THE CASE OF THE “HUSKING TRAYS”: AN EXPERIMENTAL ANALYSIS OF A TYPOLOGY SUPPOSED TO BE USED FOR BAKING BREAD Sergio Taranto1

ABSTRACT In the present research, the attention is focused on the so called “husking trays”; one of the most “international” pottery shapes used by the communities of Upper during the 7th and the first half of the 6th millennium BC. The main characteristic of these large trays is that their interior surfaces were crossed by incisions and impressions of different depths. Several hy- potheses were suggested by the scholars about how the husking trays could have been used, but these ones have remained merely theories so far. The two most shared hypotheses about their functionality have been tested through a first set of experiments.

RESUMEN En la investigación actual la atención ésta focalizada en los llamados “husking trays”, una de las formas cerámicas más “internacionales” utilizadas por las comunidades del Neolítico Final a lo largo del 7º y la primera mitad del 6º milenio BC. La principal característica de estas grandes bandejas cerámicas es que tienen una superficie interior con incisiones e impresiones de distinta profundidad. Diversas hipótesis han sido sugeridas por los científicos sobre como los husking trays pudieron haber sido utilizados, pero estas no han salido del ámbito de la teoría. Las dos hipótesis más compartidas sobre su posible función han sido objeto de una primera serie de pruebas experimentales.

Keywords: Late , Upper Mesopotamia, pottery, cereals, ovens, bread.

Palabras clave: Neolítico Final, Alta Mesopotamia, cerámica, cereales, hornos, pan.

1. Universitat Autònoma de Barcelona in joint research with l’Università degli studi di Roma la Sapienza. GRAMPO; SAPPO; LTFAPA [email protected]

Butlletí Arqueològic, V, 40 (2018), ISSN 1695-5862 (p. 57-63) 58 SERGIO TARANTO

Introduction As early as the mid-Twentieth century several fragments of a large pottery ba- sins came to the light from strata of archaeological excavations in the Northern Mesopotamia. They were large trays with wide oval bases (up to 60 cm in length and 40 cm long) and low sides (ca. 10 cm high) made of a coarsely strew-tempered clay. Both the base and the sides were extremely thick and usu- ally carried out in a rough way. Meanwhile the outer surfaces did not show any decoration, the inner surfaces were completely crossed by incisions, impressions or grooves of different depths. By these actions the surfaces could be characterized by a wide variety of patterns: parallel lines, criss-crossed lines, impressions and also combining different patterns together. These marks entirely covered the inte- rior surface of the husking trays and have always caught the curiosity of scholars.

Functional hypothesis Since the first findings of fragments of husking trays, the archaeologists (Lloyd and Safar 1945), made the attempt to imagine a function for a such peculiar pot- tery shape. Thus, they suggested that those basins were used to facilitate the de- tachment of the grains from their husk. For this reason, they gave them the name “husking trays”. Likely they supposed that these trays were used like large graters on which corrugated surface the spikes would be rubbed so on the corrugated sur- face that the cereal caryopsis should be husked. In the early 80’s, when M. Voigt was writing about the husking trays found in the excavation of Hajji Firuz Tepe in , she questioned the previous interpretation of Lloyd and Safar. She wrote: “It is unlikely that the trays were used for husking, given their size and weight and the availability of baskets for tossing grain into the wind….One possibility is that they were used to bake flat bread: when heated, their textured, matte interiors would have provided a surface on which a thin layer of dough could have been cooked and removed with a minimum of sticking” (Voigt 1983). Since then, it had begun to be doubted that the husking trays had the function that their name suggests. Today a lot of scholars tend not to believe it. Instead Voigt’s hypotheses had enough success within the scientific world. Other scholar hypotheses were put forward, but with less success.

Experimental tests To verify if the two most shared hypotheses about the functionality of the husking trays were actually feasible, experimental tests were carried out1: a “husking experiment” and a series of “baking experiments”.

