ÇATALHÖYÜK 1994 ARCHIVE REPORT

Report on the Knapped Stone Artefacts

James Conolly

1 Introduction

This report presents the results of the analyses of the chipped stone artefacts from the 1994 Çatalhöyük field season. Included in this report are the results of further analysis of the 1993 chipped stone artefacts, presented together with the results of the 1994 data.

By way of introduction, some general characteristics of the Çatalhöyük chipped stone industry are given based on the examination of previous analyses conducted in the 1960's. The results of a brief examination of the chipped stone artefacts currently housed in the Konya Archaeological Museum are then presented. Next, the methodological approach used for the analysis of the chipped stone artefacts from the 1993/4 season is described followed by the results of the analysis of chipped stone artefacts from surface collection and top-scrape contexts are then presented. The outcome of some quantitative analysis of the data from both surface and top-scraped contexts is also discussed.

It should be borne in mind that this report is neither a definitive statement on the Çatalhöyük industry, nor necessarily even a definitive statement on the 1993/4 chipped stone assemblage. Rather, this report presents my current understanding of the chipped stone industry at a very early stage in the project. The analysis of the industry in subsequent years will necessarily adopt a modified methodological approach based on the findings of this years analysis of both the chipped stone and all other aspects of material culture to further advance the interpretive contribution of the chipped stone industry to the Çatalhöyük research project.

2 Previous Analysis

Bialor (1962), Mellaart (1964) and Todd (1981) are the three primary sources of information concerning the Çatalhöyük chipped stone industry and they provide some provisional observations regarding the technological and typological characteristics of the industry. More recently, in a review of prehistoric Anatolian chipped stone industries, Balkan-Atli (1994) provides a useful brief synopsis of the main characteristics of the chipped stone artefacts at Çatalhöyük.

Perry Bialor was the first analyst of the Çatalhöyük chipped stone artefacts. His report discusses the material derived from the first season of excavation by Mellaart in 1961 and, while thorough from a typological perspective, does not address the technological characteristics of the industry in any detail. Nevertheless, Bialor's summary of the characteristics of the industry by focusing of the morphological characteristics of tool forms, provides a useful synopsis of the forms or types encountered (Bialor 1962:69):

The industry is characterised by the presence of numerous tanged arrow- and lanceheads, not very numerous awls, and some drills (there are, of course, many bone awls also), scrapers of various kinds, some which are rather well- shaped round or ovoid scrapers, laurel-leaf daggers, the typical parallel- sided blades, a couple of heavily retouched fabricators, some heavy pointed blades, several specialised implements of problematic usage, and rather scanty waste flakes.... Equally significant is what is lacking; this includes burins..., chipped axes, adzes, picks, and hoes, microliths and geometrics in any size, barbed or notched arrowheads and sickle blades in any significant amount (only a few have been provisionally identified, although the author admits to an inability to clearly identify sickle blades of obsidian when silica sheen, so omnipresent on flint and chert, is missing).

After three seasons, it had become clear to Bialor that "the great preponderance of weapons and tools necessary for the hunt" was a defining characteristic of the industry (1964:103). His initial description of the industry, however, needed to be modified as excavations progressed. For instance, in 1963, a hoard of 'microliths' was found under a floor from Level II. Although they do not appear to be microlithic tools in the geometric sense of the term and the term microtool is perhaps more appropriate, they are small segments of blades which, according to Mellaart, do not appear to be the elements of any composite tools "and one can only imagine that they were precision tools for some craftsman" (Mellaart 1964:105). Mellaart also provides some descriptions of blade tools to supplement Bialor's report. This list of tool types made on blades is, again according to Mellaart, small and consists of retouched blade segments, some of which are possible scrapers, blades with notching, and blades retouched to form points for boring or piercing. A hoard from Level VII did contain a fair number of burins, mainly of the transverse type, as well as numerous blade tools (1964:111).

Nur Balkan-Atli (1994) has briefly summarised the chipped stone assemblage from levels XII to VIII based on the findings of the first three years of excavation at Çatalhöyük. A number of different tool types are described, including piercers, utilised, retouched, notched and truncated blades, sickle elements, scrapers, utilised and retouched flakes, bifaces, chisels and projectile points. In the early levels, projectile points and scrapers and retouched flakes are the predominant tool type. Six point types are identified by Balkan-Atli: (i) short oval points; (ii) long oval points; (iii) lozenge shaped points; (iv) roughly tanged; (v) tanged and; (vi) tanged and fined points.

Changes over Time

With regards to debitage technology, Bialor (1962:69) notes that there may well be a shift from wider blades to narrower ones. Mellaart, following the excavation of earlier levels (VIII-X) briefly discusses the characteristics of chipped stone from these levels to provide comparison with the later industry described by Bialor. Some differences between the early and late levels do emerge; for instance, flint is more common in the earlier levels, although obsidian still overwhelmingly predominates the industry. With regards to debitage, it is fairly clear that flakes dominate the earliest levels (i.e. from XII to VIII). Balkan-Atli's summery of the early phases of the industry includes a tally of flake and blade debitage. This summery clearly demonstrates the predominance of flakes in the earlier levels (XII - VIII) (table 1).

