulwangdong islands: review of the site function. In Excavation 425. Seoul: Sahoi Pyoungnon. [In Korean.] report at Nambuk-Dong and Eulwang-Dong I site. S. B. Yi, S. Lim, Kang H. T., Y.-S. Choo, and K. Na. 1993. Radiocarbon dating and S. Yang, and E. Hong, eds. Pp. 269–280. Seoul: Seoul National calibration: re-calibration of dates in Northeast Asia. Journal of University Museum. [In Korean.] the Korean Archaeological Society 30:5–61. [In Korean.] ———. 2009. Archaeological approaches to documenting animals Kim, B. 2006. Political economy of wet rice cultivation in the Bronze in the Korean Peninsula. In New approaches to prehistoric agri- period in central western Korea. Journal of the Korean Archaeo- culture. S.-M. Ahn and J.-j. Lee, eds. Pp. 252–269. Seoul: Sahoi logical Society 58:40–65. [In Korean.] Pyoungnon. [In Korean.] Kim, E.-y. 2007. Resource procurement strategies in central-western Lee, Y. J., and T. S. Park. 1997. New materials on prehistoric rice Korea in the Neolithic period. In Issues in the Neolithic culture in cultivation in Korea from the Kawaji site, Gyeonggi Province. Pa- central-western Korea: proceedings of the 2007 Joint Conference of per presented at the 2nd International Academic Conference on the Seoul-Gyeonggi Archaeological Society and Korean Neolithic So- Agricultural Archaeology, Nanchang, China. ciety. Seoul-Gyeonggi Archaeological Society and Korean Neolithic Lee, Y.-J., and J.-Y. Woo. 2002. Recovery of rice grains at the Sorori Society, eds. Pp. 133–152. Seoul: National Museum of Korea. [In site. In Prehistoric agriculture in Asia and rice grains form the So- Korean.] rori Site. Chungbuk National University Museum, ed. Pp. 17–18. Kim, J. 2003. Land-use conflicts and the rate of the transition to Cheongju, Republic of Korea: Chungbuk National University agricultural economy: a comparative study of southern Scandi- Press. [In Korean.] Lee Transition to Farming in Korea S329

Lim, S. 2009. The adoption process of agriculture and socioeconomic Sample, L. L. 1974. Tongsamdong: a contribution to Korean changes during the Chulmun period in the southern Korean Pen- Neolithic culture history. Arctic Anthropology 11(2):1–125. insula. In New approaches to prehistoric agriculture. S.-M. Ahn and Sato, H. 2008. Reconstruction of Holocene sea-level change along J.-j. Lee, eds. Pp. 144–171. Seoul: Sahoi Pyoungnon. [In Korean.] the coast of Harimanada in the eastern part of the Seto Inland Lu, H., J. Zhang, K.-b. Liu, N. Wu, Y. Li, K. Zhou, M. Ye, et al. 2009. Sea, western Japan. Daiyonki-Kenkyu 47(4):247–259. [In Japanese.] Earliest domestication of common millet (Panicum miliaceum)in Seoul National University Museum. 2006. Nambukdong and Ul- East Asia extended to 10,000 years ago. Proceedings of the National wangdong sites, Yongyu Island. Seoul: Seoul National University. Academy of Sciences of the USA 106(18):7367–7372. [In Korean.] Matsui, A., and M. Kanehara. 2006. The question of prehistoric plant Shimoyama, S., and I. Nishida. 1999. Environmental changes during husbandry during the Jomon period in Japan. World Archaeology the last 10,000 years in north Kyushu, west Japan with special 38(2):259–273. reference to shoreline movements. Bulletin of the National Museum Minnis, P. E. 1985. Social adaptation to food stress: a prehistoric south- of Japanese History 81:249–266. western example. Chicago: University of Chicago Press. Shin, J. Y., and J.-j. Lee. 2009. Tracing status-dependent dietary dif- Nelson, S. M. 1992. The question of agricultural impact on socio- ferences in the Shilla period through stable isotope paleodietary political development in prehistoric Korea. In Pacific Northeast reconstruction from Imdang, Geongsan. Journal of the Korean Ar- Asia in prehistory: hunter-fisher-gatherers, farmers, and sociopolitical chaeological Society 70:84–109. [In Korean.] elites. C. M. Aikens and S.-N. Rhee, eds. Pp. 170–184. Pullman: Shin, S.-J. 2007. Retrospect and prospect of the research on the Washington State University Press. Korean Neolithic culture. In Issues in Neolithic culture in central- ———. 1993. The archaeology of Korea. Cambridge: Cambridge Uni- western Korea: proceedings of the 2007 Joint Conference of the Seoul- versity Press. Gyeonggi Archaeological Society and Korean Neolithic Society. Seoul- Norton, C. J. 2000. Subsistence change at Konam-ri: implications of Gyeonggi Archaeological Society and Korean Neolithic Society, eds. the advent of rice agriculture in Korea. Journal of Anthropological Pp. 7–38. Seoul: National Museum of Korea. [In Korean.] Research 56(3):325–348. Smith, B. D. 2001. Low-level food production. Journal of Archaeo- ———. 2007. Sedentism, territorial circumscription, and the in- logical Research 9(1):1–43. creased use of plant domesticates across Neolithic–Bronze Age ———. 2011. The cultural context of plant domestication in eastern Korea. Asian Perspectives 46(1):133–165. North America. Current Anthropology 52(suppl. 4):S471–S484. Obata, H. 2008. Description of legume seeds. Prehistoric and Ancient Son, P. K. 1987. Prehistory of the Udo, west coast of Korea. Tankuk Cultigens in the Far East 3:225–252. [In Japanese.] University Museum Series 3. Seoul: Tankuk University Press. [In O¨ zdog˘an, M. 2011. Archaeological evidence on the westward expan- Korean.] sion of farming communities from eastern Anatolia to the Aegean Stuiver, M., and T. Braziunas. 1993. Modeling atmospheric 14C in- and the Balkans. Current Anthropology 52(suppl. 4):S415–S430. fluences and 14C ages of marine samples to 10,000 BC. Radiocarbon Park, S. B., S. Lim, J.-j. Lee, and J. Kim, eds. 2001. Gado shell midden. 35(1):137–189. Daejeon, Republic of Korea: Museum of the Chungnam National Testart, A. 1982. Storage in hunter-gatherer societies. Current An- University. [In Korean.] thropology 23(5):523–537. Pearson, R. 1977. Paleoenvironment and human settlement in Japan Toizumi, T. 1999. The relationship between the marine environment and Korea. Science 197(4310):1239–1246. in the Holocene and Jomon shell middens in the Tokyo Bay area, Pemberton, R. W. and N.-S. Lee. 1996. Wild food plants in South central Japan. Bulletin of the National Museum of Japanese History Korea: market presence, new crops, and exports to the United 81:289–308. [In Japanese.] States. Economic Botany 50(1):57–70. Weber, S. A. 1998. Out of Africa: the initial impact of millets in South Piperno, D. R. 2011. The origins of plant cultivation and domesti- Asia. Current Anthropology 39(2):267–274. cation in the New World tropics: patterns, process, and new de- Weiss, E., and D. Zohary. 2011. The Neolithic southwest Asian foun- velopments. Current Anthropology 52(suppl. 4):S453–S470. der crops: their biology and archaeobotany. Current Anthropology Price, T. D., ed. 2000a. Europe’s first farmers. Cambridge: Cambridge 52(suppl. 4):S237–S254. University Press. Xu, D., J. Abe, J. Gai, and Y. Shimamoto. 2002. Diversity of chlo- ———. 2000b. Lessons in the transition to agriculture. In Europe’s roplast DNA SSRs in wild and cultivated soybeans: evidence for first farmers. T. D. Price, ed. Pp. 301–318. Cambridge: Cambridge multiple origins of cultivated soybean. Theoretical and Applied Ge- University Press. netics 105(5):645–653. Price, T. D., and O. Bar-Yosef. 2011. The origins of agriculture: new Yi, S., Y. Saito, H. Oshima, Y. Zhou, and H. Wei. 2003. Holocene data, new ideas: an introduction to supplement 4. Current An- environmental history inferred from pollen assemblages in the thropology 52(suppl. 4):S163–S174. Huanghe (Yellow River) Delta, China: climatic change and human Price, T. D., and A. B. Gebauer. 1995. New perspectives on the impact. Quaternary Science Reviews 22(5–7):609–628. transition to agriculture. In Last hunters, first farmers: new per- Yoon, M., H. Im, and S. Oh. 2004. Dongam-dong shell midden site: spectives on the prehistoric transition to agriculture.T.D.Priceand the third excavation. Seoul: National Museum of Korea. [In Korean.] A. B. Gebauer, eds. Pp. 3–20. Santa Fe, NM: School of American Yoshizaki, M. 1992. Ancient cultigen distribution. Kokogaku Janaru Research Press. 335:2–13. [In Japanese.] Purseglove, J. W. 1972. Tropical crops: monocotyledons. London: Long- Zhao, Z. 2011. New archaeobotanic data for the study of the origins man. of agriculture in China. Current Anthropology 52(suppl. 4):S295– Rowley-Conwy, P. 2004. How the West was lost: a reconsideration S306. of agricultural origins in Britain, Ireland, and southern Scandi- Zvelebil, M. 1986. Mesolithic societies and the transition to farming: navia. Current Anthropology 45(suppl. 4):S83–S113. problems of time, scale, and organisation. In Hunters in transition: ———. 2011. Westward ho! the spread of agriculture from central Mesolithic societies of temperate Eurasia and their transition to farm- Europe to the Atlantic. Current Anthropology 52(suppl. 4):S431– ing. M. Zvelebil, ed. Pp. 167–187. Cambridge: Cambridge Uni- S451. versity Press.

Current Anthropology Volume 52, Supplement 4, October 2011 S331

Advances in Understanding Early Agriculture in Japan

by Gary W. Crawford

Six episodes—the Jomon, Yayoi, Tohoku Yayoi, Satsumon and Ainu, Okhotsk, and Gusuku—of agricultural development are examined. These events involve both indigenous adaptations as well as migration and diffusion to and within the Japanese archipelago. All but Jomon subsistence adaptations began as a result of migration and diffusion. Jomon populations engaged in niche construction/ anthropogenesis that ranged from annual plant encouragement and probably management, lacquer tree (Toxicodendron verniciflua) and nut tree (Castanea crenata and Aesculus turbinata) management, and probable domestication of barnyard millet and soybean as well as cultivation of bottle gourd and hemp and possible cultivation of Perilla and adzuki. These characteristics place the Jomon in a middle ground that is neither hunting and gathering nor traditionally conceptualized agriculture. A brief comparison with China shows late Upper Paleolithic and Early Neolithic/Early Jomon similarities that can inform discussions about agricultural origins. The Okhotsk raised pigs, grew a few crops, hunted and gathered; this culture also does not fit traditional definitions of an agricultural economy. The other episodes involve forms of agriculture similar to those found in mainland East Asia.

Japan is not normally considered a region of independent or Satsumon) establishing food production in a significant por- primary agricultural origins. Yet evidence for domestication, tion of Hokkaido beginning between 1400 and 1200 cal BP. cultigens, plant management, and anthropogenesis indicate These transitions are reassessed in this paper, and two ad- that a unique form of resource production had developed in ditional transitions are considered: the expansion of agricul- Middle Holocene Japan. Furthermore, developments during ture to the central Ryukyu Islands (Okinawa) from the north the Pleistocene-Holocene transition in Japan have parallels and to the northern coast of Hokkaido (the Okhotsk culture) with developments in China where indigenous agriculture from mainland East Asia. quickly developed. The issue of primary origins is masked by the apparent contrasts between the economic systems of the Jomon and the apparent obliteration of indigenous systems Conceptualizing Agriculture by agricultural practices introduced from China and Korea. This paper explores the Jomon issue as well as subsequent My perspective does not adhere to an overly constricted def- developments in Japan. The complex processes I outlined inition of agriculture. I acknowledge that the widely held almost two decades ago still have validity today (Crawford perspective of what constitutes agriculture is Eurocentric, and 1992). At the time, I felt it constructive to approach the topic excellent discussions of this perspective can be found in Deur from the perspective of four transitions. The first involves and Turner (2005) and Vrydaghs and Denham (2007). I also small-scale Jomon practices (what I called “plant hus- reject the hunter-gatherer/farmer or hunting-gathering agri- bandry”). The second transition is tied to the development culture dichotomy, as do many others (Smith 2001, 2005). of the Yayoi culture beginning shortly after 3000 cal BP (re- The existence of a broad continuum or “middle ground” vised chronology) in southwestern Japan. Yayoi influence (Smith 2001) informs this paper. Smith, acknowledging the spread to northeastern Honshu (fig. 1) during a third phase difficulty identifying this middle ground, calls it “low-level that resulted in a production system dominated by dry crops. food production” (Smith 2001). For the Middle Holocene in The fourth step resulted in the ancestors of the Ainu (the Japan, the term “resource production” may be more appro- priate because, as will be seen later in this paper, food was Gary W. Crawford is Professor in the Department of Anthropology, not the only resource produced, and distinguishing “low” University of Toronto Mississauga (3359 Mississauga Road, from higher levels of production, particularly in Japan, is Mississauga, Ontario L5L 1C6, Canada [[email protected]]). difficult if it involves, for example, woodland management This paper was submitted 13 XI 09, accepted 1 XII 10, and (Crawford 2008). Nevertheless, this refocusing removes con- electronically published 2 VIII 11. straints on the discourse. The focus is on a broad range of

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0012$10.00. DOI: 10.1086/658369 S332 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 1. Map of sites in Japan and South Korea mentioned in text. human activities that may include cultivation, other man- domesticated plants and animals, but it does not rely on agement techniques, and anthropogenesis. Domestication is domestication as the key signal that a form of agriculture is an important issue, and it may help distinguish agriculture, present. In Zeder’s words, I embrace the ambiguity and reject but one cannot assume that phenotypic characteristics asso- pigeonholing the Jomon (in particular). This paper is not ciated with domestication will be visible in the archaeobo- concerned with whether the Jomon were hunter-gatherers or tanical record (see, e.g., Smith 2005). Some domesticated agriculturalists. plants (e.g., barnyard millet Echinochloa esculenta) probably Humans are early-succession organisms. That is, humans originated in Japan, so domestication is important to inves- flourish in nonmature or disrupted ecosystems. Early stages tigate there. Adherence to a dogmatic categorical reliance on of succession in terrestrial systems have, among other char- physical markers of domestication may mask the presence of acteristics, high production and reproduction rates and or- management, cultivation, or even domestication (see, e.g., ganisms with short life spans (annual and biennial rather than Zeder 2011). As such, this paper reviews evidence for indig- perennial; Odum 1969). Resource production (and agricul- enous domestication in Japan as well as the introduction of ture) is a specialized early-successional stage. Creation of such Crawford Early Agriculture in Japan S333 systems by humans is a form of ecological engineering, an or agriculture (Crawford 2008). Fortunately, many Japanese aspect of human adaptation that has a long history and is archaeologists feel the same way (Matsui and Kanehara 2006). not exclusively agriculture (Smith and Wishnie 2000). This A fresh evaluation is long overdue. has also been conceptualized in niche construction theory Some archaeological projects in Japan involve the regular (Laland, Odling-Smee, and Feldman 2001; Smith 2007). Some use of flotation, while most still do not. The bulk of the forms of production have been conceptualized as “agroeco- recently reported plant-specific subsistence-ecological data are systems” (Deur and Turner 2005). Evidence for specific forms from wet sites such as Shimoyakebe, Awazu, and Matsugatake of anthropogenesis need to be explored in the archaeological (fig. 1). Most reports of small numbers of exotic domesticates record because they may pertain to agriculture in its most from Jomon sites are not reliable (Nishimoto 2007), so the general sense. At the very least, Japan provides an enriching development of a mainland style agricultural system during comparative database for understanding the nature of agri- the Jomon as a result of external influences is not a viable culture, not just its origins. Japan may also be the locale to hypothesis. Nevertheless, evidence indicates that the Japanese investigate the domestication of several significant economic archipelago was not particularly isolated. Pottery from both plants. the Korean Chulmun and Chinese Neolithic has been re- ported from a few Jomon contexts. Significant interaction between the Korean Peninsula and Kyushu is evidenced at Jomon sites such as Tongsamdong in Pusan, South Korea (Sample 1974). Peach Prunus persica (not native to Japan) pits (not The existence of Jomon agriculture is still debated (Crawford AMS dated) from the Ikiriki site on the coast of Kyushu 2008; Matsui and Kanehara 2006). The lack of a clear outcome suggest early contact with the mainland (Minaki et al. 1986). of the debate is linked to theoretical stances, ethnographic I am not aware of any other reports of peach from the Jomon analogies, a tightly constrained definition of agriculture, and period. popular conceptions of the past and its role in defining Jap- The plant that is consistently recovered from northeastern anese culture (Crawford 2008; see also Mizoguchi 2002). I Japanese sites by flotation is the annual disturbance-affiliated have also argued that these conceptualizations unnecessarily barnyard grass Echinochloa crus-galli (Crawford 1983, 1997; constrain Jomon research (Crawford 2008). Concerns have D’Andrea 1995b; Yoshizaki 1997; H. Yamaguchi, unpublished overemphasized the classification of the Jomon to the extent report). The oldest site from which caryopses of these plants that many scholars reference “Jomon hunter-gatherers.” In have been recovered dates to roughly 9000–8700 cal BP at the Japanese-language literature, the Jomon is usually referred the Nakano B locality at Hakodate Airport (Crawford 1983; to as the “Jomon culture” (Jomon bunka) or the “Jomon Yoshizaki 1996). So far, populations of Echinochloa caryopses period” (Jomon jidai). The emphasis on the name “Jomon have been recovered in Hokkaido from 16 Jomon sites (every hunter-gatherers” began in the 1980s when the concept of Jomon site from which flotation samples have been obtained), affluent foragers seemed to provide a better characterization 17 Satsumon sites, and 15 Medieval and Ainu sites. On Hon- of the Jomon subsistence economy (Aikens, Ames, and Sanger shu (Tohoku), they are reported from two Jomon sites, one 1986; Koyama and Thomas 1981). This conceptualization was Yayoi site, and 11 sites dating from the ninth century and based on analogies with a narrow range of cultures in western later (H. Yamaguchi, unpublished report). It is also reported North America as well as with the Ainu of northern Japan. from wet contexts at Shimoyakebe in Tokyo. In one study of Agriculture to many Japanese archaeologists is synonymous collections from Jomon sites in the Kameda Peninsula, we with “wet rice agriculture.” The view that agriculture in Japan were able to document that the seeds increased in size by is fundamentally wet rice production is a reductionist argu- about 20% over several millennia, indicating that selection ment that ignores the fact that rice was only one of a wide was taking place (Crawford 1983, 1987). One specimen was range of crops there. The hunter-gatherer conceptualization recovered from residue adhering to the inside of the base of is also rooted in outdated and partially romanticized narra- a Middle Jomon pot that was set into the floor of a house at tives of the ancient human condition in Japan. The Ainu have Usujiri B. The specimen is a fruit with the caryopsis and become part of the folklore of this ancient condition. The portions of the pallea, lemma, and rachis. It is indistinguish- Ainu, however, have a complex history, with agriculture being able from the cultigen Echinochloa esculenta (fig. 2). Clusters part of that history (Crawford 2008). The major ethnography of about 100 seeds are reported from single contexts at the by Watanabe (Watanabe 1968, 1972) cited by many archae- late Early Jomon Hamanasuno and Middle Jomon Usujiri B ologists has a partially romantic foundation. Unfortunately, sites, while smaller clusters are reported from the Yagi site the Ainu still play a role in hunter-gatherer studies and are (early Early Jomon). Other Panicoid grasses have been re- one of the three ethnographic examples of hunter-gatherers covered from the Kameda Peninsula sites. They represent a from Northeast Asia cited by Binford (2001). I do not argue minimum of three types that range from Digitaria-like to that the Jomon had food production or agriculture in the Setaria sp. Hordeae (possibly Agropyron or Elymus). A Dig- way that the Chinese Middle Neolithic did, but the Jomon itaria-type grass cluster numbered nearly 900 specimens at should not be isolated from discussions of food production Yagi. S334 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 2. Ventral (L) and lateral (R) views of Echinochloa caryopses. Top row, Usujiri B site Echinochloa specimen; bottom row, Echinochloa escu- lenta (domesticated barnyard millet) reference. Portions of the rachis and husk are still attached.

Jomon people in the Kameda Peninsula were also exploiting two genera are quite similar but can be distinguished on the herbaceous annuals such as chenopods (Chenopodium sp.), basis of pericarp structure (Yoshikawa and Ito 2004). A sig- dock (Rumex sp.), knotweeds (Polygonum sp.), and other an- nificant number of Rhus/Toxicodendron seeds have been re- thropogenic plants including biennials (e.g., Rubus) and early covered from Jomon sites in Minamikaybe. A preliminary successional species such as silvervine (Actinidia), elderberry examination of a few specimens indicates that at least two (Sambucus), grape (Vitis), and lacquer tree Toxicodendron ver- species are in the Kameda Peninsula samples. The cross sec- niciflua and trees such as walnut (Juglans ailanthifolia)and tion of the pericarp of some conforms to the structure of T. Amur cork tree Phelodendron amurense (Crawford 1983, 1997; verniciflua (fig. 3). Botanists believe that the lacquer tree was Crawford, Hurley, and Yoshizaki 1976). introduced to Japan from China based on its growth in forests The presence of the lacquer tree (T. verniciflua) raises ques- that are not considered to be “natural,” in other words, early tions. The species has been considered an introduction from successional or secondary forests. However, it grows almost China. However, the oldest lacquer products in the world are exclusively in secondary forests in China, too (Suzuki, Yone- documented in the Kameda Peninsula of Hokkaido. Lacquer- kura, and Noshiro 2007). This species is likely indigenous to painted pottery has been AMS dated to approximately 9000 Japan, China, and probably Korea as well. Methods to dis- cal BP at Kakinoshima B (Minamikayabe Buried Cultural tinguish wood charcoal and pollen of the lacquer tree have Properties Research Team 2002). Lacquer is commonly as- been developed, and both lines of data indicate that it was sociated with Jomon burials, but it is not limited to such this species that grew around sites and was utilized (Suzuki, contexts. The lacquer tree is classified as Toxicodendron (not Yonekura, and Noshiro 2007). Both wood charcoal and pollen Rhus) verniciflua because recent genetic studies indicate a clear have been identified from Early Jomon contexts and later. separation of the two genera. Only one species of Rhus (Rhus Tools for extracting lacquer as well as a pot with remnants javanica), following this reclassification, is present in Japan of lacquer collection have been recovered from the Shimo- (Suzuki, Yonekura, and Noshiro 2007:59). The seeds of the yakebe site. Not only was the resin used for the production Crawford Early Agriculture in Japan S335

Figure 3. Toxicodendron seed from the Usujiri Shougakko site illustrating the pericarp-cross-section characteristic of Toxicodendron verniciflua. of lacquer, but the wood was used as a construction material. and Momohara 2007). Hemp is also reported in substantial The domesticated status of this tree is not clear. At present quantities (clusters of charred, fused seeds) from the Late– in China, people collect sap from both tended and untended Final Jomon occupation dating to ca. 3400 cal BP at Shi- trees. Apparently untended trees produce more sap (Suzuki, moyakebe (Sasaki, Kudo, and Momohara 2007). Japanese yam Yonekura, and Noshiro 2007:60). Dioscorea japonica, foxnut (Euryale), and prickly ash (Zan- Just over 100 km south of the Kameda Peninsula in the thoxylum sp.) are also evidenced at the Matsugasaki site (Mat- city of Aomori is the Sannai-Maruyama site. Bottle gourd sui and Kanehara 2006). The oldest reports of bottle gourd Lagenaria siceraria and burdock Arctium lappa (a cultigen (AMS dated to 11,700–10,900 cal BP), Perilla, beans (Vigna form, gobo, is apparently native to Japan) are reported (Habu sp.?), burdock, and goosefoot (Chenopodium) date to the Ini- 2004). Many beans (Fabaceae) have been recovered from the tial Jomon at the Awazu site on L. Biwa (Tsuboi 1994). Seeds Early Jomon component of Sannai-Maruyama (ca. 6000 BP), “similar to” foxtail millet as well as barnyard millet are added but their specific identification based on seed morphology has to the list by the Middle Jomon at Awazu (Matsui and Ka- been difficult to ascertain. Ancient DNA research has so far nehara 2006:264). Quite possibly this is reference to the wild identified only Glycine sp. (soybean) in the assemblage (Sa- forms: green foxtail and barnyard grass. Bottle gourd and kamoto et al. 2006). The DNA is so far similar to both wild Perilla are reported from the Middle Jomon at Shimoyakebe. (Glycine soja; revised taxonomy: Glycine max subsp. formo- Other plant remains from the site include leek/onion (Allium sana) and domesticated soybean (G. max; Sakamoto et al. sp.), elderberry, and mulberry Morus australis (Sasaki, Kudo, 2006:3). Barnyard grass phytoliths have been identified in the and Momohara 2007). Legumes, totaling about 80 specimens, clay matrix of pottery. Other plant remains similar to those are reported from most periods at Shimoyakebe, where the recovered from the Kameda Jomon sites (herbaceous plants oldest association is with the 5300–4800 cal BP walnut mid- such as sedges [Cyperaceae] and knotweeds, shrubs such as den. The original report distinguishes two types of Vigna elderberry and bramble [Rubus], and vines such as grape [Vitis based on size (Sasaki, Kudo, and Momohara 2007:41). The sp.]) and silvervine indicate the presence of anthropogenic larger seeds, ranging from about 5.8–8.3 mm long by 4.0–5.8 habitats similar to those in and around the Kameda Peninsula mm wide, have been reclassified as Glycine sp. A sample of Jomon occupations. these Glycine is AMS dated to about 5000 cal BP (table 1; A number of potential domesticates/tended plants have Kudo and Sasaki 2010). The date is consistent with AMS dates been recovered from Torihama, Fukui prefecture. The reports on walnut shell and a bean from the same level (table 1). The cite bottle gourd, hemp Cannabis sativa, beefsteak plant measurements are significantly larger than those of wild soy- (Perilla), paper mulberry Broussonetia papyrifera, burdock, bean, suggesting that selection for larger soybeans was taking and mustard family (Brassicaceae) as well as weedy annuals place by the Middle Jomon. The relationship of soybean seed and shrub fruits and nuts (Matsui and Kanehara 2006; Mo- size to its domestication is a complex issue (Lee et al. 2011). rikawa and Hashimoto 1994:87). Perilla is reported from sev- The Middle Jomon Shimoyakebe soybean is considered do- eral other Jomon sites such as Matsugasaki (ca. 5800 BP; mesticated based on its size. Japanese researchers have also Matsui and Kanehara 2006) and Shimoyakebe (Sasaki, Kudo, been examining impressions on pottery recovered mainly in S336 Current Anthropology Volume 52, Supplement 4, October 2011

Table 1. AMS dates from the Middle Jomon Shimoyakebe site

Lab no. Material Date (BP) Calibrated BP (1j)a (P p .63); 5140–5100 (.17); 5082–5040 (.21) 5160–5280 45 ע MTC05383 Walnut shell 4475 (21.) 5220–5270 ;(18.) 5180–5220 ;(58.) 4970–5065 40 ע 201266b Walnut shell 4440 (21.) 5220–5270 ;(18.) 5180–5220 ;(58.) 4970–5065 40 ע 201265b Walnut shell 4380 (19.) 5260–5300 ;(73.) 5060–5190 45 ע MTC05837 Adzuki or soybean 4515 (36.) 4930–4960 ;(60.) 4850–4890 30 ע PLD-9088 Soybean 4335 a Calibrated using Calib 6.0. b Beta Analytic; MTC, Tokyo University; PLD, Paleo Labo, Japan.

Kyushu. Flotation has not been conducted, so attention has 3.7% to 10.1%, the higher rates being associated with the turned to the scanning electron micrography examination of Late–Final Jomon (Temple 2006). These rates are not con- seed impressions on pottery using peels or casts. About 20 sistent with agricultural dependence. Temple suggests that impressions have been identified as soybean. Five are im- root and tuber consumption were the cariogenic foods re- pressions of the beans themselves, while 15 examples are soy- sponsible. The role of yams and burduck may be underesti- bean hilar areas. The oldest is from the Middle Jomon Sa- mated in the Jomon diet. kanomiba site. The beans are significantly larger (over 10 mm One of the more important Jomon resources is the wild long) than wild soybean (Obata, Sasaki, and Senba 2007). boar or pig (Sus scrofa). They were exploited so extensively Obata, Sasaki, and Senba (2007) hypothesize that there was during the Early Jomon that a mutualistic relationship may an independent domestication of one land race of soybean well have existed at the time (Anezaki 2007:306). The overall in Japan by at least 4,000 years ago. Genetic diversity among size of pigs appears to have increased from the Early to the 120 cultivars indicates that the Japanese and Korean soybeans Middle Jomon and subsequently decreased. By AD 400, their are distinct from those of China (Li and Nelson 2001:1346). size increased substantially. Middle Jomon pig burials have Interpretation of this pattern has been informed by an as- been found, and some of these burials are newborn pigs ac- sumption that soybean was domesticated once and in China, companying human infants, suggesting that there was a com- but this assumption no longer appears viable (Lee et al. 2011). plex relationship between pigs and people at the time (Anezaki More than 100 examples of beans have been recovered from 2007:306). Furthermore, pigs were moved to islands as early archaeological sites in Japan. Primarily two types are found: as 9000 BP, with the farthest offshore transport by Jomon adzuki or red bean (Vigna), and soybean (Glycine). Wild spe- people being about 200 km (to Hachijo Island; Kobayashi, cies of both genera are distributed widely in East Asia. Wild Kaner, and Nakamura 2004:88). Age and sex distributions are adzuki bean (Vigna angularis subsp. nipponensis) is found not those of a managed population; furthermore, populations throughout East Asia, and forms intermediate between wild of domesticated pigs in East Asia do not evidence any genetic and domesticated are generally confined to the Korean Pen- contribution from the Japanese wild boar (Larson et al. 2010: insula and Japan (Yano, Yasuda, and Yamaguchi 2004:S136). 7687). Yayoi, Satsumon, and later historic beans have been confirmed The orthodox view that the Jomon developed and sustained to be adzuki (V. angularis) through DNA studies (Yano, Ya- itself for millennia relatively passively in a naturally rich en- suda, and Yamaguchi 2004). The same study unsuccessfully vironment is an oversimplification if not incorrect (Crawford attempted to confirm the suspicion that the Sannai-Maru- 2008; Kobayashi, Kaner, and Nakamura 2004; Nishida 1981, yama beans were also Vigna, but subsequent research has 1983). Humans as ecological engineers and niche constructors shown that they are Glycine (Sakamoto et al. 2006). Adzuki, generally do not inform discourse on the longevity of the on the other hand, is not easily distinguished from the mung Jomon. Furthermore, critical assessments of what “natural” bean. Nevertheless, because adzuki is native to Japan, ar- means for Japan during the Holocene are lacking (Crawford chaeobotanists refer to these specimens as “adzuki type.” 2008). Jomon populations were quite large in some areas, Yoshizaki and Tsubakisaka (2001) outline a method to de- particularly northeastern Japan, where 80% of Jomon sites termine the difference between the two, but in order to do are located (Koyama 1992). In all likelihood, Jomon popu- so one requires specimens in which the embryo is well pre- lations were ecologically dominant in this region, and their served. Well-preserved specimens have been classified as ad- impact on both local and regional scales was probably sig- zuki. The oldest examples are from contexts dating to about nificant. If domestication as well as resource production arose 8000 cal BP. According to Obata, Sasaki, and Senba (2007), during the Jomon, then Holocene human ecology of the Jap- the number of adzuki reports increases through the Late Jo- anese archipelago becomes relevant to the discourse on early mon. agriculture. This appears to be the case. The extent to which Jomon populations utilized root foods/ Plant remains from Jomon sites in the Kameda Peninsula crops such as yams and burdock (gobo) is an open question. evidence a wide variety of annual herbaceous small-seeded Dental caries among selected Jomon populations range from plants as well as a number of perennial shrubs and vines. Crawford Early Agriculture in Japan S337

These plants would have thrived in and around Jomon com- have been conducted to test the idea that Jomon people man- munities. Dwellings during the Early Jomon were semisub- aged nut trees (Kitagawa and Yasuda 2004, 2008). Compar- terranean houses often more than 1 m deep and 10 m in isons of expected pollen percentages of the species with the diameter. Initial Jomon dwellings such as those at Nakano B pollen recovered from cores near archaeological sites has re- were not as large, but they were certainly numerous. The fill sulted in a number of examples where pollen of chestnuts excavated during construction of these structures would have and horse chestnuts far exceeds expectations, particularly be- provided rich habitats for plants as would have the abandoned cause both are insect pollinated. Wood remains from Jomon dwellings. These occupations are normally quite large and sites indicate a marked preference for chestnut as a construc- lasted for centuries and in some cases millennia. At other tion material (Noshiro and Sasaki 2007; Noshiro, Suzuki, and sites, such as Kakinoshima A situated on a coastal terrace of Yamada 1992; Sasaki and Noshiro 2004). The general con- the Kameda Peninsula, the land was leveled, and the extracted clusion is that in some places and times, chestnuts and horse earth was heaped around the coastal side of the terrace, re- chestnuts were indeed managed. Nut-tree management was sulting in a three-sided embankment about 450 m long and not simply for the purpose of increasing the quantity of nuts 35 m wide with an elevation of about 3 m (Minamikayabe but to provide construction material as well. Burial Cultural Properties Research Team 2003). This is prob- ably not unique. Some Late Jomon populations in Hokkaido also invested considerable labor in the creation of cemeteries Yayoi surrounded by earthworks. Plant remains distinguish at least four episodes in the During the first millennium BP, significant socioeconomic Kameda Peninsula Jomon record (Crawford 1983). The Initial changes began to sweep southwestern Japan. This period, the Jomon is characterized by low densities of herbaceous annual Yayoi, sees the development of clearly ranked society, met- plants, while shrubs and nuts predominate. The Early through allurgy, and significant interaction with regions outside Japan. Late Jomon is characterized by annual plants as well as tree This is also the period when intensive agriculture, usually (but not nuts) and shrub fruit. This period was the setting described as wet rice production, became established in Japan. in which barnyard grass appears to have gone through a se- The agricultural system was, of course, not strictly based on lection process that appears to be domestication. Nut remains wet rice production. A broad suite of plants including barley, become slightly more common later in this sequence. Chest- wheat, millet, and other dry crops was also a significant com- nut (Castanea) makes its first Kameda Peninsula appearance ponent of the new agricultural system. When and under what in the Late Jomon at the Seizan site. This tree appears to have circumstances this significant change developed is actively be- been introduced to southwestern Hokkaido by Jomon people ing investigated. In the last few years the chronology has (Yamada and Shibauchi 1997). significantly changed (Shoda 2007). The Yayoi chronology was Chestnut, horse chestnut (Aesculus), and walnut consump- originally based (uncritically) on cross-dating with the main- tion is a theme that runs through Jomon archaeology (e.g., land. A project that radiocarbon-dated residue on pottery has Kobayashi, Kaner, and Nakamura 2004). This is not the case established that the chronology was in serious need of revi- for Hokkaido, but it is for other parts of Japan. Extensive use sion. Instead of beginning about 2400 BP and comprising of nuts is considered to have been so prevalent that nut-tree Early, Middle, and Late phases, the Yayoi began ca. 2800 BP management is generally accepted. Nishida (1983) pointed (Harunaru and Imamura 2004). It begins now with an Initial out that these tree species along with other plants flourish in period. This is significant because the Yayoi involves an in- open sunlit areas. He also pointed out that plant management trusion from outside. In terms of food production, there is and plant use in the contemporary village of Mukasa beyond little difference between what was transpiring on the mainland what we traditionally consider agriculture is extensive. Some 2800 cal BP as opposed to 2400 cal BP. Nevertheless, the new nuts were harvested from the only relatively undisturbed for- chronology is an exceptional development in Yayoi archae- est in the area, but the closer to the village center, the more ology. intensive plant management and plant use becomes (Nishida Although the Yayoi culture has many characteristics of cul- 1983). In the spaces among houses and gardens are chestnut, tures that are its contemporaries in China and Korea, the persimmon Diosporus kaki, apricot Prunus armeniaca,fig(Fi- Yayoi became uniquely Japanese. Once the Yayoi was estab- cus), prickly ash (Zanthoxylum), Perilla, and ginger (Zingiber), lished in the Japanese archipelago, it developed along its own and a number of other plants were established (Nishida 1983: historical trajectory. Its history involved considerable inter- 309). Some of these plants were purposefully planted while action with the indigenous Jomon cultures (Kobayashi 2001). others were not. Nishida noticed that people selectively cleared Although 80% of the Jomon population resided in north- and retained certain plants. He also proposed that chestnuts eastern Japan, the Yayoi still encountered significant Jomon and walnuts, being sun-loving plants, were managed in groves populations. Questions surround the nature of the relation- during the Jomon and that Jomon plant use likely follow the ship between Yayoi and Jomon peoples. Both material culture pattern similar to the one he observed in and around Mukasa. studies and genetics indicate that the Jomon were not strictly Since Nishida wrote his paper, substantial pollen analyses replaced. The development of distinct Yayoi pottery com- S338 Current Anthropology Volume 52, Supplement 4, October 2011 plexes was not a simple matter and involved interaction with tions do not suggest increasing anthropogenic influences. This neighboring Jomon peoples, and the Yayoi was a fusion of pattern contrasts significantly from exploitation and settle- cultures (Kobayashi 2001). Genetic studies indicate that Jo- ment patterns of the preceding Jomon and the later Satsumon. mon genes survive “at high frequencies” in modern Japanese populations (Hammer et al. 2006). Since the 1980s, a substantial number of archaeobotanical Satsumon and Ainu studies have been conducted on sites dating to roughly 3000– 700 cal BP in northern Tohoku and Hokkaido. These sites The subsequent Satsumon culture represents a significant belong to five cultures: Tohoku Yayoi, Epi-Jomon (Zoku Jo- technological, subsistence, and settlement change (Yokoyama mon), Satsumon, Okhotsk, and Ainu. To briefly summarize, 1984; Yoshizaki 1984). We have hypothesized that the Sat- the emergence of the Tohoku Yayoi was a result of a complex sumon culture arose from the Tohoku Yayoi culture that was set of processes involving the interaction of Late and Final interacting with cultures of southwestern Japan. Significant Jomon cultures with the Yayoi cultures of southwestern Japan sociopolitical developments in southwestern Japan resulted in (Crawford and Takamiya 1990). The Tohoku Yayoi in turn the formation of the first Japanese state. As the state devel- interacted with their contemporaries in northernmost Hon- oped, so did its interests in northeastern Japan. Through a shu as well as in Hokkaido. The Epi-Jomon developed in this complex process of warfare, exchange, colonization, and the context. While the Tohoku Yayoi culture established an in- establishment of political and economic affiliations, socio- tensified dry farming system, its contemporaries in Hokkaido political and cultural developments in the northeast were not did not develop such a system until much later. Epi-Jomon static. The circumstances in the northeast were ultimately so sites are, with a few rare exceptions, a mixture of burials and hostile that the Emishi (the name applied to all non-Japanese short-term occupations (Crawford and Takamiya 1990; No- in the region at the time) left Honshu for Hokkaido. This mura and Utagawa 2003). These types of settlements are not event marks the beginning of the Hokkaido Satsumon period. known for the preceding Jomon. The short-term occupations To what extent the Satsumon merged with or eliminated the are evidenced by artifact, bone, and plant-remains concen- Epi-Jomon is not known. No one questions that the Satsumon trations (Crawford 1987; Crawford and Takamiya 1990; Sap- is ancestral to the Ainu. Many archaeologists still prefer a poro-shi Kyoiku Iinkai 1987). Often these concentrations are unilineal model, with the Jomon leading directly to the Ainu. in shallow, burned deposits. Epi-Jomon pit houses are oc- Elsewhere I have reviewed the arguments against this model casionally reported but are rare. For the first time an obvious (Crawford 2008; Crawford and Takamiya 1990); however, record of sacred art flourishes in Hokkaido. Such art is con- most archaeologists agree that the underlying factor that dis- centrated in two cave sites. Ritual offerings such as pottery tinguishes the Satsumon from its predecessors on Hokkaido sherds placed on deer scapulae have also been reported. The is food production. explanation for the Final Jomon to change in so many ways Although sporadic and fortuitous reports of crop remains has yet to be examined in any detail, but it likely relates to had been reported over the years from Hokkaido Satsumon circumstances of its contacts with the Tohoku Yayoi. Under sites, it was not until systematic flotation at the Sakushu- such circumstances, disease, intergroup hostilities, revitali- Kotoni River site on the Hokkaido University campus that zation movement(s), among others, may be in play. crop remains were clearly associated with day-to-day activities Plant remains from Epi-Jomon sites primarily represent of this culture (Crawford 1987; Sapporo-shi Kyoiku Iinkai nuts, shrub and vine fruits, and anthropogenic annual plants 1987; Yoshizaki 1990). We recovered tens of thousands of (Crawford 1987; D’Andrea 1992, 1995a). A few cultigens such charred specimens of crops, weeds, and other plants. The as rice are reported, but they are not found in significant crops at the site included barley, wheat, foxtail millet, broom- quantities. These plants were likely exchanged for other prod- corn millet, melon Cucumis melo, flax Linum usitatissimum, ucts. Not only were crops exchanged but Yayoi pottery and Perilla, soybean, Japanese red bean, and hemp. glass beads are a minor component of Epi-Jomon assem- The twelfth century in Hokkaido saw significant techno- blages. Epi-Jomon pottery is found on Tohoku Yayoi sites. logical changes. Indigenous technology is replaced to a large Extensive flotation at K135 and Mochiyazawa indicate that extent by technology from the southwest. Significantly, local Epi-Jomon resource use is characterized by significant vari- manufacturing of pottery ceases. The technology of Hokkaido ation and is distinct from the preceding Jomon periods comes to look very much like that of the rest of Japan. Other (Crawford 1987; D’Andrea 1995a; Yoshizaki 1990). Plant re- aspects of the material culture, including agriculture, re- mains from K135 suggest short-term processing events. Large mained unchanged. This new period of intensified influence concentrations of single species of nuts indicate that they were from the rest of Japan is what separates the Satsumon from processed and consumed over a short period of time. Similar the Ainu. Plant remains have been recovered by flotation from patterns are noted for annual weedy plants such as knotweeds. eight Ainu sites. The collections are primarily rice, but the At the moment there is no clear evidence of local food pro- general assumption is that the rice was obtained by trade. duction during the Epi-Jomon. At least two occupations are Other crops, including barnyard millet, were grown. Other- separated by alluvial deposits. The earlier and later occupa- wise, the crop assemblage is similar to that of the preceding Crawford Early Agriculture in Japan S339 Satsumon (Yamada 2002). Iron agricultural tools as well as Japan and China ridged fields have been found in the archaeological record dating to the seventeenth century AD (Yamada 1999). The Jomon, although exhibiting considerable regional and chronological variability, represents relatively stable and sim- ilar adaptations with few if any rivals at this degree of com- Okhotsk Culture plexity and longevity during the Holocene. This is particularly evident when compared with North China and the Yangzi The Okhotsk culture is commonly characterized as a maritime basin where cultures evolved after the Upper Paleolithic from culture that migrated from Sakhalin to the north coast of low-level food production emphasizing millets or rice as well Hokkaido ca. 1500–1400 cal BP. DNA research points to their as other resources to intensive farming that supported cen- close relationships to people of the lower Amur and Sakhalin tralized polities and urban centers by 5000 to 4000 cal BP as well as to the Ainu (Sato et al. 2007). The Okhotsk culture (Lee et al. 2007). The divergence of Japanese and Chinese disappeared a little over 1,000 years ago, presumably absorbed subsistence-economic systems after the Upper Paleolithic is by the Satsumon culture. Flotation samples have been col- the orthodox view; Chinese populations embarked on one or lected from five sites, and it is now clear that they brought more agricultural trajectories while Korean and Japanese pop- crops and domesticated pigs with them. Crops include barley, ulations settled into hunting-and-gathering modes of varying foxtail, and broomcorn millet. Weedy plants include cheno- affluence and complexity that never became centralized, hi- pod, silvervine, grape, and elderberry; some nuts (walnut) are erarchical, or heterarchical (see Pearson 2007). One problem usually also part of the plant assemblage (Yamada and Tsu- is that flotation and the recovery of other plant evidence bakisaka 1995). The earliest Okhotsk barley is distinct from (when the plant remains are not obvious, such as at wet sites) the type grown by the Satsumon, being wider and thicker. in Japan (with some notable exceptions) is favored only when Crops eventually grown by the Ainu likely included varieties investigating questions related to agriculture. In contrast, flo- tation has been rapidly adopted in China over the last decade introduced directly from the mainland, not only from the and has become standard practice there for the most part. At south. sites in Japan presumed to be hunter-gatherer occupations, plant-related data are not a high priority, so plant-related Central Ryukyu Islands: Okinawa issues are normally not empirically investigated, thereby re- inforcing preconceptions; that is, comparisons are hampered Agriculture was a relatively late and sudden development in by the dearth of research explicitly designed to answer sub- the central Ryukyu Islands that lie between Kyushu and Tai- sistence-ecological questions. As a result, this discussion is wan (Takamiya 2001, 2005). Significant research has been more in the realm of hypothesis building rather than testing accomplished in this region since 1985. Flotation samples and is meant to be exploratory. Despite the limitations of the from sites dating from the Jomon to the beginning of the data, table 2 compares the earliest evidence of settlement and Gusuku period indicate that only wild foods were collected. technological innovations as well as ecological indicators of The plant remains are mainly nuts and seeds of fruits such anthropogenesis/ecological engineering/niche construction in Japan and China. Table 2 is not meant to be comprehensive as grape and silvervine (Takamiya 2005). Results of flotation but provides key or best-available examples of the evidence. at the Yayoi-Heian period Nazakibaru site have clarified that The Upper Paleolithic throughout Northeast Asia is far less between the eighth and tenth centuries AD, wheat, barley, regionally variable than the subsequent Early and Middle Ho- rice, and foxtail millet were the principal crops and that weeds locene cultures. Cultures dating from the Late Glacial Max- such as knotweeds and sedges indicate that the crops were imum to 13,500–10,000 cal BP in Japan and parts of China grown locally; potential fields have been identified (Takamiya and the Russian Far East developed the oldest pottery vessels 2005). However, Takamiya suggests that because a rapid pop- in the world, beginning 17,800–16,000 cal BP (Boaretto et al. ulation increase is not associated with the Yayoi-Heian period, 2009; Kuzmin 2006; see also Cohen 2011) in contexts that this initial experiment with crops failed. Not until the tenth are otherwise Paleolithic. By 13,500 cal BP, differences in de- to the twelfth centuries AD (Gusuku period) did agriculture velopmental trajectories arise, but they are not what one might become permanently established. Apparently the shift to food anticipate if populations in China were developing material production at the beginning of the Gusuku period was rapid, and ecological precursors to the Neolithic (in the Western likely the result of a migration from Kyushu (Takamiya 2005). sense that includes food production), and those in Japan were Preliminary DNA studies indicate that the modern Ryukyu not. Houses, villages/hamlets, and nut exploitation are all pos- population originated in Kyushu (Shinoda 2007). Further- sibly two to three millennia earlier in Japan than in China more, the Japanese dialects of the Ryukyu Islands are believed (table 1). Grinding technology and evidence of storage may to have separated from the Japanese dialect of Kyushu some not be much later in China, although this depends on the time in the middle of the first millennium AD (Hokama interpretation and dating of remains from Nanzhuangtou (see 1977). Cohen 2011). Grass (millets and rice) exploitation and pro- ? China Yangzi Valley wet sites; smalllongwa and amounts acorn of at charredChen, Yuezhuang, walnut all and at ca. Wang Xing- 2006; 8100–7700 Shelach cal 2000; BP Zong (Crawford, et7500 al. cal 2007) BP), Kuahuqiaojiang ca. Provincial 7700–7600 Institute cal BP, ofMuseum etc. Cultural 2004) (Yan Relics 1991; Zhe- and Archaeology Xiaoshan Wang 2006; Zhao 2005) tute of Archaeology 2006); rare in later sites, all are wild phenotype BP (e.g., Xinglonggou, Yuezhuang) and Liu 2006; Lu et al. 2009) (Crawford, Chen, and Wang 2006; Jiang and Liu12,700–11,000 2006; cal Lu BP et (Cohen al. 2009) 2011) at Kuahuqiao ca. 7700–7600 cal BP (Shelach 2000; Zong et al. 2007) Houses cluster ca. 8000 cal BP but possibly 2 millennia earlier (Jiang 10,500–10,200 cal BP (Lu et al. 2009); ash pits at Nanzhuangtou, Contrast between North and South China; large quantities of acorn in Good evidence from Pengtoushan/Bashidang 9500–8600 cal BP (or to Pengtoushan/Bashidang ca. 9500–8600 cal BP (Hunan Provincial Insti- By 8100–7700 cal BP at Cishan, Xinglonggou, and Yuezhuang Certainly by ca. 8000–7700 cal BP and probably by 10,500–9000 cal BP Ditches at Xinglonggou ca. 7700 cal BP; possible wetland management 11,200–8700 cal BP (Jiang and Liu 2006); Nanzhuangtou (Cohen 2011) Yuezhuang ca. 8000–7700 cal BP (Crawford, Chen, and Wang 2006) 17,800–17,000 cal BP (Boaretto et al. 2009) Japan Jomon (Crawford 1983) chida (see Habu 2004:64) Middle Jomon (Crawford 1983);status; 6700–6300 Matsui cal and BP Kanehara at 2006) Awazu (unclear AMS-dated domesticate at Usujiriwith B, AMS-dated Hokkaido; rice cultigen at1995; associated Kazahari Matsui (Crawford and 1983; Kanehara D’Andrea 2006) et al. for detailed summary see Pearson 2006) are Late Jomon Middle Jomon (Crawford 1983) Large quantity in single pit 13,400–13,100 cal BP at Higashi-Kurostu- Initial and Early Jomon (Crawford 1983; MatsuiEarly and Jomon Kanehara (Crawford 2006) 1983) Initial Jomon Xinglonggou, (Nakano Yuezhuang B, ca. Hokkaido); 8100–7700 potentially cal domesticated BP by (Crawford, Chen, and Genus in Early Jomon Hokkaido flotation sample and Awazu; non- Rare; domesticate at Late Jomon, Kazahari (D’Andrea et al. 1995) Cishan millet ca. 10,500–10,000 cal BP; common at sites after 8000 cal Pit houses 14,000–12,800 cal BP 12,800–9500 cal BP in Kyushu (Uenohara 4) and13,400–13,100 Hokkaido cal (Nakano BP B; in Kyushu (Habu 2004:64) Kanjodori earthworks at Kiusu, Kakinoshima-A site preparation; these 13,500 BP (see Pearson 2006) 17,000–16,000 cal BP (Boaretto et al. 2009; Kuzmin 2006) Rumex Chenopodium Polygonum Echinochloa Setaria Panicum construction) Nut collection Fleshy fruits (from vines, shrubs, or trees) Small quantities recovered in InitialAnnual Jomon, plants: extensive use by Early Table 2. Japan-China earliest evidence comparison Pottery Houses Hamlets/villages Storage pits Landscape modification (beyond village Metates/querns

S340 ...... and J. Wang, unpublished manuscript) China, Institute for PlantMuseum Research, 1984) and the Gansu Provincial 2011) if not earlier; chickenZhou also 1981) reported from Cishan (Barton et al. 2009; Gansu Province 5600–4600 cal BP (Teacher’s College of Northwest Present in Late Neolithic Hemudu Variety of Paniceae usually associated with millets By 8000–7700 cal BP; small through Chinese Neolithic 8000 cal BP (probably domesticated); wild (?) at Nanzhuangtou (Cohen Nanzhuangtou (Cohen 2011); by 8000–7200 BP at Cishan and Dadiwan Hemudu (7000–6000 cal BP) at Digitaria wood, and seeds throughout Jomon Yagi [early Early Jomon]; Crawford 1983) Sakanomiba and Shimoyakebe Quantities of at least two species (Hordeae and probable 11,700–10,900 cal BP at Awazu (Matsui and Kanehara 2006) Awazu 6700–6300 cal BP (Matsui and Kanehara 2006); Torihama Houli Period, Yuezhuang siteLacquer (8000 at cal Kakinoshima-B BP; (Initial G. Jomon) W. Crawford, ca. X. 9000 Chen, cal BP; lacquer, Early Jomon (Awazu ) 6700–6300 cal BP (Matsui and Kanehara 2006) Sannai-Maruyama (ca. 6000 cal BP); large by MiddleInitial Jomon Jomon at (Awazu; Matsui and Kanehara 2006) Early Jomon (Matsugasaki; Matsui and Kanehara 2006) Boar transported offshore ca. 9000 BP Initial Jomon (e.g., Natsushima site; see Imamura 1996:61) Torihama (Initial–Early Jomon), Shimoyakebe (Late Jomon) Perilla Bottle gourd Hemp Lacquer tree Burdock Other grasses (Poaceae) Glycine Vigna Dioscorea Pig or wild boar Dog Roots/tubers: Animals:

S341 S342 Current Anthropology Volume 52, Supplement 4, October 2011 duction becomes prevalent in both North and South China early in China. Another significant difference may be the ex- between 10,400 and 8000 cal BP but not in Japan. Nut ex- tent to which ecological succession was engineered in Jomon ploitation in China before 8000 cal BP is not yet clearly doc- Japan to produce nuts and lacquer rather than cereals. The umented. Nut remains are reported in low densities from former represents a young successional stage while cereal pro- Early Neolithic sites in the north (e.g., Xinglonggou and Yue- duction represents an initial stage. zhuang) and in much higher densities in the Yangzi Valley I have tried to make it clear that the distinctions between (e.g., Kuahuqiao). The history of nut use in China is probably Jomon and Chinese Neolithic subsistence developments are much longer than these data suggest, but these reports are significantly different from the orthodox view. Plants were all we have for now. With only 1,000 years separating the being cultivated, and productive habitats such as nut-tree or- earliest evidence of Late Pleistocene–Early Holocene settle- chards appear to have been created. Domestication appears ment changes in China and Japan, cultures in both areas were to have occurred at least by the Middle Jomon. Criteria for beginning to develop along similar trajectories, although de- assessing the domesticated status of plants such as Perilla have tails of the timing suggest that some changes were occurring not been established, so much work remains. However, a earlier in Japan (houses, grinding technology, storage pits, Chinese-style agriculture is relatively late in Japan and results and nut use) while others may be earlier in China (pottery). from extensive interaction with the mainland. These devel- Houses in the form of hamlets or villages do not simply opments occurred during the Yayoi, Satsumon, Okhotsk, and represent increased sedentism. They also represent a form of Gusuku periods. Environment change after the Late Glacial habitat modification. Significant earthmoving is later in Japan Maximum and concomitant rearrangement of vegetation, (Late and Final Jomon) than in China (Early Neolithic). In large mammal extinctions, and changes in animal species dis- China, ditches around and within communities are known tributions likely influenced Early Holocene developments, but during the Early Neolithic. These would produce specific the precise nature of the relationship to subsistence and set- forms of anthropogenic habitat modification. How these hab- tlement developments have not been adequately examined. itats may relate to food production is not known. The extent We know little (empirically) about Upper Paleolithic subsis- to which anthropogenic habitats were exploited in this initial tence in Japan or China. The level of production and inten- stage is not clear either. If the Cishan millet data are an in- sification during the Jomon compared with China is difficult dication, grass exploitation and possibly cultivation predate to assess given the eventual focus on grass seeds in China and 10,400 cal BP given the level of development shortly thereafter. not in Japan, although the Jomon did not ignore grasses. A low-density representation of annual plants including at The parallels in subsistence-settlement systems and tech- least one grass, the wild ancestor of barnyard millet, is evident nology, at least from a qualitative perspective, in Early Ho- during the Initial Jomon in Hokkaido. Grass exploitation is locene China and Japan raise an important question: why did potentially present at a low level in Early Holocene Japan in the cultural trajectories differ so substantially by 8000–6000 contrast to at least some locations in China. Not until the cal BP? In particular, why did resource use in China rapidly Early Jomon do anthropogenic plants become commonly ex- involve many domestication events and the subsequently sig- ploited. Elsewhere in Japan are the earliest records of hemp, nificant economic contribution of these resources and agrar- bottle gourd, Perilla, and the lacquer tree as well as lacquer ian lifeways while the Jomon did not, instead developing a production. In fact, they are represented in the archaeological somewhat agrarian lifestyle based on a few domesticates that record much earlier than in China. Chenopodium on the other apparently did not significantly displace other resources? I do hand is present in the Chinese Early Neolithic as well as in not intend to answer the question here, but I hope that raising the Initial and Early Jomon. Legumes were being exploited it will stimulate discussion. during the Initial Jomon, too. Vigna and Glycine have been Part of the problem involves understanding that people recovered from several sites. The latter is well represented in were not only adjusting to late and post-Pleistocene environ- the archaeological record of China and Korea, but for now, mental changes but that they were also actively changing the Vigna is rare in the Chinese Neolithic record. Domesticated environment in which they were living; that is, they were soybean, adzuki, and barnyard millet may be present in the ecological engineers who made significant anthropogenic im- Middle Jomon along with a variety of managed/cultivated pact on their habitats. A food crisis or population pressure resources. The plant correlates of anthropogenesis, cultiva- on resources is not apparent in either region. If anything, tion, and selection leading to domestication are evident in packing would be more influential in Japan than in China both China and Japan by 7000 cal BP, but the level of pro- given that sea-level changes cut the archipelago from the duction appears to be much higher in China by 8000–7000 mainland, land area was reduced, and habitable territory is cal BP. However, the diversity of cultivated plants appears to much smaller in area and more constricted than in China. be higher in Japan where the focus is not on grain production, So packing does not appear to be a useful way to approach although one grain seems to have emerged as a cultigen by agricultural origins and domestication in this area, much as ca. 4500–4000 cal BP. Pigs, too, were important to the Jomon a case has yet to be made for packing being causal in eastern just as they were in China, but they appear not to be do- North America (Smith 2011). The earliest millet-producing mesticated in Jomon Japan, whereas they were domesticated sites in China are found along the ecotone between the loess Crawford Early Agriculture in Japan S343 zone where grasses and other annuals such as Chenopodium References Cited thrived and forested river valleys (Chen 2004). Such an eco- Aikens, C. M., Kenneth M. Ames, and David Sanger. 1986. Affluent tone is minimally evident in Japan and absent at the beginning collectors at the edges of Eurasia and North America: some com- of the Holocene. Nevertheless, a different balance of managed parisons and observations on the evolution of society among resources, crops not strictly for food, food crops, and fished, north-temperate coastal hunter-gatherers. In Prehistoric hunter- hunted, and gathered resources proved to be sustainable for gatherers in Japan: new research methods. T. Akazawa and C. M. thousands of years. Aikens, eds. Pp. 3–26. Bulletin 27. Tokyo: University Museum, University of Tokyo. Anezaki, T. 2007. Pig exploitation in the southern Kanto region, Japan. International Journal of Osteoarchaeology 17:299–308. Discussion Barton, L., S. D. Newsome, F.-H. Chen, H. Wang, T. P. Guilderson, and R. L. Bettinger. 2009. Agricultural origins and the isotopic The six episodes outlined in this paper document a complexity identity of domestication in northern China. Proceedings of the seen in few other places. Developments during the Jomon National Academy of Sciences of the USA 106:5523–5528. have parallels to eastern North America (Smith 2011), Papua Binford, L. R. 2001. Constructing frames of reference: an analytical method for archaeological theory building using hunter-gatherer and New Guinea (Denham 2011), and even China (Cohen 2011; environmental data sets. Berkeley: University of California Press. Zhao 2011). Post-Jomon developments that involve diffusion Boaretto, E., X. Wu, J. Yuan, O. Bar-Yosef, V. Chu, Y. Pan, K. Liu, and migration (Yayoi, Tohoku Yayoi, Satsumon, Okhotsk, Gu- et al. 2009. Radiocarbon dating of charcoal and bone collagen suku) also have parallels with parts of North America (in- associated with early pottery at Yuchanyan Cave, Hunan Province, cluding the northwest coast) and with Europe (Rowley- China. Proceedings of the National Academy of Sciences of the USA 106:9595–9600. Conwy 2011) and Oceania as well. The Jomon is the most Chen, S. 2004. Adaptive changes of prehistoric hunter-gatherers dur- ambiguous of the six episodes, with evidence of niche con- ing the Pleistocene-Holocene transition in China. PhD disserta- struction/anthropogenesis that included domestication, cul- tion, Dedman College, Southern Methodist University. tivation, and tree management and potentially wild-pig man- Cohen, David Joel. 2011. The beginnings of agriculture in China: a agement. Late Upper Paleolithic and Early Holocene multiregional view. Current Anthropology 52(suppl. 4):S273–S293. Crawford, G. 2008. The Jomon in early agriculture discourse: issues developments parallel, or are identical to, some aspects of arising from Matsui, Kanehara, and Pearson. World Archaeology these developments in China at the same time. Yet the Jomon 40:445–465. did not evolve along the same trajectory of domestication and Crawford, G. W. 1983. Paleoethnobotany of the Kameda Peninsula intensification of agroecology as did China. Jomon subsistence Jomon. Anthropology Papers, no. 73. Ann Arbor: Museum of An- adaptations were extremely successful, lasting more than thropology, University of Michigan. ———. 1987. Plant seeds excavated from the K135 site. In The K135 10,000 years. Viewed in these terms, early post-Pleistocene site. Sapporo-shi Kyoiku Iinkai, ed. Pp. 565–581. Sapporo, Japan: adaptations in China could be considered maladaptive ini- Sapporo-shi Kyoiku Iinkai. [In Japanese.] tially, and they led to domestication and agrarian lifeways ———. 1992. The transitions to agriculture in Japan. In Transitions being adaptive over the same 10 millennia. The differences to agriculture in prehistory. A. B. Gebauer and T. D. Price, eds. Pp. would be easily, if not overly simplistically, explained envi- 117–132. Monographs in World Archaeology, no. 4. Madison, WI: Prehistory. ronmentally, Japan having significantly less habitat variation ———. 1997. Anthropogenesis in prehistoric northeastern Japan. In with no xeric to mesic ecotones, potentially fewer domestic- People, plants, and landscapes: studies in paleoethnobotany.K. able resources, and differences in seasonality and associated Gremillion, ed. Pp. 86–103. Tuscaloosa: University of Alabama risks. However, environmental differences are not particularly Press. satisfying explanations at this point. Agricultural-origins re- Crawford, G. W., X. Chen, and J. Wang. 2006. Houli culture rice from the Yuezhuang site, Jinan. East Asia Archaeology 3:247–251. search in China is still relatively poorly documented, so com- [In Chinese.] paring China and Japan in any substantial way is premature. Crawford, G. W., W. H. Hurley, and M. Yoshizaki. 1976. Implications Cultures in Japan did not have an agricultural system resem- of plant remains from the Early Jomon, Hamanasuno site. Asian bling that of parts of China until the development of the Perspectives 19:145–155. Yayoi that had both wet and dry agriculture. The similarities Crawford, G. W., and H. Takamiya. 1990. The origins and impli- cations of late prehistoric plant husbandry in northern Japan. mainly are due to the ancestry of these cropping systems in Antiquity 64:889–911. mainland East Asia. By Satsumon, Okhotsk, and Gusuku D’Andrea, A. C. 1992. Palaeoethnobotany of Later Jomon and Yayoi times, locally suited systems of relatively productive agricul- cultures of northeastern Aomori and southwestern Hokkaido. PhD ture had developed. Each of these systems—included hunting, dissertation, University of Toronto. gathering, and/or fishing—and determining the contribution ———. 1995a. Archaeobotanical evidence for Zoku-Jomon subsis- tence at the Mochiyazawa site, Hokkaido, Japan. Journal of Ar- agriculture made to society (nutrition, trade, symbolism, etc.) chaeological Science 22:583–595. are difficult to assess. The Jomon will remain a special prob- ———. 1995b. Later Jomon subsistence in northeastern Japan: new lem until more research is directed to understanding the roles evidence from palaeoethnobotanical studies. Asian Perspectives 34: of legumes, grasses, and other small-seeded plants in local 195–227. economies and to the chronological and spatial variation in D’Andrea, A. C., G. W. Crawford, M. Yoshizaki, and T. Kudo. 1995. Late Jomon cultigens in northeastern Japan. Antiquity 69:146–152. tree (nut, other fruit, and lacquer) management. Denham, Tim. 2011. Early agriculture and plant domestication in S344 Current Anthropology Volume 52, Supplement 4, October 2011

New Guinea and Island Southeast Asia. Current Anthropology accessions from three countries measured by RAPDs. Crop Science 52(suppl. 4):S379–S395. 41:1337. Deur, D., and N. J. Turner. 2005. Introduction: reassessing indigenous Lu, H., J. Zhang, K.-B. Liu, N. Wu, Y. Li, K. Zhou, M. Ye, et al. 2009. resource management, reassessing the history of an idea. In Keep- Earliest domestication of common millet (Panicum miliaceum)in ing it living: traditions of plant use and cultivation on the northwest East Asia extended to 10,000 years ago. Proceedings of the National coast of North America. D. Deur and N. J. Turner, eds. Pp. 3–34. Academy of Sciences of the USA 106:7367–7372. Vancouver, British Columbia: UBC Press. Matsui, A., and M. Kanehara. 2006. The question of prehistoric plant Habu, J. 2004. Ancient Jomon of Japan. Case Studies in Early Societies. husbandry during the Jomon period in Japan. World Archaeology Cambridge: Cambridge University Press. 38:259–273. Hammer, M. F., T. M. Karafet, H. Park, K. Omoto, S. Harihara, M. Minaki, M., S. Hohjo, S. Kokawa, S. Kosugi, and M. Susuki. 1986. Stoneking, and S. Horai. 2006. Dual origins of the Japanese: com- In Ikiriki Iseki. Tarami-cho Kyoiku Iinkai, eds. Pp. 44–53. Tarami- mon ground for hunter-gatherer and farmer Y chromosomes. Jour- cho Kyoiku Iinkai, Japan: Nagasaki-ken Nishisonogi-gun Tarami- nal of Human Genetics 51:47–58. cho. [In Japanese.] Harunaru, H., and M. Imamura. 2004. Yayoi Jidai no Jitsu Nendai Minamikayabe Buried Cultural Properties Research Team. 2002. The [The real dates for the Yayoi period]. Tokyo: Gakuseisha. [In Jap- Kakinoshima B site. Minamikayabe, Japan: Minamikayabe Buried anese.] Cultural Properties Research Team. [In Japanese.] Hokama, S. 1977. Okinawa language and its history. Iwanami Koza ———. 2003. The Kakinoshima A site. Minamikayabe, Japan: Nagato 11 Hogen. Tokyo: Iwanami Shoten. [In Japanese.] Shuppansha Insatsubu. [In Japanese.] Hunan Provincial Institute of Archaeology. 2006. Pengtoushan and Mizoguchi, K. J. 2002. An archaeological history of Japan: 30,000 BC Bashidang II. Beijing: Science. [In Chinese.] to AD 700. Philadelphia: University of Pennsylvania Press. Imamura, K. 1996. Prehistoric Japan: new perspectives on insular East Morikawa, M., and S. Hashimoto. 1994. The Torihama shellmound: Asia. Honolulu: University of Hawaii Press. a Jomon time capsule. Tokyo: Yomiuri Shimbunsha. [In Japanese.] Jiang, L., and L. Liu. 2006. New evidence for the origins of sedentism Nishida, M. 1981. Human-plant relationships in the Jomon period and rice domestication in the lower Yangzi River, China. Antiquity and the emergence of food production. Bulletin of the National 80:355–361. Museum of Ethnology 6:234–255. [In Japanese.] Kitagawa, J., and Y. Yasuda. 2004. The influence of climatic change ———. 1983. The emergence of food production in neolithic Japan. on chestnut and horse chestnut preservation around Jomon sites Journal of Anthropological Archaeology 2:305–322. in northeastern Japan with special reference to the Sannai- Nishimoto, T. 2007. From the Jomon period to the Yayoi period. Tokyo: Maruyama and Kamegaoka sites. Quaternary International 123– Yuzankaku-shuppan. [In Japanese.] 125:89–103. Nomura, T., and H. Utagawa. 2003. The Epi-Jomon and Okhotsk ———. 2008. Development and distribution of Castanea and Aes- cultures. Sapporo, Japan: Hokkaido Shinbunsha. [In Japanese.] culus culture during the Jomon period in Japan. Quaternary In- Noshiro, S., and Y. Sasaki. 2007. Use of timber resources at the ternational 184:41–55. Shimo-yakebe site, Tokyo, and its characteristics in the Kanto plain Kobayashi, S. 2001. Eastern Japanese pottery during the Jomon-Yayoi of the late to latest Jomon periods. Japanese Journal of Historical transition: a study in forager-farmer interaction. Bulletin of the Botany 15:19–34. Indo-Pacific Prehistory Association 21:37–42. Noshiro, S., M. Suzuki, and M. Yamada. 1992. Species selection for Kobayashi, T., S. Kaner, and O. Nakamura. 2004. Jomon reflections. wooden artefacts by prehistoric and early historic people in the Oxford: Oxbow. Kanto Plain, central Japan. Journal of Archaeological Science 19: Koyama, S. 1992. Prehistoric Japanese populations: a subsistence- 429–443. demographic approach. In Japanese as a member of the Asian and Obata, H., Y. Sasaki, and Y. Senba. 2007. Impressions on pottery Pacific populations. K. Hanihara, eds. Pp. 187–197. Kyoto, Japan: revealed cultivation of Glycine max subsp. max (soybean) in the International Research Center for Japanese Studies. late to latest Jomon periods in Kyushu Island. Japanese Journal of Koyama, S., and D. H. Thomas, eds. 1981. Affluent foragers. Senri Historical Botany 15:97–114 Ethnological Studies, no. 9. Senri, Osaka, Japan: National Museum Odum, E. P. 1969. The strategy of ecosystem development. Science of Ethnology. 164:262–270. Kudo, Y., and Y. Sasaki. 2010. Characterization of plant remains on Pearson, R. 2006. Jomon hot spot: increasing sedentism in south- Jomon pottery excavated from the Shimoyakebe site, Tokyo, Japan. western Japan in the Incipient Jomon. World Archaeology 38:239– Bulletin of National Museum of Japanese History 158:1–26. [In 258. Japanese.] ———. 2007. Debating Jomon social complexity. Asian Perspectives Kuzmin, Y. V. 2006. Chronology of the earliest pottery in East Asia: 46:361–388. progress and pitfalls. Antiquity 80:362–371. Rowley-Conwy, Peter. Westward ho! the spread of agriculture from Laland, K. N., J. Odling-Smee, and M. W. Feldman. 2001. Cultural central Europe to the Atlantic. Current Anthropology 52(suppl. 4): niche construction and human evolution. Journal of Evolutionary S431–S451. Biology 14:22–33. Sakamoto, S., R. Ishikawa, I. Nakamura, Y. Sato, and Y. Shimamoto. Larson, G., R. Liu, X. Zhao, J. Yuan, D. Fuller, L. Barton, K. Dobney, 2006. Species identification of 6,000-years-old beans from Sannai- et al. 2010. Patterns of East Asian pig domestication, migration, Maruyama site, Aomori, Japan. Journal of Fossil Research 39:1–6. and turnover revealed by modern and ancient DNA. Proceedings Sample, T. 1974. Tongsamdong: a contribution to Korea Neolithic of the National Academy of Sciences of the USA 107:7686–7691. culture history. Arctic Anthropology 11:1–125. Lee, G.-A., G. W. Crawford, L. Liu, and X. Chen. 2007. Plants and Sapporo-shi Kyoiku Iinkai. 1987. The K135 Site. Sapporo-shi, Japan: people from the Early Neolithic to Shang periods in North China. Sapporo-shi Kyoiku Iinkai, Sapporo-shi Bunkazai Chosa Hoko- Proceedings of the National Academy of Sciences of the USA 104: kusho. [In Japanese.] 1087–1092. Sasaki, Y., Y. Kudo, and A. Momohara. 2007. Utilization of plant Lee, G.-A., G. W. Crawford, L. Liu, Y. Sasaki, and X. Chen. 2011. resources reconstructed from plant macrofossils during the latter Archaeological soybean (Glycine max) in East Asia: does size mat- half of the Jomon period at the Shimo-yakebe site, Tokyo. Japanese ter? PLoS One. Forthcoming. Journal of Historical Botany 15:35–50. Li, Z., and R. L. Nelson. 2001. Genetic diversity among soybean Sasaki, Y., and S. Noshiro. 2004. Utilization of forest resources in Crawford Early Agriculture in Japan S345

the Late Jomon Period deduced from wooden remains for water to Ainu cultures. Hokkaido Historical Museum Research Reports usage at the Shimo-yakebe Site, Tokyo, Japan. Japanese Journal of 10610405:37–43. [In Japanese.] Historical Botany 12:37–46. Yamada, G., and S. Shibauchi. 1997. Nuts excavated from Hokkaido Sato, T., T. Amano, H. Ono, H. Ishida, H. Kodera, H. Matsumura, Jomon sites: chestnuts. Bulletin of the Historical Museum of Hok- M. Yoneda, and R. Masuda. 2007. Origins and genetic features of kaido 25:17–30. [In Japanese.] the Okhotsk people, revealed by ancient mitochondrial DNA anal- Yamada, G., and Y. Tsubakisaka. 1995. The spread of cultivated crops ysis. Journal of Human Genetics 52:618–627. from the mainland. In Final Report on Research Project of the Shelach, G. 2000. The earliest Neolithic cultures of northeast China: Historical and Cultural Exchange of the North. Pp. 107–134. Hok- recent discoveries and new perspectives on the beginning of ag- kaido Historical Museum. [In Japanese.] riculture. Journal of World Prehistory 14:363–413. Yan, W. 1991. China’s earliest rice agricultural remains. Indo-Pacific Shinoda, K. 2007. The ancestors who became Japanese. Tokyo: NHK. Prehistory Association Bulletin 10:118–126. [In Japanese.] Yano, A., K. Yasuda, and H. Yamaguchi. 2004. A test for molecular Shoda, S. 2007. A comment on the Yayoi period dating controversy. identification of Japanese archaeological beans and phylogenetic Bulletin of the Society for East Asian Archaeology 1:1–7. relationship of wild and cultivated species of subgenus Ceratotropis Smith, Bruce D. 2001. Low-level food production. Journal of Ar- (genus Vigna, Papilionaceae) using sequence variation in two non- chaeological Research 9:1–43. coding regions of the Trnl and Trnf genes. Economic Botany ———. 2005. Low-level food production on the northwest coast. In 58(suppl.):S135–S146. Keeping it living: traditions of plant use and cultivation on the north- Yokoyama, E. 1984. Hajiki before the use of rokuro in Hokkaido: west coast of North America. D. Deur and N. J. Turner, eds. Pp. the formation of the Early Satsumon. Koukogaku Zasshi 70:52–75. 37–66. Vancouver, British Columbia: UBC Press. [In Japanese.] ———. 2007. Niche construction and the behavioral context of plant Yoshikawa, M., and Y. Ito. 2004. Large plant fossils from the Iwatari- and animal domestication. Evolutionary Anthropology 16:188–199. Shoukoku 4 site, Aomori City. In Itawari-Kotoni 4 site. Aomori ———. 2011. The cultural context of plant domestication in eastern Buried Cultural Relics Research Team, ed. Pp. 293–319. Aomori, North America. Current Anthropology 52(suppl. 4):S471–S484. Japan: Aomori Board of Education. [In Japanese.] Smith, E. A., and M. Wishnie. 2000. Conservation and subsistence Yoshizaki, M. 1984. Some problems in Satsumon culture research. in small-scale societies. Annual Review of Anthropology 29:493– Hokkaido Kokogaku 20:65–74. [In Japanese.] 524. ———. 1990. Plant seeds from the K135 site, 4 chome locality. In Suzuki, M., K. Yonekura, and S. Noshiro. 2007. Distribution and K135 site, 4th st. locality. Sapporo Buried Cultural Relics Research habitat of Toxicodendron vernicifluum (Stokes) F. A. Barkl. (Ana- Team, ed. Pp. 56–59, 111–112. Sapporo, Japan: Sapporoshi Bun- cardiaceae) in China. Japanese Journal of Historical Botany 15:58– kazai Chosa Hokokusho. [In Japanese.] 62. ———. 1996. Seeds of the genus Echinochloa discovered at the Na- Takamiya, H. 2001. The transition from foragers to farmers on the kano B site. In The Nakano B site (III), Hakodate City, vol. 2. island of Okinawa. In Melaka papers, vol. 5. Special issue, Indo- Hokkaido Maizou Bunkazai Centre, ed. Pp. 617–621. Sapporo, Pacific Prehistory Association Bulletin 21:60–67. Japan: Zaidanhoujin Hokkaido Maizo Bunkazai Sentaa. [In Jap- ———. 2005. Island prehistory. Naha, Okinawa, Japan: Borderink. anese.] [In Japanese.] Teacher’s College of Northwest China, Institute for Plant Research ———. 1997. The problem of Japanese millet in the Jomon period. and the Gansu Provincial Museum. 1984. Hemp and millet from International Study on the Origins of Japanese Peoples and Cultures the Majiayao culture site of Linjia at Dongxiang, Gansu. Kaogu 7: Newsletter 2:5–6. [In Japanese.] 654–655, 663. [In Chinese.] Yoshizaki, M., and Y. Tsubakisaka. 2001. Prehistoric legumes: an ar- Temple, D. H. 2006. Bioarchaeological evidence for nutritional var- chaeobotanical perspective. Mamerui Jihou [Legume Newsletter] iation among prehistoric Jomon foragers. American Journal of 24:1–9. [In Japanese.] Physical Anthropology 129(suppl. 42):175–176. Zeder, Melinda A. 2011. The origins of agriculture in the Near East. Tsuboi, K. 1994. Jomon lakes: exploring the Awazu Shellmound, Lake Current Anthropology 52(suppl. 4):S221–S235. Biwa. Tokyo: Yuzankaku. [In Japanese.] Zhao, Z. 2005. Discussion of the Xinglonggou site flotation results Vrydaghs, L., and T. Denham. 2007. Rethinking agriculture: intro- and the origin of dry farming in northern China. Antiquities of ductory thoughts. In Rethinking agriculture: archaeological and eth- Eastern Asia A:188–199. [In Chinese.] noarchaeological perspectives. T. Denham, J. Iriarte, and L. Vry- ———. 2011. New archaeobotanic data for the study of the origins daghs, eds. Pp. 1–15. Walnut Creek, CA: Left Coast. of agriculture in China. Current Anthropology 52(suppl. 4):S295– Watanabe, H. 1968. Subsistence and ecology of northern food gath- S306. erers with special reference to the Ainu. In Man the hunter. R. Lee Zhejiang Provincial Institute of Cultural Relics and Archaeology and I. DeVore, eds. Pp. 60–77. Chicago: Aldine. Xiaoshan Museum. 2004. Kuahuqiao. Beijing: Cultural Relics. [In ———. 1972. The Ainu ecosystem. American Ethnographic Society Chinese.] Monograph no. 54. Seattle: University of Washington Press. Zhou, B. 1981. Animal remains discovered at Cishan village, Wu’,an Yamada, G. 1999. Crops excavated from sites of the Ainu culture Hebei Province. Kaogu Xuebao 3:339–347. [In Chinese.] period in medieval and recent times. Bulletin of the Historical Mu- Zong, Y., Z. Chen, J. B. Innes, C. Chen, Z. Wang, and H. Wang. seum of Hokkaido 27:9–18. [In Japanese.] 2007. Fire and flood management of coastal swamp enabled first ———. 2002. The continuity of cereal farming from the Satsumon rice paddy cultivation in east China. Nature 449:459–462.

Current Anthropology Volume 52, Supplement 4, October 2011 S347

Finding Plant Domestication in the Indian Subcontinent

by Dorian Q Fuller

CAϩ Online-Only Material: Supplement A

Recent research indicates that cultivation may have begun in as many as five regions of India before the introduction of exogenous crops and cultivation systems: South India, Orissa, the Middle Ganges, Saurashtra, and the Himalayan foothills of the Punjab region. These potential centers of crop origin have been triangulated from data on the biogeography of wild progenitors and a growing archaeo- botanical database. Nevertheless, none of these centers provide unambiguous evidence for local domestication or evidence that domestication occurred entirely in the absence of introduced crops and food-production systems. One of the major lacunae is archaeobotanical evidence from hunter- gatherer sites or evidence of the transition to initial cultivation. In addition, documentation of the morphological changes accompanying domestication is available for only a few species. This paper reviews the arguments for local domestication in each of these five regions, paying particular attention to data that might document domestication processes. But an alternative hypothesis for several regions also can be considered in which agriculture arose as a result of secondary domestications of local species after an initial introduction of farming from outside. On the basis of these alternative working hypotheses, research priorities are identified for resolving these issues.

Introduction: Multiple and Overlooked In the past few years there have been arguments in favor South Asian Centers of Origin of independent agricultural origins in India (Fuller 2002, 2006, 2008a; Saraswat 2005; Tewari et al. 2006, 2008). This has arisen in part from increased attention to identifying the Almost everywhere on earth, hunter-gatherers of prehistory wild progenitors for many crops that are poorly known in have at some time given way to farmers or given up their the mainstream Western agricultural literature and that are bows for ploughs. In archaeological research, however, greater underdocumented relative to crops from the Near East, West efforts have been expended on investigating the origins of Africa, or the Americas. It is only the past few years that agriculture in those few world regions that have long been Southern Neolithic sites have joined the growing body of accepted as having been likely “centers of origin” (see Harlan archaeobotanical evidence from South Asia (Fuller et al. 1971; Harris 1990). However, as research into agricultural 2004). The Middle Ganges plain has also seen rapid growth origins has tended to focus on accepted centers of origin with a shortage of research on other regions, this has meant that in new archaeobotanical information (Saraswat 2005; Tewari it is difficult to assess local origins of agriculture whether by et al. 2008). This paper reviews the archaeological and bo- introduction or local evolution. Such blanks on the map of tanical evidence for agricultural origins in six regions of South ϩ research in turn help to reinforce the sense that there were Asia (fig. 1; for additional details for each region, see CA very few widely spaced centers of origin. One of these areas online supplement A). that has long been a blank on maps of early agriculture is the The past 25 years have seen the steady increase in archaeo- ϩ Indian Subcontinent, and this paper reviews recent empirical botanical evidence in South Asia (fig. A1 in CA online research that indicates local domestication processes. supplement A), especially by flotation, which now accounts for more than half of the reported evidence. Of particular note is the recent addition of a rich database from Southern Dorian Q Fuller is Reader in Archaeobotany, Institute of Archaeology, University College London (31-34 Gordon Square, India. It should be stressed that we still lack evidence in South London WC1H 0PY, United Kingdom [[email protected]]). This Asia for most of the transitions, for initial cultivation, and paper was submitted 13 XI 09, accepted 9 XII 10, and electronically for shifts toward morphological domestication and agricul- published 21 IX 11. tural dependence. Instead, our evidence is clearest for the

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0013$10.00. DOI: 10.1086/658900 S348 Current Anthropology Volume 52, Supplement 4, October 2011 emergence of sedentary village societies that are invariably The South Deccan Center already dependent on cultivation and usually have domesti- cated crops and livestock. The South Deccan has recently emerged as a potential center of a distinctive indigenous early agriculture (Boivin et al. 2008; A Terminological Note on “Independent Fuller, Korisettar, and Venkatasubbaiah 2001; Fuller et al. 2004). The crops that recur and dominate Southern Deccan Origins” Neolithic archaeobotanical samples (fig. 2; table A1, available As the archaeology of agricultural origins draws on natural as an Excel file in the online edition; table A2) have wild and social sciences, it is important to be aware of potential ancestors to be found in the hills of scrub woodlands of the differences of emphasis in how origins are defined. Whether region (Asouti and Fuller 2008; Fuller 2006). While these are origins are “independent” or not is ambiguous without spec- likely to have come together from local domestications in the ifying whether one’s emphasis is on the genetic derivation of Southern Deccan, these origins themselves remain elusive. a cultigen or the cultural tradition of cultivation. Table 1 Archaeobotanical evidence from sites of the Southern Neo- provides a clarification by defining secondary and primary lithic consistently indicates the predominance of mungbean origins as they are used in this paper. Primary and secondary (Vigna radiata), horsegram (Macrotyloma uniflorum), and two origins of crops in a given geographic region are differentiated millets (Brachiaria ramosa and Setaria verticillata; Fuller, Kor- on the basis of whether the plant species are introduced as isettar, and Venkatasubbaiah 2001; Fuller et al. 2004). These cultigens, in which case they are secondary, or are derived crops occur on virtually all sites (that have been well sampled) from local populations of wild progenitors, in which case they and dominate samples on sites across the region (fig. 3; table are primary. Within both primary and secondary origins we A1). The millets of the Southern Neolithic, in particular, had can make further divisions in terms of the social processes been previously overlooked. In recent times, browntop millet of agricultural origins. In cultural terms we can ask whether B. ramosa is restricted to small plots of cultivation in a few local domestications might be inspired by contacts with other isolated areas of the eastern Ghats zone in Eastern Karnataka farmers or whether the impetus to begin cultivation was truly and Andhra Pradesh (Kimata, Ashok, and Seetharam 2000). pristine and an innovation in the cultural sense. Cases of I have no unambiguous evidence that S. verticillata is culti- “additive primary origins” are those that botanists have tra- vated anywhere today, although one secondary source indi- ditionally called “secondary domestications” (e.g., Vavilov cates cultivation somewhere in Tamil Nadu (Maheshwari and 1992 [1927]:151), such as poppies in the Western Mediter- Singh 1965). It is clear that a century ago, it was a wild- ranean or European rye and oats, where local wild/weedy gathered resource in Maharashtra (Gammie 1911), while the species were added to the repertoire of existing farmers. sister species Setaria pumila (syn. Setaria glauca of many au- In this paper, and for me, independence is considered first thors) was and still is cultivated (Kimata, Ashok, and See- on the botanical side of the matrix. In this sense, several parts tharam 2000; Kobayashi and Kimata 1989; Prasada Rao et al. of South Asia can be defined as centers of primary crop or- 1987). igins. In the discussion that follows, consideration of whether Unfortunately, at present our earliest systematic archaeo- the social process was pristine or inspired or whether agri- botanical samples date only from ca. 2000 BC with some culture arrived by migration or adoption will also be consid- earlier incompletely published evidence from Watgal (M. D. ered. But the inference of such social processes from the ar- Kajale, personal communication; cf. Deveraj et al. 1995) and chaeological evidence remains less certain and more debatable Budihal (Kajale and Eksambekar 1997). Thus, while we have than biological domestication. good evidence for this package once it was fully formed and The agrobiodiversity in South Asia is high by comparative established in its region, we lack a clear sequence for its origin: standards, especially in terms of cereals and grain legumes. this must become a priority of future research (table A2). The modern agriculture of India boasts 12 domesticated millet Most members of this crop package, the two millets and crop species in addition to rice, wheats (emmer, durum, bread horsegram, could have been more widely available in the drier wheat), barley, and Job’s tears; these are augmented by a dozen savannah zones of peninsular India. These are not species that pulse species. In a series of previous publications, I have sum- form extensive stands in the way that wild wheats, barley, or marized evidence for native millet and pulse domesticates of wild rice do, but rather these species are likely to have formed India and their likely geographical and ecological contexts of local dense patches in favorable microenvironments such as origin (e.g., Asouti and Fuller 2008, chap. 3; Fuller 2002, 2006; springs and slope bases for the millets and perhaps less dis- Fuller and Harvey 2006). Tables A2, A4, and A5 (in CAϩ turbed scrub patches for the wild horsegram. Thus foragers online supplement A) and tables A1, A3, and A6 (available exploiting these species may have seen advantages in concen- as Excel files in the online edition) summarize current details trating them together and expanding these stands through for the key domesticates, while the sections below provide a clearance of other plants. At present we lack clear morpho- synthetic interpretation of the evidence. See figure A1 for a logical indicators on preserved remains for domestication, and cumulative total of sites with published archaeobotanical re- thus cultivation has only been inferred based on dominance ports from 1960 to 2005. of samples, recurrence across sites, and association with Fuller Plant Domestication in India S349

Figure 1. Synthetic map of centers of distinctive early agricultural pack- ages and possible regions of origins. Dashed lines demarcate broader spread zones connected to these early centers. Dashed arrows indicate direction of diffusion of important domesticates, while solid arrows in- dicate inferred directions of Neolithic migrations. mungbean and introduced domesticates such as wheat and domestication occurred must have been toward the sides of barley. the peninsula where the dry savannahs intergrade into de- Mungbean differs from the other crops in its particular ciduous forests, and my current suggestion is the eastern region of origin, and the core region of the Southern Neolithic Ghats north of the Krishna River, along the Northeast Kar- was, and is, too dry (Asouti and Fuller 2008; Fuller and Kor- nataka/Andhra/Maharashtra borders, where V. radiata occurs isettar 2004; cf. Tomooka et al. 2003). Thus the zone in which without Vigna mungo (Asouti and Fuller 2008; Fuller and S350 Current Anthropology Volume 52, Supplement 4, October 2011

Table 1. Defining primary and secondary origins of cultivars and cultivation

Plant origins Social process Primary: cultivation started from wild progenitors local to region Pristine: cultivation initiated by hunter-gatherers only in contact with hunter-gatherers Inspired: cultivation initiated by hunter-gatherers in contact with food producers (i.e., “stimulus diffusion”) Additive: cultivation initiated by farmers possessing other crops Secondary: crops introduced, not evolved from local wild progenitors Migration: immigrant farmers carry crops and/or livestock (i.e., “demic diffusion,” moving frontier of agriculture) Adoption: crops or livestock from another area obtained by trade with minimal immigration (i.e., “cultural diffusion,” static frontier of agriculture)

Harvey 2006). It might have occurred somewhat more widely akallu, but many sites lack these species, including the lowest wild under slightly wetter mid-Holocene conditions. We levels at Hallur that similarly date back to ca. 2000 BC. In (Boivin et al. 2008; Fuller, Boivin, and Korisettar 2007) suggest addition, they would have been poor choices for staples in north of the Krishna River because Neolithic sites in the Kur- this region without significant winter rainfall, probably re- nool district are only from the later Neolithic, mainly after quiring irrigation (Fuller, Korisettar, and Venkatasubbaiah 1700 BC. This candidate region is, however, poorly explored 2001; Kajale 1988). Wheat and barley may have been adopted for the Neolithic or Mesolithic, and thus a key gap in ar- as valued choice foodstuffs (Fuller 2005). Their appearance chaeological research can be highlighted (fig. 3). correlates with new necked-jar forms that have their precur- The mungbeans that have been found from Southern Neo- sors in the northwest and might be indicative that their initial lithic sites, however, are decidedly small seeded, like modern adoption in the south involved production of new beverages, wild populations, indicating that the seed enlargement as- such as beers (Fuller 2005). Evidence for bread (roti) con- sociated with modern domesticates had not yet evolved (Ful- sumption, inferred from the presence of plates, does not occur ler 2007a; Fuller and Harvey 2006). Domestication, by which until the Iron Age. Historical linguistic evidence for agricul- is meant a presumed reduction in germination inhibition, is ture, crop, and tree names in Dravidian languages is also inferred based on mung’s occurrence in quantity in regions opposed to the secondary, migration (“Elamo-Dravidian”) such as Bellary beyond the wild-progenitor range. When a model (see Fuller 2003b, 2007a; Southworth 2005). regional data set is considered, seed-size increase in mung- In social terms the Neolithic plant domestications of South beans took place from the end of the Southern Neolithic, after India were probably inspired. Livestock—including domes- the mid-second millennium BC and through Iron Age times ticated sheep, goat, and cattle—are present on Southern Neo- (fig. 4; cf. Fuller and Harvey 2006). Evidence from the Ganges, lithic sites from the earliest periods (i.e., from 3000 to 2600 where this species was introduced, suggests a parallel but BC; Fuller, Boivin, and Korisettar 2007). None of these species somewhat earlier evolution of large-seeded forms. Even earlier are likely to represent local domestications but instead spread large-seed specimens are reported from the eastern Harappan either with moving pastoralists or by adoption into hunter- zone, suggesting that small-seed mungbean was introduced gatherer societies. The same may also be true of ceramic tech- to the Ganges from the south. The South Indian increase in nology. In the 1960s Allchin (1963:160) pointed toward pos- mungbean size, beginning from ca. 1500 BC, correlates with sible pre-Harappan Gujarati precursors for the ceramic a period of probable agricultural intensification and diversifi- tradition of South India. Indeed, recent data suggest that cation, the same period when African crops became estab- around the southern and eastern margins of the Thar Desert, lished in South India (Fuller, Boivin, and Korisettar 2007), pastoralism and ceramics began ca. 3500 BC, indicated by and there is evidence from Hallur for little millet (Panicum evidence from Loteshwar in Gujarat (Patel 2008), and by the sumatrense), a probable domesticate of Gujarat (see below; earliest levels at Balathal, which date to before 3000 BC (Misra Fuller et al. 2004). 2005), although in the latter case there is also crop agriculture Other crops were introduced into the Southern Neolithic, based on winter Harappan crops (Kajale 1996a). In the case such as wheat and barley, by 1900 BC (table A2). They are of Loteshwar, this is plausibly a mobile pastoralist-collector not widespread or dominant crops, and this suggests that they economy. Such a way of life may have been well suited to the were adopted through processes of cultural diffusion (Boivin semiarid savanna corridor that ran down the middle part of et al. 2008; Fuller 2005) rather than the immigration of north- the peninsula (fig. 5; cf. Asouti and Fuller 2008). Historical western winter cereal growers. The near absence of other linguistic data suggest that Proto-Dravidian speakers knew members of the winter crop package (pea, chickpea, and len- livestock, knew the flora of peninsular Indian savannas, knew til) that are present in the Harappan area and in Chalcolithic the adjacent deciduous woodlands, and had techniques of seed Maharashtra is also worth noting. Nevertheless, wheat and gathering and processing (Fuller 2003b, 2007a). A geograph- barley do occur at the earliest well-sampled levels at Sangan- ical center of linguistic gravity would place them on the north- Fuller Plant Domestication in India S351

Figure 2. Southern Neolithic crop package. Line drawings of horsegram, mungbean, browntop millet, and bristly foxtail panicles showing overall plants and seeds. (Seed drawings and millet panicles by the author; horse- gram and mungbean plants from Church [1886].) ern peninsula (Southworth 2005). We can propose then that Based on current dating evidence, the early settlements in there was a secondary dispersal, perhaps partly by adoption the South Deccan predate 2500 BC while those in the North but also by migration of Proto-Dravidian speakers, focused Deccan start from 2500 to 2300 BC. Where data are available on the savannah zone between the Thar Desert margins and (e.g., Utner, Piklihal, Banahalli), South Deccan ceramic sites the South Deccan between 3500 and 2500 BC. The recent include domesticated caprines and cattle, and the early phase report of primitive pre-Chalcolithic(?) pottery from Sakshal of the Deccan ashmounds (composed largely of burnt cattle Pipri, although inadequately dated, could relate to this dung at penning sites: Allchin 1963; Johansen 2004; Korisettar, (Mishra et al. 2006). This pastoralist-collector way of life can Venkatasubbaiah, and Fuller 2001) date to this period (Fuller, be suggested to have inspired the local primary domestications Boivin, and Korisettar 2007). Subsequently in both South and of indigenous pulse and millet crops on the peninsula. North Deccan, this savanna corridor becomes increasingly Figure 3. Map of sites with archaeobotanical evidence from peninsular India, indicating those with Neolithic/Chalcolithic evidence and those with Iron Age/Early Historic evidence. Indicated are putative centers of plant domestication in region A and region C (from fig. 1), both gaps in archaeological survey and sampling. 1, Kayatha; 2, Dangwada; 3, Ujjain; 4, Nagda; 5, Navdatoli; 6, Kaothe; 7, Tuljapur Garhi; 8, Kaundinyapur; 9, Kharwada; 10, Bhagimohari; 11, Bhatkuli; 12, Naikund; 13, Paunar; 14, Adam Cave; 15, Bhokardan; 16, Daimabad; 17, Nevasa; 18, Paithan; 19, Walaki; 20, Inamgaon; 21, Terr (Thair); 22, Gopalpur; 23, Golbai Sassan; 24, Kolhapur; 25, Budihal; 26, Piklihal; 27, Veerapuram; 28, Tek- kalakota; 29, Kurugodu; 30, Sanganakallu; 31, Hiregudda; 32, Hattibel- agallu; 33, Velpumudugu; 34, Sanyasula; 35, Ramapuram; 36, Singana- palle; 37, Rupanagudi; 38, Injedu; 39, Peddamudiyam; 40, Hanumantaraopeta; 41, Hallur; 42, Koppa; 43, Kunnatur; 44, Arikamedu; 45, Adichannalur; 46, Perur; 47, Kodumanal; 48, Mangudi. Fuller Plant Domestication in India S353

Figure 4. Post-Neolithic size increase in peninsular mung and urd beans (based on data from Fuller and Harvey [2006]). Averages and standard deviation of seed length per site/phase are plotted against median age for the site/phase; maximum and minimum measurements for each site are indicated as outliers. For comparison, modern domesticated and wild ranges are shown at far right, adjusted to account for decrease in length by 20%. The increase in size appears most marked between 1400 and 1100 BC, and only after this time are populations in the putative do- mesticated range. populated by sedentary farmers, especially after 2200–2000 cultivation, whereas archaeobotanical or morphological “do- BC, and the role of a climatic aridification around this time mestication” refers to the fixation of genotypes adapted to bears some consideration in the emergence of sedentism. On cultivation and dependent on humans. The protracted tran- these sedentary sites, integrated agropastoralism is evident, sition between cultivation and domestication makes it difficult including the indigenous crop package of South India but to pinpoint how many starts of cultivation there were, and also on sites of the northern peninsula, having a strong em- genetic data can offer only a bare minimum (Allaby, Fuller, phasis on Harappan/Near Eastern winter crops (wheat, barley, and Brown 2008). lentil, pea, grasspea), which must have spread by secondary Numerous data sets and studies indicate wide genetic di- adoption to the northern peninsula, perhaps from the Ahar/ vergence between modern indica and japonica cultivars (e.g., Balathal culture zone of Rajasthan. Bautista et al. 2001; Caicedo et al. 2007; Cheng et al. 2003; Garris et al. 2005; Londo et al. 2006; Takahashi, Sato, and The Ganges Center Nakamura 2008; Tang et al. 2004; Vitte et al. 2004). There is plausibly more than one origin within each of these as well On the great floodplain system of the Ganges and its tribu- as a distinct aus-type origin in the Bangladeshi region. Nev- taries, the origins of rice takes center stage. Rice is ubiquitous ertheless, completely separate Indian and Chinese origins are on all archaeobotanically sampled sites of Mesolithic, Neo- lithic, or Chalcolithic age (fig. 6; table A3). The balance of complicated by recent data indicating that a few key domes- evidence suggests a distinct origin of cultivation of ancestral tication genes, including one that controls panicle shattering indica rice in this region but that full domestication in the (sh4) and one that causes grains to be white (rc), are shared morphological sense was not completed until domesticated across indica and japonica (Kovach, Sweeney, and McCouch japonica varieties were introduced to North India from China, 2007; Sang 2009; Sweeney and McCouch 2007). In addition, allowing for hybridization (Fuller and Qin 2009; Fuller et al. one key mutation that affects plant growth—prog1, making 2010). The necessity of hybridization was first postulated by crops tall and erect as opposed to short and spreading the Sato (1996, 2002) and more recently numerous other authors stature of wild rices—is also shared (Tan et al. 2008). In other (e.g., Sang 2009; Vaughan, Lu, and Tomooka 2008). A key to words, phylogenetic evidence from across the nuclear and unraveling this complex story is to recognize that “domes- chloroplast genomes suggests separate origins of cultivation tication” as used in phylogenetic studies refers to the sub- (wild-population subsampling), but the evidence of func- sampling process of wild populations with the beginnings of tional genes indicates partly shared domestication through Figure 5. Distribution of Neolithic/Chalcolithic sites on the Indian Pen- insula showing their correlation with the “savanna corridor” (generalized from vegetation zones and rainfall data, 40–80 cm/year; see Asouti and Fuller 2008). The geographic division is indicated between a South Indian Neolithic, reliant on a package of indigenous crops, and North Deccan Chalcolithic, where introduced winter cereals were more important than indigenous millets. Selected early/representative sites are labeled. Note the lack of sites in the North/East, which may reflect a lack of systematic investigation and excavation including in the hypothesized domestication center. Fuller Plant Domestication in India S355

