Norse Greenland Settlement: Reflections on Climate Change

Home , Eastern Settlement, Faroe Islands, History of Greenland, Sermermiut, Tusk, Western Settlement

Norse Greenland Settlement: Reﬂ ections on Climate Change, Trade, and The Contrasting Fates of Human Settlements in the North Atlantic Islands

Andrew J. Dugmore, Christian Keller, and Thomas H. McGovern

Abstract. Changing economies and patterns of trade, rather than climatic deterioration, could have critically marginalized the Norse Greenland settlements and effectively sealed their fate. Counter-intuitively, the end of Norse Greenland might not be symptomatic of a failure to adapt to environmental change, but a consequence of successful wider economic developments of Norse communities across North Atlantic. Data from Greenland, the Faroe Islands, and medieval Iceland is used to explore the interplay of Norse society with climate, environment, settlement, and other circumstances. Long term increases in vulnerability caused by economic change and cumulative climate changes sparked a cascading collapse of integrated interdependent settlement systems, bringing the end of Norse Greenland.

Introduction parison to unrelated processes of social, political, and economic change (Diamond 2005). Further- At a time when the effects of global climate change more, even if climate change can be shown to have can be seen to be taking place, there is a pressing produced a direct impact (for instance crop fail- need to assess how these might affect human soci- ure or livestock mortality), perhaps the most im- ety.1 The extent to which climate changes exacer- portant question of all is why people d0 not (or bate environmental degradation, drive settlement cannot) adapt and either avoid or mitigate the bad collapse, cause famine, spur migrations, or trigger effects of the weather. As with so many other en- conﬂ ict over resources has received widespread vironmentally related issues that cut across dis- attention (e.g., Diamond 2005; Fagan 2000, 2004; ciplinary boundaries, assessments of relative em- Linden 2006). It is clear that climate change, and phasis, sensitivity, thresholds, adaptation, and the weather it produces, can have a wide range of response are vital. This paper reconsiders the case impacts, many negative, some positive, but all with of Norse Greenland and the end of the settlements the potential to affect human security. Crucially, in the early part of the Little Ice Age (Grove 2001; however, opinions differ as to the importance to Jones 1986). For this iconic example of settlement human societies of climate on its own and in com- desertion widely associated with climate change

Andrew J. Dugmore, Institute of Geography, School of GeoSciences, University of Edinburgh Drummond Street, Edinburgh EH8 9XP, Scotland, UK Christian Keller, University of Oslo, P.O. Box 1072 Blindern, Oslo, Norway 0316 Thomas H. McGovern, Hunter Bioarchaeology Laboratory, Department of Anthropology Hunter College, City University of New York, 695 Park Ave, NewYork City 10021

WW4336.indb4336.indb 1212 44/20/07/20/07 10:15:2910:15:29 AAMM Dugmore, Keller, and McGovern: Contrasting Fates of Human Settlements in the North Atlantic Islands 13

and an inability to adapt, the paper explores both tephrochronology, an environmental dating based the nature of climate change and unrelated eco- on layers of volcanic ash (e.g., Buckland et al. nomic factors that may have played crucial if not 1991; Dugmore, Larsen, and Newton 2004; Meeker dominant roles in determining the ultimate fate of and Mayewski 2002; Thórarinsson 1967, 1975, the Norse Greenlanders. 1981). Quite remarkable new perspectives have also been gained from faunal records based on hundreds of thousands of identifi ed bones from The Norse in The North Atlantic archaeological contexts that give unrivalled in- The Norse settlement of the Faroe Islands, Iceland, sights into the subsistence and economies of the and Greenland (Fig. 1) began between the early past (e.g., Church et al. 2005; McGovern 2000; Mc- ninth and late tenth centuries A.D. and, although Govern, Perdikaris, and Tinsley 2001; Perdikaris comparatively recent, it spans a suffi ciently long 1999; Perdikaris and McGovern in press a). period to include notably different climatic ep- The Norse colonies in Greenland were es- isodes and major cultural changes (Grove 1988; tablished shortly before the year 1000 A.D, the Jones 1986). When considered as a whole, the last documented outside contact was in 1409 A.D., North Atlantic islands permit the analysis of hu- and by the sixteenth century A.D. the Norse Green- man colonization, adaptation, and long-term set- landers had disappeared in circumstances that are tlement undertaken by similar groups of people unresolved, but are thought to have involved an (Viking Age settlers from northwestern Europe) in interplay between a range of factors including cli- contrasting island environments across signifi cant matic deterioration, environmental degradation, climatic gradients (Edwards et al. 2004). A pan- maladaptation, economic marginalization, and con- Atlantic view can consider key aspects of people’s fl ict with the Inuit (Arneborg 2002, 2003a, 2003b; experiences of, and responses to, different envi- Gulløv 2004:273–275, 277; McGovern 2000; Pe- ronmental and cultural challenges and, ultimately, tersen 2000; Seaver 1996). In addition, the “Great their very different experiences of success and fail- Plague” that devastated Iceland for the fi rst time ure (Barlow et al. 1997; Karlsson 2000; McGovern in 1402–1403 A.D. created a demographic defi cit and Perdikaris 2000). Importantly, historical con- in Iceland that could be fi lled from abroad (Karls- text and the culture of the Scandinavian settlers son 1996). Alternatively, this epidemic might have of North Atlantic islands are well known, due to a reached Greenland a decade later, with terminal unique host of medieval manuscripts (see Friðrik- consequences for the Norse. sson 1994; Karlsson 2000; Kristjánsson 1993; Óla- Coherent syntheses can be fashioned that sson 1998; Vasey 1996; Vésteinsson 2000). High combine all these and other explanations (such as quality regional environmental data are available raids by Barbary pirates). In the process of fi nd- from Greenland ice cores and accurate correlation ing consensus, however, it is possible that impor- of multiple data sources (historical, archaeologi- tant insights may be lost, including the distinction cal, and environmental data from landforms, soils, between factors that could have increased the fun- sediments, and peats) is possible to very high stan- damental vulnerability of the Norse settlements to dards through the use of both radiocarbon and extinction (and made their ultimate fate inevitable)

Figure 1. The North Atlantic showing places referred to in the text.

WW4336.indb4336.indb 1313 44/20/07/20/07 10:15:3010:15:30 AAMM 14 Arctic Anthropology 44:1

and the fi nal trigger for extinction (an event, or events that under other circumstances might not have had such decisive effects). It is generally accepted that initial settlement was boosted by climatic conditions that were, on the whole, more favorable to the Norse than those of the fourteenth century A.D. onwards (Barlow et al 1997; Ogilvie and McGovern 2000). With the local onset of cooler conditions, the increasing extent of pack ice could have seriously affected ship- ping and compromised hunts for marine mammals, in particular forays into the northern hunting grounds for walrus. Hand in hand with the development of more extensive pack ice, the shortened growing seasons, reduced fodder production, longer over-wintering periods, environmental degradation, and more demanding conditions for raising domestic animals could have made the subsistence base of the Norse more precarious than it had been at the time of settlement (Buckland et al. 1996). In combination with an inability to adapt, these all- encompassing climatically driven changes could be all that really mattered, as they alone could have made the Norse Greenlanders’ way of life so marginal that it was simply a question of when the colonies failed rather than if. This inevitable de- scent to destruction might have been accelerated Figure 2. A conceptual model of Norse Greenland by cultural factors (McGovern 1994). An inability settlement highlighting the key roles of a pastoral focus, or unwillingness to adapt, to adopt an Inuit life- the impact of climate change, and an inability to adapt. style, or to embrace other radical change, may have critically reduced chances for survival, and accelerated the fi nal collapse. Maladaptation could wrong with the Norse approach to subsistence in have taken a variety of forms. There is no evidence Greenland and it could have been adequate for sur- that the Norse Greenlanders adopted Inuit Arctic- vival if other circumstances had been different; adapted technology, such as the toggle harpoon specifi cally, if the economic viability of the settle- and the kayak. This failure to acquire Inuit tech- ments had not been eroded and the Norse popula- nologies and related subsistence strategies that tion had not declined (Lynnerup 2003). might have saved the Norse could be explained by intense cultural conservatism and the oppressive control of an elite who did not wish to see change Motivations for the Norse that might have weakened their own position. In Colonization of Greenland— addition, violent confl ict with the Inuit could have Land Hunger or Hunting? further compromised any possibility of cultural transmissions and the sharing of viable survival Indications of the likely priorities of the Norse in strategies (Petersen 2000). Other aspects of mal- Greenland, and the balance between trade and sub- adaptation could have included damaging or in- sistence may be gained from the circumstances of appropriate applications by the Norse of their the initial colonization effort. A written account of own pastoralist system. This could have included Eirik the Red and the naming of Greenland, can be poorly managed grazing strategies that enhanced found in Íslendingabók [the Book of the Iceland- erosion, inappropriate mixes of livestock, and ers), a short chronicle of Iceland’s early history, poor strategies for producing meat, dairy products, that was written between 1122 and 1133 A.D. by wool, and hides, as well as the unsustainable har- Ari Þorgilsson the Learned (Hreinsson 1997; Karl- vesting of wild resources (Fig. 2). sson 2000). In it, the now famous story is told that: The argument for maladaptation and the im- The land which is called Greenland was discov- pact of climatic deterioration rests upon the as- ered and settled from Iceland. Eirik the Red was sumption that change of lifestyle was of funda- the name of a Breidafjord man who went out there mental importance. An alternative view is that from here and took land in settlement at the place these could be the wrong issues to emphasize. which has ever since been called Eiriksfjord. He There might have been nothing fundamentally gave the land a name, and called it Greenland,

WW4336.indb4336.indb 1414 44/20/07/20/07 10:15:3010:15:30 AAMM Dugmore, Keller, and McGovern: Contrasting Fates of Human Settlements in the North Atlantic Islands 15

arguing that men would go there if the land had a eleventh centuries A.D. Recent research in Ice- good name. . . . (Karlsson 2000:29) land demonstrates a clear need to radically revise earlier models that saw the initial settlement pro- Apparently on the strength of this “sales pitch” cess as a slow trickle of individual chieftains and Eirik was able to mount a very large and successful their households gradually moving inland from colonization effort, and took over some of the best coastal enclaves (established just after the well agricultural land at Brattahlið in what was to be- dated Landnám tephra of A.D. 871+2). Tephra and come the Eastern Settlement. Eirik and the other radiocarbon dates indicate that deep inland areas early chieftains are said to have divided the wilder- in north Iceland were also settled directly after the ness, providing names to fjords, mountains, and Landnám tephra fall, with indications of early per- other natural features (often naming them after manent farms as much as 100 km from the coast themselves) and thus bringing the new cultural (McGovern et al. in press; Vésteinsson, McGovern, landscape into being. The written sources for the and Keller 2002). These inland settlements at the settlement of Iceland which also mention Green- then grassy highland edge of the dense birch woodland (l´slendingabók, Landnámabók [the Book of lands (around 300 – 400 meters above sea level) Settlement], Eiríks saga rauða [the Saga of Eirik the are exactly contemporary with the better docu- Red], Grænlendingasaga [the Saga of the Green- mented coastal settlements. The inland highland landers]) stress the role of such land hungry chief- settlements in the Mývatn district of North Ice- tains seeking pastures for their imported domestic land (Fig. 1) appear to have “leapfrogged” over the animals and access to a range of wild resources to heavily wooded lowland valley systems, whose support clients and reward new followers (Hrein- ungrazed and pathless birch and willow thickets sson 1997). would have initially been unattractive farming These sources were written centuries after country. When the northern valley systems were the events described, and are clearly infl uenced by deforested and cleared for settlement in the later contemporary twelfth and thirteenth century A.D. tenth to early eleventh centuries A.D. they were to elite politics and the need to retroactively infl ate provide some of the richest farm sites in Iceland, (or create) the role of notable ancestors. As Adolf thus opening attractive settlement potentials just Friðriksson and Orri Vesteinsson (2003) demon- as Greenland was being settled. strate Landnámabók’s apparently authoritative fjord The initial, probably fl uid, phase of settle- by fjord listing of fi rst settlers in much of Iceland ment in Iceland was over by the mid tenth century. on closer analysis looks like retrospective guess- By that time the Icelandic national assembly, the work based on etymological interpretations of Althing, had been established 930 A.D. and laws place names occasionally reinforced by elements had been created to protect land holdings and of local tradition and much rearranged to suit later privileges (Karlsson 2000) The land was divided political ends. If the Icelandic literary record is a and owned, but the broad holdings of the major strongly suspect source for the history of early set- families were very large and probably were not ac- tlement in Iceland, one may suspect that it may be tually fully settled until much later (Vesteinsson, an even more ill informed, selective, and heavily Mcgovern, and Keller 2002). There are few envi- edited account of the early days of the Greenland ronmental signs of signifi cant pressures on the settlement. land having developed in Iceland by the late tenth While scholars have long debated the histori- century and the Norse colonization of Greenland cal value of these later medieval accounts of ninth (Dugmore and Buckland 1991; Dugmore and Ers- to early eleventh century settlement, the Norse kine 1994; Dugmore et al. 2000). In the most eco- colonization is still usually read in the context of logically favorable areas, the dense indigenous land-hunger, driving an expanding agricultural woodland would have been opened up with in- population to go to the very margins of the biogeo- roads and clearings, so greatly increasing its util- graphic viability of their inherited North Western ity. Even as newly cleared mid-valley farms were European farming economy (e.g., McGovern et al. occupied and the initially broad magnate hold- 1988). While the close association of later perma- ings became more fully settled, additional settle- nent farms in Greenland with patches of pasture ment in land subdivisions must have still been vegetation (noted by all fi eldworkers since the possible in Iceland, as this process of subdivision eighteenth century) does indicate the key role of and planned settlement extended into later medi- farming based on domestic stock production in eval times. Major landowners could have had pow- later time periods, we now have some reasons to erful reasons to encourage additional settlement doubt the conventional Viking Age settlement on their own estates as a way to increase their own model, ultimately based upon the written accounts power, wealth, and infl uence. In Iceland, settle- of the high Middle Ages. ment infrastructure including established farms Simple subsistence-based “land hunger” is with hayfi elds, breeding stocks of domesticated unlikely for Icelanders in the late tenth to early animals, and regional exchange mechanisms had

