ÇATALHÖYÜK 1999 ARCHIVE REPORT

Clays, Clay Balls & other Clay ‘Objects'

Kil Toplar Ve Diğer Kil Objeler

Sonya Suponcic

Abstract

Work on clay balls revealed that in the earlier levels of the South Area and the KOPAL area there were changes in fabric and form with a range of geometric shapes not found in later levels. Petrographic analysis and experimental work was undertaken.

Özeti

Kil toplar üzerine yapılan araştırmalar göstermiştir ki Güney Alanı’ninerken evrelerinde ve KOPAL alanında, kil toplar hammadde ve form açısından farklılık göstermiş ve geç evrelerde görülmeyen farklı geometrik şekillerle ortaya çıkmışlardır. Petrografik analizler ve deneysel çalışmalar da bu sezon yer almıştır.

Although work has been done on the body of material referred to here as ‘Clay Balls’ since the 1996 field season at Çatalhöyük (Hamilton 1996, Suponcic 1998) there has not been an in depth report produced discussing the current state of research on the material. This report is thus meant to both introduce the research currently underway on this topic while also reporting the specific work done during the 1999 field season.

Introduction to Clay Balls

Clay Balls are solid spherical objects made of fired or baked clay with smooth surfaces of varying colors weighing between 1 to 900 grams with diameters varying from 1 to 9 cm. They occur in every house excavated at Çatalhöyük in a variety of contexts including under, in, or near ovens, between buildings, and in refuse pits or middens. While their ubiquitous presence in domestic contexts leaves no doubt that they were part of daily practice, the function(s) and significance of the clay balls remains somewhat of a puzzle. The balls had traditionally been assumed to be projectiles or ‘sling balls’ used for hunting (Mellaart 1967). However, since beginning my research at Çatalhöyük during the 1997 field season I have been challenging this interpretation and have been conducting research related to my major working hypothesis – that the clay balls’ primary function was related to heat transfer and that they would have been used in cooking and food preparation as well as for heating rooms within houses.

Important New Finds from the Deep Sounding (South) and Kopal Trench

During the 1999 Field Season there were two very noticeable and important changes in the clay ‘ball’ materials found. These changes relate 1) to the shape of the materials found and 2) to their matrix or fabric.

Geometric Shapes

Material excavated from the upper most levels of the ‘off mound’ Kopal trench this season included not only clay ball fragments but also a range of other clay ‘objects’, all with a geometric form. Macroscopic observation with a hand lens (15X) show these objects have to the same fabric as the clay balls found in previous seasons (i.e. a fine mineral tempered paste with very few organic inclusions) (see ‘Petrographic Section’ of this report for further details). The surface treatment of these objects is also similar to the clay balls previously excavated (smooth well formed surface with some burnishing treatment). Shapes include: rectangular, conical, cubic, and ovoid objects. As excavation continued deeper layers revealed even larger numbers of clay objects with an increasingly wider variety of shapes and sizes. Of these the most distinctive were the rectangular and cubical objects which often have a yellowish-white slip (Munsell 2.5Y 8/2-8/4) and various incised patterns.

Similarly, in the deep sounding in the South area of the mound geometric shaped clay objects were found. As in the Kopal trench, these objects include rectangular, conical and ovoid shapes as well as the ‘typical’ spherical (ball) shape, but seemingly lacks the large number of rectangular and cubic objects found in the Kopal trench. Also in the south area (unit #5154, north of space 113, associated with a series of ovens from an early phase of this space) a cut was found which was presumably made into the fill next to an oven. This cut contained numerous whole clay objects including spheres, cones, and ovoids; none of these had the ‘slip’ mentioned above.

Changes in Fabric

This large quantity of ‘new’ geometric shaped objects found in the deep sounding also differ from previously found clay balls in terms of their clay fabric. Rather than a fine well sorted mineral temper, these objects have a more course, poorly sorted mineral temper which includes a wider variety of minerals (beyond the previously seen quartz, mica and feldspars). The objects also contain a larger amount of organic voids seemingly produced by the use of plant materials as temper. As work continued in both the Kopal and South, the excavated clay objects from both areas began to resemble each other in terms of both shape and fabric. It is also interesting to note that although the objects from the cut next to the northern oven in space 113 are similar in shape (cones, spheres, ovoids) to objects from the deep sounding and Kopal, they do not have the course poorly sorted mineral/organic tempered fabric mentioned above.