Figure 6. Map of northern South Asia showing representative archaeo- logical sites that track the spread of rice (after Fuller and Qin 2009) but indicating likely foci of early Gangetic cultivation. 1, Harappa; 2, Kunal; 3, Ojiyana; 4, Mahagara; 5, Koldihwa; 6, Tokwa; 7, Malhar; 8, Senuwar; 9, Lahuradewa; 10, Narhan; 11, Imlidh-Khurd; 12, Manji; 13, Charda; 14, Hulaskhera; 15, Gopalpur; 16, Golbai Sassan; 17, Sannai Tal lake (pollen core); 18, Balathal; 19, Shikarpur; 20, Surkotada; 21, Kanmer; 22, Bagasra (Gholo-daro); 23, Kuntasi; 24, Rojdi; 25, Oriyo Timbo; 26, Babor Kot; 27, Rangpur. Sites of the greater Indo-Gangetic divide region, with mixed indigenous(?)-introduced agriculture: E1, Kalibangan; E2,Ban- awali; E3, Rohira; E4, Sanghol; E5, Rupar; E6, Mahorana; E7, Balu; E8, Mitathal; E9, Daulatpur; E10, Burthana/Tigrana; E11, Laduwala. hybridization. Vaughan, Lu, and Tomooka (2008) maintain a pogon sensu stricto), in which manipulation of water levels single-origin model in which introgression from wild popu- and selection for annual high-grain-yielding plants was central lation into the crop as it spread south from China accounts (see also Cohen 2011; Zhao 2011). By contrast, the monsoon- for the genetic divergence. But this hypothesis ignores the adapted wild annual (Oryza nivara) was already high yielding archaeobotanical evidence from India, the archaeological evi- and could be readily collected from extensive wild stands dence for early contacts between northwestern South Asia (much as is postulated for wild sorghum, wheats, or barley). (not the northeast) and China, and the archaeobotanical evi- Oryza nivara probably could be effectively exploited without dence in China that increasingly suggests a slow and late agriculture but through simple management such as burning domestication process for rice (Fuller, Harvey, and Qin 2007; off competing vegetation during the dry season. Recent pal- Fuller, Qin, and Harvey 2008; Fuller et al. 2009, 2010). It is ynological studies in the Ganges plain indicate significant mi- also not clear how the “snowball model” of Vaughan and crocharcoal levels from the terminal Pleistocene warm period, colleagues (Vaughan, Lu, and Tomooka 2008) can account ca. 14,500–13,000 BP through the early and mid-Holocene, for divergence in the chloroplast genome (e.g., Chen et al. at Sannai Tal with similar evidence from the end of the Pleis- 1993; Takahashi, Sato, and Nakamura 2008; Tang et al. 2004), tocene at Lahuradewa (Singh 2005). These charcoal levels which is not transmitted in pollen. from the terminal Pleistocene are accompanied by the pres- The distinct origins of indica and japonica cultivars is as- ence of Oryza bulliform phytoliths from ca. 8000 BC, which sociated with different wild ecologies and differing early eco- show increased levels and morphological diversity (suggesting nomic systems of exploitation (Fuller and Qin 2009). Essen- mosaics with nivara, rufipogon, and perhaps other wild Oryza) tially, the process in the Lower and/or Middle Yangtze of from ca. 7000 BC (Saxena et al. 2006). Whether some of these China focused on a perennial wetland species (Oryza rufi- bulliforms represent cultivated rice (as argued by Saxena et S356 Current Anthropology Volume 52, Supplement 4, October 2011 al. 2006) or not requires further investigation, but there is no Grain size and shape evolution in Indian rice does not strong basis for linking bulliform types (from the leaves) to appear to be early or precede full domestication as postulated morphological domestication (in terms of grain size and dis- for Chinese rice (Fuller 2007a). The available morphometric persal traits; cf. Pearsall et al. 1995). evidence for Neolithic Gangetic rice (collected in Fuller et al. Charred grain finds and a direct AMS date from the site 2010) indicates that grains are on the small side, congruent of Lahuradewa put rice use and pottery back to ca. 6400 BC with morphologically wild rice and perhaps immature grains (Saraswat 2005; Saxena et al. 2006; Tewari et al. 2006). Both being present. Indeed, the measured grains from Koldihwa pottery making and rice consumption may have started even and Mahagara are consistent with those from Lahuradewa earlier, with more recent excavations and radiocarbon dates about 4,000 years earlier, all of which are smaller than later of ca. 9000–8000 BC (Tewari et al. 2008). It remains uncertain, domesticated rice grains from Iron Age or Early Historic India however, whether this rice was gathered or cultivated, and (Fuller et al. 2010). Re-sorting of samples has recently con- there is no hard morphological evidence that it is of domes- firmed that rice spikelet bases, previously overlooked, are ticated type (Fuller 2006, 2007b:399, 406; further comments present in the fine fraction of flotation samples from Ma- in CAϩ online supplement A). Tewari et al. (2006:49) argue hagara (1700–1600 BC) and that these are predominantly of that some of the grains resemble domesticated type while domesticated morphology (for earlier preliminary archaeo- others are of weedy/wild types, but the limited metrical data botanical results, see Harvey and Fuller 2005; Harvey et al. are consistent with grains from two wild species (fig. A2). 2006). Interestingly, this implies that spikelet bases were non- The recently published illustrations (Tewari et al. 2008) sug- shattering by this time, perhaps based on an introgressed gest the presence of wild immature-harvested spikelets and mutation from introduced japonica cultivars, but grains at do not show clear domesticated spikelet bases. Co-occurrence this period had not yet undergone clear size increase, which of wild taxa such as Coix and Setaria pumila (Tewari et al. occurs instead by the end of the second millennium BC into 2006) is ambiguous: these are known weeds of rice but also the Iron Age (Fuller et al. 2010). This size increase appears co-occur with wild Oryza or may be gathered as food in their to occur faster than that documented for the Lower Yangtze own right. Nevertheless, much later evidence from the region, domestication of japonica rice, although the latter began much such as Lahuradewa period 1B and Senuwar period 2 (Sar- earlier when wild-type shattering still predominated (Fuller aswat 2005), suggest that rice was a major food resource, et al. 2010). The period of grain-size increase in Indian rice probably cultivated, before crops from other regions were parallels an apparent delay in size increase in Gangetic mung- adopted. Interestingly, rice may have been the only crop in beans (Fuller and Harvey 2006) and may imply a connected the primary Neolithic of this region. Although it is suggested cause such as the advent of more intensive cultivation and that co-occurring millet S. pumila (syn. Setaria glauca auct. tillage. The available weed data from the Ganges suggest that pl.) may have been utilized (Saraswat 2004; Tewari et al. 2006), early rice cultivation was essentially dry cropping based on it is unclear whether this or later occurrences of small millets monsoon rains and seasonal flood recession but that plausible (e.g., Brachiaria ramosa from Mahagara) were actually cul- wet-field irrigated rice may have been grown by the end of tivated. Earlier cultivation of rice by ca. 3000–2600 BC is the second millennium BC and certainly by the Iron Age implied by recent finds beyond its probable wild zone at Early (Fuller and Qin 2009). Thus two key factors, the development Harappan Kunal (Saraswat and Pokharia 2003) and perhaps of irrigated cultivation techniques and the introduction of the third millennium BC in the Swat valley of Pakistan (Cos- some japonica rice genes, may in turn be linked to the evo- tantini 1987; Fuller 2006:35–36). lution of larger-grained indica rices and the more extensive Interestingly, somewhat later rice in the northwest from spread of rice (cf. Fuller and Qin 2009). Agricultural inten- Late Harappan times (after 1950 BC, e.g., at Pirak) has been sification is indicated also by the adoption of cash crops such suggested to represent japonica based on grain (Costantini as cotton and flax and may be linked to the emergence of 1979) and bulliform phytolith morphometrics (Sato 2005; cf. social hierarchy (cf. Fuller 2008b). Fujiwara 1993:157). This may imply diffusion of fully do- From as early as 2400 BC, secondary cultivars were intro- mesticated rice from East Asia via Central Asia had occurred duced including free-threshing wheat and barley (Saraswat around the start of the second millennium BC. Other evidence 2004, 2005), and these were widespread by ca. 1700 BC (Fuller for diffusion from China includes the first evidence in South 2006; Harvey 2006). Goats, sheep, and probably domestic Asia for Panicum miliaceum, Cannabis sativa, peaches and cattle also arrived in this horizon, certainly by 2000 BC. These apricots, and Longshan-style harvest knives in the Kashmir presumably came from the Indus region to the west, which Neolithic (see Fuller 2006:36; on peaches and apricots, see also shows some ceramics links. After 2000 BC, peninsular Fuller and Madella 2001:341). The diffusion of japonica rice Indian cultivars were adopted, such as mungbean and horse- at this time may be important in terms of introducing key gram (Fuller and Harvey 2006). Cash craft crops such as flax mutations (e.g., sh4, rc, prog1) into Indian rice cultivars (Fuller and cotton are found from Ganges Chalcolithic sites from ca. et al. 2010). This hybridization would have made indica rice 1400 BC (Fuller 2008b). It can be suggested that ard tillage fully domesticated and paved the way for more intensive rice was adopted in this period too. The evidence for local con- agriculture in India (Fuller and Qin 2009). tinuity within the Middle Ganges suggests that this was mainly Fuller Plant Domestication in India S357 a sequence of episodes of secondary adoption from the west eastern Andhra, such as Korisettar, Venkatasubbaiah, and Ful- (the Indo-Gangetic divide) and sometimes the south (down ler 2001, fig. 9), and as such this is most likely a case of tributaries such as the Betawa and Chambal) into a regime adoption. However, the persistence of pockets of North Dra- of primary rice cultivation. The evidence seems to favor pri- vidian– and South-Central Dravidian–speaking tribes in the mary pristine origins for rice cultivation, but dependence on hills of Jharkhand, western Orissa, and Chattisgarh may be cultivation and the emergence of sedentism occurred with the indicative of some early migration processes into parts of the secondary adoptions of additional domesticates, including region. Some apparently shared loan words between the fully domesticated rice (japonica hybrids). Interestingly, the Munda languages and Dravidian also support this (Fuller emergence of sedentism occurs between 2400 and 1800 BC, 2003b, 2007a). The earlier establishment of rice agriculture with a plausible focus on 2200–2000 BC (i.e., synchronous remains unknown, and while wild populations make primary with that in South India). domestication possible, secondary spread from the Ganges is at least equally plausible. However, as indicated in figure 4, The Greater Orissa Center the postulated center for crop origins in the upper Mahanadi River basin and the surrounding foothills remains largely The plains of the Mahanadi River and the surrounding hills unexplored: the region can boast no published excavations, constitute a plausible region for rice origins based on the although the river courses are rich in Neolithic/Chalcolithic widespread and diverse wild-progenitor populations. In the settlement mounds (Dorian Q Fuller, personal observation, early 1980s it was established by Van der Maeson (1980) that 2004; Dorian Q Fuller, personal communication from Sa- the wild progenitor of the pigeon pea—the second most im- dasiba Prabhan, Utkal University). Domestication of Cajanus portant pulse, behind chickpea, in Indian production today— is primary but presumably inspired or additive. It is not yet was native to South Orissa and the adjacent Bastar area (table clear whether this precedes the introduction of peninsular A4). However, this region remained off the map of early ag- pulses such as horsegram that could have served as a model. riculture for the lack of archaeology. Attempts to retrieve With the sparse data at present, no causal model can be pro- archaeobotanical evidence from four upland sites of central posed. and northern Orissa were unsuccessful apart from phytolith “background” noise (Harvey 2006; Harvey et al. 2006). Other The Saurashtra Center archaeobotanical sampling in 2003/2004 was successful (at Gopalpur and Golbai Sassan; table A4) and provides an ar- Archaeobotanical research on the Saurashtra Peninsula of Gu- chaeobotanical corpus for the established village societies on jarat (fig. 6; table A5 and table A6) has been mainly carried the coastal plains around Chilka Lake dating to the Late Neo- as part of excavations of sites related to the Harappan civi- lithic/Chalcolithic (i.e., 1500–1000 BC; Harvey 2006; Harvey lization (2500–2000 BC) and to the local Late/Post-Harappan et al. 2006]). cultural phase (e.g., Chanchala 1995; Kajale 1996b; Pokharia Rice appears to be the dominant find. In addition, native 2007; Reddy 2003; Weber 1991). My only firsthand experience Indian pulses are present, including the local domesticate Ca- with material is from a small flotation assemblage from recent janus cajan and introduced Macrotyloma uniflorum and both Maharaja Sayajirao University of Baroda excavations at Har- Vigna radiata and Vigna mungo (Fuller and Harvey 2006). appan Gholo-daro (Bagasra; Luddy 2008; S. Luddy and Do- The Vigna grains are still largely in the small “wild” size range rian Q Fuller, unpublished data). The case for a potential but presumably domesticated in terms of germination and primary domestication center has been made on the basis of dehiscence congruent with a dispersal from the South before the contrasts between archaeobotany in this region and that the Vigna seed-size increase of 1400–1100 BC. Millets appear of the Harappan Indus (Fuller and Madella 2001) and the to have been insignificant in this region. A couple of grains biogeographic evidence of potential wild progenitors. I have of Panicum sumatrense from Gopalpur could indicate use, or serious concerns over the reliability of some reported iden- cultivation, of this species as a cereal, but the sample size is tifications of small millet grains in previous work (which have too small to inspire confidence, and the wild progenitor is been detailed in Fuller 2002, 2003a; see also Hilu, De Wet, available in the region. These and other small millets (Echin- and Harlan 1979), and these concerns have not yet been ad- ochloa, Paspalum, Setaria sp.) are all plausible weeds of rice. dressed through adequate illustrated publication of identifi- The sum of the evidence is that agriculture and sedentism cation criteria. At issue is first, the presence and antiquity of were established by the mid-second millennium BC. These finger millet Eleusine coracana of African origin, and second, sites also appear already to have domesticated fauna (at least the presence/antiquity of Setaria italica of Chinese origin, cattle and buffalo, but sample size is very small; Kar, Basa, which is extremely difficult to separate from endemic Bra- and Joglekar 1998). This established agriculture certainly in- chiaria ramosa. In the case of finger millet, it is clear from cludes some secondary crops from the south and west and photographs that reports from some sites were misidentifi- adopted livestock. In terms of ceramics, stone-axe forms, and cations of Setaria/Brachiaria grains (see Fuller 2003a). Recent bone tools, this region is distinct from the peninsula region work on samples from Gholo-daro recovered only native In- (although probably continuous with the Neolithic of north- dian millets Panicum sumatrense (the cereal that all workers S358 Current Anthropology Volume 52, Supplement 4, October 2011 agree was important), B. ramosa, and other wild Indian Setaria the Ganges-Yamuna. Most archaeobotanical research comes spp., but no S. italica or E. coracana. from unpublished excavations, making it difficult to be con- Despite these problems, the quantitative picture of the re- fident about the cultural and chronological context of re- gion is archaeobotany dominated by small millets including ported assemblages, and these are mainly limited to presence/ P. sumatrense and native small Setaria spp. (plus at least some absence lists (table A6; fig. A3). B. ramosa) during the Mature Harappan period, 2500–2000 Samples from the Early Harappan period (in this region BC. The agriculture is thus completely different from that of from ca. 3000 to 2400 BC) are few and are limited to Lu- the Harappan-period Lower Indus and Baluchistan (see Fuller dawala, Siswal, and Kunal 1A/1B with comparison available and Madella 2001; Tengberg 1999). The winter crops that were from Harappa 1–2 (ca. 3200–2600 BC). All of these show the staples of the Harappan civilization and the Indus region are dominance of wheat and barley agriculture with Near Eastern present but extremely rare. For example, a few fragmentary winter pulses at Harappa. However, some native Indian crops wheat and barley grains could be identified at Bagasra (Luddy are present, including Vigna pulses and horsegram from Lu- 2008), while a few wheat were reported from Kuntasi, and dawala, Panicum sumatrense, another small millet (Setaria 10 grains of barley (compared with thousands of millets) were sp.), Vigna radiata from Harappa, and rice from Kunal. In found at Rojdi (Weber 1991). Lentils were also a rare find at the subsequent Mature Harappan phase, such summer crops Rojdi. Pulses in general are rare, but Vigna mungo, potentially become more common. Given the date (12400 BC), this native to the region, is present in the third millennium, while seems rather too early for most of these crops to have spread Vigna radiata and Macrotyloma are introduced in the early from peninsular India. Rice may well have become a cultivar second millennium. Introduced at the start of the second from the Middle Ganges before this time, while it is plausible millennium BC were crops of African origin, perhaps the first that the pulses and millets were already cultivated in Gujarat acceptably identified and dated in South Asia, including pearl and had spread around the eastern side of the Thar Desert. millet, sorghum, and Lablab purpureus (Fuller 2003a). However, it seems equally plausible that these indigenous The evidence is suggestive of a primary pre-Harappan ag- crops were brought independently into cultivation from wild ricultural tradition based on native monsoon-adapted crops populations in Northwest India. Additional support for a local (small millets, including P. sumatrense and V. mungo). As with derivation comes from the enlarged size of Harappan-period South India, however, there is no archaeobotanical evidence Vigna seeds from Kunal and Balu, as these appear to have for the earlier stages of this tradition or for the transition undergone size increase ahead of those elsewhere in India from collecting to farming these species (table A5). The evi- (Fuller and Harvey 2006). dence for introduced domesticated fauna with early ceramics Given the limited Early Harappan evidence and the lack of from northern Gujarat (the southern fringes of the Thar Des- Neolithic precursors, the origins in this region are obscure. ert) by ca. 3500–3000 BC, for example, at Loteshwar (Patel It seems plausible to see domestications in this region as 2008) would fit with a model in which primary local plant initially secondary, based on introduced Near Eastern crops, domestications were inspired either by the advent of intrusive after which local plants were added to the repertoire as “pri- pastoralism or by contacts with the Indus and regions to the mary additive” domesticates. west. Saurashtra may have been the first link in a chain of inspiration with the spread of pastoralism that culminated in the plant domestications of the Southern Neolithic. If there Assessing Early Near Eastern Influence: were indeed primary plant domestications in Orissa, these Some Notes on Baluchistan may also have occurred after the arrival of pastoralists from the West. Most accounts of the Neolithic of South Asia start in the northwest with the site of Mehrgarh, where I have chosen to end. This is an important and informative site (Jarrige 1982, A Center Near the Indo-Gangetic Divide? 2008; Jarrige, Jarrige, and Quivron 2006; Possehl 1999:450– 473), but it remains without good comparisons, neither local Where the plains of the Upper Ravi River and other Indus early Holocene precursors nor well-sampled contemporane- tributaries come closest to the Upper Ganges and Yamuna ous sites dated 7000–4000 BC; instead, most archaeobotanical rivers is a region that bears further consideration in terms of data come from the third millennium BC (table A6; fig. A3). crop origins. Previously I concluded that the published ar- This region clearly had early agriculture based in large part chaeobotanical record “hints at domestication of monsoonal of nonindigenous crops (glume wheat[s], bread wheat, barley millet crops that is earlier than and perhaps independent of varieties; Costantini 2008) and probably nonindigenous live- those further South on the Peninsula or in Gujarat” (Fuller stock (at least the goat; Meadow 1996). Other Near Eastern 2006:35). This Eastern Harappan zone (of “Sothi-Siswal” Har- crop domesticates also came to the Indus, although whether appan, in the terminology of Possehl 1999, 2002) has a dis- this happened at the initial dispersal or only later remains tinctive Early Harappan and Late Harappan (Baran) tradition unclear (table A6). Also, it should be noted that the barley and shows continuity with later Bronze Age developments in on South Asia probably had its origins all or at least in part Fuller Plant Domestication in India S359 from a separate domestication in the Eastern Fertile Crescent chaeobotanical recovery. What we have at present is clear (Iran), which appears to be the source of most Central, South, evidence for the nature of regional agricultural economies and East Asian barleys (Morrell and Clegg 2007; Saisho and from the period when sedentary villages emerged. Intrigu- Purugganan 2007). Nevertheless, the domesticates were not ingly, this seems to have occurred across most of India be- a single package entirely from one area of origin but were tween the later third millennium BC and the early second shuffled together along the trajectory to Baluchistan. While millennium BC (Fuller 2006). If climatic pressures are to be this region is a secondary center in terms of agricultural or- considered, it is the 2200 BC “event” and the establishment igins, it is nevertheless an important area for indigenous de- of sedentary agropastoralists that might be linked rather than velopments in terms of adding new domesticates to the econ- the beginnings of pastoralism or cultivation per se. omy: zebu cattle, quantitatively the most important domestic Only in the Middle Ganges region is there evidence for a animal for all of the previously discussed regional Neolithics pristine start of cultivation before animal domestication, (when quantified bone reports are available); cotton, certainly whereas elsewhere the spread of domestic livestock herding used for threads in the Late Neolithic/Early Chalcolithic and potting appears to have occurred among hunter-gatherer- (Moulherat et al. 2002); and probably somewhat later and fishers. In some Indian centers, cultivation began later among domesticated in the Indus alluvial plains, sesame (Fuller potting pastoralists. This was especially true around the south- 2003a) and water buffalo (Meadow and Patel 2003; Patel and ern margins of the Thar Desert and through the peninsular Meadow 1998). While all of these are primary additive do- savannah corridor. Useful comparisons can be drawn with mesticates, their importance to economic developments the Sahelian-Sudanic pastoralists of the middle Holocene in within the Indus valley, which ultimately supported the Har- Africa (see Marshall and Weissbrod 2011). The hypothesis of appan civilization, and to the spread of crops and livestock a “scheduled availability” model (sensu Marshall and Hil- to other Neolithic traditions elsewhere in India deserves em- debrand 2002) for motivating groups of pastoralist-collectors phasis. Nevertheless, the early Holocene village farming Neo- to plant and tend key species seems plausible; such a model lithic represented by Mehrgarh and its successors in the Indus may be useful for considering motivations behind the initial valley and its western borderlands is more typically Neolithic uptake of small domestic herds among grain-gathering hunt- in the Near Eastern sense but quite different from the early ers. Also, it may be that as domestic herds (and/or herders) food-producing communities to the east in the subcontinent. increased in population, territoriality emerged, which in- creased competition for or denied access to limited seasonal seed-food resources located in the forests and hills. Such hill Comparative Comments: Trajectories and forests—although rich in plant resources, including wild pro- Causation genitors of several Indian domesticates—were peripheral to the savannahs, which were attractive for herding, and moti- The accumulation of archaeobotanical data makes it clear that vation for cultivation may lie in this spatial disjunction of a number of tropical crops were domesticated in South Asia resources. These earliest stages of food production are better in prehistory, and we are now able to propose the regions considered in terms of a “primary pastoral community” with and general periods in which this occurred and whether this its comparisons to the early Nile region and Sudan (e.g., represents additive, inspired, or pristine origins of cultivation. Wengrow 2006:26–31) rather than a classic Neolithic (of the It is now clear that different trajectories into the Neolithic Levantine, European, or Chinese sort). As in this period in took place in different parts of the subcontinent, and this Egypt, occupation sites were ephemeral and characterized by suggests that different world regions may provide the most dung deposits and thin middens of ash and other waste. Such useful comparisons for assessing convergent processes in the sites have preservational problems and have been difficult to various Indian regions. What is striking is that there is no recognize until when in the Southern Deccan some of them evidence in South Asia for the emergence of sedentary, densely became monumentalized through burning, creating the Dec- populated hunter-gatherer societies comparable to the Epi- can ashmounds (see Boivin 2004; Johansen 2004; Korisettar, palaeolithic Levant (see Goring-Morris and Belfer-Cohen Venkatasubbaiah, and Fuller 2001)—which itself may relate 2011; Zeder 2011) or parts of East Asia (e.g., the early to mid- to increasing territoriality and associated social and ritual Holocene Lower Yangtze, parts of Jomon Japan; see Cohen transformations. In such early pastoral communities, herds 2011; Crawford 2011; Zhao 2011). Rather, it appears that food and especially cattle were central to systems of wealth and production began among more mobile and perhaps sparser ritual rather than land and its plant produce. Thus it need populations. This must surely call into question the relevance not be surprising that at the third millennium town of Balathal of demographic prime movers often considered for explaining in Rajasthan, the center of the settlement appears to have the beginnings of agriculture in the Near East (see, e.g., Bar- been a fortified pen with a deep stratigraphy of dung (Misra Yosef 2011; Bocquet-Appel 2011). In addition, this leaves us 2005). One challenge is to better understand the processes of with the challenge that the archaeological record is largely sedentarization and the role that local plant domestication silent before sedentism; presedentary sites have been harder may have played in this, while another even greater challenge to identify, rarely excavated, and present difficulties for ar- is to find and document the evidence for the latest hunter- S360 Current Anthropology Volume 52, Supplement 4, October 2011 gatherers and their interactions with or transformations into Deveraj, D. V., J. G. Shaffer, C. S. Patil, and Balasubramanya. 1995. early pastoralists. The Watgal excavations: an interim report. Man and Environment 20:57–74. Fujiwara, H. 1993. Research into the history of rice cultivation using plant opal analysis. In Current research in phytolith analysis: ap- Acknowledgments plications in archaeology and paleoecology. D. M. Pearsall and D. R. Piperno, eds. Pp. 147–158. MASCA Research Papers in Science The manuscript benefited from fruitful discussion at the Te- and Archaeology. Philadelphia: University Museum of Archaeology mozon symposium. Revisions were made while I was a visiting and Anthropology, University of Pennsylvania. scholar at the Research Institute for Humanity and Nature, Fuller, D. Q. 2002. Fifty years of archaeobotanical studies in India: laying a solid foundation. In Indian archaeology in retrospect, vol. Kyoto. 3. S. Settar and R. Korisettar, eds. Pp. 247–363. Archaeology and Interactive Disciplines. New Delhi: Manohar. References Cited ———. 2003a. African crops in prehistoric South Asia: a critical review. In Food, fuel and fields: progress in African archaeobotany. Allaby, R. G., D. Q Fuller, and T. A. Brown. 2008. The genetic ex- K. Neumann, A. Butler, and S. Kahlheber, eds. Pp. 239–271. Africa pectations of a protracted model for the origins of domesticated Praehistorica 15. Cologne: Heinrich-Barth Institut. crops. Proceedings of the National Academy of Sciences, U.S.A. ———. 2003b. An agricultural perspective on Dravidian historical 105(37):13982–13986. linguistics: archaeological crop packages, livestock and Dravidian Allchin, F. R. 1963. Neolithic cattle keepers of South India: a case study crop vocabulary. In Examining the farming/language dispersal hy- of the Deccan ashmounds. Cambridge: Cambridge University Press. pothesis. P. Bellwood and C. Renfrew, eds. Pp. 191–213. McDonald Asouti, E., and D. Q Fuller. 2008. Trees and woodlands of South India: Institute Monographs. Cambridge: McDonald Institute for Ar- archaeological perspectives. Institute of Archaeology Publications. chaeological Research. Walnut Creek, CA: Left Coast Press. ———. 2005. Ceramics, seeds and culinary change in prehistoric Bar-Yosef, Ofer. 2011. Climatic fluctuations and early farming in West India. Antiquity 79:761–777. and East Asia. Current Anthropology 52(suppl. 4):S175–S193. ———. 2006. Agricultural origins and frontiers in South Asia: a Bautista, N. S., R. Solis, O. Kamijima, and T. Ishii. 2001. RAPD, working synthesis. Journal of World Prehistory 20:1–86. RFLP and SSLP analyses of phylogenetic relationships between ———. 2007a. Contrasting patterns in crop domestication and do- cultivated and wild species of rice. Genes and Genetics Systems 76: mestication rates: recent archaeobotanical insights from the Old 71–79. World. Annals of Botany 100:903–924. Bocquet-Appel, Jean-Pierre. 2011. The agricultural demographic ———. 2007b. Non-human genetics, agricultural origins and his- transition during and after the agriculture inventions. Current An- torical linguistics. In The evolution and history of human popula- thropology 52(suppl. 4):S497–S510. tions in South Asia: interdisciplinary studies in archaeology, biological Boivin, N. 2004. Landscape and cosmology in the South Indian Neo- anthropology, linguistics and genetics. M. Petraglia and B. Allchin, lithic: new perspectives on the Deccan ashmounds. Cambridge eds. Pp. 389–439. Dordrecht, Netherlands: Springer. Archaeological Journal 14(2):235–257. ———. 2008a. Asia, South: Neolithic cultures. In Encyclopedia of Boivin, N., D. Fuller, R. Korisettar, and M. Petraglia. 2008. First archaeology. D. M. Pearsall, ed. Pp. 756–768. New York: Springer. farmers in South India: the role of internal processes and external ———. 2008b. The spread of textile production and textile crops in influences in the emergence and transformations of South India’s India beyond the Harappan zone: an aspect of the emergence of earliest settled societies. Pragdhara 18:179–200. craft specialization and systematic trade. In Linguistics, archaeology Caicedo, A. L., S. H. Williamson, R. D. Hernandez, A. Boyko, A. and the human past. T. Osada and A. Uesugi, eds. Pp. 1–26. Oc- Fiedel-Alon, T. L. York, N. R. Polato, et al. 2007. Genome-wide casional Paper 3. Kyoto, Japan: Indus Project, Research Institute patterns of nucleotide polymorphism in domesticated rice. PLoS for Humanity and Nature. Genetics 3:e163. Fuller, D. Q, N. Boivin, and R. Korisettar. 2007. Dating the Neolithic Chanchala, S. 1995. Harappan plant economy of Kutch, Gujarat. Geophytology 23:227–233. of South India: new radiometric evidence for key economic, social Chen, W. B., I. Nakamura, Y.-I. Sato, and H. Nakai. 1993. Distri- and ritual transformations. Antiquity 81:755–778. bution of deletion type in cpDNA of cultivated and wild rice. Fuller, D. Q, and E. L. Harvey. 2006. The archaeobotany of Indian Japanese Journal of Genetics 68:597–603. pulses: identification, processing and evidence for cultivation. En- Cheng, C., R. Motohashi, S. Tchuchimoto, Y. Fukuta, H. Ohtsubo, vironmental Archaeology 11:241–268. and E. Ohtsubo. 2003. Polyphyletic origin of cultivated rice based Fuller, D. Q, E. L. Harvey, and L. Qin. 2007. Presumed domestication? on the interspersion patterns of SINEs. Molecular Biology and Evo- evidence for wild rice cultivation and domestication in the fifth lution 20:67–75. millennium BC of the Lower Yangtze region. Antiquity 81:316– Church, A. H. 1886. Food grains of India. London: Chapman & Hall. 331. Cohen, David Joel. 2011. The beginnings of agriculture in China: a Fuller, D. Q, and R. Korisettar. 2004. The vegetational context of multiregional view. Current Anthropology 52(suppl. 4):S273–S293. early agriculture in South India. Man and Environment 29(1):7– Costantini, L. 1979. Plant remains at Pirak. In Fouilles de Pirak, vol. 27. 1. J.-F. Jarrige and M. Santoni, eds. Pp. 326–333. Paris: Boccard. Fuller, D. Q, R. Korisettar, and P. C. Venkatasubbaiah. 2001. Southern ———. 1987. Appendix B: vegetal remains. In Prehistoric and pro- Neolithic cultivation systems: a reconstruction based on archaeo- tohistoric Swat, Pakistan. G. Stacul, ed. Pp. 155–165. Rome, Italy: botanical evidence. South Asian Studies 17:171–187. Instituto Italiano per il Medio ed Estremo Orientale. Fuller, D. Q, R. Korisettar, P. C. Venkatasubbaiah, and M. K. Jones. ———. 2008. The first farmers in western Pakistan: the evidence of 2004. Early plant domestications in southern India: some prelim- the Neolithic agro-pastoral settlement of Mehrgarh. Pragdhara 18: inary archaeobotanical results. Vegetation History and Archaeobot- 167–178. any 13:115–129. Crawford, Gary W. 2011. Advances in understanding early agriculture Fuller, D. Q, and M. Madella. 2001. Issues in Harappan archaeo- in Japan. Current Anthropology 52(suppl. 4):S331–S345. botany: retrospect and prospect. In Protohistory, vol. 2 of Indian Fuller Plant Domestication in India S361

archaeology in retrospect. S. Settar and R. Korisettar, eds. Pp. 317– eds. Pp. 285–289. Pune, India: Deccan College Postgraduate and 390. New Delhi: Manohar. Research Institute. Fuller, D. Q, and L. Qin. 2009. Water management and labour in Kajale, M. D, and S. Eksamberkar. 1997. Application of phytolith the origins and dispersal of Asian rice. World Archaeology 41:86– analyses to a Neolithic site at Budihal, District Gulbarga, South 109. India. In Estado actual de los etudios de fitolitos en suelos y plantas Fuller, D. Q, L. Qin, and E. L. Harvey. 2008. Evidence for a late onset (The state of the art of phytoliths in soils and plants). A. Pinilla, of agriculture in the Lower Yangtze region and challenges for an J. Juan-Tresserras, and M. J. Machado, eds. Pp. 219–229. Madrid: archaeobotany of rice. In Past human migrations in East Asia: Centro Ciencial Medioambientales, Consejo Superior de Investi- matching archaeology, linguistics and genetics. A. Sanchez-Mazas, R. gaciones Cientificas. Blench, M. Ross, M. Lin, and I. Peiros, eds. Pp. 40–83. London: Kar, S. K., K. K. Basa, and P. P. Joglekar. 1998. Exploration at Go- Taylor & Francis. palpur, District Nayagarh, coastal Orissa. Man and Environment Fuller, D. Q, L. Qin, Y. Zheng, Z. Zhao, X. Chen, L. A. Hosoya, and 23(1):107–114. G.-P. Sun. 2009. The domestication process and domestication rate Kimata, M., E. G. Ashok, and A. Seetharam. 2000. Domestication, in rice: spikelet bases from the Lower Yangtze. Science 323:1607– cultivation and utilization of two small millets, Brachiaria ramosa 1610. and Setaria glauca (Poaceae), in South India. Economic Botany 54: Fuller, D. Q, Y.-I. Sato, C. Castillo, L. Qin, A. R. Weisskopf, E. J. 217–227. Kingwell-Banham, J. Song, S.-M. Ahn, and J. van Etten. 2010. Kobayashi, H., and M. Kimata. 1989. Millets in Maharashtra and Consilience of genetics and archaeobotany in the entangled history Orissa in India. In A preliminary report of the studies on millet of rice. Archaeological and Anthropological Sciences 2:115–131. cultivation and its agro-pastoral culture complex in the Indian Sub- Gammie, G. A. 1911. Millets of the genus Setaria in the Bombay continent II. S. Sakamoto, ed. Pp. 1–9. Kyoto, Japan: Research Team Presidency and Sind. Memoirs of the Department of Agriculture in for the Studies of Millet Cultivation in the Indian Subcontinent, India, Botanical Series 4:1–8. Kyoto University. Garris, A. J., T. H. Tai, J. Coburn, S. Kresovich, and S. McCouch. Korisettar, R., P. C. Venkatasubbaiah, and D. Q Fuller. 2001. Brah- 2005. Genetic structure and diversity in Oryza sativa L. Genetics magiri and beyond: the archaeology of the Southern Neolithic. In 169:1631–1638. Prehistory,vol.1ofIndian archaeology in retrospect. R. Korisettar Goring-Morris, A. Nigel, and Anna Belfer-Cohen. 2011. Neolithi- and S. Settar, eds. Pp. 151–238. New Delhi: Manohar. zation processes in the Levant: the outer envelope. Current An- Kovach, M. J., M. T. Sweeney, and S. R. McCouch. 2007. New insights thropology 52(suppl. 4):S195–S208. into the history of rice domestication. Trends in Genetics 23:578– Harlan, J. R. 1971. Agricultural origins: centers and noncenters. Sci- 587. ence 174:468–474. Londo, J. P.,Y.-C. Chiang, K.-H. Hung, T.-Y. Chiang, and B. A. Schaal. Harris, D. R. 1990. Vavilov’s concept of centres of origin of cultivated 2006. Phylogeography of Asian wild rice, Oryza rufipogon, reveals plants: its genesis and its influence on the study of agricultural multiple independent domestications of cultivated rice, Oryza sa- origins. Biological Journal of the Linnean Society 39:7–16. tiva. Proceedings of the National Academy of Sciences, U.S.A. 103: Harvey, E. L. 2006. Early agricultural communities in Northern and 9578–9583. Eastern India: an archaeobotanical investigation. PhD dissertation, Luddy, S. 2008. The site of Ghola Dhoro, Bagasra in the framework University College London. of cuisine and cultural identity in Harappan Gujarat: an archaeo- Harvey, E. L., and D. Q Fuller. 2005. Investigating crop processing botanical and ceramic analysis. BSc dissertation, University College London. through phytolith analysis: the case of rice and millets. Journal of Maheshwari, P., and U. Singh. 1965. Dictionary of economic plants in Archaeological Science 32:739–752. India. New Delhi: Indian Council for Agricultural Research. Harvey, E. L., D. Q Fuller, K. K. Basa, R. Mohany, and B. Mohanta. Marshall, F., and E. Hildebrand. 2002. Cattle before crops: the be- 2006. Early agriculture in Orissa: some archaeobotanical results ginnings of food production in Africa. Journal of World Prehistory and field observations on the Neolithic. Man and Environment 31: 16:99–143. 21–32. Marshall, Fiona, and Lior Weissbrod. 2011. Domestication processes Hilu, K. W., J. M. J. De Wet, and J. R. Harlan. 1979. Archaeobotanical and morphological change: through the lens of the donkey and studies of Eleusine coracana ssp. coracana (finger millet). American African pastoralism. Current Anthropology 52(suppl. 4):S397–S413. Journal of Botany 66:330–333. Meadow, R. 1996. The origin and spread of agriculture and pastor- Jarrige, J.-F. 1982. Excavations at Mehrgarh: their significance for alism in northwestern South Asia. In The origin and spread of understanding the background of the Harappan civilization. In agriculture and pastoralism in Eurasia. D. R. Harris, ed. Pp. 390– Harappan civilization: a contemporary perspective. G. L. Possehl, 412. London: University College London Press. ed. Pp. 79–84. Delhi: Oxford/IBH. Meadow, R., and A. K. Patel. 2003. Prehistoric pastoralism in north- ———. 2008. Mehrgarh Neolithic. Pragdhara 18:135–154. western South Asia from the Neolithic through the Harappan Jarrige, J.-F., C. Jarrige, and G. Quivron. 2006. Mehrgarh Neolithic: period. In Indus ethnobiology: new perspectives from the field.S.A. the updated sequence. In South Asian archaeology 2001. C. Jarrige Weber and W. R. Belcher, eds. Pp. 65–94. Lanham, MD: Lexington. and V. Lefevre, eds. Pp. 129–141. Paris: Editions Recherche sur les Mishra, S., S. G. Deo, S. Naik, P. P. Joglekar, A. Deshpande- Civilisations. Mukherjee, S. Ghatel, A. Kshirasagar, and U. Adhav. 2006. Evi- Johansen, P. G. 2004. Landscape, monumental architecture, and rit- dence of human occupation from a terrace dated to 7.8 kyr at ual: a reconsideration of the South Indian ashmounds. Journal of Sakshal Pimpri, District Bid, Maharashtra. Man and Environment Anthropological Archaeology 23:309–330. 31:19–26. Kajale, M. D. 1988. Plant economy. In Excavations at Inamgaon.M. Misra, V. N. 2005. Radiocarbon chronology of Balathal, District Udai- K. Dhavalikar, H. D. Sankalia, and Z. D. Ansari, eds. Pp. 727–821. pur, Rajasthan. Man and Environment 30:54–60. Pune, India: Deccan College Postgraduate and Research Institute. Morrell, P. L., and M. T. Clegg. 2007. Genetic evidence for a second ———. 1996a. Palaeobotanical investigations at Balathal: prelimi- domestication of barley (Hordeum vulgare) east of the Fertile Cres- nary results. Man and Environment 21:98–102. cent. Proceedings of the National Academy of Sciences, U.S.A. 104(9): ———. 1996b. Plant remains. In Kuntasi: a Harappan emporium on 3289–3294. the west coast. M. K. Dhavalikar, M. R. Raval, and Y. M. Chitalwala, Moulherat, C., M. Tengberg, J.-F. Haquet, and B. Mille. 2002. First S362 Current Anthropology Volume 52, Supplement 4, October 2011

evidence of cotton at Neolithic Mehrgarh, Pakistan: analysis of and the human past. T. Osada, ed. Pp. 213–214. Kyoto, Japan: mineralized fibres from a copper bead. Journal of Archaeological Research Institute for Humanity and Nature. Science 29:1393–1401. Saxena, A., V. Prasad, I. B. Singh, M. S. Chauhan, and R. Hassan. Patel, A. K. 2008. New radiocarbon determinations from Loteshwar 2006. On the Holocene record of phytoliths of wild and cultivated and their implications for understanding Holocene settlement and rice from Ganga Plain: evidence for rice-based agriculture. Current subsistence in North Gujarat and adjoining areas. In South Asian Science 90:1547–1552. Archaeology 1999. E. R. Raven, ed. Pp. 123–134. Groningen, Neth- Singh, I. B. 2005. Quaternary palaeoenvironments of the Ganga Plain erlands: Forsten. and anthropogenic activity. Man and Environment 30:1–35. Patel, A. K., and R. Meadow. 1998. The exploitation of wild and Southworth, F. 2005. The linguistic archaeology of South Asia. London: domestic water buffalo in prehistoric northwestern South Asia. In Routledge. Archaeozoology of the Near East III. H. Buitenhuis, L. Bartosiewicz, Sweeney, M. T., and S. R. McCouch. 2007. The complex history of and A. M. Choyke, eds. Pp. 180–199. Groningen, Netherlands: the domestication of rice. Annals of Botany 100:951–957. Centre for Archaeological Research and Consultancy, Rijksuni- Takahashi, H., Y.-I. Sato, and I. Nakamura. 2008. Evolutionary anal- versiteit Groningen. ysis of two plastic DNA sequences in cultivated and wild species Pearsall, D. M., D. R. Piperno, E. H. Dinan, M. Umlauf, Z. Zhao, of Oryza. Breeding Science 58(3):225–233. and R. A. Bentfer Jr. 1995. Distinguishing rice (Oryza sativa Po- Tan, L., X. Li, F. Liu, X. Sun, C. Li, Z. Zhu, Y. Fu, et al. 2008. Control aceae) from wild Oryza species through phytolith analysis: results of a key transition from prostrate to erect growth in rice domes- of preliminary research. Economic Botany 49(2):183–196. tication. Nature Genetics 40:1360–1364. Pokharia, A. K. 2007. Palaeobotanical investigations. In J. S. Khar- Tang, J. B., H. A. Xia, M. L. Cao, X. Q. Zhang, W. Y. Zheng, S. N. akwal, Y. S. Rawat, and T. Osada, Kanmer. A Harappan site in Hu, W. Tong, et al. 2004. A comparison of rice chloroplast ge- Kachchh, Gujarat, India. In Linguistics, archaeology and the human nomes. Plant Physiology 135:412–420. past. T. Osada, ed. Pp. 29–31. Occasional Paper 2. Kyoto, Japan: Tengberg, M. 1999. Crop husbandry at Miri Qalat, Makran, SW Indus Project, Research Institute for Humanity and Nature. Pakistan (4000–2000 B.C.). Vegetation History and Archaeobotany ———. 2008. Palaeoethnobotanical investigation. In J. S. Kharakwal, 8:3–12. Y. S. Rawat, and T. Osada. Preliminary observations on the ex- Tewari, R., R. K. Srivastava, K. S. Saraswat, I. B. Singh, and K. K. cavation at Kachchh, India 2006–07. In Linguistics, archaeology and Singh. 2008. Early farming at Lahuradewa. Pragdhara 18:347–373. the human past. T. Osada and A. Uesugi, eds. Pp. 19–22. Occasional Tewari, R., R. K. Srivastava, K. K. Singh, K. S. Saraswat, I. B. Singh, Paper 5. Kyoto, Japan: Indus Project, Research Institute for Hu- M. S. Chauhan, A. K. Pokharia, A. Saxena, V. Prasad, and M. manity and Nature. Sharma. 2006. Second preliminary report of the excavations at Possehl, G. L. 1999. Indus age: the Beginnings. Philadelphia: University Lahuradewa District Sant Kabir Nagar, U. P.: 2002–2003–2004 and of Pennsylvania Press. 2005–06. Pragdhara 16:35–68. ———. 2002. The Indus civilization: a contemporary perspective. Wal- Tomooka, N., D. A. Vaughan, H. Moss, and N. Maxted. 2003. The nut Creek, CA: Alta Mira. Asian Vigna: genus Vigna subgenus Ceratotropis genetic resources. Prasada Rao, K. E, J. M. J. De Wet, D. E. Brink, and M. H. Mengesha. Dordrecht, Netherlands: Springer. 1987. Infraspecific variation and systematics of cultivated Setaria Van der Maeson, L. J. G. 1980. India is the native home of the italica, foxtail millet (Poaceae). Economic Botany 41:108–116. pigeonpea. In Libergratulatorius in honorem H.C.D. de Wit.J.C. Reddy, S. N. 2003. Food and fodder: plant usage and changing socio- cultural landscapes during the Harappan phase in Gujarat, India. Arends, G. Boelema, C. T. de Groot, and A. J. M. Leeuwenberg, In Indus ethnobiology: new perspectives from the field.S.A.Weber eds. Pp. 257–262. Landbouwhoge School Miscellaneous Papers. and W. R. Belcher, eds. Pp. 327–342. Lanham, MD: Lexington. Wageningen, Netherlands: Veenman. Saisho, D., and M. D. Purugganan. 2007. Molecular phylogeography Vaughan, D. A., B.-R. Lu, and N. Tomooka. 2008. The evolving story of domesticated barley traces expansion of agriculture in the Old of rice evolution. Plant Science 174:394–408. World. Genetics 177:1765–1776. Vavilov N. 1992 (1927). Origin and geography of cultivated plants. Sang, T. 2009. Genes and mutations underlying domestication tran- Cambridge: Cambridge University Press. sitions in grasses. Plant Physiology 149:63–70. Vitte, C., T. Ishii, F. Lamy, D. Brar, and O. Panaud. 2004. Genomic Saraswat, K. S. 2004. Plant economy of early farming communities. paleontology provides evidence for two distinct origins of Asian In Early farming communities of the Kaimur (excavations at Se- rice (Oryza sativa L.). Molecular Genetics and Genomics 272:504– nuwar). B. P. Singh, ed. Pp. 416–535. Jaipur, India: Scheme. 511. ———. 2005. Agricultural background of the early farming com- Weber, S. A. 1991. Plants and Harappan subsistence: an example of munities in the Middle Ganga Plain. Pragdhara 15:145–177. stability and change from Rojdi. New Delhi: Oxford/IBH. Saraswat, K. S., and A. K. Pokharia. 2003. Palaeoethnobotanical in- Wengrow, David. 2006. The archaeology of early Egypt: social trans- vestigations at Early Harappan Kunal. Pragdhara 13:105–140. formations in North-East Africa, 10,000 to 2650 BC. Cambridge: Sato, Y.-I. 1996. Origin of rice and rice cultivation based on DNA Cambridge University Press. analysis. Tokyo: Japan Broadcast. [In Japanese.] Zeder, Melinda A. 2011. The origins of agriculture in the Near East. ———. 2002. Origin of rice cultivation in the Yangtze River basin. Current Anthropology 52(suppl. 4):S221–S235. In The origins of pottery and agriculture. Y. Yasuda, ed. Pp. 143– Zhao, Zhijun. 2011. New archaeobotanic data for the study of the 150. New Delhi: Lustre/Roli. origins of agriculture in China. Current Anthropology 52(suppl. 4): ———. 2005. Rice and Indus civilization. In Linguistics, archaeology S295–S306. Current Anthropology Volume 52, Supplement 4, October 2011 S363

Holocene Population History in the Pacific Region as a Model for Worldwide Food Producer Dispersals

by Peter Bellwood

Pacific prehistory (excluding Australia) since 3000 BC reflects the impacts of two source regions for food production: China from the Yangzi southward (including Taiwan) and the western Pacific (especially the New Guinea Highlands). The linguistic (Austronesian, Trans–New Guinea), bioan- thropological/human genetic, and Neolithic archaeological records each carry signals of expansion from these two source regions. A combined consideration of the multiregional results within all three disciplines (archaeology, linguistics, and biology) offers a historical perspective that will never be obtained from one discipline or one region alone. The fundamental process of human behavior involved in such expansion—population dispersal linked to increases in human population size—is significant for explaining the early spreads of food production and language families in many parts of the world. This article is concerned mainly with the archaeological record for the expansion of early food producers, Austronesian languages, and Neolithic technologies through Taiwan into the northern Philippines as an early stage in what was to become the greatest dispersal of an ethnolinguistic population in world history before AD 1500.

Introduction that populations would have migrated when they did into the remote reaches of Oceania. No stage in Austronesian dispersal involved populations who permanently abandoned all com- The organizers of the Wenner-Gren conference “The Origins mitment to food production unless forced to do so by en- of Agriculture: New Data, New Ideas” asked me to examine vironmental circumstances, no matter how significant the the dispersal of agricultural systems in the Pacific and the contributions of fishing, bird and sea-mammal hunting, and relationships among language, culture, and farming in the wild-plant consumption were in initial colonizing situations area and also to discuss to what extent I think this model of natural plenty. No claim is made that farming drove the provides for a more general understanding of the dispersal expansion via overpopulation or a desperate search for new of languages and farming in other parts of the world. At the territory, but the systematic development of food production outset, I wish to make my fundamental perspective clear. The stands forward as the fundamental cultural basis for the ex- dispersal of the Austronesian language family and its speakers pansion. commenced after 4000 BC1 among increasing populations of My topic thus focuses on dispersals of agriculture, lan- Neolithic food producers in southern China and Taiwan, in guages, and attached human populations, and not specifically a cultural situation of increasing population density, advanc- on the ultimate origins of food production systems in China ing technology (including boat construction and carpentry), or New Guinea or gradients in their development through and increasing dependence on agriculture and animal do- stages such as predomestication cultivation or low-level food mestication, also a portable food production repertoire that production. We start with populations in southern China and allowed long-distance dispersal to take place. Food production Taiwan before 3000 BC who manifested already an identifiable and the remarkable lithic and lignic technology behind it thus level of dependence on food production. underpinned the expansion, and without them it is unlikely Initially, I offer several introductory observations presented in more detail elsewhere (Bellwood 2001, 2005, 2008, 2009a, Peter Bellwood is Professor of Archaeology in the School of 2009b; Bellwood and Oxenham 2008). (a) Spreads of food Archaeology and Anthropology, Australian National University production and human migration in general are phenomena (Canberra, Australian Capital Territory 0200, Australia [peter [email protected]]). This paper was submitted 13 XI 09, 1. All dates are expressed as BC/AD and based on calibrated C14 accepted 1 XI 10, and electronically published 13 V 11. determinations.

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0014$10.00. DOI: 10.1086/658181 S364 Current Anthropology Volume 52, Supplement 4, October 2011 that have always involved demographic growth—declining by early farmers; indeed, the latter would probably have wel- populations cannot create successful colonists. Birth rates comed new members from foraging communities before the must exceed death rates over the long term. (b) Archaeology development of sufficient population density to promote re- and comparative linguistics attest to coherent movements of source competition. Well-studied farming dispersals in Eu- food-producing populations from several large homeland rope, the Andes (Heggerty and Beresford-Jones 2010), South regions in different parts of the world. (c) Food production Asia, and China required more than 3,000 years for comple- in general spread with existing food producers even when tion, and homeland genetic configurations could not possibly previous hunter-gatherer populations existed, the more so as have spread intact and hermetically sealed through such long dependence on morphologically domesticated plants and an- time spans. imals increased within the food producers’ diet. (d) Migration The validity of the demic diffusion model is strengthened and interaction have played different roles in creating cultural by analyses of early food-producing cemetery populations that and biological patterns within humanity (Bellwood 1996; see Hunley et al. 2008 for western Oceania). Migration, as a spo- indicate marked increases in birth rate following the regional radic rather than continuous occurrence on any large scale, appearances of agriculture and animal husbandry. These in- has laid down successive foundations of biological phylogeny creases occurred before the later increases in mortality that and cultural/linguistic genealogy. Interaction has continually resulted from the crowded and insanitary lifestyles associated and productively modified these foundations, normally re- with increasingly sedentary life (Bocquet-Appel 2011). In- organizing rather than suddenly replacing previous patterns. creasing populations must either seek fresh land or intensify (e) In general, major movements of languages in prehistoric production in order to survive, and the former option would societies occurred with major movements of native speakers, often have been inviting in situations surrounded by lower- not through language shift or the expansion of “trade lan- density forager populations. The historical validity of such a guages.” Ostler (2005) and Ross (2008) have discussed these model is also revealed for us by a small number of ethno- matters in detail, the latter for Austronesian, and those who historical small-scale and kinship-based tribal populations believe in trade languages as the source of any extensive an- who managed to avoid the main hazards of the colonial era. cient language landscape should consider Melanesia—home One such group were the Iban of Sarawak (Borneo), swidden for some of the most famous exchange systems in world rice farmers with long fallow land requirements who were ethnography yet also one of the most diverse linguistic regions colonizing along rivers starting perhaps in the late eighteenth in the world in terms of its high number of mutually unin- century and continuing into the early decades of the twentieth telligible precolonial languages. Exchange in Melanesia did century (Freeman 1970). The movements spanned an already- not homogenize languages. Indeed, it probably had the op- posite effect through its role in underpinning group identity. populated territory stretching more than 850 km from west- Communication relied on multilingualism, not language shift. ern Kalimantan through Sarawak to Brunei Bay. According This is one reason why the recent linguistic arguments by to Freeman, Donohue and Denham (2010) for the spread of Austronesian The main incentive behind the remarkable migrations of languages by elite dominance, as lexical replacements within the Iban has been a desire to exploit new tracts of primeval existing non-Austronesian languages, are unconvincing. (f ) forest, and the tendency has been for communities to aban- Comparative linguistics is also of absolutely fundamental im- don their land as soon as a few lucrative harvests have been portance in offering hypotheses about language-family home- reaped, and move on to fresh precincts. (Freeman 1970:76) lands, directions, and relative chronologies of subgroup dis- persal as well as protolanguage reconstructions of meanings The Iban expansion, with its consequent assimilation of relevant for archaeological inference. Linguistic reconstruc- intervening populations, can be compared with that of the tions can be deeply meaningful for historical understanding ethnographic Yanomami of the upper Orinoco as described given the importance of language as the most significant ve- by Chagnon (1992) and the Nuer of Sudan as described by hicle of human expression and interaction. Kelly (1985). All represent demic spreads of food producers My operational model for early agriculturalist dispersal is into terrain inhabited either by foragers or by other less ag- that of “demic diffusion” involving a food producer “wave gressive food producers. Interestingly, recent genetic research of advance” extending or leapfrogging, as the case may be, in Bali (Lansing et al. 2008) has revealed that modern subak and mixing with forager populations as proposed originally for Europe by Ammerman and Cavalli-Sforza (1973). This is irrigation associations have low genetic diversity and so prob- a logical mechanism for population expansion, involving con- ably formed by internal demographic growth through fis- tinuous population growth in frontier regions, fissioning, and sioning from founder populations into adjacent drainage sys- intermarriage with other communities. It is not necessarily a tems, rather than by state-enforced population movement. mechanism for population replacement. The “early farming Although not directly related to Neolithic issues, there is an dispersal” hypothesis (Bellwood 2005, 2009a, 2009b; see also interesting model here for the gradual spread of earlier farm- Renfrew 2002) does not demand extermination of foragers ing populations through demographic growth. Bellwood Holocene Population Prehistory in the Pacific Region S365 Agricultural Origins in East Asia ern China to reach ultimately as far as the Ganges Basin, and the Pacific Sumatra, and Samoa. These movements sometimes involved migration into uninhabited terrain, but more common were What is the central question? I suppose it has to be why at processes of demic diffusion into regions already occupied by the beginning of the colonial era (AD 1500 for argument’s indigenous populations. However, the New Guinea Highland sake) were the majority of populations in Southeast Asia and populations restricted their biological and cultural influences Oceania subsisting from the proceeds of food production? to the western Pacific, and while some degree of migration Continuations of foraging, hunting, and fishing from the wild might have occurred, particularly from Trans–New Guinea are irrelevant for the question posed because they have always genetic and linguistic perspectives (Mona et al. 2007; Pawley existed and always will, global warming permitting. Likewise, 2007), it was evidently not dramatic. This situation reflects all ethnographically recorded foragers manage resources in the differing capacities of the food production systems in one way or another. Predomestication cultivation and low- China and New Guinea to support internal population growth level food production (Smith 2001, 2011) are useful concepts via increased birth rates. Cereals, legumes, and pigs fueled for the beginnings of food production, and the latter will expansion from China, particularly during the period of high- always exist among ecologically marginal communities, but density population represented by the Qujialing, Liangzhu, the focus here is on the conscious production of food from and so-called Longshan cultures of the Yangzi and Yellow river combinations of wild and domesticated species in nonmar- basins (third millennium BC; Jing and Campbell 2009; Zhang ginal environments sufficiently productive to support outward and Hung 2008, 2010). Without cereals and domesticated population expansion. animals (pigs and dogs probably arrived less than 3,000 years Within the Asia-Pacific region, the inceptions of food pro- ago on the New Guinea mainland), the early New Guinea duction occurred independently in the New Guinea High- systems of food production were less expansive. lands and central China (the Yangzi and Yellow river basins). In terms of agricultural dispersals and human prehistory In both regions, some populations were reasonably dependent in general, Island Southeast Asia and Oceania have had dif- on food production by 4000 BC (Cohen 2011; Denham 2011; ferent historical trajectories. East of the Solomons, in the Fuller, Harvey, and Qin 2007; Zhao 2011). China witnessed scattered archipelagos of “Remote Oceania,” human settle- the development of cereal (rice, foxtail, and common millet), ment has occurred only within the past 3,100 years (fig. 1). legume (soybean), and pig production. Whether the Yellow Nowhere within Remote Oceania is there evidence to suggest and Yangzi river basins formed one or two “centers” is not complete population replacement—whether genetic, linguis- relevant for present purposes—I suspect only one on the evi- tic, or cultural—except to a minor degree in cases such as dence of cultural connections, but animal domestication ap- those of the Polynesian outliers. All scholars agree that the peared slightly later in the south than in the north (Yuan, first settlers in these islands were the direct ancestors of the Flad, and Luo 2008). The New Guinea Highlands, a unique populations there today, allowing for some obvious situations high-altitude and equatorial-cordilleran environment without of later interisland migration, contact, and intermarriage (e.g., geomorphic parallel anywhere in the volcanic arcs of Island Addison and Matisoo-Smith 2010). All of these populations, Southeast Asia, witnessed the development of tuber and fruit without exception, speak Malayo-Polynesian languages, and cultivation without cereals or domestic animals. all practiced noncereal food production focused on fruit/nut Chinese material culture was quintessentially “Neolithic,” and tuber cultivation in one form or another at European with pottery, sawn and ground stone, weaving, advanced car- contact, with the exception of the southern Maoris of New pentry, and boats and paddles, whereas New Guinea Highland Zealand, who lived beyond the growing-season limits of the material culture was less modified from its Pleistocene lithic sweet potato, their only significant cultivated plant. and aceramic roots, except for the addition of fully polished The area termed “Near Oceania” (New Guinea and western stone axes during the Holocene. New Guinea Highland ar- Island Melanesia), together with the much vaster extent of chaeology reveals no signs of direct contact with contem- Island Southeast Asia to the west, is rather more complex. porary Neolithic societies in Island Southeast Asia, although Modern humans presumably first reached these islands within the situation was different with respect to the islands of Mel- the past 50 kyr, and their descendants in New Guinea and anesia and lowland New Guinea. Movements of useful plants Melanesia were also mainly food producers at European con- such as sugarcane, bananas, and taro from New Guinea into tact. But it is in the region west of New Guinea that the Indonesia in the mid-Holocene are perfectly possible, as contentious debates occur. Today, no one suggests that the claimed by Donohue and Denham (2010), but are so far islands of Remote Oceania underwent totally independent undemonstrated. So, too, are pre-Neolithic movements the transitions to agriculture, but most accept that the island of other way, from Indonesia into New Guinea. New Guinea, particularly its cordilleran central highlands, Following the development of food production in central supported indigenous systems of fruit and tuber food pro- China, outward migrations of farming populations with their duction early in the Holocene (Denham 2011). The issues genes, material cultures, and languages occurred widely be- involve the Southeast Asian islands—Taiwan, the Philippines, tween 3000 and 1000 BC, when populations moved via south- Indonesia, and East Malaysia (northern Borneo). These is- Figure 1. Southeast Asia and Oceania, showing geographical divisions and the main chronological trends in Neolithic and Austronesian settlement. Bellwood Holocene Population Prehistory in the Pacific Region S367 lands, except for a few immediately west of New Guinea (Ti- . . . is quite simply that their speakers dispersed, taking their mor, Halmahera, Alor, Pantar), are occupied entirely by languages with them” (165). speakers of Austronesian languages who are biologically clinal, Comparative lexical reconstructions for PAn and its daugh- with a dramatic and steep change from Asian to Melanesian ters PMP and proto-Oceanic (POc) indicate major changes genetic ancestry within a relatively narrow geographical win- in material culture through time and space (Blust 2009; Paw- dow in the eastern Lesser Sunda islands and the Moluccas ley 2002; Ross, Pawley, and Osmond 2007). For PAn itself, (Cox 2008; Cox et al. 2010; Mona et al. 2009). they reveal an economy focused on rice cultivation, with a At present, there is no archaeological evidence that Island large vocabulary for rice in many forms and stages of growth Southeast Asia witnessed any independent development of as well as a processing and cultivating vocabulary with words agriculture, even at a very low level. We have yet to learn for millet, sugarcane, and possibly aroids. PMP added many whether early Holocene foragers here were involved in moving fruits and tubers to this vocabulary as befitted its probable around plants such as bananas, breadfruit, and sugarcane (but location in the tropical Philippines (Taiwan is mostly tem- presumably not cereals), perhaps even selecting and planting perate in latitude). POc witnessed the ultimate loss of rice some varieties. Yet even if people made such transfers, es- under equatorial climatic and day-length conditions (Paz pecially from New Guinea, well before the appearance of the 2002:280). Other reconstructions apply to words for pigs and Asian-derived ceramic Neolithic, it could be entirely irrelevant dogs (but not chickens until PMP), pottery, boats and sails, for understanding events since that time. The botanical origins fishing, and a wild placental (not marsupial) mammal fauna. of translocated plant species are not the issue under debate; The morphological and semantic integrity of these many re- rather, it is food production and human migration. constructions, as Pawley (2002:266) points out, implies con- tinuous linguistic transmission through time, not late bor- rowing. In other words, Austronesians did not start as farmers, The Austronesian-Speaking Peoples revert to foraging and lose all their farming vocabulary, and and Their Significance then readopt farming much later. Even the few centuries of My aim here is to examine the data enshrined within the naive faunal exploitation on Remote Oceanic islands did not different disciplines that support the inference of a migration reduce the food-producing vocabulary of Polynesians. Fur- of ancestral populations speaking Austronesian languages, es- thermore, the rakelike phylogeny of the main MP subgroups pecially those within the major Malayo-Polynesian subgroup, implies rapid migrational spread at least from the Philippines as outlined in figure 1. I make no apology for beginning with to as far east as western Polynesia (Gray, Drummond, and the linguistic evidence because the Austronesian-speaking Greenhill 2009; Pawley 1999). populations have a well-studied linguistic history and formed The recent claim by Donohue and Denham (2010) that the most widespread ethnolinguistic group in the world before differing geographical distributions of Malayo-Polynesian AD 1500. The Austronesian language family is a clearly de- grammatical features in Island Southeast Asia imply only a fined taxon, and all current homeland theories for it focus spread of Austronesian vocabulary, not whole languages, on Taiwan (Donohue and Denham 2010; Ross 2008). As lin- hence without significant population movement, is not con- guist Malcolm Ross notes, vincing. No recorded non-Austronesian languages survive for comparative purposes west of Halmahera and Timor, and the Of the disciplines represented...linguistics is probably the described grammatical variations could also be due to contact- closest to unanimity about Austronesian origins. All Aus- induced change operating subsequently to the period of initial tronesian languages spoken outside Taiwan belong to a sin- Malayo-Polynesian settlement. Furthermore, their claim does gle subgroup, dubbed Malayo-Polynesian by Blust (1977), not explain why Malayo-Polynesian languages are spoken whilst the thirteen Austronesian languages still spoken in across the entirety of the Philippines and Indonesia, west of Taiwan belong to several primary subgroups (Blust 1999 Timor and Halmahera, yet not spoken at all in most of the proposes nine, on phonological grounds). The logical in- interior of New Guinea. Why was New Guinea different if, ference is that proto-Austronesian (PAn) was spoken in Tai- as Donohue and Denham claim, food production existed wan, that it split initially into dialects, and that these dialects across the whole area in pre-Austronesian times? How could eventually diversified into separate languages. Speakers of Austronesian languages possibly have spread by processes that just one of the dialects, proto-Malayo-Polynesian (PMP), involved almost no population movement? Naturally, pre- left Taiwan and settled initially either on Lanyu (Orchid) Austronesian languages once occurred in Island Southeast Island, or somewhere in the Batanes Islands, or on the north Asia; the disagreement concerns not their existence but their coast of Luzon. It is speakers of languages descended from role in the formation of the modern languages of the region. PMP who have settled the huge expanse of the Austronesian- Despite the clarity of their shared ancestry in a linguistic speaking region beyond Taiwan. (cited in Bellwood et al., sense, Austronesian biological origins are expectably more forthcoming) diverse. Not all Austronesians share a recent genetic origin, Ross (2008) also comments that “the main reason why Aus- and the prehistories of some regions in eastern Indonesia and tronesian languages cover the vast territory which they occupy the western Pacific have involved a high degree of genetic S368 Current Anthropology Volume 52, Supplement 4, October 2011 indigenization and localized instances of full language shift no trace in caves or surface finds of prior hunter-gatherer (Mona et al. 2009). Nevertheless, while Donohue and Den- preceramic occupation or flaked lithic tool manufacture. Lu- ham (2010) rely on somewhat outdated molecular clock– zon, to the contrary, had Paleolithic hunters and gatherers in based mitochondrial DNA analyses to support their claimed occupation since at least 24,000 and possibly 67,000 years ago, refutation of any significant Austronesian population expan- so the first Neolithic arrivals must have interacted with these sion, the reality from current genetic research—haploid, au- groups, as Mijares has shown for the Pen˜ablanca Caves near tosomal, and increasingly at a whole-genome level—is that Tuguegarao (Mijares 2006; Mijares et al. 2010). Holocene population movement from mainland Asia via Tai- wan into Island Southeast Asia and Oceania is very strongly supported (Cox 2008; Cox et al. 2010; Friedlaender et al. 2008; The Spread of Neolithic Pottery from Kayser 2010; Kayser et al. 2008a, 2008b; Kimura et al. 2008; Taiwan into the Philippines and Tabbada et al. 2010). The genetics of human stomach bacterial Indonesia parasites (Moodley et al. 2009) provide additional confir- mation. Newly excavated ceramic data establish the development of a tradition of red-slipped plain ware pottery manufacture in southern and eastern Taiwan, emergent by at least 2200 BC Current Archaeological Perspectives: New from a prior “Middle Neolithic” tradition with both cord Research in Taiwan and the Philippines marking and red slip (Bellwood and Dizon 2008; Hung 2008; Understanding of prehistory in Taiwan and the northern Phil- table 1; fig. 3). A key site here is Chaolaiqiao, on Shanyuan ippines has recently developed very rapidly. The main break- Bay, accurately dated by AMS C14 to 2200 BC and with red- throughs have come with the established presence by at least slipped pottery forming almost 100% of the assemblage, with 2800 BC of an agricultural (rice and foxtail millet) economy virtually no cord marking or other type of decoration. By for the Dabenkeng Neolithic culture of southwestern coastal 2000 BC, this red-slipped plain ware tradition had spread to Taiwan (Tsang 2005; Tsang, Li, and Chu 2006), with the doc- previously uninhabited Batanes as documented in Reranum umentation of a sixfold or greater increase in site numbers and Torongan Caves on Itbayat. Pottery vessels in this phase during the course of the third millennium BC in eastern commonly had pedestal bases and tall everted rims and lacked Taiwan (Hung 2005:126) and also with the recovery of fine- body decoration apart from the red slip (fig. 2U,2V). Rer- grained ceramic evidence for the spread at about 2200 BC of anum and Chaolaiqiao still have some residual cord marking Neolithic material culture from Taiwan to the Batanes Islands (fig. 2R), and the close similarities in red-slipped pottery be- (previously uninhabited) and northern Luzon (Bellwood and tween these two sites raise the possibility that a direct mi- Dizon 2005, 2008; Hung 2005, 2008). This Neolithic spread gration from sites such as An Son in southeastern Taiwan to carried (not necessarily all together) red-slipped pottery with Itbayat could have occurred between 2200 and 2000 BC. specific rim forms and body shapes, pottery spindle whorls, For northern Luzon (Philippines), current research on the stone bark cloth beaters, tanged or grooved stone adzes, Feng- lowest deposits beneath the late Neolithic shell mound at tian (eastern Taiwan) nephrite, Taiwan slate knives and pro- Nagsabaran (Hung 2008; Hung et al., forthcoming) suggests jectile points, notched pebble net sinkers, domestic pigs (Sus that both red-slipped plain ware and stamped pottery ap- scrofa) and dogs, and rice (presumably domesticated but a peared together around 2000 BC (see also Ogawa 2005). A prehistoric presence of millet still remains uncertain beyond closely related tradition of red-slipped and punctate/circle- Taiwan). A large number of these items from the Batanes stamped pottery decoration is also reported from early sites Islands in the northern Philippines are shown in figure 2. A such as Achugao and Unai Bapot on Saipan in the Mariana precise archaeological homeland within the island of Taiwan Islands, western Micronesia, where initial settlement across is not yet identifiable, and it is possible that groups from 2,300 km of open sea (the first truly long-distance sea voyage different regions were involved in many individual move- in human history) occurred from the northern Philippines at ments, with the closest ceramic parallels so far being focused about 1500 BC (Butler 1995; Carson 2008; Clark et al. 2010; on the southeastern coastline. Fish bone data from Eluanbi Hung et al., forthcoming). A similar co-occurrence of red- in southern Taiwan and also from Batanes indicate that the slipped plain ware with small amounts of stamped and incised canoeborne ability to catch open-sea pelagic carnivores such decoration also commenced at ca. 1300 BC at Bukit Tengkorak as dolphinfish (Coryphaena hippurus) using trolling lures with in Sabah, here with bark cloth beaters and trapezoidal cross- stone shanks and bone points had also developed by at least sectioned adzes paralleled in Batanes, Taiwan, and Fujian 2000 BC (Hung et al., forthcoming). (Chia 2003; Jiao 2007). Bukit Tengkorak also has Talasea (Ku- In the case of the Batanes, excavations in five caves and tau/Bao) obsidian from the Bismarck Archipelago in Near rock shelters with plentiful ceramic period occupation leave Oceania, thus illuminating two-way human movement on a no doubt that humans had not previously reached these wind- remarkable scale (Bellwood 1989; Chia 2003). swept islands, protected by relatively rough sea and sometimes In other parts of central and eastern Indonesia, the dec- strong ocean currents, until the Neolithic. There is absolutely orated pottery appears later than the red-slipped plain ware, Bellwood Holocene Population Prehistory in the Pacific Region S369 similar to the situation in southeastern Taiwan and Batanes. Son in southern Vietnam (P. Bellwood, M. Oxenham, C. H. Red-slipped plain ware is dated from ca. 1500 BC at Kamassi Bui, et al., unpublished manuscript). However, this is still an and Minanga Sipakko in West Sulawesi (Simanjuntak et al. area of uncertainty that requires further research. 2008; incised pottery and obsidian from these sites is youn- The above pottery data thus point to a secure ceramic ger), and other sites in central and eastern Indonesia with sequence for much of Island Southeast Asia (except for Sa- early red-slipped plain ware include Kendeng Lembu in east- rawak and perhaps Sumatra), within which combinations of ern Java, Uattamdi in Maluku, Leang Tuwo Mane’e in Talaud, red-slipped and stamped pottery spread from Taiwan and possibly Paso in northern Sulawesi, and Madai Cave in Sabah possibly other adjacent regions of the Asian mainland into (Bellwood 1997; fig. 3). Indonesia and western Oceania via the Philippines, with lin- The above evidence comes together (table 1) to suggest that guistically close populations moving in different directions at a red-slipped plain ware tradition of clear Taiwan origin was the same time. Influences from this tradition reached the joined after 2000 BC by a very significant tradition of zoned Marianas and Lapita regions of Island Melanesia between 1500 incision with infilling by punctate or circle stamping, the and 1000 BC (fig. 3; table 1). punctate made by a multiple-toothed tool such as a tattooing Debate has recently been expressed over whether the Ne- chisel. Some sites have both plain and stamped pottery from olithic assemblages carried from Taiwan into Island Southeast the start; others appear to have an earlier horizon of plain Asia and western Oceania constituted a “package” of coher- ware only, but the picture is still obscure because so many ently related material items (Donohue and Denham 2010). If assemblages are very small and come from caves and rock there was such an Austronesian package, it was clearly poly- shelters. Perhaps the punctate and circle-stamping tradition thetic in the sense of David Clarke (1968). Rice cultivation, was introduced from mainland southern China or Hainan (Rispoli 2008; H.-C. Hung, unpublished data), although it is domesticated pigs and dogs (see below), red-slipped pottery, not possible to rule out Taiwan as the immediate source on earthenware spindle whorls, sawn and ground stone adzes present evidence because examples occur there from Yuan- with tanged or grooved butts and quadrangular to triangular shan and Yingpu contexts dating to ca. 1500–500 BC (e.g., cross sections, Taiwan slate and nephrite, notched pebble sink- Chang 1969, pls. 82D,84D). ers, stone bark cloth beaters, and a variety of shell, stone, or Similar punctate and dentate stamping is a very typical bone artifacts including bracelets, beads, and fishhooks can feature of Lapita pottery in western Melanesia (1350–750 BC; all be argued to have been carried from Taiwan into at least Green 2003) and occurs here with white lime or clay infilling the Philippines and in many cases beyond (Bellwood and of the designs, as in Luzon and the Marianas, where the Dizon 2008; Hung 2008; fig. 2). Eventually, the domestic an- greatest similarities occur (fig. 3). Punctate stamping with imals, pottery, stone adzes, bark cloth beaters, woodworking Lapita affinities is rare to absent in eastern Indonesia. In the technology, canoe construction, and fishing gear, and perhaps Lapita sites in Oceania, the stamping apparently commenced even the working of nephrite (e.g., in Maori New Zealand), everywhere at the base of the local Neolithic sequence, as in extended deep into Remote Oceania, although pottery making Luzon and possibly the Marianas, and it is likely that move- did not survive after about 2,000 years ago in Polynesia or ment through the latter islands into the Bismarck Archipelago much of Micronesia. Other items remained restricted in dis- introduced some of the Lapita decorative repertoire into Oce- tribution—artifacts of Taiwan slate, for instance, are so far ania. No claim is made that Lapita origins occurred only via reported only from Batanes, and pottery spindle whorls ap- the Marianas, because multidirectional movements through parently did not travel beyond Luzon. Indonesia are also implied by the Bukit Tengkorak obsidian My view is that no “package” concept can afford to be and could have involved populations speaking relatively un- exclusive, and there will always be a danger that people will differentiated languages still very close to PMP. read far more into the concept than is necessary. Allowance Different developments appear to have taken place in Sa- must always be made for indigenous contributions to the suite rawak, where rice husks in pottery dating from about 2200 of moving concepts and items, if and when they are required. BC onward in the cave of Gua Sireh (Beavitt, Kurui, and The problem is that such contributions are often difficult to Thompson 1996; Ipoi 1993) are associated with paddle- or comb-impressed pottery with only rare red slip and no stamp- establish, especially from pre-Austronesian Island Southeast ing. The Gua Sireh impression is paralleled in rim forms and Asia and Oceania. Hard archaeological evidence as opposed decoration in Middle Neolithic assemblages in southern Tai- to supposition does not support the claims for pre-Austro- wan (e.g., Li 1983) and Hong Kong (Meacham 1978: Sham nesian interaction networks made by Donohue and Denham Wan assemblages F and C). I have also raised the possibility (2010), for instance, or Bulbeck (2008), or in the form of that the Gua Sireh assemblage could indicate a former Aus- Terrell’s (2010) “ancient lagoons” hypothesis for early Ho- troasiatic linguistic presence in Borneo (Bellwood 1997:117, locene New Guinea. Ancient lagoons formed by high mid- 236–238), and current research on the Neolithic in southern Holocene sea levels were certainly not unique to New Guinea. Vietnam leaves this option open, especially for certain par- Populations in southeastern coastal China experienced them allels with rice chaff–tempered pottery from sites such as An as well.

Bellwood Holocene Population Prehistory in the Pacific Region S371 How Significant Were Rice and Pigs? BC and AD 1 from Gua Sireh and Lubang Angin (see above) in Sarawak and Madai and Bukit Tengkorak in Sabah. For It has been suggested that the failure of rice cultivation to Sarawak, Doherty, Beavitt, and Kurui (2000) report the dis- spread widely in perhumid equatorial eastern Indonesia or covery of rice in pottery from a total of 35 sites. Strangely, into prehistoric Oceania renders invalid any suggestion of however, Victor Paz (2002:279) identified charred remains of Neolithic movements out of Taiwan (Oppenheimer and Rich- Dioscorea alata and possibly Colocasia esculenta, but no rice, ards 2002:289). But these authors fail to note that remains of in a large quantity of late prehistoric charcoal from Madai rice and millet were universally absent from sites of the Da- Cave 1 in Sabah, even though one hearth sample from the benkeng phase in Taiwan (3500–2500 BC) until both were same site dating to ca. 2000 years ago, submitted to phytolith found in unprecedented carbonized quantities in hitherto specialist Geoff Parr, yielded rice phytoliths in quantity. Sit- unique waterlogged conditions dating to ca. 2800 BC in the uations such as this are puzzling and suggest that failure of Nanguanli sites in the Tainan Science-Based Industrial Park (Tsang 2005; Tsang, Li, and Chu 2006). In fact, the list of rice remains to survive macroscopically need not imply a total sites in Island Southeast Asia in which evidence for rice has absence. New excavations at Eluanbi in southern Taiwan have been found, particularly as a result of careful analysis of pot- also yielded rice phytoliths in a situation where charcoal is tery or phytoliths, is rapidly increasing, especially in circum- absent (T. Cheng and L. Tsuo-Ting, personal communication, stances where carbonized macroremains are absent. For ex- February 2009). As Pearsall (2003:274) notes for sites in Ec- ample, rice remains are present in pottery at Andarayan in uador, and as I have noted many times in Southeast Asian the Cagayan Valley before 1400 BC (Snow et al. 1986). Rice field conditions, charcoal can disintegrate rapidly in seasonal phytoliths have been identified in the site of Kamassi in the tropical environments with strong cycles of wetting and dry- Karama valley in western Sulawesi, but it is unclear whether ing. they relate to food production (Anggraeni, personal com- Put simply, archaeologists working in Island Southeast Asia munication). In Malaysian Borneo, rice remains including have probably failed to recognize cereal remains because of phytoliths have been reported dating variously between 2200 poor preservation conditions and lack of observational tech-

Figure 2. Artifacts from the Batanes Islands that have Taiwan affinities or actual Taiwan origins in the period 2200–800 BC. The scales apply to all items except the pottery. A, “Hoe” of volcanic rock from Torongan Cave, Itbayat, ca. 1500–2000 BC. B, Bifacially flaked hoe of volcanic rock from Anaro, Itbayat, surface find. C, Tanged adze of metamorphic rock from Sunget, Batan, 1200–800 BC (cf. Duff 1970:115 for Yuanshan par- allels and Chang 1969:165 for a Dabenkeng parallel). D, E, Taiwan slate point and point base from Anaro, surface finds (but such forms go back to the Dabenkeng early Neolithic in Taiwan; Chang 1969:54; Tsang, Li, and Chu 2006:90–91 for Youxianfang). F, Grooved adze of volcanic rock from Anaro, surface find (cf. Duff 1970:116 for Taiwan parallels). G, H, Sawn and drilled pieces of Taiwan nephrite from Anaro; these are surface finds, but similar fragments date from 1000 BC onward in the site. I, Broken adze of Taiwan nephrite, Anaro, surface find. J, K, Bone bipoint and fishing gorge (of pig canine tooth) from Anaro, first millennium BC (cf. Li 1983, pl. 102 [Eluanbi III]; Tsang, Li, and Chu 2006:175). L, Broken bark cloth beater from Anaro; this form (type III of Cameron 2006) occurs in excavated context at Nanguanli in southwest Taiwan at ca. 2800 BC (Dabenkeng phase; Tsang, Li, and Chu 2006:91). M, Notched pebble net sinker from Savidug, Sabtang Island, ca. 800 BC; this form also occurs from Dabenkeng times onward in Taiwan (Tsang, Li, and Chu 2006:113). N, O, Conus and Trochus shell bracelets (cf. Li 1983, pl. 53 [Eluanbi III, southern tip of Taiwan, ca. 1000 BC]). P, Q, Baked clay spindle whorls from Anaro, surface finds; this form occurs from Middle Neolithic on- ward in Fujian and Taiwan (Cameron 2002; Tsang, Li, and Chu 2006: 192, Sanbaozhu). R, Fine-cord-marked sherd of Taiwan Middle Neolithic type from Reranum Cave, Itbayat, ca. 1500 BC. S, Chipped operculum of Turbo marmoratus, Savidug, 1000–5000 BC (Li 1983, e.g., pl. 30, il- lustrates parallels from Taiwan preceramic onward). T, “Scoop” of Turbo marmoratus, Savidug, 1000–500 BC; for an exact parallel from Eluanbi III, see Li (1983, pl. 93). U, V, W, Batanes pottery, 1500–800 BC. (Artifacts reproduced courtesy of the National Museum of Philippines.) S372 Current Anthropology Volume 52, Supplement 4, October 2011

Table 1. Archaeological Phases of the Neolithic in Taiwan and the Northern Philippines from before 3500 BC to 500 BC

Phase Description 1 (before 3500 BC) Flaked lithics and shell tools, no evidence for Neolithic technology (such as stone sawing, grinding, use of nephrite, or pottery) 2 (3500–2200 BC) Dabenkeng and “fine-cord-marked” phases in Taiwan: appearance of Neolithic technology with use of nephrite and slate, rice and foxtail millet cultivation, and transition from cord-marked to red- slipped plain ware pottery. No known expansion south of Taiwan occurred at this time, but Sum- merhayes and Anderson (2009) raise the possibility of movement into the nearby Ryukyu Islands, southern Japan 3 (2200–1000 BC) First human settlement of Batanes and Luzon from southern Taiwan, about 2200 BC, followed by a considerable flow of material culture, especially red-slipped plain ware pottery, from Taiwan into the Philippines and onward into Island Southeast Asia. Associated with this was a spread of stamped pottery decoration from ultimate mainland Asian sources, which traveled with the red- slipped plain ware into the Philippines and onward to the Mariana Islands in western Micronesia and the Lapita zone of Island Melanesia and western Polynesia. This spread appears to have had only limited impact in Taiwan but a major impact in the northern Philippines 4 (first millennium BC) Increasing evidence for frequent Taiwan-Batanes-Philippines and across-the-South-China-Sea contacts in both directions, particularly involving Taiwan nephrite (Hung and Bellwood 2010; Hung et al. 2007) nology. Furthermore, the above-listed Borneo sites, apart and Robertson 1903–1909, vol. 38:98), we would not nec- from Gua Sireh, are far from fertile rice-growing terrain and, essarily expect to find it, even though it is grown today as a in the case of the Niah Caves (Barker 2005), supported a minor monsoon crop in upland fields. The Batanes Islands continuing hunter-gatherer population (Punan) until the Iban have no flat alluvial land and no high-level water sources that incursions of the nineteenth century. It is likely that many of could support terrace farming in the manner of the famous the Neolithic and Iron Age burials in the upper Niah stra- Ifugao terraces at Banaue in northern Luzon. My suspicion tigraphy were of people not native to the immediate area is that the Batanes formed a filter against a spread of rice in (Valentine, Kamenov, and Krigbaum 2008), and so the traces the early years of Austronesian expansion and that it became of rice found in their pots need cause no surprise. Caves in significant only when settlers reached the broad alluvial land- marginal agricultural terrain such as Niah are of questionable scapes of the Cagayan Valley and other parts of Luzon. Within relevance for any discussion of agricultural origins in Island much of the Philippines and Indonesia, Neolithic populations Southeast Asia. Real data on agricultural prehistory are far in mid-Holocene-drowned coastal landscapes (Bellwood et al. more likely to come from waterlogged alluvial and coastal 2008) that lacked good alluvial rice soils moved expectably sites such as those recently discovered in Taiwan, but such toward fruits and tubers for subsistence and doubtless en- are at present undiscovered in the islands to the south. countered these species when they were already under ex- The failure of rice cultivation to spread into Oceania is not ploitation by indigenous food gatherers. hard to explain. Dewar (2003) discusses reasons of climatic The domestication of the pig in Island Southeast Asia is variability for a relative absence (or nonsignificance) of rice currently a topic of considerable debate (Larson et al. 2007, in many parts of northern and eastern Island Southeast Asia, 2010) rendered complex by the wide distribution of native suggesting that El Nin˜o–Southern Oscillation (ENSO)-related suids in mainland Asia, western Indonesia (Sundaland), and rainfall unreliability was a major reason for its failure to spread Sulawesi. Pig bones are widespread in Neolithic sites in Taiwan into Oceania. Paz (2002), as noted above, refers to day-length and are common in the lower Neolithic layer at Nagsabaran considerations. Island Southeast Asia encompasses 30Њ of lat- in the Cagayan Valley, where Piper has identified teeth of itude, from subtropical regions with strong seasonality of rain- domesticated Sus scrofa directly AMS dated before 2000 BC fall distribution into equatorial regions that were uniformly (Piper et al. 2009). These pigs predate, perhaps by many cen- hot and perhumid all year round. We can hardly expect that turies, the introduction of the so-called Pacific clade of pigs people would always have continued to grow rice when more of southern China or northern Indochina origin into Lapita suitable fruits and tubers were available, just as farming pop- Melanesia. Unfortunately, the oldest Batanes sites, Reranum ulations moving from the Indus into the Ganga Valley at about and Torongan Caves, contain no animal bone, but pig was 3000 BC eventually adopted rice and other monsoon crops present by 1200 BC in Sunget on Batan. and allowed their West Asian winter crops to decrease in At this stage, it is not clear to what degree pigs traveled significance (Bellwood 2005:87; Fuller 2011). So far, our at- with Neolithic migrants in Island Southeast Asia or how many tempts to identify rice in phytolith samples from Batanes have separate domestications of local species occurred (Larson not been successful, but because rice was apparently not 2011). It seems that pigs were not carried from Luzon to the grown at European contact on Batan (Dampier 1687, in Blair Marianas in prehistoric times, so they did not enter the Lapita Bellwood Holocene Population Prehistory in the Pacific Region S373

Figure 3. Pottery surface decoration in Southeast Asia, 2500–1000 BC. Inset motifs of incised and stamped decoration on sherds: A, Nagsabaran, northern Luzon (courtesy Hsiao-chun Hung); B, Achugao, Saipan (drawn from a photo provided by Brian Butler; see Butler 1995, fig. 12-6b); C, Lapita, New Caledonia (redrawn with permission from Sand 1999:46). zone by this route. It is also quite possible that pre-Neolithic Academia Sinica, Taipei, using a low-vacuum scanning elec- populations translocated indigenous wild pig species to tron microscope equipped with an energy-dispersive x-ray resource-poor islands such as Flores in eastern Indonesia spectrometer (Iizuka and Hung 2005; Iizuka et al. 2005). It (Dobney et al. 2008), just as they might have translocated can be identified with confidence in terms of the chemistry marsupials out of New Guinea (Flannery et al. 1998). of its matrix and its zinc-chromite inclusion minerals. All green jade artifacts tested from Taiwan and the Philippines The Nephrite Trail are from the Fengtian source (but Luzon also has at least one separate white nephrite source). In Taiwan, nephrite (jade) tools and ornaments have been Jade working was most probably introduced into Taiwan identified from more than 100 sites dating between 3000 BC from southern China, where it was present in the Yangzi Basin and AD 500 (Hung et al. 2007). Taiwan nephrite is generally green in color and was exploited from deposits at Fengtian, as early as 5000 BC. In Taiwan, sawn nephrite adzes appear located at the northern end of the eastern rift valley of Taiwan in the Dabenkeng phase (ca. 3000 BC; Hung 2004), and the inland from the city of Hualian. This appears to have been long-lasting tradition of grooving, snapping, drilling, and pol- the only nephrite source utilized in prehistoric Taiwan. Feng- ishing nephrite later developed into the remarkable funerary tian nephrite has recently been subjected to a detailed sourcing assemblages of Beinan in southeastern Taiwan (ca. 1500–500 program by Yoshiyuki Iizuka at the Institute of Earth Sciences, BC), with pendants (some anthropomorphic), penannular S374 Current Anthropology Volume 52, Supplement 4, October 2011 earrings (some with four circumferential projections), bell- least 3000 BC in cord-marked and red-slipped pottery, and shaped and tubular beads, perforated arrowheads and spear- many of the items are illustrated in figure 2. To these can be heads, and adzes (Lien 2002). Worked nephrite occurs with added the oldest radiocarbon dates for rice and millet, do- the red-slipped pottery referred to above from Chaolaiqiao mesticated dogs, and probably pigs in Southeast Asia. (3) in southeastern Taiwan (2200 BC), and it seems likely that There is an absence of closely related Neolithic material cul- some of the earliest Neolithic settlers from Taiwan could have ture before 1500 BC in Indonesia, and there are deep and carried it to the northern Philippines. significant differences in most aspects of Neolithic (pre–Sa The oldest sites in Batanes (Reranum and Torongan caves) Huynh) material culture between Vietnam and the Philippines so far lack nephrite, but an adze of Fengtian nephrite was before 1000 BC. (4) The absence of an earlier population in found at Sunget (ca. 1200 BC), and several surface finds of the Batanes Islands implies a movement of people to establish Taiwan sawn nephrite stone adzes come from Anaro on It- colonization, not an adoption of Neolithic material culture bayat Island. Stone adzes of other metamorphic rocks from by an indigenous hunter-gatherer population (as probably Batanes and Luzon also show a preference for sawing rather happened to a degree in the Pen˜ablanca Caves in Luzon; than flaking in manufacture. More significantly, in terms of Mijares 2006). date, parts of two Fengtian nephrite bracelets have been re- From an archaeological perspective, the progression of covered from the base of Nagsabaran (ca. 2000–1500 BC) in Neolithic material culture assemblages of ultimate East Asian/ the Cagayan Valley. In Taiwan, this form is most commonly Taiwan origin through the regions settled by ancestral Aus- found in the Middle Neolithic phase of fine-cord-marked tronesian speakers required about 4,000 years to unfold from pottery before 2000 BC. For instance, 25 were recovered, in Taiwan to New Zealand, perhaps 6,000 years if commencing some cases in association with burials, from the Niuchouzi in southern China (fig. 3). Populations already resident in phase site of Youxianfang near Tainan in southwestern Tai- Island Southeast Asia and Melanesia contributed cultural cap- wan. This site is radiocarbon dated between 1850 and 1350 ital in the form of some shell-artifact technologies (especially BC (Tsang, Li, and Chu 2006) and is thus roughly contem- flaked-shell tools), tuber and fruit crops of western Pacific porary with the lower layer in Nagsabaran. Such bracelets are (especially New Guinea) origin, flaked lithic traditions (found uncommon in the post–1500 BC Beinan culture (Lien 2002). commonly mixed with Neolithic assemblages in Indonesian In later times, a flourishing movement of ornaments of caves), and even translocated species of marsupials in some Fengtian nephrite linked a number of Iron Age (ca. 300 BC– islands close to New Guinea. As to the origins of the economic AD 500) sites throughout the Philippines with Sarawak, and technological complexes that fueled Austronesian dis- southern Vietnam, peninsular Thailand, and possibly eastern persal in the first instance, we must look to southern coastal Cambodia (Hung et al. 2007). This later trade/exchange net- China (Fuller, Harvey, and Qin 2007; Hung 2008; Jiao 2007; work is not of direct concern for the earlier phases of Aus- Rispoli 2008; Zhang and Hung 2008, 2010). Food production, tronesian dispersal, but it does emphasize that Austronesian- maritime knowledge, and perhaps even domino effects from speaking communities never lost contact with Taiwan and in the movements of Yangzi Basin rice farmers all played roles this regard is very significant. Fengtian nephrite was exported here. What is clear is that the Neolithic complex that spread for upward of 2,000 years through the Philippines and onward through Island Southeast Asia and into the Pacific emanated to further regions on a fairly regular basis and presumably from southern China via Taiwan and then presumably via the by people who mostly shared an Austronesian linguistic her- Philippines and Borneo. It manifestly did not originate from itage. the south. In conclusion, it is possible to focus on Taiwan and the Out of Taiwan into the Philippines: Philippines at the end of the third millennium BC in order A Summary to identify the start of a migration, by both land and sea, of speakers of Malayo-Polynesian languages and of the carriers Four factors render a southward movement of Neolithic ma- of a broad range of Neolithic material culture with food pro- terial culture from Taiwan into the northern Philippines at duction. Both of these spreads are likely to have been two about 2000 BC a virtual certainty. (1) Strong parallels in ma- sides of one “event” involving a single ethnolinguistic and terial culture between 2200 and 1500 BC link southern Taiwan genetic population in the final resort, albeit one that was and the northern Philippines, reinforced by the movement constantly adapting and interacting, as all humans do. The of artifacts of Taiwan slate and positively sourced Taiwan migration originated more deeply in southern China and nephrite. The synchronous changes in pottery decoration that eventually encompassed warm temperate through equatorial link southeast Taiwan, the Batanes Islands, and Luzon, from latitudes in the Northern Hemisphere, out again into tem- a predominance of cord marking in Taiwan only, through a perate latitudes (New Zealand) in the Southern Hemisphere, dominance of red-slipped pottery and into a slightly later thus passing through some major zones of environmental and phase of red-slipped and stamped pottery are especially im- resource difference as well as through preexisting populations portant. (2) Taiwan has chronological priority of the artifact with their own long-established cultures and languages, par- types concerned involving an unbroken continuity since at ticularly in western Oceania. One of the results of this equa- Bellwood Holocene Population Prehistory in the Pacific Region S375 torial transition appears to have been an abandonment of uct of forager activity in the archipelagoes of Island Southeast cereal (rice and millet) cultivation in Indonesia in favor of Asia during the early Holocene millennia of sea-level rise. the fruit-and-tuber-based plant economy that characterized Neither could it have been caused only by maritime knowl- the Pacific Islands. edge or unusual concatenations of ENSO events (Anderson Why did the Austronesian dispersal occur? Within eastern et al. 2006). Such factors no doubt helped, and of course Taiwan, the archaeological record indicates a marked increase without their rafts or canoes, Austronesians would presum- in the number of archaeological sites after 2500 BC (Hung ably not have traveled far. But without the demographic im- 2008), so population growth and a need for new cultivation petus and technological advancement provided by East Asian land come to mind, given that southeastern Taiwan is a rugged food production, this dispersal could never have occurred, at area with low agricultural potential. But early Austronesians least not through the inhabited portions of Island Southeast moved on to settle new islands very rapidly in terms of both Asia. The previously empty islands in the Pacific would in archaeology and comparative linguistics (Pawley 1999). For many cases have been too small, isolated, and impoverished instance, colonists spread 8,000 km from the Batanes Islands in terrestrial resources for long-term low-technology forager to Samoa in less than 1,000 years, beyond which a slowdown settlement. Early Jomon people in Japan and early Australians occurred, perhaps due in part to the nonemergence of Pacific had rudimentary sea craft in some regions at least, but as far atolls until much later in time (Dickinson 2003). This rapid as we know, their forager economies did not lead to maritime movement to western Polynesia surely reflected a reliance on population dispersal over large distances. Neither of these both maritime and lowland agricultural resources, the latter populations attempted to colonize the Pacific. greatly reduced in extent by the drowning of the most fertile In conclusion, the Austronesian dispersal was little different alluvial and coastal soils as the sea attained its maximum mid- in a demographic and causative sense from other early farm- Holocene sea level (Bellwood et al. 2008). This would have ing dispersals across the Eurasian, African, and American con- rendered good coastal and alluvial farmland scarce in the early tinents. The maritime setting is unusual, but we have no data centuries of Austronesian migration, creating deep estuaries to suggest that Austronesians ever entirely gave up food pro- and steep coastlines against rugged island interiors, at least duction, except in equatorial rainforests in Borneo and frosty until forest clearance caused soil aggradation to build up fer- southern New Zealand. Doubtless, Austronesian food pro- tile lowlands. Advancing maritime technology also fueled the duction varied in intensity—many Austronesians are low-level Austronesian spread, with the earliest evidence for canoes and food producers in Bruce Smith’s (2001, 2011) terms today— paddles in this region coming from coastal central China but this need not be a universally ancestral condition. Inability during the early Holocene. to find prolific evidence for cultivated crops in archaeological sites does not mean that we must throw out all the compar- ative linguistic reconstructions of an early Austronesian ag- Can the Austronesian Dispersal Inform ricultural vocabulary. Naive birds, fishing, and westerly winds Us about Dispersals of Languages and doubtless assisted many individual episodes of colonization Farming in Other Parts of the World? but alone did not lead to Pacific colonization beyond Near Oceania. The reality of Austronesian dispersal via Taiwan and the Phil- ippines is powerfully reinforced by evidence from three in- dependent research areas: linguistics, archaeology, and human biology. This dispersal did not involve an autonomous origin Acknowledgments of food production on present evidence, which admittedly is The reported archaeological research in the Batanes Islands, slight; rather, it was a peripheral result of the developments Cagayan Valley, and southern Vietnam was funded by the of food production farther north in mainland East Asia and Australian Research Council, the Wenner-Gren Foundation farther east in New Guinea. There is much uncertainty here, for Anthropological Research, and the Chiang Ching-Kuo and little mileage would be gained by debating whether Aus- Foundation (Taipei). It involved crews from the National Mu- tronesian languages or peoples can be traced directly into the seum of the Philippines, University of the Philippines, Center Yangzi Basin. Domino effects could have occurred, even lo- for Archaeological Studies (Ho Chi Minh City), and Austra- calized language shift (because language shift was surely al- lian National University. A monograph on the Batanes re- ways localized under Neolithic social conditions), as well as search is in preparation, to be published by ANU E Press. direct transmission. What is clear, however, is that the Austronesian dispersal References Cited to Taiwan and beyond between 3000 and 1000 BC occurred at the same time that agricultural populations in China and Addison, D., and E. Matisoo-Smith. 2010. Rethinking Polynesian origins: a West Polynesia triple-I model. Archaeology in Oceania Taiwan attained very high demographic densities, especially 45:1–12. during the Longshan and Shang periods in China itself. Aus- Ammerman, A., and L. L. Cavalli-Sforza. 1973. A population model tronesian dispersal did not originate as an independent prod- for the diffusion of early farming in Europe. In The explanation S376 Current Anthropology Volume 52, Supplement 4, October 2011

of culture change. C. Renfrew, ed. Pp. 343–357. London: Duck- Bulbeck, D. 2008. An integrated perspective on the Austronesian worth. diaspora. Australian Archaeology 67:31–52. Anderson, A., J. Chappell, M. Gagan, and R. Grove. 2006. Prehistoric Butler, B. 1995. Archaeological investigations in the Achugao and Ma- maritime migration in the Pacific Islands: an hypothesis of ENSO tansa areas of Saipan, Mariana Islands. Micronesian Archaeological forcing. Holocene 16:1–6. Survey report no. 30. Saipan: Micronesian Archaeological Survey. Barker, G. 2005. The archaeology of foraging and farming at Niah Cameron, J. 2002. Textile technology and Austronesian dispersals. In Cave, Sarawak. Asian Perspectives 44:90–106. The archaeology of Lapita dispersal in Oceania. G. Clark, A. An- Beavitt, P., E. Kurui, and G. Thompson. 1996. Confirmation of an derson, and T. Vunidilo, eds. Pp. 177–181. Canberra: Pandanus. early date for the presence of rice in Borneo. Borneo Research ———. 2006. Trans-oceanic transfer of bark-cloth technology from Bulletin 27:29–37. South China–Southeast Asia to Mesoamerica? In Islands of enquiry. Bellwood, P. 1989. Archaeological investigations at Bukit Tengkorak G. Clark, F. Leach, and S. O’Connor, eds. Pp. 203–210. Canberra: and Segarong, southeastern Sabah. Bulletin of the Indo-Pacific Pre- Australian National University E Press. history Association 9:122–162. Carson, M. 2008. Refining earliest settlement in Remote Oceania. ———. 1996. Phylogeny and reticulation in prehistory. Antiquity 70: Journal of Island and Coastal Archaeology 3:115–139. 881–890. Chagnon, A. 1992. Yanomamo¨. Fort Worth, TX: Harcourt Brace Jov- ———. 1997. Prehistory of the Indo-Malaysian Archipelago. Rev. edi- anovich. tion. Honolulu: University of Hawaii Press. Chang, K. C. 1969. Fengpitou, Tapenkeng and the prehistory of Taiwan. ———. 2001. Early agriculturalist population diasporas? farming, Publications in Anthropology 73. New Haven, CT: Yale University languages and genes. Annual Review of Anthropology 30:181–207. Press. ———. 2005. First farmers. Oxford: Blackwell. Chia, S. 2003. The prehistory of Bukit Tengkorak. Sabah Museum ———. 2008. Archaeology and the origins of language families. In Monograph 8. Kota Kinabalu: Sabah Museum. Handbook of archaeological theories. A. Bentley, H. Maschner, and Clark, G., F. Petchey, O. Winter, M. Carson, and P. O’Day. 2010. C. Chippindale, eds. Pp. 225–243. Lanham, MD: Altamira. New radiocarbon dates from the Bapot-1 site in Saipan and Neo- ———. 2009a. La diffusion des populations d’agriculteurs dans le lithic dispersal by stratified diffusion. Journal of Pacific Archaeology monde. In La re´volution ne´olithique dans le monde. Jean-Paul De- 1(1):21–35. moule, ed. Pp. 239–262. Paris: CNRS. Clarke, D. 1968. Analytical archaeology. London: Methuen. ———. 2009b. The dispersals of established food-producing pop- Cohen, D. J. 2011. The beginnings of agriculture in China: a mul- ulations. Current Anthropology 50:621–626. tiregional view. Current Anthropology 52(suppl. 4):S273–S293. Bellwood, P., G. Chambers, M. Ross, and H.-C. Hung. Forthcoming. Cox, M. 2008. The genetic environment of Melanesia: clines, clusters Are cultures inherited? multidisciplinary perspectives on the ori- and contact. In Population genetics research progress.V.T.Koven, gins and migrations of Austronesian speaking peoples prior to ed. Pp. 45–83. New York: Nova Science. 1000 BC. In Investigating archaeological cultures: material culture, Cox, M., T. M. Karafet, J. S. Lansing, H. Sudoyo, and M. F. Hammer. variability and transmission. B. Roberts and M. van der Linden, 2010. Autosomal and X-linked single nucleotide polymorphisms eds. Dordrecht: Springer. reveal a steep Asian-Melanesian ancestry cline in eastern Indonesia. Bellwood, P., and E. Dizon. 2005. The Batanes archaeological project Proceedings of the Royal Society B 277:1589–1596. and the out of Taiwan hypothesis for Austronesian dispersal. Jour- Denham, T. 2011. Early agriculture and plant domestication in New nal of Austronesian Studies 1:1–33. Guinea and Island Southeast Asia. Current Anthropology 52(suppl. 4):S379–S395. ———. 2008. Austronesian cultural origins: out of Taiwan, via the Dewar, R. 2003. Rainfall variability and subsistence systems in South- Batanes Islands, and onwards to western Polynesia. In Past human east Asia and the western Pacific. Current Anthropology 44:369– migrations in East Asia: matching archaeology, linguistics and ge- 388. netics. Alicia Sanchez-Mazas, Roger Blench, Malcolm D. Ross, Ilia Dickinson, W. 2003. Impact of mid-Holocene hydro-isostatic high- Peiros, and Marie Lin, eds. Pp. 23–39. London: Routledge. stand in regional sea level on habitability of islands in Pacific Bellwood, P., and M. Oxenham. 2008. The expansions of farming Oceania. Journal of Coastal Research 19:489–502. societies and the role of the Neolithic demographic transition. In Dobney, K., A. Ervynck, U. Albarella, and P. Rowley-Conwy. 2008. The Neolithic demographic transition and its consequences. Jean- The transition from wild boar to domestic pig in Eurasia. In Pigs Pierre Bocquet-Appel and Ofer Bar-Yosef, eds. Pp. 13–34. Dor- and humans: 10,000 years of interaction. U. Albarella, K. Dobney, drecht: Springer. A. Ervynck, and P. Rowley-Conwy, eds. Pp. 57–82. Oxford: Oxford Bellwood, P., J. Stevenson, E. Dizon, A. Mijares, G. Lacsina, and E. University Press. Robles. 2008. Where are the Neolithic landscapes of Ilocos Norte? Doherty, C., P. Beavitt, and E. Kurui. 2000. Recent observations of Hukay 15:25–38. rice temper in pottery from Niah and other sites in Sarawak. Blair, E. H., and J. A. Robertson. 1903–1909. The Philippine Islands Bulletin of the Indo-Pacific Prehistory Association 20:147–152. 1493–1898, 55 vols. Cleveland, OH: Clark. Donohue, M., and T. Denham. 2010. Farming and language in Island Blust, R. 1977. The proto-Austronesian pronouns and Austronesian Southeast Asia: reframing Austronesian history. Current Anthro- subgrouping: a preliminary report. Working Papers in Linguistics, pology 51:223–256. University of Hawai’i 9(2):1–15. Duff, R. 1970. Stone adzes of Southeast Asia. Christchurch: Canterbury ———. 1999. Subgrouping, circularity and extinction: some issues Museum. in Austronesian comparative linguistics. In Selected papers from Flannery, T. F., P. Bellwood, J. P. White, T. Ennis, G. Irwin, K. Schu- the eighth international conference on Austronesian linguistics.E. bert, and S. Balasubramanian. 1998. Mammals from Holocene Zeitoun and P. J.-K. Li, eds. Pp. 31–94. Taipei, Taiwan: Academia archaeological deposits on Gebe and Morotai islands, Northern Sinica. Moluccas, Indonesia. Australian Mammalogy 20(3):391–400. ———. 2009. The Austronesian languages. Canberra: Pacific Lin- Freeman, D. 1970. Report on the Iban. London: Athlone. guistics. Friedlaender, J., F. Friedlaender, F. A. Reed, K. K. Kidd, J. R. Kidd, Bocquet-Appel, J.-P. 2011. The agricultural demographic transition G. K. Chambers, R. A. Lea, et al. 2008. The genetic structure of during and after the agriculture inventions. Current Anthropology Pacific Islanders. PLoS Genetics 4(1):e19, doi:10.1371/journal 52(suppl. 4):S497–S510. .pgen.0040019. Bellwood Holocene Population Prehistory in the Pacific Region S377