1. A set of experiments were performed. Here only the main results will be shown.

Butlletí Arqueològic, V, 40 (2018), ISSN 1695-5862 (p. 57-63) THE CASE OF THE “HUSKING TRAYS”: AN EXPERIMENTAL ANALYSIS OF A POTTERY TYPOLOGY 59

To limit the variables involved in the experiments, some parameters related to the main aspects of the tests (pottery, fire installations, cereal, dough, sour- dough, fuel) were arbitrarily chosen2. • Reproduction of the husking trays. The first step was to reproduce the husking trays on the basis of a precise context which showed a good con- cerning documentation (Mathieu 2002). The Syrian archaeological site of Sabi Abyad was chosen (Nieuwenhuyse 2008). On the basis of the drawings and the compositional data of husking trays found in this site, functional replicas were recreated. Six reproductions, were performed according to both the pinching and the coiling techniques with coarse clay, a large amount of vegetal bits and water3. • Reproduction of the fire installations. To test the second hypothesis it was necessary to reproduce also the three kinds of fire installations supposed to be present in that period: domed-oven, tabun-like oven4 and fireplace.

Husking Experiment Firstly, the husking trays were used to separate the grains from the husk to verify the hypothesis suggested by Lloyd, Safar and Braidwood. Thus, the husk- ing trays were used like large graters. Barley spikes were rubbed against the inner surface of the husking trays. The experimental tests have revealed that is highly unlikely that the husking trays were used in this manner because the surface eas- ily eroded; the pottery is too fragile to do an action like this one with such hard grains. In addition, the impressions/incisions did not seem to be functional, if anything they were an obstacle. Usually, the impressions and the grooves were so large that whole bits of spikes filled them without breaking up.

Baking Experiment Then the husking trays were used as trays to bake bread. Obviously, there is an infinite variety of kinds of bread in the world. The objective was also to

2. During the experiments both durum wheat and hulled barley were used according to the main common paleobotanical remains of the Late Neolitihic sites and Sabi Abyad (Ak- kermans et al. 2014); the flour was stone-grounded and was roughly sifted. For the baking tests wood-coal was used as fuel. 3. The major part of the husking tray replicas was carried out during the course of experi- mental archaeology maintained by professor C. Lemorini in the 2015/16 at the Sapienza Uni- versity of Rome. 4. The actual existence of tabun-like ovens for the Neolithic Period of the Near East was recently questioned (Balossi-Mori 2014).

Butlletí Arqueològic, V, 40 (2018), ISSN 1695-5862 (p. 57-63) 60 SERGIO TARANTO obtain a bread not completely burned and yet not so undercooked. Preferably a bread baked homogenously. To understand what happened during the experiments it is necessary to un- derline two concepts of the physical and chemical transformation involved in baking bread: 1) The crust of the bread comes out when the superficial part of the dough abruptly comes into contact with high temperatures, while the inside part of the dough turns into crumbs caused by a different kind of phe- nomena. 2) In the cereals there is a great concentration of gluten. It creates strong interconnections inside the dough and the bread became a unity: the loaf.

I baking experiment In the first set of experiments, the husking trays were used for baking bread in the domed oven (fig. 1). Even if the variables like consistency of dough (liq- uid, semiliquid or solid), cooking-time, temperatures, presence/absence of sour- dough, kinds of cereal-flour were changed, the results were always negative. While at the base of the tray the dough remained totally undercooked; so wet and tenuous that the marks of the tray were completely useless. Instead the top of the dough turned into a super-hard crust that stuck in a very strong way to the sides of the husking tray. After baking it was impossible to take out the bread from the tray without the help of metal tools and to avoid damaging the ceramic surface. All these problems were due to the fact that the heat only came from the top.

II baking experiment Bread was baked by means of the tabun-like oven in the way that the heat would have come from the bottom. The tray was put at the top of the oven (fig. 2). Also, with this kind of oven the variables were changed more times but, suf- ficient satisfactory results were never obtained. In fact, not only did the upper surface of the bread remained undercooked but, the use of this kind of oven left burning traces on the external surface of the tray that does not appear on the archaeological shards.

III baking experiment For the third set of experiments the hypothesis suggested by Voigt was tested: the husking trays were used with the simple fireplace. Once again, even though the variables were changed, the husking trays did not work either: the dough never got completely baked. The reason was that the husking tray was so large

Butlletí Arqueològic, V, 40 (2018), ISSN 1695-5862 (p. 57-63) THE CASE OF THE “HUSKING TRAYS”: AN EXPERIMENTAL ANALYSIS OF A POTTERY TYPOLOGY 61 that it prevented the oxygen to feed the burning-coals below 5; it extinguished them and the dough did not have enough heat to turn into bread. To remedy this, an attempt was made by putting another husking tray as a lid, or by wrap- ping the dough in leaves. However, these attempts did not have any success.

