ÇATALHÖYÜK 1998 ARCHIVE REPORT
Çatalhöyük Animal Bone Report
Nerissa Russell and Louise Martin
The 1998 field season at Çatalhöyük saw continued work on the faunal remains from the site. As in previous seasons, all animal bone material was studied in the laboratory dedicated to this task in the excavation house on site, using the comparative reference collection which has been built up over the last five years. Animal bone material is preserved extremely well in the Neolithic deposits at Çatalhöyük, and this factor, plus the excellent retrieval practices (whereby all deposits are sieved though a 4mm mesh), result in large quantities of material being collected. As in previous field seasons, the sheer quantity of animal bone recovered necessitated a strategy of sub-sampling. Following the system instigated in 1997, the material sampled for study was that from the 'priority' units - those contexts that were deemed to be of prime interest as a result of discussions between both excavation and laboratory staff, and hence selected above others for initial recording and analysis.
The results presented below are preliminary, and are also in most instances based on these sampled 'priority' units, rather than the complete assemblages from the areas of excavation. This report focuses on results from the four areas excavated in 1998 - the Mellaart area, the North area, the BACH area, and Çatalhöyük West (Chalcolithic) - although data are presented from all areas from which there are results.
In the 1998 season, 30,826 pieces of bone were recorded, bringing the total to 64,877 specimens recorded so far. This figure breaks down in the following way for excavation areas: 8476 from the BACH area, 30,053 from Mellaart, 14,440 from North, 8576 from Summit, 3326 from West, and 6 bone tools from the KOPAL trench.
A wide range of taxa have been identified from Çatalhöyük, including sheep and goats, cattle, wild horses and asses, red deer, roe deer, wild boar, foxes, wolves, dogs and other small carnivores, birds and fish, and this season, we can also add bear to the list of species found (see Table 1).
The pattern reported from previous seasons is repeated in that sheep and goat bones predominate in each excavation area. Table 2 shows the percentage of the sheep/goat category ranging from 52% in the Mellaart area to 81% in the North area, when counts are calculated using diagnostic zones (DZs) to standardize quantification across different taxa (following Watson 1979). Variation in the proportion of sheep/goats relates primarily to the representation of cattle and equid remains; for example, the Mellaart area has the highest percentage of equid bones and the lowest proportion of sheep/goat bones, while the North area has low counts of both cattle and equid remains, and a correspondingly high sheep/goat frequency (Table 2). Interpretation of these patterns is complex, and analysis is too preliminary as yet to gauge the effects of animal management and domestication on overall trends, but it is likely that both temporal and spatial/contextual differences are at play. Equid remains, for example, are fewer in the later phases (e.g. North, BACH, Summit and Chalcolithic Çatalhöyük West), maybe suggesting that the availability of wild equids declines through time, or that they are hunted less. On the other hand, the representation of cattle remains is likely to be more influenced by the discard practices, or particular 'installation' practices seen in different excavation areas.
The domestic/wild status of the cattle, sheep and goats at Çatalhöyük is still under investigation, with morphometric analyses awaiting larger samples of measurable bones. The issue is complicated, since the site is in a location where wild ancestors of each would have been present. A preliminary observation, however, is that there is a wide range of bone sizes for each of these taxa, beyond the expectations of sexual dimorphism, and this hints that both wild and domestic varieties were present at the site.
The ratio of sheep to goats for each area (Table 3) shows sheep to be consistently more common than goats. This is a pattern which has been observed at several PPNB sites in the Middle East (e.g. Cayonu, Lawrence 1982; Abu Hureyra, Legge 1996; Ain Ghazal, Wasse 1998; Ghoraife II, Ducos 1993) where it is argued that sheep at least are domestic. It might not be surprising, therefore, to find that the sheep at Çatalhöyük are domesticates, and indeed, sheep would probably have an advantage over goats in this environment of well-watered plains with marked seasonality (cf. Redding 1984). For both sheep and goats, however, wild-type horn cores have been found at Çatalhöyük (in the Mellaart and North areas respectively), showing the need for clarification of this issue.
