ÇATALHÖYÜK 1994 ARCHIVE REPORT

Pinarbasi excavations 1994: first report.

Trevor Watkins, Douglas Baird, Catriona Gibson and Mick Rawlings

Introduction

Location and description of the site

The site of Pinarbasi is situated at the tip of a line of limestone hills at the northwest edge of the Kara Dag massif. It is in the lands of Süleymanhaci Village, close to the northwestern boundary of Karaman province. On the Turkish 1:25,000 map series the site is on sheet N30-al at 015E 498N.

The first part of the site consists of a series of four or five rock--shelters at the foot of cliffs in an embayment which overlooks a small permanent lake and a much larger area of seasonal shallow lake, swamp and reed-beds. Several of the rock--shelters have produced surface obsidian pieces which could be dated to either the epipalaeolithic or early neolithic periods. One of the shelters has been shaped in the Byzantine period, and some of its prehistoric deposits have been pushed out of the cave. This rock-shelter has a short rock-inscription. The rock-shelters are set around a semi-circular area which slopes considerably from south-east to north-west. Across this slope there are the remains of two terrace walls. The surface of the slope is scattered with sherds predominantly of Byzantine date.

Around the corner to the south of this embayment there is a fifth probable rock--shelter, below which is a series of exposed bed-rock mortars that is, mortars which have been worked into the surface of natural limestone.

Below the embayment and the fifth rock-shelter there is a small peninsula which ends in a heap of large limestone boulders, and which projects into the lake (on the 1;25,000 map this cone of rocks is labelled Kötü Tepe - bad mound). The lake is fed by at least two springs below its surface (which give the locale its name). The spring to the west of the peninsula is the larger, but there is a second, smaller flown approximately north-east of the tip of the peninsula. The surface of the peninsula takes the form of a low ridge, growing higher as it approaches the terminal, conical heap of boulders. North of the boulders there is an expanse of exposed limestone.

The surface of the peninsula is scattered with Early Bronze Age pottery and a few, very small obsidian pieces; the surface indications were that it had been occupied in the epi-palaeolithic period, again in the Early Bronze Age. In addition, there were clear indications in the form of crudely looted tombs, of a further phase of use as a cemetery in the Roman period. Above the peninsula and several of the rock-shelters, on the top of a rocky knoll there are traces of rectilinear architecture, which may be Roman in date. Tins area has also been noted as producing some surface obsidian pieces.

Circumstances of excavation

The site bad been known as a single rock-shelter for some years, having been visited first by Dr David French. It was visited by Drs Trevor Watkins and Douglas Baird in September 1993 as part of the reconnaissance season for the Çatalhöyük Regional Survey. At that time the additional rock-shelters were noted, and the probable epi-palaeolithic open settlement below the Early Bronze Age occupation on the peninsula. It was also clear on that visit that the site was being damaged; a pit had been dug against the rock wall of the already known rock-shelter, and several. large pits had been dug and tombs looted on the peninsula. It was in the spoil from one of these tomb-looting pits, and in the exposed section. that the probable epi-palaeolithic strata were recognised.

The potential of such a site for Turkish and West -Asian prehistoric archaeology and the recent damage that it had suffered made it seem very important to bring to the special attention of the of the Directorate-General Museums and Antiquities. The report submitted suggested that the site needed more detailed assessment before a decision could be made, and the Directorate-General offered to consider an excavation permit application. The proposal made in the application was to spend the firs season on assessment of the site's archaeological potential. and then to take a further three seasons to conduct larger scale excavations, if the potential of the site warranted it. The Directorate-General recommended that the rescue excavation be carried out as a collaboration between the Karaman Museum, within whose region the site lies, and the University of Edinburgh.

Objectives of First season

The objectives of the 1994 season, therefore, were essentially concerned with obtaining information that would help in assessing the research potential of the site. There were constraints of time and human resources, since the team was also due to carry out survey work as part of the same season. As a minimum we needed to work on the rock-shelter that was being damaged and on the open site on the peninsula, implying two areas of excavation. If a third area of work could be entertained we would excavate a second area within the peninsula site in order to learn something of its extent. From the trenches we opened we wanted.

