Lecture                                                                                                                                        

Eleni ASOUTI, Andrew FAIRBAIRN, Louise MARTIN, Nerissa RUSSEL, Denise CARRUTHERS
Institute of Archaeology, University College London, 31-34 Gordon Square, London WC1H 0PY, UK. / School of Archaeology and Anthropology, Australian National University, Canberra, AU. / University of Edinburgh, UK. / Cornell University, New York, Ithaca, US.
/ e.asouti@ucl.ac.uk / andrew.fairbairn@anu.edu.au / louise.martin@ucl.ac.uk / nr29@cornell.edu / denise@permedia.ca

SUBSISTENCE ECONOMY IN CENTRAL ANATOLIA DURING THE NEOLITHIC : THE ARCHEAOBOTANICAL EVIDENCE

by Eleni Asouti and Andrew Fairbairn                                                                                            

Some central issues concerning the emergence of agriculture in Anatolia

The systematic retrieval and study of charred plant remains from several early sites in Southwest Asia (Central Anatolia (in particular the excavations at Canhasan by David French and his team) was indeed the place where the retrieval of charred plant remains by means of mechanical flotation (French 1971) was implemented for the first time worldwide, thus revolutionising our understanding of past societies not only in terms of architectural development or artefacts, but also in their full environmental and economic context.) have demonstrated the diversity of the Neolithic subsistence strategies, with accumulating evidence on the existence of complex processes of agricultural uptake and considerable regional variation in subsistence practices (cf. Miller 1991; Willcox 1999; Colledge 2001). In this context, much research has focused on the emergence of sedentary communities with complex systems of social organisation and symbolic expression, which pre-date the appearance of fully-fledged agricultural economies (cf. Cauvin 2000). In the southern and central Levant, Southeast Anatolia and Northeast Iraq, this process (beginning during the Pre-Pottery Neolithic A) saw its expression in the appearance of sedentary settlements, few of which (Jericho, Iraq ed-Dubb and Aswad I) have given definite evidence for at least some exploitation of cereal domesticates (Van Zeist and Bakker-Heeres 1982; Hopf 1983; Watkins et al. 1991; Kozlowski 1989; Rosenberg et al. 1995; Hole 1996; Colledge 2001). For the most part these early settlements had developed ‘foraging economies’, relying mainly on the gathering and hunting of a wide variety of wild plants and animals, with only minor contributions from cultivated crops (Harris 1998). (For a comprehensive review of the archaeobotanical record see Garrard 1999; Colledge 2001.)

It was within this research background that a strong inclination has developed to identify first in Southeastern Anatolia and subsequently in Central Anatolia too, a pattern of subsistence practices based on the preponderance of gathered plant foods as opposed to cultivation. Thus, for the period corresponding to the early PPNB (otherwise considered as a period of agricultural expansion) it has been argued that the Grill-Building subphase at Çayönü demonstrates exactly this, with a local process of neolithisation emerging from an essentially hunting and gathering background at the end of the 9th millennium cal BC (see Özdogan 1995, 1997a, 1997b). The large quantities of wild pulses recovered from Çayönü have been used to support this interpretation. Furthermore, the apparent importance of pulses at Cafer Höyük (where morphologically domesticated cereals did exist) was also taken to exemplify an idiosyncratic trait of the Early Anatolian Neolithic, whereby agriculture played a secondary role compared to plant gathering, even where it was practiced. The same line of argument lies behind the consideration of Asikli Höyük as a society of sedentary hunter-gatherers that managed to persist for 1000 years apparently without an overt reliance on cultivation (Esin and Harmankaya 1999). There have also been suggestions that plant domesticates played a minor role at Çatalhöyük, although the little evidence published so far points firmly in the opposite direction (Özdogan 1995, 1997a, 1997b; but see Helbaek 1964; Asouti et al. 1999; Fairbairn et al., in press).

Such an approach calls for some reconsideration of the archaeobotanical evidence. To begin with, the lack of morphological evidence for pulse domestication does not necessarily signify that they were not actually cultivated. (For a summary of the views on this issue see Miller 1991. The main morphological change in domesticated pulses is the development of a non-shattering seed pod, very rarely preserved in archaeological charred specimens. Furthermore, changes in size (i.e., to larger forms) as a result of domestication occur very gradually in pulses, resulting in a difficulty to classify intermediate forms.) Indeed, even in sites as late as Erbaba in the Beysehir area we do not find pulses with the full morphological characteristics of domesticate forms (Van Zeist and Buitenhuis 1983). (By contrast, the presence of morphologically domestic pulses in Çatalhöyük has been ascertained through finds of stores of large, well-formed seeds. Storage contexts are all important in this respect since mixed refuse deposits deriving from all kinds of ashy debris, burnt dung, etc., may well include in abundance smaller forms, which in turn are likely to represent crop processing by-products used as fodder and/or fuel.) At the same time, unlike cereals, collecting and processing of wild pulses for food consumption can prove very unproductive in terms of their food value: wild legumes do not form dense stands, and large quantities of seeds cannot therefore be easily collected (Ladizinsky 1989). (The same author has observed for example that it requires approximately 10,000 wild lentil plants (Lens orientalis, producing on average ten seeds per plant) to obtain 1kg of clean seed.) Given the substantial pulse concentrations found in Çayönü and later sites such as Cafer Höyük and Gritille, it is indeed extremely unlikely that the bulk of pulses were actually collected from the wild (Van Zeist 1972; Van Zeist and De Roller 1991–1992; De Moulins 1993, 1997; Miller, in press).

Therefore, far from arguing that Çayönü forms a par excellence case of sedentary hunter-gatherer society, we could instead view the same evidence as indicative of a local trajectory relating to agricultural production. Furthermore, drawing from the archaeobotanical record, the suggestion has been put forward recently that, during the PPNB, a distinct pattern of plant exploitation is discernible in Southeastern Anatolia characterised by the predominance of pulse crops and a marked preference for the use of wood instead of dung as fuel (Miller, in press). This has moreover been contrasted with sites in the more arid regions of the Syrian steppe (e.g., Abu Hureyra 2) where more drought-resistant crops such as barley became more widespread and the indications for the use of dung as fuel are more prominent. (Miller, in press. The absence of barley cultivation in early Çayönü has also been stressed by Van Zeist (1988). Potential indicators for the use of dung fuel in Abu Hureyra 2 include the presence and abundance of ‘weed’ seeds (notably small-seeded legumes and grasses) which accounted on average for over 85 % of the samples (De Moulins 2000). That such a ‘tradition’ (possibly instigated by environmental differences between Southeast Anatolia and the Syrian steppe) of plant management based on pulses could have had a long history in Southeastern Anatolia is further suggested by the archaeobotanical findings from the early site of Hallan Çemi with the absence of wild cereal grains (Rosenberg et al. 1995). In Çayönü, pulse and cereal domesticates are present from the early levels, including one-seeded einkorn (T. monococcum) and emmer (T. dicoccum). The presence of one-seeded domesticated einkorn indicates that this cultivar was introduced in Çayönü, since its wild progenitor (T. boeoticum ssp. aegilopoides) does not occur naturally in the area (Van Zeist 1972). Furthermore, wild cereals were not intensively harvested by the inhabitants of Çayönü, judging by their low presence and abundance compared to domesticates (ibid.; Van Zeist and De Roller 1991–1992). Thus Van Zeist and De Roller conclude that plant cultivation (with a strong emphasis on pulse crops) was practiced in Çayönü from its very beginning. It should also be noted here that wild cereal types may continue to propagate under cultivation due to spontaneous sowing and furthermore most likely occurred in early stages as ‘weeds’ alongside cultivated cereals (Willcox 1999). Based on the presently available evidence, one should therefore leave open the possibility that plant domesticates were introduced/adopted and adjusted to pre-existing subsistence routines by settled non-agrarian communities, although at a date much earlier than the Grill-Building subphase of Çayönü. Only further sampling of the round house levels at Çayönü and the full analysis of the archaeobotanical remains from early sites such as Göbekli Tepe and Nevali Çori would help in clarifying this issue.)

Plant domesticates and subsistence strategies in Early Neolithic Central Anatolia

Whether or not cereal and pulse domesticates are present in Early Neolithic Central Anatolian sites appears to be mainly a reflection of sampling policies and (to a lesser extent) preservation conditions. The more extensively sampled sites such as Çatalhöyük and Canhasan III (In both Çatalhöyük and Canhasan III mechanical flotation for the retrieval of charred plant remains has been applied (French et al. 1972; Hastorf and Near 1997).) have given the most representative assemblages in this respect, whilst Asikli and Musular (with less favourable preservation conditions) present only a slightly different picture (Van Zeist and De Roller 1995; Mihriban Özbasaran, pers. comm.). (For a preliminary list of the plant taxa retrieved from Musular see Özbasaran 2000.) Suberde was not sampled at all and the effect of this in species presence is most evident: the sole evidence of plant remains amounts to three clay impressions (see Bordaz 1977). It is interesting to note that the same phenomenon applies to the representation of fruit species which serves to illustrate the point as concerns likely sampling and preservation biases (The complete lack of fruit species at Erbaba, however, merits special attention. It is possible that it reflects the absence from the environs of Erbaba of those dryland fruit-producing species (e.g., Pistacia, Celtis, Amygdalus) ‘traditionally’ exploited by Neolithic villages located in the Anatolian plateau.).