In the Mellaart area, we continued to investigate patterns of animal bone discard. House floors and occupation layers appear to be relatively 'clean' with only small fragments of undiagnostic bone present. This is seen, for example, in the occupation layers of House 3, space 150, where the small units 3148, 3156 produced a total of 36 bone fragments, which were so highly fragmented that none could be identified to skeletal element or taxon. This is reminiscent of the pattern for House 1 (North area) floors, and perhaps shows that while food producing activities took place on these surfaces, few traces remain.
Large-scale discard of food refuse occurs primarily in the midden areas, as demonstrated by results from space 115, which is an 'external' area. This season, we examined the bone material from unit 3314 which is a large context extending horizontally across much of the exposed midden area. The breakdown of taxa in this context is shown in Table 4, where results are compared to a similarly defined 'ashy midden' deposit also in space 115, excavated in 1997 (unit 1668). While there are differences between the two units, such as the higher proportions of sheep/goat in 3314 and the much higher presence of pigs bones, the general pattern of taxonomic representation (with high percentages of the larger animals) and the signs of carcass treatment are similar. Both contexts contain highly fragmented bones, which probably results from smashing for marrow and bone grease. The fragmentation is unlikely to be a post-depositional phenomenon, since bone surfaces are fresh-looking and unweathered, suggesting relatively rapidly burial. A preliminary interpretation of such contexts is that they contain mixed and rapidly deposited material that has seen an intensive level of processing (i.e. for grease and marrow, as well as meat).
Excavations in space 112 allowed for the investigation of the fills found below a house floor, or rather, the fills used for levelling-up prior to house construction on top. A sequence of units was examined from space 112 (units 2844, 2876, 2897, 3107, 3129), which, on excavation, seemed to suggest redeposited midden, and indeed the bone from here was more weathered and fragmented than 'midden' material, perhaps indicating disturbance and movement. Table 5 shows taxonomic representation data from space 112 alongside those from space 115 midden, and House 2 fill for comparison. Straightforward comparison between these spaces is difficult since the sample size from space 112 is much smaller, but it is apparent that these contexts have fewer remains of larger mammals (cattle/equid size) than the space 115 and 117 units. There are various explanations for this. Either space 112 fills derive from types of middening that we have not yet encountered, which have fewer large mammal remains in; or they derive from middens similar to those in space 115 but have been selected or sorted so that the larger bones are not included in the levelling material; or the process of redeposition has had a biased effect on the destruction of bones, crushing or fragmenting larger bones more, rendering them unidentifiable. Whatever the case, further understanding of the nature of these fills is obviously important in the interpretation of the faunal remains from them.
Another kind of deposit which was observed this season (but not yet recorded, hence no data are available) was that between the walls of the buildings in spaces 112 and 113 (unit 3736). From the narrow space between these structural walls came a high density of animal bone material, which on first impressions seems to be from a single depositional episode. Remains are all from large animals (cattle, equids). Those belonging to cattle could potentially derive from a single individual, although bones were not in articulation, and an initial scan of the material shows that skull, horn core and scapula parts are absent. Interestingly, these 'missing' body parts are those which might be used either in structural installations (skull, horn cores) or as tools (scapulae). A preliminary interpretation of this deposit is that it may result from the consumption of a couple of large animals, and may be a foundation deposit for a building, whereby the remains of a feast were laid down between walls, with certain skeletal parts kept for other purposes.
Particularly noteworthy are the bones of a bear's claw (unit 3603 X4) found in a small room - termed an 'antechamber' to Level VII's shrine 10 during Mellaart's 1960ís campaign, but at that time unexcavated - now called space 159. The whole claw was found in articulation, with all phalanges present except one (digit 4, phalanx 3). The claw awaits secure identification, but appears to be from a right hand side hind foot, presumably of Ursus arctos.
