Barley

Joy McCorriston. Cambridge World History of Food. Editor: Kenneth F Kiple & Kriemhild Conee Ornelas. Volume 1. Cambridge, UK: Cambridge University Press, 2000.

That people do not live “by bread alone” is emphatically demonstrated by the domestication of a range of foodstuffs and the cultural diversity of food combinations and preparations. But even though many foods have been brought under human control, it was the domestication of cereals that marked the earliest transition to a food-producing way of life. Barley, one of the cereals to be domesticated, offered a versatile, hardy crop with an (eventual) tolerance for a wide range of climatic and ecological conditions. Once domesticated, barley also offered humans a wide range of valuable products and uses.

The origins of wheat and barley agriculture are to be found some 10,000 years ago in the ancient Near East. Cereal domestication was probably encouraged by significant climatic and environmental changes that occurred at the end of the glaciated Pleistocene period, and intensive harvesting and manipulation of wild cereals resulted in those morphological changes that today identify domesticated plants. Anthropologists and biologists continue to discuss the processes and causes of domestication, as we have done in this book’s chapter on wheat, and most of the arguments and issues covered there are not reviewed here. All experts agree, however, on the importance of interdisciplinary research and multiple lines of evidence in reconstructing the story of cereal domestication.

Readers of this chapter may note some close similarities to the evidence for wheat domestication and an overlap with several important archaeological sites. Nonetheless, barley has a different story to tell. Barley grains and plant fragments are regular components of almost all sites with any plant remains in the Near East, regardless of period or food-producing strategy. Wild barley thrives widely in the Near East today—on slopes, in lightly grazed and fired pastures, in scrub-oak clearings, in fields and field margins, and along roadsides. These circumstances suggest a different set of research questions about barley domestication, such as: What was barley used for? Was its domestication a unique event? And how long did barley domestication take?

In addition, there are subthemes to be considered. Some researchers, for example, have suggested that barley was not domesticated for the same reasons that led to the domestication of other cereals—such as the dwindling of other resources, seasonal shortages, a desire for a sedentary food base, or the need for a surplus for exchange. Instead, barley may have been cultivated for the brewing of ale or beer.

In another view, the very slight differences between the wild and domesticated forms, and the ease with which wild barley can be domesticated, make it difficult to believe that barley domestication did not occur more than once. Geneticists and agricultural historians have generally believed that groups of crops were domesticated in relatively small regions and spread by human migration and trade. If barley was domesticated independently in several different communities, this would indicate that the transition to farming in those areas required little innovation and took place under recurring conditions.

Finally, because of their presence in many excavated sites, ancient barleys provide some of the best archaeological evidence bearing on the general problem of the pace of plant domestication. Whether the process took place over the course of a single human lifetime or evolved over many decades or centuries remains an important issue that may never be resolved with archaeological evidence alone (Hillman and Davies 1990). But the pace of domestication lies at the heart of the debate over the Neolithic—was it revolution or evolution (Childe 1951; Rindos 1984)? Did plant domestication radically transform people’s lifestyles, or was it a gradual by-product of long-term behaviors with radical consequences noted only in retrospect? Barley is a crop that may hold answers to such questions.

Archaeological Evidence for the Domestication of Barley

Archaeological evidence points to the domestication of barley in concert with the emergence of Neolithic villages in the Levantine arc of the Fertile Crescent. Pre-Neolithic peoples, notably Natufian foragers (whose cultural remains include relatively large numbers of grinding stones, sickle blades, and storage pits), increasingly depended on plant foods and, perhaps, plant cultivation. Unfortunately, their tools point only to general plant use, and archaeologists continue to discuss which plants were actually processed (e.g., McCorriston 1994; Mason 1995). There is no evidence to suggest that the Natufians domesticated barley or any other plant.

Charred plant remains, the best evidence for domestication of specific plants, have rarely been recovered from pre-Neolithic sites. Preservation is generally poor. In the case of barley, only a few sites prior to the Neolithic contain recognizable fragments, and all examples indicate wild types. A few barley grains from Wadi Kubbaniya, an 18,000-year-old forager site in southern Egypt, were at first thought to be early examples of domesticated barley (Wendorf et al. 1979), but subsequent laboratory tests showed these to be modern grains that had intruded into ancient occupation layers (Stemler and Falk 1980; Wendorf et al. 1984). Other plant remains from Wadi Kubbaniya included relatively high numbers of wild Cyperus tubers and wild fruits and seeds (Hillman 1989).