Table 1: Presence of the major cereal and pulse domesticates in Neolithic sites of Central Anatolia
(for full references to the individual sites reports see text)

Table 2: Presence of fruit and nut species collected from the wild in Neolithic sites of Central Anatolia
(for full references to the individual sites reports see text)

Nevertheless, it is still possible to discern some divergent local trajectories in evidence for subsistence practices. We shall use as reference points the sites of Asikli and Çatalhöyük, as more representative of the regions of Cappadocia and Konya respectively. In Asikli, cereals and pulses appear to have been of almost equal importance as cultivated crops. The archaeobotanists who analysed the material have concluded that bitter vetch was cultivated, alongside lentil and pea (Van Zeist and De Roller 1995). From the pulses, most important appears to have been bitter vetch, which is widely considered (together with chickpea) to be a SE Anatolian domesticate anyway, in contrast to all the major cereal crops that were (on the basis of the currently available archaeobotanical evidence and genetic studies) introduced to both Southeastern and Central Anatolia from the Levant. (The argument for the first domestication of bitter vetch (Vicia ervilia) at Çayönü is based on the fact that it appears there for the first time in large concentrations, which suggests its status as a domesticate; large concentrations of bitter vetch have been found in several other Anatolian sites (including Canhasan III, Çatalhöyük, Erbaba and recently Asikli too) which suggests a long tradition for the exploitation of this cultivar in Anatolia (Van Zeist 1988; Van Zeist and De Roller 1995). By contrast, the earliest evidence for the domestication of emmer, einkorn and barley comes from sites in the Levant (for useful summaries of the evidence see Garrard 1999; Willcox 1999; Zohary and Hopf 2000; Colledge 2001). Zohary and Hopf (2000) consider the possibility that peas and chickpeas were brought under cultivation in Anatolia; lentils however appear to be closely associated with the onset of cereal cultivation. The genetics argument for a single-event domestication of the founder crops is developed in detail by Zohary (1996, 1999).) Further evidence for the important role of cereals in daily life is provided by the fact that processing activities (including threshing) were taking place on site (Van Zeist and De Roller 1995). The weed flora also indicates that the inhabitants of Asikli probably practiced rainfed cultivation on raised surfaces next to the Melendiz River and/or hillsides and dry areas in the valley, close to the settlement (ibid.). (The analysis of wild taxa has revealed the presence of a mixture of dryland and wetland plants including mainly Buglossoides arvensis, Taeniatherum caput-medusae, Cynodon dactylon, Carex, Eleocharis and Scirpus, accompanied by various grasses and legumes, all present in low numbers. As one of the possible reasons for the overall low abundance of weed seeds has been suggested the likely effect of harvesting practices such as uprooting (Van Zeist and De Roller 1995).) To summarise, the archaeobotanical record does indicate a fairly sedentary lifestyle based on the exploitation of locally available resources, primarily cultivated cereal and pulse crops. The regular presence of fruits such as Pistacia and Celtis throughout the sampled contexts may in part reflect the location of woodland stands close to the site, which were seasonally harvested. The lack of extensive data on fuel exploitation notwithstanding, there is still some evidence that Asikli (being in the upland zone) probably relied on the collection of pistachio and hackberry wood, supplemented by riverine species and oak, all locally available (Woldring and Cappers 2001). (E. Asouti has also examined soil micromorphological thin sections from Asikli that indicated the presence of a similar species range.) To date, no evidence has been produced for the widespread use of dung as fuel. (The low presence and abundance of wild seed taxa could be an indication that dung fuel was not used extensively. Some evidence for the burning of dung fuel has arisen from the analysis of micromorphological thin sections (W. Matthews, pers. comm.) It has to be stressed however that, in the absence of supportive evidence from quantifiable botanical macro-remains, such results cannot be used on their own as evidence for the extent to which dung fuel was used in the past. Only further archaeobotanical research could suggest something definitive.)

Fig. 1: Relative presence of major edible plant taxa (i.e., excluding possible 'weed' species) from Asikli.
Percentage presence scores have been calculated based on the data published by Van Zeist and De Roller 1995
(total no. of samples: 144)

For Celtis in particular, its over-representation in the assemblages can be explained by the fact that hackberry fruit stones are virtually indestructible and require no charring to be preserved in contrast to seed and chaff remains that in dryland sites must be charred, mineralised or desiccated to survive. Some large concentrations of Celtis in Asikli have indicated that hackberry was indeed intensively exploited. However, they are not directly comparable to the abundance values of charred cereal and pulse remains. A more realistic approach should involve comparing only charred nutshell and cereal/pulse grain as representatives of wild and domestic plant foods respectively, assuming that nuts and grains/pulses are good candidates for storable foods from each group (Jones 2000). (The same author however cautions against comparisons that disregard context-related variation in the archaeobotanical record, since cereal remains are much less likely to be adequately represented in non-storage contexts compared to nutshell. Nutshell in particular stands a better chance of being preserved through its exposure to fire as waste and/or fuel-kindling than do prime foodstuffs such as cereals and pulses (see also Legge 1989).) Such a comparison using the published information from Asikli shows the predominance of cultivars, in terms of both abundance and the relative calorific value of each class of material. (The calorific conversion is a relative one based on the conversion of charred remains to an equivalent grain size and then calculating the calorific values of each plant type based on their modern equivalents. The results presented here simply rely on the total abundance for each type of plant material, thus lacking any consideration of context-related variation (very little contextual information for the archaeobotanical assemblages has been published so far) but they are useful nonetheless for prompting discussion and also offer a more balanced perspective than presence and abundance values do.)

Fig. 2: Comparison of cultivated (cereals and pulses) and gathered (nuts and fruits) plant resources preserved by the same means (i.e. charring) based on (on the left) charred plant remain abundance and (on the right) calorific conversion from Asikli Höyük (abundance values were calculated from the data published by Van Zeist and De Roller 1995)

In Çatalhöyük, (Information on the plant remains has been compiled from Helbaek 1964; Asouti et al. 1999. A major synthesis of the first results of the renewed archaeobotanical investigations will appear shortly in print (Fairbairn et al., in press).) emmer wheat (T. dicoccum) and bread wheat (T. aestivum) are dominant, with smaller quantities of einkorn and barley. These dominate the flotation samples and the storage contexts in similar proportions. (That cereals were important in Çatalhöyük is also evident in the fact that cereal chaff and straw were extensively used in mud-brick and pottery manufacture. By contrast, wild resources such as reeds (Scirpus) and grasses were the primary raw material for basketry and mat-weaving (Fairbairn et al., in press).) Legumes at Çatalhöyük overall are less archaeologically visible than in Asikli but still form an important element of the charred plant assemblage. Bitter vetch and lentil are again the most common pulses, with pea and chickpea less so. Charred remains of Celtis, Pistacia, Prunoideae (wild plums) and Amygdalus (almond) were also present throughout the sampled sequence.

Fig. 3: Relative presence of major edible plant taxa (i.e., excluding possible 'weed' species) from Çatalhöyük.
Percentage presence scores have been calculated based on the data made available by the renewed archaeobotanical
research in Çatalhöyük (Fairbairn et al., in press; total no. of samples: 334)

Fig. 4: Comparison of cultivated (cereals and pulses) and gathered (nuts and fruits) plant resources preserved by the same means (i.e. charring) based on (on the left) charred plant remain abundance and (on the right) calorific conversion from Çatalhöyük (abundance values were calculated from data to be published soon; Fairbairn et al., in press)

Most of the wild plant seeds (mainly small-seeded legumes, reeds and wild grasses) and the cereal chaff from the flotation samples appear to be largely derived from material fed to animals and were probably charred through the burning of dung fuel. (The occurrence of many typical weed taxa and cereal chaff in the dung assemblages testifies to the integrated character of agropastoral production in Çatalhöyük (Fairbairn et al., in press).) Dung fuel was present throughout the excavated sequence. Its clear predominance however is manifested in the earliest levels and is further matched by low densities of wood charcoal (mainly of riverine species) from the same levels. In the later levels this phenomenon is reversed, with much higher densities of charcoal and overall lower archaeological visibility of dung fuel compared to the early assemblages (Fairbairn et al., in press; Asouti and Fairbairn, forthcoming). At the same time, a much wider array of tree and shrub species make their appearance most of which are associated with oak woodland vegetation (Asouti and Hather 2001).