Each of the first phalanges (from the big toe digit to the fifth digit) has small cut marks on their palmar sides, which were made with a very sharp tool, since bone is almost 'shaved' in placed. The placement of these cuts - just below the proximal articulations - might suggest that they served to remove the toes or claws from the rest of the foot, and perhaps also resulted from cutting away the more fleshy parts of the paw. It is possible that the animal's skin could have been still attached to the toes, and indeed that the claw arrived on site attached to a skin is one explanation for the presence of these bones (since no other bear bones have been found). An alternative interpretation is that the claw was an object for use in itself, or had ritual significance. One line of evidence which may support the former interpretation is that small amounts of fine-grained white plaster were found wedged between the adjacent first phalanges, and in fact were moulded to the shape of the bones. Under microscope examination, this plaster was found to resemble the wall plaster so far analysed from Çatalhöyük, rather than the floor plaster (pers. comm. Evan Koppelson). Could this indicate that the claw with attached skin was fixed to a wall by keying the bones into wet plaster? Whatever the case may be, it seems that the wall plaster is closely associated with the bear toes, and not coincidentally juxtaposed through post-depositional mixing.
Other unusual finds from space 159 include a sheep/goat femur which had been incised multiple times for the production of bone rings (in unit 3370, the matrix surrounding the bear's claw). Parallel circumscribing grooves had been made along this bone shaft, and it seems that an attempt to break one of these rings off had split the whole bone longitudinally, rendering it useless, and hence it was discarded. Also noteworthy from this same context is an articulated ankle joint (metatarsal, astragalus, calcaneum and other tarsals) of a small-medium sized equid. Future work in this space will hopefully lead us understand whether and how these finds relate to the surrounding structures.
For the North area, faunal work concentrated on the material from the fills beneath House 1 floors, material from outside House 1 to the south, and samples from House 5. Table 6 shows the taxonomic representation for each of the spaces defined for both House 1 and House 5, and external areas, while Table 7 shows data for House 1 summarized by context type.
Although the numbers of identified bones are frequently very low, it is apparent that sheep and goat bones dominate all internal areas (whether fills or floors) from both House 1 and 5. In the external areas, however, this is less marked, with space 73 to the east of House 1 having higher cattle and boar/pig representation plus a much higher proportion of dog bones. The external space 153, which is to the south of House 1, also has a lower sheep/goat percentage with equid and cattle bones being more highly represented than in any other area in the North. This pattern would seem to reflect discard practices whereby the internal areas are relatively 'clean' and have mainly incidental bone on floors and in fills.
Another observation is that there is very little material at all so far from House 5 floors (space 154), which may indicate that they are fairly clean, although it may also reflect the small areas so far sampled. Similarly, there was little bone from the House 5 bins in space 157, and that which there was is assumed to have entered with a later fill, rather than being in primary association with the features. Space 156 contrasted this picture, by having several large pieces of animal bone in the room, below a secondary (supporting) wall: one of these was a pair of sheep's horn cores attached to a skull fragment (3810), and a clay ball was found centrally between the two horn cores. This would seem to suggest a 'placement' of the pieces before construction of the supporting wall.
Excavations in the BACH area produced several different kinds of contexts for comparison (Table 8). The fills in the large central area of House 3 (space 86) are rich in bone, and are loosely termed 'midden'. They are shown to have a very high proportion of cattle remains, although this is clearly skewed by the inclusion of unit 2296 (see below), which contained 8 virtually complete cattle scapulae. The collapsed roof deposits in space 86 not surprisingly contained little bone, and its origin is as yet unclear: does it represent activities which took place on the roof, or accidental inclusions during roof construction?
The west side of the House (space 158) has middening of a different nature: the bone here overwhelmingly belongs to cattle, although interestingly while this is evident in the raw counts (NISPs) in Table 8, it is not so for the DZ calculation. This suggests that the bone was fragmented in a way that whole ends (DZs) were not present to be counted. This midden contained a disproportionate number of cattle scapulae parts, plus a red deer (Cervus elaphus) scapula (one of the first postcranial deer bones to be found on the site), which led to suggestions that special consumption events (feasts?) were responsible for the deposit, in which there was a strong practice of particular body part selection.