Fuller, D., E. Harvey, and L. Qin. 2007. Presumed domestication? more Asian than Melanesian ancestry of Polynesians. American Antiquity 81:316–331. Journal of Human Genetics 82:194–198. Fuller, D. Q. 2011. Finding plant domestication in the Indian sub- Kelly, R. 1985. The Nuer conquest. Ann Arbor: University of Michigan continent. Current Anthropology 52(suppl. 4):S347-S362. Press. Gray, R., A. Drummond, and S. Greenhill. 2009. Language phylog- Kimura, R., J. Ohashi, Y. Matsumura, M. Nakazawa, T. Inaoka, R. enies reveal expansion pulses and pauses in Pacific settlement. Ohtsuka, M. Osawa, and K. Tokunaga. 2008. Gene flow and natural Science 323:479–483. selection in oceanic human populations inferred from genome- Green, R. 2003. The Lapita horizon and traditions: signature for one wide SNP typing. Molecular Biology and Evolution 25:1750–1761. set of Oceanic migrations. In Pacific archaeology: assessments and Lansing, S., T. Karafet, J. Schoenfelder, and M. Hammer. 2008. A prospects. C. Sand, ed. Pp. 95–120. Le cahiers de l’arche´ologie en DNA signature for the expansion of irrigation in Bali? In Past Nouvelle-Cale´donie, vol. 15. Noume´a, New Caledonia: Departe- human migrations in East Asia: matching archaeology, linguistics ment Arche´ologie, Service des muse´es et du patrimoine de Nou- and genetics. Alicia Sanchez-Mazas, Roger Blench, Malcolm D. velle-Cale´donie. Ross, Ilia Peiros, and Marie Lin, eds. Pp. 376–394. London: Rout- Heggerty, P., and D. Beresford-Jones. 2010. Agriculture and language ledge. dispersals: limitations, refinements, and an Andean exception? Larson, G. 2011. Genetics and domestication: important questions Current Anthropology 51:163–192. for new answers. Current Anthropology 52(suppl. 4):S485–S495. Hung, H.-C. 2004. A sourcing study of Taiwan stone adzes. Bulletin Larson, G., T. Cucchi, M. Fujita, E. Matisoo-Smith, J. Robins, A. of the Indo-Pacific Prehistory Association 24:57–70. Anderson, B. Rolett, et al. 2007. Phylogeny and ancient DNA of ———. 2005. Neolithic interaction between Taiwan and northern Sus provides insights into Neolithic expansion in Island Southeast Luzon: the pottery and jade evidences from the Cagayan Valley. Asia and Oceania. Proceedings of the National Academy of Sciences Journal of Austronesian Studies 1(1):109–134. of the USA 104(12):4834–4839. ———. 2008. Migration and cultural interaction in southern coastal Larson, G., R. Liu, X. Zhao, J. Yuan, D. Fuller, L. Barton, K. Dobney, China, Taiwan and the Northern Philippines, 3000 BC to AD 1. et al. 2010. Patterns of East Asian pig domestication, migration, PhD thesis, Australian National University, Canberra. and turnover revealed by modern and ancient DNA. Proceedings Hung, H.-C., and P. Bellwood. 2010. Movement of raw materials of the National Academy of Sciences of the USA 107(17):7686–7691. and manufactured goods across the South China Sea after 500 Li, K.-C. 1983. Report of archaeological investigations in the O-Luan- BCE: from Taiwan to Thailand, and back. In 50 years of archaeology Pi Park at the southern tip of Taiwan. Taipei: Department of An- in Southeast Asia: essays in honour of Ian Glover. B. Bellina, L. thropology, National Taiwan University. [In Chinese.] Bacus, O. Pryce, and J. Wisseman Christie, eds. Pp. 234–243. Bang- Lien, Chao-mei. 2002. The jade industry of Neolithic Taiwan. Bulletin kok: River. of the Indo-Pacific Prehistory Association 22:55–62. Hung, H-C., M. Carson, P. Bellwood, F. Campos, P. Piper, E. Dizon, Meacham, W., ed. 1978. Sham Wan, Lamma Island. Journal mono- M. Bolunia, M. Oxenham, and C. Zhang. Forthcoming. The first graphs 3. Hong Kong: Hong Kong Archaeological Society. settlement of Remote Oceania: Luzon to the Marianas. Antiquity. Mijares, A. S., F. De´troit, P. Piper, R. Gru¨n, P. Bellwood, M. Aubert, Hung, H.-C., Y. Iizuka, P. Bellwood, K. D. Nguyen, B. Bellina, P. G. Champion, N. Cuevas, A. De Leon, and E. Dizon. 2010. New Silapanth, E. Dizon, R. Santiago, I. Datan, and J. Manton. 2007. evidence for a 67,000-year-old human presence at Callao Cave, Ancient jades map 3000 years of prehistoric exchange in Southeast Luzon, Philippines. Journal of Human Evolution 59:123–132. Asia. Proceedings of the National Academy of Sciences of the USA Mijares, A. S. B. 2006. The early Austronesian migration to Luzon: 104:19745–19750. perspectives from the Pen˜ablanca cave sites. Bulletin of the Indo- Hunley, K., M. Dunn, E. Lindstro¨m, G. Reesink, A. Terrill, M. E. Pacific Prehistory Association 26:72–78. Healy, G. Koki, et al. 2008. Genetic and linguistic coevolution in Mona, S., K. E. Grunz, S. Brauer, B. Pakendorf, L. Castrı`, H. Sudoyo, Northern Island Melanesia. PLoS Genetics 4(10):e1000239, doi: S. Marzuki, et al. 2009. Genetic admixture history of eastern In- 10.1371/journal.pgen.1000239. donesia as revealed by Y-chromosome and mitochondrial DNA Iizuka, Y., P. Bellwood, H.-C. Hung, and E. Dizon. 2005. A non- analysis. Molecular Biology and Evolution 26:1865–1877. destructive mineralogical study of nephritic artifacts from Itbayat Mona, S., M. Tomasetto-Ponzetta, S. Brauer, H. Sudoyo, S. Marzuki, Island, Batanes, northern Philippines. Journal of Austronesian Stud- and M. Kayser. 2007. Patterns of Y-chromosome diversity intersect ies 1(1):83–108. with the trans-New Guinea hypothesis. Molecular Biology and Evo- Iizuka, Y., and H.-C. Hung. 2005. Archaeomineralogy of Taiwan lution 24:2546–2555. nephrite: sourcing study of nephrite artifacts from the Philippines. Moodley, Y., B. Linz, Y. Yamaoka, H. M. Windsor, S. Breurec, J.-Y. Journal of Austronesian Studies 1(1):35–80. Wu, A. Maady, et al. 2009. The peopling of the Pacific from a Ipoi, D. 1993. Archaeological excavations at Gua Sireh (Serian) and bacterial perspective. Science 323:527–530. Lubang Angin (Gunung Mulu National Park), Sarawak, Malaysia. Ogawa, H. 2005. Typological chronology of pottery assemblages from Special monograph 6, Sarawak Museum Journal. the Lal-Lo shell midden in northern Luzon, Philippines. Journal Jiao, T. 2007. The Neolithic of southeast China. Youngstown, NY: of Southeast Asian Archaeology 25:1–30. Cambria. Oppenheimer, S., and M. Richards. 2002. Polynesians: devolved Tai- Jing, Y., and R. Campbell. 2009. Recent archaeometric research on wanese rice farmers or Wallacean maritime traders with fishing, “the origins of Chinese civilisation.” Antiquity 83:96–109. foraging and horticultural skills. In Examining the farming/lan- Kayser, M. 2010. The human genetic history of Oceania: near and guage dispersal hypothesis. P. Bellwood and C. Renfrew, eds. Pp. remote views of dispersal. Current Biology 20(4):R194–R201, doi: 287–298. Cambridge: McDonald Institute for Archaeological Re- 10.1016/j.cub.2009.12.004. search. Kayser, M., Y. Choi, M. van Oven, S. Mona, S. Brauer, R. J. Trent, Ostler, N. 2005. Empires of the word. London: Harper Perennial. D. Suarkia, et al. 2008a. The impact of the Austronesian expansion: Pawley, A. 1999. Chasing rainbows: implications of the rapid dispersal evidence from mtDNA and Y chromosome diversity in the Ad- of Austronesian languages for subgrouping and reconstruction. In miralty Islands of Melanesia. Molecular Biology and Evolution 25: Selected papers from the Eighth International Conference on Aus- 1362–1374. tronesian Linguistics. E. Zeitoun and P. J.-K. Li, eds. Pp. 95–138. Kayser, M., O. Lau, K. Saar, S. Brauer, X. Wang, P. Nu¨rnberg, R. J. Taipei: Institute of Linguistics, Academia Sinica. Trent, and M. Stoneking. 2008b. Genome-wide analysis indicates ———. 2002. The Austronesian dispersal: languages, technologies S378 Current Anthropology Volume 52, Supplement 4, October 2011

and people. In Examining the farming/language dispersal hypothesis. the Neolithic of the Karama River. In Austronesian in Sulawesi.T. P. Bellwood and C. Renfrew, eds. Pp. 251–273. Cambridge: Mac- Simanjuntak, ed. Pp. 57–76. Depok, Indonesia: Center for Pre- Donald Institute for Archaeological Research. historic and Austronesian Studies. ———. 2007. Recent research on the historical relationships of the Smith, B. D. 2001. Low-level food production. Journal of Archaeo- Papuan languages. In Genes, language and culture history in the logical Research 9:1–43. Southwest Pacific. J. Friedlaender, ed. Pp. 36–60. Oxford: Oxford ———. 2011. The cultural context of plant domestication in eastern University Press. North America. Current Anthropology 52(suppl. 4):S471–S484. Paz, V. 2002. Island Southeast Asia: spread or friction zone? In Ex- Snow, B., R. Shutler, D. Nelson, J. Vogel, and J. Southon. 1986. amining the farming/language dispersal hypothesis. P. Bellwood and Evidence of early rice cultivation in the Philippines. Philippine C. Renfrew, eds. Pp. 275–285. Cambridge: McDonald Institute for Quarterly of Culture and Society 14:3–11. Archaeological Research. Summerhayes, G., and A. Anderson. 2009. An Austronesian presence Pearsall, D. 2003. Plant food resources of the Ecuadorian Formative: in southern Japan: early occupation in the Yaeyama Islands. Bul- an overview and comparison to the central Andes. In Archaeology letin of the Indo-Pacific Prehistory Association 29:76–91. of Formative Ecuador. J. S. Raymond, R. L. Burger, and J. Quilter, Tabbada, K. A., J. Trejaut, J.-H. Loo, Y.-M. Chen, M. Lin, M. Mirazo´n- eds. Pp. 213–258. Washington, DC: Dumbarton Oaks. Lahr, T. Kivisild, and M. C. A. De Ungria. 2010. Philippine mi- Piper, P., H.-C. Hung, F. Z. Campos, P. Bellwood, and R. Santiago. tochondrial DNA diversity: a populated viaduct between Taiwan 2009. A 4000 year old introduction of domestic pigs into the and Indonesia? Molecular Biology and Evolution 27:21–31. Philippine Archipelago. Antiquity 83:687–695. Terrell, J. 2010. Social network analysis of the genetic structure of Renfrew, C. 2002. “The emerging synthesis”: the archaeogenetics of Pacific Islanders. Annals of Human Genetics 74:211–232. language/farming dispersals and other spread zones. In Examining Tsang, C.-H. 2005. Recent discoveries at a Tapenkeng culture site in the farming/language dispersal hypothesis. P. Bellwood and C. Ren- Taiwan: implications for the problem of Austronesian origins. In frew, eds. Pp. 3–16. Cambridge: McDonald Institute for Archae- The peopling of East Asia. L. Sagart, R. Blench, and A. Sanchez- ological Research. Mazas, eds. Pp. 63–73. London: RoutledgeCurzon. Rispoli, F. 2008. The incised and impressed pottery of mainland Southeast Asia: following the paths of Neolithization. East and Tsang, C.-H., K.-T. Li, and C.-Y. Chu. 2006. Footprints of ancestors: West 57:235–304. archaeological discoveries in Tainan Science-Based Industrial Park. Ross, M. 2008. The integrity of the Austronesian language family: Tainan: Tainan County Government. [In Chinese.] from Taiwan to Oceania. In Past human migrations in East Asia: Valentine, B., G. Kamenov, and J. Krigbaum. 2008. Reconstructing matching archaeology, linguistics and genetics. A. Sanchez-Mazas, R. Neolithic groups in Sarawak, Malaysia through lead and strontium Blench, M. D. Ross, I. Peiros, and M. Lin, eds. Pp. 161–181. Lon- isotope analysis. Journal of Archaeological Science 35:1463–1473. don: Routledge. Yuan, J., R. Flad, and L. Yunbing. 2008. Meat acquisition patterns Ross, M., A. Pawley, and M. Osmond. 2007. The lexicon of proto- in the Neolithic Yangzi River valley, China. Antiquity 82:351–366. Oceanic: the culture and environment of ancestral Oceanic society. Zhang, C., and H.-C. Hung. 2008. The Neolithic of southern China: 2nd edition. Canberra: Australian National University E Press. origin, development and dispersal. Asian Perspectives 47:299–329. Sand, C. 1999.Arche´ologie des origines: le lapita cale´donien. Noume´a, ———. 2010. The emergence of agriculture in southern China. An- New Caledonia: Departement Arche´ologie, Service des muse´es et tiquity 84:11–25. du patrimoine de Nouvelle-Cale´donie. Zhao, Z. 2011. New archaeobotanic data for the study of the origins Simanjuntak, T., M. J. Morwood, F. S. Intan, I. Mahmud, K. Grant, of agriculture in China. Current Anthropology 52(suppl. 4):S295– N. Somba, B. Akw, and D. Utomo. 2008. Minanga Sipakko and S306. Current Anthropology Volume 52, Supplement 4, October 2011 S379

Early Agriculture and Plant Domestication in New Guinea and Island Southeast Asia

by Tim Denham

A multidimensional conceptual framework is advanced that characterizes early agriculture as a subset of human-environment interactions. Three cross-articulating dimensions of human-environment interaction are considered that accommodate the varied expressions of early agriculture in different parts of the world: spatial scales, transformative mechanisms, and temporalities of associated phe- nomena. These ideas are applied and exemplified at two different scales of resolution—contextual and comparative—in terms of early agricultural development in the highlands of New Guinea and the dispersal of domesticates from New Guinea into Island Southeast Asia.

Recent conceptual debates on early agriculture have shifted tually at the landscape scale. The spread of domesticates from from definitions based on domestication—namely, pheno- New Guinea to Island Southeast Asia is discussed compara- typic and genotypic transformations in plants and animals or tively at broader spatial and temporal scales. The multidi- degree of dependence on domesticates (e.g., Harris 2007; mensional framework potentially has broader application to Smith 2001)—to broader understandings of the environmen- elicit commonalities and differences for the various regions tal and social contexts within which early agriculture was across the world in which agriculture has emerged, although practiced (e.g., Cauvin 2000; Ingold 2000; Marshall 2007; there is insufficient space for elaboration here. Pearsall 2007). Currently, there is neither agreement on nor universal application of a standard definition for identifying Early Agriculture: A Subset of Human- early agriculture in the past, with considerable geographical variation (e.g., Barker 2006; Bellwood 2005; Denham, Iriarte, Environment Relationships and Vrydaghs 2007). In this article, a multifaceted framework is proposed that Animal and plant exploitation, including pastoralism and ag- characterizes early agriculture as a subset of broader human- riculture, are among the most important subsets of human- environment relations, thereby incorporating both the bio- environment relations both in terms of human dependence physical and social realms. The intensive, discursive nature of and environmental change. The interpretation of agriculture the Temozon conference contributed greatly to the emergence in the past should consider the multiplicity of factors that of these ideas and to my understanding of research into early converge in any given human-environment interaction, in- agriculture across the globe. I am indebted to both the or- cluding those of the biophysical realm—such as climate, en- ganizers for inviting me and to other participants for their vironment, and the biology of cultivated plants and tended stimulating company. livestock—as well as the social realm—namely, people and My intention here is to develop a conceptual framework various facets of their practices, including cultures, societies, that can incorporate different emphases in our understanding and technologies. Consequently, and inherently, agriculture of early agriculture for different regions of the world rather has multiple socioenvironmental dimensions that are mutu- than to propose a prescriptive definition designed for uni- ally transformative; namely, they are historical, and each acts versal conformity. The purpose is to characterize rather than on and changes the other through time (e.g., from Timex to to define. The multidimensional model is exemplified at two Timex ϩ 1 in fig. 1). It is proposed and demonstrated here that different scales of resolution. The emergence of agriculture in the totality of these dimensions is essential to characterize the the Upper Wahgi Valley of New Guinea is discussed contex- emergence of agriculture in any given historicogeographical context. Tim Denham is Monash Research Fellow, School of Geography and An inclusive conception of agriculture as a subset of Environmental Science, Monash University (Clayton Campus, human-environment interactions sheds critical light on the Victoria 3800, Australia [[email protected]]). This nature of attempts to elicit a singular definition or an ultimate paper was submitted 13 XI 09, accepted 1 XII 10, and electronically “cause” of agricultural development in the past. Any singular published 25 V 11. definition of agriculture prioritizes one epiphenomenon

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0015$10.00. DOI: 10.1086/658682 S380 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 1. Schematic representation of the multiple dimensions of early agriculture, hereby described as a subset of human-environment inter- actions. Each dimension (a to c) is characterized in figure 2. above others. For example, definitions focused on the iden- the early Holocene. For instance, indigenous plants were evi- tification of domesticated animals or plants in the archaeo- dently domesticated in other regions much later (e.g., India botanical record or on inferences regarding dependence of [Fuller et al. 2004], parts of Africa [Kahlheber and Neumann people on domesticates fail to fully acknowledge that the mor- 2007], and North America [Asch and Hart 2004]). Any cli- phogenetic transformation of species are not uniform in the matic explanation soon shifts sideways from the biophysical past or the present. Numerous factors affect the relative pro- realm to the social realm in order to account for the observed pensity of a species to accumulate anthropically selected traits spatiotemporal variations. The social realm necessitates a through time; some are biological (Ladizinsky 1998), while much broader consideration of how people in different locales others are environmental (e.g., Pearsall 2007) and social, tech- engaged with their environments such that some developed nical, or practical (e.g., Denham 2007; Marshall 2007). Mul- agriculture and others did not. In logical terms, what may tiple dimensions of domesticatory relationships influence initially be characterized as a cause—whether climatic ame- their archaeological visibility. lioration, environmental degradation, or social transforma- The recursivity of human-environment interactions hinders tion—soon becomes a relatively widespread precondition, the interpretation of a singular or ultimate cause of early which in turn then becomes a relatively benign context. agriculture. The positing of any biophysical or social phe- Three cross-articulating dimensions of human-environ- nomenon as a primary cause is arbitrary because any expla- ment interaction are relevant to the characterization of early nation inevitably folds back into the duality of human- agriculture in any historicogeographical context (fig. 2): (a) environment relations (drawing on Giddens 1984). It is spatial scales of analysis, type of method, and lines of evidence; effectively impossible to determine causation within a recur- (b) transformative mechanisms and archaeological expres- sive spiral; transformation in human-environment interac- sions of agriculture; and (c) temporalities of associated phe- tions is continuous, multifaceted, and multicausal. To ex- nomena. Each dimension is briefly discussed below. emplify, if climatic amelioration and stabilization at the beginning of the Holocene are posited as the ultimate cause Articulating Space and Place for the emergence of agriculture across the globe (e.g., Richer- son, Boyd, and Bettinger 2001), this explanation fails to ac- There are considerable variations in the spatial scale of analysis count for the restricted number of locations in which this through which early agriculture is inferred. The scale of anal- actually occurred. Why did agriculture emerge in some places ysis adopted has implications for the ways in which evidence and not in others, and why is there so much temporal var- is used—whether conflated, low resolution, or in particular, iation? high resolution—and tends to be associated with a specific Any account of ultimate causation is soon beset by qual- methodology—either comparative or contextual, respectively ifiers, such as resource availability and species’ susceptibility (fig. 2a). Characterizations of early agriculture at the conti- to domestication. Although there are geographical variations nental and subcontinental scales often conflate data from dif- in the availability and susceptibility of animal and plant re- ferent locales and of slightly different ages to draw a general sources to domestication, these do not solely account for why comparative picture of agricultural development (e.g., Bell- agriculture emerged in some places and not in others during wood 2005; Renfrew 2002). Others draw on locally generated, Denham Early Agriculture and Plant Domestication S381

Figure 2. Three dimensions of early agriculture: a, correspondences be- tween spatial scale, lines of evidence, and methods; b, four domains, or contexts, within which to consider the mutually transformative nature of domesticatory relationships between animals/plants and people; and c, temporal considerations. heavily contextualized evidence to characterize early agricul- Thomas 1996), they are not necessarily incompatible. A key tural development within given landscapes or places (e.g., aim of archaeological research is “to use archaeological data Denham and Haberle 2008; Pearsall 2007) and emphasize to gain an understanding of the indeterminate relations be- intraregional variability. tween large-scale processes and individual lives” (Hodder Although there are some conceptual tensions between com- 1999:175). For those seeking an understanding of early ag- parative and contextual approaches (cf. Renfrew 1973, 2002; riculture, reconstructions at the local level need to be situated S382 Current Anthropology Volume 52, Supplement 4, October 2011 within broader regional, interregional, and continental pro- tuated nature of archaeological finds and the need to place cesses. Highly specific and contextual information can be cau- fragmentary finds into chronological-geographic sequences tiously situated within broader historical and geographical (e.g., Scarre 1988). Interpretations adopt various lines of rea- processes, although the converse is more problematic because soning—from uniformitarian to Occam’s razor to historical of the lack of resolution and specificity in conflated data sets. materialist to postprocessualist—to place archaeological finds in time, namely, to temporalize them, inferring their temporal extension and position in a sequence. Although discontinu- Transformative Mechanisms and Archaeological Expression ities can be recognized in prehistory, such as the abandonment Agriculture is based on the management of plants and animals of a crop or technology, these can be hard to determine with for human exploitation. Agriculture is predicated on varying confidence because of the absence of evidence, which is not degrees of human intervention in the life cycle of plants, which evidence of absence. in turn yield intended and unintended consequences in terms Second, rates of change are rarely considered, which in part of plant biology and the utility of managed plants for people. can be a function of the records and partly a function of The nature of the mutually transformative interaction be- perspective. For example, domestication is a process that op- tween animals/plants and people is differentially expressed in erates at variable rates for different species and subspecies in given historicogeographical contexts; it is not restricted to the different historicogeographic contexts. Some species—per- biological domain even though issues of domestication have haps those with annual life cycles subject to intensive human been the focus of most debate. selection and a high degree of genetic isolation—would be Four domains have a bearing on the character of agricul- anticipated to accumulate traits resulting from human man- ture, namely, how agriculture is expressed archaeologically in agement at a relatively rapid rate. Conversely, these same spe- any historicogeographical context (fig. 2b; cf. Sayer 1984): (1) cies not subject to the same degrees of human selection and biology: degrees of domestication (phenotypic and genotypic genetic isolation, as well as other species with longer life cycles change), gene expression, and phenotypic elasticity; (2) social (such as trees and some animals), would be anticipated to world: demography, dependence, Neolithic traits, orientation accumulate traits at a slower rate. The rate of accumulation to resources, sedentism, and sociopolitical change; (3) envi- of domestication traits within an organism is a function of ronment: human and natural influences, transformation, rate the human-environment context, namely, a function of the of change, sensitivity to change; and (4) technology: trans- multiple domains associated with a domesticatory relation- location, propagation, cultivation, harvesting, processing, and ship. Thus, experimental farming can yield high rates of storage. change in cereals within decades (Hillman and Davies 1990), Academic discourses draw variably on different domains perhaps because of the high degrees of human selection and and lines of evidence to construct arguments for or against genetic isolation, whereas archaeobotany suggests the accu- early agriculture in different parts of the globe. In part, these mulation of these traits actually occurred over thousands of positions reflect the definitions of agriculture adopted, which years in Southwest Asia (Tanno and Willcox 2006), perhaps are usually either inherited from research in Eurasia or de- because of continual genetic interaction between wild and veloped to suit the available evidence within a region. These cultivated stock as much as to selection through management domains are variably expressed, articulated, and aligned in practices (see Jones and Brown 2007). It follows that the different instances of early agriculture. Particular factors seem accumulation of domestication traits in trees can be antici- to be important and correspond in some regions whereas in pated to occur over centuries or millennia (Yen 1996). others they do not; however, they are all relevant. Genotypic and phenotypic changes are a continuum of change along which measures of domesticity, as opposed to wildness, are determined. Variations in the rate of domesti- Temporalities of Associated Phenomena cation, whether measured genetically or phenotypically (as is Debates concerning early agriculture tend to project unilinear customary in archaeobotany), can be anticipated to vary de- or multilinear trajectories from the past to the present; pending on whether a plant is propagated sexually or clonally; namely, they are teleological. Time is viewed as continuous; the accumulation of selected somatic mutations in the latter processes are viewed as cumulative (e.g., Richerson, Boyd, is a qualitatively different type of process to the accumulation and Bettinger 2001), as if they lead somewhere significant of selected mutations through sexual reproduction (Yen other than solely toward the present. Three aspects of time, 2003). Additionally, genotypic and phenotypic traits resulting and the temporality of things (namely the temporal extension from human management should not be anticipated to ac- of something, or its being in time; Thomas 1996), are signif- cumulate at the same rate within a species, or between species, icant. Although time and temporality are implicit to any dis- especially given latent issues of gene expression and pheno- cussion of early agriculture and subsequent transformations, typic elasticity for some plants under cultivation (Gremillion they are rarely made explicit (fig. 2c). and Piperno 2009). The generation of phenotypic varieties in First, the temporality of things is usually assumed to be some plants, such as bananas and yams, need not correspond continuous, or semicontinuous. In part this reflects the punc- to genotypic change but may solely represent a phenotypic Denham Early Agriculture and Plant Domestication S383 response or the differential expression of a gene due to the tation of each within the Upper Wahgi Valley (Denham 2009; environment of cultivation and growth. Denham and Haberle 2008). In this landscape, a teleological Third, the archaeology of early agriculture has tended to unfolding is at one level unavoidable because agriculture did view time in the abstract, namely, chronologically, and not emerge from previous foraging practices at some point in the from the perspective of lived experience or experiential time. early Holocene. However, if the island of New Guinea is taken Issues of plant domestication are generally considered from as a whole, a unilinear or multilinear characterization does the perspective of how many years before traits x and y be- not hold, because multiple types, including ambiguous types come apparent in the archaeobotanical record; they are rarely of plant exploitation, were being practiced up to the modern considered from the perspective of how these traits accu- era. mulated through the day-to-day activities of people and were The multidisciplinary evidence from a relatively restricted passed on from generation to generation. The time of lived region or landscape can be relied on as the primary context experience is a precondition for the constitution of chrono- for interpreting the emergence and transformation of agri- logical time (Heidegger 1962) and processes thereby inferred. culture (Denham 2008; Denham, Fullagar, and Head 2009; Existential aspects of time are glossed or avoided because Denham and Haberle 2008). The multisite records from the they can be considered to be attempts to get inside the minds Upper Wahgi Valley are the most detailed in New Guinea for of people in the past. Although partly true, this is always the eliciting plant exploitation and associated landscape changes case, because discussion of domestication often considers in the past, primarily because of numerous archaeological and whether traits were intentionally or unintentionally accu- paleoecological investigations at wetlands bearing evidence of mulated, namely, to understand the intent, or mindset, of the past manipulation for plant exploitation, including early ag- people involved. The intentionality of the domesticatory pro- riculture (figs. 4, 5; table 1; reviewed in Denham and Haberle cess requires a consideration of experiential time whether to 2008 and supplemented by Coulter et al. 2009; Denham, Ha- understand the deliberate selection of a taro corm or cereal berle, and Pierret 2009; Denham et al. 2009; Sniderman, Finn, grain or the qualitatively different temporal perspective of and Denham 2009). The evidence from these sites of food planting a tree. The former yields within a year, whereas the production will eventually be complemented by the multi- latter may take decades before yielding and requires an in- disciplinary results (once complete) from the excavations of tergenerational perspective (Ingold 2000; Terrell 2002). proximal occupation sites along an altitudinal gradient on the valley wall (Christensen 1975; Donoghue 1989). Contextual: Early Agriculture in the Highlands of New Guinea Transformative Mechanisms and Domestication Histories The phenotypic or genotypic (application of ancient DNA) transformation of plants from wild to domestic forms has Landscape Scale of Analysis not been clearly charted through time and across space for In New Guinea, as elsewhere, there are high degrees of regional any plant in the New Guinea region despite variable claims variability in the nature of plant exploitation practices across (Denham 2004b; Donoghue 1989; Golson introduction in the island (e.g., in major crop plants; Bourke and Harwood Christensen 1975; Haberle 1995; Lentfer 2009; Yen 1996). As 2009; Kennedy and Clarke 2004), the nature of cultivation a result, the case for early agricultural development in New practices (Bourke 2001), and the degree of reliance on cul- Guinea has leaned heavily on archaeological evidence of tech- tivation (Roscoe 2002; Terrell 2002), with similar variability nologies and past practices (Denham 2005a, 2006, 2007, 2009; likely to have characterized the recent and more distant pasts Denham and Haberle 2008; Golson 1982) and environmental (see fig. 3; Denham 2005b). Consequently, it may not be transformation (geomorphology and paleoecology: Denham, meaningful to conflate evidence over broad sociospatial Golson, and Hughes 2004; Denham et al. 2003; Golson and scales—from across the highlands or lowlands, for the island Hughes 1980; Haberle 1994, 2003, 2007; Hope and Haberle of New Guinea, or for Near Oceania—together into a single 2005). There is limited understanding of the social world macrochronology or macrointerpretation of plant exploita- inhabited by early agriculturalists (Golson and Gardner 1990; tion in the past. By so doing, plant exploitation practices and Haberle and Chepstow-Lusty 2000) largely because few oc- suites of domesticates that never co-occurred may be erro- cupation sites dating to the early Holocene have been exca- neously brought together into a single historical narrative; vated in the highlands, and most of these have not been namely, the whole can be misleadingly coherent and much published in full (except White 1972). greater than the sum of the parts. Types of archaeological information often considered to The interpretation of the emergence and transformation of accompany early agricultural development in other parts of agriculture in Highland New Guinea has sought to avoid tel- the world are equivocal or absent in the New Guinea record eological interpretations by focusing on the expanding rep- (e.g., Diamond 2002; Piggott 1954). In the absence of cem- ertoire of plant exploitation practices (including various forms eteries or ossuaries, demographic signatures of early agricul- of cultivation) toward the present and the spatial manifes- ture have been inferred using crude proxies, such as number . Denham Early Agriculture and Plant Domestication S385 of archaeological sites (Haberle and Chepstow-Lusty 2000). terms of current understandings of how plant exploitation Several Neolithic traits often associated with early agriculture practices changed through time. For the Upper Wahgi Valley, (e.g., pottery and domesticated animals) are absent. The re- these hypotheses can be formulated with respect to bananas, lationship between sedentism and early agriculture is equiv- taro, and a yam (e.g., Denham 2009). Each crop plant had ocal; there are questionable claims for Pleistocene-aged set- been considered to be of lowland derivation and brought to tlements in the highlands at NFX (Watson and Cole 1977) the highlands by people (Hope and Golson 1995; Yen 1995), and Wan˜elek (Bulmer 1977), but most settlements postdate although they potentially grew naturally in the highlands dur- 4000 cal BP. Archaeobotanical investigations at occupation ing the early Holocene (Denham, Haberle, and Lentfer 2004; sites are few (excepting Christensen 1975) and incomplete, Haberle 1993). Only two early steps in the domestication and consequently shed little light on plant-food consumption process can be hypothesized from the archaeological and ar- preferences and practices. In a similar vein, sociopolitical chaeobotanical evidence: planting and cultivation. transformations are poorly known, although the rise of pu- First, planting of these three plants from wild stock is hy- tatively patrilineal, highly territorial descent groups and the pothesized to have occurred in the early Holocene before big-man form of leadership have been inferred from agri- mounded cultivation dating to 7000–6500 cal BP on the wet- cultural history (Denham and Haberle 2008; Golson and land margin at Kuk (radiocarbon dating for Kuk is presented Gardner 1990). in Denham et al. 2003). Some form of swidden cultivation Given the limited knowledge of the biological and social is likely to have preceded more intensive forms of cultivation domains in the early and mid-Holocene, the following dis- (Denham and Haberle 2008). It is still unclear whether initial cussion of the transformation of plant exploitation through planting was through seed or vegetative propagation, although time focuses on people’s orientation to plants within the Up- the latter is the dominant characteristic of highlands agri- per Wahgi Valley as inferred from the cross-articulation of culture and some other forms of plant exploitation in New archaeobotanical (plants), archaeological (technology/prac- Guinea. Indeed, domestication has led to the development of tice), and paleoecological (environment) evidence. numerous cultivars that are no longer able to reproduce sex- Archaeobotanical and paleoecological research provides in- ually under normal growing conditions. Consequently, veg- formation on food plants that were available locally at Kuk etative propagation is generally considered to have always and in the Upper Wahgi Valley as well as for the processing been the dominant mode of plant reproduction under cul- and inferred cultivation of some staples (Denham 2005b, table tivation in New Guinea (Denham 2005b), although some 2). Evidence for the presence of a food plant does not indicate plants are grown from seed. Planting encouraged the devel- that it was exploited in the past even if its use has been opment of “cultiwild” populations, namely, cultivated or documented ethnographically (Powell 1982). However, on- managed populations of effectively wild plants, which may site evidence of food plants in association with archaeological still interbreed with any wild populations growing in the vi- features representing either foraging or a form of agriculture cinity (following De Langhe et al. 2009). is more suggestive of potential exploitation, even if adventi- Second, an increased reliance on vegetative propagation and tious. By contrast, archaeobotanical evidence for processing the systematization of agricultural practices in the highlands, or cultivation enables more direct comparison with, and the exhibited with the development of mounded cultivation by direct integration of plants into, the plant exploitation chro- at least 7000–6500 cal BP and of ditched field systems by ca. nology for the region (e.g., Denham 2009; Denham and Ha- 4000 cal BP, led to increasing degrees of domestication (Ca- berle 2008). At present, significant archaeobotanical evidence ballero 2004) in managed stands and the creation of an array is limited to bananas (Musa spp.), karuka Pandanus (Pan- of cultivars through anthropic selection and increasing genetic danus julianettii/iwen/brosimos), taro (Colocasia esculenta), isolation of favored plants. Genetic isolation prevented in- and a yam (Dioscorea sp.; Denham, Haberle, and Lentfer 2004; breeding between cultivated and wild stock and was achieved Denham et al. 2003; Donoghue 1989; Fullagar et al. 2006, through a combination of intentional and unintentional prac- 2008). tices, including plant translocation, environmental degrada- Because of evidential deficiencies, the domestication his- tion, and vegetative propagation of cultivated stock. Trans- tories of significant crop plants can only be hypothesized in location brought managed plants from lower altitudes to the

Figure 3. Maps depicting the variability in plant exploitation across Papua New Guinea. Top, contour map of Papua New Guinea; middle, geograph- ical distribution of major crop plants today (from Bourke and Harwood 2009); bottom, interpretation of the geographical distribution of major crop plants following the exclusion of recently introduced crops. Crop plants likely to have been introduced within the last 500 years comprise sweet potato (Ipomoea batatas), cassava (Manihot esculenta), and taro kongkong (Xanthosoma sagittifolium). . Denham Early Agriculture and Plant Domestication S387

floor of the Upper Wahgi Valley, in which wild stands of the although the dispersal of plants considered to be domesticated same species were rare or absent. The degradation of forest in the New Guinea vicinity sheds a clearer light on plant to grasslands on the floor of the Wahgi Valley in the vicinity exploitation practices, effectively the orientation of people to of Kuk by ca. 7000 cal BP would have removed the habitats plant resources, within these regions (Denham 2010; Dono- of wild stands of banana, taro, and yam, if present, thereby hue and Denham 2010). Currently, knowledge of the social further isolating cultivated stock. The focus on vegetative and technological domains of early plant exploitation through propagation isolated cultivars genetically through clonal re- Island Southeast Asia is limited; they can only be implied production, albeit with somatic mutation, thereby partially through broadscale and low-resolution (both spatially and selecting against sexually reproduced plants and increasing chronologically) comparisons of the dispersal and nondis- the isolation of cultivated stands from wild gene pools. Some persal of crop plants from New Guinea. cultivars in New Guinea are still interfertile with wild pop- Traditionally, the suite of domesticates considered indige- ulations, especially some diploid banana cultivars, and inter- nous to New Guinea included a range of highland and lowland breeding is encouraged in some areas (Kennedy and Clarke plants, most of which were insignificant or absent outside the 2004); however, this would be dependent on the survival of region. Highland domesticates included the karuka Pandanus local pollen sources and pollinators in heavily altered envi- complex (Pandanus julianettii/iwen/brosimos complex), edible ronments. pitpit (Setaria palmifolia), and Rungia (Rungia klossii), These domestication scenarios are currently hypotheses; the whereas lowland domesticates included species of marita Pan- purpose is heuristic and not intended to suggest that these danus (Pandanus conoideus), Canarium spp., and Terminalia processes were restricted to the Upper Wahgi Valley region. spp., as well as sago (Metroxylon sagu; Barrau 1955:46). Spe- Rather, similar types of process were probably widespread, if cies with broad altitudinal ranges included sugarcane (Sac- variable, across New Guinea and plausibly in adjacent regions. charum officinarum; Simmonds 1976b:104–108) and Austral- The net effects of these domesticatory relationships are visible imusa bananas (Musa spp.; Simmonds 1976a:211–215). The in the phylogenies that shed light on the natural distributions, domestication locus of Pueraria lobata (Watson 1964), a tu- geodomestication pathways, and anthropic spread of major berous plant formerly cultivated in the highlands, is ques- crop plants in New Guinea and Island Southeast Asia. tionable. Interpretations of origin and domestication have largely been based on centers of greatest genetic diversity and the presence of ancestral wild forms from which domesticated forms arose (Yen 1985, 1991). Domestication processes in Comparative: Dispersal of Domesticates New Guinea, as elsewhere, focused on decreasing toxicity (e.g., and Plant Management in Island taro), decreasing seed size in some fruits (e.g., some bread- Southeast Asia fruit, bananas), and increasing the edible portions of most At present, there are no multidisciplinary records of plant root crops, fruits, and nuts. exploitation elsewhere in New Guinea comparable to those The application of genetic tools to modern-day plant pop- from the Upper Wahgi Valley. Tentative comparisons can be ulations has been revolutionary and suggests that a whole made with similar types of records from locales within neigh- range of important pantropical food plants underwent initial boring regions of northern Australia (Denham, Fullagar, and or separate domestication in the New Guinea region (table Head 2009) and Island Southeast Asia (Barton and Denham, 2; Kennedy and Clarke 2004; Lebot 1999). Many of these food forthcoming). These comparisons suggest that the plant ex- plants were previously thought to have been domesticated in ploitation mosaics characteristic of New Guinea may have Southeast Asia, and potentially the locus of domestication extended possibly into northern Australia before European may change again as additional cultivated and wild popula- arrival (Denham, Donohue, and Booth 2009; Jones and Mee- tions are included within the analysis. Of particular signifi- han 1989) and into Island Southeast Asia before the advent cance for understanding the history of agriculture in New of Austronesian speakers (Barker 2006; Barton and Denham, Guinea, Island Southeast Asia, and beyond are four globally forthcoming). At present these mosaics are largely invisible, significant starch-rich plants (elaborated in Denham 2010).

Figure 4. Depictions of (a) significant archaeological and paleoecological sites on the island of New Guinea; (b) significant archaeological and paleoecological sites in the western highlands of Papua New Guinea (shaded box in a); (c) the locations of Ambra, Kuk, Warrawau, and Wurup sites in the Upper Wahgi Valley (box in b); and (d) the relative locations of Kuk, Warrawau, and Wurup sites along an altitudinal transect from the floor to the upper walls of the Upper Wahgi Valley (reproduction of Denham and Haberle 2008, fig. 1). S388 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 5. Relative intersite chronology for early and mid-Holocene re- mains at wetland agricultural sites in the interior of New Guinea. This table is an updated version of Denham 2007, table 5.3. aThe dates at Kuk are based on tephrochronology and radiocarbon dates (see sections in Denham 2003b). bThe relative dates of mounded paleosurfaces at Mu- gumamp and Warrawau are based on tephrochonology (i.e., the lie of R ash in paleosurface features) and on radiocarbon dates for a paleochannel at the latter (Denham 2003a). cDitches at Warrawau, Kana, Tambul, and Haeapugua have been radiocarbon dated (Denham 2003a). dThe dating of “intensive cultivation” at Kana and Ruti is based on tephrochronology (i.e., the lie of R ash) and should be considered provisional at both sites (Denham 2003a). eThe stratigraphic relationship between R ash depo- sition and the earliest Phase 3 ditch at Kuk is uncertain (Denham, Golson, and Hughes 2004).

Bananas (Musa spp.) East African cultivars (AAA), as well as the yellow Cavendish banana (AAA), which is the most widely grown and consumed Bananas (Musa spp.) are an important cash and subsistence plantation cultivar today (e.g., Kennedy 2008). The enormous crop in the tropics and subtropics (De Langhe et al. 2009). diversity of modern Eumusa-derived banana cultivars rep- The most significant fruiting bananas are derived from species resents complex geodomestication pathways, including mul- of Eumusa section, principally Musa acuminata (genome A) and Musa balbisiana (genome B). Formerly, Eumusa bananas tiple subspecific and specific domestications, progressive par- were presumed to be Southeast Asian domesticates, whereas thenocarpy and seed suppression, the creation of diploids and bananas of sections Australimusa and Ingentimusa were con- triploids, interspecific and intersubspecific hybridization, and sidered indigenous to the New Guinea region (Simmonds somatic mutation (Perrier et al. 2009). 1976a; Stover and Simmonds 1987; Yen 1973). Interpretations suggesting early domestication of Musa spp. According to recent research, the initial stages of the do- in the New Guinea region have received some archaeobotan- mestication of most Eumusa cultivars can be traced to Musa ical corroboration. Phytoliths of Eumusa type date to ca. acuminata ssp. banksii populations, which are indigenous to 10,000 years at Kuk Swamp in the highlands, with subsequent the New Guinea region; parthenocarpy is inferred to have banana cultivation inferred from high Musaceae phytolith arisen in this species first (Perrier et al. 2009). Musa acuminata frequencies, including those of Eumusa type, in association ssp. banksii contributes to the genome of several different with archaeological features from 7,000 to 6,500 years ago groups of banana cultivars, including some of the Pacific plan- (Denham et al. 2003; Lentfer 2009). Multidisciplinary evi- tains (AAB), Western and Central African plantains (AAB), dence suggests the dispersal of bananas westward from New Denham Early Agriculture and Plant Domestication S389

Table 1. Summary of wetland archaeological excavations and evidence for prehistoric agriculture in the interior of New Guinea

Altitude Main field Site namea (m) Location seasons Principal publications Tambul 2,170 Upper Kaugel Valley 1976 Golson 1997 Mogoropugua 1,890 Tari Basin 1980 Ballard 1995:193–195; Golson 1982:121 Minjigina 1,890 Upper Wahgi Valley 1967 Golson 1982:121; Lampert 1970; Powell 1970:172–174 Ambra Crater 1,760 Upper Wahgi Valley 1999 Sniderman et al. 2009 Haeapugua 1,650 Tari Basin 1991–1992 Ballard 1995, 2001 Kindengb 1,600 Upper Wahgi Valley 1968 Unpublished Warrawau (Manton’s) 1,590 Upper Wahgi Valley 1966, 1977 Golson 1982:121, 2002; Golson et al. 1967; Lampert 1967; Powell 1970:142–146 Kuk 1,560 Upper Wahgi Valley 1972–1977, 1998–1999 Bayliss-Smith and Golson 1992a, 1992b, 1999; Denham 2003a, 2003b, 2004a, 2005a; Denham et al. 2003; Denham, Golson, and Hughes 2004; Golson 1977, 1991; Golson and Hughes 1980 Mugumamp 1,560 Upper Wahgi Valley 1977 Harris and Hughes 1978 Kana 1,480 Middle Wahgi Valley 1993–1994 Muke and Mandui 2003 Ruti Flats 480 Lower Jimi Valley 1983–1985 Gillieson, Gorecki, and Hope 1985; Gil- lieson et al. 1987; Gorecki and Gillieson 1984, 1989 Note. This table is an updated version of Denham 2007, table 5.1. a Other wetland sites have been inspected by archaeologists, although none was investigated in detail. For example, the site at Kotna (1,580 m) in the Upper Wahgi Valley was village land under drainage for coffee. The site was visited by Jack Golson and John Muke in 1988, at which time they sought permission to record features exposed in drain walls. Permission was refused, but while waiting they were able to look at some stretches of drain wall, in which ditches comparable to those of Phase 5 at Kuk were exposed (Jack Golson, personal communication, 2002). b The archaeological finds at Kindeng have not been cross-correlated with those at other wetland sites (Jack Golson, personal communication, 2001).

Guinea occurred within a pre-Austronesian time frame (Den- sumed to have occurred in New Guinea, alternative origins ham and Donohue 2009; Donohue and Denham 2009, 2010). in East Asia have been proposed (Daniels and Daniels 1993) Musa acuminata ssp. banksii–derived cultivars spread west- and are yet to be fully investigated. ward into eastern Island Southeast Asia, where they hybridized The initial stages of sugarcane domestication were proposed with other species and subspecies to produce more robust to comprise the anthropic selection and domestication of a triploid cultivars that subsequently became widely dispersed wild ancestor of Saccharum robustum in New Guinea (Sim- throughout Southeast Asia and Africa, potentially to Kot Diji monds 1976b:104–108). Lebot advanced this scenario in his in Pakistan by ca. 4000 cal BP (Fuller and Madella 2009) and evaluation of the molecular evidence to suggest that to Cameroon by ca. 2500 cal BP (Mbida Mindzie et al. 2001; S. robustum is the most likely precursor of sugarcane and cf. Neumann and Hildebrand 2009). These processes are sug- was domesticated in New Guinea where human selection gestive of cultivation of bananas in parts of Island Southeast of chewing plants with sweet juice and low fibre produced Asia before the advent of Austronesian language speakers from the S. officinarum clones. Cultivars were subsequently dif- ca. 4000 cal BP. ferentiated in numerous distinct morphotypes via vegetative propagation and selection of somatic mutants. (Lebot 1999: Sugarcane (Saccharum officinarum) 622–623) Although widely perceived to be a snack food, sugarcane has Lebot (1999) concludes that “S. officinarum cultivars are de- been documented as a staple in parts of the eastern highlands rived from introgressions between wild forms of S. robustum of New Guinea (Daniels and Daniels 1993), and its impor- and S. spontaneum in Melanesia” (623). Lebot envisages a tance in other regions in the past should be considered, es- similar domestication scenario for Saccharum edule, a plant pecially as fodder. Sugarcane is an interspecific cultivar pre- cultivated in New Guinea for its aborted inflorescences. While dominantly derived from the hybridization of Saccharum agreeing that initial domestication of S. robustum occurred in robustum and Saccharum spontaneum. The scenarios of sug- New Guinea, Grivet et al. (2004) consider that the resultant arcane domestication are in some respects similar to that for cultivar dispersed westward to Southeast Asia, where it hy- Eumusa bananas, namely, initial domestication in New bridized primarily with wild populations of S. spontaneum to Guinea with subsequent westward dispersal and interspecific produce S. officinarum as well as with other species to produce hybridization in Southeast Asia (Grivet et al. 2004). Although other cultivars. the origin and domestication of sugarcane has long been pre- Archaeobotanical evidence of any antiquity for sugarcane S390 Current Anthropology Volume 52, Supplement 4, October 2011

Table 2. Plant domesticates from the New Guinea region that are significant food plants outside Melanesia

Botanical name Common name Reference Alocasia macrorrhiza Giant taro Lebot 1999 Artocarpus altilis Breadfruit Zerega, Ragone, and Motley 2004 Musa spp. (Australimusa section) Fe’i bananas Sharrock 2001 Colocasia esculenta Taro Lebot et al. 2004 Cyrtosperma chamissonis Giant swamp taro Lebot 1999 Dioscorea alata Greater yam Malapa et al. 2005 Musa spp. (Eumusa section) Bananas and plantains Perrier et al. 2009 Metroxylon sagu Sago Kjær et al. 2004 Saccharum officinarum Sugarcane Grivet et al. 2004 Note. Some plants have undergone independent domestication events elsewhere (e.g., taro and aerial yam). is nonexistent; the putative find from Yuku rock shelter in Southeast Asia and New Guinea were prevented through re- the New Guinea Highlands (Bulmer 1975:31) can be dis- productive isolation, whether geographically or culturally de- counted given the uncertainty of its identification (Yen 1998: termined (Lebot 1999:624). 31) and disturbance to the site (T. Denham, unpublished Archaeobotanical and paleoecological research indicates research). The only approximate chronological information that taro was potentially exploited during the Pleistocene in for the origin of sugarcane is linguistic and tenuous; a term Island Melanesia (Loy, Spriggs, and Wickler 1992) and on for sugarcane, “*CebuS,” reconstructs to proto-Austronesian, Borneo (Barton and Paz 2007). At ca. 10,000 cal BP, taro was namely, the languages on Taiwan before the subsequent dif- exploited at Kuk Swamp in the highlands of New Guinea ferentiation of Austronesian languages on Taiwan ca. 5,500 (Fullagar et al. 2006), and taro pollen was present in northern years ago (Blust 1984–1985; Pawley 2007). If the linguistic Australia (Haberle 2005) and lowland New Guinea (Haberle reconstructions are sufficiently specific, both botanically and 1995). Despite the relative ubiquity of the plant, the potential chronologically, which is doubtful, then several stages in the exploitation of taro during the Pleistocene and early Holocene domestication of sugarcane occurred before Austronesian lan- is significant because of the processing required to remove guage dispersal from Taiwan (Donohue and Denham 2010), acridity and increase the edibility of wild types. including initial domestication of S. robustum in New Guinea, westward movement of the derived cultivar to Island South- Greater Yam (Dioscorea alata) east Asia, and subsequent hybridization with S. spontaneum. The natural distribution and locus of domestication for greater yam (Dioscorea alata) are unknown. Current pan- Taro (Colocasia esculenta) tropical distributions of D. alata cultivars are generally ac- Wild-type taro (Colocasia esculenta var. aquatilis), the pre- cepted to result from human agency, namely, vegetative prop- cursor to cultivated taro (C. esculenta), has a pantropical dis- agation of clones, even though the plant can become a tribution extending from northeastern India to mainland persistent weed once established in a region. Morphological, Southeast Asia, Indonesia, New Guinea, and northern Aus- enzymatic, and physicochemical characteristics are not solely tralia (Matthews 1995:108–114). The biogeographic and hu- due to somatic mutation and asexual reproduction; some sex- man processes that created this wild-type distribution are un- ual reproduction is represented (Lebot et al. 1998). known (Matthews 1991). Some regions, however, such as New Limited intraspecific differentiation of D. alata cultivars Guinea, have endemic species-specific pests (Tarophagus spp., reflects geography, morphology, isozymes, and physicochem- Matthews 2003) and potentially species-specific pollinators ical characteristics, but there are few correlations among at- (Matthews 1995), suggesting a long antiquity for the plant tributes (Lebot 1999:624–625; Lebot et al. 1998; Malapa et al. beyond that involving human management. 2005). For example, the high degree of morphological vari- The locus of taro domestication has been variably deter- ation within D. alata represents phenotypic elasticity and not mined based on different types of analysis, that is, from north- genotypic variation (Malapa et al. 2005). Effectively, widely eastern India (Kurvilla and Singh 1981) to eastern Indonesia dispersed cultivars are clones with a narrow genetic base (Le- (Lebot and Aradhya 1991). Recent research suggests inde- bot 1999:625). pendent domestications of taro in Southeast Asia and New Several authors consider the New Guinea region to be the Guinea (Irwin et al. 1998; Lebot et al. 2004). The gene pools locus of initial D. alata domestication because it exhibits the of diploid cultivars in Southeast Asia and New Guinea are greatest morphological variation and genetic diversity (Cour- clearly distinguishable, and each exhibits relatively low genetic sey 1972, 1976; Lebot 1999; Martin and Rhodes 1977). As De diversity (Lebot et al. 2004). Based on current evidence, the Candolle noted (1884:13), however, the determination of or- interbreeding and intermixing of diploid taro populations in igin should be based on the elimination of all artificial Denham Early Agriculture and Plant Domestication S391 forms (namely cultivars) and not on the diversity of cultivars. localized exchange networks between islands resulted in the Perhaps the most compelling evidence for Wallacea and Sahul net transfer of ideas and things over vast regions. The plants being the locus of origin of D. alata is the genetic proximity would have been moved and planted rather than solely being of this species to two other yam species derived from the tradable commodities exchanged via long-distance voyaging. same regions, Dioscorea nummularia and Dioscorea transversa At present, it is not clear whether crop plants spread together (Malapa et al. 2005:928). Although circumstantial, the New with practices of cultivation and processing or whether cul- Guinea region seems to be the place of D. alata origin and tivars were introduced and adopted into preexisting plant domestication from which cultivars were dispersed clonally exploitation mosaics across Island Southeast Asia. The time across the globe. depth of these practices and dispersals across Island Southeast There is no definitive archaeobotanical evidence for D. alata Asia is similarly enigmatic, although they seemingly predate predating ca. 3500 cal BP (Paz 2005). Nonspecific yam res- 4000 cal BP. idues dating to the early Holocene have been identified on In sum, it is proposed that people in parts of Island South- stone tools in Island Melanesia (Barton and White 1993) and east Asia practiced forms of cultivation before Austronesian at Kuk Swamp (Fullagar et al. 2006). The reporting of “Dios- language dispersal. At present the history of these presumably corea sp., possibly D. alata” (Barton 2005:66) at Niah cave in nascent agricultural practices is unknown, and it is unclear Borneo should be treated as provisional. Despite the ambi- whether they were in situ developments or were introduced guities of archaeobotanical, botanical, and genetic evidence, from Asia or New Guinea. Arguably, people inhabiting parts the proposed domestication of D. alata in the New Guinea of Island Southeast Asia had comparable orientations to plant region is currently the most plausible interpretation. Clones resources to those documented for parts of New Guinea, in- were subsequently dispersed over a wide geographical area, cluding plant management, vegetative propagation, and cul- including Island Southeast Asia, Southeast Asia, and Africa tivation. as well as eastward into the Pacific. Conclusions Implications for Understanding Plant Domestication Mosaics In this article, a multidimensional model of early agriculture in Island Southeast Asia has been applied at two different scales of analysis: a contex- There is no doubt that the New Guinea region was a major tual application at the landscape scale and a comparative ap- center of plant domestication (Lebot 1999). It is also impor- plication at the regional scale. The emergence and transfor- tant to note that the crop domestication histories and dis- mation of agriculture in the highlands of New Guinea is persals outlined above do not follow similar historicogeo- discussed with respect to multidisciplinary evidence from the graphical pathways. These plants are unlikely to have spread Upper Wahgi Valley. In contrast to previous publications, westward from New Guinea as part of a coherent which have emphasized environmental and technological (horti)cultural package, although dispersal of crop associa- transformations, the focus here has been on the changing tions may have potentially occurred from Island Southeast nature of domesticatory relationships through time. A similar Asia to other regions, such as to parts of mainland Southeast theme was applied to understanding the dispersal and non- Asia, South Asia, and Africa. Bananas and sugarcane are sug- dispersal of crop plants from New Guinea to Island Southeast gestive of westward dispersal from New Guinea with subse- Asia; the focus was on crop-plant domestication and dispersal quent hybridizations in Southeast Asia. Taro is suggestive of with a view to eliciting an impression of plant exploitation independent domestications in New Guinea and Southeast mosaics that are, in the main, currently invisible to archae- Asia with geographical and cultural isolation between cultivar ological and paleoecological research. gene pools. Greater yam is suggestive of domestication in New The variability in plant exploitation practices and crop Guinea with subsequent widespread dispersal of clonally re- plants witnessed for New Guinea in the recent past and re- produced cultivars. Presumably some plants did not move, constructed for the distant past are likely to apply to parts of such as taro and some yams (e.g., Dioscorea bulbifera; Lebot Island Southeast Asia before any Austronesian influence on 1999), because they were relatively ubiquitous resources across the region. Plant exploitation mosaics are likely to have in- New Guinea and Island Southeast Asia, were subject to vary- cluded parts of Island Southeast Asia and New Guinea for ing local forms of management and domestication, or were much of the Holocene as well as other parts of Near Oceania less significant than alternatives and can therefore be assumed and potentially northern Australia. Food plants (and plants not to have been highly prized trade items (Denham 2010). used for other purposes) were incorporated in varying degrees In the absence of archaeobotanical verification, there seems of domestication into these plant exploitation mosaics as evi- to have been a westward movement of bananas and sugarcane denced by the westward dispersal of Musa bananas and a precursor from New Guinea before the arrival of Austronesian domesticated form of Saccharum robustum (sugarcane pre- language speakers from Taiwan after 4000 cal BP. These dis- cursor), which indicates that the region’s inhabitants had an persals were facilitated by interisland interaction within Island orientation to plant resources that plausibly included forms Southeast Asia (Bulbeck 2008; Donohue and Denham 2010); of cultivation. S392 Current Anthropology Volume 52, Supplement 4, October 2011

The framing of early agriculture within a broader set of ing-farming transitions in Island Southeast Asia. G. Barker and M. human-environment relationships is intended to free discus- Janowski, eds. Cambridge: McDonald Institute Monograph. Barton, H., and V. Paz. 2007. Subterranean diets in the tropical rain sion from conceptually restrictive debates regarding defini- forests of Sarawak, Malaysia. In Rethinking agriculture: archaeo- tions and causation. The multidimensional perspective out- logical and ethnoarchaeological perspectives. T. P. Denham, J. Iriarte, lined here is designed to be an enabling framework for and L. Vrydaghs, eds. Pp. 50–77. Walnut Creek, CA: Left Coast. interpretation rather than to be prescriptive. Viewing early Barton, H., and J. P. White. 1993. Use of stone and shell artefacts agriculture through the multidimensional lens of human- at Balof 2, New Ireland, Papua New Guinea. Asian Perspectives 32: 169–181. environment interactions provides an integrated perspective Bayliss-Smith, T. P., and J. Golson. 1992a. A Colocasian revolution to understand how agriculture emerged and transformed in in the New Guinea Highlands? insights from Phase 4 at Kuk. the past. Instead of prioritizing one epiphenomenon of ag- Archaeology in Oceania 27:1–21. riculture, most often domestication, the generic framework ———. 1992b. Wetland agriculture in New Guinea Highlands pre- encompasses biological, environmental, social, and technical history. In The wetland revolution in prehistory. B. Coles, ed. Pp. 15–27. Exeter, UK: Prehistoric Society and Wetland Archaeological domains. Although illustrated with respect to the evidence Research Project. from New Guinea and Island Southeast Asia, the concepts ———. 1999. The meaning of ditches: deconstructing the social have broader applicability. The comparative application of this landscapes of drainage in New Guinea, Kuk, Phase 4. In The pre- generic framework would enable commonalities and differ- history of food: appetites for change. C. Gosden and J. Hather, eds. ences, both of kind and emphasis, to be determined from the Pp. 199–231. London: Routledge. Bellwood, P. 2005. First farmers. Oxford: Blackwell. diverse multifaceted expressions of early agriculture in dif- Blust, R. 1984–1985. The Austronesian homeland: a linguistic per- ferent regions of the world. spective. Asian Perspectives 20:45–67. Bourke, R. M. 2001. Intensification of agricultural systems in Papua New Guinea. Asia Pacific Viewpoint 42:219–236. Bourke, R. M., and T. Harwood, eds. 2009. Food and agriculture in Papua New Guinea. Canberra, Australian Capital Territory: ANU Acknowledgments E Press. Bulbeck, D. 2008. An integrated perspective on the Austronesian Many thanks to Ofer Bar-Yosef and Doug Price, as well as to diaspora: the switch from cereal agriculture to maritime foraging the Wenner-Gren Foundation, for the invitation to “The Or- in the colonisation of Island Southeast Asia. Australian Archaeology igins of Agriculture: New Data, New Ideas” symposium. The 67:31–51. ideas in this article were stimulated by participants at the Bulmer, S. 1975. Settlement and economy in prehistoric Papua New Guinea: a review of the archaeological evidence. Journal de la So- symposium and derive from several significant collaborations, cie´te´ des Oce´anistes 31:7–75. most especially with Huw Barton, Mark Donohue, Jack Gol- ———. 1977. Between the mountain and the plain: prehistoric set- son, and Simon Haberle. I would like to acknowledge their tlement and environment in the Kaironk Valley. In The Melanesian profound contributions to my thinking and research. I also environment. J. H. Winslow, ed. Pp. 61–73. Canberra: ANU Press. human-plant interaction: an evolu- thank Mike Bourke, Edmond De Langhe, Carol Lentfer, Ka- Caballero, J. 2004. Patterns in tionary perspective. Paper presented at the International Society tharina Neumann, and Xavier Perrier for giving me permis- of Ethnobiology, Ninth International Congress, Canterbury. sion to cite what was at the time unpublished research. Thanks Cauvin, J. 2000. The birth of the gods and the origins of agriculture. to Uri Gilad for assistance with figure 3 and Kara Rasmanis Cambridge: Cambridge University Press. for the production of the other figures. Christensen, O. A. 1975. Hunters and horticulturalists: a preliminary report of the 1972–4 excavations in the Manim Valley, Papua New Guinea. Mankind 10:24–36. References Cited Coulter, S., T. P. Denham, C. Turney, and V. Hall. 2009. The geo- chemical characterisation and correlation of locally distributed late Asch, D. L., and J. P. Hart. 2004. Crop domestication in prehistoric Holocene tephra layers at Ambra Crater and Kuk Swamp, Papua eastern North America. In Encyclopedia of plant and crop science. New Guinea. Geological Journal 44:568–592. R. M. Goodman, ed. Pp. 314–319. New York: Dekker. Coursey, D. G. 1972. The civilizations of the yam: interrelationships Ballard, C. 1995. The death of a great land: ritual, history and sub- of man and yams in Africa and the Indo-Pacific region. Archaeology sistence revolution in the Southern Highlands of Papua New Guinea. and Physical Anthropology in Oceania 7:215–233. PhD thesis, Australian National University, Canberra. ———. 1976. Yams: Dioscorea spp. (Dioscoreaceae). In Evolution of ———. 2001. Wetland drainage and agricultural transformations in crop plants. N. W. Simmonds, ed. Pp. 70–74. London: Longman. the southern highlands of Papua New Guinea. Asia Pacific View- Daniels, J., and C. Daniels. 1993. Sugarcane in prehistory. Archaeology point 42:287–304. in Oceania 28:1–7. Barker, G. 2006. The agricultural revolution in prehistory. Oxford: De Candolle, A. 1884. Origin of cultivated plants. London: Kegan Oxford University Press. Paul. Barrau, J. 1955. Subsistence agriculture in Melanesia. Noumea, New De Langhe, E., L. Vrydaghs, P. de Maret, X. Perrier, and T. P.Denham. Caledonia: South Pacific Commission. 2009. Why bananas matter: an introduction to the history of ba- Barton, H. 2005. The case for rainforest foragers: the starch record nana domestication. Ethnobotany Research and Applications 7:165– at Niah cave, Sarawak. Asian Perspectives 44:56–72. 177. Barton, H., and T. P. Denham. Forthcoming. Prehistoric vegeculture Denham, T. P. 2003a. Archaeological evidence for mid-Holocene ag- and social life in Island Southeast Asia and Melanesia. In Why riculture in the interior of Papua New Guinea: a critical review. cultivate? anthropological and archaeological perspectives on forag- Special issue, Archaeology in Oceania 38:159–176. Denham Early Agriculture and Plant Domestication S393

———. 2003b. The Kuk morass: multi-disciplinary investigations of Fairbairn, S. O’Connor, and B. Marwick, eds. Pp. 139–154. Terra early to mid Holocene plant exploitation at Kuk Swamp, Wahgi Australis 28. Canberra: ANU E Press. Valley, Papua New Guinea. PhD thesis, Australian National Uni- Denham, T. P., J. Iriarte, and L. Vrydaghs, eds. 2007. Rethinking versity, Canberra. agriculture: archaeological and ethnoarchaeological perspectives.Wal- ———. 2004a. Early agriculture in the highlands of New Guinea: nut Creek, CA: Left Coast. an assessment of Phase 1 at Kuk Swamp. In A Pacific odyssey: Denham, T. P., K. Sniderman, K. Saunders, B. Winsborough, and A. archaeology and anthropology in the western Pacific: papers in honour Pierret. 2009. Contiguous multi-proxy analyses (x-radiography, di- of Jim Specht. V. Attenbrow and R. Fullagar, eds. Pp. 47–57. Records atom, pollen and microcharcoal) of Holocene archaeological fea- of the Australian Museum, suppl. 29. Sydney: Australian Museum. tures at Kuk Swamp, Upper Wahgi Valley, Papua New Guinea. ———. 2004b. The roots of agriculture and arboriculture in New Geoarchaeology 24:715–742. Guinea: looking beyond Austronesian expansion, Neolithic pack- Diamond, J. 2002. Evolution, consequences and future of animal and ages and indigenous origins. World Archaeology 36:610–620. plant domestication. Nature 418:700–707. ———. 2005a. Agricultural origins and the emergence of rectilinear Donoghue, D. 1989. Carbonised plant fossils. In Plants in Australian ditch networks in the highlands of New Guinea. In Papuan pasts: archaeology. W. Beck, A. Clarke, and L. Head, eds. Pp. 90–100. cultural, linguistic and biological histories of Papuan-speaking peo- Tempus 1. St. Lucia: University of Queensland. ples. A. Pawley, R. Attenborough, J. Golson, and R. Hide, eds. Pp. Donohue, M., and T. P. Denham. 2009. Banana (Musa spp.) do- 329–361. Pacific Linguistics 572. Canberra: RSPAS, ANU. mestication in the Asia-Pacific region: linguistic and archaeological ———. 2005b. Envisaging early agriculture in the highlands of New perspectives. Ethnobotany Research and Applications 7:293–332. Guinea: landscapes, plants and practices. World Archaeology 37: ———. 2010. Island Southeast Asia during the mid-Holocene: re- 290–306. framing Austronesian history. Current Anthropology 51:223–256. ———. 2006. The origins of agriculture in New Guinea: evidence, Fullagar, R., J. Field, T. P. Denham, and C. Lentfer. 2006. Early and interpretation and reflection. In Blackwell guide to archaeology in mid-Holocene processing of taro (Colocasia esculenta) and yam Oceania: Australia and the Pacific Islands. I. Lilley, ed. Pp. 160– (Dioscorea sp.) at Kuk Swamp in the highlands of Papua New 188. Oxford: Blackwell. Guinea. Journal of Archaeological Science 33:595–614. ———. 2007. Thinking about plant exploitation in New Guinea: ———. 2008. Stone tools and early agriculture at Kuk Swamp, Papua towards a contingent interpretation of agriculture. In Rethinking New Guinea. In “Prehistoric Technology” 40 years later: functional agriculture: archaeological and ethnoarchaeological perspectives.T.P. studies and the Russian legacy. L. Longo and N. Skakun with M. Denham, J. Iriarte, and L. Vrydaghs, eds. Pp. 78–108. Walnut Saracino and M. Dalla Riva, eds. Pp. 395–398. BAR S1783. Oxford: Creek, CA: Left Coast. Archaeopress. ———. 2008. Environmental archaeology: interpreting practices-in- Fuller, D. Q., R. Korisettar, P. C. Venkatasubbaiah, and M. K. Jones. the-landscape through geoarchaeology. In Handbook of landscape 2004. Early plant domestication in southern India: some prelim- archaeology. B. David and J. Thomas, eds. Pp. 468–481. Walnut inary archaeobotanical results. Vegetation History and Archaeo- Creek, CA: Left Coast. botany 13:115–129. ———. 2009. A practice-centred method for charting the emergence Fuller, D. Q., and M. Madella. 2009. Banana cultivation in South and transformation of agriculture. Current Anthropology 50:661– Asia and East Asia: a review of the evidence from archaeology and 667. linguistics. Ethnobotany Research and Applications 7:333–351. ———. 2010. From domestication histories to regional prehistory: Giddens, A. 1984. The constitution of society. Cambridge: Polity. using plants to re-evaluate early and mid-Holocene interaction Gillieson, D., P. Gorecki, J. Head, and G. Hope. 1987. Soil erosion between New Guinea and Southeast Asia. Food and History 8:3– and agricultural history in the central highlands of Papua New 22. Guinea. In International geomorphology, pt. 2. V. Gardiner, ed. Pp. Denham, T. P., and M. Donohue. 2009. Pre-Austronesian dispersal 507–522. London: Wiley. of banana cultivars west from New Guinea: linguistic relics from Gillieson, D., P. Gorecki, and G. Hope. 1985. Prehistoric agricultural eastern Indonesia. Archaeology in Oceania 44:18–28. systems in a lowland swamp, Papua New Guinea. Archaeology in Denham, T. P., M. Donohue, and S. Booth. 2009. Revisiting an old Oceania 20:32–37. hypothesis: horticultural experimentation in northern Australia. Golson, J. 1977. No room at the top: agricultural intensification in Antiquity 83:634–648. Denham, T. P., R. Fullagar, and L. Head. 2009. Plant exploitation on the New Guinea Highlands. In Sunda and Sahul. J. Allen, J. Golson, Sahul: from colonisation to the emergence of regional specialisa- and R. Jones, eds. Pp. 601–638. London: Academic Press. tion during the Holocene. Quaternary International 202:29–40. ———. 1982. The Ipomoean revolution revisited: society and sweet Denham, T. P., J. Golson, and P. J. Hughes. 2004. Reading early potato in the Upper Wahgi Valley. In Inequality in New Guinea agriculture at Kuk (Phases 1–3), Wahgi Valley, Papua New Guinea: Highland societies. A. Strathern, ed. Pp. 109–136. Cambridge: Cam- the wetland archaeological features. Proceedings of the Prehistoric bridge University Press. Society 70:259–298. ———. 1991. The New Guinea Highlands on the eve of agriculture. Denham, T. P., and S. G. Haberle. 2008. Agricultural emergence and Bulletin of the Indo-Pacific Prehistory Association 11:82–91. transformation in the Upper Wahgi Valley during the Holocene: ———. 1997. The Tambul spade. In Work in progress: essays in New theory, method and practice. Holocene 18:499–514. Guinea Highlands ethnography in honour of Paula Brown Glick.H. Denham, T. P., S. G. Haberle, and C. Lentfer. 2004. New evidence Levine and A. Ploeg, eds. Pp. 142–171. Oxford: Lang. and interpretations for early agriculture in highland New Guinea. ———. 2002. Gourds in New Guinea, Asia and the Pacific. In Fifty Antiquity 78:839–857. years in the field: essays in honour and celebration of Richard Shutler Denham, T. P., S. G. Haberle, C. Lentfer, R. Fullagar, J. Field, M. Jr.’s archaeological career. S. Bedford, C. Sand, and D. Burley, eds. Therin, N. Porch, and B. Winsborough. 2003. Origins of agricul- Pp. 69–78. New Zealand Archaeological Journal Monograph 25. ture at Kuk Swamp in the highlands of New Guinea. Science 301: Auckland, New Zealand: Auckland Museum. 189–193. Golson, J., and D. Gardner. 1990. Agriculture and socio-political Denham, T. P., S. G. Haberle, and A. Pierret. 2009. A multi-disci- organisation in New Guinea Highlands prehistory. Annual Review plinary method for the investigation of early agriculture: learning of Anthropology 19:395–417. lessons from Kuk. In New directions in archaeological science.A. Golson, J., and P. J. Hughes. 1980. The appearance of plant and S394 Current Anthropology Volume 52, Supplement 4, October 2011

animal domestication in New Guinea. Journal de la Socie´te´des Ingold, T. 2000. The perception of the environment: essays on livelihood, Oce´anistes 36:294–303. dwelling and skill. London: Routledge. Golson, J., R. J. Lampert, J. M. Wheeler, and W. R. Ambrose. 1967. Irwin, S. V., P. Kaufusi, K. Banks, R. de la Pena, and J. J. Cho. 1998. A note on carbon dates for horticulture in the New Guinea High- Molecular characterisation of taro (Colocasia esculenta) using lands. Journal of the Polynesian Society 76:369–371. RAPD markers. Euphytica 99:183–189. Gorecki, P., and D. S. Gillieson. 1984. The highland fringes as a key Jones, M. K., and T. Brown. 2007. Selection, cultivation and repro- zone for prehistoric developments in Papua New Guinea: a prog- ductive isolation: a reconsideration of the morphological and mo- ress report. Bulletin of the Indo-Pacific Prehistory Association 5:93– lecular signals of domestication. In Rethinking agriculture: archae- 103. ological and ethnoarchaeological perspectives.T.P.Denham,J. ———. 1989. A crack in the spine: prehistory and ecology of the Jimi- Iriarte, and L. Vrydaghs, eds. Pp. 36–49. Walnut Creek, CA: Left Yuat River, Papua New Guinea. Townsville, Queensland: James Coast. Cook University. Jones, R., and B. Meehan. 1989. Plant foods of the Gidjingal: eth- Gremillion, K. J., and D. Piperno. 2009. Human behavioral ecology, nographic and archaeological perspectives from northern Australia phenotypic (developmental) plasticity, and agricultural origins: in- on tuber and seed exploitation. In Foraging and farming: the evo- sights from the emerging evolutionary synthesis. Current Anthro- lution of plant exploitation. D. R. Harris and G. C. Hillman, eds. pology 50:615–619. Pp. 120–135. London: Unwin Hyman. Grivet, L., C. Daniels, J. C. Glaszman, and A. D’Hont. 2004. A review Kahlheber, S., and K. Neumann. 2007. The development of plant of recent molecular genetics evidence for sugarcane evolution and cultivation in semi-arid West Africa. In Rethinking agriculture: ar- domestication. Ethnobotany Research and Applications 2:9–17. chaeological and ethnoarchaeological perspectives. T. P. Denham, J. Haberle, S. G. 1993. Late Quaternary environmental history of the Iriarte, and L. Vrydaghs, eds. Pp. 320–346. Walnut Creek, CA: Left Tari Basin, Papua New Guinea. PhD thesis, Australian National Coast. University, Canberra. Kennedy, J. 2008. Pacific bananas: complex origins, multiple disper- ———. 1994. Anthropogenic indicators in pollen diagrams: prob- sals? Asian Perspectives 47:75–94. lems and prospects for late Quaternary palynology in New Guinea. Kennedy, J., and W. Clarke. 2004. Cultivated landscapes of the South- In Tropical archaeobotany: applications and new developments.J.G. west Pacific. RMAP Working Paper 50. Canberra, Australian Cap- Hather, ed. Pp. 172–201. London: Routledge. ital Territory: RSPAS ANU. ———. 1995. Identification of cultivated Pandanus and Colocasia in Kjær, A., A. S. Barfod, C. B. Amussen, and O. Seberg. 2004. Inves- pollen records and the implications for the study of early agri- tigation of genetic and morphological variation in the sago palm culture in New Guinea. Vegetation History and Archaeobotany 4: (Metroxylon sagu; Arecaceae) in Papua New Guinea. Annals of 195–210. Botany 94:109–117. ———. 2003. The emergence of an agricultural landscape in the Kurvilla, K. M., and A. Singh. 1981. Karyotypic and electrophoretic highlands of New Guinea. Archaeology in Oceania 38:149–158. studies on taro and its origin. Euphytica 30:405–413. ———. 2005. A 22 Ka pollen record from Lake Euramoo, wet tropics Ladizinsky, G. 1998. Plant evolution under domestication. Dordrecht: of northeast Queensland, Australia. Quaternary Research 64:343– Kluwer. 356. Lampert, R. J. 1967. Horticulture in the New Guinea Highlands: C14 ———. 2007. Prehistoric human impact on rainforest biodiversity dating. Antiquity 41:307–309. in Highland New Guinea. Philosophical Transactions of the Royal ———. 1970. Appendix 5: archaeological report of the Minjigina site. In The impact of man on the vegetation of the Mount Hagen Society B 362:219–228. region, New Guinea. J. M. Powell. PhD thesis, Australian National Haberle, S. G., and A. Chepstow-Lusty. 2000. Can climate influence University, Canberra. cultural development? a view through time. Environmental History Lebot, V. 1999. Biomolecular evidence for plant domestication in 6:349–369. Sahul. Genetic Resources and Crop Evolution 46:619–628. Harris, D. R. 2007. Agriculture, cultivation and domestication: ex- Lebot, V., and K. M. Aradhya. 1991. Isozyme variation in taro (Col- ploring the conceptual framework of early food production. In ocasia esculenta (L.) Schott) from Asia and Oceania. Euphytica 56: Rethinking agriculture: archaeological and ethnoarchaeological per- 55–66. spectives. T. P. Denham, J. Iriarte, and L. Vrydaghs, eds. Pp. 16– Lebot, V., M. S. Prana, N. Kreike, H. van Heck, J. Pardales, T. Okpul, 35. Walnut Creek, CA: Left Coast. T. Gendua, et al. 2004. Characterisation of taro (Colocasia esculenta Harris, E. C., and P. J. Hughes. 1978. An early agricultural system (L.) Schott) genetic resources in Southeast Asia and Oceania. Ge- at Mugumamp Ridge, Western Highlands Province, Papua New netic Resources and Crop Evolution 51:381–392. Guinea. Mankind 11:437–445. Lebot, V., B. Trilles, J. L. Noyer, and J. Modesto. 1998. Genetic re- Heidegger, M. 1962. Being and time. J. Macquarrie and E. Robinson, lationships between Dioscorea alata L. cultivars. Genetic Resources trans. Oxford: Blackwell. and Crop Evolution 45:499–509. Hillman, G. C., and M. S. Davies. 1990. Measured domestication Lentfer, C. 2009. Tracing domestication and cultivation of bananas rates in wild wheats and barley under primitive cultivation, and from phytoliths: an update from New Guinea. Ethnobotany Re- their archaeological implications. Journal of World Prehistory 4: search and Applications 7:247–270. 157–222. Loy, T., M. Spriggs, and S. Wickler. 1992. Direct evidence for human Hodder, I. 1999. The archaeological process: an introduction. Oxford: use of plants 28,000 years ago: starch residues on stone artefacts Blackwell. from northern Solomon Islands. Antiquity 66:898–912. Hope, G. S., and J. Golson. 1995. Late Quaternary change in the Malapa, R., G. Arnau, J. L. Noyer, and V. Lebot. 2005. Genetic di- mountains of New Guinea. Antiquity 69:818–830. versity of the greater yam (Dioscorea alata L.) and relatedness to Hope, G. S., and S. G. Haberle. 2005. The history of the human D. nummularia Lam. and D. transversa Br. as revealed with AFLP landscapes of New Guinea. In Papuan pasts: cultural, linguistic and markers. Genetic Resources and Crop Evolution 52:919–929. biological histories of Papuan-speaking peoples. A. Pawley, R. At- Marshall, F. 2007. African pastoral perspectives on domestication of tenborough, J. Golson, and R. Hide, eds. Pp. 541–554. Pacific the donkey: a first synthesis. In Rethinking agriculture: archaeolog- Linguistics 572. Canberra, Australian Capital Territory: RSPAS, ical and ethnoarchaeological perspectives. T. P. Denham, J. Iriarte, ANU. and L. Vrydaghs, eds. Pp. 371–407. Walnut Creek, CA: Left Coast. Denham Early Agriculture and Plant Domestication S395

Martin, F. W., and A. M. Rhodes. 1977. Intra-specific classification Simmonds, N. W. 1976a. Bananas: Musa (Musaceae). In Evolution of Dioscorea alata. Tropical Agriculture 54:1–13. of crop plants. N. W. Simmonds, ed. Pp. 211–215. London: Long- Matthews, P. J. 1991. A possible wild type taro: Colocasia esculenta man. var. aquatilis. Indo-Pacific Prehistory Association Bulletin 11:69–81. ———. 1976b. Sugarcanes: Saccharum (Gramineae-Andropogo- ———. 1995. Aroids and Austronesians. Tropics 4:105–126. neae). In Evolution of crop plants. N. W. Simmonds, ed. Pp. 104– ———. 2003. Taro planthoppers (Tarophagus spp.) in Australia and 108. London: Longman. the origins of taro (Colocasia esculenta) in Oceania. Archaeology Smith, B. D. 2001. Low-level food production. Journal of Archaeo- in Oceania 38:159–176. logical Research 9:1–43. Mbida Mindzie, C., H. Doutrelpont, L. Vrydaghs, R. L. Sewennen, Sniderman, J. M. K., J. Finn, and T. P. Denham. 2009. A late Holocene R. J. Swennen, H. Beeckman, E. de Langhe, and P. de Maret. 2001. palaeoecological record from Ambra Crater in the highlands of First archaeological evidence of banana cultivation in central Africa Papua New Guinea and implications for agricultural history. Ho- during the third millennium before present. Vegetation History and locene 19:449–458. Archaeobotany 10:1–6. Stover, R. H., and N. W. Simmonds, eds. 1987. Bananas. 3rd edition. Muke, J., and H. Mandui. 2003. In the shadows of Kuk: evidence Tropical Agriculture Series. Harlow, UK: Longman. for prehistoric agriculture at Kana, Wahgi Valley, Papua New Tanno, K.-I., and G. Willcox. 2006. How fast was wild wheat do- Guinea. Archaeology in Oceania 38:177–185. mesticated? Science 311:1886. Neumann, K., and E. Hildebrand. 2009. Early bananas in Africa: the Terrell, J. E. 2002. Tropical agroforestry, coastal lagoons and Holocene state of the art. Ethnobotany Research and Applications 7:353–362. prehistory in Greater Near Oceania. In Proceedings of the Inter- Pawley, A. 2007. The origins of early Lapita culture: the testimony national Area Studies Conference VII: vegeculture in Eastern Asia of historical linguistics. In Oceanic explorations: Lapita and Western and Oceania. Y. Shuji and P. J. Matthews, eds. Pp. 195–216. Osaka, Pacific settlement. S. Bedford, C. Sand, and S. P. Connaughton, eds. Pp. 17–49. Canberra, Australian Capital Territory: ANU. Japan: National Museum of Ethnology. Paz, V. 2005. Rock shelters, caves, and archaeobotany in Island South- Thomas, J. 1996. The cultural context of the first use of domesticates east Asia. Asian Perspectives 44:107–118. in continental Central and Northwest Europe. In The origins and Pearsall, D. M. 2007. Modelling prehistoric agriculture through the spread of agriculture and pastoralism in Eurasia. D. R. Harris, ed. palaeoenvironmental record: theoretical and methodological is- Pp. 310–322. London: University College London Press. sues. In Rethinking agriculture: archaeological and ethnoarchaeolog- Watson, J. B. 1964. A previously unreported root crop from the New ical perspectives. T. P. Denham, J. Iriarte, and L. Vrydaghs, eds. Pp. Guinea Highlands. Ethnology 3:1–5. 210–230. Walnut Creek, CA: Left Coast. Watson, V. D., and J. D. Cole. 1977. Prehistory of the eastern highlands Perrier, X., F. Bakry, F. Carreel, C. Jenny, J.-P. Horry, V. Lebot, and of New Guinea. Seattle: University of Washington Press. I. Hippolyte. 2009. Combining biological approaches to shed light White, J. P. 1972. Ol Tumbuna: archaeological excavations in the east- on the evolution of edible bananas. Ethnobotany Research and ern central highlands, Papua New Guinea. Terra Australis 2. Can- Applications 7:199–216. berra, Australian Capital Territory: Department of Prehistory, Re- Piggott, S. 1954. The Neolithic cultures of the British Isles. Cambridge: search School of Pacific Studies, Australian National University. Cambridge University Press. Yen, D. 1973. The origins of Oceanic agriculture. Archaeology and Powell, J. M. 1970. The impact of man on the vegetation of the Physical Anthropology in Oceania 8:68–85. Mount Hagen region, New Guinea. PhD thesis, Australian Na- ———. 1985. Wild plants and domestication in Pacific islands. In tional University, Canberra. Recent advances in Indo-Pacific prehistory. V. N. Misra and P. Bell- ———. 1982. The history of plant use and man’s impact on the wood, eds. Pp. 315–326. New Dehli: Oxford and IBH. vegetation. In Biogeography and ecology of New Guinea, vol. 1. J. ———. 1991. Domestication: the lessons from New Guinea. In Man L. Gressitt, ed. Pp. 207–227. The Hague, Netherlands: Junk. and a half: essays in Pacific anthropology and ethnobiology in honour Renfrew, C. 1973. Before civilization: the radiocarbon revolution and of Ralph Bulmer. A. Pawley, ed. Pp. 558–569. Auckland: Polynesian prehistoric Europe. Harmondsworth, UK: Penguin. Society. ———. 2002. “The emerging synthesis”: the archaeogenetics of farm- ———. 1995. The development of Sahul agriculture with Australia ing/language dispersals and other spread zones. In Examining the as bystander. Antiquity 69:831–847. farming/language dispersal hypothesis. P. Bellwood and C. Renfrew, ———. 1996. Melanesian arboriculture: historical perspectives with eds. Pp. 3–16. Cambridge: McDonald Institute for Archaeological emphasis on the genus Canarium.InSouth Pacific indigenous nuts. Research. B. R. Evans, R. M. Bourke, and P. Ferrar, eds. Pp. 36–44. Canberra, Richerson, P. J., R. Boyd, and R. L. Bettinger. 2001. Was agriculture Australian Capital Territory: ACIAR. impossible during the Pleistocene but mandatory during the Ho- locene? a climate change hypothesis. American Antiquity 66:387– ———. 1998. Subsistence to commerce in Pacific agriculture: some 411. four thousand years of plant exchange. In Plants for food and Roscoe, P. 2002. The hunters and gatherers of New Guinea. Current medicine. H. D. V. Pendergast, N. L. Etkin, D. R. Harris, and P. J. Anthropology 43:153–162. Houghton, eds. Pp. 161–183. Kew, UK: Royal Botanic Gardens. Sayer, A. 1984. Method in social science: a realist approach. London: ———. 2003. The nature of domestication in agricultural formation. Hutchinson. Paper presented at World Archaeological Congress 5, Washington, Scarre, C. 1988. Past worlds: the Times atlas of archaeology. London: DC. The Times. Zerega, N. J. C., D. Ragone, and T. J. Motley. 2004. Complex origins Sharrock, S. 2001. Diversity in the genus Musa: focus on Australi- of breadfruit (Artocarpus altilis, Moraceae): implications for hu- musa. INIBAP annual report 2000. Pp. 14–19. Montpellier, France: man migrations in the Pacific. American Journal of Botany 91:760– INIBAP. 766.

Current Anthropology Volume 52, Supplement 4, October 2011 S397

Domestication Processes and Morphological Change Through the Lens of the Donkey and African Pastoralism

by Fiona Marshall and Lior Weissbrod

Little is known about the beginnings and spread of food production in the tropics, but recent research suggests that definitions that depend on morphological change may hamper recognition of early farming in these regions. The earliest form of food production in Africa developed in arid tropical grasslands. Animals were the earliest domesticates, and the mobility of early herders shaped the development of social and economic systems. Genetic data indicate that cattle were domesticated in North Africa and suggest domestication of two different African wild asses, in the Sahara and in the Horn. Cowpeas and pearl millet were domesticated several thousand years later, but some intensively used African plants have never undergone morphological change. Morphological, genetic, ethno- archaeological, and behavioral research reveals relationships between management, animal behavior, selection, and domestication of the donkey. Donkeys eventually showed phenotypic and morpho- logical changes distinctive of domestication, but the process was slow. This African research on domestication of the donkey and the development of pastoralism raises questions regarding how we conceptualize hunter-gatherer versus food-producer land use. It also suggests that we should focus more intently on the methods used to recognize management, agropastoral systems, and domesti- cation events.