Given that we wanted to maintain a closely quantified sampling system, and that we were likely to be working with a small-scale bone and lithics assemblage, it was clear that the amounts of deposits removed would be small and that trenches would be correspondingly small in area. We did not aim to uncover areas of such a size as to reveal coherent traces of structures.

Method of excavation

All digging was done with hand-(trowels or small mattocks) by members of the tearn; no workmen were employed. As much material as could be handled was processed through the flotation and wet-sieving machine. Typically, whole earth samples were of 60 or 120 litres. Much of the rest of the material was dry-sieved (with or without quantification), using either a 3 or a 5 mm mesh.

Material floated off the whole earth samples was caught in two sieves, one of 1 mm the other of 0. 3 mm mesh. The wet-sieving mesh was of 1 mm. After drying, the heavy residues were sieved into three fractions, >5 mm, >3 mm and > 1 mm. The heavy residues were sorted at the excavation headquarters by members of the team. All the >5 mm residues was sorted, and 20% of the > 3 mm was sampled. In some samples 5 % of the > 1 mm residue was sorted, but a good deal of the finer residues were left unsorted. All the unsorted residues were stored until a decision on whether they are needed em be made in the light of the study of the first season's materials. The floated material has not been examined at all in the field. All of this material is to be examined during the winter in England. Most of the animal bone was given a preliminary sort in the field, and the diagnostics from a selected set of contexts will be studied during the winter. The lithic assemblages were mostly examined and analysed in the field. A small selection of typical pieces is to be drawn over the winter in England.

Acknowledgements

We are extremely grateful to the Director-General for Museums and Antiquities, Professor Dr Engin Özgen, for the permit to excavate at Pinarbasi, and to Cengiz Topal of the Karaman Museum for his collaboration with us. Without his assistance it is doubtful whether the work could have been started so successfully. We owe a deep debt of gratitude also to the funders of the work., the British Institute of Archaeology at Ankara, the British Academy, and various funds in the University of Edinburgh. The constant assistance, advice and support of the staff of the British Institute in Ankara was invaluable, and we are happy to have this opportunity to express our thanks to Dr David Shankland and his staff.

It was a good indication of the multidisciplinary of the Çatalhöyük Research Project that Dr Neil Roberts and palaeoenvironmental team made a high priority of working alongside us during the excavation. It is certainly of great advantage to the Pinarbasi excavations that there will be a long core from the palaeo-lake bed, and it should be of equal advantage to the interpretation of the material from the core that the excavated sites are so close and look likely to cover a long and very interesting period from the climatological and environmental point of view.

Finally, as director of the first season of work- at Pinarbasi, the author wishes to acknowledge his admiration and appreciation of the team who worked so hard and so well together to make the season a success:- Dr Douglas Baird (co-director), Dr Nicola Murray (archaeo-zoologist), Mark Nesbitt (archaeo-botanist), Nick Rawlings and Catriona Gibson (trench supervisors), Müge Sevketloglu, Marcia Taylor, Christopher Stevens and Seona Anderson.

Summary of results

For a number of readers it may be helpful to m=arise the results of the first season of excavation in a few words at this point.

A considerable depth of archaeological deposit was found to exist both in the rock-shelter (called Area B,) and in the open area of settlement on the peninsula (called Area A). In neither area was the bottom of the archaeological stratigraphy reached.

In the rock-shelter prehistoric archaeological material was found immediately below the present surface. The complex features were extremely coherent and consisted of a bowl-shaped hearth, pits, stone settings, layers of living debris and part of a large, stone-built, curvilinear structure. The cultural assemblage associated with these strata seems to be Neolithic in date, being either contemporary with the later levels at Çatalhöyük or somewhat earlier than them (i.e. dating to around 6,000 BC or earlier in uncalibrated radiocarbon terms). There are indications in the lithic assemblage that there is in all probability an earlier neolithic and an epi-palaeolithic occupation further down in the stratigraphy. The quality of preservation and quantities of both animal bone and carbonised plant materials was very high.

