ÇATALHÖYÜK 1997 ARCHIVE REPORT

Çatalhöyük Pottery Report 1997

Jonathan Last

In 1997 excavations took place in five areas: BACH, KOPAL, MELL, NORTH and SUMMIT. This report provides a detailed discussion of the ceramics from all areas except SUMMIT, the material from which has been studied separately. However some comments on the typology and dating of this assemblage are offered as well. In general terms the earliest pottery is that from MELL (levels VIII-VII in Mellaart's phasing), followed by NORTH (VII-VI), BACH (probably VI), SUMMIT (IV) and KOPAL (?Late Neolithic/Early Chalcolithic). Hence the different assemblages provide snapshots of activity on and around the mound at several time points, although interpretation is limited by their small size.

1 Mellaart Area

Excavations in the Mellaart area continued from the 1996 campaign in Spaces 115, 116 and 117 and were initiated during 1997 in Spaces 109, 112 and 113. The latter are all level VII rooms largely excavated by Mellaart and which led him to characterise level VII as 'virtually aceramic'. The excavation of midden deposits in 1995-6 has shown that pottery, while not common, is certainly present in level VII, but the 1997 campaign confirmed that it is extremely scarce in building fill/make-up deposits. Space 109 yielded just seven sherds in both vegetable- and mineral-tempered fabrics while Space 112 had the same number of pieces and the same mix of fabrics. Space 113 produced 12 sherds, the majority of which are vegetable-tempered - including the diagnostics: two rims (1883.S2 and 2006.S2, both holemouth forms) and a base (1883.S1, a slightly unusual 'developed' form). Mean sherd size of these assemblages is small, equivalent to that from Space 106 in 1996 and lower than that from the NORTH area. No joins were recognized across units. It appears that only residual material occurs in these deposits; formal discard was reserved for the midden areas.

The spaces in which work began in 1996, physically and stratigraphically lower than the 109-112-113 row, produced the majority of the sherds from the Mellaart area in 1997, a total of 111. Only five small sherds came from the fill of the small space 116, all vegetable-tempered, to add to the 16 of 1996, which were much larger on average - though that was largely attributable to one big base sherd. The majority of the 23 sherds from this space (two more were found in 1997 in flotation samples collected the previous season from units 1664 and 1820) have been 3D-recorded as an experiment and the assemblage is discussed here as a whole.

In fact the material shows little significant patterning, probably due to the small numbers and because all the deposits considered here are generalised room fill. The horizontal distribution shows an even spread across the space, not indicative of the sherds accumulating from any particular direction. The vertical distribution is more interesting, with three groups of finds separated by zones with no sherds (Fig 27): above 1006.50m; between 1006.40m and 1006.10m; and below 1005.90m (probably - since only one of the 1997 sherds was 3D-recorded). At one level these appear to reflect distinct depositional horizons with assemblages of different attributes. The largest sherds come from the upper group - although this reflects the influence of the big base mentioned above, and the standard deviation of this group is accordingly high. However, the other groups also testify to a decrease in mean sherd size with depth, reflected in the difference between the 1996 and 1997 assemblages. This suggests, perhaps, that different sources of material were drawn upon for the filling of the building, which was therefore not a single event. On the other hand the plausible link between 1612.X2 and 1655.X3 as sherds of the same vessel from different parts of the fill (horizontally and vertically) apparently contradicts this. But the group cannot be considered in isolation - 1649.X1 from the middle of the room fill joins to sherds in 106 and 117, implicating adjacent spaces in the same broad fill process. Perhaps we should consider planned levelling or terracing operations mixing material from different sources. Clearly more data is needed on the speed of the filling process and the origin of the material within it, which will also have a bearing on the important question of building abandonment. It can, however, be stated with some confidence that the sherds from 116 have little to do with events or activities which took place within that space. Those events could be inferred only (and with caution) from material lying on or near surfaces.

Thirty sherds came from midden 115 to add to the 87 from 1996 (and six from flotation of 1996 units in 1997). This space probably dates to level VIII in Mellaart's terms, and accordingly all but two of the 1997 sherds were vegetable-tempered. Eight rims and one base were found, with the majority of the former being holemouth forms. This contrasts with the 1996 assemblage which had more open bowl forms. Also different from 1996 were the preservational characteristics with the sherds being somewhat smaller on average, but also thinner. The mixed nature of the deposits and their formation processes probably accounts for this variation. Nevertheless the recognition in 1996 of numerous sherd joins within the midden continued this season, with links between 2340.S4 (from a charcoal deposit) and 2344.S1 (coprolite layer), as well as between 2340, 2709 and 2739 (both fill). In addition, joins between 1996 and 1997 contexts were noted: these linked unit 2739 with 1513, and units 2340/2344 with 1533.