Clay Object Production

Preliminary Petrographic Examination

In order to begin to understand the methods of production of clay objects, samples were taken for petrographic analysis in both 1998 and 1999. The small number of samples taken in 1998 (13) were used to determine if petrographic analysis would be a beneficial method for obtaining production information. After preliminary examination of the 13 samples as well as a more in-depth examination of 3 of these samples it is clear that petrographic analysis can provide useful information related to the production techniques of these clay objects.

The 13 samples, which represent various contexts and temporal phases of the site, exhibit slight variations in color and texture when examined macroscopically. Examination with a hand lens (15X) shows very similar fabrics consisting of well to moderately sorted mineral temper (quartz and mica) with very few organic voids. However, after thin sections were made and examined using a polarizing petrographic microscope at least 3 different ‘types’ were visible. These types were determined on the basis of the size, sorting and variety of mineral and organic inclusions.

Petrographic Results

The 3 thin sections (one representing each of the three types mentioned above) chosen for more in-depth analysis show clearly the variation in clay ball fabric. Sample 2588M1 has a large amount (75%) of inclusions with very little clay fabric. This dramatically contrasts with the 20-30 % and the 10-20 % of inclusions seen in samples 2502M1 and 2261M1, respectively. The differences between the inclusions in these samples was not limited to their amount, it also extends to the sorting and size of these inclusions. Sample 2588M1 inclusions were poorly sorted with minerals ranging from 50 µm – 800 µm while sample 2502M1 had moderate to poorly sorted inclusions from 75 µm – 500 µm. At the other end of the spectrum, sample 2261M1 had well sorted inclusions with a smaller size range (only 50 µm – 350 µm).

Moving beyond the size and amount of the mineral inclusions, I would like to address the types of inclusions present and their relative abundance. All samples had similar mineral inclusions which include: quartz, feldspar, biotite mica, hornblende, and basalt. The relative percentages of these minerals were also fairly similar – in samples 2261M1 and 2502M1 quartz content was 50% of all inclusions while in sample 2588M1 the quartz content was only slightly higher at 67%. Feldspar content ranged from 7% to 17% (highest in sample 2502M1). In samples 2502M1 and 2588M1 hornblende and biotite mica content was relatively low at 3% each in both samples. However, in sample 2261M1 these minerals were found in much higher amounts, each made up 17% of total inclusions.

In terms of sphericity and roundness, the samples seemed to be relatively similar with nearly all inclusions being angular or sub-angular. There was only one case in which the inclusions were sub-rounded – the basalt and calcareous rock inclusions in sample 2588M1. Sorting, however, varied greatly between samples. Sample 2261M1 was the most well sorted, with all inclusions being well sorted (with the exception of quartz which was moderately sorted). Inclusions in sample 2502M1 ranged from poor to well sorting while sample 2588M1 was predominantly poorly or moderately sorted.

The analysis of the clay matrix is much less quantitative then the above description of the inclusions. Color was very different in all three samples. The colors ranged from light yellow-brown (sample 2261M1) in PPL to dark brown-black (sample 2588M1), and in XPL the colors ranged from golden brown (sample 2261M1) to dark brown (2588M1). It also seems important to note that of all three samples, one had a very low birefringence in XPL (sample 2588M1). In two samples (2588M1 and 2261M1) the inclusions nearest the surface of the clay balls were much more closely compacted. There were also areas in sample 2261M1 and 2502M1 with ‘inclusion-free’ clay aggregates. These clay aggregates are composed of fine material well below the 50µm course/fine ratio. One final result of my analysis which is related to the fabric of the clay balls is the presence of rounded vesicular voids in two of the samples (2502M1 and 2261M1).