In the open settlement area on the peninsula the uppermost deposit was a surprisingly well preserved occupation of Early Bronze Age date. which included mud-brick walls, made floors, and burials cut down into the earlier deposits. Below there was a mixed deposit, the result of several millennia of exposure of the top of the lower deposit to natural processes, especially burrowing animals, and of the digging of features in the Early Bronze Age settlement. Below that again were undisturbed, loamy deposits of unknown depth (but certainly more than half a metre), dating to the epi-palaeolithic period.

The lithic assemblage from Area A represents an industry on a very small scale, made mostly in obsidian but also in flint and chert and including an impressive number of very recognisable microlithic types. The animal bone assemblage was becoming well preserved and quite copious in the deepest levels reached, but carbonised plant remains were still fairly meagre.

The excavations into the deeper levels were confined to a narrow sounding of only three square metres. Nevertheless, in that small space there occurred the crouched inhumation of a juvenile and the tantalising trace of a wall made of white, clay-like marl material.

Area A: the open settlement area

Catriona Gibson

Introduction

It was decided to open a small sounding close to the E end of this part of the site, adjacent to the track and very close to one of the looting pits, in whose section two main members of stratigraphy had been observed in 199-3. The trench in this part of the site was arbitrarily named Area A.

Area A was opened on the 25th August 1994, and section-drawing was completed on the 11th September. The trench size was a 3 by 3m square. A western extension to the trench was made on the 5th September, 1994, with dimensions of 1.5m N-S and 1m W-E. The purpose of this extension was to expose and excavate an Early Bronze Age pithos burial that was lying in the comer of the original trench. Although the whole trench was excavated to begin with, a decision was taken on 6th September 1994 to go down in a 1x3m area only, up against the Eastern section. This sondage went down 0.92m from the beginning to the end of the season. At the end of the available time the deposit showed no signs of ending, and two possible structural features had to he left for the 1995 season.

Stratigraphic sequence

Provisionally the stratigraphy of the site has been divided up into nine phases. The lowest three phases seem to be epi-palaeolithic in date. They produced significant quantities of well-preserved animal bone and diagnostic chipped stone, both in obsidian and flint, in particular a substantial number of microlithic pieces. In addition a decorated shaft straightener was found in this phase, reminding us in its decoration of a very similar piece in the same soft, fine-grained stone and of another shaft straightener with different decoration, both found on the surface last year.

The earliest phase so far encountered consists of dark humic loamy layers, associated with three possible structures. Some groundstone tools, particularly smoothers and pounder fragments, came from this phase. Parts of a stony curvilinear feature were found, and two further possible structures formed from smallish and medium very well compacted and stones and pebbles concreted together were also noted. Another potential structural element, which may belong in either this phase or the next phase is a yellowish-white marl structure, made of at least three rectangular blocks of marl sitting in the east facing section of the 1m sondage. This feature / structure has not yet been excavated, since it ended exactly on the section line, nor is it yet clear how much deeper it goes.

The next phase consisted of a grave with a small child inhumation, as well as another small subcircular cut (possible pit) and fill running up and into the western baulk. Ale grave was sub-ovoid in plan with vertical sides and contained the remains of a crouched child inhumation, which was lying on its left side with arms folded. The body itself was aligned N-S with its feet towards the south. Its grave goods included a nicely rounded bun-shaped lump of red ochre placed just to the south of the inhumation's folded arms. Against the eastern side of the grave lay what appears to be a damaged adze re-used as a smoothing implement, a part of a sub-spherical basalt object, and a fragment of a basalt grinding stone. The bones were unfortunately in quite a friable condition, and the body had been severely disturbed, presumably by burrowing animals, but efforts were made to lift it carefully in a matrix of soil, and recover most of the fragments in. the excavation work-rooms.