Table 1: Çatalhöyük debitage, levels XII - VIII (from Balkan-Atli 1994:127)

Bialor (1962:73) produced a similar table for levels (VIII-II) based on the debitage recovered from the first season on excavation which reveals a steady increase in the proportional amount of blades recovered through the sequence (table 2).

Table 2: Çatalhöyük debitage, levels VIII-II (from Bialor 1962:73)

These changes in the proportion of blades to flakes suggest that there may be some significant technological changes in the industry from earlier to later levels. Given this, as well as the other minor changes in typology noted above, it may be possible to provisionally date assemblages of chipped stone artefacts derived from various contexts (e.g. 10X10 sample squares or section samples) across the mound by the calculation of a simple blade index.

3 Examination of the Konya Museum Collection

Collections of chipped stone excavated in the 1960's currently housed in the Konya Archaeological Museum were briefly examined in 1993 and 1994. In 1993 the collection was inventoried and assessed for its suitability for further study while in 1994 the collection was again briefly examined, this time to investigate possible differences between shrine/non-shrine contexts (as defined by Mellaart) as well as investigate temporal differences in the chipped stone technology and typology. Time limitations prohibited the detailed recording of technological or typological attributes of the material examined. Nevertheless, a number of provisional observations regarding differences between shrine/non-shrine and early/late contexts were made.

An cursory examination of the material from four shrines between levels II and VIII reveal characteristics which distinguish them from non-shrine contexts (or at least, those structures not identified as shrines). Overall, shrines seem to contain cores, blades, and 'formal' tools (predominantly projectile points, large scrapers and large retouched flakes which could be choppers or butchery implements) rather than debitage and irregular retouched tools. For instance, shrine A.II.1 (described in Mellaart 1963:45) contained three complete obsidian prismatic blade cores, three blade core fragments and one blade core tablet, together with (approximately) fifteen long blades, five or six large heavy retouched flakes, one or two projectile point fragments and a few (six or seven) smaller retouched flakes. There was also a microlithic (or microtool) 'hoard' buried under the floor. The hunting shrine of level V (F.V.1) (described in ibid. 1966:187) contained primarily blades and blade fragments (c. 75%). The majority of the blades were extremely regular (i.e. prismatic), and about half are retouched. The flakes tend to be large, heavy and retouched. There are two (possible) core tablets and two irregular flake cores. No mention of projectiles being found is made in the text. Shrine A.VI.1 (described in ibid. 1963:50) contained one large obsidian core fragment possibly re-used as a wedge, two large heavy retouched flakes, five obsidian projectile points all with broken tips and tangs, ten regular obsidian blade fragments, one large complete blade and one burnt flint scraper together with nine other smaller retouched flakes. A shrine from level eight (E.VIII.39), however, contained mostly thinning flakes with no blades. Seven large bifaces (possibly projectile preforms) appear to be the source of the thinning flakes -- that is, the bifaces were thinned in situ. This room also contained five large retouched flakes with one or two larger pieces (cleavers?) as well as a flint scraper.

In contrast I felt that the rooms I examined did not, on the whole, contain as many retouched pieces, projectile points, cores, or 'formal' tools (large heavy retouched flakes and scrapers). For instance, two rooms from level VIII (E.VIII.10 and 29) contain predominantly unretouched irregular flakes, without any bifaces or projectile points; E.VIII.29 contained only one irregular flint scraper and low numbers of irregularly retouched flakes whereas the only retouched implement in E.VIII.10 was one large flake amongst c.150 unretouched flakes. Rooms examined in from level VI (E.VI.70, 71, 72, 74, and 75) by and large also conform to the pattern. For example, E.VI.70, 72 and 74 contained a number of items including one or two scrapers as well as a projectile point (in 72) although the majority of the items are unretouched blades and flakes. E.VI.71, however, contained seven projectile points, approximately seven core fragments as well as a dozen or so large heavy flakes. This room is described as part of a cluster of rooms (E.VI.71- 77) that were "evidently houses" for they "lacked any of the special decoration such as distinguishes the shrines from ordinary dwellings" (1966:172). The collection of lithic artefacts found in E.VI.71 just described, however, distinguishes this room from the others.