The question of whether understanding of the beginnings of (Kahlheber and Neumann 2007; Marshall and Hildebrand food production is being constrained by definitions and meth- 2002; for North America, Smith 2001, 2011). ods of detection that focus on morphological change rather In their approach to definitions and the question of whether than management is becoming a major theme in studies of morphological change is an effective marker of domestication, the origins of agriculture. Recent research in the humid tropics Jones and Brown (2007) focus on selection processes and of southeastern Asia and the Pacific suggests that definitions timing rather than on region. They contend that under certain that depend on morphological change hamper recognition of circumstances, practices of cultivation and protective tending early farming in these areas (Bayliss-Smith 2007; Denham could have resulted in stable long-term systems of food pro- 2007, 2011). This perspective has so far centered on plants duction that depended on plants and animals lacking dis- of the humid tropics that have a history of long-term culti- tinctively domestic morphological and genetic characteristics. vation in agricultural systems but lack morphological change Reproductive isolation and morphological change, Jones and (Denham 2007; Kahlheber and Neumann 2007; Yen 1989). Brown (2007) go on to suggest, are linked with later stages Another feature of both humid and arid tropical agricultural of agricultural development, when human populations ex- practices that has strained conceptions of early agricultural panded and people removed plants and animals from their systems is the variety of economic activities—including fish- wild ranges. ing, gathering, hunting, cultivation, and herding—that may There is a growing appreciation, however, of differences be combined in complex and diverse subsistence systems among species in time elapsed before domestication processes are readily detectable and of variability in the sensitivity of methods that can be brought to bear on any given taxon. In Fiona Marshall is Professor in the Department of Anthropology, a detailed study of the domestication of goats in western Asia, Washington University, Saint Louis (1 Brooking Drive, Saint Louis, Missouri 63130, U.S.A. [[email protected]]). LiorWeissbrod Zeder (Zeder 2008; Zeder and Hesse 2000) used regional and is a postdoctoral researcher at the Zinman Institute of Archaeology, age- and sex-based variability in animal size to document early University of Haifa (Mount Carmel, Haifa 31905, Israel herd management, which was followed by diminution in size. [[email protected]]). This paper was submitted 13 XI 09, In the absence of clear morphological indicators, evidence for accepted 2 XII 10, and electronically published 8 VI 11. management—culling, corralling, and milking—has also been

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0016$10.00. DOI: 10.1086/658389 S398 Current Anthropology Volume 52, Supplement 4, October 2011 key to a better understanding of early phases of domestication us to return to the larger question of Africa’s contribution to of the horse (Outram et al. 2009). The discovery by Rossel understanding variability in early agricultural systems world- et al. (2008) that donkeys used by Egyptian pharaohs for wide. In most of Africa, pastoralism is considered the earliest transport at approximately 5000 cal BP (historic date 3000 form of agriculture, followed by plant cultivation and adop- BC; table 1) remained morphologically wild 1,000 years after tion of mixed herding-cultivation systems. they were thought to have been first domesticated further emphasizes possibilities for underestimating the timing of do- mestication of large mammals and draws attention to species- Early Food Production in Africa specific pathways to domestication (see also Zeder 2011). In the light of these different emphases on global, regional, and taxon-specific impacts of late morphological change on Africanists have built up a picture of the beginnings of food general understanding of early food production, we evaluate production in which early dependence on domestic animals current perspectives on the beginnings of food production in and increasing reliance on mobility guided the development Africa, a continent that represents the world’s largest tropical of social and economic systems of the Early Holocene and landmass. We reexamine evidence of early animal and plant resulted in late domestication of African plants. Specific domesticates and employ ethnoarchaeological data on donkey themes that have emerged include locally and socially con- management and breeding behavior to examine species- tingent responses to large-scale climatic change, domestica- specific domesticatory practices that influenced selection and tion of cattle for food and donkeys for transport, intensive the likelihood of morphological change. These analyses allow hunting and possible management of Barbary sheep, long-

Table 1. Key African animal and plant domesticates, with summaries of sites, date ranges, and arguments for management or domestication processes

Taxon, region Site, period (cal BP) Argument for management/domestication Barbary sheep: Acacus Mountains (Libya) Uan Afuda Uan Tabu; tenth to ninth millennium Caprine dung in shelter suggests penning; plant re- mains suggest foddering Cattle: West Nile (Egypt), arid grasslands Nabta Playa; eleventh to tenth millennium Cattle out of wild range; provisioning of water from well Modern African cattle Sheko, Ethiopia; Ndama, Mali Modern cattle genetics: mitochondrial, microsatellite, Y-chromosome, Y2 data distinctive of Africa; geo- graphic patterning Donkeys: Nile Valley (Egypt) Maadi, Merimde, Hierakonpolis; Donkey bones, some smaller than wild ancestor’s mid-seventh millennium Abydos; ∼5000 cal BP Donkey skeletons of same size as wild ancestor’s but with pathologies distinctive of transport use Nubian/Somali wild ass, donkeys Sahara, Horn Modern donkey and wild ass genetics: mitochondrial, genetic variability, ancient DNA Sorghum: Sahel Um Direiwa; seventh millennium Intensive use of grains, no morphological change, 20,000 grindstone fragments Sahel Qasr Ibrm; early second millennium Morphological change Pearl millet: West African Sahel KN05, Birimi, Gajiganna, Dhar Tichitt; Morphological change, size decrease late third/early fourth millennium Cow pea: West Africa savanna, forest B-sites, Birimi Early fourth millennium Morphological change Teff: Ethiopia, high altitude Aksum, Ona Nagast; early second millennium Morphological change, no size decrease Roselle, baobab, shea butter: West African park savanna Early second millennium Common in sites; heavily used today, but no mor- phological change Oil palm, canarium, yams: West African tropical forests Early second millennium Common in sites, heavily used, no morphological change; yams: no preservation, morphological change? Marshall and Weissbrod Domestication Processes and Morphological Change S399 term reliance on a broad range of wild plants and animals, the Acacus Mountains of Libya, and Wadi Howar (Garcea and late domestication of African plants. 2004, 2006; Po¨llath and Peters 2007; fig. 1). In this review of the African evidence, we see domestication During the Early and mid-Holocene, diverse hunter-gath- as a microevolutionary process that transformed animal and erer groups lived close to permanent water in widely separated plant communities and human societies (see Clutton-Brock regions of northeastern Africa, from the Acacus to Lake Vic- 1992), but we examine rather than assume relationships be- toria (Caneva 1988; Garcea 2006; Holl 2005; Prendergast and tween domestication and long-term genetic and morpholog- Lane 2010). Ethnoarchaeological research suggests that this ical change (see also Vigne et al. 2011). We follow Zeder (2009, social and economic variability played a significant role in 2011; Rindos 1984) in emphasizing long-term coevolutionary pathways to food production in Africa. Recent hunter-gath- relations between people, animals, and plants, but unlike erers with long-term investment in hive and trap construction Rindos (1984), we also highlight the intentional role that and delayed-return social systems and limited sharing have individuals played in selection (Hildebrand 2003b; Marshall historically been able to accommodate more easily property- and Hildebrand 2002). Pastoralism is also an important con- rights issues arising out of time investment in agriculture than cept for discussions of the beginnings of food production in have those with highly egalitarian norms (Brooks, Gelburd, Africa, and this, we argue, differs from herding or simple and Yellen 1984; Dale, Marshall, and Pilgram 2004; Marshall keeping of animals because pastoralists rely on moving live- 2000; Smith 1998; Woodburn 1982). Moreover, cattle herding stock to pasture and emphasize the social and symbolic role requires significantly greater commitment than cultivation be- of domestic animals (Dyson-Hudson and Dyson-Hudson cause foragers can tend crops intermittently and accommo- 1980; Smith 2005; Spear and Waller 1993). This does not date them into flexible hunter-gatherer schedules, whereas necessarily imply, however, a diet heavily based on domestic animal herds require protection against predators and con- animals. Historically, African pastoralists prioritized the needs stant attention (Dale, Marshall, and Pilgram 2004; Marshall of their herds in scheduling activities and locating settlements 2000). As a result, Africanists have hypothesized that do- (McCabe 2004; Western and Dunne 1979), but they usually mestication of cattle is more likely to have been undertaken relied on a broad range of complementary subsistence strat- and pastoralism adopted in regions of northeastern Africa egies ranging from seasonal cultivation, fishing, hunting, and that were occupied by complex rather than highly mobile gathering to food exchange or trade (Dyson-Hudson and egalitarian hunter-gatherers (Marshall and Hildebrand 2002). Dyson-Hudson 1980; Evans-Pritchard 1940; Schneider 1979). Arguments that complex or delayed-return systems of so- As a result, it is overly simplistic to rely on high proportions cial organization existed in the Acacus, the Sudanese Nile of domestic animal bones to differentiate pastoral from Valley, and some other regions of the African Early to mid- hunter-gatherer or farming sites. Multiple lines of evidence Holocene are based on elaboration of material culture, in- are necessary, including households oriented to mobility— cluding manufacture of ceramics and storage facilities in these with slope, soil, and vegetation characteristics organized areas and highly patterned use of rock-shelter sites and local around the needs of domestic herds (Western and Dunne landscapes (Barich 1987; di Lernia 1999, 2001; Garcea 2004; 1979)—animal pens, dung deposits (Shahack-Gross, Mar- McDonald 2008). Significant investment in living spaces and shall, and Weiner 2003; Shahack-Gross, Simons, and Ambrose limited movement are indicated by hut construction at Nabta 2008), milk residues (see Evershed et al. 2008), livestock- Playa in the Acacus Mountains and the northern Sudanese focused rock art, and ritual livestock burials (di Lernia 2006). Nile Valley and by isotopic analyses at Gobero in Niger and Acacus sites (Barich 1987; Garcea 2006; Sereno et al. 2008; Tafuri et al. 2006). In the central Sahara, the Sudanese Nile Domesticatory Settings: Climatic and Social Variability and Valley, and the Acacus, human burials are common (Caneva Subsistence Intensification 1988; Honegger 2004; Sereno et al. 2008). Garcea (2004) and Large-scale climate change forms the backdrop to the begin- di Lernia (1999, 2001) argue that their presence in the Late nings of food production in northeastern Africa (Kro¨pelin et Acacus phase (ca. 10,250–9600 to 9890–9440 cal BP) may al. 2008). Hunter-gatherer communities deserted most of the relate to group identities and rights to land. northern interior of the continent during the arid glacial max- North African hunter-gatherers of the Early and mid- imum and took refuge along the North African coast, the Holocene employed highly diverse subsistence as well as social Nile Valley, and the southern fringes of the Sahara (Barich systems. Wild cattle (Bos primigenius) were hunted along the and Garcea 2008; Garcea 2006; Kuper and Kro¨pelin 2006). Mediterranean coast and the Nile Valley, and small numbers During the subsequent Early Holocene African humid phase, of wild ass (Equus africanus) were also present in many sites from the mid-eleventh to the early ninth millennium cal BP, (Alhaique and Marshall 2009; Gautier 1987a; Marshall 2007). ceramic-using hunter-gatherers took advantage of more fa- Barbary sheep (Ammotragus lervia) were the most common vorable savanna conditions to resettle much of northeastern animal hunted across North Africa at this time (di Lernia 2001; Africa (Holl 2005; Kuper and Kro¨pelin 2006). Evidence of Gautier 1987a; Saxon et al. 1974). In the Late Acacus sites of domestic animals first appeared in sites in the Western Desert Ti-n-Torah, Uan Tabu, and Uan Afuda, intensive exploitation of Egypt, the Khartoum region of the Nile, northern Niger, of wild cereals (e.g., Echinochloa, Panicum, Setaria, Digitaria, S400 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 1. Map of North Africa showing the location of sites mentioned in the text. and Pennisetum) is associated with heavy grindstone use (di di Lernia 2001; Garcea 2006). Di Lernia (2001) argues that Lernia 1999; Garcea 2001; Mercuri 2001; fig. 1). A similar set dense dung deposits in these rock shelters differ from natural of wild grass seeds were harvested, processed, and stored in the dung accumulations characterized by loose and scattered pellet eastern Sahara during the late tenth and early ninth millennia matrices and result instead from use of shelters for corralling at Nabta Playa, site E-75-6 (Wasylikowa et al. 1993; Wendorf animals. Micromorphological analyses of the “dung layer” sed- and Schild 1998; for radiocarbon dates, see table 2). Along the iments suggest trampling and indicate the presence of spher- Sudanese Nile, a variety of wild mammals were hunted in con- ulites common in caprine dung, and studies of the plant re- junction with fishing for large deepwater fish and intensive mains indicate a selected range of plant species suggestive of grindstone use (Caneva 1988; Haaland 1987). foddering (Castelletti et al. 1999; di Lernia 2001; Mercuri 1999). Interestingly, Livingstone Smith (2001) notes that hunter-gath- Taming of Barbary sheep. There has been a recurrent suggestion erer pottery of Late Acacus levels at Uan Afuda is dung tem- that some North Africans penned and culled Barbary sheep pered, a characteristic of later pastoral ceramics. The number herds during early phases of the Holocene (di Lernia 1998, of Barbary sheep remains declines in later sites, however, and 2001; Garcea 2006; Saxon et al. 1974; table 1). Earlier arguments there are no dung deposits that suggest subsequent emphasis for management without morphological change were based on on Barbary sheep (di Lernia 1999; Garcea 2001, 2004). Taken young male–dominated culling profiles from the sites of Tamar together, the micromorphological and archaeological evidence Hat and Haua Fteah on the Mediterranean coast (Saxon et al. for dung accumulation resulting from penning of Barbary sheep 1974; Smith 2008; fig. 1). More recent evidence is based on the in the Late Acacus rock shelters is suggestive, but additional presence of dung accumulations in the rear of rock-shelter sites faunal data and dung deposits are needed from open-air sites. occupied by complex hunter-gatherers during the tenth and early ninth millennia cal BP in the Libyan Acacus at Uan Afuda, Domestication of African cattle? The evidence for taming of Uan Tabu, and Fozzigiaren (Cremaschi and Trombino 2001; wild cattle during the Early Holocene provides an interesting Marshall and Weissbrod Domestication Processes and Morphological Change S401

Table 2. Radiocarbon dates for key African sites with evidence relating to management and domestication

Age (BP) Calendrical Taxon, site Uncalibrated Calibrated date (BC) Material Lab ID References Barbary sheep: ,Plants, GX 20753, Di Lernia 1998:117 ,7506–7726 ,9455–9675 ,110 ע Uan Afuda Late Acacus 8555 dung GX 18104 2001:419 6768–7060 8717–9009 100 ע 8000 Cattle: Charcoal, SMU-858, Wendorf and Schild ,8781–10,014 ,10,730–11,963 ,380 ע Nabta Playa E-79-8 9829 charcoal SMU-928 2001:98 8346–8796 10,295–10,745 150 ע 9350 Ostrich egg, SMU-257, Schild and Wendorf ,7191–7476 ,9140–9425 ,80 ע Nabta Playa E-75-6 8290 charcoal Gd 6257 2001:52 6465–6747 8414–8696 110 ע 7770 Sorghum, morphologi- cally wild: Shell, T-3697, Haaland 1987 ,4790–5018 ,6739–6967 ,90 ע Um Direiwa 6010 shell T-4045 4337–4549 6286–6498 110 ע 5600 Domestic cowpea: Cotyledon TO 11883 D’Andrea et al. 2007:689 1625–1865 3574–3814 60 ע B-sites 3410 Oil palm, morphologi- cally wild: Endocarp GX-29105 D’Andrea, Logan, and 1626–1737 3575–3686 40 ע B-sites 3380 Watson 2006:203 Domestic pearl millet: Seed OxA16919 Manning et al. 2011:317 2484–2570 4433–4519 33 ע Karkarichinkat 4011 -Grain Pa-1157 Summarized in Neu 1691–1949 3640–3898 100 ע Dhar Tichitt 3500 mann 2003:75 Grain TO-8172 D’Andrea, Klee, and 1526–2028 3475–3977 200 ע Birimi 3460 Casey 2001:343 Note. All calibrated BP ages and calendrical dates were calculated with IntCal 09, OxCal v4.1 (Bronk Ramsey 2009), with 68.2% confidence (1-j range). Note that only the earliest and latest dates from each site are reported here. parallel to that for management of Barbary sheep. Wendorf Close and Wendorf (1992) and Gautier (1984b, 1987b) also and colleagues (Gautier 1987b; Wendorf and Kro´lik 2001; argued, largely on the basis of a well and a watering basin at Wendorf and Schild 1998; Wendorf, Schild, and Close 1984) site E-75-6, that the repeated presence of water-dependent have argued that seasonally settled hunter-gatherers of the North African B. primigenius in Western Desert sites during Nabta Playa region (fig. 1) domesticated African cattle in the the tenth and ninth millennia cal BP (table 2) reflected range Western Desert of Egypt during the eleventh to tenth millen- extension facilitated by management and watering of cattle nium cal BP (reviews of arguments in Gifford-Gonzalez 2005; (table 1). Bos cranial remains in a human grave at El Barga table 2). Domestic sheep and goats, on the other hand, were in northern Sudan further support the presence of cattle in introduced to Africa from southwestern Asia during the early the region during the early ninth millennium cal BP (Ho- eighth millennium cal BP and postdate the appearance of negger 2005:247–248). The earliest evidence of small domestic cattle at all sites except Uan Muhaggiag (Gautier 2001; Lin- cattle from the central Sahara dates, however, to the eighth seele 2010; Linseele et al. 2010). The independent domesti- millennium BP (at Ti-n-Torha and Uan Muhaggiag; Gautier cation of African cattle has been tied to arid episodes, the 1987b; fig. 1). desire of hunter-gatherers for increased short-term predict- To date, the strongest evidence for domestication of cattle ability in food resources, and the difficulty of intensifying in Africa comes from a series of major studies of the genetic plant foods under these conditions (Marshall and Hildebrand characteristics and biodiversity of contemporary cattle breeds. 2002). Bos remains are ubiquitous in sites of the Nabta and Changing genetic approaches are reviewed by Larson (2011). Bir Kiseiba regions (fig. 1) from the eleventh to the tenth Initial analyses of maternal mitochondrial DNA (mtDNA) millennium cal BP (table 2) but in very small numbers, pre- showed that African cattle shared a distinctively higher fre- cluding detailed analyses of morphometric change or recon- quency of the T1 mitochondrial haplogroup than is common struction of culling profiles (Gautier 2001). Linseele (2004) in other regions and a large proportion of unique haplotypes has demonstrated, however, that size decrease is not a useful (Bradley et al. 1996). These findings are consistent with an indicator of domestication in northeastern Africa because the independent African domestication, although the possibility size of African Bos primigenius varied regionally and tem- of a demographic expansion of Near Eastern cattle in Africa porally and because ancient Egyptian longhorn cattle over- could not be ruled out (Bradley and Magee 2006; but see lapped in size with some wild cattle populations. Achilli et al. 2008). Recent analysis of single-nucleotide poly- S402 Current Anthropology Volume 52, Supplement 4, October 2011 morphisms from whole-genome sequences derived from show, however, that these animals were not yet morpholog- small numbers of cattle demonstrate that African breeds di- ically distinguishable from the African wild ass. verged early from the European taurine cattle (Decker et al. Recent studies of genetic variability in modern donkeys 2009). New analyses of high-resolution interspersed multi- suggest that prehistoric pastoralists may have domesticated locus microsatellites on the male-specific region of the Y chro- donkeys on the fringes of the Sahara. Beja-Pereira and col- mosome demonstrate the existence of an African subfamily leagues (2004; also Vila´, Leonard, and Beja-Pereira 2006) doc- in taurine cattle of the Y2 haplogroup (Pe´rez-Pardal et al. ument the existence of two different haplogroups or clades 2009). Associated analyses also indicate that neither the ge- of domestic donkeys. Their genetic-diversity data suggest two netic diversity in the African mtDNA T1 haplogroup nor the domestication events, both in northeastern Africa. Kimura et diversity in the Y2 haplogroup is consistent with the bottle- al.’s (2010) recent analysis of ancient DNA from the Nubian neck that would have been required to fix these haplotypes donkey (Equus africanus africanus) and the Somali wild ass from Near Eastern taurine cattle (Pe´rez-Pardal et al. 2010; see (Equus africanus somaliensis) demonstrates that the Nubian also Bovine HapMap Consortium 2009). Taken together with wild ass was the ancestor of modern donkey Clade I but that data on variation in autosomal microsatellites (rapidly evolv- the ancestor of donkeys of Clade II is currently unknown. ing regions of the nuclear genome) and other data on Y- This research also documents the ancient distribution of the chromosome variability in African cattle breeds (Bradley and Nubian wild ass and Clade I donkeys from the Atbara River Magee 2006; Hanotte et al. 2002), the genetic data as a whole and Red Sea Hills in Sudan and northern Eritrea across the point strongly to an independent African domestication of Sahara to Libya, a geographic distribution that suggests that cattle (Pe´rez-Pardal et al. 2009, 2010). prehistoric pastoralists domesticated Clade I donkeys (Kimura Ethnographic studies suggest, however, that genetic and et al. 2010). However, domestication by pastoralists or farmers phenotypic change may have been slow in early northeastern- of the northern Nile Valley during late prehistoric/early Pre- African cattle and that neither morphological nor genetic dynastic times is also a possibility. studies are likely to detect the early phases of this process. Given recurrent cycles of drought and disease, contemporary The Herding-Hunting Mosaic and the Spread of Pastoralism African pastoralists manage their herds for maximum growth In the central Sahara, cattle became common in the eighth by keeping a high proportion of females in herds (Dahl and to sixth millennium cal BP at sites such as Ti-n-Torha, Uan Hjort 1976). However, the main intentional selective processes Muhaggiag, Uan Telocat, Adrar Bous, Gobero, Enneri Bar- acting on African cattle are culling and castration, which affect dague´, and Wadi Howar (Clark et al. 2008; di Lernia 2006; males rather than females (Dahl and Hjort 1976; Ryan et al. Garcea 2004; Gautier 1987b; Jesse et al. 2007; Roset 1987; 2000). Natural selection in the form of drought and disease Sereno et al. 2008; fig. 1). The main advantages for hunter- often play a larger role in mortality than culling (Mutundu gatherers of herding cattle over intensification of plant re- 2005), multiple bulls are common in herds, offtake is low sources or reliance on hunting and gathering are thought to (4%–8%), and culling often takes place after sexual maturity have been decreased reliance on local rainfall and increased (Ryan et al. 2000). Such processes, together with some intro- predictability in daily access to cattle herds for blood, meat, gression with wild bulls, are likely to have worked against and ceremonial purposes (Jesse et al. 2007; Marshall and Hil- rapid morphological change in early pastoral herds and to debrand 2002). Foraging continued, but the intensity of the have resulted in a postmanagement lag in morphological new human-animal relationship would have required own- change. ership patterns and schedules oriented to animal care and transformation of hunter-gatherer societies. Dependence on Domestication of the donkey. It has long been suggested that wild calories could have been somewhat reduced, however, ancient Egyptians domesticated the donkey (Equus asinus), by milking, a practice that archaeologists have tended to as- although the Near East has also been considered a possible sume was adopted after herding for blood and meat and with area of origin. Egyptian Predynastic sites have yielded the some difficulty (but see Linseele 2010). earliest potential domestic donkeys, which date to the mid- Different genetic bases for lactase persistence in Europe and seventh millennium cal BP (historic date 4600–4400 BC; Africa show coevolution between people and cattle and the Boessneck and von den Driesch 1990; table 1). Some faunal strong selective advantage conferred by drinking milk (Tish- elements from these sites, zooarchaeologists argue, exhibit size koff et al. 2006). Interestingly, recent research has documented decrease relative to the wild ass (Boessneck and von den lactase persistence among some contemporary African Driesch 1990), but widespread morphological change was hunter-gatherers. Tishkoff et al. (2006, supplementary infor- slow to develop in ancient Egypt. Evidence of bone pathol- mation) note that lactase persistence could be selected for by ogies from early dynastic donkey burials at Abydos (fig. 1) delaying weaning of infants and, moreover, that the trait is demonstrates that by approximately 5000 cal BP (historic date also adaptive for digestion of certain roots and barks. This 3000 BC), First Dynasty Egyptian kings were using donkeys suggests several pathways to lactase persistence among hunter- to carry heavy loads (Rossel et al. 2008). Rossel et al. (2008) gatherers and raises the question of whether African herders Marshall and Weissbrod Domestication Processes and Morphological Change S403 milked their cattle earlier and incorporated dairy products bones and relatively complete ceramic pots (Jesse et al. 2007; into their diets with fewer digestive difficulties than previously fig. 1). As far south as Kenya by the middle of the fifth thought. However, milking scenes depicted in prehistoric Af- millennium cal BP, large stone circles such as those at Jarigole rican rock art and in Saharan ceramics have so far not pro- were constructed as centers for human burial rituals by south- duced dates or residues that bear on the antiquity of milking ward-migrating herders (Marshall, Grillo, and Arco 2011; Nel- in Africa (Jesse et al. 2007; Marshall 2000). son 1995). Hunter-gatherers, however, continued to flourish Oscillating periods of aridity and humidity resulted in pe- after the movement of herders into these regions (Lane et al. riods of increased mobility and occasional depopulation of 2007; Lesur, Vigne, and Gutherz 2007). the Sahara (di Lernia 2002; Garcea 2004; Kro¨pelin et al. 2008). In the eighth to seventh millennia cal BP, herders combined Domestication of African Plants livestock keeping with hunting and collection of wild grain in regions such as the Acacus Mountains (Gautier 1987b). At The earliest evidence for domestication of indigenous African Adrar Bous and other sites near lowland lakes, herders also plants with morphological change dates only to the beginnings fished and collected shellfish (Gifford-Gonzalez 2005; Smith of the fourth millennium cal BP (table 1). Although many 1992; fig. 1). Cattle-focused rock art attests to the symbolic Holocene hunter-gatherers of northeastern Africa relied importance of cattle for Saharan herders (Holl 2004; Smith heavily on wild Saharan cereals, high mobility and repeated 1992, 2005). Hunter-gatherers also flourished during this pe- abandonment of the region seem to have impeded long-term riod at sites such as Dakleh Oasis (McDonald 2008) and directional selection and morphological and genetic change. Amekni (Camps 1969; fig. 1), creating a mosaic of hunters Instead, selection processes culminated in morphological and herders across northeastern Africa (fig. 1). change once Saharan herders settled in the southern reaches Through the mid-Holocene, grasslands became more arid, of the Sahara and more humid Sahelian regions and estab- precipitation became increasingly unpredictable, and desert lished more permanent settlements in areas that were still regions of the Sahara expanded. Northeastern Africans re- within or close to the edge of the wild range of Saharan sponded to these pressures by heightening mobility, relying species. on introduced sheep and goats, and decreasing use of wild Sahelian herders—who also hunted, gathered, and fished— cereals (Barich 2002; di Lernia 2002; Garcea 2004; Gautier integrated cultivation of domestic pearl millet Pennisetum 1987a). It was during this period that the donkey was do- glaucum into their subsistence economies in one or two do- mesticated (Rossel et al. 2008). Their use would have made mestication events documented at or after 3898–3640 cal BP increased residential mobility and dispersal of settlements at sites west of Lake Chad, including Karkarichinkat Nord from water possible and would have facilitated long-distance (KN05), Dhar Tichitt, Birimi, and Gajiganna (D’Andrea, Klee, migrations (Marshall 2007). and Casey 2001; Fuller 2007; Kahlheber and Neumann 2007; Significant expansion of the geographic distribution of the Manning et al. 2011; fig. 1, table 1). Morphologically, this is dotted-wavy-line ceramic motif and distinctive human mor- evidenced by changes in seed shedding and shape, although tuary practices in the early seventh millennium cal BP reflect increases in seed size were delayed (D’Andrea, Klee, and Casey the southward movement of pastoralists, long-distance con- 2001). Fuller (2007) argues that the appearance of domestic tacts among Saharan groups, and elaboration of pastoralist pearl millet in India in the mid-fourth millennium cal BP ideologies (Jesse et al. 2007; Keding, Lenssen-Erz, and Pas- indicates a somewhat earlier African domestication and rapid toors 2007; Smith 1992; Wendorf and Kro´lik 2001). Just as dispersal. Recent research has also shown that the cow pea in the Mediterranean and western Europe, however, the tra- Vigna unguiculata was also an early-fourth-millennium mor- jectories of small immigrant groups may have varied greatly phological domesticate, dating to ca. 3898–3475 cal BP at the (O¨ zdog˘an 2011; Rowley-Conwy 2011). Domestic stock appear Kintampo B-sites in the grasslands of central Ghana to the south in the Sudanese Sahel by the early seventh mil- (D’Andrea et al. 2007; table 2). By contrast, African rice Oryza lennium cal BP at Esh Shaheinab and Kadero (Gautier 1984a, glaberrima was domesticated in the inland Niger delta of the 1984b) and by the mid-fifth millennium cal BP in Kenya Niger bend region by the early second millennium cal BP. On (Marshall and Hildebrand 2002). Similarly, Saharan lithics the eastern side of the continent, domestic teff Eagrostis tef and other traces of Saharan herders are first found in the West and finger millet Eleusine coracana were cultivated by Aksum- African Sahel by approximately 4500 cal BP (Jousse et al. ite populations in the Ethiopian highlands by the beginnings 2008; Linseele 2010; Smith 1992). Di Lernia (2006) argues of the second millennium cal BP (historic date AD 150–350; that the widespread ritual burial of cattle across the Sahara D’Andrea 2008). The oil-seed noog Guizotia abyssinica is also at the end of the seventh millennium BP represents a social present in Late Aksumite contexts (D’Andrea 2008). D’Andrea response to rapid aridification. Cattle burials and associated (2008) points out, however, that morphological change is dif- ritual activity are a prominent feature of site E-96-1 at Nabta ficult to identify in the small-seeded cereal teff, which was (Wendorf and Kro´lik 2001). At Djabarona 84/13, in the mid- selected for reliable production under arid conditions rather dle of Wadi Howar from the beginning of the sixth millen- than for increased seed size. In humid forested southwestern nium cal BP, more than a thousand pits are filled with cattle Ethiopia, Hildebrand (2003a, 2003b, 2007) has documented S404 Current Anthropology Volume 52, Supplement 4, October 2011 varied selection processes leading to domestication of yams their agricultural systems and fished and hunted a wide range Dioscorea cayenensis and ensete Ensete ventricosum. In these of wild-animal foods (Casey 2005; Neumann 2005; Plug and and other areas of Africa, domestic plants are thought to have Voigt 1985; van Neer 2000). been advantageous to pastoral hunter-fishers for risk mini- This brings to the fore the question raised at the outset of mization and greater predictability (D’Andrea et al. 2007; whether such diverse subsistence strategies fit current con- Kahlheber and Neumann 2007; Marshall and Hildebrand ceptions of agricultural systems. Kahlheber and Neumann 2002). (2007:339) are doubtful whether “farming” is an appropriate Although morphological change occurred in a range of term for some of these ways of life. Smith’s (2001, 2011) term domesticated African plant taxa, it has been suggested that a “low-level food production” has been used in the region, but number of African savanna plants were cultivated or inten- it does not fully capture the complexities of African settings. sively managed over the long term in ways that did not lead The question of whether the Kintampo should be considered to morphological domestication (reviews in Marshall and Hil- “foragers,” “farmers,” or something else has also been re- debrand 2002; Neumann 2005). Haaland (1999) and Abdel- viewed by Casey (2005) and by D’Andrea and colleagues Magid (1989) argued, largely on the basis of the ∼30,000 (D’Andrea, Logan, and Watson 2006:216–218; D’Andrea et grindstones that were unearthed at the site of Um Direiwa, al. 2007), who argue that although there are clear-cut cases for cultivation of sorghum Sorghum bicolor in Sudanese sites of foragers or farmers in Africa, there are many others that dating to the seventh millennium cal BP (table 1). Mecha- defy simple categorization. Hildebrand’s (2003a) ethno- nisms that they suggested for late morphological change in- graphic research among the Sheko of southwestern Ethiopia clude continued outcrossing between cultivated and wild pop- and the literature on use of weedy greens in Africa (Etkin ulations and harvesting through beating into baskets or 1994; Fleuret 1979; Marshall 2001 and references therein) uprooting. This has led to arguments that sorghum was not provide ample evidence that such subsistence strategies have morphologically domesticated until it was removed from its long-term trajectories in many parts of Africa and cannot be wild African range (Haaland 1999; but see Fuller 2003). Al- dismissed as transitory. though mechanisms exist that may have caused late mor- phological change in African cereals and harvesting of wild grains was at times intensive, there is no macrobotanical evi- Ethnoarchaeological Insight into dence or indication of landscape modification that supports Management, Selection Processes, and claims for cultivation of African grains before the early fourth Domestication of the Donkey millennium cal BP. In the wetter tropical regions, there is evidence of long- One approach to better addressing conceptual problems pre- term use of a number of forest taxa without morphological sented by questions of late morphological change and the change. Long-term use of oil palm Elaeis guineensis and in- diversity of economic systems in Africa is to consider path- cense trees Canarium schweinfurthii has been documented ways to domestication for particular species in light of the across the humid tropics of Africa (D’Andrea, Logan, and potential for morphological change, or lack thereof, in specific Watson 2006; Mercader et al. 2006). This pattern is not con- social and environmental contexts. The question that we ad- fined to forests, however. D’Andrea, Logan, and Watson dress here is how the behavior of the African wild ass and (2006:216–217) argue that Kintampo people living in the management of donkeys by herders and small-scale farmers grasslands of central Ghana employed a system of arboricul- in Africa contribute to selection processes and the likelihood ture that did not rely on management strategies that would of development of archaeological signatures of domestication result in morphological change. Kahlheber and Neumann in the donkey. This analysis focuses on aspects of the biology (2007) also note that a number of west African park savanna and behavior of the donkey and its use as a transport animal species, such as baobab Adsonia digitata and the shea-butter that influence management practices in extensive pastoral and tree Vitellaria paradoxa, were protected and encouraged but agricultural systems and are relevant (sensu Wylie 2002) to never domesticated. Other wild plants that are still protected ancient settings for domestication. It is often argued, for in- and sometimes actively sown in many different African en- stance, that sociability and the presence of a dominance hi- vironments include weedy green species ranging in status from erarchy are desirable characteristics for potential domesti- crops to semidomesticated or wild (Kahlheber and Neumann catability (Clutton-Brock 1992; Diamond 1997). African wild 2007; Marshall 2001). Kahlheber and Neumann (2007:333) ass do not, however, fit this profile. The extant Somali wild point out that in the West Africa Sahel, reliance on morpho- ass, or dibokali, is solitary or forms groups with weak short- logically wild park savanna species became more evident when term associations. It also lacks a pronounced dominance hi- economies diversified and populations concentrated close to erarchy (Klingel 1974; Moehlman 2002). This social system water 2,000 years ago. In many regions of Africa, Iron Age profoundly influences donkey behavior under human man- agriculturalists relied on a particularly broad range of re- agement. sources, and farmers incorporated diverse domestic crops and Recent ethnoarchaeologial research on donkey use and managed plants, cattle, sheep, goats, dogs, and donkeys into management among Maasai households in Kajiado District Marshall and Weissbrod Domestication Processes and Morphological Change S405 of southern Kenya provides the first detailed information on Table 3. Donkey holdings of eight Maasai households orga- selection processes in a pastoral social and economic context. nized according to settlements and individual caretakers During 2006, Lior Weissbrod lived in Maasai communities in the study area and collected interview and participant obser- Male Female vation data from 26 women from eight households spread Settlement, women Juvenile Adult Old Juvenile Adult Old Total among six different pastoral settlements (table 3). The study S1: focused on use and daily management, herd composition, A2114 mortality, and breeding behavior. After a 2-year period of B112 severe drought (2004–2006), the donkey holdings of house- C21115 S2: holds participating in the study were reduced but still totaled A1124 65. B1113 Donkeys were not regarded as food. They were considered C2 2 women’s animals, important for transport but without the D1113 symbolic status of cattle. Women were the caretakers of don- E11 S3: keys and used them to carry household goods during resi- A134 dential moves, to collect water, and to take intermittent trips B1 23 to trading centers. Donkeys also carried meat, firewood, and C213 water for large ceremonies. During the dry season, women D112 went long distances for water every other day, returning with E112 S4: a typical load of 50 L per donkey. Children herded household A1 1 donkeys with the calves, but during the wet season, donkeys B11 were free ranging. Many families penned donkeys within the C11 settlement thorn fence or in calf enclosures at night for pro- D11 tection against predators. E1 1 S5: Our data show that the use of donkeys in Kajiado enhanced A11 the flexibility and stability of local herding systems (see also B1113 Marshall 2007; Marshall and Weissbrod 2009). Families in the C1135 study area who did not own donkeys could not move as a D11 whole away from permanent sources of water and were unable E1 12 F11 to make optimum use of available grazing. Donkeys were, G347 nevertheless, managed less than other livestock. Marshall S6: (2007) previously noted that the ability of donkeys to dig for A112 water and to protect themselves from predators more suc- cessfully than other livestock was associated with low levels of management, which might result in low levels of selection. same way that they do cattle or value color distinctions ideo- Our data show that behavior was a factor but that the level logically. of use of donkeys in the study area ultimately determined the The dynamics of wild ass mating systems, based on short- degree to which donkeys were herded and penned. lived associations that occur when females move through male In addition to management practices, we also collected in- territories, influence donkey breeding in the domesticated en- formation on reproduction and desired characteristics of don- vironment. Maasai women stressed their concern with the keys that might be selected for through strategic breeding. aggressive behavior of jacks during mating. Even when they Women that we talked to particularly valued strength and wanted to keep a female from breeding with an especially calmness in a donkey. Some also mentioned the importance aggressive jack, women said that they found it impossible to of disease and drought resistance, although they noted that keep the male away. They also noted that estrus jennies might donkeys were less vulnerable to these hazards than other live- go astray without warning in search of males. They are often stock. We found, however, that participants in the study made lost this way, and we documented a number of cases in which no attempt at all to influence mate choice among donkeys or wandering females, as well as males, were cared over a long to breed for particular characteristics. The ancestry of a par- term by women in distant settlements. Lack of selection be- ticular donkey was unknown except for the female parent. By cause of the difficulty of controlling donkey breeding is, there- contrast, research on cattle genealogies shows that Maasai fore, likely wherever a premium is placed on “wild” charac- herders memorize these in great detail for several generations teristics of the donkey, such as strength, rather than on docility (Ryan et al. 2000). The lack of strategic breeding of donkeys and productivity for food. The relatively high proportion of is influenced by donkey behavior and herd compositions but males in herds (one male : two females) is another factor that is also related, at least in part, to the fact that Maasai herders makes control over breeding logistically difficult. Because do not use donkeys as symbols of social transactions in the donkey owners kept small herds specifically for transport, they S406 Current Anthropology Volume 52, Supplement 4, October 2011 weighed the breeding advantages of females against the su- intentional control over breeding, however. Mohammed men- perior transport potential of males. The strength of males was tions that copulation might occur anywhere and was actively greatly favored, and so was their consistent availability for discouraged in the market center (Mohammed 1991). transport use. Overview of management and selection. In order to con- Herd growth and mortality patterns also contributed to sider patterns of directional selection, it is useful to examine patterns of selection in domestic donkey herds. Pastoral factors that affect the likelihood of genetic drift, intentional Maasai donkeys had, on average, a foal every 2 years. Mortality selection, and reproductive isolation in donkeys managed by resulted from predation by hyenas, disease, and drought. pastoralists and small-scale farmers. Culling of male donkeys Herds grew relatively slowly, and additional animals were re- by Ethiopian villagers and castration of male donkeys by cruited to herds through gifts, loans, and purchase. Socially Maasai pastoralists were important factors affecting selection. based loans or exchanges of cattle are deeply woven into the These practices ensured that males with desired traits, such fabric of Maasai society (Ryan et al. 2000). To a lesser extent, as strength or size, remained in the breeding pool. Females, this system is also used for donkeys, and social exchange is on the other hand, were never culled, and management of a mechanism of selection and gene flow. Animals entering or donkeys was minimal. None of the donkey owners that we leaving a herd through loans were carefully selected and pre- studied tried to ensure a diverse set of breeding males, to dominantly female. In some cases, however, exchanges were breed select females or males, or to keep records of parentage. involuntary, resulting from donkeys running away. We argue that these management practices are influenced by In the wet season, herds of donkeys made up of animals wild ass and donkey courtship and breeding behavior and from different settlements in the same neighborhoods range have significant consequences for long-term directional se- freely. This practice and the system of intentional and un- lection and domesticatory processes. The data also indicate intentional loans maintain gene flow among settlements. Pur- that different sets of functional and symbolic considerations chases were rarer than loans also but recruited animals to affect Maasai practices of cattle and donkey management and slow-growing herds and maintained intentional selection on are associated with differing levels of selective pressure and an interregional scale. Men purchased animals when visiting control of gene flow. In our study area, people also bred or markets, and the strength and price of the donkey were major obtained cattle for ideal coat colors and conformation, and considerations affecting purchases. Young male donkeys were it is possible that without this additional symbolic motivation, cheaper than others, and purchases were one male : two fe- functional reasons for breeding donkeys were not enough to males. There was no intentional culling of donkeys, and don- overcome significant practical difficulties. As research on keys were not eaten, but small, slow-growing, or aggressive mammals such as the fur fox (Belyaev 1979; Trut 1999) and males were removed from the breeding pool through castra- the guinea pig (Ku¨nzl et al. 2003) has shown, without selective tion. We recorded six castrated males (40% of the males stud- breeding, retention of individual animals with desired traits ied), and castration of male donkeys was a more important and culling of others, directional selection may be very slow factor affecting the direction of intentional selection than cull- or fail to occur even in the absence of gene flow from wild ing or selective breeding. populations. Very few studies of donkey management and selection have From a wider perspective, there are related issues that work been conducted in settled agricultural villages. Mohammed’s against genetic drift as a major factor driving genetic and (1991) and Wilson’s (1991) Ethiopian research can, however, morphological change in donkeys. In both the Maasai and be used for comparison with the Maasai pastoral study. They Ethiopian Arsi cases, donkeys from numerous households focused on Ethiopian farmers of the central and southern grazed unsupervised in mixed herds, allowing uncontrolled highlands who used donkeys to transport grain to market and genetic exchange among neighborhood populations. Donkeys for hauling household firewood and water. Most families in were loaned among broad social networks in both regions, the study areas kept one to two donkeys, usually female (Mo- and the frequency with which donkeys were taken to market hammed 1991; Wilson 1991). Donkeys were also loaned to in Ethiopia also provided a wider setting for interbreeding family and friends. In the Awassa region, males were rare (100 among donkeys from different areas. We argue, however, that females : 1 male); in other regions the number of males was in both the pastoral Maasai and Arsi farmer cases, low levels higher (73 females : 27 males). Where males were more com- of formal management and lack of intentional selective breed- mon, they were usually less than 4 years old. Mohammed ing are linked to donkey biology and behavior, the use of (1991) notes that male donkeys were not castrated. We infer donkeys for transport, and the fact that donkeys are not often that low proportions of males in herds indicated male culling, eaten. Male culling plays a significantly greater role in animals although donkey eating was not discussed. People in Awassa that are primarily managed for meat—including cattle, sheep, did not supervise donkeys when they were not using them, and goat—than it does in donkeys. Although culling and and Mohammed (1991) documents minimal donkey man- castration affect donkey selection, they are outweighed by lack agement and poor animal nutrition in this area. Because of of directional selection in breeding and consistent gene flow the danger presented by hyenas, however, people often among donkeys over significant distances. brought donkeys inside their houses at night. There was no The data for Maasai pastoralists and Ethiopian Arsi farmers Marshall and Weissbrod Domestication Processes and Morphological Change S407 also suggest that the potential for gene flow from the wild is Do Holocene Pastoralists in Africa Fit likely in both settings but marginally less so in agricultural Conceptions of Early Agricultural villages. The Maasai villages studied lie outside the historic Systems in Other Regions? range of the wild ass. But it is easy to see that had they not, the runaway tendencies of estrus females would have made After examining evidence for the beginnings and spread of the prevention of introgression difficult. Like contemporary food production in Africa and analysis of the way that man- herders valuing strength and endurance in their donkeys, his- agement and behavioral factors affect the likelihood of mor- toric Beja pastoralists of Sudan and Eritrea intentionally en- phological change in one large mammal—the donkey—we couraged interbreeding among donkeys from domesticated return to consideration of whether African pastoralism fits and wild settings (Baker 1867; Murray 1935). During the current conceptions of early agricultural systems developed 1950s, Nicolaisen (1963) also recorded capture and taming for other regions. We start by considering the question of of wild or feral animals by Tuareg pastoralists of the central whether recognition of early food production in tropical Sahara. regions of Africa has been hampered by concepts of domes- It is possible, therefore, to begin to identify separate con- tication that rely on morphological change by focusing on texts for the domestication process of donkeys in Africa. We donkeys, cattle, Barbary sheep, African cereals, and West Af- predict that ancient Saharan pastoralists reduced the number rican tropical tree crops. of breeding males in herds through culling and castration in Some evidence suggests that complex hunter-gathers may order to cope with practical difficulties resulting from court- have attempted to manage cattle in the northeastern Sahara ship and breeding behavior in donkeys. Isolation from wild and, for a time, Barbary sheep in the Libyan Acacus. There ancestors would have been possible in some pastoral settings is no doubt that short-term participation in domesticatory as a result of mid-to-Late Holocene climate change, range relations are difficult to recognize archaeologically, but nev- fragmentation, and pastoral settlement in island or marginal ertheless evidence for management of Barbary sheep is sug- ecosystems. Wild asses may also have been removed from gestive rather than conclusive. In contrast, genetic data offer their wild range by pastoral dispersals into the high-altitude a measure of support for the hypothesis of cattle domesti- cation in Africa. The sociality of wild Bovini, however, and Ethiopian highlands and other regions, such as southern Su- the expectation that wild cattle were used mainly for food dan and northern Kenya, outside the historic range of the suggests strong selection and a pathway to domestication— wild ass. characterized by a postmanagement lag rather than late mor- Selection for morphological change would have been slow phological change and fewer problems with identification of until donkeys were removed from close proximity to the wild early domesticates—different from that discussed for the don- ass and interbreeding between local donkey populations was key. restricted. It would appear that reproductive isolation of cap- Ethnoarchaeological data on the donkey reveal relations tive wild asses from free-living populations is somewhat more among selection processes and slow genetic and morpholog- likely to have occurred in ancient urban settings such as the ical change and illuminate conditions under which biology Predynastic and Dynastic Egyptian towns of the Nile Valley, and human management influenced domestication and the with permanent walls and high densities of protected agri- likelihood of late morphological change. The biological and cultural land. Gene flow would still have been possible, how- behavioral reality of donkeys in current domesticatory settings ever, given the narrowness of the Nile agricultural belt and in Africa is that females actively seek out mates, territorial the mobility of pack donkeys. An appreciation for the ad- males are reproductively aggressive, and high proportions of vantages of strong animals may also have made interbreeding males are advantageous for transport use. These factors in- between captive and wild asses desirable for both villagers and teract to make reproduction difficult to control and gene flow pastoralists. likely among donkeys of different households and villages, The lack of morphological change evident in the Abydos along trade routes, and between tame animals and wild asses. donkeys as late as 5000 cal BP (3000 BC; Rossel et al. 2008) Archaeological and genetic data suggest that pastoral so- demonstrates that size decrease was not generally established cieties of the Sahara or the Horn of Africa played an important until well after this period. It is also conceivable that mor- role in the early development of stable and long-term systems phological change did not occur until donkeys were taken of management of morphologically wild donkeys. Morpho- across the Red Sea to Yemen or other regions of Asia. Which- logical change was late, and mechanisms for this probably ever the case, donkeys are a classic example of a species that included creation of built environments of the Nile Valley, was used to carry loads for millennia as a domesticate but late agriculturally modified landscapes, the high mobility of with late morphological change. We conclude that slow mor- Saharan pastoralists, and ecological fragmentation created by phological change in domesticated donkeys can be explained climatic changes of the mid-Holocene. by low levels of selection, high potential for interbreeding Although an appreciation of the likelihood of delayed mor- between founder populations, and potential for introgression phological change and biases against identification of do- with the wild. mestic donkeys is novel, Africanists have long discussed the S408 Current Anthropology Volume 52, Supplement 4, October 2011 question of whether the lack of morphological change resulted of morphological change and the potential for interpretive in bias against recognition of cultivation of early cereal crops. bias. These data are strongest with regard to African tree crops. There is mounting evidence for long periods of intensive use Despite this, however, there is little evidence that archaeol- of wild cereal grasses by Early Holocene hunter-gatherers and ogists have ignored early agriculture in the humid tropics of early herders of the Sahara without evidence of domestic Africa. There is, in fact, no archaeological evidence that the traits. This has been related to a lack of continuous directional humid tropical forests were heavily populated by African selection as a result of increased aridity and pastoral mobility. hunter-gatherers during the Early Holocene, and there are Morphological changes in well-known African cereals such as few traces of intensification in these regions until after they pearl millet and pulses such as cowpeas occur relatively late were settled by food producers (see D’Andrea, Logan, and and in conjunction with pastoral sedentization in better- Watson 2006; D’Andrea et al. 2007; Mercader et al. 2006 and watered locales within the semiarid Sahel and in the more references therein). Nevertheless, as Africanist paleoethno- humid West African woodlands after the fourth millennium botanists have pointed out, much work remains to be done cal BP. Recent research in more humid regions of West Africa on the nature of agricultural systems dating to the past several has revealed, however, a number of tended and managed tree thousand years in the humid tropics (D’Andrea, Logan, and crops, such as incense, baobab, and the shea-butter tree, that Watson 2006; Hildebrand 2007; Kahlheber and Neumann were heavily used during the Holocene but remain morpho- 2007). logically wild to this day. This is typical of tropical tree crops We conclude that there is no indication of significant worldwide and common in weedy greens. regional-scale biases that would have affected current inter- It is worth reiterating at this point that identification of pretations of the sequence of plant and animal domestication management of plants and animals before genetic or mor- in Africa or geographic patterns of the timing and spread of phological change is inherently problematic, and the longer food production. The larger patterns, as we see them, are that the period before morphological change occurs in a particular some complex hunter-gatherers of the Early Holocene in plant, animal, or setting, the greater the difficulties that arise. North Africa successfully managed cattle, developed pastoral It is clear that there are at least three axes of variability in social and subsistence systems, and spread over vast areas of morphological responses of plants and animals to selection the Sahara. Other such groups in North Africa may have during coevolutionary relations with humans. We have found experimented with management of Barbary sheep, but this it useful here to conceive of this temporal and spatial variation was short-lived. Later, during the mid-Holocene, there is evi- in terms of a “postmanagement lag” before morphological dence that donkeys were domesticated by African pastoralists change, as opposed to “late morphological change” or “re- in the Sahara and the Horn of Africa and possibly by Pre- gionally clustered variability.” dynastic Egyptians in towns along the Nile. These animals Our review suggests that all these forms of variability exist remained morphologically wild for long periods. The earliest in Africa. The available data appear to accord with Jones and plant domesticates in Africa are associated with decreased Brown’s (2007) suggestion that a long, stable period of man- mobility as pastoralists moved into better-watered locales agement without morphological change or a normal “mor- within the semiarid Sahel and into West Africa. It can also phological lag” is common to many domesticates worldwide. be shown, however, that in some humid tropical regions of In Africa, however, it is not clear that their corollary—that Africa, clusters of species existed with a long history of cul- population expansion leads to removal of plants and animals tivation or tending by established agricultural communities from their wild range and morphological change—holds true. and with biological traits amenable to management but no Instead, heightened mobility related to climatic changes and traces of morphological domestication. increased aridity ultimately led to the movement of some African patterns of food production were distinctive. An- species out of their wild ranges. Furthermore, early African imals were domesticated before plants, herding populations cereals appear to have been domesticated within their wild became more mobile than their forager ancestors, the sub- ranges and intensified on the edge of these regions. Increas- sistence system was characterized by a few morphologically ingly settled pastoral communities and management practices wild domesticates (e.g., the donkey), a wide range of wild that maintained directional selection seem to have been more resources in ecodiverse combinations continued in use, and important factors affecting domestication of these crops than mosaics of hunter-gatherers and herders occupied varied reproductive isolation. regions. Pastoralism developed early in the arid topics, We focused above on the possibility of biases against the whereas the beginning of farming based on domesticated recognition of early agriculture in tropical regions. We do not, plants was late. however, see a cluster of taxa subject to late morphological These African data are informed by and provide perspec- change in the arid or high-altitude subtropics of Africa; here, tives on pathways to food production in other regions. In species-specific analyses of the likelihood of late morpholog- discussions at the Wenner-Gren conference in Temozo´n in ical change are crucial. We agree with Denham (2007), how- 2009, Meadow (2009) and Fuller (2009; also see Fuller 2006) ever, that the biology of many species of the African humid argued that South Indian patterns of early pastoralism and tropics increases the likelihood of a lack or significant delay subsequent domestication of local millets and pulses are rem- Marshall and Weissbrod Domestication Processes and Morphological Change S409 iniscent of Africa. Similarly, pastoralism has long been con- Aldenderfer, Mark S. 2003. Moving up in the world: archaeologists sidered an early phenomenon in the Andes (Aldenderfer 2003; seek to understand how and when people came to occupy the Andean and Tibetan plateaus. American Scientist 91(6):542–550. Browman 1974; Mengoni-Gon˜alons and Yacobaccio 2006) Alhaique, Francesca, and Fiona Marshall. 2009. Preliminary report and the Zagros (Abdi 2003; Hole 1996). Mobile pastoralism on the Jebel Gharbi fauna from site SJ-00-56 (2000 and 2002 is also a major theme in data emerging on the beginnings of excavations). Africa 64(3–4):498–507. food production in central Asia (Frachetti and Benecke 2009; Baker, Samuel W. 1867. The Nile tributaries of Abyssinia. London: Outram et al. 2009). In addition, Belfer-Cohen and Goring- Macmillan. Barich, Barbara E., ed. 1987. Archaeology and environment in the Morris (2011) and Goring-Morris and Belfer-Cohen (2011) Libyan Sahara, 1978–1983. Cambridge Monographs in African Ar- document African-like mosaics of hunter-gatherer and early- chaeology 23, BAR International Series 368. Oxford: British Ar- food-producer settlement in the Levant during the Early Ho- chaeological Reports. locene. Evidence is also mounting that shows continued re- ———. 2002. Cultural responses to climatic changes in North Africa: liance on wild resources and ecodiverse strategies pursued by beginning and spread of pastoralism in the Sahara. In Droughts, food, and culture: ecological change and food security in Africa’s later small-scale food producers or low-level farmers of the Amer- prehistory. Fekri A. Hassan, ed. Pp. 209–223. New York: Kluwer icas and subtropical and tropical regions (Denham 2011; Fritz Academic/Plenum. 2007; Piperno and Pearsall 1998; Smith 2001, 2011) and per- Barich, Barbara E., and Elena A. A. Garcea. 2008. Ecological patterns haps even temperate regions of Asia (Crawford 2011; Lee in the Pleistocene and Holocene in the Jebel Gharbi, northern 2011; Zhao 2011). Understanding ways in which specific Libya: chronology, climate and human occupation. African Ar- chaeological Review 25(1–2):87–97. strands such as these contribute to larger similarities and dif- Bayliss-Smith, Tim. 2007. The meaning of ditches: interpreting the ferences in the warp and weft of data on the beginnings of archaeological record from New Guinea using insights from eth- agriculture requires attention to methods of detection of early nography. In Rethinking agriculture: archaeological and ethnoar- phases of domestication, information on specific social con- chaeological perspectives. Tim Denham, Jose´Iriarte, and Luc Vry- texts, and regionally focused and temporarily expansive re- daghs, eds. Pp. 126–148. Walnut Creek, CA: Left Coast. Beja Pereira, Albano, Philip R. England, Nuno Ferrand, Steve Jordan, search. These kinds of data are only just beginning to emerge Amel O. Bakhiet, Mohammed A. Abdalla, Marjan Mashkour, Jordi from Africa, which, as this summary demonstrates, has much Jordana, Pierre Taberlet, and Gordon Luikart. 2004. African origins to contribute to unraveling patterns of variability in global of the domestic donkey. Science 304(5678):1781. pathways to food production. Belfer-Cohen, Anna, and A. Nigel Goring-Morris. 2011. Becoming farmers: the inside story. Current Anthropology 52(suppl. 4):S209– S220. Belyaev, Dmitri K. 1979. Destabilizing selection as a factor in do- Acknowledgments mestication. Journal of Heredity 70(5):301–308. Boessneck, Joachim, and Angela von den Driesch. 1990. Tierreste We are grateful to Ofer Bar-Yosef, Douglas Price, the Wenner- aus der vorgeschichtlichen Siedlung von El-Omari bei Heluan/ Gren Foundation, and our companions in Temozo´n for the Untera¨gypten. In El Omari. Fernand Debono and Bodi Mortensen, eds. Pp. 99–107. Mainz, Germany: von Zabern. genesis of this paper and for challenging and enriching our Bovine HapMap Consortium. 2009. Genome-wide survey of SNP views about domestication and the spread of food production. variation uncovers the genetic structure of cattle breeds. Science The donkey ethnoarchaeological research could not have been 324(5926):528–532. conducted without the expert knowledge and gracious sup- Bradley, Daniel G., David E. MacHugh, Patrick Cunningham, and port of the people of Kajiado. We are also grateful to the Ronan T. Loftus. 1996. Mitcochondrial diversity and the origins of African and European cattle. Proceedings of the National Acad- Kenya National Museums and Dr. Purity Kiura for their sup- emy of Sciences of the USA 93(10):5131–5135. port and to the government of Kenya for permission to un- Bradley, Daniel G., and David Magee. 2006. Genetics and origins of dertake research. This paper has benefited greatly from the domestic cattle. In Documenting domestication: new genetic and thoughtful comments of reviewers. Research was supported archaeological paradigms. Melinda A. Zeder, Daniel G. Bradley, Eve by National Science Foundation grants BCS-0447369 and Emshwiller, and Bruce D. Smith, eds. Pp. 317–328. Berkeley: Uni- versity of California Press. BCS-0536507. Bronk Ramsey, Christopher. 2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51(1):337–360. References Cited Brooks, Alison S., Diane E. Gelburd, and John E. Yellen. 1984. Food production and culture change among the !Kung San: implications Abdel-Magid, Anwar. 1989. Plant domestication in the middle Nile for prehistoric research. In From hunters to farmers. J. Desmond Basin: an archaeobotanical case study. Cambridge Monographs in Clark and Steven A. Brandt, eds. Pp. 293–310. Berkeley: University African Archaeology 35, BAR International Series 532. Oxford: of California Press. British Archaeological Reports. Browman, David L. 1974. Pastoral nomadism in the Andes. Current Abdi, Kamyar. 2003. The early development of pastoralism in the Anthropology 15(2):188–196. central Zagros Mountains. Journal of World Prehistory 17(4):395– Camps, Gabriel. 1969. Amekni: ne´olithique ancien du Hoggar. Paris: 448. Arts et me´tiers graphiques. Achilli, Alessandro, Anna Olivieri, Marco Pellecchia, Cristina Uboldi, Caneva, Isabella, ed. 1988. El Geili: the history of a middle Nile en- Licia Colli, Nadia Al-Zahery, Matteo Accetturo, et al. 2008. Mi- vironment 7000 B.C.–A.D. 1,500. Cambridge Monographs in Af- tochondrial genomes of extinct aurochs survive in domestic cattle. rican Archaeology 29, BAR International Series 424. Oxford: Brit- Current Biology 18(2):R157–R158. ish Archaeological Reports. S410 Current Anthropology Volume 52, Supplement 4, October 2011

Casey, Joanna. 2005. Holocene occupations of the forest and savanna. ———. 1999. Early Holocene pre-pastoral cultures in the Uan Afuda In African archaeology: a critical introduction. Ann B. Stahl, ed. Pp. cave, wadi Kessan, Tadrart Acacus (Libyan Sahara). In Wadi Te- 225–248. Oxford: Blackwell. shuinat: palaeoenvironment and prehistory in south-western Fezzan Castelletti, Lanfredo, Elisabetta Castiglioni, Michela Cottini, and (Libyan Sahara). Mauro Cremaschi and Savino di Lernia, eds. Pp. Mauro Rottoli. 1999. Archaeobotanical analysis of charcoal, wood 123–154. Florence, Italy: all’Insegna del Giglio. and seeds. In The Uan Afuda Cave: hunter-gatherer societies of ———. 2001. Dismantling dung: delayed use of food resources Central Sahara. Savino di Lernia, ed. Pp. 131–148. Florence, Italy: among Early Holocene foragers of the Libyan Sahara. Journal of all’Insegna del Giglio. Anthropological Archaeology 20(4):408–441. Clark, J. Desmond, Diana C. Crader, Elena A. A. Garcea, Diane ———. 2002. Dry climatic events and cultural trajectories: adjusting Gifford-Gonzalez, Andrew B. Smith, and Martin A. J. Williamson, middle Holocene pastoral economy of the Libyan Sahara. In eds. 2008. Adrar Bous: the archaeology of a granitic ring complex Droughts, food, and culture: ecological change and food security in in central Sahara, Niger. Annales in Archaeology. Tervuren, Bel- Africa’s later prehistory. Fekri A. Hassan, ed. Pp. 225–250. New gium: Royal Africa Museum. York: Kluwer Academic. Close, Angela E., and Fred Wendorf. 1992. The beginnings of food ———. 2006. Building monuments, creating identity: cattle cult as production in the eastern Sahara. In Transitions to agriculture in a social response to rapid environmental changes in the Holocene prehistory. Anne B. Gebauer and T. Douglas Price, eds. Pp. 63– Sahara. Quaternary International 151(1):50–62. 72. Madison, WI: Prehistory Press. Dyson-Hudson, Rada, and Neville Dyson-Hudson. 1980. Nomadic Clutton-Brock, Juliet. 1992. The process of domestication. Mammal pastoralism. Annual Review of Anthropology 9:15–61. Review 22(2):79–85. Etkin, Nina L. 1994. The cull of the wild. In Eating on the wild side. Crawford, Gary W. 2011. Advances in understanding early agriculture Nina L. Etkin, ed. Pp. 1–21. Tucson: University of Arizona Press. in Japan. Current Anthropology 52(suppl. 4):S331–S345. Evans-Pritchard, E. E. 1940. The Nuer. Oxford: Oxford University Cremaschi, Mauro, and Luca Trombino. 2001. The formation process Press. of the stratigraphic sequence of the site and their palaeoenviron- Evershed, Richard P., Sebastian Payne, Andrew G. Sherratt, Mark S. mental implications. In Uan Tabu in the settlement history of the Copley, Jennifer Coolidge, Duska Urem-Kotsu, Kostas Kotsakis, Libyan Sahara. Elena A. A. Garcea, ed. Pp. 15–23. Florence, Italy: et. al. 2008. Earliest date for milk use in the Near East and south- all’Insegna del Giglio. eastern Europe linked to cattle herding. Nature 455(7212):528– Dahl, Gudrun, and Anders Hjort. 1976. Having herds: pastoral herd 531, doi:10.1038/nature07180. growth and household economy. Stockholm: Department of Social Fleuret, Anne. 1979. The role of wild foliage plants in the diet: a case Anthropology, University of Stockholm. study from Lushoto, Tanzania. Ecology of Food and Nutrition 8(2): Dale, Darla, Fiona Marshall, and Tom Pilgram. 2004. Delayed-return 87–93. hunter-gatherers in Africa? historic perspectives from the Okiek Frachetti, Michael, and Norbert Benecke. 2009. From sheep to (some) and archaeological perspectives from the Kansyore. In Hunter- horses: 4500 years of herd structure at the pastoralist settlement gatherers in theory and archaeology. George M. Crothers, ed. Pp. of Begash (southeastern Kazakhstan). Antiquity 83(322):1023– 340–375. Carbondale: Center for Archaeological Investigations, 1037. Southern Illinois University. Fritz, Gayle J. 2007. Keepers of Louisiana’s levees: early mound build- D’Andrea, A. Catherine. 2008. T’ef (Eagrostis tef ) in ancient agri- ers and forest managers. In Rethinking agriculture: archaeological cultural systems of highland Ethiopia. Economic Botany 62(4):547– and ethnoarchaeological perspectives. Tim Denham, Jose´Iriarte, and 566. Luc Vrydaghs, eds. Pp. 189–209. Walnut Creek, CA: Left Coast. D’Andrea, A. Catherine, Stefanie Kahlheber, Amanda L. Logan, and Fuller, Dorian Q. 2003. African crops in prehistoric South Asia: a Derek J. Watson. 2007. Early domesticated cowpea (Vigna ungui- critical review. In Fuel, food and fields: progress in African archaeo- culata) from central Ghana. Antiquity 81(313):686–698. botany. Katharina Neumann, Ann Butler, and Stefanie Kahlheber, D’Andrea, A. Catherine, Marlies Klee, and Joanna Casey. 2001. Ar- eds. Africa Praehistorica 15. Pp. 239–271. Cologne: Heinrich Barth chaeobotanical evidence for pearl millet (Pennisetum glaucum)in Institut. sub-Saharan Africa. Antiquity 75(288):341–348. ———. 2006. Agricultural origins and frontiers in South Asia: a D’Andrea, A. Catherine, Amanda L. Logan, and Derek J. Watson. working synthesis. Journal of World Prehistory 20(1):1–86. 2006. Oil palm and prehistoric subsistence in tropical West Africa. ———. 2007. Contrasting patterns in crop domestication and do- Journal of African Archaeology 42(1):195–222. mestication rates: recent archaeobotanical insights from the Old Decker, Jared E., J. Chris Pires, Gavin C. Conant, Stephanie D. Mc- World. Annals of Botany 100(5):903–924. Kay, Michael P. Heaton, Kefei Chen, Alan Cooper, et al. 2009. ———. 2009. Discussion. Wenner-Gren Foundation Symposium 141 Resolving the evolution of extant and extinct ruminants with high- on the Beginnings of Agriculture, Temozo´n, Mexico, March 6–13. throughput phylogenomics. Proceedings of the National Academy Garcea, Elena A. A. 2001. Cultural adaptations at Uan Tabu from of Sciences of the USA 106(44):18644–18649. the upper Pleistocene to the late Holocene. In Uan Tabu in the Denham, Tim. 2007. Early to mid-Holocene plant exploitation in settlement history of the Libyan Sahara. Elena A. A. Garcea, ed. Pp. New Guinea: towards a contingent interpretation of agriculture. 219–235. Florence, Italy: all’Insegna del Giglio. In Rethinking agriculture: archaeological and ethnoarchaeological ———. 2004. An alternative way towards food production: the per- perspectives. Tim Denham, Jose´Iriarte, and Luc Vrydaghs, eds. Pp. spectives from the Libyan Sahara. Journal of World Prehistory 18(2): 78–108. Walnut Creek, CA: Left Coast. 107–154. ———. 2011. Early agriculture and plant domestication in New ———. 2006. Semi-permanent foragers in semi-arid environments Guinea and Island Southeast Asia. Current Anthropology 52(suppl. in North Africa. World Archaeology 38(2):197–219. 4):S379–S395. Gautier, Achilles. 1984a. The fauna of the Neolithic site of Kadero Diamond, Jared. 1997. Guns, germs, and steel: the fates of human (central Sudan). In Origin and early development of food-producing societies. New York: Norton. cultures in north-eastern Africa. Lech Krzyzaniak and Michał Ko- di Lernia, Savino. 1998. Cultural control over wild animals during busiewicz, eds. Pp. 317–319. Poznan: Polish Academy of Sciences the Early Holocene: the case of Barbary sheep in central Sahara. and Poznan Archaeological Museum. In Before food production in North Africa. Savino di Lernia and ———. 1984b. Quaternary mammals and archaeozoology of Egypt Giorgio Manzi, eds. Pp. 113–126. Forlı´, Italy: ABACO. and the Sudan: a survey. In Origin and early development of food- Marshall and Weissbrod Domestication Processes and Morphological Change S411

producing cultures in north-eastern Africa. Lech Krzyzaniak and molecular signals of domestication. In Rethinking agriculture: ar- Michał Kobusiewicz, eds. Pp. 43–56. Poznan: Polish Academy of chaeological and ethnoarchaeological perspectives. Tim Denham, Jose´ Sciences and Poznan Archaeological Museum. Iriarte, and Luc Vrydaghs, eds. Pp. 36–49. Walnut Creek, CA: Left ———.1987a. The archaeological sequence of the Acacus. In Ar- Coast. chaeology and environment in the Libyan Sahara, 1978–1983.Bar- Jousse, He´le`ne, Henriette Obermaier, Michel Raimbault, and Joris bara E. Barich, ed. Pp. 283–308. Cambridge Monographs in Af- Peters. 2008. Late Holocene economic specialization through rican Archaeology 23, BAR International Series 368. Oxford: aquatic resource exploitation at Kobadi in the Me´ma, Mali. In- British Archaeological Reports. ternational Journal of Osteoarchaeology 18(6):549–572. ———. 1987b. Prehistoric men and cattle in North Africa: a dearth Kahlheber, Stefanie, and Katharina Neumann. 2007. The develop- of data and a surfeit of models. In Prehistory of arid North Africa: ment of plant cultivation in semi-arid West Africa. In Rethinking essays in honor of Fred Wendorf. Angela E. Close, ed. Pp. 163–187. agriculture: archaeological and ethnoarchaeological perspectives.Tim Dallas, TX: Southern Methodist University Press. Denham, Jose´Iriarte, and Luc Vrydaghs, eds. Pp. 320–346. Walnut ———. 2001. The Early to Late Neolithic archeofaunas from Nabta Creek, CA: Left Coast. and Bir Kiseiba. In The archaeology of Nabta Playa, vol. 1 of Ho- Keding, Birgit, Tilman Lenssen-Erz, and Andreas Pastoors. 2007. locene settlement of the Egyptian Sahara. Fred Wendorf and Ro- Pictures and pots from pastoralists: investigations into the pre- muald Schild, eds. Pp. 609–635. New York: Kluwer Academic. history of the Ennedi highlands in NE Chad. Sahara 18:23–46. Gifford-Gonzalez, Diane P. 2005. Pastoralism and its consequences. Kimura, Birgitta, Fiona B. Marshall, Shanyuan Chen, So´nia Rosen- In African archaeology: a critical introduction. Ann B. Stahl, ed. Pp. bom, Patricia D. Moehlman, Noreen Tuross, Richard C. Sabin, et 187–224. Oxford: Blackwell. al. 2010. Ancient DNA provides insights into African wild ass Goring-Morris, A. Nigel, and Anna Belfer-Cohen. 2011. Neolithi- phylogeny and donkey domestication. Proceedings of the Royal So- zation processes in the Levant: the outer envelope. Current An- ciety B: Biological Sciences 278(1702):50–57, doi:10.1098/rspb thropology 52(suppl. 4):S195–S208. .2010.0708. Haaland, Randi. 1987. Socio-economic differentiation in the Neolithic Klingel, Hans. 1974. Soziale organisation und Verhalten des Grevy- Sudan. Cambridge Monographs in African Archaeology 20, BAR Zebras (Equus grevyi). Zeitschrift fu¨r Tierpsychologie 36(1–5):37– International Series 350. Oxford: British Archaeological Reports. 70. ———. 1999. The puzzle of the late emergence of domesticated Kro¨pelin, Stefan, Dirk Verschuren, Anne-Marie Le´zine, Hilde Eg- sorghum in the Nile Valley. In The prehistory of food: appetites for germont, Christine Cocquyt, Pierre Francus, Jean-Pierre Cazet, et change. Chris Gosden and Jon Hather, eds. Pp. 397–418. London: al. 2008. Climate-driven ecosystem succession in the Sahara: the Routledge. past 6000 years. Science 320(5877):765–768. Hanotte, Olivier, Daniel G. Bradley, Joel W. Ochieng, Yasmin Verjee, Ku¨nzl, Christine S., Sylvia Kaiser, Edda Meier, and Norbert Sachser. Emmeline W. Hill, and J. Edward O. Rege. 2002. African pasto- 2003. Is a wild mammal kept and reared in captivity still a wild ralism: genetic imprints of origins and migrations. Science animal? Hormones and Behavior 43(1):187–196. 296(5566):336–339. Kuper, Rudolph, and Stefan Kro¨pelin. 2006. Climate-controlled Ho- Hildebrand, Elisabeth A. 2003a. Ensete, yams and honey: ethnoar- locene occupation in the Sahara: motor of Africa’s evolution. Sci- chaeological approaches to the origins of horticulture in southwest ence 313(5788):803–807. Ethiopia. PhD dissertation, Washington University, Saint Louis. Lane, Paul, Ceri Ashley, Oula Seitsonen, Paul Harvey, Sada Mire, ———. 2003b. Motives and opportunities for domestication: an eth- and Frederick Odede. 2007. The transition to farming in eastern noarchaeological study in southwest Ethiopia. Journal of Anthro- Africa: new faunal and dating evidence from Wadh Lang’o and pological Archaeology 22(4):358–375. Usenge, Kenya. Antiquity 81(311):62–81. ———. 2007. A tale of two tuber crops: how attributes of enset and Larson, Greger. 2011. Genetics and domestication: important ques- yams may have shaped prehistoric human-plant interaction in tions for new answers. Current Anthropology 52(suppl. 4):S485– southwest Ethiopia. In Rethinking agriculture: archaeological and S495. ethnoarchaeological perspectives. Tim Denham, Jose´Iriarte, and Luc Lee, Gyoung-Ah. 2011. The transition from foraging to farming in Vrydaghs, eds. Pp. 273–298. Walnut Creek, CA: Left Coast. prehistoric Korea. Current Anthropology 52(suppl. 4):S307–S329. Hole, Frank. 1996. The context of caprine domestication in the Zag- Lesur, Jose´phine, Jean-Denis Vigne, and Xavier Gutherz. 2007. Ex- ros. In The origins and spread of agriculture and pastoralism in ploitation of wild mammals in south-west Ethiopia during the Eurasia. David R. Harris, ed. Pp. 263–281. Washington, DC: Smith- Holocene (4000 BC–500 AD): the finds from Moche Borago Shel- sonian Institution. ter (Wolayta). Environmental Archaeology 12(2):139–159. Holl, Augustin F. C. 2004. Saharan rock art. Walnut Creek, CA: Linseele, Veerle. 2004. Size and size change of the African aurochs Altamira. during the Pleistocene and Holocene. Journal of African Archae- ———. 2005. Holocene “aquatic” adaptations in north tropical Af- ology 2(2):165–185. rica. In African archaeology: a critical introduction. Ann B. Stahl, ———. 2010. Did specialized pastoralism develop differently in Af- ed. Pp. 174–186. Oxford: Blackwell. rica than in the Near East? an example from the West African Honegger, Matthieu. 2004. Settlement and cemeteries of the Meso- Sahel. Journal of World Prehistory 23(2):43–77, doi:10.1007/s10963- lithic and Early Neolithic at El-Barga, Kerma region. Sudan and 101-9033-3. Nubia 8:27–32. Linseele, Veerle, Elena Marinova, Wim Van Neer, and Pierre Ver- ———. 2005. Kerma and the origin of the African Neolithic. Genava meersch. 2010. Sites with Holocene dung deposits in the Eastern 53:239–249. Desert of Egypt: visited by herders? Journal of Arid Environments Jesse, Friederike, Birgit Keding, Nadja Po¨llath, Marianne Bechhaus- 74(7):818–828. Gerst, and Tilman Lenssen-Erz. 2007. Cattle herding in the south- Livingstone Smith, Alexandre. 2001. Pottery manufacturing pro- ern Libyan Desert. In Atlas of cultural and environmental change cesses: reconstruction and interpretation. In Uan Tabu in the set- in arid Africa. Olaf Bubenzer, Andreas Bolten, and Frank Darius, tlement history of the Libyan Sahara. Elena A. A. Garcea, ed. Pp. eds. Pp. 46–49. Africa Praehistorica 21. Cologne: Heinrich Barth 113–152. Florence, Italy: all’Insegna del Giglio. Institut. Marshall, Katie, Ruth Pelling, Tom Higham, Jean-Luc Schwenniger, Jones, Martin, and Terry Brown. 2007. Selection, cultivation and and Dorian Q Fuller. 2011. 4500-year old domesticated pearl millet reproductive isolation: a reconsideration of the morphological and (Pennisetum glaucum) from the Tilemsi Valley, Mali: new insights S412 Current Anthropology Volume 52, Supplement 4, October 2011

into an alternative cereal domestication pathway. Journal of Ar- Pp. 185–188. Edinburgh: Centre for Tropical Veterinary Medicine, chaeological Science 38(2):312–322. University of Edinburgh. Marshall, Fiona. 2000. The origins and spread of domestic animals Murray, George W. 1935. Sons of Ishmael: a study of the Egyptian in East Africa. In The origins and development of African livestock: Bedouin. London: Routledge. archaeology, genetics, linguistics and ethnography. Roger M. Blench Mutundu, Kennedy K. 2005. Domestic stock age profiles and herd and Kevin C. MacDonald, eds. Pp. 191–221. London: University management practices: ethnoarchaeological implications from College London Press. Maasai settlements in southern Kenya. Archaeofauna 14:83–92. ———. 2001. Agriculture and use of wild and weedy greens by the Nelson, Charles M. 1995. The work of the Koobi Fora Field School Piik ap Oom Okiek of Kenya. Economic Botany 55(1):32–46. at the Jarigole Pillar site. Kenya Past and Present 27:49–63. ———. 2007. African pastoral perspectives on domestication of the Neumann, Katharina. 2003. The late emergence of agriculture in sub- donkey: a first synthesis. In Rethinking agriculture: archaeological Saharan Africa: archaeobotanical evidence and ecological consid- and ethnoarchaeological perspectives. Tim Denham, Jose´Iriarte, and erations. In Food, fuel, and fields. Katharina Neumann, Ann Butler, Luc Vrydaghs, eds. Pp. 371–407. Walnut Creek, CA: Left Coast. and Stefanie Kahlheber, eds. Pp. 71–92. Africa Praehistorica 15. Marshall, Fiona, Katherine Grillo, and Lee Arco. 2011. Prehistoric Cologne: Heinrich Barth Institut. pastoralists and social responses to climatic risk in East Africa. In ———. 2005. The romance of farming: plant cultivation and do- Sustainable lifeways: cultural persistence in an ever-changing envi- mestication in Africa. In African archaeology: a critical introduction. ronment. Naomi Miller, Katherine Moore, and Kathleen Ryan, eds. Ann B. Stahl, ed. Pp. 249–275. Oxford: Blackwell. University of Pennsylvania Museum of Archaeology and Anthro- Nicolaisen, Johannes. 1963. Ecology and culture of the pastoral Tuareg. pology Series. Philadelphia: University of Pennsylvania Press. Copenhagen: Copenhagen National Museum. Forthcoming. Outram, Alan K., Natalie A. Stear, Robin Bendrey, Sandra Olsen, Marshall, Fiona, and Elisabeth Hildebrand. 2002. Cattle before crops: Alexei Kasparov, Victor Zaibert, Nick Thorpe, and Richard P. Ev- the beginnings of food production in Africa. Journal of World ershed. 2009. The earliest horse harnessing and milking. Science Prehistory 16(2):99–143. 323(5919):1332–1335. Marshall, Fiona, and Lior Weissbrod. 2009. The consequences of O¨ zdog˘an, Mehmet. 2011. Archaeological evidence on the westward women’s use of donkeys for pastoral flexibility: Maasai ethnoar- expansion of farming communities from eastern Anatolia to the chaeology. In Tracking down the past: ethnohistory meets archaeo- Aegean and the Balkans. Current Anthropology 52(suppl. 4):S415– zoology. Gisella Grupe, George McGlynn, and Joris Peters, eds. Pp. S430. 59–79. Documenta Archaeobiologiae 7. Rahden/Westfalen, Ger- Pe´rez-Pardal, Lucı´a, Luis J. Royo, Albano Beja-Pereira, Shanyuan many: Leidorf. Chen, Rodolfo J. C. Cantet, Amadou Traore´, Ino Cˇ urik, et al. 2010. McCabe, J. Terrence. 2004. Cattle bring us to our enemies: Turkana Multiple paternal origins of domestic cattle revealed by Y-specific ecology, politics, and raiding in a disequilibrium system. Ann Arbor: interspersed multilocus microsatellites. Heredity 105(6):511–519, University of Michigan Press. doi:10.1038/hdy2010.30. McDonald, Mary A. 2008. Emerging social complexity In the mid- Pe´rez-Pardal, Lucı´a, Luis J. Royo, Albano Beja-Pereira, Ino Cˇ urik, Holocene Egyptian Western Desert: site 270 and its neighbors in Amadou Traore´, Iva´n Ferna´ndez, Johann So¨lkner, et al. 2009. Y- southeastern Dakhleh Oasis. In The oasis papers 2: proceedings of specific microsatellites reveal an African subfamily in taurine (Bos the Second International Conference of the Dakhleh Oasis Project. taurus) cattle. Animal Genetics 41(3):232–241. Marcia F. Wiseman, ed. Pp. 83–106. Oxford: Oxbow. Piperno, Dolores R., and Deborah M. Pearsall. 1998. The origins of Meadow, Richard. 2009. Bovid domestication and the development agriculture in the lowland Neotropics. San Diego, CA: Academic of pastoralism in South, Southeast and East Asia with a focus on Press. cattle and water buffalo. Paper presented at Wenner-Gren Foun- Plug, Ina, and Elisabeth A. Voigt. 1985. Archaeozoological studies of dation Symposium 141 on the Beginnings of Agriculture, Temo- Iron Age communities in southern Africa. Advances in World Ar- zo´n, Mexico, March 6–13. chaeology 4:189–238. Mengoni-Gon˜alons, Guillermo L., and Hugo Yacobaccio. 2006. The Po¨llath, Nadja, and Joris Peters. 2007. Holocene climate change, domestication of South American camelids: a view from the south- human adaptation and landscape degradation in arid Africa as central Andes. In Documenting domestication: new genetic and ar- evidenced by the faunal record. In Atlas of cultural and environ- chaeological paradigms. Melinda A. Zeder, Daniel G. Bradley, Eve mental change in arid Africa. Olaf Bubenzer, Andreas Bolten, and Emshwiller, and Bruce D. Smith, eds. Pp. 228–244. Berkeley: Uni- Frank Darius, eds. Pp. 64–67. Africa Praehistorica 21. Cologne: versity of California Press. Heinrich Barth Institut. Mercader Julio, Raquel J. Marti, Jayne Wilkins, and Kent D. Fowler. Prendergast, Mary E., and Paul J. Lane. 2010. Middle Holocene fish- 2006. The eastern periphery of the Yoruba cultural sphere: ceramics ing strategies in East Africa: zooarchaeological analysis of Pundo, from the lowland rain forests of southwestern Cameroon. Current a Kansyore shell midden in northern Nyanza (Kenya). Interna- Anthropology 47(1):173–184. tional Journal of Osteoarchaeology 20(1):88–112. Mercuri, Anna Maria. 1999. Palynological analysis of the Early Ho- Rindos, David. 1984. The origins of agriculture: an evolutionary per- locene sequence. In The Uan Afuda Cave: hunter-gatherer societies spective. New York: Academic Press. of central Sahara. Savino di Lernia, ed. Pp. 149–181. Florence, Italy: Roset, Jean-Pierre. 1987. Ne´olithisation, ne´olithique et post-ne´oli- all’Insegna del Giglio. thique au Niger nord-oriental. Bulletin de l’Association franc¸aise ———. 2001. Preliminary analyses of fruits and seed from the Early pour l’e´tude du Quaternaire 32(4):203–214. Holocene sequence. In Uan Tabu in the settlement history of the Rossel, Stine, Fiona Marshall, Joris Peters, Tom Pilgram, Matthew Libyan Sahara. Elena A. A. Garcea, ed. Pp. 161–188. Florence, D. Adams, and David O’Connor. 2008. Domestication of the don- Italy: all’Insegna del Giglio. key: new data on timing, processes and indicators. Proceedings of Moehlman, Patricia D. 2002. Equids: zebras, asses and horses: status the National Academy of Sciences of the USA 105(10):3715–3720. survey and conservation plan. Gland, Switzerland: International Rowley-Conwy, Peter. 2011. Westward ho! the spread of agriculture Union for Conservation of Nature. from central Europe to the Atlantic. Current Anthropology Mohammed, A. 1991. Management and breeding aspects of donkeys 52(suppl. 4):S431–S451. around Awassa, Ethiopia. In Donkeys, mules and horses in tropical Ryan, Kathleen, Karega-Munene, Samuel M. Kahinju, and Paul N. agricultural development. Denis Fielding and R. Anne Pearson, eds. Kunoni. 2000. Ethnographic perspectives on cattle management Marshall and Weissbrod Domestication Processes and Morphological Change S413

in semi-arid environments: a case study from Maasailand. In The Roger M. Blench and Kevin C. MacDonald, eds. Pp. 163–190. origins and development of African livestock: archaeology, genetics, London: University College London Press. linguistics, and ethnography. Roger M. Blench and Kevin C. Mac- Vigne, Jean-Denis, Isabelle Carre`re, Franc¸ois Briois, and Jean Gui- Donald, eds. Pp. 462–477. London: University College London laine. 2011. The early process of mammal domestication in the Press. Near East: new evidence from the pre-Neolithic and Pre-Pottery Saxon, Earl C., Angela E. Close, C. Cluzel, V. Morse, and N. J. Neolithic in Cyprus. Current Anthropology 52(suppl. 4):S255–S271. Shackleton. 1974. Results of recent investigations at Tamar Hat. Vila´, Carles, Jennifer A. Leonard, and Albano Beja-Pereira. 2006. Libyca 22:49–91. Genetic documentation of horse and donkey domestication. In Schild, Romuald, and Fred Wendorf. 2001. Combined Prehistoric Documenting domestication: new genetic and archaeological para- Expeditions’s radiocarbon dates associated with Neolithic occu- digms. Melinda A. Zeder, Daniel G. Bradley, Eve Emshwiller, and pation in the southern Western Desert of Egypt. In The archaeology Bruce D. Smith, eds. Pp. 342–353. Berkeley: University of Cali- of Nabta Playa, vol. 1 of Holocene settlement of the Egyptian Sahara. fornia Press. Fred Wendorf and Romuald Schild, eds. Pp. 51–56. New York: Wasylikowa, Krystyna, Jack R. Harlan, John Evans, Fred Wendorf, Plenum. Romuald Schild, Angela Close, Halina Kro´lik, and Rupert Housley. Schneider, Harold K. 1979. Livestock and equality in East Africa. 1993. Examination of botanical remains from Early Neolithic Bloomington: Indiana University Press. houses at Nabta Playa, Western Desert, Egypt, with special ref- Sereno, Paul C., Elena A. A. Garcea, He´le`ne Jousse, Christopher M. erence to sorghum grains. In The archaeology of Africa: food, metals Stojanowski, Jean-Franc¸ois Salie`ge, Abdulaye Maga, Oumarou A. and towns. Thurstan Shaw, Paul Sinclair, Bassey Andah, and Alex Ide, et al. 2008. Lakeside cemeteries in the Sahara: 5000 years of Okpoko, eds. Pp. 154–164. London: Routledge. Holocene population and environmental change. PLoS ONE 3(8): Wendorf, Fred, and Halina Kro´lik. 2001. Site E-96-1: the complex e2995, doi:10.1371/journal.pone.0002995. structures or shrines. In The archaeology of Nabta Playa, vol. 1 of Shahack-Gross, Ruth, Fiona Marshall, and Steve Weiner. 2003. Geo- Holocene settlement of the Egyptian Sahara. Fred Wendorf and Ro- ethnoarchaeology of pastoral sites: the identification of livestock muald Schild, eds. Pp. 503–520. New York: Plenum. enclosures in abandoned Maasai settlements. Journal of Archaeo- Wendorf, Fred, and Romuald Schild. 1998. Nabta Playa and its role logical Science 30(4):439–459. in northeastern African prehistory. Journal of Anthropological Ar- Shahack-Gross, Ruth, Allison Simons, and Stanley A. Ambrose. 2008. chaeology 17(2):97–123. Identification of pastoral sites using stable nitrogen and carbon ———. 2001. Site E-75-9: the excavation of an El Adam (?)Early isotopes from bulk sediment samples: a case study in modern and Neolithic settlement at Nabta Playa. In The archaeology of Nabta archaeological pastoral settlements in Kenya. Journal of Archaeo- Playa,vol.1ofHolocene settlement of the Egyptian Sahara.Fred logical Science 35(4):983–990. Wendorf and Romuald Schild, eds. Pp. 97–110. New York: Plenum. Smith, Abigail C. 2008. Identifying a possible domestication: a mor- Wendorf, Fred, Romuald Schild, and Angela E. Close, eds. 1984. phometric study of Barbary sheep (Ammotragus lervia). Poster Cattle keepers of the eastern Sahara: the Neolithic of Bir Kiseiba. presented at the Society of Africanist Archaeologists 19th Biennial Dallas, TX: Institute for the Study of Earth and Man, Southern Meeting, Frankfurt, September 8–11. Methodist University. Smith, Andrew B. 1992. Pastoralism in Africa: origins and development Western, David, and Thomas Dunne. 1979. Environmental aspects ecology. London: Hurst. of settlement site decisions among pastoral Maasai. Human Ecology ———. 1998. Keeping people on the periphery: the ideology of social 7(1):75–98. hierarchies between hunters and herders. Journal of Anthropological Wilson, R. Trevor. 1991. Equines in Ethiopia. In Donkeys, mules and Archaeology 17(2):201–215. horses in tropical agricultural development. Denis Fielding and R. ———. 2005. African herders. Walnut Creek, CA: Altamira. Anne Pearson, eds. Pp. 185–188. Edinburgh: Centre for Tropical Smith, Bruce D. 2001. Low level food production. Journal of Ar- Veterinary Medicine, University of Edinburgh. chaeological Research 9(1):1–43. Woodburn, James. 1982. Egalitarian societies. Man, n.s., 17(3):431– ———. 2011. The cultural context of plant domestication in eastern 451. North America. Current Anthropology 52(suppl. 4):S471–S484. Wylie, Alison. 2002. Thinking from things. Berkeley: University of Spear, Thomas, and Richard Waller, eds. 1993. Being Maasai: ethnicity California Press. and identity in East Africa. London: Currey. Yen, Douglas E. 1989. The domestication of environment. In Foraging Tafuri, Mary Anne, R. Alexander Bentley, Giorgio Manzi, and Savino and farming: the evolution of plant exploitation. David R. Harris di Lernia. 2006. Mobility and kinship in the prehistoric Sahara: and Gordon C. Hillman, eds. Pp. 55–75. London: Unwin Hyman. strontium isotope analysis of Holocene human skeletons from the Zeder, Melinda A. 2008. Domestication and early agriculture in the Acacus Mts. (southwestern Libya). Journal of Anthropological Ar- Mediterranean Basin: origins, diffusion and impact. Proceedings of chaeology 25(3):390–402. the National Academy of Sciences of the USA 105(33):11597–11604. Tishkoff, Sarah A., Floyd A. Reed, Alessia Ranciaro, Benjamin F. ———. 2009. Macroevolutionary theory and the study of culture Voight, Courtney C. Babbitt, Jesse S. Silverman, Kweli Powell, et change. Journal of Archaeological Research 17(1):1–63. al. 2006. Convergent adaptation of human lactase persistence in ———. 2011. The origins of agriculture in the Near East. Current Africa and Europe. Nature Genetics 39(1):31–40, doi:10.1038/ Anthropology 52(suppl. 4):S221–S235. ng1946. Zeder, Melinda A., and Brian Hesse. 2000. The initial domestication Trut, Lyudmila. 1999. Early canid domestication: the farm-fox ex- of goats (Capra hircus) in the Zagros Mountains 10,000 years ago. periment. American Scientist 87(2):160–169. Science 287(5461):2254–2257. van Neer, Wim. 2000. Domestic animals from archaeological sites in Zhao, Zhijun. 2011. New archaeobotanic data for the study of the central and west-central Africa. In The origins and development of origins of agriculture in China. Current Anthropology 52(suppl. 4): African livestock: archaeology, genetics, linguistics, and ethnography. S295–S306.

Current Anthropology Volume 52, Supplement 4, October 2011 S415

Archaeological Evidence on the Westward Expansion of Farming Communities from Eastern Anatolia to the Aegean and the Balkans

by Mehmet O¨ zdog˘an

The beginnings of the Neolithic way of life in Europe and the role played by the Anatolian Peninsula in this process are much-debated issues that involve a number of distinct topics. In this debate, it should not be overlooked that distinct from Europe, at least a portion of the Anatolian plateau had been part of the “Neolithic world” for at least 4,000 years before the appearance of the earliest claimed Neolithic culture in Europe. Accordingly, in viewing the interaction between southeastern Europe and the Aegean with the Anatolian Peninsula, the core area of primary Neolithization has to be considered.

The Neolithic way of life, after stabilizing in the region of its and Chi 2010) or considered in the view of conventional origin for several millennia, was rather suddenly dispersed to hypothetical frameworks, which has led to a rather chaotic other regions by the beginning of the seventh millennium BC. admixture of conventional models with a random selection The mode and the pace of this expansion have been much of new data. It is necessary to begin with a conspectus ques- debated. The problem has been unresolved because of the tioning the history of the problem. lack of research in the peripheral areas of primary Neolithi- zation and in particular in the contact zone between Anatolia Preamble: A Prelude to the Problem and southeastern Europe. During the last 2 decades, there have been a number of new excavations in this critical zone The question of how farming began in Europe has been much that for the first time have provided concrete evidence re- debated since the first half of the last century (Tringham garding the expansion of the Neolithic way of life. Until re- 1971). Because the question pertains to both understanding cently, the Neolithic period of western Anatolia was known, the foundations of European civilization in addition and sci- aside from excavations such as at Hacılar, largely from sherds entific concerns, it has been a stimulus to politics as well collected from the surface. Thus, all comparisons between (O¨ zdog˘an 2004, 2007a). It seemed evident in the early years Anatolia and the Balkans had to have their focus on stylistic of research that the area key to the resolution of this problem features of the pottery assemblage. Now, though data collec- was southeastern Europe—the Aegean and the Balkans. As tion is still in a preliminary stage, there is a wide range of can be seen in Rowley-Conwy (2011), defining the process available data covering settlement organizations, architectural of Neolithization in western parts of Europe is less problem- features, subsistence patterns, lithic technologies, and burial atic, because there the discussions are based solely on solid customs. Many of these sites are still being excavated, and evidence. In the southeastern parts of Europe, however, ar- thus, the results are available mainly in the form of prelim- chaeological biases bear not only on political concerns but inary reports. However, even a preliminary assessment of what even the origin of agriculture. Accordingly, before presenting has been published is enough to give insight into a newly the recent evidence from western Turkey, for the sake of clarity appearing picture. A brief survey of recent literature indicates it is necessary to include an overview of changing trends that these new data have been either overlooked (Anthony pertaining to the origin of Early Neolithic cultures. There has been so much controversy regarding the begin- Mehmet O¨ zdog˘an is Professor in Prehistorya Anabilim Dalı, ning of the Neolithic way of life in southeastern Europe that Edebiyat Faku¨ltesi, I˙stanbul U¨ niversitesi (34134 Istanbul, Turkey the question has occasionally been distorted. In the midst of [[email protected]]). This paper was submitted 13 XI 09, accepted ongoing discussions, it is worth keeping in mind that the 9 XII 10, and electronically published 11 VIII 11. initial concern is to understand how and when the Neolithic

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0017$10.00. DOI: 10.1086/658895 S416 Current Anthropology Volume 52, Supplement 4, October 2011 way of life began in southeastern Europe. The quest to answer to 3200 BC. At the same time, it seemed evident that the these questions has followed two distinct trajectories, consid- Neolithic communities of southeastern Europe had been de- ering Neolithization either as the result of a local development rived from the Near East. As the presence of Neolithic cultures or as being introduced from the East. Uncompromising con- in Anatolia was beyond consideration, various models were frontation between these two approaches remains an enduring envisaged to connect littoral areas of the eastern Mediterra- dilemma. Nevertheless, it is now evident that the emergence nean with the Balkans—bypassing Anatolia. In this context, of the Neolithic way of life in southeastern Europe was very Childe noted in 1957 that “no recognizable archaeological complex and did not develop along a single line but from milestones mark an ancient route across Anatolia from the different types of simultaneous occurrences. Orient to Europe” (Childe 1964 [1957]:36) and “Sakc¸ego¨zu¨ As noted, the resolution of Neolithic origins in southeastern (Coba Ho¨yu¨k) further north seem[s] to prove that Meso- Europe is embedded in the Neolithic culture’s interaction with potamian farmers with slings and gaily painted pots colonized the Near East, where an uneven distribution of excavated sites the Orontes valley; indeed they penetrated right to the Med- due to research strategies has been the primary agent to ham- iterranean coasts at least at Ugarit (Ras Shamra)” (Childe 1964 per reaching a consensus. The excitement triggered by the [1957]:217). Some years later, Seton Lloyd also commented transformation of the meaning of the term “Neolithic” from that “the region more correctly described as Anatolia shows technology to imply a new way of living based on food pro- no sign whatever of habitation during the Neolithic period” duction had a significant impact in designing field projects, (Lloyd 1956:53–54) and “various phases were later found, first both in the Near East and in southeastern Europe, consid- in north Syria and then at Taurus and Mersin beyond the erably increasing the number of Neolithic excavations. In the Turkish frontier . . . testifying to the westward and northward initial stages, Neolithic research was focused mainly on two extension of the great Chalcolithic province . . . barrier still geographic regions, one in the semiarid regions of the Near existed and beyond it Anatolia remained unpopulated” (Lloyd East, mainly along the so-called Fertile Crescent, and the other 1956:58–61). in southeastern Europe. The former was considered to be the While culture historians were trying to define routes be- core area or the main formative zone of sedentary farming tween the Levant and the Aegean, others working in south- communities, and the latter was considered to be where the eastern Europe were more engaged in searching for parallels European Neolithic originated. with the Neolithic sequence of the Near East. In this respect, The geographic locations of the research projects in the one of the active discussions of the time was the one triggered Near East were selected based on the following considerations: by V. Milojcˇic´, a prominent archaeologist on the prehistory (a) regions where environmental restrictions would necessi- of the Balkans who argued for the presence of an “Aceramic tate consuming cereals; (b) zones of the natural habitat of the Neolithic” horizon in Thessaly (Milojcˇic´ 1960) that was al- primary components of the Neolithic package, mainly the most identical to the cultural stage that had just been iden- regions where wild cereals and wild forms of domestic animals tified in the Near East. Needless to say, this was a highly were present; and (c) areas where the political context did paradoxical model, because Milojcˇic´ believed in both the Near not hamper archaeological research. Given such consider- Eastern origins of the Balkan Neolithic and the very young ations, research on the Neolithic Period was concentrated in dates ascribed to it. Nevertheless, given the dates for the Bal- the Levant and, to a lesser degree, along the eastern arm of kan and the Anatolian cultures, numerous parallels were put the Fertile Crescent. The work in the Near East was research forth between these two regions, the most notable being the oriented, focused on understanding various aspects of the association of Troy I with the Vincˇa culture. As narratives of Neolithic process. In southeastern Europe, however, excavat- changing perceptions regarding the Neolithization model of ing a Neolithic site was considered somewhat more prestigious Europe have been published extensively (Bogucki 1996; O¨ z- then working on other periods. By the second half of the dog˘an 1995; Renfrew 2002; Sherratt 2004), here I shall note twentieth century, the number of excavated Neolithic sites in only a few relevant stages. southeastern Europe had reached 200. By the third quarter The implementation of 14C dating, in particular for the of the twentieth century, there had been considerable knowl- early cultures of southeastern Europe, led to the creation of edge accumulated on both ends of the Anatolian Peninsula, a “mental fault line” between the early Balkan and Anatolian which until then had been largely outside the sphere of interest cultures. The 14C revolution came at a time when numerous of Neolithic research. On a general level, Anatolia was dis- Neolithic sites were being excavated in southeastern Europe, cussed mainly as the connection between the Near Eastern revealing that the Balkan cultures were 2,000 to 3,000 years Neolithic cultures and the Balkans, but even in this context, older than assumed. While the cultural sequence in the Bal- the question was whether Anatolia was a barrier or a cultural kans was going through revolutionary changes, almost no bridge (French 1986; Tringham 2000). prehistoric excavations were taking place on the Anatolian In this respect, it should not be forgotten that before the plateau, and the conventional dates established before the 14C implementation of 14C dating, the beginning of the Neolithic revolution, particularly from sites such as Alis¸ar with a long Period in southeastern Europe, including that of mainland sequence, were not questioned for decades (Mellaart 1960; Greece, was believed to be rather late, sometime around 3400 O¨ zdog˘an 1996). The discrepancy among the dates of the early O¨ zdog˘an Westward Expansion of Farming S417 cultures of these regions, now spanning almost 3,000 years, began once more to be considered as the source area of Eu- coupled with the uneven geographic distribution of excava- ropean Neolithic cultures. It is of interest to note that ar- tions, eliminated Anatolia as the source area of Balkan cul- guments based on ethnolinguistic and/or biogenetic studies tures. This, along with contemporary political concerns, led were more instrumental than conventional comparisons based to a new way of thinking based on disregarding endemic on artifact assemblages in silencing antidiffusionistic ap- movements between regions, conveniently known as the “an- proaches (Renfrew 2002). tidiffusionist” approach. Through some decades, the discus- During the last decade or so, the picture has almost com- sion of the origin of the European Neolithic developed along pletely changed; what we now see is a revival of the old Bal- this line. It was hypothesized that there was an autochthonous kano-Anatolian cultural complex based on more concrete evi- development in the Balkans devoid of any impact from the dence. All of the previous explanations, whether diffusionist East. Thus, the concept of a “source area” for the European or antidiffusionist, are now regarded as overly simplistic. Even Neolithic cultures became firmly attached to the Balkans. if there are still considerable lacunae in our knowledge, it is Mention of possible diffusion from the East was considered evident that the real picture was far more complex than ever an embarrassment. Concerning this interim stage, the ques- envisaged. tion of how staple crops such as wheat, which do not have wild ancestors in southeastern Europe, appeared in the early Defining the Problem Neolithic sites remained a subject of unresolved debate. In spite of the prevailing tendencies, a few scholars continued The Neolithic of southeastern Europe has been rather well to consider the Eastern origins of European civilization. studied through hundreds of excavations evenly distributed Among them, Garas¸anin, with her “Balkano-Anatolian cul- in the Balkans and in the Aegean. Even though considerable tural complex” (Garas¸anin 1981, 1997), and Theocharis, in- amounts of data have been accumulated—in addition to bi- sistent on Anatolia as the source of the Aegean Neolithic ases that are deeply rooted in the history of research, as noted (Theocharis 1973), need to be acknowledged. above—the development of a clear vision of the Neolithiza- The recovery of early sites—Hacılar first, followed by C¸a- tion process has also been obscured by working with mis- talho¨yu¨k, Can Hasan, Su¨berde, and Erbaba on the Anatolian conceived definitions. Among them, overlooking the differ- plateau—initiated a number of discussions questioning both ence between the appearance of the “Neolithic way of life” the place of Anatolia in the process of Neolithic formation and “Neolithic elements” has considerably obstructed an- in the Near East and its role as the source of the European swering the question “how” Neolithic dispersal happened. A Neolithic. In this debate the insight and intuition of James Neolithic way of life implies a village life with all of its social Mellaart (1975) and David French (1986) stand as landmarks network and regulations; it can hardly be imitated unless the in noting the significant importance of the Anatolian Pen- settlers were village dwellers previously. On the other hand, insula in the process of Neolithization at such an early date commodities can travel and technologies can be learned. In and with such little evidence. In this respect, Fritz Schach- this respect, for example, Neolithic culture appears in the ermeyr should also be acknowledged as one of the few scholars Balkans or in Cyprus as a way of living in villages, regardless of his time to perceive the connections between Aegean, An- of their size. However, farther to the west, in the Mediter- atolian, and Near Eastern assemblages (Schachermeyr 1976). ranean littoral where certain commodities or technologies oc- However, it is interesting to note that in spite of the flam- cur, there is hardly any evidence of village life in the sense of boyant finds from C¸atalho¨yu¨k and elsewhere, Anatolia for a Anatolia or the Balkans. long time was still not included in theoretical discussions of The next question to be asked is “when”; the answer, in the emergence of Neolithic cultures. spite of all the available 14C dates, is still far from being re- As new excavations commenced in Anatolia, the antidif- solved. The main bias of this line of thought is considering fusionist views went through a gradual stage of transformation the Neolithization of southeastern Europe as an instantaneous (Harris 2003; O¨ zdog˘an 1995, 2007b; Runnels 2003). In this event. We now realize that it was an extended process that respect, excavations at C¸atalho¨yu¨k and Hacılar caused con- took place in installments and that lasted for more than a siderable excitement with their colorful artifactual finds; how- thousand years. Compressing the events that had such a long ever, the main impetus in bringing Anatolia into the theo- history into one arbitrary horizon has seriously jeopardized retical framework of the Neolithic studies was the data on any ability to make sense out of the absolute dates. the natural environment provided by the natural scientists One other obstacle that deserves to be noted here is the taking part in the Cayo¨nu¨ excavations. Nevertheless, until the “conceptual” border between the East and the West (O¨ zdog˘an commencement of excavations at Nevali C¸ ori, Hallan C¸emi, 2007a). Archaeology as a science, in Anatolia and in the Bal- and Asikli, the conventional view of disregarding Anatolia kans, has developed in different directions. Different and at from the primary zone of neolithization endured. At first the the same time opposing political regimes in these regions have presence of early Neolithic settlements on the plateau was created a barrier to sharing knowledge. For decades, scholarly taken as a late reflection of the Levantine Neolithic, connoted contacts and the flow of data between Anatolia and south- as an “area of secondary Neolithization.” Later still, Anatolia eastern Europe was minimal if any did pass through. More- S418 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 1. Core area of Neolithic formation and sites mentioned in the text. 1, Franchti; 2, Sesklo; 3, Argitsa-Magula; 4, Krainitsi; 5, Lepenski Vir; 6, Vlassac; 7, Gradesnitsa; 8, Koprivets; 9, Karanovo; 10, As¸ag˘ı Pınar; 11, Hoca C¸es¸me; 12, Ug˘urlu-Zeytinlik; 13, Yarımburgaz; 14, Yenikapı; 15, Fikirtepe; 16, Pendik; 17, Ilıpınar; 18, Mentes¸e; 19, Barc¸ın; 20, Demirci Ho¨yu¨k; 21, Kec¸ic¸ayırı; 22, Aktopraklık; 23, Troya; 24, Ege Gu¨bre; 25, Ulucak; 26, Yes¸ilova; 27, C¸ukuric¸i Ho¨yu¨k; 28, C¸ine-Tepecik; 29, Peynir- c¸ic¸eg˘i; 30, O¨ ku¨zini; 31, Bademag˘acı; 32, Ho¨yu¨cek; 33, Kuruc¸ay; 34, Ha- cılar; 35, Su¨berde; 36, Erbaba; 37, C¸atalho¨yu¨k; 38, Can Hasan; 39, As¸ıklı; 40, Gelveri; 41, Kaletepe; 42, Tepecik-C¸iftlik; 43, Ko¨s¸k Ho¨yu¨k; 44, Yu- muktepe. Gray shading indicates the primary zone of Neolithization, that is, the core area of the Neolithic. over, these regions have developed distinct schools of social Europe Aegean are not limited to what I have noted above. science. Anatolian archaeology has traditionally been Meso- If paucity of Neolithic research is the prime obstacle in An- potamia centric, almost totally uninterested in Europe, in- atolia, the picture in the Balkans is blurred by an abundance cluding the Turkish part of Thrace. On the other hand, most of data in which the information either remains as a chaotic of the archaeology in southeastern Europe concentrated on bulk or is of questionable context. It is rather regrettable that a local scale. Thus, even simple concepts such as the definition a great number of excavations in southeastern Europe have of culture or of a site differed considerably between these been conducted as small soundings with no clearly defined regions. This, coupled with the problems of following the stratigraphy or architectural features. Even with extensive ex- outcomes of recent archaeological work in the other region, cavations, publication of the data has been extremely selective made any reasonable dialogue almost impossible between An- and not always presented in an objective manner. Likewise, atolia and southeastern Europe. the bulk of 14C dates are from unclear contexts or processed Constraints on drawing a comprehensive picture between in unreliable labs. However, in spite of all these obstacles, with the Neolithic cultures of Near East Anatolia and southeastern an unbiased approach, the recent evidence provides grounds O¨ zdog˘an Westward Expansion of Farming S419 to develop a new perspective and allows reformulation of Neolithic reached southeastern Europe, this became a new questions. It is too early to draw a conclusive picture, but at core area for the expanding zones. Components of initial least now with the new information pouring in from the expansion, whether an endemic movement or the movement western parts of Turkey, it is possible to redefine the problem of ideas and commodities, can clearly be observed only in and to ask proper questions. Accordingly, the following are this region not only because of its immediate proximity to essential questions that need to be asked before any general the core area but also because there is no need to sort out statements on how and when the Neolithic way of life began the impact of local assemblages, as is the case in much of in southeastern Europe can be made. Europe.