This contrasts with space 88, where unit 2268 seems to have an absence of scapulae, although the more even balance between sheep/goat and cattle remains (in the NISP count) means that this is not exactly a 'mirror image' of unit 2250. There are several scapulae in unit 2289,which underlies 2268, but has not yet been recorded.
Table 9 shows data for the BACH middens according to the different depositional types (primary, bricky midden and black midden). It is interesting to note that equid remains are only present in the bricky midden deposits, which in fact have the highest diversity of taxa represented. The other main observation is that bone fragmentation varied between deposit type, if the difference between NISP and DZ is seen as an approximate measure of fragmentation (with lower fragmentation giving more equal counts). Following this, the primary midden deposits have less fragmented remains that either the bricky or black/ashy deposits.
Many of the 2296 bones directly overlay a phytolith layer, which was seen very clearly as the bones were lifted (had the bones aided the preservation of this matter, or were they associated?). The boar maxilla, however, lay directly over an ashy burnt layer, seeming to represent grasses that had been only very briefly fired and charred. Most of the scapulae lay flat with their dorsal sides facing upwards. Two, however, were almost upright (x2 and x22), abutting the east-west wall, and it seems they must have been leaning against this wall when it fell and slumped. The positions of the bones seem to imply intentional placement, rather than dumping. Surface weathering was slight, indicating that they had not been exposed for any length of time; the scapula blades were generally intact, suggesting a lack of disturbance, and none was burnt.
In terms of bone modification, closer examination of all the material is needed in future, but so far there is evidence that one of the scapula spines was intentionally removed, and one glenoid cavity (articular end) is worn on all sides, as if from use, which might suggest their use as tools. The boar maxilla had been cut in two places, firstly along the medial-saggital plane, bisecting the skull, and secondly at the front of the jaw, just behind the canine (tusk). The large piece of cattle skull was chopped in a transverse manner, just in front of where the horn cores would be, which might suggest either that people were trying to get to the brain, or that they were removing the horn cores.
There are many possible explanations for this deposit. The cluster might represent a mixture of installation elements (i.e. horn cores and skull fragments that may have been inserted into the architecture) and tools or raw material for tool manufacture (scapulae). Is it possible that these elements were all 'active' in the house in some way, and hence dismembered and placed together on abandonment? A closer examination of the scapulae blades for traces of working or use will be interesting in this respect; and it is worth mentioning that such scapula tools are already known from the North Houses 1 and 5. Another possibility is that the cluster represents the remains of a special consumption event, or feast. This builds on the ideas mentioned above that people in House 3 may have preferentially received scapulae at communal cooking events, for consumption in the house, and that this particular deposit represents some kind of abandonment feast. What may argue against this idea is that firstly, other areas so far excavated do not show strong body part selection of other skeletal elements; secondly, the cattle horn cores are clearly not meat bearing bones, and also that so far, no meat-removal marks have been observed on the scapulae.
Mention should also be made of the cattle bukranium (unit 3524 x1) uncovered in the middle of space 86 in House 3. This skull has both horn cores present, both left and right frontals are complete, but the lacrimals and nasals have been removed, along with the rest of the skull, leaving a symmetrical flat piece of skull shortened at the front. On the basis of horn core size and shape, the skull seems to be from a male Bos primigenius, and the horn core texture indicates an adult.
It is interesting to note the total absence of cervid bones (and even antler fragments) from the West mound, and there may be a number of possible explanations for this. It may indicate a decline in woodlands in the area, or that the inhabitants of Çatalhöyük West did not frequent wooded areas for hunting purposes. Alternatively, they may not have engaged in antler working, and hence may not have imported these body parts (which are by far the most common deer elements found on the east mound).
Among the equid remains from Çatalhöyük West, several bones were observed as belonging to 'large' equids, which is likely to mean Equus caballus. If these identifications are confirmed, this would greatly alter our knowledge of the distribution of wild horses in the Holocene.