With regards to temporal changes I was restricted to investigating technological characteristics of the industry, such as examining the changes in debitage proportions over time, rather than typological changes. This is simply because the most obvious tool to examine morphological changes over time are projectile points and they are, more often than not, removed from the bags and stored separately without their context immediately apparent. It is possible that there are changes in other types of tools, such as scrapers, but this would first involve examining the parameters of the type and devising a suitable means of delineating variation (such as angle, or extent of retouch, or blank size/shape), for which there was insufficient time. Nevertheless, technology is an obliging yardstick to examine temporal changes and there is sufficient data to make some provisional observation. At Çatalhöyük the most evident change in technology has already been noted; the shift from a flake dominated industry to one in which obsidian prismatic blades play a greater role. The major jump occurs sometime between levels VII and VI (E.VII.28 contains c.80 flakes with no regular blades whereas blades comprise approximately one-half of the lithic assemblage in the houses in level VI). After level VI, blades seem to become more common. The assemblage from F.V.I (described above) contains approximately 70% blades. By level II, to judge from A.II.1, blades predominate the assemblage. It is difficult at this point to assess the accuracy of this 'blade index' for assigning dates beyond a gross early/middle/late framework to assemblages of unknown temporal context. Nonetheless, it is a start and as further analysis is conducted on the Konya Museum collections a refined means of dating assemblages, both by technological and typological indices, will no doubt emerge.

4 Methodological approach to the 1993/4 Analysis

Introduction

One can conveniently divide the manner in which chipped stone assemblages are studied into three areas (Nishiaki 1992:39): (i) typological studies, which attempt to establish morphologically defined types of tools, or interpret the results of typological classification; (ii) technological studies which deal with manufacturing methods and techniques and; (iii) functional studies, which involve the study of how stone tools were used. The methodological approach employed in the 1993 season at Çatalhöyük was an attempt to strike a balance between time and the recording of the lithic industry in as much detail as possible; when faced with thousands of artefacts and a limited amount of time, the level of recording can not be as detailed as one would ultimately wish. Recognising this from the start, I attempted to record the chipped stone industry in sufficient detail as to be able to reconstruct its basic technological and typological characteristics, identify and explore spacial and temporal variation, and pinpoint areas where more in-depth analysis would be profitable. The chipped stone artefacts from top-scraped sample were examined in more detail than the material derived from surface contexts. No functional analysis of Çatalhöyük chipped stone tools has yet been conducted, although this facet of chipped stone analysis will be introduced at some point in the future.

Technological Analysis

A technological analysis of both subsets of data (i.e. surface samples and top scrape samples) was conducted. This involved an initial separation of material by raw material in order provide a breakdown of the relative contribution of obsidian and chert to the total industry. Following this, all artefacts were then classified into debitage categories. Each debitage category is defined by a combination of technological characteristics which can then be used to reconstruct the technological characteristics of the assemblage (Nishiaki 1992:39). I have provided a brief description of the debitage categories used in the analysis (adopted from Inizan et al. 1992; Ataman 1989):

Core: Cores are blocks of raw material pieces from which flakes and blades and other debitage products have been intentionally removed.

Core Tablet: Pieces removed from the striking platform of a core in order to rejuvenate a platform (for instance, to alter the striking angle). Remnants of the striking platform as well as the scars from earlier removals originating from the striking platform will be present.

Flake: A general term used to describe the removals from a core. A single interior surface (bulbar, or ventral surface) must be present to qualify as a flake. No standardised morphology is by the term flake, although convention dictates that if a flake's length is two times its width, it is called a blade. Incomplete pieces of debitage which can be identified as being originally flakes, are termed flake fragments. Broken flakes are flakes with incomplete margins but with a complete proximal end.

Blade: A flake in which the length is twice that of the width. If broken, then the term blade fragment is used. Pieces which are broken so that the length is no longer twice that of the width, yet can be fairly confidently identified as originally being blades, are also termed blade fragments. The term blade is used to describe complete blades, proximal blade fragments those fragments with the proximal (butt) end intact, and distal blade fragments with the distal end intact. A plunging blade is a blade that has 'plunged' into the core on removal and thus exhibits the distal end (or portion thereof) the core from which it was struck.

Crested Blade: Blades which exhibit the crests that were produced to create a ridge to guide the direction of debitage so that a blade is removed. A crest is usually formed on a core so that an opening blade can be struck from a core and subsequent blade removals are then guided by the scars from the initial crested blade. Hence, crested blades are associated with the initial shaping of cores, and the early stages of blade debitage.

Chips: Very small (usually <1 cm) flakes or flake fragments, usually too small to serve as tool blanks, produced either in core reduction or tool manufacture.

Shatter: Large irregular pieces with no clear interior (bulbar) surface, or removal scars. Produced by shattering in the debitage process.

Pièces Esquillée: Pieces with crushing on one or both ends, often with small removal scars. Possibly bipolar cores for the production of small flakes, but more likely the crushing and scarring is a result of use in a manner similar to that of a wedge or chisel.

This relatively simple process of analysis is a fundamental step towards the reconstruction of the technological characteristics of an assemblage. With a sufficiently large and representative lithic assemblage to work with, and a detailed examination of debitage, the technological analyses of debitage (and tools) can aid in the reconstruction of lithic chaine operatoires (Inizan et al. 1992).

To this end, the top-scraped sample was subjected to further analysis involving the taking of width, thickness and length measurements, as well as the description of specific attributes such as butt types, scar patterns and scar counts, and descriptions of the edge, profile and cross-section to provide additional information concerning the debitage process and specific techniques of manufacture.