WW4336.indb4336.indb 1515 44/20/07/20/07 10:15:3010:15:30 AAMM 16 Arctic Anthropology 44:1

been created, and were well established by the mid ration by hunters and scouts must have pre-dated tenth century A.D. (McGovern et al. in press). The the full scale (and high risk) recorded trans-Atlantic concentration of power and privilege in the hands Landnám settlements by whole households and of a small elite had yet to happen and the disas- their domestic animals. Giving a new country an trous civil confl icts of the thirteenth century A.D. attractive name is not the same thing as exploring Sturlungar Period were still way into the future it, and it is hard to imagine any of the articulate, (Karlsson 2000:79– 82). In places such as south authoritative, and hard-headed women featured in Iceland, abandonment of settlement sites in the the sagas as household managers and infl uential upland margin had not yet begun, and lowland leaders in their own right consenting to uproot kin woodlands devoted to charcoal production were and servants, farming gear, family wealth, and pre- yet to be “worked out” (Dugmore et al. 2006). In cious domestic stock solely on the word of a (so short there seems to have been little in the way far) failed upstart like Eirik. of a mass “land hunger” and few environmental Did groups of hunters precede the Norse “push factors” encouraging people to leave Ice- farmers, providing detailed knowledge of land and land in search of living space in Greenland. resources and a vital initial economic base, thus Any “push factors” existing in late tenth cen- signifi cantly contributing to the success of the later tury A.D. Iceland were probably social rather than Landnám before being conveniently forgotten by environmental. By 930 A.D. the goðorð or chief- later scribes? Before the farmers came, were there tainships, were established, courts set up, land brief but vital initial phases of exploration, exploi- claimed (if not settled), and the rules of competi- tation, and occupation undertaken by Norse hunt- tion established which would last until the civil ers in Iceland in the ninth century A.D. and in wars of the mid thirteenth century A.D., which Greenland in the tenth century A.D.? brought on the twilight years of the Icelandic Com- In southwest Iceland, the discovery of walrus monwealth. Relatively late colonizers in Iceland bones (some newborn, suggesting a resident breed- who were not part of the older (now landed) chief- ing walrus population) under the streets of modern tainships, and who probably did not fi t into the Reykjavik, followed by fi nds of three nearly com- evolving set of regional góði to bóndi [chieftain to plete, expertly extracted walrus tusks, datable to independent farmer] relationships, were in a sense just after the deposition of the Landnám tephra of thrown back into a pre-Viking Age situation of be- 871+2 A.D., and left beneath the bench of a long ing excluded from the obvious sources of prestige hall less than 500m away, confi rm the place name and wealth, land and infl uence over clients. evidence suggesting multiple walrus hunting areas By the mid tenth century A.D., Iceland was thus along the nearby Reykjanes peninsula (McGovern, not so much over-populated as “over-chieftained.” Perdikaris, and Tinsley 2001). Walrus colonies in Sources of wealth that could gain prestige and Iceland would have provided an immediate re- honor for ambitious individuals and lineages were turn in low-bulk, high- value Viking age prestige largely earmarked if not fully developed and con- goods for hunters, while large colonies of nesting testing these claims could end in defeat and out- sea birds (recalled as initially “unwary and easily lawry (as suggested in the very late and somewhat killed” by later texts) would have provided provi- unreliable saga of Eirik the Red ). In the late tenth sions. Did mid to late tenth century hunters (prob- century A.D., Greenland was therefore attractive to ably well below chiefl y status) provide guides or Icelanders not because of chronic land or subsis- at least reliable travel reports for the well recorded tence shortage in Iceland, but because of a short- chiefl y settlers? In any case, the Icelandic walrus age of ways to gain prestige and honor in Iceland. populations of the south coast do not seem to have Did the decision to settle Greenland spring en- survived after the tenth century A.D. tirely from Eirik’s failed aggrandizement in Ice- The Norse discovery of Greenland’s massive land, or was there background in an earlier “un- populations of walrus and other marine mammals, told saga” relating to hunting and Arctic prestige plus fur bearing animals and rare items such as goods rather than to the land-hungry chieftain falcons and narwhal tusks would have thus pre- farmers so extensively featured in the later written sented attractive economic opportunities to any sources now familiar to us? Icelandic hunters by the mid tenth century A.D. The Íslendingabók records that early Norse visi- tors to southern Greenland found still-recognizable Pioneering Hunters? ruins and artifacts of “The Skrælings of the same Was there an “unoffi cial,” undocumented, and type as those who inhabited Vínland” which must perhaps socially unacceptable earlier occupation refer to the Dorset Paleoeskimo, but not to direct of Iceland and Greenland less tied to agricultural contact, apparently supporting the archaeologi- land and the demands of chieftainship? In several cal evidence indicating Dorset withdrawal to the recent workshops, Adolf Friðriksson has hypoth- north in this period (Gulløv 2004) and suggesting esized that some period of pre-agricultural explo- an empty adaptive niche ca. 950 –1000 A.D. While

WW4336.indb4336.indb 1616 44/20/07/20/07 10:15:3010:15:30 AAMM Dugmore, Keller, and McGovern: Contrasting Fates of Human Settlements in the North Atlantic Islands 17

much research is currently ongoing in Greenland tinued nevertheless throughout the history of Norse and Arctic Canada, it is probably still fair to say Greenland. Fragments of the dense maxillary bone that we do not yet fully understand the dynamics surrounding the deep-rooted walrus tusks are con- of the centuries-long Norse-Dorset contact period centrated on high status sites in both Western and of ca. 1000 –1200 A.D., but evidence is growing Eastern Settlements (McGovern et al. 1996; Church that this contact was probably of very different et al. in press), but some maxillary chips are found character from the Thule-Norse interactions after in virtually every archaeofauna collection known 1200 A.D. in Greenland (Gulløv 1997, 1999, 2000; from Norse Greenland, including those from small Schledermann and McCullough 2003; Sutherland far inland sites (McGovern 1994; McGovern et al. 2000). As in Iceland, there is at present no direct 1996). Final extraction of the ivory and its pre- archaeological evidence for a Norse hunter’s set- paration for export was thus a dispersed winter tlement predating the arrival of Eirik’s farmers, but “cottage industry” on most farms and a major there may be some hints of their presence surviv- household chore on the larger magnate farms. Con- ing in Eirik the Red’s own saga. centrations of walrus tusk processing debris ac- Walrus are found throughout Greenlandic tually increase through time in current stratifi ed waters, but have historically concentrated in the bone collections and there is certainly no evidence Disko Bay area half way up the west coast where that the intensity of the hunt slackened in the later rich feeding grounds and multiple hauling-out lo- medieval period. Recent archaeological research cations [uglit] still produce some of the largest con- by the SILA Greenland Research Centre of the Na- centrations of walrus in the eastern Arctic. The tional Museum of Denmark at the newly discov- distance from Iceland to the west coast of Green- ered trading center at Sandhavn near Kap Farvel land, and the prime hunting grounds of the Disko in southern Greenland serves to underline the im- region, combined with the need to trap fur bear- portance of trade and export in Norse Greenland ing animals in the wintertime required permanent (Gulløv 2004) and many authors have suggested bases in Greenland, and these were rapidly es- that the huge storage buildings on the bishop’s tablished (Figure 1). Both written sources and ar- manor at Gardar played a role as warehouses for chaeology (Gad 1970; McGovern 1985a) confi rm export (Gulløv 2004:242–243; McGovern 1985b; that Disko Bay was the area of the medieval Norse Nørlund 1930). Norðrsetur [northern hunting grounds]. Walrus penis bones (bacula) are also regu- These attractions, plus the opportunities to larly recovered from farm sites far from the Norðr- pioneer a new wave of settlement and become lead- setur, and rows of complete walrus and narwhal ers in a new land most probably provided the pow- skulls were interred inside the consecrated church- erful motivation necessary for launching a major yard dike at the cathedral at Gardar (Krogh 1983: and hazardous colonization effort. In tenth century 163; McGovern 1985b). While walrus ivory itself Iceland the initial fl eet assembled by Eirik the Red is extremely rare in Norse Greenlandic artifact (apparently 25 ocean-going ships), and subsequent collections, small objects made from the peg-like fl eets of ships and colonists must have represented post-canines were commonly used for buttons a very signifi cant capital investment. The timing and craft work, including small walrus, polar bear, was probably critical. Iceland had been colonized and sea bird fi gurines pierced for suspension (Fitz- suffi ciently early in the Viking Age, that its own hugh and Ward, eds 2000). Walrus hunting and the potential for providing prestige trade items was ef- Norðrsetur voyages were thus socially and perhaps fectively exhausted, yet Icelandic settlement was even ritually embedded in the society of Norse suffi ciently well established to mount a substan- Greenland, and there is every indication that the tial colonization effort and, above all, the European Norse Greenlanders highly valued the hunt for rea- market for ivory and fur was buoyant and favorable sons sometimes transcending coldly calculated in the late tenth and early eleventh centuries. economics. Arctic hunting certainly did not decline with With this hunting orientated perspective in the establishment of farming communities in the mind, subsistence in Greenland could be seen as inner fjords of southwest Greenland. There are sev- primarily the means to underpin an export econ- eral lost poems describing hunting trips to the omy, and not simply an end in its self. Previous Norðrsetur, and surviving accounts record the num- paleoeconomic reconstructions viewing the Norðr- ber of weeks of travel required to reach the distant setur hunt as a curious adjunct to the main busi- hunting grounds from the Western Settlement ness of household subsistence (McGovern 1985) (modern Nuuk district) and the larger Eastern Set- may have actually had the economic emphasis pre- tlement (modern Qaqortoq and Narsaq districts) cisely backwards. If we view the Norðrsetur hunt further south. These long summer voyages de- as the original and continuing core reason for the prived the farming communities to the south of existence of Norse Greenland, then the subsequent vital adult labor and scarce boats for one of the development of this export economy becomes a busiest parts of the agricultural year, but they con- key issue.

