The Impact of Environmental Conditions on Ghassulian Settlement (ca. 4,500-3,900 BC): A GIS- based Spatial Analysis

Thesis submitted in partial fulfillment of the requirements for the degree of “DOCTOR OF PHILOSOPHY”

by

Rona Winter-Livneh

Submitted to the Senate of Ben-Gurion University of the Negev

June 2013

Beer-Sheva

The Impact of Environmental Conditions on Ghassulian Settlement (ca. 4,500-3,900 BC): A GIS based Spatial Analysis

Thesis submitted in partial fulfillment of the requirements for the degree of “DOCTOR OF PHILOSOPHY”

by

Rona Winter-Livneh

Submitted to the Senate of Ben-Gurion University of the Negev

Approved by the advisors Approved by the Dean of the Kreitman School of Advanced Graduate Studies

June 2013

Beer-Sheva

This work was carried out under the supervision of Prof. Isaac Gilead In the Department of Bible, Archaeology and the Ancient Near East The Faculty of Humanities and Social Sciences Prof. Tal Svoray In the Department of Geography and Environmental Development The Faculty of Humanities and Social Sciences

Research-Student's Affidavit when Submitting the Doctoral Thesis for Judgment

I Rona Winter-Livneh whose signature appears below, hereby declare that (Please mark the appropriate statements):

_X_I have written this Thesis by myself, except for the help and guidance offered by my Thesis Advisors.

X The scientific materials included in this Thesis are products of my own research, culled from the period during which I was a research student.

X This Thesis incorporates research materials produced in cooperation with others, excluding the technical help commonly received during experimental work. Therefore, I am attaching another affidavit stating the contributions made by myself and the other participants in this research, which has been approved by them and submitted with their approval.

Date:__ _21/06/13__ Student's name: Rona Winter-Livneh_ Signature: ______

Table of Contents

Acknowledgments 1

Abstract 2

1. Introduction 4

1.1 General review on Ghassulian culture research 4

1.1.1 The Ghassulian culture 4

1.1.2 The research history 5

1.1.3 The research method 7

1.2 General review on the research papers 9

1.2.1 Where to live? – Settlement patterns in the northern Negev 9

1.2.2 All that lives must die – Cemeteries and settlement patterns in the coastal plain and the Shephella 11

1.2.3 My house is [not] your house – architectural symmetry and patterns of structures 15

2. Published papers 18

2.1 Settlement patterns, social complexity and agricultural strategies during the Chalcolithic period in the Northern Negev, Israel 18

2.2 Secondary burial cemeteries, visibility and land tenure: A view from the southern Levant Chalcolithic period 30

2.3 Shape reproducibility and architectural symmetry during the Chalcolithic period in the southern Levant 47

3. Discussion and Conclusions 63

4. References 68

77 5. תקציר

Acknowledgments

Funding for this research was provided by the Pratt Foundation PhD Fellowship Program at Ben- Gurion University of the Negev.

I acknowledge the Geological Survey of Israel for kindly providing the DEM.

This research would not have been possible without the support of many people; First and foremost I thank both my advisors Prof. Isaac Gilead and Prof. Tal Svoray.

Prof. Isaac Gilead’s critical readings and excellent editorial commentary improved this document substantially. I am grateful for his uncompromising standards, professionalism and advice throughout the analysis and writing stages.

Prof. Tal Svoray proved to be an excellent advisor on method and data analysis. Tal introduced me to GIS analysis and encouraged me along the path of GIS archaeology. I am grateful for his faith in my abilities.

A great debt of gratitude is due to Prof. Steve Rosen for his critical reading, generous help, valuable advice and continuous encouragement in all stages of the research.

I would also like to thank Prof. Kenneth Kvamme and Prof. Lynne Goldstein for their help with less accessible literature.

The professional help of the Israel Antiquities Authority deputy director, Dr Uzi Dahari, Dr Peter Fabian, Dr. Edwin van den Brink, Dr. Hagit Torge, and all members of the Israel Antiquities Authority southern regional office is acknowledged.

I acknowledge Paola Ronzino who prepared the first version of the coastal plain cemeteries dataset.

Janet Levy and Karni Golan are thanked for their kind help in previous drafts.

The members of the GI-Lab at Ben-Gurion University of the Negev are thanked for providing useful information concerning GIS procedures.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 1 Abstract

This study explores spatial aspects of the Ghassulian settlements of the Chalcolithic period (ca. 4500-3900 B.C.) in relation to their immediate environments. The objective of the study is to illuminate the social and economic aspects which determined the spatial behavior of these communities. From the methodological aspect, this is a GIS (Geographic Information Systems) based research which includes geo-statistical analyses such as: Moran’s I spatial autocorrelation analysis, Ripley’s K-function, and Kernel density spatial analysis. Other analytical methods include Viewshed analysis, Continuance Symmetry Measure and a new method, Shape Reproducibility, which was developed for the purpose of this research. The conclusions are based on intra- and inter-site spatial variability and they shed new light on a number of issues concerning aspects of Ghassulian social and economic organization.

The results of the research were published in three papers, which deal with settlement distribution related to different scales and across diverse geographical regions. The first paper focuses on the Northern Negev where numerous Chalcolithic period sites have been discovered. Within this region of interest I have examined the distribution of ca. 150 Chalcolithic sites located on the banks of the two main drainage systems of the northern Negev, Nahal Beer-Sheva and Nahal Besor. Analyzing the relations between the sites clusters locations and the physical attributes of the nearby riverbeds which includes their topography, the wadi volume, flow accumulation, the wadi gradient and its riverbanks gradient. The second paper explores Chalcolithic settlement patterns in the coastal plain and Shephella, where many Ghassulian burial sites are located. It analyses the spatial interrelationships between 48 burial sites and habitation sites and their surroundings, and considers the social implications. In this paper I have measured the size of the area from which a burial site is visible and the area from where the neighboring habitation site is visible. I have measured how much these two areas overlap and complement one another. Next I measured how unique or distinguished are these measurements to this specific spatial pattern. The third paper explores the spatial and architectural attributes of Ghassulian structures and their inter- and intra-sites variability. This paper focuses on four large Chalcolithic settlements, with a relatively detailed site plan published by the sites excavators. Using these plans I have examined the symmetry of each structure and its shape reproducibility in relation to the rest of the structures within the same site. Next, I compared the results to the same measurements of structures in the other sites.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 2 This study shows that in the northern Negev the location of sites clusters along the wadis is not random and is determined according to the topographical conditions of the region’s riverbanks, indicating that the community did not require any use of artificial elements such as diversion walls or dams for agricultural prosperity. Within the coastal plain and the Shephella the location of burial sites and their spatial relation with the neighboring settlement indicate cemeteries played a role in claiming land tenure and access rights. The analyses of architectural remains results in similar levels of intra-site heterogeneity between Ghassulian sites which could suggest intensification of economic differentiation, perhaps due to increased reliance on agriculture. All three papers indicate the significance of agricultural activity for the Chalcolithic period communities, the intensity of agricultural production and the growing reliance on it. This could explain the spatial patterning of the northern Negev settlement, the relation between the location of burial sites and their neighboring habitation sites, as well as the level of architectural heterogeneity observed within the main settlements of the Chalcolithic period.

Keywords: settlement patterns; prehistoric architecture; secondary burial; agricultural strategies; Chalcolithic period; Ghassulian culture; southern Levant; spatial analysis; GIS

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 3 1. Introduction

1.1 General review on Ghassulian culture research

1.1.1 The Ghassulian culture

The Chalcolithic period and its main cultural entity, the Ghassulian (ca. 4500-3900 BC after Gilead 2007), follow the establishment of the Neolithic period sedentary agricultural societies and precede the first urban societies of the Early Bronze Age. Thus, the Ghassulian can be regarded as a transition between these two different socio-economic systems. The social and cultural attributes of the Ghassulian communities during this time of transition is one of the primary interests of the current research of the Chalcolithic period (Ben-Tor 1992: 398; Gilead 1988). Changes in settlement patterns, mortuary practices, craft production and symbolic expression characterize this time span. The exact nature of the social and economic organization is, however, debated. Several cultural entities existed in the southern Levant during the Chalcolithic period, of which the Ghassulian culture was the most prominent (Gilead 2011: 13). This culture features assemblages broadly similar to those found at the upper levels of Teleilat Ghassul, the Ghassulian type-site (Gilead 2011; North 1959, 1961). These include specific sets of artifact-type categories such as V-shape bowls, churns, cornets, sickle blades, microliths, basalt bowls and copper and ivory artifacts. Geographically, the Ghassulian is distributed over the Jordan valley, the Shephella and the central hill country, the central and coastal plain, the Dead Sea basin and the northern Negev (Rowan and Golden 2009: 14-20).Ghassulian communities, mainly of sedentary farmers, practiced mixed agriculture; cultivation of cereals, pulses and fruit trees, herding of sheep-goats and raising of pigs and cattle (Gilead 1988; Levy 1995; Rowan and Golden 2009). Craft specialists such as ivory carvers and coppersmiths are worth noting (Gilead 1988; Golden 2009; Levy 1995; Rowan and Golden 2009). The nature of the socio-economic organization of these communities is controversial and there is very little agreement on the meaning of some of this period’s major material remains, such as settlement patterns, burials, and architecture. Scholars who follow Service (1962) suggest a hereditary chiefdom society with centers that coordinate social, economic and religious activities (Gibson and Rowan 2006; Golden 2009; Levy 1986, 1995, 2006). Others find no indication of a hereditary hierarchical society, and argue that the archaeological evidence reflects a lesser level of complexity (Epstein 1998; Gilead 1988, 1993, 1995; Gophna and Tsuk

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 4 2005; Hermon 2008; Rosen 1993). In this study, I address social organization from different perspectives and scales by using GIS-based methodology.

1.1.2 The research history

The earliest studies on the Chalcolithic period, conducted during the interwar years, focused on defining the chronological framework of this period and its cultural characteristics. These early studies consist of intensive archaeological activity, which include large scale excavations at sites such as Teleilat Ghassul (Mallon 1929; Mallon et al. 1934; Neuville and Mallon 1931) and the Nahal Besor sites (MacDonald 1932). In addition, other studies have focused on sorting and classifying artifacts (mostly ceramics) into typological categories and accordingly dividing the chronological frame into sub-phases (e.g., Albright 1931, 1932; Frankfort 1924; Koeppel 1932; Shipton 1939; Wright 1937). Since the early 1950’s renewed excavations were conducted at sites that were previously excavated such as Teleilat Ghassul (Bourke 1997; Bourke 2001; Bourke et al. 2000; Hennessy 1969; Lee 1973; Lovell 2002; North 1961), alongside many new sites such as Perrot’s excavation at Abu Matar and Bir es-Safadi (Perrot 1955, 1984), Horvat Beter (Dothan 1959), cemetery sites in the coastal plain of Israel (Perrot and Ladiray 1980), the Shephella (Van den Brink 2005), En Gedi (Ussishkin 1980), northeastern Sinai (Oren and Gilead 1981) Shiqmim (Levy 1987), Gilat (Levy 2006), Grar (Gilead 1995), and the Golan Heights (Epstein 1998). In addition, many archaeological surveys were conducted within large areas of Israel, Gaza Strip, the West Bank, Sinai and Jordan. These include Nahal Beer-Sheva-Besor survey (Alon and Levy 1980), the Golan Heights survey (Epstein 1998), and the MacDonald survey in Jordan (MacDonald et al. 2001), to name a few. These important excavations and surveys profoundly enlarged the existing database of this period. Over the years, an interdisciplinary approach has prevailed within both the theoretical and the methodological domains of archaeological research. The use of anthropological models for analyzing cultural social and economical aspects as part of theoretical-orientated research has increased. One example is Levy’s research on the Chalcolithic settlement in the northern Negev (Levy 1981, 1987; Levy and Alon 1982). In these works he used the anthropological models developed by Service (1962) and Fried (1967) and their archaeological correlates as listed by Renfrew (1973a). Methodologically, aside of ceramics and flint specialists, other specialists such

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 5 as geologists, physical anthropologists, paleobotanical, and zooarcheaologists were added to the excavation research. As a result, the integration between environmental and archaeological data has become an inseparable part of excavation research analysis in sites such as Shiqmim (Levy 1987), Abu Hamid (Dollfus and Kafafi 1986, 1988), Grar (Gilead 1995), and Teleilat Ghassul (Bourke 2002; Bourke et al. 2000). This interdisciplinary approach has produced new and diverse information – including spatial data - that provide significantly larger and more detailed datasets. These enable a more comprehensive analysis focusing on various and different aspects of the cultural entities of the Chalcolithic period (e.g., Gilead and Goren 1989; Goldberg and Bar-Yosef 1982; Goldberg and Rosen 1987; Goren 1987; Katz et al. 2007; Rosen 1987a; e.g., Rosen and Weiner 1994). Nonetheless, the means through which factors such as local adaptation to the environment, territoriality, visible patterns, or pure chance have been explored based on this carefully recorded spatial information are limited to flat distribution maps (e. g. Epstein 1998; Finkelstein and Gophna 1993; Gophna and Portugali 1988; Yekutieli 2002). Flat distribution maps, which constitute the majority of the 20th century spatial archaeological data, can assist in highlighting and identifying broad spatial directions or trends (Wheatley and Gillings 2002: 6). In many incidences, however, when there is a need to describe or explain processes related to complex spatial phenomena these distribution maps are limited. The analysis or synthesis is restricted to the visual appraisal of these maps, looking for similarities, trends and differences. However in most cases, we are interested in understanding more than just ‘where’ things are. Flat distribution maps display the location of things but provide very little information as to what they are or regarding their temporal status and their relations with other locations. Traditional strategies to overcome these problems are producing multiple distribution maps and/or use of symbols. These solutions, although perhaps plausible with small datasets, however once datasets become larger and more complex, researchers will find themselves dealing with hundreds of different plots or a confusing large collection of symbols. Additional problems arise if there is a need to integrate spatial information outside the artifactual sphere with environmental factors such as types of soils, lithology, and topography or with social factors such as territorial boundaries or the need to best capture the last ray of the midsummer sunset (Wheatley and Gillings 2002: 7- 8). Use of a Geographic information system (GIS) enables to progress beyond the routine capabilities of static distribution mapping and manual overlay. It enables analyzing a large and

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 6 varied volume of information, examining the statistical significance of any relationship between different types of data and identifying any possible patterning within a specific dataset. It enables visual summaries and provides a firm platform upon which more sophisticated exploratory and statistical examination could be performed.

1.1.3 The research method

The present study uses a GIS platform to explore Chalcolithic settlement spatial relations. As GIS is used by a growing number of disciplines (Coppock and Rhind 1991). As a result, defining GIS depends on its highly varied group of users (Maguire 1991: 9). Computer scientists, geographers, geologists, biologists, zoologists, sociologists, economists, and archaeologists (to name a few) have all introduced definitions to GIS. Although all of which are valid, they display the difference according to the specific function which emphasizes each discipline’s requirements (Maguire 1991: 10-11). Nonetheless, in its most general sense GIS is defined as follows, by two of the most popular textbooks:

“…a powerful set of tools for collecting, sorting, retrieving at will, transforming and displaying spatial data from the real world.” (Burrough 1986: 6) “…an information technology which stores analyses, and displays both spatial and non-spatial data.” (Parker 1988: 1547)

The use of GIS in archaeological research started during the 1980’s in North America (Kvamme 1995: 4; 2006). Since the early 1990s there has been a gradual increase in publications describing the use of GIS in European archaeology as well (Richards 1998: 336). During the past decade, an increase in the number of GIS based archaeological studies is evident in the southern Levant. The EAIS project (Egyptian Antiquity Information Services) of the antiquity authority of Egypt was initiated in 2000, developing a large GIS database of all archaeological sites in Egypt (Amin 2002). Similar projects that include the acquisition and management of data from all known archaeological sites and cultural heritage sites have been initiated and conducted in Jordan by Savage (Savage 1990); in Israel and Jordan, within the Arava region by Bienkowski and van der Steen from the University of Manchester (Bienkowski and Van der Steen 2006); in the Negev region Haiman is focusing on ancient water systems (Haiman 2006). Other examples for the integration of GIS within archaeology includes research which focuses on a more specific period

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 7 or smaller areas such as the research of Comer (2000) who examined the relation between the environment and Pre-Pottery Neolithic (PPN) sites in Beida, Jordan using remote sensing technology; Parslow (2006) who examined the social interaction between Natufian site inhabitants in the southern Levant based on lithics artifacts; Mesika (2006) who studied the spatial distribution of artifactual assemblages at Meggido; and Ackermann (Ackermann et al. 2008) who studied Byzantine and Early Islamic agricultural terraces systems in the southern Shephella. Studies implementing GIS for Chalcolithic period research have been introduced only recently. Pierce (2006) reconstructed the population size in Nahal Besor area, in the northern Negev, Israel, based on agricultural production of cereals. Fletcher (2008) used cost-distance, Moran’s I, Getis- Ord and Nearest Neighbor analyses on Chalcolithic period sites to examine their distribution patterns along the wadis at various scales. He also examines different environmental aspects such as location of terrain suitable for grazing, and how vegetation may have been distributed during this period, assuming precipitation amount was twice the current average amount. These studies are important since they integrate different types of data sets, and use GIS for exploring complex and time consuming questions, effectively and quickly. In the present study the use of GIS includes geo-statistical tools such as Moran’s I autocorrelation analysis (Moran 1948) and Ripley’s K-function (Bevan and Conolly 2006). These analyses recognize and quantitatively characterize patterns of spatial dependence between sites. Namely, they reveal the spatial distribution of sites in relation to other sites within the particular landscape surroundings (in the northern Negev, for example). In addition, Kernel density analysis (Silverman 1986) was used to construct an estimation of the relative density of habitation sites within the coastal plain and the Shephella regions. Other methods of analysis include Viewshed analysis (Wheatley 1995a; Wheatley and Gillings 2002) which measures how much can be seen from a burial ground in relation to the area size that is visible from a non-burial location. All analyses were carried out using ESRI ArcGIS 9.3. Symmetry measurement was applied to architectural data from four Ghassulian sites using Continuous Symmetry Measure method (CSM) (Zabrodsky and Avnir 1995), and quantifying the similarly between these architectural shapes was accomplished by developing a new method - Shape Reproducibility (SR). These methods, applied on Chalcolithic data, contribute to current research by producing objective significant results concerning the spatial patterns of the Ghassulian society. The results, in some cases, validate previous assumptions which are a subject of debate in the current research of the Chalcolithic

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 8 period, and in other cases produce new information concerning Ghassulian spatial behavior. Generally speaking, the results of this study provide additional understanding concerning Ghassulian social and economic aspects. In the following sections each of the research papers is reviewed shortly according to their chronological order of publishing, highlighting each paper’s main points, its contribution to current research, and its relation with the remaining papers.

1.2 General review on the research papers

1.2.1 Where to live? – Settlement patterns in the northern Negev

The first paper, titled: “Settlement patterns, social complexity and agricultural strategies during the Chalcolithic period in the Northern Negev, Israel” (chapter 2.1) explores the spatial patterns of the large number of northern Negev Chalcolithic settlements. According to the Israel Antiquity Authority GIS MENORA database, which has never been used before for such an analysis, of the 974 Chalcolithic sites in Israel, 353 sites are located in the Northern Negev; covering an area of 1252 km2. This phenomenon is extraordinary. Until the fifth millennium this region was poorly or sparsely populated. Only during the Byzantine period, thousands of years later, was the Northern Negev settled intensively again (Haiman 1995; Rubin 1990; Shershevski 1991). The distribution of sites within the northern Negev clearly indicates that Chalcolithic settlements were clustered on the banks along the main drainage systems of the northern Negev. The profusion of Chalcolithic settlements in the northern Negev is surprising considering the semiarid climate of this region. Alon and Levy (1980) suggest that dams and diversion walls along the banks indicate that irrigation systems were constructed. Accordingly it is argued that the organization of labor needed for such water management systems is indicative of a chiefdom society (Levy, 1986: 103). However, the dating of these dams has never been established and it is more plausible that they are from the Byzantine period (Ackermann et al., 2008: 932; Gilead, 1988: 421). The interactions between local communities and their most important environmental resource – the wadis – have been overlooked in most cases and only few have studied this phenomenon thoroughly. These include phytolith studies of Rosen and Weiner (1994) on cereals from sites of Shiqmim and Gilat, and the study of Katz et al. (2007) on the phytoliths from the site of Grar. Spatial studies on the relation between the settlement pattern and the environment include Levy’s Ph.D. research (1981), and later publications in which he reconstructs a two-tiered hierarchical

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 9 chiefdom society based on spatial analysis of the settlement distribution within the northern Negev. By using cluster analysis he identifies four site clusters in the area of Nahal Beer-Sheva – Nahal Besor, and divides the northern Negev into ‘territories’ of chiefdoms based on constructing weighted Thiessen polygons around what he considered settlement centers (Levy 1981: 267-275; 1986: 99-100; 2006). An additional study which readdresses the spatial–social issue of the northern Negev settlement patterns was published by Fletcher (2008). He uses GIS-based analysis tools such as Average Nearest Neighbor Distance, Getis-Ord General G, and Moran’s I. The results of his analyses contradict Levy’s results, and “indicate that settlement patterns were random.’’ (Fletcher 2008: 2053). He thus concludes that ‘‘it is extremely difficult to support a dynamic regional chiefdom-level society’’ (Fletcher 2008: 2056). Herein, I readdress Fletcher and Levy’s spatial-social controversy. To this end, I first reexamine the raw data published in these papers. Levy based his analysis on 75 sites he and David Alon found. However since 1981 many more sites have been uncovered. Fletcher studied 406 Chalcolithic sites in 2008. A close inspection of the table of sites that he used (Fletcher 2008: 2049) indicates that 53 sites out of the 403 were in fact parts of other sites already counted. This means that only 353 of the list items are actually sites, and they are the basis of our analyses. This reexamination of the list of sites was done by using Israel Antiquity Authority GIS MENORA database. Apparently, dozens of sites are registered twice or even three times. A reexamination of the sampling method and the analytical procedures published in Fletcher’s analyses was conducted using Moran’s I analysis and an additional method named Ripley’s K- function, on the updated data. These analyses allow to quantitatively characterize the patterns of spatial dependence between sites (clustered or random). By repeating the same sampling conditions of grid shape as ‘blocks’ as described in Fletcher’s autocorrelation analyses (Fletcher 2008: 2052-2053), I demonstrate the weakness of using this type of sampling method for the problem at hand (see chapter 2.1 p. 289). Instead, I propose an alternative and more appropriate sampling method; delineation of buffers along the wadi slopes and beds. This method is based on the natural curves of the wadi, a real phenomenon in contrast to sampling grid blocks that are arbitrary in both their location and size. This sampling method produces a constant average distance of each buffer sub-unit from the wadi axis. As a result, a negligible (near zero) variance of distance within the buffers is attained. Thus, the effects of ‘‘proximity to resources/environment’’ (Fletcher, 2008: 2053) on site distribution are neutralized in this sampling

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 10 method. This reexamination has produced significant results, indicating the settlements are clustered within the wadis local environment. Next, by using General Linear Model (GLM) I was able to analyze the relation between the apparent site clustering within the wadis and the environmental characteristics of the wadis. The wadi’s environmental variables include: wadi volume, wadi gradient, wadi depth index (which characterizes the area of the wadi banks), flow accumulation and two additional integration variables, all of which were produced and calculated using the Digital Elevation Model (DEM) layer. This GLM analysis produced significant results which indicate a subtle reciprocal relationship between the settlement pattern and the environment. Thus, Chalcolithic site location is affected by the nature of the riverbanks. There is preference to settle near areas where the riverbank is more shallow and leveled. Areas which are relatively narrow and steep were settled only if the flow accumulation within these areas is greater than the amount of flow accumulation in areas where the riverbank is shallow and leveled.

1.2.2 All that lives must die – Cemeteries and settlement patterns in the coastal plain and the Shephella

While many sites have been found within the northern Negev, burial sites seem to flourish within a different, separate region, - the coastal plain and the Shephella, covering an area of 2484 km2. Interpretation of the distribution of Ghassulian burial sites has been problematic in the past, primarily due to what seemed to be a lack of habitation sites in the central coastal plain. It was therefore suggested that the burial sites there were used by nomadic pastoralists (Perrot 1984; Perrot and Ladiray 1980); however the discovery of cemeteries at Shiqmim have invalidated this suggestion. Levy and Alon (1987: 348) suggest, on the basis of ethnographic data, that secondary burials characterize sedentary rather than semi-nomadic societies. During recent years many habitation sites have been discovered in these regions (Gophna and Portugali 1988; Van den Brink 2008; Van den Brink and Gophna 2005). Thus, it is currently agreed that the burial caves were used by sedentary communities living nearby (Khalaily and Marder 2010: 19-20; Porat et al. 2006: 58-59; Van den Brink 2011: 45-46; Van den Brink and Gophna 2005: 168-169; Yannai and Porath 2006: 1-2). The second paper published as part of the present study, titled: “secondary burial cemeteries, visibility and land tenure: a view from the southern Levant Chalcolithic period”,

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 11 explores the spatial relation between Chalcolithic period settlement and burial sites within the coastal plain and Shephella region. In the first stage of this study I used the IAA MENORA database to select all Chalcolithic sites within the study area. Next I used Kernel density spatial analysis to map the dense areas of burial sites and compare them to the dense area locations of all non-burial sites within the study area. This required using vector grids as the sampling tool. The grids that cover the research area are equal in their overall size but differ in cell size. To avoid any possibility that the correlation results are biased by the sampling grid cell size, I resampled the research area by using three cell sizes. Then, I calculated the zonal statistics on the kernel density raster layer within each cell of each grid. Generating vector grids and calculating zonal statistics were both performed by Hawth’s analysis tools version 3.26 (Beyer 2004). A correlation test was conducted on the zonal statistics results from all the sampling grids between the burial site kernel density layers and the non-burial site kernel density layers. However, since most areas of the kernel density maps are empty of sites, that is, equal zero, the correlation used grid cells with a zonal statistical average (of the non-burial site kernel density layer or the burial site kernel density) >0. This indicated that the location of burial sites tends to be far away from areas where non-burial sites are located. Secondary burials placed within an off-site location are an innovation of the Ghassulian society. Secondary burials during the Chalcolithic period usually involve ossuaries. In such cases the body is first interred in habitation sites in places such as pits, silos and below floors walls (Gilead 1988: 428), and later on, after a culturally determined length of time, the remains are retrieved and re- interred at a new location (Schroeder 2001: 79). In this respect, secondary burial, which usually involves a primary burial, is perceived here as an exclusive class or type of mortuary practice. During preceding periods most burials were primary ones, and most of them (either primary or secondary) were located within the habitation or household surroundings (e.g. Kuijt 1996). Ghassulian secondary burials are known only from cemetery sites, mostly from burial caves. The caves can be found either in kurkar ridges or limestone hills and most of them are located within the coastal plain and Shephella region (Van den Brink 1998, 2005). The contribution of the spatial aspect to the study of mortuary practices has been recognized and demonstrated by many researchers (e.g. Bradley 1998; Cannon 2002; Clarke 1977; Goldstein 2002; Hutchinson and Aragon 2002; Renfrew 1976; Williams 1999). Despite the growing number of Ghassulian secondary burial sites, the spatial aspect of these burials has received relatively little

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 12 attention (but see Joffe, 2003). Moreover, little attention has been paid to the role these burial sites play in structuring and organizing past landscape, a role that most probably goes far beyond their use as funerary sites. Previous studies propose that the formal cemeteries of the Chalcolithic period located at sites distinct from habitation sites reflect a concern for territoriality (Levy, 1986, 1995). In addition, Joffe (2003) suggests that the visibility of cemeteries, like Shiqmim in the northern Negev and Adeimeh near Teleilat Ghassul in Jordan, is a statement of territorial claim while the majority of cave burials in the coastal plain area that ‘‘stand out for their invisibility on the living landscape’’ reflect a less public praxis (Joffe, 2003: 51). Golden (2009) suggests that the burial caves served Ghassulian elites who used the remote and concealed locations to protect the wealth deposited in the tombs. It should be noted, however, that none of these propositions has been examined yet. The objective of this study is to examine the relationship between the non-burial sites and the burial caves which may reveal how and why their specific locations were determined and the way this is related to the socio-economic aspects of this period. After establishing the burial/non-burial distribution patterns, I reviewed ethnographic literature where off-site secondary burial is practiced to find its social and economic implications. The ethnographic data demonstrate that locations of burial grounds and cemeteries play an important role in establishing or preserving land tenure (Bloch 1989; Bradley 1991; Earle 1991; Fleming 1973; Goldstein 1981; Renfrew 1973b; Saxe 1970). This enables to formulate the hypothesis that Ghassulian secondary burials were used in a manner similar to those observed in the ethnographic record, as landscape markers in a land tenure system. To examine this, a simple set of assumption and constraints were used. These assumptions are based on social science studies and archaeological landscape research. According to these studies, visibility is considered within many cultures as one of the means to establish territorial rights (Brighenti 2007; Eason and Stamps 1992; Foucault 1995 (1975); Llobera 2003, 2007; Renfrew 1976; Wheatley 1995b; Yekutieli 2006). Therefore, I regard areas which are visible from habitation sites as most likely to have been included as parts that are controlled by the sites. Furthermore, terrain which is visible from a burial site can also be regarded as a controlled area. This assumption does not suggest that the burial sites were intended to be used as actual watch towers of any kind. It is those areas, within the landscape, from which the burial sites locations are visible, that are of particular interest. What influences the society is the likelihood of recognizing and remembering the burial site as a landscape feature, a physical marker in the landscape.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 13 To test this hypothesis I used GIS viewshed analysis tools. This was done to determine whether the areas visible from burial sites were organized so that they could supplement or enhance the visible area seen from habitation site (contributing viewsheds) and thus be established together (the burial location and the habitation location) a continuous visible landscape within the habitation site surroundings. The study group included a sample of 24 burial sites and 24 of their nearest neighboring habitation sites. The analysis consisted of mapping and measuring viewsheds, the areas within the landscape which are visible from each of the burial sites, and comparing them to the areas within the landscape which are visible from its (the burial site’s) nearest neighboring habitation site. Next, I examined how much of the viewshed of each pair of sites (the burial and the nearest neighboring habitation site) overlap and what is the contribution of the burial site to the broadening of the area viewed from the habitation site. To evaluate the significance of the results based on examining the study group, I compared them to the viewshed measurements obtained from a matched control group, that is, from pairs of non-burial sites that are located within similar geographical conditions. Examining the percentage of the contributing area visible from the burial sites shows that it is almost double the size of the visible area (observing area) surrounding the habitation site, in comparison to the matching control group. Thus, this study indicates that burial caves were located in areas which are visible from (observed by) a significantly larger part of the landscape than their nearest settlements, and this large area completes the visibility of the terrain within the habitation near surroundings. Meaning, more area around the habitation site is under visible control. Based on the ethnographical record these results could suggest that Chalcolithic cemeteries played a role in claiming land tenure. The Chalcolithic community practiced secondary burial which maintained their social identity and group cohesion alongside asserting and defining rights over land use by means of lineal ties to ancestors. This was marked by the location of a nearby hill where burial caves were quarried.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 14 1.2.3 My house is [not] your house – architectural symmetry and patterns of structures

While the first two published papers examined the spatial behavior in large areas, of 2484 km2 and 1252 km2, next I chose to tackle the subject using a different scale of analysis, of the spatial patterns that are present within the sites, covering areas between 1350 m2 and 5500 m2. This provides an additional complementary perspective of the spatial behavior of the Ghassulian communities. Thus, the third paper: “Shape reproducibility and architectural symmetry during the Chalcolithic period in the southern Levant” confronts Chalcolithic settlement data of a small scale. It explores the spatial patterns of architectural shape within four important Ghassulian sites. Architecture is a visible cultural manifestation that derives from social behavior, influences it and provides the framework for social interaction and community organization (Byrd 1994: 643; Ingold 2000: 175-178; Wilson 1988: 21). Different attributes of the dwelling structure such as shape, size, building material and decoration have significance beyond their immediate function. It is also evident that dwellings are subject to spatio-temporal changes (e.g. Flannery 1972; Goring- Morris and Belfer-Cohen 2008; Kempinski and Reich 1992). The Chalcolithic architectural unit is usually composed of a single rectangular room often termed “broad-room” or “broad house” (e.g. Gilead 1988:416; Porath 1992:41; Rowan and Golden 2009: 29). The Chalcolithic architecture is for some considerably varied and indicates “the presence of central authority” (Levy 1986: 88), and the “the development of elite residential complexes” (Bourke 2002: 22). For others, it is not indicative of “structural hierarchy” (Gilead 1988: 418) or suggests an egalitarian community (Epstein 1998:7). To examine the above mentioned opinions, I tried to accurately quantify architecture in order to characterize and compare different sites and structures. Reviewing recent literature concerning traditional architecture shows that social, cultural and economic factors determine a house’s shape (Allison 2002; Carsten and Hugh-Jones 1995; Donley 1982; Hillier and Hanson 1984; Ingold 1995, 2000; Kent 1990; King 1980; Lau 2010; Rapoport 1969, 1982; Wilson 1988). This suggests that by measuring and comparing architectural shapes within settlements, it is possible to illuminate social, cultural and economic aspects of Ghassulian communities. Beside shape of structures, other attributes such as symmetry can be measured to help in evaluating cultural and social aspects. Symmetry is a key element in architecture which signals

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 15 balance, since pressure on a structure or building is distributed equally if there is symmetry. Studies of flint tool symmetry – such as Acheulian handaxes – show that the difference between early, less symmetrical artifacts and later, more symmetrical artifacts, is well established and regarded as an indication of more elaborated production techniques and increased skill (Saragusti et al. 2005; Saragusti et al. 1998; Wynn 1985). Moreover, studies have shown that manufacturing techniques involve social dynamics, and the technical knowledge is directly related to social knowledge. A progress in technology is driven by cultural accumulation of knowledge. This could suggest that difference between less symmetrical prehistoric structures and more symmetrical structures might reveal aspects concerning the technology and skills which characterize the societies and their cultural contexts (e.g. Bridgeman 2002; Ingold 1990; Schiffer and Skibo 1987). To explore the structures and their social implications, I applied two methods on Ghassulian site architectural ground plan data. I used Continuous Symmetry Measure (CSM), a method which was originally developed for analyzing flint artifacts and ceramic vessels to objectively compare between shape symmetries of houses (Saragusti et al. 2005; Saragusti et al. 1998). I also developed a new method, Shape Reproducibility (SR), which enables to compare between the shapes of these ancient buildings. The study consists of four Chalcolithic sites (three Ghassulian and one Golanian for comparison), and an additional comparative control site from the Early Bronze Age (Arad). By using CSM the symmetrical level of each structure was measured in order to find and compare between each of the sites’ average symmetry level. By using SR the relative similarity between each structure and each one of the remainder of the structures in the studied sample produced a similarity matrix table from which the average similarity within and between sites of any particular group of structures was calculated. Symmetry analysis results show that changes in symmetry levels of prehistoric architecture over time do not support the idea that accumulation of knowledge over time results in technological progress. On the one hand, it is plausible that the symmetry of prehistoric structures fails to reflect social changes. On the other, it may, at least in some cases, indicate that architectural techniques and traditions are diverse and do not develop unilinearily. The SR analysis results show that the Golan site is characterized by an intra-site structural homogeneity. The shape of structures here was deliberately or undeliberately restricted to one shape and the result is a particular settlement pattern. Thus, this site lacks any indication of socio-

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 16 economic differentiation. Partial heterogeneity was revealed in the Ghassulian sites (Shiqmim, Teleilat Ghassul, and Abu Hamid), suggesting a certain level of differentiation between the dwellings within each of the settlements. This low-level heterogeneity may signify variability in a number of socio-economic aspects, such as intensity of production, craft specialization and technological innovation. Different activities or different production intensities probably resulted in different storage requirements for different structures, therefore the structures were of different shapes. A relatively high level of heterogeneity was observed among the Early Bronze Age structures at the Arad site. This suggests a significant visible dissimilarity which is a result of structures having different functions. Public structures (such as the temple complex) and dwellings were also designed differently to express socio-economic distinction.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 17 2. Published papers

2.1 Settlement patterns, social complexity and agricultural strategies during the Chalcolithic period in the Northern Negev, Israel

This study was published in Journal of Archaeological Science 37: 284-294 (2010). A copy of the article follows.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 18 Journal of Archaeological Science 37 (2010) 284–294

Contents lists available at ScienceDirect

Journal of Archaeological Science

journal homepage: http://www.elsevier.com/locate/jas

Settlement patterns, social complexity and agricultural strategies during the Chalcolithic period in the Northern Negev, Israel

Rona Winter-Livneh a,*, Tal Svoray b, Isaac Gilead a a Archaeological Division, Ben-Gurion University of the Negev, Beer-Sheva 84105, Israel b Department of Geography and Environmental Development, Ben-Gurion University of the Negev, Beer-Sheva 84105, Israel article info abstract

Article history: Previous studies have reported a Chalcolithic site distribution pattern in the Northern Negev as clustered, Received 26 January 2009 exhibiting a number of characteristics that imply a chiefdom level of organization. However, a recent Received in revised form spatial analysis suggests that in some areas of the Northern Negev, settlement patterns were essentially 12 September 2009 random and that there is no evidence for regional chiefdom organization. We examine this controversy by Accepted 21 September 2009 closely inspecting the methods of spatial analyses employed by previous researchers, by introducing an additional multiscalar spatial technique, namely Ripley’s K-function, and by using updated and modified Keywords: data. Our results indicate that settlement distributions were essentially clustered, even in small areas along Settlement patterns Agricultural strategies the wadis. Examining the relation between these spatial-dependent distributions and the landscape Chiefdom society surroundings revealed that particular physiographic characteristics of the wadis contribute to increscent in GIS site clusters. Furthermore, a general linear model analysis suggests that the distribution of Chalcolithic Spatial autocorrelation sites is determined primarily by environmental factors rather than factors related to political organization. General linear model Ó 2009 Elsevier Ltd. All rights reserved. Chalcolithic period Southern Levant