Typological Analysis

Following the recording of debitage attributes, the artefacts were then macroscopically examined for evidence of retouching or use. Those that exhibited characteristics consistent with use or intention retouch were, in the case of the surface sample, were measured and described by a term suited to the morphological characteristics of the blank and type and location of retouch (such as flake side scraper, retouched blade, blade with notch, etc.) similar to the descriptive categories used at previously at Çatalhöyük, Çayönü and Can Hasan III (Bialor 1962; Mellaart 1964; Redman 1982; Ataman 1989). In order to obtain more extensive information concerning irregularly retouched tools, which are by far the most common tool category at Çatalhöyük, an attribute analysis involving the classification of a retouched piece by its principle characteristics was conducted on all retouched pieces from the top- scrape sample. These characteristics were adopted from Inizan et al. and include the description of retouch localisation, postion, distribution, delineation, extent, angle and morphology on each separate blank.

5 General Characteristics of the Industry

The Çatalhöyük chipped stone industry can be characterised in general terms as consisting regular obsidian prismatic blades and flakes in varying proportions together with a small number of irregular blades. Sieved surface derived samples indicate that debris (chips and shatter) occur, on the east mound, in quantities only second to blades or, on the west mound, as the third (after flakes and blades) most frequent debitage production. Small amount of crested blades, plunging blades, core tablets, fragments are also present. Importantly, the obsidian is almost exclusively non-cortical and the small handful of pieces that exhibit cortex are always secondary, rather than primary, flakes suggesting that core preparation, or at least the 'de- cortexing' of obsidian nodules was conducted elsewhere (presumably at the raw material sources). Similarly, flint pieces rarely have any cortical surfaces, although slightly more frequently than obsidian. A brief foray into the alluvium around Çatalhöyük confirmed suspicions that no raw material suitable for knapping is in the immediate vicinity. There are some gravel beds to the east of the site which may have nodules of flint, but this has not been fully explored. In any event, the extreme predominance of non-cortical flint flakes would suggest that the flint sources are far enough away to make the removal of cortex an economizing practice prior to exporting flint to Çatalhöyük.

A small number of cores have been found over the last two seasons, and they vary from small, amorphous, exhausted flake cores, to regular single platform prismatic blade cores. Obsidian cores are primarily obsidian single platform blade cores with platform angles between approximately 70 and 90 degrees with a facetted striking platform and uni-directional scars. These are undoubtedly the source of the regular obsidian blades which appear in considerable numbers in the later sequences. There is one irregular obsidian flake core that does not have any prepared platforms and may have been used as a wedge. A lesser number of flint cores were found, all amorphous flake cores with no prepared platforms. No opposed platform cores of the Asiklihoyuk variety were found, nor cores similar to the Can Hasan III blade cores, nor flint blade cores similar to those at Hacilar, although obsidian blade cores from Çatalhöyük resemble obsidian blade cores from Hacilar levels IX-I (Mellaart 1970:448-449).

The fairly comprehensive attribute analysis undertaken on all retouched pieces found in the 1993 and 1994 field seasons described above will, when complete, provide an extensive descriptive account of the nature of the retouched pieces at Çatalhöyük. Until such time that this is completed, a more general descriptive analysis of the tools, in a manner similar to Can Hasan III, will suffice. The general descriptions of the industry as given by Bialor, Mellaart and Balkan-Atli have not been altered by the re-analysis of the museum objects, nor by the acquisition of new data.

With regards to tools, well made projectile points of varying shape and size and well formed oval flake scrapers are frequently encountered. However, the vast majority of the tools are irregular blade and flake based tools common to PPN and PN industries elsewhere in Anatolia and the Levant such as borers, scrapers, denticulates, and notches. Retouched blades (either on one or both lateral margins and either inverse, direct, bifacial or alternate) are numerous. Other blade based tools such as burins, tangs are less common, but make an appearance.

The most distinctive element in the Çatalhöyük chipped stone industry, certainly the one aspect which in the past has received the most attention, is the high numbers of extremely well made projectile points and/or daggers. The past two seasons have seen the recovery of a number of point fragments from surface collection or top-scrape contexts. Most are fragments and are impossible to describe beyond, for instance, whether the point was unifacially or bifacially worked. In 1993 there were five projectiles or large enough fragments to identify whether or not the point was tanged, notched, etc. Of these five, three (one flint, two obsidian) appear to have been tanged, unshouldered and bifacially worked. There was also one very large (157.0mm x 25.3mm x 13.1mm) complete bifacial double-pointed obsidian projectile point found in 1993. This year, a further fourteen fragments and two near-complete projectiles were recovered from top-scrape samples together with one fragment from the surface collection. For those projectiles from which key morphological characteristics could be determined, nine of the points were tanged or probably tanged with the remainder were too incomplete to determine. Three were shouldered and three were not shouldered, with the remainder to fragmentary to classify. Fifteen of the seventeen fragments appear to be bifacially worked, the remaining two have covering or near-covering retouch on one surface and short retouch on the opposite. Typological groupings for Çatalhöyük projectiles have been devised both by Balkan-Atli (1994) for levels XII-VII and Bialor (1962:70), who discusses the major source of variability in the 1960's material (principally based on the presence or absence of tangs, shoulders, basal rounding, and variation in the location of retouch). All the point recovered over the previous two season fall within the variation described in these typologies.