WW4336.indb4336.indb 1717 44/20/07/20/07 10:15:3110:15:31 AAMM 18 Arctic Anthropology 44:1

Barrett, Locker, and Roberts 2004a, 2004b, Barrett, Norse Greenlanders and Nicholson, and Cerron-Carrasco 1997, Barrett et al. Wider Developments in the 2003, Bond 1998). By the early 1100s A.D. the Nor- wegian crown and church had begun to fully orga- North Atlantic Region nize the stock fi sh trade based on the rich winter Changing Trade fi sheries of Lofoten and Vesterålen An earlier intra- Norse trade in dried marine fi sh predated the FEH In Scandinavia, the Norse traded furs from the in both Iceland and Faroes by at least a century Sami in the north and the Finns in the east, both (Brewington 2007; McGovern et al. in press; Perdi- to sell on the growing European market and to ex- karis and McGovern in press a and b). Early coastal port to the Middle East by way of the great Rus- settlements in North West Iceland may well have sian rivers. The famous late ninth century A.D. controlled local pre-FEH fi sh production and by account made by the North Norwegian chieftain 1250 A.D. the region was a major center in the pro- Ohthere to King Alfred is still a key source con- duction of dried fi sh for export and local consump- cerning the acquisition and marketing of such Arc- tion (Amundsen et al 2005; Edvardsson et al. 2004; tic prestige goods by Nordic traders in the early Edvardsson and McGovern 2005). After the FEH Viking age. The Icelandic and Greenlandic wal- what was to prove a major and long-lasting engine rus ivory exports thus fall neatly into the pattern for economic development increasingly connected of early Viking low-bulk, high-value trade in pres- Iceland and Norse Scotland to the rapidly expand- tige goods. After the 1240s A.D., Hansa merchants ing urban markets of Europe. in collaboration with Novgorod and other Russian By the thirteenth century A.D. this trade in city states developed the fur trade from the Bal- low-value, large-bulk consumption goods was of- tic region northwards into the White Sea; elephant fering an effective alternative to the prestige trade ivory from Africa began to provide unbeatable com- in ivory and fur. This development was symptom- petition to walrus ivory in European markets and, atic of the “commercial revolution” in Europe dur- perhaps most importantly, religious art increasing the thirteenth century (Spufford 2006:12). In ingly moved away from the use of ivory (Roesdahl other words, alternative ways to generate trade 2005). Other developments could have eroded the goods and wealth were becoming available in Ice- trade position of the Greenlanders: the develop- land and the draw may have been responsible for ment of hemp ropes, for example, may have effec- a critical hemorrhage of people from Greenland tively replaced a market for cables made from wal- (Lynnerup 2003). rus hide. Add increasing operational diffi culties in Greenland (caused by colder, stormier weather and more sea ice) to a fundamental erosion of the Norse Thule Contact with Dorset and Norse Greenlanders’ economic position, then their situa- While nearly 300 years of Dorset-Norse contact in tion could have become dire. In addition, the Black Greenland and the eastern arctic remains incom- Death of 1347–1351 A.D. and subsequent plagues pletely understood, it seems to have had fairly heavily depleted the population in Europe (Golt- limited impacts on either population. Certainly fried 1983), and in Norway the loss of 30 to 50 per- neither the Norse nor the Dorset seem to have sig- cent of the population led to an economic collapse; nifi cantly expanded their settlement areas after the value of silver dropped to roughly one-third of the period of the Norse Landnám and the probable its previous value, relative to the price of land and times of initial Norse-Dorset contact around 1000 – food. Hence the value of ivory relative to stockfi sh 1100 A.D. This was not the case when the Thule must also have been subject to a signifi cant drop. people expanded eastwards into Arctic Canada and The development and expansion of the trade Greenland, as they replaced both Dorset and Norse in dried Atlantic cod around 1100 A.D. was to have —culturally if not biologically—throughout the re- widespread economic impacts throughout Europe, gion before 1500 A.D. The timing and motivation which probably did not work to the advantage of for the Thule expansion is still an area of intense the Norse Greenlanders. Thanks to a comprehen- and productive research across the Arctic, but a sive zooarchaeological synthesis carried out by growing consensus sees the Thule migration reach- James Barrett and colleagues (Barrett, Locker, and ing northern Greenland around 1200 A.D.(Gulløv Roberts 2004a; 2004b) we now know that prior to a 2004; Schledermann and McCullough 2003). Rob- “Fish Event Horizon” (FEH) around 950 –1050 A.D. ert McGhee (1984, 2003) has updated his earlier virtually no marine fi sh bones appear on any in- stress upon warm climate and intensive bowhead land site in Britain, but after the horizon they be- whale hunting, arguing that a major cause for the come common, as suggested by the historical re- Thule people’s long journey eastwards was in fact cord. In the Norse-occupied Northern and Western an effort to track down the source of the metal and Isles, “fi sh middens” densely packed with fi sh bone other desirable trade goods that had been trickling showing evidence of preparation for export fi rst be- westward over the previous centuries. Extensive re- come common ca 1050 –1100 A.D. (Barrett 2000, cent archaeological and ethnographic research in

WW4336.indb4336.indb 1818 44/20/07/20/07 10:15:3110:15:31 AAMM Dugmore, Keller, and McGovern: Contrasting Fates of Human Settlements in the North Atlantic Islands 19

the western Arctic has indicated cyclical patterns probably communal boat drives aimed at the mil- in episodes of intensifi ed whaling associated with lions of migrating harp (Phoca groenlandicus) intensifi ed warfare (e.g., Mason 1998). Thus, the and hooded seals (Cystophora cristata) arriving Thule reaching Greenland in the thirteenth century along the coast of West Greenland in spring. These A.D. may well have been far more experienced and hugely abundant migratory species were rare or skilled in low-intensity warfare than once believed, absent in the rest of the Viking Age North Atlantic, and so might have had combat skills superior to and they presented a far richer resource, and one those of the Norse Greenlanders, who by 1200 A.D. much less likely to be depleted by large scale ex- were neither Viking warriors nor medieval men ploitation, than the small resident harbor (Phoca at arms, but peasant farmers. On the west coast of vitulina) and grey seal (Halichoerus gryphus) colo- Greenland, Thule winter settlements spread rap- nies of Iceland or the British Isles. Norse archaeo- idly southwards. Mathiassen’s excavations at the fauna in Greenland from the earliest phases have unusually large and nucleated Thule settlement at far more seal bones than appear in any of the other Sermermiut in the Disko Bay region (Mathiassen North Atlantic bone collections (Perdikaris and 1958) produced extensive collections of Norse arti- McGovern in press c), and there is a general ten- facts (church bell fragments, woolen cloth, substan- dency for the relative percentages of seal bones to tial amounts of iron, and draughtsmen converted increase through time—topping 70% of the total into tops—all types also recovered from later exca- in some smaller sites (McGovern et al 1996, Mc- vations in Inuit contexts in Greenland). As Sermer- Govern 2000). Isotopic evidence from human re- muit lies directly across the Norse sailing route to mains clearly indicates an increasing marine com- the Norðrsetur its many residents could have posed ponent in Norse diet in Greenland, suggesting that a serious threat to Norse hunting parties if hostili- the seal hunt and other marine foraging played a ties broke out. In these circumstances, the smaller progressively more vital role in subsistence (Ar- Inuit winter settlements which reached the outer neborg, et al. 1999). Norse migratory seal hunt- fjords of Godthaabsfjord by ca. 1300 (Gulløv 1997, ing in Greenland was thus both highly productive 1999) occupied the seasonal sealing grounds of the and capable of expansion and intensifi cation, and Norse Western Settlement, could also have pre- the Norse communal seal hunters were not only sented an additional potential threat to Norse sea- able to provision their communities without Inuit sonal subsistence sealing. technology, but also capable of continuing to ex- The Norse did not adopt Inuit ice hunting pand seal catches as long as labor and boats were techniques and thus did not take many ringed seal available. The Norse seal hunters were thus no (Phoca hispida) (McGovern 2000), but perhaps op- less competent than the Inuit, but they made far timizing subsistence was not the key issue through greater use of pooled communal labor and care- the long decades of apparent success until the very fully scheduled targeting of seasonal concentra- last years of the settlement when a terminal crisis tions of migratory seals rather than more individ- may have developed too quickly to permit any adap- ualistic year-round sealing with harpoons from tive change. The economic perspective raises the kayak or ice edge. Communal seal harvesting, possibility that an Inuit lifestyle might not have which probably involved most active adults in been chosen by the Norse in the thirteenth and the community and distributed seal meat to every fourteenth centuries because it was not needed, farm (even far inland), must have had a signifi cant and the main contact issue may have been the po- social dimension, hedged about with tradition tential disruption of the Norðrsetur connection and representing an annual expression of commu- and potential loss of Norse hunters and scarce nal solidarity and cooperation (McGovern 1985a, boats during periods of confl ict. As a result the 1991). key problems facing the Norse Greenlandic settle- The Faroe Islands, where hunting of the pi- ments could have been those of limited popula- lot whale (Globicephala melus) has occurred for tion, meeting labor requirements, and economic at least 500 years and probably since the Viking viability; even quite limited out migration driven Age and has provided a signifi cant part of the sta- by changing trade or sporadic casualties in low- ple diet for centuries, may provide a social as well intensity warfare could have had very serious con- as economic model for communal seal hunting in sequences for long term viability. Greenland (Gibson-Lonsdale 1990; Guttesen 1996). In times of subsistence crisis, whales have provided a key resource to the Faroese, a point vividly Adaptation, Population, and the demonstrated during the Second World War when Possibility of a Rapid Final Crisis in 1941 A.D. a record catch of 4,325 pilot whales provided a vital food supply (Guttesen 1996:90). While the Norse middens do not contain harpoons The immediate and total mobilization of all active or the ringed seals usually taken with such gear, community members for a pilot whale drive re- the Norse Greenlanders were certainly highly com- mains a key element in modern Faroese society, en- petent seal hunters who made use of nets and hanced rather than diminished by the availability

WW4336.indb4336.indb 1919 44/20/07/20/07 10:15:3110:15:31 AAMM 20 Arctic Anthropology 44:1

of cell phones, email, and motorized transport, and collaboration evident in the Norðrsetur hunt and totally undeterred by tourist sensibilities. The and home processing of walrus tusks well have modern Faroese pilot whale drive is now primar- also underwritten many components of the subsis- ily about community solidarity and visible expres- tence economy. Ironically the capacity of a well- sion core social values through collective action; integrated system to cope with variation may have but still, no dead whale goes uneaten. meant that when a crisis suffi cient to overwhelm The viability of Norse subsistence may thus the community did develop, it could have resulted be related to the degree of interdependence be- in a catastrophic collapse. tween farms and scale of the organization of labor The subsistence issue is closely linked to resources required to maintain effective commu- that of population, and population levels in Norse nal cooperation. A well-integrated network widen- Greenland might have been signifi cantly lower ing the geographical spread of producers and the than previously assumed (Lynnerup 2000, 2003). range of food sources exploited could ensure effec- Lynnerup’s “low population” model, reaching a tive buffering against shortage, but would require total of between 2,000 to 3,000 about 1300 A.D. collaboration between households and probably is controversial, and signifi cantly lower than the between districts. The presence of seal and seabird peak population range of 4,000 to 6,000 estimated bones at inland sites and caribou bones in coastal by others (e.g., Berglund 1986; Keller 1986; Mc- locations outwith their probable range all indicate Govern 1979; Meldgaard 1965). Higher or lower economic integration well above the household population estimates and their changes through level, suggesting pooling of labor at least season- time have a number of key implications. On the ally (McGovern 1992, 1985b). Labor may have been one hand, if communal hunting activities make a in high demand for local farming activities as well. major contribution to subsistence and the wild re- In Greenland, Norse fl ocks were nearly evenly split sources they tap into are effectively “unlimited,” between sheep and goats throughout the settlement, then higher population levels could be strongly and sheep were probably being managed largely benefi cial and low or falling populations a prob- for milk and not for signifi cant surplus wool pro- lem. On the other hand, if pastoralism was criti- duction (Mainland and Halstead 2005). In con- cal, productive land limited, and land degradation trast, Icelandic archaeofauna show a shift in fl ocks through climate change and soil erosion signifi - to nearly all sheep by c. 1200 A.D., many of which cant, then lower or falling populations may have seem to have been managed primarily for wool relieved or avoided a resource driven constraint. production (McGovern et al in press). Dairy herds In this context, the intensive exploitation of small require much more constant human contact and landholdings around isolated farms can be read labor investment throughout the summer than in quite different ways if this was part of a rotat- fl ocks of infertile ewes and castrated rams or weth- ing or shifting exploitation of the landscape (lower ers, which were traditionally left to fend for them- or more concentrated populations?), as opposed selves in the Icelandic highlands in summer. The to a strategy demanded by a fully occupied land- extensively developed highland saeter and “dairy scape with no scope for movement (higher or more farms” documented in the Eastern Settlement in dispersed populations?). In other words intensive Greenland (Albrethsen and Keller 1986) proba- exploitation of resources at any one site—in the bly refl ect this more labor-intensive herding pat- absence of regional perspectives based on multi- tern aimed at maximizing provisions rather than site/whole-landscape approaches—may be inter- wool production. In Greenland as in the rest of preted in quite different ways as either indications the North Atlantic, collection of winter fodder for of “desperation” or “highly effi cient maximum ex- stalled domestic animals was a labor-intensive ploitation (perhaps followed by recovery?).” process critical to stock survival (Amorosi et al. Returning to the question of climate change, 1998). Deeper snow and longer winters in Green- we now have two quite different ideas to con- land would increase the winter fodder demands sider. On the one hand, climate change could relative to Iceland, and the rugged terrain and scat- have been the principle challenge that drove the tered patches of upland pasture in Greenland to- Norse Greenlanders to extinction. From this point gether with the rarity of extensive areas of highly of view maladaptation and a subsistence crisis productive wet meadow that were only partially were key problems. The area simply became too compensated by extensive irrigation channels (Ar- marginal for them to endure with their chosen neborg 2005; Keller 1983; Krogh 1974, 1982:103, lifestyle. 144) would tend to have reduced the labor effi - Alternatively, climate change might have been ciency of harvesters. much less signifi cant, infl uencing the timing, but If the Norse settlements did function in large not necessarily the ultimate fate of the settlements. integrated units with signifi cant interdependence, Basic subsistence might have always been possi- rather than a group of effectively independent semi- ble, but economic marginalization, and the pres- self suffi cient units, the same community solidarity ence of alternative options back in Iceland (or in-