Overall, apart from fuel, there is little evidence to suggest that there were marked changes in plant-based subsistence at Çatalhöyük over time. The archaeobotanical record shows a remarkable continuity in crop usage and presence throughout the period sampled. However, this picture of continuity and stability in agricultural practices comes together with evidence suggesting that reliable crop cultivation is unlikely to have taken place in the immediate environs of Çatalhöyük, at least not in terms of optimal production (i.e., risk buffering in the medium and/or long term). Geomorphological investigations have indicated that the site was founded on a low-lying alluvial delta/floodplain, lacking substantial raised dry surfaces (Roberts et al. 1996, 1999; Roberts and Boyer 1999). The soils of the area were saturated for much of the year and the spring floods (either annual or at longer intervals) would have inundated large areas of the alluvial basin for prolonged periods, thus destroying any cereal crops that had not been planted on raised locations. On environmental grounds, the most optimal source for crops appear to be the Neogene terraces and the low hills flanking the Konya Plain, some 10–12km to the south of Çatalhöyük. (The question of the location of arable fields at Çatalhöyük has been the subject of intense debate within the project. The weed flora contains both wetland and dryland species, but since many of them were probably incorporated in the archaeobotanical record via dung burning it is likely that they do not represent an accurate reflection of cropping practices per se (although the occurrence of many dryland weed seeds co-varies with that of cereal crops; Fairbairn et al., in press). Recent phytolith studies have also provided some indications for the occurrence of dryland cereal crops (Arlene Rosen, pers. comm.) In objection to this interpretation, it has been pointed out that the site is in a similar situation to settlements occupying the Mesopotamian lowlands, where successful cropping occurred in association with the annual floods (cf. Charles 1988; Potts 1997). The environment at Çatalhöyük however is qualitatively different. The site lies in the middle of the Çarsamba fan, which is deposited in a non-outlet basin (the Pleistocene Konya palaeolake) and is characterised by permanent high water levels. Alluvial sediments accumulated gradually on the flat marl lake floor and there is an absence of evidence for the development of levées and extensive dry raised surfaces. Fine-grained sediments gradually prograded towards the centre of the basin to the north, where they settled in the form of heavy clays. At the same time, the whole area slopes from SW to NE, resulting in a higher water table in the northern part of the plain. Overall, save modern drainage works, drainage in the south and west of the Çumra area (i.e., closer to the foothills) is much better than in the east and north (where Çatalhöyük is located) thus leading to the formation of extensive marshes and backswamps. Presently the Neogene terraces south of Çumra are classified as the best agricultural soils for rainfed cultivation in the plain (Driessen and De Meester 1969; Roberts et al. 1999).

Still, one cannot claim that the floodplain was not used at all for cropping. It is entirely possible that certain pulse crops were spring-sown on freshly exposed patches of alluvium after spring floods had retreated (Fairbairn et al., in press). From an archaeobotanical point of view, only the analysis of the crop stores retrieved in large numbers by Hans Helbaek in the 1960s will furnish some positive indications on the issue of field location (through the examination of the associated weed floras). However, if cereals were indeed grown in the immediate environs of Çatalhöyük, the way of cropping seriously challenges currently established assumptions about Neolithic cereal cultivation. It would imply one or all of the following:

Spring-sown cereal varieties had developed; this would have been much earlier than we know at the moment (see Oates and Oates 1977). Winter-sown cereals need more stable, drier conditions and (while they could have survived the occasional prolonged flood) they could not persist in soils that were wet through much of the autumn, winter and spring. Permanent high water levels prevent germination (Hook 1984:268) and lead to cessation of growth and/or death through hypoxia or anoxia in non-adapted plants such as the cereal crops (Trought and Drew 1981; papers in Kozlowski 1984).

Some form of flood control existed, at least locally (i.e., in the form of drainage ditches or embankments none of which is evidenced in Çatalhöyük); this again would be much earlier than we have evidence from elsewhere in Southwest Asia (cf. Oates and Oates 1977).

Cropping was dispersed and very mobile; this would also be a high risk strategy and prone to low yields or even total crop failure in the event of particularly wet winters or pronounced spring floods.)

The geomorphological evidence together with the results of charcoal analysis (indicating the regular collection of firewood from wet riparian forests, oak woodland and woodland steppe alike) (Asouti and Hather 2001; Fairbairn et al., in press) suggest that the Neolithic community of Çatalhöyük routinely exploited on a seasonal basis widely dispersed territories (including the marshes, the floodplain, the hills and the steppe) for pasture, food, fodder and fuel, a pattern of resource use which could have been well-suited to the extreme environmental gradients and resource instability characterising the semi-arid Konya Plain. In this sense Çatalhöyük appears to be different from Asikli, although much of this difference could be explained as a result of the contrasting environments of the Konya Plain and the Cappadocian highlands respectively.

The origins of agriculture in Central Anatolia

As to the origins of cereal and pulse cultivation introduced in Central Anatolia during the Early Neolithic the preceding discussion has hopefully demonstrated that agriculture is likely to have arrived in the region from the southeast. The crop assemblages of Asikli and, to a lesser extent, Çatalhöyük (characterised by the greater importance of pulses) indicate clear links with the Southeast Anatolian Neolithic complex. (This is not to deny differences that have been observed in other aspects of daily life and material culture between Southeast Anatolia and the Early Central Anatolian sequence, as for example in architecture, artistic expression and lithic industries. The range of crops utilised at Asikli and Çatalhöyük are typical of those used in the Neolithic across Central Anatolia and the presence of emmer, einkorn, bitter vetch, chickpea and lentil presents obvious links to Southeast Anatolia and the Levant. One difference is the occurrence of naked six-row barley which is present at most sites in the region as well as Hacilar (Helbaek 1970) and Ilipinar (Van Zeist and Waterbolk-van Rooijen 1995) but is notably absent from contemporary sites in the southeast (see tables in Garrard 1999). Barley crops are known to perform better than wheats in harsher conditions (Zohary and Hopf 2000) such as increased aridity/salinity and short growing seasons. So this preference may reflect differences in environmental conditions between Central Anatolia and the southeast, although we should not discount cultural explanations including likely culinary preferences or barley’s traditional use as a fodder crop (Zohary and Hopf 2000).) Further than this, Çatalhöyük provides an example of how the construction of permanent dwellings and the use of plant domesticates probably did not go hand in hand with reduced seasonal mobility and considerations about the availability of prime agricultural land, as it appears to be the case with Asikli. An explanation of this phenomenon should therefore also address the question of the origins of the community of Çatalhöyük. Three possibilities are open to discussion:

1. The site was settled by agriculturalist-colonisers

If this is correct then our assumptions about site selection by groups we view as primarily agriculturalists have to be re-addressed. The optimal agricultural location would be at the southern edge of the Çarsamba floodplain on the Neogene terraces and the low-sloping hills. The geomorphological evidence shows that Çatalhöyük was in all likelihood far from this. Another possibility is that the site was initially settled and farmed and then, being overwhelmed by flooding, saw the transference of farming to the hills. This hypothesis suffers because the soils underlying the alluvium are highly calcareous (hence unsuitable for arable exploitation) (Calcareous marls cannot support crops unless broken up by terracing or deep ploughing, due to poor root penetration. The same holds for heavy clays.) and the radiometric dates for the onset of alluviation pre-date our earliest settlement dates. (c. 7,480 cal BC for the onset of the earliest excavated phases and 7,550 cal BC for the start of the alluvial deposition (Cessford 2001; Roberts et al. 1999).)

2. Acceptance and use of crops/new technologies by a population already using the area and fitted into a pre-existing resource round

In this case the non-optimal farming location can be explained as being the traditional ancestral home, people being bound to it by familiarity, belief and history. The planting of crops was adjusted into an established cultural landscape, for centuries part of tribal or kin territories used for hunting and other resource extraction. Pinarbasi A shows that mobile hunting groups occupying seasonal campsites were present in the area, inhabiting wetland locales before Çatalhöyük was established, although it has furnished no evidence for the exploitation of plant resources apart from firewood gathering (Watkins 1996; Asouti, forthcoming; Mark Nesbitt, pers. comm.).

3. A combination of the above

This appears to be the most likely scenario. First, as noted before it is possible to discern in the patterns of plant resource exploitation a ‘tradition’ of mobility and diversification, which would accord with our evidence for occupation prior to Çatalhöyük. Pinarbasi has given the strongest indications for such a pattern of mobility in the Konya Plain during the Neolithic which furthermore persisted into later periods, whilst the results of surface surveys in the Konya Plain itself have produced the first tangible evidence for a long tradition of settlement on wetland environments in this area. We have argued for the existence of comparable patterns in Çatalhöyük with the great dispersion of resources and their predicted seasonal exploitation. In this sense, there might be some support for ideas envisaging the dissolution of early conservative societies such as Asikli, followed by their dispersion and re-assembly elsewhere. However, it is also possible to argue for the presence of a ‘foreign’ element too, bringing into this area pre-existing notions of exploitable resources from elsewhere. Thus, the preference for dung fuel over firewood which is so marked in the early levels of Çatalhöyük (otherwise difficult to interpret on purely environmental and/or functional grounds) (Asouti and Fairbairn, forthcoming; Fairbairn et al., in press) might reflect the attitude of a ‘frontier group’ with entrenched beliefs about fuel use, and is much reminiscent of patterns of fuel exploitation familiar from the northern Levant (Miller, in press). Such an inclusive interpretation could benefit from further investigations and comparisons with other classes of archaeological evidence. However, a data-informed understanding of the processes involved will be achieved only when a) The core of Çatalhöyük is excavated and comprehensively sampled, and b) Other early sites that exist in this area are investigated in a comparable way (i.e., with full archaeobotanical and palaeoenvironmental analyses).