One other notable tool class which deserves mention are the large flake based tools with varying degrees of retouch which appear in several of the room contexts excavated by Mellaart, and four of which we found in a cache in the 1993 season. These seem to be very large cutting or butchery tools rather than scraping tools by virtue of the low angles partial retouch that provides for a sturdy and sharp cutting, as opposed to scraping, edge. The four that were found in 1993 all have either facetted butts and/or evidence of platform preparation and are fairly regular in shape suggesting some effort went into removing a large, thin, long, regular blank. A microwear analysis would certainly confirm the function of this fairly common tool.

6 Surface Sample

2,565 chipped stone artefacts of were recovered from 2x2 m surface collection squares across the east and west mound. All soil from surface collection units was sieved through a 5 mm mesh according good recovery of smaller pieces of debitage. Following the methodology defined above, all artefacts from surface collection units were initially classified by raw material, classified by debitage category, and then separated into pieces which exhibited use damage and/or intentional modification (i.e. retouch) and those which did not. For the sake of simplicity these two categories were differentiated by the terms 'tools' and 'debitage'. Tools were described by a simple common descriptive term (e.g. flake scraper, retouched blade, piercer) of the sort widely employed by other lithic analysts from sites in the Turkey (e.g. Can Hasan III, Hacilar, Asiklihöyük). As well as providing a means of defining the technological and typological characteristics of the industry, I hoped that this level of analysis, although less exhaustive than that given to the top- scrape derived material, would nevertheless reveal some manner of patterning in the spacial distribution of debitage and tool types. For instance, I thought the examination of the distribution of those pieces of debitage that are the by-products of reduction sequences (e.g. exhausted cores, chips and shatter), may aid in identifying manufacturing areas. An issue such as this becomes significant if one considers that the excavations in the 1960's apparently failed to locate any areas which appeared to relate to a lithic manufacturing area. Indeed, Ian Todd (1981:81) suggested that a systematic surface collection may be useful in locating such an area. On a more general level, the identification of surface variability in the type and frequency of lithic debitage and tool types may aid in the identification of areas which were occupied at different periods.

Raw Material

Of the 2,565 pieces of chipped stone from surface collection contexts, 2,411 were obsidian (93.7%), and 154 (6.0%) were fine grained crypo-crystalline silicious rock (chert or flint, although the term flint will be used). Eight other pieces which looked as if they had been struck in a manner consistent with other chipped stone could not be confidently identified, although it three were almost certainly basalt and five appear to be quartzite. One large flake of completely clear, white, glass-like, obsidian was found on the west mound which does not correspond to any descriptions of known obsidian raw material in Anatolia.

Debitage

Debitage from the two mounds is dominated by blade fragments, flake fragments and chips and shatter with the various other debitage products contributing rather less. It is clear from the surface collection that flint and obsidian were used in different ways and for different things; a breakdown of raw material by debitage category demonstrates the different uses of the two materials (graphs 1, 2, 3 and 4). The coding for the various debitage categories is as follows: B = complete blade; BF = blade fragment; PBF = proximal blade fragment; DBF = distal blade fragment; IB = irregular complete blade; IBF = irregular blade fragment; IPBF = irregular proximal blade fragment; F = complete flake; FB = broken flake; FF = flake fragment; C = complete core; CF = core fragment; CT = core tablet; CB = crested blade; PB = plunging blade; CP = chip; SH = shatter; PE = pièce esquillées; ? = unclassifiable debris. Note that a '?' following a debitage category indicates that it is probably of that category.

In contrast to flint where flakes are in the majority, the manufacturing of obsidian blades appears to be the primary focus of obsidian production. Obsidian blades compose 33.9% of the total debitage from both east and west mounds, while flint blades contribute 0.03% of the total. There is a significant discrepancy between the number of obsidian blade fragments and proximal or distal blade fragments. Where sufficiently detailed recording permits comparison, this appears to be a phenomena noted elsewhere on Neolithic sites in Anatolia and Syria such as Abu Hureyra (Nishiaki 1992; Conolly 1993) and Can Hasan III (Ataman 1989). The meanings for this are unclear, although it may be that the proximal ends of blades were discarded and did not find their way into general use or circulation. Both obsidian and flint chips and shatter are common suggesting that some form of core reduction and tool manufacture was occurring on both the east and west mounds.

Tools

A variety of different tool forms were recovered from the surface collection, all of which correspond to the categories of tools described by both Mellaart and Bialor, such as various retouched and worn blades and flakes, scrapers, bifaces, projectile points, burins, drills and piercing tools. Drawn examples of tools of this type can be found following the text. By far the most frequent tools from both the east and west mounds are retouched blades, which may have retouch on one or both lateral margins either directly or inversely. Retouched flakes are the second most frequent category, followed by worn blades, which are characterised by one or more damaged edges but no clear signs of retouch. Notched blades, notched flakes, flake scrapers, piece esquill‚es, bifaces of unknown function, and a small number of piercers, burins and projectile fragments are also present. A final category, retouched piece, was used to describe pieces with clearly exhibited signs of retouch but I was unable to determine either the function or the original blank. The counts of individual categories are given in graphs 5, 6 and 7.