WW4336.indb4336.indb 2020 44/20/07/20/07 10:15:3210:15:32 AAMM Dugmore, Keller, and McGovern: Contrasting Fates of Human Settlements in the North Atlantic Islands 21

deed elsewhere in Scandinavia –especially in post ability of the remaining population to continue an plague-ravaged Norway) may have fatally under- integrated approach to resource exploitation and mined the rationale and means for occupying exploitation of all the hitherto available resources Greenland, and drawn people away from the is- due to shortages of labor (e.g., to continue sealing, land. For these reasons population contraction hunting caribou, managing domestic livestock, col- might have reached the point where subsistence lecting fodder, and so on). A cascading reduction and survival did become impossible. At a crit- of available resources could then outstrip popula- ical level of sensitivity perhaps any number of tion decline. Collapse may then result because of events from extreme weather to disease or confl ict critical gaps between labor availability and key could have provided the fi nal push to collapse and tasks, mismatches between supplies and demand, extinction. the losses of knowledge and skills, and perhaps If the Norse did have a well-integrated and most importantly of all a collapse of the social communal approach to living in Greenland, this structures that provided stability and enabled the may have contributed to a situation where a sud- community to tackle collective tasks. This could den crisis developed out of control and resulted have rapidly reinforced the initial population in social collapse and the rapid extinction of the shock, forcing a continued decline below thresh- remaining population. The idea can be explored olds of viability, and initiated total collapse. in a simple conceptual model of population and Taking this trade-orientated viewpoint, Norse resources (Figure 3). The resources available to Greenland would shift from being a case of mal- support the population are enhanced from their adaptation and an example of the impact of cli- initial levels by a combination of introduced do- mate change, to being one of a series of examples mesticated livestock, infrastructure (e.g., the cre- of episodic European exploitation of the Polar re- ation of home fi elds, irrigation systems, and con- gions where success or failure was primarily destruction of shielings), interdependence, and the termined by economic circumstances. The ruins harnessing of materials from a wide area. A sud- of Norse Greenland could therefore exist for fun- den population decline could interfere with the damentally similar reasons to the traces left by

Figure 3. A “collapse model” for Greenland. See text for details.

WW4336.indb4336.indb 2121 44/20/07/20/07 10:15:3210:15:32 AAMM 22 Arctic Anthropology 44:1

whalers, sealers, and miners elsewhere in the Arc- ninth century, and developed even at times when tic and Antarctic (Sugden 1982), and it could be as climate change is not thought to be an issue (Dug- misleading to read the record of Norse Greenland more et al. 2000, in press; Ogilvie and McGovern primarily in relation to periods of warmth or cold, 2000; Thórarinsson 1961). As a result we need to calm or storm as it would be to do so with these be able to disentangle the roles of climate change other short-lived excursions into polar regions. and nonclimatic factors in landscape degradation. One point of view, to put it crudely, is simply that people are very bad for the environment and hu- Is Climate Change Anything More man settlement and degradation always and inevitably go hand in hand. This observation is impor- Than The “Executor Of The Unfi t”? tant for our interpretation of settlement decline It could be a mistake to rule out climate change as in Greenland because, if true, it could strengthen a cause of real and substantial negative impacts the case for maladaptation and its potential role in on otherwise viable settlement in the Atlantic Is- promotion of crises. This “ultra Malthusian” view, lands. Iceland in particular provides an important that environmental degradation by the Norse in comparison to Greenland, because Icelandic settle- Iceland was simply and inevitably a result of hu- ment has also experienced major episodes of expan- man over exploitation cannot, however, offer a sion and contraction and these have at least in part complete explanation, because we have evidence been associated with major climatic challenges for environmentally aware and responsive prac- and widespread environmental impacts, although, tices that have managed impact, avoiding total once again, the extent to which these impacts were catastrophe. solely the result of climate change is debated (Karl- sson 2000). Even in medieval times Icelandic settlements supported many more people than Green- Environmental Interaction of land (at least 10 and possibly more than 30 times as many people), and unlike Greenland, Iceland the Norse in the Tenth through could not be argued to exist just as some extreme the Fifteenth Centuries A.D. form of olden-day trading base camp. It was a place that people sought out, and where people Woodland Decline in Iceland—An stayed, because it was potentially a good place to Ecological Catastrophe or a Managed live. The idea of climate change having signifi cant Landscape Fit for Purpose? cultural impacts is encouraged by notable anec- In Iceland today, woodlands are but small rem- dotes of farm sites disappearing beneath advanc- nants of the pre-historic forests and this can be ing glaciers (Grove 1988). Environmental degrada- viewed as a result of inappropriate land manage- tion has occurred on a large scale with extensive ment (Eysteinsson and Blöndal 2003). Indeed changes to vegetation cover (the reduction of wood- there are many historical examples of confl ict as a lands to a fraction of their former extent) and wide- result of disputes over this declining resource, as spread soil erosion (the loss of perhaps 20,000 km 2 through much of Icelandic history native wood- of soil) (Arnalds et al. 1997; Dugmore et al. 2005; land was the key source of charcoal, vital for the Thórarinsson 1961). Climate change has been ar- effective maintenance of tools and farm operations. gued to play a key role in this change (and so ex- Perhaps, however, a key point when considering ert unwelcome pressures upon Icelandic society) the fate of Icelandic woodlands is that they have because of a clear relationship between tempera- actually survived into the modern era (if in a much ture and vegetation cover. In Iceland the boundar- reduced form) —a better record than that of Rapi ies of ecological zones (such as the upland limit of Nui/ Easter Island (Flenley and Bahn 2003). In Ice- tree growth) are generally determined by tempera- land, measures were taken to conserve declining ture; if temperature changes then boundaries will woodland resources when they had reached the change (Ólafsdóttir, Schlyter, and Haraldsson point that further reductions could have been po- 2001). Temperatures decline and grass growth is tentially catastrophic. In southern Iceland, for ex- affected. In addition, cold wet periods can criti- ample, lowland woodlands were cleared rapidly cally affect other activities such as fodder collec- (Hallsdóttir 1987). At the farm of Stóra Mörk (c. tion, the chance of successfully over wintering 30 km west of Þórsmörk), for example, tephra lay- livestock, and lambing, producing high neo-natal ers in peat sections show that woodlands in the fatalities for sheep (Amorosi et al. 1997). area of the present home fi elds were cleared be- Despite these clear associations with climate, tween 870 and 920 A.D., and probably within it is apparent from the environmental record that the fi rst 30 years of settlement (Mairs et al. 2006) some environmental degradation, such as soil ero- (Fig. 1). Up valley, charcoal production pits (also sion, began very soon after settlement in the late preserved in aggrading soil profi les containing

WW4336.indb4336.indb 2222 44/20/07/20/07 10:15:3210:15:32 AAMM Dugmore, Keller, and McGovern: Contrasting Fates of Human Settlements in the North Atlantic Islands 23

N Farms abandoned by end of 13th century

MARKARFLJÓT WoodlandÞÓRSMÖRKÞÓÞÓRSMÖRKÓRSMÖRSSMÖRÖRKRRK K ES CharcoalLANGLLA GANESANE pits 10th- conserved STAKKHSTSTATAAKKKHKK 13th century KHHOLTL WoodsMÖRKMÖRK EYVINDAHOLT cleared late 9th century

DALURDDALUALUURUR KEY 100m contours

10 00m Rivers ÐAR AR Extent of Markarfljót RAG HAM Vegetated pasture SELJALANDSEELJA Cultivated land Landholding boundaries 50 0m 05km 1010 0m

Figure 4. Woodland change in southern Iceland. By the end of the thirteenth century A.D. the progressive loss of woodland in Markarﬂ jót valley had reached Þórsmörk; here farm sites were abandoned, and the surrounding woodlands survived through seven centuries to the present, while they continued to be used as a vital and major source of charcoal for farms in the region (Dugmore et al. 2006; Church et al. in press). This can be inter- preted as an example of the medieval Norse successfully responding to a developing environmental issue, and has implications for how we view the Norse Greenlanders’ responses to similar issues at similar times.

more tephra layers) record the use of birch wood dence from the sites themselves, to have been es- (Betula pubescens) between 920 –1341 A.D. (Dug- tablished early and abandoned before the four- more et al. 2006; Church et al. in press) (Fig. 4). teenth century (Tómasson 1996). Recent envi- No production pits have been found in this area ronmental work has shown that the occupation that post date the fourteenth century, and there of these sites coincided with major localized ep- is no birch woodland in the area today. At the in- isodes of soil erosion (Dugmore et al. 2006). Af- land margins of the valley isolated areas of wood- ter the Þórsmörk farm sites were abandoned, the land survive to this day in Þórsmörk, and this area affected areas began to regenerate and soils be- was intensively managed for charcoal production gan to reform. Once again, precise dating control through to early modern times. There is a pattern is given by the distribution of well-dated layers of of progressive woodland destruction and continued volcanic ash. Abandonment of these farm sites, ly- survival in isolated areas close to the upland eco- ing at the upland limit of settlement in the region, logical margin in Iceland in general and in other might be argued to be a consequence of the local- southern valleys such as Þjórsádalur in particular ized destruction of grazing lands associated with (Fig. 1) The implication from the environmental the ﬁ rst introduction of domesticated animals. Im- record is that actions were taken to conserve re- portantly, however, destruction was limited to up- sources before they were completely destroyed. land grazing areas. Within sheltered valleys con- In Þórsmörk, where the woodlands survived, taining woodlands today, there is evidence that there lie the ruins of four farm sites known, on the woodland survived throughout the period of occu- basis of both the written record and artifact evi- pation. One possibility is that farm abandonment,