Acknowledgements

We wish to thank the CANeW organisers for inviting us to contribute to the works of the Round Table. Ian Hodder and Christine Hastorf have facilitated our research in Çatalhöyük that marked the beginning of our active involvement in the archaeobotany of Neolithic Central Anatolia. The ideas and arguments presented in this paper have benefited from numerous discussions and communications with Hijlke Buitenhuis, Susan Colledge, Douglas Baird, Füsun Ertug, Dorian Fuller, Louise Martin, Wendy Matthews, Naomi Miller, Julie Near, Mihriban Özbasaran, Neil Roberts, Arlene Rosen and Henk Woldring. The final responsibility for the opinions expressed in this paper lies of course with us.

References

Asouti, E., forthcoming. Woodland vegetation and fuel exploitation at the prehistoric campsite of Pinarbasi, south-central Anatolia, Turkey: the evidence from the wood charcoal macroremains.
Asouti, E. A. Fairbairn, forthcoming. Animal dung versus firewood: a history of fuel exploitation at the Neolithic site of Çatalhöyük
Asouti, E., A. Erkal, A. Fairbairn, C. Hastorf, A. Kennedy, J. Near and A. Rosen, 1999. Archaeobotany and related plant studies. In Çatalhöyük 1999 Archive Report. Edited by I. Hodder and the Çatalhöyük Research Trust. Online: http://catal.arch.cam.ac.uk/Archive_rep99
Asouti, E. and J. Hather, 2001. Charcoal analysis and the reconstruction of ancient woodland vegetation in the Konya Basin, south-central Anatolia, Turkey: results from the Neolithic site of Çatalhöyük East. Vegetation History and Archaeobotany 10, 23–32
Bordaz, J., 1977. Beysehir–Sugla Basin, 1976. Anatolian Studies 27, 32–33
Cauvin, J., 2000. The birth of the gods and the origins of agriculture. Cambridge: Cambridge University Press
Cessford, C., 2001. A new dating sequence for Çatalhöyük. Antiquity 75, 717–725
Charles, M., 1988. Irrigation in lowland Mesopotamia. Bulletin on Sumerian Agriculture 4, 1–39
Colledge, S., 2001. Plant exploitation on Epipalaeolithic and early Neolithic sites in the Levant. Oxford: British Archaeological Reports, International Series 986
Driessen, P. and T. de Meester, 1969. Soils of the Çumra area, Turkey. Wageningen: Centre for Agricultural Publishing and Documentation
Esin, U. and S. Harmankaya, 1999. Asikli. In M. Özdogan and N. Basgelen [eds.]. Neolithic in Turkey, the cradle of civilization. New discoveries. Istanbul: Arkeoloji ve Sanat Yayinlari, 115–132
Fairbairn, A., E. Asouti, J. Near and D. Martinoli, in press. Macro-botanical evidence for plant use at Neolithic Çatalhöyük. Vegetation History and Archaeobotany
French, D., 1971. An experiment in water-sieving. Anatolian Studies 21, 59–64
French, D., G. Hillman, S. Payne and R. Payne, 1972. Excavations at Can Hasan III 1969–1970. In E. Higgs [ed.]. Papers in economic prehistory. Cambridge: Cambridge University Press, 181–190
Garrard, A., 1999. Charting the emergence of cereal and pulse domestication in Southwest Asia. Environmental Archaeology 4, 67–86
Harris, D., 1998. The spread of Neolithic agriculture from the Levant to western-central Asia. In A. Damania, J. Valkoun, G. Willcox and C. Qualset [eds.]. The origins of agriculture and crop domestication. Aleppo: ICARDA, 64–82
Hastorf, C. and J. Near, 1997. Archaeobotanical archive report. In Çatalhöyük 1997 Archive Report. Edited by I. Hodder and the Çatalhöyük Research Trust. Online: http://catal.arch.cam.ac.uk/Archive_rep97
Helbaek, H., 1964. First impressions of the Çatal Hüyük plant husbandry. Anatolian Studies 14, 121–123
Helbaek, H., 1970. The plant husbandry of Hacilar. In J. Mellaart [ed.]. Excavations at Hacilar. Edinburgh: Edinburgh University Press, 189–244
Hole, F. 1996., The context of caprine domestication in the Zagros region. In D. Harris [ed.]. The origins and spread of agriculture and pastoralism in Eurasia. London: UCL Press, 263–281
Hook, D., 1984. Adaptations to flooding with fresh water. In T.T. Kozlowski [ed.]. Flooding and plant growth. London: Academic Press, 265–294
Hopf, M., 1983. The plants found at Jericho. In K. Kenyon and T. Holland [eds.]. Excavations at Jericho V: the pottery phases of the tell and other finds. London: British School of Archaeology in Jerusalem, 580–621
Jones, G., 2000. Evaluating the importance of cultivation and collecting in Neolithic Britain. In A. Fairbairn [ed.]. Plants in Neolithic Britain and beyond. Oxford: Oxbow, 79–84
Kozlowski, S.K., 1989. Nemrik 9, a PPN site in northern Iraq. Paléorient 15, 25–31
Kozlowski, T.T. [ed.], 1984. Flooding and plant growth. London: Academic Press.
Ladizinsky, G., 1989. Origin and domestication of the Southwest Asian grain legumes. In D. Harris and G. Hillman [eds.]. Foraging and farming: the evolution of plant exploitation. London: Unwyn Hyman, 374–389
Legge, A., 1989. Milking the evidence: a reply to Entwhistle and Grant. In A. Milles, D. Williams and N. Gardner [eds.]. The beginnings of agriculture. Oxford: British Archaeological Reports, International Series 496, 217–242
Miller, N., 1991. The Near East. In W. van Zeist, K. Wasylikova and K.-E. Behre [eds.]. Progress in Old World palaeoethnobotany. Rotterdam: Balkema, 133–160
Miller, N., in press. Tracing the development of the agropastoral economy in Southeastern Anatolia and Northern Syria. Paper presented at the International Workshop, ‘The transition from foraging to farming in Southwest Asia’, September 7–11, 1998, Groningen, the Netherlands
Moulins, D. de, 1993. Les restes de plantes carbonisées du Cafer Höyük. Cahiers de l’Euphrate 7, 191–234
Moulins, D. de, 1997. Agricultural changes at Euphrates and steppe sites in the mid-8th to the 6th millennium BC. Oxford: British Archaeological Reports, International Series 683
Moulins, D. de, 2000. Abu Hureyra 2: plant remains from the Neolithic. In A. Moore, G. Hillman and A. Legge [eds.]. Village on the Euphrates. From foraging to farming in Abu Hureyra. Oxford: Oxford University Press, 399–416
Oates, D. and J. Oates, 1977. Early irrigation agriculture in Mesopotamia. In G. de G. Sieveking, I. Longworth and K. Wilson [eds.]. Problems in economic and social archaeology. London: Duckworth, 109–135
Özbasaran, M., 2000. The Neolithic site of Musular-Central Anatolia. Anatolica 26, 129–151
Özdogan, M., 1995. Neolithization of Europe: a view from Anatolia. Part 1: The problem and the evidence from East Anatolia. Porocilo o raziskovanju paleolitika, neolitika in eneolitika v Sloveniji 22, 25–61
Özdogan, M., 1997a. The beginning of Neolithic economies in Southeastern Europe: an Anatolian perspective. Journal of European Archaeology 5(2), 1–33
Özdogan, M., 1997b. Anatolia from the last glacial maximum to the Holocene climatic optimum: cultural formations and the impact of the environmental setting. Paléorient 23, 25–38
Potts, D., 1997. Mesopotamian civilization: the material foundations. London: Athlone Press
Roberts, N., S. Black, P. Boyer, W. Eastwood, H. Griffiths, H. Lamb, M. Leng, R. Parish, M. Reed, D. Twigg and H. Yigitbasioglu, 1999. Chronology and stratigraphy of late Quaternary sediments in the Konya Basin, Turkey: Results from the KOPAL project. Quaternary Science Reviews 18, 611–630
Roberts, N. and P. Boyer, 1999. Uncovering the Neolithic cultural landscape around Çatalhöyük, Turkey. Unpublished research report
Roberts, N., P. Boyer and R. Parish, 1996. Preliminary results of geomorphological investigations at Çatalhöyük. In I. Hodder [ed.]. On the surface: Çatalhöyük 1993–1995. Cambridge and London: McDonald Institute for Archaeological Research and British Institute of Archaeology at Ankara, 19–40
Rosenberg, M., M. Nesbitt, R. Redding and T. Strasser, 1995. Hallan Çemi Tepesi: some preliminary observations concerning early Neolithic subsistence behaviors in Eastern Anatolia. Anatolica 21, 3–12
Trought, M. and M. Drew, 1981. Alleviation of injury to young wheat plants in anaerobic solution cultures in relations to the supply of nitrate and other inorganic nutrients. Journal of Experimental Botany 32, 509–522
Watkins, T., 1996. Excavations at Pinarbasi: the early stages. In I. Hodder [ed.]. On the Surface: Çatalhöyük 1993–1995. Cambridge and London: McDonald Institute for Archaeological Research and British Institute of Archaeology at Ankara, 47–58
Watkins, T., A. Betts, K. Dobney, M. Nesbitt, R. Gale and T. Molleson, 1991. Qermez Dere, Tel Afar. Interim Report No 2, 1989. University of Edinburgh: Department of Archaeology
Willcox, G., 1999. Agrarian change and the beginnings of cultivation in the Near East: evidence from wild progenitors, experimental cultivation and archaeobotanical data. In C. Gosden and J. Hather [eds.]. The prehistory of food. London: Routledge, 478–500
Woldring, H. and R. Cappers, 2001. The origin of the ‘wild orchards’ of Central Anatolia. Turkish Journal of Botany 25, 1–9
Zeist, W. van, 1972. Paleobotanical results of the 1970 season at Çayönü, Turkey. Helinium 12(1), 3–19
Zeist, W. van, 1988. Some aspects of early Neolithic plant husbandry in the Near East. Anatolica 25, 49–68
Zeist, W. van and J. Bakker-Heeres, 1982. Archaeobotanical studies in the Levant. 1. Neolithic sites in the Damascus Basin: Aswad, Ghoraife, Ramad. Palaeohistoria 24, 165–256
Zeist, W. van and H. Buitenhuis, 1983. A palaeobotanical study of Neolithic Erbaba, Turkey. Anatolica 10, 47–89
Zeist, W. van and G.-J. de Roller, 1991–1992. The plant husbandry of Aceramic Çayönü, SE Turkey. Palaeohistoria 33–34, 65–96
Zeist, W. van and G.-J. de Roller, 1995. Plant remains from Asikli Höyük, a Pre-Pottery Neolithic site in Central Anatolia. Vegetation History and Archaeobotany 4, 179–185
Zeist, W. van and W. Waterbolk-van Rooijen, 1995. Floral remains from Late-Neolithic Ilipinar. In J. Roodenberg [ed.]. The Ilipinar excavations I. Five seasons of fieldwork in NW Anatolia, 1987–91. Leiden: Nederlands Historisch-Archeologisch Instituut te Istanbul, 159–166
Zohary, D., 1996. The mode of domestication of the founder crops of Southwest Asian agriculture. In D. Harris [ed.]. The origins and spread of agriculture and pastoralism in Eurasia. London: UCL Press, 142–158
Zohary, D., 1999. Monophyletic vs. polyphyletic origin of the crops on which agriculture was founded in the Near East. Genetic Resources and Crop Evolution 46, 133–142
Zohary, D. and M. Hopf, 2000. Domestication of plants in the Old World (3rd edition). Oxford: Clarendon Press