One of the most extraordinary prefarming archaeological sites to be discovered in recent years is Ohalo II, on the shore of the Sea of Galilee, which yielded quantities of charred plant remains, including hundreds of wild barley grains (Kislev, Nadel, and Carmi 1992). About 19,000 years ago, foragers camped there, and the remains of their hearths and refuse pits came to light during a phase of very pronounced shoreline recession several years ago. Excavators believe that the charred plants found in the site were the remains of foods collected by Ohalo II’s Epi-Paleolithic foragers. If so, these foragers exploited wild barley (Hordeum spontaneum Koch.), which was ancestral to domesticated barley.

Despite the generally poor preservation of plant remains, there are two Natufian sites with evidence suggesting that foraging peoples there collected some wild barley just prior to the beginnings of agriculture. Wadi Hammeh, a 12,000-year-old Early Natufian hamlet overlooking the Jordan Valley (Edwards 1988), contained charred seeds of wild barley (Hordeum spontaneum), wild grasses, small legumes, crucifers, and a range of other plants, as yet unidentified (Colledge, in Edwards et al. 1988). The seeds were scattered among deposits in several round houses, somewhat like the scatter of plant remains at another Natufian site, Hayonim Cave, where Early and Late Natufian dwellers had constructed round houses, possibly seasonal dwellings, within the cave (Hopf and Bar Yosef 1987). To be certain that disturbances had not carried later seeds down into Natufian layers (a problem at Wadi Kubbaniya and at another Natufian site, Nahel Oren), excavators had the charred seeds individually dated. Wild lupines found with wild almonds, wild peas, and wild barley (Hordeum spontaneum) suggest that the Natufian inhabitants collected plants that could be stored for later consumption.

Although there is little doubt that some foraging groups collected wild barley, evidence for the beginnings of barley domestication is far more controversial. Until recently, archaeologists were convinced that farmers (as opposed to foragers) had occupied any site containing even a few barley grains or rachis fragments with the morphological characteristics of domestic cereals. Thus, the identification of tough-rachis barley in the Pre-Pottery Neolithic A (PPNA) levels at Jericho (Hopf 1983: 609) implied that the earliest Neolithic occupants domesticated barley in addition to wheats and legumes.1 A tough rachis inhibits seed dispersal, and wild barleys have brittle rachises that shatter when the seeds mature. Each segment of the rachis supports a cluster of three florets (flowers), from which only one grain develops. If a tough rachis fails to shatter, the seeds remain on the intact stalk, vulnerable to predators and unable to root and grow. Yet a tough-rachis crop is more easily and efficiently harvested by humans. Through human manipulation, the tough-rachis trait, which is maladaptive and scarce in the wild (Hillman and Davies 1990: 166-7), dominates and characterizes domesticated barley. At Jericho, the tough-rachis finds suggested to archaeologists that barley domestication had either preceded or accompanied the evident Neolithic practices of plant cultivation using floodwater manipulation in an oasis habitat.

However, at Netiv Hagdud, a contemporary (PPNA) site to the north of Jericho, Neolithic settlers seem to have practiced barley cultivation (Bar-Yosef et al. 1991), and the recovery there of rich archaeobotanical remains has forced archaeobotanists to rethink the significance of a few barley remains of the domesticated type in Early Neolithic sites throughout the Near East. Built on an alluvial fan in a setting not unlike that of Jericho, Netiv Hagdud proved to contain the foundations of almost a dozen large oval and small round structures. Some of these were probably houses with rock platform hearths and grinding equipment (Bar-Yosef et al. 1991: 408-11). The charred plant remains from the site included thousands of barley grains and rachis fragments, and the opportunity to examine these as an assemblage led to a surprising discovery.

Although a number of fragments clearly displayed a domesticated-type tough rachis (Kislev and Bar Yosef 1986), archaeobotanists realized that as an assemblage, the Netiv Hagdud barley most closely resembles modern wild barley. Even among a stand of wild barley, approximately 12 percent of the spikes have a tough rachis (Zohary and Hopf 1993: 62) because this characteristic regularly appears as the result of mutation and self-fertilization to generate a pair of recessive alleles in a gene (Hillman and Davies 1990: 168). At Netiv Hagdud, the thousands of barley remains (including low numbers of tough-rachis examples) were collected, possibly from cultivated stands, by Early Neolithic people who appear from available evidence to have had no domesticated crops, although some scholars suggest that harvesting timing and techniques might still make it difficult to distinguish between wild and semidomesticated barley (Kislev 1992; Zohary 1992; Bar-Yosef and Meadow 1995). That evidence also implies that occasional domesticated-type barley remains at other Neolithic sites, including Tell Aswad (van Zeist and Bakker Heeres 1982), may actually belong to an assemblage of purely wild barley.