IV baking experiment In the fourth and last set of experiments, on the basis of an ethnographic comparison6 (Balossi - Mori 2014) the technique of preheating the tray be- fore baking the bread was adopted. The first step was to heat the domed oven where an empty husking tray was left for 10 minutes to 400ºC to reach a high temperature and allowed it to retain the heat. After that the husking tray was boiling, it was pulled out from the oven, sprinkled with flour and filled with a solid dough. Coming into contact with the high temperature of the vessel, the dough turns into a thin crust. Immediately after, the husking tray was put into the oven again, so that the dough continued its baking. It received the heat from the top (the temperature inside the oven) and also from below (thanks to the heat the tray had already previously retained). The dough turns into bread perfectly. The inferior surface of the dough turned into crust. In this way not only did the incisions/impressions of the husking tray leave evident marks on the surface of the bread but, they seem to have had the function to facilitate the detachment of the bread from vessel.

Conclusions If the husking trays were used in this way every characteristic of this pottery shape makes sense: • The wide base and low sides qualify the husking tray as a perfect pottery shape to cook food in a domed oven because its shape allows food to be evenly distributed so that it could reach a similar temperature at the same time and cook in a homogenous way. • The extreme thickness of the walls would serve to retain the heat that the tray would release when the bottom coals extinguish, allowing the dough to continue cooking; • The coarse, plant tempered clay would minimize the thermal stress. • The oval shape of the husking tray allowed a more homogenous baking (overall all the central part of the dough) than if it was circular.

5. This phenomenon was noted when the tray was lifted from the fireplace: the coals below the vessels remained black while the surrounding coals turned into grey ash. 6. The idea was suggested to me by professor F. R. Balossi.

Butlletí Arqueològic, V, 40 (2018), ISSN 1695-5862 (p. 57-63) 62 SERGIO TARANTO

In conclusion, this set of experimental tests allows us to believe that the husking trays could be perfectly used to bake bread. Moreover, it suggests that the incisions/impressions could be anti-adhesive arrangements. Their function- al purpose would be to create a bumpy surface which would help avoid the dough from adhering well to the vessel. Moreover, this hypothesis also fits well with the archaeological data. This first analysis needs further experimental tests which will be able to support and specify these data7.

Acknowledgments I would like to thank Professor C. Lemorini and Professor F. R. Balossi for all of the suggestions and the availability shown during my MA thesis research.

REFERENCES Akkermans et al. 2014: Excavations at Late Neolithic , , The 1994-1999 field seasons, Papers on Archaeology from The Leiden Museum of Antiquities, 11. Balossi, R. F.; Mori, L. 2014: “Bread, baking moulds and related cooking techniques in the ”, Food and History, Revue de l’Institut Européen d’Histoire de l’Alimentation, 12 (3), p. 39-55. Lloyd, S.; Safar, F. 1945: “Tell Hassuna: Excavations by the Government Directorate General of Antiquities in 1943 and 1944”, Journal of Near Eastern Studies, 4, p. 255-289. Mathieu, J. R. 2002: Experimental archaeology. Replicating past objects, behaviours, and processes, BAR International Series, 1035. Nieuwenhuyse, O. P. 2008: Plain and Painted Pottery: The Rise of Late Neolithic Ceramic Styles on the Syrian and Northern Mesopotamian Plains, Papers on Archaeology of the Leiden Museum of Antiquities, 3, Turnhout, Brepols Publishers. Taranto S. (in press): “The role of the Husking Trays in the Late Neolithic communities of the Nort- hern Mesopotamia” , Proceedings of Broadening Horizons 5, June 5-8, 2017, In West & East. The Archaeological Journal of the Post Graduate School of Archaeology of the Universities of Trieste, Udine and Venezia Ca’ Foscari. Voigt, M. 1983: Hajji Firuz Tepe, Iran: The Neolithic Settlement, Philadelphia, University of Penn- sylvania Press.

7. For an explanation of the experimental tests related with the archaeological issue see Taranto (in press).

Butlletí Arqueològic, V, 40 (2018), ISSN 1695-5862 (p. 57-63) THE CASE OF THE “HUSKING TRAYS”: AN EXPERIMENTAL ANALYSIS OF A POTTERY TYPOLOGY 63

Figure 1. First baking experiment with Figure 2. Second baking experiment with domed oven. tabun-like oven.

Butlletí Arqueològic, V, 40 (2018), ISSN 1695-5862 (p. 57-63)