1. Introduction patterns, overlooked are the interactions between societies and their most apparent environmental resource, i.e., the wadis. The Northern Negev is the part of the Southern Levant with Presently, there is a noticeable disagreement among scholars as probably the largest number of sites from the Chalcolithic period to the nature of social organization and subsistence economy of fifth (ca. 4500–3900 BC after Gilead, 2007). This intensive settlement is millennium Chalcolithic societies. Several scholars suggest that the unique in this part of the country and only during the Byzantine data reflect a chiefdom society with centers that coordinate social, period, thousands of years later, was the Northern Negev desert economic and religious activities (Gibson and Rowan, 2006; Levy, settled intensively again. One of the most apparent phenomena of 1981, 1986, 1995; Levy and Alon, 1983; Levy et al., 2006). The the distribution of Chalcolithic sites in the Northern Negev is the chiefdom society is characterized by ‘‘. clearly defined territorial fact that many of them are located along the banks of the main boundaries or borders’’ (Levy, 1986: 87), as well as other attributes wadis (riverbeds, dry for most of the year) (Fletcher and Winter, such as craft specialization, population density, religious, public 2008; Gilead, 1988; Levy, 1986). Although the concentration along labor etc., that characterize chiefdom in the archaeological record wadis is acknowledged in many studies, it is often considered as (Renfrew, 1973). However, other scholars suggest that Chalcolithic a static attribute, rather than a significantly dynamic feature of the sites represent different entities, either of several time spans or of archaeological data (e.g., Alon, 1961; Alon and Levy, 1980; Gazit, different cultural traditions, which may be remnants of rural 1996; Gilead and Goren, 1986; Gophna, 1990; Levy and Alon, 1983). communities with a low degree of social complexity (Gilead, 1988, As a result, while Chalcolithic period sites and artifact assemblages 1993, 1995, 2007; Hermon, 2008; Rosen, 1993). This is based on the of the Northern Negev serve as firm databases for exploring aspects observation that the variability of the archaeological record is not such as subsistence economy, social organization, and settlement pronounced enough to support the chiefdom model. It seems that the Chalcolithic archaeology of the Northern Negev fits better the type of social organization Maisels (2001: 156–168) calls ‘‘Stratified and Augmented Household’’. Alon and Levy (1980) suggest that dams and diversion walls * Corresponding author at: Archaeological Division, Ben-Gurion University of the Negev, PO Box 653, Beer-Sheva 84105, Israel. Tel.: þ972 8 9491932. along the banks of Nahal (i.e., wadi) Beer-Sheva indicate that irri- E-mail address: [email protected] (R. Winter-Livneh). gation systems were constructed. The organization of labor needed

0305-4403/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.jas.2009.09.039 R. Winter-Livneh et al. / Journal of Archaeological Science 37 (2010) 284–294 285 for such water management systems is indicative of a chiefdom society (Levy, 1986: 103). However, these water systems have never been studied in the detailed manner they deserve, and their date has never been established. Some even claim, on comparative grounds, that it is more plausible to date the dams to the Byzantines period (Ackermann et al., 2008: 932; Gilead, 1988: 421). It is well known that the Byzantine settlement is characterized by profusion of agricultural installations including dams and diversion walls in the wadis of the Negev (Evenari et al., 1971). Rosen and Weiner (1994), in their phytolith study, show that cereals from sites of Shiqmim and Gilat were grown under conditions of high water availability, accomplished by placing the fields in floodplains. Katz et al. (2007), however, concluded, based on phytolith samples from the site of Grar, that crops had been dry farmed, implying that features of the landscape other than flood plains were used as fields during the Chalcolithic in the Southern Levant. Levy (1981) reconstructs a two-tiered hierarchical chiefdom society by using spatial analyses such as nearest neighbor analysis, cluster analysis, and Thiessen polygons to study the settlement Fig. 1. Study area; distribution of Chalcolithic sites in the Northern Negev. distribution patterns. By using cluster analysis he identifies four site clusters in the area of Nahal Beer-Sheva – Nahal Besor, and divides the Northern Negev into ‘territories’ of chiefdoms based on con- patterns of sites and clarify the attributes of wadis that most influ- structing weighted Thiessen polygons around what he considered enced these patterns. settlements centers (Levy 1981: 267–275; 1986: 99–100; Levy et al., 2006). In a more recent attempt to characterize the intensive Chal- 2. Study area and dataset colithic settlement of the Northern Negev, Fletcher (2008) re- addresses the spatial–social issue by using Average Nearest Neighbor The study area is delineated by a convex hull drawn around 353 Distance, Getis-Ord General G, and Moran’s I analyses. The results, he Chalcolithic sites, covering an area of 1252 km2 (Fig. 1). The area is argues, indicate that ‘‘autocorrelation indices for areas smaller than a depression both topographically and tectonically. It rises nearly the entire Northern Negev indicate that settlement patterns were imperceptibly from 150 m a.m.s.l. near the Mediterranean coastal random.’’ (Fletcher 2008: 2053). He thus concludes that ‘‘it is ridges in the north-west, to 400 m a.m.s.l. in the inland foothill zone extremely difficult to support a dynamic regional chiefdom-level at its eastern extremity. The area includes a number of broad and society’’ (Fletcher 2008: 2056). deep main wadis. The western part is dominated by Nahal Besor The studies reviewed above highlight the difficulty in exploring (Wadi Gaza) and the eastern part is dominated by Nahal Beer-Sheva. and interpreting two of the most conspicuous, though little Reworked Pleistocene aeolian loess and sand deposits cover about explored, aspects of Chalcolithic settlement: (i) the preferred 4/5 of the area. The major drainages of the Besor and Beer-Sheva location of sites; and (ii) the relation of sites to the adjacent wadis. catchment basins include the Central Negev Highlands, the These research gaps could be better understood through tackling Northern Negev Upland and the southern part of the Judean Desert the following unanswered questions: What were the limiting (Levy and Goldberg, 1987; Nir, 1970; Singer, 2007). Climatic condi- factors that influenced the spatial distribution of sites? What area tions in the Northern Negev are semi-arid. Rain falls during winter surrounding the wadi did the Chalcolithic settlers prefer? How did months only, with the three months December, January, and Chalcolithic societies use their landscape? February accounting for two thirds of the annual rainfall. Annual In the current study we explore the environmental matrix and its average rainfall varies between 200 and 300 mm, depending on potential in terms of subsistence, and how it influences the distri- specific location (Shmueli and Gradus,1979). The vegetation, as part bution of sites. Also explored are latent questions such as: How did of a Mediterranean to Irano-Turanian transition zone, is inter- people use particular landscapes? How did they adapt to limiting spersed with patches of shrubs dominated by Sarcopoterium spi- factors of the environment? And what were the social interactions nosum and other herbaceous understory species, mainly Thymelaea of human groups in this specific landscape? The framework of our hirsute, Noaea mucronata and Atractylis serratuloides (Danin et al., analytical approach originates in Butzerian contextual archaeology 1975; Wright et al., 2006). The environmental, geophysical data, of (Butzer 1982: 12) rather than in anthropological archaeology. Thus, the study area was measured and quantified on the basis of the distribution pattern is not tested against an anthropological or a contour-based DEM layer with 25 m vertical and 10 m horizontal social model (e.g., Renfrew’s chiefdom model), but rather against resolutions (Hall, 1993). the attributes of the immediate environment. We use two tech- Since Levy’s (1981) pioneering spatial study, which covered 75 niques of spatial analysis to characterize quantitatively the spatial Chalcolithic sites, hundreds of new archaeological sites were patterning of the sites: (i) we reexamine Moran’s I procedures that discovered and added to the Israel Antiquities Authority (IAA) were used by Fletcher (2008) in light of a modified and updated database. The IAA dataset table that was used in the most recent Northern Negev database and introduce an alternative methodo- research on Chalcolithic sites in the Northern Negev (Fletcher, logical procedure; and (ii) an additional method, Ripley’s K-function 2008), included 406 sites. However, recent examination and data (Ripley, 1976), which is designed to analyze the distribution of quality assessment of the IAA database with the help of the points in a multiscalar approach and is used here to examine MENORA system (see below) revealed that a large number of sites whether site distribution patterns can be identified in small-scale are registered twice or even three times according to the site name analysis. Following the quantification of site spatial patterns, the rubric, while, according to the site identity number, no duplicates relationships between these patterns and the environmental attri- are detected. These duplicates are archaeological site extensions, butes of wadis is statistically explored. These allow us to elucidate which are part of the legal site designation process of the IAA and how landscape surroundings contribute to distance-defined can accidentally lead to a wrong impression concerning the amount 286 R. Winter-Livneh et al. / Journal of Archaeological Science 37 (2010) 284–294 of sites in the Northern Negev area. The site extension information is listed within each site file in the IAA MENORA database system and this database is installed in the IAA server and is maintained and updated constantly by the GIS Department of the IAA. There- fore, the MENORA database is currently the most reliable source of computerized information for archaeological sites in Israel (Uzi Dahari, personal communication), and was not used by previous authors. To avoid the double listing, every site within the Northern Negev area was validated by us against the MENORA database. A close inspection of the dataset table that was used by Fletcher (2008) has indicated that 53 sites out of the 406 were in fact site extensions rather than defined archaeological sites. That means that only 353 of the list items are actually sites, and they are the basis of our analyses.

3. Methods

3.1. Spatial analyses

We applied Moran’s I autocorrelation analysis and Ripley’s K- Fig. 2. Large-scale analysis of Northern Negev sites. Each cell is 500 500 m. function to the database of 353 Chalcolithic sites. These analyses are intended to recognize and quantitatively characterize patterns large blocks was 150–200 km2, medium blocks were approximately of spatial dependence between the sites. In other words, they aim 80 km2, small blocks ranged from 15 to 20 km2, while the smallest to reveal the spatial distribution of the sites in relation to other sites blocks were ca. 4 km2. These intervals are all part of the reanalysis within the particular landscape surroundings. The analyses also procedure and were required according to Fletcher’s rationale ‘‘in allow investigation of the contradicting results (clustered vs order to try to overcome the natural clustering that occurs along the random) presented in previous studies by repeating the same wadi courses.’’ (Fletcher, 2008: 2053). conditions as described in Fletcher’s autocorrelation analyses Along with the reexamination of Fletcher’s data sampling and (Fletcher 2008: 2052–2053). organization, we examined the data by using an alternative Moran’s I spatial autocorrelation analysis (Moran, 1948) sampling method for Moran’s I autocorrelation multiscalar analysis. measures the covariation of juxtaposed map values of a given This approach includes using natural sub-units, rather than ‘blocks’ variable (Griffith, 1987). In other words, it measures the observed of grid shape. For this purpose, we used buffer areas of 500 m radius similarity based on a simultaneous measurement of both observed around each of the major wadi axes (Nahal Beer-Sheva, Nahal Besor, locations and observed values. The index value expresses the Nahal Grar and Nahal Sekher). Each buffer area was divided into spatial relationship between location and value; it is large when sub-units (grid cells/polygons) using editorial operators (Fig. 4). The pairs of points are near each other, and decreases as distance average area of a sub-unit cell was 1 km2. The largest sampling area between points grows (i.e., inverse distance). After Wheatley and included 149 Chalcolithic sites, within a total area of 116.7 km2 (of Gillings (2002); Eq. (6.6), Moran’s I is defined in Eq. (1): all the major wadis). Smaller-scale sampling areas included Nahal P P 2 n n Besor (sampling area of 31.3 km ); Nahal Beer-Sheva (sampling area n i 1 j 1 Wijðxi xÞ xj x 2 2 I ¼ P ¼ P ¼ P (1) of 49 km ); Nahal Grar (sampling area of 22.1 km ); Nahal Sekher n n n 2 2 i¼1 j¼1 Wij i¼1ðxi xÞ (sampling area of 14.1 km ). Dividing the wadi into equal-sized sub- units neutralizes the effect of wadi proximity by generating zero where n is the number of instances, i ¼ 1, ., n, each of which is associated with a known coordinate,P andP is divided by the aggre- n n gate of all the spatial weights ( i¼1 j¼1);x is the mean of all attributes and Wi;j is the spatial weight matrix between each pair of neighbors (instances i and j). Moran’s index varies between values near þ1.0 (clustered pattern) and 1.0 (dispersed pattern). The statistical significance (Z-score value) determines whether the observed pattern is just one of many possible versions of randomness or whether it indicates a significant finding of positive or negative autocorrelation (Griffith, 1987; Kvamme, 1990a,b; Mitchell, 2005; Wheatley and Gillings, 2002). Reexamining Fletcher’s (2008) Moran’s I autocorrelation anal- yses required data organization and sampling in two sizes of blocks grid cells: 250 250 m and 500 500 m The grids were produced by using Hawth’s Analysis Tools free software operators (Beyer, 2004). In each of these two sampling block cell areas, the number of sites was counted to generate the observed value, and grid cell centroids were used as the observed value location for autocorre- lation. The largest sampling block area was of 1252 km and includes all 353 sites (Fig. 2). Various scales of analysis were tested for multiscalar patterns examination, in the form of block grid sampling in Nahal Besor, Nahal Grar and Nahal Beer-Sheva (Fig. 3). Fig. 3. Various scales of analyses using the blocks sampling method (after Fletcher In this multiscalar analysis, four scales were analyzed: the area of 2008: 2052, Fig. 5). R. Winter-Livneh et al. / Journal of Archaeological Science 37 (2010) 284–294 287

Ripley’s K-function method was first introduced to archaeolo- gists by Bevan and Conolly (2006). It calculates the average cumu- lative frequency of points at a given radius over a range of distances (Bevan and Conolly, 2006: 221; Schwarz and Mount, 2006: 179). The observed value is calculated for different distances as in Eq. (2): sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiP P n n A i 1j 1 js1 kði; jÞ LðrÞ¼ ¼ ¼ (2) pnðn 1Þ

where r is the distance, n is the total number of features, A represents the total area of the features, and Ki,j is a weight. Ki,j is modified according to the given edge correction method. If there is no edge correction, then the weight is equal to 1 when the distance between i

Fig. 4. Sub-units (i.e., grid cells) along the wadis, based on a buffer of 500 m in length. variance of distance within the buffers sub-units. Furthermore, apart from testing the effect of different sampling grid shapes (i.e., blocks/curved buffers) on the autocorrelation results, a test on the effect of block sampling location was required for result verifi- cation. Thus, one sampling block (‘Besor – small’), was tested in the exact location that was addressed by Fletcher (2008), and 250 m west of the original location (Fig. 5). The Moran’s I tests were applied by using inverse distance without limiting any distance band.

Fig. 5. ‘Besor – Small’ area grids (250 250 m) on two different locations: the blue grid location is according to Fletcher’s study, the yellow grid was placed 250 m west of the blue grid’s location. Fig. 6. Nahal Besor (a) and Nahal Beer-Sheva (b) sub-areas, based on the wadi axis. 288 R. Winter-Livneh et al. / Journal of Archaeological Science 37 (2010) 284–294

sub-area was tested by Ripley’s K-function while using several different distance increments (neighborhood radius size) and several different numbers of distance bands (the number of times to incre- ment the neighborhood size). Ripley’s Edge Correction Formula was used as the boundary correction method for all tests. All spatial analyses were carried out using ArcGIS 9.2.

3.2. General linear model

To test if the environmental characteristics of the wadis signifi- cantly affected the variation in the quantities of sites, correlation analyses and multiple regression models were applied. Six envi- ronmental variables, extracted from the DEM, were measured at each sub-unit sampling area (1 km2) at the Nahal Beer-Sheva (n ¼ 41) and Nahal Besor (n ¼ 32) 500 m polygonal buffer layers (Fig. 4). Wadi volume was calculated using 3D analyst tool of ArcGIS Fig. 7. 3D illustration of Wadi volume surface area. The graphic striped area represents 9.2 from the DEM. The volume is the space between the terrain w70% of the total sampling sub-unit surface area, which is used in calculating the Wadi volume variable. surface and a horizontal reference plane set at a particular height above the surface. In order to prevent extreme values of individual elevation points in the terrain surface of each particular sub-unit and j is less than r, otherwise it equals zero (Fortin,1999; Gatrell et al., from influencing the volume calculation, an average of 70% of the 1996; Mitchell, 2005). The observed average frequency of each given total sub-unit surface area was measured (Fig. 7). This required radius is compared with the average cumulative frequency of the a reference plane set at 10 m in Nahal Beer-Sheva and 15 m in Nahal expected random (L(r) ¼ 0) distribution of points in the same given Besor. To validate measurement compatibility, the two wadi volume radius. Observed values that are higher than expected (L(r) > 0) surface areas were compared by means of a t-test analysis which suggest clustering, while lower observed values suggest dispersed resulted in p > 0.3815. This result indicates that the surface volume distribution (L(r) < 0). The statistical significance of the observed areas (of 70%) as taken from the two wadis are not significantly values is based on comparing them with the highest or the lowest different. Wadi gradient is the slope gradient calculated along the cumulative frequency (confidence envelope) generated by Monte- axis between the most western and the most eastern points of the Carlo simulations. Monte-Carlo simulation randomly places wadi in a given sub-unit. Wadi depth index was calculated by sub- a number of points equal to the number of the observed values. Each tracting the average wadi bottom elevation from the average sub- set of random placements is a permutation and the confidence unit surface elevation. The wadi depth index is based on the envelope (range) is created according to the number of permutations. assumption that the shallower and more level the wadi bank area, We used 99 permutations for creating the confidence envelope. the lower its depth index, while the deeper and narrower the wadi A clustered pattern, if it is statistically significant, has observed values bank area, the higher the depth index (Fig. 8). Wadi volume is higher than the high-confidence envelope values. A dispersed influenced both by the gradient of the wadi axis itself and by the pattern is statistically significant if the observed values are lower than gradient of the riverbanks. Wadi depth index excludes wadi gradient, the lowest confidence envelope values. therefore it expresses only the riverbank topographical conditions, The largest area analyzed included all 353 site locations in the whether they are steep or level, with no implications for the actual Northern Negev (1252 km2). Smaller sampling areas of analysis volume of the wadi. The same volume could result from an area that included Nahal Besor and Nahal Beer-Sheva. These samplings sub- has an extremely steep riverbank or one that has very level river- areas were created using four equal buffer zones with 500, 1000, banks. Flow accumulation was calculated based on the 8D approach 2000, 3000 m distance off the axis of only two of the major wadi (Ackermann et al., 2008; Svoray, 2004) using the DEM layer with the riverbeds, Nahal Besor and Nahal Beer-Sheva (Fig. 6a,b). Each spatial analysis hydrology tools of ArcGIS. Interaction variables

Fig. 8. Sub-units section illustration of the index of Wadi depth: (a) shallow and leveled wadi surface area section that will result in a low index of depth; (b) a deep and narrow wadi area that will result in a relatively high depth index. R. Winter-Livneh et al. / Journal of Archaeological Science 37 (2010) 284–294 289

Table 1 Results of Moran’s I alternative sampling methodological 2 Indices of Moran’s I spatial analyses results in the Northern Negev (1252 km ). approach, which involved 500 m buffer unit areas around Nahal Grid of analysis Number of sites Index value Z-score Besor and Nahal Beer-Sheva, indicate highly clustered pattern 500 m 353 0.02 62.34 results. Clustered results are also indicated from including all the 250 m 353 0.01 83.23 major wadis (Besor, Beer-Sheva, Grar and Sekher) in the calculation (‘union wadis’ Table 2). Although most tests (23 out of 35) indicate clustering, 12 tests were not indicative (Table 2), all of which were included flow accumulation interaction with the component of wadi part of the reanalysis procedure as described by Fletcher (2008). This depth index, and flow accumulation interaction with wadi gradient. phenomenon can be explained by the relatively small number of These interaction variables were created by multiplying the flow observed values (n < 30) that were used in these specific incidents. accumulation values with the wadi depth index and the accumula- Fig. 9 shows Ripley’s K-function results of site distribution tion values by the wadi gradient. patterns. The results indicate a clustered pattern for a maximum In the first stage, we carried out correlation tests to clarify the radius distance band of 15 km with statistical significance of p < 0.01. relation between wadi volume and the quantity of sites in Nahal Fig.10 provides the results of site distribution in the two study blocks Beer-Sheva and Nahal Besor. This was followed by correlating wadi that were used by Fletcher (2008), ‘Besor – large’ and ‘Beer-Sheva – volume and wadi gradient, in addition to analyzing the relation large’. Results indicate a clustered pattern on both study areas, as between the residuals scores of wadi volume as predicted by the well as in smaller block areas of ‘Besor – medium’ and ‘Beer-Sheva – gradient and the quantities of sites. In the second stage, five of the medium’ (Table 2) with statistical significance of p < 0.01, for aforementioned variables were processed using general linear a maximum radius distance band of 3 km. model (GLM) (Johnston, 1978). The analyzed variables were: wadi Fig. 11 shows results of Ripley’s K-function spatial analysis for all gradient, flow accumulation, wadi depth index, the interaction sub-area buffers in Nahal Besor and Nahal Beer-Sheva (Table 2). The between wadi depth index and accumulation, and the interaction buffer areas of 3000 m and 2000 m indicate cluster patterns at between wadi gradient and flow accumulation. The statistical a maximum cumulative radius distance of 4 km for each of the analyses implemented correlation tests and GLM with STATISTICA 7. wadis. Indication for clustered patterns have also been generated in the buffer areas of 1000 m of each wadi, for a maximum cumulative 4. Results distance of 3 km and in buffer areas of 500 m for a maximum cumulative distance of 2 km. All analyses are products of distance Table 1 shows the results of repeating Fletcher’s (2008) auto- increments of 50 m, with statistical significance of p < 0.01. correlation analyses with the updated modified dataset at the Statistical analyses results in Table 3 indicate a significant largest scale of analysis. This table shows two grid cell sizes for the correlation between wadi volume and the quantity of sites (i.e., site autocorrelation analysis and the results indicate that during the clusters). However, while in Nahal Beer-Sheva the correlation was Chalcolithic period, in the Northern Negev, sites are clustered. positive (r ¼ 0.32, p < 0.05) indicating that site clusters’ locations However, it should be noted that the results in Table 1 are limited were correlated with the increscent of volume; in Nahal Besor, the both here and in Fletcher’s analyses (Fletcher, 2008: 2053, Table 1) correlation was found to be negative (r ¼0.39, p < 0.05), indi- for the following reason. A significant result can be obtained when cating that in this wadi area site clusters’ locations correlated with a dataset is very large. In the case of a 250 m2 grid there are more the decline of the wadi volume measurements. Importantly, though than 20,000 observed values. Moreover, in a cell size of 250 250 m correlation results of r ¼ 0.32 can be stated as imperfect due to the or 500 500 m, only one or two sites may be found (three sites effect of other variables, similar exceptions can be found (e.g., occur only twice in our entire dataset). Actually, in more than 99% of Lyman, 2003). As we mentioned earlier, however, the volume of these grid cells, not even one site was found. Therefore, the dataset wadis is influenced by the gradient of their banks as well as by the variance is minimal and cannot be used for Moran’s I calculations. gradient of the wadi axis. The volume of upstream areas of the Table 2 reports the results of spatial analyses of areas smaller than polygons has therefore minimal effect on the total polygon volume. the entire Northern Negev that were conducted in this research. From this perspective, it could be argued that the correlations Reexamining smaller scale according to the ‘block’ areas sampling reported above stem from the gradient of the wadi bank (i.e., method, such as the ‘Besor – large’/‘Besor – medium’, as well as the polygon) rather than its actual volume. Indeed, the correlation ‘Beer-Sheva – large’/’Beer-Sheva – medium’ areas using Moran’s I between wadi volume and gradient was negative (Beer-Sheva: technique, yields highly significant Z-score values, which indicate r ¼0.63, p < 0.05; Besor: r ¼0.45, p < 0.05). Therefore, correla- clustered patterns for both 250 250 m grid cell and 500 500 m tion between the residual scores of the volume as predicted by the grid cells. It is worth noting that reexamining areas, such as ‘Besor – gradient of each polygon and the site clusters in each polygon was very small’ (‘vsmall’) has produced meaningless results because the calculated. Surprisingly, while in the Nahal Besor the correlation dataset of a maximum area of 4 km2 in a 500 500 m grid allows between the volume and the number of sites remains negative only 20 observed values, while the minimum input features (r ¼0.43, p < 0.05), in the Nahal Beer-Sheva area, the volume (observed values) for the analysis is 30. This problem is also apparent correction returned insignificant results (r ¼ 0.09, p < 0.54). This in Beer-Sheva ‘small’ and ‘vsmall’ areas, Shiqmim areas, and Nahal suggests that changes in the volume of wadis played a major role in Grar areas (Table 2). This problem emphasizes the need for using determining the location of sites, especially in the Nahal Besor area. other techniques, such as the Ripley’s K-function, to better address Table 4 shows the results of multiple regression analysis for both small-scale areas of analysis and small numbers of sites. Additional wadis. It illustrates that two out of the five wadi environmental reexamination of the block grid sampling method used for Moran’s I features listed above could explain close to 25% (R ¼ 0.48, n ¼ 73, autocorrelation analysis by Fletcher (2008) addresses the block grid p < 0.05) of the variance in clustered sites. Interestingly, accumu- sampling location aspect and its influence on the test results. While lation alone is insignificant within the GLM results, while the results of ‘Besor – small’ block 250 250 m grid area in the original interaction variable of accumulation and wadi depth show signifi- location (the location used by Fletcher) do not indicate a clustered cant results, suggesting that these variables depend on wadi depth pattern, moving the sampling block 250 m to the west have changed within this explanation regression model. The beta value of the the results to Moran’s I index of 0.02 and Z score of 2.6, which indi- interaction variable of accumulation and depth indicates that the cates a clustered pattern (Table 2; Fig. 5). larger the interaction variable, the larger the number of sites 290 R. Winter-Livneh et al. / Journal of Archaeological Science 37 (2010) 284–294

Table 2 Indices of Moran’s I and Ripley’s K multiscalar analyses results.

Study area na No. of polygons n s 0 Polygon area Area (km2) Moran’s I Ripley’s K (km2) Avg. SD Index z-score ib r (km)c Besor ‘blocks’ 500 m grid Large 94 560 56 0.25 140 0.04 15.01 C 3 Medium 79 276 50 69 0.03 8.02 C 2 Small 46 60 26 15 0 1.35 Vsmall 20 20 9 5

Besor ‘blocks’ 250 m grid Large 94 2070 76 0.06 130 0.02 24.88 Medium 79 1104 70 69 0.02 15.96 Small 46 209 43 14 0.02d 0.77d Vsmall 20 60 18 4

Beer-Sheva ‘blocks’ 500 m grid Large 46 735 33 0.25 183 0.01 7.94 C 3 Medium 38 340 25 85 0.01 3.7 C 2 Small 21 80 15 20 Vsmall 15 18 10 4.5

Beer-Sheva ‘blocks’ 250 m grid Large 46 2829 35 0.06 173 0.01 12.36 Medium 38 1292 29 80 0.01 6.68 Small 21 300 19 18.7 Vsmall 15 55 13 3.5

Grar ‘blocks’ 500 m grid Large 12 462 12 0.25 115.5 Medium 10 264 10 66 Small 4 78 4 19.5 Vsmall 1 9 1 2.5

Shiqmim ‘blocks’ 500 m grid Small 21 80 17 0.25 20 Vsmall 8 20 6 5

500 m buffer unites Besor 81 31 25 1.01 0 31.3 0.12 2.94 Beer-Sheva 45 49 21 1 0 49 0.11 3.34 Grar 10 22 10 1.01 0 22.1 Sekher 13 14 10 1.01 0 14.1 Union wadis 149 116 66 1.01 0 116.7 0.2 10.15

Besor Ripley’s K buffer 3000 m 132 191.2 C 4 2000 m 120 129.4 C 4 1000 m 105 67.2 C 3 500 m 81 35 C 2

Beer-Sheva Ripley’s K buffer 3000 m 72 367 C 4 2000 m 64 246 C 4 1000 m 56 126 C 3 500 m 51 65 C 2

a Number of sites. b i refers to the indication of pattern, C indicates clustered. c r is the maximum radius of distance. d The same block area has been moved 250 m to the west and produced Moran’s index of 0.02 and z-score of 2.6. expected. Alternatively, it suggests that an increscent in the accu- of the Northern Negev. Other scholars (Fletcher, 2008; Levy, 1981; mulation correlates with the increscent in site numbers only in Levy et al., 2006) have shown contradicting results regarding the places where the wadi depth is great. Furthermore, negative beta spatial nature of these societies. We tackled this contradiction by results of the depth index suggest that Chalcolithic period societies revisiting previous methods, using updated site data, considering preferred to be closer to areas were the wadi riverbanks were alternative methods, and combining an additional spatial analysis shallow and level. That is, the shallower the wadi riverbank, the that is more appropriate to the problem at hand. The use of an out- larger the number of sites expected. of-date dataset and the imposition of sampling grid blocks that are arbitrary in both their location and size (Table 2; Fig. 5) has led 5. Discussion Fletcher (2008) to conclude that settlement distribution is random. He explains that he used ‘blocks’ ‘‘.in order to try to overcome the Quantifying settlement spatial dependence is of great impor- natural clustering that occurs along the wadi courses in order that tance in understanding the Chalcolithic societies and their envi- any pattern that is apparent will reflect criteria other than prox- ronmental and socio-economic contexts. According to our results, imity to resources/environment’’ (Fletcher, 2008: 2053). Our during the Chalcolithic period settlement distributions indicate alternative, the delineation of buffers along the wadis, is based on clustered patterns in the Northern Negev as well as in sub-regions a constant average distance of each buffer sub-unit from the wadi R. Winter-Livneh et al. / Journal of Archaeological Science 37 (2010) 284–294 291

relation between points of location and their observed values can be traced (Conolly and Lake, 2006: 158). In settlement pattern studies, as in most fields of research, the concept of randomness is always assumed to be the null hypothesis (Hodder and Orton, 1976: 53; Schwarz and Mount, 2006: 170). Only when evidence for a pattern is found are alternative explanations sought (Cliff and Ord, 1981; Goodchild, 1986; Hodder and Orton, ExpectedK 1976:174;Wheatley and Gillings, 2002: 127). Fletcher’s (2008) ObservedK Moran’s I analyses produce near-zero, weak relationships and insignificant results. Such results are not indicative of randomness; Cumulative frequency LwConfEnv they only reflect the null hypothesis, i.e., the baseline assumption of HiConfEnv the tests (Wheatley and Gillings, 2002: 128). The fact that the results of Tables 1 and 2 differ markedly from the results of Fletcher (2008) underlines the need for selecting an appropriate method for 500 1500 2500 3500 4500 5500 6500 7500 8500 9500 10500 11500 12500 13500 14500 the problem at hand. This point has already been emphasized by Distance Kvamme (1990b), who revisited the problematic use of Moran’s I analysis by Whitley and Clark (1985). The use of Ripley’s K-function Fig. 9. Northern Negev site distribution pattern: the observed K distribution of sites was, therefore, needed to address this problem. Results of the (ObservedK) are well above the high-confidence envelope line (HiConfEnv) marking Ripley’s K-function indicated clustered patterns along Nahal Beer- a clustered distribution. Sheva and Nahal Besor. Together, our results suggest that Chalco- lithic sites were not randomly distributed along the wadis and that axis. As a result, this produces negligible (near zero) variance of their spatial signature is determined by specific factors. distance within the buffers. Thus, the effects of ‘‘proximity to We argue that these Northern Negev clustered patterns are resources/environment’’ (Fletcher, 2008: 2053) on site distribution primarily determined by the environmental features within the in this sampling method have been neutralized. Using the natural wadis’ local landscape. The statistical results in Table 3 show that curves of the wadi is more justified as it is based on sampling a real changes in wadi volume correlate with site clusters. Further indi- phenomenon. Furthermore, near-zero spatial autocorrelation cation of the immediate adjacent wadi characteristics as factors results in Moran’s I autocorrelation analysis suggest that no clear influencing the settlement distribution is provided by the GLM analysis results in Table 4. A simple test shows a positive correlation between the wadi volume and the settlement clusters (r ¼ 0.32, a p < 0.05) in Nahal Beer-Sheva, and a negative correlation between the two in Nahal Besor (r ¼0.39, p < 0.05). It can be argued that this opposite correlation is due to the fact that Nahal Beer-Sheva is a tributary of Nahal Besor. That is, much of Beer-Sheva water flow reaches Nahal Besor, and therefore topographical characteristics of the riverbanks (such as wadi volume) have little to do with the site ExpectedK distribution patterns because Nahal Besor enjoys much higher ObservedK water flows. To test this argument we calculated flow accumulation and examined its association with the site clusters. LwConfEnv Culumative frequency GLM results (Table 4) show that flow accumulation alone is HiConfEnv statistically insignificant (p > 0.1) in explaining the variability in site clustering. Nevertheless, the difference between adjacent wadis located within the same environment of the Northern Negev 500 1000 1500 2000 2500 implies that variables such as flow accumulation may depend on the Distance level of another independent variable. Thus, interaction effects were added to the GLM. And, indeed, the interaction between flow b accumulation and wadi depth shows a significant result, which indicates that the flow accumulation dependence on wadi depth is on the gradient of the wadi banks. In other words, increase in the interaction variable suggests that site clusters increase where the riverbank area is narrow and relatively steep, mainly when flow accumulation in this specific area increases as well. In areas with relatively low flow accumulation values, increase in site clusters will ExpectedK show only if the riverbank is relatively shallow and level. This ObservedK finding bears a resemblance to the opposite correlation results reported above. Surprisingly, although wadi depth index was Culumative frequency LwConfEnv HiConfEnv already included in the interaction variable discussed above, GLM results also indicate that the wadi depth index alone negatively correlated with site clustering (Table 4). To simplify, this finding suggests that in small-scale areas, where 500 1000 1500 2000 2500 flow accumulation values hardly vary, there is a preference for lower Distance depth index values (i.e., shallow and level riverbanks) in both wadis’ areas. To some extent, this can imply that in both wadi settings, Fig. 10. ‘Beer-Sheva – large’ (a) and ‘Besor – large’ (b): the observed K distribution sites (ObservedK) are well above the high-confidence envelope line (HiConfEnv) marking a subtle reciprocal relationship exists between the subsistence– a clustered distribution. settlement system and its environment. It is, therefore, most likely 292 R. Winter-Livneh et al. / Journal of Archaeological Science 37 (2010) 284–294

Nahal Beer-Sheva Nahal Besor y yc c neu n e u q qerf e r f

ev evit i tam 3000 m. a lu ul m uC uC 004 057 00 0 0081 0512 0052 05 002 0553 0 0 0 0011 05 008 0512 0 0582 0 0 0 5 0 04 57 0 0 5 0 1 41 8 9 41 5 23 53 9 1 2 3 3 1 2 3 Distance Distance yc ycn ne e uqe u qerf r f ev e v i i talumuC t 2000 m. alumuC

004 057 0011 0541 0 05 00 05 0023 0553 009 004 057 0 05 0081 0 00 0 0023 0 0 0 0 5 5 55 0 8 12 5 8 11 4 12 5 82 9 1 2 2 3 1 2 3 3 Distance Distance ycn ycn e eu uqerf e uqerf qer f f evit v ital 1000 m. a mu u luC muC

004 05 00 0 00 0512 0 0582 00 05 0011 0541 0 0 0052 058 5 0 0 5 7 1 41 8 52 4 7 8 12 1 1 1 2 Distance Distance yc ycne n euqerf e euqerf u qerf e v vi itam 500 m. tam ul u l uC uC 00 055 008 0 0031 05 0081 003 0 008 05 0031 0551 0 50 5 08 3 5 5 0 1 1 1 1 Distance Distance

ExpectedK LwConfEnv ObservedK HiConfEnv

Fig. 11. Nahal Beer-Sheva and Nahal Besor Ripley’s K-function results. The observed K distribution sites (ObservedK) are well above the high-confidence envelope line (HiConfEnv) marking a clustered distribution.

that the wadi resource (of thick fertile soils due to accumulation of adaptive strategy is indicated, by locating sites closer to areas where sediments, water and vegetation establishment), was not over- the riverbank was most level and shallow and, thus, could more exploited by complex irrigation systems such as dams or diversion effectively, and with the least effort, be exploited for agriculture walls along the banks, as claimed by Levy (1986). A more subtle (Butzer, 1982). The key in establishing sedentary agriculture in semi-arid Table 3 conditions is maximum utilization of the meager rainfall (Evenari Statistical analysis correlation results. et al., 1961; Lavee et al., 1997). Topography has an important role in Variables Nahal Beer-Sheva Nahal Besor water availability as well as in the accumulated depth of loess (n ¼ 41) (n ¼ 32) deposits for practicing agricultural fields, as well as on vegetation rprp cover (e.g., Carmel and Kadmon, 1999; Davis and Goetz, 1990; Site clusters vs wadi volume 0.32 <0.05 0.39 <0.05 Jobbagy et al., 1996; Svoray et al., 2008). Our results show that Wadi volume vs gradient 0.63 <0.05 0.45 <0.05 during the Chalcolithic period the landscape was used, most a Residual scores vs gradient 0.09 <0.54 0.43 <0.05 probably, for dry farming based on topographical features of the a Residual score of the wadi volume as predicted by the gradient of each polygon. wadis rather than for constructing irrigation facilities. Only during R. Winter-Livneh et al. / Journal of Archaeological Science 37 (2010) 284–294 293

Table 4 GLM multiple regression summary of dependent variables of sites within Nahal Beer-Sheva and Nahal Besor.