____________________________________

ANIMALS REMAINS FROM THE CENTRAL ANATOLIAN NEOLITHIC

by Louise Martin, Nerissa Russell and Denise Carruthers                                                              

Introduction

This paper aims to explore temporal and spatial patterning in animal-related subsistence practices in the Central Anatolian Neolithic, by both presenting new results and reviewing existing available faunal data. Of the many aspects of human-animal interaction that could be explored, we primarily focus on the question of the appearance of domesticated animals at sites. In many respects, the question posed by Payne thirty years ago (1972), as to whether the Central Anatolian sites represent villages of hunter-gatherers, incipient cultivators and domesticators, or fully fledged farmers, has still not been fully answered, and indeed the situation may well be more complicated than implied by those discrete terms. This paper does not aim to classify sites on the basis of faunal remains, but rather to suggest ways that the inhabitants of sites may have procured animals, and to discuss any trends in the zooarchaeological patterning.
Animal bone remains do not always provide clear indications of whether animals were wild and hunted, domestic and herded, or whether there were less clear-cut and maybe more ‘transitional’ practices taking place. Even in situations where large well-retrieved faunal assemblages are present, that cover long time sequences and provide good data (e.g. on osteometrics, age and sex estimates for creating cull patterns), zooarchaeological patterns are often ambiguous and open to more than one interpretation. In particular, the study of Neolithic animal remains from Central Anatolia is complicated by two factors. First, the area lies within the natural distribution zone of (at least) four of the major domesticates: cattle, sheep, goat, and pigs (Uerpmann 1987), and therefore the appearance of domesticates cannot be assessed by tracing the arrival of ‘new’ species to a site. Second, the lack of earlier (pre-Neolithic) comparative material produces a gap in knowledge as to the nature of truly ‘wild’ fauna, for example in terms of faunal distributions, or the morphology and size ranges of animals.
In addition, there has been great debate about what actually characterises domestication, although most would now concur that it is a two phase process (e.g. Clutton-Brock 1992; Smith 1995). The first phase represents a change in the way humans interact with wild animals, for example, a shift to managing wild herds through controlled or selective culling. This need not result in separating stock from wild herds, or observable morphological changes, but represents an important change in the human–animal relationship. The second phase involves a genetic isolation of some livestock from wild populations, subsequent to the new management practices, which at some stage should result in morphological and behavioural changes in those animals. It is only after the stock has been separated from a wild breeding pool that true ‘biological domestication’ has taken place. It should be noted that the first phase need not always lead to the second in a linear fashion; loose herd management may well be a long-term procurement method in itself in some societies, while in others, closely herded animals may be intentionally bred with wild stock to gain particular characters (and hence not truly separated from the wild). For zooarchaeology, the challenge is to be able to identify these different stages of the domestication process through analyses of animal morphological change (including size change), culling strategies, and other contextual evidence.
For Central Anatolia, data from recent field-projects at Asikli Höyük, Pinarbasi A and B, and Çatalhöyük East provide valuable insights and it is on these that discussion will focus, although patterns from Canhasan III, Suberde and Erbaba will also be mentioned. Interpretations of the Central Anatolian evidence are, of course, aided by a rapidly increasing knowledge of the animal domestication processes in eastern Anatolia, the northern Levant and Zagros (e.g. Peters et al. 1999; Zeder 1999).

The animals

For the purposes of this paper, only certain taxonomic categories have been selected for discussion in order to facilitate comparisons between sites. These categories are cattle (wild Bos primigenius and potentially also domestic Bos taurus), equids (including the wild horse Equus caballus, the half ass Equus hemionus and European wild ass Equus hydruntinus), deer (including red deer Cervus elaphus, fallow Dama dama and roe Capreolus capreolus), boar/pig (Sus scrofa), sheep (including Ovis orientalis and Ovis aries), goat (including Capra aegagrus and Capra hircus) and hare (Lepus sp.). It should be noted that sites often include a wide range of other mammalian, bird and fish taxa which will not be discussed here.

Pinarbasi A

The two sites of Pinarbasi A and B are located at the southern edge of the Konya Plain, north-west of the volcanic massif of Kara Dag, within a ridge of limestone hills (Watkins 1996). The sites are approximately 25km south-east of Çatalhöyük, and 12km north-west of Canhasan III. Site A is an open-air camp-site dating to the mid- to late-9th millennium cal BC (ECA I) (The breakdown of the Early Central Anatolian (ECA) into 5 stages follows Matthews’ division. and as such is the earliest excavated site in Central Anatolia. Pinarbasi B is a later ECA III site which will be discussed below.

Fig. 1: The relative proportions of selected mammalian taxa (by NISP) from Pinarbasi A (left) and Pinarbasi B (right)

Figure 1 shows the relative proportions of selected animal taxa from Site A (left histogram) alongside Site B (right histogram), by the Number of Identified Specimens (NISP). (The faunal assemblages from Pinarbasi A and B are the subject of current doctoral research by Denise Carruthers, University of Edinburgh. The remains from site A derive from dated contexts ABR, ABU and ABJ (earliest to latest).) Site A has an extremely small sample of these taxa (n=47) but despite this it is interesting to note that each of the taxonomic categories are represented, and although sheep/goat are present in the highest numbers, they are not as dominant as in later sites, (When taken as a proportion of the total assemblage (including all mammals, birds and fish) sheep and goat constitute 23 % of Pinarbasi A and 55 % of Pinarbasi B.) and other taxa (pig/boar, cattle and hare) are each represented. This relative breadth of taxa at Pinarbasi A is more notable considering that these ‘main animals’ make up less than 50 % of the complete assemblage: a wide array of other animals were found, including fox, beaver, rodents, fish and birds, some of which may have been food animals, others not.