Reflecting the different absolute quantities of raw material, there are differences in the use of obsidian and flint for tools between the east and west mound. For instance, the proportion of obsidian to flint in retouched or visibility utilised tools differs between the east and west mound (graphs 8 and 9). On the east mound, 7% of obsidian and 18% of flint exhibits signs of use or retouch, compared to 15% the obsidian and 14% of the flint on the west mound. This suggests that obsidian was more intensively used on the west mound, whereas flint was marginally less intensively used. It may be that (although this is highly speculative) this relates to changes in raw material acquisition methods between the two different periods the mounds were occupied.

Overall, in all tool categories obsidian implements outnumber their flint counterparts. However, flint does seem to have been selected for some tools more often than others, perhaps because of its different physical characteristics (graph 10, 11 and 12). For instance, excluding those retouched pieces of unknown function, flint seems to be restricted to retouched flakes, flake scrapers, and pierces, in addition to a small number of retouched blades.

Obsidian blades are clearly favoured for tool blanks; as a group, obsidian blade tools comprise over 60% of the total number of tools from the surface collection. On the other hand, obsidian flake tools comprise approximately 20% of the total number of obsidian tools, while flint flake tools form over 60% of the total number of flint tools. It is clear that the treatment of obsidian and flint, both in terms of production and in the choice of blanks for use as tools, was different. As noted above, flint is more intensively utilised than obsidian on the east mound. This phenomena has been noted from other Central Anatolian early Neolithic sites, most notably Can Hasan III, where Ataman (1989:74) attributes the differences to either flint being conserved more than obsidian, flint being imported in a finished or near-finished form, or that the debitage process of obsidian and flint differs. I would provisionally suggest that the first and last explanations are the more likely reasons for the observed differences between the two materials. The alluvial Konya plain is not an area in which flint can be readily found. Nur Balkan-Alti (1994:37) suggests that central and south- western flint sources are located near Ankara, the south-western Taurus and possibly around Beysehir. Whether these areas contain tabular sources suitable for the production of large blade cores in unknown. Large flint blades do occur at Çatalhöyük, found both in the 1993 surface sample and from the excavations in the 1960s, and it is possible that these were imported as ready made items. The small irregular flint flakes that are common in the surface sample were likely produced on amorphous flint flake cores from flint cobbles, some of which may have been obtained from the local area. While it is difficult at this point to define the differences between flint and obsidian exploitation any further, it is perhaps sufficient to note that differences in the manner that these two materials were used do exist.

Spacial Variability

Absolute frequencies of specific debitage categories were calculated for each surface collection unit and the results plotted in a three-dimensional graph in Microsoft Excel. This allows an immediate visual assessment to be made of debitage distribution patterns (graphs 13, 14, 15, 16, 17 and 18). Tool counts were thought to be too low to derive any meaningful patterning in spacial distribution. Although the patterns in all debitage distributions are slight, two notable examples with may relate to spacial or temporal patterning are the distribution of production debris (chips and shatter) which shows a slight tendency to cluster in the north-west area of the east mound, and the distribution of blade fragments occur more frequently towards to eastern side of the mound. It is as of yet unclear whether the distribution of chips and shatter relates to post-depositional action (such as run-off) or is a reflection of production, discard, or other activity. It has been argued (Behm 1985) that vegetation considerably prohibits the movement of small chipped stone artefacts, even in areas of severe seasonal precipitation. If the dense vegetational cover of Çatalhöyük was the same in antiquity as it is currently, then the downslope movement of chipping debris was likely minimal and the concentrations can be interpreted as cultural in origin. The (slightly) higher frequency of obsidian blades in the east, given the increased use of blades in the later levels of the mound, may indicate that this area was occupied later than other areas of the east mound.

7 Top-Scrape Sample

Debitage

2,596 chipped stone artefacts of obsidian, flint and small quantities of quartz, basalt and other crypo-crystalline rock have been analysed from top- scraped units. Drawings of a sample of debitage products can be found following the text.

Following the technological analysis outlined above, all chipped stone artefacts from top-scraped units were sorted into debitage categories, regardless of whether they exhibited signs of retouch. A breakdown of this analysis (graph 19 -- coding as for surface debitage) clearly demonstrates the preponderance of blade and flake fragments, although blade fragments are the more common debitage product recovered (n=737 vs. n=614). Small overall proportions of chips in the assemblage are more likely a function of the recovery technique (visual pick-up while digging, with no sieving). The surface sample, as a sieved sample, consists predominantly of chips and had the top scraped sample also been sieved I would expect far higher numbers of this category.