The majority of the sherds in 1997 came from Space 117. 77 sherds were found to add to the 58 from 1996, most of them from deposits excavated early in the season (1868, 1873, 1889). Ceramics were less common lower down in the fill, with only two sherds coming from the lowest 'orange layer' (which may be a surface); both of them were very abraded (one may be a potdisc). Unlike 115 the statistics for 117 are very similar to those from 1996. Sherds were again both larger in size and weight and thicker on average than those from 115. All the rims from this space were of bowls and all the bases were simple, rounded forms. This also follows the trend seen in 1996.

The most interesting finds in 1997 were associated with the 'bin' in the south-west corner of 117, which was filled with a large number of clay balls (1889). 1889 also produced relatively large pieces of three vessels: the first (1889.X15, S8, S9, 2322.S2, S4, S5, 2329.S2) includes some pieces from the dump deposit adjacent to the bin, although it is unclear how dispersed the fragments of this potsherd actually were - it may just have been a problem with defining the limits of the different deposits. The reconstructed sherd forms the lightly burnished rim of a simple bowl 130mm in diameter, with mottled surfaces. The second vessel (1889.X72, X113, 1868.S2) links the bin with an overlying dump unit; it consists of part of a large, coarse light-faced base about 140mm in diameter with signs of use-related abrasion on the underside. The third piece (1889.X114) is a large body sherd apparently not related to either of the other two vessels. Analysis of the other finds from this deposit may provide information on how the material got there, but the presence of animal bones suggests the assemblage represents discard rather than de facto refuse (abandoned material in storage), and the fragmentary nature of the pottery here would tend to confirm that interpretation. No sherd joins across units were noted among the rest of the 117 assemblage but joins were found within the larger units, e.g. dump 1873 produced 19 sherds from a minimum of 11 vessels. Moreover 2329.S1, 1873.S1 and 1873.S6 have a distinctive fabric with fine striations on the exterior and almost certainly came from the same vessel.

A real problem from the MELLAART area is the fragmentary nature of the assemblage, which has precluded the reconstruction of any complete profiles. The rims and bases found so far imply simple profiles, largely open bowls, and the limited technology (thick walls and abundant vegetable inclusions suggesting the use of rather non-plastic clay) would not suggest complex forms were readily achievable. However a few body sherds, such as 1868.S17, show a change of angle, suggesting a more complex profile. The large body sherd from 1889 (X114) does not have a regular curve to it either, but this may just reflect irregularities introduced during manufacture rather than a deliberate shaping.

The vegetable-tempered sherds of level VIII are often burnished but it is hard to tell by eye whether they are also slipped. In some cases, however, the surfaces show less vegetable matter than is apparent within the paste, which may be indicative of the application of a vegetable-free slip. This was assessed for Space 115 (1996-7), with the following results:

Hence perhaps a quarter to a half of vessels had slips applied to the exterior surface (rather less to the interior). This may have served to make the porous, often poorly-fired vessels more durable; the uneven surface treatments in the early levels do not give the impression of a decorative purpose like the glossier slips and burnishes of the later levels - when firing conditions, and therefore colours, were better controlled. However, there is a distinct class of early 'fineware', with less vegetable matter in the paste, or none at all, and careful high burnish. In 1997 examples came from 2009 (Space 112), 2075 (Space 109), 2078 (117), 2369 (113/115), 2726 (117) and 2739 (115).

How the burnish was applied is uncertain but in general, especially for the mineral-gritted pots of later levels, the visible burnish lines suggest a hard tool (stone, bone or ceramic) rather than a soft cloth. Two possible potdiscs came from the MELLAART excavations this year, 2766.S1 (38mm diameter) and 1868.S4 (27mm) - although rather more were found in 1996. These seem too small to have been lids or stoppers, given the known mouth diameters of vessels, and they may have been burnishing tools. No such rounded sherds have been found in the later levels, however, when burnishing was more extensive.

Other surface treatments are not common. Unusually, two sherds from close to the base of different vessels within 2029 (2029.S1, S2) had an unusual hard yellowish deposit on the interior surface; it is not clear what this represents but could perhaps be a resin aimed at waterproofing the pots.