R ¼ 0.4829616 R2 ¼ 0.2332519 Adjusted R2 ¼ 0.17848421 t(73) p-level

F(5,70) ¼ 4.2589 p < 0.00194 SE of estimate: 1.7654

n ¼ 73 Beta SE of beta B SE of B Intercept 2.617587 1.002542 2.61095 0.011037 Gradient 0.095157 0.228035 0.040917 0.098055 0.41729 0.677744 Accumulation 0.368951 0.276705 0.000001 0 1.33337 0.186733 Depth index 0.629776 0.258372 0.399932 0.164076 2.43748 0.017334 Interaction: accumulation and depth 1.129543 0.400969 0 0 2.81704 0.006294 Interaction: accumulation and gradient 0.05716 0.232907 0 0 0.24542 0.806849

the Iron Age, can clear archaeological evidence be found for irri- distribution in the Northern Negev has been recognized as clustered. gation facilities in the Negev region (Evenari et al.,1961; Lavee et al., This pattern was verified in our updated study and it is attributed to 1997; Shanan, 2000). the topographic characteristics of the landscape. Our results support previous suggestions that during the Chal- Our analyses also indicate that Chalcolithic site location is highly colithic period agricultural practice in the Southern Levant included affected by the nature of the riverbanks. In the case of Nahal Beer- both high water availability conditions (Rosen and Weiner, 1994), Sheva, preferred sites were relatively shallow and level riverbanks, which was accomplished not by building diversion walls and dams, while in Nahal Besor, clusters were associated with relatively but by placing the fields in floodplain areas (Hillel, 1994), and close narrow and steep areas, with an increase in flow accumulation. This to riverbank topographical features other than floodplains (Katz difference cautiously supports the division of the society, during et al., 2007) that are less flat or level, but offer a greater amount of this period, into two different entities: ‘Beer-Sheva cluster’ and flow accumulation. ‘Besor cluster’ (Gilead, 1995, 2007). Dry farming, which is not based on artificial irrigation facilities Furthermore, by providing another perspective for evaluating such as dams, diversion walls or canals, is universally and cross- the issues of water management and wadi resource exploitation of culturally recorded in agricultural small-scale societies in semi-arid the Northern Negev Chalcolithic sites, our results support previous and arid regions across the globe (El-Fouly et al., 2007; Kanitkar suggestions of dry farming (Katz et al., 2007), and flood-irrigation et al., 1968; Mahmood, 1995; Oweis and Hachum, 2005) and in agricultural practice (Rosen and Weiner, 1994). We therefore current Bedouin societies in the Negev (Meir and Tsoar, 1996). conclude that the exploitation of the resources of wadi environ- Archaeologically, dry-farming techniques have rarely been studied ments, such as soils, water and vegetation, was enhanced by the use in such early societies in the Southern Levant. Data from this study of natural runoff sink areas rather than by the construction of improve our knowledge about the agricultural strategies of the early irrigation systems such as dams or diversion walls along the wadis. farmers, perhaps a necessary step on the way to full-scale irrigated Water management and the use of wadi resources were therefore farming. much simpler, adjusted to the most basic agricultural strategy. This The settlement pattern of the Chalcolithic societies in the strategy consists of locating sites in parts of Nahal Beer-Sheva that Northern Negev is associated with small-scale societies with are shallow and level riverbanks with low accumulation values. In a relatively low level of social and economic complexity. The water Nahal Besor the sites are located in areas where the riverbanks are needed to grow crops was not dependent on diversion walls, canals relatively deep and steep but with high accumulation values. This or dams. Enough water was available naturally by placing fields in implies that varied agricultural practices were used during the shallow and level areas in the floodplains of Nahal Beer-Sheva. In the Chalcolithic period in the Northern Negev. case of Nahal Besor, the sites clustered in relatively steep and narrow areas that are associated with greater amounts of flow accumula- Acknowledgements tion. This difference between the Nahal Beer-Sheva and Nahal Besor sites in relation to wadi volume suggests different agricultural The authors wish to thank the Pratt Foundation PhD Fellowship strategies. Furthermore, these findings support the hypothesis that Program at Ben-Gurion University of the Negev. We thank the during the Chalcolithic period there were two cultural entities, Geological Survey of Israel for providing the DEM. The professional which are referred to as the ‘Beer-Sheva cluster’ and the ‘Besor help of the Israel Antiquities Authority deputy director, Dr Uzi cluster’(Gilead, 1995, 2007). This dichotomy, based until now solely Dahari, and of the members from the Israel Antiquities Authority on typology of artifacts and the variability between the Besor sites southern regional office is acknowledged. We thank also Prof. and Beer-Sheva sites, has been acknowledged since the 1950s (e.g., Kenneth Kvamme for help with less accessible literature, and for Gophna, 1979; Perrot, 1955; Roshwalb, 1981). Our study suggests providing useful advice on the subject of autocorrelation analysis. that differences in the microenvironments of wadis are behind the Uri Gilead is thanked for his helpful information regarding the DEM economic and cultural differences between the Beer-Sheva and the and Prof. Steve Rosen for his helpful comments and advice. The Besor clusters of sites. members of the GI-Lab at Ben-Gurion University are thanked for providing useful information concerning GIS procedures. 6. Conclusions References The settlement distribution patterns of past human activities across a landscape provide important information not only regarding Ackermann, O., Svoray, T., Haiman, M., 2008. Nari (calcrete) outcrop contribution to the way people used the landscape, but also about the adaptation of ancient agricultural terraces in the southern Shephelah, Israel: insights from communities to environmental limiting factors, and about social and digital terrain analysis and a geoarchaeological field survey. Journal of Archaeological Science 35, 930–941. economic organization (Clarke, 1977; Conard, 2001; Holl and Levy, Alon, D., 1961. Early settlements along the Nahal Grar and Nahal Patish. M’befnim 1, 1993; Silbernagel et al., 1997). For decades, Chalcolithic site 87–96. 294 R. Winter-Livneh et al. / Journal of Archaeological Science 37 (2010) 284–294

Alon, D., Levy, T.E., 1980. Preliminary note on the distribution of Chalcolithic sites on Katz, O., Gilead, I., Bar, P., (Kutiel), Shahack-Gross, R., 2007. Chalcolithic agricultural the Wadi Beersheva and lower Wadi Besor drainage system. Israel Exploration life at Grar, northern Negev, Israel: dry farmed cereals and dung-fueled hearths. Journal 30, 140–147. Paleorie´nt 31, 101–116. Bevan, A., Conolly, J., 2006. Multiscalar approaches to settlement pattern analysis. Kvamme, L.K., 1990a. GIS algorithms and their effects on regional archaeological In: Lock, G., Molyneaux, B.L. (Eds.), Confronting Scale in Archaeology: Issues of analysis. In: Allen, K.M.S., Green, S.W., Zubrow, B.W.E. (Eds.), Interpreting Space: Theory and Practice. Springer, New York, pp. 217–234. GIS and Archaeology. Taylor & Francis, London, pp. 112–125. Beyer, H.L., 2004. Hawth’s Analysis Tools for ArcGIS. Available at. http://www. Kvamme, L.K., 1990b. Spatial autocorrelation and the Classic Maya Collapse revis- spatialecology.com/htools/. ited: refined techniques and new conclusions. Journal of Archaeological Science Butzer, K.W., 1982. Archaeology as Human Ecology. Cambridge University Press, 17, 197–207. Cambridge. Lavee, H., Poesen, J., Yair, A., 1997. Evidence of high efficiency water-harvesting by Carmel, Y., Kadmon, R., 1999. Effect of grazing and topography on long-term ancient farmers in the Negev Desert, Israel. Journal of Arid Environments 35, vegetation changes in a Mediterranean ecosystem in Israel. Plant Ecology 145, 341–348. 243–254. Levy, T.E., 1981. Chalcolithic Settlement and Subsistence in the northern Negev Clarke, D.L. (Ed.), 1977. Spatial Archaeology. Academic Press INC, London. Desert, Israel. University of Sheffield, Ann Arbor. Cliff, A.D., Ord, J.K., 1981. Spatial Processes, Models and Applications. Pion, London. Levy, T.E., 1986. The Chalcolithic period. Biblical Archaeologist 49, 82–108. Conard, N.J., 2001. Advances and problems in the study of Paleolithic settlement Levy, T.E., 1995. Cult, metallurgy and rank societies – Chalcolithic period (CA 4500– systems. In: Conard, N.J. (Ed.), Settlement Dynamics of the Middle Paleolithic 3500 BCE). In: Levy, T.E. (Ed.), Archaeology of Society in the Holy Land. Facts on and Middle Stone Age. Kerns Verlag, Tu¨ bingen, pp. 7–20. File, New York, pp. 226–243. Conolly, J., Lake, M., 2006. Geographic Information Systems in Archaeology. Cam- Levy, T.E., Alon, D., 1983. Chalcolithic settlement patterns in the northern Negev bridge University Press, Cambridge. desert. Current Anthropology 24, 105–107. Danin, A., Orshan, G., Zohary, M., 1975. The vegetation of the northern Negev and Levy, T., Goldberg, P., 1987. The environmental setting of the northern Negev. In: the Judaean desert of Israel. Israel Journal of Botany 24, 118–172. Levy, T. (Ed.), Shiqmim vol. I. BAR International Series 356, Oxford, pp. 1–22. Davis, F.W., Goetz, S., 1990. Modeling vegetation pattern using digital terrain data. Levy, T.E., Burton, M.M., Rowan, Y.M., 2006. Chalcolithic hamlet excavations near Landscape Ecology 4, 69–80. Shiqmim, Negev desert, Israel. Journal of Field Archaeology 31, 41–60. El-Fouly, M.M., El-Sayed, A.A., Fawzi, A.F.A., Shaaban, S.H.A., 2007. Nutritional status Lyman, R.L., 2003. The influence of time averaging and space averaging on the of oil olives groves grown under dry framing conditions in North Western Coast application of foraging theory in zooarchaeology. Journal of Archaeological of Egypt. International Journal of Food, Agriculture and Environment 5, Science 30, 595–610. 216–219. Mahmood, A., 1995. Economics of Barani (rainfed) farming and farm household Evenari, M., Shanan, L., Tadmor, N., Aharoni, Y., 1961. Ancient agriculture in the production behaviour in Pakistan. The Pakistan Development Review 34, 901–912. Negev. Science 133, 979–996. Maisels, C.K., 2001. Civilizations of the Old World, the Early Mesopotamia, India, Evenari, M., Shanan, L., Tadmor, N., 1971. The Negev: The Challenge of a Desert. Formative Histories of Egypt, the Levant and China. Routledge, London and Harvard University Press, Cambridge. New York. Fletcher, R., 2008. Some spatial analyses of Chalcolithic settlement in southern Meir, A., Tsoar, H., 1996. International borders and range ecology: the case of Israel. Journal of Archaeological Science 35, 2048–2058. Bedouin transborder. Human Ecology 24, 39–64. Fletcher, R., Winter, R., 2008. Prospects and problems in GIS applications in Mitchell, A., 2005. The ESRI Guide to GIS Analysis. Vol. 2. ESRI Press. studying Chalcolithic archaeology in southern Israel. Bulletin of the American Moran, P., 1948. The interpretation of statistical maps. Journal of the Royal Statistical Schools of Oriental Research 352, 1–28. Society 10B, 243–251. Fortin, M.J., 1999. Spatial statistics in landscape ecology. In: Klopatek, M.J., Nir, D., 1970. The Geomorphology of Israel. Aqademon, Jerusalem. Gardner, H.R. (Eds.), Landscape Ecological Analysis: Issues and Applications. Oweis, T., Hachum, A., 2005. Water harvesting and supplemental irrigation for Springer, New York, pp. 253–279. improved water productivity of dry farming systems in West Asia and North Gatrell, A.C., Bailey, T.C., Diggle, P.J., Rowlingson, B.S., 1996. Spatial point pattern Africa. Agricultural Water Management 80, 57–73. analysis and its application in geographical epidemiology. Transactions of the Perrot, T., 1955. Excavations at Tell Abu Matar, near Beersheba. Israel Exploration Institute of British Geographers 21, 256–274. Journal 5, 17–40. 73–84, 167–189. Gazit, D., 1996. Map of Urim (125). Israel Antiquities Authority, Jerusalem. Renfrew, C., 1973. Monuments, mobilization and social organization in Neolithic Gibson, S., Rowan, Y.M., 2006. The Chalcolithic in the central highlands of Palestine: Essex. In: Renfrew, C. (Ed.), The Explanation of Cultural Change: Models in a reassessment based on new examination of Khibet es-Sauma’a. Levant 38, Prehistory. Gerald Duckworth and Company Ltd, London, pp. 539–558. 85–108. Ripley, B.D., 1976. The second-order analysis of stationary point processes. Journal Gilead, I., 1988. The Chalcolithic period in the Levant. Journal of World Prehistory 2, of Applied Probability 13, 255–266. 397–443. Rosen, A.M., Weiner, S., 1994. Identifying ancient irrigation: a new method using Gilead, I., 1993. Sociopolitical organization in the northern Negev at the end of the opaline phytoliths from Emmer wheat. Journal of archaeological Science 21, Chalcolithic period. In: Biran, A., Aviram, J. (Eds.), Biblical Archaeology Today, 125–132. 1990, Supplement. Israel Exploration Society, Jerusalem. Rosen, A.S., 1993. Metals, rocks, specialization, and the beginning of urbanism in the Gilead, I., 1995. Grar: A Chalcolithic Site in the northern Negev. Ben-Gurion northern Negev. In: Biran, A., Aviram, J. (Eds.), Biblical Archaeology Today, 1990. University Press, Beer Sheva. Israel Exploration Society, Jerusalem, pp. 41–56. Gilead, I., 2007. The Besorian: a pre-Ghassulian cultural entity. Paleorie´nt 33, 33–49. Roshwalb, A.F., 1981. Protohistory in the Wadi Ghazzeh: a typological study based Gilead, I., Goren, Y., 1986. Stations of the Chalcolithic period in Nahal Sekher, on the Macdonald excavations. PhD dissertation. University of London, London. northern Negev. Paleorie´nt 12, 83–90. Schwarz, K.R., Mount, J., 2006. Integrating spatial statistics into archaeological data Goodchild, M.F., 1986. Spatial Autocorrelation. Geo Books, Norwich. modeling. In: Mehrer, M.W., Wescott, K.L. (Eds.), GIS and Archaeological Site Gophna, R., 1979. The settlement of north-western Negev during the Chalcolithic Location Modeling. Taylor and Francis, New-York, pp. 167–189. period. In: Shmueli, A., Grados, Y. (Eds.), The Land of the Negev I. Tel-Aviv, pp. Shanan, L., 2000. Runoff, erosion, and the sustainability of ancient irrigation 203–208. systems in the Central Negev desert. In: Hassan, A.M., Slaymaker, O., Gophna, R., 1990. The Early Bronze Age I settlement at ’En Besor oasis. Israel Berkowicz, M.S. (Eds.), The Hydrology–Geomorphology Interface: Rainfall, Exploration Journal 40, 1–11. Floods, Sedimentation, Land Use. The Hebrew University of Jerusalem, Jer- Griffith, D., 1987. Spatial Autocorrelation: A Primer. Association of American usalem, pp. 75–106. Geographers, Washington, DC. Shmueli, A., Gradus, Y., (Eds.), 1979. The Land of the Negev, part 1. Ministry of Hall, J.K., 1993. The GSI digital terrain model (DTM) project completed. GSI Current Defence Publishing House, Tel Aviv (Hebrew). Research 8, 47–50. Silbernagel, J., Martin, S.R., Gale, M.R., Chen, J., 1997. Prehistoric, historic, and Hermon, S., 2008. Socio-economic Aspects of Chalcolithic (4500–3500 BC) Societies present settlement patterns related to ecological hierarchy in the Eastern Upper in the southern Levant. In: BAR International Series 1744. Archaeopress, Oxford. Peninsula of Michigan, USA. Landscape Ecology 12, 223–240. Hillel, D., 1994. Rivers of Eden: The Struggle for Water and the Quest for Peace in the Singer, A., 2007. The Soils of Israel. Springer, Berlin. Middle East. Oxford University Press, Oxford. Svoray, T., 2004. Integrating automatically processed SPOT HRV Pan imagery in Hodder, I., Orton, C., 1976. Spatial Analysis in Archaeology. Cambridge University a DEM-based procedure for channel network extraction. International Journal Press, Cambridge. of Remote Sensing 25, 3541–3547. Holl, A., Levy, T.E. (Eds.), 1993. Spatial Boundaries and Social Dynamics: Case Studies Svoray, T., Karnieli, A., Dedieu, G., 2008. Satellite evidence of a topographic sink– from Food-Producing Societies. International Monographs in Prehistory, Ann source system in a small semi-arid watershed. International Journal of Remote Arbor. Sensing 29, 609–616. Jobbagy, E.G., Paruelo, J.M., Leon, R.J.C., 1996. Vegetation heterogeneity and diversity Wheatley, D., Gillings, M., 2002. Spatial Technology and Archaeology. Taylor and in flat and mountain landscapes of Patagonia (Argentina). Journal of Vegetation Francis, London. Science 7, 599–608. Whitley, D.S., Clark, W.A.V., 1985. Spatial autocorrelation test and the Classic Maya Johnston, R.J., 1978. Multivariate Statistical Analysis in Geography: A Primer on the collapse: methods and inferences. Journal of Archaeological Science 12, 377–395. General Linear Model. Longman, London. Wright, J.P., Jones, C.G., Boeken, B., Shachak, M., 2006. Predictability of ecosystem Kanitkar, N.V., Sirur, S.S., Gokhale, D.H., 1968. Dry Farming in India. Indian Council of engineering effects on species richness across environmental variability and Agricultural Research, New Delhi. spatial scales. Journal of Ecology 94, 815–824. 2.2 Secondary burial cemeteries, visibility and land tenure: A view from the southern Levant Chalcolithic period

This study was published in Journal of Anthropological Archaeology 31: 423-438 (2012). A copy of the article follows.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 30 Author's personal copy

Journal of Anthropological Archaeology 31 (2012) 423–438

Contents lists available at SciVerse ScienceDirect

Journal of Anthropological Archaeology

journal homepage: www.elsevier.com/locate/jaa

Secondary burial cemeteries, visibility and land tenure: A view from the southern Levant Chalcolithic period ⇑ Rona Winter-Livneh a, , Tal Svoray b, Isaac Gilead a a Dept. of Bible, Archaeology and Ancient Near Eastern Studies, Ben-Gurion University of the Negev, Israel b Dept. of Geography and Environmental Development, Ben-Gurion University of the Negev, Israel article info abstract

Article history: Off-site secondary burial cemeteries in the southern Levant are an innovation of the Chalcolithic period. Received 4 January 2011 Ethnographic studies suggest that location of burial places was one of the means used to establish and Revision received 26 February 2012 socially mediate ownership over the landscape. The current research examines whether the spatial pat- Available online 3 May 2012 tern of burial sites during the Chalcolithic period in the southern Levant represents a land tenure system. Spatial analyses of burial and habitation sites located mainly along the central Israeli coastal area indicate Keywords: that Ghassulian communities avoided locating their burial sites near habitation sites concentrations. Chalcolithic Viewshed analysis indicates that the areas observed from burial sites are significantly larger than those Secondary burial observed from habitation sites, and that their location scattered across an area which increases rather Land tenure Visibility than overlaps the size of area visible from the habitation sites. Furthermore, our results show that com- Viewshed munities which wanted to claim land tenure over a larger territory use burial sites locations in order to GIS maximize the observed area within habitation sites surroundings. Southern Levant Ó 2012 Elsevier Inc. All rights reserved.

Introduction ella and the central hill country, the central and southern coastal plain, the Dead Sea basin and in the northern Negev (Gilead, Mortuary practice, a common cultural aspect encountered by 2007, p. 35; Gilead, 2009, p. 345). Surveys and studies indicate a archaeologists, is an important feature in understanding past soci- pronounced population growth during this period (Finkelstein eties social and economic systems, ideologies, and worldviews (e.g. and Gophna, 1993; Gophna and Portugali, 1988; Gophna and Tsuk, Binford, 1971; Carr, 1995; Chapman et al., 1981; Hodder, 1982; 2005; Khalaily and Marder, 2010). This is apparent not only from Kuijt, 1996; Metcalf and Huntington, 1991; O’Shea, 1984). Data ob- the increase in the number of known sites but also from the in- tained from archaeological excavations of burial sites usually in- crease in site size and from the expansion into areas which were cludes artifacts, installations, structures, and bones. While these unpopulated or sparsely populated prior to the Chalcolithic period, remain the traditional focus of archaeological study and interpreta- such as the semi-arid Northern Negev (Golden, 2009,p.9;Khalaily tion, additional important information can be derived from study- and Marder, 2010, p. 15; Lev-Tov Chattah and Smith, 2006, p. 472; ing the location of burial sites in the landscape (Clarke, 1977; Lev-Tov et al., 2003, p. 122; Rowan and Golden, 2009, pp. 27–28; Goldstein, 1981, 2002; Williams, 1999). Significant changes in Winter-Livneh et al., 2010, p. 284). mortuary practices were introduced during the Chalcolithic period The Chalcolithic communities, mainly of sedentary farmers, in the southern Levant. The most prominent entity of the Chalco- practiced mixed agriculture; cultivation of cereals, pulses and fruit lithic period is the Ghassulian culture (ca. 4500–3900 BC cal.), trees, herding of sheep-goats and raising of, pigs and cattle (Gilead, which features assemblages broadly similar to those found at the 1988a; Levy, 1995; Rowan and Golden, 2009). Craft specialists such upper levels of Teleilat Ghassul, the Ghassulian type-site (Gilead, as ivory and copper producers are worth noting (Gilead, 1988a; 2011; North, 1959). They include specific sets of artifact-type cat- Golden, 2009; Levy, 1995; Rowan and Golden, 2009). The nature egories such as V-shaped bowls, churns, cornets, sickle blades, of the socio-political organization of the Chalcolithic communities microliths, basalt bowls, and copper and ivory artifacts. Geograph- is controversial. There are scholars who follow Service (1962) and ically, the Ghassulian is distributed in the Jordan valley, the Sheph- suggest a hereditary chiefdom society with centers that coordinate social, economic and religious activities (Gibson and Rowan, 2006; ⇑ Golden, 2009; Levy, 1986, 1995, 2006). Other scholars find the Corresponding author. Address: Dept. of Bible, Archaeology and Ancient Near archaeological evidence to reflects a lesser level of complexity Eastern Studies, Ben-Gurion University of the Negev, P.O.B. 653, Beer-Sheva 84105, Israel. (Epstein, 1998; Gilead, 1988a, 1993, 1995; Gophna and Tsuk, E-mail address: [email protected] (R. Winter-Livneh). 2005; Hermon, 2008; Rosen, 1993). The society consisted of rural

0278-4165/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jaa.2012.03.002 Author's personal copy

424 R. Winter-Livneh et al. / Journal of Anthropological Archaeology 31 (2012) 423–438 communities whose social and economic activities were influenced mark of the Chalcolithic period and is marginal in earlier time by groups of seniors household heads (Gilead, 1988a, p. 436), sim- spans. ilar to what Maisels (2001, pp. 156–168) calls ‘‘augmented and Although beyond the scope of this paper, it should be empha- stratified households’’. sized that the very concept of ‘‘secondary burial’’ has been dis- It is within this geographic, demographic, economic and social cussed and critiqued in recent anthropological studies (e.g. contexts that profound changes in mortuary practices took place, Chénier, 2009; Weiss-Krejci, 2004, 2011). The too common and and most notable is the establishment of off-site cemeteries of sec- unqualified use of the term ‘‘secondary burial’’ masks an impres- ondary burials in ossuaries (Gilead, 1988a; Levy, 1986). In such buri- sive cross-cultural heterogeneity and hides pronounced differences als the body is interred in habitation deposits such as pits, silos and of a wide range of burial practices, their symbols, and their social below floors walls (Gilead, 1988a, p. 428), and later on, after a cul- and economic implications. This is valid also for the archaeology turally determined length of time, the remains are retrieved and of the southern Levant. The pre-Chalcolithic practice of skull re- re-interred in a new location (Schroeder, 2001, p. 79). In this respect, moval and reburial on the one hand, and the Chalcolithic practice secondary burial, which usually involves a primary burial, is per- of collecting bones after natural decarnization and placing in ossu- ceived here as an exclusive class or type of mortuary practice. Dur- aries, on the other hand, have very little in common. The origins, ing preceding periods most burials were primary ones, and most of consequences and meanings of each of these practices differ con- them are located within the habitation or household surroundings siderably. Yet, most archaeologists working in the southern Levant (e.g. Kuijt, 1996). Chalcolithic period secondary burials are known will use the same term – secondary burial – to describe them. only from cemetery sites, mostly from burial caves. The goal of this Beside the profound differences described above, most of the study is to explore and analyze the spatial aspects of burial sites and pre-Chalcolithic secondary burials as well as primary burials, occur their relation to the location of the habitation sites. in intramural, subfloor and other residential contexts, excluding Kfar HaHoresh (Goring-Morris (2000). ‘Uyun al-Hammam, for Pre-Chalcolithic and Chalcolithic secondary burials in the Levant example, where the earliest documented secondary skull removal has been recently uncovered, is an open-air site, covering an area A relatively small numbers of secondary burials have been re- of 1000–1500 m2. The excavators documented ‘‘potential trampled ported from the southern Levant prior to the Chalcolithic period. earth surface, ash dumps and several discrete refuse middens’’, The earliest selective skull removal has been recently reported ‘‘The burials are dug into pre-existing Epipaleolithic deposits’’ from the Middle Epipalaeolithic period (c. 23,000–11,600 cal BP) (Maher et al., 2011, p. 2). at ‘Uyun al-Hammam in northern Jordan (Maher et al., 2011). During early stages of the Natufian occupation at Hayonim cave, The practice of skull removal continued into the Late Natufian (c. the dwellings structures were built in front of the cave while the 13,000–10,300 BC) as has been demonstrated at Hayonim Cave burials were located to the rear. In the later stages, the burials were (Belfer-Cohen, 1988, p. 300; Belfer-Cohen, 1991, p. 171), Hayonim placed in the front while the dwelling were located in the inner Terrace (Valla, 1986), Nahal Oren (Stekelis and Yizraeli, 1963), and part of the cave (Belfer-Cohen, 1988, p. 297). Though some sort Eynan (Perrot and Ladiray, 1988) (see Fig. 1). During the Pre-Pot- of separation can be recognized between the domestic and the bur- tery Neolithic A (PPNA) (ca. 10,300–8500 BC) and Pre-Pottery Neo- ial activity, both are placed within the same site, within the same lithic B (PPNB) periods (ca. 8500–6250 BC), skull removal became cave (Belfer-Cohen, 1991, p. 171). Natufian burial and habitation more frequent (Kuijt, 1996, p. 319; Kuijt, 2001, p. 84; Kuijt and structures were within the confines of the same site is also appar- Goring-Morris, 2002, p. 376), although, by far, the majority of buri- ent at Eynan (Perrot and Ladiray, 1988). Clearly, during these early als are primary (Peterson, 2010, p. 254; Verhoeven, 2002, Tab. 1B, periods burial or sacred activities were not differentiated spatially p. 7). During the Pottery Neolithic (PN) period (ca. 6250–4500 BC), from everyday or domestic activities, into which they were secondary burials (including skull manipulation) became rare strongly integrated. (Verhoeven, 2002, p. 7). Although secondary burials and skull Although during the PPN and PN periods skull removal prolifer- manipulation were practiced during the Epipalaeolithic and Neo- ated as mentioned above, interment was still carried out within the lithic periods, off-site cemeteries are virtually unknown, excluding perimeter of the sites. At the PPNB site of ‘Ain Ghazal for example, Kfar HaHoresh as suggested by Goring-Morris (2000) (Fig. 1). three skulls (two adults and one child) placed in a row, were recov- During the Chalcolithic period the number of secondary burials ered from beneath the floor of the southeast corner of a house. In as well as off-site cemeteries is unprecedented. While it is reason- another room, within the same structure, a skull of an adolescent able to assume that during the Pre-Pottery and Pottery Neolithic was placed beneath the southwest corner of the floor (Rollefson, periods secondary burials did not exceed a few hundred (Hershko- 1986, pp. 50–51). vitz and Gopher, 1990,p.12), during the Chalcolithic period sec- There are three exceptions in which primary burial, skull re- ondary burials probably reached well into the thousands. This is moval and secondary treatment were observed in non-residential based on the fact that during this period more than 50 formal bur- sites: At the Natufian site of Hilazon Tachtit (Grosman and Munro, ial sites are known in this region, most of which are caves (see be- 2007; Grosman et al., 2008), at the PPNB site of Kfar HaHoresh low). An estimation of the number of individuals buried within (Goring-Morris, 2000; Goring-Morris and Horwitz, 2007) and at each burial cave has reached in some cases well into the hundreds. Nahal Hemar (Arensburg and Hershkovitz, 1989). These are not In the Peqi’in cave, for example, it is estimated that remains of residential sites, but the burials there are not in contexts that 600–1000 individuals were recovered (Gal et al., 2011, p. 204; Lev- can be regarded as representing cemeteries. Tov et al., 2003, p. 123 with reference therein). Remains of 174 The Natufian site at Hilazon Tachtit cave is unique in many as- individuals were uncovered at the recently excavated burial caves pects. It contains a burial of a woman interpreted as a shaman site of Horvat Qarqar South (Yosef Nagar personal communication). (Grosman et al., 2008, p. 17668), with a high concentration of tor- Considering the recent serious damage caused to the site, the pos- toises shells and remains of three aurochs which are interpreted as sibility that about 500 individuals were buried at the caves cannot remains of a feast (Munro and Grosman, 2010, p. 15365). However, be excluded (Peter Fabian personal communication). Moreover, the flint and faunal assemblages recovered at the site ‘‘represent a these figures should be related to the time duration of the Chalco- broad spectrum of everyday activities’’ (Grosman and Munro, 2007, lithic period. While periods such as the PPNA, PPNB or PN lasted p. 12). Moreover, ‘‘the raw materials used for tools manufacture approximately 2000 years each, the Chalcolithic period is about were collected from the wadi below the cave -confirming that 600 years (Gilead, 2011). Thus, secondary burial practice is a hall- the Natufians were using local environments rather than visiting Author's personal copy

R. Winter-Livneh et al. / Journal of Anthropological Archaeology 31 (2012) 423–438 425 the area only for specific burial activities.’’ (Grosman and Munro, Many off-sites cemeteries consist of several adjacent caves, e.g. 2007, p. 12). Giv’atayim – seven burial caves; Sha’ar Efrayim – seven burial The PPNB site of Kfar HaHoresh produced a number L-shaped caves; Qula – 10 burial caves; Shoham (North) – six burial caves; walls and lime-plastered floors below which a large number of Ben Shemen – six burial caves; Palmahim – 11 burial caves. There burials were found. The burials were both primary and secondary, are natural karstic caves such as Peqi’in, Nahal Qanah, alongside totaling more than 60 individuals, including three plastered skulls. artificially hewn caves such as Azor, Hadera, Bene-Berak, Giv’ata- Animal carcasses were associated with some of the burials. Fifteen yim, Tel-Aviv, Ma’abarot, and Shoham North (Van den Brink, of the primary burials show evidence of post-depositional head re- 1998, 2011a)(Fig. 1). moval (Goring-Morris, 2000; Goring-Morris et al., 1998; Goring- The hewn caves can be found either in kurkar ridges or lime- Morris and Horwitz, 2007; Horwitz and Goring-Morris, 2004, pp. stone hills. They are of oval, irregular-oval, or semicircular form 168–169 and see references within). This site was interpreted by quarried into the bedrock surface, some of which feature pillars the excavator as a ‘‘regional funerary center for nearby communi- as roof supports and additional niches. They are only few meters ties’’ (Goring-Morris, 2000, p. 109), Others, however, doubt it in size (e.g. Ma’abarot: 4 8 m; Tel Ifshar: diameter of c. 1.5 m; (Gafinkel, 2006, p. 114). In any case, the site yielded evidences of Et-Taiyiba: c. 4.5 6 m) but there are also larger chambers (e.g. daily activities as well as areas for lime plaster production, flint Shoham (north) Cave 3: c. 21 18 m and Cave 4: c. 15 12 m). knapping and midden deposits rich in burnt organic remains and Their maximum height varies between 1.5 and 2.5 m. Several ash (Horwitz and Goring-Morris, 2004, pp. 166–167). Another phases of use, within the Chalcolithic period, were reported from PPNB site, Nahal Hemar, is a non-residential site with numerous several caves (e.g. Ma’abarot, Shoham (north), Azor, and Ben She- unique artifacts as well as burials. It is interpreted as a special stor- men). The stratigraphy and the typology, mainly of bowls, within age place for cultic objects (Bar-Yosef, 1986). The sites of Hilazon burial sites such as Azor and Ben Shemen indicate change over Tachtit and Kfar HaHoresh lack remains of domestic architecture. time (Nativ and Gopher, 2011; Perrot and Ladiray, 1980). It seems, Nonetheless, the heterogeneity of the archaeological remains indi- therefore, that these caves were used recurrently for secondary cates that both domestic and burial activities took place within burials and maybe ‘‘on other occasions’’ too (Van den Brink, these special sites. 2005, p. 184). The pre-Chalcolithic mortuary practices indicate that secondary Chalcolithic secondary burials are characterized by the inten- burials were far less frequent during these early periods. In most sive use of ossuaries, ceramic or stone receptacles for depositing cases the secondary burial practice consists of skull removal. More- disarticulated bones that signify a secondary treatment. The over, both primary and secondary burial are generally un-differen- Ghassulian ossuaries vary greatly in shape: there are rectangular tiated spatially. In addition, the burials tend to be located within or ‘chest-shapes’ or open-tub ossuaries which are usually made of adjacent to dwelling areas. It seems that long term coexistence of stone (Perrot and Ladiray, 1980, pp. 28–29: Group I), ‘house- the living and the dead was an important attribute of Neolithic ide- shaped’ ossuaries (Perrot and Ladiray, 1980, pp. 28–35: Group II), ology. During the Chalcolithic period secondary burials shifted rounded krater ossuaries, ‘urns’ or ‘domed jars’ (Perrot and Ladiray, from within habitation sites to off-site locales, explicitly and solely 1980, pp. 36–37: Group III). Considerable variability of the ossuar- dedicate to the dead. ies within and between burial caves is apparent. In most cases Ghassulian burial caves are the most common form of off-site several categories of ossuaries can be found in the same burial mortuary grounds in the southern Levant, although four sites with cave. Some burial caves include both ceramic and stone ossuaries burial structures are also known at Shiqmim (Levy, 1987), Palma- (e.g. Giv’atayim, Qula). Others, however, show less variability in him North (Gorzalczany, 2006), Kissufim Road (Goren and Fabian, the different types of ossuaries within the same cave (e.g. Mazor 2002) and Adeimeh (Neuville, 1930). The remarkable change in the West, in Khalaily and Marder, 2010, p. 15; Milevski, 2007). spatial aspect of mortuary behavior is very obvious but detailed Ossuary fragments have also been recovered from a few locational analyses of this phenomenon are yet lacking. It is the Chalcolithic sites, not in burial caves (e.g., Megiddo, Yehud, Tel goal of our study to shed some light on this aspect, and investigate Lod, see Table Exc. 1.2. in Van den Brink (2005)). what does the spatial change of the secondary burials from within The ossuaries are accompanied by ceramic vessels of various habitation sites to the off-site locations represent and what may types (such as ‘V-shaped’ bowl holemouth jars, goblets, cornets), have stimulate this change. few flint artifacts and, less frequently basalt vessels, hematite A brief description of the layout and geographical distribution maceheads, figurines, ivory and copper objects, all of which are of the cemeteries and their contents follows. This overview is com- known from habitation sites. In addition, recent excavations indi- plemented by discussing current interpretations of the mortuary cate that flint artifacts and ceramic vessel, including ossuaries were practices and their relevance to a better understanding of social locally produced (Cohen-Weinberger, 2011, p. 64; Khalaily, 2011, and economic issues of the Chalcolithic period. p. 62). Use of upright limestone slabs also known as mazzebot (ste- lae), have been identified in several burial caves such as Giv’atayim (Sussman and Beit-Arieh, 1966) and the mortuary site of Kissufim Chalcolithic secondary burial cemeteries in the Levant Road (Fabian and Goren, 2002 with references therein). Most burial caves are distributed along the coastal plain, be- Sukenik (1937) was the first to excavate a Ghassulian burial tween the modern towns of Hadera in the north and Gedera in cave that was quarried into the kurkar ridge at Hadera and was the south, and slightly further inland in the hills of the Shephella. filled with ceramic ossuaries, pottery and human bones. Since then, Very few burial caves have been found outside these boundaries, many burial caves have been discovered. In addition to the assem- and worth mentioning is the site of Peqi’in in the Upper Galilee blage of cemeteries excavated by Perrot and Ladiray (1980) at Azor (Gal et al., 1997, 1999). In a recent survey carried out further in- and Ben Shemen, worth noting are Nahal Qanah (Gopher and Tsuk, land, in the karstic regions of Jordan, no Chalcolithic burial caves 1991), Qula (Milevski, 2001), Sha’ar Efrayim (Van den Brink, were found (Lovell, 2009). 2011a), Shoham North (Van den Brink and Gophna, 2005), Ma’aba- Of special interest are the burial caves of Peqi’in (Gal et al., rot (Porath, 2006), Et-Taiyiba (Yannai and Porath, 2006), Peki’in 1997) and Nahal Qanah (Gopher and Tsuk, 1991), both of which (Gal et al., 1999), Palmahim (Gophna and Lifshitz, 1980), Giv’ata- contain mortuary assemblage with some unique findings. The most yim (Sussman and Ben-Arieh, 1966) and Horvat Qarqar South richly decorated ossuaries are those of Peqi’in, including painted (Fabian, personal communication). faces on façades and lids with human facial features and three Author's personal copy

426 R. Winter-Livneh et al. / Journal of Anthropological Archaeology 31 (2012) 423–438 dimensional heads (Gal et al., 1999). In Nahal Qanah cave, in addi- the wealth deposited in the tombs (Golden, 2009, p. 68). It should tion to V-shapes, jars, holemouth and churns, a unique group of be noted, however, that these propositions have not been examine eight gold and electrum rings was recovered (Gopher and Tsuk, yet. 1991). The significance of the spatial aspect to mortuary analysis Mortuary practices have frequently been used to reconstruct has been recognized and demonstrated by many researchers the nature of Ghassulian social complexity. Haas and Nathan (e.g. Bradley, 1998; Cannon, 2002; Clarke, 1977; Goldstein, 2002; (1973), who studied burials in the Judean desert caves, stated that Hutchinson and Aragon, 2002; Renfrew, 1976; Williams, 1999). the mortuary behavior reflects social hierarchy. Perrot and Ladiray Despite the growing number of Ghassulian secondary burial sites, (1980) interpret each of the coastal plain burial caves as a family the spatial component of these burials has received relatively little burial ground of a generally egalitarian society. Levy and Alon treatment by scholars (excluding Joffe, 2003). Moreover, very little (1985) and Levy (1986) based on his analysis of Shiqmim cemetery attention has been paid to the role these burial sites play in struc- I stone circles, attributes the burials to a chiefdom society. An turing and organizing past landscape, a role that most probably examination of the Shiqmim dataset have shown that the analysis goes far beyond their use as funerary sites. The archaeological evi- is problematic and that the nature of the stone circles do not sup- dence just presented indicates that the construction of these port the chiefdom model (Gilead, 1988b, p. 148). Moreover, since numerous caves involved considerable knowledge, effort, and most Chalcolithic burials lack prestige grave goods, the idea of costs. Their distribution along the kurkar ridges and the limestone clear social stratification cannot be maintained (Gilead, 1988a,p. hills, as well as the fact that they were recurrently used, shows that 429; Rowan and Golden, 2009, p. 68). Prestige artifacts of gold the exact location of such sites was preserved for generations by and electrum were found only at Nahal Qanah, a secondary burial community members. Their location outside and away from the cave (Gopher and Tsuk, 1996), but their relation to individual or settlements seems intuitively to imply attempts to conceal group burials as suggested by Gopher and Tsuk (1996, pp. 223– the burials, and therefore may have involved only a segment of 226) cannot be confirmed. Golden (2009, pp. 68–70) argues that the communities. But additional data indicates that this spatial burial caves in which ‘‘valuable luxury goods’’ are found, such as as well as conceptual knowledge was most probably shared among complex metal, ivory or basalt objects, indicate that the burial cave members of the communities. This is based on the frequency of was used by members of powerful elite lineages. Similarly, Gal individuals buried within each cave and the variability of the ossu- et al. (2011), argue that the finding from Peqi’in burial cave reflect aries within and between burial caves. Moreover, most Chalcolithic social hierarchy, and only individuals of higher social status were burials and cemeteries lack prestigious grave goods and are usually buried within the cave (Gal et al., 2011, p. 205). However, copper, accompanied by pottery vessels, flint artifacts and other objects, ivory or basalt objects are found in many habitation sites as well, similar to those known from habitation sites. The deposition of where no claims for activities of an elite can be established. Fur- funerary offerings similar to the domestic vessels, and the evidence thermore, considering the quantity of individuals buried within indicating local manufacture of ossuaries, pottery vessels, and flint some burial caves, such as Peqi’in, where at least 600 individuals artifacts, suggest that the caves were probably associated with the were buried, the quantity of such items is negligible, consisting nearby settlements. Understanding the relationship between the of one head of a small ivory figure, ten ‘violin-shaped’ figurines, habitation sites and the burial caves may reveal how and why their fragments of basalt bowls on fenestrated bases, four copper objects specific locations were determined and the way it is related to the (Gal et al., 1997, pp. 150–153), and 190 beads (Bar-Yosef Mayer socio-economic organization during this period. et al., 2004). Joffe (2003, p. 45) argues that the burials represent a religio-political ritual led by shamans. The role of shamanism Secondary burial in recent societies in the ritual behavior in the Chalcolithic period is acknowledged in other contexts too (Gilead, 2002, pp. 116–122). At this point, it is helpful to overview some social and economic Interpretation of the distribution of Ghassulian burial site has implications derived from studying ethnographic parallels where been problematic in the past, primarily due to what seemed to secondary burial in cemeteries is practiced. These examples, how- be a lack of residential sites in the central coastal plain. It was ever, are not used here as a direct analogy but merely to provide a therefore suggested that the burial sites there were used by noma- general interpretive framework and to broaden our interpretive dic pastoralists (Perrot, 1984; Perrot and Ladiray, 1980). The dis- horizons (and see Peregrine, 2001; Wylie, 1985). One of the best coveries of cemeteries at Shiqmim have invalidated this ethnographic description of secondary burials is the well-known suggestion. Levy and Alon (1987, p. 348) suggest, on the basis of ‘‘Feast of the Dead’’ (Thwaites, 1898, pp. 279–305), practice by ethnographic data, that secondary burials characterize sedentary the Hurons. The Hurons are native Americans who lived in Ontario rather than semi-nomadic societies. During recent years many until 1649 (Kidd, 1953; Tooker, 1964; Trigger, 1969, 1976). Accord- habitation sites have been discovered in these regions (Gophna ing to the Jesuit priest, Jean de Brébeuf, who witnessed and re- and Portugali, 1988; Van den Brink, 2008; Van den Brink and corded the ‘‘Feast of the Dead’’ during the spring of 1636, every Gophna, 2005). Thus, it is currently agreed that the burial caves 10–12 years all the contents of the graves of individuals who had were used by sedentary communities living nearby (Khalaily and died unexceptionally and who were not too old or too young at Marder, 2010, pp. 19–20; Porath, 2006, p. 58–59; Van den Brink, the time of death, were transported from their individual graves 2011a, pp. 45–46; Van den Brink and Gophna, 2005, pp. 168– and placed in a large ossuary for communal reburial. The ossuary 169; Yannai and Porath, 2006, pp. 1–2). was a large pit into which all the corpses or bones were deposited Previous studies propose that the formal cemeteries of the Chal- and arranged around three large kettles intended for the use of the colithic period located at sites distinct from habitation sites reflect deceased souls (Kidd, 1953, pp. 372–375; Robb, 2007, pp. 290– a concern for territoriality (Levy, 1986, 1995). In addition, Joffe 291; Thwaites, 1898, pp. 279–305; Trigger, 1976, pp. 88–91). (2003) suggests that the visibility of cemeteries, like Shiqmim in Nonetheless, secondary burial practice is by no means a legacy the northern Negev and Adeimeh near Teleilat Ghassul in Jordan, of past societies. Up-to-date descriptions of secondary burial prac- is a statement of territorial claims while the majority of cave buri- tices can be found even in daily newspapers, such as in the case of als in the coastal plain area that ‘‘stand out for their invisibility on Ambohimirary, Madagascar (Bearak, 2010). A different but a rele- the living landscape’’ reflect a less public praxis (Joffe, 2003, p. 51). vant example of reburial of bones in recent societies can be found Golden (2009) suggests that the burial caves served Ghassulian in Hawaii: a patent that has recently been published consists of an elites who used the remote and concealed locations to protect apparatus and method for removing the flesh from the bone prior Author's personal copy