The latest deposit within this phase is a dark brown sandy deposit. This layer is important because it appears to be the latest deposit witch contains no ceramic finds (with the slight exception in the northern part of the trench, where there was observed animal disturbance). it also seals the burial and assures us of its early date. The line drawn between Phases VII and VIII may in fact be defined as one of long duration. We may be dealing with an occupational hiatus lasting thousands of years and a soil horizon of up to 0.23m that may be the soil build-up between the two main periods of occupation.

This next phase, VI in the sequence, is made up of one possible building phase, a badly degraded mud brick horizon containing evidence of mud brick collapse with a few defined mud bricks as well as one possible curvilinear structure. A few sherds of pottery as well as chipped stone debitage and animal bone came from this horizon. There is in fact quite a change, and not only in soil types. The layers from this point upwards are much lighter and grittier/siltier and generally less humic and loamy than those in the earlier phases. Another major difference is that the next phases, up to the surface, contained quantities of pottery, both wheel-turned and hand.

Phase VI includes several incidences of disturbance of the lower deposits, including a pithos burial and what may be a cist burial. The probable cist only projected a short distance into the trench on its west side, and was not excavated this season. The pithos burial was set into a very tight fitting, roughly oval cut, measuring 1.15m x 0.61m. It was in fact made up of parts of three separate pithoi. Within the main, almost complete pithos lay the remains of a burial. Its feet pointed towards the eastern, base, end, while other fragments including ribs and finger bones and fragments of other sherds of pottery were oriented more towards the western rim end of the vessel. Ale bones are those of a juvenile, and were much disordered by animal disturbance. Associated with the pithos burial, and lying, immediately above it, were some fine red painted pottery sherds and a complete goblet in fine, red slipped and burnished Early Bronze Age fabric. (Our knowledge of the pottery from the site is still fairly slight, as is understanding of the Early Bronze Age chronology of the Konya plain.)

Within the next phase, V, we can assign a fragmentary plaster floor surface with an underlying pea-gritty make-up, which covered the W area of the trench. It sealed both the pithos burial and the cist, although it was very fragmentary in the N. The following phase, IV, consisted of a deposit of yellow sandy silt with flecks of plaster. This context was notable for its low density of finds, especially pottery. However, it was associated with a stone-lined pit, possibly a post-setting, with one veer large squarish stone set at its base.

Phase III consisted of a double coursed mud brick wall, composed of ovoid-shaped mud bricks roughly 0.35m in length and forming a structure surviving to a height of roughly 0.20m. Finds within Phase III consisted of animal bone, wheel-turned and hand-made pottery and obsidian debitage, comprising a good deal of derived epi-palaeolithic material from this later disturbance.

The next phase of activity was represented by a single feature, consisting of a small cut to accommodate a hand-made, medium coarse, plain-lipped, straight-sided vessel. The stone setting for the pot consisted of very tightly packed almost concreted sub-angular pebbles. Other sherds of pottery have been thrown in with this pot. It was provisionally dated to the Early Bronze Age, although it could be later, and had a rounded base with almost ledge-like handles situated very close to the rim.

Finally, immediately below the surface soil, there was a linear feature roughly 0.65m wide and running W-E across the trench. It appears as a shallow ditch with stoney fill, although it could be a very badly degraded and robbed wall. Amongst the finds were high densities of fine and coarse ware pottery, chipped stone and two bronze coins, one of which is definitely, Roman and could date this feature to the Roman period.

Area B: the rock-shelter

Mick- Rawlings

Introduction

The rock-shelter where there was damage was arbitrarily named Area B. It consists of an are of more or less level ground under an overhang of rock. A somewhat larger area has been walled off with a curved wall composed in part of very massive boulders and in part of smaller stones. This animal enclosure has a doorway formed of two upright stones. The damage consisted of a conical pit against the rock face in the middle of the arc of the sheltered area. Our sounding in this area was set close to and south of the damaged area. Past of its purpose was to recover information as close as possible to the damaged area. Its main objective was to provide a means of assessment of the stratigraphic sequence and research potential of (one of) the rock-shelters.

