The emergence of cooking in Southwest Asia

There has been surprisingly little systematic study by prehisto­ rians of how in the distant past people cooked and consumed food. There are many unanswered questions. For example, how did cooking emerge and affect human evolution, how did it change with the advent of farming, when did kitchens first appear and who built the earliest known ovens? Research on Palaeolithic and Neolithic food preparation and consumption is now beginning to suggest answers to such questions.

C ooking and dining, together with the ability to create and use symbols, distinguish human beings from other creatures, but little is known about the earliest technology and social behaviour associ ated with these activities. The research described here is part of a project con cerned with the evolution and social sig nificance of early prehistoric cooking and dining technologies in Southwest Asia. My interest in this topic grew out of earlier research on common but little studied food-processing artefacts such as ground stone pounding and milling tools, and now encompasses a wider range of artefacts and features associated with food preparation, especially hearths and other fire features.
What is cooking?
The anthropologist Claude Levi-Strauss asserted that human beings transform nature into culture by cooking: raw food is nature, cooked food culture. 2 Cooking opens up unlimited possibilities for new cultural rules. Furthermore, dining turns culture (food) into social life (meals), whereas cuisine changes meals into feasts, social life into hierarchy.3 Cooking is defined here as any tradition in which heat is used, at least sometimes, in food prepa ration. Thus, it implies the use of any com bination of food-preparation techniques, provided that one of them is heating. In this article the focus is on techniques used rel atively close to the time and place of con sumption, so some activities involved in producing food, such as harvesting and winnowing grain, are not included. Tech niques of food preparation vary enor mously, but most can be grouped into four broad types of procedure: changing the physical structure of a fo od; using liquids but not heat; drying and applying dry heat; and applying liquids and heat together.4 food preparation need not have been used only for foods but are likely to have served other purposes too, such as heating and craft production.  Palaeolithic features are often simple flat fireplaces or firepits, but occasionally they are pit hearths with stone borders at the top or rim, or with fills containing small peb bles and fire-cracked rocks. Pebble borders and fi lls could have been used for blocking drafts, for supporting items to be heated, or for heating pebbles and moving them to containers or even the insides of whole animals. 8 There is little dramatic change in the basic technologies of groundstone arte facts and hearth construction from the Middle to the Upper Palaeolithic, but these artefacts and features appear more often and in larger numbers. The range of varia tion in the physical construction of indi vidual hearths in the Upper Palaeolithic is somewhat greater than in the Middle Palaeolithic, from fire patches to stone-bor dered or pebble-filled pit hearths (as at the c. 22,000-year-old site of Uwaynid 18 in Jordan, Fig. 1).9 In the Upper Palaeolithic there are indi cations that, at some sites, social groups larger than those of the Middle Palaeolithic may have engaged in cooking at the same time. At these sites, carefully arranged clusters or lines of fire features are fre quently present. For example, at Abu Noshra I in the southern Sinai, there are four stone-bordered fireplaces and pit hearths and one very large firepit that con tained only bones of wild ass, which is interpreted as evidence for the animal hav ing been roasted whole. 10 These patterns hint at more structured cooperation and coordination of cooking and food sharing by larger groups, which would have encouraged more rapid trans mission of cultural information. We can not be certain that multiple hearths were used contemporaneously. But even if they were established at different times, by the end of these occupations hunter-gatherers would have had available to them a larger set of on-site cooking facilities than we see in most Middle Palaeolithic camp sites. People living at sites with multiple hearths would have been able to produce more pre pared food at one time. Thus, the evidence from Southwest Asia possibly supports the idea that large-scale cooking (or feasting)11 first took place at such Upper Palaeolithic sitesY Arrangements for cooking appear to Archaeological analysis offood customs involves the examination of the technolog ical capacities of relevant artefacts and fea tures such as stone tools and hearths; the spatial organization of food-preparation activities in archaeological sites; plant and animal remains; and human skeletons. It is essential to analyze variations in the con struction and form of fire features. 5 It is also important to remember that the func tions of tools and features associated with ,·-. . Th e black features consist of a rectangular hearth in the centre; a large boulder mortar south of the hearth, with its circular use-surfa ce shown in black; ground stone artefacts, mainly pestles, west of the hearth and boulder; and a burial just outside the structure on the eastern side.
become more spatially structured in Early Epipalaeolithic sites between 21,750 and 12,750 BC. The most striking technological change is the appearance of mortars with deep grinding holes and elongated pes tles.13 Thus, at Ein Gev I in Israel, evidence was found of a hut that had been repeatedly occupied, with a simple hearth in the mid dle of the floor and two pestles and a large stone mortar close by, evidently cached there as site furniture.14 Similar features, including a stone worktable, were found at Ohalo n in Israel. 15 In these sites, as in ear lier ones, the features and artefacts associ ated with cooking and dining that have been preserved look remarkably utilitar ian. There is no indication that they were decorated or otherwise constructed for purposes of conspicuous display. This situation changes at the beginning of the Natufian (Late Epipalaeolithic) period about 12,750 BC.