R. Winter-Livneh et al. / Journal of Anthropological Archaeology 31 (2012) 423–438 427 to burial (Nihipali, 2007). A recent cross-cultural survey shows that people buried’’ (Glazier, 1984, p. 142). Accordingly, burials location 61 (32.7%) out of 186 current societies practice secondary burial, became a fundamental part of recording the relationship between and its intensity is similar in all regions of the world (Schroeder, the lineages and the land. ‘‘That is, a grave site now establishes a 2001, pp. 85, 87, Tab. 1). visible connection between a particular territory and forebearers Such a widely distributed phenomenon has triggered different buried within it, there by forging new and socially valued links be- ideological, sociological and economic interpretations. Hertz tween the land and its claimants’’ (Glazier, 1984, p. 144). (1960 [1907]) wrote one of the important synthesis of secondary Other societies in Kenya, which strongly identify ancestors’ burials and influenced many archaeologists and anthropologists. graves with particular land, are the Nilotic, Kenyan Luo. The Luo His study of the Dayak of Borneo (Old Ngaju) stressed that the are the third largest ethnic group in the country. They consist of transfer of the remains to their final burial place is not a mere segmentary lineages, practicing mixed farming and herding. change of place, but rather a release of the deceased from the iso- According to Shipton (1984, pp. 123–124) ancestral graves are lation into which he was plunged with death and the reunification the most important fixed points in a Luo’s life, and the land associ- of his soul, embedded in the bones, with his ancestors who reside ated with them is regarded with considerable reverence’’. Popula- in the communal ossuary (Hertz, 1960, p. 55). Hertz compared the tion growth during colonial years has promoted rising concern Dayak rituals to parallel practices in other societies, such as the towards landholdings, and involved ’’bitter feuds over fields and above mentioned. He emphasized that secondary burial among boundaries, using ancestral graves as the most important fixed the Hurons is a rite performed by the larger community and not points’’ (Shipton, 1984, p. 127). by the family (Hertz, 1960, pp. 70–71). Another example can be found in the central Ugandan region of Other scholars, however, argued that Hertz failed to address Buganda. During pre-colonial times, the lands of Buganda were important economic aspects of the rites, suggesting that secondary owned by the 52 clans (Green, 2006, p. 373). The Ganda clans of burial is based upon accumulation of wealth needed to perform the Buganda, for example, established, during the 16th-century, a rela- burial rites (Metcalf, 1981; Miles, 1965; Poyil, 2009). Nevertheless, tionship of obligation between the ancestors and the leaders the same argument applies to any of a whole host of non-second- (chiefs) and their followers (Mair, 1933). The presence of graves ary memorial ceremonies for the dead that are conducted months was reported as an additional dimension to the manner in which and sometimes even years after the primary funeral. Thus, the out- the individual Ganda clans remembered their ancestors (Hanson, standing feature of this rite, moving the deceased remains from the 2003, p. 34). The clan’s primary estate, a public property, consisted temporary grave to its final resting place, remains unexplained. mainly of two features – a sacred site and burial grounds (Kodesh, 2008, p. 204). Burial location and land tenure in recent societies Meggitt (1965a,b) studied the Mea-Enga in highland New Gui- nea and argues that pressure on agricultural land was linked to In order to understand the shift from pre-Chalcolithic burials descent groups’ members who held the rights to high quality agri- within the settlement to Chalcolithic off-settlement burials, we ad- cultural land. Therefore, the descendants of the dead seek to affirm dress here the wider issues of practical and symbolic aspects of the their group membership by ritual that would connect them to the location of burial sites in recent societies. Many ethnographic stud- group ancestors. ‘‘On the one hand, rituals regularly reaffirm the ies indicate that locations of cemeteries are associated with land cohesion and continuity of the patrilineal group; on the other, tenure rights of communities. Goldstein (1981, p. 61) following the dogma in itself implies title to land by relating living members Saxe (1970), regards territorial claims as ‘‘...group rights to use of the group to a founding ancestor who is believed to have first se- and/or control crucial but restricted resources’’. In order to clarify lected that locality for settlement’’ (Meggitt, 1965b, p. 131). An- and simplify this claim of territory, a claim that is implemented other ethnographic example from Papua-New Guinea is the by actually observing the landscape, and not to be mixed with community of the Rumu. The Rumu practice secondary burial that other territorial behavior, that typically is associated with areas consists of ‘centralized clan ossuaries’ which are burial caves. which are carefully marked (using fencing) and/or defended by or- These burial caves are known as ‘‘...ancestral claims to territory’’ ganized warfare, we use here the term land tenure. Land tenure, (David et al., 2008, p. 163). As a result, the Rumu landscape is based on Adler’s definition (1996, p. 338), indicates a strategy to understood by community members ‘‘not only by reference to spir- socially define the realms and rights to access and use the natural its...but as spiritscapes’’ (David et al., 2008, p. 165). resources scatted in the landscape. The resources may range from The ethnographic data provide valuable information concern- arable lands to grazing pastures, from transportation roads to fruit ing the location of burials and their territorial implications. It trees. Accordingly, the relocation of the deceased into his/her final demonstrates that locations of graves and cemeteries play an resting place determines and reinforces the rights of the commu- important role in establishing or preserving land tenure. They nity over a given area (Bloch, 1989; Bradley, 1998; Earle, 1991; provide direct evidence for land tenure and territorial behavior. Fleming, 1973; Goldstein, 1981; Renfrew, 1973b; Saxe, 1970). Naturally, not all societies use burials as a physical marker of The following ethnographic examples demonstrate the role of the land tenure. Other markings of the landscape such as walls, locations of ancestors’ graves as markers of groups’ rights over cer- rock piles, stone mounds sometimes in lines, ditches, hedge- tain territories. rows, rock art or other non-burial monument such as shrines, This point was stressed by Bloch (1982) in his analysis of the are known in the ethnographic and archaeological research to Merina people of central Madagascar: ‘‘The presence of the tombs have been use as means for declaring land tenure (Adler, of a certain deme [kin-group] in an area is what makes this area the 1996; Bintliff, 1982; Gleason, 1994; Johnson and Earle, 1987; ancestral land of that deme’’ (Bloch, 1982, p. 213). Moreover, in Stone, 1994). Mozambique, ancestral burial location signify a legitimate claim The present study, thus, aims to find whether Ghassulian sec- for land acceptable by state administration (Unruh, 2006, pp. ondary burials were used in a manner similar to those observed 767–768). in the ethnographic record, as landscape markers in a land tenure In Mbeere, a Bantu people of east central Kenya, lineages have system. In addition, we discuss the implication of our analyses re- customarily exercised corporate control over the land. Overall rural sults in relation to previous studies and their reconstructions of development in their region promote seeking critical indicators for Ghassulian social organization. We use GIS base analyses in order proving ones land ownership, reinforcing ‘‘the importance of iden- to assess the possibility that the innovation of secondary off-site tifying the forebearers, who, in many cases were among the first cemeteries during the Chalcolithic may have been, at least partly, Author's personal copy

428 R. Winter-Livneh et al. / Journal of Anthropological Archaeology 31 (2012) 423–438 motivated by a necessity to enhance communities’ control over its understanding the off-site location of Ghassulian secondary buri- landscape to establish land tenure. als. Being similar to secondary burials of traditional societies, it im- plies that it was one of the ways to manifest the rights of Land tenure, control and visibility Ghassulian communities over the landscape.

But how can one actually evaluate a social behavior such as con- trol? In archaeological landscape research, as well as in social stud- Dataset and study area ies, visibility or field of visibility is considered by many cultures as one of the means to establish territorial rights (Brighenti, 2007; There are 974 Chalcolithic sites according to the Israel Antiquity Llobera, 2003, 2007; Wheatley, 1995). Broadly speaking, the com- Authority (IAA) GIS MENORA database, which provides a solid and plex phenomenon of the field of visibility can be understood as a relatively well researched and updated body of data. A thorough relevant symbol of control or recognition for the social relation search of the database and additional archival records resulted with within the studied culture (Brighenti, 2007, p. 324). Accordingly, 53 burial sites (5.4%) (Fig. 2). Van den Brink (Van den Brink, 1998, many studies have shown the implications and approaches to the 2005) has recently prepared a list of 47 such sites. His list includes subject of the field of visibility, as a mode to enforce control or sur- the site of ‘Qurnat Haramiya (MizpeAfek)’ (van den Brink, 2005: veillance (Brighenti, 2007; Eason and Stamps, 1992; Foucault, 1995 Table Exc. 1.1) which lacks clear evidence of Ghassulian secondary (1975), pp. 195–199; Renfrew, 1976; Yekutieli, 2006). burials since no ossuary fragments were found as is stressed by It has been demonstrated that the location of features (such as the site excavator (Torge personal communication). Furthermore, settlements or cemeteries) can be explained by visible control over burial cave sites Nos. 25–28 in the list of van den Brink (2005 the landscape (Antrop, 1988; Sevenant and Antrop, 2007). A recent Table Exc. 1.1), located in the vicinity of Shoham, were registered example of an early application of visibility control was demon- by us as one burial site named ‘‘Shoham North’’. The additional bur- strated by Yekutieli (2006). According to his analysis, during EB ial sites included in our inventory are the recently discovered burial II-III and the Roman periods ‘‘people in a desolate place such as sites of Horvat Qarqar South and Palmahim North. Although these 53 the Judean Desert were capable of using the panoptic effect burial sites are spread over a large area of 21,000 km2, we restrict our through an intelligent management of their natural environment’’ study to the burial caves of the coastal plain and the Shephella (ca. (Yekutieli, 2006, p. 85). 2484 km2) since here more than 50% of the sites (n = 29) are concen- GIS environment provides the ability of assigning concepts such trated within 10% of the total area. as control or recognition into a quantitative value by using view- To study the spatial relationship between the burial sites and all shed analysis. Viewshed are considered to represent cognitive per- other sites within this study area, we selected out of the IAA MENO- ceptual acts which structured and organized the location of RA database the burial sites within this study area and an additional cultural features (such as settlements and burial sites) as well as group of their neighboring habitation sites. The habitation sites have their inhabited landscape (Wheatley and Gillings, 2000) p. 3). been chosen according to two criteria: (1) The assessments of the Viewshed have been frequently assigned as an integral part of burial sites excavators and researchers who consider them to be re- the landscape and proven to be a useful tool in understanding spa- lated to the burial sites. Porath (2006, pp. 58–59), for example, tial distribution of archaeological sites (Fry et al., 2004; Kvamme, writes: ‘‘it seems that these burial sites [Ma’abarot and Tel Ifshar] 1990; Llobera, 2003, 2007, 2010; Sevenant and Antrop, 2007; served the communities who lived in the lower basin of Nahal Alex- Wheatley, 1995; Wheatley and Gillings, 2000). ander’’. (2) Their proximity, in terms of nearest neighbor analysis, to Specifically, we start with examining the burial sites spatial pat- the burial site in the cases the excavator did not mention a possible tern in relation to the rest of the sites spatial pattern. That is, quan- relationship. Five burial sites (Palmahim, Palmahim North, Benaya, titatively sustain the assumption that Chalcolithic burial grounds Yavneh, Tel Lod) were excluded from our sample since they are rel- were indeed located in an off-site location away from habitation atively isolated compared with other burial site in the research sites areas. Subsequently, in order to examine whether secondary study area. Thus, the dataset consists of a sample of 24 burial sites burial are related to land tenure system, we use a simple set of and a sample of 24 of their nearest neighboring habitation sites. Both assumption and constraints on our viewshed analyses. It is reason- are referred to below as the study group (Fig. 3). able to propose that areas which are visible from habitation sites We wish to note that in comparison with the burial sites most are most likely to have been included as parts that are controlled of the paired habitation sites within the study group have not been by the sites. Furthermore, terrain which is visible from a burial site excavated and mainly consist of survey data. This raises the ques- can also be regarded as a controlled area. However, since areas tion whether the paired sites are contemporaneous? In the coastal which are visible from a particular observation point represent plain and the Shephella region the sites of the Chalcolithic period the areas from which the site (habitation or burial) is visible, our are of the Ghassulian culture and sites of cultural entities such as assumption does not suggest that the burial sites were intended the Besorian, the Timnian or the Golanian are unknown in this part to be used as actual watch towers of any kind. It is those areas, of the country. Moreover, during recent years and thanks to sal- within the landscape, from which the burial sites locations are vis- vage excavations carried out by the IAA in this area, it becomes ible, that are of particular interest. What influences the society is more and more evident that there are more habitation sites than the likelihood of recognizing and remembering the burial site as previously known and they are all Ghassulian. A good recent exam- a landscape feature, a physical marker in the landscape. The ques- ple can be found in Van Den Brink excavations at the site of Namir tion is if such landscape features could be recognized. Were they road in Tel-Aviv (Van den Brink, 2006, 2011b). visible from the surrounding arable lands near the habitation sites? To evaluate the significance of the results based on examining the Could both the habitation sites and the burial sites be observed at study group, we created a matched control group. The results obtained the same time (overlapping viewsheds)? And if so, to what extent? from studying the spatial relationship within the study group were Subsequently, we will examine whether the areas visible from bur- compared to the results which were gained from analyzing the ial sites were organized so that they could complete or enhance the matched control group. The latter consists of a random selection of visible area seen from habitation site (contributing viewsheds) and non-burial sites (n = 24) out of the main database of such sites thus establishing together (the burial location and the habitation (n = 921) and within the research area, named here sample 1 (com- location) a continuous visible landscape within the habitation sites parable to the area of the burial sites dataset). An additional match- surrounding. Revealing the extent of the visible landscape helps ing set of 24 sites, named here sample 2 (comparable with the Author's personal copy

R. Winter-Livneh et al. / Journal of Anthropological Archaeology 31 (2012) 423–438 429

Fig. 1. Location of sites mention in the text: 1. Eynan; 2. Peqi’in; 3. Hayonim Cave/Hayonim Terrace; 4. Nahal Oren; 5. Kefar HaHoresh; 6. Mishmar Ha’emeq; 7. Hadera, 8. Tel Ifshar, 9. Ma’abarot; 10. Sha’ar Efrayim; 11. Et-Taiyiba; 12. Nahal Qanah; 13. Qurnat Haramiya (Mizpe Afek); 14. Bene-Berak; 15. Giv’atayim; 16. Tel-Aviv; 17. Azor; 18. Shoham North; 19. Tel Lod; 20. Qulah; 21. Ben Shemen; 22. Palmahim North; 23. Palmahim; 24. Yavneh; 25. Benaya; 26. Teleilat Ghassul; 27. Adeimeh; 28. Horvat Qarqar South; 29. Kissufim Road; 30. Shiqmim; 31 Nahal Hemar. habitation sites dataset), was selected based on the site proximity to and 25 m horizontal resolutions (Hall, 1993). The research study sample 1 sites, by using nearest neighbor analysis and limited to the area includes the coastal plain and the Shephella (Fig. 3). The coastal same ranges of distances that were measured between the burial plain is topographically a low land belt that stretches from the Med- sites and the non-burial sites of the study group. To validate measure- iterranean Sea shore to the eastern edge of the Shephella. This land- ment compatibility the distances between the two datasets were scape is relatively wide in the south and center of Israel (20–40 km), compared by means of a t-test analysis which has resulted in and narrow in the north of the country. The Shephella is located east p > 0.503. An additional way to validate measurement compatibility of the coastal plain, slightly inland in the low piedmont foothills. It is was to compare the two datasets of elevations by using the t-test a narrow, hilly region which separates the southern coastal plain which resulted in p > 0.05. These results indicate that the elevation from the Judean Mountains. It extends from the vicinity of Beer She- and distances between the sites of the study group and the elevation va in the south to the environs of Lod in the north. Its eastern higher and distances between sites of the matched control group are similar. part (380–500 m) consists of flat hilltops. The western lower part Topographic information on the research area was measured (130–270 m) contains broad and leveled valleys, frequently attain- and quantified using a contour-based DEM layer with 10 m vertical ing the aspect of foothill plains. The valleys and plains, and Author's personal copy

430 R. Winter-Livneh et al. / Journal of Anthropological Archaeology 31 (2012) 423–438

Fig. 2. Locations of Chalcolithic burial (n = 53) and non-burial (n = 921) sites in the Fig. 3. Location of burial sites sample (n = 24) and their nearest neighbor habitation Southern Levant (Israel). sites (n = 24). frequently even the plateau-like hilltops, are covered with fine- grained, alluvial sediments (Singer, 2007, pp. 23–26).

Methodology

Our methodology consists of two stages: (i) Kernel density spa- tial analysis; (ii) viewshed analysis. A simple way to explain the methodology is to describe it as a process in which we first map the locations of all burial sites within the study area and compare them to the locations of all non-burial sites in order to statistically sustain the assumption that during this period secondary burial migrate to off-site locations. The second stage involves mapping and measuring viewsheds, the areas within the landscape which are visible from (or observe) each of the burial sites and compare it to the areas within the landscape which are visible from (or ob- serve) it’s nearest neighboring habitation site. Next, we examine how much of the viewshed of each pair of sites (the burial and the nearest neighboring habitation site) overlap and what is the Fig. 4. Illustration of habitation site (yellow) and burial site (red) viewsheds, within a limited radius of 5 km each. The overlapping area is marked in orange color. The contribution of the burial site to the broadening of the area viewed contributing area is marked with dashed contour line colored yellow. (For from the habitation site (Fig. 4). Furthermore, in order to evaluate interpretation of the references to color in this figure legend, the reader is referred the significance of the viewshed measurements, we compare them to the web version of this article.) to the viewshed measurements abstained from a matched control group, that is, from pairs of non-burial sites that are located within similar geographical conditions. The following is a detailed points (e.g. sites) in the neighborhood around them. The aim of description of these methodological processes. using this spatial analysis is to identify locations with concentra- tions of burial sites, and concentrations of non-burial sites. Kernel Spatial analysis density creates a continuous raster surface output from the point (i.e., site) dataset. In this surface output the cell value is highest Kernel density is a method used to construct an estimation of a at the location of the point and diminishes with increasing distance relative density from a given dataset. It calculates the density of from the point, reaching zero at the search radius distance from the Author's personal copy

R. Winter-Livneh et al. / Journal of Anthropological Archaeology 31 (2012) 423–438 431

Fig. 5. Sampling grid of 2 2 km cell size spread over the kernel density map of non-burial sites using search radius of 3 km. point. The kernel function is based on the quadratic function de- scribed by Silverman (1986, p. 76, Eq. 4.5). Since the site datasets have no field value other than location, NONE has been specified, and therefore the volume under the surface equals one. The density at each output raster cell is calculated by adding the values of all Fig. 6. The relative density locations of (a) habitation sites; (b) burial sites; (c) habitation and burial sites. the kernel surfaces where they overlay the raster cell center. Increasing the search radius will not greatly change the calculated density values. Although more points will fall inside the larger neighborhood, this number will be divided by a larger area when calculating density. The main effect of a larger radius is that den- sity is calculated considering a larger number of points, which can be further away from the raster cell, this results in a more gen- eralized output raster (Mitchell, 2005; Silverman, 1986). To examine the correlation between burial sites density pattern and habitation sites density pattern, kernel density analysis was applied to the burial site dataset (n = 29) and habitation site data- set (n = 117). All are located in the study area, along the coast plain and in the Shephella (ca. 2484 km2). We used 1 km, 2 km, and 3 km search radius. We generated vector grids using the sampling tools

Fig. 7. Mean and standard error bar of viewshed areas of burial and habitation sites. Table 1 Correlation coefficients of kernel density of burial sites versus kernel density of non- burial sites. All results are of statistical significance of p < 0.05. of Hawth’s analysis tools version 3.26 (Beyer, 2004). All sampling grids covering the research area are equal in their size, but differ Grid (cell size) in their cell sizes. To avoid any possibility that the correlation re- 1 1km 2 2km 3 3km sults will be influence from the sampling grid cell size, we repeated Kernel density (search radius) the sampling of the research area by using three cell sizes of 1km 0.515 0.462 0.496 1 1km(n = 2484), 2 2km(n = 632), and 3 3km(n = 275). 2km 0.592 0.528 0.638 Then, we calculated the zonal statistics of the kernel density raster 3km 0.749 0.726 0.785 layers within each cell of each of the grids (Fig. 5). A correlation Author's personal copy

432 R. Winter-Livneh et al. / Journal of Anthropological Archaeology 31 (2012) 423–438 test was conducted on the zonal statistics results from all the sam- same procedures were attributed to the matched control group data- pling grids between the burial site kernel density layers and the sets. The overlapping visible areas and the contribution visible areas non-burial site kernel density layers. However, since most areas of the two groups were compared by means of t-test analyses (for of the kernel density maps are empty of sites, that is, equal zero, additional information on overlapping viewshed or co-visibility the correlation was made only on grid cells with a zonal statistical see Llobera, 2007, 2010). Our analysis is based on viewshed area average (of the non-burial site kernel density layer or the burial relations between a single burial site and a single habitation site site kernel density) greater than zero. The Kernel density and Rip- (its nearest neighbor). This minimizes confounding variables, and ley’s K-function analyses were carried out using ArcGIS 9.3. enables to examine the phenomenon significance. Further research should include testing viewshed area relations between several Viewshed analysis sites and each burial site (one-to-many relationship) to test the likelihood of other assumption that these burial sites may have To test our ethnographically based assumption concerning the served more than the nearest site. All viewshed analyses were gen- potential role of the burial sites location in terms of the enhancing erated using the ModelBuilder Iteration processing tool of ArcGIS controlled visible landscape areas of Ghassulian communities, we 9.3. first calculated how much can be seen from each burial location in relation to the area size that is visible from each of the habita- Results tion site by using viewshed analysis. Viewshed analysis calculation, given a digital elevation model Statistical analyses (Table 1) indicate significant negative corre- (DEM), requires a straight line be interpolated between the observe lation between kernel density maps of the burial sites and kernel point (i.e., site) and every other cell (pixel) within the elevation density maps of the non-burial sites. Results indicate that the size model. The elevation of all the cells which occur on the straight line of the grid cell has relatively little effect on the correlation results. between the observing and the target cells can then be obtained in Using different search radius, however, shows that the greater the order to ascertain whether or not the cell exceeds the height of the search radius distance is the stronger the correlation it will have. line. The result of each of these calculations is either positive The important outcome is that in all cases correlation are negative, (coded 1) for a visible cell or negative (coded 0) for a cell that is indicating that non-burial site clusters locations are correlated not visible. Calculated for the entire raster layer it produces a bin- with decline in burial sites density estimation. In other words, ary image with those areas of the landscape that have a direct line the location of burial sites tends to stay away from areas with of sight from the observed point and those areas with no line of non-burial sites (Fig. 6). sight (Wheatley, 1995, p. 171; Wheatley and Gillings, 2002, p. 205). Fig. 7 shows a descriptive graph of the mean and standard error Furthermore, we examined whether the visibility of the sur- of the visibility area size from the study group datasets. Within a rounding area from each burial site interact with the neighboring maximum radius distance of 5 km the average size of visible area habitation site by adding additional visible area within its sur- observed from burial sites (11.4 km2) is clearly larger than the rounding landscape. At first, we calculated the visible area (view- average size of visible area observed from habitation sites shed analysis) around each incident from the study group (7.5 km2). The two dataset areas were compared using a student datasets and each incident from the matched control group datasets. t-test analysis and has resulted in p < 0.04, indicating a significant The target area was limited to a maximum radius of 5 km distance, difference. This result indicates that the community burial sites excluding the sea. A vertical distance (offset) of 2 m was added to had larger areas of visible control. each observation point height value in all incidences in each of the Fig. 8 shows that the contributing visible area within the study datasets group in order to simulate human visual ability over the group is substantially larger (92%) than the contributing visible topography. A comparison between the visible areas was made area within the matched control group (46%). A comparison of the by means of a student t-test analysis. In the second stage, a buffer two contributing areas using student t-test analysis has resulted zone of 5 km radius distance was interpolated around each of the in p < 0.004 indicating the areas are significantly different. This re- habitation sites. Within each of these buffer areas, we calculated sult indicates that within a radius of 5 km from the habitation site the contributing and the overlapping viewshed areas of each burial an average of 92% contribution of visible area from the nearest bur- site to its nearest habitation site (Fig. 4). Overlapping viewshed area ial site is added to the habitation site visible area (for examples see is the normalized percentage of visible area (viewshed) observed Fig. 10a and b). from both the specific burial site location and its nearest neighbor- Fig. 9 shows that an average of 37% overlapping visible areas is ing habitation site: produced within the study group datasets and 25% overlapping vis- Vhi \ Vbi ible areas are produced within the matched control group dataset. Oi ¼ 100 ð1Þ Vhi These overlapping areas were compared by means of student t-test analysis, resulting in p < 0.18. This indicates that overlapping areas where O is the overlapping visible area of the i0 couple of burial site i measured within the study group datasets and the matched control viewshed (Vb) and habitation viewshed (Vh) sites. Note that the raw group datasets are similar, while their contributing areas are signif- overlapping area (the numerator of equation number 1) is divided icantly different. by the raw area of the habitation site, which normalizes C to per- centage units. The Contributing area is the normalized percentage of additional visible area contributed by the burial site to the visible Discussion area from the habitation site within its buffer zone area of 5 km dis- tance radius: Given the locations of secondary burial sites and their relation to habitation sites, our study shows a pattern of visibility control ðVh [ Vb ÞVh C ¼ i i i 100 ð2Þ over the landscape. This pattern probably reflects attempts to so- i Vh i cially constitute and mediate ownership overlarge areas of the 0 where Ci is the contributingof visible area of the i burial site view- landscape as part of a land tenure system. shed Vh to the i0 habitation site. Note that the raw contributing area Correlation analyses between the locations of the relatively (the numerator of equation number 2) is divided by the raw area of dense clusters of non-burial sites and burial sites have produced the habitation site, which normalizes C to percentage units. The significant negative correlations. These results indicate that the Author's personal copy

R. Winter-Livneh et al. / Journal of Anthropological Archaeology 31 (2012) 423–438 433

ever, it is difficult to explain the establishment of off-site cemeter- ies during the Chalcolithic period solely as a mean to reaffirm social identity. As demonstrated above, secondary burial rituals as a mean of social integration were performed during the Neo- lithic period within the habitation sites. Moreover, there are ethno- graphic indications that mortuary practices signify important economic and political aspects beyond social integration (Chénier, 2009; Metcalf, 1981; Metcalf and Huntington, 1991; Miles, 1965; Poyil, 2009). As Goldstein (2008, p. 189) notes: ‘‘...in contrast to primary burial, secondary burial may have lit- tle to do with death per se. The rite may come a year or even several years later, and it is triggered not by the death of the individual being afforded the treatment but by some other Fig. 8. Mean and standard error bar of the percentage contributing areas of the event.’’ study group sites and matched control group sites.

There are archaeological records of secondary burials in the southern Levant both before and after the Chalcolithic period. Mor- tuary practices of the PPNB period, for example, indicate that to reaffirm social identity secondary burials, and probably the rites involved, were carried out inside habitation sites (e.g. Kuijt, 1996, 2001, 2008; Kuijt and Goring-Morris, 2002). The spatial as- pect of Chalcolithic period secondary burials cannot be overem- phasized. It is practically the first time in the archaeology of the southern Levant that cemeteries shift in a systemic manner from within settlements to off-site locations that are exclusively dedi- cated to burial. Secondary burial in ossuaries within burial caves was practices in this region during the Roman period (Hachlili, 2005, p. 522). Most of the ossuaries have been found within Jerusalem and its Fig. 9. Mean and standard error bar of the percentage overlapping areas of the study group sites and matched control group sites. surroundings, especially during the period of city prosperity until its destruction in 70 BC (Rahmani, 1978, p. 110). The burial caves were quarried into limestone hills or arable lands. A cave with location of burial sites tends to avoid the location of non-burial ossuaries usually represent burials of a family (Regev, 2002, pp. sites, i.e., clusters of burial sites will most likely be placed away 59–60) in communities of farmers or other sedentary settlers. In from habitation sites. This substantiates our assertion that Chalco- such communities the collective identification is primarily with lithic secondary burial grounds are off-site cemeteries, located the ancestors and their land property (Regev, 2002, p. 50; Rubin, away from habitation sites. This represents a phenomenon that dif- 1994, p. 258). fers considerably from the secondary burials of pre-Chalcolithic During the Chalcolithic period an increased demand for arable times, when burials were within the habitation sites (but see above land due to population growth and reliance on agriculture may the claims of Goring-Morris concerning Kfar HaHoresh). have stimulate a need to signal claims over arable lands. Arable The relationship between the habitation sites and the burial lands were undoubtedly claimed also in later periods, but such caves is apparent when analyzing the spatial pattern. Our results claims were made in different ways (Adler, 1996; Earle, 1991, show that the size of the terrain seen from burial sites (field of vis- 2000). Territorial behavior, in regard to exploitation of resources, ibility), is significantly larger than the size of the terrain seen from is an aspect that was first introduced to archaeological prehistoric habitation sites (Fig. 7). Furthermore, we found that these larger research by Vita-Finzi and Higgs (Bintliff, 1999, p. 505). Accord- viewshed areas that are seen from the burial sites are located ingly, territorial behavior is assumed to have existed in all prehis- mostly adjacent to the habitation sites. When examining the per- toric and historic periods. Some scholars argue that humans are centage of the visible contributing area (Fig. 4) from the burial sites territorial by nature (Ardrey, 1966; Cohen, 1976) or that some pat- our results show it almost doubles the size of visible area sur- terns of territorial control are apparent even among bands of hun- rounding the habitation site, in contrast to the match control group ter-gatherers (Cashdan, 1983, 1989; Dayson-Hudson and Smith, (Fig. 8). This indicates that the larger viewshed area, seen from the 1978; King, 1975; Peterson, 1975; Wilson, 1975). burial sites, primarily affected the contributing visible area and not The idea of territorial control is also significant in Saxe’s the overlapping visible areas (Fig. 9). Burial caves have been located hypothesis 8 (1970, p. 119) which is based on a hypothesis set in areas which are visible from a significantly larger part of the forth by Meggitt (1965a, p. 279). Goldstein’s (1981, p. 61) restate- landscape than their nearest settlements, thus enlarging the size ment of Saxe’s hypothesis 8 suggests that burial grounds represent of the terrain that can be viewed. We suggest that this pattern indi- corporate groups that control crucial resources. ‘‘This corporate cates that Chalcolithic cemeteries played a role in claiming land control is most likely to be attained and or legitimized by means tenure, similar to what have been described in the ethnographical of lineal decent from the dead...of the critical resources passing record. from parent to offspring.’’ Goldstein notes that Saxe changed Ethnographic and archaeological studies that follow Hertz’s Meggitt’s concepts of ‘‘’land’ to ‘vital resources’ and ‘agnation’ to (1960) view argue that the role of secondary burial rites is to affirm ‘lineal descent’’’ (Goldstein, 1981, p. 60). Meaning that the nature the social identity and establish community cohesion (Hertz, 1960; of the crucial or vital but restricted resources has been generalized Kidd, 1953, pp. 372–375; Kuijt, 2008, p. 186; Robb, 2007, pp. 290– in order to make the hypothesis cross-culturally applicable. In the 291; Thwaites, 1898, pp. 279–305; Trigger, 1976, pp. 88–91). How- archaeological research ‘crucial but restricted resources’ refers Author's personal copy

434 R. Winter-Livneh et al. / Journal of Anthropological Archaeology 31 (2012) 423–438

Fig. 10. Examples of two burial sites and their nearest neighboring habitation sites viewsheds areas: (a) Hadera; (b) Giv’atayim.

mostly to land or agricultural land (Bintliff, 1999; Earle, 2000; found with the description of Jean de Brébeuf (1593–1649) who Gerritsen, 2003). participate in the Huron’s secondary burial ceremony known as The Ghassulian way of claiming rights over the land was prob- the ‘‘Feast of the Dead’’ in 1636: ‘‘...the old men and notables of ably based also on gatherings of the community during secondary the country assemble, to deliberate in a definite way on the time burials ceremonies, when land rights may have been negotiated, at which the feast shall be held to the satisfaction of the whole defined and asserted. The specific part of the terrain that was cho- country and of the foreign nations that may be invited to it’’ (Thwa- sen for the burial cave, be it a particular kurkar ridge or limestone ites, 1898, p. 279). Moreover, the archaeological evidence indicates hill, turned into a landmark which was understood by the local that the number of individuals buried within each cave during the communities. The burial caves locations created ‘‘spiritscapes’’ Chalcolithic has reached well into the hundreds in several cases. (David et al., 2008; McNiven, 2004), landmarks which are identi- Thus, it is highly probable that the location of the burial caves fied with a spiritual and political power, similar to the way Gros- was known to the entire community. In addition, there are no indi- man (2003, p. 577) interprets the role of the Natufian Hilazon cations that a foreigner population penetrated this region. Tachtit cave. Changing recurrently the landscape with burial caves One of the archaeological primary interests in the study of mor- defined patterns of group relationship and rules of inheritance that tuary remains is due to the fact that mortuary behavior relates clo- related to land use and/or access rights. sely to the organization of society (Binford, 1971; Goldstein, 1976; The pronounced variability of the caves as well as the nature of O’Shea, 1984; Saxe, 1970; Ucko, 1969). There are researchers who the social units that used them is still poorly understood. Although view the Chalcolithic burial grounds as reflecting ‘‘elite lineages, it has been argued that meanings and ideas behind burial practices extended families and kin groups that emerged at the top of a nas- may differ from those of the living community (e.g. Bloch, 1971; cent social hierarchy’’ (Golden, 2009, p. 70), or ‘‘...reflect social Hodder, 1982; Ucko, 1969), it is possible that each cave and its hierarchy...individuals buried in the Peqi’in cave were of higher receptacles represent the social persona of the group that its mem- social status among members of the Chalcolithic community/com- bers are buried in it. An example of a recent attempt to identify so- munities.’’ (Gal et al., 2011, p. 205). Levy (1986, 1995), citing Saxe/ cial entities is the study of Nativ and Gopher (2011) who suggest Goldstein hypothesis, argues that the formal burial ground signi- that there were two regional ‘cemetery groups’ a southern and a fies a ‘‘hereditary chiefdom society’’ (Levy, 1995, p. 235). His asser- northern one. tion is based on the anthropological models of Service (1962) and We wish to emphasis that the territorial behavior of the com- Fried (1967) and on their archaeological correlates listed by Ren- munity was primarily aimed toward regulating intra-community frew (1973a). However, the archaeological evidence from cemeter- affairs rather than reaction to foreign intruders as is the case of ies supporting the idea of a socially ranked society hardly exists the Megalithic monuments of Europe (e.g. Renfrew, 1976). This is (see introduction; Gilead, 1988a, 1988b; Perrot and Ladiray, based on indications that most, if not all community members par- 1980, p. 131). Furthermore, Ghassulian settlement patterns and ticipated in the mortuary ceremonies. A good example for the house shapes and sizes, cast doubts on Levy’s chiefdom model number of the participants in secondary burial ceremony may be (Bourke, 2001, p. 151; Gopher and Tsuk, 1990, p. 15). Golden’s Author's personal copy

R. Winter-Livneh et al. / Journal of Anthropological Archaeology 31 (2012) 423–438 435 and Gal et al. propositions that the burial caves were only for elite paying little attention to patterns of burial caves locations and lineages/higher social status members, implies the untenable prop- their relationship to the habitation sites and the landscape. It is osition of ‘too many chiefs – no Indians’. Moreover, state societies not argued that the spatial aspects of the burial sites are more were preceded by societies that were not egalitarian, yet the chief- important than other aspects of mortuary behavior. It is rather sug- dom default model and its ubiquitous use should not be used gested that they are important for a better understanding of the mechanically to define them (Bawden, 1989; Crumley, 1995; Hole, Ghassulian secondary burials. 1983, pp. 322–328; Yoffee, 1993). In any case, corporate lineal de- The link between secondary burial caves and land tenure has cent groups are not necessarily equivalents to the chiefdom stage never been previously considered. Our study, however, is by no of Service (Knight, 1990, p. 4). Based on ethnographic data, it is means exhaustive. We have chosen to concentrate only on the spa- more likely that corporate groups (clans or lineages), who own or tial distribution of burial caves, and its relation with neighboring control lands, which are often marked by special boundary cere- habitation sites, excluding several sites such as – Teleilat Ghassul. monies and/or cemeteries, represent local groups (tribes), rather This may raise the question whether or not this model of land ten- than chiefdoms, that depend on agriculture, and settled ‘‘perma- ure system can be applied to all the Ghassulian cemeteries that are nently or seasonally close to fields or other prime subsistence re- not burial caves. Unfortunately there are only four Ghassulian sources’’ (Earle, 2000, p. 46; Johnson and Earle, 1987). above ground cemeteries: Shiqmim (Levy, 1987), Palmahim North O’Shea and Milner (2002, p. 200) state that tribal social organi- (Gorzalczany, 2006), Kissufim Road (Goren and Fabian, 2002) and zation ‘‘is a means of predictably organizing people within a de- perhaps Adeimeh (Neuville, 1930), near the type-site Teleilat Ghas- fined territory’’, and that ‘‘...the flexibility inherent to tribal sul. Currently, the comparison between these two burial forms is organization can be seen as both a major adaptive benefit...and problematic mainly because burial caves are frequent while above as a potential seedbed for evolutionary change’’. Kinship and ideol- ground cemeteries are very rare. ogy provide the ‘‘...floor plan that describes the cosmological origin To conclude, in this study we have tried to analyze the spatial and interrelationships among the people and their land’’ (O’Shea inter-relationships between secondary burial caves and the neigh- and Milner, 2002, p. 201). Natural features and built structures, boring habitation sites. We found that the cemeteries and their such as burial mounds, provide the active representation of ‘‘inten- neighboring habitation sites are distributed in a pattern which sig- tioned and conventionalized use of material markers to designate nificantly increases the area that can be monitored. This distribu- identity, boundaries and ownership.’’ (O’Shea and Milner, 2002, tion may reflect a system of landmarks associated with p. 207). Based on the present evidence it seems that Chalcolithic negotiating land tenure. During the Neolithic period secondary burial grounds represent local corporate groups (similar to tribal burials were means to enhance social cohesion. During the Chalco- systems) rather than chiefdoms. These local corporate groups prac- lithic period another function was added: to define and regulate ticed secondary burial which maintained their social identity and the rights to use and access land and its resources. group cohesion alongside asserting and defining rights over land use by means of lineal ties to ancestors. This was marked by the location of a nearby hill where burial caves were quarried. These Acknowledgments lineal ties were socially articulated and symbolized by the second- ary burial of the deceased relatives in formal cemeteries. The loca- This study was carried out in the framework of the research tion of burial caves on hill sides suggests a signaling mechanism of project ‘‘Feasts of the Dead’’ (1110/09), financed by the Israel Sci- visibility, from the areas near the habitation sites. The above men- ence Foundation. The authors wish to thank the Pratt Foundation tioned Rumu from Papua New Guinea is a good example of demon- Ph.D. Fellowship Program at Ben-Gurion University of the Negev. strating the close link between secondary burial in caves and land The Geological Survey of Israel kindly provided the DEM. We also tenure. The landscape in this particular case is even defined as thank Uzi Dahari, Peter Fabian, Edwin van den Brink, Hagit Torge ‘‘spiritscape’’. We therefore suggest that the location of the Chalco- and Lynne Goldstein for their important contribution. We thank lithic secondary burial grounds reflects both the affirmation of the Paola Ronzino who prepared the first version of our dataset. Steve corporate group’s social identity and the link of this identity to a Rosen, Janet Levy and four anonymous readers read previous drafts particular area within the surrounding landscape that is the source and made important comments that improved the paper consider- of vital subsistence resources. ably. Any mistakes, however, remain our responsibility alone.