Excavations in Area B commenced on 27/8/94 and continued until 11/9/94. An original trench measuring c. 4.5m x 2m was laid out along the east-west axis of the main site grid. 'The eastern end of the trench abutted the rock-face within the walled enclosure, and the south-west corner of the trench was established as 341.00E/210.00N. On 7/9/94 the trench size was reduced to enable a deeper sounding to be carried out. This deeper trench was c. 3m x 1m and was aligned east-west adjacent to the northern baulk of the main trench, with a south-west corner established at 342.00E/211.00N.

A preliminary, phasing of the excavated area has resulted in the attribution of five distinct and separate phases, which are briefly summarised below from earliest to latest.

Phase: V: A series of tips and fills within an area enclosed by a curving dry-stone wall

IV: A layer of greyish bone-rich deposits

III: A 'fire installation' bonded to the rock face at the eastern end of the trench

II: Two small pits with an upper fill of medium-sized stones and a secondary fill of charcoal

I: Modern chaff/dung and preliminary clearance

Phase V, the earliest so far reached was recorded within the deeper sounding excavated after the reduction of the trench on 7/9/94. At the base of this reduced area was a curving wall made up of large blocks of limestone. At least two courses were present and at the western end of the exposed part of the wall was a single large limestone orthostat which bad been recorded within the trench area at the start of the excavation as it was projecting slightly above the pre-excavation ground surface. To the north of the orthostat was a gap in the wall, with some smaller stones at the base of the excavated area here. This may be part of an entrance but could equally be the result of stone-robbing activity or erosion. A close examination of the south-facing section at this point indicates a slight change in the colour of the deposits here. The wall shows no sign of any bonding material having been used. It seems to be a dry-stone construction and may be a retaining wall, holding up the material outside the curve and providing a defined area at a lower level within the wall.

The fill within the wall was a series of tips and dumps of material, dipping quite steeply in from the inner edge of the wall towards the centre of the area enclosed by the wall. Some of the thinner tip-lines were very rich in charcoal; other, thicker tips were more silty but still had a high ash and charcoal content. No floor was reached before the end of excavation.

The lower of the excavated deposits showed a quite distinct differentiation, visible during excavation, between the area enclosed by the curving wall and the area immediately outside the wall. Few artefacts, especially chipped stone, were recovered from within the enclosed area whereas outside the wall some quite large pieces of obsidian tools were found.

The next phase comprised a layer of grey ashy material visible throughout the whole trench except where it was removed by features of Phases II and III. This material was up to 0. 15m in depth and was excavated from most of the trench. The large quantity and splintered nature of the animal bone from this deposit, and the copious amounts of charcoal, may suggest a build-up of occupation material but there were no surfaces defined within the deposit.

Phase III consisted of a 'fire installation' or hearth recorded in the north-eastern corner of the trench. This had a base of baked/fired clay which was actually bonded to the rock- face at the east and was probably the initial lining of the feature. The whole installation was placed within a cut and consisted of a series of ashy or silty fills with at least one instance of re-lining indicated by another deposit of baked/fired clay.

The following phase comprised two shallow, sub-circular or ovoid pits, with upper fills of medium-sized stones, some of which were burnt. The easternmost of these pits cut through the upper edge of the Phase III fire installation and was 0.35m deep adjacent to the northern trench baulk. Around the south-eastern side of this pit was a ring of upright stones which formed a lining at surface level and which was butted by the upper fill of stones. Underlying this stone layer was a thin layer of charcoal which ran below the lining of upright stones. A layer of brownish silts formed the basal layer of the pit. The westernmost pit was much shallower and consisted only of the upper fill of stones and an underlying thin and intermittent band of charcoal. These two features were probably small hearths or cooking pits.

The latest identified phase within the trench area was merely the surface sweepings and preliminary cleaning of a shallow spit of deposit in order to define potential features and the nature of the deposits.