Defining the Western Border of the Primary Defining “Neolithic Packages” and Trajectories Zone of Neolithization For a long time the concept of the “Neolithic package” was The area of primary Neolithization, or the core area of Neo- limited to cultivated plants, domesticated animals, ground- lithic formation, which long was considered to be restricted stone artifacts, and pottery, and it ignored not only other to the regions in and around the Fertile Crescent, is now elements such as prestige or cult objects but also more spe- firmly extended to cover southeastern sections of the central cifically architecture, the arrangement of settlements, and the Anatolian basin. Along with the excavations at Pınarbas¸ı, way of life. It now seems evident that there were different As¸ıklı Ho¨yu¨k, and Kaletepe (O¨ zdog˘an and Bas¸gelen 2007), a types of Neolithic packages—though almost all contained cul- number of surveys conducted in the southeastern section of tivated plants, domesticated animals, groundstone artifacts, inner Anatolia have pushed its beginning to considerably ear- and pottery—that reflect different social structures and ways lier dates, implying that here the process of Neolithization of living. Thus, to define patterns, it is necessary to make a was more or less parallel to that of the Fertile Crescent (fig. more detailed analysis of the assemblages and to work out 1). Even though it is not possible to define the western and trajectories of Neolithic expansion (O¨ zdog˘an 2010). northern boundaries of the primary Neolithic formation zone, it is clear that it did not extend too much in western and The Evidence northern directions. What was there in those times is far from clear; even in the most intensively surveyed areas, material Regretfully, there is very little new data from Greece and from indicative of the presence of a Mesolithic/Epipaleolithic ho- the Balkans in the last 2 decades on the initial stages of the rizon, as is the case in Greece and southern Bulgaria, is re- Neolithic period. There are numerous publications, though stricted to the coastal areas, suggesting that the inner parts each with a distinct view on the Aegean (Runnels 2001), might be devoid of population (O¨ zdog˘an 2005, 2008). As Greece (Perle`s 2001), and Bulgaria (Nikolov 2003, 2004; Ni- revealed by excavated coastal sites such as O¨ ku¨zini or Franch- kolov, Bacvarov, and Kalchev 2004; Todorova 1995). On the thi, the Mesolithic assemblages bear no indication of contact other hand, there is now an unprecedented inflow of new with the contemporary Neolithic cultures nor any element information from the western and northwestern parts of Tur- that could be taken as a forerunner of the Neolithic cultures key (fig. 1). The geographic distribution and the status of that appeared later in those areas. This strongly implies that research of sites relevant to the subject matter of this article up to a certain period, Neolithic cultures remained within the are listed below as a gazetteer. The primary information on formation zone without expanding their area. these sites, with extensive illustrations and references, are to be found in O¨ zdog˘an and Bas¸gelen (2007). Here I note only Defining the Contact Zone the most recent publications. This is an issue that until recently has been largely overlooked. Eastern Thrace Most of the literature dealing with the Neolithization process in Europe has focused either on the northern and western Hoca C¸es¸me: small mound (Bertram and Karul 2005); As¸ag˘ı sections of the Balkan peninsula or central Europe, where Pınar: mound, excavations continuing (Karul et al. 2003); there existed a strong Mesolithic/Epipaleolithic substratum Yarımburgaz Cave and I˙stanbul Yenikapı: submerged site and (Bogucki 1996; Pinhasi 2003; Zvelebil and Lillie 2000). What necropolis excavations continuing (Kızıltan 2007); I˙stanbul happened in these areas with the advent of Neolithic cultures C¸ekmece survey (Aydıngu¨n 2009). is apt to be different from regions that were either devoid of habitation or very sparsely inhabited, such as the western parts Eastern and Southern Marmara of Anatolia, mainland Greece, and southern Bulgaria (Gatsov and O¨ zdog˘an 1994; O¨ zdog˘an 2007b, 2008). Evidently, western Fikirtepe: flat site; Pendik: flat site and necropolis; Ilıpınar: parts of the Anatolian Peninsula, the immediate periphery of mound and necropolis (Roodenberg and Alpaslan-Rooden- the primary zone of Neolithization, formed the main “contact berg 2008); Mentes¸e: mound; Barc¸ın: small mound, excava- zone” during the initial stage of expansion. Later, after the tions continuing (Roodenberg, van As, and Alpaslan-Rooden- S420 Current Anthropology Volume 52, Supplement 4, October 2011 berg 2008); Aktopraklık: mound and necropolis, excavations penski Vir and Vlassac in the Iron Gates of the Danube. continuing. However, for each of these cases there have been others claim- ing that neither of these sites were Aceramic and that some Coastal Aegean pottery was always present, including whole pots in photo- graphs from Lepenski Vir (Garas¸anin and Radovanovic¸2001; Ulucak: mound, excavations continuing (Abay 2005; C¸ ilin- Reingruber 2005). A similar case can be seen at Hacılar in girog˘lu and Abay 2005); Yes¸ilova: mound, excavations con- western Anatolia. Mellaart claimed to find an Aceramic tinuing; Ege Gu¨bre: mound, excavations continuing; C¸ukuric¸i mound below the alluvial plain on the outskirts of the pottery Ho¨yu¨k: mound, excavations continuing (Horejs 2008); C¸ ine- mound of Hacılar (Mellaart 1970:3–7). Much later at C¸ atal- Tepecik: mound, excavations continuing (Gu¨nel 2003); Bes¸- ho¨yu¨k, Mellaart noted that an Aceramic layer was recovered parmak Latmos: cave site, excavations continuing (Peschlow- at the bottom of the deep sounding (Mellaart 1989:316). Duru Bindokat 2003); Ug˘urlu-Zeytinlik: small mound, excavations (1989), reexcavating Mellaart’s Aceramic mound at Hacılar, continuing (Harmankaya and Erdog˘u 2003); Cos¸kuntepe: found some coarse sherds in situ, and at C¸atalho¨yu¨k the pres- small mound, excavations continuing (Seeher 1990; Takaog˘lu ence of a Pre-Pottery horizon has not been confirmed, al- 2005). though pottery is reported to be extremely rare both at basal Hacılar and in the deep layers of C¸atalho¨yu¨k. A similar claim Coastal Mediterranean has been made at Ulucak following the 2009 season, reaching a Pre-Pottery level, phase VI, with red-coated floors and walls Peynirc¸ic¸eg˘i: cave site, excavations continuing (S. Yaylalı, un- (C¸evik 2010); the available dates of the overlaying phase are published manuscript, 2007); Suluin: cave site, excavations as early as the end of the eighth millennium BC. continuing (Tas¸kıran 2008). There have been other suggestions regarding the presence of Pre-Pottery Neolithic sites in various surveys conducted in Inner West Anatolia western Anatolia, including Kec¸ic¸ayırı, located on the Phry- Kuruc¸ay: mound; Ho¨yu¨cek: mound (Duru 2008); Badema- gian highlands in the eastern part (Efe 1997, 2005), and C¸alca g˘acı: mound, excavations continuing (Duru 2008); Kec¸ic¸ayırı: (O¨ zdog˘an and Gatsov 1998), based on lithic typology, but flat hilltop site, excavations continuing (Efe 2005); Demirci some pottery is also present at both sites. For the other claims Ho¨yu¨k: mound; Dedecik-Heybelitepe: small mound, exca- for Pre-Pottery sites, mainly from the littoral areas of both vations continuing. Turkey and Greece, there is so little lithic material available As the above list documents, over the last 2 decades a total that it is not possible to say with any confidence whether they of 26 Neolithic sites have been excavated in the western parts represent a Mesolithic or a younger horizon. of Turkey. Thirteen of these are currently being excavated. Even though the evidence does not support the existence The increased number of excavated sites is certainly of sig- of an Aceramic horizon in the regions west of the central nificance; however, what is more important is the fact that Anatolian basin, it still should not be excluded as a possibility. most of these sites are multilayered mounds with Neolithic As noted above, pottery is extremely rare in the basal layers deposits 4–8 m thick, and they are being extensively excavated; of all known early Neolithic sites, including those such as most of the excavations are conducted by competent multi- C¸atalho¨yu¨k and Bademag˘acı, which contain a long sequence disciplinary teams. Along with the increase in the number of of development. The fact that the earliest horizons with pot- excavations, there are now numerous surface surveys in var- tery have dates as early as the eighth millennium implies that ious parts of the region providing not only new evidence but the first detectable expansion of the Neolithic culture took also making it possible to define distribution patterns. Here place either at the beginning of the Pottery Neolithic period, I will restrict myself to presenting a conspectus on certain when pottery vessels were rare, or just before it, more or less selected issues that contribute to developing a new insight on between 7400 and 7100 cal BC. Nevertheless, it would not be the dispersal of a Neolithic way of life. at all surprising if there were some sporadic movements to the west at a slightly earlier date. A Dilemma: The Aceramic Horizon Multiple Trajectories and Maritime The beginnings of a Neolithic way of life in the western parts Expansion of Anatolia, as in the Balkans, are far from clear; the evidence for this incipient stage is very scanty and at the same time The second controversy involves the route of this initial dis- open to controversy. At the root of this problem lies the persal. Given the recent information from the Anatolian pla- question of whether a Pre-Pottery or Aceramic Neolithic teau, we have always considered a land route rather than the phase was present. Even though the question had its origins maritime route suggested by Childe in earlier years. Recently, in parallels with the Near East, it is deeply embedded in however, Perle`s (2003, 2005), after analysis of early Neolithic Milojcˇic´’s claims for an Aceramic horizon in Thessaly (Mi- assemblages in Greece, suggested that there are some elements lojcˇic´ 1960). Similar claims were later put forward for Le- in the Aegean that seems to have arrived directly from the O¨ zdog˘an Westward Expansion of Farming S421 eastern Mediterranean, bypassing Anatolia. The recovery of pansion, there is a succession of rapid and rather massive Early Neolithic sites in Cyprus and in some of the Aegean movements of the Neolithic way of life extending to temperate islands (Sampson 2005; Vigne et al. 2011) strongly supports Europe. Now it is evident that this process, at least in the the interpretation of Perle`s. In this discussion, the recent ex- western parts of Turkey and in the adjacent areas of the Bal- cavations at Ege Gu¨bre in I˙zmir, located on the coastal plain kans, was not an instantaneous happening. There are at least adjacent to the Aegean, presents an interesting case. The ear- two definable stages bringing with them different assemblages liest Neolithic occupation, with dates around 6250 cal BC, in and following different trajectories. What is being normally addition to typical central Anatolian elements has round stone referred to in the literature as the Early Neolithic occupation buildings and quantities of impresso pottery that are alien to of Europe is related to the second stage. Consideration of inland Anatolian assemblages. Likewise, the basal layers of these two distinct movements as a single process has been Hoca C¸es¸me, dated to 7637 BP (Bln 4639; i.e., roughly to the source of considerable bias in assessments of the beginning 6400 cal BC), is also located on the coastal plain by the Ae- of the Neolithic in Europe. Here I consider them under dif- gean. This layer contains both monochrome pottery almost ferent headings. identical to that from central Anatolia and round architecture As noted above, the evidence for the initial expansion of not found in central Turkey. A naviform core and related the Neolithic communities, whether by land or by sea, seems blades were recovered in 2007 in a surface survey conducted to be rather sparse and sporadic. This first stage appears to along the northern coast of the Sea of Marmara (Aydıngu¨n have continued up to the second half of the seventh millen- 2009). The total absence of these cores—so common in the nium, around 6500–6400 BC, just to the time of C¸ atalho¨yu¨k Levant—anywhere in the Anatolian plateau also seems to sup- layer 6. port long-distance maritime connections in the Neolithic. Immediately afterward, there are the first indications of a Discussion of what parts of the Neolithic package came more substantial expansion indicated by the numerous sites west by which route is a rather tricky issue, full of traps, in the lakes district and in the inner parts of western Anatolia. particularly if the discussion is based on single items that are What triggered this movement is beyond the scope of this uncommon in an assemblage. Accordingly, here again the article, although some sort of a social turbulence in the main general picture should be more helpful in developing insight core area (O¨ zdog˘an 1997, 2005) coupled with the unstable even though there will always be some contradictions because environmental conditions seemingly related to the 8.2 kyr BP of the presence or absence of certain diagnostic elements. Here climatic event (Alley and A´ gu´stsdo´ttir 2005; Berger and Gui- at least it is certain that the sea was intensively used since the laine 2008; Roberts and Rosen 2009) seem to be the main earliest stages of Neolithization, as evidenced by the presence agencies in this movement. In parts of Anatolia in close prox- of early sites in Cyprus and now also in Crete by the inflow imity to the central Anatolian basin, this era is much better of central Anatolian obsidian (S¸evketog˘lu 2008) and the early documented than in the coastal regions of the Aegean, al- distribution of Melos obsidian. Accordingly, it is no surprise though current work in the earlier levels of the sites such as that there was considerable interaction between the eastern C¸ukuric¸i Ho¨yu¨k, Ulucak, and Yes¸ilova are now yielding ample Mediterranean and the Aegean along a coastal maritime route evidence of the presence of this second stage in the littoral (van Andel 2005). In this respect, round-plan buildings, as areas of the Aegean. best known from Cypriot sites, and the so-called impresso This horizon is best represented by monochrome pottery pottery stand as clear indicators of a maritime package, and with well-finished and burnished surfaces very much akin to both are conspicuously lacking at inland sites. Thus, it seems the so-called dark-faced burnished wares of the Near East. possible to surmise that both the land route over the Anatolian Similarly, the vessel forms range from variations of hole- plateau and the maritime route following the Anatolian coasts mouth jars to semiglobular cups, some having ledge or cres- were operating simultaneously and that coastal sites along the cent lugs, with small knobs not being uncommon. Decoration Aegean incorporated a mixture of both assemblages. The geo- is rather rare; very seldom are there vessels with red bands, graphic location of some sites such as Kec¸ic¸ayırı or C¸ alca on incised patterns, or relief bands. Even though most of the high plateaus suggests that some early Neolithic groups were pottery is dark brown, reddish brown, or black, occasionally following the plateaus or mountains and not the alluvial plains there are also pieces with a fine burnished whitish or light as conventionally considered. At our current level of knowl- creamy slip. The Anatolian origin of this pottery is indis- edge, it is not possible to define the trajectories of endemic putable. It is possible to find analogous material from a large movements with any precision; however, patterning the dis- geographic zone extending from southeastern Anatolia to the tribution of Neolithic elements indicates that there was more Central Plateau. The assemblage, besides pottery, is rather than one movement, each following a different route. limited; the chipped stone industry is characterized by the extensive presence of pressure flaking, microbladelets, pris- The First Congregate Movement of matic and/or bullet cores, round scrapers, backed bladelets, Neolithic Communities and chipped discs. Bone polishers, spoons, and hooks are among the few definable prestige objects. Architectural details Following what I have described as the initial stage of ex- as well as the layout and the organization of the settlements S422 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 2. Some components of the Neolithic package. a, b, Bone spoons, As¸ag˘ı Pınar; c, pintaderra, As¸ag˘ı Pınar; d, e, applied bull motifs on red- slipped ware, Hoca C¸es¸me; f, M-shaped figurine, Hoca C¸es¸me; g, h, applied bull motifs on red-slipped ware, Tepecik-C¸ iftlik. also reflect what is known from farther east in Anatolia. Sub- terrazzo technique (Hauptmann and Yalc¸ın 2000) of the east, sistence, unlike most early Neolithic areas in central and east- occur in some sites, mainly at Bademag˘acı and Hoca C¸es¸me, ern Anatolia, was mainly dependant on farming, with little seemingly as “social memory.” On the other hand, during this evidence of hunting. Sheep and goat were more common then stage some specific tools that require special craftsmanship, cattle. such as flint arrow points, disappear in both regions, sug- Comparing the assemblages from both regions, it is evident gesting that the skilled craftsman had also joined moving that every component found in the newly settled areas has a groups (O¨ zdog˘an 2002). All of a sudden, with the disap- counterpart in the primary zone of Neolithization, though in pearance of arrow points, clay sling missiles become a major varying degrees of intensity. However, there is much that re- component of the assemblage in both regions. mained in the core area that did not move along with the This stage is best documented in northwestern Turkey, expansion, indicating some sort of a selection. Resorting the mainly along the southern and eastern parts of the Marmara assemblage indicates that only certain categories in the orig- region, as the Archaic and Classic phases (O¨ zdog˘an 1997, inal package were carried by the farmers, and these were 2005) of the Fikirtepe culture. This cultural stage is known mainly the utilitarian components of the culture (O¨ zdog˘an from over 25 sites. Ilıpınar (Roodenberg, van As, and Alpas- 2010). The expansion model includes all variations of do- lan-Roodenberg 2008) and Aktopraklık (Karul 2007) have mesticated animals and cultivated plants—not only cereals securely solved the chronological position of the Fikirtepe but in particular legumes and lentils—groundstone artifacts culture, placing it between 6450 and 6100 cal BC. including celts and building techniques but without most of As the details of the Fikirtepe culture and its development the cult and prestige objects. This strongly implies that the have been extensively published (O¨ zdog˘an 1997; O¨ zdog˘an and groups that were on the move did not include the ruling elite Bas¸gelen 2007), these will not be repeated here. However, the or the clergy, and it was more as a segregated movement of differences between the inland and coastal sites of this culture simple farmers or herdsman (O¨ zdog˘an 2008). Going into are important for this discussion. As noted above, there is an detailed analyses of the sites, the picture is far more complex extensive presence of Mesolithic or Epipaleolithic sites in the than what I have presented here as a generalization; some coastal areas of eastern Marmara and along the Black Sea components such as painted lime floors, reminiscent of the littoral (Gatsov and O¨ zdog˘an 1994). On the other hand, far- O¨ zdog˘an Westward Expansion of Farming S423

Figure 3. Figurines. a,As¸ag˘ı Pınar layer 7 partially steatophic figurine; b, Hacılar, Anatolian-type steatophic figurine; c,As¸ag˘ı Pınar layer 7 Anatolian-type steatopygic figurine; d,As¸ag˘ı Pınar layer 4 flat-bodied figurine. ther inland on the Anatolian plateau, this stage is conspicu- such as Pendik and Fikirtepe have intramural burials beneath ously absent. There was always an argument that this might the floors of the huts. Even more interesting are the recently be the result of survey biases. However, the difference in the found cremation burials at Yenikapı in Istanbul (Kızıltan composition of Fikirtepe assemblages between the areas where 2010), in the Fikirtepe and Yarımburgaz 4 horizons, a custom a Mesolithic substratum is known and other areas where it completely alien to Early Neolithic Anatolia. Likewise, sub- is not found strongly implies that the survey results were more sistence at sites such as Ilıpınar is primarily dependent on or less correct. In both areas, most of the assemblage and in farming, with almost no evidence of hunting or fishing (Bui- particular the pottery—in ware, shape, decoration, and in tenhuis 1995). Fikirtepe, Pendik, and I˙stanbul Yenikapı dis- sequential development—is identical. On the other hand, play a mixed pattern of hunting, fishing, mollusk collecting, those sites located inland, such as Ilıpınar or Mentes¸e, have and farming (Boessneck and von der Driesch 1979). Accord- rectangular buildings arranged like the Anatolian settlements, ingly, it is possible to assume that while inland sites represent while those along the coastal areas have round or oval wattle- an endemic movement, those in the coastal areas merged with and-daub huts with semisunken floors. This latter form is indigenous groups without signs of hostility. best known from Fikirtepe, Pendik, I˙stanbul Yenikapı, and How far west in the Balkans these groups penetrated is Aktopraklık (O¨ zdog˘an and Bas¸gelen 2007). Another striking difficult to say at present. In the western parts of the Marmara difference between the two areas is seen in burial customs. region and in eastern Thrace, the presence of this stage is Inland sites such as Ilıpınar and Aktopraklık have extramural firmly attested at As¸ag˘ı Pınar in layer 8, well stratified below cemeteries (Alpaslan-Roodenberg 2008), while coastal sites a Karanovo I–related level. In our surveys in eastern Thrace, S424 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 4. Anthropomorphic vessels from eastern Thrace. a–c,As¸ag˘ı Pınar layer 3; d, Toptepe layer 5. this type of material has been recorded from Kaynarca (O¨ z- ticipate that ongoing excavations in the Aegean parts of Tur- dog˘an 1986) at Gelibolu and Ug˘urlu-Zeytinlik Mevkii on key will contribute to defining this stage. Go¨kc¸eada Island (Harmankaya and Erdog˘u 2003), indicating Subsistence patterns during this stage are of crucial im- its presence in northwestern Turkey. Farther west, in the Ae- portance not only for this local region but also for defining gean, at least some of the material from Agios Petros (Ev- the beginning of agriculture and domestication in Europe. stratiou 1985) and from mainland Greece, Pre- and Proto- Even though natural scientists have been participating in al- Sesklo stage, can be considered to be related, though with most every excavation conducted during the last 2 decades, notable differences. Farther into the Balkans, the picture is at present only lists of species—plants, animals, fishes, or much less clear, especially in light of the controversy regarding mollusks—and general overviews are published, with few spe- the so-called Proto-Starcˇevo and/or Ko¨ro¨s cultures. It is not cific details. What can be noted at this stage is that even the possible to present a brief comment within the spectrum of first migrant groups arrived with the full range of domesti- this article. cated farm animals, cultivated cereals, and legumes. The re- In Bulgaria, the presence of an monochrome horizon pre- cent evidence indicating that not only fully domesticated ceding the Karanovo I painted-pottery horizon has been much sheep and goat but also cattle and pigs arrived in the Marmara argued (Stefanova 1996; Todorova 2003). It seems very likely region presents a new problem regarding how these groups that the dark-burnished pottery at sites such as Koprivets and were able to move such long distances together with those Krainitsi represent the rapid but rather thin expansion of the animals through Anatolia. In this context, the recovery of prepainted-pottery stages all the way up to the Danube. The dairy product remnants in Fikirtepe pottery is challenging presence of similar dark-burnished pottery found stratified expectations (Evershed et al. 2008). The preliminary assess- below Karanovo I horizon at As¸ag˘ı Pınar supports this view. ment of Fikirtepe that documents a mode of subsistence seems This stage of Neolithic expansion, which with some res- rather similar to what has been described as a “low-level food ervation we have been calling the “monochrome phase,” is production” model (Smith 2001, 2011). In this respect, the of critical importance even if it is not as notable as the next extensive presence of various fruits and nuts (mostly almonds phase because it sets the foundation of the Neolithization and seeds of wild pear) in many of the sites is worth consid- process in Europe. With the exception of northwestern parts ering, particularly in light of the evidence for planted trees at of Turkey, this phase is still poorly understood. We can an- Yenikapı, suggesting that tree cultivation might have been O¨ zdog˘an Westward Expansion of Farming S425

Figure 5. Cult vessels. a,As¸ag˘ı Pınar layer 6; b,As¸ag˘ı Pınar layer 3; c, Fikirtepe. more important and had an earlier beginning then previously occupied have no trace of earlier occupation; even in previ- assumed (see also Weiss and Zohary 2011). ously settled regions, with some exceptions, the site location preferences of the newcomers were different. This new wave The Second and More Rapid Massive is at best characterized by red-slipped and burnished pottery displaying mostly S-curved profiles, tubular lugs, plastic dec- Expansion of the Neolithic Way of Life oration in relief, anthropomorphic or zoomorphic vessels, It is not possible to comment on the intensity of the initial steatopygic figurines, pintaderas, and so forth (figs. 3–5). Re- wave of the Neolithic that I described above. It seems to have cent excavations at sites in the eastern parts of central Ana- been mainly oriented toward the eastern parts of the Marmara tolia, such as Tepecik-C¸iftlik and Ko¨s¸k Ho¨yu¨k (O¨ zdog˘an and region, as indicated by the numbers of settlements recorded Bas¸gelen 2007), strongly imply that this assemblage must orig- in that area. Farther to the west, the evidence is more sparse inate in that region. and difficult to trace. The second impulse, on the contrary, Even if this wave was much more massive than the earlier is easier to observe because it must have been a more massive one, it is also evident that there was a preference for alluvial and a rapid movement. Questions such as exactly when it valleys and large and well-watered intermountain plains as began or whether it was an uninterrupted continuum of the well as for avoiding high plateaus. The lake district and the earlier wave are difficult to answer. But at the end of the alluvial valleys extending from central Anatolia to the Aegean seventh millennium BC, numerous Neolithic settlements, big coastline are among the most intensively settled regions, where and small, appeared almost instantly over almost all of western recent surveys have revealed over a hundred sites of this ho- Turkey and in most of the Balkan Peninsula, sharing more rizon (O¨ zsait 1991). On the other hand, the eastern sections or less similar assemblages (fig. 2). Most of the areas now of the Marmara region, most densely inhabited by the pre- S426 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 6. Black-burnished wares with spiral decoration. a,As¸ag˘ı Pınar layer 7; b, Yenikapı Neolithic burial; c, sherds from Gelveri in Central Anatolia. vious group, seems to have remained beyond the reach of this viously occupied by the Fikirtepe culture remained outside new wave. of this wave. There, all over the eastern Marmara and around Excavations at sites such as Kuruc¸ay,Ho¨yu¨cek, Bademag˘acı, the Bosporus, Yarımburgaz 4 culture, possibly developing Ulucak, Yes¸ilova, and Hoca C¸es¸me have revealed information from the Classical Fikirtepe, became established. Yarımburgaz on the sequential development of this stage, indicating that 4 culture (O¨ zdog˘an, Miyake, and O¨ zbas¸aran-Dede 1991) still painted decoration was extremely rare or even absent at the produced pottery in dark colors, although there are some red- very beginning. As the issue of painted decoration surfaces slipped wares, but the decoration is either deeply incised or in most of the discussions in the Balkans (Nikolov 1987, 2002, excised, or occasionally it is executed in the so-called fur- 2003; Schubert 1999), it is worth presenting some details. In chenstich technique. Yarımburgaz culture seems to continue the older layers of all the sites mentioned above, the pottery developing locally into what is known as Yarımburgaz 3, dis- assemblage is dominated by very fine lustrously burnished red playing very fine black-burnished pottery with incised spiral or jet black vessels. When painted decoration first appears, it and notenkopf-like motifs. Major sites of this culture, besides is extremely rare and displays thick bands in red that are Yarımburgaz, include Ilıpınar layer 8, Aktopraklık, and I˙stan- mostly made not by paint but by using different colored slips. bul Yenikapı (fig. 6). The pottery of Yarımburgaz culture, This earlier stage, being rather common along the Aegean along with some of its other elements, is highly reminiscent coast of Turkey, is either absent in Bulgaria or it has not been of the early Linear cultures in the Danubian area (O¨ zdog˘an reported in the publications. The final stage of painted dec- 1989). Whether these similarities are accidental or imply some oration with a vast range of geometric designs is found all genetic relationship remains a controversy. During the rescue over the Balkans under different cultural denominations such excavations carried out in I˙stanbul Yenikapı in 2008, a number as Karanovo I–II, Kremikovci, Gradesnitsa, Starcevo, and so of burials of Yarımburgaz 4 culture were recovered. The pres- forth. Nevertheless, differences among these groups are not ence of a vessel with deeply incised spiral designs in the graves as significant as suggested in the literature. conspicuously points to connections with the Danubian area. During the rapid expansion of red-slipped pottery—with As no pottery of Yarımburgaz type has been found either in or without painting—in the eastern Marmara, regions pre- eastern Thrace or in Bulgaria, any connection with the north- O¨ zdog˘an Westward Expansion of Farming S427 ern Balkans must have been along the Black Sea coastline (Biagi, Shennan, and Spataro 2005). Thus, I confined extrap- through the Danube River valley. olations to a general level. The border between the red-slipped and painted cultural After an overview of previous discussions on the nature of groups and the Yarımburgaz distribution is more or less clear. this expansion, it now seems clear that all arguments, no In eastern Thrace, As¸ag˘ı Pınar is the easternmost excavated matter how contradictory, were correct. There was endemic site of this culture, and the surveys have revealed some other movement, migration, and colonization by both land and sea; sites 50 km farther to the east. Because the contemporaneity there were “frontiers” merging with local communities, ex- of As¸ag˘ı Pınar and Yarımburgaz is firmly attested by 14C dates, pansion by exchange of knowledge and/or commodities, and the border must lie immediately to the west of Istanbul, im- to a degree, local development. Of the latter, transformation plying that the second wave entered Thrace not through the of the mud-brick architecture of the Anatolian plateau to Bosporus but from the western parts of the Sea of Marmara, wooden posts or wattle-and-daub structures in the wooded probably following the coastal plains along the Aegean. areas of Marmara should be considered as a fine example of adaptation to local environment. Some of the mobile groups were strictly farmers and others had a mixed economy with Concluding Remarks intensive hunting, but there were others who were highly dependent on marine resources. What is clear is the fact that Recent archaeological evidence from the western parts of Tur- it was a dynamic era, and there was motivation to move or key strongly implies that the arrival of a Neolithic way of life to migrate that was not common in other periods. What is was not a single instantaneous event and on the contrary was also clear at this stage is the sustained relationship between a multifarious process that lasted for more than 1,000 years. the newly settled areas and the original homeland. It is now possible to speak of separate waves of Neolithiza- I find this continuing connection to be extremely significant tion—each with its own mode, pace, and trajectory—yet some in understanding the mode of Neolithic expansion. If it had may have taken place simultaneously. With this article I tried taken place as a rapid and organized movement, then sooner to present a conspectus of the recent evidence from the west- or later the developments in the new areas would have been ern and northwestern parts of Turkey and its implications for different from those of the homeland. On the contrary, from our understanding of the initial stages of Neolithization in the earliest stage of the Pottery Neolithic up to the so-called southeastern Europe, including the Aegean. One other point Vincˇa period in the Balkans, there is an apparent parallelism needs to be emphasized. In the earlier stages of the Neolithic in the primary cultural traits between Anatolia and the Bal- expansion, the western parts of Turkey were the frontier, de- kans that is defined by Garas¸anin as the “Balkano-Anatolian veloping as a periphery. However, as the Neolithic culture culture complex” (Garas¸anin 2000). This implies that moving expanded its coverage to southeastern Europe, those parts groups somehow sustained contact to keep track of what was became the core or the primary zone of the European Neo- happening in their original homeland, a pattern described as lithic. For that reason, understanding both historical roles is “chain migration” by Anthony (1997:24). essential to understanding what and how Neolithization de- As I have noted, the evidence is still very fragmentary, and veloped in Europe. the data are very recent and need time to develop. Here I I could have written this article using same evidence in have offered some generalizations, and they should be un- another way by defining the distinct compositions of different derstood on that level. Every excavated site has various pieces Neolithic packages. Then the picture would have been slightly of evidence that do not fit with the general statements in this different. As an example, there are some Neolithic packages article. Perhaps the Neolithic era can best be described as a that do not include the “way of village life.” Only commodities complex mosaic with a strong social motivation to migrate. seem to have been transferred—most probably by some sort of an exchange. There are others where all sorts of status or prestige objects are either absent or present in minimal Acknowledgments amounts, possibly indicating that these groups represent sim- ple farmers on the move. It all leads to the conclusion that My work has been going on in northwestern Turkey since the latter half of the eighth millennium and lasting up to the 1980 with the support from Istanbul University research sixth millennium witnessed a very dynamic succession of funds, the Department of Antiquities, and the Institute for events that cannot be explained by a single model. Through- Aegean Prehistory. My collaboration with Prof. Hermann Par- out the article, I intentionally avoided going into the discus- zinger in 1993 has also provided the support of the German sion of absolute chronology; as noted above, with the available Archaeological Institute, Berlin. dates regularly published by the Central Anatolian Neolithic E-Workshop,1 especially from the Balkans, working out a pre- References Cited cise chronology that is not problematic is still not possible Abay, E. 2005. Neolithic settlement at Ulucak Ho¨yu¨k and its cultural relations with neighbour regions in western Anatolia. In How did 1. http://www.canew.org. farming reach Europe? Anatolian-European relations from the second S428 Current Anthropology Volume 52, Supplement 4, October 2011

half of the 7th through the first half of the 6th millennium cal BC. D. Urem-Kotsu, K. Kotsakis, et al. 2008. Earliest date for milk use C. Lichter, ed. Pp. 75–84. Byzas 2. Istanbul: Ege. in the Near East and southeastern Europe linked to cattle herding. Alley, R. B., and A. M. A´ gu´stsdo´ttir. 2005. The 8k event: cause and Nature 7180:1–4. consequences of a major Holocene abrupt climate change. Qua- Evstratiou, N. 1985. Agios Petros: a Neolithic site in the northern ternary Science Reviews 24:1123–1149. Sporades. BAR International Series 241. Oxford: British Archae- Alpaslan-Roodenberg, S. 2008. The Neolithic cemetery: the anthro- ological Reports. pological view. In Life and death in a prehistoric settlement in north- French, D. 1986. Anatolia: bridge or barrier? Tu¨rk Tarih Kongresi west Anatolia. The Ilıpınar excavations, volume III: with contribu- (Kongreye Sunulan Bildiriler) 9(1):117–118. tions on Hacılartepe and Mentes¸e. J. Roodenberg and S. Garas¸anin, M. 1981. Consideration sur les rapports du sud-est eu- Alpaslan-Roodenberg, eds. Pp. 35–68. Leiden: Instituut voor het ropeen et de l’Anatolie aux e´poques neolithiques et eneolithiques. Nabıje Oosten. Archaeologia Iugoslavica 20–21:7–12. Anthony, D. 1997. Prehistoric migration as social process. In Migra- ———. 1997. Der spa¨te Balkanisch-Anatolische Komplex: ein Ru¨ck- tions and invasions in archaeological explanation. J. Chapman and blick nach vier Jahrzehnten. Starinar 48:15–31. H. Hamerow, eds. Pp. 21–32. BAR International Series 664. Ox- ———. 2000. Zum Begriff des Balkanisch-Anatolischen Komplexes ford: British Archaeological Reports. des spa¨ten Neolithikums. In Karanovo: Beitra¨ge zum Neolithikum Anthony, D. W., and J. Y. Chi, eds. 2010. The lost world of old Europe: in Su¨dosteuropa. S. Hiller and V. Nikolov, eds. Pp. 343–347. Vienna: the Danube Valley, 5000–3500 BC. New York: New York University Phoibos. Institute for the Study of the Ancient World. Garas¸anin, M., and I. Radovanovic¸. 2001. A pot in house 54 at Aydıngu¨n, S¸. 2009. Early Neolithic discoveries at Istanbul. Antiquity Lepenski Vir I. Antiquity 75(287):118–125. 83(320): project gallery. http://www.antiquity.ac.uk/projgall/ Gatsov, I., and M. O¨ zdog˘an. 1994. Some Epi-Paleolithic sites from Aydingun310. Accessed June 2009. NW Turkey: Ag˘ac¸lı, Domalı and Gu¨mu¨s¸dere. Anatolica 20:97–120. Berger, J.-F., and J. Guilaine. 2008. The 8200 cal BP abrupt envi- Gu¨nel, S. 2003. Batı Anadolu Bo¨lgesi Tariho¨ncesi Do¨nemlerine Yeni ronmental change and the Neolithic transition: a Mediterranean Katkılar. Belleten 67:719–738. perspective. Quaternary International 200(1–2):31–49. Harmankaya, S., and B. Erdog˘u. 2003. Prehistoric sites of Go¨kc¸eada, Bertram, J.-K., and N. Karul. 2005. From Anatolia to Europe: the Turkey. In From village to cities: early villages in the Near East.M. ceramic sequence of Hoca C¸es¸me in Turkish Thrace. In How did O¨ zdog˘an, H. Hauptmann, and N. Bas¸gelen, eds. Pp. 459–479. farming reach Europe? Anatolian-European relations from the second Istanbul, Turkey: Arkeoloji ve Sanat Yayınları. half of the 7th through the first half of the 6th millennium cal BC. Harris, D. R. 2003. Paradigms and transitions: reflections on the C. Lichter, ed. Pp. 117–129. Byzas 2. Istanbul: Ege. study of the origins and spread of agriculture. In The widening Biagi, P., S. Shennan, and M. Spataro. 2005. Rapid rivers and slow harvest: the Neolithic transition in Europe: looking back, looking seas? new data for the radiocarbon chronology of the Balkan Pen- forward. A. Ammerman and P. Biagi, eds. Pp. 43–58. Boston: Ar- insula. In Prehistoric archaeology and anthropological theory and chaeological Institute of America. education. L. Nikolova and J. Higgins, eds. Pp. 41–52. Reports of Hauptmann, A., and U¨ . Yalc¸ın. 2000. Lime plaster, cement and the Prehistoric Research Projects 6–7. Sofia: Agatho. first puzzolanic reaction. Pale´orient 26:61–68. Boessneck, J., and A. von den Driesch. 1979. Die Tierknochenfunde Horejs, B. 2008. Metalworkers at the C¸ukuric¸i Ho¨yu¨k? an Early aus der Neolithischen Siedlung auf dem Fikirtepe bei Kadıko¨y am Bronze Age mould and a “Near Eastern weight” from western Marmarameer. Munich: Institut fu¨r Domestikationsforschung. Anatolia. In Metals and societies: studies in honour of Barbara S. Bogucki, P. 1996. The spread of early farming in Europe. American Ottaway. T. L. Kienlin and B. Roberts, eds. Bonn: Universita¨ts- Scientist 84:242–253. forschungen zur pra¨historischen Archa¨ologie. Buitenhuis, H. 1995. The faunal remains. In The Ilıpınar excavations Karul, N. 2007. Aktopraklık. In Anadolu’da Uygarlıg˘ın Dog˘us¸u ve I: five seasons of fieldwork in NW Anatolia, 1987–91. J. Roodenberg, Avrupa’ya Yayılımı: Tu¨rkiye’de Neolitik Do¨nem: Yeni Kazılar, Yeni ed. Pp. 151–156. Istanbul: Nederlands Historisch-Archaeologisch Bulgular, 2 vols. M. O¨ zdog˘an and N. Bas¸gelen, eds. Pp. 387–392. Instituut. Istanbul: Arkeoloji ve Sanat Yayınları. C¸evik, O¨ . 2010. Ulucak Ho¨yu¨k 2009 Yılı Kazı Sonuc¸ları. Paper read Karul, N., Z. Eres, M. O¨ zdog˘an, and H. Parzinger, eds. 2003. As¸ag˘ı at 32nd International Symposium of Excavations, Surveys and Pınar I: Einfu¨hrung, Forschungsgeschichte, Statigraphie und Archi- Archaeometry, Denizli, May. tektur. Mainz am Rhein: von Zabern. Childe, G. 1964 (1957). The dawn of European civilization. New York: Kızıltan, Z., ed. 2007. Gu¨n Is¸ıg˘ında˙ Istanbul’un 8000 Yılı: Marmaray, Vintage. Metro ve Sultanahmet Kazıları. Istanbul: Vehbi Koc¸Vakfı-I˙stanbul C¸ilingirog˘lu, A., and E. Abay. 2005. Ulucak Ho¨yu¨k excavations: new Arkeoloji Mu¨zeleri. results. Mediterranean Archaeology and Archaeometry 5(3):5–21. ———. 2010. Marmaray-Metro Projeleri Kapsamında Yapılan, Yeni- Duru, R. 1989. Were the earliest cultures at Hacılar really Aceramic? kapı, Sirkeci ve U¨ sku¨dar Kazıları. In ˙Istanbul Arkeoloji Mu¨zeleri 1: In Anatolia and the ancient Near East: studies in honor of Tahsin Marmaray-Metro Kurtarma Kazıları Sempozyumu Bildiriler Kitabı. O¨ zgu¨c¸. K. Emre et al., eds. Pp. 99–104. Ankara: Tu¨rk Tarih Kurumu U. Kocabas¸, ed. Pp. 1–16. Istanbul: Istanbul Arkeoloji Mu¨zeleri. Basımevi. Lloyd, S. 1956. Early Anatolia. London: Penguin. ———. 2008. MO¨ 8000’den MO¨ 2000’e Burdur-Antalya Bo¨lgesi’nin Mellaart, J. 1960. Anatolia and the Balkans. Antiquity 34:270–278. Altıbin Yılı. Antalya, Turkey: Suna-I˙nan Kırac¸ Akdeniz Medeni- ———. 1970. Excavations at Hacılar I. Edinburgh: Edinburgh Uni- yetleri Aras¸tırma Enstitu¨su¨. versity Press. Efe, T. 1997. 1995 Yılında Ku¨tahya, Bilecik ve Eskis¸ehir I˙llerinde ———. 1975. The Neolithic of the Near East. London: Thames & Yapılan Yu¨zey Aras¸tırması. Aras¸tırma Sonuc¸ları Toplantısı 14(2): Hudson. 215–232. ———. 1989. Neolithic chronology at C¸atal Hu¨yu¨k?In Anatolia and ———. 2005. The Neolithisation in inland northwestern Anatolia. ancient Near East: studies in honor of Tahsin O¨ zgu¨c¸.K.Emreetal., In How did farming reach Europe? Anatolian-European relations eds. Pp. 315–318. Ankara: Tu¨rk Tarih Kurumu Basımevi. from the second half of the 7th through the first half of the 6th Milojcˇic´, V. 1960. Pra¨keramisches Neolithikum auf der Balkanhal- millennium cal BC. C. Lichter, ed. Pp. 107–115. Byzas 2. Istanbul: binsel. Germania 38:320–340. Ege. Nikolov, V. 1987. Beitra¨ge zu den Beziehungen zwischen Vorderasien Evershed, R. P., S. Payne, A. G. Sherratt, M. S. Copley, J. Coolidge, und Su¨dosteuropa aufgrund der fru¨hneolithischen bemalten Ker- O¨ zdog˘an Westward Expansion of Farming S429

amik aus dem Zentralbalkan. Acta Praehistorica et Archaeologica O¨ zdog˘an, M., and I. Gatsov. 1998. The Aceramic Neolithic Period 19:7–18. in western Turkey and in the Aegean. Anatolica 24:209–232. ———. 2002. Nochmals u¨ber die Kontakte zwischen Anatolien und O¨ zdog˘an, M., Y. Miyake, and N. O¨ zbas¸aran-Dede. 1991. An interim dem Balkan im 6. Jt. v. Ch. In Mauerschau. Festschrift fu¨r Manfred report on the excavations at Yarımburgaz and Toptepe in eastern Korfmann. R. Aslan et al., eds. Pp. 673–678. Remshalden-Grun- Thrace. Anatolica 17:59–121. bach: Greiner. O¨ zsait, M. 1991. Nouveaux sites contemporains de Hacılar en Pisidie ———. 2003. The Neolithic and Chalcolithic periods in northern Occidentale. De Anatolia Antiqua 1:59–118. Thrace. TU¨ BA-AR 6:21–83. Perle`s, C. 2001. The early Neolithic in Greece. Cambridge: Cambridge ———. 2004. Dynamics of the cultural processes in Neolithic Thrace. World Archaeology. In Prehistoric Thrace: proceedings of the International Symposium ———. 2003. An alternate (and old-fashioned) view of Neolithis- in Stara Zagora. V. Nikolov, K. Bacvarov, and P. Kalchev, eds. Pp. ation in Greece. Documenta Praehistorica 30:99–113. 18–25. Sofia, Bulgaria: Institute of Archaeology with Museum- ———. 2005. From the Near East to Greece: let’s reverse the focus: BAS, Regional Museum of History, Stara Zagora. cultural elements that didn’t transfer. In How did farming reach Nikolov, V., K. Bacvarov, and P. Kalchev, eds. 2004. Prehistoric Thrace: Europe? Anatolian-European relations from the second half of the proceedings of the International Symposium in Stara Zagora. Sofia, 7th through the first half of the 6th millennium cal BC. C. Lichter, Bulgaria: Institute of Archaeology with Museum-BAS, Regional ed. Pp. 275–290. Byzas 2. Istanbul: Ege. Museum of History, Stara Zagora. Peschlow-Bindokat, A. 2003. Fru¨he Menschenbilder: die pra¨histo- O¨ zdog˘an, M. 1986. Prehistoric sites in the Gelibolu Peninsula. Anado- rischen Felsmalereien des Latmos-Gebirges. Mainz: von Zabern. lu Aras¸tırmaları 10:51–66. Pinhasi, R. 2003. A new model for the spread of the first farmers in ———. 1989. Neolithic cultures of northwestern Turkey: a general Europe. Documenta Praehistorica 30:1–47. appraisal of the evidence and some considerations. In Neolithic of Reingruber, A. 2005. The Argissa Magoula and the beginning of the southeastern Europe and its Near Eastern connections.S.Bo¨ko¨nyi, Neolithic in Thessaly. In How did farming reach Europe? Anatolian- ed. Pp. 201–215. Budapest: Varia Archaeologica Hungarica II. European relations from the second half of the 7th through the first ———. 1995. Neolithization of Europe: a view from Anatolia. 1. half of the 6th millennium cal BC. C. Lichter, ed. Pp. 155–171. The problem and the evidence of East Anatolia. Procilo 20:25–61. Byzas 2. Istanbul: Ege. ———. 1996. Pre-Bronze Age sequence of central Anatolia: an al- Renfrew, C. 2002. The emerging synthesis: the archaeogenetics of ternative approach. In Vom Halys zum Euphrat: Thomas Beran zu farming/language dispersals and other spread zones. In Examining Ehren. U. Magen and M. Rashad, eds. Pp. 185–202. Mu¨nster: the farming/language dispersal hypothesis. P. Bellwood and C. Ren- Ugarit. frew, eds. Pp. 3–16. Cambridge: McDonald Institute for Archae- ———. 1997. The beginning of Neolithic economies in southeastern ological Research. Europe: an Anatolian perspective. Journal of European Archaeology Roberts, N., and A. Rosen. 2009. Diversity and complexity in early 5(2):1–33. farming communities of southwest Asia: new insights into the ———. 2002. On arrows and sling missiles: what happened to the economic and environmental basis of Neolithic C¸atalho¨yu¨k. Cur- arrows? In Mauer Schau: Festschrift fu¨r Manfred Korfmann I.R. rent Anthropology 50(3):393–394. Aslan et al., eds. Pp. 437–444. Stuttgart: Greiner. Roodenberg, J., and S. Alpaslan-Roodenberg, eds. 2008. Life and death ———. 2004. Heritage and nationalism in the Balkans and Anatolia in a prehistoric settlement in northwest Anatolia: the Ilıpınar ex- or changing patterns: what has happened since Hasluck? In Ar- cavations, volume III, with contributions on Hacılartepe and Mentes¸e. chaeology, anthropology and heritage in the Balkans and Anatolia. Leiden: Nederlands Instituut voor het Nabıje Oosten. D. Shankland, ed. Pp. 389–405. Istanbul: ISI˙S. Roodenberg, J., A. van As, and S. Alpaslan-Roodenberg. 2008. Barcın ———. 2005. The expansion of the Neolithic way of life: what we Hu¨yu¨k in the plain in Yenis¸ehir (2005–2006): a preliminary note know and what we do not know. In How did farming reach Europe? on the fieldwork, pottery and human remains of the prehistoric Anatolian-European relations from the second half of the 7th through levels. Anatolica 34:53–66. the first half of the 6th millennium cal BC. C. Lichter, ed. Pp. 13– Rowley-Conwy, Peter. 2011. Westward ho! the spread of agriculture 27. Byzas 2. Istanbul: Ege. from central Europe to the Atlantic. Current Anthropology ———. 2007a. Amidst Mesopotamia-centric and Euro-centric ap- 52(Suppl. 4):S431–S451. proaches: the changing role of the Anatolian Peninsula between Runnels, C. 2001. The Stone Age of Greece from the Paleolithic to the East and the West. Anatolian Studies 57:17–24. the advent of the Neolithic: addendum 1995–1999. In Aegean pre- ———. 2007b. Von Zentralanatolien nach Europa: die Ausbreitung history: a review. T. Cullen, ed. Pp. 225–254. Boston: Archaeological der neolithischen Lebensweise. In Vor 12.000 Jahren in Anatolien: Institute of America. die a¨ltesten Monumente der Menschheit. C. Lichter, ed. Pp. 150– ———. 2003. The origins of the Greek Neolithic: a personal view. 160. Karlsruhe: Badisches Landesmuseum. In The widening harvest: the Neolithic transition in Europe: looking ———. 2008. An alternative approach in tracing changes in de- back, looking forward. A. Ammerman and P. Biagi, eds. Pp. 121– mographic composition: the westward expansion of the Neolithic 132. Boston: Archaeological Institute of America. way of life. In The Neolithic demographic transition and its con- Sampson, A. 2005. New evidence from the early productive stages sequences. J. Bocquet-Appel and O. Bar-Yosef, eds. Pp. 139–178. in the Aegean Basin from the 9th to the 7th millennium cal BC. Heidelberg: Springer. In How did farming reach Europe? Anatolian-European relations ———. 2010. Westward expansion of the Neolithic way of life: sort- from the second half of the 7th through the first half of the 6th ing the Neolithic package into distinct packages. In Near Eastern millennium cal BC. C. Lichter, ed. Pp. 131–141. Byzas 2. Istanbul: archaeology in the past, present and future: heritage and identity, Ege. vol. 1 of Proceedings of the 6th International Congress of the Ar- Schachermeyr, F. 1976. Die A¨ga¨ische Fru¨hzeit: Forschungsbericht u¨ber chaeology of the Ancient Near East, 5 May–10 May 2009, “Sapienza,” die Ausgrabungen im letzten Jahrzehnt und u¨ber ihre Ergebnisse fu¨r Universita`di Roma. P.Matthiae et al., eds. Pp. 883–897. Wiesbaden, unser Geschichtsbild, vol. 1. Vienna: O¨ sterreichischen Akademie der Germany: Harrassowitz. Wissenschaften. O¨ zdog˘an, M., and N. Bas¸gelen, eds. 2007. Anadolu’da Uygarlıg˘ın Schubert, H. 1999. Die bemalte Keramiks des Fru¨hneolithikums in Dog˘us¸u ve Avrupa’ya Yayılımı. Tu¨rkiye’de Neolitik Do¨nem: Yeni Ka- Su¨dosteuropa, Italien und Westanatolien. Rahden/Westphalia: Lei- zılar, Yeni Bulgular. Istanbul: Arkeoloji ve Sanat Yayınları. dorf. S430 Current Anthropology Volume 52, Supplement 4, October 2011

Seeher, J. 1990. Cos¸kuntepe-Anatolisches Neolithikum am Nordo- in Bulgarian prehistory. In Prehistoric Bulgaria.D.W.Baileyand stufer der A¨ ga¨is. Istanbuler Mitteilungen 40:9–15. I. Panajotov, eds. Pp. 79–98. Monographs in World Archaeology S¸evketog˘lu, M. 2008. Early settlements and precurement of raw ma- 22. Madison, WI: Prehistory. terials: new evidence based on research at Akanthou-Arkosykos ———. 2003. Neue Angaben zur Neolithisierung der Balkanhalbin- (Tatlısu-C¸iflikdu¨zu¨), northern Cyprus. TU¨ BA-AR 11:63–72. sel. In Morgenrot der Kulturen: fru¨he Etappen der Menschheitsge- Sherratt, A. 2004. Fractal farmers: patterns of Neolithic origin and schichte in Mittel-und Su¨dosteuropa: festschrift fu¨r Nandor Kalicz. dispersal. In Explaining social change: studies in honor of Colin E. Jerem and P. Raczky, eds. Pp. 83–88. Budapest: Archaeolingua. Renfrew. J. Cherry, C. Scarre, and S. Shennan, eds. Pp. 53–63. Tringham, R. 1971. Hunters, fishers and farmers of eastern Europe. McDonald Institute Monographs. Cambridge: Cambridge Uni- London: Hutchinson. versity Press. ———. 2000. Southeastern Europe in the transition to agriculture Smith, B. 2001. Low-level food production. Journal of Archaeological in Europe: bridge, buffer, or mosaic. In Europe’s first farmers.T. Research 9(1):1–43. D. Price, ed. Pp. 19–56. Cambridge: Cambridge University Press. Smith, Bruce D. 2011. The cultural context of plant domestication van Andel, T. 2005. Coastal migrants in a changing world? an essay in eastern North America. Current Anthropology 52(Suppl. 4): on the Mesolithic in the eastern Mediterranean. Journal of the S471–S484. Israel Prehistoric Society 35:381–396. Stefanova, T. 1996. A comparative analysis of pottery from the “monochrome Early Neolithic Horizon” and “Karanovo I Hori- Vigne, Jean-Denis, Isabelle Carre`re, Franc¸ois Briois, and Jean Gui- zon” and the problems of the Neolithization of Bulgaria. Proc¸ilo laine. 2011. The early process of mammal domestication in the o Razjskovanju Paleolita, Neolita in eneolita v Sloveniji 23:15–38. Near East: new evidence from the Pre-Neolithic and Pre-Pottery Takaog˘lu, T. 2005. Cos¸kuntepe: a Neolithic quern production site in Neolithic in Cyprus. Current Anthropology 52(Suppl. 4):S255– NW Turkey. Journal of Field Archaeology 30:419–433. S271. Tas¸kıran, H. 2008. Suluin excavation: 2007. Paper presented at the Weiss, Ehud, and Daniel Zohary. 2011. The Neolithic southwest Asian 30th International Symposium of Excavations, Surveys and Ar- founder crops: their biology and archaeobotany. Current Anthro- chaeometry, May, Ankara. pology 52(Suppl. 4):S237–S254. Theocharis, D. R. 1973. Neolithic Greece. Athena: National Bank of Zvelebil, M., and M. Lillie. 2000. Transition to agriculture in eastern Greece. Europe. In Europe’s first farmers. T. D. Price, ed. Pp. 57–92. Cam- Todorova, H. 1995. The Neolithic, Eneolithic and transitional period bridge: Cambridge University Press. Current Anthropology Volume 52, Supplement 4, October 2011 S431

Westward Ho! The Spread of Agriculture from Central Europe to the Atlantic

by Peter Rowley-Conwy

Recent work on the four major areas of the spread of agriculture in Neolithic western Europe has revealed that they are both chronologically and economically much more abrupt than has hitherto been envisaged. Most claims of a little agriculture in Late Mesolithic communities are shown to be incorrect. In most places, full sedentary agriculture was introduced very rapidly at the start of the Neolithic. “Transitional” economies are virtually absent. Consequently, the long-term processes of internal development from forager to farmer, so often discussed in Mesolithic-Neolithic Europe, are increasingly hard to sustain. The spread of agriculture by immigration is thus an increasingly viable explanation. The crucial role of boats for transport and of dairying for the survival of new farming settlements are both highlighted. Farming migrations were punctuated and sporadic, not a single wave of advance. Consequently, there was much genetic mixing as farming spread, so that agricultural immigrants into any region carried a majority of native European Mesolithic genes, not Near Eastern ones.

been “sharpened up,” rendering them chronologically more Westward ho! abrupt. They thus appear more as archaeological events than (This was the Thames watermen’s cry, indicating direc- as processes. tion of travel to prospective passengers. By Shakespeare’s “Process” implies continuity within the local group. time, it was a more general expression of intent to travel “Event” raises the possibility of immigration. In this article I westward, often by boat. [Twelfth Night, act 3, scene 1, will argue that agriculture was introduced by immigration line 134]) more often than is currently believed: the four spread events Introduction: Event and Process considered here were all probably migrations. This position is remarkably similar to that adopted by O¨ zdog˘an (2011) in his consideration of Anatolia and the Balkans; at the Wenner- In this article I will consider the spread of agriculture from Gren conference from which these articles stem, “The Origins central Europe to the Atlantic (fig. 1). This involves four major of Agriculture: New Data, New Ideas,” it seemed almost that “spread events”: the Cardial of the western Mediterranean, they were two halves of the same article. the Linienbandkeramik (LBK) of the interior, the Trægtbæ- On a broader scale, the conference threw two things into gerkultur (TRB) of southern Scandinavia, and the Neolithic sharp relief for me. The first is the contrast between the of Britain and Ireland. lengthy and complex origins of the Near Eastern agricultural Archaeologists often regard the appearance of agriculture system—as discussed by Goring-Morris and Belfer-Cohen in any region as a slow process of transition: the Later Meso- (2011) and Zeder (2011)—and the rapid subsequent spread flithic may have had a few domestic animals or plants, while of that system across Europe. It seems that “origins” and the Neolithic may still have involved nomadic foraging. In “dispersals” are becoming ever more sharply differentiated as this article I will, however, argue that recent work has de- archaeological work proceeds. The second is the fact that the molished the basis for thinking that these were slow transi- origin and spread of the Chinese agricultural system shows tions. Instead, full sedentary agriculture appeared rapidly in some similarity to the Near Eastern system, the complex or- most places. The evidence for these four spread events has igins (Cohen 2011; Zhao 2011) contrasting with its spread through Korea (Lee 2011) and Japan (Crawford 2011). Similar Peter Rowley-Conwy is Professor in the Department of Archaeology, explanations may thus be emerging for the spread of agri- Durham University (South Road, Durham DH1 3LE, United culture at both ends of Eurasia: in temperate zones away from Kingdom [[email protected]]). This paper was the centers of agricultural origins. At a still wider level, the submitted 13 XI 09, accepted 1 XI 10, and electronically published conference served (for me, at least) to emphasize the differ- 13 V 11. ence between the mode and speed of the spread of the East

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0018$10.00. DOI: 10.1086/658368 Figure 1. Map showing the major farming “spread events” discussed in this article. Dates are in calibrated years BP. A color version of this figure is available in the online edition of Current Anthropology. Rowley-Conwy The Westward Spread of Agriculture S433 and West Asian cereal/pulse/domestic-animal farming systems domestic pigs of Near Eastern origin (Larson et al. 2007), and on the one hand and the Southeast Asian root/fruit system rye and oats were domesticated in the Iron Age. (Denham 2011) on the other: the former appears on current Indigenism, however, receives support from human ge- evidence to spread much faster than the latter. netics. Early work used modern human blood groups in Eu- rope to argue for large-scale immigration in the Neolithic, the famous “wave of advance” hypothesis (Ammerman and The Debate: Immigrants or Indigenes Cavalli-Sforza 1984). Recent work on mitochondrial DNA This debate has a long and complex history in Europe. Most (mtDNA) indicates the opposite, that most Europeans are scholars of course argue for a combination of the two but descended from pre-Neolithic hunter-gatherers. This is be- nevertheless tend to favor one over the other. In the middle cause mtDNA lineages began to diverge before the Neolithic. twentieth century, immigration was the standard view (Childe The most recent studies have involved the complete mito- 1957). In the wake of the radiocarbon revolution, British chondrial genome and indicate that this divergence started scholars began suggesting indigenous developments even ∼15,000 years ago (Soares et al. 2009, 2010). though radiocarbon dating had not altered the relative dating Indigenism has therefore largely become “adoptionism.” of early agriculture in Europe: in Britain it was still later than Local hunter-gatherers provided the bulk of the human genes in the near continent, and it remained earliest in the Balkans. in Neolithic and later cultures but acquired the elements of Indigenism, however, appeared an attractive new idea forming farming from neighbors who had already “gone agricultural.” part of a broader research agenda seeking indigenous origins for megalithic tombs and Bronze Age chiefdoms. British re- Migrationism searchers applied indigenist arguments to early agriculture in wide areas of Europe (Barker 1985; Dennell 1983), and recent Migrationism has received support from linguistics. However, postprocessualists have followed this pattern (Thomas 1999; Renfrew’s (1987) argument that agriculture was carried into Whittle 1996). Continental reactions have been mixed. South Europe by Indo-European speakers has met with considerable Scandinavians have mainly concurred (Jensen 1982) as more opposition. Mallory (1989:121, 126–127) pointed out that recently have many Iberians (Arias 1999); but most German words associated with wheels, carts, and traction are related archaeologists mistrust Anglophone overviews and espouse in most Indo-European languages, and because such tech- migrationism (Gronenborn 1999, 2007). nology did not exist before ca. 5500 BP, this suggested an Indo-European dispersal after that date. Anthony (2007) ar- gues that if Proto-Indo-European was spoken by Europe’s Indigenism earliest farmers around 8500 BP, the common terminology Wild einkorn is present in the Balkans; aurochs and wild boar for wagons would mean that it remained a single language were widespread in Europe; some have suggested that the wild spread across most of Europe for 3,000 years, diversifying species of sheep, barley, and lentils were present as well. Var- only after 5500 BP, a most unlikely scenario. A more recent ious arguments for local domestication have been put forward suggestion is that Basque, normally regarded as unrelated to (Barker 1985:252–253; Dennell 1983:159–163; Whittle 1996: any other language, may in fact have connections with other 67; Zvelebil 1995, 2008:31). The latest reiteration is by Barker linguistic isolates, including some languages spoken in the (2006:336), who adds wild goat. mountains of the Caucasus, and with Burushaski, spoken in Current research is unkind to these suggestions. The ge- northern Pakistan. Some of the suggested connections refer netics of einkorn, barley, and lentils all indicate a Near Eastern to domestic cattle, sheep, and goats, to cultivated cereals, and origin (Badr et al. 2000; Heun et al. 1997; Ladizinsky 1999). to milking and tillage. This could imply that Basque is Eu- Modern “wild” sheep on Mediterranean islands are in fact rope’s sole remnant of a pre-Indo-European language family feral (Poplin 1979), and recent DNA studies support a Near that spread with the first agriculturalists (Bengtson 2009). Eastern origin (Meadows et al. 2007). The “goats” mentioned Most European Neolithic archaeologists do not involve them- by Barker (2006:336) are actually ibex (Capra ibex), which selves in the linguistic debate, probably feeling that it can have never been domesticated and have nothing to do with contribute little to the elucidation of the archaeological rec- the origin of domestic goats. Capra aegagrus was domesticated ord. in the Near East (Naderi et al. 2008). Metrical evidence is Genetics also provides some support for immigration. The against the European domestication of aurochs (Rowley- divergence dates for most European mtDNA lineages (as men- Conwy 2003a); the genetic evidence agrees, demonstrating tioned above) fall in the Late Paleolithic. But these studies Near Eastern origins for the domestication of cattle (Bollon- allow perhaps 15% of modern European mtDNA to derive gino and Burger 2007; Edwards et al. 2007; Troy et al. 2001). from Neolithic immigrants (Soares et al. 2010), so the initial Attempts to make Europe part of a vastly expanded Near introduction of agriculture was probably by immigration. The Eastern “agricultural hearth” (cf. Barker 2006:384) have thus Y chromosome provides a complication. Chikhi et al. (2002) failed. Some secondary domestications did occur: wild boar suggest that as much as 50%–65% of modern European Y were domesticated in Europe but only after the arrival of chromosomes descend from Near Eastern ancestors. This does S434 Current Anthropology Volume 52, Supplement 4, October 2011 not mean that such a proportion of males immigrated in the riculturalists, but agriculturalists who carried largely “Meso- Neolithic, however, because the movements cannot be dated lithic” genes from elsewhere in Europe. Small-scale alterna- (Chikhi et al. 2002:11012–11013). A recent estimate for the tives to the wave of advance are envisaged as “infiltration,” immigrant Neolithic Y chromosome proportion is that it “trickle,” or “creep” migrations. Perhaps more applicable to might have been similar to that for mtDNA (Soares et al. larger movements is “leapfrog colonization” moving beyond 2010). the farming frontier into available space (Rowley-Conwy Y chromosome patterns are particularly difficult to dis- 2004b; Zvelebil 2000). Several likely instances will be men- entangle. Zvelebil (2000:70) aptly describes the modern pat- tioned below. tern as an “incremental palimpsest.” Migrations of all kinds have occurred in the past. Armies have high Y chromosome counts and are highly mobile. Roman and Ottoman armies, Sharpening the Agricultural Spreads to mention but two, penetrated huge areas of Europe and Improved excavation and dating everywhere in Europe has involved diverse males—auxiliaries and janissaries recruited put migrationism firmly back on the agenda. The following soldiers not of Roman or Turkish ethnicity. The difficulties sections show how this has happened with regard to the four of establishing what may be Neolithic are shown by the at- spread events plotted in figure 1. tempt by King and Underhill (2002) to correlate immigrant Y chromosomes and Neolithic anthropomorphic figurines. The Cardial of the West Mediterranean Where the correlation is positive (e.g., northern Italy, south- ern France), Neolithic immigration is proposed, but where Twenty years ago, the northern Italian Neolithic was thought figurines are absent (e.g., eastern Spain), the authors resort to start around 7000 BP (uncalibrated), its subsistence based to historically recorded Greek or Phoenician migrations 5,000 largely on wild foods. The adoption of agriculture from the years later to account for the immigrant Y chromosomes now south was slow, sheep and cereals becoming predominant only present in eastern Spain. This seemingly random invocation later in the period (Barker et al. 1987). The Neolithic farther of such disparate migrations does not inspire confidence— west (fig. 2) was, however, much earlier, with dates of BP (uncalibrated) at Chaˆteauneuf-les-Martigues 240 ע why not Napoleon’s invasion army of 1808? It ignores the 7520 BP (uncalibrated) at Verdelpino (Guilaine 150 ע archaeological unity of the Cardial phenomenon (see below), and7970 and the notion that “Y chromosome lineages p figurines” is 1979). Furthermore, sheep were reported in Mesolithic con- remarkably reminiscent of the “pots p people” archaeological texts at several sites: at Chaˆteauneuf, preceding the early 14C theories of half a century ago. date just quoted (Ducos 1977), Gazel and Dourgne (Geddes Genetics will no doubt be taken much further by the ad- 1980, 1985), and Nerja (Boessneck and von den Driesch dition of ancient DNA to the discussion. This is problematic 1980). The role of cereal cultivation was unclear, some emmer because of the possibility of contamination, but it is beginning wheat perhaps appearing at the start of the Neolithic (Lewth- to occur. It is likely that there will be surprises. One study of waite 1986a). 24 LBK skeletons revealed that six of them (25%) carried an This appeared to be a classic case of agricultural adoption mtDNA type that is now present in just 0.2% of modern by hunter-gatherers. The “filter model” suggested that agri- Europeans. If this tiny sample is representative, modern Eu- culture moved along preexisting Mesolithic interconnections. ropeans cannot therefore be descended from this farming Foragers in northwest Italy adopted only certain agricultural population (Haak et al. 2005). elements—initially only sheep—so the other items were “fil- tered out” and not available for transmission to the west. These transitional economies developed into full agriculture Integrationism later in the Neolithic (Lewthwaite 1986a,1986b). Our explanations must now rest on two major foundations: Recent work has transformed this. Early Neolithic agri- most Neolithic genes were native, but the major domesticates culture in northern Italy is now known to comprise four were exotic. Small-scale rather than continent-wide migra- cereals—emmer, einkorn, free-threshing wheat, and barley— tions are the best way to integrate these into one model. and five pulses—lentil, pea, broad bean, bitter vetch, and grass Agriculture in a region may have been introduced by im- pea—from the start, indicating a rapid transmission of full migrants, but that does not mean that the immigrants carried agriculture from the south (Rottoli and Castiglione 2009). mainly Near Eastern genes (Richards 2003; Rowley-Conwy The earliest Neolithic in northwest Italy at Arene Candide is 2004b; Zvelebil 2005). The LBK, for example, originated in dated to ∼7700 cal BP (Maggi and Nisbet 2000), and the the Carpathian Basin; the population that moved westward fauna is dominated by domestic animals from the start emerged there carrying a complex mix of European and Near (Rowley-Conwy 1997). In Portugal, Cisterna dates to ∼7400 Eastern mtDNA and no doubt picking up more as it moved. cal BP (Zilha˜o 2009, forthcoming), Caldeira˜o to ∼7300 cal The same is potentially true of all the spreads shown in figure BP (Zilha˜o 1992). The claimed earlier Neolithic dates in be- 1. tween have been discounted: Zilha˜o (1993:47) has pointed One integrationist scenario is therefore migration by ag- out that the sites were not excavated to modern standards Rowley-Conwy The Westward Spread of Agriculture S435

Figure 2. Map of the western Mediterranean, showing the location of sites mentioned in the text. A color version of this figure is available in the online edition of Current Anthropology. and were disturbed by burrowing animals. “Mesolithic” sheep (Geddes 1995); significantly, four cereal grains in Mesolithic are no longer accepted because of these problems (Guilaine contexts are discounted as intrusive (Marinval 1995:72). and Manen 2007:25–26) and the difficulty of distinguishing Nothing remains of “transitional” economies. the bones of domestic sheep and goat from wild ibex (Capra The Epicardial of the Iberian interior has only been studied ibex) and chamois (Rupicapra rupicapra), especially when ju- very recently. Major rivers such as the Ebro and Tagus were venile (Rowley-Conwy 2004a). likely routes for colonization, though mating with local Meso- Continuity from the Mesolithic is now regarded as mini- lithic people may also have occurred (Arias 1999:414; Guilaine mal. Long considered a “cave Neolithic,” the Early Cardial and Manen 2007:43). Very early Neolithic dates have been has recently produced major open-air settlements at La Draga claimed: 8000 cal BP at Mendandia (Alday 2007) and 7800 (Bosch, Chinchilla, and Tarru´s 2000) and Mas d’Is (Bernabeu cal BP at La La´mpara (Rojo et al. 2006:53, 60). They have, et al. 2003), each with several structures. Farther southwest, however, been criticized by Zilha˜o (forthcoming): pottery was the earliest houses so far known are at Castelo Belinho, dating present at Mendandia, but all the animals were wild (Alday to the later Early Neolithic, after ∼6900 cal BP (Gomes 2008). 2007), while at La La´mpara the dated bone was not identified Sa˜o Pedro de Canaferrim has produced a substantial open- as to species. Directly dated cereal grains at La La´mpara start air settlement dating to before 7000 cal BP (Simo˜es 1999), so at ∼7200 cal BP (Rojo et al. 2006; Stika 2005), the same as earlier houses are likely to be found. Faunal assemblages con- tain a majority of domestic species from the start—La Draga at Los Barruecos on the Tagus (Cerillo and Gonza´lez 2006). has 190% domestic animals (Palomo et al. 2005). Cultigens There was a delay of several centuries before agriculture spread predominate among the plants: the eastern Spanish Cardial to the northern coast, perhaps because the region was more economy comprises the same four cereals and five pulses as densely settled with hunter-gatherers (Arias 2007:62). Cattle in northern Italy (Zapata et al. 2004, table 2); La Draga pro- appear at Arenaza at ∼6900 cal BP (Arias 2007:60) and emmer duced a huge sample of over half a million wheat grains at El Miro´n at ∼6400 cal BP (Pen˜a-Chocarro et al. 2005). (Buxo´, Rovira, and Sau¨ch 2000, fig. 103). Margineda has both Cultivation seems to have become restricted to the cereals as Mesolithic and Neolithic layers (Guilaine and Martzluff 1995). it spread northwest: just three peas were found at Margineda Domestic animals first appear at the start of the Neolithic (Marinval 1995), and no legumes at all were found at La S436 Current Anthropology Volume 52, Supplement 4, October 2011

La´mpara (Stika 2005) or El Miro´n (Pen˜a-Chocarro et al. colonizing voyages would be late summer, after the harvest 2005). was gathered and before winter-sown crops were planted (Broodbank and Strasser 1991:241; Case 1969:178). The fea- sibility of such colonizing voyages is underlined by the spec- Cardial Colonists and Colonization tacularly early evidence of Neolithic colonizers on Cyprus Most researchers now regard the Cardial phenomenon as a presented at the conference (Vigne 2011). rapid colonization by boat (see the influential article by Zilha˜o Survival for the first year, until the newly established cereal 2001). This has been reinforced by the recent definition of fields begin producing, is a further critical variable. Along the the earliest Cardial phase, the “Impressa,” at Pendimoun and Mediterranean coasts hunting would presumably have been Arene Candide (Binder and Maggi 2001). This has subse- particularly important, but this would have been chancy, and quently turned up also at Pont de Roque-Haute and Peiro any food supplies that could be carried with the colonists Signado along the coast in France (Guilaine, Manen, and would have been extremely valuable. Cereals are heavy. Recent Vigne 2007). These two sites are only 3 km apart, but the work suggests an alternative: dairy products. The milking of Peiro Signado ceramics resemble those at Arene Candide, sheep and goat by Cardial people has been suggested on zoo- while Pont de Roque-Haute is more similar to Giglio farther archaeological grounds (Rowley-Conwy 1997:168, 2000). around the Italian coast (Manen 2007:160–163). Pendimoun Analysis of lipids in ceramics has recently demonstrated the resembles sites in eastern peninsular Italy (Binder et al. 1993: use of dairy products in Anatolia back to the ninth millen- 227–228). This suggests multiple leapfrog colonizations, and nium cal BP (Evershed et al. 2008). Because this was the the recent discovery of an Impressa site at El Barranquet in source area of the Cardial agricultural regime, it is likely that Spain extends this apparent Ligurian colonization to 900 km dairying spread with the other items and practices. Dairying from Arene Candide (Bernabeu et al. 2009). Even this early, might be critical in one particular way, because pregnant or the fauna is dominated by domestic animals (Vigne and Car- lactating animals can be driven (Case 1969:177). Such animals re`re 2007). could therefore presumably be carried by boat without ill In most areas there is a radiocarbon gap between Mesolithic effects. The presence of some lactating animals in a newly and Neolithic (Guilaine and Manen 2007, their fig. 2; Skeates established settlement would be invaluable, providing food 2003). In southern Portugal there was considerable Mesolithic each day for several months, thus bridging the gap until other occupation that, bypassed by the Cardial colonization, con- aspects of the economy “kicked in.” tinued until after 7000 cal BP (Zilha˜o 2003, forthcoming). In conclusion, the Cardial phenomenon is an immeasurably The Mesolithic diet was largely coastal, that of the Neolithic sharper event than was understood 20 years ago. In its new terrestrial (Lubell et al. 1994). DNA from the two populations guise it conforms with what we would expect from a migra- differs, although neither carry the Near Eastern lineage im- tion: cultural derivation from northwest Italy, not the local plicated in carrying agriculture into Europe (Chandler, Sykes, Mesolithic; a very rapid spread, with the transplantation of and Zilha˜o 2005). This is consistent with Cardial immigration the entire agricultural system; and the means in place to assure involving farmers of indigenous Mesolithic ancestry (cf. the its spread and survival. “integrationist” model suggested above). The practicalities of maritime colonization must be con- The LBK of Central Europe sidered. Broodbank and Strasser (1991) discuss the Early Neo- lithic colonization of Crete. Forty colonizing humans would The LBK was accepted as appearing relatively abruptly much need 5–10 breeding pairs of each animal species and 250 kg earlier than the Cardial, so recent developments have been of grain, weighing in total some 15–20 tons, to establish an less revolutionary. In a landmark article, Quitta (1960:163– agricultural economy. Such a cargo might be carried in 10– 164, fig. 3) argued that the cultural uniformity of the LBK 15 boats each carrying 1–2 tons (Broodbank and Strasser indicated a rapid immigration along two routes: the Elbe and 1991:241) or fewer boats each making several voyages. A 110- the Danube/Upper Rhine. Subsequent workers mostly accept m log boat comes from the Cardial lake settlement of La migration (e.g., Bogucki 2003; Lu¨ning 1989; Scarre 2002), Marmotta near Rome (Fugazzola, Delpino, and Mineo 1995). though some advocate indigenism (e.g., Price 2000a; Whittle A reconstruction was routinely able to sail 30 km in a day 1996). Gronenborn (2003:81) argues for a migration covering with a crew of 10 and plenty of space for cargo (Tichy 1999). 800 km in 100 years and involving a “multi-faceted combi- Boats made of animal hides stretched over a frame would nation of migrations, adaptations and acculturations” (2007: have been an even better option: they have greater cargo 73). Faunal assemblages usually comprise 160%–80% do- capacity but are light enough to be carried by their own crew mestic animals, with cattle and pigs being predominant (Case 1969). The efficacy of such boats for transport is shown (Do¨hle 1997). Botanical assemblages testify to the overwhelm- by the Irish curragh (fig. 3). Whatever kind of boat they used, ing importance of cultigens (Bogaard 2004; Lu¨ning et al. Cardial people took cattle, sheep, goat, and pig to Sardinia 1997). (e.g., Filiestru: Levine 1983) and also (but without the cattle) The logistics of rapid migration must be considered. Early to Corsica (e.g., Basi: Vigne 1988:153). The optimal time for discussions envisaged shifting cultivation and temporary fields Rowley-Conwy The Westward Spread of Agriculture S437

Figure 3. Transport by Irish curragh. Top, pigs ready for embarking; the animals have their feet tied to prevent movement. Photo by T. Mason, from Jones (2004, fig. 157). Bottom, short-distance towing method; the cow has been tied up to prevent it from struggling. Photo by and courtesy of John Waddell. A color version of this figure is available in the online edition of Current Anthropology. moving across Europe on a broad front (Childe 1929:46; Clark discrete siedlungskammer, or “settlement cells” (Lu¨ning 1989), 1952:92–93). Shifting cultivation has subsequently come un- sometimes widely dispersed (fig. 4). These siedlungskammer der heavy criticism (e.g., Lu¨ning 1980, 2000; Rowley-Conwy start with a few houses close to rivers; later, the houses mul- 1981, 2003b; Sherratt 1980), though some still espouse it tiply, and settlement spreads away from the rivers (Kruk 1980; (Whittle 1996:160–162, 1997:22). Bogaard (2004) has con- Lu¨ning 1989; Stehli 1989). vincingly shown on the basis of weed floras that LBK crops The forest facing the LBK immigrants was thick, with much were intensively cultivated in small autumn-sown fixed plots. understory and undergrowth (Kreuz 2008). Heading off to Distribution is patchy, which is not suggestive of a broad- found a new siedlungskammer with the entire agricultural front migration: settlement largely follows loess soils, forming package would be a challenging enterprise. We must dispense Figure 4. West central Europe at the time of agricultural colonization. The northern part of the map is based on Verhart (2000, fig. 1.15). Loess areas and settlement cells to the southeast are from the more general map in Clark (1952, fig. 45). Where these disagree I have followed Verhart. La Hoguette and Limburg pottery sites from Lu¨ning, Kloos, and Albert (1989, fig. 2), van Berg (1990), and Lefranc (2008, fig. 5). Several Limburg findspots in the Hesbaye cluster are omitted for clarity. A color version of this figure is available in the online edition of Current Anthropology. Rowley-Conwy The Westward Spread of Agriculture S439 with preconceptions derived from the managed woodlands Elsewhere, the immigrants absorbed hunter-gatherers. and regulated rivers of our own times. Choked undergrowth Strontium isotopes in human skeletons suggest that more and sprawling waterways and mires would face anyone at- females than males were of nonlocal origin in the LBK cem- tempting to move through the landscape. In the open, cattle eteries at Schwetzingen and Dillingen (Price et al. 2001). This and pigs cluster into manageable herds, but in woodland they may indicate hunter-gatherer women marrying into farming disperse and rapidly become uncontrollable. How then could communities (Bentley et al. 2002). This pattern is not repeated an 800-km migration be achieved in 100 years? everywhere, and farming practices such as long-range pas- The clue may lie in Quitta’s (1960:163) stressing of the toralism are complicating factors (Bentley et al. 2003; Bickle importance of the Elbe and Danube as routes. The siedlung- and Hofmann 2007). However, the loss of only a few women skammer germinate near rivers (see above). Perhaps the rivers could destabilize low-density foraging populations and might themselves were the highways along which Neolithic colonists prompt the remainder to adopt farming if the unmarried moved. If maritime colonization was feasible for the Cardial males developed “cattle envy.” If this suggests peaceful inter- (see above), riverine colonization is a viable hypothesis for action, the same was not true farther north: in the Hesbaye the LBK. Many central European rivers form a nexus that region, several LBK sites were fortified, apparently against would facilitate this (fig. 4). From the Elbe, travel to the Weser Mesolithic attack (Keeley and Cahen 1989). would be difficult. But from there, short overland moves pro- “La Hoguette” and “Limburg” ceramics are problematic. vided access to the Ruhr and Rhine and from there to the La Hoguette sherds appear on the earliest LBK sites and also Meuse and subsequently to the Marne and Seine. The Danube farther west, beyond the LBK distribution, with Mesolithic similarly runs close to the Rhine and Neckar. Considerable flints (fig. 4). Limburg ceramics are later, with an analogous sections of these river systems must have been navigable in distribution on and beyond the earliest farming settlements the Neolithic. farther west (Lefranc 2008; Lu¨ning, Kloos, and Albert 1989; A 10-m hide boat could carry several people, a couple of van Berg 1990). These groups may be partially acculturated dogs, two cows, two calves, and their bedding (Case 1969). hunter-gatherers, so their ceramics on LBK sites would in- After a day or so, the animals would become restive because dicate contact (Gronenborn 1990:178; Zvelebil 2005). Eco- of thirst, but on a river, with overnight stops, this would not nomic evidence is very limited; the La Hoguette site of Bad be a problem. Recent Irish curraghs are smaller (up to ca. 6 Cannstatt has provided nine deer bones (and 39 antler frag- m), but they can effectively move cattle and pigs, which can ments) and two caprine teeth (von den Driesch, in Kalis et be carried inside the boat for longer moves or towed behind al. 2001, table 3). Dating also is imprecise: there is no clear for shorter hops (fig. 3). Early autumn, between harvesting evidence that La Hoguette pottery predates the LBK. The and autumn sowing, was an agricultural “down time” and suggestion that La Hoguette represents “local Mesolithic peo- would be the optimal time for movement. As with the Cardial, ples who had begun to practice horticulture and herding sev- the first year in a new settlement would be critical. Hunting eral hundred years before the arrival of LBK” (Price et al. would no doubt be important, but once again dairy products 2001:595) goes further than the evidence currently allows. An may have played a key role. A cow that had given birth in added complexity is that La Hoguette ceramics are derived late spring would be lactating in early autumn and might also from the Epicardial of southern France. Lefranc (2008) and be again in calf. This would be a way of transporting a con- van Berg (1990) argue that Limburg has a similar origin, tinuously productive food supply. Bogucki (1984) has argued though Lu¨ning, Kloos, and Albert (1989) disagree. Both ce- that the ceramic sieves found in LBK contexts were used in ramic styles have turned up in Cardial contexts at Gazel (Gui- dairying. laine and Manen 2007:46). The Rhoˆne-Saoˆne system is part LBK contacts with local foragers took a variety of forms. of the river nexus mentioned above (fig. 4), thus providing In much of its area, the LBK is at the head of a sequence of a potential route to the north. Could the La Hoguette people later cultures derived from it. This is not the case for the be “Epi-epi-cardial” farming immigrants? Villeneuve-Saint-Germain (VSG) of northwestern France (fig. 1). Scarre (2002:401) regards the VSG as the ultimate west- Beyond the Northern Frontier ward extension of colonizing farmers. But the VSG did not give rise to later cultures descended from it. After a couple The northern edge of the LBK was the longest-lived forager- of centuries it disappeared, replaced by a more widespread farmer boundary in Neolithic Europe, lasting from ∼7500 cal local Neolithic. Agriculturalized foragers appear to have ab- BP to ∼6000 cal BP. To the north were hunter-gatherers: sorbed the immigrants (Scarre 2002). The picture near the Swifterbant in the Low Countries and Ertebølle in southern mouth of the Vistula is different. The recent discovery of Scandinavia and the Baltic. Artifactual signs of contact occur several post-LBK Stroke Ornamented Pottery (stichbandker- soon after the arrival of adjacent Neolithic groups. LBK ar- amik) settlements indicates agriculture here around 7000 cal rowheads occur in the Swifterbant region from 7500 cal BP BP (Czerniak 2007). But this colonization failed, and the (Louwe Kooijmans 2003). Adzes also spread north (see fig. economy of the area reverted to hunting and gathering after 4) into the same area (Verhart 2000, fig. 1.15) and through this apparently brief and abortive penetration. Poland (Bogucki 2008), reaching the Baltic by 6700 cal BP S440 Current Anthropology Volume 52, Supplement 4, October 2011

(Hartz, Lu¨bke, and Terberger 2007). The Swifterbant regions ization. As archaeologists, we like imported artifacts because began making pottery around 7000 cal BP (Louwe Kooijmans they are (a) identifiable and (b) considered important. But 2007), the Ertebølle following around 6700 cal BP (Andersen how important were they to their Mesolithic recipients? Large 2007; Hartz, Lu¨bke, and Terberger 2007), although the ce- quantities are known, many getting as far as Denmark (fig. ramic style owed more to northeastern Baltic types (Hallgren 5). It is often assumed that they were sought after merely 2004). Things were not all one-way: Bogucki (2008) argues because they were exotic, engendering competition and de- that T-axes of antler in the LBK derive from Ertebølle pro- stabilization in hunter-gatherer societies (e.g., Fischer 2002; totypes. Thomas 1996; Zvelebil 1996, 1998). But virtually all are stray These artifactual contacts did not, however, lead to the finds (Klassen 2002:308–309; Verhart 2000:33); there is no diffusion of agriculture. A few domestic animals are sporad- archaeological reason to assume that they were regarded as ically claimed, but they rarely stand up to scrutiny. Mesolithic in any way special by their Mesolithic users. caprines were claimed at Deby in Poland (Domanska 1989). The LBK remains a sharp archaeological event, still best Objections were soon raised (Kozlowski 1990), and these interpreted as a migration. The speed of the westward mi- claims are no longer accepted (Domanska 2003:590). Some gration contrasts spectacularly with the 1,500-year pause be- domestic cattle and pigs were claimed at the Ertebølle site of fore agriculture spread north into the Swifterbant/Ertebølle Dabki on the Baltic (Ilkiewicz 1989). No criteria demonstrat- sphere. The huge lag between the spread of artifacts and ag- ing domestication were ever put forward, but the specimens riculture suggests that the artifacts did not have any significant are still sometimes mentioned (Zvelebil 2004). A recent re- catalyzing effect. The way to destabilize and Neolithize hunter- examination has, however, demolished these claims: domestic gatherers is not to sell them axes but to encroach on their cattle appear at ∼6200 cal BP (Kabacinski, Heinrich, and Ter- territory and steal their women. berger 2009). A few cattle bones at Rosenhof at ∼6700 cal BP were claimed to be domestic (Nobis 1975). On metrical Southern Scandinavia grounds these are probably aurochs (Rowley-Conwy 1985, 2003a), but they are still often cited as domestic (Hartz 2005; Southern Scandinavia is characterized by the Late Mesolithic Hartz, Heinrich, and Lu¨bke 2000, 2002; Schmo¨lke 2005; Zve- Ertebølle shell middens. The hunter-gatherer population may lebil 2004). Strong support for their identification as aurochs have been as high as about 1 individual per 2 km2;thisis has come from two recent studies. Isotopes show that these similar to that of recent sedentary groups in California and animals’ diet was like that of contemporary aurochs, not later the Northwest Coast and much greater than typical densities domestic cattle (Noe-Nygaard, Price, and Hede 2005). Their elsewhere (Rowley-Conwy 1983). For this reason alone, a sub- DNA matches the European aurochs lineage, not the imported stantial degree of indigenism is espoused by virtually all who domestic Near Eastern lineage (Scheu et al. 2008). There can consider the appearance of agriculture (e.g., Andersen 1973, be little doubt that these animals were wild. The earliest do- 2007; Fischer 2002; Larsson 2007; Price 2000b; Rowley-Conwy mestic cattle from this area are those from Wangels, dating 1999; Zvelebil 2008). to after ∼6100 cal BP, when caprines and cereals also occur Yet even here the case can be made for some migration. (Hartz, Lu¨bke, and Terberger 2007:586; Price and Noe- Chronological sharpness is the crucial issue: if the change to Nygaard 2009). agriculture is gradual, it can be portrayed as an indigenous In the Netherlands, a few bones of domestic animals are development, merely acquiring the domestic crops and ani- claimed at ∼6700 cal BP: 20 sheep/goat at Brandwijk L30 mals from elsewhere; but if it is rapid, migration becomes (Raemaekers 1999, table 3.49), and 15 more, with three do- more likely. Numerous axes of continental Neolithic origin mestic pigs and 15 domestic cows, at Hardinxveld De Bruin are found in Mesolithic Denmark (fig. 5). It is sometimes (Oversteegen et al. 2001). The criteria for separating the cattle argued that these caused social developments toward Neo- and pigs into wild and domestic have not been published, lithization, such as the sedentary occupation of large coastal but the sheep/goat (if contemporary with the rest of the ma- base camps (Fischer 2002). The counterargument (see above) terials) are clearly domestic. Louwe Kooijmans (2003:621) is that the exotic artifacts had little impact: increasing pop- regards these as likely imports of joints of meat, not as locally ulation density, base camps, and territoriality as indicated by reared animals. Cereal grains have not been found, but one the appearance of cemeteries are better explained as a Meso- barley grain comes from Doel in Belgium in a context dated lithic response to increasingly maritime conditions (Rowley- to 6600 cal BP (Crombe´ and Vanmontfort 2007:269); there Conwy 1998, 1999). These developments in fact all took place is, however, later occupation at this site (Crombe´, Perdaen, much earlier than the arrival of agriculture in central Europe and Sergant 2005:55), and the grain itself has not been directly (Larsson 2007; Rowley-Conwy 1999). Coexistence between dated (Philippe Crombe´, e-mail, February 2009). Cereal ag- foragers and farmers and a gradual transition to agriculture riculture is first well attested after 6100 cal BP at Swifterbant are suggested at Lo¨ddesborg (Jennbert 1984). This site, how- S3 (Cappers and Raemaekers 2008). ever, has major stratigraphic problems and is written off by There is thus a disparity between artifacts and agriculture: Scandinavian scholars (see Rowley-Conwy 2004b:87), al- 1,500 years of artifact exchange led to no economic Neolith- though it is still occasionally cited (Zvelebil 2004, fig. 4.2). Figure 5. Late Mesolithic and Early Neolithic southern Scandinavia. Shoe-last axes of Late Mesolithic date imported from farmers to the south from Klassen (2002, fig. 20.1). Trægtbægerkultur and other pottery in Norway from Østmo (2007, fig. 1); Early Neolithic thin-butted axes im- ported from Denmark or southern Sweden from Hinsch (1955, fig. 7). A color version of this figure is available in the online edition of Current Anthropology. S442 Current Anthropology Volume 52, Supplement 4, October 2011

The appearance of agriculture was quite abrupt. In Den- searchers have recently located this gene in 13 archaeological mark there was a rapid change in settlement pattern. Some skeletons, the oldest being Middle Neolithic (directly dated sites did continue to be occupied, but these were seasonal to 5308–4980 cal BP; Malmstro¨m et al. 2008). If this gene special-purpose camps (Rowley-Conwy 1983; Skaarup 1973). was present earlier (and elsewhere in Europe), this would have Mesolithic base camps such as Ertebølle (Andersen and Jo- made the daily food production from a lactating animal even hansen 1986) or Bjørnsholm (Andersen 1991) show reduced more valuable to migrants. levels of Neolithic occupation and may have become special- This agricultural spread apparently stopped just as abruptly purpose fishing camps. Neolithic “base camps” were in inland as it started and did not extend west of Oslo Fjord. Atypical agricultural areas. The shift to interior settlement at the start ceramics and Early Neolithic thin-butted flint axes imported of the Neolithic was abrupt (Larsson 1986; Nielsen 1985). A from the farmers do occur across southern Norway (fig. 5), classic example is the island of Bornholm, where even the and in earlier times these were assumed to indicate farming earliest Neolithic occupation has a markedly inland distri- (Hinsch 1955). Cereal pollen has been claimed at Kotedalen, bution (Nielsen 2009, fig. 6). The same happens on Gotland dated to ∼5800–5000 cal BP (Hjelle 1992). No charred grains (O¨ sterholm 1989). The pointed butted axe, characteristic of were present (Soltvedt 1992), however, and domestic animals the earliest phase of the Early Neolithic, has a markedly inland were completely absent (Hufthammer 1992, 1995). The pollen distribution both in eastern Denmark (Nielsen 1977, fig. 7) record has been strongly criticized (Prescott 1995, 1996; and southern Sweden (Hernek 1988, figs. 4–7). Rowley-Conwy 1999) because of the discrepancy between pol- The available economic evidence also suggests an abrupt len and zooarchaeological evidence: the first domestic-animal change. There is no evidence for domestic animals in Meso- bones appear at Skipshelleren and Srivarhelleren after ∼4200 lithic contexts, contra Thomas (1996:314) and Zvelebil (1996: cal BP (Prescott 1995). 334). One domestic cow, directly dated to ∼6900–6600 cal The agricultural arrival in southern Scandinavia thus ap- BP, is claimed from Lollikhuse (Sørensen 2009), but the iden- pears sharp. Gradualist views of Late Mesolithic developments tification of the tooth remains in doubt (S. Sørensen, e-mail, can be discounted despite the spread of shoe-last axes beyond March 2009). The earliest domestic cattle in Denmark are 16 the farming frontier. Western Norway presents a similar pat- specimens from A˚ konge; this site is dated to after 5900 cal tern: axes and ceramics were in circulation for over a mil- BP and is transitional to the Neolithic (Gotfredsen 1998). lennium beyond the farming boundary. Southern Scandinavia Neolithic economies are dominated by domestic species. Iso- had dense hunter-gatherer populations, which made a high topic analysis of human bones shows an abrupt transition genetic contribution to later agricultural populations. I sug- from a Late Mesolithic marine diet to a terrestrial Neolithic gested long ago that ecological events might have destabilized one (Tauber 1981); even farmers living close to the coast ate the hunter-gatherer economy around 6000 cal BP (Rowley- few marine foods (Richards and Koch 2001). Conwy 1984). This idea has not found much favor, but Bon- This sharp change suggests that some migration occurred sall et al. (2002) have argued that other ecological factors may even if most Neolithic mtDNA was local. The TRB in fact have been active at the same time. If this is so, these ecological spreads a long way north very rapidly (Knutsson and Knuts- factors achieved in a generation what 1,500 years of trading son 2003), Skogsmossen (Hallgren et al. 1997) being near the axes from farmers failed to do. northern limits of cultivation (fig. 5). Farming here seems to “overreach” itself, and subsequently it retreated, being re- Ireland and Britain placed by the Pitted Ware culture, termed “Middle Neolithic” even though they were hunter-gatherers (Eriksson 2004). Since the later 1960s, Britain has been the homeland of grad- Farming probably reached southeastern Norway as well, ualistic perspectives on agricultural origins. This is one of the marked by the TRB pottery in figure 5, though organic pres- few models that has successfully crossed from “processual” ervation is poor (Østmo 2007; Østmo and Skogstrand 2006). to “postprocessual” archaeology and has indeed been taken This farming spread must have been by boat. There were further by the latter. The currently dominant model argues no native aurochs on Zealand (Aaris-Sørensen 1980), so the that for much of the Neolithic, Britain and Ireland remained early cattle at A˚ konge were definitely imported. Farther north, effectively “Mesolithic,” based on nomadic hunting and gath- agriculture was probably carried by boat up the coasts, an ering. Cultivated cereals and domestic animals were “special” easier method of travel than overland (see above). Baltic cross- foods that were hardly utilized on a daily basis but kept in ings would require longer open-water voyages than in the unoccupied stores for occasional consumption (Richmond Cardial or LBK. Irish curraghs can, however, make substantial 1999; Thomas 1996, 2008). voyages and weather considerable seas (Hornell 1938, sec. 5: Against this, some have argued that the transition was 17–21), and a large one has even crossed the Atlantic (Severin abrupt. Unoccupied cereal stores are an unlikely notion: burnt 1978). The role of dairy products in such moves has been houses full of charred cereal remains, as for example at Bal- stressed above. If fresh milk (rather than yogurt, cheese, etc.) bridie, are most simply interpreted as normal domestic struc- was to be consumed, the consumers must have the lactose- tures, more and more of which are turning up (Rowley- tolerant gene present in some modern peoples. Swedish re- Conwy 2004b). Cereal agriculture has long been written down, Rowley-Conwy The Westward Spread of Agriculture S443 but a very large amount of evidence is now available (Jones dated cow bones from anywhere in Britain and Ireland are and Rowley-Conwy 2007) that is fully comparable to the evi- currently two specimens from Late Mesolithic Ferriter’s Cove dence for LBK cultivation (Bogaard and Jones 2007). Neo- dating to ∼6300 cal BP; a slightly younger one comes from lithic faunas with a predominance of wild species persistently Kilgreany Cave (Woodman, Anderson, and Finlay 1999; refuse to make themselves known (Rowley-Conwy 2003b). Woodman and McCarthy 2003). These may have been im- Hambledon Hill, with Britain’s largest Neolithic fauna dated ported as joints of meat rather than as live cattle (Tresset to ∼5500 cal BP, has just been published—and it is dominated 2003:26), but at all events they indicate connections with the by domestic animals (Legge 2008). Dairying could again be continent in the Late Mesolithic, a time when no such con- very important (contra Thomas 2008:70–71); it has long been nections can be seen in Britain (Sheridan 2007:466). Dates argued for on zooarchaeological grounds by Legge (1981) and for the conventional Neolithic in the west of Ireland are some- is supported by lipid analysis of Neolithic ceramics (Copley what curious. The Carrowmore megalithic cemetery has pro- et al. 2003), notably at the recently excavated Early Neolithic duced very early radiocarbon dates, some before 6000 cal BP timber hall at Crathes (Soberl and Evershed 2009). (Burenhult 2001). The early dates have been criticized (Bergh Stable isotope analysis supports an abrupt transition to 1995:98–110). Sheridan (2003:12) nevertheless accepts that terrestrial (i.e., agricultural) foods at the start of the Neolithic construction started at ∼6000 cal BP; if correct, that makes even in such agriculturally unpromising areas as western Scot- them among the earliest megalithic tombs in Britain and Ire- land (Schulting and Richards 2002). The Ce´ide Fields, a 12- land. And the causewayed enclosure at Magheraboy, just 2 km2 agricultural field system in the west of Ireland, are Early km from the Carrowmore cemetery, has been dated to ∼6000 Neolithic (Caulfield, O’Donnell, and Mitchell 1998). An cal BP, also among the earliest in Britain and Ireland (Danaher “abruptist” view is thus a viable alternative to the gradualist 2007:104). orthodoxy. The evidence discussed here suggests that the appearance In keeping with this, migrations are being suggested. Pride of agriculture in Britain and Ireland was about as abrupt as of place among migrants goes to the Orkney vole, a subspecies radiocarbon dating is currently capable of demonstrating. Be- (Microtus arvalis orcadensis) quite different than the voles of cause both are islands, a degree of migration is not just a the rest of Britain. Orkney voles have inhabited Orkney since viable but an inevitable explanation. Further, we must face some time during the Early Neolithic, as a directly dated the surprising possibility that Ireland “went agricultural” be- specimen from Links of Noltland shows; it is not clear whether fore Britain; if true, it would be the biggest leapfrog migration they arrived at the start of the Neolithic or somewhat later. in Europe and the ultimate testimony to the importance of The mtDNA of modern voles shows that they are probably the hide boat. Thomas argues that the rapidity of the process derived from voles in the Bay of Biscay region (Thaw et al. indicates either “a massive, coordinated seaborne invasion” 2004). One or more pairs of voles must have been stowaways or local adoption by hunter-gatherers, because leapfrog mi- on an Early Neolithic voyage from Biscay to Orkney—perhaps gration could not achieve the same result so fast (Thomas in animal bedding. This is the longest individual voyage pro- 2008:65). But this is a false dichotomy: leapfrog migration posed here for a curragh-type boat (fig. 6). A new study most closely accounts for the visible patterns (see above). The suggests that while such a voyage would be difficult in a sailing couple of centuries spanned by the dates is entirely compatible boat, a paddled boat could make the trip from Brittany around with this and is similar to that documented in the other the west of Ireland to Orkney in less than 2 weeks (Callaghan spreads discussed above. and Scarre 2009). Sheridan (2010) proposes that the Early Neolithic was in- Conclusion: Lurches of Advance troduced by several different migrations (fig. 6): the Carinated Bowl tradition may derive from northern France, reaching The four major spreads of agriculture (fig. 1) are all com- Scotland around 5900 cal BP; simple megaliths and their as- patible with the immigration of at least a considerable pro- sociated ceramics suggest connections from southern Brittany portion of farmers. The proposed scenario differs from the to western Scotland and Ireland as early as 6000 cal BP (Sher- “wave of advance” in two important ways: first, each involved idan 2003); and northwest France and southwest England farmers carrying many genes of European Mesolithic origin, show similarities in ceramics and funerary monuments, the and second, the movement is sporadic and punctuated, not “Trans-Manche West” connection (Sheridan 2007; Sheridan continuous. We must replace the monolithic “wave of ad- et al. 2008). Tresset (2003) argues that the animal economy vance” concept with a series of local and disparate “lurches of southern England is so similar to that of northern France of advance.” A similar scenario may also be appropriate for that direct import is likely. Sheridan (2010) states that no southeastern Europe (see O¨ zdog˘an 2011). areas of Britain or Ireland are perfect cultural analogues of The incoming farmers interacted with local foragers in a any area of Europe, so complete cultural transplantations are wide variety of ways. Sometimes the foragers were rapidly not proposed. The mixings and blending inherent in piece- overwhelmed and disappeared as a separate group. All the meal leapfrog migration are the likely causes of this pattern. four major spreads can be interpreted in this way, though A special place may be reserved for Ireland. The earliest with differing proportions of “local” and “incoming” genes Figure 6. Map of Britain, Ireland, and the adjacent continent showing sites mentioned in the text and the connections described. A color version of this figure is available in the online edition of Current Anthropology. Rowley-Conwy The Westward Spread of Agriculture S445 being present in the subsequent Neolithic cultures. In some anonymous reviewers are thanked for their comments, which cases—for example the VSG—the incomers overreached have substantially improved this article. themselves and were subsumed into a Neolithic largely “local” in character. And in two cases—the lower Vistula and eastern References Cited Sweden—farming did not root itself after its first arrival, failed initially, and was displaced temporarily by renewed hunting Aaris-Sørensen, K. 1980. Depauperation of the mammalian fauna of and gathering. the island of Zealand during the Atlantic period. Videnskabelige Meddelelser fra Dansk Naturhistorisk Forening 142:131–138. If the scenarios put forward here have any merit, various Alday, A. 2007. Me´solithique et ne´olithique au Pays Basque d’apre`s lines of inquiry need following up with more energy than l’abri de Mendandia (8500–6400 BP): l’e´volution de l’industrie hitherto. Boats and dairying are two themes I have cham- lithique, le proble`medelace´ramique et les strate´gies d’occupation. pioned; the latter is now being more generally examined, L’Anthropologie 111:39–67. and more thought should perhaps be given to the former. Ammerman, A., and L. L. Cavalli-Sforza. 1984. The Neolithic tran- sition and the genetics of populations in Europe. Princeton, NJ: Finally, the rapidity of the spreads in each area raise the long- Princeton University Press. discredited specter of environmental contexts for some of the Andersen, S. H. 1973. Overgangen fra ældre til yngre stenalder i “lurches of advance,” as Bonsall et al. (2002) have suggested Sydskandinavien set fra en mesolitisk synsvinkel. In Bonde-Veide- for Scandinavia and Britain. mann: bofast-ikke bofast i nordisk forhistorie. P. Simonsen and G. Above all, the migrationist scenarios suggested here may S. Munch, eds. Pp. 26–44. Tromsø Museums Skrifter 14. Tromsø, Norway: Universitetsforlaget. account for one thing: why we so rarely see long-term “tran- ———. 1991. Bjørnsholm: a stratified køkkenmødding on the central sitional” stages between foraging and farming. Now we see Limfjord, North Jutland. Journal of Danish Archaeology 10:59–96. foragers, now we see farmers; but in Europe we have singularly ———. 2007. Shell middens (“køkkenmøddinger”) in Danish pre- failed to catch foragers in the act of becoming farmers. The history as a reflection of the marine environment. In Shell middens long-term developmental processes we have expected for de- in Atlantic Europe. N. Milner, O. E. Craig, and G. N. Bailey, eds. Pp. 31–45. Oxford: Oxbow. cades have not materialized. Farmers can evidently trade axes Andersen, S. H., and E. Johansen. 1986. Ertebølle revisited. Journal with foragers for centuries or longer without destabilizing of Danish Archaeology 5:31–61. them or leading them to adopt farming. “Processes” there Anthony, D. W. 2007. The horse, the wheel, and language: how Bronze- undoubtedly are, but we need to look inside the standard Age riders from the Eurasian steppes shaped the modern world. deviation of a radiocarbon date to see them in action. Princeton, NJ: Princeton University Press. Arias, P. 1999. The origins of the Neolithic along the Atlantic coast of continental Europe: a survey. Journal of World Prehistory 13: 403–464. ———. 2007. Neighbours but diverse: social change in north-west Acknowledgments Iberia during the transition from the Mesolithic to the Neolithic This article is dedicated to the memory of Humphrey Case (5500–4000 cal BC). In Going over: the Mesolithic-Neolithic tran- sition in north-west Europe. A. Whittle and V. Cummings, eds. (1918–2009), who pioneered the idea of agricultural coloni- Special issue, Proceedings of the British Academy 144:53–71. zation by boat and who read this article and offered me the Badr, A., K. Mu¨ller, R. Scha¨fer-Pregl, H. El Rabey, S. Effgen, H. H. benefit of his comments just before he died. Leslie Aiello, Ibrahim, C. Pozzi, W. Rohde, and F. Salamini. 2000. On the origin Ofer Bar-Yosef, Laurie Obbink, and Doug Price are all thanked and domestication history of barley (Hordeum vulgare). Molecular for the magnificent conference, “The Origins of Agriculture: Biology and Evolution 17:499–510. Barker, G. 1985. Prehistoric farming in Europe. Cambridge: Cam- New Data, New Ideas,” at the Hacienda Temozon. Numerous bridge University Press. colleagues provided a great deal of help, sent references, and ———. 2006. The agricultural revolution in prehistory: why did for- gave me access to their unpublished work. In particular I agers become farmers? Oxford: Oxford University Press. thank Joan Bernabeu Auban, John Bengtson, Peter Bogucki, Barker, G., P. Biagi, L. Castelletti, M. Cremaschi, and R. Nisbet. 1987. Dan Bradley, Go¨ran Burenhult, Ramon Buxo´ Capdevila, Sussistenza, economia ed ambiente nel Neolitico dell’Italia setten- trionale. In Atti del XXVI Riunione Scientifica: il Neolitico in Italia, Humphrey Case, Philippe Crombe´, Thomas Cucchi, Mar- vol. 2. Pp. 103–117. Florence: Istituto Italiano di Preistoria e Pro- garita Diaz-Andreu, Keith Dobney, Louisa Gidney, Anders tostoria. Gotherstro¨m, Jean Guilaine, Fredrik Hallgren, Daniel Helmer, Bengtson, J. D. 2009. Basque and the other Mediterranean languages. Lawrence Keeley, Lars Larsson, Mats Larsson, Anna Linder- Mother Tongue 14:157–175. holm, Roberto Maggi, Mario Mineo, Tom Moore, Hilary Bentley, R. A., R. Krause, T. D. Price, and B. Kaufmann. 2003. Human mobility at the Early Neolithic settlement of Vaihingen, Germany: Murray, Poul Otto Nielsen, Einar Østmo, Christopher Pres- evidence from strontium isotope analysis. Archaeometry 45:471– cott, Martin Richards, Chris Scarre, Jeremy Searle, Stephen 486. Shennan, Alison Sheridan, Robin Skeates, Bart Vanmontfort, Bentley, R. A., T. D. Price, J. Lu¨ning, D. Gronenborn, J. Wahl, and Leo Verhart, Jean-Denis Vigne, Alsadair Whittle, Joa˜o Zilhao, P. D. Fullagar. 2002. Prehistoric migration in Europe: strontium and Marek Zvelebil. Carleton Jones is thanked for permission isotope analysis of Early Neolithic skeletons. Current Anthropology 43:799–804. to reproduce the photograph of pigs being loaded into cur- Bergh, S. 1995. Landscape of the monuments: a study of the passage raghs in figure 3. John Waddell is thanked for the photograph graves in the Cu´il Irra region, Co. Sligo, Ireland. Arkeologiska Un- of the cow towed behind a curragh in the same figure. Two derso¨kningar, Skrifter no. 6. Stockholm: Riksantikvariea¨mbetet. S446 Current Anthropology Volume 52, Supplement 4, October 2011