The majority of fabrics, however, contain abundant vegetable matter. The precise nature of this is unclear but it is more likely to be agricultural chaff than chopped-up grass: the occasional inclusion of snail shells and rodent bones (see 1996 report) may indicate that this material had previously been stored. In 1997 large shell inclusions were noted in the fabrics of 2308.S1 and S2 (probably not the same vessel). More conclusive was the presence of a carbonised seed of einkorn (provisional identification by Julie Near) within a body sherd from 1889, which is probably from the same vessel as 1889.X15. The incorporation of the agricultural product within the body of the pot provides a symbolic link to go with the spatial link between pottery and fire installations (Mellaart pers. comm.) or with storage facilities (as in B.1 and Space 117), other contexts also associated with grain.

Mineral inclusions within the vegetable-tempered pottery can be summed up as sparse to rare very fine or fine white calcareous inclusions, presumably naturally included within the potting clay. The rarer mineral-gritted sherds appear to have been tempered with poorly-sorted sand, given the mixed nature of the material but the presence in most cases of some quartz. Grog was not generally noticed in either the MELLAART or NORTH assemblages but occasional clay pellets appeared in a few sherds from e.g. 2175 (NORTH), 2213 (BACH) and 2739 (MELLAART).

Fig 27: Sherds in Space 116

2. NORTH

Excavations in the NORTH area were concerned with burials and wall removal for a large part of the season and little pottery was found, in fact only 44 sherds from the whole area. Despite the small size, however, it is an interesting assemblage because both vegetable-tempered and mineral-tempered pottery is represented. The assemblage recovered from within Building 1 in 1995-6 was exclusively mineral-gritted and considered to equate with Mellaart's levels VI-V; the presence of several vegetable-tempered sherds in the make-up for B.1 and the fill of the space below (152) suggests that the lower building dates to Mellaart's level VII; B.1, directly above it, should therefore be considered as level VI. This is potentially an interesting sequence since it covers the transition from vegetable to mineral temper which was not fully seen in the Mellaart area, where excavations began in level VII with a mixed assemblage. The transition from VII to VI, when vegetable-tempered pottery disappeared entirely, can now be seen in the north.

In all 18 vegetable- and 25 mineral-tempered sherds were found, all generally small in size: the average dimension of mineral-gritted sherds was 26.4mm (mean weight: 5.2g), smaller than those from the BACH area (see below); the mean size of vegetable-tempered sherds, which are considerably thicker on average (9.3 compared to 5.2mm), was 38.0mm (15.0g). Vegetable-tempered pottery came from grave fills 1956, 2175 and 2194 below Space 71, where it can be considered residual; sub-floor packing in Space 111 (1969); infill/make-up in 71 (2162, 2180, 2181, 2517, 2539); and room fill in Space 152 below (2503, 2518, 2556, 2558, 2562). Mineral-gritted pottery came from mudbrick walls in Space 70 (2148, 2150, 2151); foundation fill/midden in 70 (2165, 2198) and 71 (2120, 2166, 2174, 2181); the fill of Space 73 outside B.1 (2160, 2519); and the fill of 152 (2533, 2535, 2542). From the packing/foundation fill of Space 71 the proportion of vegetable-tempered pottery was 27.3% and the mean density of sherds one every 233L. From the fill of 152 the proportion of vegetable-tempered was 62.5% and the mean density one sherd every 824L. At the level of the Space, therefore, the different fabrics are mixed, but in fact no individual unit had sherds of both types except for 2181. This may indicate separate contexts (or chronologies) of use, although it should be pointed out that only eight units produced two or more sherds!

The figures above suggest both the gradual nature of the change to sandy fabrics and the general lack of ceramic material in fill and make-up deposits. Understanding the social context of technological change depends, of course, on finding larger assemblages (which will probably come, as in the Mellaart area, from extra-mural midden deposits) as well as in situ material on floors and in features, particularly to assess the use contexts of pottery vessels of different technologies. It is to be hoped that, in the future, comparisons between Space 152 - when it is fully excavated - and 70/71 may provide some enlightenment in this direction.