WW4336.indb4336.indb 2323 44/20/07/20/07 10:15:3210:15:32 AAMM 24 Arctic Anthropology 44:1

while related to the increasing loss of grazing land, 2005). The record is not, however, one of univer- could in fact have been driven by a more general sal continuity and endurance as there are nota- communal need to conserve a diminishing supply ble examples of both localized extinction (walrus of woodland. Once the woodlands had been de- colonies no longer occupy the coasts of Iceland) stroyed down valley, people could have recognized and acts of species extinction (the last Great Auk the importance of conserving the surviving rem- (Pinguinus impennis) was killed in 1844 A.D. on nant in Þórsmörk. The farms (and presumably their a small island of Eldey off the coast of Iceland) associated livestock) were removed and the wood- (Bárðarson 1986). The key point is, however, that land, now under the control of lowland landowners there is empirical zooarchaeological evidence for and the church, survived (Dugmore et al. 2006). practices that began in Viking Age Iceland, at the A key point of this example is that even time Icelanders were colonizing Greenland that when exploitation of a resource and related envi- have proved sustainable over more than 1,000 ronmental degradation appears ill-conceived (the years. near-total destruction of native woodland), successful corrective conservation measures were em- ployed in at least some areas before the resource Grazing Management was totally lost. Signifi cantly for the understand- The effective management of grazing is another ing of events in Greenland, and these actions were key sign of adaptation. The notion of sustainabil- taking place in Iceland in the thirteenth century, ity does occurs in Norse records of medieval land showing that the Norse at this time were capable management as contained in passages of the Old of responding to an environmental challenge in Icelandic law code, Grágás (Dennis, Foote, and a constructive and ultimately semi sustainable Perkins 1993; Simpson et al. 2001). The code was manner. replaced in the 1280s A.D. after the end of the Ice- landic Commonwealth period, but offers key in- Animal Conservation and Extinction sights into the principles of early land management in Iceland, principles that could have been The ability to practice a sustainable exploitation similar to those practiced in contemporary Norse of wild resources could have been key to effective Greenland. Grazing rights were owned, managed, subsistence in Norse Greenland. In the Faroe Is- and regulated. Stocking levels were determined by lands (Fig. 1) a substantial pilot whale hunt has the stage at which livestock began to lose weight. been sustained for at least fi ve centuries, but ex- If the animals became thinner throughout the year, ploitation takes place within the confi nes of the ar- then stocking levels were deemed to be too high. chipelago, harvesting an extensive pelagic popu- This is strong evidence that upland environmental lation. The scales are such that it could have been degradation was not simply a result of over exploi- impossible for an island-based kill to compromise tation and a “tragedy of the commons.” the viability of the pilot whale population as a whole in the wider North Atlantic. Land based birds offer a more sensitive examples because of their potential vulnerability. Around Lake Mývatn The Ability To Adapt in Iceland (Fig. 1) harvesting of duck eggs is an im- A consideration of case studies of Faroese and portant local farm activity, with some 10,000 eggs early Icelandic use of marine mammal, bird, wood- a year collected. New evidence shows that duck land, and grazing resources argues for fl exibil- eggs were being exploited over 1,000 years ago, ity and innovation and signifi cant degrees of sus- and although many different species of adult birds tainable management within landscapes fi t for were being killed and eaten as well, zooarchaeo- purpose. We have examples of how to determine logical data indicate that waterfowl hunting was sustainable practice (monitoring the weight of live- avoided, and the Ptarmigan (Lagopus mutus) was stock), evidence of management tools (ownership the focus of hunting efforts (McGovern et al. in and regulation of grazing, legal sanctions for mal- press ) . A thousand years after exploitation began practice), examples of sustainable exploitation duck populations still fl ourish (because some eggs (duck eggs and pilot whales), and examples of cor- are always left in the nests to hatch) and, signifi - rective measures taken to conserve declining re- cantly, ptarmigan still endure in the heath and sources (woodlands for charcoal production). But uplands as well. Elsewhere in the Atlantic this in addition we do have examples of overexploi- view of sustainable Norse practices is reinforced: tation and extinction (the Icelandic walrus and in the Faroe Islands, puffi n (Fratercula arctica) Great Auk) and we have evidence of major envi- make up a signifi cant population of the early ar- ronmental issues (extensive soil erosion) and a his- chaeofauna at the site of Sandur, showing that they tory marked by repeated examples of population were intensively exploited for food, but again pop- decline and hardship. So why did people “get it ulations still fl ourish on the islands (Church et al. wrong”?

WW4336.indb4336.indb 2424 44/20/07/20/07 10:15:3310:15:33 AAMM Dugmore, Keller, and McGovern: Contrasting Fates of Human Settlements in the North Atlantic Islands 25

was children’s work, and so may not have drawn Problems Of Prediction the adult workforce away from other tasks. Possible explanations for why apparently good Second, although monitoring the weight of practices and proven abilities to adapt and re- livestock brought down off the hills in the autumn spond effectively to new challenges have produced is an effective way to maximize yield, the practice very mixed results, brings us back to the question is probably insensitive to the onset of accelerated of climatic impacts, especially the consequences environmental costs as it is probable that landscape of the Little Ice Age climate changes (Grove 1988; degradation would begin before livestock weight Meeker and Mayewski 2002). As the climate record reduction became apparent. As a result, landscape through the last two centuries of the Norse settle- problems could develop before the animal mon- ments in Greenland includes signifi cant variability itoring mechanism highlighted the issue. In this it is diffi cult to fasten on any one particular epi- case draw-down of natural capital could have been sode of unfavorable weather as the critical event as an immediate consequence of unforeseen (and un- these repeated frequently. Climate change could be managed) changes. Archaeological surveys have re- most readily translated into effects that people re- vealed examples of water-management strategies, sponded to badly by changes that are substantial, both to harness snow melting and for irrigation, unpredictable and repeated, as opposed to changes but most of these measures were focused on the that are gradual and predictable. Whether “bad” home-fi elds, leaving the rangeland more or less un- becomes “catastrophic” depends on how human touched. Boundary turf walls in the mountain pas- societies can cope, and this is a refl ection of so- tures, established to control the pastures (Einars- cial attitudes and structure, economic and politi- son, Hansson, and Vésteinsson 2002) may some- cal systems, as well as demographic fundamentals times have prevented, and some times (ironically) such as the potential labor force. Although a sud- have accelerated water erosion. den episode of bad weather in any one year may Thirdly, the nature of soil erosion in Iceland be diffi cult, a sequence of bad years could rapidly has distinctive features that can effectively prop- change hardship into crisis and then catastrophe, agate initially limited impacts. Over most of the as the ability to buffer any negative impacts of landscape vegetated at the time of Norse settlement change are exhausted. Extreme events capture the soil cover occurred in the form of a deep (Ͼ50cm) imagination, and it is easy to accept the argument sediment of aeolian origin with intercalated layers that major, abrupt changes may be signifi cant, but of volcanic ash (Dugmore et al. 2000; Thórarin- what about the lower order changes, gradual, but sson 1961). In the south of Iceland, lowland pre- signifi cant shifts in climate that erode prosperity settlement soil covers were commonly several me- and draw down both natural and social capital ters deep. Once trees have been removed, the root over a number of years? zone is effectively confi ned to upper layers of the Returning to the example of rangeland man- soil, and does not extend down into the underly- agement, problems in Iceland may well have de- ing substrait of coarse-grained sediments. Soil ero- veloped even though people understood environ- sion develops by exploiting breeches in the vegeta- mental dynamics and had sound management tion cover that allow frost, wind, and rain to move mechanisms in place, because of climate changes sediment. Bare patches of exposed substrait de- that unpredictably affected the onset and length of velop, bordered by steep, eroding faces of soil that the growing season. Although suffi cient biomass under-cut and eat into the surrounding vegetated could grow through the summer to support the live- areas. Soil erosion then proceeds as a loss of soil stock grazing on the hills, a delayed start to the area. Once established, erosion is resistant to sta- growing season could have resulted in grazing be- bilization, as even low grazing intensities will tend fore any growth began, compromising the stability to keep bare soil exposed on the eroding slopes, of vegetation cover (Simpson et al. 2001). permitting the continued effective action of the el- The problem created by short phases of ements. It is useful to think of soil erosion having heightened impacts on vegetation could have been two key stages: a triggering event, which results in compounded by three other factors. First,, mini- the breeching of the vegetation cover (and is usu- mal (if any) shepherding of livestock utilizing the ally related to grazing), and a propagation phase rangelands, while economizing on labor demands which is driven by the intensity of frost action, and freeing people for other tasks (such as fi shing rainfall, slope wash and wind acting on exposed or textile production), could have signifi cantly in- soil. A simple return to the conditions prevailing creased potential vulnerability to climatic impacts. before the onset of soil erosion does not necessar- Without day-to-day management of livestock graz- ily result in a cessation of erosion, and in that re- ing (to utilize the best or most robust areas of vege- spect soil erosion is best regarded as a threshold tation, and reduce pressure on vulnerable, eroding change that has the form of a catastrophe cusp. areas) regeneration would be diffi cult to promote. Although erosion could have been triggered Often in the Norse world, however, shepherding by a mismatch between land use and the growing

WW4336.indb4336.indb 2525 44/20/07/20/07 10:15:3310:15:33 AAMM 26 Arctic Anthropology 44:1

season, the initial damage may not have been im- sure is developed by determining the long term mediately apparent, as the monitoring mecha- mean of a time series of data, say for example, an- nisms seem to have focused on the livestock. But nual temperature. For each year the deviation from once erosion started, it would not necessarily sta- the long term mean is calculated, then, starting at bilize even if climate conditions returned to their the oldest data point, Year 1 is presented as simply previous norm. the deviation from the mean, Year 2 is defi ned by that year’s deviation plus the deviation from Year 1, Year 3 is defi ned by that year’s deviation plus Misleading Memories the deviations from both Year 1 and Year 2, and so People may fail to appreciate that signifi cant cli- on. This is illustrated for different proxy records mate change has occurred, or be resistant to taking in Figures 5, 6, and 7. Cumulative analysis is not effective counter measures until too late, if their used to assess climate changes such as the North experiences of the past are an inappropriate guide Atlantic Oscillation because the weather in any to the future. A quantifi cation of “unexpected” or one year is independent of what happened in the “unpredictable” change may be usefully gained previous year, but when considering people and from measures of cumulative deviation. This mea- the environment cumulative measures are most

Figure 5. The ␦ 18O time series from GISP2 (LHS scale), a proxy temperature record, shown against human isotopic data (RHS scale showing proportion of diet from marine sources) (Arneborg et al. 1999; Mayewski et al. 1993; Mayewski and White 2002; data from Greenland Ice Sheet Project 2. http://www.gisp2.sr.unh.edu/): Com- parisons of ␦ 18O time series for GISP2 and meteorological records for West Greenland for the period 1860 A.D. to the present demonstrate a statistical relationship between the ␦ 18O time series and mean winter air temperatures across West Greenland (Barlow 2001; Dawson et al. 2003). The lower line is the 5 year running mean of deviations from the long term mean. The upper line is the same data presented as cumulative deviations (see text for explanation). The points represent isotopic data on Norse Greenlanders showing the proportion of marine food consumed. The change in the cumulative deviation —marking a shift to cooler condition—also marks the time of a distinct shift in the isotopic records of diet. Closed circles data from the Eastern Settlement, open circles data from the Western Settlement.

WW4336.indb4336.indb 2626 44/20/07/20/07 10:15:3310:15:33 AAMM Dugmore, Keller, and McGovern: Contrasting Fates of Human Settlements in the North Atlantic Islands 27

Figure 6. The Na+ concentration (sea salt) time series, a proxy record of storminess, from GISP2, shown against human isotopic data —lower RHS scale showing proportion of diet from marine sources (Arneborg et al. 1999; Meeker and Mayewski 2002; data from Greenland Ice Sheet Project 2. http://www.gisp2.sr.unh.edu/): The Na+ in the Greenland icecap derives principally from sea salt in the North Atlantic (Meeker and Mayewski 2002), so temporal variations in sea salt concentrations in the GISP2 record represent a proxy record for winter storminess in the North Atlantic. The upper line (top LHS scale) is the 5 year running mean of deviations from the long term mean. The lower line (RHS scale) is the same data presented as cumulative deviations (see text for explanation). Storminess is an indicator of regional circulation change in the North Atlantic, and perhaps unsur- prisingly this regional measure of change does not show a strong correlation with evidence of dietary changes in Greenland but the most signiﬁ cant shift occurs within a short time of the ﬁ nal extinction of the Norse settlement. Closed circles data from the Eastern Settlement, open circles data from the Western Settlement.

instructive. For example, bad weather one year ten when memories mislead; knowing when it is does have a continuing effect through its impact the time to change is perhaps the most obvious on vegetation; rain and wind year after year have statement of how to deal with climate change, but cumulative effects on the development of soil ero- perhaps the most difﬁ cult to resolve. How long is sion. Crucially, the measure of cumulative devi- a new situation going to last? Do we just continue ation clearly shows when trajectories of change in the ways we always have, or do we take difﬁ - alter. This is important not only for the environ- cult decisions to alter what we do? Trying to man- ment, but also for people’s perception of it. Our age a subsistence economy in a world of changing views of what is likely to happen next year are climates is a game where the player is blindfolded, based on knowledge accumulated over previous since there is no way of knowing what the future years. Today, just as in the past, we use experi- has in store. Still, decisions have to be made. In a ences and memory to anticipate what will happen historical perspective, some of them will inevita- next. Inappropriate actions may be taken most of- bly prove to be wrong.