Bernabeu, J., L. Molina, M.A. Esquembre, J. Ramo´n, and J. Boronat. Cappers, R. T. J., and D. C. M. Raemaekers. 2008. Cereal cultivation 2009. La cera´mica impresa Mediterra´nea en el origen Neolı´tico de at Swifterbant? Neolithic wetland farming on the North European la penı´nsula Ibe´rica? In De Me´diterrane´e et d’Ailleurs ...:me´langes Plain. Current Anthropology 49:385–402. offerts a`Jean Guilaine. M. Barbaza et al., eds. Pp. 83–95. Toulouse: Case, H. J. 1969. Neolithic explanations. Antiquity 43:176–186. Archives d’e´cologie pre´historique. Caulfield, S., R. G. O’Donnell, and P. I. Mitchell. 1998. 14C dating Bernabeu, J., T. Orozco, A. Dı´ez, M. Go´mez, and F. J. Molina. 2003. of a Neolithic field system at Ce´ide Fields, County Mayo, Ireland. Mas d’Is (Pena`guila, Alicante): aldeas y recintos monumentales Radiocarbon 40:629–640. del Neolı´tico Inicial en el Valle del Serpis. Trabajos de Prehistoria Cerillo, E., and A. Gonza´lez. 2006. El Neolı´tico Antiguo en la cuenca 60:39–59. media Tajo: estado actual de los conocimientos. In Do Epipaleolı´tico Bickle, P., and D. Hofmann. 2007. Moving on: the contribution of ao Calcolı´tico na Penı´nsula Ibe´rica: actes do IV Congresso de Ar- isotope studies to the Early Neolithic of central Europe. Antiquity queologia Peninsular. N. F. Bicho and H. Verı´ssimo, eds. Pp. 183– 81:1029–1041. 195. Promontoria Monogra´fica 4. Faro: Universidade do Algarve. Binder, D., J.-E. Brochier, H. Duday, D. Helmer, P. Marinval, S. Chandler, H., B. Sykes, and J. Zilha˜o. 2005. Using ancient DNA to Thie´bault, and J. Wattez. 1993. L’abri Pendimoun a` Castellar examine genetic continuity at the Mesolithic-Neolithic transition (Alpes Maritimes): nouvelles donne´es sur le complexe culturel de in Portugal. In III Congreso del Neolı´tico en la Penı´nsula Ibe´rica. la ce´ramique imprime´e Me´diterrane´enne dans son contexte stra- P. Arias, R. Otan˜o´n, and C. Garcı´a-Monco´, eds. Pp. 781–786. Mon- tigraphique. Gallia Pre´histoire 35:177–251. ografı´as del Instituto Internacional de Investigaciones Prehisto´ricas Binder, D., and R. Maggi. 2001. Le Ne´olithique ancien de l’arc liguro- de Cantabria 1. Santander: Universidad de Cantabria. provenc¸al. Bulletin de la Socie´te´ Pre´historique Franc¸aise 98:411–422. Chikhi, L., R. A. Nichols, G. Barbujani, and M. A. Beaumont. 2002. Boessneck, J., and A. von den Driesch. 1980. Tierknochenfunde aus Y genetic data support the Neolithic demic diffusion model. Pro- vier Su¨dspanischen Ho¨hlen. In Studien u¨ber fru¨he Tierknochen- ceedings of the National Academy of Sciences of the USA 99:11008– funde von der Iberischen Halbinsel 7. J. Boessneck and A. von den 110013. Driesch, eds. Pp. 1–83. Munich: Institut fu¨r Palaeoanatomie, Do- Childe, V. G. 1929. The Danube in prehistory. Oxford: Clarendon. mestikationsforschung und Geschichte der Tiermedizin der Uni- ———. 1957. The dawn of European civilization. 6th edition. London: versita¨t Mu¨nchen. Routledge & Kegan Paul. Bogaard, A. 2004. Neolithic farming in central Europe: an archaeo- Clark, J. G. D. 1952. Prehistoric Europe: the economic basis. Cam- botanical study of crop husbandry practices. London: Routledge. bridge: Cambridge University Press. Bogaard, A., and G. Jones. 2007. Neolithic farming in Britain and Cohen, David Joel. 2011. The beginnings of agriculture in China: a central Europe: contrast or continuity? In Going over: the Meso- multiregional view. Current Anthropology 52(suppl. 4):S273–S293. lithic-Neolithic transition in north-west Europe. A. Whittle and V. Copley, M. S., R. Berstan, S. N. Dudd, G. Docherty, A. J. Mukherjee, Cummings, eds. Special issue, Proceedings of the British Academy V. Straker, S. Payne, and R. P. Evershed. 2003. Direct chemical 144:357–375. evidence for widespread dairying in prehistoric Britain. Proceedings Bogucki, P. 1984. Ceramic sieves of the Linear Pottery culture and of the National Academy of Sciences of the USA 100(4):1524–1529. their economic implications. Oxford Journal of Archaeology 3:15– Crawford, Gary W. 2011. Advances in understanding early agriculture 30. in Japan. Current Anthropology 52(suppl. 4):S331–S345. ———. 2003. Neolithic dispersals in riverine interior central Europe. Crombe´, P., Y. Perdaen, and J. Sergant. 2005. La ne´olithisation de la In The widening harvest: the Neolithic transition in Europe: looking Belge: quelque refle´xions. Me´moires de la Socie´te´ Pre´historique forward, looking back. A. Ammerman and P. Biagi, eds. Pp. 249– Franc¸aise 36:47–66. 272. Colloquia and Conference Papers 6. Boston: Archaeological Crombe´, P., and B. Vanmontfort. 2007. The Neolithisation of the Institute of America. Scheldt basin in western Belgium. In Going over: the Mesolithic- ———. 2008. The Danubian-Baltic borderland: northern Poland in Neolithic transition in north-west Europe. A. Whittle and V. Cum- the fifth millennium BC. In Between foraging and farming.H. mings, eds. Special issue, Proceedings of the British Academy 144: Fokkens, B. Coles, A. L. van Gijn, J. P. Kleijne, H. H. Ponjee, and 263–285. C. G. Slappendel, eds. Special issue, Analecta Praehistorica Leidensia Czerniak, L. 2007. The north-east frontier of the post-LBK cultures. 40:51–65. In The Lengyel, Polga´r and related cultures in the Middle/Late Neo- Bollongino, R., and J. Burger. 2007. Neolithic cattle domestication lithic in central Europe. J. K. Kozlowski and P. Raczky, eds. Pp. as seen from ancient DNA. In Going over: the Mesolithic-Neolithic 233–248. Krako´w: Polish Academy of Arts and Sciences. transition in north-west Europe. A. Whittle and V. Cummings, eds. Danaher, E. 2007. Monumental beginnings: the archaeology of the N4 Special issue, Proceedings of the British Academy 144:165–187. Sligo Inner Relief Road. NRA Scheme Monographs 1. Dublin: Na- Bonsall, C., M. G. Macklin, D. G. Anderson, and R. W. Payton. 2002. tional Roads Authority. Climate change and the adoption of agriculture in north-west Denham, Tim. 2011. Early agriculture and plant domestication in Europe. European Journal of Archaeology 5:9–23. New Guinea and Island Southeast Asia. Current Anthropology Bosch, A., J. Chinchilla, and J. Tarru´s. 2000. El poblat lacustre Neolitic 52(suppl. 4):S379–S395. de la Draga: excavacions de 1990 a 1998. Monografies del CASC Dennell, R. 1983. European economic prehistory. New York: Academic 2. Girona, Spain: Museu d’Arqueologia de Catalunya. Press. Broodbank, C., and T. F. Strasser. 1991. Migrant farmers and the Do¨hle, H.-J. 1997. Husbandry and hunting in the Neolithic of central Neolithic colonization of Crete. Antiquity 65:233–245. Germany. In Proceedings of the 7th ICAZ Conference.M.Kokabi Burenhult, G. 2001. The illustrated guide to the megalithic cemetery and J. Wahl, eds. Special issue, Anthropozoologica 25/26:441–448. of Carrowmore, Co. Sligo. Tjo¨rnarp, Sweden: G. Burenhult. Domanska, L. 1989. Elements of a food-producing economy in the Buxo´, R., N. Rovira, and C. Sau¨ch. 2000. Les restes vegetals de llavors Late Mesolithic of the Polish lowland. In The Mesolithic in Europe. i fruits. In El poblat lacustre Neolitic de la Draga: excavacions de C. Bonsall, ed. Pp. 447–455. Edinburgh: Donald. 1990 a 1998. A. Bosch, J. Chinchilla, and J. Tarru´s, eds. Pp. 129– ———. 2003. Interactions between the Late Mesolithic hunter-gath- 140. Monografies del CASC 2. Girona, Spain: Museu d’Arqueologia erers and farming communities in northern Poland. In Mesolithic de Catalunya. on the move. L. Larsson, H. Kindgren, K. Knutsson, D. Loeffler, Callaghan, R., and C. Scarre. 2009. Simulating the western seaways. and A. A˚ kerlund, eds. Pp. 588–591. Oxford: Oxbow. Oxford Journal of Archaeology 28:357–372. Ducos, P. 1977. Le mouton de Chaˆteauneuf-lez-Martiques. In Rowley-Conwy The Westward Spread of Agriculture S447

L’e´levage en Me´diterranne´e occidentale. J.-L. Miege, ed. Pp. 77–85. Guilaine, J., and M. Martzluff. 1995. Les excavacions a la Balma de Paris: CNRS. la Margineda (1979–1991). 3 vols. Andorra: Edicions del Govern Edwards, C. J., R. Bollongino, A. Scheu, A. Chamberlain, A. Tresset, d’Andorra. J.-D. Vigne, J. F. Baird, et al. 2007. Mitochondrial DNA analysis Haak, W., P. Forster, B. Bramanti, S. Matsumura, G. Brandt, M. shows a Near Eastern Neolithic origin for domestic cattle and no Ta¨nzer, R. Villems, et al. 2005. Ancient DNA from first European indication of domestication of European aurochs. Proceedings of farmers in 7500-year-old Neolithic sites. Science 310:1016–1018. the Royal Society of London B: Biological Sciences 274:1377–1385. Hallgren, F. 2004. The introduction of ceramic technology around Eriksson, G. 2004. Part-time farmers or hard-core sealers? Va¨sterbjers the Baltic Sea in the 6th millennium. In Coast to coast: arrival.H. studied by means of stable isotope analysis. Journal of Anthropo- Knutsson, ed. Pp. 123–142. Coast to Coast 10. Uppsala: Uppsala logical Archaeology 23:135–162. University. Evershed, P., S. Payne, A. G. Sherratt, M. S. Copley, J. Coolidge, D. Hallgren, F., U. Djerw, M. af Geierstam, and M. Steineke. 1997. Urem-Kotsu, K. Kotsakis, et al. 2008. Earliest date for milk use in Skogsmossen, an Early Neolithic settlement site and sacrificial fen the Near East and southeastern Europe linked to cattle herding. in the northern borderland of the Funnel-Beaker culture. Tor 29: Nature 455:528–531. 49–111. Fischer, A. 2002. Food for feasting? an evaluation of explanations of Hartz, S. 2005. Aktuelle Forschungen zur Chronologie und Soed- the Neolithisation of Denmark and southern Sweden. In The Neo- lungsweise der Ertebølle und fru¨hesten Trichterbecherkultur in lithisation of Denmark: 150 years of debate. A. Fischer and K. Kris- Schleswig-Holstein. Bodendenkmalpflege in Mecklenburg-Vorpom- tiansen, eds. Pp. 343–393. Sheffield: Collis. mern, Jahrbuch 2004 52:61–81. Fugazzola Delpino, M. A., and M. Mineo. 1995. La piroga neolitica Hartz, S., D. Heinrich, and H. Lu¨bke. 2000. Fru¨he Bauern an der del lago di Bracciano (“La Marmotta 1”). Bullettino de Paletnologia Ku¨ste. Neue 14C-Daten und aktuelle Aspekte zum Neolithisierung- Italiana 86:197–266. sprozeß im norddeutschen Ostseeku¨stengebiet. Praehistorische Geddes, D. 1980. De la chasse au troupeau en Me´diterrane´e occidentale: Zeitschrift 75:129–152. les de´buts de l’e´levage dans le bassin de l’Aude. Archives d’ecologie ———. 2002. Coastal farmers: the neolithisation of northernmost pre´historique 5. Toulouse: Ecole des hautes e´tudes en sciences Germany. In The Neolithisation of Denmark: 150 years of debate. sociales. A. Fischer and K. Kristiansen, eds. Pp. 321–340. Sheffield: Collis. ———. 1985. Mesolithic domestic sheep in west Mediterranean Eu- Hartz, S., H. Lu¨bke, and T. Terberger. 2007. From fish and seal to rope. Journal of Archaeological Science 12:25–48. sheep and cattle: new research into the process of Neolithisation ———. 1995. La fauna mesolı´tica i neolı´tica de la Balma de la in northern Germany. In Going over: the Mesolithic-Neolithic tran- Margineda. In Les excavacions a la Balma de la Margineda (1979– sition in north-west Europe. A. Whittle and V. Cummings, eds. 1991), vol. 3. J. Guilaine and M. Martzluff, eds. Pp. 83–97. An- Special issue, Proceedings of the British Academy 144:567–594. dorra: Edicions del Govern d’Andorra. Hernek, R. 1988. Den spetsnackiga yxan av flinta. Fornva¨nnen 83: Gomes, M. V. 2008. Castelo Belinho (Algarve, Portugal) and the first 216–223. southwest Iberian villages. In The Early Neolithic in the Iberian Heun, M., R. Scha¨fer-Pregl, D. Klawan, R. Castagna, M. Accerbi, B. peninsula. M. Diniz, ed. Pp. 71–78. British Archaeological Reports, Borghi, and F. Salamini. 1997. Site of einkorn wheat domestication International Series 1858. Oxford: Archaeopress. identified by DNA fingerprinting. Science 278:1312–1314. Goring-Morris, A. Nigel, and Anna Belfer-Cohen. 2011. Neolithi- Hinsch, E. 1955. Traktbegerkultur-Megalitkultur: en studie av Øst- zation processes in the Levant: the outer envelope. Current An- Norges eldste, neolitiske gruppe. Universitetets Oldsaksamlings A˚r- thropology 52(suppl. 4):S195–S208. bok 1951–1953:10–177. Gotfredsen, A. B. 1998. En rekonstruktion af palæomiljøet omkring Hjelle, K. L. 1992. Pollenanalytiske undersøkelser innenfor boplassen tre senmesolitiske bopladser i Store A˚ mose, Vestsjælland—baseret i Kotedalen. In Kotedalen: en boplass gjennom 5000 A˚r, vol. 2. K. pa˚pattedyr—og fugleknogler. Geologisk Tidsskrift 1998(2):92–103. L. Hjelle, A. K. Hufthammer, P. E. Kaland, A. B. Olsen, and E. C. Gronenborn, D. 1990. Mesolithic-Neolithic interactions: the lithic Soltvedt, eds. Pp. 91–122. Bergen, Norway: Historisk Museum. industry of the earliest Bandkeramik culture site at Friedberg- Hornell, J. 1938. British coracles and Irish curraghs. London: Society Bruchenbru¨cken, Wetteraukreis (West Germany). In Contributions for Nautical Research. to the Mesolithic in Europe. P. M. Vermeersch and P. V. Peer, eds. Pp. 173–182. Leuven: Leuven University Press. Hufthammer, A. K. 1992. De osteologiske undersøkelsene fra Ko- ˚ ———. 1999. A variation on a basic theme: the transition to farming tedalen. In Kotedalen: en boplass gjennom 5000 Ar, vol. 2. K. L. in southern central Europe. Journal of World Prehistory 13:123– Hjelle, A. K. Hufthammer, P. E. Kaland, A. B. Olsen, and E. C. 210. Soltvedt, eds. Pp. 9–64. Bergen, Norway: Historisk Museum. ———. 2003. Migration, acculturation and culture change in western ———. 1995. Tidlig husdyrhold i Vest-Norge. Arkeologiske Skrifter temperate Eurasia, 6500–5000 cal BC. Documenta Praehistorica 30: fra Arkeologisk Institutt, Bergen Museum 8:203–219. 79–91. Ilkiewicz, J. 1989. From studies on cultures of the 4th millennium ———. 2007. Beyond the models: “Neolithisation” in central Eu- B.C. in the central part of the Polish coastal area. Przeglad Ar- rope. In Going over: the Mesolithic-Neolithic transition in north- cheologiczny 36:17–55. west Europe. A. Whittle and V. Cummings, eds. Special issue, Pro- Jennbert, K. 1984. Den produktiva ga˚van. Acta Archaeologia Lun- ceedings of the British Academy 144:73–98. densia, Series in 4Њ, 16. Lund: University of Lund. Guilaine, J. 1979. The earliest Neolithic in the west Mediterranean: Jensen, J. 1982. The prehistory of Denmark. London: Methuen. a new appraisal. Antiquity 53:22–30. Jones, C. 2004. The Burren and the Aran Islands: exploring the ar- Guilaine, J., and C. Manen. 2007. From Mesolithic to Early Neolithic chaeology. Cork: Collins. in the western Mediterranean. In Going over: the Mesolithic-Neo- Jones, G., and P. Rowley-Conwy. 2007. On the importance of cereal lithic transition in north-west Europe. A. Whittle and V. Cummings, cultivation in the British Neolithic. In The origins and spread of eds. Special issue, Proceedings of the British Academy 144:21–51. domestic plants in Southwest Asia and Europe. S. Colledge and J. Guilaine, J., C. Manen, and J.-D. Vigne. 2007. Pont de Roque-Haute: Conolly, eds. Pp. 391–419. University College London Institute of nouveaux regards sur la ne´olithisation de la France Me´diterrane´enne. Archaeology Publications. Walnut Creek, CA: Left Coast. Toulouse: Centre de recherche sur la pre´histoire et la protohistoire Kabacinski, J., D. Heinrich, and T. Terberger. 2009. Dabki revisited: de la Me´diterrane´e. new evidence on the question of the earliest cattle use in Pom- S448 Current Anthropology Volume 52, Supplement 4, October 2011

erania. In Mesolithic horizons, vol. 2. S. B. McCartan, R. Schulting, ———. 1986b. The transition to food production: a Mediterranean G. Warren, and P. Woodman, eds. Pp. 548–555. Oxford: Oxbow. perspective. In Hunters in transition. M. Zvelebil, ed. Pp. 53–66. Kalis, A. J., J. Meurers-Balke, K. van der Borg, A. von den Driesch, Cambridge: Cambridge University Press. W. Ra¨hle, U. Tegtmeier, and H. Thiemeyer. 2001. Der La Hoguette- Louwe Kooijmans, L. 2003. The Hardinxveld sites in the Rhine/ Fundhorizont in der Wilhelma von Stuttgart-Bad Cannstatt. In Meuse Delta, the Netherlands, 5500–4500 cal BC. In Mesolithic on Zeit-Ra¨ume: Gedenkschrift fu¨r Wolfgang Taute, vol. 2. B. Gehlen, the move. L. Larsson, H. Kindgren, K. Knutsson, D. Loeffler, and M. Heinen, and A. Tillman, eds. Pp. 649–672. Bonn: Deutsche A. A˚ kerlund, eds. Pp. 608–624. Oxford: Oxbow. Gesellschaft fu¨r Ur- und Fru¨geshichte and Rudolf Habelt. ———. 2007. The gradual transition to farming in the Lower Rhine Keeley, L. H., and D. Cahen. 1989. Early Neolithic forts and villages basin. In Going over: the Mesolithic-Neolithic transition in north- in NE Belgium: a preliminary report. Journal of Field Archaeology west Europe. A. Whittle and V. Cummings, eds. Special issue, Pro- 16:157–176. ceedings of the British Academy 144:287–309. King, R., and P. A. Underhill. 2002. Congruent distribution of Neo- Lubell, D., M. Jackes, H. Schwarcz, M. Knyf, and C. Meiklejohn. lithic painted pottery and ceramic figurines with Y-chromosome 1994. The Mesolithic-Neolithic transition in Portugal: isotopic and lineages. Antiquity 76:707–714. dental evidence of diet. Journal of Archaeological Science 21:201– Klassen, L. 2002. The Ertebølle culture and Neolithic continental 216. Europe. In The Neolithisation of Denmark: 150 years of debate.A. Lu¨ning, J. 1980. Getreideanbau ohne Du¨ngung. Archa¨ologisches Kor- Fischer and K. Kristiansen, eds. Pp. 303–317. Sheffield: Collis. respondenzblatt 10:117–122. Knutsson, H., and K. Knutsson. 2003. Stone Age transitions: Neo- ———. 1989. Fru¨he Bauern in Mitteleuropa im 6. und 5. Jahrtausend lithization in central Scandinavia. Documenta Praehistorica 30:49– v. Chr. Jahrbuch des Ro¨misch-Germanischen Zentralmuseums Mainz 78. 35:27–93. Kozlowski, S. K. 1990. Pre-pottery Neolithic in Kuyavia? In Har- ———. 2000. Steinzeitliche Bauern in Deutschland. Bonn, Germany: vesting the sea, farming the forest: the emergence of Neolithic societies Habelt. in the Baltic region. M. Zvelebil, L. Domanska, and R. Dennell, Lu¨ning, J., A. Jockenho¨vel, H. Bender, and T. Capelle. 1997. Deutsche eds. Pp. 141–149. Sheffield: Sheffield Academic Press. Agrargeschichte: Vor- und Fru¨hgeschichte. Stuttgart: Ulmer. Kreuz, A. 2008. Closed forest or open woodland as natural vegetation Lu¨ning, J., U. Kloos, and S. Albert. 1989. Westliche Nachbarn der in the surroundings of Linearbandkeramik settlements? Vegetation bandkeramischen Kultur: La Hoguette und Limburg. Germania History and Archaeobotany 17:51–64. 67:355–420. Kruk, J. 1980. The Neolithic settlement of southern Poland. British Maggi, R., and R. Nisbet. 2000. Alberi da foraggio e scalvatura Neo- Archaeological Reports, International Series 93. Oxford: Archaeo- litica: nuovi dati dalle Arene Candide. In La neolitizzazione tra press. Orient e Occidente. A. Pessina and G. Muscio, eds. Pp. 289–308. Ladizinsky, G. 1999. Identification of the lentil’s wild genetic stock. Udine, Italy: Museo Friulano do Storia Naturale. Genetic Resources and Crop Evolution 46:115–118. Mallory, J. P. 1989. The Indo-Europeans: language, archaeology and Larson, G., A. Albarella, K. Dobney, P. Rowley-Conwy, J. Schibler, myth. London: Thames & Hudson. A. Tresset, J.-D. Vigne, et al. 2007. Ancient DNA, pig domesti- Malmstro¨m, H., A. Linderholm, L. Dale´n, K. Lide´n, J. Stora˚, P. Mol- cation, and the spread of the Neolithic into Europe. Proceedings nar, T. Gilbert, E. Willerslev, G. Holmlund, and A. Go¨therstro¨m. of the National Academy of Sciences of the USA 104(39):15276– 2008. Allele frequencies of the lactase gene in Scandinavian Neo- 15281. lithic populations: hunter-gatherers vs. farmers. In Migration in Larsson, L. 2007. Mistrust traditions, consider innovations? the prehistory: DNA and stable isotope analysis of Swedish skeletal ma- Mesolithic-Neolithic transition in southern Scandinavia. In Going terial. A. Linderholm, ed. Theses and Papers in Scientific Archae- over: the Mesolithic-Neolithic transition in north-west Europe.A. ology 10. Stockholm: Stockholm University. Whittle and V. Cummings, eds. Special issue, Proceedings of the Manen, C. 2007. La production ce´ramique de Pont de Roque-Haute: British Academy 144:595–616. synthe`se et comparaisons. In Pont de Roque-Haute: nouveaux re- Larsson, M. 1986. Neolithization in Scania: a Funnel Beaker per- gards sur la ne´olithisation de la France Me´diterrane´enne.J.Guilaine, spective. Journal of Danish Archaeology 5:244–247. C. Manen, and J.-D. Vigne, eds. Pp. 151–166. Toulouse: Centre Lee, Gyoung-Ah. 2011. The transition from foraging to farming in de recherche sur la pre´histoire et la protohistoire de la Me´diter- prehistoric Korea. Current Anthropology 52(suppl. 4):S307–S329. rane´e. Lefranc, P. 2008. Nouveaux e´le´ments pour une pe´riodisation de la Marinval, P. 1995. Recolleccio´i agricultura de l’epipaleolı´tic al neolı´tic ce´ramique de la Hoguette sur le site rubane´ d’Ittenheim (Bas- antic: ana`lisi carpolo`gica de la Balma de la Margineda. In Les Rhin). Bulletin de la Socie´te´ Pre´historique Franc¸aise 105(2):299– excavacions a la Balma de la Margineda (1979–1991), vol. 3. J. 308. Guilaine and M. Martzluff, eds. Pp. 65–82. Andorra: Edicions del Legge, A. J. 1981. Aspects of cattle husbandry. In Farming practice Govern d’Andorra. in British prehistory. R. Mercer, ed. Pp. 169–181. Edinburgh: Uni- Meadows, J. R. S., I. Cemal, O. Karaca, E. Gootwine, and J. W. Kijas. versity Press. 2007. Five ovine mitochondrial lineages identified from sheep ———. 2008. Livestock and Neolithic society at Hambledon Hill. breeds of the Near East. Genetics 175:1371–1379. In Hambledon Hill, Dorset, England: excavation and survey of a Naderi, S., H.-R. Rezaei, F. Pompanon, M. G. B. Blum, R. Negrini, Neolithic monument complex and its surrounding landscape, vol. 2. H.-R. Naghash, O¨ . Balkız, et al. 2008. The goat domestication R. Mercer and F. Healey, eds. Pp. 536–585. Swindon: English Her- process inferred from large-scale mitochondrial DNA analysis of itage. wild and domestic individuals. Proceedings of the National Academy Levine, M. 1983. La fauna di Filiestru (trincea D). In La Grotta di of Sciences of the USA 105(46):17659–17664. Filiestru a Bonu Ighinu, Mara (SS). D. Trump, ed. Pp. 109–131, Nielsen, P. O. 1977. Die Flintbeile der Fru¨hen Trichterbecherkultur tables 12–27. Quaderni 13. Sassari, Italy: Ministero per i Beni in Da¨nemark. Acta Archaeologica 48:61–138. Culturali e Ambientali. ———. 1985. De første Bønder: nye fund fra den tidligste Lewthwaite, J. 1986a. From Menton to the Mondego in three steps: Trægtbægerkultur ved Sigersted. Aarbøger for Nordisk Oldkyn- application of the availability model to the transition to food pro- dighed og Historie 1984:96–126. duction in Occitania, Mediterranean Spain and southern Portugal. ———. 2009. Den tidligneolitiske bosættelse pa˚Bornholm. In Plads Arqueologia 13:95–119. og Rum i Trægtbægerkulturen. A. Schu¨lke, ed. Pp. 9–24. Nordiske Rowley-Conwy The Westward Spread of Agriculture S449

Fortidsminder Series C, 6. Copenhagen: Det Kongelige Nordiske Richards, Martin. 2003. The Neolithic invasion of Europe. Annual Oldskriftselskab. Review of Anthropology 32:135–162. Nobis, G. 1975. Zur fauna des ellerbekzeitlichen Wohnplatzes Ro- Richards, M. P., and E. Koch. 2001. Neolitisk kost: analyser af senhof in Ostholstein I. Schriften des Naturwissenschaftlichen Vere- kvælstof-isotopen 15N i menneskeskeletter fra yngre stenalder. ins fu¨r Schleswig-Holstein 45:5–30. Aarbøger for Nordisk Oldkyndighed og Historie 1999:7–17. Noe-Nygaard, N., T. D. Price, and S. U. Hede. 2005. Diet of aurochs Richmond, A. 1999. Preferred economies: the nature of the subsistence and early cattle in southern Scandinavia: evidence from 15N and base throughout mainland Britain during prehistory. British Ar- 13C stable isotopes. Journal of Archaeological Science 32:855–871. chaeological Reports, British Series 290. Oxford: Archaeopress. O¨ sterholm, I. 1989. Bosa¨ttningsmo¨nstret pa˚Gotland under Stena˚ldern. Rojo, M. A., M. Kunst, R. Garrido, and I. Garcı´a. 2006. La neoliti- Theses and Papers in Archaeology 4. Stockholm: University of zacı´on de la Meseta Norte a la luz del C-14: ana´lisis de 47 data- Stockholm Institute of Archaeology. ciones absolutas ine´ditas de los yacimientos dome´sticos del Valle Østmo, E. 2007. The northern periphery of the TRB: graves and de Ambrone, Soria, Espan˜a. Archivo de Prehistoria Levantina 26: ritual deposits in Norway. Acta Archaeologica 78(2):111–142. 39–100. Østmo, E., and L. Skogstrand. 2006. Nye funn av traktbegerkeramikk Rottoli, M., and E. Castiglioni. 2009. Prehistory of plant growing ved Oslofjorde: Børsebakke og Vøyenenga. Viking 69:69–90. and collecting in northern Italy, based on seed remains from the Oversteegen, J. F. S., L. H. van Wijngaarden-Bakker, R. Maliepaard, Early Neolithic to the Chalcolithic (c. 5600–2100 cal BC). Vege- and T. van Kolfschoten. 2001. Zoogdieren, vogels en reptielen. In tation History and Archaeobotany 18:91–103. Hardinxveld-De Bruin: een kampplaats uit het Laat-Mesolithicum Rowley-Conwy, P. 1981. Slash and burn in the temperate European en het begin van de Swifterbant-cultuur (5500–4450 v. Chr.). L. L. Neolithic. In Farming practice in British prehistory. R. Mercer, ed. Kooijmans, ed. Pp. 209–297. Rapportage Archeologische Monu- Pp. 85–96. Edinburgh: University Press. mentenzorg 88. Amersfoort, Holland: Rijksdienst voor het Oudh- ———. 1983. Sedentary hunters: the Ertebølle example. In Hunter- eidkundig Bodemonderzoek. gatherer economy in prehistory. G. N. Bailey, ed. Pp. 111–126. Cam- O¨ zdog˘an, Mehmet. 2011. Archaeological evidence on the westward bridge: Cambridge University Press. expansion of farming communities from eastern Anatolia to the ———. 1984. The laziness of the short-distance hunter: the origins Aegean and the Balkans. Current Anthropology 52(suppl. 4):S415– of agriculture in Denmark. Journal of Anthropological Archaeology S430. 3:300–324. Palomo, A., J. F. Gibaja, R. Pique´, M. San˜a, A. Bosch, and J. Tarru´s. ———. 1985. The origins of agriculture in Denmark: a review of 2005. La caza en la yacimiento Neolı´tico lacustre de La Draga some theories. Journal of Danish Archaeology 4:188–195. (Banyoles, Girona). In III Congreso del Neolı´tico en la Penı´nsula ———. 1997. The animal bones from Arene Candide: final report. Ibe´rica. P. Arias, R. Otan˜o´n, and C. Garcı´a-Monco´, eds. Pp. 135– In Arene Candide: functional and environmental assessment of the 143. Monografı´as del Instituto Internacional de Investigaciones Holocene sequence. R. Maggi, ed. Pp. 153–277. Memorie dell’Is- Prehisto´ricas de Cantabria 1. Santander: Universidad de Cantabria. tituto Italiano di Paleontologia Umana, n.s. 5. Rome: Ministero Pen˜a-Chocarro, L., L. Zapata, M. J. Iriarte, M. Gonza´lez, and L. G. per i Beni Culturali e Ambientali. Straus. 2005. The oldest agriculture in northern Atlantic Spain: ———. 1998. Cemeteries, seasonality and complexity in the Ertebølle new evidence from El Miro´n Cave (Ramales de la Victoria, Can- of southern Scandinavia. In Harvesting the sea, farming the forest: tabria). Journal of Archaeological Science 32:579–587. the emergence of Neolithic societies in the Baltic region. M. Zvelebil, Poplin, F. 1979. Origine du Mouflon de Corse dans une nouvelle L. Domanska, and R. Dennell, eds. Pp. 193–202. Sheffield: Sheffield perspective pale´ontologique: par marronage. Annales Ge´ne´tique et Academic Press. Se´lection Animale 11:133–143. ———. 1999. Economic prehistory in southern Scandinavia. In Prescott, C. 1995. From Stone Age to Iron Age: a study from Sogn, World prehistory: studies in memory of Grahame Clark. J. Coles, R. western Norway. British Archaeological Reports, International Se- M. Bewley, and P. Mellars, eds. Special issue, Proceedings of the ries 603. Oxford: Hadrian. British Academy 99:125–159. ———. 1996. Was there really a Neolithic in Norway? Antiquity 70: ———. 2000. Milking caprines, hunting pigs: the Neolithic economy 77–87. of Arene Candide in its west Mediterranean context. In Animal Price, T. D. 2000a. Europe’s first farmers: an introduction. In Europe’s bones, human societies. P. Rowley-Conwy, ed. Pp. 124–132. Oxford: first farmers. T. D. Price, ed. Pp. 1–18. Cambridge: Cambridge Oxbow. University Press. ———. 2003a. Early domestic animals in Europe: imported or locally ———. 2000b. The introduction of farming in northern Europe. In domesticated? In The widening harvest: the Neolithic transition in Europe’s first farmers. T. D. Price, ed. Pp. 260–300. Cambridge: Europe: looking forward, looking back. A. Ammerman and P. Biagi, Cambridge University Press. eds. Pp. 99–117. Colloquia and Conference Papers 6. Boston: Ar- Price, T. D., R. A. Bentley, J. Lu¨ning, D. Gronenborn, and J. Wahl. chaeological Institute of America. 2001. Prehistoric human migration in the Linearbandkeramik of ———. 2003b. No fixed abode? nomadism in the northwest Eu- central Europe. Antiquity 75:593–603. ropean Neolithic. In Stones and bones: formal disposal of the dead Price, T. D., and N. Noe-Nygaard. 2009. Early domestic cattle in in Atlantic Europe during the Mesolithic-Neolithic interface 6000– southern Scandinavia. In From Bann Flakes to Bushmills. N. Finlay, 3000 BC. G. Burenhult and S. Westergaard, eds. Pp. 115–144. S. McCartan, N. Milner, and C. Wickham-Jones, eds. Pp. 198– British Archaeological Reports, International Series 1201. Oxford: 210. Oxford: Oxbow. Archaeopress. Quitta, H. 1960. Zur frage der a¨ltesten Bandkeramik in Mitteleuropa. ———. 2004a. From Arene Candide to the Atlantic: the Bernabo` Pra¨historische Zeitschrift 38:1–38, 153–188. Brea excavations and early domestic animals in the west Medi- Raemaekers, D. C. M. 1999. The articulation of a new Neolithic: the terranean. In Dalle Arene Candide a Lipari. P. Pelagatti and G. meaning of the Swifterbant culture for the process of Neolithisation Spadea, eds. Special issue, Bollettino d’Arte, not numbered:123– in the western part of the North European Plain (4900–3400 BC). 132. Archaeological Studies Leiden University. Leiden: Faculty of Ar- ———. 2004b. How the West was lost: a reconsideration of agri- chaeology, Leiden University. cultural origins in Britain, Ireland, and southern Scandinavia. Cur- Renfrew, C. 1987. Archaeology and language: the puzzle of Indo- rent Anthropology 45(suppl.):S83–S113. European origins. London: Cape. Scarre, C. 2002. Pioneer farmers? the Neolithic transition in western S450 Current Anthropology Volume 52, Supplement 4, October 2011

Europe. In Examining the farming/language dispersal hypothesis.P. Soria, central Spain. Vegetation History and Archaeobotany 14:189– Bellwood and C. Renfrew, eds. Pp. 395–407. Cambridge: Mc- 197. Donald Institute for Archaeological Research. Tauber, H. 1981. 13C evidence for dietary habits of prehistoric man Scheu, A., S. Hartz, U. Schmo¨lke, A. Tresset, J. Burger, and R. Bol- in Denmark. Nature 292:332–333. longino. 2008. Ancient DNA provides no evidence for independent Thaw, S., M. Jaarola, J. Searle, and K. M. Dobney. 2004. The origin domestication of cattle in Mesolithic Rosenhof, northern Germany. of the Orkney vole Microtus arvalis orcadensis: a proxy for recon- Journal of Archaeological Science 35:1257–1264. structing human movements. In Atlantic connections and adap- Schmo¨lcke, U. 2005. Neue archa¨ozoologische Untersuchungen zur tations: economies, environments and subsistence in lands bordering Mecklenburger Bucht und zum Jasmunder Bodden. Bodendenk- the North Atlantic. R. A. Housley and G. Coles, eds. Pp. 114–119. malpflege in Mecklenburg-Vorpommern, Jahrbuch 2004 52:145–153. Symposia of the Association for Environmental Archaeology 21. Schulting, R., and M. P. Richards. 2002. The wet, the wild and the Oxford: Oxbow. domesticated: the Mesolithic-Neolithic transition on the west coast Thomas, J. 1996. The cultural context of the first use of domesticates of Scotland. European Journal of Archaeology 5:147–189. in continental central and northwest Europe. In The origins and Severin, T. 1978. The Brendan voyage. London: Hutchinson. spread of agriculture and pastoralism in Eurasia. D. R. Harris, ed. Sheridan, A. 2003. French connections I: spreading the marmites Pp. 310–322. London: University College London Press. thinly. In Neolithic settlement in Ireland and western Britain.I. ———. 1999. Understanding the Neolithic. London: Routledge. Armit, E. Murphy, E. Nelis, and D. Simpson, eds. Pp. 3–17. Oxford: ———. 2008. The Mesolithic-Neolithic transition in Britain. In Pre- Oxbow. historic Britain. J. Pollard, ed. Pp. 58–89. Oxford: Blackwell. ———. 2007. From Picardie to Pickering and Pencraig Hill? new Tichy, R. 1999. Monoxylon II: plavba po 8000 letech (with English information on the “Carinated Bowl Neolithic” in northern Brit- translation: Monoxylon expeditions). Hronov 1999:167–215. ain. In Going over: the Mesolithic-Neolithic transition in north-west Tresset, A. 2003. French connections II: of cows and men. In Neolithic Europe. A. Whittle and V. Cummings, eds. Special issue, Proceed- settlement in Ireland and western Britain. I. Armit, E. Murphy, E. ings of the British Academy 144:441–492. Nelis, and D. Simpson, eds. Pp. 18–30. Oxford: Oxbow. ———. 2010. The Neolithisation of Britain and Ireland: the “big Troy, C. S., D. E. MacHugh, J. F. Balley, D. A. Magee, R. T. Loftus, picture.” In Landscapes in transition. B. Finlayson and G. Warren, P. Cunningham, A. T. Chamberlain, B. C. Sykes, and D. G. Bradley. eds. Pp. 89–105. Levant Supplementary Series 8. Oxford: Oxbow. 2001. Genetic evidence for Near-Eastern origins of European cattle. Sheridan, A., R. Schulting, H. Quinnell, and R. Taylor. 2008. Revis- Nature 410:1088–1091. iting a small passage tomb at Broadsands, Devon. Devon Archae- van Berg, P.-L. 1990. Ceramique du Limbourg et Neolithisation en ological Society Proceedings 66:1–26. Europe du Nord-Ouest. In Rubane´et Cardial. D. Cahen and M. Sherratt, A. 1980. Water, soil and seasonality in early cereal culti- Otte, eds. Pp. 161–208. E´ tudes et recherches arche´ologiques de vation. World Archaeology 11:313–330. l’Universite´ de Lie`ge 39. Lie`ge: Universite´ de Lie`ge. Simo˜es, T. 1999. Osı´tio Neolı´tico de Sa˜o Pedro de Canaferrim, Sintra. Verhart, L. 2000. Times fade away: the Neolithization of the southern Trabalhos de Arqueologia 12. Lisbon: Instituto Portugueˆs de Ar- Netherlands in an anthropological and geographical perspective.Ar- queologia. chaeological Studies Leiden University. Leiden: Faculty of Ar- Skaarup, J. 1973. Hesselø-Sølager: Jagdstationen der su¨dskandinav- chaeology, Leiden University. ischen Trichterbecherkultur. Copenhagen: Akademisk. Vigne, Jean-Denis. 1988. Les mammife`res post-glaciaires de Corse: Skeates, R. 2003. Radiocarbon dating and interpretations of the e´tude arche´ozoologique. XXVIe supple´ment a` Gallia Pre´histoire. Mesolithic-Neolithic transition in Italy. In The widening harvest: Paris: CNRS. the Neolithic transition in Europe: looking forward, looking back.A. Vigne, Jean-Denis, and I. Carre`re. 2007. Les verte´bre´s terrestres et Ammerman and P. Biagi, eds. Pp. 157–187. Colloquia and Con- l’exploitation des ressources animales. 2007. In Pont de Roque- ference Papers 6. Boston: Archaeological Institute of America. Haute: nouveaux regards sur la ne´olithisation de la France Me´di- Soares, P., A. Achilli, O. Semino, W. Davies, V. Macaulay, H.-J. Ban- terrane´enne. J. Guilaine, C. Manen, and J.-D. Vigne, eds. Pp. 181– delt, A. Torroni, and M. Richards. 2010. The archaeogenetics of 214. Toulouse, France: Centre de recherche sur la pre´histoire et la Europe. Current Biology 20:174–183. protohistoire de la Me´diterrane´e. Soares, P., L. Ermini, N. Thomson, M. Mormina, T. Rito, A. Ro¨hl, Vigne, Jean-Denis, Isabelle Carre`re, Franc¸ois Briois, and Jean Gui- A. Salas, S. Oppenheimer, and M. B. Richards. 2009. Correcting laine. 2011. The early process of mammal domestication in the for purifying selection: an improved human mitochondrial mo- Near East: new evidence from the pre-Neolithic and Pre-Pottery lecular clock. American Journal of Human Genetics 84:1–20. Neolithic in Cyprus. Current Anthropology 52(suppl. 4):S255–S271. Soberl, L., and P. Evershed. 2009. Organic residue analysis of pottery Whittle, A. 1996. Europe in the Neolithic: the creation of new worlds. from Warren Field Timber Hall and the Crathes Castle Overflow Cambridge World Archaeology. Cambridge: Cambridge University Car Park site. In A tale of the unknown unknowns: a Mesolithic pit Press. alignment and a Neolithic timber hall at Warren Field, Crathes, ———. 1997. Moving on and moving around: Neolithic settlement Aberdeenshire. H. K. Murray, J. C. Murray, and S. M. Fraser, eds. mobility. In Neolithic landscapes. P. Topping, ed. Pp. 15–22. Neo- Pp. 93–97. Oxford: Oxbow. lithic Studies Group Seminar Papers 2. Oxford: Oxbow. Soltvedt, E. C. 1992. Makrofossilundersøkelserne fra Kotedalen. In Woodman, P., and M. McCarthy. 2003. Contemplating some awful(ly Kotedalen: en boplass gjennom 5000 A˚r, vol. 2. K. L. Hjelle, A. K. interesting) vistas: importing cattle and red deer into prehistoric Hufthammer, P. E. Kaland, A. B. Olsen, and E. C. Soltvedt, eds. Ireland. In Neolithic settlement in Ireland and western Britain.I. Pp. 123–137. Bergen: Historisk Museum. Armit, E. Murphy, E. Nelis, and D. Simpson, eds. Pp. 31–39. Sørensen, S. A. 2009. Lollikhuse, a site from the transitional phase Oxford: Oxbow. between the Mesolithic and Neolithic in Denmark. In Mesolithic Woodman, P. C., E. Anderson, and N. Finlay. 1999. Excavations at horizons, vol. 2. S. B. McCartan, R. Schulting, G. Warren, and P. Ferriter’s Cove, 1983–85: last foragers, first farmers in the Dingle Woodman, eds. Pp. 541–547. Oxford: Oxbow. Peninsula. Bray: Wordwell. Stehli, P. 1989. Merzbachtal: Umwelt und Geschichte einer band- Zapata, L., L. Pen˜a-Chocarro, G. Pe´rez-Jorda´, and H.-P. Stika. 2004. keramischen Siedlungskammer. Germania 67:51–76. Early Neolithic agriculture in the Iberian Peninsula. Journal of Stika, H.-P. 2005. Early Neolithic agriculture in Ambrona, Province World Prehistory 18:283–325. Rowley-Conwy The Westward Spread of Agriculture S451

Zeder, Melinda A. 2011. The origins of agriculture in the Near East. Zvelebil, M. 1995. Hunting, gathering, or husbandry? management Current Anthropology 52(suppl. 4):S221–S235. of food resources by the late Mesolithic communities of temperate Zhao, Zhijun. 2011. New archaeobotanic data for the study of the Europe. MASCA Research Papers in Science and Archaeology origins of agriculture in China. Current Anthropology 52(suppl. 4): 12(suppl.):79–104. S295–S306. ———. 1996. The agricultural frontier and the transition to farming Zilha˜o, J. 1992. Gruta do Caldeira˜o: o Neolitico antigo. Trabalhos de in the circum-Baltic region. In The origins and spread of agriculture Arqeologia 6. Lisbon: Instituto Portugueˆs do Patrimo´nio Arqui- and pastoralism in Eurasia. D. R. Harris, ed. Pp. 323–345. London: tecto´nico e Arqueolo´gico, Departamento de Arqueologia. University College London Press. ———. 1993. The spread of agro-pastoral economies across Med- iterranean Europe: a view from the far west. Journal of Mediter- ———. 1998. Agricultural frontiers, Neolithic origins, and the tran- ranean Archaeology 6:5–63. sition to farming in the Baltic basin. In Harvesting the sea, farming ———. 2001. Radiocarbon evidence for maritime pioneer coloni- the forest: the emergence of Neolithic societies in the Baltic region. zation at the origins of farming in west Mediterranean Europe. M. Zvelebil, L. Domanska, and R. Dennell, eds. Pp. 9–27. Sheffield Proceedings of the National Academy of Sciences of the USA 98(24): Archaeological Monographs 10. Sheffield: Sheffield Academic 14180–14185. Press. ———. 2003. The Neolithic transition in Portugal and the role of ———. 2000. The social context of the agricultural transition in demic diffusion in the spread of agriculture across west Mediter- Europe. In Archaeogenetics: DNA and the population prehistory of ranean Europe. In The widening harvest: the Neolithic transition in Europe. C. Renfrew and K. Boyle, eds. Pp. 57–79. Cambridge: Europe: looking forward, looking back. A. Ammerman and P. Biagi, McDonald Institute. eds. Pp. 207–223. Colloquia and Conference Papers 6. Boston: ———. 2004. Who were we 6000 years ago? in search of prehistoric Archaeological Institute of America. identities. In Traces of ancestry: studies in honour of Colin Renfrew. ———. 2009. The Early Neolithic artifact assemblage from the Gal- eria da Cisterna (Almonda karstic system, Torres Novas, Portugal). M. Jones, ed. Pp. 41–60. Cambridge: McDonald Institute. In De Me´diterrane´e et d’Ailleurs . . . : me´langes offerts a`Jean Gui- ———. 2005. Looking back at the Neolithic transition in Europe. laine. M. Barbaza, ed. Pp. 821–831. Toulouse, France: Archives European Journal of Archaeology 8:183–194. d’e´cologie pre´historique. ———. 2008. Innovating hunter-gatherers: the Mesolithic of the ———. Forthcoming. Time. In The dynamics of Neolithisation in Baltic. In Mesolithic Europe. G. Bailey and P. Spikins, eds. Pp. 18– Europe: studies in honour of Andrew Sherratt. Oxford: Oxbow. 59. Cambridge: Cambridge University Press.

Current Anthropology Volume 52, Supplement 4, October 2011 S453

The Origins of Plant Cultivation and Domestication in the New World Tropics Patterns, Process, and New Developments

by Dolores R. Piperno

CAϩ Online-Only Material: Supplement A

The New World tropical forest is now considered to be an early and independent cradle of agriculture. As in other areas of the world, our understanding of this issue has been significantly advanced by a steady stream of archaeobotanical, paleoecological, and molecular/genetic data. Also importantly, a renewed focus on formulating testable theories and explanations for the transition from foraging to food production has led to applications from subdisciplines of ecology, economy, and evolution not previously applied to agricultural origins. Most recently, the integration of formerly separated disciplines, such as developmental and evolutionary biology, is causing reconsiderations of how novel phenotypes, including domesticated species, originate and the influence of artificial selection on the domestication process. It is becoming clear that the more interesting question may be the origins of plant cultivation rather than the origins of agriculture. This paper reviews this body of evidence and assesses current views about how and why domestication and plant food production arose.

Introduction evidence bearing on the prehistory of Neotropical agriculture (Piperno and Pearsall 1998). Since then, much more evidence has been generated from It has long been recognized that numerous New World plant all the contributing disciplines. New archaeological sites have domesticates—more than half, in fact, of all American crops been discovered, excavated, and analyzed using the full com- and many of the staples that supported indigenous peoples plement of now-available archaeobotanical techniques. Im- when Europeans arrived—originated in Neotropical forests portant early sites that were first identified and studied more (e.g., Harris 1972; Sauer 1950). In the past few decades, a than 30 years ago and for which little to no botanical infor- large corpus of archaeobotanical, paleoecological, and mo- mation was available have seen reexcavation and applications lecular/genetic information has become available from Cen- of microfossil research. Phytoliths and in some cases mac- tral and South America that has led to a significantly increased rofossils from other early human occupations have been re- understanding of the geography and chronology of tropical visited during the past 10 years with the use of expanded and food production. This information has established the low- improved modern reference collections, and some of these land Neotropical forest as an early and independent cradle of remains have been directly dated. Starch grain data from stone agricultural origins. In a volume published in 1998, Deborah tools and human teeth are becoming available from a growing Pearsall and I reviewed and synthesized the known archae- number of sites, and construction of large modern reference ological, paleocological, ethnographic, and molecular/genetic collections has allowed identification of some domesticated root and seed crops. Major collecting efforts of important economic taxa such as the Cucurbitaceae have been under- Dolores R. Piperno is Senior Scientist and Curator of Archaeobotany taken, adding significant knowledge about modern landrace and South American Archaeology in the Program for Human Ecology diversity in Cucurbita and Lagenaria, along with distributional and Archaeology in the Department of Anthropology at the Smithsonian National Museum of Natural History (10th Street and ranges of wild species. Last but not least, newer molecular Constitution Avenue, Washington, DC 20560, U.S.A. [pipernod@ and other information has refined and in some cases revised si.edu]). This paper was submitted 13 XI 09, accepted 9 XII 10, and the geography of plant domestication and also caused us to electronically published 4 VIII 11. reconsider how we should view and come to an understanding

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0019$10.00. DOI: 10.1086/659998 S454 Current Anthropology Volume 52, Supplement 4, October 2011 of the domestication process. In this paper, I review the pres- seasonal tropical forest is the native vegetation (e.g., Doebley ently available information and assess current views of how 2004; Matsuoka et al. 2002; van Heerwaarden et al. 2011). and why plant food production and domestication arose.1 In many cases studied where multiple domestications were discussed, single domestications are now indicated (e.g., man- The Geography of Domestication ioc, maize, the important pejibaye palm; Le´otard et al., forth- coming; Matsuoka et al. 2002; Olsen 2002; Olsen and Schaal 1999; Rodrigues, Filho, and Clement 2004). On the other Agriculture may originate in discrete centers or evolve over hand, a double domestication of the lima bean, one event in vast areas without definable centers. the southern Ecuador/northern Peru Andes leading to the (Jack R. Harlan 1971:468) large-size lima and another in the humid tropical lowlands It is useful to begin by assessing what is currently known leading to the small-seeded sieva bean, is again indicated by about the geography of Neotropical plant domestication (for the newest molecular data (Motta-Aldana et al., forthcoming). a summary, see fig. 1). Scholars such as David Harris, Carl Furthermore, lowland western Mexico is definitively revealed Sauer, Jack Harlan, and others of their generations knew that as an origin area for the sieva, and it is possible that another tropical forest agriculture probably developed independently independent domestication event occurred in the lowlands of long before European arrival, and they for the most part Central or South America (Motta-Aldana et al., forthcoming). understood whether Central America (largely Mesoamerica) More than one domestication of an important Central Amer- or South America was the original home of the major crops. ican tree crop, S. purpurea, also appears likely (Miller and More precise area origins for New World crops, and whether Schaal 2005). single or multiple domestications or even hybrid origins oc- An Old World plant, the bottle gourd, was transported to curred in some, was often less clear for a number of reasons. and well used throughout tropical America during the pre- Wild congenerics that potentially could be progenitor species Columbian era. Erickson et al. (2005), working with modern on the basis of shared morphological attributes or apparent and ancient DNA, established conclusively that prehistoric lack of hybridization barriers were often broadly distributed, plants originated in Asia, not Africa, as was previously as- and they sometimes occurred in both Central America and sumed. The archaeological rinds they studied bore the mark South America (e.g., Manihot [manioc]; Cucurbita [squashes of domesticated plants. The investigators concluded that hu- and gourds]; Zea [teosinte], the ancestor of maize; Ipomoea mans, not ocean currents, probably carried the domesticated [sweet potato]; Xanthossoma [malanga or cocoyam, New plant from Asia during the initial colonization of the New World representatives of the taro family]; Dioscorea [yams]; World. Phaseolus lunatus [sieva beans]; Gossypium [cotton]; many Importantly, molecular data have also elucidated in more tree crops; Piperno and Pearsall 1998). detail some of the different processes involved in the do- The analysis of protein and DNA-based molecular markers mestication of plants (see also “The How of Plant Domes- has significantly elucidated these issues, most recently for tication”). For example, interesting cases of hybrid origins of some major root, seed, and tree crops such as manioc, various Mesoamerican polyploid tree crops were confirmed in the species of Cucurbita, chayote (Sechium edule), P. lunatus (sieva important genus Leucaena, whereby human-mediated sym- beans), Leucaena spp. (guaje), Spondias purpurea (jocote or patry in “backyard gardens” of previously separated wild taxa, Mexican plum), Bactris gasipaes (pejibaye or peach palm), extensive predomestication cultivation, and subsequent spon- peanuts (Arachis hypogaea), South American cotton, and oth- taneous hybridization led to the emergence of different do- ers (e.g., Bertoti de Cunha et al. 2008; Cross, Saade, and mesticated species in the genus (Hughes et al. 2007). The Motley 2006; Hughes et al. 2007; Le´otard et al., forthcoming; most widely cultivated Mesoamerican cactus, Opuntia ficus- Milla, Isleib, and Stalker 2005; Miller and Schaal 2005; Motta- indica, probably arose by the same means (Griffith 2004; Aldana et al., forthcoming; Olsen 2002; Olsen and Schaal Hughes et al. 2007). The importance of these processes and 1999; Robledo, Lavia, and Seijo 2009; Rodrigues, Filho, and backyard, also called “dooryard,” cultivation in contributing Clement 2004; Sanjur et al. 2002; Westengen, Huama´n, and to the stock of plants domesticated prehistorically was pro- Heun 2005; fig. 1). Furthermore, it has been conclusively posed long ago by Edgar Anderson (1954). We should re- shown that the wild ancestor of maize is not native to the member that other major American crops, such as peanuts, semiarid Mexican highlands but rather to lower, warmer, and are products of the hybridization of two different species and moister habitats of Guerrero and Michoaca´n states, where the how, when, and where of the predomestication hybridi- zation events that led to their initial emergence are not well 1. The following definitions are used in this paper. “Cultivation” and understood. “farming” refer to the preparation of plots specified for plant propagation Figure 1 shows known or postulated geographic zones of and repeated planting and harvesting in such plots. A “cultivated plant” domestication for some Neotropical crops on the basis of or “cultivar” refers to those that are planted and harvested, regardless of their domesticated status. “Domesticated species” are those that have current molecular, archaeological, and ecological evidence (for been genetically altered through artificial selection such that phenotypic a more complete list of crops native to the Neotropics, see characteristics distinguish them from wild progenitors. Harlan 1992). In some cases, designations of the circled and Piperno Plant Cultivation and Domestication in the New World Tropics S455 other localities on the maps as origin areas for particular crops Of course, knowing that crop plant origins were geograph- should be treated as hypotheses that require testing with ad- ically widespread does not necessarily lead to a conclusion ditional molecular and archaeological data. For example, that food production was independently developed wherever more precise domestication localities for yam Dioscorea trifida, a particular crop originated. Rather, the spread of a crop or leren Calathea allouia, cocoyam Xanthossoma sagittifolium, crops into new regions may have inspired receiving cultures and achira Canna edulis within their likely origin areas of to grow their own native plants. This issue is further com- northern South America may be forthcoming once relevant plicated by the fact that we do not know when some of the molecular studies have been undertaken. There is a major crops displayed in figure 1 were initially brought under cul- root crop for which neither Central American nor South tivation. How many truly independent developments of low- American origins have been conclusively demonstrated. Sweet land food production were there? Given that lowland north- potato’s presumed wild ancestor is Ipomoea trifida, a poorly ern and southern South American domestication zones are demarcated taxon naturally distributed from Mexico through separated by large distances and that a number of plants native northern South America (Piperno and Pearsall 1998). Some to these areas were taken under cultivation and domesticated molecular work based on amounts of diversity in modern by the middle of the eighth millennium BP, it is difficult to landraces suggests that it may be Central American in origin see how the northern and southern lowland regions do not (Zhang et al. 2000), although the oldest archaeological evi- form at least two to three independent areas of food pro- dence for it by far derives from western South America (Pip- duction (e.g., D1, D3, and D4 in fig. 1B). These questions erno and Pearsall 1998; Shady, Haas, and Creamer 2001). will be further illuminated in the near future as more data Centers of present diversity do not always accurately pinpoint are accumulated. a center of origin, and additional molecular work focusing In sum, Harlan’s (1971) idea that peoples over a wide geo- on the construction of a rigorous phylogeny of wild and do- graphic area were simultaneously engaging in early cultivation mesticated sweet potato is needed to resolve the issue. and domesticatory relationships with plants and considerably Even with the existing uncertainties, it is clear that crop influenced the early development of some domesticates after origins were spatially diffuse. Using the conventional defi- the plants left their native areas seems particularly relevant in nition of a center—a circumscribed region where agriculture the light of current data. Opportunities for such kinds of began and out of where it spread—Mesoamerica, more pre- processes to occur would have been even greater if, as in the cisely Mexico, may still qualify, although it is clear now that Old World (e.g., Fuller 2007; Jones and Brown 2007; Weiss, maize, the common bean (Kwak, Kami, and Gepts 2009) and Kislev, and Hartmann 2006; Willcox, Fortnite, and Herveux other Phaseolus species (the tepary bean), various tree crops, 2008), protracted periods of predomestication cultivation chile peppers, cotton, chayote, and others were domesticated and/or spread of predomesticated crops occurred in the Neo- in different and sometimes ecologically dissimilar regions of tropics. This issue is discussed in more detail below. the country (see also Piperno, forthcoming). In South Amer- Finally, molecular and botanical studies together with an ica, an attempt to designate a single circumscribed center or increasing amount of archaeobotanical data, summarized be- core area of agriculture would clearly not be supported. The low, tell us that the wild ancestors of many important crop origins of major and now minor crops are spread from the plants are native to the seasonal tropical forest, those for- northern parts to the southern parts of the continent, west mations that receive a 4–7-month-long period every year dur- and east of the Andes, mostly in seasonal types of lowland ing which little to no rain falls. Annual precipitation in these tropical forest for major root and seed crops but also in low- areas averages about 1,200–1,800 mm a year, soils are less land wet forests and midelevation moist forest habitats. weathered and thus more highly fertile than in ever-wet (asea- Patterns indicating multiple areas of domestication become sonal) forest, and the prolonged dry season enabled early even more accentuated when highland crops such as Phas- farmers to efficiently clear vegetation and prepare plots for eolus, Amaranthus spp., Chenopodium spp., and the various planting with the simple use of fire. Seasonally dry tropical Andean tubers are added to the picture (the Andean complex forests do not carry the distinction of their rain forest relatives, also includes additional species of Cucurbita and Capsicum; but their prominent position in Neotropical agricultural or- Piperno, forthcoming; Piperno and Dillehay, forthcoming). igins is clear. Crops that would eventually become major staple foods or condiments and that were commonly grown together after Early Food Production and Its Cultural food production was established—such as maize, manioc, var- and Ecological Contexts ious squashes, chile peppers, sweet potato, and Phaseolus and Canavalia beans—had spatially disparate areas of origin and The New World was colonized before 15,000 BP, by which did not at first spread together. In fact, some of the earliest time human populations had traversed most of the western and most widespread anchors of lowland food production are South American landmass, probably by moving along the now minor, even disappearing, foods (below). Early patterns Pacific Coast, and found themselves in what is now southern of dispersals probably did not involve significant population Chile (Dillehay et al. 2008). (All dates given in the text are movements or diffusion of crops in packages. in calibrated 14C years; uncalibrated and calibrated determi- S456 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 1. Postulated domestication areas for various tropical crops in Central and South America. Open circles are archaeological and paleo- ecological sites with early domesticated crop remains. The numbers in parentheses after a taxon indicate that more than one independent do- mestication occurred. The possible area of origin for the sieva bean ex- tends into the Pacific lowlands north of the oval area. The oval here and those in B labeled D1–D4 designate areas where it appears that more than one or two important crops may have originated. Arrows point to approximate areas and are not meant to denote specific domestication locales (for more details and sources used in the figure, see Piperno and Pearsall 1998; Piperno 2006a; see also Cross, Saade, and Motley 2006; Hughes et al. 2007; Matsuoka et al. 2002; Miller and Schaal 2005; Motta- Aldana et al. forthcoming; Olsen and Schaal 1999; Plowman 1984; Ro- drigues, Filho, and Clement 2004; Sanjur et al. 2002; Wessel-Beaver 2000; Westengen, Huama´n, and Heun 2005). Modern vegetation zone guides are (a) 1, tropical evergreen forest; 2, tropical semievergreen forest; 3, tropical deciduous forest; 4, savanna; 5, low scrub/grass/desert; 6, mostly cactus scrub and desert; and (b) 1, tropical evergreen forest (TEF); 2, tropical semievergreen forest (TSEF); 3, tropical deciduous forest (TDF); 4, mixtures of TEF, TSEF, and TDF; 5, mainly semievergreen forest and drier types of evergreen forest; 6, savanna; 7, thorn scrub; 8, caatinga; 9, cerrado; 10, desert. nations for specific dates from sites discussed are found in ended at about 11,400 BP, human settlement of the Neotropics table 1.) The earliest indisputable evidence of human occu- began to change from these sparsely distributed and short- pation in what is now the lowland Neotropical forest comes term occupations. People “settled into” their landscapes, stay- from sites distributed from Belize to eastern Brazil that were ing for longer and/or more frequently returning to specific first occupied at about 13,000 BP (Piperno and Pearsall 1998; locations, and they frequently manipulated and altered their Ranere and Cooke 2003). Beginning soon after the Pleistocene environments by creating clearings in forests and/or burning Piperno Plant Cultivation and Domestication in the New World Tropics S457

Figure 1. (Continued) them. They developed tool kits indicating that for the first (Dillehay et al. 2007; Piperno and Dillehay 2008). The veg- time the exploitation of plants was as important an economic etation of all of these areas was humid tropical forest with strategy as hunting had been (e.g., Gnecco and Aceituno 2006; the exception of the Vegas, Ecuador, sites located at an ecotone Mora 2003; Ranere and Cooke 2003; Ranere et al. 2009). between forest and scrub vegetation. Table 1 contains detailed Archaeobotanical information indicates that food production information on crop plant occurrence and chronology. As- began in a number of localities in tropical Central and South sociated published references and other information about America during the early Holocene (between 11,000 and 7600 the sites involved can be found in CAϩ online supplement BP), not long after the Neotropical climate and vegetation A. Figure 2 displays site locations. The reader should refer to underwent profound changes associated with the end of the the citations given above and in CAϩ supplement A in dis- Pleistocene (discussed in more detail below). cussions that follow below. The best evidence currently comes from the Central Balsas The relevant archaeobotanical data are in the main part Valley of southwestern Mexico (Piperno et al. 2009; Ranere from microfossils, namely, starch grains recovered from nu- et al. 2009), central Pacific and western Panama (Dickau 2010; merous securely dated stone tools and human teeth and phy- Dickau, Ranere, and Cooke 2007; Piperno 2006c, 2009; Pip- toliths retrieved from the same stone tools and/or closely erno et al. 2000), the sub-Andean and premontane zones (elevation between 1,000 and 1,600 m) of the Cauca and Porce associated sediments. In several cases, phytoliths and starch valleys in Colombia (e.g., Aceituno and Castillo 2005; Bray grains have been directly dated. Macrobotanical information et al. 1987; Gnecco and Aceituno 2006), the Colombian Am- is available from Colombia and northern Peru, and paleo- azon (Cavelier et al. 1995; Mora 2003; Piperno and Pearsall ecological efforts allied with archaeological work provide pol- 1998), southwestern Ecuador (Pearsall 2003; Pearsall, Chan- len, phytolith, and charcoal data indicating crop presence and/ dler-Ezell, and Chandler-Ezell 2003; Pearsall, Chandler-Ezell, or substantial human vegetational modification near sites. Ta- and Zeidler 2004; Piperno 2009; Piperno and Stothert 2003; ble 1 also includes early macrobotanical data from the arid Zarillo et al. 2008), and the Zan˜a Valley of northern Peru coast of Peru, where crops grown were largely from elsewhere Table 1. Crop plant occurrence and chronology

Site Age Crop plants Mexico: Guerrero State: (BP (by 8960–8940, 8850–8840, Maize (Phy and SG-GS), Cucurbita (Phy 40 ע Xihuatoxtla Shelter By 7920 and 8780–8630 cal BP) Tabasco State: (BP (7204–6904 cal BP) Maize (Phy, Po 47 ע San Andre´s 6208 Panama: Central Pacific Panama: Aguadulce Rock Shelter By ca. 8600 cal BP Cucurbita moschata, leren, bottle gourd (Phy), Arrowroot (Phy, SG-GS) ,(BP (7740–7640 cal BP) Maize, manioc (SG-GS 60 ע 6910 (BP (7922–7754 cal BP) Maize (Phy 81 ע 7061 By ca. 5700 cal BP Dioscorea trifida (SG-GS) (BP (7804–7631 cal BP) Maize (SG-GS, Phy, Po 90 ע Cueva de los Ladrones 6860 (cal BP) Maize (SG-GS 7584–7779) 110 ע Cerro Mangote 6810 Western Panama: (BP (7554–7381 cal BP) Arrowroot, maize (SG-GS 120 ע Chiriqui Rock Shelters 6560 Ca. 5600 cal BP Manioc (SG-GS) (BP (7779–7584 cal BP) Maize (SG-GS 270 ע Hornito 6270 Colombia: (BP (be- Maize (SG, Phy-GS;Po 80 ע and 5880 120 ע Middle Porce Valley Between 6280 tween 7321–7032 and 6799–6597 cal BP) Middle Cauca Valley: ([?BP (8493–8313 cal BP) Dioscorea (SG-GS [D. trifida 90 ע El Jazmin 7590 Between ca. 7000 and 5000 BP (ca. 8000–6000 Maize (Po) cal BP) Middle Cauca Valley, Calima Region: BP C. cf. moschata (Phy); Persea americana 140 ע and 7830 120 ע El Recreo 7980 (M, [Cul?]); (9001–8674 and 8903–8508 cal BP) Cucurbitaceae (M) (180BP(16138–5721 cal BP) Maize (Po ע Hacienda Lusitania 15150 (BP (7771–7349 cal BP) Maize (Po 230 ע Hacienda El Dorado 6680 Upper Cauca Valley: BP (11,058–10,706 cal BP) Bottle gourd (M, Phy), Cucurbita (Phy 100 ע San Isidro 9530 [Cul.?]), P. americana (M), Maranta cf. arundinacea (SG-GS [Cul.?]) Colombian Amazon: Middle Caqueta´ Region: (BP (9107–8884 cal BP) Cucurbita, leren, bottle gourd (Phy 60 ע Pen˜a Roja 8090 (BP (15539–5351 cal BP) Maize, manioc (Po 40 ע Abeja 14695 Southwestern Ecuador: Las Vegas Sites: (BP Cucurbita ecuadorensis (Phy 250 ע and 9320 40 ע OGSE-80 and OGSE-67 Between 10,130 (11,750–10,220 cal BP) (BP (11,060–10,950, 10,780–10,220 Leren, bottle Gourd (Phy 250 ע 9320 cal BP) (BP (8015–7945 cal BP) Maize (Phy 60 ע 7170 (BP (6850–6810 cal BP) Maize (Phy 180 ע 15820 Valdivia Sites: Real Alto Ca. 4300 BP (ca. 5000 cal BP) Leren, achira, arrowroot, maize, manioc (SG, Phy-GS; Phy), jack bean, cotton (M) ,BP (5260–5000 cal BP) Arrowroot, maize, manioc, jack bean 40 ע Loma Alta 4470 Capsicum (SG-Cer) Ca. 5500–4400 BP (ca. 6500–5200 cal BP) Maize, Cucurbita, Achira (Phy) Ecuadorian Amazon: Ayauchi Ca. 5300 BP (ca. 6000 cal BP) Maize (Po, Phy-Lake sediments) Eastern Amazon: (?) BP (6662–6464 cal BP) Slash-and-burn cultivation 90 ע Geral, Brazil 5760 Ca. 3350 BP (ca. 3800 BP) Maize (Po, Phy-Lake sediments) Piperno Plant Cultivation and Domestication in the New World Tropics S459

Table 1. (Continued) Site Age Crop plants Northern Peru: (BP (10402–10253 cal BP) C. moschata (M 50 ע Zan˜a Valley 9240 (BP (8630–8580 cal BP) Arachis sp. (M 40 ע 7840 (BP (6301–6133 cal BP) Cotton (M 60 ע 5490 Ca. 7500 BP (ca. 8500 cal BP) Manioc (M) BP (9403–8784 cal BP) C. moschata, Arachis, Phaseolus, Inga 180 ע 8210 feuillei (SG-HT) BP (7950 cal BP) Coca (Erythroxylum novagranatense var 50 ע 7120 truxillense;M) (BP (11015–10885 cal BP; BGS 2426) Cucurbita (Phy 65 ע Siches 9533 BP (10243–10306 cal BP; BGS 60 ע and 9222 2475) Southern Coastal Peru: Paloma Ca. 7800 BP (ca. 8800 cal BP) Bottle gourd (M) (BP (5900–5740 cal BP) Cucurbita ficifolia (M 40 ע 5070 By 5300–4700 BP (6500–5700 cal BP) Phaseolus lunatus, Cucurbita spp., guava (Psidium guajava;M) (BP (6440–6310 cal BP) P. lunatus (M 57 ע Chilca 1 5616 By 4400 BP (5400 cal BP) Cucurbita, Achira, Jicama (Pachyrhizus ahipa), jack bean (BP (8445–8395 cal BP) Bottle gourd (M 50 ע Quebrada Jaguay 7660 Southeastern Uruguay: -BP (4800–4540 cal BP) Maize, Phaseolus (SG-GS); maize, Cucur 40 ע Los Ajos 4190 bita (Phy) Note. Dates indicate the first appearance of crops in the different contexts from each site studied. GS p recovered from grinding stones; HT p recovered from human teeth; Cer p recovered from food residues in ceramic pots; if none of these context designations is listed, the botanical remains were recovered from sediments. M p macrobotanical; Phy p phytoliths; SG p starch grains; Po p pollen. Bold printed 14C dates indicate that the determinations were made directly on the botanical material. Other 14C dates are on other materials (usually wood charcoal, sometimes shell and human bone) from closely associated contexts. Laboratory numbers are listed for previously unpublished radiocarbon dates. Cul? p uncertainty as to whether remains are from wild or cultivated/domesticated plants. and their first appearance provides a minimum date for their 7600 BP. Pollen evidence from the Colombian Amazon in- domestication. dicates that manioc arrived there before 5800 BP. It is possible The earliest crops were Calathea allouia (leren) and Ma- that peanuts and manioc moved north together from a com- ranta arundinacea (arrowroot), both grown for their tubers; mon area of origin (fig. 1B). Chile peppers were well dispersed Cucurbita moschata, Cucurbita ecuadorensis, and possibly Cu- in southern Central America and South America by 6000– curbita argyrosperma squash; bottle gourd; maize; manioc; 5000 BP (Perry et al. 2007). peanuts; avocado; and pacay (Inga feullei), another tree crop. Recent studies in the Central Balsas River Valley of Mexico, The first three—leren, arrowroot, and C. moschata squash— maize’s postulated cradle of origin, document the presence along with the bottle gourd are persistently present in north- of maize phytoliths and starch grains at 8700 BP, the earliest ern South America and Panama between 10,200 and 7600 BP, date recorded for the crop (Piperno et al. 2009; Ranere et al. underlining their probable northern South American origins. 2009). A large corpus of data indicates that it was dispersed In many cases, different kinds of archaeobotanical remains into lower Central America by 7600 BP and had moved into provide mutually supporting evidence for the same crop spe- the inter-Andean valleys of Colombia between 7000 and 6000 cies from the same site and specific context (table 1). It is BP. Given the number of Cauca Valley, Colombia, sites that obvious now that the earliest crop complexes were neither demonstrate early maize, it is likely that the inter-Andean seed, tree, nor root crop based but rather mixtures of these valleys were a major dispersal route for the crop after it en- different elements. tered South America (table 1; fig. 2). Furthermore, directly During the middle eighth millennium BP, the first signs of dated starch grains from food residues in early Valdivia ce- major crop movements northward from their area(s) of origin ramics once again affirm maize and other crop presence in in southern South American can be seen. Macrofossil and domestic contexts in these Early Formative (6000–5000 BP) starch grain data show that peanuts moved into the Zan˜a occupations (Zarillo et al. 2008; see table 1 for other Valdivia Valley of northern Peru by 8500 BP. Macrofossils of manioc crop plant occurrences). (identification confirmed by starch grains isolated directly An abundance of artifactual evidence also speaks to early from root remains) also occur there by about 8500 BP, and traditions of dedicated plant exploitation and cultivation. starch grains from the plant are present in central Panama at Stone implements used for plant processing (“edge-ground S460 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 2. Location of sites with early food production in Central and South America discussed in the text. The sites are shown with open circles. Sites with an open triangle are important examples of pre-6000 BP crop diffusion into southern coastal Peru (Chilca 1, Paloma, and Quebrada Jaguay) and regions of South America where agriculture as a whole had been poorly documented (e.g., Geral, located in the interior of the Am- azon, and possibly Los Ajos, Uruguay, by ca. 4000 BP). They are not discussed in the text but are listed in table 1. cobbles,” quern stone bases, hand milling stones) are common amples, which by 7300 BP are beautifully polished, are among in sites with early food-producing activities, and they are often the most typical implements documented in early preceramic present from the earliest Holocene onward in sites that were occupations there. Stone axes presumably used for creating occupied at that time. Residues studied from these tools have openings in the forest are also found in the Cauca, Porce commonly produced starch grains and phytoliths from a va- Valley, and Colombian Amazon occupations at the same time. riety of cultigens (table 1). In the Porce and Middle Cauca In the way of other economic remains documented, a num- valleys of Colombia, the Colombian Amazon, and the Zan˜a ber of wild plant taxa, including a variety of palms and other Valley in Peru, finely made stone hoes firmly dated to from tree fruits (e.g., Spondias, Erythrina, Byrsonima), yams other 9500–7300, 8000, and 7000 BP, respectively, occur in sites than Dioscorea trifida, Calathea species other than C. allouia, with evidence of early food production (table 1) and in others and Zamia have been identified. It is possible that some of for which botanical data are not available (Aceituno and Cas- them, particularly a few of the palm genera, were cultivated, tillo 2005; Gnecco and Aceituno 2006; Herrera et al. 1992; but this cannot be empirically demonstrated with either the Mora 2003; Salgado 1995). The Cauca region Colombian ex- macrobotanical data or the microbotanical data. Current in- Piperno Plant Cultivation and Domestication in the New World Tropics S461 formation similarly cannot be employed to adequately quan- sources (see papers this issue). For a number of reasons, the tify changes in the proportions of wild and cultivated plant high biodiversity of tropical forests does not translate into foods as early periods of cultivation ensued. Bone often does habitats with abundant and stable wild resources for hunters not survive in the humid tropics, but in the few sites with and gatherers. Useful calorie-rich plant species occur in low faunal records (e.g., Las Vegas and the Zan˜a Valley), both number and are widely dispersed in space, and animal protein large and small animals were hunted. The totality of evidence is similarly far from plentiful, with the effect that natural indicates broad-spectrum subsistence orientations shortly be- resources would generally be expected to support small groups fore and at the beginning of food production. of mobile foragers (Piperno and Pearsall 1998:52–78). Re- With evidence accumulating rapidly now, other questions source abundance would have been better, at least seasonally, such as how early cultivated and domesticated plants moved in localities of early food production where permanent lakes (through movements of people, objects, or cultural knowl- occurred, such as in an area of the Central Balsas, Mexico, edge) will take on increasing importance. This issue cannot studied recently (Piperno et al. 2007, 2009; Ranere et al. 2009). be discussed in any detail here, but considering the present However, most early food producers did not have access to evidence, simple down-the-line forms of exchange that did lakes. Piperno and Pearsall (1998:323) concluded that in the not involve significant population shifts or movements of Neotropics, early farming occurred “in the most optimal material culture may best account for early crop plant dif- zones of the most optimal types of forests and, in this sense, fusion. It should be stressed that different scenarios may have they represent resource abundance,” adding that “nonetheless, been true for later cases of crop diffusion through the Neo- as long as people are not starving or otherwise experiencing tropics not discussed here, when Formative-period societies frequent and severe shortfalls of food, food abundance per were established and population numbers were much higher se may have little relevance.” There is no reason to alter this than they were during the early and early middle Holocene. view. It is clear that unlike in the Near East and China (see papers this issue), Neotropical food production did not originate and When Did Neotropical Farming Become take hold in the context of large or fairly large permanent a Significant Subsistence Practice? and nucleated villages situated in major river valleys. Rather, intensive foot surveys and excavations in the Balsas region of The evidence increasingly indicates that it would be a mistake southwestern Mexico, central and western Panama, south- to assume that Neotropical food production during the entire western Ecuador, the Cauca and Porce regions, Colombia, 11,000–7000-BP period was a casual undertaking practiced and northern Peru show that between 11,000 and 7000 BP, by people who are better called foragers than farmers (see sites are typically rock shelters and/or limited clusters of small also Iriarte 2007, 2009 for discussions of this issue). Multi- open-air occupations that were located beside secondary wa- faceted archaeobotanical data—now including starch grain tercourses and seasonal streams whose small stretches of al- and phytolith evidence from the calculus of human teeth— luvium likely were used for planting gardens. Settlements were along with settlement pattern, landscape modification (in typically less than 1 ha in size, and many may have been some cases resulting from slash-and-burn cultivation), and occupied seasonally. Settlement organization was similar to artifactual information indicate that by 8800–7600 BP in the modern tropical hamlets and hamlet clusters, where one to Zan˜a Valley, Peru; central Panama; probably the Central Bal- a few nuclear families composed the residential community. sas, Mexico; Tabasco, Mexico; and possibly other regions, a Early expressions of permanent settlements and seminu- significant number of dietary calories and nutrients came cleated communities are found in the Zan˜a Valley, Peru, where from crop plants. For example, more than 70% of the starch occupations between 10,000 and 7600 BP are small circular grains recovered from the teeth of nine different Zan˜a Valley houses located 200–400 m apart with stone foundations and individuals dated from 8800–7700 BP were from four crops: stone-lined storage pits. The deep and dense Las Vegas, Ec- Phaseolus, Cucurbita moschata, peanuts, and Inga feuillea,a uador, middens, situated near the Pacific coast, where a wide tree (Piperno and Dillehay 2008). In Panama, the Colombian variety of marine and estuary resources (e.g., mollusks, fish, Amazon, and southwestern Ecuador, root crops such as leren and crabs) as well as plants and terrestrial animals were ex- and arrowroot that were significant and reliable sources of ploited, may also have been occupied on a permanent basis. carbohydrates are ubiquitous components of early crop plant It was not until substantially later in time that settlements assemblages. Cucurbita, another ubiquitous early plant, pro- in these and other regions were positioned to exploit the rich vided high-quality proteins and fats. Furthermore, the starch, bottomland of significant river courses. The earliest such evi- phytolith, and macrobotanical evidence from a number of dence comes from the Valdivia culture of southwest Ecuador regions makes it clear that squashes (including the fruit flesh and dates to 5500 BP (e.g., Pearsall, Chandler-Ezell, and in the Zan˜a Valley) were routinely consumed, had undergone Chandler-Ezell 2003; Raymond 2008). In other regions, this artificial selection for different traits related to fruit edibility, development took place at about 4000–3400 BP. Another re- and thus were not primarily used as little-modified nondietary lated and often-discussed facet of early farming is the view plants. that it should originate in zones of plentiful wild food re- In the Central Balsas and San Andre´s regions of Mexico S462 Current Anthropology Volume 52, Supplement 4, October 2011 and in central Panama, forest clearance from slash-and-burn to reassess how plant domestication occurred. Fundamental cultivation is documented by pollen, phytolith, and charcoal new insights about the origin of novel traits and adaptive records from lake sediment cores beginning at 7600–7200 BP evolution are arising from the fields of evolutionary devel- (Piperno et al. 2007; Pohl et al. 2007; Pope et al. 2001). Small opmental biology (evo-devo), molecular biology, and devel- irrigation canals are present in the Zan˜a Valley beginning at opmental plasticity. In addition, a constant stream of detailed 7700 BP (Dillehay, Eling, and Rossen 2005). In a number of archaeobotanical investigations has altered our views about regions, then, agricultural intensification began before 7000 the trajectory of plant domestication following the advent of BP. In central Panama; the Zan˜a Valley, Peru; and the Cauca systematic cultivation (defined here as the cycle of planting and Porce regions, Colombia, it is also evident that site num- and harvesting in plots prepared for this purpose). I start first ber and artifact density significantly increased around 7600– with new insights from the nonarchaeological spheres of re- 7000 BP. These trends are likely the result of an increase in search. human carrying capacity made possible by effective systematic Gene regulation/expression acting during an organism’s de- food production. velopment and phenotypic (developmental) plasticity are now In summary, these people appear to have been committed widely viewed as significant forces, indeed by some investi- horticulturists and slash-and-burn cultivators who, while still gators as the primary forces, in evolutionary diversification integrating planting with collecting and hunting, were taking and the origin of novel traits (e.g., Carroll 2005; Kruglyak important steps along the pathway to full-scale agriculture. and Stern 2007; Pigliucci 2005; West Eberhard 2003). There The appearance of large sedentary and nucleated villages, is good reason to believe that these research foci should be- which postdates 6000 BP throughout the Americas, should come standard elements of domestication studies (Gremillion no longer be considered a necessary backdrop for the occur- and Piperno 2009a). Regulatory genes are not protein-coding rence or recognition of effective and productive agriculture or functional genes but rather act to switch other genes on in the Americas. This is all the more true when it is considered or off or change when and where in an organism they are that indigenous Neotropical farmers still live in—and more active or increase/decrease their effects, usually during early commonly did so in the recent past—small seminucleated ontogeny. This is the process of gene expression (e.g., Krug- and shifting communities. The idea that the appearance of lyak and Stern 2007). Novel phenotypes may then often result sedentary village life and all of its trappings should be the from reorganizations of existing genomes, not by the spread measure of when farming began in the Americas emerged or appearance of mutations, and new phenotypic variation from faulty comparisons with Near Eastern pre- and post- can rapidly arise in populations without a corresponding ge- agricultural trajectories, which were a product of ecological netic change. and demographic circumstances very different from those as- Developmental plasticity, recently given a complete treat- sociated with the beginnings of Neotropical food production ment in a major book (West-Eberhard 2003), refers to the (Piperno and Pearsall 1998). As also stressed by Iriarte (2007), inherent capacity of organisms to rapidly produce novel her- Vrydaghs and Denham (2007), and Denham (2011), agri- itable phenotypes through one of several available develop- cultural origins in different parts of the world should be stud- mental pathways in direct response to changes in their en- ied using the totality of evidence relevant to the region being vironments. This capacity should be particularly important investigated. in plants, which cannot simply get up and move to places more to their liking when physical and biotic conditions The How of Plant Domestication change and become less favorable. Gene expression during plant development orchestrates this process, resulting in dif- Phenotypic innovation depends on developmental inno- ferent phenotypic pathways to adulthood. The new pheno- vation, and developmental innovation spans a broader field types have the potential to become fixed (stabilized) by genetic of possibilities than does mutation alone. assimilation/accommodation if the new ecological conditions (Mary Jane West-Eberhard 2003:144) are maintained over multiple generations (West-Eberhard 2003). An important source of genetic variation that may Allowing for cryptic variants and novel phenotypes from enable the creation of novelties through developmental- new epistatic combinations to arise during domestication, mediated mechanisms is “cryptic genetic variation,” so named it is easy to imagine that maize was domesticated from because it is unexpressed in the phenotype of standing pop- teosinte. ulations and thus normally hidden from selection (e.g., Gib- (John Doebley 2004:56) son and Dworkin 2004; Lauter and Doebley 2002). However, it may be set in motion or “released” by interactions between genes (epistasis) or by perturbations from the external en- Evolutionary Development, Gene Expression, and vironment, rapidly resulting in new phenotypes. Lauter and Developmental Plasticity Doebley (2002) have shown that maize’s wild ancestor pos- Recent developments in evolutionary biology and archaeo- sesses a significant degree of cryptic variation that is associated botanical data gathering and interpretation indicate the need with important traits such as polystichous (many-rowed) cobs Piperno Plant Cultivation and Domestication in the New World Tropics S463 and nonshattering ears. Studies of this kind are needed in versification that was possible after domestication. Maize is a many other crop plants. classic example of a domesticated species that was capable of Crop evolution has usually been characterized using the adapting successfully to many different environmental cir- assumption that phenotypic change is driven by selection for cumstances, and prehistoric peoples across the New World rare mutants that are deleterious in wild plants or by selection created many different races from it, neither surprising in for new random mutations that appeared after cultivation light of how much variation and plasticity are present in its began. Associated single trait–single gene models are also wild ancestor. deeply rooted in domestication studies. However, it is in- Additional research will provide more detailed information creasingly recognized that (1) inheritance of domesticated about how gene expression and multiple possible develop- traits is often complex (simple Mendelian segregation is not mental trajectories influenced the creation of domesticated indicated) and does not involve straightforward human se- phenotypes, and this type of work will be important for an- lection on mutations that occurred as rare variants in wild imal domestication as well (e.g., Dobney and Larson 2006). ancestral or early cultivated populations; (2) some “domes- What has so far been largely neglected in domestication stud- tication genes” controlling crucial phenotypic attributes are ies and should be an important part of future investigations in fact regulatory genes, for example, in maize (tb1 and tga1, is the role of what West-Eberhard (2003) calls “environmental controlling plant architecture and the formation of naked induction” in creating new phenotypes from ancestral pop- grains, respectively), rice (sh4 and qSH1, controlling shatter- ulations (Gremillion and Piperno 2009a). This is a crux of ing), and tomato (fw2.2, underwriting fruit size increase); (3) the phenotypic plasticity argument as developed by West- these and other identified genes control developmental path- Eberhard—that alternative phenotypes can be induced di- ways in specific immature tissues and organs that lead to adult rectly by environmental change and subsequently genetically phenotypic variability; (4) phenotypic outcomes are signifi- assimilated/accommodated (genetically stabilized) under the cantly influenced by the external environment as well as by proper conditions. It is known, for example, that maizelike regulatory and other gene-gene interactions; and (5) consid- phenotypes in plant architecture can be induced in teosinte erable cryptic variation occurs in at least one crop progenitor, by environmental stresses such as lowered temperatures and teosinte, that provides preexisting genetic material that can light intensity (West-Eberhard 2003). Given that crop do- translate into multiple developmental trajectories (e.g., Doe- mestication first occurred shortly after the environmental per- bley 2004, 2006; Jones and Brown 2007; Lauter and Doebley turbations that marked the end of the Pleistocene, human 2002). environmental modification was significant starting during Doebley (2004) describes what happens in crosses between the early Holocene, and human care for/selection of novel teosinte and maize as a result of gene actions that alter the phenotypes in secure ecological niches—cultivated fields— trajectories of plant development: “The dichotomies of single would lead to the genetic assimilation necessary to stabilize versus paired spikelets, shattering versus nonshattering ears, the new phenotypes, the role of natural- and human-induced soft versus hard glumes, and two- versus multiranked ears environmental induction as a substitute for and/or comple- are striking when one compares maize and teosinte. However, ment to artificial selection in engineering some of the major in F2 families, these discrete classes blur into a continuum of phenotypic steps in crop evolution should be investigated (see phenotypes, and novel phenotypes and interactions appear” Gremillion and Piperno 2009a for possible examples relating (54). All of this means that the domestication process involved to Chenopodium seed attributes and maize branching and simple unconscious/conscious selection targeted at specific inflorescence architecture). mutations controlling discrete (single) traits and more com- plex plant-people-genetic-environmental interactions that of- Predomestication and Nondomestication Cultivation ten played out during plant ontogeny and involved humans “reconfiguring” (Lauter and Doebley 2002:341) the preexist- Newer and more highly detailed archeobotanical information ing pool of variability into new combinations. Present evi- indicates that rather than a rapid appearance of domesticated dence suggests that in some crops, the latter may have been plants shortly after cultivation began, protracted periods of more important than the former. Gene expression appears to predomestication cultivation (PDC) and even instances of have been very important, and it probably gave early culti- nondomestication cultivation (NDC) of cereals and pulses vators a great deal of phenotypic variation to work with in occurred in the Old World. PDC goes hand in hand with some crops and traits, especially as hybridization between what appears to have been a nonsimultaneous development crops and wild ancestors was initially common. This may have of the suite of traits that make up the domestication syndrome either slowed down or speeded up the domestication process. (e.g., Cohen 2011; Dillehay et al. 2007; Fuller 2007; Piperno Furthermore, preexisting variation for plasticity would have and Dillehay 2008; Weiss, Kislev, and Hartmann 2006; Weiss enabled crops to more quickly adapt to new ecological cir- and Zohary 2011; Willcox, Fortnite, and Herveux 2008; Zeder cumstances as they were brought from their natural habitats 2011). Some important implications from the accumulated into cultivated fields and dispersed out of their areas of origin, data are obvious. Domestication took longer after cultivation and it likely in part determined the amount of landrace di- was initiated and was a more complex process than was S464 Current Anthropology Volume 52, Supplement 4, October 2011 thought, involving different kinds and thresholds of (often Coming full circle back to the discussion that initiated this competing) natural and artificial selection pressures from re- section, the complex inheritance of and multifactorial influ- gion to region and cultivar to cultivar. The cultivation of ences on domestication traits may have played roles in slowing major crop species may have arisen independently more than domestication. Furthermore, the effects of gene expression once. Productive and stable food production could be based and possible environment-on-phenotype influences at the on- on morphologically wild plants. And if, as seems likely, PDC set of plant cultivation and after may have led to pools of plants spread, defining a single geographically localized area interesting phenotypic diversity in early crops that people as the cradle of origin for the domesticated plant would prove picked through and discouraged or encouraged over a pro- difficult because multiregional processes involving gene flow tracted period before the domesticated examples that we iden- were at work (Allaby, Fuller, and Brown 2008; Brown et al. tify archaeologically emerged. 2009; Jones and Brown 2007). I previously suggested that the commonplace focus on do- mestication as the preeminent event in human/plant rela- The Why of Food Production tionships is perhaps misplaced and that the crucial shift we may want to understand is the origins of plant cultivation, Human Behavioral Ecology and Agricultural Origins when people began to repeatedly sow and harvest plants in plots prepared for this purpose (see Piperno 2006a for a more Piperno and Pearsall (1998) and Piperno (2006a) offered an detailed discussion). Given the evidence brought to light more account of why and when Neotropical agricultural origins recently, I see no reason to change this view. Less attention occurred that is rooted in human behavioral ecology (HBE), has been paid to PDC and NDC issues in the New World, especially optimal foraging theory, and they suggested that but some highland and lowland data already suggest that they HBE models had potential relevance in other regions of the were components of early food production. At Guila´ Naquitz world. Explanations such as population pressure and changing Cave, highland Mexico, a type of morphologically wild runner social relationships (e.g., competitive feasting) used by some bean (Phaseolus spp.) was present between ca. 10,600 and investigators to account for agricultural beginnings elsewhere 8500 BP in quantities that suggested to the investigators that (Piperno and Pearsall 1998:10–18) did not and still do not people were artificially increasing their density by cultivating appear to be important in the Neotropics. Before discussing them (Flannery 1986). The plant was never domesticated. In the relationships between agricultural origins, HBE, and other the Zan˜a Valley, Peru, hulls of peanuts (the nuts themselves explanations for the transition to food production, I should did not survive) dated to 8500 BP do not exhibit some features reiterate my belief that the near synchroneity of food pro- of modern domesticated plants, nor do they conform to a duction origins in at least seven widely dispersed and eco- known wild species (Dillehay et al. 2007). Predomestication logically disparate regions of the world when or shortly after cultivation of this plant and its transport out of its area of the world’s fauna and flora were experiencing profound shifts origin in southern South America before it acquired the full driven by the end-Pleistocene ecological perturbations should package of domesticated characteristics is thus indicated. Pea- cause us to look for common underlying processes that may nut nut starch grains recovered from human teeth, on the have been influential in the transition wherever it occurred. other hand, are identical to those from modern varieties of Put at its simplest, it is difficult to believe that near-synchro- Arachis hypogaea and are unlike starch in modern wild pea- nous origins of such a major economic transformation were nuts closely related to A. hypogaea analyzed so far (Piperno a coincidence, an accidental convergence of disparate region- and Dillehay 2008). More work is needed on wild peanut ally specific underlying factors in widely dispersed and dif- starch, but the suite of phenotypic traits that characterize ferent social systems. This is not to say that identifying specific domesticated peanuts may not have developed all at once. aspects of Neolithic developments on a region-by-region basis In fact, as discussed in detail elsewhere (Piperno 2006a: is not important, only that we should in a deliberate way look 162–163; Piperno and Pearsall 1998), PDC and NDC sce- for underlying commonalities rather than immediately em- narios for agriculture, even if they were not called that, have phasizing putatively unique, supposedly causal local or re- long seemed to a number of investigators to be particularly gional variables that may turn out to be more apparent than well suited to the Neotropics (e.g., Harlan 1992; Harris 1989). real. Even today, many horticultural plots still contain morpho- The basic concepts and uses of HBE in archaeology are by logically wild plants that do not change their phenotypic char- now well known and discussed and will not be reviewed here acteristics even after many years of persistent cultivation. in any detail (for recent literature reviews, see Bird and Crops grown for their belowground organs (e.g., Calathea O’Connell 2006; Lupo 2007). In brief, HBE is concerned with and Dioscorea spp.) and many palms and other fruit trees are the adaptive plasticity of the human phenotype in response among the most stubborn taxa. Instances of PDC and NDC to variation in its particular ecological and social environ- should be sought for more widely in the New World by using ment. It assumes that behavior is shaped by evolutionary multifaceted archaeobotanical data sets and allied paleoeco- forces and asks “why certain patterns of behavior have logical evidence whenever possible. emerged and continue to persist” (Bird and O’Connell 2006: Piperno Plant Cultivation and Domestication in the New World Tropics S465