Dining and social bonds in the Natufian
In the Natufian period between 12,750 and 10,050 BC, hunter-gatherers inhabited base camps, characterized archaeologically by groups of stone houses (Fig. 2)/6 and also burials, groundstone tools, rock-cut basins, pits, and the remains of plants (some of which may have been cultivated) and ani mals, all of which suggests that sites were occupied year-round and that the people enjoyed diverse diets. Many mortars, pestles, grinding tools and stone vessels have been found in the base camps. The stone artefacts are strik ingly similar from one base camp to another in the frequencies of different typesY They were labour intensive to produce, curated (i.e. carefully kept and stored) and sometimes decorated with carvings or paint. Similarly, Natufian hearths are more elaborate than any earlier ones, and they include pits carefully lined with flat stone slabs.18 Overall, the Late Epi palaeolithic ground stone technology suggests a new formality in food sharing and an element of social ritual surrounding it (Fig. 3). What under lay this development? Most scholars agree that Natufian base camps represent settle ments that were occupied year-round, while foraging territories were becoming quite restricted.19 This situation would have interfered with the classic means of conflict resolution practised by mobile hunter-gatherers, namely fission (i.e. the departure of one party to the conflict to other foraging territories). In short, because Natufian groups were more sedentary than earlier hunter-gatherers, they had to find other means of social integration. It seems likely that social rules surrounding food consumption grew more formalized in order to enhance social cohesion.
Feasting may have been part of this pro cess, although we have no direct evidence for it. Settings for eating and drinking undoubtedly varied according to season and occasion, and areas where food was prepared have been found in both houses and outdoor contexts. Because base camps were probably occupied year-round, we can infer that in winter or bad weather some cooking and dining took place under shelter. Natufian houses never exceeded 28m2 in area, so indoor processing and cooking could have involved only a few individuals. Sets of mortars and pestles found in pairs in caches suggest either that two people were involved in pounding and mixing foodstuffs or that pounded foods were deemed to require two separate sets. 20 Simple tastes: cooking and dining in the PPNA By the Pre-Pottery Neolithic A (PPNA) period (10,050-9000 BC), there is evidence for domesticated barley and wheat at a few sites, indicating that the transition from foraging to farming was under way. With rare exceptions, PPNA groundstone arte facts utterly lack the diversity, workman ship and decoration that characterized the Natufian. The same is true of hearths, which were often merely areas of ash or burnt stones. Clearly, a different set of atti tudes to food and dining were at work. The PPNA villagers seem to have had no interest in conspicuous displays in preparing and serving food. Despite the profound eco nomic changes that accompanied the so-called agricultural revolution, cultural practices surrounding meals seem to have been remarkably undramatic.
There are at least three possible reasons for this: the medium of decoration or dis play may have been perishable materials rather than stone; other items of material culture, such as figurines or human skulls, may have been used to convey social mes sages at meals; simplicity may have been regarded as the proper aesthetic of cooking and dining. But storage features are likely to be a partial exception to the pattern. Stone-lined bins, which are relatively rare in Natufian sites, are considerably more frequent in PPNA sites.21 Kitchen and cupboard: the Pre Pottery Neolithic B Areas of food-related activity were rela tively unspecialized in the Natufian and the PPNA, and boundaries between houses and communal spaces seem to have been fluid. During the two millennia ofthe PPNB (9000-6950 BC) this situation changed pro foundly. In the Early and Middle PPNB there is evidence that food preparation was governed by much more structured spatial rules. Milling, cooking and storage now took place in areas near house entrances, a border zone between community space and the individual household (e.g. Fig.   4). 22 These activities became highly visible and public, affording opportunities for social contacts between households. Nev ertheless, individual households appear to have continued to control their own facil ities for food preparation and storage, and meals were probably centred on the house hold. In the Late PPNB, intensification and privatization of storage, milling, cooking and dining seem to have taken place, and these activities became more secluded from the village as a whole (Fig. 5).
