Kuro-Araxes wares: Reconstructing Ancient Technology (KARAT)

On-site experimentation at TBY in 2009-10 was based on the raw material analyses conducted earlier, which determined that one of the two main sources of materials used in KKW was valley rendzina from the site itself. Accordingly, experimental re-creation of the KKW production sequence began with the preparation of raw material based on soil excavated at the site, largely derived from decayed mudbrick, in the course of the archaeological field season. Inclusions—organics, grog, and naturally occurring nonplastics—were prepared and added according to these same analyses.
Through trial and error, and in view of the extreme leanness of the soil, the following procedure was determined for clay preparation:

1) The finest siftings from the relevant excavated strata were mixed with water into a thin slip. When the mixture was completely hydrated through mixing, it was allowed to settle briefly . Following gentle agitation, the mixture was poured off, leaving only the larger particles in the bottom of the bucket. The resulting slip was left to settle over night.

2) After pouring off the extra water, the slip with clay particles and minute inclusions was decanted. Some of the larger inclusions that remained at the bottom of the buckets were saved to be reintroduced to the clay at a later stage.

3) Step 2 was repeated twice in order to isolate increasingly finer particles.The ratio of sifted raw material collected to usable clay by volume was 1:4.8. Recent granulometric studies of the TBY matrix (O. Ackerman p.c.) suggest that this leanness may be a product of pretreatment of the soil used for the mudbrick; thus, soil obtained directly from the lake or river cuts would have been richer in clay.

4) The mixture was left to settle for a day and some of the upper most layer of clay was skimmed off and set aside to make the sig.

5) The buckets of good clay were then laid out to dry to a wedgeable consistency.

6) The clay was then wedged and divided into four different clay bodies:
- Straw + Clay (with inclusions)
- Grog + Clay (with inclusions)
- Grog + Straw + Clay (with inclusions)
- Only Clay (with inclusions)

7) The percentages of straw, grog, and inclusions were the same through out the four different clay bodies.
The straw was roughly a 1% addition
The grog was roughly a 3% addition
The inclusions, which represented a range of sizes obtained in Step 2, were roughly a 12% addition

Grog was prepared from discarded KKW sherds by pounding and sifting until the required size was attained. Organic inclusions were made of cut dried grass (a hand blender was used to mimic the effect of a threshing floor).

A number of insights afforded by the experimental work may be introduced here:

a. The use of onsite soil saves transport labor but results in poor-quality clay (see below) that requires considerable time and effort for preparation.

b. The poor raw-material to clay ratio indicates a deficiency of actual clay in the siftings. Is it possible that some grinding method was used in antiquity to make the clay finer, thus increasing plasticity? As demonstrated by relatively predictable percentages of straw, grog, silt, and inclusions, these people had a control over their clay body that suggests at least a modicum of testing. If so, this would correspond with the want of finer particles for the sig.

c. The green strength of this clay is almost too low, resulting in a very weak clay body. The reason, at least in part, may be related to particle packing issue and more precisely a lack of particle range. A body is strongest when its components have many different particle sizes, and are gradually increase in size. In the case of this valley rendzina clay, it seems the clay particles are large and perhaps they do not vary in size. Possibly, better results could be obtained by ageing the clay.

d. Grog-making is very tedious and time consuming. Given the time and effort, grog might be considered a precious commodity.


Vessel Formation

Both additive and reductive methods of construction were explored. The additive methods included slump molding, coil building, slab building, pinching, and press molding. The reductive methods included trimming and scraping. The reductive methods were only used once a rough form was made by one of the additive techniques.

A large amount of full-length straw was wedged into the large vessel clay body to make a green ware press mold. The omphalos base was formed into this press mold. If bisque ware were use as molds, and bisque ware can make good molds, then perhaps what have been classified as bisqued ‘bowls’ are actually molds. If partially broken bowls were used as molds they would be classified as incomplete vessels, or perhaps they would have deteriorated into shards. If this is the case, there would be no “molds” to find, only more broken bowls or shards. Could a mortar, for food or other use, have been used as a mold instead and therefore “molds” have been found, but because of their other use have not been thought to have been molds?

Several different tools made from different material were used to burnish both raw vessels and slipped vessels. The green stalk of a palm frond produced a clean, smooth, well burnished, uniform surface on both raw vessels and slipped vessels. A water-smoothed piece of flint proved the best burnishing tool, producing surfaces that matched KKW quality.

Burnishing was made more difficult by the weak green strength of this clay. It was a fine line between applying too much pressure and consequently breaking the vessel, and not applying enough pressure to achieve a properly burnished surface. If the clay was a little on the wet side, and too much pressure was applied while burnishing, small, directional burnishing marks would form rather than a smooth, tight, flawless surface. Some KKW shards exhibit such a surface. There is a window of optimal moisture content within which to achieve the desired burnished effect. When the test surfaces were almost bone dry, a sea-smoothed piece of flint was used to burnish them to a glossy sheen leaving very slight directional burnishing marks.


Using locally available combustibles, test tiles are fired in open firing pits. Preliminary tests show that while the needed temperatures can be achieved, a mechanism for the introduction of oxygen was probably used in order to achieve the red coloration. Various methods were successfully used to achieve black coloration, including the application of dung to the test tiles.