References Conclusions Adler, A.M., 1996. Land tenure, Archaeology and the ancestral Pueblo social The distribution of secondary burial caves in the landscape may landscape. Journal of Anthropological Archaeology 15, 337–371. reflect an element of land tenure system. The burial caves were vis- Antrop, M., 1988. Invisible connectivity in rural landscapes. In: Schreiber, K.F. (Ed.), Proceedings of the Second International Seminar of IALE, Connectivity in ible from larger areas than the areas that are visible from the hab- Landscape Ecology, vol. 29. Münsterische Geographische Arbeiten, Münster, pp. itation sites and their location completes rather than overlaps the 57–62. area visible from the habitation sites. This may suggest a need to Ardrey, R., 1966. The Territorial Imperative. Atheneum, New York. Arensburg, B., Hershkovitz, I., 1989. Artificial skull ‘‘treatment’’ in the PPNB period: enlarge the size of the terrain controlled by a community, the Nahal Hemar. In: Hershkovitz, I. (Ed.), People and Culture in Change: source of vital subsistence resources, mainly arable lands. Second- Proceedings of the Second Symposium on Upper Paleolithic, Mesolithic, and ary burial was adopted by the Ghassulian communities during a Neolithic Populations of Europe and the Mediterranean Basin. British time of demographic growth and intensification of crop growing Archaeological Reports International Series 508, pp. 115–133. Bar-Yosef Mayer, D.E., Porat, N., Gal, Z., Shalem, D., Smithline, H., 2004. Steatite and animal husbandry, activities that required large areas of arable beads at Peqi’in: long distance trade and pyro-technology during the land. Thus, it is likely that secondary burials have been used not Chalcolithic of the Levant. Journal of Archaeological Science 31, 493–502. only to enhance social cohesion but also to socially mediate and as- Bar-Yosef, O., 1986. Excavation in Nahal Hemar Cave. Mitekufat Haeven. Journal of the Israel Prehistoric Society 19, 83–84. sert land use responsibilities and rights of access. Bawden, G., 1989. The Andean state as a state of mind. Journal of Anthropological During the last two decades the archaeological research con- Research 45, 327–332. cerning the burial practices of the Chalcolithic period have focused Bearak, B., 2010. Dead join the living in a family celebration. The New York Times A7. on social, cultural and symbolic aspects. Most of the research has Belfer-Cohen, A., 1988. The Natufian graveyard in Hayonim Cave. Paleorient 14, been devoted to the study of ossuaries and other funerary objects, 297–308. Author's personal copy

436 R. Winter-Livneh et al. / Journal of Anthropological Archaeology 31 (2012) 423–438

Belfer-Cohen, A., 1991. The Natufian in the Levant. Annual Review of Anthropology Gilead, I., 1988a. The Chalcolithic period in the Levant. Journal of World Prehistory 20, 167–186. 2, 397–443. Beyer, H.L., 2004. Hawth’s Analysis Tools for ArcGIS. . the Israel Prehistoric Society 21, 145–150. Binford, R., 1971. Mortuary practices: their study and potential. American Antiquity Gilead, I., 1993. Sociopolitical organization in the northern Negev at the end of the 36, 6–29. Chalcolithic period. In: Biran, A., Aviram, J. (Eds.), Biblical Archaeology Today Bintliff, J., 1982. Settlement patterns, land tenure and social structure: a diachronic 1990, Supplement. Israel Exploration Society, Jerusalem. model. In: Renfrew, C., Shennan, S. (Eds.), Ranking, Resource and Exchange. Gilead, I., 1995. Grar: A Chalcolithic Site in the Northern Negev. Ben-Gurion Cambridge University Press, Cambridge, pp. 106–111. University Press, Beer-Sheva, Beer Sheva. Bintliff, J., 1999. Settlement and territory. In: Barker, G. (Ed.), Companion Gilead, I., 2002. Religio-magic behavior in the Chalcolithic period of Palestine. In: Encyclopedia of Archaeology. Routledge, London, pp. 505–545. Ahituv, S., Oren, E. (Eds.), Aharon Kempinski Memorial Volume, vol. XV. Ben- Bloch, M., 1971. Placing the Dead: Tombs, Ancestral Villages, and Kinship Gurion University of the Negev Press, Beer-Sheva, pp. 103–128. Organization in Madagascar. Seminar, London. Gilead, I., 2007. The Besorian: a Pre-Ghassulian cultural entity. Paléorient 33, 33–49. Bloch, M., 1982. Death, women, and power. In: Bloch, M., Parry, J. (Eds.), Death and Gilead, I., 2009. The Neolithic-chalcolithic transition in the southern Levant: late the Regeneration of Life. Cambridge University Press, Cambridge, pp. 211–230. sixth-fifth millennium culture history. In: Shea, J.J., Lieberman, D.E. (Eds.), Bloch, M., 1989. Ritual, History and Power. Althone, London. Transitions in Prehistory: Essays in Honor of Ofer Bar-Yosef. Oxbow Books, Bourke, J.S., 2001. The Chalcolithic period. In: MacDonald, B., Adams, R., Bienkowski, Cambridge, pp. 335–355. P. (Eds.), The Archaeology of Jordan. Sheffield Academic Press, Sheffield, pp. Gilead, I., 2011. Chalcolithic Culture History: Ghassulian and Other Entities in the 107–162. Southern Levant. In: Lovell, L.J., Rowan, M.Y. (Eds.), Culture, Chronology and the Bradley, R., 1998. The Significant of Monuments: On the Shaping of Human Chalcolithic. Theory and Transition. Oxbow Books, Oxford, pp. 12–24. Experience in Neolithic and Bronze Age Europe. Routledge, London. Glazier, J., 1984. Mbeere ancestors and the domestication of death. Man 19, 133– Brighenti, A., 2007. Visibility. A catagory for the social sciences. Current Sociology 147. 55, 323–342. Gleason, K.L., 1994. To bound and to cultivate: introduction. In: Miller, N., Gleason, Cannon, A., 2002. Spatial narratives of death, memory and transcendence. K. (Eds.), The Archaeology of Garden and Field. University of Pennsilvenia Press, Archeological Papers of the American Anthropological Association 11, 191–199. Philadelphia, pp. 1–24. Carr, C., 1995. Mortuary practices: their social, philosophical-religious, Golden, J., 2009. Dawn of the Metal Age. Equinox, London. circumstantial, and physical determinants. Journal of Archaeological Method Goldstein, L., 1976. Spatial Structure and Social Organization: Regional and Theory 2, 105–200. Manifestations of Mississippian Society. Unpublished PhD Dissertation, Cashdan, E., 1983. Territoriality among human foragers: ecological models and an Northwestern University. application to four Bushman groups. Current Anthropology 24, 47–66. Goldstein, L., 1981. One-dimensional archaeology and multi-dimensional people: Cashdan, E., 1989. Hunters and gatherers: economic behavior in bands. In: Plattner, spatial organization and mortuary analysis. In: Chapman, R., Kinnes, I., S. (Ed.), Economic Anthropology. Stanford University Press, Stanford, pp. 21–48. Randsborg, K. (Eds.), The Archaeology of Death. Cambridge University Press, Chapman, R., Kinnes, I., Randsborg, K., 1981. The Archaeology of Death. Cambridge New York, pp. 53–69. University Press, Cambridge. Goldstein, L., 2002. Afterword Visible death: mortuary site and mortuary landscape Chénier, A., 2009. Bones, people and communities: tensions between individual and in diachronic perspective. Archeological Papers of the American corporate identities in secondary burial ritual. Nexus: The Canadian Student Anthropological Association 11, 201–205. Journal of Anthropology 21, 27–40. Goldstein, L., 2008. Comments: The regeneration of life, Kuijt, I. Current Clarke, D.L. (Ed.), 1977. Spatial Archaeology. Academic Press INC., London. Anthropology 49 (2), 189–190. Cohen-Weinberger, A., 2011. Provenance study of selected Chalcolithic and Early Gopher, A., Tsuk, T., 1990. Ancient Gold-Rare Finds from the Nahal Qanah Cave. Bronze Age I pottery from Sha’ar Efrayim. ‘Atiqot 66, 63–66. Jerusalem. Cohen, E., 1976. Environmental orientations: a multidimensional approach to social Gopher, A., Tsuk, T., 1996. The Nahal Qanah Cave: Earliest Gold in the Southern ecology. Current Anthropology 17, 49–70. Levant. Monograph Series of the Institute of Archaelology, vol. 12. Institute of Crumley, C.L., 1995. Heterarchy and the analysis of complex societies. Archeological Archaelology, Tel Aviv University, Tel Aviv, 250 pp. Papers of the American Anthropological Association 6, 1–5. Gophna, R., Lifshitz, S., 1980. A Chalcolithic burial cave at Palmahim. ‘Atiqot 14, 1–8. David, B., Pivoru, M., Pivoru, W., Green, M., Barker, B., Weiner, J., Simala, D., Kokents, Gophna, R., Portugali, J., 1988. Settlement and demographic processes in Israel’s T., Arabo, L., Dop, J., 2008. Living landscapes of the dead: archaeology of the costal plain from the Chalcolithic to the Middle Bronze Age. Bulletin of the afterworld among the Rumu of Papua New Guinea. In: David, B., Thomas, J. American Schools of Oriental Research 269, 11–28. (Eds.), Handbook of Landscape Archaeology. Left Coast Press, San Francisco, pp. Gopher, A., Tsuk, T., 1991. Ancient Gold-Rare Finds from the Nahal Qanah Cave. 158–166. Jerusalem, Israel Museum. Dayson-Hudson, R., Smith, E.A., 1978. Human territoriality: an ecological Gophna, R., Tsuk, T., 2005. Chalcolithic settlements in the western Samaria foothills. reassessment. American Anthropologist 80, 21–41. Tel Aviv 32, 3–19. Earle, T.K., 1991. Property rights and the evolution of chiefdoms. In: Earle, T.K. (Ed.), Goren, Y., Fabian, P., 2002. Kissufim Road: A Chalcolithic Mortuary Site, vol. 16. Chiefdoms: Power, Economy and Ideology. Cambridge University Press, Israel Antiquities Authority, Jerusalem. Cambridge, pp. 71–99. Goring-Morris, N., 2000. The quick and the dead: the social context of Aceramic Earle, T.K., 2000. Archaeology, property, and prehistory. Annual Review of Neolithic mortuary practices as seen from Kfar HaHoresh. In: Kuijt, I. (Ed.), Life Anthropology 29, 39–60. in Neolithic Farming Communities: Social Organization, Identity, and Eason, P.K., Stamps, J.A., 1992. The effect of visibility on territory size and shape. Differentiation. Plenum Press, New York, pp. 103–136. Behavioral Ecology 3, 166–172. Goring-Morris, N., Burns, R., Davidzon, A., Eshed, V., Goren, Y., Hershkovitz, I., Epstein, C., 1998. The Chalcolithic Culture of the Golan. IAA Reports No. 4, Kangas, S., Kelecevic, J., 1998. The 1997 season of excavations at the mortuary Jerusalem. site of Kfar HaHoresh, Galilee, Israel. Neo-Lithics 3 (98), 1–4. Finkelstein, I., Gophna, R., 1993. Settlement, demographic, and economic patterns in Goring-Morris, N., Horwitz, L.K., 2007. Funerals and feasts during the Pre-Pottery the highlands of Palestine in the Chalcolithic and Early Bronze periods and the Neolithic B of the Near East. Antiquity 81, 902–919. beginning of urbanism. Bulletin of the American Schools of Oriental Research Gorzalczany, A., 2006. Chalcolithic burial site at Palmahim North. Qadmoniot 132, 295, 1–22. 87–94. Fleming, A., 1973. Tombs for the living. Man 8, 177–193. Green, E.D., 2006. Ethnicity and the politics of land tenure reform in central Uganda. Foucault, M., 1995 (1975). Discipline and Punish: The Birth of the Prison, second ed., Commonwealth and Comparative Politics 44, 370–388. Vintage Books, New York. Grosman, L., 2003. Preserving cultural traditions in period of instability: the late Fried, M.H., 1967. The Evolution of Political Society. Random House, New York. Natufian of the hilly Mediterranean zone. Current Anthropology 44, 571–579. Fry, G., Skar, B., Jerpåsen, G., Bakkestuen, V., Erikstad, L., 2004. Locating Grosman, L., Munro, D.N., 2007. The scared and the mundane: domestic activities at archaeological sites in the landscape: a hierarchical approach based upon a Late Natufian burial site in the Levant. Before Farming 4, 1–14. landscape indicators. Landscape and Urban Planning 67, 97–108. Grosman, L., Munro, D.N., Belfer-Cohen, A., 2008. A 12,000-year-old shaman burial Gafinkel, Y., 2006. The burials of Kfar HaHoresh – a regional or local phenomenon? from the southern Levant (Israel). Proceedings of the National Academy of Mitekufat Haeven. Journal of the Israel Prehistoric Society 36, 109–116. Sciences 105, 17665–17669. Gal, Z., Shalem, D., Smithline, H., 2011. The Peqi’in cave: a Chalcolithic cemetery in Haas, N., Nathan, H., 1973. An attempt at a social interpretation of the Chalcolithic Upper Galilee, Israel. Near Eastern Archaeology 74, 196–206. burials in the Nahal Mishmar caves. In: Aharoni, Y. (Ed.) Excavations and Gal, Z., Smithline, H., Shalem, D., 1997. A Chalcolithic burial cave in Peqi’in, Upper Studies: Essays in Honour of Professor Shemuel Yeivin. Tel-Aviv University, Galilee. Israel Exploration Journal 47, 145–154. Institute of Archaeology, Tel-Aviv, pp. 143–153 (Hebrew). Gal, Z., Smithline, H., Shalem, D., 1999. New iconographic aspects of Chalcolithic art: Hachlili, R., 2005. Jewish Funerary Customs, Practices and Rites in the Second preliminary observations on finds from the Peqi’in cave. ‘Atiqot 47, 1–16. Temple Period. Brill, Leiden. Gerritsen, F., 2003. Local Identities: Landscape and Community in the late Hall, J.K., 1993. The GSI digital terrain model (DTM) project completed. GSI Current Prehistoric Meuse-Demer-Scheldt Region. Amsterdam University Press, Research 8, 47–50. Amsterdam. Hanson, H.E., 2003. Landed Obligation: The Practice of Power in Buganda. Social Gibson, S., Rowan, Y.M., 2006. The Chalcolithic in the central highlands of Palestine: History in Africa Series. Heinemann, Portsmouth, New Hampshire. a reassessment based on new examination of Khibet es-Sauma’a. Levant 38, 85– Hermon, S., 2008. Socio-economic Aspects of Chalcolithic (4500–3500 BC) Societies 108. in the Southern Levant. BAR International Series 1744. Archaeopress, Oxford. Author's personal copy

R. Winter-Livneh et al. / Journal of Anthropological Archaeology 31 (2012) 423–438 437

Hershkovitz, I., Gopher, A., 1990. Paleodemography, burial customs, and food- McNiven, I., 2004. Saltwater People: spiritscapes, maritime rituals and the producing economy at the beginning of the Holocene: A perspective from the archaeology of Australian indigenous seascapes. World Archaeology 35, 329– southern Levant. Mitekufat Haeven. Journal of the Israel Prehistoric Society 23, 349. 9–48. Meggitt, M.J., 1965a. The Lineage System of the Mae Enga of New Guinea. Oliver and Hertz, R., 1960. In: Needham, R., Needham, C. (Trans.), Death and the Right Hand. Boyd, Edinburgh. Cohen and West, London. Meggitt, M.J., 1965b. The Mea-Enga of the Western Highlands. In: Lawrence, P., Hodder, I., 1982. Symbolic and Structural Archaeology. Cambridge University Press, Meggitt, M.J. (Eds.), Gods, Ghosts and Men in Melanesia. Oxford University Cambridge. Press, New York. Hole, F., 1983. Symbols of religion and social organization at Susa. In: Young, T.C., Metcalf, P., 1981. Meaning and materialism: the ritual economy of death. Man 16, Smith, P.E.L., Mortensen, P. (Eds.), The Hilly Flanks and Byond. Oriental Institute 564–578. of the University of Chicago, Chicago, pp. 315–334. Metcalf, P., Huntington, R., 1991. Celebrations of Death. Cambridge University Press, Horwitz, L.K., Goring-Morris, N., 2004. Animals and ritual during the Levantine Cambridge. PPNB: a case study from the site of Kfar Hahoresh, Israel. Anthropozoologica 39, Miles, D., 1965. Socio-economic aspects of secondary burial. Oceania 35, 165–178. 161–174. Hutchinson, D.L., Aragon, L.V., 2002. Collective burials and community memories: Milevski, I., 2001. Qula, area K. Excavations and Surveys in Israel 113, 62–63. interpreting the placement of the dead in the Southeastern and Mid-Atlantic Milevski, I. 2007. Mazor (west).In: Hadashot Arkheologiyot , vol. 119. . Archeological Papers of the American Anthropological Association 11, Mitchell, A., 2005. The ESRI Guide to GIS Analysis, vol. 2, ESRI Press. 27–54. Munro, D.N., Grosman, L., 2010. Early evidence (ca. 12,000 B.P.) for feasting at a Joffe, H.A., 2003.‘‘Slouching toward Beersheva: Chalcolithic mortuary practices in burial cave in Israel. Proceedings of the National Academy of Sciences 107, local and regional context. In: Nakhai, B.A. (Ed.), The Near East in the 15362–15366. Southwest: Essays in Honor of William G. Dever. American Schools of Nativ, A., Gopher, A., 2011. The cemetry as a symbol: a reconsideration of Oriental Research, Boston, MA, pp. 45–67. Chalcolithic burial caves in the southern Levant. Cambridge Archaeological Johnson, A.W., Earle, T.K., 1987. The Evolution of Human Societies. Stanford Journal 21, 229–245. University Press, Stanford. Neuville, R., 1930. La ne´cropole me´galithique D’el-Adeimeh. Biblica 9, 249–265. Khalaily, H., 2011. The flint assemblages from Sha’ar Efrayim burial caves 1–3. Nihipali, K., 2007. Steamatory for Preparing a Body for Burial, Patent Application ‘Atiqot 66, 55–62. Publication, vol. 2007/0143974 A1, United States, pp. 1–4. Khalaily, H., Marder, O., 2010. Prehistoric investigations along the cross-Israel North, R.S.J., 1959. ‘‘Ghassulian’’ in Palestine chronological nomencalture. Biblica highway: state of research. Atiqot 64, 5–23. 40, 541–555. Kidd, E.K., 1953. The excavation and historical identification of Huron ossuary. O’Shea, J., 1984. Mortuary Variability: An Archaeological Investigation. Academic American Antiquity 18, 359–379. Press, New York. King, G.E., 1975. Society and territory in human evolution. Journal of Human O’Shea, J., Milner, C.M., 2002. Material indicators of territory, identity and Evolution 5, 323–332. interaction in a prehistoric tribal system. In: Parkinson, W.A. (Ed.), The Knight, V.J., 1990. Social organization and the evolution of hierarchy in southeastern Archaeology of Tribal Societies, vol. 15. International Monographs in Chiefdoms. Journal of Anthropological Research 46, 1–23. Prehistory, Ann Arbor, Mich., pp. 200–226. Kodesh, N., 2008. Networks of knowledge: clanship and collective well-being in Peregrine, N.P., 2001. Cross-cultural comparative approaches in archaeology. Buganda. Journal of African History 49, 197–216. Annual Review of Anthropology 30, 1–18. Kuijt, I., 1996. Negotiating equality through ritual: a consideration of Late Natufian Perrot, J., 1984. Structures d’habitat, mode de vie et environnement. Les villages and Prepottery Neolithic A period mortuary practices. Journal of souterrains des pasteurs de Beershe´va, dans le sud d’Israël, au IVe mille´naire Anthropological Archaeology 15, 313–336. avant l’e‘re chre´tienne. Paleorient 10, 75–96. Kuijt, I., 2001. Place, death, and the transmission of social memory in early Perrot, J., Ladiray, D., 1980. Tombes a‘ ossuaires de la re´gion cotie‘re Palestinienne, agricultural communities of the Near Eastern Pre-Pottery Neolithic. au IVe mille´naire avant l’e‘re chre´tienne. Association Paléorient, Paris. Archeological Papers of the American Anthropological Association 10, 80–99. Perrot, J., Ladiray, D., 1988. Les hommes de Mallaha (Eynan) Israel, les sepulchres. Kuijt, I., 2008. The regeneration of life: Neolithic structures of symbolic Memoires et Travaux du Centre de Recherche Française de Jerusalem, No. 7. remembering and forgetting. Current Anthropology 49, 171–197. Association Paléorient, Paris. Kuijt, I., Goring-Morris, N., 2002. Foraging, farming and social complexity in the Pre- Peterson, J., 2010. Domesticating gender: Neolithic patterns from the southern Pottery Neolithic of the Southern Levant: a review and synthesis. Journal of Levant. Journal of Anthropological Archaeology 29, 249–264. World Prehistory 16, 361–440. Peterson, N., 1975. Hunter-gatherer territoriality: the perspective from Australia. Kvamme, K.L., 1990. One-sample tests in regional archaeological analysis: new American Anthropologist 77, 53–68. possibilities through computer technology. American Antiquity 55, 367–381. Porath, Y., 2006. Chalcolithic burial sites at Ma’abarot and Tel Ifshar. ‘Atiqot 53, 45– Lev-Tov Chattah, N., Smith, P., 2006. Variation in occlusal dental wear of two 64. Chalcolithic populations in the Southern Levant. American Journal of Physical Poyil, M., 2009. Farewell ritual and transmigrating souls: secondary funeral of the Anthropology 130, 471–479. Attappadi Kurumbas. Anthropologist 11, 31–38. Lev-Tov, N., Gopher, A., Smith, P., 2003. Dental evidence for dietary practices in the Rahmani, L.Y., 1978. Ossuaries and bone-gathering in the late second temple period. Chalcolithic period: the findings from a burial cave in Peqi’in (Northern Israel). Qadmoniot 11, 102–112 (Hebrew). Paleorient 29, 121–134. Regev, E., 2002. Family burial in herodianic Jerusalem and its environs and the Levy, T.E., 1986. The Chalcolithic period. Biblical Archaeologist 49, 82–108. social organization of immigrants and sectarians. Qatedrah 106, 35–60 Levy, T.E. (Ed.), 1987. Shiqmim I, Studies Concerning Chalcolithic Societies in the (Hebrew). Northern Negev Desert, Israel. BAR International Series 356, Oxford. Renfrew, C., 1973a. Before Civilization. Alfred Knopf, New York. Levy, T.E., 1995. Cult, metallurgy and rank societies – Chalcolithic period (CA. 4500– Renfrew, C., 1973b. Monuments, mobilisation, and social organization in Neolithic 3500 BCE). In: Levy, T.E. (Ed.), Archaeology of Society in the Holy Land. Facts on Wessex. In: Renfrew, C. (Ed.), The Explanation of Culture Change: Models in File, New York, pp. 226–243. Prehistory. Duckworth, London, pp. 539–558. Levy, T.E. (Ed.), 2006. Archaeology, Anthropology and Cult: The Sanctuary at Gilat, Renfrew, C., 1976. Megaliths, territories and population. In: De Laet, S.J. (Ed.), Israel. Equinox Publishing, London. Acculturation and Continuity in Atlantic Europe. De Tempel, Brugge, pp. 198– Levy, T.E., Alon, D., 1985. Shiqmim: a Chalcolithic village and mortuary center in the 220. northern Negev. Paleoriént 11, 71–83. Robb, J., 2007. Burial treatment as transformations of bodily ideology. In: Laneri, N. Levy, T. E., Alon, D. 1987. Excavations in Shiqmim cemetery 3: final report on the (Ed.), Preforming Death: Social analyses of Funerary Traditions in the Ancient 1982 excavations. In: Levy, T.E. (Ed.), Shiqmim I, Studies Concerning Near East and Mediterranean. The Oriental Institute of the University of Chalcolithic Societies in the Northern Negev Desert, Israel. Oxford. BAR Chicago, Chicago, pp. 287–298. International Series, vol. 356, pp. 333–355. Rollefson, G.O., 1986. Neolithic ‘Ain Ghazal (Jordan): ritual and ceremony, II. Llobera, M., 2003. Extending GIS-based visual analysis: the concept of visualscapes. Paléorient 12, 45–52. International Journal of Geographical Information Science 17, 25–48. Rosen, A.S., 1993. Metals, rocks, specialization, and the beginning of urbanism in the Llobera, M., 2007. Reconstructing visual landscapes. World Archaeology 39, northern Negev. In: Biran, A., Aviram, J. (Eds.), Biblical Archaeology Today, 1990. 51–69. Israel Exploration Society, Jerusalem, pp. 41–56. Llobera, M., 2010. Archaeological visualization: towards an archaeological Rowan, M.Y., Golden, J., 2009. The Chalcolithic period of the Southern Levant: a information science (AISc). Journal of Archaeological Method and Theory synthetic review. Journal of World Prehistory 22, 1–92. (Online). Rubin, N., 1994. Secondary burials in the Mishnaic and Talmudic periods: a Lovell, L.J., 2009. Chalcolithic caves discovered east of the River Jordan. Antiquity. proposed model of the relationship of social structure to burial practice. In: Maher, L.A., Stock, J.T., Finney, S., Heywood, J.J.N., Miracle, P.T., Banning, E.B., 2011. A Singer, I. (Ed.), Graves and Burial Practices in Israel in the Ancient Period. Yad Unique Human-Fox Burial from a Pre-Natufian Cemetery in the Levant (Jordan). Izhak Ben-Zvi, Jerusalem, pp. 248–269 (Hebrew). PLoS ONE 6, e15815. Saxe, A., 1970. Social Dimensions of Mortuary Practices. Unpublished Ph.D. Mair, L.P., 1933. Baganda land tenure. Journal of International African Institute 6, Dissertation, University of Michigan. 187–205. Schroeder, S., 2001. Secondary disposal of the dead: cross-cultural codes. World Maisels, C.K., 2001. Civilizations of the Old World, the Early Mesopotamia, India, Cultures 12, 77–93. Formative Histories of Egypt, the Levant and China. Routledge, London and New Service, E.R., 1962. Primitive Social Organization: An Evolutionary Perspective. York. Random House, New York. Author's personal copy

438 R. Winter-Livneh et al. / Journal of Anthropological Archaeology 31 (2012) 423–438

Sevenant, M., Antrop, M., 2007. Settlement models, land use and visibility in rural Van den Brink, E.C.M., 2011a. A Chalcolithic and Early Bronze Age I burial ground landscapes: two case studies in Greece. Landscape and Urban Planning 80, 362– near Sha’ar Efrayim in the Sharon plain. ‘Atiqot 66, 1–53. 374. Van den Brink, E.C.M., 2011b. Tel Aviv, Namir Road preliminary report. Hadashot Shipton, P., 1984. Lineage and locality as antithetical principles in East African Arkheologiyot, Excavations and Surveys in Israel 123. systems of land tenure. Ethnology 23, 117–132. Van den Brink, E.C.M., Gophna, R., 2005. Shoham (North): Late Chalcolithic Burial Silverman, B.W., 1986. Density Estimation for Statistics and Data Analysis. Chapman Caves in the Lod Valley, Israel. IAA Reports. Israel Antiquities Authority, and Hall, London. Jerusalem. Singer, A., 2007. The Soils of Israel. Springer, Berlin. Verhoeven, M., 2002. Transformations of society: the changing role of ritual and Stekelis, M., Yizraeli, T., 1963. Excavations at Nahal Oren. Israel Exploration Journal symbolism in the PPNB and the PN in the Levant, Syria and south-east Anatolia. 13, 1–12. Paléorient 28, 5–13. Stone, G.D., 1994. Agricultural intensification and perimetrics: ethnoarchaeological Weiss-Krejci, E., 2004. Mortuary representations of the noble house. A cross- evidence from Nigeria. Current Anthropology 35, 317–324. cultural comparison between collective tombs of the ancient Maya and dynastic Sukenik, E.L., 1937. A Chalcolithic necropolis at Hadera. Journal of the Palestine Europe. Journal of Social Archaeology 4, 368–404. Oriental Society 17, 15–30. Weiss-Krejci, E., 2011. The formation of mortuary deposits. Implication for Sussman, V., Ben-Arieh, S., 1966. Ancient burials in Giv’atayim. ‘Atiqot 3, understanding mortuary behavior of past populations. In: Agarwal, C.S., 27–39. Glencross, A.B. (Eds.), Social Bioarchaeology. Wiley-Blackwell, pp. 68–106. Thwaites, R.G. (Ed.), 1898. The Jesuit Relation and Allied Documents: Travels and Wheatley, D., 1995. Cumulative viewshed analysis: a GIS-based method for Explorations of the Jesuit Missionaries in New France 1610–1791, vol. X. investigating intervisibility, and its archaeological application. In: Lock, G., Burrows, Cleveland. Stancic, Z. (Eds.), Archaeology and Geographical Information Systems: A Tooker, E., 1964. An Ethnography of the Huron Indians, 1615–1649. Smithsonian European Perspective. Taylor & Francis, London, pp. 171–185. Institute, Washington, DC. Wheatley, D., Gillings, M., 2000. Vision, perception and GIS: developing enriched Trigger, B.C., 1969. The Huron: Farmers of the North. Holt, Rinehart, & Winston, New approaches to the study of archaeological visibility. In: Lock, G. (Ed.), Beyond York. the Map. IOS Press, Oxford. Trigger, B.C., 1976. The Children of Aataentsic: A History of the Huron People to Wheatley, D., Gillings, M., 2002. Spatial Technology and Archaeology. Taylor and 1660. McGill-Queen’s University, Montreal. Francis, London. Ucko, P., 1969. Ethnography and archaeological interpretation of funerary remains. Williams, H.M.R., 1999. Placing the dead: investigating the location of wealthy World Archaeology 1, 262–280. barrow burial in seventh century England. In: Rundkvist, M. (Ed.), Grave Unruh, J.D., 2006. Land tenure and the ‘‘Evidence Landscape’’ in developing Matters: Eight Studies of First Millennium AD Burials in Crimea, England and countries. Annals of the Association of American Geographers 96, 754–772. Southern Scandinavia. Archaeopress, Oxford, pp. 57–86. Valla, F.R., 1986. De nouvelles structures Natoufiennes sur la Terrace d’Hayonim. Wilson, E.O., 1975. Sociobiology: The New Synthesis. Harvard University Press, Mitekufat Haeven 19, 90–92. Cambridge. Van den Brink, E.C.M., 1998. An index to Chalcolithic mortuary caves in Israel. Israel Winter-Livneh, R., Svoray, T., Gilead, I., 2010. Settlement patterns, social complexity Exploration Journal 48, 165–173. and agricultural strategies during the Chalcolithic period in the Northern Negev, Van den Brink, E.C.M., 2005. Chalcolithic burial caves in coastal and inland Israel. In: Israel. Journal of Archaeological Science 37, 284–294. van den Brink, E.C.M., Gophna, R. (Eds.), Shoham (North): Late Chalcolithic Wylie, A., 1985. The reaction against analogy. Advances in Archaeological Method Burial Caves in the Lod Valley, Israel. Israel Antiquities Authority, Jerusalem, pp. and Theory 8, 63–111. 175–189. Yannai, E., Porath, Y., 2006. A Chalcolithic burial cave as Et-Taiyiba. ‘Atiqot 53, 1–44. Van den Brink, E.C.M., 2006. Tel Aviv, Namir Road. Hadashot Arkheologiyot, Yekutieli, Y., 2006. Is somebody watching you? Ancient surveillance systems in the Excavations and Surveys in Israel 118. southern Judean Desert. Journal of Mediterranean Archaeology 19, 65–89. Van den Brink, E.C.M., 2008. A new fossile directeur of the Chalcolithic landscape in Yoffee, N., 1993. Too many chiefs? (or, safe texts for the ‘90s). In: Yoffee, N., Sherratt, the Shephelah and the Samarian and Judaen hill countries: stationary grinding A. (Eds.), Archaeological Theory: Who Sets the Agenda? Cambridge University facilities in bedrock. Israel Exploration Journal 58, 1–23. Press, Cambridge, pp. 60–76. 2.3 Shape reproducibility and architectural symmetry during the Chalcolithic period in the southern Levant

This study was published in Journal of Archaeological Science 40: 1340-1353 (2013). A copy of the article follows.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 47 Author's personal copy

Journal of Archaeological Science 40 (2013) 1340e1353

Contents lists available at SciVerse ScienceDirect

Journal of Archaeological Science

journal homepage: http://www.elsevier.com/locate/jas

Shape Reproducibility and architectural symmetry during the Chalcolithic period in the southern Levant

Rona Winter-Livneh a,*, Tal Svoray b, Isaac Gilead a a Dept. of Bible, Archaeology and Ancient Near Eastern Studies, Ben-Gurion University of the Negev, Israel b Dept. of Geography and Environmental Development, Ben-Gurion University of the Negev, Israel article info abstract

Article history: Architecture reflects social aspects of past communities. Structure attributes such as shape, size, building Received 5 July 2012 material and decoration, provide valuable information beyond their immediate structural function. Received in revised form However, while attributes such as size can be measured and therefore objectively compared between 6 October 2012 structures, the comparison of shape between structures is based on subjective observations. In the Accepted 8 October 2012 current study we use two quantification methods for analyzing prehistoric shape-based architectural data: (1) we developed a new method, Shape Reproducibility (SR), based on objective computerized Keywords: procedure for analyzing the similarity and difference between shapes of ancient buildings; and (2) we Chalcolithic fl Architecture use Continuous Symmetry Measure (CSM), a method which was originally developed for analyzing int Shape Reproducibility artifacts and ceramic vessels to objectively compare between shape symmetry. Applying these methods Symmetry to settlement data of the Chalcolithic period enables quantification of the level of architectural similarity GIS within and between different sites and their comparison to architectural data of later periods, such as the Southern Levant Early Bronze Age II urban center at Arad. Our CSM results suggest that the symmetry of architecture does not increase through time. Our SR findings demonstrate that in the main cultural Chalcolithic entity, the Ghassulian, the architecture of different sites could not be distinguished from one site to the other. In addition, we demonstrate that the architecture of the Chalcolithic sites in the Golan Heights is homogeneous and significantly differs from other Chalcolithic sites, while Ghassulian intra-site vari- ability is higher. In comparison with Arad, however, this variability is relatively low and limited. These results suggest that status differentiation or hierarchical social organization cannot be indicated from Ghassulian architecture. Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction population size, technology and subsistence economy (e.g. Allison, 2002; Banning, 2010, Banning and Byrd, 1987; Binford, 1990; Building is among the prime activities carried out by humans Carsten and Hugh-Jones, 1995; Flannery, 1972; Hillier and Hanson, since earliest times (Bar-Yosef, 1992: 31). Architecture is a visible 1984; Hodder, 1994; Lévi-Strauss, 1963; Rapoport, 1969, 1982; cultural manifestation that influences social behavior and provides Wilson, 1988). the framework for social interaction and community organization The first structures in the southern Levant were made of (Byrd, 1994: 643; Ingold, 2000:175e178; Wilson, 1988: 21). It is perishable materials leaving practically no traces, and their exis- also evident that dwellings are subject to spatio-temporal changes tence is inferred on the basis of the spatial distribution of other (e.g. Flannery, 1972; Goring-Morris and Belfer-Cohen, 2008; finds, such as in the case of the Early Epipalaeolithic site of Ohalo II Kempinski and Reich, 1992). Different attributes of structure such (ca. 21,500e20,500 B.C.) (Bar-Yosef, 1992:31;Goring-Morris and as shape, size, building material and decoration have significance Belfer-Cohen, 2008: 249e250; Nadel and Werker, 1999). Later on, beyond their immediate function. They provide invaluable infor- during Natufian cultural phase (ca. 13,000e9600 B.C.), the exis- mation about the social aspects of past societies, as well as evidence tence of post-holes indicates some sort of roofing, as at Ein Gev concerning modes of adaptation to environments, changes in I and III (Arensburg and Bar-Yosef, 1973; Martin and Bar-Yosef, 1979). In other sites architectural remains consist of several walls made of undressed stones that probably supported wooden poles, * Corresponding author. Tel.: þ972 8 9491932. or a few freestanding walls, oval or rounded in shape (Valla, 1988). E-mail address: [email protected] (R. Winter-Livneh). Pre-Pottery Neolithic A (ca. 9600e8500 B.C.) architecture consists