WW4336.indb4336.indb 2727 44/20/07/20/07 10:15:3410:15:34 AAMM 28 Arctic Anthropology 44:1

Figure 7. The proxy record of sea ice from GISP2 –RHS scales, bold fi gures cumulative scale (Dugmore et al. in press; Mayewski et al. 1993) shown against human isotopic data, RHS scale showing proportion of diet from marine sources (Arneborg et al. 1999): The upper line is the 5 year running mean of deviations from the long term mean (data from Greenland Ice Sheet Project 2. http://www.gisp2.sr.unh.edu/). The lower line is the same data presented as cumulative deviations (see text for explanation). The most signifi cant shift of the cumulative measure (most pronounced change in over 400 years, when past experience of anticyclonic conditions/sea ice extent were becoming the poorest indicator of future trajectories of change) occurs within a short time of the fi nal extinction of the Norse settlement. Closed circles data from the Eastern Settlement, open circles data from the Western Settlement. The extent of sea ice from GISP2 Cl - excess timeseries: GISP2 ice core chemistry has been used to derive an estimate of past changes in sea ice extent around Greenland using the chloride (Cl -) time series (both chloride excess and Cl -/Na+). Mayewski et al. (1993). Sea ice in the Greenland Sea is most extensive during April and May and the expansion of sea ice is associated with regionally increased surface air pressure . As a result changes in sea ice extent have been inferred from the non-sea salt Cl - record as it is thought to refl ect anticyclonic activity. Caution is needed in the interpretation of the chloride (Cl -) time series, as sea ice expansion over the Greenland Sea can coincide with sea ice contraction in the Davis Strait and along the West Greenland coast and vica versa. Between 1270 A.D. and the twentieth century A.D. the pattern of the GISP2 Cl - excess timeseries is broadly in agreement with historical reconstructions (e.g. Ogilvie, 1984, 1992), with a period of reduced sea ice cover between ca 1050 –1420 A.D. followed by a marked increase in sea ice extent that culmi- nated during the late nineteenth century A.D. (Mayewski et al. 1993).

WW4336.indb4336.indb 2828 44/20/07/20/07 10:15:3410:15:34 AAMM Dugmore, Keller, and McGovern: Contrasting Fates of Human Settlements in the North Atlantic Islands 29

Isotopic data on the diet of Norse Greenland- nerability, because people exploit a range of plants ers (Arneborg et al 1999) has been plotted on the and animals that could have wider or different same graphs as temperature proxy data (Fig. 5), sensitivities to climatic change than those who fo- storminess data (Fig. 6) and sea ice proxy data cus on a small area or specifi c parts of an ecosys- (Fig. 7). Although the fi ve-year running mean of tem. Broad-based strategies of exploitation may, temperature deviation shows little correspondence however, promote another sort of vulnerability to the isotopic record, the biggest shift in the cumu- by limiting labor availability for any one partic- lative record coincides with a decisive shift in diet. ular activity and so lessening its effi ciency or re- This suggests fi rst that, , cumulative measures could turn. Commercial opportunities can connect com- be the best guide to the type of climate change that munities across wide areas, but in doing so they has the most signifi cance for people and, second, may also create increased potential for disruption that the Norse in Greenland did respond to this as a result of climate change, or indeed also for de- change with a shift in diet. The storminess record mographic or ecological changes in remote mar- in Figure 6 shows what could have been a trigger kets. This can happen because there is a greater for the end of the Norse settlement. The notable chance that climate change may affect one part of change in storminess in 1425 A.D. comes after the interconnection system, and so have cascading the last recorded contact with the settlers. A sim- effects outside the areas actually affected by the ilar relationship is suggested by the sea ice proxy changing weather. (Fig. 7). In Greenland, and Iceland the mid fi f- Questions of scale are crucial. We can, for teenth century A.D. was a time when past experi- example, see how the fate of farms at the upland ence became an increasingly inappropriate guide limits of settlement may have been affected by to the future, and this increasing uncertainty about changes in the lowlands, so that despite localized the future, in combination with the economic environmental degradation (soil erosion), the key changes that had made the colonies increasingly environmental change that determined abandon- marginal, might have signaled the end of the Norse ment was most probably degradation elsewhere settlements. No one knows exactly how long the (loss of woodland) and the needs that arose from last Norse Greenlanders survived. It is possible, that (conservation of surviving woodland at the however, that towards the end the social order may upland margin). Likewise, but on a much larger have broken down, and with it the communal ef- scale, commercial changes in Europe could have forts necessary for survival, resulting in a total col- been a key driver in creating sensitivities to cli- lapse of the society. If so, fi nal extinction may have mate change on the other side of the Atlantic in come about very quickly. Norse Greenland. We conclude that shifting market conditions very far from Norse Greenland may have contributed to its economic marginalization, Conclusions through the opening of the Mediterranean trade Studies of past environments offer valuable in- routes and the re-introduction of elephant ivory sights into the relationships between people and on the European market. climate change, and offer instructive examples Asymmetric thresholds or catastrophe cusps of previous outcomes. As many alternative ideas may be crossed even in the presence of “sustain- come into play about the impact of climate change, able” and responsive management strategies where the ways in which people respond to the chal- changes are unpredictable and environmental im- lenges these changes pose, and the signifi cance of pacts are not immediately signifi cant or apparent. coincident but wholly unrelated cultural changes, Cumulative measures of change are good analyti- it is important to have rich, multi-layered data sets cal tools because they help focus attention on the that include a wide range of environmental and cul- implications for both cultural and environmental tural records. High resolution, seasonal, annual, “memories” when the past is no longer an effective and sub-decadal chronologies are essential, as are guide to the future. many different scales of enquiry. The human di- One widely held view is that the impact of mension needs to be evaluated on seasonal times- climate change, the failure of their pastoral sub- cales, and spatial resolutions extending from in- sistence base, and an inability to adapt were key dividual settlement sites and land holdings, to factors in the end of Norse settlement in Green- groups of settlements in regional landscapes, and land (Fig. 2). Alternatively, as we argue here, un- trade and socio-political relationships across con- favorable economic changes and falling popula- tinental scales. tions might actually have been the key factors in Crucial distinctions have to be made between increasing the settlements’ vulnerability to extinc- climate effects on subsistence and cash economies. tion (Fig. 8). Through much of the settlements’ his- Where subsistence is based on a wide range of re- tory the subsistence base of Norse Greenlanders sources (such as marine and terrestrial food sup- could have been robustly based on hunting. Ulti- plies) this broad base may reduce one sort of vul- mately, however, critical numbers of people were

WW4336.indb4336.indb 2929 44/20/07/20/07 10:15:3410:15:34 AAMM 30 Arctic Anthropology 44:1

Figure 8. A conceptual model of Norse Greenland settlement highlighting the key roles of hunting, trade and social collapse.

necessary to maintain this way of life. Against a Petter Gleditch at the International Peace Research background of economic marginalization and pop- Institute, Oslo (PRIO), triggered the writing of the ulation decline, we conclude that cumulative paper through his invitation to participate in a changes in climate and their relation to differing workshop on Human Security and Climate Change. cultural and environmental “memories,” combined The comments of two anonymous referees were with a cascading collapse of an integrated interde- most helpful in reﬁ ning the ﬁ nal paper. pendent settlement system as a result of depopulation could have been the key factors in triggering the end of Norse Greenland. Endnote 1. This paper is based on the keynote presentation Acknowledgments. We gratefully acknowledge the to the conference on Human Security and Climate support of the Leverhulme Trust (through the pro- Change held in Oslo, June 21–23, 2005, a meeting gram Landscapes circum Landnám), the National was organized by the Centre for the Study of Civil Geographic Society Committee for Research and War, International Peace Research Institute, Oslo Exploration, CUNY Northern Science and Educa- (PRIO), the Centre for International Environmen- tion Center, VECO Polar Services, the Greenland tal and Climate Research at the University of Oslo Museum and Archives, and the United States Na- (CICERO), and the Global Environmental Change tional Science Foundation (both Archaeology and Human Security Program (GECHS), and spon- Program and Arctic Social Science Program). Nils sored by the United Nations.

WW4336.indb4336.indb 3030 44/20/07/20/07 10:15:3410:15:34 AAMM Dugmore, Keller, and McGovern: Contrasting Fates of Human Settlements in the North Atlantic Islands 31

eds. Pp. 15–24. Hertingen/Basel: Folio Verlag, References Cited Dr. G. Wesselkamp. Albrethsen, Svend E. and Christian Keller Arneborg, Jette, Jan Heinemeier, Niels Lynnerup, Hen- 1986 The Use of the Saeter in Medieval Norse Farm- rik L Nielsen, Niels Rud, and Arny E. Sveinbjörnsdóttir ing in Greenland. Arctic Anthropology 23(1–2): 1999 Change of Diet of the Greenland Vikings De- 91–107. termined from Stable Carbon Isotope Analysis Amorosi, Thomas, Paul C. Buckland, Andrew J. Dug- and 14C Dating of Their Bones. Radiocarbon more, Jon H. Ingimundarson, and Thomas H. McGovern 41(2):157–168. 1997 Raiding the Landscape: Human Impact in the Barlow, Lisa. K. North Atlantic. Human Ecology 25:491–518. 2001 The Time Period A.D. 1400 –1900 in Central Amorosi, Thomas, Paul C. Buckland, Kevin J. Edwards, Greenland Ice Cores in Relation to the North Ingrid L. Mainland, Thomas H. McGovern, Jon P. Sadler, Atlantic Sector. Climatic Change 48:101–119. and Peter Skidmore Barlow, Lisa K., Jon P. Sadler, Astrid E. J. Ogilvie, Paul C. 1998 They Did Not Live by Grass Alone: The Politics Buckland, Thomas Amorosi, Jon H. Ingimundarson, and Paleoecology of Animal Fodder in the North Peter Skidmore, Andrew J. Dugmore, and Thomas H. Atlantic Region. Environmental Archaeology McGovern 1:41–54. 1997 Interdisciplinary Investigations of the End of Amundsen, Colin, Sophia Perdikaris, Thomas H. Mc- the Norse Western Settlement in Greenland. The Govern, Yekaterina Krivogorskaya, Matthew Brown, Holocene 7:489–99. Konrad Smiarowski, Shaye Storm, Salena Modugno, Bárðarson, Hjálmar R. Malgorzata Frik, and Monica Koczela 1986 Birds of Iceland. Reykjavik: H. R. Bárðarson. 2005 Fishing Booths and Fishing Strategies in Medi- 336 pp. eval Iceland: An Archaeofauna from the Site of Akurvík, North-West Iceland. Environmental Barrett, James, H. Archaeology 10 (2):141–198. 2003 Culture Contact in Viking Age Scotland. In Con- tact, Continuity, and Collapse: The Norse Colo- Arnalds, Ólafur, Elin F. Thorarinsdóttir, Sigmar Metu- nization of the North Atlantic. James Barrett, ed. salemsson, Asgeir Jonsson, Einar Gretarsson, and Arnor Pp. 73–113. Studies in the Early Middle Ages, Arnason 5. Turnhout: Brepols. 1997 Soil Erosion in Iceland. Reykjavik: Icelandic SCS and the Agricultural Research Institute. 2005 Economic Intensiﬁ cation. Viking and Medieval Orkney, Scotland: Excavations at Quoygrew. In Arneborg, Jette Viking and Norse in the North Atlantic: Select 2005 Greenland Irrigation Systems on a West Nordic Papers from the Proceedings of the 14th Viking Background: An Overview of the Evidence of Congress, Tórshavn, Faroe Islands 2001. Andras Irrigation Systems in Norse Greenland c. 980 – Mortensen and Símun Arge, eds. Pp. 264 –284. 1450 AD. In Water Management in Medieval Ru- Tórshavn: Annales Societatis Scientarium ral Economy. J. Klápšteˇ, ed. Ruralia V. Pp. 137– Faeroensis Supplementa, 44. 145. Památky Archaeologické. Supplementum, 17. Prague: Institute of Archaeology. Barrett, James H., R. Beukens, I. A Simpson, P. Ashmore, S. Poaps, and J. Huntley 2003a Norse Greenland Archaeology: The Dialogue 2000 What Was the Viking Age and When Did It Hap- between the Written and the Archaeological pen? A View from Orkney. Norwegian Archaeo- Record. In Vinland Revisited: The Norse World logical Review 33:1–39 at the Turn of the First Millennium. Selected papers from the Viking Millennium Interna- Barrett, James H., A. M. Locker, and C. M. Roberts tional Symposium 15 –24 Septem ber, 2000, 2004a The Origin of Intensive Marine Fishing in Medi- Newfoundland and Labrador. S. M. Lewis, ed. eval Europe: The English Evidence. Proceedings Pp. 111–123. St. JohnЈs: Newfoundland Historic of the Royal Society B 271:2417–2421. Sites Association. 2004b “Dark Age Economics” Revisited: The English 2003b Norse Greenland –Reﬂ ections on the Discus- Fish Bone Evidence A.D. 600 –1600. Antiquity sion of Depopulation. In Contact, Continuity, 78(301):618– 636. and Collapse: The Norse Colonization of the North Atlantic. James Barrett, ed. Pp. 163–181. Barrett, James H., R. Nicholson, and R. Cerron-Carrasco Studies in the Early Middle Ages, 5: Turnhout: 1997 Fish Trade in Norse Orkney and Caithness: A Brepols. Zooarchaeological Approach. Antiquity 71: 616– 638. 2002 Inhospitable Regions and Marginality– Settlement at the End of the World. In Centre, Berglund, Joel Region, Periphery Medieval Europe, vol. 3. 1986 The Decline of the Norse Settlements in Green- G. Helmig, B. Scholkmann, and M. Untermann, land. Arctic Anthropology 23(1–2):109–135.