Other Early Neolithic sites with remains of barley include Gilgal I, a PPNA site with cultural remains similar to those at Netiv Hagdud and Jericho. It is still unclear whether the “large amounts of oat and barley seeds” recovered in the mid-1980s from a silo in one of the Neolithic houses were domesticated or wild types (Noy 1989: 13). Tell Aswad, formerly nearer to Mediterranean oak forests and set beside a marshy lakeshore in the Damascus Basin, has also yielded a mix of rachis fragments, predominantly wild-type but with some domesticated-type (van Zeist and Bakker Heeres 1982: 201-4). This same mix of wild and domesticated types was found in early levels at other early sites, including Ganj Dareh on the eastern margins of the Fertile Crescent (van Zeist et al. 1984: 219).

Over time, the percentages of wild-type and domesticated-type barley remains were inverted at sites in the Damascus Basin (van Zeist et al. 1984: 204). Tell Aswad was the earliest occupied of these sites in a 2,000-year sequence of nearly continuous residence there (Contenson 1985), and when it was abandoned, farmers had already settled nearby, at Tell Ghoraifé. Although the barley remains at Ghoraifé included many domesticated-type specimens, it was from evidence of a later occupation at Tell Aswad (8,500 to 8,900 years ago) and the nearby site of Ramad (occupied from 8,200 years ago) that archaeobotanists could unequivocally distinguish between fully domesticated barley and low numbers of wild barley in the same assemblages. The archaeo-botanists suggested that the apparent mix of wild and domesticated types in earlier deposits may indicate an intermediate stage in the domestication process (van Zeist and Bakker-Heeres 1982: 184-5, 201-4).

There has never been much doubt that remains of barley from northern Levantine sites indicate plant collection or incipient stages of cultivation without domestication 10,000 years ago. Wild barley (Hordeum spontaneum Koch.) is well represented among the plant remains from Mureybit, an Early Neolithic village on the Euphrates River in the Syrian steppe (van Zeist and Bakker-Heeres 1984a: 171). Contemporary levels at Qeremez Dere, also in the steppe, contain abundant remains of fragmented wild grass grains, some of which “seem most probably to be wild barley” (Nesbitt, in Watkins, Baird, and Betts 1989: 21). Plant remains suggest that cereal cultivation was of little importance as one moved further north and east during the first centuries of Neolithic occupation. Cereals were present only as traces in the northern Iraq site of Nemrik 9 along the Tigris River (Nesbitt, in Kozlowski 1989: 30), at M’lefaat (Nesbitt and Watkins 1995: 11), and not at all at Hallan Çemi (Rosenberg et al. 1995) nor at earliest Çayönü (van Zeist 1972).

Available evidence now seems to suggest that domesticated barley appeared in the second phase of the early Neolithic—Pre-Pottery Neolithic B (PPNB)—several hundred years after wheat farming was established.2 By the PPNB (beginning around 9,200 years ago), several different forms of domesticated barley, two-row and six-row (see the section “Taxonomy”), appeared among plant remains from Neolithic sites. Jericho and Ramad had both forms from about 9,000 years ago (van Zeist and Bakker-Heeres 1982: 183; Hopf 1983: 609). Barley does not seem to have been among the earliest crops in the Taurus Mountains—at PPNB sites such as Çayönü (van Zeist 1972; Stewart 1976) and Çafar Höyük (Moulins 1993). At Neolithic Damishliyya, in northern Syria (from 8,000 years ago), and Ras Shamra on the Mediterranean coast (from 8,500 years ago), domesticated two-row barley was present (van Zeist and Bakker-Heeres 1984b: 151, 159; Akkermans 1989: 128, 1991: 124). In Anatolia, domesticated barley also characterized the first appearance of farming, for example, at Çatal Hüyük (Helbaek 1964b). This is also the case in the eastern Fertile Crescent, where domesticated plants, barley among them, first showed up around 9,000 years ago—at Jarmo (Helbaek 1960: 108-9) and Ali Kosh (Helbaek 1969).