144). The emphasis on a flexible phenotype means that HBE’s influence on food procurement decisions (reviewed in Pip- engine for change resides in human decision making. HBE is erno and Pearsall 1998). The following lines of evidence were embedded in archaeological theory and empirical testing to used to argue that optimal foraging and the relative energetic a degree such that it is viewed along with dual inheritance efficiency of resource sets available to foragers during the late theory as a primary and distinct research tradition in evo- Pleistocene and early Holocene played a major role in Neo- lutionary archaeology (e.g., Shennen 2008). As such, it has tropical agricultural origins (details in Piperno 2006a; Piperno been and will continue to be a focus of discussions, both and Pearsall 1998). First, a large set of paleoecological data incisive and off the mark. Examples of the latter (Smith 2009; ranging now from the Central Balsas region of southwest Zeder and Smith 2009) make faulty claims that explanations Mexico to Bolivia (e.g., Burbridge, Mayle, and Killeen 2004; derived from HBE are analogous to covering law models or Piperno et al. 2007) indicates that the shift from foraging to that HBE ignores the active roles of foragers in altering en- food production began within contexts of rapid and signifi- vironments (see Gremillion and Piperno 2009b for additional cant changes of climate, vegetation, and fauna occurring at comments). Investigators in fact are drawn to HBE by its the close of the Pleistocene. Large increases in temperature, focused, hypothesis-driven, problem-solving approaches that, precipitation, and atmospheric CO2 levels resulted in vege- as mentioned, incorporate human agency and dynamic hu- tational transitions from savanna-like/thorny scrub growth to man/environmental relationships. These elements constitute seasonal types of tropical forest across the Neotropics. Many good science and are productive means for “eliminating prob- now-extinct megafauna were replaced by the smaller, fewer, lematic answers and identifying and pursuing more promising and more dispersed tropical forest fauna found in modern ones” (Bird and O’Connell 2006:171). environments. Although HBE has been used to study a broad array of Second, on the basis of robust sets of ethnographic and economic and social issues, “foraging theory” applications ecological data, these environmental perturbations would that address subsistence decisions are most common to this have significantly lowered the overall efficiency of food pro- point. At the heart of foraging theory is the assumption that, curement for hunters and gatherers in those zones where all things being equal, more efficient food procurement strat- open-land types of vegetation gave way to forest when the egies should be favored by natural selection over those less Pleistocene ended. For example, the big game and open-land efficient, a largely unquestioned premise in nonhuman animal plants that disappeared were almost certainly higher-ranked studies. The simplest version of foraging models is called the resources compared with those offered by the tropical forest, “optimal diet” model, or the “diet breadth” model (DBM). in which animals were far fewer and smaller, carbohydrates It employs a straightforward currency—energy—to measure were limited and spatially dispersed, and many plants were the costs and benefits of alternative resource sets and assumes toxic and required extensive processing before they were con- that humans will have a goal of optimizing the energetic sumed. Decreasing foraging efficiency as dietary breadth ex- returns of their subsistence labor. Recent research indicating panded to incorporate these low-ranked resources was prob- that spatial memory in women is preferentially engaged and ably an important selection pressure acting on human food most accurately expressed for calorie-rich foods further but- procurement strategies during the early Holocene. tresses the likelihood that natural selection shaped a human Third, on the basis of these factors, along with the costs proclivity for efficient plant foraging (New et al. 2007). As of plant cultivation estimated from modern small-scale trop- reviewed in detail elsewhere (Piperno 2006a; Piperno and ical farming, it appears that early Neotropical food-producing Pearsall 1998:15–18), there are a number of good reasons why strategies probably were more energetically efficient, not more the DBM can productively be applied to agricultural origins costly, than full-time foraging (see Piperno 2006a forade- and dispersals. For one thing, its sometimes counterintuitive tailed discussion of this issue for the tropical forest). Finally, predictions regarding resource choice bring new insight to following the DBM, people would have initiated the culti- issues such as the onset of “broad-spectrum” subsistence strat- vation of some plants when the net return from this strategy egies and the links between diet breadth, technology, and exceeded the return from full-time hunting and gathering. In resource intensification. Another strength of the model is its light of shifts in vegetational formations and available re- ability to elucidate from patterns in the archaeological record sources indicated by paleoecological records, the ca. 11,000– important processual questions relating to human economic 9000-BP time period should have been highly relevant for the decision making and its determinants vis-a`-vis the focal point initiation of food production. It should be noted that all of of energetic constraints. Importantly, in this regard, paleo- the above points are especially relevant to areas that today ecological data can serve as objective informants on past shifts support or would support in the absence of human distur- in the availability of different subsistence resources and as bance highly seasonal types of tropical forest, where the end- proxies of changing returns to labor from foraging, adding Pleistocene environmental perturbations would have im- valuable information that can be used in conjunction with or pacted foraging return rates most strongly. in the absence of archaeological subsistence records. Comparing these points with archaeological data (see A range of ethnographic information indicates that in the above; Piperno and Pearsall 1998), it appears that predictions Neotropical forest, as elsewhere, energetic efficiency is a major of optimal foraging and the DBM are well supported and S466 Current Anthropology Volume 52, Supplement 4, October 2011 effectively account for when and where food production HBE and Coevolution emerged in the lowland Neotropics. As has been discussed in There is little disagreement that coevolutionary forces and detail elsewhere, other issues such as why some plants and food production origins were entwined. As the general process not others out of the many available were singled out for unfolded, dependency of humans on certain plants increased human manipulation can be elucidated using HBE (Piperno and vice versa, and both the plants and people involved ex- 2006a; see below). The point made above about early farmers’ perienced fitness increases. However, when coevolution is op- energetic efficiency should be stressed. Although many in- erationalized at more detailed levels of analysis, discordance vestigators assume that early farmers everywhere experienced between the model and empirical archaeobotanical data be- diminishing returns to their labor, studies in the southwestern comes evident. This was initially discussed by Piperno and eastern United States and the Near East have also failed (2006a), and more examples can now be added. to support that notion when it was properly tested using The Cucurbitaceae and Marantaceae, each with important appropriate measures of resource costs and benefits (e.g., Bar- examples of early domesticated plants, are large families with low 2002; Gremillion 2004; Simms and Russell 1997). The many edible species that respond to a human presence by energetic efficiency of early farming compared with the last hunting and gathering should be reevaluated in other regions quickly colonizing disturbed areas near living places. This of the world. Especially in the cases where foraging theory makes them particularly suitable for testing the most complete appears to effectively account for the initiation of food pro- and prominent explication of how agricultural transitions ex- duction, it would not be surprising to see a revision of pre- emplify coevolution, that of Rindos (1984). The fruit rinds viously assumed cost/benefit equations for foraging versus of most Cucurbitaceae genera produce high numbers of rec- farming. ognizable phytoliths, and Marantaceae seeds and tubers are The DBM or combinations of DBM and other HBE theories similarly rich in diagnostic phytoliths. These durable micro- have recently been productively applied to predict under what fossils provide a record of exploitation unbiased by the pres- circumstances and how intensively foragers should become ervation factor. Marantaceae tubers also produce many di- farmers in a number of world regions (for other examples, agnostic starch grains, which would be expected to occur on see, e.g., Barlow 2002; Bird and O’Connell 2006; Gremillion stone tools that were used to process them. The sites of Xi- 2004; Kennett and Winterhalder 2006). In the eastern United huatoxtla, Mexico; San Isidro, El Recreo, and Pen˜a Roja, Co- States (Gremillion 2004), it was found that some of the pre- lombia; Siches, Ecuador; and Zan˜a Valley, Peru, can be added dictions of the DBM (e.g., that immediate resource returns to those from central Panama and Vegas, Ecuador, as examples provide the best measure of their utility) did not fit the ar- of where only Cucurbita (often along with bottle gourd) is chaeological data. In this case, the research led to a better recorded among plants from this family. In the sequences grasp of the variables most important in agricultural transi- from Tehuacan and Guila´ Naquitz, Mexico, where macro- tions (e.g., risk reduction, energetic considerations of pro- botanical preservation was excellent, one additional genus of cessing certain key resources with regard to the seasonal avail- the Cucurbitaceae, Apodanthera, occurred with Cucurbita and ability of others), and refinements in how HBE models should bottle gourd, but rarely, and it was considered a possible be applied in this regional circumstance. Students of HBE see modern intrusive. Students of foraging theory would im- this as one of its strongest features. If a good fit between mediately notice that the fruits and seeds of wild Cucurbita theory and data is not evident, then one moves on to find a are among the largest in the family and therefore would likely better-supported theoretical approach. provide higher return rates relative to other cucurbit taxa. A similar example of highly selective exploitation is found with regard to the Marantaceae. Maranta arundinacea and a HBE and Risk few species of Calathea, usually solely Calathea allouia,are HBE is a highly flexible research program by no means limited the only Marantaceae taxa represented in the tropical sites. to simple energetic efficiency models and currencies. For ex- There are numerous plants from a variety of other families ample, risk sensitivity, an often-used explanation for why for- with edible seeds and underground organs that should be agers turned to farming, is increasingly being incorporated visible in phytolith and starch grain records had they been into HBE research (see papers in Kennett and Winterhalder exploited early on with any persistency, but they are not re- 2006). It is evident that under a number of circumstances, corded. Moreover, just a few species of wild grasses out of energetically efficient diets are also risk-minimizing diets and the hundreds that were available to human foragers were ma- that attention to risk avoidance does not necessarily override nipulated in pre-Columbian Mexico: two species of Setaria attention to energetic efficiency. It is worth repeating here (Austin 2006) and Zea mays ssp. parviglumis (Balsas teosinte). that there is also considerable ethnographic evidence that Only the latter was demonstrably domesticated. A correlation among tropical foragers and horticulturists, food sharing in between grass domestication and seed size is also evident, as the context of extensive household exchange is the most im- teosinte has the largest grains of any annual Mexican grass. portant strategy for mitigating risk (see Piperno and Pearsall Therefore, Neotropical archaeobotanical data better sup- 1998:239–241 for a detailed discussion of these issues). port predictions from foraging theory that stress the impor- Piperno Plant Cultivation and Domestication in the New World Tropics S467 tance of deliberate directed human choice rather than the 2008. Genetic relationships among Arachis hypogaea L. (AABB) slowly unwinding reciprocal plant/human interactions in- and diploid Arachis species with AA and BB genomes. Genetic Resources and Crop Evolution 55:15–20. volving experimentation with numerous different taxa that Bird, D. W., and J. F. O’Connell. 2006. Behavioral ecology and ar- would be characteristic of coevolution. Efficiently procuring chaeology. Journal of Archaeological Research 14:143–188. food, especially in the tropical forest, would necessarily entail Bray, W. B., L. Herrera, M. Cardale Schrimpff, P. Botero, and J. G. paying serious attention to and deriving the most plant cal- Monsalve. 1987. The ancient agricultural landscape of Calima, ories, proteins, and fats from a relatively small set of the Colombia. In Pre-Hispanic agricultural fields in the Andean region. W. M. Denevan, K. Mathewson, and G. Knapp, eds. Pp. 443–481. available wild species. Long periods of experimentation and British Archaeological Reports International 359. Manchester: mutualistic relationships with a high diversity of species do British Archaeological Reports International. not appear to have been associated with squash, Marantaceae, Brown, T. A., M. K. Jones, W. Powell, and R. G. Allaby. 2009. The and grass domestication in part because energetic efficiency complex origins of domesticated crops in the Fertile Crescent. was probably a major determinant of subsistence decisions. Trends in Ecology & Evolution 24:103–109. Burbridge, R. E., F. E. Mayle, and T. J. Killeen. 2004. Fifty-thousand- year vegetation and climate history of Noel Kempff Mercado Na- Conclusions tional Park, Bolivian Amazon. Quaternary Research 61:215–230. Bush, M. B., M. C. Miller, P. E. DeOlivera, and P. A. Colinvaux. Discussions regarding the origins and dispersals of agriculture 2000. Two histories of environmental change and human distur- are increasingly based on robust empirical information as new bance in eastern lowland Amazonia. Holocene 10:543–553. and more precise data from all the contributing disciplines Bush, M. B., D. R. Piperno, and P. A. Colinvaux. 1989. A 6,000 year history of Amazonian maize cultivation. Nature 340:303–305. of study become available. With gene expression, phenotypic Carroll, S. B. 2005. Endless forms most beautiful: the new science of plasticity, and HBE among the important additions to the evo devo. New York: Norton. how and why of food production, the level of analysis has Castillo, N., and F. J. Aceituno. 2006. El bosque domesticado, el been raised to incorporate the phenotype of both the plants bosque cultivado: un proceso milenario en el valle medio del Rı´o and humans involved. Given the immense importance that Porce en el Noroccidente colombiano. Latin American Antiquity 17:561–578. gene expression and developmental plasticity are finding in Cavalier, I., C. Rodrı´guez, L. F. Herrera, G. Morcote, and S. Mora. evolutionary biology, it is likely that they will contribute much 1995. No solo de caza vive el hombre: ocupacio´n del bosque Ama- to our understanding of how people domesticated plants. zo´nico, holoceno temprano. In Ambito y ocupaciones tempranas de Human behavioral ecology similarly is likely to have growing la Ame´rica tropical. I. Cavalier and S. Mora, eds. Pp. 27–44. Bogota´, importance in agricultural origin studies. Microfossil data Colombia: Instituto Colombiano de Antropologı´a, Fundacio´n Er- igaie. have been shown to be essential lines of evidence in docu- Chandler-Ezell, K., D. M. Pearsall, and J. A. Zeidler. 2006. Root and menting the transition from foraging to farming in the Neo- tuber phytoliths and starch grains document manioc (Manihot tropics. As new sites are discovered and older ones reexam- esculenta), arrowroot (Maranta arundinacea), and lleren (Calathea ined, much more information can be expected on early sp.) at the Real Alto site, Ecuador. Economic Botany 60:103–120. agriculture in the New World’s tropical regions. Cohen, D. J. 2011. The beginnings of agriculture in China: a mul- tiregional view. Current Anthropology 52(suppl. 4):S273–S293. Cross, H., R. Lira Saade, and T. J. Motley. 2006. Origin and diver- sification of chayote. In Darwin’s harvest. T. J. Motley, N. Zerega, Acknowledgments and H. Cross, eds. Pp. 171–194. New York: Columbia University Press. I thank Doug Price and Ofer Bar-Yosef for inviting me to Denham, T. 2011. Early agriculture and plant domestication in New attend the Temozon conference. I learned much from the Guinea and Island Southeast Asia. Current Anthropology 52(suppl. 4):S379–S395. stimulating discussions and papers presented there. Dickau, R. 2010. Comparing subsistence and the transition to ag- riculture in western and central Panama through the integration References Cited of macrobotanical, microbotanical, and paleoenvironmental in- formation. In Integrating zooarchaeology and paleoethnobotany: a Aceituno, F. J., and N. Castillo. 2005. Mobility strategies in Co- consideration of issues, methods, and cases. A. M. VanDerwarker lombia’s middle mountain range between the early and middle and T. M. Peres, eds. Pp. 99–134. New York: Springer. Holocene. Before Farming 2005(2), art. 2. Dickau, R., A. J. Ranere, and R. G. Cooke. 2007. Starch grain evidence Allaby, R., D. Q. Fuller, and T. A. Brown. 2008. The genetic expec- for the preceramic dispersals of maize and root crops into tropical tations of a protracted model for the origins of domesticated crops. dry and humid forests of Panama. Proceedings of the National Proceedings of the National Academy of Sciences of the USA 105: Academy of Sciences of the USA 104:3651–3656. 13982–13986. Dillehay, T. D., H. H. Eling Jr., and J. Rossen. 2005. Irrigation canals Anderson, E. 1954. Plants, man, and life. London: Melrose. in the Peruvian Andes. Proceedings of the National Academy of Austin, D. F. 2006. Fox-tail millets (Setaria: Poaceae): abandoned Sciences of the USA 102:17241–17244. food in two hemispheres. Economic Botany 60:143–158. Dillehay, T. D., P. J. Netherly, and J. Rossen. 1989. Middle preceramic Barlow, R. K. 2002. Predicting maize agriculture among the Fremont: public and residential sites on the forested slope of the Western an economic comparison of farming and foraging in the American Andes, Northern Peru. American Antiquity 54:733–759. Southwest. American Antiquity 67:65–88. Dillehay, T. D., C. Ramirez, M. Pino, M. B. Collins, J. Rossen, and Bertoti de Cunha, F., P. M. Nobile, A. A. Hoshino, A. Moretzsohn, J. D. Pino-Navarro. 2008. Monte Verde: seaweed, food, medicine, M. de Carvalho, C. A. Lopes, R. Catalina, and M. A. Gimenes. and the peopling of South America. Science 320:784–789. S468 Current Anthropology Volume 52, Supplement 4, October 2011

Dillehay, T. D., J. Rossen, T. C. Andres, and D. E. Williams. 2007. bridization and the origin of crops. Proceedings of the National Preceramic adoption of peanut, squash, and cotton in northern Academy of Sciences of the USA 104:14389–14394. Peru. Science 316:1890–1893. Iriarte, J. 2007. New perspectives on plant domestication and the Dillehay, T. D., J. Rossen, P. Netherly, and A. Karathanasis. 2010. development of agriculture in the New World. In Rethinking ag- Early Holocene coca production in northern Peru. Antiquity 84: riculture. T. Denham, J. Iriarte, and L. Vrydaghs, eds. Pp. 167– 939–953. 188. Walnut Creek, CA: Left Coast. Dobney, K., and G. Larson. 2006. Genetics and animal domestication: ———. 2009. Narrowing the gap: exploring the diversity of early new windows on an elusive process. Journal of Zoology 269:261– food-production economies in the Americas. In Rethinking the 271. origins of agriculture. M. N. Cohen, ed. Special issue, Current An- Doebley, J. 2004. The genetics of maize evolution. Annual Review of thropology 50:677–680. Genetics 38:37–59. Iriarte, J., I. Holst, O. Marozzi, C. Listopad, E. Alonso, A. Rinder- ———. 2006. Unfallen grains: how ancient farmers turned weeds knecht, and J. Montan˜a. 2004. Evidence for cultivar adoption and into crops. Science 312:1318–1319. emerging complexity during the mid-Holocene in the La Plata Erickson, D. L., B. D. Smith, A. C. Clarke, D. H. Sandweiss, and N. Basin, Uruguay. Nature 432:614–617. Tuross. 2005. An Asian origin for a 10,000 year-old domesticated Jones, M., and T. Brown. 2007. Selection, cultivation, and repro- plant in the Americas. Proceedings of the National Academy of ductive isolation. In Rethinking agriculture. T. Denham, J. Iriarte, Sciences of the USA 102:18315–18320. and L. Vrydaghs, eds. Pp. 36–49. Walnut Creek, CA: Left Coast. Flannery, K. V. 1986. The research problem. In Guila´ Naquitz: archaic Kaplan, L., and T. F. Lynch. 1999. Phaseolus (Fabaceae) in archae- foraging and early agriculture in Oaxaca, Mexico. K. V. Flannery, ology: AMS radiocarbon dates and their significance for pre-Co- ed. Pp. 3–18. Orlando, FL: Academic Press. lumbian agriculture. Economic Botany 53:262–272. Fuller, D. 2007. Contrasting patterns in crop domestication and do- Kennett, D. J., and B. Winterhalder, eds. 2006. Behavioral ecology and mestication rates: recent archaeological insights from the Old the transition to agriculture. Berkeley: University of California Press. World. Annals of Botany 100:903–924. Kruglyak, L., and D. L. Stern. 2007. An embarrassment of switches. Gibson, G., and I. Dworkin. 2004. Uncovering cryptic genetic var- Science 317:758–759. iation. Nature Reviews Genetics 5:681–690. Kwak, M., J. A. Kami, and P. Gepts. 2009. The putative Mesoamerican Gnecco, C., and J. Aceituno. 2006. Early humanized landscapes of domestication center of Phaseolus vulgaris is located in the Lerma- northern South America. In Paleoindian archaeology: a hemispheric Santiago Basin of Mexico. Crop Science 49:554–563. perspective. J. E. Morrow and C. Gnecco, eds. Pp. 86–104. Gaines- Lauter, N., and J. Doebley. 2002. Genetic variation for phenotypically ville: University of Florida Press. invariant traits detected in teosinte: implications for the evolution Gremillion, K. J. 2004. Seed processing and the origins of food pro- of novel forms. Genetics 160:333–342. duction in eastern North America. American Antiquity 69:215– Le´otard, G., A. Duputie´, F. Kjellberg, E. J. P. Douzery, C. Debain, J.- 233. J. de Granville, and D. McKey. Forthcoming. Phylogeography and Gremillion, K. J., and D. R. Piperno. 2009a. Human behavioral ecol- the origin of cassava: new insights from the northern rim of the ogy, phenotypic (developmental) plasticity, and agricultural ori- Amazon basin. Molecular Phylogenetics and Evolution. gins: insights from the emerging evolutionary synthesis. In Re- Lupo, K. D. 2007. Evolutionary foraging models in zooarchaeological thinking the origins of agriculture. M. N. Cohen, ed. Special issue, analysis: recent applications and future challenges. Journal of Ar- Current Anthropology 50:615–619. chaeological Research 15:143–189. Matsuoka, Y., Y. Vigouroux, M. M. Goodman, J. Sanchez, E. Buckler, ———. 2009b. Parallel causation does not make a prime mover and and J. Doebley. 2002. A single domestication for maize shown by historical particulars do not make an explanation: response. In multilocus microsatellite genotyping. Proceedings of the National Rethinking the origins of agriculture. M. N. Cohen, ed. Special issue, Academy of Sciences of the USA 99:6080–6084. Current Anthropology 50:709–711. Milla, S. R., T. G. Isleib, and H. T. Stalker. 2005. Taxonomic rela- Griffith, M. P. 2004. The origins of an important cactus crop, Opuntia tionships among Arachis sect. Arachis species as revealed by AFLP ficus-indica (Cactaceae): new molecular evidence. American Journal markers. Genome 48:1–11. of Botany 91:1915–1921. Miller, A., and B. Schaal. 2005. Domestication of Mesoamerican cul- Harlan, J. R. 1971. Agricultural origins: centers and non-centers. tivated fruit tree, Spondias purpurea. Proceedings of the National Science 174:468–474. Academy of Sciences of the USA 102:12801–12806. ———. 1992. Crops and man. 2nd edition. Madison, WI: American Mora, S. 2003. Early inhabitants of the Amazonian tropical rain forest: Society of Agronomy and Crop Science Society of America. a study of humans and environmental dynamics. Latin American Harris, D. R. 1972. The origins of agriculture in the tropics. American Archaeology Reports, vol. 3. Pittsburgh: University of Pittsburgh. Scientist 60:180–193. Mora, S., L. F. Herrera, I. Cavelier, and C. Rodrı´guez. 1991. Cultivars, ———. 1989. An evolutionary continuum of people-plant interac- anthropic soils and stability. Latin American Archaeology Reports, tion. In Foraging and farming: the evolution of plant exploitation. vol. 3. Pittsburgh, PA: University of Pittsburgh. D. R. Harris and G. Hillman, eds. Pp. 11–26. London: Unwin Motta-Aldana, J. R., M. L. Serrano-Serrano, J. H. Torres, G. Castillo- Hyman. Villamizar, D. G. Debouck, and M. I. Chaco´n. Forthcoming. Do- Herrera, L., W. Bray, M. Cardale de Schrimpff, and P. Botero. 1992. mestication patterns in wild lima beans (Phaseolus lunatus L.) from Nuevas fechas de radiocarbono para el preceramico en la cordillera the Americas. In Harlan II: biodiversity in agriculture: domestication, occidental de Colombia. In Archaeology and environment in Latin evolution, and sustainability. A. Damania and P. Gepts, eds. Cam- America. O. R. Ortiz-Troncoso and T. Van Der Hammen, eds. Pp. bridge: Cambridge University Press. 145–164. Amsterdam: Universiteit van Amsterdam. New, J., M. M. Krasnow, D. Truxaw, and S. J. C. Gaulin. 2007. Spatial Holst, I., J. E. Moreno, and D. R. Piperno. 2007. Identification of adaptations for foraging: women excel and calories count. Pro- teosinte, maize, and Tripsacum in Mesoamerica by using pollen, ceedings of the Royal Society B 274:2679–2684. starch grains, and phytoliths. Proceedings of the National Academy Olsen, K. M. 2002. Population history of Manihot esculenta (Eu- of Sciences of the USA 104:17608–17613. phorbiaceae) inferred from nuclear DNA sequences. Molecular Hughes, C. E., R. Govindarajulu, A. Robertson, D. L. Filer, S. A. Ecology 11:901–911. Harris, and C. Donovan Bailey. 2007. Serendipitous backyard hy- Olsen, K. M., and B. Schaal. 1999. Evidence on the origin of cassava: Piperno Plant Cultivation and Domestication in the New World Tropics S469

phylogeography of Manihot esculenta. Proceedings of the National Piperno, D. R., A. J. Ranere, I. Holst, J. Iriarte, and R. Dickau. 2009. Academy of Sciences of the USA 96:5586–5591. Starch grain and phytolith evidence for early ninth millennium Pearsall, D. M. 1978. Phytolith analysis of archaeological soils: evi- BP maize from the Central Balsas River Valley, Mexico. Proceedings dence for maize cultivation in Formative Ecuador. Science 199: of the National Academy of Sciences of the USA 106:5019–5024. 177–178. Piperno, D. R., and K. Stothert. 2003. Phytolith evidence for early ———. 2000. Paleoethnobotany: a handbook of procedures. 2nd edi- Holocene Cucurbita domestication in southwest Ecuador. Science tion. San Diego, CA: Academic Press. 299:1054–1057. ———. 2003. Plant food resources of the Ecuadorian Formative: an Plowman, T. 1984. The origin, evolution, and diffusion of coca, Er- overview and comparison to the Central Andes. In Archaeology of ythroxylum sp., in South and Central America. In Pre-Columbian Formative Ecuador. J. S. Raymond and R. L. Burger, eds. Pp. 213– plant migration. D. Stone, ed. Pp. 126–163. Cambridge, MA: Pea- 257. Washington, DC: Dumbarton Oaks. body Museum, Harvard University. Pearsall, D. M., K. Chandler-Ezell, and A. Chandler-Ezell. 2003. Iden- Pohl, M. E. D., D. R. Piperno, K. O. Pope, and J. G. Jones. 2007. tifying maize in Neotropical sediments and soils using cob phy- Microfossil evidence for pre-Columbian maize dispersals in the toliths. Journal of Archaeological Science 30:611–627. Neotropics from San Andre´s, Tabasco, Mexico. Proceedings of the Pearsall, D. M., K. Chandler-Ezell, and J. A. Zeidler. 2004. Maize in National Academy of Sciences of the USA 104:6870–6875. ancient Ecuador: results of residue analysis of stone tools from the Pope, K. O., M. E. D. Pohl, J. G. Jones, D. L. Lentz, C. Von Nagy, Real Alto site. Journal of Archaeological Science 31:423–442. F. J. Vega, and I. R. Quitmyer. 2001. Origin and environmental Perry, L., R. Dickau, S. Zarillo, I. Holst, D. M. Pearsall, D. R. Piperno, setting of ancient agriculture in the lowlands of Mesoamerica. M. J. Berman, et al. 2007. Starch fossils and the domestication and Science 292:1370–1373. dispersal of chile peppers (Capsicum sp.) in the Americas. Science Ranere, A. J., and R. G. Cooke. 2003. Late glacial and early Holocene 315:986–988. occupation of Central American tropical forests. In Under the can- Pigliucci, M. 2005. Expanding evolution. Nature 435:565–566. opy: the archaeology of tropical rain forests. J. Mercader, ed. Pp. Piperno, D. R. 1988. Phytolith analysis: an archaeological and geological 219–248. Piscataway, NJ: Rutgers University Press. perspective. San Diego, CA: Academic Press. Ranere, A. J., D. R. Piperno, I. Holst, R. Dickau, and J. Iriarte. 2009. ———. 1990. Aboriginal agriculture and land usage in the Amazon Preceramic human occupation of the Central Balsas Valley, Mexico: Basin, Ecuador. Journal of Archaeological Science 17:665–677. cultural context of early domesticated maize and squash. Proceed- ———. 2006a. A behavioral ecological approach to the origins of ings of the National Academy of Sciences of the USA 106:5014–5018. plant cultivation and domestication in the seasonal tropical forests Raymond, J. S. 2008. The process of sedentism in northwestern South of the New World. In Foraging theory and the transition to agri- America. In Handbook of South American archaeology. H. Silver- culture. D. Kennett and B. Winterhalder, eds. Pp. 137–166. Berke- man and W. H. Isbell, eds. Pp. 79–90. New York: Springer. ley: University of California Press. Richardson, J. B., III. 1973. The preceramic sequence and the Pleis- ———. 2006b. Identifying manioc (Manihot esculenta Crantz) and tocene and post-Pleistocene climate of northwest Peru. In Vari- other crops in pre-Columbian tropical America through starch ations in anthropology: essays in honor of John C. McGregor.D.W. grain analysis: a case study from Panama. In Documenting do- Lathrap and J. Douglas, eds. Pp. 199–211. Urbana: Illinois Ar- mestication: new genetic and archaeological paradigms.M.Zeder, chaeological Survey. E. Emschwiller, D. Bradley, and B. S. Smith, eds. Pp. 46–67. Berke- Rindos, D. 1984. The origins of agriculture: an evolutionary perspective. ley: University of California Press. Orlando, FL: Academic Press. ———. 2006c. Phytoliths: a comprehensive guide for archaeologists Robledo, G., G. I. Lavia, and G. Seijo. 2009. Species relations among and paleoecologists. Lanham, MD: AltaMira. wild Arachis species with the A genome as revealed by FISH map- ———. 2009. Identifying crop plants with phytoliths (and starch ping of rDNA loci and heterochromatin detection. Theoretical and grains) in Central and South America: a review and an update of Applied Genetics 118:1295–1307. the evidence. Quaternary International 193:146–159. Rodrigues, D. P., S. A. Filho, and C. R. Clement. 2004. Molecular ———. Forthcoming. New archaeobotanical information on early marker-mediated validation of morphologically defined landraces cultivation and plant domestication involving microplant remains. of Pejibaye (Bactris gasipaes) and their phylogenetic relationships. In Harlan II: biodiversity in agriculture: domestication, evolution, Genetic Resources and Crop Evolution 51:871–882. and sustainability. A. Damania and P. Gepts, eds. Cambridge: Cam- Rosas, R. V., T. Andres, and M. Nee. 2004. The Goldman cucurbit bridge University Press. collecting expedition in Peru, 2004. http://www.cucurbit.org/Peru/ Piperno, D. R., K. Clary, R. G. Cooke, A. J. Ranere, and D. Weiland. PeruReport.html. 1985. Preceramic maize in central Panama: phytolith and pollen Salgado, H. 1995. El precera´mico en el can˜o´n del rı´o Calima, cor- evidence. American Anthropologist 87:871–878. dillera occidental. In A´mbito y ocupaciones tempranas de la Ame´rica Piperno, D. R., and T. D. Dillehay. 2008. Starch grains on human Tropical. I. Cavelier and S. Mora, eds. Pp. 73–82. Bogota´: Fun- teeth reveal early broad crop diet in northern Peru. Proceedings of dacio´n Erigaie, Instituto Colombiano de Antropologı´a. the National Academy of Sciences of the USA 105:19622–19627. Sandweiss, D. H. 2003. Terminal Pleistocene through mid-Holocene ———. Forthcoming. Agricultural origins and their social impli- archaeological sites as paleoclimatic archives for the Peruvian coast. cations in South America. In The Cambridge prehistory. C. Renfrew Palaeogeography, Palaeoclimatology, Palaeoecology 194:23–40. and P. Bahn, eds. Cambridge: Cambridge University Press. Sanjur, O., D. R. Piperno, T. C. Andres, and L. Wessell-Beaver. 2002. Piperno, D. R., J. E. Moreno, I. Holst, M. Lachniet, J. G. Jones, A. Phylogenetic relationships among domesticated and wild species J. Ranere, and R. Castanzo. 2007. Late Pleistocene and Holocene of Cucurbita (Cucurbitaceae) inferred from a mitochondrial gene: environmental history of the Iguala Valley, central Balsas watershed implications for crop plant evolution and areas of origin. Pro- of Mexico. Proceedings of the National Academy of Sciences of the ceedings of the National Academy of Sciences of the USA 99:535– USA 104:11874–11881. 540. Piperno, D. R., and D. M. Pearsall. 1998. The origins of agriculture Sauer, C. O. 1950. Cultivated plants of South and Central America. in the lowland Neotropics. San Diego, CA: Academic Press. In Handbook of South American Indians. J. Steward, ed. Pp. 487– Piperno, D. R., A. J. Ranere, I. Holst, and P. Hansell. 2000. Starch 543. Bureau of American Ethnology, Bulletin 143, vol. 6. Wash- grains reveal early root crop horticulture in the Panamanian trop- ington, DC: U.S. Government Printing Office. ical forest. Nature 407:894–897. Shady Solis, R., J. Haas, and W. Creamer. 2001. Dating Caral, a S470 Current Anthropology Volume 52, Supplement 4, October 2011

preceramic site in the Supe Valley on the central coast of Peru. Weiss, E., M. E. Kislev, and A. Hartmann. 2006. Autonomous cul- Science 292:723–726. tivation before domestication. Science 312:1608–1610. Shennan, S. 2008. Evolution in archaeology. Annual Review of An- Weiss, E., and D. Zohary. 2011. The Neolithic Southwest Asian foun- thropology 37:75–91. der crops: their biology and archaeobotany. Current Anthropology Simms, S., and K. W. Russell. 1997. Bedouin hand harvesting of 52(suppl. 4):S237–S254. wheat and barley: implications for early cultivation in South- Wessel-Beaver, L. 2000. Evidence for the center of diversity of Cu- western Asia. Current Anthropology 38:696–702. curbita moschata in Colombia. Cucurbit Genetics Cooperative Re- Smith, B. D. 2009. Core conceptual flaws in human behavioral ecol- port 23:54–55. ogy. Communicative and Integrative Biology 2:533–534. West-Eberhard, M. J. 2003. Developmental plasticity and evolution. Stothert, K. 1985. The preceramic Las Vegas culture of coastal Ec- Oxford: Oxford University Press. uador. American Antiquity 50:613–637. Westengen, O. T., Z. Huama´n, and M. Heun. 2005. Genetic diversity Stothert, K., D. R. Piperno, and T. C. Andres. 2003. Terminal Pleis- and geographic pattern in early South American cotton domes- tocene/early Holocene human adaptation in coastal Ecuador: the tication. Theoretical and Applied Genetics 110:392–402. Willcox, G., S. Fornite, and L. Herveux. 2008. Early Holocene cul- Las Vegas evidence. Quaternary International 109–110:23–43. tivation before domestication in northern Syria. Vegetational His- van Heerwaarden, J., J. Doebley, W. H. Briggs, J. C. Glaubitz, M. M. tory and Archaeobotany 17:313–325. Goodman, J. J. Sa´nchez Gonza´lez, J. Ross-Ibarra, B. S. Gaut, and Zarillo, S., D. M. Pearsall, J. S. Raymond, M. A. Tisdale, and D. J. L. Eguiarte. 2011. Genetic signals of origin, spread, and intro- Quon. 2008. Directly dated starch residues document Early For- gression in a large sample of maize landraces. Proceedings of the mative maize (Zea mays L.) in tropical Ecuador. Proceedings of the National Academy of Sciences of the USA 108:1088–1092. National Academy of Sciences of the USA 105:5005. Vrydaghs, L., and T. Denham. 2007. Rethinking agriculture: intro- Zeder, M. A. 2011. The origins of agriculture in the Near East. Current ductory thoughts. In Rethinking agriculture. T. Denham, J. Iriarte, Anthropology 52(suppl. 4):S221–S235. and L. Vrydaghs, eds. Pp. 1–15. Walnut Creek, CA: Left Coast. Zeder, M. A., and B. D. Smith. 2009. A conversation on agricultural Weir, G. H., and J. P. Dering. 1986. The lomas of Paloma: human- origins: talking past each other in a crowded room. Current An- environment relations in a central Peruvian fog oasis: archaeo- thropology 50:681–691. botany and palynology. In Andean archaeology: papers in memory Zhang, D. P., J. Cervantes, Z. Huama´n, E. Carey, and M. Ghislain. of Clifford Evans. R. Matos, S. A. Turpin, and H. H. Eling Jr., eds. 2000. Assessing genetic diversity of sweet potato (Ipomoea batatas Pp. 18–44. Los Angeles: Institute of Archaeology, University of (L.) Lam.) using cultivars from tropical America using AFLP. Ge- California. netic Resources and Crop Evolution 47:659–665. Current Anthropology Volume 52, Supplement 4, October 2011 S471

The Cultural Context of Plant Domestication in Eastern North America

by Bruce D. Smith

The timing and sequence of the independent domestication of indigenous eastern North American seed plants (Cucurbita pepo, Helianthus annuus, Iva annua, Chenopodium berlandieri) and the sub- sequent development of a crop complex are discussed within a broader environmental and cultural context. The settlements that have yielded the earliest record of eastern domesticates are all small and situated in resource-rich lower-order river valley corridors within oak-savannah and oak-hickory forest regions. Well-preserved floral and faunal assemblages indicate continued substantial reliance on a wide range of wild species with no evidence of resource depletion. Similarly, there is no indication of landscape packing in terms of high site density in these resource-rich river valleys, calling into question developmental models of domestication and agricultural origins that rely on population pressure or resource imbalance as causal factors.

primary among these being that the spatial and temporal Our archaeology is that of hunting and gathering groups, parameters of two developmental milestones are reasonably no more complex than a simple band or bands . . . that well established: first, the initial domestication of the region’s late in the Archaic period began to experiment with simple four domesticated seed plants and, second, the formation of gardening. a distinctive crop complex based on these indigenous do- (Kay 1983:64) mesticates. This well-defined temporal and spatial framework Introduction for initial plant domestication and the formation of a dis- tinctive crop complex provides a good foundation for a regional-scale consideration of the general environmental and Over the past 2 decades, geneticists and archaeologists have cultural context within which domestication and the initial made impressive advances worldwide in documenting the transition to food production occurred. In this article I briefly temporal and spatial context of domestication of a growing outline the temporal and spatial parameters of initial plant variety of different species of plants and animals, as well as domestication and the subsequent establishment of a dis- identifying their wild progenitors (Zeder et al. 2006). When combined at regional scales of analysis, these species-level tinctive crop complex in eastern North America and then initiatives have in turn significantly increased our understand- employ this spatial-temporal framework to focus on a con- ing of the pace and sequence of domestication of different sideration of the general environmental setting and socio- species within the world’s independent centers of domesti- cultural characteristics of the societies that made the initial cation and agricultural origin, including eastern North Amer- transition from hunting and gathering to food production in ica (Smith 2006a). the region by looking at their levels of technology, settlements Even though eastern North America temporally lags behind and subsistence economies, and regional-scale networks of many of the world’s other centers of agricultural origin and interaction. only witnessed the domestication of a few species that remain important in today’s world economies (e.g., sunflower and The Temporal and Spatial Context of Cucurbita pepo squash), it nonetheless provides an important comparative case study situation in a number of respects, Plant Domestication in Eastern North America

Based on several morphological changes (e.g., seed size in- Bruce D. Smith is Curator of North American Archaeology, Department of Anthropology, National Museum of Natural History, crease, reduction in seed coat thickness) that are associated Smithsonian Institution (P.O. Box 37012, Washington, DC 20013- with the “adaptive syndrome of domestication” and that have 7012, U.S.A. [[email protected]]). This paper was submitted 13 XI 09, been documented in well-preserved seed specimens recovered accepted 27 XII 10, and electronically published 11 VIII 11. from a number of archaeological sites in eastern North Amer-

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0020$10.00. DOI: 10.1086/659645 S472 Current Anthropology Volume 52, Supplement 4, October 2011

Table 1. Earliest occurrence of indigenous domesticated seed crops in eastern North America

Age (AMS-calibrated calendar years BP) Age (radiocarbon Laboratory Archaeological site Domesticated plant species Intercept 1j age range years BP) sample no. and provenience (b47293 Phillips Spring (unit K2 75 ע Pepo squash (Cucurbita pepo ssp. ovifera) 5025 5290–4870 4440 (b45050 Hayes (level 14 60 ע Sunflower (Helianthus annus) 4840 4860–4830 4265 (b216463 Napoleon Hollow (feature 20 40 ע Marshelder (Iva annua) 4400 4420–4290 3920 Chenopod (Chenopodium berlandieri): (b253114 Riverton (feature 1 40 ע Naked” 3800 3830–3700 3490“ (b11348 Cloudsplitter (F. S. 1361 150 ע Thin testa 3700 3900–3490 3450 (b253117 Riverton (feature 8A 40 ע Thin testa 3690 3810–3640 3440 (b11347 Newt Kash (El 1114 150 ע Thin testa 3640 3840–3460 3400

ica, at least four indigenous seed-bearing plants were brought 6 centuries, at 3800 BP. Earlier evidence for chenopod do- under domestication in the region over a span of about 1,300 mestication will quite likely be recovered in the future, how- years, from 5000 to 3700 BP (table 1; fig. 1). Maize (Zea ever, and I expect that initial domestication of the four eastern mays), the first Mesoamerican domesticate to reach eastern domesticates identified so far in terms of morphological North America, did not arrive for another 1,200 years, at ca. changes as well as perhaps others will fall into the 5-century 200 BC (Riley et al. 1994).1 In addition to these four seed span between 5000 and 4500 BP. Rather than restricting this plants that exhibit morphological changes, three other species discussion to the societies and settlements of this 500-year that lack such changes have also been identified as probably period, however, it is worthwhile to expand the time span of having been the subject of deliberate planting and harvesting consideration another 11 centuries, from 5000 to 3400 cali- of stored seed stock based on their relative abundance in seed brated calendar years BP, in order to both increase the number assemblages of this time period: erect knotweed (Polygonum of archaeological sites that have yielded domesticates and in- erectum), little barley (Hordeum pusillum), and maygrass clude several sites that provide substantial information. This (Phalaris caroliniana; Smith 2006a; Yarnell 2004). time span also correlates closely with the cultural period often The genetic and archaeological evidence for the domesti- termed the “Late Archaic” (Emerson, McElrath, and Fortier cation of these eastern seed plants has been discussed at length 2009; Sassaman and Anderson 1996; Smith 1986). by a number of researchers over the past several decades The currently available evidence of indigenous domesti- (Reisberg and Harter 2006; Smith 2006a, 2006b; Yarnell 2004), cated plants for the 5000–3400-BP time period has been re- but in comparison, relatively little attention has been given covered from a total of seven archaeological sites, all of which to what the societies were like that brought these plants under are located within the resource-rich oak-savannah and oak- domestication. The general temporal and spatial parameters hickory forest regions of eastern North America (Delcourt of initial domestication and the emergence of early low-level and Delcourt 1981; fig. 1).2 These small settlements, which food production economies in the region, as indicated in table are scattered across five states and are separated by up to 1 and figure 1, provide a good starting point for such a con- 1,000 km, represent the primary data sets for considering the sideration. early history of low-level food production in the region. In- As shown in table 1, morphological changes indicating the formation regarding these seven sites, however, is relatively initial domestication of three of the four eastern crop plants limited, and an expanded consideration of a larger sample of are first documented in assemblages dated to between 5000 and 4400 BP, with the earliest evidence for domestication of 2. Of the seven archaeological sites in eastern North America that have the fourth species (Chenopodium) not appearing for another provided early evidence of domesticated plants, only six are profiled here in detail. Newt Kash Hollow is not discussed, given the limited amount 1. Discussions of the independent domestication of plants in different of information that is currently available regarding the site. The domes- regions of the New World rarely consider the bottle gourd (Lagenaria ticated Chenopodium berlandieri documented for Newt Kash (table 1) siceraria) because it was a “utilitarian” domesticate rather than a food consisted of thin-testa fruits extracted from an unprovenienced but di- plant. Highly prized for its strong, lightweight fruits that made excellent rectly dated human coprolite from the site (Smith and Cowan 1987). A containers and vessels of various shapes and sizes, the bottle gourd was number of other sites that date to this time period and that have been carried from Asia to the Americas either by ocean currents or more likely identified as having produced domesticates (Yarnell 2004)—for example, by Paleoindian colonists (along with another utilitarian domesticate, Jernigan II, Peter Cave, and Iddens—are not included in this discussion Canis familiaris), reaching the New World by 10,000 BP (Erickson et al. either because the species in question was the bottle gourd, a utilitarian 2005). It reached eastern North America by 7300 BP (Doran, Dickel, and domesticate with deep time depth, or because the specimens recovered Newsom 1990) and is frequently recovered in association with eastern did not exhibit morphological markers of domesticated status (e.g., thin- North American domesticates, particularly before the development of rind wild Cucurbita gourd rind fragments and maygrass, sunflower, and ceramic vessels in the region. chenopod seeds that fall into the wild size range). Smith Plant Domestication in Eastern North America S473

Figure 1. Oak-savannah and oak-hickory forest regions of eastern North America at ca. 5000 BP (after Delcourt and Delcourt 1981). The Late Archaic settlements that have provided the earliest evidence of domes- ticated plants in eastern North America are shown.

Late Archaic period sites from the oak-savannah and oak- ervation of plant remains in these sites is excellent because hickory forest zones will provide a broader and clearer general of an absence of soil moisture in the archaeological deposits. profile of the societies that initially domesticated plants in Between them, Phillips Spring, Napoleon Hollow, Riverton, eastern North America. The numerous small-scale societies and Hayes are all located in river valley settings, with pres- that populated the blank portions of figure 1 were similar in ervation of early domesticates resulting from either carbon- many respects to those represented by the seven sites shown, ization or, in the case of Phillips Spring, wet rather than dry and while excavations at many of these other sites have not soil conditions. yielded domesticates, they have produced considerable ma- The recovery of early domesticates from both riverine and terial cultural evidence of technology, subsistence and settle- more upland settings has engendered a very interesting and ment patterns, and likely mechanisms and networks of ongoing discussion regarding the specific habitat setting of broadscale regional interaction (Jefferies 1996). initial plant domestication in eastern North America: is the presence of early domesticates in more upland sites largely a The Late Archaic Settlements That Have function of good preservation of plant remains in dry rock- Yielded the Earliest Evidence of shelters, or does it indicate in fact that the initial domesti- cation of Cucurbita gourds, marshelder, and chenopod—all Domesticates early successional floodplain colonizers (see discussion be- Scattered across the interior riverine midlatitudes of the oak- low)—occurred outside of their natural riverine habitats? At savannah and oak-hickory forest zones between 5000 and the present time the available archaeological evidence best 3400 BP, the seven sites shown in figure 1 are often thought supports a river valley context of initial domestication, with of as comprising two distinct groups based on their environ- the earliest evidence for all four eastern domesticates recov- mental setting and the factors that contributed to the pres- ered from sites in stream valley settings. Rather than being ervation of archaeobotanical specimens. Marble Bluff at the situated along the main trunk of the Mississippi Valley, how- western end of this distribution of sites and Cloudsplitter and ever, or any of its major tributaries (e.g., the Missouri, Ohio, Newt Kash at the eastern end are dry rockshelters situated in and White rivers, etc.), all seven of the sites discussed here rugged upland settings in proximity to small streams. Pres- are instead situated in proximity to smaller primary (Illinois), S474 Current Anthropology Volume 52, Supplement 4, October 2011 secondary (Wabash, Buffalo), and tertiary (Pomme de Terre, reaching 47 inches in depth and producing a “full range” of Duck, Red) tributaries of the Mississippi. In addition, al- artifact types and 405 catalog numbers (Fritz 1986:27, 1997), though the seven sites that have produced the earliest do- the long time span of likely occupation of the site, the low mesticates all share a generally similar river valley setting, they frequency of temporally diagnostic material culture, and the are dispersed across a broad area of the interior midlatitudes. often disturbed nature of Ozark rockshelter deposits in gen- While acknowledging the challenges inherent in determin- eral have made it difficult to either distinguish Archaic ma- ing the geographical range of wild progenitor populations at terials and features from those of subsequent occupational 5000 BP, genetic profiling of modern wild populations of both episodes or to accurately characterize the nature of the Late Cucurbita pepo var. ovifera squash and sunflower indicate that Archaic occupation of the settlement. In addition, most of initial domestication of these two species probably occurred the domesticated sunflower, chenopod, marshelder, and C. along the western end of the oak-savannah and oak-hickory pepo squash specimens recovered from Marble Bluff came forest zone (Reisberg and Harter 2006; Smith 2006a). Because from an isolated charred storage deposit located in a crevice neither of the other two eastern domesticates—Iva annua and against the back wall of the shelter and could not be easily Chenopodium berlandieri—is cultivated today in the region, associated with other cultural features or artifact assemblages. any effort through genetic analysis to determine more spe- In general, Late Archaic small-scale societies in the Ozarks cifically where they were brought under domestication will have been characterized as being largely autonomous, with necessitate comparison of modern wild populations with do- settlement systems centered on multiseasonal multiple-activ- mesticated specimens from archaeological contexts. ity base camps that are situated on the terraces of major While future genetic and archaeological research may es- streams and smaller shorter-term seasonally occupied rock- tablish with greater spatial resolution those geographical lo- shelters in smaller tributary drainages and upland settings cations where marsheleder, chenopod, and other eastern crop (Fritz 1986:39; Sabo, Waddell, and House 1982:64). Subsis- plants were most likely initially domesticated, it is possible tence economies centered on white-tailed deer (Odocolieus now to identify with considerable confidence the specific hab- virginianus) and smaller species (raccoon, turkey, squirrels), itats within which domestication occurred. Four and perhaps as well as on hickory nuts and acorns. more local seed-bearing plants—all aggressive river valley col- onizers of disturbed soil settings—were initially domesticated Phillips Spring (23Hi16) and were being cultivated in stream valley and upland settings across a broad area of the Mississippi drainage catchment Located in Hickory County, Missouri, the Phillips Spring site (midlatitude oak- and hickory-dominated forest zones) by centers on a small artesian spring situated on the 1b terrace 5,000–3,500 years ago. At the same time, it is surprising that of the Pomme de Terre River, a tributary of the Osage, which if you look past the dating and documentation of the early in turn flows into the Missouri River. Excavation units in the domesticates from these seven sites, there is relatively little water-saturated anaerobic sediments adjacent to the spring additional published information that is currently available exposed a “squash and gourd zone” (unit K2) that contained regarding the actual settlements or their inhabitants. This rel- abundant plant remains (hickory, walnut, abundant acorns, ative absence of information regarding the seven sites in ques- grape, elderberry, ragweed), including bottle gourd rind frag- tion is the result of a number of factors, including the limited ments and 125 uncarbonized C. pepo seeds and seed fragments nature of excavations, complex stratigraphy, substantial dis- (Kay 1983, 1986; Kay, King, and Robinson 1980). Based on turbance and mixing of habitation layers in multicomponent their size and a direct date of 5025 calibrated calendar years sites (see Cowan 1984:323; Fritz 1986), and lack of published BP on one of the seeds (table 1), this large assemblage provides site reports. Drawing from available primary references, brief the earliest evidence for the domestication of this species in cameo descriptions of each of the Late Archaic settlements eastern North America. that have yielded early domesticates are provided below. This unit K squash and gourd zone, which is the oldest of the six stratified living surfaces documented at Phillips Spring, was exposed to only a very limited horizontal extent during Marble Bluff (3SE1) excavation, and little information is available regarding ar- Excavated by a University of Arkansas crew in 1934, Marble chitectural features and patterns of artifact distribution. The Bluff is a 7-m-wide habitation zone that extends for about subsequent six Sedalia-phase occupational episodes at Phillips 120 m along a west- to southwest-facing overhanging bluff Spring, which occurred over a period of about 500 years (Kay located just above Mill Creek, a tributary of the Buffalo River, 1983:54–55), however, yielded abundant artifacts, pit features, which in turn flows into the White River, in Searcy County, hearths, and discrete residential-area midden scatters, which Arkansas (Fritz 1986:11, 24, 27; 1997). A large alluvial terrace Kay (1983:61) considers to represent a series of warm-season is located less than 0.5 km from Marble Bluff, and an esti- (spring to fall, and perhaps through the winter) base en- mated 30 ha of arable land was available within 2 km of the campments that centered around the artesian spring. Along settlement (Fritz 1986:30, 36). with more or less sedentary river valley settlements like Phil- Although cultural deposits at Marble Bluff are described as lips Spring, smaller family groups occupied outlying sites, Smith Plant Domestication in Eastern North America S475 including caves and rockshelters, on an intermittent seasonal a 20-mi stretch of the Wabash River, a tributary of the Ohio basis. Although squash and bottle gourd were grown, oak River, in Crawford and Lawrence counties, Illinois. One of mast and hickory nuts played a major dietary role, with white- these, the Riverton site, yielded evidence of four domesticated tailed deer, mussels, and fish being the most important prey plants: sunflower, chenopod, C. pepo squash, and bottle species. gourd. The Riverton site is a large, deeply stratified midden that Napoleon Hollow extends over an area of about 2 acres (ca. 470ft# 220 ft) on a T-0 terrace of the Wabash River. In 1961, five# 5 -ft5 Located where a small tributary valley known as Napoleon excavation units in three separate locations exposed 44 inches Hollow joins the Illinois River, which in turn flows into the of cultural deposits, including abundant features and well- Mississippi, this large, deeply stratified site is known primarily preserved material culture assemblages. In 1963, a large block for its extensive and well-documented middle woodland oc- excavation unit (unit X) exposed, just below the plow zone, cupations (Wiant and McGimsey 1986). A block excavation a group of 10 clay floors, along with associated pit and hearth into the colluvial fan emanating from the bluff on the north features, artifacts, and extensive midden lenses (fig. 2). side of the hollow where it joins the Illinois Valley, however, Roughly rectangular in shape, the Riverton clay floors ranged also exposed a series of stratified Middle and Late Archaic in size from about 100 to 200 ft2 and from 4 to 6 inches in period occupations (Wiant, Farnsworth, and Hajic 2009). depth, and they are thought to have been built within a short One of the pit features associated with a Late Archaic Tit- time of each other. Like the numerous similar features doc- terington phase (feature 20) contained abundant artifacts and umented in Late Archaic contexts across the eastern wood- plant remains, including Cucurbita rind fragments and Chen- lands (Sassaman and Ledbetter 1996; Smith 1986:27), these opodium berlandieri, sunflower, and ragweed seeds, along with clay floors are thought to be prepared house floors even 44 carbonized marshelder achenes. Based on their large size though associated post holes are not always observed. and a direct date of 4400 calibrated calendar years BP, this A series of eight radiocarbon dates from unit X indicate a substantial assemblage of Iva annua achenes provides the ear- relatively short-term occupation dating to ca. 3800–3700 cal- liest evidence for the domestication of this species (Asch and ibrated calendar years BP (Smith and Yarnell 2009; Winters Asch 1985:161; table 1). 1969). The Riverton material culture assemblage included a Given the number and variety of stone tools and the wide range of chipped and groundstone lithic tools (projectile amount of debris recovered during excavation, the Tittering- ton component at Napoleon Hollow is considered to have been a seasonally occupied river valley base camp (Wiant, Farnsworth, and Hajic 2009). In a broad regional consider- ation of the Titterington phase and other neighboring and generally contemporaneous Late Archaic societies, Cook (1986:175) characterizes their subsistence system as having a heavy reliance on white-tailed deer, shellfish, fish, waterfowl, small mammals, hickory, walnut, and oaks, with seed plants— including chenopod, ragweed, sunflower, marshelder, and Cu- curbita—playing only a minor role. Their annual cycle in- volved seasonal multigroup nucleated base camps and smaller single-group settlements dispersed across a range of environ- ments: many Titterington-phase sites with preserved feature pop- ulations have a recurrent pattern: either there is a single cluster of nonoverlapping shallow-basin hearths, or large roasting pits or there are linear clusters of such features. I assume that the single-cluster sites are occupied by one or more extended families while the multiple-cluster sites rep- resent simultaneous use by several extended families. (Cook 1986:184)

Riverton (Cw-170) In his landmark study of the Late Archaic Riverton culture, Howard Winters (1969) excavated three large shell-midden Figure 2. Clay house floors and other features uncovered in unit sites (Robeson Hills, Swan Island, and Riverton) located along X of the Riverton site (after Winters 1969:96). S476 Current Anthropology Volume 52, Supplement 4, October 2011 points, knives, side and end scrapers, hammerstones, manos, Although relatively few intact Late Archaic deposits were un- metates, axes, etc.), along with a variety of well-preserved bone covered during the site’s excavation in 1978 because of dis- tools (antler projectile points, awls, gravers, fleshing tools, turbance both by subsequent Early Woodland period occu- etc.; Winters 1969). pations and modern vandalism, a 35-m2 area along the back Soil samples taken from features associated with the clay wall of the shelter yielded evidence of a number of inter- floors at Riverton by Richard Yarnell subsequently yielded mittent occupational episodes between 4500 and 3000 BP. A domesticated sunflower, C. pepo squash, bottle gourd, and surface hearth, a basin-shaped pit, and a few ash lenses were two distinct cultigen varieties of domesticated chenopod uncovered, along with three post molds that suggested a small (Smith and Yarnell 2009; Yarnell 2004). Although seeds of a enclosed area near the rear wall of the overhang. small number of wild species were also recovered (e.g., per- The archaeobotanical assemblage from Late Archaic con- simmon, elderberry, Polygonum), the archaeobotanical assem- texts at Cloudsplitter is dominated by nuts, with various hick- blage primarily consisted of carbonized fragments of black ory species and black walnut accounting for 90% of the nuts walnut and hickory nuts and acorns (Yarnell 2004). A diverse recovered (Cowan 1984:338). In addition to the seeds of a faunal assemblage was dominated by white-tail deer, with range of wild plant species, five domesticates are represented significant representation of raccoon, beaver, and turkey, in low numbers: bottle gourd, C. pepo squash, marshelder, along with waterfowl and other birds, 13 species of fish, and sunflower, and chenopod. Several uncarbonized and well-pre- 37 species of freshwater mussels (Parmalee 1969). served thin-testa fruits of domesticated chenopod recovered Given the floodplain location of Swan Island and Riverton, during excavation yielded a direct AMS calibrated intercept along with seasonality indicators, Winters (1969:137) iden- of 3700 BP (Smith and Cowan 1987; table 1). White-tailed tified them as spring-fall and summer occupations, respec- deer dominated the faunal assemblage, followed by turtles, tively, while Robeson Hills, with its higher elevation above small to medium mammals (e.g., raccoon, squirrel, opossum), the Wabash Valley, was characterized as a winter settlement. and turkey. With lithic debris reflecting retouch of tools rather Alternatively, all three sites could be considered as generally than manufacture, a relative lack of formal tools, a limited comparable in terms of representing larger river valley mul- number of features, and the absence of any evidence of forest tiseasonal macroband base camps linked to numerous out- clearance, Cloudsplitter was identified as reflecting a pattern lying shorter-term, more limited activity sites (Smith 1986). of repeated short-term seasonal (likely fall) occupational ep- isodes during the Late Archaic as part of a seasonal round Hayes (40ML139) that involved larger and longer-occupied river valley settle- ments (Cowan 1984:359–364). Situated on a T-1 terrace at the confluence of Caney Creek and the Duck River, a tributary of the Tennessee River, in Marshall County, Tennessee, the Hayes site is a large (ca. 900- Setting the Stage: Middle Holocene m2) stratified multicomponent midden with up to a 1.7-m- Environmental Change and River thick Middle Archaic stratum of freshwater gastropod shells Valley Stabilization (stratum III) overlaid by more than 1 m of Late Holocene alluvial deposits, including a Late Archaic shell-free habitation The Late Archaic sites briefly described above span more than layer (stratum II; Crites 1987; Klippel and Morey 1986). A 1,500 years and are scattered across a large geographical area flotation sample recovered from square 989N920E, level 14 of eastern North America. In addition, with the exception of (130–140 cm below ground surface), in a small block exca- Riverton, they have yielded only limited information regard- vation unit located on the T-1 terrace (Crites 1987:16), con- ing the way of life of their inhabitants. At the same time, tained six complete domesticate-size sunflower seeds, one of however, they all do reflect a range of shared characteristics which yielded an AMS radiocarbon date of 4840 calibrated and fit a general overarching profile that comes into clearer calendar years BP (Crites 1993; table 1). Although both Ben- focus when viewed within a broader environmental and cul- ton and Ledbetter projectile point/knives were also recovered tural context. from the same natural stratum as the sunflower seeds (Crites To a considerable extent, the answer to why plant domes- 1993:146), a fuller description or characterization of the Late tication in eastern North America lagged behind a number Archaic occupation of the Hayes site is not available. of the other centers of domestication worldwide by 4,000 years or more can be found in the “prerequisite” Mid-Holocene climatic and environmental changes that reshaped the interior Cloudsplitter (15Mf-36) river valley landscapes of the region. By about 6500–6000 BP, Situated about 100 m upslope from a semipermanent stream many of the river systems across the interior midlatitudes had that flows 250 m southward to join the Red River in Menifee shifted from an Early Holocene pattern of episodic pulses of County, Kentucky, the Cloudsplitter rockshelter extends 55 sediment removal and river incision to a sustained phase of m along a westerly facing overhang that protects an area of aggradation and stabilization (Knox 1983; Schuldenrein 1996: about 15 m in width (Cowan 1984:315; Cowan et al. 1981). 3, 9–10; Smith 1986:22–23; Styles 1986), resulting in the for- Smith Plant Domestication in Eastern North America S477 mation of river valley meander belt topography with asso- because of post-Columbian human impacts and introductions ciated oxbow lakes and backswamps along some stream of exotics on the one hand and the well-documented pro- courses and shallow-water and shoal habitats in others. This pensity of the species in question to both hitch a ride on shift in seasonal stream flow patterns and the stabilization of human transport and to invade and colonize both natural river valley landscapes in turn resulted in a significant increase and anthropogenic disturbed soil situations (Smith 2006b)on in the abundance and diversity of floodplain plant and animal the other. In addition, the taxonomies of Chenopodium, Iva, resources available for human exploitation, while at the same and C. pepo in eastern North America remain challenging at time an apparent decrease in effective precipitation resulted the species level, making the use of modern distribution stud- in a deterioration of upland resources (Schuldenrein 1996; ies sometimes problematic. Smith 1986:24). Not surprisingly, these changes in river valley Surveys carried out in the 1980s document the present-day and upland environments were accompanied by a broadscale distribution of C. berlandieri and Iva annua in riverine hab- intensification of human occupation of river and stream valley itats across much of the oak-savannah and oak-hickory forest corridors (Brown 1985). As Joseph Schuldenrein (1996:3) regions shown in figure 1 (Smith 2006b), and neither species notes, the Mid-Holocene was the “‘window of adjustment’ grows today with any frequency south of the fall line onto during which postglacial environments stabilized, stream the Atlantic coastal plain, west into the plains, east into the channels adjusted to renascent floodplains, hill and slope sed- Appalachians, or very far north into the Northeast or Great imentation rates diminished, and critical resource zones Lakes. Similarly, free-living populations of C. pepo var. ovifera emerged,” and human societies “mapped onto” these favor- have been located growing in riverine settings in many areas able resource and subsistence zones. of the oak-savannah and oak-hickory forest regions, as well The stabilization and resource enrichment of river valley as down onto the coastal plain (Smith 2006b), and the extent corridors was not uniform across the eastern woodlands, how- to which the distribution of this wild cucurbit extended into ever, and outside of this interior midlatitude riverine zone, the Northeast remains a topic of considerable interest (Fritz Mid-Holocene climatic and environmental changes were quite 1999). Although the sunflower has a geographical range that different (Schuldenrein 1996; Smith 1986). While meander encompasses a broad area of North America, present-day wild belt systems were established in many of the tributaries of populations along the western edge of the oak-savannah forest the Mississippi River by 6000 BP, for example, such is not zone in figure 1 have been identified as the specific and sole the case in the lower alluvial valley of the Mississippi, where source of all modern domesticated sunflowers (Rieseberg and open swamps emerge on the margins of prograded deltas Harter 2006). (Schuldenrein 1996:9). In addition, within the interior mid- At the same time that Mid-Holocene river valley stabili- latitude riverine zone, substantial and sustained “mapping” zation and enrichment were important in determining when onto river valley corridors does not appear to have been uni- plant domestication occurred in eastern North America and versal, nor does it reflect any filling in or “packing” of these the geographical distribution of wild progenitors within the enriched resource zones (Claassen 1996:240–242; Smith 1986: resource-rich river valley corridors of the oak-savannah and 22). oak-hickory forest regions played a role in determining where domestication occurred, the answer to why some plants were domesticated and not others can be found, to a considerable The Geographical Range and Floodplain degree, in the functional ecological profile of the wild pro- Niche of Eastern Domesticated genitors (Bogaard et al. 1999; Charles, Jones, and Hodgson Seed Plants 1997). Three of the four species brought under domestication in eastern North America—marshelder, chenopodium, and Just as the Mid-Holocene emergence of stable and enriched C. pepo gourds—are floodplain “weeds,” aggressive pioneers river valley landscapes within the Mississippi River catchment of the disturbed and exposed soil situations created on an plays a significant prerequisite role in the timing of the initial annual basis by spring floods (Smith 2006b). Their abundant domestication of seed plants in eastern North America, the seeds (a single C. berlandieri plant can produce 50,000 seeds; geographical range and habitat of their wild ancestors is also a single C. pepo plant can produce 5,000) are dispersed by of obvious importance in delineating the region where initial floodwaters, and they colonize the sandy banks and backwater domestication took place. In general, the present-day geo- margins that are exposed each year by receding floodwaters graphical range of the wild species from which the eastern within the constantly shifting landscape of river floodplains. domesticates were derived is fairly well established. Using Early successional species, they cannot compete with other modern distribution data to estimate the ca. 5000-BP geo- plants for long in undisturbed locations, but frequently they graphical range of the wild ancestors of the eastern lineage can be found in recently disturbed floodplain soil situations, of Cucurbita pepo squash, sunflower, marshelder, and Chen- both natural and man-made (e.g., construction sites, backyard opodium berlandieri is complicated, however, by a range of gardens, dumps, fields lacking herbicides, etc.). As a result, factors. Principal among these is the extent to which natural all three are preadapted in a number of respects to both river valley habitats and plant communities have been altered flourish in garden plots created for them and provide a sub- S478 Current Anthropology Volume 52, Supplement 4, October 2011 stantial return at minimal cost if humans play the role of regarding how and where to draw phase boundaries in both dispersal agent and scatter their seeds in areas naturally ex- space and time (McElrath 1993:150). To some extent, the posed by receding floodwaters (Smith 2009b). The substantial relatively large size of these Late Archaic phases and the dif- expansion of “natural” stands by casual scattering of seeds of ferent opinions regarding where to draw phase boundaries is all three of these floodplain weeds into appropriate exposed a reflection of limited available archaeological information. soil settings, described up into the 1700s in Louisiana (Smith But it also indicates that while exhibiting geographical vari- 2006b, 2009b), may have considerable time depth in the re- ation in diagnostic artifact categories (primarily projectile gion, along with a number of other niche construction efforts points), these societies were also generally quite similar in by human societies in eastern North America (Smith 2011), terms of their stable long-term adaptation to eastern decid- and may have paved the way to the establishment of more uous forest ecosystems: their technology, their economies, permanent and stable garden plots. their annual cycle, and their settlements all follow the same In contrast, while the sunflower frequently grows in dis- general pattern. turbed soil settings in river valley landscapes and generally conforms to the floodplain weed profile of the other three Technology eastern domesticates in that it is also an early successional species favoring open disturbed soil settings, it can be found With the exception of the Mount Nebo phase of northwestern thriving in many locations where the other three would not Missouri and adjacent states at the western edge of the oak- grow. Like marshelder, chenopod, and C. pepo, however, sun- savannah forest zone, which has yielded small sherds of fiber- flower was well adapted to take full advantage of the new tempered ceramics of uncertain vessel forms (Reid 1983:29– growing opportunities that humans would offer. 32), none of the Late Archaic (5000–3400 BP) societies of the oak-savannah and oak-hickory forest regions had pottery ves- sels. Cooking was done over open fires or in pits using heated The Societies That Domesticated and stones (as reflected by abundant fire-cracked rock and basin- Cultivated Eastern Seed Plants between shaped features). Artifact assemblages are dominated by 5000 and 3400 BP and Developed a chipped-stone tools and debitage, with chert and quartzite Crop Complex by 3800 BP raw material used to manufacture a wide range of formal and expedient hunting and processing tools, including bifacially Briefly described above, the seven archaeological sites that flaked knives, drills, and spear points (the bow and arrow was have yielded all of the archaeobotanical evidence for the do- not present in the east during the Late Archaic), along with mestication and cultivation of seed-bearing plants in eastern unifacial end scrapers and side scrapers. Groundstone tools North America before 3400 BP extend over a considerable include heavy-duty three-quarter-grooved axes; manos; and span of time and a sizable geographical area. They also vary hammer, grinding, and nutting stones. Bone-tool assemblages, considerably in terms of the amount of information each has when preserved, are dominated by white-tailed deer skeletal provided regarding the way of life of the societies that oc- elements, including ulna awls, metatarsal and humerus flesh- cupied them. In spite of these challenges of space and time ing tools, antler projectile points and flakers, and bone pins. and variable archaeological information, however, it is pos- Of the seven sites briefly described above, Riverton provides sible to develop a general profile of these Late Archaic period one of the largest and most comprehensively described artifact societies. assemblages; Howard Winters’s (1969) landmark functional In large part, this is possible because of the broadscale analysis of both lithic and bone-tool categories offers an ex- patterns of continuity in material culture assemblages, settle- cellent general profile of Late Archaic technology. Numerous ment types, and subsistence patterns that characterize the Late other more recent comprehensive analyses of assemblages Archaic societies that occupied the midlatitude riverine en- from Late Archaic sites have also been published (e.g., Cook vironments of the oak-savannah and oak-hickory forest 1976; Emerson, McElrath, and Fortier 2009; Fortier 1984). regions (Emerson, McElrath, and Fortier 2009). These soci- Other material culture categories, including basketry and eties are often partitioned in space and time by phase des- cordage, are preserved only occasionally in dry rockshelters ignations (e.g., the Riverton phase, the Titterington phase), and caves (Fritz 1986; Scholtz 1975). which group together sites that are geographically and tem- porally close to each other and that have similar material Settlements culture assemblages. As more sites are excavated and larger material culture inventories become available for comparative Riverton also provides one of the best windows on what the analysis, phase designations often become more tightly settlements of the societies that cultivated seed plants in east- bounded in terms of the geographical areas and the time spans ern North America between 5000 and 3400 BP looked like. they encompass. In the oak-savannah and oak-hickory forest The unit X block excavation exposed a grouping of 10 roughly regions during the Late Archaic, phase designations at the rectangular clay floors ranging in size from 100 to 200 ft2 that present time are rather large in scale, and opinions differ are thought to have supported small house structures (fig. 2). Smith Plant Domestication in Eastern North America S479

Numerous shallow-basin-shaped pit features, a few hearths, cators that can be employed to establish what part of the year and sheet-midden deposits were associated with the Riverton a site was occupied, but other than substantial evidence of clay floors, along with abundant artifact assemblages. Clay cold-season house structures (which are lacking at Late Ar- floors of similar size and suspected function have been de- chaic sites), there are no good indicators of a settlement being scribed at Middle and Late Archaic period sites across the occupied during the winter. eastern United States, and, along with more infrequently doc- Taking into consideration all of these potential complica- umented post mold patterns, they provide evidence of hab- tions in describing any individual Late Archaic settlement in itation structures that were relatively small and that did not detail (including the seven sites that have provided the earliest reflect a substantial investment of labor (Sassaman and Led- evidence of domesticated plants in the region), a number of better 1996). generally similar settlement pattern models have been pro- In the far more frequent absence of any evidence at all of posed over the past 4 decades that, while relatively low in house structures in Late Archaic contexts, the spatial signature resolution, are still of considerable interpretive value and are of basic domestic units (likely extended families) usually con- reasonably applicable to many of the Late Archaic phases of sists of a cluster of pit and hearth features (Fortier 1983). the oak-savannah and oak-hickory forest regions (Smith 1986: Determining the number of such family units that made up 24–25). These settlement systems encompass two different a settlement is complicated by a number of factors. Frequently, settlement types, the larger of which can be labeled the “river as was the case at Phillips Spring, Napoleon Hollow, Riverton, valley base camp.” Situated on lower terraces of river and and Hayes, excavation does not expose the full extent of the stream valleys, base camp settlements are sometimes char- occupation. In addition, even when there is little overlap or acterized as permanent year-round settlements or more fre- superposition of feature clusters, it is difficult to determine quently as semipermanent to permanent, summer-fall, low- which features (and family units) were present contempo- water seasonal occupations reoccupied on an annual basis raneously as opposed to reflecting sequential habitation ep- over long periods of time by a number of related extended isodes over a number of years. At the same time, the overall families. Phillips Spring, Napoleon Hollow, Riverton, and areal extent of Late Archaic settlements can be misleading in Hayes could be placed in this general settlement category that river valley topography can play an important role is based on their river valley location and occupational history. determining site boundaries. The overall size of the Riverton During the season of the year when the lower terrace lo- site, for example, may in large part be dictated by the areal cation of these base camps was subject to flooding or isolation extent of Groundhog Hill, on which the settlement was lo- by rising waters, groups are thought to have moved into the cated. In contrast, the Go Kart North site, just east of St. uplands, either to nearby functionally distinct valley-edge wet- Louis, dating to 4000 BP and like Napoleon Hollow, assigned season residential base camps (e.g., Robeson Hills; Winters to the Titterington phase, was situated on a topographically 1969) or smaller dispersed short-term occupation sites oc- unconstrained bank of what was an active stream channel and cupied by a single extended-family unit. Marble Bluff, Cloud- extends linearly for more than 175 m, leading Thomas Em- splitter, and Newt Kash fit this settlement profile. Short-term erson to suggest that, floodplain topography permitting, Late limited activity settlements could also, of course, be located Archaic groups moving through an annual cycle may return in river valley locations (e.g., Go Kart North), and both base not to specific points on the landscape but to more general camps and shorter-term sites could vary considerably in du- “locales” (Emerson 1984:343–344; Emerson and McElrath ration, range of activity sets, and number of domestic units 1983). from year to year. Go Kart North, which is the only fully exposed Late Archaic One of the most interesting aspects of the larger semiper- settlement in the interior Riverine midlatitudes, also provides manent to permanent multifamily river valley base camp set- clear evidence of Late Archaic settlement structure (Fortier tlements is the lack of any internal organizational structure 1983). Even though it extended for more than 175 m along above the level of extended-family domestic units. Artifact the edge of the Hill Lake Meander and included 131 discrete assemblages, feature clusters, and structural evidence (clay hearth and pit features and lithic activity areas organized into floors, post mold patterns, when present) all show little dif- four discrete and “socially segmented” clusters, it is charac- ferentiation across settlements, and while some linearity and terized not as a large settlement of any duration but rather adjacency can be discerned in the patterning of domestic unit as a short-term seasonal occupation involving a relatively lim- placement, there is no evidence of centrality (e.g., integration ited range of activities, including animal butchering, food around a central shared or corporate open area), nor is there preparation, and lithic manufacture. any indication of storage above the extended-family level. This To add to the difficulties often inherent in establishing the picture of associated but largely autonomous extended-family number of basic domestic units that make up a Late Archaic units is also reflected in corporate mortuary sites (Charles settlement at any one point in time, it is also difficult to and Buikstra 1983:121–122), which when present both in- establish whether settlements were occupied year round on a dicate sustained long-term shared “ownership” of a society’s permanent basis or annually during certain seasons of the resource catchment area and the absence of any within-group year. There are a wide variety of different seasonality indi- ascribed status differentiation. S480 Current Anthropology Volume 52, Supplement 4, October 2011

This is a rather homogeneous picture of generally similar tion likely plays a role to some extent in the low seed-to-nut technology, societal organization, settlements, and seasonal ratios documented in Late Archaic archaeobotanical assem- rounds over a broad geographical area and with phase bound- blages across the interior midlatitudes (Johannessen 1984), aries defined largely in terms of variation in the shape of but it is also clear that human utilization of plant resources projectile points and the presence or absence of other dis- was relatively narrowly focused, with a concentration on hick- tinctive tool types. It is also evident in the subsistence econ- ory and walnuts, along with acorns, which could be gathered, omies of Late Archaic societies within the oak-savannah and processed, and stored over long periods with a minimum of oak-hickory forest regions. effort (Ash, Ford, and Asch 1972:27–28).

Subsistence Economies Regional Interaction The Late Archaic societies that first domesticated and culti- Given the relatively limited importance of seeds generally in vated eastern seed-bearing plants in the interior midlatitudes the diet of Late Archaic societies in the interior midlatitudes of eastern North America between 5000 and 3700 BP shared and the substantial distances that separated the settlements what was in many respects a very similar basic subsistence that have yielded early evidence of domesticated plants, it economy. White-tailed deer was the single most important would be reasonable to wonder how, exactly, the seeds of prey species, and it invariably dominates faunal assemblages domesticates were moved from society to society. There is no from Late Archaic sites. This is not surprising in that this clear answer to this question beyond the simple suggestion species combines large body size with a very high reproductive that potential (Smith 2009b), and it is the primary prey species in the exchange networks that moved information, innova- faunal assemblages in the region throughout the Holocene. tions, various raw materials, and finished artifacts around In sites with good bone preservation, smaller mammals— the Southeast [and adjacent areas] during the 5000–2500- including raccoons, opossums, rabbits, and squirrels—are also BP time period apparently consisted of innumerable mul- frequently present in Late Archaic faunal assemblages, as well tidirectional, reciprocal, down-the-line exchanges between as turkey. Not surprisingly, river valley settlements with good trading partners (often lineage leaders) of both nearby and preservation also contain abundant evidence of fish and fresh- distant communities. (Smith 1986:30) water bivalves, as well as waterfowl. In situations where pres- ervation and recovery are good, this basic pattern of faunal The extent to which settlements within the oak-savannah utilization can be expanded, and the finer-grain differences and oak-hickory forest regions were linked by these complex between Late Archaic economies in different parts of the oak- webs of interaction is reflected in the broad geographical dis- savannah and oak-hickory forest zones come into clearer fo- tribution of a range of different artifact forms, most notably cus (Phillips and Brown 1983; Styles and McMillian 2009). projectile points but also including a number of other cate- In a similar manner, Late Archaic sites throughout these gories of objects. By mapping the geographical distribution oak- and hickory-forest regions have archaeobotanical assem- of distinctively carved and engraved bone pins, for example, blages invariably dominated by carbonized acorns and nut Richard Jefferies (1996:228) has recently outlined a “lower fragments of hickory and black walnut (Simon 2009). Rep- Ohio–central Mississippi Valley interaction network” that he resenting an abundant and highly nutritious food source that argues “defines a socially bounded area extending over several can be stored for a year or more, hickory and black walnuts hundred kilometers of the midcontinent, within which groups and acorns occur in varying frequency in plant assemblages of increasingly sedentary hunter-gatherers interacted and ex- across the region, with acorns more abundant at Phillips changed information that facilitated their survival.” While the Spring, for example, in the oak-savannah forest zone and area outlined by Jefferies includes both the Riverton and Na- black walnut predominating at Riverton, with Yarnell (2004) poleon Hollow sites (fig. 3), a similarly regionally scaled dis- discussing the excellent habitat the Wabash River Valley rep- cussion by Cook (1986) outlines an area of interaction based resented for this species. As Munson (1986) suggested more on shared hafted biface types that in turn encompasses three than 3 decades ago, it is possible that human niche construc- archaeological phases (Titterington, Sedalia, and Nebo Hill; tion efforts involving the selective culling of forest trees to fig. 3) and includes the Napoleon Hollow and Phillips Spring encourage nut- and mast-bearing species, which is well doc- sites. Dale McElrath (1993:150) concludes that “it seems likely umented in the early historic period (Foster, Black, and that the Titterington-Sedalia-Nebo Hill cultural entities in- Abrams 2004), may have deep time depth in eastern North teracted to a significant degree and can be viewed as repre- America (Smith 2009b, 2011) and may extend into the Late senting a culture (TSN culture).” McElrath (1993:150) also Archaic. notes the occurrence of the distinctive Benton projectile point Although seeds of a wide variety of different species have at the Titterington phase Go Kart North site in the American been recovered from Late Archaic contexts—particularly of Bottom, indicating a clear link to other Late Middle Archaic fruits, including grape, persimmon, hackberry, and plum— and Late Archaic sites across the Southeast that have yielded overall seed counts are invariably low. Differential preserva- Benton points, including the Hayes site. Smith Plant Domestication in Eastern North America S481

Figure 3. Location of Archaic period interaction zones identified by Cook (1986) and Jefferies (1996).

Although linked by a number of artifact forms (and ar- the very earliest stages of a transition from hunting and gath- chaeologists), Phillips Spring, Napoleon Hollow, Riverton, ering to food production, to look at the societies involved in and the Hayes site span a considerable period of time and the initial efforts to domesticate local species and the sub- space, and such linkages are meant not to suggest either direct sequent emergence of a coherent crop complex (Smith and or contemporaneous interaction between these settlements Yarnell 2009), and to consider the general cultural and en- but rather to indicate that the societies in question were all vironmental contexts of this major shift in human history. involved in broadscale networks of simple down-the-line in- As this period of initial experimentation with domestication teraction and exchange that existed over long periods of time. comes into clearer focus in the other independent centers of People and ideas, as well as seeds and other lightweight com- agricultural origin worldwide, it will certainly be interesting modities, moved along these networks in ways that are as yet to compare these as yet less well-documented developmental not very well understood. But it does seem clear that between trajectories with what is known about the societies that do- 5000 and 3500 BP, the oak-savannah and oak-hickory forest mesticated seed plants in eastern North America. regions were inhabited by a large number of small autono- In the oak-savannah and oak-hickory forest regions of the mous societies, some if not all of which were experimenting East, the initial domestication and early cultivation of indig- to various degrees with the cultivation of local seed plants and sharing their success and failure, as well as their seed enous seed plants (ca. 5000–3400 BP) were carried out by stores, along well-established networks of interaction. small-scale societies, each consisting of perhaps a half dozen or more related extended-family units. Situated along and Summary Discussion tethered to the second- and third-order tributary river valley corridors of the Mississippi River catchment, these societies Eastern North America provides a rare opportunity to view followed an annual cycle that linked multiple-family-unit S482 Current Anthropology Volume 52, Supplement 4, October 2011 semipermanent to permanent settlements in river valley lo- be made. Claassen (1996) makes a major point in this regard, cations with a range of other short-term multiple-family and observing that there are many resource-rich river and stream single-family river valley and upland occupations. valley settings in the region that did not witness substantial Clay floors and post mold patterns, along with spatially human occupation during the Late Archaic period. If human discrete feature clusters, provide occasional glimpses of the population growth led to an increase in human occupation spatial structure and size of these settlements, which lack any of river valleys and greater competition over resources, why indication of ascribed status differentiation or corporate or- do some valley segments remain empty? Similarly, why do ganization above the level of extended-family domestic units. subsistence economies remain stable and generally compa- The relative frequency and within-site distribution of a range rable over long time spans and across geographical areas? In of different chipped-stone, groundstone, and bone tools al- addition, in a number of areas where extensive surveys have lows for the identification of different activity sets, from lithic documented the size and spacing of Late Archaic river valley and bone-tool manufacture and maintenance to food pro- settlements having a long history of occupation, resource cessing. catchment zones turn out to be quite substantial. The three Although a large range of different species of plants and large river valley settlements identified by Winters in the Wa- animals have been identified in the archaeobiological assem- bash Valley that make up the Riverton culture, for example blages of these settlements, their subsistence economies were (Riverton, Swan Island, Robeson Hills), are spaced at 10-mi generally comparable over a broad area of the interior riverine intervals, and he estimates that the resource catchment area midlatitudes in terms of centering on a limited number of on one side of the river for all three settlements was 500 mi2 plant resources (e.g., hickory, walnut, oaks) and a consistent (Winters 1969:110). list of terrestrial animal species and species groups, including The debate about the role of population growth, landscape the white-tailed deer, several smaller terrestrial species (e.g., packing, and resource in the initial domestication of plants turkeys, raccoons, rabbits, squirrels), and aquatic resources and animals worldwide will no doubt continue for a sub- (bivalves, snails, and fish). Although seeds of a variety of stantial period of time. I would argue, however, that eastern different species played a small role in the economy of these North America, arguably the best-documented regional case groups, based on consistently low seed/nut ratios, the initial study currently available, does not provide much support for domestication and early cultivation of local species did take general models, including those of human behavioral ecology place within their river corridor natural habitat zones rather (Smith 2009a) that incorporate environmental downturn, ex- than in upland settings based on the earliest recovery of all ternal environmental stress, population growth, landscape four eastern domesticates from river valley settlements (table packing, constricted resource zones, and carrying-capacity 1). Subsistence economies remained stable over long periods imbalance or resource scarcity in explaining the initial do- of time, reflecting general long-term successful adaptations to mestication process. Based on the archaeological information the resource-rich river valley corridors of the oak-savannah now available, small societies in eastern North America first and oak-hickory forest regions. These societies may have been domesticated local seed plants and developed initial crop modifying their environments in a range of different ways, complexes in resource-rich river valley environments within from differential culling of trees to expanding natural stands a larger context of stable long-term adaptations and broad- of floodplain seed plants and establishing “orchards” of fruit- scale niche construction efforts that were carried out in the and berry-producing species, and initial domestication of absence of any carrying-capacity challenges or seriously com- eastern seed plants could well have taken place within a pressed and compromised resource catchment areas. It is in- broader behavioral context of human niche construction teresting in this regard to reconsider the discussions of the (Smith 2007, 2009b, 2011). role resource-rich environments and initial domestication There has been considerable discussion over the past several made more than 15 years ago by Price and Gebauer (1995). decades about the extent to which population growth during the Late Archaic period resulted in the filling up or “packing” of eastern North American landscapes. This, it is argued, re- References Cited sulted in a reduction in resource catchment zones and in- creased territoriality and competition over resources, which Asch, D. L., and N. B. Asch. 1985. Prehistoric plant cultivation in in turn resulted in a range of adaptive responses, including, west-central Illinois. In Prehistoric food production in North Amer- ica. R. I. Ford, ed. Pp. 149–203. Anthropological Papers, no. 75. perhaps, the domestication of plants (Brown 1985; Jeffries Ann Arbor: Museum of Anthropology, University of Michigan. 1996; Smith 1986). Alternatively, I would argue that even Asch, N. B., R. I. Ford, and D. L. Asch. 1972. Paleoethnobotany of though site density certainly increases in the Late Archaic and the Koster site: the Archaic horizons. Illinois State Museum Reports corporate cemeteries and deep-midden deposits reflect the of Investigations, no. 24. Springfield: Illinois State Museum. establishment of long-term human utilization and “owner- Bogaard, A., C. Palmer, M. Charles, and J. G. Hodgson. 1999. A FIBS approach to the use of weed ecology for the archaeobotanical ship” of sections of river valley corridors and adjacent up- recognition of crop rotation regimes. Journal of Archaeological Sci- lands, the case for landscape packing and resource compe- ence 26:1211–1224. tition playing a causal role in plant domestication has yet to Brown, J. A. 1985. Long-term trends in sedentism and the emergence Smith Plant Domestication in Eastern North America S483

of complexity in the American Midwest. In Prehistoric hunter- settlement forests of east-central Alabama. Human Ecology 32:27– gatherers. T. D. Price and J. A. Brown, eds. Pp. 201–234. Orlando, 47. FL: Academic Press. Fritz, G. 1986. Prehistoric Ozark agriculture: the University of Ar- Charles, D., and J. Buikstra. 1983. Archaic mortuary sites in the kansas rockshelter collections. PhD dissertation, University of central Mississippi drainage: distribution, structure, and behavioral North Carolina, Chapel Hill. implications. In Archaic hunters and gatherers in the American Mid- ———. 1997. A three-thousand-year-old cache of crop seeds from west. J. L. Phillips and J. A. Brown, eds. Pp. 117–145. Orlando, Marble Bluff, Arkansas. In People, plants, and landscapes.K.J. FL: Academic Press. Gremillion, ed. Pp. 42–62. Tuscaloosa: University of Alabama Charles, M., G. Jones, and J. G. Hodgson. 1997. FIBS in archaeo- Press. botany: functional interpretation of weed floras in relation to hus- ———. 1999. Gender and the early cultivation of gourds in eastern bandry practices. Journal of Archaeological Science 24:1151–1161. North America. American Antiquity 64:417–430. Claassen, C. P. 1996. A consideration of the social organization of Jefferies, R. 1996. The emergence of long distance trade networks in the Shell Mound Archaic. In Archaeology of the Mid-Holocene the southeastern United States. In Archaeology of the Mid-Holocene Southeast. K. Sassaman and D. Anderson, eds. Pp. 235–258. Southeast. K. Sassaman and D. Anderson, eds. Pp. 222–234. Gainesville: University Press of Florida. Gainesville: University Press of Florida. Cook, T. G. 1976. Koster: an artifact analysis of two Archaic phases Johannessen, S. 1984. Plant remains. In The Go-Kart North site.A. in west-central Illinois. Archaeological Program Prehistoric Re- Fortier, ed. Pp. 166–181. American Bottom Archaeology FAI-270 cords 1. Evanston, IL: Northwestern University. Site Reports 9. Urbana: University of Illinois Press. ———. 1986. A dispersed harvesting economy: the Titterington Kay, M. 1983. Archaic period research in the western Ozark Highland, phase. In Foraging, collecting, and harvesting: Archaic period sub- Missouri. In Archaic hunters and gatherers in the American Midwest. sistence and settlement in the eastern woodlands. S. W. Neusius, ed. J. L. Phillips and J. A. Brown, eds. Pp. 41–70. Orlando, FL: Ac- Pp. 175–200. Occasional Paper, no. 6, Center for Archaeological ademic Press. Investigations. Carbondale: Southern Illinois University. ———. 1986. Phillips Spring: a synopsis of Sedalia phase settlement Cowan, C. W. 1984. From foraging to incipient food production: and subsistence. In Foraging, collecting, and harvesting: Archaic subsistence change and continuity on the Cumberland Plateau of period subsistence and settlement in the eastern woodlands.S.W. eastern Kentucky. PhD dissertation, University of Michigan, Ann Neusius, ed. Pp. 275–288. Occasional Paper, no. 6, Center for Arbor. Archaeological Investigations. Carbondale: Southern Illinois Uni- Cowan, C. W., H. E. Jackson, K. Moore, A. Nickelhoff, and T. L. versity. Smart. 1981. The Cloudsplitter rockshelter, Menifee County, Ken- Kay, M., F. B. King, and C. K. Robinson. 1980. Cucurbits from Phillips tucky: a preliminary report. Southeastern Archaeological Conference Spring: new evidence and interpretations. American Antiquity 45: Bulletin 24:60–76. 807–822. Crites, G. D. 1987. Middle and Late Holocene ethnobotany of the Klippel, W. E., and D. F. Morey. 1986. Contextual and nutritional Hayes site (40ML139): evidence from unit 990N918E. Midconti- analysis of freshwater gastropods from Middle Archaic deposits at nental Journal of Archaeology 12:3–32. the Hayes site, middle Tennessee. American Antiquity 51:799–813. ———. 1993. Domesticated sunflower in fifth millennium B.P. tem- Knox, J. C. 1983. Responses of river systems to Holocene climates. poral context: new evidence from middle Tennessee. American In The Holocene, vol. 2 of Late Quaternary environments of the Antiquity 58:146–148. United States. H. E. Wright, ed. Pp. 26–41. Minneapolis: University Delcourt, P. A., and H. R. Delcourt. 1981. Vegetation maps for eastern of Minnesota Press. North America. In Geobotany II. R. Romans, ed. Pp. 123–165. McElrath, D. L. 1993. Mule road: a newly defined Late Archaic phase New York: Plenum. in the American Bottom. Illinois Archaeology 5:148–157. Doran, G. H., D. Dickel, and L. Newsom. 1990. A 7,290 year-old Munson, P. 1986. Hickory silviculture: a subsistence revolution in bottle gourd from the Windover site, Florida. American Antiquity the prehistory of eastern North America. Paper presented at the 55(2):354–360. Conference on Emergent Horticultural Economies of the Eastern Emerson, T. E. 1984. The Dyroff and Levin sites. American Bottom Woodlands, Carbondale, IL. Archaeology FAI-270 Site Reports 9. Urbana: University of Illinois Parmalee, P. W. 1969. Appendix 1: animal remains from the Archaic Press. Riverton, Swan Island, and Robeson Hills sites, Illinois. In The Emerson, T. E., and D. L. McElrath. 1983. A settlement-subsistence Riverton culture. H. Winters. Pp. 139–144. Report of Investigations, model of the terminal Late Archaic adaptation in the American no. 13. Springfield: Illinois State Museum. Bottom, Illinois. In Archaic hunters and gatherers in the American Phillips, J. L., and J. A. Brown, eds. 1983. Archaic hunters and gatherers Midwest. J. L. Phillips and J. A. Brown, eds. Pp. 219–242. Orlando, in the American Midwest. Orlando, FL: Academic Press. FL: Academic Press. Price, T. Douglas, and Anne Birgitte Gebauer, eds. 1995. Last hunters, Emerson, T. E., D. L. McElrath, and A. C. Fortier. 2009. Archaic first farmers. Santa Fe, NM: School of American Research. societies. Albany: State University of New York Press. Reid, K. L. 1983. The Nebo Hill phase: Late Archaic prehistory in Erickson, D., B. D. Smith, A. C. Clarke, D. H. Sandweiss, and N. the Lower Missouri Valley. In Archaic hunters and gatherers in the Tuross. 2005. An Asian origin for a 10,000-year-old domesticated American Midwest. J. L. Phillips and J. A. Brown, eds. Pp. 11–39. plant in the Americas. Proceedings of the National Academy of Orlando, FL: Academic Press. Sciences of the USA 102:18315–18320. Rieseberg, L. H., and A. V. Harter. 2006. Molecular evidence and the Fortier, A. 1983. Settlement and subsistence at the Go-Kart North evolutionary history of the domesticated sunflowers. In Darwin’s site: a Late Archaic Titterington occupation in the American Bot- harvest: new approaches to the origins, evolution, and conservation tom, Illinois. In Archaic hunters and gatherers in the American of crops: a broad taxonomic and geographic survey. T. Motley, ed. Midwest. J. L. Phillips and J. A. Brown, eds. Pp. 243–260. Orlando, Pp. 31–48. New York: Columbia University Press. FL: Academic Press. Riley, T. C., A. Waltz, A. Bareis, A. Fortier, and K. Parker. 1994. ———. 1984. The Go-Kart North site. American Bottom Archaeology Accelerator mass spectrometry (AMS) dates confirm early Zea FAI-270 Site Reports 9. Urbana: University of Illinois Press. mays in the Mississippi River Valley. American Antiquity 59:490– Foster, H. T., B. A. Black, and M. D. Abrams. 2004. A witness tree 497. analysis of the effects of Native Americans on the pre-European Sabo, G., D. B. Waddell, and J. H. House. 1982. A cultural resource S484 Current Anthropology Volume 52, Supplement 4, October 2011

overview of the Ozark–St. Francis National Forests, Arkansas. Re- construction by small-scale societies in North America. In The port submitted to the U.S.D.A. Forest Service, Ozark–St. Francis subsistence economies of indigenous North American societies.B.D. National Forests. Fayetteville: Arkansas Archaeological Survey. Smith, ed. Pp. 593–609. Washington, DC: Rowman & Littlefield. Sassaman, K., and D. Anderson, eds. 1996. Archaeology of the Mid- Smith, B. D., and C. W. Cowan. 1987. Domesticated Chenopodium Holocene Southeast. Gainesville: University Press of Florida. in prehistoric eastern North America: new accelerator dates from Sassaman, K., and R. J. Ledbetter. 1996. Middle and Late Archaic eastern Kentucky. American Antiquity 52:355–357. architecture. In Archaeology of the Mid-Holocene Southeast. K. Sass- Smith, B. D., and R. A. Yarnell. 2009. Initial formation of an indig- aman and D. Anderson, eds. Pp. 75–96. Gainesville: University enous crop complex in eastern North America at 3800 B.P. Pro- Press of Florida. ceedings of the National Academy of Sciences of the USA 106:6561– Scholtz, S. C. 1975. Prehistoric piles: a structural and comparative 6566. analysis of cordage, netting, basketry, and fabric from Ozark Bluff Styles, B. 1986. Aquatic exploitation in the Lower Illinois River Valley: shelters. Arkansas Archaeological Survey Research Series, no. 9. the role of paleoecological change. In Foraging, collecting, and har- Fayetteville: Arkansas Archaeological Survey. vesting: Archaic period subsistence and settlement in the eastern Schuldenrein, J. 1996. Geoarchaeology and the Mid-Holocene land- woodlands. S. W. Neusius, ed. Pp. 145–174. Occasional Paper, no. scape history of the greater Southeast. In Archaeology of the Mid- 6, Center for Archaeological Investigations. Carbondale: Southern Holocene Southeast. K. Sassaman and D. Anderson, eds. Pp. 3–27. Illinois University. Gainesville: University Press of Florida. Styles, B., and R. B. McMillan. 2009. Archaic faunal exploitation in Simon, M. L. 2009. A regional and chronological synthesis of Archaic the Prairie Peninsula and surrounding regions of the mid-conti- period plant use in the mid-continent. In Archaic societies.T.E. nent. In Archaic societies. T. E. Emerson, D. L. McElrath, and A. Emerson, D. L. McElrath, and A. C. Fortier, eds. Pp. 81–114. C. Fortier, eds. Pp. 39–80. Albany: State University of New York Albany: State University of New York Press. Press. Smith, B. D. 1986. The archaeology of the southeastern United States: Wiant, M. D., K. B. Farnsworth, and E. Hajic. 2009. The Archaic from Dalton to de Soto, 10,500–500 B.P. In Advances in world period in the Lower Illinois River Basin. In Archaic societies.T.E. archaeology, vol. 5. F. Wendorf and A. Close, eds. Pp. 1–92. Or- Emerson, D. L. McElrath, and A. C. Fortier, eds. Pp. 229–286. lando, FL: Academic Press. Albany: State University of New York Press. ———. 2006a. Eastern North America as an independent center of Wiant, M. D., and C. R. McGimsey. 1986. Woodland period occu- plant domestication. Proceedings of the National Academy of Sci- pations of the Napoleon Hollow site in the Lower Illinois Valley. ences of the USA 103:12223–12228. Kampsville Archeological Center Research Series, vol. 6. Kamps- ———. 2006b. Rivers of change. 3rd edition. Tuscaloosa: University ville, IL: Center for American Archeology. of Alabama Press. Winters, H. 1969. The Riverton culture. Report of Investigations, no. ———. 2007. Niche construction and the behavioral context of plant 13. Springfield: Illinois State Museum. and animal domestication. Evolutionary Anthropology 16:188–199. Yarnell, R. I. 2004. Riverton plant remains and terminal Archaic ———. 2009a. Core conceptual flaws in human behavioral ecology. crops. In Aboriginal ritual and economy in the eastern woodlands. Communicative and Integrative Biology 2:533–534. A. M. Cantwell, L. A. Conrad, and J. E. Reyman, eds. Pp. 123– ———. 2009b. Resource resilience, human niche construction, and 130. Illinois State Museum Scientific Papers, vol. 30. Springfield: the long-term sustainability of Pre-Columbian subsistence econ- Illinois State Museum. omies in the Mississippi River Valley corridor. Ethnobiology 29: Zeder, M. A., E. Emshwiller, B. D. Smith, and D. Bradley. 2006. 167–183. Documenting domestication: the intersection of genetics and ag- ———. 2011. Shaping the natural world: patterns of human niche riculture. Trends in Genetics 22:139–155. Current Anthropology Volume 52, Supplement 4, October 2011 S485

Genetics and Domestication Important Questions for New Answers

by Greger Larson

The recent ability to extract genetic data from archaeological remains of wild and domestic animals has opened up a new window onto the history and process of domestication. This article summarizes the impact of that new perspective derived from both modern and ancient DNA and presents a discussion of the validity of both the methods and conclusions. In general I address the use of post hoc conclusions and the lack of starting hypotheses to inform what we know about domestication from a genetics perspective. I use three case examples (dogs, goats, and pigs) to exemplify fundamental aspects of the genetic data we still do not understand before specifically commenting on the use of molecular clocks to date domestication and the necessity of thinking about domestication as a process. I conclude on a positive note with a brief discussion about the future relationship between genetics and domestication.