Modes of Consumption: Daily Household Refuse Versus Feasting Remains
Remains of daily consumption are characterized by highly fragmented bone resulting from processing not only for marrow, but probably also for bone grease or broth. It is hard to know how much meat was actually eaten on a daily basis, but these deposits probably accumulated relatively slowly, and are likely to have been more disturbed by dogs and other scavengers as they lay exposed. What we suggest are feasting deposits stand out as containing large pieces of bone, apparently the result of marrow-cracking but not bone grease processing. There is also usually a greater proportion of taxa larger than sheep-size. A sheep can be consumed by a household, but a larger animal can only be utilized fully by a larger group, unless the meat is dried or smoked for storage. Feasting remains are more likely to be deposited quickly, both because the larger amount of bone is more of a problem, and because they may have ritual associations that demand special treatment. These remains therefore tend to exhibit greater integrity: we often find several pieces of the same animal.
Table 12 shows the identified taxa for the units that we have tentatively identified as deriving from daily household waste or feasting remains. (Other units show signs of being redeposited, or of containing material from several sources.) There is not a clear pattern of differing taxa between these two groups of units, in part because the feast in units 3036/3037 in the North area included one or more sheep. Nevertheless, there is clearly a lot of variation among the units that no doubt results from many factors, including differing household hunting success or access to domestic animals as well as postdepositional mixing of bone from different events. Table 13 and Table 14 do indicate that the feasting units have larger pieces (as measured by length and especially by weight) of certain taxa, principally cattle and sheep. Unit 1873 is excluded from these tables because it contains a mixed deposit, with what appear to be feasting remains spread across the bottom of the unit and the remains of daily consumption in the upper part.
The greater size of the relatively rare cervid remains in the feasting units can probably be attributed to the nearly complete scapula in 2250 and a large piece of antler in 3036/3037. It is noteworthy that equids do not seem to be involved in feasting in the units so far recorded. Observations in the field suggest that they may well have played a role in some feasts in the Mellaart area.
As noted above, we are starting to perceive different modes of consumption at Çatalhöyük. In the future, we hope to move beyond the crude dichotomy between daily meals and feasting to a more nuanced consideration of the varying contexts of meat consumption. Although we have only been able to allude to this very briefly in this report, we are also beginning to trace differing discard practices in the bone.
As we have noted in previous years, sheep and goat predominate in numbers, but cattle and equids would have made a substantial contribution to the diet given their larger size. These animals would all have been present in the plain, whether as wild or domesticated animals. The forest animals ó deer, boar, bear ó are relatively rare, and there is some indication that, with only occasional exceptions, only certain body parts are collected and brought back or imported through exchange to the site from woodlands further afield. This is the first time bear bones have been identified from Çatalhöyük, and their rarity suggest that the bears were not common in the Konya Plain. More likely, they inhabited mountain areas, and the paw or skin alone was brought to the site.
Lawrence, B. 1982 Principal food animals at Cayonu, in (eds) L. Braidwood and R. Braidwood Prehistoric village archaeology in south east Turkey, 175-199. Oxford: British Archaeological Reports, International Series 138.
Legge 1996 The beginning of caprine domestication in Southwest Asia, in (ed) D. Harris The Origins and Spread of Agriculture and Pastoralism in Eurasia, 238-262. London: UCL Press.
Redding, R. 1984 Theoretical Determinants of a Herder's Decisions: Modelling Variation in the Sheep/Goat Ratio. in Animals and Archaeology 3: Early Herders and their Flocks (eds) J. Clutton-Brock and C. Grigson, 223-241. Oxford: British Archaeological Reports, International Series 202.
Wasse, A. 1994 Pastoral adaptations in the central and southern Levant during the Neolithic: the sheep and goat bones from 'Ain Ghazal, Jordan, unpublished BA report, Institute of Archaeology, University College London.
Watson, J. 1979 The Estimation of the Relative Frequencies of Mammalian species: Khirokitia 1972. Journal of Archaeological Science 6, 127-137.
© Çatalhöyük Research Project and individual authors, 1998