No sherd joins were noted in the fill/packing material or with the 1996 assemblage from B.1. Diagnostic pieces from the NORTH area were few in number, with just seven rims which could be identified by shape. From the make-up in Spaces 70 and 71 were two mineral-gritted simple holemouths (2120.S1, 2165.S1) and one vegetable-tempered bowl (2162.S1); from Space 152 came one mineral-gritted bowl (2533.S1), one vegetable-tempered holemouth (2503.S1) and two vegetable-tempered bowls, including a unique sherd with a pierced lug coming straight off the rim (2556.S1). The only other example of such a placement was a piece from Space 117 in the Mellaart area (see 1996 report), although in that case the handle projected above the rim while here it sits flush with the rim. Perhaps limited innovation in forms was part of the trend towards a new technology through level VII. In general, however, the model of a simple transition from open bowl forms of level VII to holemouth jars of level VI is not contradicted by these finds.

The only base junction was a dark-burnished mineral-gritted piece from 2165, just 2mm thick (2165.S3, S4). This is typical of the B.1 assemblage from 1996, and also resembles 2255.X3 from the nearby Building 3 in the BACH area.

3. BACH

A limited excavation season by Ruth Tringham (UC Berkeley) in the BACH area, directly south-east of the NORTH area, was hindered by the presence of a number of late Roman burials. The season produced only 156 potsherds from building fill and burials, as well as two complete vessels from one of the graves.

i. The Burials

a. Vessels from 2237

Two vessels were found with this burial. The first is an unguentarium in a pinkish, moderately gritty fabric with a yellow slip. It stands 250mm high and has a number of characteristics which suggest it is later than the Hellenistic period: these include a simple base, a weak shoulder, and the maximum diameter occurring low down on the body.

The second vessel is a small oil lamp, with blackening around the spout (wick) providing evidence of use. The body has a maximum diameter of 45mm and the lamp stands 19mm high. It is made in a buff fabric with faint traces of a red slip.

b. Material from fills

In all 67 sherds (41.7%) came from the burial fills, of which 22 are Classical (Hellenistic/Roman/Byzantine) and 39 Neolithic. The remainder are Chalcolithic or of uncertain date. Both the Neolithic and Classical material was generally small and probably redeposited with the burial fills rather than incorporated as fresh material. The mean size of Classical sherds was 27.1mm, while Neolithic sherds averaged 22.1mm (mean weight 3.7g).

The Classical material included little that is specifically diagnostic. Slipped wares of Hellenistic type came from 2206, 2212 and 2226, but these may be considerably earlier than the burials, given the likely date of the grave goods.

Noticeable among the residual Neolithic material is a relatively high proportion of red slip sherds (7 pieces), which are generally late in the Neolithic sequence. Five rim sherds were present, of which two could be characterised by form; both of these are open bowls, one (red slipped) with a slightly angular profile, which is also a late type (2235.S1). The most unusual piece is a body sherd with raised, applied decoration in a curvilinear form (2212.S6). It has no parallels among either the Classical or prehistoric material so far discovered at Catalhoyuk and is not slipped or burnished, but appears to be handmade in a fabric which would not be out of place in a late Neolithic or early Chalcolithic assemblage.

ii. Building 3

The building fill produced 67 Neolithic sherds and 19 Classical pieces, as well as five probable Chalcolithic sherds and three of uncertain date. The late sherds are not restricted to the uppermost surface units, but they are absent from the lowest spits within the fill and midden deposits (2222, 2228, 2229, 2239, 2242, 2246, 2247, 2249, 2252, 2255). They occur in all quadrants except the south-east corner (2209, 2216, 2229, 2247) and hence show the disturbed nature of the upper fill across virtually the whole of Building 3. These sherds are smaller than those from the burial fills (mean length 19.8mm) and have perhaps been introduced by soil animals or other post-depositional alteration of Neolithic deposits.

In contrast the Neolithic sherds, though still small, are larger than the residual pieces from the grave fills (mean length 29.3mm, weight 6.5g). There are also typological differences, with proportionately fewer red slip sherds (all from the uppermost spits or those with Classical contamination), but a number of vegetable-tempered sherds of the MELL type (all from lower spits or midden deposits). This suggests that while the occupation of B.3 is probably contemporary with the adjacent B.1 (NORTH), there are residual traces of a later occupation (late Neolithic/early Chalcolithic) which has perhaps been eroded since the Classical activity. The colluvial deposits visible in the KOPAL trench (see 1996 report) suggest significant mound erosion since the Byzantine period; the shallow depth of the surviving graves in BACH might give some indication of how much has been lost. On the other hand it is unlikely that three or more building phases (necessary to fill the gap between the B.1 material and the red slip sherds from the surface) have been lost, so the nature and extent of post-level VI occupation on this northern part of the mound remains in question.