These results can inform us on the production of clay balls as well as on the level of craft specialization and the social organization of production. The variation in the amount of inclusions and the similarity of the inclusions may indicate that there are different recipes for clay balls. Since the minerals inclusions in all clay ball samples are so similar in composition and relative percentage, people making these balls were likely obtaining the raw materials from the same source. What is different though is the amount of tempering material they were using as is evident in the course/fine ratios of the samples. Since there was such a wide variety, ranging from 75% inclusions to 15%, the idea of multiple producers with distinct recipes is most strongly supported. Moreover, the relative size and shape of the inclusions also lends support to the idea that individual producers were involved to greater or lesser extents in working the raw materials used in clay ball production prior to mixing of temper with clay.

In the scenario described above, individual producers would have traveled to an area with a sandy tempering source to obtain materials to be used as inclusions. An equally likely alternative is that the producers used remnants or small pieces of debitage from stone grinding activities as temper for the clay balls. They would have prepared these well sorted deposits by crushing them (more so in samples 2502M1 and 2261M1 since the inclusions were dramatically smaller then in sample 2588M1). These tempering agents would have been mixed with fine grained wet clay, thus resulting in the vesicular voids seen in samples 2502M1 and 2261M1 which are characteristic of very wet clay production. Evidence of mixing of the clay with the tempering agents can be seen in these same two samples which have clay aggregates visible within which there are no tempering materials. It is therefore likely that producers were using a fine clay source to which they added ground tempering agents without completely incorporating these inclusions into the overall clay ball fabric.

One final note should be made about the production of the clay balls which is related to the close packing of clay and inclusions toward the surface of the balls in samples 2588M1 and 2261M1. This packing is likely a product of the pressure created after the initial forming of the clay balls, possibly during the semi-dry or leather hard stage. This observation may indicate that the tempered clay mixture was formed into spherical shaped objects while in a very wet state as postulated above. These balls would have then been left to dry for some time and then smoothed by hand-pounding before firing. An alternative idea is that that the clay and tempering agents were mixed while the clay was quite wet and that later, after a bit of drying, hand sized lumps of this mixture were pulled apart and formed into spheres which were compacted on the surface by the action of beating the balls by hand in order to shape and smooth their surfaces. Both of these explanations would account for the compaction of materials visible on the surface of the balls while also explaining the vesicular voids remaining on their interior.

In order to help understand the production of the clay balls I conducted experimental work during the 1999 field season. These experiments began with locating raw materials which were likely utilized during the Neolithic (clay and mineral tempering materials). A clay source from the Kopal trench believed to have been available during the Neolithic was used for these experiments. The clay is grey ( Munsell Gley 1 6/ - 5/) with no visible mineral inclusions. There was considerable question as to whether this grey clay could produce the brown and orange colors common in most clay balls (commonly 7.5YR – 10YR). The clay was removed from the Kopal trench in a moist state and then dried in a shady area for nearly one week, producing hard crumbly clay chunks. These clay chunks were then rehydrated with water and smoothed by hand. Mineral tempering materials were obtained from Karadag by the stone specialist (Adnan Baysal) for his own research. Debitage fragments from these large stone samples were further crushed using a large basalt fragment. This debitage was added to the wet clay, mixed by hand and left to dry in a plastic basin for 4 days.

After drying, the mixture was formed into clay spheres ranging from 3-8cm in diameter. Cones, ovoids, and cuboid shapes were also formed using this mixture. These were left to dry in a shaded area for 3-4 days and then fired using a wood fueled fire which reached a maximum temperature of 500°C. The objects were first placed on the exterior of the fire for approximately one hour and then placed directly into the fire for one hour.

During firing all objects broke or fractured into several fragments which was likely due to insufficient tempering material which would have helped lessen the thermal shock. Insufficient drying was also partly responsible for the breakage. On the positive side, it was observed that the color of the objects after firing ranged from a dark brownish grey (Munsell 7.5YR 4/1) to a light orange brown (Munsell 10YR 6/3) – colors which are consistent with the majority of the clay balls and objects found at Çatalhöyük. These samples are currently at Middle Eastern Technical University where they will be analyzed along with archaeological samples of clay objects using X-ray Diffraction (XRD) and petrographic techniques in order to compare these modern experimental objects with those from actual archaeological contexts. It is hoped that a comparison of these modern and ancient samples will provide further information about the production techniques and technologies utilized during the Neolithic at Çatalhöyük.