So-called ‘broad spectrum’ assemblages such as this are characteristic of PPNA sites in the southern Levant (e.g. Tchernov 1998; Davis 1985) where they have been interpreted as a response to diminishing large game resources around early permanently settled villages, where sedentary communities exerted hunting pressure on large mammal populations. This is not a convincing explanation for Pinarbasi A, however, which is interpreted as a small seasonal campsite occupation (Carruthers, forthcoming), and hence pressure on local resources is unlikely to have been great. (So far, there is evidence for spring occupation from newborn caprine bones, but other seasons certainly cannot be ruled out.) Rather than reflecting resource stress, the Pinarbasi A assemblage could result from relatively opportunistic hunting in the various environments surrounding the site, such as the lakes, marshes, and plains. It is possible also that the smaller animals such as the carnivores may have been hunted for their fur or skins, and birds for feathers, so the bones may not represent meat consumption alone.
Most of the animals present are interpreted as being wild, but there remains a question about the status of the caprines (the sheep/goat category). As yet, morphometric data are inconclusive as to whether they are wild or domestic, and hence suggestions remain speculative. One the one hand, it is not inconceivable that the caprine bones belong to herded domesticates, since in eastern Anatolia there are claims for domestic sheep and goats at c. 8500 cal BC (Peters et al. 1999), by the time of the occupation of Pinarbasi A.. Hunter-gatherer groups could have adopted them and integrated them into their ways of life, meaning the site could possibly represent a combined hunting and herding encampment.
On the other hand, as will be n below, there are no claims for domesticates in the Central Anatolian region at this date, which would make their presence at Pinarbasi A isolated, and may lend weight to an argument for them being wild. This interpretation finds some support in the fact that only sheep have been identified within the ‘sheep/goat’ category, and not goats (although sample sizes are extremely small); wild sheep would have been available in the local plains and lower foothills, while wild goats are likely to have occupied more distant upland habitats. (Although Uerpmann (1987:113–118) and Wasse (2001) caution against using modern habitats to predict early Holocene goat distributions, and note that in the past wild goat had more extensive territory.) Tentatively, therefore, Pinarbasi A does not m to have the ‘package’ of sheep and goats characteristic of sites outside core domestication areas where domesticated animals were introduced (see Horwitz et al. 1999). Clearly, until more morphometric data are available, the status of the Pinarbasi A caprines remains unknown, and whether the site represents a hunter-gatherer, or hunter-herder occupation cannot yet be determined.

Asikli Höyük

The other ECA I site that has so far yielded fauna is the sedentary village of Asikli Höyük in the Cappadocia area, where Buitenhuis’ analysis shows very different results to those from Pinarbasi A. Figure 2 is based on Buitenhuis’ breakdown of taxa through the phases of the site (Buitenhuis 1997:657, Table 3) and demonstrates a clear dominance of caprines throughout the sequence – they constitute over 70 % of each phase. The other taxa represented are cattle, equid, pig/boar and hare, each present in relatively low proportions, and cervids which are relatively rare. The equids and deer represent wild fauna, while Buitenhuis reports that the domestication status of cattle and pigs is as yet uncertain (1997:659); so far, the sheep and goat have been the focus of interest.

Fig. 2: The relative proportions of selected mammalian taxa (by NISP) from the levels of Asikli Höyük (following Buitenhuis 1997)

Both the sheep and goat remains from Asikli Höyük are morphologically indistinguishable from their wild counterparts, showing no size differences from regional wild comparatives (Buitenhuis 1997:660, Figure 3). What is notable, however, about Asikli is the dominance of sheep and goats compared to contemporary sites (e.g. in eastern Anatolia), and while this may be due to its location providing an abundance of available wild caprines, it also hints at a more focused procurement strategy. Buitenhuis finds support for this idea in the sheep and goat cull patterns, which show the majority of animals to have been killed before 4 years of age, with most in the 2.5–4 years range (Buitenhuis 1997:659), and also in the large number of perinatal bones. This combined evidence leads him to suggest that there was some form of control over the kills, and that early attempts at keeping animals may have resulted in the high infant mortality (since the hunting of newborns would be an unlikely strategy).
There is no straightforward interpretation of each of these lines of evidence. Distinguishing between selective hunting practices (which might, for example, target particular animal age groups for meat), and loose herd management practices (which might involve the culling of the same age groups, but with the goal of herd reproduction) is extremely difficult through zooarchaeological methods, primarily because the cull patterns would appear similar (e.g. Collier and White 1976). Nevertheless, the Asikli pattern does seem very focused, and as such hints at some form of selection. The number of juvenile deaths (12 % from perinatal to 6 months of age) at Asikli is also difficult to interpret. While the hunting of pregnant mothers in the spring cannot be ruled out, as Buitenhuis states (1997:659) it seems doubtful as a hunting strategy. As an intentional herd management practice, the culling of very young animals has only been proposed as a dairying strategy (Payne 1973, but see McCormick 1992), but there is no evidence for milking in the Early Neolithic, and therefore this would be an extremely unlikely explanation. Rather, Buitenhuis’ interpretation of these as natural deaths occurring under conditions of some human control, where people had access to these weak young animals and appeared to be consuming them, (This interpretation is based on the peri-natal and juvenile bones being found disarticulated and amongst food middens (Buitenhuis 1997:659).) seems reasonable. (12 % may not be a particularly ‘high’ figure for infant and juvenile mortality, even when compared to flocks under modern herd management conditions (see Payne 1973:301).)
The exploitation of sheep and goats at Asikli has therefore been interpreted as one of herd-management (Buitenhuis 1997), or proto-domestication (Esin 1998), terms which depend on the definition of domestication being used. As Buitenhuis has succinctly stated (Vigne, Buitenhuis and Davis 1999:58), from an anthropological point of view, the caprines at Asikli would appear to be an ‘appropriated’ resource where animals were under human control, while from a biological perspective, they were clearly not yet modified morphologically.
Two recent studies confirm that size change, or morphological change, does not necessarily accompany the early stages of animal management and manipulation. First, at Ganj Dareh in the Zagros, in the heart of wild goat country, the intensive selective culling of goats did not result in their size change (Zeder and Hesse 2000). Perhaps more challenging to zooarchaeological assumptions is the example of Shillourokambos on Cyprus, where Vigne and colleagues have demonstrated that the human importation of pig, fallow deer, sheep, goats and cattle onto the island by the end of the 9th millennium cal BC, did not result in a size change for either sheep or goats (Vigne et al. 2000).
These studies strongly suggest that morphological change can lag behind the changes in human–animal interaction that could be defined as animal management or herd control, maybe for long periods of time. Size change as a marker of domestication may not only require a population to be isolated (in breeding terms) from wild progenitors, as they clearly were in Cyprus, but also to be kept (in part at least) in an anthropogenic environment where strong selective pressures are at play. At Asikli Höyük, we can conclude that those combined conditions did not yet exist.
If it is accepted that the population of Asikli was involved in some form of production of sheep and goats, this leads to the question of whether such methods were ‘home grown’ or developed elsewhere. In the absence of earlier sites local to Asikli, this is very difficult to discuss. Buitenhuis, however, has convincingly argued that while local domestication may be envisagable in this area where wild caprine ancestors are abundant, the lack of any observable development through the long sequence at Asikli rather argues against it (Vigne, Buitenhuis and Davis 1999). Indeed, at sites where we do have convincing evidence for indigenous domestication (e.g. Ganj Dareh (Zeder and Hesse 2000); Mehrgarh (Meadow 1984)) there are certainly visible changes suggestive of internal developments of herding practices. This hints at an external origin for the ideas of herd management practices at Asikli, with eastern Anatolia a likely source since there is evidence for the domestication of sheep at Nevali Çori in the mid-late 9th millennium cal BC (Peters et al. 1999), and evidence of the diffusion of husbandry practices outside that area (e.g. Tell Halula, Saña Seguí and Helmer 1999). But wherever the ideas of animal management came from at Asikli, there is little doubt that they were carried out on local wild stock.

Çatalhöyük East

Analysis of the faunal remains from ECA II Çatalhöyük East is still ongoing and results are therefore preliminary, but it is clear from work to date that on-site deposits are generally dominated by sheep and goats, contrary to Perkins’ results from the 1960s excavations at the site (Perkins 1969). Figure 3 (Figure 3 shows relative proportions quantified by Diagnostic Zones (DZs) following Watson 1979. This method discounts horncores, antlers and other non-standard skeletal elements which may create biases in representation between species (which is a problem of using NISP).) shows a breakdown of selected taxa by area and groups of levels, with KOPAL being an offsite trench, South Pre Level XII representing the earliest on-site deposits excavated, South XII–VII representing a continuation of the occupation sequence (following Mellaart’s 1967 Levels), and North/BACH/Summit covering the latest levels (VII–V) so far studied by the renewed excavation project. (For results of the new dating project for the Çatalhöyük East levels.)

The on-site areas show a roughly consistent pattern, with cattle at less than 20 %, equids varying somewhat through the South area sequence but never constituting more than 15 %, and cervids, boar/pig and hare making up relatively small proportions. Sheep and goats make up roughly 80 % of the earlier South area deposits and the North/BACH/Summit deposits, and a slightly lower 65 % of South Level XII–VII deposits. This strongly contrasts the off-site KOPAL area where cattle dominate and there are relatively high proportions of cervids and boar/pig, with far less sheep and goats, suggesting that the site-edge area sampled saw very different preparation, consumption and discard activities to the onsite areas.