0305-4403/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jas.2012.10.007 Author's personal copy

R. Winter-Livneh et al. / Journal of Archaeological Science 40 (2013) 1340e1353 1341 of oval and subcircular structures that were either freestanding or 2005, 1998). It should be noted that though there are many semi-subterranean (Finlayson et al., 2011a, 2011b; Kuijt and aspects of symmetry such as symmetry of rotation, treatable Finlayson, 2009; Kuijt and Goring-Morris, 2002: 373). These symmetry, etc., the archaeological study of symmetry is usually structures were made of a stone foundation with mud brick limited to bilateral symmetry, meaning that the shape does not superstructure. Worth mentioning is the large Neolithic tower of change upon undergoing a reflection. Bilateral symmetry or Jericho. This unique tower is 8.25 m in height and 8 m in diameter reflection symmetry in archaeological studies is referred to simply made of undressed stone with a staircase built inside (Kenyon, as “symmetry”. 1957; Kenyon and Holland, 1981). During the Pre-Pottery Symmetry appears in the form of artifacts, buildings and built Neolithic B period (ca. 8500e6400 B.C.) the transition from oval/ environments all over the world (Wynn, 2002: 390). Many studies rounded to rectangular structures is evident. Constructions during regard the degree and nature of symmetry as cultural attributes or this period were mainly of mudbrick on stone foundations as manifestations of cultural progress (e.g. Bridgeman, 2002; Lycett, (Banning and Byrd, 1988; Bar-Yosef, 1992; Goring-Morris and 2008; Oakley, 1972; Shennan, 2006; Simao, 2002; Wynn, 1985). Belfer-Cohen, 2003, 2008; Kuijt and Goring-Morris, 2002; Kuijt Others argue that symmetry is related to the evolution of human et al., 2011; Moore, 1985; Rollefson, 1997). Stone and mudbrick cognition (Stout and Chaminade, 2007; Toth, 1990; Wynn, 2002), or rectangular architecture continued well into the Pottery Neolithic link it to functional effectiveness (Jones, 1980; Machin et al., 2007; (ca. 6400e4500 B.C.) and Chalcolithic (4500e3900 B.C.) periods Mitchell, 1996), to sexual display (Kohn and Mithen, 1999), or to (Ben-Shlomo and Garfinkel, 2009; Garfinkel and Ben-Shlomo, aesthetics (Hodgson, 2011; Schick and Toth, 2001: 282). There are, 2002; Garfinkel et al., 2009; Porath, 1987, 1992). In addition to however, studies that show that symmetry could result from the rectangular above ground structures, subterranean structures coincidental factors such as type of raw material, resharpening were uncovered in Chalcolithic sites of the northern Negev such as (McPherron, 2000; Nowell, 2000, 2002), or post depositional Abu Matar, Bir es-Safadi (Perrot, 1984), and Horvat Beter (Dothan, processes which involve environmental disturbances that damage 1959; Rosen and Eldar, 1993). Preservation of mudbrick walls is stone tools (Grosman et al., 2011). poor and in many Chalcolithic sites preserved were only fragments Indeed, symmetrical attribute signaling safer, more effective, and of walls, in many cases without side or parallel walls. Bricks were more predictable artifacts or buildings than asymmetrical made by hand, of local silts (Porath, 1992: 44), and after their ones (Liu and Kersten, 2003; van der Helm, 2002; Vetter et al., collapse they disintegrated and could not be distinguished from the 1994; Wagemans, 1995). However, unlike earlier periods, when natural sediment. Along with other destructive processes, walls are manufacturing of Acheulean handaxes was associated with different under-represented in comparison to pits and other installations hominins, and with butchering effectiveness, during later periods (Gilead, 1995: 30). the intentional concern for symmetry seems to be detached from Architectural studies are usually based on the analysis of the evolution of cognitive, adaptive or functional factors. Thus, it is structures at different sites in order to explain the similarity or reasonable to study architectural symmetrically of later periods e variability of the shape patterns. In many cases the researchers the Chalcolithic and Early Bronze Age periods in our case e which define a ‘typical’, frequent or ‘average ’ house shape that charac- are much too short for evolutionary change, as a manifestation of terizes specific cultures/periods/regions. Such as the Four Room culture change and variability (Bridgeman, 2002: 403, Hodgson, House or the Israelite house during the Iron age (Faust and 2011: 38). Symmetry is a key element in architecture which Bunimovitz, 2003), or later examples such as the Arab-Islamic signals balance, since pressure on a structure or building is distrib- House (Ron, 1998) or the Black Tent (Manderscheid, 2001). Anal- uted equally if there is symmetry. In the study of symmetry of yses of architectural shapes are commonly based on the prehistoric flint tools, such as Acheulean handaxes for example, the researcher’s skill, intuition, and subjective evaluation which result difference between early, less symmetrical artifacts and later, more in biased and sometimes inaccurate conclusions and may lead to symmetrical artifacts, is well established and regarded as an indi- equivocal results. Although there are few exceptions (e.g. Dickens, cation of more elaborated production techniques and increased skill 1977; Fletcher, 1977) most studies of past architectural shapes (Saragusti et al., 2005, 1998; Wynn, 1985). Studies have shown that lacks formal quantitative methods. Below we introduce two manufacturing technique involve social dynamics, and the technical methods for objective and accurate quantification for character- knowledge is directly related to social knowledge (e.g. Dobres, 2010; izing and comparing between shapes applied here to prehistoric Dobres and Hoffman,1994,1999; Schiffer and Skibo,1987; Torrence, architectural data. 1989; van der Leeuw, 1993; Wright, 1993). This is necessarily Quantitative analyses of artifact shapes have been carried out for mediated by culture (Dobres, 2010: 106). In addition a progress in more than half a century (Clarke, 1968: 525e534), and they have technology is driven by cultural accumulation of knowledge increased significantly during recent years (e.g. Durham et al., 1995; (Bridgeman, 2002; Ingold, 1990, e.g. Schiffer and Skibo, 1987). In Gero and Mazzullo, 1984; Gilboa et al., 2004; Grosman et al., 2011, studying the symmetry of prehistoric architecture we, therefore, 2008; Karasik, 2010, Karasik and Smilansky, 2008; 2011; Leese and expect that difference between less symmetrical structures and Main, 1983; Liming et al., 1989; Saragusti et al., 2005, 1998). Such more symmetrical structure might reveal aspects concerning the studies are also based on advance computing along with a variety of technology and skills which characterize the societies and their technologies such as 3D and laser scanning. These studies, however, cultural contexts. focus mostly on pottery vessels and lithics, while architectural Beside the degree of symmetry, studies have shown that the remains are left behind. These studies have introduced a number of shape itself of a house is determined by social or economic factors important mathematical methods for quantifying shape attributes (Allison, 2002; Carsten and Hugh-Jones, 1995; Donley, 1982; Hillier such as symmetry, roughness and deformation. and Hanson, 1984; Ingold, 1995, 2000; Kent, 1990b; King, 1980; Lau, Continuous Symmetry Measure (CSM) is a versatile method 2010; Rapoport, 1969, 1982; Wilson, 1988). Nevertheless, there are which was originally developed to distinguish molecules from each others who argue that the main factors are environmental or other by their degree of shape chirality (dissymmetry) (Zabrodsky physical. These factors include: climate (e.g. Correa, 1982; Fitch and and Avnir, 1995). This tool was first used in archaeology for Branch, 1960; Givoni, 1969; Herzog, 1980; Mauss and Beauchat, measuring the degree of symmetry of Lower Paleolithic handaxes 1979; Sozen and Gedik, 2007); topography or land scarcity (e.g. (Saragusti et al., 1998). It has been demonstrated that symmetry of Alexander, 1964; Sopher, 1964); technology and building materials handaxes and pottery vessels increases with time (Saragusti et al., (e.g. Aalen, 1966; Agorsah, 1985; Laksmi, 2006; Rumana, 2007). Author's personal copy

1342 R. Winter-Livneh et al. / Journal of Archaeological Science 40 (2013) 1340e1353

A good archaeological example for the examination of house shape Ghassulian (Teleilat Ghassul, Shiqmim, Abu Hamid) and Golanian and its possible implication on past societies’ social and economic (Rasam Harbush) architecture (Fig. 1), because they provide rela- organization can be found in the transition from circular huts to tively well defined architectural remains, that have been frequently rectangular houses. Flannery (1972) argued that this transition studied during recent decades. reflects population growth and intensification of production during The Chalcolithic architectural unit is usually composed of Pre-Pottery Neolithic times, and the development of privatized a single room, located on the narrow side of a rectangular or storage, which is more effective in rectangular houses (Flannery, trapezoidal walled courtyard (e.g., Teleilat Ghassul, Fasa’el, Meser, 1972:38e46, 2002:418e422). Following Wills (1992: 169), who Golan sites). These rooms are rectilinear in shape and often termed suggests that reduced sharing, more restricted land tenure, and “broad-room” or “broad house” (e.g. Gilead, 1988:416, Porath, growing privatization of storage greatly increased the economic 1992:41; Rowan and Golden, 2009: 29) indicating that the options of early farmers, Flannery now stresses that archaeologists entrance to these structures was located in one of the long walls. should expect to find “a lot more variation in house size, house Installations such as hearths and silos are usually located in front of shape, storage facilities.and other features.” (Flannery, 2002: these structures, sometimes enclosed by courtyard walls. The small 423). On the other hand Saidel (1993) argues that the change from structures were probably used for sleeping and storage while other round to rectilinear dwellings indicates social change which results daily activities were carried out in the courtyards or adjacent to the from the “combined effect of anticipated mobility, agriculture and dwelling unit. Although structures are variable, neither size small craft production resulted in an increased mode of production differentiation nor size hierarchies are apparent (Gilead, 1988: and a change in social organization.” (Saidel, 1993: 96). Hunter- 417e418). Structures in a site may be organized in different ways Anderson (1977) study sought a functional interpretation to (Banning, 2010). In the Chalcolithic, there are sites where the round and rectilinear houses. Accordingly, within round dwellings architectural units are clustered, such as in Teleilat Ghassul and few of the activities were shared by the group or recognized as Shiqmim. In other sites, such as Gilat (Levy, 2006) or Grar (Gilead, a person’s identity, meaning there was low level of social or task- 1995), the units are less clustered. In the Golan sites the struc- role differentiation within such society. While, within rectilinear tures are arranged in lines, or as labeled by Epstein “house building the architectural differentiation of internal space have chains”(1998:6e8). helped coping with the spatial and social problems inherent when One of the great challenges of Chalcolithic period research has multiple contemporary activities needed to be carried out within been to interpret the architectural remains in terms of social and a single building. The segmentation of internal space and the economic organization. Levy (1986a), for one, suggests that “the making of rooms prevented interference and disturbances between layout of Shiqmim, for example, may indicate the presence of contemporary activities (Hunter-Anderson, 1977: 305). a single decision maker.” and that “.a similar pattern can be Here, by using CSM, we examine change in symmetry through observed at Teleilat Ghassul in the Jordan Valley.” (Levy, 1986b: 11). time. We ask whether symmetry increases in later periods, Thus, the ”Chalcolithic settlements.are characterized by planned whether similar levels of symmetry can quantified in different sites villages” (Levy,1995: 229), that indicate “.the presence of a central of the same cultural entity, in trying to evaluate whether archi- authority” (Levy, 1986a: 88) and “the emergence of a new social tecture building and planning techniques have been the product of organization, the first ‘chiefdoms’ in Palestine” (Levy, 1995: 238). In accumulated knowledge over time. By using SR our aim is to addition he suggests that Shiqmim’s architecture “preview many of examine the variability of architectural shapes within and between the architectural features that are fully developed at the northern settlements in order to illuminate, social, cultural and economic Negev urban site at Arad in the following Early Bronze Age.” (Levy aspects of late prehistoric communities in the southern Levant. and Alon, 1985: 78). Gilead (1988), on the other hand, suggests that More specifically we attempt to determine the socio-economic “.structures and settlement.were unplanned and there is no correlates of the dwellings forms, whether they can be traced in clearly observed structural hierarchy” (Gilead, 1988: 418). Epstein different sites or cultures, and of the degree shape variation within (1998: 7), on the basis of her excavations in the Golan, writes that a single cultural entity. “.none of the [mostly structural] evidence can be interpreted as indicating status differentiation“ and that their standardized 2. Architecture of the Chalcolithic period in the southern dwellings indicate that “the community was egalitarian”. She also Levant suggests that “the Golan structures resemble those of Teleilat Ghassul” (Epstein, 1977: 58). Porath also notes the similarity Several cultural entities existed in the southern Levant during between the Chalcolithic domestic architecture at different sites, the Chalcolithic period, of which the Ghassulian culture (ca. 4500e and states that “their basic plan was similar” (1992: 40). Bourke 3900 B.C.) was the most prominent (Gilead, 2011; Lovell, 2001; (2001), based on a close examination of Mallon’s Tulayl 1 settle- Rowan and Golden, 2009). The Ghassulian culture features ment plan, argues that “there is considerable variation in the size, assemblages broadly similar to those found at the upper levels of shape and elaboration of construction in individual dwellings.” Teleilat Ghassul, the Ghassulian type-site (North, 1961). This culture (Bourke, 2001: 120) and that the proposition that it may “reflect the extends from the Northern Negev and the Dead Sea basin, to the development of elite residential complexes.is not unreasonable.” Shephela, the coastal plain, and the Jordan valley and the Galilee. (Bourke, 2002: 22). Banning (2010) use space syntax analysis on Other Chalcolithic cultural entities are the Besorian, the Timnian late NeolithiceChalcolithic settlements to find “variation in the and the Golanian (Epstein, 1998; Gilead, 2011; Rosen, 2011). Since built environment over this period” which he suggests indicate the Timnian architecture is poor and it is geographically limited to “political and economic inequalities beyond those determined by the arid zones of the Southern and Central Negev, the Aravah, gender, age, talent, or ability”. “a degree of socio-economic southern Jordan and Eastern and Southern Sinai (Rosen, 2011), it is ranking” (Banning, 2010: 79). Thus there is little agreement on not included in our analysis. Also excluded is the Besorian since it the meaning of structural patterning in this period, at least partially predates the Ghassulian (Gilead, 2007) and its architectural because there is no objective means of analysis and comparison. remains are meager. The Golanian sites are located in the Golan The characterization of patterns of settlement layouts, and Heights, and are contemporary with the Ghassulian, but feature conclusions concerning their implications, should be treated with a remarkably different and locally manufactured ceramic assem- reserve mainly due to the fact that most settlement layouts are blage (Epstein, 1998). The following analysis will concentrate on fragmentary and come only from a small section of the site. This is Author's personal copy

R. Winter-Livneh et al. / Journal of Archaeological Science 40 (2013) 1340e1353 1343

Fig. 1. Map of the southern Levant showing the locations of the studied sites and other Chalcolithic sites.

due mainly to the poor preservation of mudbrick. The probability 3. Methodology that smaller and ephemeral sites are underrepresented while large sites may be exceptional rather than typical cannot be excluded. The methods used here comprise four elements: (i) data Even when a larger part of sites had been excavated and the acquisition and digitization of architectural ground-plans from site architecture is relatively well-preserved (e.g. Teleilat Ghassul e reports; (ii) application of the Continuous Symmetry Measure Tulayel 1 and 3 which were excavated between 1929 and 1938), method (CSM) to evaluate structures symmetry; (iii) application of stratigraphic uncertainties and errors are common, and we cannot Shape Reproducibility (SR) for quantifying the similarly between be sure that all the building attributed to a phase or stratum were structure shapes; and (iv) evaluating the results in terms of Chal- contemporary (Banning, 2010:53e54). colithic period social organization. We use Matlab and Geographic Nevertheless, structures at Teleilat Ghassul are according to Information System (GIS) platform for these analyses (for previous Epstein, on the one hand, similar to those of the Golan sites, and GIS based analysis on Chalcolithic period dataset see: Fletcher, thus reflect a society that is undifferentiated in terms of status. 2008; Fletcher and Winter, 2008; Fletcher et al., 2008; Pierce, According to Levy, on the other hand, Teleilat Ghassul is similar to 2006; Winter-Livneh et al., 2010, 2012). Shiqmim, and both represent a chiefdom society. Finally, according to Bourke, Teleilat Ghassul demonstrates significant variability 3.1. Dataset which implies a hierarchical society, although not a chiefdom. Here, we present a detailed analysis of the architectural remains which The analysis presented here is based on a sample of 99 structural provides additional important information that may illuminate units, retrieved from four Chalcolithic sites (Teleilat Ghassul, Abu further some of these social and cultural issues. Hamid, Shiqmim, and Rasam Harbush) and one Early Bronze Age Author's personal copy

1344 R. Winter-Livneh et al. / Journal of Archaeological Science 40 (2013) 1340e1353 site (Arad) which is used here as a comparative dataset (Fig. 1; In Tulayl 3, two phases of occupation with architectural remains Table 1). Teleilat Ghassul (Fig. 2a) is located in the southern Jordan were excavated (Koeppel et al., 1940: Plan I and II). The PBI exca- Valley, some three kilometers north-east to the Dead Sea and vations are of considerable importance and they are a cornerstone consists of several differentiated hillocks (Tulayl) which cover an in the research of Chalcolithic architecture. However, we excluded area of approximately 25 ha. Each hillock contains a cluster of Tulayl 3 from our analysis since the stratigraphy of each architec- dwellings which includes a small structured unit located on the tural phase is very problematic, and later expeditions found it narrow side of a walled courtyard. Excavations started in the late difficult to define the stratigraphic horizons clearly (Bourke, 2002: 1920s (Mallon et al., 1934), and it has been intensively researched 2e3, 2007). In addition, we excluded architectural elements that since (see Bourke, 2001:107e111, 2002:2e5; Lovell, 2001). are too fragmentary. Only structures that more than 80% of their Abu Hamid (Fig. 2b) is located in the central Jordan Valley, contours were preserved are included in the dataset. 0.5 km east of the Jordan River. The excavations of the site started in The digitization process is based on the manually drawn 1986 by Dollfus and Kafafi (Dollfus and Kafafi, 1986; Lovell et al., ground-plans published by the excavators. Each ground-plan was 2007: 51) and three main levels of occupation were unearthed. scanned (Fig. 5a.i) and transformed into a GIS platform as layers for Phase II is a subject of dispute between Garfinkel (1999: 158) who further modifications. The procedure consists of defining the layout defines it as Ghassulian, and Lovell et al. (Lovell et al., 2007: 63, 74) of each illustrated unit (i.e. room). This was done by identifying the who suggest it is pre-Ghassulian. Thus, we limit our analysis to the corners of the structure with each shape’s points (vertices) remains of phase III, the undisputed Ghassulian phase at Abu (Fig. 5a.ii), and initiating straight line segments between them that Hamid. The settlement includes rectangular houses, often unicel- form a closed polygonal surface, an accurate presentation of the lular, built either completely in mud bricks or in some cases on architectural shape (Fig. 5a.iii). To enable comparisons between the a stone foundation. The dwellings are separated from each other by shapes and to avoid size effects, we normalized the original large spaces, sometimes delimited by long, stone walls (Lovell et al., structures sizes into the same equivalent absolute size (Fig. 5b). 2007: 57). This was done by dividing each edge length by the square root of The northern Negev site of Shiqmim (Fig. 2c), about 18 km the shape’s area. south-west of Beer-Sheva, is one of the largest Chalcolithic sites in the area. Seven seasons of excavations directed by Levy and 3.2. Continuous Symmetry Measure (CSM) Alon during 1979e1993 uncovered a Chalcolithic village with subterranean structures and a nearby cemetery. The settlement CSM was first introduced to archaeological research by Saragusti consists of above ground rectangular dwellings structures as et al. (1998) and is described in detail by Zabrodsky and Avnir well as underground structures in the northern part of the site. The (1995). In our study we use this method to measure the distance above structures can be divided into two main sizes: small between locations of points in the original architectural shape and (ca. 2.5 5.5 m) and large (5.5 10 m). They are built of mud bricks their locations as designated by the nearest mirror-symmetrical with stone foundation, associated with rectangular courtyards. shape (bilateral symmetry). These points are the boundary points Small finds within these structures indicate the domestic nature of or vertices that provide and design the shapes’ general boundary food preparation and consumption and other activities. Some of the line. The number of points is determined according to the nature of buildings are associated with copper working (Levy, 1987; Levy and the architectural shape providing a small number of vertices which Alon, 1982, 1985). represent the cornerstone locations. Next, we find the reflection Rasam Harbush, in the Golan Heights, is the largest of a number line that will cause the minimal move of points to create the nearest of Golanian sites (Fig. 2d). The site was excavated by Epstein (1977, mirror symmetrical shape. The line is determined from the aver- 1998) and it consists of rectangular dwellings, built of local stone, aged sum of each pair of coordinate points. According to this attached to rectangular courtyards. The site dwellings are posi- reflection line, each subset of points is duplicated. Redrawing the tioned one next to the other from east to west forming together outlines according to these points, results in a polygon of a mirror- several parallel lines of dwellings which are referred by Epstein as symmetrical shape (Fig. 4). This “foldingeunfolding” method ‘house chains’ (1998:6e8). (explained in detail by Zabrodsky and Avnir [1995]; and see also The Early Bronze Age town of Arad (ca. 3000e2650 B.C.) is Saragusti et al., 1998: Appendix 1) is repeated for each set of two located in the northern Negev, about 30 km east of Beer-Sheva points. The polygon of the mirror-symmetrical shape closest to the (Fig. 3). Arad is a large fortified urban settlement that consists of original shape is chosen for comparison with the original shape. residential units separated by streets and alleys. Dwellings are of Symmetry measurement (Saragusti et al., 1998:819)isdefined in many sizes, the smallest ca. 50 sq m and the largest 150 sq m. There Equation (1):

are also public structures in Arad, including a water system and ˇ a temple complex (Amiran and Ilan, 1996; Ben-Tor, 1973; Herzog, Xn 1980). SðGÞ¼ Pi Pi (1) The areas covered by the plans do not represent the extension of i the entire sites but rather the parts subjected to our analyses.

Repeated excavations at Teleilat Ghassul have been largely confined Where P is the location of the n vertices of the original shape i ˇ 2 fi to the hillock known as Tulayl 1 of which some 3,500 m with rich con guration, and Pi are the corresponding points in the artifactual assemblages and mudbrick architecture were unearthed. symmetric configuration. Note that the size is normalized in order

Table 1 The dataset table includes four Chalcolithic sites and one Early Bronze Age II site.

Site Excavated area size (m2) Number of structures Culture Date Period Reference Teleilat Ghassul (Tulayl 1) 3500 21 Ghassulian 4500e3900 B.C. Chalcolithic Mallon et al., 1934: Fig. 12 Shiqmim 4500 15 Ghassulian 4500e3900 B.C. Chalcolithic Levy, 1987: Fig. 6.2 Abu Hamid (phase III) 1350 7 Ghassulian 4500e3900 B.C. Chalcolithic Lovell et al., 2007: Fig. 2 Rasam Harbush (Golan site 12) 4500 18 Golanian 4500e3900 B.C. Chalcolithic Epstein, 1998: Plan 1a Arad (strata II-I) 5500 38 Early Bronze II 3000e2650 B.C. Early Bronze II Amiran and Ilan, 1996: Plate 70 Author's personal copy

R. Winter-Livneh et al. / Journal of Archaeological Science 40 (2013) 1340e1353 1345

Fig. 2. Sites ground-plans: (a) Teleilat Ghassul (after Mallon et al., 1934: Fig. 12); (b) Abu Hamid (after Lovell et al., 2007: Fig. 2); (c) Shiqmim (after Levy, 1987: Fig. 6.2); (d) Rasam Harbush (Epstein, 1998: Plan 1a). Author's personal copy

1346 R. Winter-Livneh et al. / Journal of Archaeological Science 40 (2013) 1340e1353

to avoid size difference influencing our results. The resulting S(G) value equals zero if the shape has the desired absolute symmetry, i.e., represents the highest symmetrical quality. The symmetry value, S(G), increases as the shape departs from its nearest symmetric configuration, i.e., is more asymmetrical or indicates a low symmetrical quality.

3.3. Shape Reproducibility (SR)

To illuminate aspects such as continuity, repetition or repro- ducibility, and standardization of building activity, we developed the Shape Reducibility measure (SR) which measures the similarity of architectural shapes within and between different sites. SR measures how much each structure’s shape is similar to another shape. The result for each architectural shape is always relative to the shape (structure) it is compared to. A simple way to explain the idea of SR is to describe it as a process in which the original ground- plans of two different structures are normalized and placed one above the other. Then, by rotating one of them, we seek to find the optimal orientation which will provide the largest overlapping area resulting in the smallest residual area. Thus, we can measure these residuals and compare them to residuals of other structures measured in the same technique. The smaller the residual is, the similar the structures are. The important aspect of SR is the ability to measure exactly how much a given shape is similar to another Fig. 3. Arad ground plan (Amiran and Ilan, 1996: Plate 70). shape rather than describing it in subjective terms. Below we describe each step of the SR methodological process (Fig. 5).

3.3.1. Normalized area The first stage of the process is to normalize the area of all shapes to an equal arbitrary one (Fig. 5b). This stage is critical because it allows direct comparison of structure’s shapes with different areas size.

3.3.2. Central point Next, we calculate the central point for each shape. This allows us to position each pair of structures one on top of the other according to their shape centers. This is done while the original orientations of the structures are left unchanged (Fig. 5c).

3.3.3. Rotation matrix Next, one of the shapes is rotated by 5 at a time and the residual area between the two shapes is calculated for all possible orien- tation from 0 to 360 (Fig. 5d). The turning of the shape by an angle about a fixed point is defined after Arfken (1985: 195) by the rotation matrix in Equation (2): cos q sin q Rq ¼ (2) sin q cos q

Where R is the matrix that rotates around a given vector coordi- nates system by a counter clockwise angle of q relative to a fixed set of axes. When the orientation in which the residual area between the two shapes is the smallest, it is selected to represent the SR value. We repeated this process for all possible pairs of structures units (n ¼ 99).

3.3.4. Similarity matrix

Fig. 4. Continuous Symmetry Measure e “foldingeunfolding” of the shapes: (a) The After collecting SR values for all possible pairs of structures, data original outline of the shape. (b) Through the center of each pair of points a reflection were arranged into a similarity matrix (i.e., table) and the values of line (dashed) is positioned and the shape is divided accordingly (yellow). (c) The structures from the same site were clustered together. Similarity mirror-symmetrical shape (yellow) produced by redrawing the outline according to matrix is illustrated in Fig. 5e where 0 or blue indicate maximum the shape’s points (blue) and the reflection line. (d) Measuring the distance (yellow) similarity and 1 or red indicate minimum similarity. This arrange- between the original shape vertex coordinates (red) and the mirror-symmetrical fi shape’s vertex coordinates (blue). (For interpretation of the references to color in ment simpli es the calculation of three different measurements this figure legend, the reader is referred to the web version of this article.) that were used: (i) intra-site similarity e the mean of the similarity Author's personal copy

R. Winter-Livneh et al. / Journal of Archaeological Science 40 (2013) 1340e1353 1347

indices at the site identity diagonal; (ii) inter-site similarity e the means of the similarity indices that lay outside the site identity diagonal; and (iii) intra-site deviation e to measure whether the intra-site level for each site is lower than the inter-site levels the mean of the intra-site similarity index the intra-site mean of this site was subtracted from its corresponding inter-site mean values. In addition, we classify the structure shapes of the Chalcolithic period in order to locate the most typical or prototype architectural shape among them. This is achieved by calculating the mean of each column of the similarity matrix, which equals the mean similarity of each individual structure with the rest of the struc- tures included in the study. Next, we sort structures by their mean from the one which provided the lowest SR mean value (the most typical structure) to the one which provided the highest SR mean value (the least typical structure). In summary, this new method provides a quantitative and automatic approach that enables this present study, and future research, to objectively and systemati- cally evaluate the similarity of prehistoric architectural shapes within and between sites.

4. Results

CSM mean values of all the studied sites indicate a significant difference between Teleilat Ghassul and the Ghassulian sites of Shiqmim and Abu Hamid (student t-test p < 0.03) (Fig. 6). Namely, the higher value of CSM of Teleilat Ghassul indicates greater levels of asymmetry than in the other two Ghassulian sites. Similar CSM values have been found between the Golanian site Rasam Harbush and the Ghassulian sites of Shiqmim and Abu Hamid (student t-test p > 0.05), which indicate similar symmetrical levels to the archi- tecture of sites of different cultural entities. In addition, the results indicate a significant difference between Chalcolithic symmetry values and that of Early Bronze Age Arad (student t-test p < 0.001), suggesting that the degree of asymmetry in Arad is significantly higher than that of the Chalcolithic sites. Our intra-site analysis indicate a significant difference between the Ghassulian sites (Teleilat Ghassul, Shiqmim, Abu Hamid) and the Golanian site Rasam Harbush (student t-test p < 0.01) (Fig. 7). This suggests that the homogeneity of architectural shapes at the Golanian site of Rasam Harbush is significantly higher than that of the Ghassulian sites. Additional significant difference was found between Arad and all Chalcolithic sites (student t-test p < 0.00005). Even after excluding Rasam Harbush, a significant difference between Arad and the Ghassulian sites was found (student t-test p < 0.003). This suggests that Ghassulian sites are more homoge- nous architecturally compared to that of the Early Bronze Age. Next, we explore the architectural uniqueness of the sites by normalizing their intra-site SR value (discussed above) and comparing it with their inter-site SR values (as described in Fig. 5d). In Fig. 8 the SR mean value of each site is converted to zero (a dashed orange line), while the mean values of its inter-site

center (ii) rotating one of the shapes (red) 5 at a time until completing a full turn (iii) finding the structure orientation (yellow) which provides the smallest remnant areas (¼SR value) comparing to all possible orientation. (e) Similarity matrix: the illustration shows the similarity index between all structures at three different sites (AeC). Notice that the values at the diagonal are zero (maximum similarity or similarity between a structure and itself) and that the squares along the diagonal of the similarity within Fig. 5. Shape Reproducibility method summery: (a) Data acquisition process (Teleilat same-site structure are lower than the similarity found in other areas of the matrix, Ghassul on top Rasam Harbush on bottom): (i) scanned illustration of the original which are between different sites. (i) Calculation of intra-site similarity, based on the ’ published architecture (ii) vertexes (blue) represent the structure shape s corners mean of the similarity indices at the site identity diagonal; (ii) inter-site similarity, (iii) closed polygonal surface of the architecture shape based initiating straight line based on the mean of the squares outside the site identity diagonal, and can be used to ’ segments between the vertexes; (b) Normalizing of the areas: (i) the structure s shapes compare the similarity among multiple sites; (iii) to test whether intra-site similarity is polygons in their original size (ii) the polygonal surface after normalizing its size; lower than inter-site similarity, the intra-site similarity mean (at the figure mark for (c) Center point: (i) calculating the center point for each polygon (ii) placing each pair site ‘B’) is subtracted from the corresponding column (the second column at the of structures one on top of the other according to their center points. (d) Rotation illustration). (For interpretation of the references to color in this figure legend, the process: (i) a pair of structures one (blue) on top of the other (red) according to their reader is referred to the web version of this article.) Author's personal copy

1348 R. Winter-Livneh et al. / Journal of Archaeological Science 40 (2013) 1340e1353

Fig. 9 illustrates the similarities between the Chalcolithic sites by calculating the mean similarity indices between each pair of sites. The smaller the SR value is, the more it resembles the paired site. These results demonstrate exactly how much the Ghassulian sites Shiqmim, Teleilat Ghassul and Abu Hamid are similar to one another and differ from Rasam Harbush and Arad. In Fig. 10 the Chalcolithic period architectural shapes are dis- played and classified in decreasing order according to their level of similarity with all the rest of the Chalcolithic period structures within our database. The data sample (n ¼ 61) includes four Chal- colithic sites. According to this analysis the most typical dwelling shape is found at Teleilat Ghassul: structure 23 depicted in Fig. 12 of Mallon et al. (1934), with a width:length ratio of 1:2.2 (Fig. 10: top left corner). The least typical structure shape, a very narrow rect- angle, on the other hand, is found in Rasam Harbush (Epstein 1998: plan 1a, structure 7) with a ratio of 1:4.5 (Fig. 10: rightmost on bottom row).

