ÇATALHÖYÜK 1997 ARCHIVE REPORT

Çatalhöyük 1997 Lithic Report

James Conolly

The first section of this report describes the patterns and characteristics of the lithic artefacts recovered during the course of the 1997 season. Following this, there is a general comparison between the flint and obsidian artefacts from Çatalhöyük and other Neolithic sites in Central Anatolia so that common characteristics in lithic technology and typology may be identified over a wider region.

Characteristics of the 1997 material.

In marked contrast to previous seasons, 1997 produced a large volume of knapped stone and a sampling strategy was adopted, partially lead by the 'priority unit' sampling system. As such, this report is not a comprehensive account of all the material recovered in 1997, but a description of the material sampled. Where justified, patterns and correlations between various lithic attributes and contexts discovered in the sample have been extrapolated to the level of general patterns that may apply on a site-wide level.

1172 flint and obsidian artefacts were recorded from dry-sieve samples, plus a further 716 from wet-sieved or floated samples of 4mm fraction. This report only considers the former, as they were more comprehensively described and recorded in the field. The weight for all the dry sieved material is just over 3240 grams, inflated by the recovery of a large hoard with many large and heavy flakes of obsidian (described below). As in previous years, the vast majority of this material was obsidian, with only a few flint artefacts found (n=18 or 1.5% by count, g=80.17 or 6.8% by weight).

The range of debitage and tools recovered this season was comparable to previous years. Blades and flakes, plus several cores, shatter, and small chips were collected, as was a range of retouched debitage. This season's assemblage of tools once again shows the enormous variety of forms retouched obsidian and flint may occur. A variety of small irregularly retouched flakes, large well-shaped flakes, blades with irregular retouched to blades with fine distal shaping, pièces esquillées, a typologically diverse array of projectiles on both blades and flakes, plus a remarkable flint dagger with a bone hilt were amassed from the excavations in the Mellaart, Bach, and North areas. Table 1 and Table 2 outline the range of debitage and tools collected in the 1997 season. Further details concerning the particular characteristics on the various categories of debitage are given in the following two sections, but as the material from this year is not significantly different from previous years, only notably different or unusual elements will be referred to in detail.

Cores and core preparation products

A total of eight cores and core fragments were recorded, all obsidian, three of which were small bipolar pièce esquillée-like cores with significant battering, three multi-platform/multi-sequence flake, and the remaining two blade cores. One of the latter appears to have been a core for the production of non-prismatic blades, the other a fragment of a prismatic blade core. Table 3 presents the major characteristics of these artefacts.

In addition, three small crested-blade fragments were analysed, all with lengths of just over 20mm and a thickness under 10mm. No core tablets were detected in the samples examined this season.

Debitage

As already noted, a wide range of debitage classes were recovered, consistent with what has been recovered in previous years. The vast majority of these tend to be small unretouched flakes of obsidian and flint in varying degrees of completeness (i.e. they range from complete to fragmentary examples) with platforms varying between small plain to larger facetted examples. Closer examination of flake techno-morphological attributes show that a proportion of flakes can be assigned to one of a variety of known, or inferred, reduction operations.

For instance, examination of non-metric variables allows inferences to be made concerning productive activities. Ventral profiles and edge shape of obsidian flakes show that straight edged and irregular edges are both independently and concurrently the most frequent characteristics, whereas strongly concave profiles and irregular edges are independently the most common, with straight edged and irregular to sub-parallel the most usual joint occurrence. The significance of this becomes more apparent when profiles are considered by platform angle, where strongly concave profiles are most closely associated with low angled platforms. At the same time, analysis of small-sized debitage suggest that low platform angles (less than 45 degrees), convex profiles, and expanding edges are the characteristics encountered preferentially on 'thinning' flakes - those flakes produced during the so-called process of 'thinning-out' bifaces. Given the high frequency of 'bifaces' in the tool assemblage such flakes can be expected in the assemblage.

The 1997 season also dramatically increased the number of large obsidian flakes found within caches. One cache of 50 large, heavy, flakes found in Space 113 are at the same time contextually circumscribed and technologically distinct. The total weight of all obsidian in the cache was 2309.07 grams (which included 16 small flakes and chips of <1 gram each), whereas the mean weight of the main body of flakes was 47.02 grams. The mean length of the cached flakes is 77mm, width 54mm, and thickness 14mm. Four of the cache also had evidence of residual cortical surfaces. Most were missing their proximal ends, so details concerning their method of reduction were difficult to deduce. However, the 10 pieces that do retain evidence of detachment show very high platform angles, flat to facetted striking platform remnants, and a mix of diffuse to very prominent bulbs of percussion and in sum suggest a direct hard-hammer approach to removal. The range of morphology of the flakes - from thin and elongated to thick and fat examples - are suggestive of a relatively non-structured and opportunistic removal strategy where the objective was to obtain only a relatively large flake that could be further used for whatever its particular shape was suited for. Indeed, the type of retouch on the flakes shows a lack of desire to significantly change edge morphology through retouch, but to instead use the pre-existing profile as the basis for a working edge.