Further Investigation of Clay Object Use

Although many theories have been advanced regarding the use of clay balls and clay objects the main working hypothesis is that these objects were used as ‘heat transfer devices’ both for various forms of cooking and food preparation and also as room warmers. According to this hypothesis, when used for food preparation and/or for heating an area of a house the clay balls would have been placed into a fire and heated for a prolonged period of time and then removed from the fire and used in any number of ways. For cooking, the balls could have been placed into a basket with some starchy plant (such as tubers) and/or meat and brought to a ‘boil’ – in essence they would have used these clay balls as ‘stone’ or ‘pot’ boilers. According to the same heat transfer idea, the clay balls may have also been used for pit cooking, in much the same way as modern Native North American Indian tribes use stones for roasting various meats and plants (Wandsnider 1997, Freiberg and Stenholm 1991, Yen 1975). If used this way, after heating the clay balls would be placed into a pit, which has been dug into the soil. Inside this pit along with the clay balls would be placed food (either plants and/or meats) and various types of moist greenery. The pit would then be covered with soil and left for 15-24 hours. After the elapsed time the heat from the balls would have slowly steamed the food making it palatable. Finally, if used to heat rooms, after being placed in the fire (presumably in a hearth located on the roof) the balls would be brought down into the room and placed in a small pit or in the interior hearth. There these heated clay balls would presumably give off enough heat to keep the small rooms warm for several hours.

Investigations into these various heat transfer hypotheses have continued since 1997 and include organic residue analysis, heating/firing temperature investigations, and contextual analysis of clay ball contexts. Although numerous samples have been taken for both organic residue and firing temperature investigations, these analyses have not yet been completed. However analysis of the context of the clay balls to this point support the heat transfer hypothesis. The balls are found in or near ovens and in at least one case from the 1999 season (unit # 5004) clay ball fragments were found to line the base of ovens, illustrating that those living at Çatalhöyük certainly felt the usefulness of these objects as insulators and heat transfer devices. As mentioned earlier, there were numerous pits (units 4915, 4913, 4941, 4290) found during the 1999 field season which contain both whole and fragmented clay balls along with little else in the fill. Each pit was found in the north central area of the room and each was close to an interior oven or hearth. This certainly leads credence to the idea that these balls were used in association with the oven/hearths. Unit number 5154, which is from the north of space 170 in the south area, contained a large amount of clay balls as well as geometrically shaped clay objects (as referred to earlier in the ‘Important New Finds from the Deep Sounding (South) and Kopal Trench’ section of this archive report). The close association of these clay objects with this oven, again supports the idea of these being used for cooking or some other activity which involves the use of the oven.

Newly Found Geometric Objects

As discussed in the ‘Important New Finds from the Deep Sounding (South) and Kopal Trench’ section in addition to spheres (balls) other geometric shaped clay objects were found this season in both the South and Kopal areas. The round shape of the spheres is particularly well suited for stone boiling and basket cooking, however the non-rounded shape of the clay objects would seemingly not be as effective. My research into these clay objects continues to focus on their use as heat transfer objects. Current theories include their use in food preparation in pits (as was proposed above for clay balls) and/or their use as ‘briquettes’ or as a base for grilling large sections of meat. However, the questions related to these objects are numerous and are still under investigation. Why do these objects (particularly the rectangular fragments) have this slip while it does not occur on the other shapes? Were all these objects used for the same purpose or do they have different functions? What is the relationship between the Kopal and deep sounding South area? These are all important questions which we are now attempting to answer.