Cores, core fragments and core by-products (plunging blades, core tablets and crested blades) are rare but present, suggesting that some knapping activity was conducted at Çatalhöyük. Similarly unretouched shatter, a by-product of knapping, also suggests that some production was occurring on site, although there is no evidence thus far which would indicate a 'workshop' such as the famous obsidian emporia at Phylakopi (Torrence 19186) or the knapping area at Jarmo (Braidwood 19XX:XX).

There are few differences between the west and east mounds in terms of debitage counts (graphs 20 and 21), suggesting that production techniques and organisation are broadly similar between the two. In so far as debitage counts may identify large scale shifts in either the organisation of production or large technical changes, such as a shift to flake from blade based production, no major breaks in tradition are apparent between the east and west mound.

There are, however, some differences in the distribution of debitage categories between the various top-scrape contexts which may relate to temporal and/or spacial variation across the surface of the east mound. A breakdown of debitage category by 10X10 unit (graphs 22 to 44) reveals differences, particularly in the absolute count and/or proportion of blades, in the individual assemblages. Overall blade fragments contribute 28.3% of the total debitage from the east and west mounds and, again, almost 30% of the east mound debitage. Across the surface of the east mound, however, the contribution of blade fragments to the total debitage from individual 10X10 samples can vary enormously. A simple index that measures the relative contribution of blades to a 10X10 sample was calculated for each 10X10 unit by dividing the sum of blade fragments and proximal blade fragments by the sum of the blade fragments, proximal blade fragments and flakes [i.e. (BF+PBF)/(BF+PBF+F+FB+FF)]. A high blade index, therefore, indicates a higher proportion of blades to flakes. Map 1 displays the differences in the proportion of blade fragments to flakes by 10X10 sample across the mound. Notably, the block of sixteen 10X10 squares (1020,1170) on the northern part of the main mound has, on average, a lower blade index than the southern part of the main mound. Similar indexes can be calculated for levels XII to II using the data compiled by both Balkan-Atli and Bialor. These indexes show a clear trend in the increased use of blades in later contexts. Consequently, the differences between the northern and southern areas of the east mound may related to temporal differences in occupation. Apart from squares 940,1040 and 1040,1040, however, the indexes calculated for all other 10X10 sample assemblages are lower than 60% which, if the level indexes are believed, would mean that most of the mound was not occupied after sometime between levels VI and V. This seems unlikely and I am going to assume that my index values are somewhat lower than those calculated by Balkan-Atli and Bialor (using the debitage collected by Mellaart) because my samples have more flakes and flake fragments owing to a more thorough collection of debitage. There is still a wide difference between the northern and southern areas of the mound, however, and I would provisionally suggest that this does have to do with temporal differences in occupation; the northern part of the mound seems to have been abandoned earlier in the Neolithic, and later occupation became focused on the southern part of the main mound.

A correspondence analysis was performed on 24 of the 10X10 top-scraped squares using debitage category counts in an attempt to define and delineate relationships between the separate squares and debitage categories. As a correspondence analysis allows one to examine relationships both between objects and contexts and within the object and context groups, it was hoped that this would provide some insight into what debitage categories naturally group together on the basis of frequencies and associations within the separate 10X10's, and what 10X10's group together on the basis of similar frequencies of individual categories of debitage. The analysis was performed using IASTATS, an in-house archaeological statistical program written by faculty members of the University of London Institute of Archaeology. The results of this analysis are discussed in Appendix Two.

Obsidian Blades

Obsidian blades are, on the whole, very regular in form. Over 60% had a trapezoid cross-section and over 90% had, or likely had, unidirectional scar patterns (graphs 45 and 46). These blades are commonly referred to a prismatic blades and their regularity of form suggests that they were removed from their cores using a pressure, rather than an indirect punch, technique. Obsidian blade fragments are marginally more frequent on the east mound (34.8%) than on the west (32.9%). Obsidian blade fragments from the east mound are also marginally smaller and more standardised than on the west mound. A width by thickness index was calculated for each obsidian blade fragment and the means and standard deviations calculated for all blade fragments from the east and west mound (graphs 47 and 48). The smaller the value of this index, the thicker the blade is in relation to its width and wide discrepancies between assemblages in this index may indicate different techniques or methods used to manufacture blades. There are very small discrepancies between the east and west mound, both in the width/thickness index and absolute measurements, suggests that there are few differences in terms of obsidian blade fragment size and, by implication, few differences in blade production techniques between the east and west mound.

Often, lithic analysts make a distinction between blades and bladelets. Implicit in the distinction of the two size differences in describing blades is that there are two modal types of production occurring -- one which produces larger blades, and one which produces smaller blades. If this can not be demonstrated, the term bladelet is only useful as descriptive term which refers to blades which fall below an arbitrarily set limit. In this case, it is arguably better to use descriptive statistics to describe the range of variation in blade size. One manner of doing this is to use a frequency distribution of blade widths or a distribution of a width by thickness index. Range and frequency of size variation can thus be displayed, and if a bimodal frequency curve is the result, then a more argument for the presence of two types of production, as well as the definition of their metric characteristics, can be made. Frequency distributions of obsidian blade widths and thicknesses from the east and west mounds and the two mounds combined are in given in graphs 49 to 54. The distributions of all six graphs are not bi- modal and, thus, no separate bladelet production or technique seems to have existed. All frequency distributions, however, are slightly positively skewed. This could be attributed to the fact that a larger core produces a certain amount of thick blades and, as the core decreases in size, blades of smaller thickness and width are removed. When the blades become between 8 and 10 mm wide and between 2 and 3 mm thick, the core is close to exhaustion and is soon abandoned.