As with several of the other caches of large obsidian artefacts (such as the flakes in 1993, and blades in 1996, as well as Mellaart's numerous examples) there is no evidence for any cores suitable for manufacturing flakes of this kind within the known assemblage. It would seem increasingly probable that this type of deposit, given their technological idiosyncrasies, are pre-formed imports from the obsidian sources in Central Anatolia brought to Çatalhöyük and, without significant modifications to their original state, cached.

Unlike previous years, few blades were recorded in the samples analysed during the 1997 season. Nothing can be added to that which is already understood about the form of blade production at Çatalhöyük, except to reiterate that blade use is unknown in the earliest levels of the settlement, gradually increasing in importance, until it dominates the later phases of occupation.

Retouched Debitage.

The 251 retouched flakes and blades from this season's work are, for the most part, very similar to those recovered in previous years insofar as they show a wide range of different retouch delineations. As in previous years, a sample of retouched debitage was examined along the lines of the Wembach Module, but only a small summary is provide here as analysis is ongoing. Table 5 shows the general tool classes of the retouched debitage. It cannot be considered representative of the full range of tools encountered.

Of more interest are the cached flakes from Space 113, where 22 show evidence of retouch to a greater or less degree. Table 6 outlines this in more detail, where the variability of their edge modification can be seen.

In sum, these pieces show a variety of different uses, where the only common denominator would appear to be the presence of relatively irregular retouch. This, together with the fact that less than 50% of the cache actually shows modification, emphasises the disparate nature of this group of objects. It is interesting to compare this to the Building 1 cache discovered in 1996, where the objects are highly standardised. Nevertheless, all the known caches can be linked by the shared characteristic of their technological otherness, suggesting that they represent deposits of material that have been brought to Çatalhöyük in a pre-formed condition.

The other retouched object of note is the large near-complete (save for the tip) hilted flint dagger found in Unit 2210 from the Bach excavation area. Plano-convex is cross-section, this implement shows a remarkable skill of production and intimate knowledge of the properties and tolerances of the raw material, which is probably tabular flint, but of unknown origin. Retouch is parallel-oblique and, with the exception of the tang, is restricted solely to the dorsal face, and the edges of the dagger are finely serrated. It is now in three pieces (one 2210.X7 and two 2210.X8), together with a bone handle in the shape of a boar's head (2210.X9). The total length of the extant blank is 176mm, with a width and thickness of 35.9mm and 8.1mm respectively, with a tang of approximately 38mm length. Macroscopic striations on the reverse side suggest that grinding has increased the thinness of the piece.

Contextual Patterns.

Material was examined from 92 separate units in 1997. At the time of writing, only 74 of these units have been recorded in the Unit Sheet database. Table 7 shows the quantity, weight and (by dividing by the number of litres of sieved soil) the density of lithic material recovered from each unit.

Perhaps of greater interest is the examination of lithic patterning by Space, where differences in the quantities of material are also evident (table 8). Of interest here is the higher density seen in Space 113, influenced by the cache of large heavy flakes. Note also that Space 115, 116 and 117, although considerably different from each other in terms of quantities of lithics recovered, show far less division when densities are calculated on the basis of quantities of sieved earth.

Finally, because of the addition to this year's units recording sheets, some exploration of lithic patterning along unit interpretative categories can be conducted. For instance, table 9 examines the density of lithic material (weight of lithics by litres of sieved soil) across these categories. Of interest here are the higher densities in room backfill, pit fill, hearth, dump layers, and cleaning, which contrast with the lower densities seen in walls, floor packing, grave pits and fills, and (perhaps surprisingly) dumps.

To explore this patterning further, the different interpretative categories were divided into their respective Spaces, and densities of lithic material examined along these grounds (table 10).

Some of the patterns identified in the table 9 are now seen in more detail, particularly the high densities of material in the Space 113 levelling deposit, pit fill of Space 117 and the building fill or dump of Space 109 (note that densities in bold indicate duplicate calculations, because of the system of competing possible interpretations for any single unit). Unfortunately these patterns cannot be explored statistically because of the low numbers, but it nevertheless demonstrates the usefulness of the interpretative category for exploring artefactual patterning.