Fig. 3: The relative proportions of selected mammalian taxa (by Diagnostic Zones) from the areas/levels of Çatalhöyük East

The equids and deer are assumed to be wild game, from steppes and woodlands respectively. Whether the boar/pig remains represent wild or domestic animals is uncertain as yet, although the large size of bones and teeth tentatively suggests they were wild, as do their low and slightly decreasing proportions through the sequence. Their low representation is fairly surprising, since they would probably find a suitable habitat in the backswamps and riverine environments in the vicinity of Çatalhöyük.

Sheep and goats

Preliminary analysis of the sheep and goat remains suggest that they are mostly from domesticates. Figure 4 is taken from Aydinoglugil (2001) and shows on the left Buitenhuis’ (1997) plot of the sheep astragalus size from Asikli Höyük compared to those from other Anatolian Neolithic sites and modern wild sheep comparatives; on the right, the same measurements are plotted from the Çatalhöyük East sheep astragali, showing them to be mostly smaller than those from Asikli.

Wild sheep in the Konya Plain would be expected to have a similar size range to wild sheep in Cappadocia, hence a likely interpretation of the smaller sized animals from Çatalhöyük is that they are domesticates. Their size suggests that these animals were genetically isolated from wild sheep, unlike the Asikli animals, and they are shown to be present from the earlier deposits on site (South). This could indicate that the groups who first founded the settlement had domestic sheep. That there are some much larger specimens in Figure 4 shows that wild sheep were also hunted in smaller numbers.

Fig. 4: Sheep astragalus size from (left) Asikli compared to other comparative sites and wild reference material (after Buitenhuis 1997) and (right) Çatalhöyük East (after Aydinoglugil 2001)

Goats are present in much lower proportions than sheep, and consequently there are far fewer measurements on goat bones than there are on sheep. The impression, however, is the same – that they are mainly domesticates with some wild animals.
Animal penning features have been identified in the South area through micromorphological analysis of deposits by Wendy Matthews (1999). Matthews finds interbedded layers of trampled herbivore dung strongly suggestive of in situ stabling, rather than collected dung, and the presence of at least four perinatal sheep/goat skeletons in these deposits shows that birthing ewes or she-goats may have been kept on site, between houses. As well as birthing mothers, the evidence of shed deciduous dentition of sheep/goats shows older juvenile animals to have been kept in the penning area too, for some part of the year at least.
Results of broader flock management issues are not yet available, but it is interesting to consider that there were possibly environmental constraints to herding close to the site at certain times of year. Palaeoenvironmental work around Çatalhöyük East is beginning to suggest large-scale spring flooding of the plain surrounding the site (see Balter 2001), and if this were the case, it may have necessitated seasonal pasturing at distances from the site in spring. Since spring is also the birthing time for caprines, splitting of flocks, with some birthing mothers kept on site in penning areas (above) may have been one solution.

Cattle

Turning to the question of cattle at Çatalhöyük East, our preliminary interpretation of the new (1990s–2000s project) evidence differs from the earlier ideas of Perkins who found the site to be a centre of cattle domestication (Perkins 1969). Perkins identified a decrease in the size of cattle at Level VI, interpreted as biological domestication, and he argued for ‘proto-domestication’ from the earliest levels of the occupation. By contrast, no size diminution is visible from the new data, and the sizes of cattle throughout the available sequence have the same range as those from Asikli Höyük, which are likely to be wild Bos primigenius (Martin, Russell and Buitenhuis in prep). (Considering the evidence shows that the inhabitants of Asikli Höyük had not genetically isolated sheep and goats from wild stock, it is unlikely that they would have done so with the much larger and more difficult to control wild aurochs.) In addition, all the cattle horncores appear to be from morphologically wild animals. A detailed exploration of cull patterns throughout the sequence is needed to assess whether there is evidence of selective culling or herd management or control. So far, there is a tentative indication of a shift between the South and Kopal areas (which show roughly equal proportions of adult males and females to have been killed) and the North, BACH and Summit areas (which have slightly more smaller animals, presumably females), but it is too early to say whether this represents an ‘appropriation’ phase. As yet, the morphological evidence indicates only that the cattle were not biologically domesticated.
In this respect, it is interesting to consider that the highly symbolic treatment of cattle at Çatalhöyük East (The renewed excavation project continues to find evidence for the special treatment of cattle skeletal parts (see archive reports on the Çatalhöyük web site (http://catal.arch.cam.ac.uk/catal/), and Russell and Martin 2000).) related to morphologically wild animals that were probably hunted in some way, rather than herded. The people of Çatalhöyük were demonstrably familiar with animal husbandry and breeding techniques – as seen with sheep and goats – and practiced them for hundreds of years, and yet seem not to have applied these fully to wild cattle. In the southern Levant and Cyprus, there are claims for the experimental domestication and control of cattle by this date (Von den Driesch and Wodtke 1997; Vigne et al. 2000), but there is little evidence so far for such practices in Central Anatolia. This rather challenges the idea that domestication is inevitable once the knowledge is in place. Indeed, perhaps in some cases wild resources were maintained as such for particular purposes, such as hunting, feasting and ritual display, which may have been socially important.

Pinarbasi B

Figure 1 shows the relative proportions of taxa from ECA III Pinarbasi B, which was occupied during the second half of the 7th millennium cal BC, overlapping with the latter part of the Çatalhöyük East sequence. Here, it is interesting to note that sheep and goat remains are more dominant (65 %) within the selected mammalian taxa than they were at Pinarbasi A, and from a much larger assemblage (n=2022), that by contrast to the earlier occupation is mostly made up of the larger mammals. Morphometric analysis shows sheep to be of small size, similar to those believed to be domesticates. (Analysis of the goat remains is not yet complete.) Cull patterns also show an emphasis on young animals, and in addition there are possible penning deposits. In short, all evidence convincingly points towards Pinarbasi B as a herding site, with hunting and other activities taking place (Carruthers, forthcoming). In relation to the issues raised above about the possible need for seasonal pasturing away from Çatalhöyük East, seasonality evidence from Pinarbasi B will be interesting in order to understand whether the small site was visited repeatedly throughout the year, or only during limited times.

Canhasan III, Suberde and Erbaba

How do other Central Anatolian Neolithic sites with available faunal reports fit in with the above picture? The large sedentary ECA II site of Canhasan III in the Karaman Plain is described as having cattle, boar/pig, sheep, goat, equids and cervids, but no quantified data are yet available although Payne states that the main meat source was cattle (Payne 1972). (Whether this was calculated from quantification of relative proportions (e.g. NISP, MNI) or potential meat weights is not stated.) Until fuller publication, the status of the sheep, goat, cattle and pigs remains unknown. Given the date of Canhasan III, overlapping with the latter part of the occupation at Asikli Höyük and the first half of Çatalhöyük East, the site is key to understanding the development of the Neolithic in Central Anatolia, and it would be interesting to compare quantification and morphometric data with those from other sites.
The ECA II site of Suberde is broadly coeval with Canhasan III, and situated at the border of the Konya Plain and the Taurus Mountains. It has a more detailed faunal report, and quantification is given for sheep, goat, boar/pig, cattle and red deer, while a range of other taxa are mentioned as present (Perkins and Daly 1968). Sheep and goat dominate the assemblage, and Perkins and Daly interpret them as wild, since they decline slightly through the sequence. As Payne (1972) convincingly points out, however, the sheep cull profile which shows all the animals to have been culled between 3 months and 3 years is strongly suggestive of herd management/domestication practices. (The same age pattern was interpreted by Perkins and Daly (1968) as resulting from ‘drive hunting’, in which the old and young would have been already taken by predators.) Morphometric data are not available. Boar/pig is interpreted as being wild on the basis of size, and cattle are seen as hunted, following Perkins and Daly’s well-known and controversial argument of a ‘schlepp effect’ seen in the skeletal part representation (1968, but see Turner 1989). In sum, while the original faunal report concluded that Suberde was primarily a hunting village, this interpretation is open to question on several scores, and there are reasons to believe that the sheep are under human control, if not fully domestic. Morphometric data would shed light on this question, and help to fill a gap in knowledge of the developments in livestock husbandry on the Central Anatolian plateau at this key date.
Finally, the later ECA III site of Erbaba, lying to the east of Lake Beysehir, produced a faunal assemblage originally studied by Perkins (1973) and reported by Bordaz (Bordaz and Bordaz 1983:87). Cattle were found to increase between the lower and upper levels, and are interpreted as domestic, although this should be treated with caution since no supporting morphometric data were given. Sheep and goats dominate the assemblage, and it is suggested they were used for secondary products (Bordaz and Bordaz 1983), but again, there are no data presented to substantiate this claim. Part of the Erbaba assemblage was revisited by Makarewicz (1999) who re-examined the Level III material only (i.e. the earliest level). She concurred with the previous study that sheep and goats remains were most abundant, and that cattle, boar/pig and a suite of other taxa were present in much lower numbers. She also suggests that most of the cattle, sheep and goats were domestic on size grounds, but sample sizes of metrics for cattle and goats are too small to produce conclusive results, and no metrical data for cattle are presented to support the claim. The status of the remains of boar/pig are uncertain. In sum, while the presence of the full suite of domesticates may not be particularly surprising at this date, there is still not convincing data to demonstrate that biologically domesticated cattle were present in Central Anatolia.