Introduction with genetic methods meant they could not say why. The sexy conclusions and the high impact that have often been gen- erated from these kinds of studies (Vila et al. 1997 has been The hullabaloo really began in 1997. That year, an article cited more than 300 times in 13 years) combined with the appeared in Science with the title: “Multiple and Ancient Or- power to begin sentences with the words “in direct contrast igins of the Domestic Dog” (Vila et al. 1997). The use of to long-held beliefs” have led to a flood of domestication population-level DNA sequence data to reveal insights into genetics papers. animal domestication was not entirely novel. The year before, In this essay I will review not only the conclusions of a Bradley et al. (1996) explored the dynamics of African and number of publications in this vein but also the more general European cows, but the high-profile nature of the dog article paradigms that geneticists have operated under in order to definitively consummated the marriage between genetics and guide their research. I will then apply a thought experiment domestication. The article created a stir for two reasons. First, to demonstrate how little is known regarding the fundamen- it demonstrated the power of population genetic analysis to tals of genetic data before addressing specific questions related reveal details that were previously beyond the scope of an to molecular clocks, the process of domestication, and the analysis based on either morphology or DNA restriction pat- ramifications that domestication studies can have on other terns. Sequences of As, Ts, Cs, and Gs possessed a degree of fields. First, however, in order to critique the validity of resolution that could not be matched by bones or differing genetics-based assertions, it is worth discussing briefly the molecular fragment lengths. methods such studies employ and the strengths and weak- Second, the primary conclusion of the article, that dogs nesses therein. were domesticated 135,000 years ago, simultaneously sparked the imaginations of science journalists and in equal measure A Genetics Primer infuriated zooarchaeologists who knew that the oldest bones that could be safely ascribed to fully domestic dogs were no The basic modus operandi of these studies is as follows. Hun- more than 10,000–12,000 years old (Clutton-Brock 1995). dreds if not thousands of (typically modern) samples of a The estimated age of 135,000 years was nonsense. Archae- given species are collected from as many different geographic ologists knew the dates were wrong, but a lack of familiarity locations, breeds, or populations as possible. These samples are usually derived from tissue, hair, or feathers, and each Greger Larson is Fellow and Principal Investigator in the Department sample is bathed in a series of chemicals into order to isolate of Archaeology, Durham University (South Road, Durham DH1 3LE, the DNA. United Kingdom [[email protected]]). This paper was The extracted DNA possesses millions of copies of the en- submitted 13 XI 09, accepted 2 XII 10, and electronically published tire genome of the organism as well as many more copies of 13 VI 11. the mitochondrial genome housed within the mitochondria

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0021$10.00. DOI: 10.1086/658401 S486 Current Anthropology Volume 52, Supplement 4, October 2011 organelle found in virtually every cell. Domestication genet- samples. Not every individual, however, will possess a unique icists have historically ignored the multibillion base pair nu- sequence. Numerous samples will often carry the identical clear genomes of their samples and instead focused on the sequence of base pairs along the entire fragment. The set of 16,000 base pair mitochondrial genome. As sequencing tech- samples that share such identical sequences are said to possess nology becomes cheaper and faster, however, the nuclear ge- the same haplotype. Although the word “haplotype” has dif- nome is becoming increasingly accessible. Still, the mito- ferent meanings depending on the type of genetic marker chondrial genome remains attractive for several reasons. amplified by geneticists, in the studies discussed below, a First, it does not recombine. That is, the only changes that “haplotype” simply refers to a unique combination of base occur in the sequence of a mitochondrial genome are the pairs across the amplified fragment. It is with the haplotypes result of mutations. Thus, modeling its linear evolution is far that the next stage of analysis begins. simpler than having to think about and account for diploid A set of haplotypes can be used as raw material to build genomes that hybridize and swap genes every generation. Sec- either a phylogenetic tree or a haplotype network. Both kinds ond, although most of the 16,000 base pairs that make up of diagrams visually depict the relationships between the hap- the mitochondrial genome code for genes crucial to an or- lotypes. Of the two, networks are the easiest to understand. ganism’s basic survival, there is a small region, typically less They typically consist of circles and lines in which each circle than 2,000 base pairs, between these protein-coding genes represents a haplotype, and the size of the circle corresponds where replication begins when a new mitochondrial genome to the number of individuals in the data set that possess that is being built. This region, known as either the control region haplotype. Larger circles result when more individuals are all or the d-loop, is not part of the translation or transcription identical across the sequenced region, and the smallest circles process and is thus not under the same restrictive selective represent those individuals that possess haplotypes not shared regime to which the rest of the genome is subjected. by any other individual. Two circles connected by a line differ Mutations, or errors that occur during replication, occur by a single substitution no matter where in the sequence that all the time. It is the fate of those mutations that differ. If a substitution sits, and hash marks are often placed across the mutation occurs in a protein-coding gene that negatively in- line to indicate additional substitutions. These kinds of figures fluences the functional properties of that gene, then the mi- are often simple enough to be drawn by hand, but software tochondria carrying that mutation will not be replicated as programs also exist to first identify the haplotypes in a data often (if at all), and the mutation will disappear. This means set and to draw the corresponding network depicting how that although mistakes during replication happen all the time closely or distantly related all the haplotypes are from one and in an unbiased fashion across the genome, rates of sub- another (fig. 1). stitution, or rather the potential of that mutation to survive, Phylogenetic trees are generally more complex than net- are distinctly nonrandom. Because the control region does works because they employ models of evolution to infer the not code for a protein and thus there are no ramifications evolutionary relatedness of the haplotypes. The evolutionary (good or bad) for a new mutation, far more of those mu- models place differing weights on the kinds of substitutions tations become incorporated into the sequence. Evolution is encoded in the sequence, and these weights alter the math- simply change through time, and because mutations in this ematical distances between the haplotypes. Because trees are region generally do not affect the organism, the control region depictions based on those distances, different models can pro- evolves at a faster rate than the rest of the genome, where duce differently shaped trees or trees with the same shape but most mutations are deleterious and are selected out before with differing levels of statistical support for the branches. they can become incorporated and sequenced by researchers. Adding additional individuals from previously unsampled This relatively speedier evolution allows for differences be- populations that possess novel haplotypes can also alter the tween populations and species to build up over short periods shape of the networks and trees. of time, thus enabling geneticists to differentiate between con- Once networks and trees have been generated, two addi- specific individuals and to draw inferences from the sequences tional terms are often used to discuss their shapes. A “hap- regarding the demographic history of the species. Though logroup” on a network is a cluster of closely related haplotypes mitochondrial DNA (mtDNA) data almost always distinguish that together create an easily recognizable group that is more between species and can often distinguish populations below or less (it is hoped more) differentiated from other haplotypes the species level, some populations, such as dog breeds, do and haplogroups. A “clade” on a tree is more or less the same not possess diagnostic mtDNA signatures. thing, although its technical definition is a group of haplotypes Once the DNA has been extracted, geneticists select and that are more closely related to each other than any one is amplify a fragment rich in variability, typically a few hundred to any other haplotype. In human-relationship terms, this base pairs of the control region of the mitochondrial genome, would be called a family. Because a single tree is just one of from every individual. As stated above, this region evolves many that could possibly be drawn from the data, phyloge- quickly, and as a result, geneticists expect to find differences neticists prefer to generate numerous trees to see how fre- (substitutions) at numerous positions along the sequenced quently the same patterns of clades and haplogroups appear. fragment, the patterns of which differentiate the individual The more often they appear, and the more robust they are Larson Genetics and Domestication S487

Figure 1. Example of a phylogenetic tree on the left (rooted with a don- key, Equus asinus) and an unrooted network on the right, both of which were generated using mitochondrial control region haplotypes of modern horses. The colors and letters associated with the clades on the tree correspond to the colored and lettered haplogroups on the network. (A color version of this figure is available in the online edition of Current Anthropology.) Note the correlations between the relative positions of the haplogroups on the network and on the tree. The tree and network are adapted respectively from Vila et al. (2001) and Jansen et al. (2002). to different parameter values within separate models of evo- structure of trees derived from the data. These studies do not lution, the more confidence phylogeneticists have that the follow the textbook scientific method that begins with a fal- relationships are “real” and not just artifacts of the data (fig. sifiable assumption and dichotomous easily defined expec- 1). tations and ends with a comparison between the generated At this stage—armed with a network, a tree, or both— data and the expected result. This is not necessarily a bad geneticists are ready to begin the process of interpreting the thing. Scientific enquiry, especially at the early stages of data images and gleaning the implications for our understanding gathering using a newly available technique, is often inves- of how, when, where, and how many times domestication of tigative and explorative. Newly derived categories of data can- a particular species has occurred. Having laid out the basic not be expected to be acquired or interpreted within the con- methods of these studies, what follows is a short critique of fines of explicitly stated hypotheses. Given the relatively short the ways in which inferences and conclusions have been period of time that population-level sequence data have been drawn. available, it is perhaps no surprise that within the field of domestication genetics, there has been little explicit hypothesis The Appeal of the Post Hoc Narrative testing. My own work on pigs is no exception. The majority of studies of the ilk I describe above do not Though many domestication studies strive to interpret the contrast their observed data with an expected result. For the data within the context of what is already known about their most part, there are no expectations regarding the shapes of study animals, many of them are content to report only de- networks, the number of haplotypes or haplogroups, or the scriptive accounts of the generated data. Not every study has S488 Current Anthropology Volume 52, Supplement 4, October 2011 done this, of course, but the general trend was recently ex- wild animals only recently transported there (Naderi et al. emplified by a 2007 article, the title of which was “Large- 2007). If true, the number of shared haplotypes between truly Scale Mitochondrial DNA Analysis of the Domestic Goat Re- wild and truly domestic would be 0. veals Six Haplogroups with High Diversity” (Naderi et al. This observation has not gone unnoticed, and the near 2007). This particular article does, in fact, present insights universal lack of shared haplotypes between dogs and wolves regarding goat domestication, but the plainly descriptive title has been exploited as a means to identify recent hybrids by hints at the nature of many of these studies that generate data observing stereotypically dog haplotypes in modern wolves in theoretical vacuums. (Randi and Lucchini 2002). Still, beyond the use of this ob- In order to elucidate this trend, I present three case studies servation as a conservation tool, no one has yet questioned that focus on dogs, goats, and pigs. All three employed the why wild and domestic animals of these species share so few general methodology discussed above, and taken together they haplotypes. demonstrate the potential and limits of genetic domestication studies. I have chosen these three because populations of the A Thought Experiment Involving respective wild ancestors—wolves, bezoars, and wild boar— Haplotypes remain extant, thus allowing for a comparative analysis of the genetic patterns found in both wild and domestic animals. We know that dogs and goats are derived exclusively from For studies of animals whose wild ancestors are either extinct wolves and bezoars, respectively. Thus, the earliest domestic (e.g., cows and camels) or uncertain (e.g., sheep), the genetic populations must have shared 100% of their mitochondrial differences between the wild and domestic forms can only be haplotypes with their wild counterparts. Given this, the ques- revealed by generating DNA sequences from archeological tion must be why and how has the shared proportion dropped material. A large number of studies have attempted to do just to virtually 0%. One explanation could be that the original that with respect to cows (Edwards et al. 2007), but the first wild populations that gave rise to domestic stocks are now ancient sheep and camel DNA article are still forthcoming. extinct and the sampled extant wild populations in these stud- Because wolves, bezoars, and wild boar are still around, ies were not involved in the domestication process. This could geneticists are able to sample them and place both wild and be especially true for wolves, which have suffered a long his- domestic variants into the same network or tree. Though tory of persecution. Under this scenario, however, the ex- more recent studies have been published, a 2002 study of dogs pected networks and trees would generate haplogroups that (Savolainen et al. 2002) is instructive. This study typed more consist of either wild or domestic animals. In the dog study, than 600 domestic dogs and nearly 40 wolves, numbers that the tree did in fact demonstrate that some clades consisted in 2002 were relatively large. Two articles focused on goats only of wolves or dogs consistent with the extirpation sce- were published in 2007 and 2008 (Naderi et al. 2007, 2008), nario, but the majority of clades contained haplotypes of both but in the intervening 5 years, the acceptable standard for dogs and wolves even if that was because some of the hap- sample numbers had increased, and these studies analyzed lotypes were shared (Savolainen et al. 2002). In goats, though 2,430 domestic goats and 473 wild bezoars. several bezoar-only clades are evident, every single domestic The first statistic normally generated in these articles is the goat sample is found within a cluster of bezoars, though again, number of haplotypes found among all the samples. In these none of the haplotypes are shared (Naderi et al. 2008). These cases, the authors identified 110 unique haplotypes in dogs patterns do not fit any simple scenario of domestication that and 17 in wolves. The goat studies, based on fourfold more focuses on demographic patterns of limited sampling from domestic samples and tenfold more wild samples, identified wild populations and periodic bottlenecks for both wild and a total of 1,783 unique haplotypes in both populations. The domestic animals. issue of what those numbers mean and whether they are The explanation above rests on an assumption that though significant is difficult to answer for the simple reason that no the control region of mtDNA does evolve quickly relative to one knows how many haplotypes to expect from a given both other genes in the mitochondrial genome and the nu- number of populations or individuals. clear genome of the organism, it is traditionally not thought Even without this understanding, a comparative approach to be fast enough for mutations to accumulate over the rel- can be used across species to ask new questions that will form atively short time frame of domestication (10,000 years). If the basis of future studies. The first question worth asking is true, this would mean that haplotypes found in modern wild how many haplotypes are found in both the wild and domestic and domestic animals have not changed since the beginning samples. In canines, out of a total of 127 haplotypes, only of the Holocene and that the observed substitutions not only one was identified in both dogs and wolves. In caprines, of occurred long before domestication but also reflect popula- 1,783 unique haplotypes, only three were shared by both wild tion structuring that resulted from a long-term lack of gene and domestic goats, and those three were found only on the flow between geographically partitioned groups. This as- island of Sicily, where the status of the goats and the timing sumption may not always hold, however, and a series of ar- of their arrival is uncertain. According to the authors, the ticles has suggested that substitution rates are not fixed (Ho domestic goats found on this island could be ancestors of and Larson 2006; Ho et al. 2005). Larson Genetics and Domestication S489

These authors demonstrated that the evolutionary rate de- the same haplotype as an indigenous wild boar that was never rived from a data set is dependent on the time depth of the part of a domestication process, male domestic pigs would most recent common ancestor of the studied sample set. A have to mate with female wild boar, and the piglets would data set consisting of a group of humans known to have had have to be incorporated into the domestic stock. The opposite a common ancestor on the order of hundreds or thousands scenario is common practice in many cultures, especially in of years will possess a great deal more variation than what New Guinea, where females are often left tied to a stick at would be expected using standard evolutionary rates. When the edge of a village overnight and are subsequently impreg- the data set is increased to include chimpanzees and other nated by feral males from the forest. In this case the resulting primates, the date of the most recent common ancestor is piglets retain their mother’s domestic mitochondrial signa- pushed back to a scale of millions of years, and the evolu- ture. tionary rate tumbles. This so-called time dependency of evo- Second, an argument that assumes a high degree of hy- lutionary rates could result from the retention of slightly del- bridization must explain why so many populations of indig- eterious mutations over a sufficient time frame to be included enous wild boar—including those in India and on islands in population-level data sets. Over longer time frames, those such as Japan, the Ryukyu chain, and Taiwan—retain their mutations are eliminated, which would then reduce the ob- genetic distinctiveness (Larson et al. 2005) and why domestic served variability in the data set and give the appearance of pigs introduced to these areas have not acquired the local a slower evolutionary rate. Additional studies of different spe- DNA haplotypes. Neither the model that associates shared cies have thus far confirmed the phenomenon (Burridge et haplotypes with independent domestication nor the model al. 2008) even if a fully satisfactory explanation remains elu- that assumes all instances of shared haplotypes are the result sive. of recent hybridizations explains the data. The truth, of What this might mean is that we should not necessarily course, probably lies somewhere in the middle, although ob- expect wild and domestic haplotypes to be identical. Instead, serving and describing DNA evidence is only the first step to wild and domestic individuals that shared a common ancestor uncovering it. around the time of the origins of domestication would possess This issue touches on a second key difference between the substitutions that have accumulated since they split. So long wild boar and wolf and bezoar data sets, and for this dis- as this pattern was generalizable across different animal do- cussion it is worth explaining another common term. “Phy- mesticates, this would explain why wild and domestic dog logeography” is the study of the association of phylogenetic and goats fail to share any common haplotypes. signals with the geographical provenance of the samples. A Pigs, however, are different. The pig data contradict the strong phylogeographic signal is the result of a high degree dog and goat data in at least two key ways. First, wild boar of reciprocal correlation between a geographic region and a and domestic pigs share at least 17 haplotypes (Larson et al. specific haplogroup or clade. If an analysis of a wild popu- 2005, 2007a, 2007b, 2010). This could easily result if the pig lation demonstrates that hypothetically, highly differentiated data were based on shorter sequences than dogs or goats, thus haplogroups are found in Spain, Italy, and Greece but that reducing the chances of finding substitutions that differentiate animals carrying all three types are present in northern Eu- individuals, which would lead to a reduction in the number rope, geneticists would be tempted to suggest that the strong of overall haplotypes. The number of base pairs amplified for phylogeographic pattern in southern Europe suggests a ge- dogs, goats, and pigs, however, is 582, 469, and 662, respec- netic differentiation that took place in refugial regions during tively. All else being equal, pigs should therefore possess more ice ages and a mixing of haplotypes when those populations total haplotypes and fewer shared haplotypes between wild migrated north after a climactic amelioration. By assigning and domestic animals. This is not the case. colors to specific haplogroups or clades and by pinning the Not only are the two most frequent domestic haplotypes colors onto a map, geneticists are able to ascertain the relative found in Europe also found in European wild boar, more strength of the phylogeographic signal. than 15 haplotypes are found in both East Asian wild boar A strong signal is desirable because it allows authors to and Chinese domestic breeds, an additional haplotype was pinpoint hypothetical centers of domestication. Unfortu- shared by Indian wild boar and domestic pigs, and another nately, most wild animals involved with domestication lack was shared by wild boar from Vietnam and domestic and feral strong signals, at least when the data sets consist only of pigs found in Island Southeast Asia (Larson et al. 2007b, mtDNA. (As sequencing techniques become cheaper, data sets 2010). The most obvious explanation for this pattern is that studies that interrogate and analyze nuclear genomes [e.g., it results not from distinct instances of domestication but that vonHoldt et al. 2010] may reveal more geographically pro- like the dog scenario discussed above, the shared haplotypes scribed and genetically distinct populations of wild and do- are the result of recent hybridizations between introduced mestic populations.) Three wolves, for example, sampled from domestic pigs and indigenous wild boar. Though this expla- China, Mongolia, and Saudi Arabia all possessed the same nation cannot be ruled out, there are two significant factors mitochondrial haplotype, as did individual wolves from Tur- that make it less likely. First, mtDNA is passed solely along key, Sweden, and Portugal (Vila et al. 1997). Thus, assigning the maternal line. Thus, in order for domestic pigs to share an origin to dogs who possessed haplotypes closely related to S490 Current Anthropology Volume 52, Supplement 4, October 2011 these wolves is problematic or at least lacking in precision. people began migrating to Australia, the Americas, and other Modern bezoars are significantly more geographically circum- parts of the world. A modern analysis of this kind that did scribed than modern wolves, and though some haplotypes not take into account the historical migrations of people to seem to be found only in small regional pockets, almost all the New World would conclude that the United States of of the full complement of haplotypes were identified in be- America was the origin of all humans. That is to say, highly zoars sampled exclusively between Turkey and Iran (Naderi diverse regions can result not just from a legacy of origination et al. 2008). but also from migration into the region by genetically diverse In many ways this makes sense. Wolves migrate long dis- populations. Thus, demonstrating that a region is particularly tances during their lifetimes, and thus different haplotypes diverse without also offering nongenetic evidence suggesting are expected to be present at many locations across the Old the region was in fact a center of domestication is problematic and New Worlds. Humans, too, have been responsible for the at best. All of the data related to cows is used in this argument movement of both wild and domestic animals, thus smearing (Troy et al. 2001) as well as in the argument that dogs were and blurring any phylogeographic pattern that may have ex- first domesticated in China (Savolainen et al. 2002). In the isted in the Pleistocene. In addition, the history of the ice former case, archaeological evidence also supports a Near ages, as described in the hypothetical above, has also played Eastern origin of cattle domestication and a subsequent Neo- a role by forcing populations apart where they begin to di- lithic migration into Europe, suggesting that the genetic in- versify before reuniting them. This has been shown to play a terpretation is correct. In the latter case, however, a supporting role in yaks, herds of which often contain individuals with narrative based on archaeology remains elusive, and a recent highly variable and differentiated haplotypes (Ho et al. 2008). publication using genetic data from a large number of geo- The overall effect of these homogenizing forces should lead graphically isolated wolves and domestic breeds concluded to a modern-day situation in which no wild population retains that because Near Eastern wolves also played a large role in a strong phylogeographic signal. Yet unlike virtually every the domestication of dogs (vonHoldt et al. 2010), China was other wild animal involved in domestication and in defiance likely not the sole center of dog domestication. of both their natural migratory ability and a long history of human-assisted transport and reproductive meddling, wild Confidence-Free Molecular Clocks boar do. This strong phylogeographic signal allows for a rel- atively straightforward identification of centers of origin. For Numerous attempts have also been made to place an inde- instance, a handful of wild boar collected in India are all pendent time frame on the history of domestication using positioned in a haplogroup that is significantly different than genetic data sets. In many cases, the authors of these articles all other groups. These haplotypes are only present in South have concluded or at least implied that animal domestication Asia, and thus when an Indian pig identified as domestic also began hundreds of thousands of years ago (Ho and Larson possessed the same signature, the most parsimonious expla- 2006). Perhaps the most famous of these attempts was the nation was that these wild boar were likely involved in do- Vila et al. (1997) publication that pushed dog domestication mestication (Larson et al. 2005). Of course, it is possible that back more than 100,000 years. The authors were able to con- domestic pigs derived from a separate population in a dif- clude this by first determining that the average mitochondrial ferent place were transported to India and then mated with genetic difference between wolves and coyotes was 7%. By an indigenous female wild boar, thus producing a litter that borrowing a date of one million years for the last shared was retained by humans. The data cannot differentiate be- common ancestor between the two species, they established tween these two scenarios, but the strength of the phylo- a rate of 1% per 135,000 years. As discussed above, because geography at least allows the suggestion of an independent most of the clades on their tree contained both wolves and Indian domestication to be made, which can then be further dogs, the authors estimated the divergence between wolves investigated and corroborated by archaeological or historical and the one clade that only contained dogs. The figure, 1%, sources. meant that dogs and wolves last shared a common ancestor When phylogeographic signals are weak, suggestions re- 135,000 years ago (Vila et al. 1997). garding the geography of domestication rest on more subtle There were a number of assumptions made during this arguments. The most popular one is based on a determination exercise, and the decision to not bracket the estimate with of the genetic variability present in domestic animals in dif- error bars gave an unwarranted impression of precision. The ferent regions. The more variation a region possesses relative conclusion proved intriguing, however, and for several years to the total diversity evident in an entire data set, the more all genetic animal domestication articles included highly sus- likely that region was a center of origin because, as the ar- pect molecular clock analyses (Ho and Larson 2006). Very gument goes, only a subset of the total diversity is generally few (if any) efforts have been made to combine all the possible transported by people away from the center. This is certainly sources of error associated with these kinds of analyses in an true of human diversity, which is far higher on the African effort to confidently ascertain the precision of the estimates. continent than it is in Australia, or at least it was until the More recently, another source of error, the time dependency fifteenth century, when large numbers of genetically diverse of molecular clocks discussed above, has only added to the Larson Genetics and Domestication S491 error, although the variable-rate issue does go some way to- regarding how the data were derived. They demonstrated that ward explaining the discrepancy between the dates derived counterintuitively, multiple origin scenarios of crop evolution from molecules and from archaeology. In most cases, genet- are more likely to give the superficial impression of a single icists applied evolutionary rates derived from data sets whose origin than a single origin scenario. This result demonstrates most recent common ancestor existed millions of years ago that our intuitions are not always valid and that we should to data sets of populations whose common ancestor was far therefore simulate data sets based on our assumptions of what more recent. By applying a slow clock to a data set that is supposed to happen to see what other mistakes we might possessed substantial variation, they significantly overesti- be making when divining the “obvious” story from the shapes mated the time it would take to produce that variation, thus of networks and trees. pushing the timing of domestication deep into the past. Tak- ing this effect into account only removes a single source of error, however, and the combination of all the others suggests A Note on the Process that the error bars almost certainly encompass the present The Allaby, Fuller, and Brown (2008) article is also noteworthy day. Thus, molecular clock efforts so far simply lack the pre- because it does not ignore the long history of domestication. cision to date Holocene phenomena. And this is true even Domestication, like speciation, is not an event. Geneticists after putting aside the issue of what a domestication date know this as well as archaeologists, but for a multitude of actually means. reasons including convenience, we often use the word “event” The production of large mtDNA data sets to glean insights and describe wild and domestic as complementaries, that is, into domestication is no longer novel, and the number of opposites that possess no intermediate form (Dobney and species left to investigate in this manner is dwindling. The Larson 2006). This fallacy is maintained largely because pro- general approach has a great deal of merit, and this first stage cesses are messier than events, and an event mind frame is a of sequence generation is necessary to understanding how a necessary fudge that must be assumed before analyses such genetics-based approach can help us to understand domes- as molecular clocks can be applied. None of the attempts to tication. But it is just the first stage. If sequence data have place a molecular time frame on the history of domestication thus far failed to revolutionize our understanding of the pat- differentiates between the beginning and the end of the pro- terns and processes of domestication, I suspect this is because cess. Instead, a single date estimate is gleaned that is intended we may have hoped that the data themselves would be easily to be interpreted as the year in which wild became domestic. interpretable and provide robust conclusions. Without start- If we are to embrace the process (see Denham 2011; Mar- ing hypotheses about what the data sets would generate, how- shall and Weissbrod 2011; Piperno 2011; Vigne 2011; and ever, easy interpretations were only possible if the trees and networks revealed something immediate and obvious. When Zeder 2011), we have to think differently about both the they did not, we have been left to either simply describe what questions and the data sets. As the Allaby, Fuller, and Brown we see or tell post hoc stories sometimes using shaky as- (2008) study demonstrated, an approach that replaces or at sumptions. least supplements the mitochondrial genome with the full I am confident, however, that the next stage will achieve a nuclear genome has enormous benefits. By looking at the great deal more. Far from asking how many times was species genome of the organism, which contains the genes that code X domesticated, we should be asking why are so few haplo- for the differences between wild and domestic individuals, it types shared between wild and domestic animals? What cli- becomes possible not just to understand what genes are matic conditions or landscape contexts are necessary to pro- changing but precisely how those changes affect the total an- duce x number of clades or haplogroups? What exactly imal. Using these kinds of data sets, we can start asking deeper constitutes a high level of diversity? Is it appropriate to com- questions that focus not on the where and when but on the pare levels of diversity between species? By focusing not on how. In other words, it may soon be possible to identify the how to interpret the data but instead on how many ways the genetic alterations that took place between the first steps of data set can be generated and under what conditions and domestication (fig. 2) and today. parameters, we can begin to replace post hoc explanations The Belyaev fox-farm experiments that began in Siberia in with a hypothesis-testing framework. Again, this is not to say 1959 revealed that by selecting solely for tameness, it was that post hoc narratives are inferior; they are a vital prereq- possible to produce, in relatively few generations, a population uisite to further understanding, but they are limited in the of foxes that looked and acted like domestic dogs (Trut 1999). degree to which they can ultimately inform the history of That much is well known. Two other aspects of these exper- domestication. iments have been less well publicized. First, the farm exper- Hypothesis-driven research in this vein is already yielding imented not just with foxes but also with populations of rats, fascinating new conclusions. An article published by Allaby, beavers, and other animals. Second, the goals were to produce Fuller, and Brown (2008) employed simulations to reveal that both extremely tame animals and extremely aggressive ones post hoc narratives used to support a rapid transition from as well. An anecdote from these later revelations stated that wild to domestic crops were based on a false assumption the Soviet army was ready to deploy large numbers of the S492 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 2. Citizen Dog cartoon by Mark O’Hara (reprinted by permission of Universal Press Syndicate, all rights reserved) that presents a more nuanced (and possibly more accurate) depiction of the process of do- mestication than is often built into genetic models of the same phenom- enon. most aggressive beavers on the Soviet borders in the event the region in a variety of wild jungle fowl. An alignment of the U.S. military ever dared a land invasion. four different wild species revealed that although the majority The two colonies of tame and aggressive rats are now in of the domestic genome was identical to the wild red jungle residence in Leipzig, Germany. By first crossing individuals fowl, the gene responsible for producing yellow legs showed from both groups and then measuring 45 separate physio- a far greater identity to the same region found in gray jungle logical and behavioral traits, a recent study (Albert et al. 2009) fowl. This result resolved the paradox of how yellow legs, a was able to identify two specific quantitative trait loci asso- trait never seen in red jungle fowl, could be so prevalent in ciated with tameness. This kind of study represents an im- domestic chickens, but in so doing it also revealed a somewhat portant first step in revealing links between genetics and be- unexpected conclusion that chickens are not derived from a havior and begins to test the hypothesis that a small number single ancestor (Eriksson et al. 2008). This revelation opens of genes are ultimately responsible for the large behavioral up an entirely new set of questions related to the process of and phenotypic differences that divide wild and domestic an- domestication, the frequency of hybridization and the crea- imals (Dobney and Larson 2006; Stricklin 2001). tion of hybrid domestic animals, and the debate over the Despite a lack of access to parallel populations bred ex- degree of human intentionality in selecting for specific traits plicitly for this purpose, a number of geneticists have already at various stages. developed a long history of insights into the genetic archi- Perhaps the best bet we have for using genetics to unravel tecture underlying domestic phenotypic traits. These kinds of the big questions surrounding domestication is to look for studies have generally been focused on single traits, many of the newly identified changes that underlie key traits in the which are commercially important. Geneticists first type a bones of domestic animals found in archaeological contexts. large number of known variable positions across the genome A recent study on coat colors in pigs demonstrated that the in two populations of animals, one that possesses one variant pattern of mutations that cause coat colors—including red, of a trait, such as a white coat, and one population that has black, and white spotted—are the result of a strong selection a different coat color. A comparison of the regions of differ- pressure away from the camouflage coat colors selected for ence and similarity across the genome allows the geneticists in the wild (Fang et al. 2009). The suggestion is that coat to focus their search, and from there they use similar methods color variation has been a feature of domestication from the to isolate the fragment of DNA that possesses the causative very beginning of the phenomenon. Armed with the causative mutation(s) underlying the trait. Actually identifying the mu- mutations, this hypothesis can be tested by screening ancient tation is often more difficult, although on occasion, such as bones for the genetic variants that underlie the specific coat in traits for muscle growth in pigs, a single mutation was colors. This method has already been used on both ancient pinpointed (Van Laere et al. 2003). mammoth (Rompler et al. 2006) and horse (Ludwig et al. Occasionally these types of studies reveal insights into the 2009) remains, the latter of which revealed an explosion in history of domestication. After identifying the gene respon- the number of coat colors in horses around the fifth millen- sible for yellow legs in chickens, geneticists then sequenced nium BP. Larson Genetics and Domestication S493 Proxies for Domestication’s grations and instances of animal transport that have taken Ramifications place since the Neolithic, it is a certainty that domestic animal populations originally introduced into a new region have sub- Ancient DNA techniques will no doubt be employed in future sequently been replaced, perhaps several times over. A tem- studies to type phenotypic traits in subfossil material. As dis- poral perspective is thus a necessity for any study that pretends cussed above, uncovering a strong phylogeographic signal in to a robust conclusion regarding the long-term history of domestic animals using an alignment of neutrally evolving population movements. DNA (the control region of the mitochondrial genome) has been rare, though wild boar possess an inexplicably strong Conclusion relationship between the phylogenetic placement and their geographic provenance. Given the relatively short period of time over which genetic In his book Guns, Germs, and Steel (Diamond 1997), Dia- methodologies have been applied to domestication questions, mond discusses the universal tendency for populations that it is perhaps no surprise that the initial claims are now being have acquired agriculture and domestic animals to first de- tempered. This is the nature of youth. Practitioners of a new velop a large population and then to move (see also Bellwood technique with the promise of novel data sets have the benefit 2011). Diamond recounts migrations of people armed with of knowing that every result is potentially revolutionary. Jour- a suite of domestic crops overtaking indigenous hunter-gath- nalists and academic journals alike are delighted to publish erers in, among other places, Europe, East Asia, sub-Saharan the rapidly generated conclusions of the new method, and Africa, and New Zealand. The routes and timings of these the more often the new studies overturn conventional wisdom migrations are often contentious, but given the fact that do- or directly contradict decades of findings based on more tra- mestic animals were always a key part of the migratory pack- ditional methodologies, the better. age, the genetic signals derived from their remains can act as As the field eases beyond its teenage brashness, my position a proxy for human migration. is that there is now time to take stock and to begin questioning Because wild boar indigenous to Europe possess such a the assumptions on which many of our early studies were divergent haplotype from those native to the Near East, a based. The massive data sets that will be generated as part of short fragment (less than 85 base pairs) of DNA was enough the high throughput sequencing revolution will reveal fine- to ascertain the genetic legacy of an ancient pig bone. The scale structure at the population level and new genes impor- wild or domestic status of pig bones was determined using a tant in the domestication process. I suspect the new tech- morphological analysis (though of course many remains could nologies will also generate insights not just into DNA not be confidently assigned to either category) after which sequences but RNA sequences as well. Insights at this level the diagnostic fragment was amplified. Not surprisingly, the of organization will facilitate an understanding of not only bones identified as wild in European Mesolithic and Neolithic what genes were key but in which tissues and when they are contexts were European in origin. The domestic bones from active. These kinds of studies will have significant ramifica- a number of sites stretching from Romania through Germany tions not just for domestication but also for the nature of to France, however, displayed a Near Eastern signature. Al- evolutionary change. In addition, a focus on simulation and though this pattern conformed to expectations based on the modeling will reveal how demographic changes affect the pat- known history of the Neolithic migrations into Europe, what terns in population genetic data, which will better allow us was a surprise was the speed with which the Near Eastern to chose which of several competing scenarios best explains lineages were replaced by domestic pigs of European origin, the early history and process of domestication. Finally, im- first in Europe itself and then in the Near East (Larson et al. proved sequencing techniques will allow for an essential tem- 2007a). European wild boar are now the primary (if not sole) poral component to be layered onto the data, and thus with progenitors of European domestic pigs, although whether this any luck a complete understanding of the hows, whens, process was initiated independently of the Near Eastern pig wheres, and maybe even the whys of domestication will be domestication or whether it was kick-started by the intro- within our grasp. duction of Near Eastern pigs remains an open question. The discussions that took place at the “The Beginnings of Beyond demonstrating the use of genetics to reveal the Agriculture: New Data, New Ideas” Wenner-Gren Foundation patterns of movement among a key domestic animal and Symposium in March 2009 in Mexico went a long way toward hence the movement of their human herders, the study by solidifying my impression that the big questions are increas- Larson et al. (2007a) also underlined the dangers of inferring ingly knowable. First, highly precise data regarding the specific historical patterns based on modern data alone. All modern temporal, geographic, and ecological circumstances in which continental pigs in Europe possess European-specific mito- domestic plants and animals became integrated into human chondrial haplotypes. But they only do so today because the settlements are accumulating at an unprecedented pace. This Near Eastern–specific pigs originally brought into Europe level of detail is allowing researchers to piece together the have been completely replaced, leaving no descendants in specific order of events (on a region-by-region basis) that first modern pig populations. Given the number of human mi- set the stage and then allowed for domestication to take place S494 Current Anthropology Volume 52, Supplement 4, October 2011

(e.g., Zeder 2011). Second and equally impressive, the theo- Jansen, T., P. Forster, M. A. Levine, H. Oelke, M. Hurles, C. Renfrew, retical framework for understanding the process of domes- J. Weber, and K. Olek. 2002. Mitochondrial DNA and the origins of the domestic horse. Proceedings of the National Academy of tication at the macrolevel is becoming ever more refined. Sciences of the USA 99:10905–10910. These structures (see Denham 2011) will allow us to place Larson, G., U. Albarella, K. Dobney, P. Rowley-Conwy, J. Schibler, the new data into a scaffold that will facilitate a genuine A. Tresset, J.-D. Vigne, et al. 2007a. Ancient DNA, pig domesti- comprehension of the bigger themes of global domestication cation, and the spread of the Neolithic into Europe. Proceedings on top of their specific regional narratives. These are exciting of the National Academy of Sciences of the USA 104:15276–15281. Larson, G., T. Cucchi, M. Fujita, E. Matisoo-Smith, J. Robins, A. times. Anderson, B. Rolett, et al. 2007b. Phylogeny and ancient DNA of Sus provides insights into Neolithic expansion in Island Southeast Asia and Oceania. Proceedings of the National Academy of Sciences of the USA 104:4834–4839. Larson, G., K. Dobney, U. Albarella, M. Y. Fang, E. Matisoo-Smith, References Cited J. Robins, S. Lowden, et al. 2005. Worldwide phylogeography of wild boar reveals multiple centers of pig domestication. Science Albert, F. W., O. Carlborg, I. Plyusnina, F. Besnier, D. Hedwig, S. 307:1618–1621. Lautenschlager, D. Lorentz, et al. 2009. Genetic architecture of Larson, G., R. R. Liu, X. B. Zhao, J. Yuan, D. Fuller, L. Barton, K. tameness in a rat model of animal domestication. Genetics 182: Dobney, et al. 2010. Patterns of East Asian pig domestication, 541–554. migration, and turnover revealed by modern and ancient DNA. Allaby, R. G., D. Q Fuller, and T. A. Brown. 2008. The genetic ex- Proceedings of the National Academy of Sciences of the USA 107: pectations of a protracted model for the origins of domesticated 7686–7691. crops. Proceedings of the National Academy of Sciences of the USA Ludwig, A., M. Pruvost, M. Reissmann, N. Benecke, G. A. Brockman, 105:13982–13986. P. Castanos, M. Cieslak, et al. 2009. Coat color variation at the Bellwood, Peter. 2011. Holocene population history in the Pacific beginning of horse domestication. Science 324:485. region as a model for worldwide food producer dispersals. Current Marshall, Fiona, and Lior Weissbrod. 2011. Domestication processes Anthropology 52(suppl. 4):S363–S378. and morphological change: through the lens of the donkey and Bradley, D. G., D. E. MacHugh, P. Cunningham, and R. T. Loftus. African pastoralism. Current Anthropology 52(suppl. 4):S397–S413. 1996. Mitochondrial diversity and the origins of African and Eu- Naderi, S., H. R. Rezaei, F. Pompanon, M. G. Blum, R. Negrini, H.- ropean cattle. Proceedings of the National Academy of Sciences of R. Naghash, O. Balkiz, et al. 2008. The goat domestication process the USA 93:5131–5135. inferred from large-scale mitochondrial DNA analysis of wild and Burridge, C. P., D. Craw, D. Fletcher, and J. M. Waters. 2008. Geo- domestic individuals. Proceedings of the National Academy of Sci- logical dates and molecular rates: fish DNA sheds light on time ences of the USA 105:17659–17664. dependency. Molecular Biology and Evolution 25:624–633. Naderi, S., H.-R. Rezaei, P. Taberlet, S. Zundel, S.-A. Rafat, H.-R. Clutton-Brock, J. 1995. A natural history of domesticated mammals. Naghash, M. A. El-Barody, O. Ertugrul, and F. Pompanon. 2007. Cambridge: Cambridge University Press. Large-scale mitochondrial DNA analysis of the domestic goat re- Denham, Tim. 2011. Early agriculture and plant domestication in veals six haplogroups with high diversity. PLoS One 2:e1012, doi: New Guinea and Island Southeast Asia. Current Anthropology 10.1371/journal.pone.0001012. 52(suppl. 4):S379–S395. Piperno, Dolores R. 2011. The origins of plant cultivation and do- Diamond, J. 1997. Guns, germs, and steel: the fates of human societites. mestication in the New World tropics: patterns, process, and new New York: Norton. developments. Current Anthropology 52(suppl. 4):S453–S470. Dobney, K., and G. Larson. 2006. Genetics and animal domestication: Randi, E., and V. Lucchini. 2002. Detecting rare introgression of new windows on an elusive process. Journal of Zoology 269:261– domestic dog genes into wild wolf (Canis lupus) populations by 271. Bayesian admixture analyses of microsatellite variation. Conser- Edwards, C. J., R. Bollongino, A. Scheu, A. Chamberlain, A. Tresset, vation Genetics 3:31–45. J. D. Vigne, J. F. Baird, et al. 2007. Mitochondrial DNA analysis Rompler, H., N. Rohland, C. Lalueza-Fox, E. Willerslev, T. Kuznet- shows a Near Eastern Neolithic origin for domestic cattle and no sova, G. Robeder, J. Bertranpetit, T. Schoneberg, and M. Hofreiter. indication of domestication of European aurochs. Proceedings of 2006. Nuclear gene indicates coat-color polymorphism in mam- the Royal Society B: Biological Sciences 274:1377–1385. moths. Science 313:62. Eriksson, J., G. Larson, U. Gunnarsson, B. Bed’hom, M. Tixier-Boi- Savolainen, P., Y. P. Zhang, J. Luo, J. Lundeberg, and T. Leitner. 2002. chard, L. Stromstedt, D. Wright, et al. 2008. Identification of the Genetic evidence for an East Asian origin of domestic dogs. Science yellow skin gene reveals a hybrid origin of the domestic chicken. 298:1610–1613. PLoS Genetics 4:e1000010, doi:10.1371/journal.pgen.1000010. Stricklin, W. R. 2001. The evolution and domestication of social Fang, M., G. Larson, H. S. Ribeiro, N. Li, and L. Andersson. 2009. behaviour. In Social behaviour in farm animals. L. J. Keeling and Contrasting mode of evolution at a coat color locus in wild and H. W. Gonyou, eds. Pp. 83–110. New York: CABI. domestic pigs. PloS Genetics 5:e1000341. Troy, C. S., D. E. MacHugh, J. F. Bailey, D. A. Magee, R. T. Loftus, Ho, S. Y. W., and G. Larson. 2006. Molecular clocks: when times are P. Cunningham, A. T. Chamberlain, B. C. Sykes, and D. G. Bradley. a-changin’. Trends in Genetics 22:79–83. 2001. Genetic evidence for Near-Eastern origins of European cattle. Ho, S. Y. W., G. Larson, C. J. Edwards, T. H. Heupink, K. E. Lakin, Nature 410:1088–1091. P. W. H. Holland, and B. Shapiro. 2008. Correlating Bayesian date Trut, L. N. 1999. Early canid domestication: the farm-fox experiment. estimates with climatic events and domestication using a bovine American Scientist 87:160–169. case study. Biology Letters 4:370–374. Van Laere, A. S., M. Nguyen, M. Braunschweig, C. Nezer, C. Collete, Ho, S. Y. W., M. J. Phillips, A. Cooper, and A. J. Drummond. 2005. L. Moreau, A. Archibald, et al. 2003. A regulatory mutation in Time dependency of molecular rate estimates and systematic over- IGF2 causes a major QTL effect on muscle growth in the pig. estimation of recent divergence times. Molecular Biology and Evo- Nature 425:832–836. lution 22:1561–1568. Vigne, Jean-Denis, Isabelle Carre`re, Franc¸ois Briois, and Jean Gui- Larson Genetics and Domestication S495

laine. 2011. The early process of mammal domestication in the 1997. Multiple and ancient origins of the domestic dog. Science Near East: new evidence from the pre-Neolithic and pre-Pottery 276:1687–1689. Neolithic in Cyprus. Current Anthropology 52(suppl. 4):S255–S271. vonHoldt, B. M., J. P. Pollinger, K. E. Lohmueller, E. J. Han, H. G. Vila, C., J. A. Leonard, A. Gotherstrom, S. Marklund, K. Sandberg, Parker, P. Quignon, J. D. Degenhardt, et al. 2010. Genome-wide K. Liden, R. K. Wayne, and H. Ellegren. 2001. Widespread origins SNP and haplotype analyses reveal a rich history underlying dog of domestic horse lineages. Science 291:474–477. domestication. Nature 464:898–902. Vila, C., P. Savolainen, J. E. Maldonado, I. R. Amorim, J. E. Rice, R. Zeder, Melinda A. 2011. The origins of agriculture in the Near East. L. Honeycutt, K. A. Crandall, J. Lundeberg, and R. K. Wayne. Current Anthropology 52(suppl. 4):S221–S235.

Current Anthropology Volume 52, Supplement 4, October 2011 S497

The Agricultural Demographic Transition During and After the Agriculture Inventions

by Jean-Pierre Bocquet-Appel

An abrupt increase in fertility has been recorded in data from 200 cemeteries and ethnographic data ranging from the Meso-Neolithic Eurasian center in the Levant to the arctic circle in the North American continent in the twentieth century AD. This shift has been called, synonymously, the Neolithic demographic transition or the agricultural demographic transition (ADT). It is interpreted as the effect on fertility of an abrupt change in maternal energetics that occurs during the transition from a mobile forager economy to a farming economy in any period, whether prehistoric or historical. The primeval prehistoric ADT was a loop of retroactions capable of rapidly raising the rate of population growth and in which the population was both the cause and the effect of the demographic shift. During the eighteenth century AD, new areas of demographic change appeared across this agricultural population area that were characterized by a drop in mortality and then in fertility and were determined by the introduction of new rules of hygiene along with medical and contraceptive techniques. This shift represents the contemporary demographic transition (CDT). The CDT occurred in reverse symmetry with the ADT. A unique phenomenon occurred at the margins of the residual area of the forager system with a quasi coincidence of the effects of both the ADT and the CDT.

During the transition from forager to food-producer econ- preted as expressing a density-dependent demographic pat- omies, the signal of a relatively abrupt shift has been detected tern in which when the population density increases, its in paleoanthropological data from about 200 cemeteries in growth rate, although still positive, decreases asymptotically. Europe and North Africa, the Levant, and North America, as The coincidence of this demographic signal with the eco- well as in archaeological data from Europe and South America nomic transition from foragers to food producers has led us (Bandy 2005, 2008; Bocquet-Appel 2002; Bocquet-Appel and to consider the signal as the signature of a major demographic Dubouloz 2004; Bocquet-Appel and Naji 2006; Bocquet- phenomenon initially named the Neolithic demographic tran- Appel and Paz de Miguel Ibanez 2002; Buikstra, Konigsberg, sition (NDT). The idea of the NDT, either named as such or and Bullington 1986; Guerrero, Naji, and Bocquet-Appel under another name, has also been expressed independently 2008; Kohler and Glaude 2008). In the cemetery data, this by other researchers (Binford and Chasko 1976; Livi-Bacci signal is characterized by a relatively abrupt increase in the 1992; Simoni et al. 2000). As we will see below, through a proportion of immature skeletons, which, as is now well return to certain neglected historical data, the NDT occurred known, mainly expresses an increase in the birthrate (and not only during the prehistoric period but also during his- beyond this in fertility) among populations and not an in- torical times that saw a transition from a nomadic forager crease in mortality (Johansson and Horowitz 1986; McCaa economy to a farming economy (or to the modern Western 2002; Sattenspiel and Harpending 1983). This transition is an economy). What has been called the NDT up to now can indication of a demographic explosion relative to the pre- therefore be generalized as the agricultural demographic tran- ceding forager period. In the archaeological data, the signal sition (ADT), a term that will be employed synonymously of demographic change shows that the rate of increase in with NDT in this article. In the first part of this article, four archaeological remains diminishes as their density increases. signals of the ADT produced at different periods and in dif- This, known as a two-stage pattern (Bandy 2008), is inter- ferent regions from prehistoric times to the twentieth century AD and obtained from archaeological and ethnographic sources are juxtaposed in order to highlight the unique pattern Jean-Pierre Bocquet-Appel is Professor at E´ cole Pratique des Hautes E´ tudes and Research Director for the Centre National de la Recherche of the ADT and to understand the causes of its variation. In Scientific (Upr2147, CNRS, 44, rue de l’Amiral Mouchez, 75014 Paris, the second part, the metabolic load model of maternal en- France [[email protected]]). This paper ergetics developed by Valeggia and Ellison (2004) is put for- was submitted 13 XI 09, accepted 27 XII 10, and electronically ward to explain the rapid increase in fertility during the ADT. published 27 VII 11. The third part addresses various questions such as the impact

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0022$10.00. DOI: 10.1086/659243 S498 Current Anthropology Volume 52, Supplement 4, October 2011 of mortality, which is not directly observable, and the self- presses the kinetics of the opening of the age pyramid and, catalyzing nature of the ADT process. In the final part, the beyond, of an increase in birth and fertility rates. contradictory impacts of the ADT and of the recent contem- When this increase in the birth/fertility rate levels off, dur- porary demographic transition are brought into perspective ing the agricultural period, the plateau indicates the stabili- through a projection involving ethnographic populations liv- zation of the fertility “phase” of the ADT in the geographical ing on the fringes of the world economy. area of the sampled cemeteries. The degree of change in fer- tility during the ADT can then be summarized by two figures:

the value of 15p5—the relative youth of the population—at Signatures of the ADT in Archaeological the plateau and time dt taken to reach it. In most cases, using and Ethnographic Data an absolute chronology would make it difficult to detect the ADT and its phenomenological unity because of the distance Different signals of the ADT have occasionally been obtained, between cemeteries in space and time, which a relative chro- from prehistoric cemeteries for the oldest examples and from nology does allow (for a discussion, see Bocquet-Appel 2002, demographic surveys of ethnographic groups for the more 2008a; Bocquet-Appel and Naji 2006). The use of relative recent. But these signals were not understood as representing chronology makes it possible to bring all the data together different expressions of the unique demographic phenome- whatever their absolute chronology. ADT images are tech- non of the ADT occurring at different times and in different nically less difficult to obtain from ethnographic data directly, however scarce such information may be. places. While the signal of the ADT is easily visible in the Four signals of the ADT are shown in figures 1 and 2. The ethnographic demographic data—expressed by a steep in- first three were obtained from paleoanthropological cemetery crease in the fertility index over a relatively short period of data and the fourth from ethnographic data. These images a few decades—this is not the case in cemeteries, where, in range geographically from the Meso-Neolithic Eurasian center general, the signal is not directly visible to the archaeologist. in the Levant to the arctic circle in the North American con- To bring it out, an archaeometric procedure to detect de- tinent in the twentieth century AD. mographic change in the space-time cemetery data has to be The first signal represents the ADT in the Levant (fig. 1A). used. This provides the underlying trend of an unconventional Its profile shows, during the forager period, first of all, a high demographic indicator expressing how the age pyramid in a value for the demographic indicator p during a period that living population expands with the rate of the transition from 15 5 corresponds in absolute chronology to the Natufian forager a forager economy to a horticulture-farming economy. As system (with p p 0.30 toward dt p 2,600 years). The pro- repeatedly stated, this indicator is represented by the pro- 15 5 file then slopes continuously downward until it falls below portion of 5- to 19-year-old skeletons in a cemetery (named the floor value of a demographically stationary population p to conform with demographic notation) relative to in- 15 5 (fig. 1A, horizontal dotted line) when the profile is interpreted dividuals aged 5 years plus, because skeletons under 5 years in terms of growth rate. The profile then curves back upward, of age are notoriously underrepresented in cemeteries. When once more crossing the stationary value atdt p 0 , which the proportion is high, the age pyramid forms the obtuse corresponds to the beginning of the Pre-Pottery Neolithic A angle of a young population; when it is low, it forms the acute cultural horizon in absolute chronology. At dt p 2,600 angle of a relatively older population. Beyond the opening of years—the limit of the available data—the expected plateau the age pyramid, this indicator expresses the birthrate, the Ӎ of 15p5, with the high value 0.400, is still not reached. The growth rate, and the fertility rate in a stable population model ADT in the Levant corresponds to a relatively slow but con- (Bocquet-Appel 2002; Bocquet-Appel and Naji 2006, fig. 6). tinuous increase in birthrate and fertility. During these 2,600 In order to extract a demographic pattern that is common years, the estimated total fertility rate (TFR; Bocquet-Appel among the space-time data scattered across the map, a tempo and Naji 2006, fig. 6) increases from 4.5 to 10 children per of the economic transition in relative chronology—relative to woman at the end of her reproductive life, that is to say an the economic transition locally—is used, symbolized by dt, increase of 2.1 additional children per millennium. instead of an absolute (historical) chronology. When the in- The second signal shows the ADT in the North American troduction of the farming system is positioned in a locality Southwest (NASW; fig. 1B; Kohler and Glaude 2008), which atdt p 0 , the effect of the relative chronology is to arrange corresponds in absolute chronology to the middle of the first the values of the demographic indicator 15p5 in a new reference millennium AD. In that region, the ADT starts at dt p 250 frame in such a way that whendt ! 0 , the indicator informs years after the introduction of the intensive farming system ≥ p p on forager demography, whiledt 0 informs on horticul- until the profile reaches a plateau value 15p 5 0.30 at dt turist-farmer demography. As we shall see below, the signal 850 years. of the ADT is characterized by a relatively abrupt increase in The third signal shows the ADT at the intercontinental the paleodemographic indicator 15p5 over the tempo of the geographical scale of the Northern Hemisphere (North Amer- relative chronology dt during the economic transition, which ica, North Africa, and Western Europe; fig. 1C) obtained from can be represented graphically by a profile. This profile ex- 133 cemeteries, five being already included in the NASW pro- Bocquet-Appel The Agricultural Demographic Transition S499

Figure 1. Three archaeological signals of the agricultural demographic transition obtained from the paleodemographic indicator for cemetery

data (15p5, vertical axis) plotted on the relative chronology (dt, horizontal axis). A, The Levant epicenter (N p 16 ; Guerrero, Naji, and Bocquet- Appel 2008); B, the North American Southwest (N p 49 ; Kohler and Glaude 2008, with permission); C, the Northern Hemisphere, covering America, Europe, and North Africa (N p 133 , Bocquet-Appel 2008a). In this figure the data are not redundant except for five sites that are common to the last two graphs. The horizontal axis (dt) represents the tempo of the economic transition from foragers to horticulturist-farmers (top) or to intensive farming (bottom). The line indicating the trend is obtained with the Loess local linear fit procedure. Up to a constant, the trend represents the variation of the birthrate in a stable population model. If the trend is interpreted in terms of the growth rate r,the

horizontal dotted line indicates the expected value of 15p5 for a stationary populationr p 0 .

file but not including the Levant. This profile is roughly flat nographic data, the question may arise of the relative im- ! Ӎ during the forager period (dt 0 ), with15p 5 0.22 , until portance of data-point dispersion in the paleodemographic p dt 1,000. It then dips slightly for approximately 600 years indicator 15p5 in the three first graphs of figure 1. There are during the period preceding the ADT. With the introduction at least three reasons: (i) an effect of binomial sampling of p of the agricultural system atdt 0 , the demographic indi- the frequencies 15p5 because of the cemetery sample size, even cator 15p5 in the profile rises relatively abruptly, reaching a though this effect has been limited by keeping only cemeteries p p ϩ plateau 15p 5 0.28 atdt 800 years. with at least 50 skeletons (5 years old); (ii) an archaeological Before moving on to the signal of the ADT in the eth- sampling bias determined by the excavated areas in the cem- S500 Current Anthropology Volume 52, Supplement 4, October 2011

Figure 1. (Continued) eteries, which are seldom excavated in their entirety; and (iii) and not spatially and temporally scattered localities like the extreme and pathological local situations affecting 15p5 because cemeteries. From 1900 to 1950, the birthrate remains at 40 of sudden events (massacres, epidemics, accidents), which add per 1,000. It then rises steeply to 48 per 1,000 within 10 years to the ADT effect. Finally, the dispersion of demographic before collapsing during the contemporary demographic tran- phenomena and their indicators in space and time is a normal sition (CDT) that follows, with its widespread use of contra- situation; it is rather the reverse that is less common, that in ception. As will be seen below, the demographic information spite of the wide dispersion, the force and visibility of an a from the ethnographic literature does not distinguish between priori unpredictable trend in the data reflects an exceptional the impact of the ADT and the CDT on populations, which demographic phenomenon. are grouped together as “transitional populations” (except in The fourth signal shows the ADT in the ethnographic data Binford and Chasko 1976). from Alaska and Canada (fig. 2; Nunamiut Eskimo: Binford Some comments are in order here concerning the variability and Chasko 1976; Canadian Indian and James Bay Indian: of the ADT signature in prehistoric cemeteries. This variability Romaniuk 1981), with the actual birthrate during the tran- provides information regarding the potential causes of the sition as a variable, not the proportion of immature, as used ADT and its tempo and mode of expansion relative to ag- in the cemetery data. Unlike the three preceding archaeolog- ricultural epicenters. It should be noted that both the Lev- ical signals reconstituted in relative chronology, this profile antine and Northern Hemisphere profiles dip during the last can be directly represented in absolute time, for the data forager period preceding the ADT for about dt p 600–800 reflect a succession of demographic states at the same lo- years (fig. 1). In the Levantine profile, this period corresponds cality—like a time-lapse film being shot in the same place— to the final Natufian. Compared with the previous period, it Bocquet-Appel The Agricultural Demographic Transition S501

Figure 1. (Continued) is characterized by “smaller social units,” “increasing mobil- effect of a regional intrusion that could be either populational, ity” (Goring-Morris and Belfer-Cohen 2011), and biological bringing with it a new economic system (with its technical hardship (Belfer-Cohen, Schepartz, and Ahrensburg 1991; and cultural tool kit) and demographic regime, or only cul- Smith and Horwitz 2007), all of these being attributed to the tural (technical tool kit only). In the overall image of the impact of the rapid onset of the Younger Dryas (Goring- Northern Hemisphere, the onset of the ADT coincides with Morris and Belfer-Cohen 2011). The profile shows the pop- the introduction of the new farming system atdt p 0 , but ulation’s response in terms of fertility to this environmental this is not the case in the NASW, where it occurs at dt p change. In light of the metabolic load model (see below Val- 250 years. Finally, it is noticeable that in the archaeological eggia and Ellison 2004), an increase in mobility must be ex- data, if the profiles are interpreted in terms of growth rate pected to induce an increase in the energy expenditure of via paleodemographic estimators (Bocquet-Appel 2002), they mothers. This increasing mobility perhaps also reveals a re- all show a positive rate except during the short Natufian pre- duction in the high-calorie food items—wild cereals and other ADT period in the Levant, as indicated above. leguminous plants in the diet, relative to the low-calorie food Two criticisms have occasionally been made against the p items obtained by hunting. In any case, as will be seen below, 15 5 values, one taphonomic/cultural and the other statistical. The even with no change of food, the increase in energy expen- taphonomic/cultural criticism is that the low p values ob- diture induced by an increase in mobility has the effect of 15 5 served in the forager cemeteries could be a data artifact re- decreasing fertility in mothers, evidence also noted 40 years sulting from the mobility of foragers, who may have failed ago by Sussman (1972). There is a noticeably high degree of to bury immature skeletons systematically in their cemeteries. coincidence between the interpretation of the archaeological data and the observed demographic profile in the Levant. For that criticism to be valid for any time and place, it would All four images show the same signal of a relatively abrupt be necessary to provide an explicit link between selective buri- increase in the demographic indicator. In comparison with als omitting 5–19-year-olds—and not other age classes—and the others, the Levantine pattern confirms the prediction of mobility, which has never been produced up to now to the 1 a much slower ADT tempo in the zones of primary invention best of my knowledge. This assertion is therefore unproven. of the Neolithic tool kit (technical, social, and political) com- 1. However, an increase in the frequency of 5–19-year-olds inversely pared with the secondary expansion zones (Bocquet-Appel proportional to the cemetery sizes and producing a systematically higher 2002). The ADT in the NASW and the Northern Hemisphere 15p5 value in small cemeteries than in large ones has been observed in a shows a relatively abrupt demographic shift, suggesting the sample of 68 European and North African cemeteries extracted from the S502 Current Anthropology Volume 52, Supplement 4, October 2011

The statistical criticism is that the size of the 15p5 sample, which is appreciably smaller for the forager group than for the farmers, casts doubt on the equal representativeness of the two groups and therefore on the quality of the profile estimation in the forager zone. To remove this doubt, the 15p5 sampling points must be understood as having been deter- mined by the corresponding demographic densities of the ADT with a pre-ADT period at low forager density and there- fore a low density of sampled points followed by an ADT period with a high farmer density and therefore a higher density of sampled points. The unequal sampling of 15p5 points for the two periods reflects the corresponding demographic densities in the profiles from which the sampling comes, which are themselves unequal. There is nothing that can be done about this. Figure 2. Signal of the agricultural demographic transition with Explaining the Agricultural Demographic the demographic data (birthrate) for Canadian Indians (Ro- maniuk 1981) and Nunamuit Eskimo (Binford and Chasko Transition: The Metabolic Load Model 1976). The line indicating the trend is obtained with the Loess local linear fit procedure. What is called the ADT here is in fact the positive effect on fertility of a relatively abrupt change in maternal energetics that occurs mainly during the transition from a mobile forager In the context of a transition from a mobile forager econ- economy to a farming economy in any period, whether pre- omy to a farming economy with (i) a reduction in energy historic or historical (Bocquet-Appel 2008a). Let us recall that intake in the proportion of low-calorie and null-carbohydrate for a fixed reproductive duration of roughly 35 years, the food items (tissues from hunted animals, !150 kcal; carbo- fertility level can be expressed by the duration of the birth hydrate, 0) relative to high-calorie and carbohydrate food interval. During the fertile life of a mother, when the duration items (1300 kcal; carbohydrate, 10; see table 1; in Eurasia: of the birth interval increases, the number of children born lentils, millet, peas, rice, wheat; in America: beans, cheno- decreases. The duration of the birth interval is inversely pro- podium, hickory nuts, sunflower, maize; for a summary, see portional to fertility. This duration is a function of the energy Pickersgill 2007) and (ii) a reduction in energy expenditure balance (energy status and energy balance). The energy bal- in the physical activity involved in mobility and in the ma- ance is determined by energy expenditure (on necessary milk ternal stress of child transportation (Bleek 1928; Blurton Jones production and physical activity) and postpartum energy in- 1986, 1987, 1989, 1994; Lee 1972, 1979), an increase is to be take (mother’s diet). The relationship between energy balance expected in what might be called the fertility energetics, rep- and fertility can be explained by the relative metabolic load resented by the (inverse) duration of postpartum amenor- model, in which the relative energy change in the energy rhoea and vice versa. This suggests that there is a function balance during gestation is the determinant variable (Dufour connecting natural fertility to the energy balance in all pri- and Sauther 2002; Ellison 1991; Ellison et al. 1993; Huffman mates and even in mammals. Another effect of fertility en- et al. 1987; Lunn et al. 1984; Valeggia and Ellison 2004; see ergetics must be noted: its impact on the onset of menstru- also Hurtado and Hill 1990). The nursing frequency is a nec- ation, which is influenced by body fat (Baker 1985; Biro et essary but not sufficient signal determining the duration of al. 2003). The consequence of this impact is to decrease the postpartum amenorrhea. age of the onset of reproductive life and, probably, to extend its duration. For the same reasons of fertility energetics given literature (Bocquet-Appel 2002). This sampling effect has been deter- above, it can be inferred that during the ADT, menstruation mined to be an error of interpretation on the part of the archaeologists in teenagers in farming economies began earlier than in mo- investigating small cemeteries in which no children are buried, which bile foragers, which, together with the smaller birth interval, were wrongly indexed in the publications under “series” instead of “cem- would tend to increase fertility (see also Cohen 2008). The eteries,” making them unidentifiable. To archaeologists, a natural cem- etery, whatever its size, must contain children even though probabilis- birth interval is equal to the duration of postpartum amen- tically cemeteries with no children are perfectly possible (see orrhoea and the possible duration of the postpartum sex ta- Bocquet-Appel 2008b). In the sample of 68 cemeteries, this anomaly boo, the latter being significantly longer, on average, among caused those with fewer than 50 skeletons (5 yearsϩ) to be eliminated ethnographic farmers than foragers (Bocquet-Appel 2008a; (i.e., 29 Neolithic cemeteries out of 62 [47%] against only 1 Mesolithic Saucier 1972). A minimal estimate of the postpartum taboo cemetery out of 6 [16%]; see Bocquet-Appel 2002). If a taphonomic/ from the semiquantitative data in Murdock’s ethnographic cultural bias truly exists, the paleodemographic indicator 15p5 is not related to the economic subdivision of cemeteries but probably to normative atlas (Gray 1999) gives 7.2 months for the former and 11.4 prejudices among archaeologists. months for the latter (Bocquet-Appel 2008b). From the be- Bocquet-Appel The Agricultural Demographic Transition S503

Table 1. Nutrients database for selected game animals and domesticated plants in Eurasia

Food kcal/g Protein (g) Carbohydrate (g) Lipid (g) Game: Eurasia: Bisona 223–146 20–19 0 16–7 Rabbit (wild)a 114 22 0 2 Reindeer (caribou)a 127 23 0 0 Deera 157 22 0 0 Goat 143–109 27–21 0 3–2 Horsea 133–110 22 0 3 Roe deer 120 22 0 4 Wild boara 122 22 0 3 Mesoamerica and North America: Antelope (Americana)a 114 22 0 2 Beavera 146 24 0 5 Raccoon, roasteda 255 29 0 14 Squirrela 120 21 0 3 Plants: Eurasia: Lentils 338 24 56 2 Milleta 378 11 73 4 Peas (seeds)a 341 25 60 1 Rice (short grain)a 358 6 79 1 Wheat (seeds) 342 13 67 3 Mesoamerica and North America: Bean common (Phaseolus vulgaris)a 339 22 61 1 Chenopodium berlandierib 347 Cucurbita pepo ssp. (summer)a 16 1 3 0 Hickory (dried)a 657 13 18 64 Maize (corn flour, yellow) 375 6 83 1 Maranta (arrowroot)a 65 4 13.3 .2 Sunflower (seeds)a 595 24 19 47 Sources. Unless otherwise noted, the data are from http://www.tabledescalories.com/aliments-boissons-o-p1.html. Note. All nutrient quantities are in grams per 100 g of food. a http://www.nal.usda.gov/fnic/foodcomp/search/. b Gremillion 2006. ginning of postpartum amenorrhea and any overlapping post- 171 populations of Jorgensen’s western North American In- partum sex taboo, the duration of the birth interval is not dians (WNAI; Bocquet-Appel 2008a; Jorgensen 1980, 1999). the sum of these two lengths of time but the duration of When the usual constraints of ethnographic data are taken whichever is longer—postpartum amenorrhea or the post- into account, WNAI data have produced reasonably satisfac- partum taboo—plus 9 months of gestation. It can therefore tory semiquantitative information on mobility and food but be seen that in the birth interval, if the fertility energetics not on fertility. The only available demographic variable is partially determine total fertility, a second part is determined density. These data show that the variation in demographic by the social use of the postpartum taboo, which should not density is not correlated with the variation of venison or a priori be subject to variations in the energy balance. In the fishery product consumption (low-calorie food) but is cor- ethnographic data, therefore, there are two contradictory ef- related with variation in the consumption of agricultural fects, neither being directly visible, on the total duration of products (high-calorie food) and/or with mobility (energy the birth interval: the average duration of postpartum amen- expenditure). On the whole, demographic density in WNAI orrhoea (fertility energetics) is longer among foragers than data, taken as a proxy for fertility, is distributed as expected among horticulturist-farmers, while the estimated average with the metabolic load model. Nevertheless, to link demo- (minimum) duration of the postpartum taboo is, on the con- graphic density to fertility, the following conditions must be trary, shorter, of (11.4 Ϫ 7. 2 p) 4.2 months, among foragers met: the population is subdivided into subpopulations (with than among farmers. All in all, these two contradictory effects same date of origin for the subdivisions), migrations are mar- increase the birth interval duration among horticulturist- ginal, the subpopulations have the same age distributions, and farmers because of their longer postpartum taboo despite mortality is constant. The whole is similar to the Island Model shorter postpartum amenorrhoea. in population genetics. Under that Island Model, the within- An attempt to evaluate the predictions of the metabolic subdivision demographic density is determined by the input load model has been made with the ethnographic sample of to the subpopulation (i.e., by fertility). But these stringent S504 Current Anthropology Volume 52, Supplement 4, October 2011 conditions are certainly not simultaneously present in the critical analysis of the data (Bentley, Goldberg, and Jasienska WNAI data, which make them rather unsatisfactory. We must 1993; Bentley, Jasienska, and Goldberg 1993) has shown that therefore rely on the results of Valeggia and Ellison (2004) they were frequently approximate. After their audit, Bentley, and on Binford and Chasko’s (1976) relatively detailed eth- Jasienska, and Goldberg (1993) reached the conclusion that nographic, demographic, and food data. the average number of children of postmenopausal mothers (also called TFR) among intensive farmers is higher than Discussion among nonfarmers but with wide within-group variance, making it impossible to predict the fertility level for a known Four main issues are discussed here. Why give the name of subsistence group. Bentley and colleagues’ study (Bentley, ADT to this abrupt demographic historical shift in human Goldberg, and Jasienska 1993; Bentley, Jasienska, and Gold- fertility and not sedentarization or NDT? Is ADT an inevitable berg 1993) was relatively influential in demographic anthro- transitional stage in the demography of mobile foragers in pology and, coupled with the extreme scarcity of ethnographic contact with the new farming system (or world economy)? data on natural fertility, brought matters to a conclusion of With no direct data on mortality, how can it nevertheless be a sort that was apparently final. When projected onto the represented from existing ecological demographic models? impact of farming on forager fertility, Bentley et al.’s conclu- What is the scenario of the original emergence, during pre- sion could be interpreted, wrongly, as showing the farming historic times, of the new parameter values for maternal en- impact as negligible at best and dubious at worst. ergetics that generated the ADT? However, besides the arguable grouping of populations,2 the actual data of Bentley and colleagues are not without 3 Neolithic, Agricultural, or Sedentarization Demographic minor errors. Specifically, the impact of the ADT cannot be Transition? deduced by comparing averages of fertility indexes for eth- nographic forager/farmer groups who are alien to each other. As the biological cause underlying the considerable increase The ADT is a within-population shift. To detect the impact in individual female fertility was determined by a change in of the ADT, what needs to be compared are not the absolute the metabolic load with, first of all, a reduction in the energy values of the averages of fertility indexes between subsistence expenditure induced by the sedentarization of foragers, why groups but the rates of change within those same populations not refer to this historically abrupt change in fertility as a when they are experiencing an economic transition. The ques- sedentarization demographic transition rather than an NDT tion that should be raised is, among the transitional popu- or an ADT? Sedentariness in itself, although it moves the lations (Bentley, Jasienska, and Goldberg 1993, table 4), how cursor of fertility energetics forward, has had a limited de- many experienced a shift in fertility that (i) was positive, mographic impact as far as can be judged from the archae- indicating the ADT; (ii) was negative, indicating the CDT; or ological data in the Levant and the shell middens (see below). (iii) was both positive and negative in succession? It is to be Moreover, not only is the carrying capacity of the sedentary expected that all the transitional populations experienced i or forager system rapidly attained, judging by the Natufian ex- iii? But the transitional data gathered by Bentley, Jasienska, ample, but also the system itself is unstable: it generates de- and Goldberg (1993) are heterogeneous and cannot be ex- mographic growth that cannot sustain itself beyond the spe- ploited directly. cific biogeographical local conditions. Its expansion capacity In ethnographic surveys, when demographic quantitative is almost nonexistent. This is not the case with the farming data exist, the shift can be evaluated simply by the within- system, where both the expansion capacity and the potential population ratio of the fertility index before and during (or carrying capacity per unit area are of considerable orders of after) the economic shift. When this ratio (which is a rate) magnitude culminating in a farming system that is still the is greater than 1, the ADT has occurred. But such data at two primary sector of the world economy. Finally, NDT, although points in time for the same group are rare. The authors of it clearly indicates the chronology of this major historical ethnographic surveys have generally given impressionistic ac- change in human evolution, is hardly usable outside an ar- chaeological context, whereas the phenomenon that it is sup- 2. Because of the usual scarcity of forager data, Bentley, Jasienska, and posed to name is still observable in the twentieth century. Goldberg (1993) mention the creation of the “forager-horticulturists” group “partly to increase the sample size for the foraging category” (272, n. 19). If we consider only the “foragers” in this group (and not the “forager/horticulturalist” aggregate), where the number of mothers is ≥ The ADT: An Inevitable Transitional Stage? 30, average fertility among foragers is curiously closer to that of farmers than that of horticulturists. More than 20 years ago, authors seeking to measure the fer- 3. Lapps, Sa´pmi, Sweden, population reported as foragers instead of tility differential between foragers, horticulturists, and farmers herders (P. Sko¨ld, “Aspects on the demographic development in life and death in Sa´pmi: demographic aspects on the history of the Sami in in natural-fertility populations from ethnographic data northern Sweden,” unpublished manuscript, p. 2), which shows the dif- (Campbell and Wood 1988) observed an absence of significant ficulties involved in gathering reliable information, often from ill-defined differences in average fertility between these three groups. A sources in the ethnographic literature. Bocquet-Appel The Agricultural Demographic Transition S505 counts (i.e., qualitative accounts of demographic shifts using taneously. Binford and Chasko’s study did not receive the descriptions of populations growing or decreasing since con- coverage it deserved, perhaps because it was concerned with tact). These demographic shifts proceed along with the usual Eskimo populations in Alaska, which are far removed in space associated economic changes, notably sedentism (or semi- and time from the invention of agriculture, and also because sedentism), together with food support provided by a gov- of the contradictory demographic impacts on the Nunamiut ernmental or religious agency and the keeping of registers on of the ADT and CDT, which occurred almost simultaneously individuals. It is therefore legitimate to take this information during a single generation.6 I will not keep the authors’ term into account. In order to work only with populations in which “first demographic transition” to refer to the abrupt increase changes are due to their fertility and not to their mortality, in Nunamiut fertility because this closes the door to other immunized populations were excluded. Finally, if we consider hypothetical demographic transitions in human history before as minor a possible migratory effect on individuals of neigh- the invention of agriculture, such as, possibly, when tools and boring populations because of the attraction/repulsion of the fire were invented. point of contact with the world economy (places surveyed), Another effect on fertility of a change that can be attributed then the demographic change in a population, described in to maternal energetics occurred marginally during episodes an apparently impressionistic way, can be reasonably attrib- of sedentism among populations that continued to forage for uted to a variation due to fertility. aquatic resources during the Holocene, such as the southern By using the quantitative and qualitative criteria described Scandinavian Ertebølle (Rowley-Conwy 1984, 1998), the above, the nature of the demographic changes experienced American Northwest Coast Indians (Ames and Maschner by Bentley, Jasienska, and Goldberg’s (1993, table 4) transi- 1999), and the Calusa of southern Florida (Widner 1988). tional populations can reasonably be inferred: ADT, CDT, or With aquatic resources, there is no qualitative change in en- unknown (see table 2). Among the 15 so-called transitional ergy intake for the caloric density compared with hunter- populations for which the information was present, seven gatherers. But in becoming sedentary or semisedentary (i.e., populations out of seven experienced the ADT (Nunamiut, in reducing their relative energy expenditure), these foragers Asmat, Tiwi, Hall Beach Eskimos, Wainwright Eskimos, partially affected their maternal energetics and their influence Northern Territory Aborigines, Athapascan), six out of seven on fertility, prompting an upward trend. Such was the bio- experienced the CDT (Nunamiut, Asmat, Tiwi, Wainwright logical cause of demographic growth among sedentary shell- Eskimos, Shipibo-Conibo, Athapascan; uncertain: Navajo), midden foragers relative to mobile foragers. and five out of six successively experienced both the ADT Postpartum abstinence is a historical creation that has cer- and CDT transitions (Nunamiut, Asmat, Tiwi, Wainwright tainly not always existed. The great apes do not abstain from Eskimos, Athapascan; see table 2). The ADT is thus indeed sex. During the ADT, with the rise in fertility (energetics), a necessary stage in the transition from nomadic foragers to the birth interval decreased considerably at the same time. sedentary farmers. Fertility among nomadic foragers always This leads us to assume that during the first phase of rapid increases with the impact of the agricultural (or world) sys- population growth, the inhibiting mechanism of postpartum tem, whatever its absolute values, which can be higher or abstinence did not yet exist and that it was invented during lower from regional population to regional population for or after the primeval ADT. It was incorporated among the uncontrolled reasons,4 the first two candidates being the post- positive checks of the Malthusian model as a new density- partum taboo and, perhaps, the maternal energetics deter- dependence mechanism along with those that must have ap- mined by the regional economic system in its ecological set- peared consecutively to the new demographic situation. I cur- ting.5 rently do not see which data would make it possible to test If the metabolic load explanatory model is true, then var- this assumption. Future bioanthropological indicators on iation in maternal energetics was the cause of the ADT in the skeletons will perhaps make it possible to estimate, from the initial transition zones from foragers to producers in ancient cemeteries, the ages at which a mother was pregnant and (pre)historic periods and also, repeatedly, throughout history therefore to open the door to comparisons of variations be- and up to the present, whenever populations have been ab- tween populations in prehistoric economic systems. sorbed at the frontiers of expansion of what is currently the world economy. This recent effect was detected and detailed for the first time, to the best of my knowledge, by Binford and Chasko (1976) among the Nunamiut, for which we have Expressing Mortality “transitional” information on reproduction and food simul- The glaring data gap in the ADT signal is the invisible impact of mortality in cemetery data. It is masked by the well-known 4. See also other cases in Bentley, Goldberg, and Jasienska (1993, n. confusing effect of the birthrate in cemetery data (Johansson 10). 5. Bentley, Goldberg, and Jasienska (1993) indicate that they did not retain populations where the postpartum taboo lasts longer than 6 6. The Tiwi (Australia) were also affected by the impacts in rapid months (n. 16). But in the reference publication on the 15 transitional succession of the ADT and CDT during the years from 1929 to 1996 nonagricultural societies (table 4), these data are missing. (Peterson and Taylor 1998). S506 Current Anthropology Volume 52, Supplement 4, October 2011 and Horowitz 1986; Sattenspiel and Harpending 1983). In the can be inferred (see Bocquet-Appel 2008a). The underlying ethnohistorical demographic data, the impact of mortality is trend toward a decline in health during the shift to agriculture either extremely high, corresponding to the introduction of (Bocquet-Appel, Naji, and Bandy 2008, fig. 4; Cohen and new pathogens during contact, or artificially low, with the Armelagos 1984; Cohen and Crane-Kramer 2007) is an in- introduction of medical services in contemporary times. dication that tends to support this assumption. The assump- There is no natural reference for ADT mortality other than tion of a rapid return to a rising mortality rate is also sup- the epidemic crisis by contact or its eradication. Therefore, ported by the density-dependent demographic model (Blum, on this point, we have to rely on indirect data and theory. Bonneuil, and Blanchet 1992; Lee 1987; Reher and Ortega Mobility spares foragers from a lot of problems that arise Osona 2000; for a discussion, see Wilson and Airey 1999; for from sedentary village life and the growth in local population mammalians, see Sibly and Hone 2003). As I said above, a density and determined permanent promiscuity (with small density-dependent pattern has been detected in the archae- commensal rodents, feces and the absence of latrines, lack of ological data representing the increase in site density in Peru clean drinking water, contamination from humans and ani- (Bandy 2005, 2008). Finally, unless we assume an exponential mals living in close proximity in enclosed spaces). The mor- population increase over a relatively long duration, eventually tality rates inherited from foragers eventually rise, particularly reaching a cosmic number, we have to postulate that a rising in children under 5 years of age, with reduced breast-feeding. death rate soon—within a few generations—followed the rise Candidate germs by “epidemiological inference” from current in birthrate, producing the historical growth rate typical of preindustrialized areas and those with poor health facilities agricultural populations (1–2 per 1,000).

Table 2. Identification of impacts of the agricultural demographic transition (ADT) and the contemporary demographic transition (CDT) in transitional populations

Populationa Source Remarks ADT CDT Navajo, Ramah Morgan 1973 Agriculturist ... ? Lapps, Sweden Fraccaro 1959 Reindeer herders (P...... Sko¨ld, unpublished manuscript, p. 2) Sioux-Ojibwa, Great Boas 1894 Poor data ...... Plains Nunamiut, Alaska Binford and Chasko 1976 ... Yes; mother-child ratio: Yes 1965–1969 : 352.9; 1945–1949 : 192.3 Asmat, Irian Jaya Van Arsdale 1978 Semisedentary (Van Ars- Yes Yes dale 1978:435, 437) Karkar, Papua New Stanhope and Horna- Domesticated pig (Hor- ...... Guinea brook 1974 nabrook, Serjeantson, and Stanhope 1977: 380) Tiwi, Bathurst and Mel- Gray 1985:25 ... Yes Yes; TFR: 1967–1971 : ville Islands 5.7; 1972–1976 : 4.4 Aborigines, Australia Kirk 1981 Poor data ...... Eskimos, Hall Beach McAlpine and Simpson Sedentary (McAlpine and Probable; TFR: 1930– ... 1976 Simpson 1976:115); 1971 : 11.1 very small sample of 15 postmenopausal mothers in 1971 Eskimos, Wainwright Milan 1970 ... Yes Yes Eskimos, Kuskokwim Hrdlicka 1936 No comparative data lo- ...... cally Aborigines, Australia Jones 1965:242, 1972:253 F. L. Jones, personal cor- Probable; TFR: 1960– No respondence, October 1962 : 4.20; 1967– 2009; Rowley 1972 1968 : 6.60 Shipibo-Conibo, Ucayali Hern 1977:356 Swidden agriculture ... Yes River Athapascan, Old Crow Roth 1981:415 ... Yes; 1916–1940 : 4.7; Yes 1941–1950 : 7.37 Aborigines, Northern Jones 1965, 1972 Duplicate information on ...... Territory, Australia Aborigines, Australia Note. TFR p total fertility rate. a From the list made by Bentley, Jasienska, and Goldberg (1993:783). Bocquet-Appel The Agricultural Demographic Transition S507

Maternal Energetics and the Emergence of the ADT produced not one of the innumerable Malthusian crashes that occurred during prehistory but what became the ADT because The scenario of the ADT at its onset might have begun with the resources of Natufian foragers in the process of becoming a steady trend toward demographic densification of the for- farmers were able to support this growth. ager’s world (Cohen 1977; in North America, see Doran 2007: The initial colonization of North America, perhaps over a 40, table 3.1), which was at quite a low population density few hundred years, by mobile hunter-gatherers who “must (Bocquet-Appel et al. 2005) because of the demographic fra- have had high reproductive rates while maintaining a very gility of ungulate herds (Stiner et al. 2008), producing pockets mobile lifestyle” (Surovell 2000:494) might seem to contradict of sedentary (or semisedentary) foragers (Bar-Yosef and Bel- this energetics-based fertility model. Surovell’s simulations fer-Cohen 1989). These are located in zones that were eco- show that annual total mobility (residential ϩ logistic) is logically favorable to a broadening of the food spectrum, in- substantially lower, with higher rather than lower residential cluding cultigens (Flannery 1969; Stiner 2004), in Eurasia at mobility producing the relatively lower energy expenditure the end of the Pleistocene (Belfer-Cohen and Bar-Yosef 2000; that is compatible with high fertility (Surovell 2000). Weiss et al. 2004) in Mesoamerica during the early Holocene (between 10,000 and 7000 BP; see Piperno 2011). A brief calculation shows that wild megafauna could not continue to Concluding Remarks fuel continuous human demographic growth.7 This socio- Starting in the continental epicenters of prehistoric agricul- natural configuration has the effect of raising maternal en- tural invention, in the 12–5-ka BP window, the ADT ex- ergetics and fertility soon followed by a mortality transition panded populations geographically up to the marginal and toward these same high values for the reasons of density de- peripheral limits of the forager system of the twentieth century pendence indicated above. AD. At an evolutionary scale, as determined by maternal en- The population was thus both the cause and the effect of ergetics, the ADT has produced an increase in the human the demographic shift: the cause—because as it increased the metapopulation that results from a (very high) fertility gain pressure on the carrying capacity of the hunter-gatherer pro- over (high) mortality as well as a massive cultural complex- duction system, the population also increased the probability ification, the whole being unprecedented. The ADT generated of a systemic transition—and the consequence—because as the demographic regime of the agricultural population, also soon as the new economic system appeared, the population called traditional or preindustrial populations. tended to expand toward the limits of the new carrying capacity During the eighteenth century AD, nearly 12 ka after the (Wood 1998) of the horticulturist-farming system. The pri- onset of the ADT, new areas of demographic change appeared meval prehistoric ADT was a positive feedback loop capable of across this agricultural population area that were considerably rapidly raising the growth rate locally or regionally (Bocquet- more abrupt than at the time of the previous demographic Appel 2008a; Bocquet-Appel and Naji 2006). This certainly transition, located in France and Massachusetts (Smith 1972; explains why the tempo of the ADT signatures in the peripheral Temkin-Greener and Swedlund 1978). In these zones, the new zones of expansion in Europe and North Africa are three to shift is initially characterized by a drop in mortality and in four times faster (400–800 years dt) than in the original Lev- fertility, in rapid succession or actually merging, determined antine epicenter (12,800 years dt). These primary centers were by the introduction of new rules of hygiene and medical and the sources of expansions proceeding according to different contraceptive techniques correlated with but with no direct patterns (continuous geographic front, discontinuous, leaps by link to the Industrial Revolution that was then primarily lo- individual families or individuals as in Amazonia today). In the cated in England (Dupaˆquier 1998:13, fig. 3). This shift rep- Levant, the doubling of fertility over 3,000 years was initially resents the CDT. Whether in the order of variation of de- determined by the transgression by sedentary Natufian foragers mographic variables ([a] mortality shift, followed by or of the evolutionary demographic regulation of mobile foragers simultaneously with [b] fertility shift) or in the direction of to the local carrying capacity, producing unprecedented de- the shift (decreasing and not increasing), the CDT occurred mographic growth. In certain groups, this demographic growth in reverse symmetry with the ADT. In the latter, as it was caught up by the extremely rapid expansion of the CDT, a 7. A simple example shows how the size of a bison metapopulation has unique phenomenon occurred at the margins of the residual to increase when the size of the human predatory population is increasing. area of the forager system with the quasi coincidence of the Meriwether Lewis reported that their expedition of 43 people consumed 1 effects of the ADT and CDT, the latter canceling the effects bison/day (Lewis and Clark 1805), making 365 bison/year for 43 humans or up to 3,560,000 bison/year for 430,000 humans. To simplify, if this of the former. predation is funneled in the death rate d of a stationary population—which corresponds to the inverse of the life expectancy at birth e0—then the size of the bison metapopulation corresponding to this death rate is P p # p d ee00. If we accept13.5 years for bison in the wild (Wilson, Hills, Acknowledgments and Shapiro 2008), then for a population of 43, 4,300, and 430,000 human predators, one obtains a bison population of 4,927, 492,750, and 49,275,000 Thanks to Bryan Hayden (Simon Fraser University), Claude bison, respectively, which is increasingly unrealistic. Masset and Lyliane Rosetta (Centre National de la Recherche S508 Current Anthropology Volume 52, Supplement 4, October 2011

Scientific, Paris), and two anonymous reviewers for their com- ———. 1994. A reply to Dr. Harpending. American Journal of Phys- ments. Thanks also to Frank Jones (Australian National Uni- ical Anthropology 93:391–397. Boas, F. 1894. The half-blood Indian: an anthropometric study. Pop- versity and University of Queensland) and K. Morgan (McGill ular Science Monthly 45:761–770. University) for the information they kindly provided. I am Bocquet-Appel, J. P. 2002. Paleoanthropological traces of a Neolithic grateful to Ofer Bar-Yosef and T. Douglas Price for their in- demographic transition. Current Anthropology 43(4):637–650. vitation to the stimulating conference “The Origins of Ag- ———. 2008a. Explaining the Neolithic demographic transition. In riculture: New Data, New Ideas” and to the Wenner-Gren The Neolithic demographic transition and its consequences.J.P.Boc- quet-Appel and O. Bar-Yosef, eds. Pp. 35–56. Dordrecht: Springer. Foundation for its support. ———. 2008b. The Neolithic demographic transition, population pressure, and cultural change. Comparative Civilizations Review 58: 36–49. References Cited Bocquet-Appel, J. P., P. Y. Demars, L. Noiret, and D. Dobrowsky. 2005. Estimates of Upper Palaeolithic meta-population size in Eu- Ames, K. M., and H. D. G. Maschner. 1999. The peoples of the north- rope from archaeological data. Journal of Archaeological Science 32: west coast: their archaeology and prehistory. London: Thames & 1656–1668. Hudson. Bocquet-Appel, J. P., and J. Dubouloz. 2004. Expected palaeoan- Baker, E. R. 1985. Body weight and the initiation of puberty. Clinical thropological and archaeological signal from a Neolithic demo- Obstetrics and Gynecology 28(3):573–579. graphic transition on a worldwide scale. Documenta Praehistorica Bandy, M. S. 2005. New World settlement evidence for a two-stage 31:25–33. Neolithic demographic transition. Current Anthropology Bocquet-Appel, J. P., and S. Naji. 2006. Testing the hypothesis of a 46(suppl.):S109–S115. worldwide Neolithic demographic transition: corroboration from ———. 2008. Global patterns of early village development. In The American cemeteries. Current Anthropology 47(2):341–365. Neolithic demographic transition and its consequences. J. P. Bocquet- Bocquet-Appel, J. P., S. Naji, and M. Bandy. 2008. Demographic and Appel and O. Bar-Yosef, eds. Pp. 333–358. Dordrecht: Springer. health changes during the transition to agriculture in North Amer- Bar-Yosef, O., and A. Belfer-Cohen. 1989. The origins of sedentism ica. In Recent advances in paleodemography: data, techniques, pat- and farming communities in the Levant. Journal of World Prehis- terns. J. P. Bocquet-Appel, ed. Pp. 277–292. Dordrecht: Springer. tory 40(3):447–498. Bocquet-Appel, J. P., and M. Paz de Miguel Ibanez. 2002. Demografia Belfer-Cohen, A., and O. Bar-Yosef. 2000. Early sedentism in the de la diffusion neolitica en Europe y los datos paleoantropologicos. Near East: a bumpy road to village life. In Life in Neolithic farming Saguntum 5:23–44. communities: social organization, identity, and differentiation. Ian Buikstra, J. E., L. W. Konigsberg, and J. Bullington. 1986. Fertility Kuijt, ed. Pp. 19–37. New York: Plenum. and the development of agriculture in the prehistoric Midwest. Belfer-Cohen, A., L. A. Schepartz, and B. Ahrensburg. 1991. New American Antiquity 51(3):528–546. biological data for the Natufian population in Israel. In The Na- Campbell, K. L., and J. W. Wood. 1988. Fertility in traditional so- tufian culture of the Levant. O. Bar-Yosef and F. R. Valla, eds. Pp. cieties. In Natural human fertility: social and biological mechanisms. 411–424. Ann Arbor, MI: International Monograph in Prehistory. P. Diggory, S. Teper, and M. Potts, eds. Pp. 39–69. London: Mac- Bentley, G. R., T. Goldberg, and G. Jasienska. 1993. The fertility of millan. agricultural and non-agricultural traditional societies. Population Cohen, M. N. 1977. The food crisis in prehistory. New Haven, CT: Studies 47(2):269–281. Yale University Press. Bentley, G. R., G. Jasienska, and T. Goldberg. 1993. Is the fertility ———. 2008. Implications of the NDT for world wide health and of agriculturists higher than that of nonagriculturists? Current An- mortality in Prehistory. In The Neolithic demographic transition thropology 34(5):778–785. and its consequences. J. P. Bocquet-Appel and O. Bar-Yosef, eds. Binford, L. R., and W. J. Chasko. 1976. Nunamiut demography his- Pp. 481–499. Dordrecht: Springer. tory: a provocative case. In Demographic anthropology: quantitative Cohen, M. N., and G. J. Armelagos, eds. 1984. Paleopathology at the approaches. E. B. W. Zubrow, ed. Pp. 63–143. Albuquerque: Uni- origins of agriculture. Orlando, FL: Academic Press. versity of New Mexico Press. Cohen, M. N., and G. M. M. Crane-Kramer, eds. 2007. Ancient health: Biro, F. M., A. W. Lucky, L. A. Simbartl, B. A. Barton, S. R. Daniels, skeletal indicators of agricultural and economic intensification. R. Striegel-Moore, S. S. Kronsberg, and J. A. Morrison. 2003. Pu- Gainesville: University Press of Florida. bertal maturation in girls and the relationship to anthropometric Doran, G. H. 2007. A brief continental view from Windover. In changes: pathways through puberty. Journal of Pediatrics 142(6): Ancient health: skeletal indicators of agricultural and economic in- 643–646. tensification. M. N. Cohen and G. M. M. Crane-Kramer, eds. Pp. Bleek, D. F. 1928. The Naron: a bushman tribe of the central Kalahari. 35–51. Gainesville: University Press of Florida. Publications of the School of African Life and Language, University Dufour, D. L., and M. L. Sauther. 2002. Comparative and evolu- of Cape Town. Cambridge: Cambridge University Press. tionary dimensions of the energetics of human pregnancy and Blum, A., N. Bonneuil, and D. Blanchet, eds. 1992. Mode`les de la lactation. American Journal of Human Biology 14(5):584–602. de´mographie historique. Congre`s et colloques 11. Paris: Insitut Na- Dupaˆquier, J. 1998. Introduction. In Histoire des populations de tional d’E´ tudes De´mographiques, Presses Universitaires de France. l’Europe II: la re´volution de´mographique 1750–1914.J.P.Bardet Blurton Jones, N. 1986. Bushman birth spacing: a test for optimal and J. Dupaˆquier, eds. Pp. 7–17. Paris: Fayard. birth intervals. Ethology and Sociobiology 7:91–105. Ellison, P. T. 1991. Reproductive ecology and human fertility. In ———. 1987. Bushman birth spacing: direct tests of some simple Applications of biological anthropology to human affairs.C.G.N. predictions. Ethology and Sociobiology 8:183–203. Mascie-Taylor and G. W. Lasker, eds. Pp. 14–54. Cambridge: Cam- ———. 1989. The costs of children and the adaptive scheduling of bridge University Press. births: towards a sociobiological perspective on demography. In Ellison, P. T., C. Panter-Brick, S. F. Lipson, and M. R. O’Rourke. The sociobiology of sexual and reproductive strategies. A. R. Rasa, 1993. The ecological context of human ovarian function. Human C. Vogel, and E. Voland, eds. Pp. 265–283. London: Chapman & Reproduction 8(12):2248–2258. Hall. Flannery, K. 1969. Origins and ecological effects of early domesti- Bocquet-Appel The Agricultural Demographic Transition S509

cation in Iran and the Near East. In The domestication and ex- Livi-Bacci, M. 1992. A concise history of world population. Oxford: ploitation of plants and animals. P. J. Ucko and G. W. Dimbleby, Blackwell. eds. Pp. 73–100. Chicago: Aldine. Lunn, P. G., S. Austin, A. M. Prentice, and R. G. Whitehead. 1984. Fraccoaro, M. 1959. Fertility differential in two Lappish populations. The effect of improved nutrition on plasma prolactin concentra- American Journal of Human Genetics 11:92. tions and postpartum infertility in lactating Gambian women. Goring-Morris, A. Nigel, and Anna Belfer-Cohen. 2011. Neolithi- American Journal of Clinical Nutrition 39:227–235. zation processes in the Levant: the outer envelope. Current An- McAlpine, P. J., and N. E. Simpson. 1976. Fertility and other de- thropology 52(suppl. 4):S195–S208. mographic aspects of the Canadian Eskimo communities of Ig- Gray, A. 1985. Limits for demographic parameters of aboriginal pop- loolik and Hall Beach. Human Biology 48:113–138. ulations in the past. Australian Aboriginal Studies 1:22–27. McCaa, R. 2002. Paleodemography of the Americas: from ancient Gray, J. P. 1999. A corrected ethnographic atlas. World Cultures 10(1): times to colonialism and beyond. In The backbone of history.R. 24–136. H. Steckel and J. C. Rose, eds. Pp. 94–124. Cambridge: Cambridge Gremillion, K. J. 2006. Central place foraging and food production University Press. on the Cumberland Plateau, eastern Kentucky. In Behavioral ecol- Milan, F. A. 1970. The demography of selected circumpolar popu- ogy and the transition to agriculture. D. J. Kennett and B. Winter- lations. In The human biology of circumpolar populations.F.A. halder, eds. Pp. 41–62. Berkeley: University of California Press. Milan, ed. Pp. 13–36. International Biological Programme 21. Guerrero, E., S. Naji, and J. P. Bocquet-Appel. 2008. The signal of Cambridge: Cambridge University Press. the Neolithic demographic transition in the Levant. In The Neo- Morgan, K. 1973. Historical demography of a Navajo community. lithic demographic transition and its consequences.J.P.Bocquet- In Methods and theories of anthropological genetics. M. H. Crawford Appel and O. Bar-Yosef, eds. Pp. 57–80. Dordrecht: Springer. and P. L. Workman, eds. Pp. 263–314. Albuquerque: University of Hern, W. M. 1977. High fertility in a Peruvian Amazon Indian village. New Mexico Press. Human Ecology 5(4):355–368. Peterson, N., and J. Taylor. 1998. Demographic transition in a hunter- Hornabrook, R. W., S. Serjeantson, and J. M. Stanhope. 1977. The gatherer population: the Tiwi case, 1929–1996. Australian Aborig- relationship between socioeconomic status and health in two inal Studies 1:11–27. Papua New Guinean populations. Human Ecology 5(4):369–382. Pickersgill, B. 2007. Domestication of plants in the Americas: insights Hrdlicka, A. 1936. Fecundity of Eskimo women. American Journal from Mendelian and molecular genetics. Annals of Botany 100: of Physical Anthropology 22:91–95. 925–940. Huffman, S. L., A. Chowdhury, H. Allen, and L. Nahar. 1987. Suckling Piperno, Dolores R. 2011. The origins of plant cultivation and do- patterns and post-partum amenorrhoea in Bangladesh. Journal of mestication in the New World tropics: patterns, process, and new Biosocial Science 19:171–179. developments. Current Anthropology 52(suppl. 4):S453–S470. Hurtado, A. M., and K. R. Hill. 1990. Seasonality in a foraging society: Reher, D. S., and J. A. Ortega Osona. 2000. Malthus revisited: ex- variation in diet, work effort, fertility and sexual division of labor ploring medium-range interaction between economic and de- among the Hiwi of Venezuela. Journal of Anthropological Research mographic forces in historic Europe. In Population and economy: 46(3):293–346. from hunger to modern economic growth. T. Bengtsson and O. Saito, Johansson, S. R., and S. Horowitz. 1986. Estimating mortality in eds. Pp. 183–212. Oxford: Oxford University Press. skeletal populations: influence of the growth rate on the inter- Romaniuk, A. 1981. Increase in natural fertility during the early stages pretation of levels and trends during the transition to agriculture. of modernization: Canadian Indians case study. Demography 18(2): 157–172. American Journal of Physical Anthropology 71(2):233–250. Roth, E. A. 1981. Demography and computer simulations in historic Jones, F. L. 1965. The demography of the Australian Aborigines. village population reconstruction. Journal of Anthropological Re- International Social Science Journal 17(2):232–245. search 37(3):279–301. ———. 1972. Fertility patterns among Aboriginal Australians. Hu- Rowley, C. D. 1972. Outcasts in white Australia. Melbourne: Pelican. man Biology in Oceania 1(4):245–254. Rowley-Conwy, P. 1984. The laziness of the short-distance hunter: Jorgensen, J. G. 1980. Western Indians: comparative environments, the origins of agriculture in western Denmark. Journal of Anthro- languages, and cultures of 172 western American Indian tribes.San pological Archaeology 3:300–324. Francisco: W. H. Freeman. ———. 1998. Cemeteries, seasonality, and complexity in the Er- ———. 1999. Codebook for Western Indian data. World Cultures tebølle of southern Scandinavia. In Harvesting the sea, farming the 10(2):144–293. forest: the emergence of Neolithic societies in the Baltic region.M. Kirk, R. L. 1981. Aboriginal man adapting: the human biology of Zvelebil, M. Dennell, and L. Domanska, eds. Pp. 193–202. Shef- Australian Aborigines. Oxford: Clarendon. field: Sheffield Academic. Kohler, T. A., and M. Glaude. 2008. The nature and timing of the Sattenspiel, L., and H. Harpending. 1983. Stable populations and Neolithic demographic transition in the North American South- skeletal age. American Antiquity 48:489–498. west. In The Neolithic demographic transition and its consequences. Saucier, J. F. 1972. Correlates of the long post-partum taboo: a cross J. P. Bocquet-Appel and O. Bar-Yosef, eds. Pp. 81–106. Dordrecht: cultural study. Current Anthropology 13(2):238–258. Springer. Sibly, R. M., and J. Hone. 2003. Population growth rate and its Lee, R. D. 1972. Population growth and the beginning of sedentary determinants: an overview. In Wildlife population growth rates.R. life among the !Kung bushmen. In Population growth: anthropo- M. Sibly, J. Hone, and T. H. Clutton-Brock, eds. Cambridge: Cam- logical implications. B. Spooner, ed. Pp. 329–342. Cambridge, MA: bridge University Press. MIT Press. Simoni, L., F. Calafell, D. Pettener, J. Bertranpetit, and G. Barbujani. ———. 1979. The !Kung San: men, women, and work in a foraging 2000. Geographic patterns of mtDNA diversity in Europe. Amer- society. Cambridge: Cambridge University Press. ican Journal of Human Genetics 66(1):262–278. ———. 1987. Population dynamics of humans and other animals. Smith, D. S. 1972. The demographic history of colonial New England. Demography 24(4):443–465. Economic History Review 32:165–183. Lewis, M., and W. Clark. 1805. The journals of the Lewis and Smith, P., and L. K. Horwitz. 2007. Ancestors and inheritors: a bioan- Clark expedition. July 13. http://lewisandclarkjournals.unl.edu/ thropological perspective on the transition to agropastoralism in index.html. the southern Levant. In Ancient health: skeletal indicators of agri- S510 Current Anthropology Volume 52, Supplement 4, October 2011

cultural and economic intensification. M. N Cohen and G. M. M. Valeggia, C., and P. T. Ellison. 2004. Lactational amenorrhoea in well- Crane-Kramer, eds. Pp. 207–222. Gainesville: University Press of nourished Toba women of Formosa, Argentina. Journal of Biosocial Florida. Science 36:573–595. Stanhope, J. M., and R. W. Hornabrook. 1974. Fertility patterns of Van Arsdale, P. W. 1978. Population dynamics among Asmat hunter- two New Guinea populations: Karkar and Lufa. Journal of Biosocial gatherers of New Guinea: data, methods, comparisons. Human Science 6:439–452. Ecology 6:435–467. Stiner, M. 2004. Thirty years on the “broad spectrum revolution” Weiss, E., W. Wetterstrom, D. Nadel, and O. Bar-Yosef. 2004. The and Paleolithic demography. Proceedings of the National Academy broad spectrum revisited: evidence from plant remains. Proceedings of Sciences of the USA 98(13):6993–6996. of the National Academy of Sciences of the USA 101(26):9551–9555. Stiner, M. C., J. E. Beaver, N. D. Munro, and T. D. Surovell. 2008. Widner, R. J. 1988. The evolution of the Calusa: a non-agricultural Modeling Paleolithic predator-prey dynamics and the effects of chiefdom on the southwest Florida coast. Tuscaloosa: University of hunting pressure on prey “choice.” In Recent advances in palaeo- Alabama Press. demography: data, techniques, patterns. J. P. Bocquet-Appel, ed. Pp. Wilson, C., and P. Airey. 1999. How can a homeostatic perspective 143–178. Dordrecht: Springer. enhance demographic transition theory? Population Studies 53(2): Surovell, T. A. 2000. Early Paleo-Indian women, children, mobility, 117–128. and fertility. American Antiquity 65(3):493–508. Wilson, M. C., L. V. Hills, and B. Shapiro. 2008. Late Pleistocene Sussman, R. W. 1972. Addendum: child transport, family size, and northward-dispersing Bison antiquus from the Bighill Creek For- increase in human population during the Neolithic. Current An- mation, Gallelli Gravel Pit, Alberta, Canada, and the fate of Bison thropology 13(2):258–267. occidentalis. Canadian Journal of Earth Sciences 45(7):827–859. Temkin-Greener, H., and A. C. Swedlund. 1978. Fertility transition Wood, J. W. 1998. A theory of pre-industrial population dynamics: in the Connecticut Valley: 1740–1850. Population Studies 32(1): demography, economy, and well-being in Malthusian systems. 27–41. Current Anthropology 39:99–135. Current Anthropology Volume 52, Supplement 4, October 2011 S511

Erratum

In the paper “Holocene population history in the Pacific re- gion as a model for worldwide food producer dispersals” by Peter Bellwood, published in this supplement and electron- ically published on May 13, 2011 (http://www.jstor.org/stable/ 10.1086/658181), there is an error in the first paragraph of the section titled “The Spread of Neolithic Pottery from Tai- wan into the Philippines and Indonesia.” Near the end of the first paragraph of the section, in the final sentence, the phrase “sites such as An Son in” was mistakenly inserted. The sen- tence should read, “Reranum and Chaolaiqiao still have some residual cord marking (fig. 2R), and the close similarities in red-slipped pottery between these two sites raise the possibility that a direct migration from southeastern Taiwan to Itbayat could have occurred between 2200 and 2000 BC.” The pub- lisher regrets this error.

᭧ 2011 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved. 0011-3204/2011/52S4-0023$10.00. DOI: 10.1086/662554