Few typologically diagnostic elements were found. All the rims were from bowl forms but while this would superficially suggest a post-level V date, it needs to be treated with caution. Only three rim sherds came from undisturbed contexts, one of which was part of a vegetable-tempered vessel (2249.S1) while the other two were joining pieces (2227.S1, 2228.S1). This latter vessel comprises the only evidence for non-residual material: the joining sherds lie across two quadrants, while a number of other large sherds, all from the three south-eastern squares within B.3, may be (non-joining) parts of this vessel (2209, 2216, 2239, 2255.X3). It is a straight-sided open-mouthed form with a very slightly inturned rim and a diameter of 125mm. With its dark brown burnished exterior (and thin base, if 2255.X3 is the same) this pot closely resembles a number of vessels from B.1, and is the best evidence that B.3 and B.1 are indeed broadly contemporary - as the stratigraphy would suggest.

The south-eastern corner from which this vessel derived also produced the greatest number of Neolithic sherds within B.3 (though the volume of soil excavated may have been greater here) and larger pieces on average than in the rest of the building. It remains to be seen whether this pattern is reflected in the distribution of any in situ material preserved on surfaces lower down.

4. SUMMIT

The SUMMIT area produced a large number of sherds: I looked at 2162 pieces, of which 1302 (60.2%) were Neolithic (this excludes fabric samples and some refitted sherds). Much of this material, however, comes from late pits and contaminated contexts: units with >50% post-Neolithic material include 1759 (??), 1760 (surface), 1761 (fill), 1769 (cut feature), 1773 (fill), 1776 (arb. layer), 1777 (pit), *1787 (pit), 1794 (layer), *1796 (layer), *1797 (layer) and 2613 (fill) - starred units have at least 30% Early Chalcolithic sherds. These include a mixture of Neolithic material, some units having primarily dark-faced sherds and holemouth rims characteristic of levels VI-V (e.g. 1760, 1777, 1787, 2600, 2604) while others have red slip bodies, footed bases and angular bowl profiles characteristic of levels IV-III (e.g. 1758, 1762, 1773, 1796). Spatial analysis of these units may provide information on the formation processes of different deposits in the SUMMIT area.

Of the Classical material Hellenistic slipped or painted wares came from 1755, 1760, 1762 and 2605, but the majority of the assemblage are coarse wares which could date anywhere in the Classical period. The large portion of a red slip handled jar found in 1797 is probably also Hellenistic, however.

Uncontaminated Neolithic contexts include: 1756, 1765, 1778, 1792 (floors); 1763, 1798, 1799, 2616, 2619, 2633, 2636, 2662 (fill), 1775, 2641 (arbitrary layers); 1795, 2643-44, 2673, 2676 (platform); 2602, 2607, 2655, 2671 (midden); 2604, 2659 (layers); 2611, 2618, 2622, 2632 (infilling); 2612, 2621, 2638, 2654 (pit); 2620, 2625, 2629, 2635, 2660, 2669 (wall). It is noticeable that these assemblages include no developed base forms nor angular bowl rims (which both become common from level III), and very few red slip bodies. Nevertheless open bowl rims continue to outnumber holemouth forms in these deposits, so the in situ assemblage can probably be assigned to Mellaart's level IV. The reconstructed vessels from pit F.101 (mostly excavated in 1996) include level III forms - angular profiles and red slipped surfaces - but also a number of relatively large holemouth vessels (including necked forms) and unusual features like vertically placed handles. This may be a rather specialised assemblage, but could well be somewhat later than the rest of the Neolithic features.