WW4336.indb4336.indb 3131 44/20/07/20/07 10:15:3410:15:34 AAMM 32 Arctic Anthropology 44:1

Bond, Julie M. Keller, Paul A. Mayewski, Thomas H. McGovern, Kerry- 1998 Beyond the Fringe? Recognising Change and Anne Mairs, and Guðrún Sveinbjarnardóttir Adaptation in Pictish and Norse Orkney. In Life In The Role of Climate in Settlement and Land- on the Edge: Human Settlement and Marginal- press scape Change in the North Atlantic Islands: ity. C. M. Mills and G. Coles, eds. Pp. 81–90. An Assessment of Cumulative Deviations in Symposia of the Association for Environmental High-Resolution Proxy Climate Records. Human Archaeology No.13, Oxbow Monograph, 100. Ecology. Oxford: Oxbow Books. Dugmore, Andrew J. and Paul C. Buckland Brewington, Seth D. 1991 Tephrochronology and Late Holocene Soil Ero- 2007 Viking-Age to Late-Norse Archaeofauna from sion in South Iceland. In Environmental Change Undir Junkarinsfl øtti, Faroe Islands: A Prelimi- in Iceland: Past and Present. Judith K. Maizels nary Report. In Life on the Edge: Maritime Ar- and Christopher Caseldine, eds. Pp 147– 61. chaeology from the North Atlantic Rim of Eu- Dordrecht: Kluwer. rope. Stephen Wickler and Colin P. Amundsen, Dugmore, Andrew J., Mike J. Church, Paul C. Buckland, eds. Oxford: British Archaeological Reports. Kevin J. Edwards, Ian T. Lawson, Thomas H. McGovern, Buckland, Paul C., Thomas Amorosi, Lisa K. Barlow, Eva Panagiotakopulu, Ian A. Simpson, Peter Skidmore, Andrew J. Dugmore, Paul A. Mayewski, Thomas H. and Guðrún Sveinbjarnardóttir McGovern, Astrid E. J. Ogilvie, Jon P. Sadler, and Peter 2005 The Norse Landnám on the North Atlantic Is- Skidmore lands: An Environmental Impact Assessment. 1996 Bioarchaeological and Climatological Evidence Polar Record 41:21–37. for the Fate of Norse Farmers in Medieval Green- Dugmore, Andrew J., Mike J. Church, Kerry-Ann Mairs, land. Antiquity 70 (267):88–96. Thomas H. McGovern, Anthony J. Newton, and Guðrún Buckland, Paul C., Andrew J. Dugmore, David W. Perry, Sveinbjarnardóttir Diana Savory, and Guðrún Sveinbjarnardóttir 2006 An Over-Optimistic Pioneer Fringe? Environ- 1991 Holt in Eyjafjallasveit, Iceland. A Palaeoecologi- mental Perspectives on Medieval Settlement cal Study of the Impact of Landnám. Acta Ar- Abandonment in Þórsmörk, South Iceland. In chaeologica 61:252–271. The Dynamics of Northern Societies. Bjarne Grønnow, Jette Arneborg, and Hans C. Gulløv, Church, Mike J., Simun V. Arge, Seth Brewington, eds. Pp. 333–344. PNM, Publications from the Thomas H. McGovern, James Woollett, Sophia Perdikaris, National Museum, Studies in Archaeology and Ian T. Lawson, Gordon T. Cook, Colin Amundsen, History, 10. Copenhagen: National Museum of R. Harrison, Kate Krivogorskaya, and E. Dunbar Denmark. 2005 Puffi ns, Pigs, Cod, and Barley: Palaeoeconomy at Undir Junkarinsfl øtti, Sandoy, Faroe Islands. Dugmore, Andrew J. and Camilla C. Erskine Environmental Archaeology 10(2):179–197. 1994 Local and Regional Patterns of Soil Erosion in Southern Iceland. Münchener Geographische Church, Mike. J, Andrew J. Dugmore, Kerry-Anne Mairs, Abhandlungen 12:63–79. Andrew R. Millard, Gordon T. Cook, Gudrun Sveinbjar- nardóttir, Philippa A. Ascough, and Katy H. Roucoux Dugmore, Andrew J., Guðrún Larsen, and Anthony J. In Charcoal Production during the Norse and Early Newton press Medieval Periods in Eyjafjallahreppur, Southern 2004 Tephrochronology and its Application to Late Iceland. Radiocarbon. Quaternary Environmental Reconstruction, with Special Reference to the North Atlantic Islands. Dawson, Alastair G., Lorne Elliott, Paul A. Mayewski, In Tools for Constructing Chronologies. Caitlin Peter Lockett, Sean Noone, Kieran Hickey, Tom Holt, E. Buck and Andrew R. Millard, eds. Pp. 173– Peter Wadhams, and Ian Foster 188. London: Springer-Verlag. 2003 Late Holocene North Atlantic Climate “See- saws,” Storminess Changes, and Greenland Ice Dugmore, Andrew J., Anthony J. Newton, Guðrún Sheet (GISP2) Palaeoclimates. The Holocene Larsen, and Gordon T. Cook 13(3):381–392. 2000 Tephrochronology, Environmental Change, and the Norse Colonization of Iceland. Environmen- Dennis, A., P. Foote, and R. Perkins (trans.) tal Archaeology 5:21–34. 1993 Laws of Early Iceland, Gragas: The Codex Re- gius of Gragas with Material from Other Manu- Edwards, Kevin J., Buckland, Paul C., Dugmore, An- scripts. 2 vols. Winnipeg: University of Mani- drew J., McGovern, Thomas H., Simpson, Ian A. Sveinb- toba Press. jarnardóttir, and Guðrún Sveinbjarnardóttir 2004 Landscapes Circum-Landnám: Viking Settle- Diamond, Jared ment in the North Atlantic and Its Human 2005 Collapse: How Societies Choose to Fail or Sur- and Ecological Consequences: A Major New vive. London: Allen Lane. Research Programme. In Atlantic Connections Dugmore, Andrew J., Douglas M. Borthwick, Mike J. and Adaptations: Economies, Environments Church, Alastair Dawson, Kevin J. Edwards, Christian and Subsistence in Lands Bordering the North

WW4336.indb4336.indb 3232 44/20/07/20/07 10:15:3510:15:35 AAMM Dugmore, Keller, and McGovern: Contrasting Fates of Human Settlements in the North Atlantic Islands 33

Atlantic. Rupert Housley and Geraint M. Coles, Grove, Jean M. eds. Pp. 260 –271. Oxford: Oxbow. 2001 The Onset of the Little Ice Age In History and Climate: Memories of the Future? Phil D. Edvardsson, Ragnar, Sophia Perdikaris, Thomas H. Jones, Astrid E. J. Ogilvie, Trevor D. Davies, McGovern, Noah Zagor and Max Waxman and Keith R. Briffa, eds. New York: Kluwer 2004 Coping with Hard Times in North-West Iceland: Academic/Plenum. Zooarchaeology, History, and Landscape Ar- chaeology at Finnbogastaðir in the 18th century. Gulløv, Hans-Christian Archaeologica Islandica 3:20 – 48. 1997 From Middle Ages to Colonial Times: Archaeo- logical and Ethnohistorical Studies of the Thule Edvardsson, Ragnar and Thomas H. McGovern Culture in South West Greenland 1300 –1800 2005 Archaeological Excavations at Vatnsfjörður A.D. Meddelelser om Grønland, Man and Soci- 2003–04. Archaeologica Islandica 4:16–30. ety, 23. Einarsson, Árni, O. Hansson, and Orri Vésteinsson 1999 The Eskimo Cultures of Greenland and the 2002 An Extensive System of Medieval Earthworks Medieval Norsemen: A Contribution to History in Northeast Iceland. Archaeologia Islandica and Ethnohistory. In Proceedings of the Interna- 2:61–73. tional Congress on the History of the Arctic and Eysteinsson, Þröstur and Sigurður Blöndal Sub-Arctic Regions. I. Sigurdsson, ed. Pp. 54 – 2003 The Forests of Iceland at the Time of Settlement: 71. Reykjavik: University of Iceland. Their Utilisation and Eventual Fate. In Vinland 2000 Natives and Norse in Greenland. In Vikings: The Revisited: The Norse World at the Turn of the North Atlantic Saga. William W. Fitzhugh and First Millennium. Shannon Lewis Simpson, ed. Elisabeth I. Ward, eds. Pp.318–327. Washing- Pp. 411– 417. St John’s: Historic Sites Associa- ton, D.C.: Smithsonian Institution Press. tion of Newfoundland and Labrador, Inc. Gulløv, Hans-Christian (ed.) Fagan, Brian 2004 Grønlands Forhistorie. Copenhagen: Gyldendal. 2000 The Little Ice Age. New York: Basic Books. Guttesen, Rolf 2004 The Long Summer: How Climate Changed Civi- 1996 Hvalbøur: Largest Old Farming Village. In The lization. London: Granta Books. Faeroe Islands Topographic Atlas. Rolf Gut- tesen, ed. Pp. 88–92. Copenhagen: Kort and Fitzhugh William W. and Elisabeth I. Ward (eds.) Matrikelstyrelsen. 2000 Vikings: The North Atlantic Saga. Washington, D.C.: Smithsonian Institution Press. Hallsdóttir, Margrét 1987 Pollen Analytical Atudies of Human Inﬂ uence Flenley, John and Paul Bahn on Vegetation in Relation to the Landnám Tephra 2003 The Enigmas of Easter Island. Oxford: Oxford layer in Southwest Iceland. Lundqua Thesis 18: University Press. Lund: Department of Quaternary Geology, Lund University. Friðriksson, Adolf 1994 Sagas and Popular Antiquarianism in Icelandic Hreinsson, Viðar (ed) Archaeology. Aldershot, Avesbury: Samson 1997 The Complete Sagas of the Icelanders. Vols. I–V. Publishers. Reykjavík: Leifur Eiriksson Publishing. Friðriksson, Adolf and Orri Vésteinsson Jones, Gwyn 2003 Creating a Past. A Historiography of the Settle- 1986 The Norse Atlantic Saga: Being the Norse Voy- ment of Iceland. In Contact, Continuity, and ages of Discovery and Settlement to Iceland, Collapse: The Norse Colonization of the North Greenland, and North America. Oxford: Oxford Atlantic. James Barrett, ed. Pp. 139–161. Studies University Press. in the Early Middle Ages 5: Turnhout: Brepols. Karlsson, Gunnar Gad, Finn 1996 Plague without Rats: The Case of Fifteenth 1970 The History of Greenland, Volume 1. London: Century Iceland. Journal of Medieval History G. Hurst and Co. 22:263–284. Gibson-Lonsdale, J. J. Karlsson, Gunnar 1990 Pilot Whaling in the Faroe Islands: Its History 2000 Iceland’s 1100 Years: History of a Marginal Soci- and Present Signiﬁ cance. Mammal Review ety. Reykjavik: Mál og menning,. 20(1):44 –52. Keller, Christian Gottfried, Robert S. 1983 Gård og seter på Grønland –et forsøk på å analy- 1983 The Black Death: Natural and Human Disaster sere ressurstilgangen ved help av satelittbilder. in Medieval Europe. New York: Free Press. In Hus, gård og bebyggelse. Guðmundur Ólafs- son, ed. Pp. 59– 66. Föredrag från det XVI nord- Grove, Jean M. iska arkeologmøtet, Island 1982. Þjóðminjasafn 1988 The Little Ice Age. London: Methuen. Íslands. Reykjavík.