The finds

Pottery

Catriona Gibson

With the exception of some of the latest, slightly contaminated contexts in both Areas A and B, the majority of the pottery from Pinarbasi is hand made, and can be divided in a simple manner into coarse and fine wares. The pottery from the Area B rock-shelter site will be summarised first, since it appears in only a few contexts. The only diagnostic sherd from Area B came from a late context; it was a rim sherd from a large straight-sided bowl with band-slipped decoration and it suggests an Early Bronze Age date. Most of the few non-diagnostic sherds from this context are fine wares, and smoothed and polished on their external surfaces. Another context produced one sherd with a brown painted body band decoration on top of a creamy white slip, possibly running horizontally across the vessel.

From one of the late pit fills, four body sherds of medium coarse ware were recovered. All were reduced and badly fired and showed a contrast with Area A in their fabrics. Only large grey sub-rounded pebble inclusions were noted as temper, with no usage of white or black grits or of organics. The sherds were all friable, handmade and in a reduced state, and have been assigned to a pottery neolithic date on the basis of the associated chipped stone tool morphology, context type, as well as the resemblance of the pottery itself to some of the Çatalhöyük ceramics.

From Area A, characteristic Early Bronze Age vessels were found throughout the phases from the surface to Phase VI. The assemblage displays a remarkable level of homogeneity throughout the sequence, both in shape, form, fabric type and surface finish. There tends to he a 40:60 ratio between coarser wares to finer wares throughout the sequence. The finer wares are more usually finished with good surface treatments, and this tends to be in the form of a red paint wash or else a red slip and burnish. The coarser wares tend to fall into the "pithos-type" category, and their surface treatment ends with a cursory smoothing, hardly extending to a polish. Decoration other than a straight wash or slip is unusual.

The majority of sherds throughout the sequence are fired in relatively uncontrolled firing conditions. Although most of the vessels have been exposed to oxidising conditions, many have smoke-clouding on their exterior surfaces, and over 75% of them have black cores in their vessel walls from incomplete firing. The general colour of the vessels fabric pastes ranges from cream to an orange-buff. The finer wares tend to be lighter coloured and better fired (perhaps as a result of their finer walls). The inclusions selected for tempering the vessels also remain constant, and similar temper types are chosen for both the coarse and fine fabrics, with less denser and smaller quantities of better levigated and prepared organics and inorganics selected for the finer variety. One slight disparity between the two wares is that micacaeous inclusions (<0.5mm) are more commonly added to the finer wares, while organic grass tempered inclusions are more frequently used in the coarse wares.

A more or less 50% to 50% ratio between open and closed vessel shapes can be noted in the sequence, with a slightly heavier bias on jars and jugs in the earlier Area A levels. Open vessel morphology includes straight-sided, plain rimmed bowls, and even more V-shaped and globular-like bowls, generally with plain or tapering rims. Two complete open vessels from Area A include a fine ware goblet with a hollow ompholos base made from a creamy fabric with a good red paint wash on all surfaces. The second vessel was a fine V-shaped cup with a tapering rim, and well finished with a red slip. The bowl types find parallels with the globular bowls from Koca Höyük and Kurtaba, and the V-shaped bowls are similar to those from Akcasehir and Sarlak, as illustrated by Mellaart in his publications of survey work on the Konya plain.

The closed vessels also display several "typical" Early Bronze Age forms. They include holemouth vessel types, a tradition surviving from the Late Chalcolithic, necked jars, bag shaped jars, pithoi and barrel-shaped jars with plain rims and basket handles. Many of these vessels are constructed with coarser fabrics and are less carefully fired, being more associated with utilitarian functions such as cooking and storage. These closed vessel types also display parallels with other sites in the Konya Plain such as Kurtbaba. There is good evidence, therefore, to propose that the later levels found in Area A can be equated with sequences of Early Bronze Age date and that the latest features in Area B are of early ceramic neolithic date.

Chipped stone

Douglas Baird

The objective of the season was to recover useful quantities of all the chipped stone. Most of the contexts excavated were wet-sieved in quantity, and a number of quite useful chipped stone samples were obtained. These were examined and described, but not drawn. Drawing of some key tool types is scheduled to take place between seasons.