Conclusions

Discussion of faunal patterning at the sites of Pinarbasi A and B, Asikli Höyük and Çatalhöyük East has been the main focus of this paper. If these sites are seen as small windows into a potentially more complex and diverse situation of animal procurement in Neolithic Central Anatolia, a few concluding points can be made.
1) Pinarbasi A yields the only evidence so far for subsistence practices at a small-scale seasonally occupied Early Neolithic camp site in Central Anatolia, but whether the site was occupied by hunter-gatherers, or hunter-herders has yet to be ascertained.
2) Asikli Höyük shows evidence for the ‘appropriation’ of animal resources (sheep and goats) by human groups, but not full domestication, and it has been suggested that the ideas behind the development are likely to have come from elsewhere, perhaps eastern Anatolia.
3) Çatalhöyük East provides evidence for full mixed farming and livestock, with hunting, trapping, collecting and gathering alongside. Domesticated sheep and goats are present from the earliest known deposits, suggesting that the earliest inhabitants had herd livestock. Cattle, however, are morphologically wild, although it cannot be ruled out that there were selective culling practices in the latter part of the sequence.
4) With respect to the appearance of domestic sheep at Çatalhöyük East, it is interesting to consider where they came from. It seems unlikely that their domestication was a local development. Rather, they might have been adopted by indigenous Central Anatolian groups who were in contact with communities to the east. Alternatively, populations moving into the area may have brought them with them. How they arrived is a question that can only be addressed through a broader discussion of archaeological evidence that is beyond the scope of this paper. But the patterns of temporal change in the Central Anatolia Neolithic do point to co-operation between communities, even if the actual processes of change are unclear (following Vigne, Buitenhuis and Davis 1999:59).
5) What factors may have led to the difference in sheep size observed between Asikli Höyük and Çatalhöyük East? Firstly, while the Asikli caprines were not isolated from wild stock, the Çatalhöyük stock appear to have been. Asikli seems to be situated more centrally in the range of wild sheep in the hills, and hence herd control in this area would not have required removing animals from their habitat. The Çatalhöyük sheep, on the other hand, were probably brought in to the area, serving both to isolate them from wild stock and introduce them to new environments, which were potentially highly impacted by human activity (agricultural and otherwise). It is the introduction of animals into anthropogenic environments that Tchernov and Horwitz (1991) have argued pushes animal body size down, through selective pressures that are not necessarily consciously applied by humans. If this was the case, while there may have been different caprine herding strategies between Asikli and Çatalhöyük (and further research will shed light on this), this was not necessarily the case: similar animal management practices could have resulted in different animal sizes in the two areas.
6) Finally, it is notable that both at Asikli Höyük and Çatalhöyük East, there is little sign of internal changes in animal management through time, with respect to sheep and goats at least. At both sites, the relative proportions of caprines do not show strong trends of increase (or decrease) through time, as if they became more (or less) important. At these early stages of analyses, it tentatively appears that sheep size does not decrease through the sequences either. If these patterns hold up with time, it would suggest a fairly remarkable degree of stability in animal procurement practices, regardless of changes evident in other spheres of life. This contrasts the pattern seen in the southern Levant, for example, where many Early Neolithic sites have observable changes in herding practices through their sequences. A fuller understanding of the issues of animal herding practices and domestication, social change and stability, will be forthcoming with more detailed research at the Central Anatolian Neolithic sites.

References

Aydinoglugil, B., 2001. Reviewing the sheep bones from Çatalhöyük in the light on information on middle Anatolian Neolithic sites and the Ovis orientalis in Konya, Bozdag. Unpublished MSc dissertation, Institute of Archaeology, University College London
Balter, M., 2001. Did plaster hold Neolithic society together? Science 294, 2278–2281
Bordaz, J. and L. Bordaz, 1983. Erbaba: the 1977 and 1978 seasons in perspective. Turk Arkeoloji Dergisi 26, 85–93
Buitenhuis, H., 1997. Asikli Höyük: a ‘protodomestic’ site. Anthopozoologica 25-6, 655–662
Carruthers, D., forthcoming. Hunting and herding in Anatolian prehistory: the 9th and 7th millennium sites at Pinarbasi. Unpublished PhD thesis, University of Edinburgh
Clutton-Brock, J., 1992. The process of domestication. Mammal Review 22/2, 79–85
Collier, S. and J. White, 1976. Get them young? Age and sex inferences on animal domestication in archaeology. American Antiquity 41, 96–102
Davis, S., 1985. A preliminary report of the fauna from Hatoula: a Natufian/Khiamian (PPNA) site near Latroun, Israel. In M. Lechevallier and A. Ronen [eds.]. Le site Natoufien-Khiamien de Hatoula, près de Latroun, Israel: fouilles 1980–1982. Jerusalem: Cahiers du C. R. F. J. 1., 71–98
Esin, U., 1998. Hunted animals at Asikli and the environment. In P. Anreiter, L. Bartosiewicz, E. Jerem and W. Meid [eds.]. Man and the animal world: studies in archaeozoology, archaeology, anthropology and palaeolinguistics – in memorium Sándor Bökönyi. Budapest: Archaeolingua, 215–226
Horwitz, L., E. Tchernov, P. Ducos, C. Becker, A. von den Driesch, L. Martin and A. Garrard, 1999. Animal domestication in the southern Levant. Paléorient 25/2, 63–80
Makarewicz, C., 1999. The faunal remains of Neolithic Erbaba: processes of domestication and the herding economy. Unpublished BA Thesis, Department of Anthropology, Brandeis University
Martin, L., N. Russell and H. Buitenhuis, in prep. The question of cattle domestication at Neolithic Çatalhöyük (Turkey) revisited
Matthews, W. 1999. Micromorphology archive report.
Online: http://catal.arch.cam.ac.uk/catal/Archive_rep99/matthews99.html
McCormick, F., 1992. Early faunal evidence for dairying. Oxford Journal of Archaeology 11, 201–209
Meadow, R., 1984. Animal domestication in the Middle East: a view from the Eastern margin. In J. Clutton-Brock and C. Grigson [eds.]. Animals and archaeology 3: early herders and their flocks. Oxford: BAR International Series 202, 309–337
Mellaart, J. ,1967. Çatal Hüyük: a Neolithic town in Anatolia. London: Thames and Hudson
Payne, S., 1972. Can Hasan III, the Anatolian Aceramic and the Greek Neolithic, in E. Higgs [ed.]. Papers in economic prehistory. Cambridge: Cambridge University Press, 191–194
Payne, S., 1973. Kill-off patterns in sheep and goats: the mandibles from Asvan Kale. Anatolian Studies 23, 281–303
Perkins, D., 1969. Fauna of Çatal Hüyük: evidence for early cattle domestication in Anatolia. Science 164, 177–179
Perkins, D., 1973. The beginnings of animal domestication in the Near East. American Journal of Archaeology 77(3), 278–282
Perkins, D. and P. Daly, 1968. A hunters’ village in Neolithic Turkey. Scientific American 219, 96–106
Peters, J., D. Helmer, A. von den Driesch and M. Saña Seguí, 1999. Early animal husbandry in the northern Levant. Paléorient 25/2, 27–47
Russell, N. and L. Martin, 2000. Neolithic Çatalhöyük: preliminary zooarchaeological results from the renewed excavations. In M. Mashkour, A. Choyke, H. Buitenhuis and F. Poplin [eds.]. Archaeozoology of the Near East IV. Groningen: ARC Publicatie 32, 164–170
Saña Seguí, M. and D. Helmer, 2000. The process of animal domestication in the north of the Euphrates Valley (Syria): socio-economic implications. In G. del Olmo Lete and J.-L. Montero Fenollós [eds.]. Archaeology of the Upper Syrian Euphrates – the Tishrim dam area. Barcelona: Editorial Ausa, 257–269
Smith, B., 1995.The emergence of agriculture. New York: Scientific American Library
Tchernov, E., 1998. An attempt to synchronize the faunal changes with the radiometric dates and the cultural chronology in southwest Asia. In H. Buitenhuis, L. Bartosiewicz and A. Choyke [eds.]. Archaeozoology of the Near East III. Groningen: ARC Publication 18, 7–44
Tchernov, E. and L. Horwitz, 1991. Body size diminution under domestication: unconscious selection in primeval domesticates. Journal of Anthropological Archaeology 10, 54–75
Turner, A., 1989. Sample selection, schlepp effects and scavenging: the implications of the terrestrial mammal assemblage