5. KOPAL

Analysis of the material from the KOPAL trench excavated through colluvial deposits on the northern slope of the mound in 1996 proved the extent of mound alteration since the Classical period. Work in 1997 concentrated on a sounding in the field immediately north of the site which was aimed to pick up the interdigitation of colluvial and alluvial deposits. However, the 1996 trench was also reopened and some further material was collected, notably from the surface of the intact archaeological deposits beneath the colluvium. This focussed on the stretch between 1260-1270m N. Hand-collected material from the colluvium (Unit 2400) produced 268 Classical, 59 Neolithic, 5 Chalcolithic and 17 uncertain sherds, including one piece with a roughly scored surface and a few organics in the paste which represents the only EB sherd found so far at Çatalhöyük. Part of an Early Chalcolithic incised potstand was also found (2400.20). The Classical material mainly comprised later Roman coarse wares (definite Hellenistic wares were absent), while the Neolithic sherds included a number of red slipped bodies and nine rims, of which six were open forms and two of these angular bowls (e.g. 2400.3). These features and the presence of developed base forms, including a ring-base (2400.6), indicate that much of this material is late Neolithic, i.e. levels III-II. The presence of a small but consistent Chalcolithic assemblage is interesting, especially given the inclusion of large amounts of Chalcolithic pottery in some of the Classical features in the SUMMIT area; no in situ material of this period has yet been found on the east mound, but some kind of occupation or activity, perhaps subsequently largely eroded, must be implied.

Material from the top of the archaeology (2401) does not clarify the picture greatly, since it contains a variety of material from typical later Neolithic burnished sherds (c levels IV-II) to possibly Early Chalcolithic light brown slips (no painted sherds were found) as well as a rim sherd in a vegetable-tempered fabric typical of the MELL area, but in a rather unusual (for those levels) angular form. The most unusual find, however, was a sherd with incised and impressed decoration, the impressions infilled in white and the sherd partly painted red (2401.3). The only comparable piece is a possible pot-lid from 1787 in the SUMMIT area, with incised and infilled decoration. However, the latter comes from a deposit with a large amount of Chalcolithic pottery and indeed is formed in a typical gritty orange Chalcolithic fabric. The sherd from 2401, in contrast, has a light brown fabric with a thin grey core, and some vegetable matter. This sherd is wholly alien to the Çatalhöyük Neolithic ceramic tradition in fabric and decoration, and if it can truly be assigned to a late Neolithic deposit, must be an imported piece. It is perhaps more likely to be early Chalcolithic but is by no means typical of Çatalhöyük West either. Test excavation to clarify the nature and date of activity on the northern slope of the east mound would clearly be useful.

The off-site KOPAL trench produced a series of deposits over a depth of some 3m. Beneath the topsoil, machining encountered fragmentary architectural remains of late Classical date. The pottery from this deposit (2403) includes a number of large handled vessels in orange/brown fabrics, some with cream slips (e.g. 2403.2). An abraded red slip Neolithic rim sherd was also found. Beneath the alluvium a carbonate-rich silt (2407), which may represent a buried land surface, produced part of a large, necked holemouth pot (2407.1) and a number of body sherds (e.g. 2407.2). In date these may best be assigned to Mellaart's levels IV or III. The size of some of the pieces does not suggest the deposit arrived there through colluvial processes (but see below). Beneath this was a thick buried soil (2410) which produced a miniature red slipped lug (2410.1). This is reminiscent of the small 'vestigial' knobs found on some Chalcolithic vessels but it is in a late Neolithic fabric (possibly therefore level II-I?). Backswamp deposit 2404, which underlay this, produced two Neolithic burnished sherds, not especially diagnostic but certainly later than level VI.

The dark silt 2412 was the lowest deposit to produce ceramic finds but they were of a rather unusual nature. Only one potsherd was found. It has plain, unburnished surfaces but in fabric is somewhat reminiscent of the incised sherd from 2401, also including vegetable matter - which is extremely rare in post-level VI Neolithic deposits but does sometimes occur in the Early Chalcolithic. Associated with this sherd were a number of fired ceramic bricks and other objects. Two of these fragments were clearly from cuboid objects and had a cream slip reminiscent of west mound pottery; one piece even has traces of red painted decoration over this slip (2412.3). Another piece recalls Neolithic clay balls but also has a slip, which is never found on the east mound clay balls. Perhaps these objects are broken potstands, but they differ in surface treatment and fabric from the 'classic' incised Early Chalcolithic potstands, such as that found in deposit 2400. Whatever their function they seem to represent a specialised deposit, perhaps associated with some off-site industrial activity (could the cuboids even be crucibles?). The date is even more of a problem, because if they are indeed Early Chalcolithic and in a deposit sealed by backswamp clay 2404, all the Neolithic material above them (including the large pot fragment in 2407) must be redeposited, colluvial material. It would also imply that the earliest deposits accumulating on the marl sands north of the mound are considerably later (by a millennium or more) than the first occupation of the mound itself.

Figure 27: Sherds in Space 116