In area B the chipped stone pieces that are chronologically diagnostic indicate that all the deposits so far excavated are neolithic. One or two sherds of pottery from the later phases and the obsidian point fragments in the latest feature(s) suggest that the latest occupation is pottery neolithic, close to the date of Çatalhöyük itself, around 6,000 to 5,600 BC. The earlier deposits also have point types and bi-face fragments that indicate a 7th or early 6th millennium date, including the fills of the large curvilinear stone structure.

The technology of the chipped stone does indicate that there is probably an earlier, aceramic neolithic component in the rock shelter, that is, something a little earlier than the currently known material from Çatalhöyük. Microliths and elements of a small-scale, bladelet-based industry are also present in these neolithic deposits in the rock shelter. It is very interesting to note that they are different from those in the epi-palaeolithic site in area A, and are presumed to derive from an epi-palaeolithic occupation in or around the rock shelter. The epi-palaeolithic in the rock shelter is therefore likely to be of a different date from that of the contexts in area A.

In area A, there is no material of the neolithic date of the material in the rock shelter. We have strata which have produced an epi-palaeolithic assemblage. The sequence of deposits so far excavated in this area have yielded what appears to be (on the basis of very small samples) a relatively uniform assemblage, characterised by quite distinctive microlithic types. Unlike the neolithic in >the rock-shelter, quite a lot of flint was used alongside the obsidian. The products of the industry are of very small size, and the raw material was clearly intensively and exhaustively utilised, brought as it was over considerable distances. Classic geometric forms are very rare but there was a tendency to produce a microlith of distinctive character which was an elongated, asymmetric triangle or arched backed piece.

Animal bones

The animal bone material recovered both from wet- and dry-sieving is generally in extremely good condition, and in most of the contexts sampled in this season seems to be present in very good quantities. Some preliminary analysis of some samples was undertaken during the season, but without access to good reference collections and more detailed study it is not possible to make definitive statements. Good samples of material of neolithic date from the rock-shelter were available from a very early stage, the ratio of diagnostic to non-diagnostic bone being very encouraging. There are animals of all sizes, ranging from Bos and equids down to small birds, small mammals, tortoise and snake, and including ovicaprids, fox and hare. There are bones of larger birds, up to the size of goose, presumably taken as water migrants on the lake, but rather fewer fish bones than might have been expected. The equids are particularly prominent, and Equus hydrontinus is apparently well represented.

More difficulty was experienced with Area A. In the Early Bronze levels, to judge from the amounts of residual epi-palaeolithic chipped stone present, the bone samples are likely to be very mixed in date. In the epi-palaeolithic levels that were encountered first, the amounts of bone were rather small and it was mostly heavily splintered. In the deeper epi-palaeolithic deposits very good bone samples were retrieved. These have not yet been sorted even in a preliminary manner, but it is quite clear that they represent a wide range of species across the whole size spectrum from Bos and Equus hemionus down to small mammals, birds of all sizes and a few fish bones.

It is of interest to note that a single Dentalium shell, the species very much used in the epi-palaeolithic Natufian culture of Israel and western Jordan, was recovered, but from a mixed context in Area A. Four Mediterranean sea-shells from the epi-palaeolithic deposits in the same area further attest interesting long-distance contacts and exchanges.

Charcoal and carbonised seeds

Large whole earth samples were processed throughout the season, and some very good results were obtained. All the flot samples are to be examined during the winter: none was examined during the excavation period.

Area B yielded very large quantities of carbonised plant remains from an early stage in the excavation. Most of the material visible in the sieve as it was floated off seemed to be of wood charcoal., but seeds are certainly also present. In Area A carbonised plant remains were much harder to come by. In the latest samples to be processed, from the loamy deposits at the bottom of the sounding in Area A, there were small amounts of very small pieces of carbonised material but there were also large quantifies of fine root material There should be some material for radiocarbon dating, but, as is not uncommon with epi-palaeolithic sites, carbonised plant remains are not easily found unless particular deposits can be located.