WW4336.indb4336.indb 3333 44/20/07/20/07 10:15:3510:15:35 AAMM 34 Arctic Anthropology 44:1

1986 Nordboerne på Grønland 985–1350. Bidrag til en Mayewski, Paul A. and Frank White demograﬁ sk økologisk diskusjon. Universitetets 2002 The Ice Chronicles. Hanover and London: Uni- Oldsaksamling Årbok 1984/5 145–157. Oslo. versity of New England Press. Kristjánsson, J. McGhee, Robert 1993 Icelandic Manuscripts, Sagas, History, and Art. 1984 Contact between Native North Americans and Reykjavík: The Icelandic Literary Society. the Medieval Norse: A Review of the Evidence. American Antiquity 49(1):4 –26. Krogh, Knud J. 1974 Kunstvanding –hemmeligheden bag Grønland- McGhee, Robert spispens hundrede køer. Nationalmuseets Ar- 2003 New Perspectives on the Thule Migration and bejdsmark, Copenhagen. Norse Contact. Paper presented at the NARP 1982 Erik den rødes Grønland. Nationalmuseets For- (Nordic Arctic Research Program) meeting, lag, Copenhagen. Akureyri Iceland May 23rd, 2002. Linden Eugene McGovern, Thomas H. 2006 The Winds of Change: Climate, Weather, and the 1979 The Palaeoeconomy of Norse Greenland: Ad- Destruction of Civilizations. New York: Simon aptation and Extinction in a Tightly Bounded and Schuster. Ecosystem. Unpublished PhD dissertation, De- partment of Anthropology, Columbia University. Lynnerup, Niels 2000 Life and Death in Norse Greenland. In Vikings: 1985a The Arctic Frontier of Norse Greenland. In The North Atlantic Saga. William W. Fitzhugh The Archaeology of Frontiers and Boundaries. and Elisabeth I. Ward, eds. Pp. 285–295. Wash- S. Green and S. Perlman, eds. Pp. 275–323. ington, D.C.: Smithsonian Institution Press. New York: Academic Press. Lynnerup, Niels 1985b Contributions to the Paleoeconomy of Norse 2003 Palaeodemography of the Greenland Norse. In Greenland. Acta Archaeologica 54:73–122. Vinland Revisited: The Norse World at the Turn 1991 Climate, Correlation, and Causation in Norse of the First Millennium. Shannon Lewis Simp- Greenland. Arctic Anthropology 28(2):77–100. son, ed. Pp.133–145. St. John’s: Historic Sites Association of Newfoundland and Labrador, Inc. 1992 Bones, Buildings, and Boundaries: Paleoeco- nomic Approaches to Norse Greenland. In Norse Mainland, Ingrid and Paul Halstead and Later Settlement and Subsistence in the 2005 The Economics of Sheep and Goat Husbandry North Atlantic. Christopher D. Morris and James in Norse Greenland. Arctic Anthropology Rackham, eds. Pp.157–186. Glasgow: Glasgow 42(1):103–120. University Press. Mairs, Kerry-Anne, Mike J. Church, Andrew J. Dugmore, 1994 Management for Extinction in Norse Greenland. and Guðrún Sveinbjarnardóttir In Historical Ecology: Cultural Knowledge and 2006 Environmental Change and Long-Term Settle- Changing Landscapes. Carole Crumley, ed. ment: The Mörk and Dalur Farm Complexes of Pp. 127–154. Santa Fe: School of American Eyjafjallasveit, South Iceland. In The Dynamics Research. of Northern Societies. Bjarne Grønnow, Jette Ar- neborg, and Hand C. Gulløv, eds. Pp. 363–372. 2000 The Demise of Norse Greenland. In Vikings: The PNM, Publications from the National Museum, North Atlantic Saga. William W. Fitzhugh and Studies in Archaeology and History, 10. Copen- Elisabeth I. Ward, eds. Pp. 327–39. Washington, hagen: National Museum of Denmark. D.C.: Smithsonian Institution Press. Mason, Owen K. McGovern, T. H., G. F. Bigelow, T. Amorosi, and 1998 The Contest between Ipuitak, Old Bering Sea D. Russell and Birnirk Polities and the Origin of Whaling 1988 Northern Islands, Human Error and Environ- during the First Millenium A.D. along the Ber- mental Degradation: A Preliminary Model for ing Strait. Journal of Anthropological Archaeol- Social and Ecological Change in the Medieval ogy 17(3):240 –325. North Atlantic. Human Ecology 16(3):45–105. Mathiassen, Therkel McGovern T. H., T. Amorosi, S. Perdikaris , and 1958 The Sermermiut excavations 1955. Meddelelser J. Woollett om Grønland 161:3. 1996 Zooarchaeology of Sandnes V51: Economic Change at a Chieftain’s Farm in West Greenland. Mayewski, Paul A., L. D. Meeker, M. C. Morrison, M. S. Arctic Anthropology 33(2):94 –122. Twickler, S. Whitlow, K. K. Ferland, D. A. Meese,. M. R. Legrand, and J. P. Stefferson McGovern, Thomas H. and Sophia P. Perdikaris 1993 Greenland Icecore “Signal” Characteristics: An 2000 The Vikings’ Silent Saga: What Went Wrong Expanded View of Climate Change. Journal of with the Scandinavian Westward Expansion? Geophysical Research 98:839– 847. Natural History Magazine 109:51–56.

WW4336.indb4336.indb 3434 44/20/07/20/07 10:15:3510:15:35 AAMM Dugmore, Keller, and McGovern: Contrasting Fates of Human Settlements in the North Atlantic Islands 35

McGovern Thomas H., Sophia Perdikaris, and Clayton Perdikaris, Sophia and Thomas H. McGovern Tinsley In Viking Age Economics and the origins of com- 2001 Economy of Landnám: The Evidence of Zooar- press b mercial cod fi sheries in the North Atlantic. In chaeology. In Approaches to Vinland. A. Wawn The North Atlantic Fisheries in the Middle Ages and Thorunn Sigurdardottir, eds. Pp. 154 –165. and Early Modern Period: Interdisciplinary Ap- Sigurdur Nordal Institute Studies, 4. Reykjavik: proaches in History, Archaeology, and Biology. Nordal Institute Studies. Louis Sickling and Darlene Abreu-Ferreira, eds. Netherlands: Brill Publishers. McGovern, Thomas H., Sophia Perdikaris, Árni Einars- son, and Jane Sidell Perdikaris, Sophia and Thomas H. McGovern In Inland Cod and Sustainable Waterfowl: Wild In Codfi sh and Kings, Seals and Subsistence: Norse press Species Use in Viking Age Mývatnssveit, N. press c Marine Resource Use in the North Atlantic. Iceland. Environmental Archaeology 11(1). In Human Impacts on Marine Environments. Torben Rick and Jon Erlandson, eds. Berkeley: Meeker, L. D. and Paul A. Mayewski UCLA Press Historical Ecology Series. 2002 A 1400 Year Long Record of Atmospheric Cir- culation over the North Atlantic and Asia. The Petersen, Hans Christian Holocene 12:257–266. 2000 The Norse Legacy in Greenland. In Vikings: The Meldgaard, Jørgen North Atlantic Saga. William W. Fitzhugh and 1965 Nordboerne i Grønland. Copenhagen. Elisabeth I. Ward, eds. Pp. 340–349. Washing- Munksgaard. ton, D.C.: Smithsonian Institution Press. Nørlund, Povl Roesdahl, E. 1930 Norse Ruins at Gardar: The Episcopal Seat of 2005 Walrus Ivory: Demand, Supply, Workshops, and Mediaeval Greenland. Meddelelser om Grøn- Greenland. In Viking and Norse in the North At- land, 76(1). Copenhagen. lantic: Select Papers from the Proceedings of the 14th Viking Congress, Tórshavn, 2001. Andras Ogilvie, Astrid E. J. Mortensen and Símun Arge, eds. Pp. 182–191. 1984 The Past Climate and Sea-Ice Record from Ice- Tóshavn Faroe Islands: Annales Societatis Sci- land, Part 1: Data to A.D. 1780. Climatic Change entarium Faeroensis Supplementa, 44. 6:131–152. Schledermann P. and K. M. McCullough 1992 Documentary Evidence for Changes in the Cli- 2003 Inuit-Norse Contact in the Smith Sound Re- mate of Iceland. In Climate since A.D. 1500. gion. In Contact, Continuity, and Collapse: The R. S. Bradley and P. D. Jones, eds. Pp. 92–117. Norse Colonization of the North Atlantic. James London: Routledge. Barrett, ed.. Pp. 183–207. Turnhout: Brepols. Ogilvie, Astrid E. J. and Thomas H. McGovern 750 –1400 2000 Sagas and Science: Climate and Human Impacts Seaver, K. A. in the North Atlantic. In Vikings: The North At- 1996 The Frozen Echo. Greenland and the Explora- lantic Saga. William W. Fitzhugh, and Elisabeth tion of North America, ca. A.D. 1000 –1500. I. Ward, eds. Pp. 385–393. Washington, D.C.: Stanford: Stanford University Press. Smithsonian Institution Press. Simpson, Ian A., Andrew J. Dugmore, Amanda M. Olafsdóttir, Rannveig, Paul Schlyter, and Harald Thomson, and Orri Vésteinsson Haraldsson 2001 Crossing the Thresholds: Human Ecology and 2001 Simulating Icelandic Vegetation Cover during Historical Patterns of Landscape Degradation. the Holocene: Implications for Long Term Land Catena 42:175–192. Degradation. Geografi ska Annaler 83A:203– 215. Spufford, P. 2006 The Merchant in Medieval Europe. London: Ólasson, Vésteinn Thames and Hudson. 1998 Dialogues with the Viking Age. Reykjavík: Heimskringla, Mál og Menning Academic Sugden, David E. Division. 1982 Arctic and Antarctic. London: Blackwell. Perdikaris, Sophia Sutherland, Patricia D. 1999 From Chiefl y Provisioning to Commercial Fish- 2000 The Norse and Native North Americans. In ery: Long term Economic Change in Arctic Nor- Vikings: The North Atlantic Saga. William W. way. World Archaeology 30(3):388– 402 Fitzhugh and Elisabeth I. Ward, eds. Pp. 238– 47. Washington, D.C.: Smithsonian Institution Perdikaris, Sophia and Thomas H. McGovern Press. In Chieftains, Cod Fish, and Walrus: Intensifi cation press a in the N Atlantic. In New Perspectives on Inten- Tómasson Þórdur sifi cation. T. Thurston, ed. New York: Plenum 1996 Þórsmörk. Land og saga. Reykjavík: Mál og Press. mynd.

WW4336.indb4336.indb 3535 44/20/07/20/07 10:15:3510:15:35 AAMM 36 Arctic Anthropology 44:1

Thórarinsson, Sigurdur Vasey, Daniel. E. 1961 Uppblástur á Íslandi í ljósi öskulagarannsókna 1996 Population Regulation, Ecology, and Political [Wind Erosion in Iceland: A Tephrochrono- Economy in Preindustrial Iceland. American logical Study]. Ársrit Skógræktarfélags Íslands Ethnologist 23(2):366–392. 1961:17–54. Vésteinsson, Orri 1967 The Eruptions of Hekla in Historical Times. The 2000 The Christianization of Iceland: Priests, Power, Eruption of Hekla 1947–1948, vol. 1. Reykjavik: and Social Change 1000 –1300. New York: Ox- Pp. 1–170. Visindefélag Islendiga and the Mu- ford University Press. seum of Natural History. Vésteinsson, Orri, Thomas H. McGovern, and Christian 1975 Katla og annáll Kötlugosa. Árbók Ferðafélags Keller Íslands. 1975:125–149. 2002 Enduring Impacts: Social and Environmental Aspects of Viking Age Settlement in Iceland and 1981 The Application of Tephrochronology in Ice- Greenland, Archaeologica Islandica 2:98136. land. In Tephra Studies. Steve Self and R. S. J. Sparks, eds., Pp. 109–134. Dordrecht: Reidel.

WW4336.indb4336.indb 3636 44/20/07/20/07 10:15:3610:15:36 AAMM