The PPNB as a whole was a period of technological innovation in food prepa ration. 2 3 Detailed studies of PPNB ground stone artefacts show that these tools are more diverse in form than the PPNA ver sions. Large milling tools (grinding slabs and handstones) dominate PPNB food processing equipment (Fig. 6). Many of them are considerably larger than their PPNA counterparts and would have permit ted cooks to process more food in a given operation. Many grinding slabs were so large and heavy that they were essentially immovable. For example, the mean weight of 26 complete grinding slabs from the village site of Beidha in southern Jordan was 26.74kg; some slabs weighed as much as 52 kg, and some slabs were set into floors as fixed features. The handstones used with these slabs vary from small disc shape tools that could be operated with one hand, to large oval and loaf-shape hands tones that demanded the use of two hands and weighed up to 2. 5 kg. PPNB groundstone tools also include small mor tars and pestles, and limestone pebbles with cup-holes (often with carbon residues on the interior and possibly used as lamps).
Vessels, usually made of limestone but sometimes of basalt, are of much finer workmanship and are more diverse in size and shape than those of the PPNA. They are simple, well made and usually undeco rated (Fig. 7). They are typically simple bowls or platters that lack spouts, handles or lids. Platters are a PPNB innovation and they are the most abundant type of vessel, especially in the Middle and Late PPNB. At Beidha, 34 of the 73 stone vessels found were platters. They are large and shallow, oval or rectangular in plan, often with thin walls. They range in diameter from 30 cm to 1 m and could hold more food at one time than most PPNA bowls. The smaller platters would have been portable, but those from Beidha weigh between 5 kg and 10kg and are unlikely to have been trans ported very far, if a t all. A few of the platters at Beidha had traces of burning on the exte rior base and were probably used in cook ing. Other PPNB vessels were made of cordage, basketry (sometimes with water proof asphalt linings). wood, plaster and early versions of pottery.
In the Late PPNB, some villages grew to unprecedented sizes of 12-15 ha. Houses, often with two storeys and complex plans, reached 160m2 in area, as much as four times larger than those of the Early PPNB. 24 There is also evidence of larger assem blages of milling tools per house and of more specialized cooking facilities such as closed ovens, mealing bins (pits and stone installations designed to hold grinding slabs in place); plaster vessels and experi ments with pottery making. Food prepara tion within houses was more secluded and sometimes took place in specialized rooms that functioned as kitchens. Storage facil ities become larger and more elaborate, and sometimes occupied whole rooms. All these changes testify to intensified produc tion of prepared fo ods for larger groups of people. The houses were larger and more complex and there was an increasing emphasis on privacy, including the con cept of private property.

Conclusion
Our knowledge of the technologies and social and cultural dimensions of food preparation in early prehistoric times is still very limited. But the evidence briefly described here, especially that derived from study of the groundstone artefacts, shows that significant changes in cooking and dining took place in Southwest Asia through the many millennia from the Pal aeolithic to about 7000 BCbefore the development and spread of pottery pro duction later in the Neolithic added a rev olutionary new technology to the means available for storing, preparing and serving fo od. 5. Many archaeologists use the term "hearth" to refer to any evidence for repeated use of fire, but it is more helpful to distinguish between ancient fire fe a tures according to their construction and the frequency of their use. Seven types of fire feature can thus be distinguished: • firepatches : flat patches of fired earth representing single burning events or lacking evidence of repeated burning • fireplaces: flat patches of earth, repeat edly burned, with burnt soil and ash, but no stones on them • firepits: hollowed out depressions with burnt deposits and ash, but no stones in the fill or at the edges • stone-bordered fireplaces: repeatedly burned fireplaces with stones marking the edges • stone-filled pit hearths: firepits with stones or pebbles inside them for distrib uting heat or supporting items to be heated • stone-bordered pit hearths: firepits with stones around the upper edges • stone-lined pit hearths: firepits with stone slabs lining the whole interior. China (Zhoukoutien] , Hungary (Vertesszi:illi:is) and Britain (Beeches Pit). But all these tantalizing hints are ambig uous: the fire features and burnt material are not all in situ and some cannot defi nitely be linked to human activity. In Southwest Asia, no firm evidence for fire control has been found at Lower Palaeo lithic sites. 7. Early fireplaces have been discovered at Middle Palaeolithic sites in France, Israel, Jordan, Lebanon, Syria, Turkey, Iraq and Iran; see for example S. James, "Hominid use of fire in the Lower and Middle Pleis