Summary of results and significance

In neither of the excavation areas was the base of the archaeological deposit reached. However, both areas showed very high potential. In terms of the objectives of the first season, we now know that there is a depth and fine quality to the stratigraphy in at least one rock-shelter and in the open site on the peninsula.

In the rock-shelter there were deposits of extraordinary fine detail and lack of contamination immediately below the surface, and the promise of more, well stratified remains of earlier periods below. We have a quite unexpected neolithic period deposit in the rock--shelter, and the interesting prospect of working in approximate chronological parallel with the excavations at Çatalhöyük . The indications from the rock-shelter of a stratigraphy that may stretch from the ceramic neolithic back to the epi-palaeolithic form a quite unexpected bonus.

In Area A we have much more coherence in the Early Bronze Age remains than would have been expected from a surface deposit of only about0.30m depth. The Early Bronze settlement clearly merits investigation in its own right. The Area A excavations have shown fairly conclusively that there was a period of epi-palaeolithic occupation on the isthmus, and the first indications of burials and structures make that a most important site in both Turkish and West Asian prehistory.

In general, the demonstration of the existence of an epi-palaeolithic occupation on the Anatolian plateau is of considerable importance, both environmentally and archaeologically. More particularly, the idea that the Area A epi-palaeolithic site may be a permanent village settlement is an hypothesis that will be exciting to test. The apparent differences of lithic industry between the residual epi-palaeolithic industry in the rock-shelter and that of the open area on the peninsula, suggesting that there may be two epi-palaeolithic phases represented on the site needs to be reconsidered once there is more material and greater stratigraphic depth known from the rock-shelter site. And the chronological depth that the site offers to the context of Çatalhöyük itself is a further prospect.

We have good evidence that the sites possess extremely well preserved and generally prolific sources of botanical and zoological material. These are of great interest and very high potential for resolving questions about the subsistence economy and changes in economy through time over a very critical period, from the epi-palaeolithic well into the neolithic periods. They will also tell us a great deal about the local environment and environmental change during the late pleistocene and early holocene periods, which is a subject of considerable interest and some debate at present with regard to the Anatolian plateau. The close proximity and the scale (10.78m) of the core obtained by the palaeoenviromnental team should also be borne in mind in this context.

Much of the archaeological significance of the site will only be established once radiocarbon dates are available. However, the preliminary indications of date from the chipped stone assemblages seem to be quite robust, and the carbonised plant remains for accelerator dating are available.

The next steps

Further excavation of both sites (not to mention the other rock-shelter sites) is clearly justified in terms of their importance for research on the early stages of sedentism in permanent village communities and the transition to cultivation and herding that followed. In view of the richness of features on both the sites sampled this year, and the richness of the cultural and environmental / economic data, work will necessarily be slow, and processing of all the residues will be a major task.

We shall need a somewhat larger team and more time. The methodology of controlled, quantified sampling is working well and will be pursued for future seasons. In the rock shelter site we should open a wider area, covering perhaps half of the available surface in the rock shelter. It was clear from small-scale experiments this year that much of the material in the rock-shelter has lain exactly where it was deposited, and excavation of contexts in terms of individual square metres seems justified as a means of defining small-scale, highly localised spatial variation in the deposits.

It will also be important to reach further down into the stratigraphy during the coming season, as we need to know as soon as possible more about the scale, continuity and nature of the earlier deposits in order to plan future research for that area. We have to be aware that the deposit may be of considerable depth and great importance for West Asian prehistory, and plans for its investigation should not be made from year to year.

In Area A we also need to open up a considerably larger area of excavation, and allow for the fact that there is a significant Early Bronze Age occupation. to be explored before the epi-palaeolithic site can be reached. As in Area B, we should plan to capitalise on the area of our small sounding in order learn as soon as possible what is the stratigraphic depth of deposit that we are dealing with.

© Çatalhöyük Research Project and individual authors, 1994