5. Discussion Fig. 6. Continuous Symmetry Measure results: shown are the mean s.e.m. of the CSM value of each site (* is for p < 0.03, and ** is for p < 0.001). We have presented here a new method, Shape Reproducibility, to examine and identify intra- and inter-site similarities of struc- similarity to the remaining sites vary between 0.15 and 0.05 tures on the basis of prehistoric architectural ground-plans. We (below or above the site’s mean value). Accordingly, sites which are found that Ghassulian communities built relatively similar struc- located above the zero line indicate dissimilarity to the specific site tures which could not be distinguished from one site to another. while sites located below the zero line indicate similarity to the site. The Golanian community shaped their structures differently, thus These results display a similar pattern of both higher and lower producing unique structure shapes which are significantly different inter-site SR means for all Ghassulian site. These results indicate from those of the Ghassulian. In addition, we found that the intra- that the architectural shapes within these sites cannot be distin- site variability of the architectural shapes of the Ghassulian culture guished from one another. But comparing them to Rasam Harbush is significantly higher than the intra-site variability level of Gola- or Arad indicates they could be distinguished from them. nian culture. However, in comparison with Early Bronze Age Arad, Rasam Harbush on the other hand displays a different pattern Ghassulian intra-site variability is significantly lower, demon- since it consists only of higher inter-site normalized mean values. strating that Chalcolithic architecture is more homogenous than This indicates that it is different from the other sites, suggesting the that of the Early Bronze Age. Furthermore, we found that the degree architecture here is unique. Comparing Arad with the SR mean of symmetry in Chalcolithic architecture is higher than in the values of the Ghassulian sites demonstrates a pattern that is not architecture of the Early Bronze Age. distinguishable from these sites, excluding Rasam Harbush. That is, The CSM results indicate that the symmetry of structures if we were to mix the architecture shapes of each Ghassulian site correlates negatively with time (Fig. 6). It seems that the change with those of Arad it is highly unlikely that we could distinguish from Chalcolithic to Early Bronze Age did not necessarily lead to between them, but if we were to mix the Rasam Harbush archi- more symmetrical structures. This may be interpreted in different tectural shapes with those of Arad we could, in high probability, ways. First, the low levels of symmetrical quality observed in Tel- distinguish them. eilat Ghassul and Arad are not indicative of architectural incom- petence or limitations, but that there was probably no particular technological or physical need for higher symmetry when building these structures. Another possibility is that building materials and local environmental variables have influenced the architectural symmetry. The possibility that corralling played a role in a number of the sites (Epstein, 1998:16e17) cannot be excluded. However, it seems that this role was limited since the remains of human habitations predominate the archaeological record. CSM analysis of architectural remains could be perhaps more effective when applied to longer chronological time spans or to larger geographical areas in order to detect increase in the symmetry. In previous studies, the symmetry of flint artifacts such as Acheulean handaxes, a tool type that was produced during hundreds of thousands of years was examined. It is possible that several hundreds of years are a too short time period to identify a gradual change or increase in the symmetry of prehistoric architecture. Our SR results demonstrate that the intra-site similarity of the Golanian site of Rasam Harbush is higher than that of the Ghas- sulian sites (Fig. 7). This finding is in line with previous studies suggesting a high homogeneity level and architectural standardi- zation of the Golanian architecture (Epstein, 1977, 1998). Further- Fig. 7. Shape Reproducibility intra-site results: shown are the mean s.e.m. of the SR more, we demonstrate that Golanian architecture consists of values within each of the sites (* is for p < 0.03 and *** is for p < 0.00005). unique rectangular proportions which differ from those at the Author's personal copy

R. Winter-Livneh et al. / Journal of Archaeological Science 40 (2013) 1340e1353 1349

Fig. 8. Shape Reproducibility inter-site results: shown are the normalized mean s.e.m. SR values of the intra-site similarity of each site (¼0, and marked with dashed orange line) in comparison with its inter-site normalized mean s.e.m. SR values. Inter-site mean that is lower than 0 indicate similarity to the site (undistinguished), while a higher than 0 result indicate difference (unique). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 9. Shape Reproducibility inter-site results: shown are the normalized mean SR values of each site. The lower the SR value is the more the site resemble to the comparative site. Author's personal copy

1350 R. Winter-Livneh et al. / Journal of Archaeological Science 40 (2013) 1340e1353

Fig. 10. Classification of the Chalcolithic period structures units according to the similarity of each structure with the rest of the structures. The shapes are organized in a decreasing order starting at the top left with the most typical structure (low SR mean) ending at the bottom right with the least typical structure (high SR mean). The dataset includes structures from Teleilat Ghassul (TG), Shiqmim (SH), Abu Hamid (AH), and Rasam Harbush (RH) (n ¼ 61).

other sites (Figs. 8 and 9). Although this was recognized previously perception and the ways of life.” (Rapoport, 1969: 12), and is and supports the claim that the Golanian differs from the Ghas- “governed by a society’s ideas, its forms of economic and social sulian (Epstein, 1998: 334, Gilead, 2011:15e16), it contradicts the organization, its distribution of resources and authority, its activi- statements that Rasam Harbush dwellings are similar to those ties and the beliefs and values which prevail at any one period of found in Teleilat Ghassul (Epstein, 1977; Porath, 1992). Even though time.” (King, 1980: 1). The explanation of the similarity level should Teleilat Ghassul displays a relatively high level of intra-site vari- be, therefore, sought within the social and economic aspects of ability, the SR results indicate that Rasam Harbush architecture is Chalcolithic communities. significantly different (student t-test p < 0.03). The architectural One of the primary interests of the current research in the study uniqueness of Rasam Harbush might reflect its distinctive and of Chalcolithic period is in the organization level of the society. different cultural and socio-economic character. There are researchers who view the archaeological evidence as The Ghassulian sites demonstrate a certain degree of architec- reflecting social hierarchy (Gal et al., 2011; Golden, 2009; Levy, tural intra-site variation. In addition, inter-site comparisons indi- 1986a, 1995). Bourke (2002) suggests that Chalcolithic site struc- cate that Ghassulian architecture, and its intra-site variability ture reflects variability which can imply the existence of residential levels, are similar in the three sites we studied (Figs. 8 and 9). This elite. Levy (1986a: 88) suggests that Chalcolithic architectural intra-site variability level, however, is significantly lower in variability implies central authority and a single decision maker. comparison with the Early Bronze Age site of Arad. These results Based on Service (1962) and Fried (1967) anthropological models, contradict previous assumptions that Ghassulian architecture he argues that Chalcolithic architecture signifies the appearance of consist of considerable variation in shape (Bourke, 2001, 2002; “hereditary chiefdom society” (Levy, 1995: 235). However, the Levy, 1986a, 1995; Levy and Alon, 1985). Nonetheless, the relatively results of our study do not support the idea of hierarchically low degree of variability in shapes during the Chalcolithic period stratified communities. implies a similarity that should be explained. One possibility is that Our results indicate that in the case of Early Bronze Age Arad the this level of similarity could be the outcome of physical or envi- level of intra-site variability is high and may signify a stratified ronmental factors such as topography or building material. This, society. This variability results from the different types of struc- however, is unlikely since the studied settlements here are located tures, including residential structures and public structures, such as within different environments which consist of different topo- the probable temple complex. Architectural variability of Arad is graphical conditions and different available local materials. More- significantly higher than that of the Ghassulian sites and might over, based on anthropological studies, although the physical suggests that the level of organization in the latter is less complex environment could exerts great constraints over the builders and than currently suggested. There is no significant difference the options they had, it appears that these environmental factors between structure forms which can represent habitations of are secondary to the socio-economic ones (Allison, 2002; Ingold, community members of a distinct social status. In addition, Epstein 1995, 2000; King, 1980; Rapoport, 1982). Accordingly, house form (1998: 7) suggests that the very standardized Golanian architecture is considered a “.direct expression of changing values, images, indicate that the structures were built by an egalitarian community. Author's personal copy

R. Winter-Livneh et al. / Journal of Archaeological Science 40 (2013) 1340e1353 1351

According to our findings this assertion cannot be excluded. The social changes. On the other hand, it may, at least in some cases, Golanian site of Rasam Harbush lacks any visible presentation of indicate that architectural techniques and traditions are diverse social differentiation, neither in the architecture nor in other and do not develop unilinearily. aspects of the material culture. It even suggests that the local To conclude, Building activity patterns in prehistory are part of community was less differentiated socially than were the other an expressive and complex system of social and economical Ghassulian communities. meaning. Future studies should reveal further micro and macro Architectural variability may imply an increase in socio- scale social aspects of the builders and dwellers in the southern economic options (e.g. Flannery, 1972, 2002; Kent, 1990a, 1990b; Levant prehistory. In addition studies should address related Rapoport, 1969, 1982; Wilk and Rathje, 1982). Previous studies have important issues such as how the house forms within different emphasized that the Ghassulian subsistence economy witnessed settlements are organized spatially. Moreover, there is a need to an intensification of agricultural production, greater investment objectively measure and compare the similarity not only between in craft production (in ceramics, flint artifacts, basalt working, individual structures, but also between ways clusters of structures and ivory carving) and technological innovations, notably metal- are spatially distributed within each settlement. Future studies lurgy (e.g. Gilead, 1988; Golden, 2009; Levy, 1986a; Rowan and should also aim at finding whether different building material may Golden, 2009). Moreover, Ghassulian economy is also linked to an have influenced the symmetrical properties of architecture. increase in storage (Bourke, 2002; Golden, 2009; Levy, 2003), that Answering these questions can help us to better understand past may have been combined with the appearance of semi- social and economic realities. subterranean features (Gilead, 1988), and a land tenure system (Winter-Livneh et al., 2012). These features may suggest a connec- Acknowledgments tion between the economic structure of Ghassulian society and the fi levels of intra-site variability, levels which are signi cantly higher The authors wish to thank the Pratt Foundation PhD Fellowship fi than the level of the Golanian site of Rasam Harbush and signi - Program at Ben-Gurion University of the Negev. We thank Steve cantly lower than the level found in Early Bronze Age Arad. Rosen and two anonymous readers who read previous drafts and made important comments that improved the paper consid- 6. Conclusions erably. We thank the members of the GI-Lab at Ben-Gurion University of the Negev for providing useful information concern- Ghassulian architecture exhibits a certain degree of variability, ing GIS procedure. high in relation to that of Rasam Harbush and low in relation to that of Early Bronze Age Arad. The available evidence suggests that social hierarchy is not apparent architecturally in Ghassulian References society. Nonetheless the relatively higher level of architectural Aalen, F.H.A., 1966. The evolution of the traditional house in western Ireland. The differentiation within Ghassulian settlement than the Golanian Journal of the Royal Society of Antiquaries of Ireland 96, 47e58. settlement might reflect socio-economic intensification. In addi- Agorsah, K., 1985. Archaeological implications of traditional house construction tion, Ghassulian architectural shapes cannot be distinguished from among the Nchumuru of northern Ghana. Current Anthropology 26, 103e115. Alexander, C., 1964. Notes on the Synthesis of Form. Harvard University Press, one site to the other, while those of the Golanian site Rasam Har- London. bush are unique and therefore distinguishable from all the other Allison, P. (Ed.), 2002. The Archaeology of Household Activities. Routledge, London sites we examined. This may support the contention that the and New York. Amiran, R., Ilan, O., 1996. Early Arad II. The Israel Museum and The Israel Explo- Golanian is a cultural entity which differs from the Ghassulian. We ration Society, Jerusalem. suggest that the differences between Ghassulian and Golanian Arensburg, B., Bar-Yosef, O., 1973. Human remains from Ein Gev I, Jordan Valley, architecture probably results from the different nature of the social Israel. Paléorient 1, 201e206. and economic characteristics of each of these two cultural entities. Arfken, B.G., 1985. Mathematical Methods for Physicists, third ed. Academic Press, Boston. The SR method helps in distinguishing different architectural Banning, E.B., 2010. Houses, households, and changing society in the Late Neolithic patterns that probably reflect socio-economic features of the past and Chalcolithic of the Southern Levant. Paléorient 36, 49e87. communities we studied. The Golan site of Rasam Harbush is char- Banning, E.B., Byrd, F.B., 1987. Houses and the changing residential unit: domestic architecture at PPNB ’Ain Ghazal, Jordan. Proceedings of the Prehistoric Society acterized by an intra-site structural homogeneity. The shape of 53, 309e325. structures here was deliberately or undeliberately restricted to Banning, E.B., Byrd, F.B., 1988. Southern Levantine pier houses: intersite architec- basically one shape and the resulted is a particular settlement type. tural patterning during the Pre-Pottery Neolithic B. Paléorient 14, 65e72. Bar-Yosef, O., 1992. Building activities in the prehistoric periods until the end of the Thus, Rasam Harbush lacks any indication of socio-economic Neolithic period. In: Kempinski, A., Reich, R. (Eds.), The Architecture of Ancient differentiation. Partial heterogeneity was revealed in Ghassulian Israel. Israel Exploration Society, Jerusalem, pp. 31e39. sites such as Shiqmim and Teleital Ghassul and suggests a certain Ben-Shlomo, D., Garfinkel, Y., 2009. Sha’ar Hagolan and new insights on Near Eastern proto-historic urban concepts. Oxford Journal of Archaeology 28, level of differentiation between the dwellings within each of the 189e209. settlements. This low-level heterogeneity may signify variability in Ben-Tor, A., 1973. Plans of dwellings and temples in Early Bronze Age Palestine. EI a number of socio-economic aspects, such as intensity of production, 11, 92e98. Binford, L., 1990. Mobility, housing and environment: a comparative study. Journal craft specialization and technological innovation, different activities of Anthropological Research 46, 119e152. or different production intensities resulted in structures with for Bourke, J.S., 2001. The Chalcolithic period. In: Macdonald, B., Adams, R., example, different storage requirements, and thus of different Bienkowski, P. (Eds.), The Archaeology of Jordan. Sheffield Academic Press, fi e shapes. A relatively high level of heterogeneity was observed among Shef eld, pp. 107 163. Bourke, J.S., 2002. The origins of social complexity in the southern Levant: the structures at Arad. This suggests a significant visible dissimi- new evidence from Teleilat Ghassul, Jordan. Palestine Exploration Quarterly larity which is a result of structures having different function. Public 134, 2e27. structures (such as the temple complex) and dwellings were Bourke, J.S., 2007. The Late Neolithic/Early Chalcolithic transition at Teleilat Ghas- sul: context, chronology and culture. Paléorient 33, 15e32. designed differently to express socio-economic distinction. Bridgeman, B., 2002. Artifacts and cognition: evolution or cultural progress? Changes in symmetry levels of prehistoric architecture over Behavioral and Brain Sciences 25, 403. time do not support the idea that accumulation of knowledge over Byrd, F.B., 1994. Public and private, domestic and corporate: the emergence of the southwest Asian village. American Antiquity 59, 639e666. time results always in technological progress. On the one hand, it is Carsten, J., Hugh-Jones, S., 1995. About the House: Levy-Strauss and Beyond. plausible that the symmetry of prehistoric structures fails to reflect Cambridge University Press, Cambridge. Author's personal copy

1352 R. Winter-Livneh et al. / Journal of Archaeological Science 40 (2013) 1340e1353

Clarke, D.L., 1968. Analytical Archaeology, first ed. Methuen & Co Ltd., London. Goring-Morris, A.N., Belfer-Cohen, A., 2008. A roof over one’s head: developments Correa, C., 1982. Architecture in a warm climate. Architecture in Development, in Near Eastern architecture across the EpipalaeolithiceNeolithic transition. In: 31e35. Bocquet-Appel, J., Bar-Yosef, O. (Eds.), The Neolithic Demographic Transition Dickens, P., 1977. An analysis of historical house-plans: a study at the structural and Its Consequences. Springer, Berlin, pp. 239e286. level (micro). In: Clarke, D.L. (Ed.), Spatial Archaeology. Academic Press, London, Grosman, L., Sharon, G., Goldman-Newuman, T., Smikt, O., Smilansky, U., 2011. pp. 33e45. Studying post depositional damage on Acheulian bifaces using 3-D scanning. Dobres, M.-A., 2010. Archaeologies of technology. Cambridge Journal of Economics Journal of Human Evolution 60, 398e406. 34, 103e114. Grosman, L., Smikt, O., Smilansky, U., 2008. On the application of 3-D scanning Dobres, M.-A., Hoffman, R.C., 1994. Social agency and the dynamics of prehistoric technology for the documentation and typology of lithic artifacts. Journal of technology. Journal of Archaeological Method and Theory 1, 211e258. Archaeological Science 35, 3101e3110. Dobres, M.-A., Hoffman, R.C., 1999. The Social Dynamics of Technology: Practice, Herzog, Z., 1980. Functional interpretation of the broadroom and longroom house Politics, and World Views. Smithsonian Institute Press, Washington, DC. types. Tel Aviv 7, 82e89. Dollfus, G., Kafafi, Z., 1986. Abu Hamid, Jordanie. Premiers resultats. Paleoriént 12, Hillier, B., Hanson, J., 1984. The Social Logic of Space. Cambridge University Press, 91e100. Cambridge. Donley, L.W., 1982. House power: Swahili space and symbolic markers. In: Hodder, I., 1994. Architecture and meaning: the example of Neolithic houses and Hodder, I. (Ed.), Symbolic and Structural Archaeology. Cambridge University tombs. In: Pearson, M.P., Richards, C. (Eds.), Architecture and Ordered Press, Cambridge, pp. 63e73. Approaches to Social Space. Routledge, London, pp. 73e86. Dothan, M., 1959. Excavations at Horvat Beter (Beersheba). ’Atiqot 2, 1e71. Hodgson, D., 2011. The first appearance of symmetry in the human lineage: where Durham, P., Lewis, S., Shennan, S., 1995. Artefact matching and retrieval using the perception meets art. Symmetry 3, 37e53. Generalised Hough Transform. In: Wilcock, J., Lockyear, K. (Eds.), Computer Hunter-Anderson, R., 1977. A theoretical approach to the study of house form. In: Applications and Quantitative Methods in Archaeology 1993. BAR International Binford, L. (Ed.), For Theory Building in Archaeology. Academic Press, New York, Series 598, pp. 25e30. pp. 287e316. Epstein, C., 1977. The Chalcolithic culture of the Golan. Biblical Archaeologist 40, Ingold, T., 1990. Society, nature and the concept of technology. Archaeological 56e62. Review from Cambridge 9, 5e17. Epstein, C., 1998. The Chalcolithic Culture of the Golan. IAA Reports No. 4. Jerusalem. Ingold, T., 1995. Building, dwelling, living: how animals and people make them- Faust, A., Bunimovitz, S., 2003. The four room house: embodying Iron age Israelite selves at home in the world. In: Strathern, M. (Ed.), Shifting Contexts: Trans- society. Near Eastern Archaeology 66, 22e31. formations in Anthropological Knowledge. Routledge, London, pp. 57e80. Finlayson, B., Kuijt, I., Mithen, S., Smith, S., 2011a. New evidence from southern Ingold, T., 2000. The Perception of the Environment: Essays on Livelihood, Dwelling Jordan: rethinking the role of architecture in changing societies at the begin- and Skill. Routledge, London. ning of the Neolithic process. Paléorient 37, 123e135. Jones, P.R., 1980. Experimental butchery with modern stone tools and its relevance Finlayson, B., Mithen, J.S., Najjar, M., Smith, S., Maricevic, D., Pankhurst, N., for Paleolithic archaeology. World Archaeology 12, 153e165. Yeomans, L., 2011b. Architecture, sedentism, and social complexity at Pre- Karasik, A., 2010. A complete, automatic procedure for pottery documentation and Pottery Neolithic A WF16, Southern Jordan. PNAS 108, 8183e8188. analysis. In: Computer Vision and Pattern Recognition Workshops (CVPRW), Fitch, J.M., Branch, D.P., 1960. Primitive architecture and climate. Scientific American 2010 IEEE Computer Society Conference on, 2010, pp. 29e34. 203, 134e144. Karasik, A., Smilansky, U., 2008. Scanning technology as a standard archaeological Flannery, K.V., 1972. The origins of the village as a settlement type in Meso- tool for pottery analysis: practice and theory. Journal of Archaeological Science america and the Near East: a comparative study. In: Ucko, P.J., Tringham, R., 35, 1148e1168. Dimbleby, G.W. (Eds.), Man, Settlement and Urbanism. Duckwort, London, Karasik, A., Smilansky, U., 2011. Computerized morphological classification of pp. 23e53. ceramics. Journal of Archaeological Science 38, 2644e2657. Flannery, K.V., 2002. The origins of the village revisited: from nuclear to extended Kempinski, A., Reich, R. (Eds.), 1992. The Architecture of Ancient Israel. Israel households. American Antiquity 67, 417e433. Exploration Society, Jerusalem. Fletcher, R., 1977. Settlement studies (micro and semi-micro). In: Clarke, D.L. (Ed.), Kent, S.,1990a. A cross-cultural study of segmentation, architecture, and the use of space. Spatial Archaeology. Academic Press, London, pp. 47e162. In: Kent, S. (Ed.), Domestic Architecture and the Use of Space: an Interdisciplinary Fletcher, R., 2008. Some spatial analyses of Chalcolithic settlement in Southern Cross-Cultural Study. Cambridge University Press, Cambridge, pp. 127e152. Israel. Journal of Archaeological Science 35, 2048e2058. Kent, S. (Ed.), 1990b. Domestic Architecture and the Use of Space: an Interdisci- Fletcher, R., Winter, R., 2008. Prospects and problems in GIS applications in plinary Cross-Cultural Study. Cambridge University Press, Cambridge. studying Chalcolithic archaeology in Southern Israel. Bulletin of the American Kenyon, K.M., 1957. Digging Up Jericho. Ernest Benn, London. Schools of Oriental Research 352, 1e28. Kenyon, K.M., Holland, T.A., 1981. Excavation at Jericho. In: The Architecture and Fletcher, R., Winter, R., Gilead, I., 2008. Patterning human behaviour in Chalcolithic Stratigraphy of the Tel, vol. 3. British School of Archaeology at Jerusalem, London. southern Palestine: differing scales of analysis. In: Posluschny, A., Lambers, K., King, A.D. (Ed.), 1980. Building and Society: Essays on the Social Development of the Herzog, I. (Eds.), Layers of Perception: Proceedings of the 35th International Built Environment. Routledge and Kegan Paul Ltd., London. Conference on Computer Applications and Quantitative Methods in Archae- Koeppel, R., Mallon, R., Neuville, R., 1940. Teleilat Ghassul II. Pontifical Biblical ology (CAA) Berlin, Germany, April 2e6, 2007. Dr. Rudolf Habelt GmbH, Bonn, Institute, Rome. pp. 257e265. Kohn, M., Mithen, S., 1999. Handaxes: products of sexual selection? Antiquity 73, Fried, M.H., 1967. The Evolution of Political Society. Random House, New York. 518e526. Gal, Z., Shalem, D., Smithline, H., 2011. The Peqi’in cave: a Chalcolithic cemetery in Kuijt, I., Finlayson, B., 2009. Evidence for food storage and predomestication Upper Galilee, Israel. Near Eastern Archaeology 74, 196e206. granaries 11,000 years ago in the Jordan Valley. PNAS 106, 10966e10970. Garfinkel, Y., 1999. Neolithic and Chalcolithic Pottery of the Southern Levant. Kuijt, I., Goring-Morris, N., 2002. Foraging, farming, and social complexity in the Institute of Archaeology, Hebrew University (Qedem 39), Jerusalem. Pre-Pottery Neolithic of the Southern Levant: a review and synthesis. Journal of Garfinkel, Y., Ben-Shlomo, D., 2002. Sha’ar Hagolan architecture in its Near Eastern World Prehistory 16, 361e440. context. In: Garfinkel, Y., Miller, A.M. (Eds.), Sha’ar Hagolan 1: Neolithic Art in Kuijt, I., Guerrero, E., Molist, M., Anfruns, J., 2011. The changing Neolithic household: Context. Oxbow Books, Oxford, pp. 71e84. household autonomy and social segmentation, Tell Halula, Syria. Journal of Garfinkel, Y., Ben-Shlomo, D., Kupermaan, T., 2009. Large-scale storage of grain Anthropological Archaeology 30, 502e522. surplus in the sixth millennium BC: the silos of Tel Tsaf. Antiquity 83, 309e325. Laksmi, G.S., 2006. Local material which influence the traditional house’s form Gero, J., Mazzullo, J., 1984. Analysis of artifact shape using Fourier series in closed against the tropical climate, Toraja traditional house, Indonesia. In: iNTA form. Journal of Field Archaeology 11, 315e322. Conference 2006 e Harmony in Culture and Nature. Gilboa, A., Larasik, A., Sharon, I., Smilansky, U., 2004. Towards computerized Lau, F.G., 2010. House form and Recuay culture: residential compounds at Yayno typology and classification of ceramics. Journal of Archaeological Science 31, (Ancash, Peru), a fortified hilltop town, AD 400e800. Journal of Anthropological 681e694. Archaeology 29, 327e351. Gilead, I., 1988. The Chalcolithic period in the Levant. Journal of World Prehistory 2, Leese, M.N., Main, P.L., 1983. An approach to the assessment of artefact dimension 397e443. as descriptors of shape. In: Haigh, J.G.B. (Ed.), Computer Applications in Gilead, I., 1995. Grar: a Chalcolithic Site in the Northern Negev, Beer-Sheva. Ben- Archaeological Sciences. University of Bradford, Bradford, pp. 171e180. Gurion University Press, Beer Sheva. Lévi-Strauss, C., 1963. Structural Anthropology. Basic Books, New York. Gilead, I., 2007. The Besorian: a Pre-Ghassulian cultural entity. Paleoriént 33, 33e49. Levy, T.E., 1986a. The Chalcolithic period. Biblical Archaeologist 49, 82e108. Gilead, I., 2011. Chalcolithic culture history: Ghassulian and other entities in the Levy, T.E., 1986b. Social archaeology and the Chalcolithic period: explaining social southern Levant. In: Lovell, J.L., Rowan, M.Y. (Eds.), Culture, Chronology and the organizational change during the 4th millennium in Israel. Michmanim 3, 5e20. Chalcolithic: Theory and Transition. Oxbow Books, Oxford, pp. 12e24. Levy, T.E. (Ed.), 1987. Shiqmim I, Studies Concerning Chalcolithic Societies in the Givoni, B., 1969. Man, Climate and Architecture. Elseveir Publishing Company Northern Negev Desert, Israel. BAR International Series 356. Oxford. Limited, Amsterdam. Levy, T.E., 1995. Cult, metallurgy and rank societies e Chalcolithic period (ca. 4500e Golden, J., 2009. Dawn of the Metal Age. Technology and Society during the 3500 BCE). In: Levy, T.E. (Ed.), Archaeology of Society in the Holy Land. Facts on Levantine Chalcolithic. Equinox, London. File, New York, pp. 226e243. Goring-Morris, A.N., Belfer-Cohen, A., 2003. Structures and dwellings in the Upper Levy, T.E., 2003. The Chalcolithic of the Southern Levant. In: Richard, S. (Ed.), Near and Epipalaeolithic (ca. 42e10 K BP) Levant: profane and symbolic uses. In: Eastern Archaeology: a Reader. Winona Lake Ind, Eisenbrauns, pp. 257e267. Soffer, O., Vasl’ev, S., Kozlowski, J. (Eds.), Perceived Landscapes and Built Envi- Levy, T.E. (Ed.), 2006. Archaeology, Anthropology and Cult: the Sanctuary at Gilat, ronments, vol. 1122. BAR International Series, Oxford, pp. 65e81. Israel. Equinox Publishing, London. Author's personal copy

R. Winter-Livneh et al. / Journal of Archaeological Science 40 (2013) 1340e1353 1353

Levy, T.E., Alon, D., 1982. The Chalcolithic mortuary site near Metzad Aluf, northern Rosen, S.A., Eldar, I., 1993. Horvat Beter revisited: the 1982 salvage excavations. Negev desert: a preliminary study. Bulletin of the American Schools of Oriental ’Atiqot 22, 13e27. Research 248, 37e59. Rowan, M.Y., Golden, J., 2009. The Chalcolithic period of the southern Levant: Levy, T.E., Alon, D., 1985. Shiqmim: a Chalcolithic village and mortuary center in the a synthetic review. Journal of World Prehistory 22, 1e92. northern Negev. Paleoriént 11, 71e83. Rumana, R., 2007. Traditional house of Bangladesh: typology of house according to Liming, G., Hongjie, L., Wilcock, J., 1989. The analysis of ancient Chinese pottery and material and location. In: Virtual Conference on Sustainable Architectural porcelain shapes: a study of classical profiles from the Yangshao culture to the Design and Urban Planning, Asia Sustainability Net.upc.edu., 15the24th, Qing dynasty using computerized profile data reduction, cluster analysis and September, 2007. fuzzy boundary discrimination. Computer Applications and Quantitative Saidel, B.A., 1993. Round house or square? Architectural form and socio-economic Methods in Archaeology, 363e374. organization in the PPNB. Journal of Mediterranean Archaeology 6, 65e108. Liu, Z., Kersten, D., 2003. Three-dimensional symmetric shapes are discriminated Saragusti, I., Karasik, A., Sharon, I., Smilansky, U., 2005. Quantitative analysis of more efficiently than asymmetric ones. Journal of Optical Society of America 20, shape attributes based on contours and section profiles in artifact analysis. 1331e1340. Journal of Archaeological Science 32, 841e853. Lovell, J., 2001. The Late Neolithic and Chalcolithic Periods in the Southern Levant: Saragusti, I., Sharon, I., Katzenelson, O., Avnir, D., 1998. Quantitative analysis of the New Data from the Site of Teleilat Ghassul, Jordan. In: BAR International Series symmetry of artefacts: Lower Paleolithic handaxes. Journal of Archaeological 974. Oxford. Science 25, 817e825. Lovell, J.L., Dollfus, G., Kafafi, Z., 2007. The ceramics of the Late Neolithic and Schick, K.D., Toth, N., 2001. Making Silent Stones Speak: Human Evolution and the Chalcolithic: Abu Hamid and the burnished tradition. Paléorient 33, 51e76. Dawn of Human Technology. Weidwnfeld and Nicolson, London. Lycett, J.S., 2008. Acheulean variation and selection: does handaxe symmetry fit Schiffer, M.B., Skibo, M., 1987. Theory and experiment in the study of technological neutral expectations? Journal of Archaeological Science 35, 2640e2648. change. Current Anthropology 28, 595e622. Machin, A.J., Hosfield, R.T., Mithen, S., 2007. Why are some handaxes symmetrical? Service, E.R., 1962. Primitive Social Organization. Random House, New York. Testing the influence of handaxe morphology on butchery effectiveness. Journal Shennan, S., 2006. From cultural history to cultural evolution: an archaeological of Archaeological Science 34, 883e893. perspective on social information transmission. In: Wells, J.C.K., Strickland, S., Mallon, A., Koeppel, R., Neuville, R., 1934. Teleilat Ghassul I. Institute Biblique Laland, K. (Eds.), Social Information Transmission and Human Biology. Taylor & Pontifical, Roma. Francis, New York, pp. 173e189. Manderscheid, A., 2001. The Black Tent in its easternmost distribution: the case of Simao, J., 2002. Tools evolve: the artificial selection and evolution of Paleolithic the Tibetan Plateau. Mountain Research and Development 21, 154e160. tools. Behavioral and Brain Sciences 25, 419. Martin, G., Bar-Yosef, O., 1979. Ein Gev III, Israel, 1978. Paléorient 5, 219e220. Sopher, D., 1964. Landscape and seasons-man and nature in India. Landscape XIII, Mauss, M., Beauchat, H., 1979. Seasonal Variations of the Eskimo. Routledge & Kegan 14e19. Paul, London. Sozen, M.S., Gedik, G.Z., 2007. Evaluation of traditional architecture in terms McPherron, S.P., 2000. Handaxes as a measure of the mental capabilities of early of building physics: old Diyarbakir houses. Building and Environment 42, hominids. Journal of Archaeological Science 27, 655e663. 1810e1816. Mitchell, J.C., 1996. Studying biface utilisation at Boxgrove: Roe deer butchery with Stout, D., Chaminade, T., 2007. The evolutionary of neuroscience of toolmaking. replica handaxes. Lithics 16, 64e69. Neuropsychologia 45, 1091e110 0. Moore, A.M.T., 1985. The development of Neolithic societies in the Near East. Torrence, R., 1989. Time, Energy and Stone Tools. Cambridge University Press, Advances in World Archaeology 4, 1e69. Cambridge. Nadel, D., Werker, E., 1999. The oldest ever brush hut plant remains from Ohalo II, Toth, N., 1990. The prehistoric roots of a human concept of symmetry. Symmetry: Jordan Valley, Israel (19,000 BP). Antiquity 73, 755e764. Culture and Science 1. North, R., 1961. Ghassul 1960: Excavation Report. Pontificium Institutum Biblicum, Valla, F.R., 1988. Aspects du sol de l’abri 131 de Mallaha (Eynan). Paléorient 14, Roma. 283e296. Nowell, A., 2000. The Archaeology of Mind: Standardization and Symmetry in van der Helm, P.A., 2002. Natural selection of visual symmetries. Behavioral and Lithics and Their Implications for the Study of the Evolution of the Human Brain Sciences 25, 422e423. Mind, Ph.D. dissertation, University of Pennsylvania. van der Leeuw, S.E., 1993. Giving the potter a choice: conceptual aspects of Nowell, A., 2002. Coincidental factors of handaxe morphology. Behavioral and Brain pottery technique. In: Lemonnier, P. (Ed.), Technological Choices: Trans- Sciences 25, 413e414. formations in Material Cultures Since the Neolithic. Routledge, London, Oakley, K., 1972. Skill as a human possession. In: Washburn, S.L., Dolhinow, P. (Eds.), pp. 238e288. Perspectives on Human Evolution. Holt, Rinehart and Winston, New York, Vetter, T., Poggio, T., Bültoff, H.H., 1994. The importance of symmetry and virtual pp. 14e50. views in three-dimensional object recognition. Current Biology 4, 18e23. Perrot, J., 1984. Structures d’habitat, mode de la vie et envirnment: Les villages Wagemans, J., 1995. Detection of visual symmetries. Spatial Vision 9, 9e32. souterrains des pasteurs de Be’ersheva dans le Sud d’Israel, au IV millénaires Wilk, R.R., Rathje, L.W., 1982. Household archaeology. American Behavioral Scientist avant l’ère chrétienne. Paleoriént 10, 75e92. 25, 617e639. Pierce, G., 2006. GIS and the Chalcolithic of the Northern Negev. In: Paper delivered Wills, W.H., 1992. Plant cultivation and the evolution of risk-prone economies in the at the Innovations in the Research of the Chalcolithic Beer Sheva Culture prehistoric American Southwest. In: Gebauer, A.B., Price, T.D. (Eds.), Transitions conference, 2006. to Agriculture in Prehistory. Monographs in World Archaeology No. 4. Prehis- Porath, Y., 1987. Dwellings in the Chalcolithic period. In: Katzenstein, H., Netzer, E., tory Press, Madison, pp. 153e176. Kempinski, A., Reich, R. (Eds.), The Architecture of Ancient Israel. Israel Explo- Wilson, P.J., 1988. The Domestication of the Human Species. Yale University Press, ration Society, Jerusalem, pp. 37e44. New Haven and London. Porath, Y., 1992. Domestic architecture of the Chalcolithic period. In: Kempinski, A., Winter-Livneh, R., Svoray, T., Gilead, I., 2010. Settlement patterns, social complexity Reich, R. (Eds.), The Architecture of Ancient Israel. Israel Exploration Society, and agricultural strategies during the Chalcolithic period in the Northern Jerusalem, pp. 40e48. Negev, Israel. Journal of Archaeological Science 37, 284e294. Rapoport, A., 1969. House Form and Culture. In: Foundations of Cultural Geography Winter-Livneh, R., Svoray, T., Gilead, I. , 2012. Secondary burial cemeteries, visibility Series. Prentice-Hall, Englewood Cliffs, N.J. and land tenure: a view from the southern Levant Chalcolithic period. Journal of Rapoport, A., 1982. The Meaning of the Built Environment: a Non-verbal Commu- Anthropological Archaeology 31, 423e438. nication Approach. SAGE, Beverly Hills. Wright, R., 1993. Technological styles: transforming a natural material into Rollefson, G.O., 1997. Changes in architecture and social organization at Neolithic a cultural object. In: Kingery, W.D., Luber, S. (Eds.), History from Things: Essays ’Ain ghazal. In: Gebel, H.G.K., Kafafi, Z., Rollefson, G.O. (Eds.), The Prehistory of on Material Culture. Smithsonian Institution Press, Washington, DC. Jordan II. Perspective from 1997. Ex Oriente, Berlin, pp. 287e308. Wynn, T., 1985. Piaget, stone tools and the evolution of human intelligence. World Ron, F., 1998. The Palestinian Arab House and the Islamic “Primitive Hut”. Muqarnas Archaeology 17, 32e43. 15, 157e177. Wynn, T., 2002. Archaeology and cognitive evolution. Behavioral and Brain Sciences Rosen, S.A., 2011. Desert chronologies and periodization systems. In: Lovell, J.L., 25, 389e438. Rowan, M.Y. (Eds.), Culture, Chronology and the Chalcolithic: Theory and Zabrodsky, H., Avnir, D., 1995. Continuous symmetry measures. Journal of American Transition. Oxbow Books, Oxford, pp. 71e83. Chemical Society 117, 462e473. 3. Discussion and Conclusions

Taken together the results of the papers published suggest that intensive agricultural production dictated spatial patterns of the Ghassulian society within the Chalcolithic period. This is evident from three different scales of analyses, and from different environmental settings and different data types. The largest regional scale is of 2484 km2 which includes 29 burial sites and 117 non- burial sites located within the coastal plain and the Shephella region. A smaller regional scale of inter-site analysis covered some 1252 km2 and included 353 sites (non-burial) located in the semi- arid zone of the northern Negev. The smallest scale is of intra-site analysis based on structure spatial patterns within four Chalcolithic settlements, each of them covering an area between 1350 m2 and 4500 m2. There is plenty of archaeological evidence indicating the importance of agricultural activity during the Chalcolithic period. These include sickle blades (Gilead et al. 1995; Gilead et al. 2004; Rosen 1987b, 1997; Rosen and Levy 1987; Vardi et al. 2010), grinding stones or facilities (Rowan et al. 2006; Van den Brink 2008), botanical remains (Katz et al. 2007; Kislev 1987; Rosen 1987a; Rosen and Weiner 1994) and dental pathologies (Lev-Tov Chattah and Smith 2006; Lev-Tov et al. 2003; Smith 1995; Smith and Horowitz 1998). Although few explored the relation between agriculture and the wadis (Katz et al. 2007; Rosen 1987a; Rosen and Weiner 1994), no study has explored which of the wadis attributes determined actual location of a site. Previous studies who did explore settlement pattern within the northern Negev setting presented a contradiction between clustered pattern (Levy 1981) and random pattern (Fletcher 2008). Moreover, according to Levy (1981; 1995) it was suggested that clustered patterns reflect the location of several central sites which coordinate social and economical activities. The results of this study indicate that the patterns were indeed clustered, however, GLM statistical analysis produced significant results (R=0.48, p<0.05), indicating that two out of five wadi environmental features which characterize both Nahal Besor and Nahal Beer-Sheva drainage systems relate to the site clusters. One is the interaction between flow accumulation and depth index (reflecting whether the riverbanks are shallow and leveled or narrow and steep), and the other is the depth index alone. The results support both high water availability conditions (Rosen and Weiner, 1994), which was accomplished not by building diversion walls and small dams (Levy 1986), but by placing the fields in floodplain areas (Hillel, 1994), close to riverbank topographical features other than

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 63 floodplains (Katz et al., 2007) that are less flat or leveled, but offer a greater amount of flow accumulation which probably were needed to achieve agricultural prosperity. The uses of wadi resources were much simpler, adjusted to the most basic agricultural strategy. The present research’s strategy consists of locating sites in areas in Nahal Beer-Sheva, near riverbanks that are shallow and leveled, with low accumulation values. In Nahal Besor the sites are located in areas where the riverbanks are relatively deep and steep but with high accumulation values. These areas were probably found by the local community as most suitable for growing crops. The difference between Nahal Beer-Sheva and Nahal Besor supports the hypothesis that during the Chalcolithic period there were two cultural entities, which are referred to as the ‘Beer- Sheva cluster’ and the ‘Besor cluster’ (Gilead, 1995, 2007). This dichotomy was based until now solely on typology of artifacts and the variability between the Besor sites and Beer-Sheva sites, and has been acknowledged since the 1950s (e.g., Gophna, 1979; Perrot, 1955; Roshwalb, 1981). This study suggests that differences in the microenvironments of wadis are behind the economic and cultural differences between the Beer-Sheva and the Besor clusters of sites. In addition the results suggest that the settlement pattern of the Chalcolithic societies in the northern Negev is associated with small-scale societies with a relatively low level of social and economic complexity which practice dry farming. Within the coastal plain and the Shephella settlement patterns are located within a region of many burial grounds. The challenge is, therefore, to explore the spatial relation between the settlement patterns and their neighboring burial grounds. Archaeological research concerning mortuary practices during the Chalcolithic period is mostly focused on social, cultural and symbolic aspects. Most of the research has been devoted to the study of ossuaries and other funerary objects (e.g. Gal et al. 2011; Gilead 1988; Golden 2009; Gopher and Tsuk 1991; Gorzalczany 2006; Haas and Nathan 1973; Levy 1995; Levy and Alon 1987; Milevski 2002; Nativ and Gopher 2011; Perrot and Ladiray 1980), paying little attention to patterns of burial cave locations and their relationship to the habitation sites and the landscape. Though spatial aspects of burial sites are not more important than other aspects of mortuary behavior, they are definitely important for better understanding of Ghassulian secondary burial. The comparison between viewshed areas of burial sites and non-burial sites indicates that the average size of visible area observing burial sites is significantly larger that the area observing the non-burial sites. The contributing visible area within the study group is significantly larger (92%)

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 64 than the contributing visible area found within the matched control group (46%). The comparison between the overlapping viewshed areas found within the study group and within the matched control group show no significant difference between the two, indicating that the large visibility areas seen from the burial sites primarily affected the contributing visible area and not the overlapping visible areas. This pattern indicates that Chalcolithic cemeteries played a role in claiming land tenure, similar to what has been described in the ethnographical record. Increased demand for arable land due to population growth and reliance on agriculture prosperity may have stimulated a need to signal claims over arable lands. Mortuary practices relate closely to the organization of society (Binford 1971; Goldstein 1976; O'Shea 1984; Saxe 1970; Ucko 1969). Accordingly, Levy (1986, 1995), citing Saxe/Goldstein hypothesis, argues that the formal burial ground signifies a ‘‘hereditary chiefdom society’’ (Levy, 1995, p. 235). But, corporate lineal decent groups are not necessarily equivalents to the chiefdom stage of Service (Knight, 1990, p. 4). Based on ethnographic data, it is more likely that corporate groups (clans or lineages), who own or control lands, which are often marked by special boundary ceremonies and/or cemeteries, represent small local groups (tribes), rather than chiefdoms, that depend on agriculture, and settled ‘‘permanently or seasonally close to fields or other prime subsistence resources’’ (Earle, 2000, p. 46; Johnson and Earle, 1987). The Ghassulian way of claiming rights over the land was probably based also on gatherings of the community during secondary burial ceremonies, when land rights may have been negotiated, defined and asserted. The specific part of the terrain that was chosen for the burial cave, be it a particular kurkar ridge or limestone hill, turned into a landmark which was understood by the local communities. The burial cave locations created ‘‘spiritscapes’’ (David et al. 2008; McNiven 2004), landmarks which are identified with spiritual and political powers. Recurrently changing the landscape with burial caves defined patterns of group relationship and rules of inheritance that related to land use and/or access rights. Agriculture activity during this period may also relate to patterns of a much smaller scale of analysis. Architectural variability in Ghassulian settlements reflects different levels of intra-site heterogeneity which was probably due to intensification of agricultural production. Architectural remains analyzed using SR intra-site analysis generated significant results of lower level of variability within the Golanian site Rasam Harbush, and a significantly higher level