Most importantly for us to understand at this point is if these new shapes along with the courser fabric are related to temporal factors, or are they simply related to the fact that these objects are found in contexts that are possibly ‘off site’? If Kopal and the deep sounding in the South area were both off-mound during the time of their use, then these objects may be related to an activity(ies) which are specifically done away from domestic spaces. The change in fabric (from a fine grained well sorted mineral fabric to the mineral/organic poorly sorted fabric found in the lower South levels as well as in Kopal) may be due to temporal changes. If this is the case, then it indicates that Kopal and the deep sounding are related temporally. However, this course fabric may be the ‘recipe’ used for these clay objects for either functionally or social reasons. It is difficult to determine the answers to these questions definitively, since there are very few even tiny fragments of clay ball available from these contexts with which we can compare. Again, this raises the same question – are the clay balls and clay objects used for similar purposes, with the clay objects being used in early time periods followed by a change to the spherical shape later in time? Or are the clay balls used domestically while the clay objects have a different use/meaning, one possibly related to off-site activities? The one thing which may help us understand and answer these questions is again related to unit # 5154 in the South area, space 170. The cut filled with clay balls and objects next to the northern oven in this room indicate that there is an association of clay balls with clay objects and that they ‘belong’ together in the house, particularly next to (or near) the oven. The balls and objects from this unit do not have the course temper seen in the deep sounding and Kopal objects, indicating that the change in fabric is due to a temporal change and not to the clay objects’ function.

Ongoing Laboratory Analysis

Finally, I would like to mention that samples were taken during the 1999 field season for further types of analysis which will help in answering the many questions which are raised throughout this report. Clay samples from ovens, bricks, and the Kopal trench were taken for clay mineralogy analysis using XRD. This analysis, which is currently underway by myself in collaboration with Professors Asuman Turkmenoglu of Middle Eastern Technical University’s Geological Engineering Department and with Ay Melek Özer of the Middle Eastern Technical University’s Archaeometry Department, will help inform us on the types of clays that were used during various building and daily life activities at Çatalhöyük. With this analysis I hope to see if the same or different clay sources were used for building ovens, making bricks, and for clay ball and object production. I also hope to see if these clay sources change through time or within certain areas of the site (for example were the people living contemporarily in the North area, the Bach area, and the South area all using the same clay source or were they each exploiting different sources?).

During the next 8 months petrographic analysis will continue with an additional 75 samples of clay balls and clay objects as well as experimental objects produced during the 1999 field season, which are also being formed into thin sections and analyzed at Middle Eastern Technical University (by myself in collaboration with Prof. Dr. Turkmenoglu). These analyses will further explain the production techniques of the clay balls and objects and give valuable information into the pyrotechnology and technological skill of Neolithic ‘potters’. Related to this topic is of course experimental work, which will continue during the two upcoming study seasons.

Additionally, samples were collected and exported for organic residue analysis and will undergo Gas Chromatography/Mass Spectrometry (GC/MS) in order to detect any organic residues present in the clay ball and object surfaces and/or interiors. It is hoped that this analysis will not only detect the presence of organic residues but will also be able to determine the type of organic material present. This is a crucial step in the determination of the clay ball and objects’ function as well as in helping to answer the questions related to the possible varied function of balls from the other geometric clay objects.

Bibliography

Freiberg, S.D. and N.A. Stenholm 1991 Surprise in a small package: a Lomatium roasting pit near Lake Wenatchee. Archaeology in Washington v. 3, p. 67-73.

Hamilton, N. (with a contribution by Mehmet Uluceviz) 1996 Figurines, clay balls, small finds and burials In On the Surface: Çatalhöyük 1993-95. McDonald Institute for Archaeological Research and British Institute of Archaeology: Cambridge.

Mellaart, J. 1967 Çatalhöyük: A Neolithic Town in Anatolia. Thames and Hudson: London.

Suponcic, S. 1998 Clay Ball Archive Report 1998. In Çatalhöyük Archive Report 1998. http://www.catalhoyuk.com/Archive_rep98/ar98_16.html

Yen, D.E. 1975 Indigenous Food Processing in Oceania. In Gastronomy: The Anthropology of Food and Food Habits. Mouton Publishers: The Hague, p. 147-168.