When present, a sample of the remnant striking platforms of blades was examined in order to examine technique of removal from the core and the nature of the core striking platform. A majority of the blades had ground or facetted dorsal surfaces immediately below the butt in order to remove the lip and a majority of a sample of proximal blade fragments had punctiform butts (53.6%) or linear butts (24.1%) suggesting a pressure technique for blade removal. This corresponds to the cores which have been recovered which also suggest that debitage was removed by a form of pressure, in which the blade is not so much struck from the core, but pulled off by the pressure exerted by a crutch of some sort.

Retouched Pieces

Of the 2,596 pieces of chipped stone analysed from the top-scraped squares, 1,055 exhibited signs of use or retouch. These tools are consistent with the types described by both Bialor and Mellaart in the various preliminary reports from Çatalhöyük, as well as the tools derived from surface collection contexts. A variety of blade tools, flake tools, bifaces, and irregularly retouched pieces were found, in addition to a small number of projectile point fragments and what appears to be the hilt of a pressure-flaked fine grained yellow flint dagger. Illustrations of a sample of these tools can be found at the end of the text.

Tools on blades consist of denticulates, notches, blades with regularly and irregularly retouched edges, worn blades and blades with tangs as well as the projectile points. The worn blades category describes those pieces which do not exhibit signs of retouch, yet one or both edges exhibit considerable damage. These pieces could have functioned as sickle elements, although the absence of gloss on obsidian blades prohibits positive identification. Flake tools consist of pieces with concave, convex, and rectilinear edges, with working edge angles varying from low to abrupt, notched flakes, flakes with tangs, and irregularly retouched edges. Flakes with semi-abrupt to abrupt retouch are common, suggesting use as scrapers. Biface fragments, some of which were likely projectile points but are too fragmentary to identify were also found. A small number of well formed drills, also noted by Bialor were also found. Burins are rare but do make an appearance, albeit infrequent. PiŠce esquill‚es are common and vary from well formed nearly square examples with crushing and scar removals on two opposed ends, to more irregular pieces although the crushing and scarring on opposed ends still occurs. The square examples are occasionally so regular that they resemble gun-flints. A number of irregular pieces of shatter had also had been retouched, more often than not irregularly, as had some core fragments. Graph 55 provides a breakdown of tool types from the total 10x10 sample from the past two seasons. An attribute analysis of all retouched tools is currently being studied, the results of which will further define and delineate the nature of the retouched tools.

8 Conclusions and General Comments

Three general goals were pursued during the 1993/4 season with regards to the chipped stone artefacts. The first and most important of these was to begin to acquire an understanding of the technological and typological characteristics of the Çatalhöyük chipped stone industry as a whole. The second goal was to examine the chipped stone artefacts from surface collected contexts in order to contribute to the identification of spacial variation across the surface of the mound. The third goal was to examine the artefacts from the top scraped contexts in order to contribute to the identification of possible temporal variability in the occupation of different areas of the mound.

With regards to the first objective, a basic understanding of the technological and typological characteristics of the industry has been reached, although the information gained from the 1993 and 1994 analysis will be used to adjust the methodological approach so that a more comprehensive understanding can be achieved in the 1995 and future seasons. Surface variation across the mound is as of yet inconclusive with regards to identifying discrete areas of production or deposition, although higher concentrations of obsidian production debris were noted in specific areas off the main mound surface. In addition to this, the collections in the Konya Museum were examined and some provisional observations concerning temporal changes and differences between shrine/non-shrine contexts. For instance, initial forays into examining the differences between shrine and non-shrine contexts suggests that there may be differences in the nature of tool forms, debitage and importantly, knapping activities and behaviours.

As such, some significant steps have been made these last two seasons. Nonetheless, my understanding of the industry is still at a basic level, and I await the examination of further samples to better define and explore both technological and typological issues that have only been rudimentarily raised here. Further work, some of which is ongoing, involving the further definition of the nature of the irregularly retouched tools, a detailed examination of the cores and fragments to examine specific techniques of debitage and, when additional data has been acquired, the reconstruction of typological sequences of tools -- both through time and possibly across space -- and, eventually operational sequences of reduction, will add to the interpretive possibilities of the knapped stone artefacts. What has been presented here is a brief report on some of the results of the analyses conducted thus far. It is by no means a exhaustive report on all facets of the industry but it is, I hope, a start.

List of Figures

© Çatalhöyük Research Project and individual authors, 1994