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 65 of variability within the Early Bronze Age site Arad in comparison to Ghassulian sites. This result indicates a higher level of heterogeneity of the Ghassulian sites in comparison to Rasam Harbush, and lower level of Ghassulian heterogeneity in comparison to Arad. In addition, SR inter-site results produced both lower inter-site normalized mean values for the Ghassulian site when comparing them with each other, and higher inter-site normalized mean values when comparing them to Rasam Harbush or Arad. This result indicates that Ghassulian architecture and intra-site variability are similar in the three sites we studied. These results challenge previous assumptions that Ghassulian architecture consists of considerable variation in shape (Bourke 2001, 2002, Levy 1986a, 1995, Levy and Alon 1985). Nonetheless, the relatively low degree of variability in shapes can be explained based on the understanding that house form is a “…direct expression of changing values, images, perception and the ways of life…” (Rapoport 1969: 12), and is “governed by a society’s ideas, its forms of economic and social organization, its distribution of resources and authority, its activities and the beliefs and values which prevail at any one period of time.” (King 1980: 1). The explanation of the similarity level should be therefore sought within the social and economic aspects of Chalcolithic communities. I mentioned earlier the abundance of evidence indicating that the Ghassulian subsistence economy witnessed an intensification of agricultural production, alongside greater investment in craft production (in ceramics, flint artifacts, basalt working, and ivory carving) and technological innovations, notably metallurgy (e.g. Gilead 1988, Golden 2009, Levy 1986a, Rowan and Golden 2009). Increased reliance on agriculture is also evident from extraordinary amount of sites in the northern Negev which are located based on their proximity to suitable fields. In the coastal plain and the Shephella region burial site locations are determined by the need to assert and define rights over agricultural land. Ghassulian economy is also linked to an increase in storage (Bourke 2002, Golden 2009, Levy 2003), that may have been combined with the appearance of semi-subterranean features (Gilead 1988). Accordingly, it is likely that the economic organization of Ghassulian communities reflects the levels of intra-site variability. Quantifying Ghassulian settlement patterns is of great importance in understanding Ghassulian societies and their environmental and socio- economic contexts. GIS based research enables to determine patterns of site distributions and their correlation with environmental features.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 66 The distribution of sites within two distinct regions and the patterns of intra-site architecture seemingly have little in common. Yet, I suggest that these three aspects are fundamentally intertwined and posit the Ghassulian obligation to agricultural prosperity. The settlement patterns most probably represent small local groups (tribes), rather than chiefdoms that depend on agriculture. Locating their sites near suitable areas for growing crops, and their use in land tenure system was simple, adjusted to the most basic agricultural strategy. Additional support for the level of agricultural production can be found with the relative heterogeneity within intra-site architectural patterns. To conclude, each of the research papers presented here explores a different Ghassulian spatial pattern, and thus is able to stand on its own. Nonetheless the published papers complete one another by retrieving information on Ghassulian communities’ socio-economic behavior based on settlement spatial relation patterns. The papers complete one another by using multiscalar analyses, and complementary geographical regions. The study is GIS based and accordingly the published papers include advance use of variety of methodological tools and approaches. This study is not, of course, a definite conclusion of Chalcolithic period research. Discoveries of new sites, cemeteries, structures as well as past landscape forms will no doubt contribute towards a better understanding of the Ghassulian society and the mechanisms that determined its spatial distribution. The archaeological database used in this research is only an available sample; when and if enlarged, it will add to and modify the way we portray this past society. Studies of sites and cemeteries locations and their inter-relationships, along with analyses of structures variability, are indispensable elements of future research.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 67 4. References

Ackermann, O., T. Svoray, and M. Haiman. 2008. Nari (calcrete) outcrop contribution to ancient agricultural terraces in the Southern Shephelah, Israel: insights from digital terrain analysis and a geoarchaeological field survey. Journal of Archaeological Science 35:930. Albright, W. F. 1931. Recent progress in the Late Prehistory of Palestine. Bulletin of the American Schools of Oriental Research 42:13-15. —. 1932. The Chalcolithic Age in Palestine. Bulletin of the American Schools of Oriental Research 48:10-13. Allison, P. Editor. 2002. The Archaeology of Household Activities. London and New York: Routledge. Alon, D., and T. E. Levy. 1980. Preliminary note on the distribution of Chalcolithic sites on the Wadi Beer-Sheva and lower Wadi Besor drainage system. Israel Exploration Journal 30:140-147. Amin, N. 2002. "The development of a GIS center of excellence within the SCA," in UNESCO Virtual Congress: http://www.cultnat.org/download/PdfConf/The_developmentof_aGiscenter.pdf. Ben-Tor, A. Editor. 1992. The Archaeology of Ancient Israel. New Haven: Yale University Press. Bevan, A., and J. Conolly. 2006. "Multiscalar approaches to settlemnt pattern analysis," in Confronting Scale in Archaeology: Issues of Theory and Practice. Edited by G. Lock and B. L. Molyneaux, pp. 217-234. New-York: Springer. Beyer, H. L. 2004. "Hawth's Analysis Tools for ArcGIS." Available at http://www.spatialecology.com/htools/. Bienkowski, P., and E. Van der Steen. 2006. "Wadi Arabah Project," http://www.wadiarabahproject.man.ac.uk/opening/index.htm. Binford, R. 1971. Mortuary practices: Their study and potential. American Antiquity 36:6-29. Bloch, M. 1989. Ritual, History and Power. London: Althone. Bourke, J. S. 1997. "The "pre-Ghassulian" Sequence at Teleilat Ghassul," in The Prehistory of Jordan, II. Perspectives from 1997. Edited by Z. K. H.G.K Gebel, and G.O. Rollefson, pp. 395-417. Berlin: Ex Oriente. —. 2001. "The Chalcolithic period," in The Archaeology of Jordan. Edited by B. MacDonald, R. Adams, and P. Bienkowski, pp. 107-162. Sheffield: Sheffield Academic Press. —. 2002. The origins of social complexity in the southern Levant: new evidence from Teleilat Ghassul, Jordan. Palestine Exploration Quarterly 134:2-27. Bourke, J. S., J. Lovell, R. Sparks, P. Seaton, L. Mairs, and J. Meadows. 2000. Preliminary report on a second and third season of renewed excavations at Teleilat Ghassul by the University of Sydney, 1995/1997. Annual of the Department of Antiquities Jordan 44:37- 89. Bradley, R. 1991. Ritual, time and history. World Archaeology 23:209-219. —. 1998. The Significant of Monuments: On the Shaping of Human Experience in Neolithic and Bronze Age Europe. London: Routledge. Bridgeman, B. 2002. Artifacts and cognition: Evolution or cultural progress? Behavioral and Brain Sciences 25:403. Brighenti, A. 2007. Visibility. A catagory for the social sciences. Current Sociology 55:323-342.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 68 Burrough, P. A. 1986. Principles of GIS for Land Resources Assessment. Oxford: Clarendon Press. Byrd, F. B. 1994. Public and private, domestic and corporate: The emergence of the southwest Asian village. American Antiquity 59:639-666. Cannon, A. 2002. Spatial narratives of death, memory and transcendence. Archeological Papers of the American Anthropological Association 11:191-199. Carsten, J., and S. Hugh-Jones. 1995. About the House: Levy-Strauss and Beyond. Cambridge: Cambridge University Press. Clarke, D. L. Editor. 1977. Spatial Archaeology. London: Academic Press INC. Comer, D. 2000. "Three-dimensional mapping of environmental and prehistoric cultural features in the Beidha region with spaceborne and airborne imagery," in http://gis.esri.com/library/userconf/proc00/professional/papers/PAP124/p124.htm. Coppock, J. T., and D. W. Rhind. 1991. "The history of GIS," in Geographical Information Systems: Principles and Applications. Edited by D. J. Maguire, M. F. Goodchild, and D. W. Rhind, pp. 21-43. London: Longman. David, B., M. Pivoru, W. Pivoru, M. Green, B. Barker, J. Weiner, D. Simala, T. Kokents, L. Arabo, and J. Dop. 2008. "Living landscapes of the dead: archaeology of the afterworld among the Rumu of Papua New Guinea," in Handbook of landscape archaeology. Edited by B. David and J. Thomas, pp. 158-166. San Francisco: Left Coast Press. Dollfus, G., and Z. Kafafi. 1986. Abu Hamid, Jourdain. premiers resultats. Paleoriént 12:91-100. —. 1988. Abu Hamid, Village du 4e Millenaire de la Vallee du Jourdain. Amman: Economic Press. Donley, L. W. 1982. "House power: Swahili space and symbolic markers," in Symbolic and Structural Archaeology. Edited by I. Hodder, pp. 63-73. Cambridge: Cambridge University Press. Dothan, M. 1959. Excavations at Horvat Beter (Beersheba). 'Atiqot 2:1-71. Earle, T. K. 1991. "Property rights and the evolution of chiefdoms," in Chiefdoms: Power, Economy and Ideology. Edited by T. K. Earle, pp. 71-99. Cambridge: Cambridge University Press. Eason, P. K., and J. A. Stamps. 1992. The effect of visibility on territory size and shape. Behavioral Ecology 3:166-172. Epstein, C. 1998. The Chalcolithic Culture of the Golan. Jerusalem: IAA Reports No. 4. Finkelstein, I., and R. Gophna. 1993. Settlement, demographic, and economic patterns in the Highlands of Palestine in the Chalcolithic and Early Bronze periods and the beginning of urbanism. Bulletin of the American Schools of Oriental Research 289:1-22. Flannery, K. V. 1972. "The origins of the village as a settlement type in Mesoamericaa and the Near East:A comparative study," in Man, Settlement and Urbanism. Edited by P. J. Ucko., R. Tringham., and G. W. Dimbleby., pp. 23-53. London: Duckwort. Fleming, A. 1973. Tombs for the living. Man 8:177-193. Fletcher, R. 2008. Some spatial analyses of Chalcolithic settlement in Southern Israel. Journal of Archaeological Science 35:2048-2058. Foucault, M. 1995 (1975). Discipline and Punish: The Birth of the prison, 2 edition. New York: Vintage Books. Frankfort, H. 1924. Studies in Early Pottery of the Near East I. London: Royal Anthropological Institute of Great Britain and Ireland. Fried, M. H. 1967. The Evolution of Political Society. New York: Random House.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 69 Gal, Z., D. Shalem, and H. Smithline. 2011. The Peqi'in cave: a Chalcolithic cemetery in Upper Galilee, Israel. Near Eastern Archaeology 74:196-206. Gibson, S., and Y. M. Rowan. 2006. The Chalcolithic in the central highlands of Palestine: a reassessment based on new examination of Khibet es-Sauma'a. Levant 38:85-108. Gilead, I. 1988. The Chalcolithic period in the Levant. Journal of World Prehistory 2:397-443. —. 1993. "Sociopolitical organization in the northern Negev at the end of the Chalcolithic period," in Biblical Archaeology Today 1990, Supplement. Edited by A. Biran and J. Aviram. Jerusalem: Israel Exploration Society. —. 1995. Grar: A Chalcolithic Site in the Northern Negev. Beer-Sheva. Beer Sheva: Ben-Gurion University Press. —. 2007. The Besorian: a Pre-Ghassulian cultural entity. Paleoriént 33:33-49. —. 2011. "Chalcolithic culture history: Ghassulian and other entities in the Southern Levant," in Culture, Chronology and the Chalcolithic. Theory and Transition. Edited by L. J. Lovell and M. Y. Rowan, pp. 12-24. Oxford: Oxbow Books. Gilead, I., and Y. Goren. 1989. Petrographic analyses of fourth millennium B.C. pottery and stone vessels from the Northern Negev, Israel. Bulletin of the American Schools of Oriental Research 275:5-14. Gilead, I., D. Hershman, and O. Marder. 1995. "The flint assemblages from Grar," in Grar: A Chalcolithic Site in the Northern Negev. Edited by I. Gilead, pp. 223-280. Beer Sheva: Ben-Gurion University of the Negev Press. Gilead, I., O. Marder, H. Khalaily, P. Fabian, Y. Abadi, and Y. Yisrael. 2004. The Beit Eshel Chalcolothic flint workshop in Beer Sheva: a preliminary report. Journal of The Israel Prehistoric Society 34:245-263. Goldberg, P., and O. Bar-Yosef. 1982. "Environmental and archaeological evidence for climatic change in the Southern Levant," in Palaeoenvironments and Human Communities in the Eastern Mediterranean Region in Later Prehistory. Edited by J. L. Bintliff and W. Van Zeist, pp. 399-414. Oxford: BAR International Series 133. Goldberg, P., and A. M. Rosen. 1987. "Early Holocene palaeoenvironments of Israel," in Shiqmim I, Studies Concerning Chalcolithic Societies in the Northern Negev Desert, Israel. Edited by T. E. Levy, pp. 23-33. Oxford: BAR International Series 356. Golden, J. 2009. Dawn of the Metal Age. London: Equinox. Goldstein, L. 1976. Spatial Structure and Social Organization: Regional Manifestations of Mississippian Society. Unpublished Ph.D dissertation, Northwestern University. —. 1981. "One-dimensional archaeology and multi-dimensional people: Spatial organization and mortuary analysis," in The Archaeology of Death. Edited by R. Chapman, I. Kinnes, and K. Randsborg, pp. 53-69. New York: Cambridge University Press. —. 2002. Visible death: mortuary site and mortuary landscape in diachronic perspective. Archeological Papers of the American Anthropological Association 11:201-205. Gopher, A., and T. Tsuk. 1991. Ancient Gold-Rare Finds from the Nahal Qanah Cave. Jerusalem: Israel Museum. Gophna, R., and J. Portugali. 1988. Settlement and demographic processes in Israel's costal plain from the Chalcolithic to the Middle Bronze Age. Bulletin of the American Schools of Oriental Research 269:11-28. Gophna, R., and T. Tsuk. 2005. Chalcolithic settlements in the western Samaria foothills. Tel Aviv 32:3-19.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 70 Goren, Y. 1987. Petrography of Chalcolithic Period Pottery Assemblages from Southern Israel, Unpublished M. A. thesis. Hebrew University, Jerusalem. Goring-Morris, A. N., and A. Belfer-Cohen. 2008. "A roof over one's head: developments in Near Eastern architecture across the Epipalaeolithic-Neolithic transition," in The Neolithic Demographic Transition and its Consequences. Edited by J. Bocquet-Appel and O. Bar-Yosef, pp. 239-286. Berlin: Springer. Gorzalczany, A. 2006. Chalcolithic burial site at Palmahim North. Qadmoniot 132:87-94. Haas, N., and H. Nathan. 1973. "An attempt at a social interpretation of the Chalcolithic burials in the Nahal Mishmar caves," in Excavations and Studies: Essays in Honour of Professor Shemuel Yeivin. Edited by Y. Aharoni, pp. 143-153. Tel Aviv. Haiman, M. 1995. Agriculture and nomad-state relations in the Negev desert in the Byzantine and Early Islamic periods. BASOR 297:29-53. —. 2006. "Ancient Desert Agriculture Systems Revived," http://www.mnemotrix.com/adasr/about.html. Hennessy, J. B. 1969. Preliminary report on a first season of excavation at Teleilat Ghassul. Levant 1:1-24. Hermon, S. 2008. Socio-economic Aspects of Chalcolithic (4500-3500 BC) Societies in the Southern Levant. BAR International Series 1744. Oxford: Archaeopress. Hillier, B., and J. Hanson. 1984. The Social Logic of Space. Cambridge: Cambridge University Press. Hutchinson, D., and L. Aragon. 2002. Collective Burials and Community Memories: Interpreting the Placement of the Dead in the Southeastern and Mid-Atlantic United States with Reference to Ethnographic Cases from Indonesia. Archeological Papers of the American Anthropological Association 11:27-54. Ingold, T. 1990. Society, nature and the concept of technology. Archaeological Review from Cambridge 9:5-17. —. 1995. "Building, dwelling, living: how animals and people make themselves at home in the world," in shifting Contexts: Transformations in Anthropological Knowledge. Edited by M. Strathern, pp. 57-80. London: Routledge. —. 2000. The Perception of the Environment: Essays on Livelihood, Dwelling and skill. London: Routledge. Joffe, H. A. 2003. "Slouching toward Beersheva: Chalcolithic mortuary practices in local and regional context," in The Near East in the Southwest: Essays in Honor of William G. Dever. Edited by B. A. Nakhai, pp. 45-67. Boston, MA: American Schools of Oriental Research. Katz, O., I. Gilead, P. Bar (Kutiel), and R. Shahack-Gross. 2007. Chalcolithic agricultural life at Grar, northern Negev, Israel: dry farmed cereals and dung-fueled hearths. Paleoriént 31:101-116. Kempinski, A., and R. Reich. Editors. 1992. The Architecture of Ancient Israel. Jerusalem: Israel Exploration Society. Kent, S. Editor. 1990. Domestic Architecture and the Use of Space: An Iterdisciplinary Cross- Cultural Study. Cambridge: Cambridge University Press. Khalaily, H., and O. Marder. 2010. Prehistoric investigations along the cross-Israel highway: State of research. 'Atiqot 64:5-23. King, A. D. Editor. 1980. Building and Society: Essays on the Social Development of the Built Environment. London: Routledge and Kegan Paul Ltd.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 71 Kislev, M. 1987. "Chalcolithic plant husbandry and ancient vegetation at Shiqmim," in Shiqmim I, Studies Concerning Chalcolithic Communities in the Northern Negev, Israel. Edited by T. Levy, pp. 251-279. Oxford: B.A.R International Series 356. Koeppel, R. 1932. Uferstudien am toten meer naturwissenschaftliches zur pentapolis und zur dentung von Tell Ghassul. Biblica 13:6-27. Kuijt, I. 1996. Negotiating equality through ritual: A consideration of Late Natufian and Prepottery Neolithic A period mortuary practices. Journal of Anthropological Archaeology 15:313-336. Kvamme, L. K. 1995. "A view from across the water: the north American experience in archaeological GIS," in Archaeology and Geographical Information Systems: A European Perspective. Edited by G. Lock and Z. Stancic, pp. 1-14. London: Taylor and Francis. —. 2006. "There and back again: revisiting archaeological location modeling," in GIS and Archaeological Site Location Modeling. Edited by M. W. Mehrer and K. L. Wescott, pp. 3-38. Boca Raton: Taylor & Francis. Lau, F. G. 2010. House form and Recuay culture: Residential compounds at Yayno (Ancash, Peru), a fortified hilltop town, AD 400-800. Journal of Anthropological Archaeology 29:327-351. Lee, J. R. 1973. Chalcolithic Ghassul: New Aspects and Master Typology. Jerusalem: Hebrew University. Lev-Tov Chattah, N., and P. Smith. 2006. Variation in occlusal dental wear of two Chalcolithic populations in the Southern Levant. American Journal of Physical Anthropology 130:471-479. Lev-Tov, N., A. Gopher, and P. Smith. 2003. Dental evidence for dietary practices in the Chalcolithic period: The findings from a burial cave in Peqi'in (Northern Israel). Paleorient 29:121-134. Levy, T. E. 1981. Chalcolithic Settlement and Subsistence in the Northern Negev Desert, Israel, University of Sheffield: Ann Arbor. —. 1986. The Chalcolithic period. Biblical Archaeologist 49:82-108. —. Editor. 1987. Shiqmim I, Studies Concerning Chalcolithic Societies in the Northern Negev Desert, Israel. Oxford: BAR International Series 356. —. 1995. "Cult, metallurgy and rank societies - Chalcolithic period (CA. 4500-3500 BCE)," in Archaeology of Society in the Holy Land. Edited by T. E. Levy, pp. 226-243. New York: Facts on File. —. Editor. 2006. Archaeology, Anthropology and Cult: The Sanctuary at Gilat, Israel. London: Equinox Publishing. Levy, T. E., and D. Alon. 1982. The Chalcolithic mortuary site near Metzad Aluf, northern Negev desert: A preliminary study. Bulletin of the American Schools of Oriental Research 248:37-59. —. 1987. "Excavations in Shiqmim cemetery 3: final report on the 1982 excavations," in Shiqmim I, Studies Concerning Chalcolithic Societies in the Northern Negev Desert, Israel. Edited by T. E. Levy, pp. 333-355. Oxford: BAR International Series 356. Llobera, M. 2003. Extending GIS-based visual analysis: The concept of visualscapes. International Journal of Geographical Information Science 17:25-48. —. 2007. Reconstructing visual landscapes. World Archaeology 39:51-69.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 72 Lovell, J. 2002. Shifting subsistence patterns: some ideas about the end of the Chalcolithic in the southern Levant. Paleorient 28:89-102. MacDonald, B., R. Adams, and P. Bienkowski. Editors. 2001. The Archaeology of Jordan. Sheffield: Sheffield Academic Press. MacDonald, E. 1932. Prehistoric Fara, Beth Pelet II. London: The British School of Archaeology in Egypt. Maguire, D. J. 1991. "An overview and definition of GIS," in Geographical Information Systems: Principles and Applications. Edited by D. J. Maguire, M. F. Goodchild, and D. W. Rhind, pp. 9-20. London: Longman. Mallon, A. 1929. Notes sur quelques sites du Ghor oriental. Biblica 10:94-99, 214-232. Mallon, A., R. Koeppel, and R. Neuville. 1934. Teleilat Ghassul I. Roma: Institute Biblique Pontifical. McNiven, I. 2004. Saltwater People: spiritscapes, maritime rituals and the archaeology of Australian indigenous seascapes. World Archaeology 35:329-349. Messika, N. 2006. Archaeological applications, Tel Aviv University. http://www.nataliemessika.co.il/. Milevski, I. 2002. A new fertility figurine and new animal motifs from the Chalcolithic in the Southern Levant: finds from cave K-1 at Quleh, Israel. Paleorient 28:133-141. Moran, P. 1948. The interpretation of statistical maps. Journal of the Royal Statistical Society 10B:243-251. Nativ, A., and A. Gopher. 2011. The cemetry as a symbol: a reconsideration of Chalcolithic burial caves in the southern Levant. Cambridge Archaeological Journal 21:229-245. Neuville, R., and A. Mallon. 1931. Les débuts de l'âge des metaux dans les grottes du désert de Judée. Syria XII:24-47. North, R. 1959. "Ghassulian" in Palestine chronological nomencalture. Biblica 40:541-555. —. 1961. Ghassul 1960: Excavation Report. Roma: Pontificium Institutum Biblicum. O'Shea, J. 1984. Mortuary Variability: An Archaeological Investigation. New York: Academic Press. Oren, D. E., and I. Gilead. 1981. chalcolithic sites in northeastern sinai. Tel Aviv 8:25-44. Parker, H. D. 1988. The unique qualities of a geographic information system: a commentary. Phtogrammetric Engineering and Remote Sensing 54:1547-1549. Parslow, A. C. 2006. Social Interaction in the Prehistoric Natufian:Generating an Interactive Agency Model Using GIS, PhD. Dissertation. University of Toronto. http://carlaparslow.googlepages.com/home. Perrot, J. 1955. Excavations at Tell Abu Matar, near Beersheba. Israel Exploration Journal 5:17- 40, 73-84, 167-189. —. 1984. Structures d'habitat, mode de la vie et envirnment: Les villages souterrains des pasteurs de Be'ersheva dans le Sud d'Israel, au IV millénaires avant l’ère chrétienne. Paleoriént 10:75-92. Perrot, J., and D. Ladiray. 1980. Tombes à Ossuaires de la Région Côtière Palestinienne au IVe Millénaire avant L'ère Chrétienne. (Mémoires et Travaux de Center de Recherches Préhistorques Français de Jérusalem 1). Paris. Pierce, G. 2006. "GIS and the Chalcolithic of the Northern Negev." paper delivered at the Innovations in the Research of the Chalcolithic Beer Sheva Culture conference, 2006. Porat, N., A. S. Rosen, E. Boaretto, and Y. Avni. 2006. Dating the Ramat Saharonim Late Neolithic desert cult site. Journal of Archaeological Science 33:1341-1355.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 73 Porath, Y. 1992. "Domestic architecture of the Chalcolithic period," in The Architecture of Ancient Israel. Edited by A. Kempinski and R. Reich, pp. 40-48. Jerusalem: Israel Exploration Society. Rapoport, A. 1969. House Form and Culture. Foundations of Cultural Geography Series. Englewood Cliffs, N.J.: Prentice-Hall. —. 1982. The Meaning of the Built Environment: A Non-verbal Communication Approach. Beverly Hills: SAGE. Renfrew, C. 1973a. Before Civilization. New York: Alfred Knopf. —. 1973b. "Monuments, mobilisation, and social organization in Neolithic Wessex," in The Explanation of Culture Change: Models in Prehistory. Edited by C. Renfrew, pp. 539- 558. London: Duckworth. —. 1976. "Megaliths, territories and population," in Acculturation and Continuity in Atlantic Europe. Edited by S. J. De Laet, pp. 198-220. Brugge: De Tempel. Richards, J. 1998. Recent trends in computer applications in archaeology. Journal of Archaeological Research 6:331-382. Rosen, A. M. 1987a. "Phytolith studies at Shiqmim," in Shiqmim I, Studies Concerning Chalcolithic Societies in the Northern Negev Desert, Israel. Edited by T. E. Levy, pp. 243-249. Oxford: BAR International Series 356. Rosen, A. M., and S. Weiner. 1994. Identifying ancient irrigation: a new method using opaline phytoliths from Emmer wheat. Journal of archaeological Science 21:125-132. Rosen, S. A. 1987b. "The potentials of lithic analysis in the Chalcolithic of the northern Negev," in Shiqmim I, vol. I. Edited by T. E. Levy, pp. 295-312. Oxford: BAR International Series 356 (i). —. 1993. "Metals, rocks, specialization, and the beginning of urbanism in the northern Negev," in Biblical Archaeology Today, 1990. Edited by A. Biran and J. Aviram, pp. 41-56. Jerusalem: Israel Exploration Society. —. 1997. Lithic After the Stone Age. Walnut Creek: AltaMira Press. Rosen, S. A., and T. E. Levy. 1987. "The chipped stone industry at Shiqmim: typological considerations," in Shiqmim, vol. I. Edited by T. E. Levy, pp. 281-294. Oxford: BAR International Series 356 (i). Rowan, M. Y., and J. Golden. 2009. The Chalcolithic period of the Southern Levant: A synthetic review. Journal of World Prehistory 22:1-92. Rowan, Y. M., T. E. Levy, D. Alon, and Y. Goren. 2006. "The ground stone industry: Stone bowls, grinding slabs, palettes, spindle whorls, maceheads and related finds," in Archaeology, Anthropology and Cult: The Sanctuary at Gilat. Edited by T. Levy, pp. 575–684. London: Equinox Press. Rubin, R. 1990. The Negev as a Settled Land. Jerusalem: Yad Ben-Zvi and the Israel Exploration Society (Hebrew). Saragusti, I., A. Karasik, I. Sharon, and U. Smilansky. 2005. Quantitative analysis of shape attributes based on contours and section profiles in artifact analysis. Journal of Archaeological Science 32:841-853. Saragusti, I., I. Sharon, O. Katzenelson, and D. Avnir. 1998. Quantitative analysis of the symmetry of artefacts: Lower Paleolithic handaxes. Journal of Archaeological Science 25:817-825.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 74 Savage, H. S. 1990. "Modelling the Late Archaic social landscape," in Interpreting Space: GIS and Archaeology. Edited by M. S. K. Allen, W. S. Green, and B. W. E. Zubrow, pp. 330- 355. London: Taylor and Francis. Saxe, A. 1970. Social Dimensions of Mortuary Practices. Unpublished Ph.D. dissertation, University of Michigan. Schiffer, M. B., and M. Skibo. 1987. Theory and experiment in the study of technological change. Current Anthropology 28:595-622. Schroeder, S. 2001. Secondary disposal of the dead: cross-cultural codes. World Cultures 12:77- 93. Service, E. R. 1962. Primitive Social Organization: An Evolutionary Perspective. New York: Random House. Shershevski, J. 1991. Byzantine Urban Settlement in the Negev Desert. Beer-Sheva 5. Beer- Sheva: Ben-Gurion University of the Negev. Shipton, G. M. 1939. Notes on the Meggido Pottery of Strata VI-XX. Chicago: University of Chicago Press. Silverman, B. W. 1986. Density Estimation for Statistics and Data Analysis. London: Chapman and Hall. Smith, P. 1995. "People of the Holy Land from prehistory to the recent past," in The Archaeology of Society in the Holy Land. Edited by T. E. Levy, pp. 58-74. London: Facts on File. Smith, P., and L. K. Horowitz. 1998. "Culture, environment and Disease: Paleo-anthropological findings for the Southern Levant," in Digging for Pathogens. Edited by C. Greenblatt, pp. 201-239. Rehovot-Philadelphia. Ucko, P. 1969. Ethnography and archaeological interpretation of funerary remains. World Archaeology 1:262-280. Ussishkin, D. 1980. The Ghassulian shrine at En Gedi. Tel Aviv 7:1-44. Van den Brink, E. C. M. 1998. An index to Chalcolithic mortuary caves in Israel. Israel Exploration Journal 48:165-173. —. 2005. "Chalcolithic burial caves in coastal and inland Israel," in Shoham (North): Late Chalcolithic Burial Caves in the Lod Valley, Israel. Edited by E. C. M. van den Brink and R. Gophna, pp. 175-189. Jerusalem: Israel Antiquities Authority. —. 2008. A new fossile directeur of the Chalcolithic landscape in the Shephelah and the Samarian and Judaen hill countries: Stationary grinding facilities in bedrock. Israel Exploration Journal 58:1-23. —. 2011. A Chalcolithic and Early Bronze Age I burial ground near Sha'ar Efrayim in the Sharon plain. 'Atiqot 66:1-53. Van den Brink, E. C. M., and R. Gophna. 2005. Shoham (North): Late Chalcolithic Burial Caves in the Lod Valley, Israel. IAA Reports. Jerusalem: Israel Antiquities Authority. Vardi, J., A. Golan, D. Levy, and I. Gilead. 2010. Tracing sickle blade levels of wear and discard patterns: a new sickle gloss quantification method. Journal of Archaeological Science 37:1716-1724. Wheatley, D. 1995a. "Cumulative viewshed analysis: A GIS-based method for investigating intervisibility and its archaeological application," in Archaeology and Geographic Information Systems: A European Perspective. Edited by G. Lock and Z. Stancic, pp. 171-185. London: Taylor and Francis.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 75 —. 1995b. "Cumulative viewshed analysis: a GIS-based method for investigating intervisibility, and its archaeological application," in Archaeology and Geographical Information Systems: A European Perspective. Edited by G. Lock and Z. Stancic, pp. 171-185. London: Taylor & Francis. Wheatley, D., and M. Gillings. 2002. Spatial Technology and Archaeology. London: Taylor and Francis. Williams, H. M. R. 1999. "Placing the dead: investigating the location of wealthy barrow burial in seventh century England," in Grave Matters: Eight Studies of First Millennium AD Burials in Crimea, England and Southern Scandinavia. Edited by M. Rundkvist, pp. 57- 86. Oxford: Archaeopress. Wilson, P. J. 1988. The Domestication of the Human Species. New Haven and London: Yale University Press. Wright, G. E. 1937. The Pottery of Palestine from the Earliest Times to the End of the Early Bronze Age. Wynn, T. 1985. Piaget, stone tools and the evolution of human intelligenece. World Archaeology 17:32-43. Yannai, E., and Y. Porath. 2006. A Chalcolithic burial cave as Et-Taiyiba. 'Atiqot 53:1-44. Yekutieli, Y. 2002. "Settlement and subsistence patterns in North Sinai during the fifth to third millennia BCE," in Egypt and the Levant: Interrelations from the 4th Through the Early 3rd Millennium BCE. Edited by E. C. M. van den Brink, pp. 422-433. London: Leicester University Press. —. 2006. Is somebody watching you? Ancient surveillance systems in the southern Judean Desert. Journal of Mediterranean Archaeology 19:65-89. Zabrodsky, H., and D. Avnir. 1995. Continuous symmetry measures. Journal of American Chemical Society 117:462-473.

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 76 המרחביים המאפיינים את פריסת אתרי היישוב בנגב, את הקשר בין מיקומם של אתרי קבורה ואתרים שכנים במרכז הארץ, כמו גם את מידת ההטרוגניות היחסית של מבנים שמאפיינת את האתרים מרכזיים של התקופה הכלקוליתית.

מילות מפתח: דפוסי התיישבות; ארכיטקטורה פרהיסטורית; קבורה משנית; אסטרטגיות חקלאיות; התקופה הכלקוליתית; התרבות הע'סולית; דרום הלבנט; ניתוחים מרחביים; ממ"ג

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 78 5. תקציר

עבודה זו עוסקת בבחינת הקשר בין הדפוסים המרחביים של ההתיישבות הע'סולית ולסביבה הקרובה בתקופה הכלקוליתית )0044-0044 לפנה"ס( . מטרת המחקר היא לעמוד על מאפייני הסביבה אשר השפיעו על התנהגותם המרחבית של קהילות התקופה הכלקוליתית וקבעו את אופייה החברתי והכלכלי. המחקר מתבסס על ממ"ג )מערכות מידע גאוגרפיות(, ועושה שימוש במגוון כלים לניתוחים גיא-וסטטיסטיים: Moran’s I spatial autocorrelation analysis, Ripley’s K-function, Kernel density spatial analysis, כלי לניתוח שדה ראיה viewshed analysis, לצד שימוש ב Continuance Symmetry Measure ושיטה חדשה שפיתחתי במחקר זה, Shape Reproducibility. מסקנות המחקר מבוססות על השונות המרחבית תוך- ובין-אתרית המאירה תובנות חדשות בנוגע למספר אספקטים חברתים וכלכליים של ההתיישבות הע'סולית.

תוצאות המחקר פורסמו בשלושה מאמרים המציגים אספקטים שונים של תפוצת ההתיישבות בשלושה סדרי גודל ובאזורים גיאוגרפיים שונים. המאמר הראשון מתרכז בנגב הצפוני שם נמצא המספר הגדול ביותר של אתרים כלקוליתיים. במחקר זה בחנתי את תפוצתם של כ004- אתרים כלקוליתיים הנמצאים על גדות שני אגני הניקוז המרכזיים בנגב הצפוני, נחל באר-שבע ונחל הבשור, ואת הקשר בין מקומם לבין מאפיינים פיסיים של אפיקי הנחלים, הכוללים את הטופוגרפיה של הנחלים, נפח הנחלים, הצטברות זרימה, שיפוע אפיק הנחל, ושיפוע גדות הנחל. המאמר השני בוחן את דגמי היישוב הכלקוליתיים במישור החוף והשפלה, אזור בו נפוצים אתרי הקבורה המשנית, ובודק את הקשר בין 48 אתרי קבורה ואתרי מגורים לסביבתם הקרובה ואת המשמעות החברתית. במחקר זה מדדתי מאלו אזורים ניתן להבחין באתר הקבורה ומאלו אזורים ניתן לראות את אתר המגורים הקרוב ביותר ובחנתי עד כמה שטחי אלו חופפים או משלימים אחד את השני, ועד כמה הדבר ייחודי אך ורק לדפוס המרחבי בו ממוקמים בשכנות אתרי הקבורה לאתרי היישוב. המאמר השלישי עוסק במאפיינים החברתיים והאדרכליים של המבנים הע'סוליים ובשונותם התוך- והבין-אתרית. המאמר מתרכז בארבעה יישובים כלקוליתיים שחופריהם פרסמו תוכניות מפורטות יחסית של המבנים שנחשפו בהם. במאמר זה בחנתי את הדימיון הצורני והסימטריה של כל מבנה ביחס לכל המבנים האחרים באותו האתר ובשלב השני השוואתי את סך כל המבנים מכל אתר לסך כל המבנים באתרים האחרים שבדקתי.

ממחקר זה עולה כי מיקומם של מקבצי אתרים לאורך הוואדיות בנגב הצפוני אינו מקרי ומתבסס על התנאים הטופוגרפיים של גדות הנחל, דבר היכול ללמד כי הקהילה הע'סולית לא הזדקקה לאמצעים מלאכותיים כגון קירות הטייה או סכרים קטנים על מנת להגיע לתנובה חקלאית רבה. במישור החוף והשפלה מיקומם של בתי קברות והקשר המרחבי שלהם עם אתרי מגורים מלמד כי לבתי הקברות תפקיד בהצהרת בעלות על קרקעות ועל שליטה בדרכי הגישה אליהן.

מניתוחם הצורני של שרידים ארכיטקטוניים התברר שרמות ההטרוגניות התוך-אתרית בין אתרים הע'סוליים דומות, עובדה היכולה לרמז על התחזקות שונות כלכלית, אולי כתוצאה מהתגברות התלות בחקלאות. שלושת המאמרים מצביעים על חשיבותה הרבה של הפעילות החקלאית בחיי קהילות התקופה הכלקוליתית, על אינטנסיביות התפוקה החקלאית ועל התלות הרבה בתוצרתה. נתונים אלה יכולים להסביר את הדפוסים

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 77

הצהרת תלמיד המחקר עם הגשת עבודת הדוקטור לשיפוט אני החתום מטה מצהיר7ה בזאת: )אנא סמן(:

X___ חיברתי את חיבורי בעצמי, להוציא עזרת ההדרכה שקיבלתי מאת מנחה7ים.

__X_ החומר המדעי הנכלל בעבודה זו הינו פרי מחקרי מתקופת היותי תלמיד7ת מחקר.

_X__ בעבודה נכלל חומר מחקרי שהוא פרי שיתוף עם אחרים, למעט עזרה טכנית הנהוגה בעבודה ניסיונית. לפי כך מצורפת בזאת הצהרה על תרומתי ותרומת שותפי למחקר, שאושרה על ידם ומוגשת בהסכמתם.

תאריך __1400..10.4 שם התלמיד7ה ___רונה וינטר-לבנה______חתימה______

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 82

העבודה נעשתה בהדרכת פרופ' יצחק גלעד במחלקה למקרא, ארכיאולוגיה ומזרח קדום בפקולטה למדעי הרוח והחברה

פרופ' טל סבוראי במחלקה לגיאוגרפיה ופיתוח סביבתי בפקולטה למדעי הרוח והחברה

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 81 השפעת התנאים הסביבתיים על התיישבות הע'סולית )0,044-0,044 לפנה"ס(: היבטים מרחביים תוך שימוש במערכות מידע גאוגרפיות

מחקר לשם מילוי חלקי של הדרישות לקבלת תואר "דוקטור לפילוסופיה"

מאת

רונה וינטר-לבנה

הוגש לסינאט אוניברסיטת בן גוריון בנגב

אישור המנחה: פרופ' יצחק גלעד תאריך: 4/74/71400

אישור המנחה: פרופ' טל סבוראי תאריך: 4/74/71400

תמוז תשע"ג יוני 1400

באר שבע

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 80

השפעת התנאים הסביבתיים על התיישבות הע'סולית )0,044-0,044 לפנה"ס(: היבטים מרחביים תוך שימוש במערכות מידע גאוגרפיות

מחקר לשם מילוי חלקי של הדרישות לקבלת תואר "דוקטור לפילוסופיה"

מאת

רונה וינטר-לבנה

הוגש לסינאט אוניברסיטת בן גוריון בנגב

יוני 1400 תמוז תשע"ג

באר שבע

Rona Winter-Livneh’s PhD Dissertation | Ben-Gurion University of the Negev Page 79