An excellent review of archaeobotanical remains (Zohary and Hopf 1993: 63-4) tracks the subsequent spread of two-row and six-row barley around the Mediterranean coast, across the Balkans, and into temperate Europe. Barley was one of the fundamental components of the Neolithic economic package introduced (and modified) with the spread of Near Eastern farmers across Anatolia and into various environmental zones of Europe (Bogucki 1996), as well as into northern Egypt (Wetterstrom 1993), Central Asia (Charles and Hillman 1992; Harris et al. 1993), and South Asia. By 8,000 years ago, barley agriculture had reached the foothills of the Indus Valley, where it supported farmers at Mehrgarh, one of the earliest settlements in South Asia (Jarrige and Meadow 1980; Costantini 1984).

During these first several thousand years, domesticated barley also spread into the steppes and desert margins of the Near East, expanding farming practices into an ecological niche where risk was high and barley offered what was probably the best chance for crop survival. At Abu Hureyra, in the Syrian steppe near Mureybit, six-row barley appeared earlier than its two-row barley ancestor (Hillman 1975), suggesting that fully domesticated barley crops were introduced to that site (Hillman and Davies 1990: 206), although recent paleoecological models suggest that wild barley grew nearby (Hillman 1996).

By the end of the PPNB, sites with barley were fairly common in very marginal farming zones—including, for example, El Koum 2 (Moulins, in Stordeur 1989: 108), Bouqras (van Zeist and Water-bolk-van Rooijen 1985), and Beidha (Helbaek 1966). At Nahal Hemar cave, in the dry Judean Hills, a single kernel of domesticated barley recovered was presumably imported to the site some 9,000 years ago (Kislev 1988: 77, 80). The spread of barley not only indicates the success of a farming way of life but also offers important archaeological indicators to complement botanical and ecological evidence of the domestication and significance of this vital crop plant.

Botanical Evidence for the Domestication of Barley

Barley grains and barley rachis internodes recovered from archaeological sites show the telling morphological characteristics of wild and domesticated forms, but botanical studies of modern barleys have also provided critical evidence for an interdisciplinary reconstruction of barley domestication. Archaeologists can examine ancient morphology but not ancient plant behavior. It was changes in both characteristics that established barley as a domesticated plant. Not only did the rachis become tough, but the ripening period of barley narrowed, and a single-season seed dormancy became established. Botanical studies have revealed relationships between species and varieties of barley and the precise nature of the changes that must have occurred under domestication.

Taxonomy

As with wheats, taxonomic classification of barleys has changed with expanding scientific knowledge of genetic relationships. Genetic evidence now indicates much closer relationships among what morphologically once appeared to be distinctive species (Briggs 1978: 77). Yet it is the morphological criteria, easily seen, that offer farmers and archaeologists a ready means by which to classify barleys (but see Hillman and Davies 1990, and Hillman et al. 1993, for alternative experimental approaches). Because archaeologists have had to rely largely on morphological criteria to detect the beginnings of domestication, the old species names remain convenient terms for distinguishing between what are now considered barley varieties (new species names in parentheses follow).

Barleys belong in the grass tribe Triticeae, to which wheats and ryes (barley’s closest crop relatives) also belong. There are 31 barley species (almost all wild), and nearly three-fourths of them are perennial grasses (Bothmer and Jacobsen 1985; Nilan and Ullrich 1993). Despite the diversity of wild barleys that can be identified today, most botanists and geneticists concur that all domesticated types most probably have a single wild ancestor, Hordeum spontaneum Koch. (H. vulgaresubsp. spontaneum) (Harlan and Zohary 1966; Zohary 1969).This plant crosses easily with all domesticated barleys. The major morphological difference between wild and domesticated barley lies in the development of a tough rachis in the domesticate.

Once farmers had acquired a domesticated tworow form of barley, they selectively favored the propagation of a further morphological variant, the six-row form. Barleys have three flowers (florets) on each rachis segment (node). In the wild and domesticated two-row forms, however, only the central floret develops a grain. Thus, one grain develops on each side of a rachis, giving the spike the appearance of two grains per row when viewed from the side. In the six-row form, Hordeum hexastichum L. (H. vulgare subsp. vulgare), the infertility of the lateral florets is overcome: Nodes now bear three grains each, so the spike has three grains on each side. This gives an appearance of six grains per row in side view. A general evolutionary trend in the grass family has been the reduction of reproductive parts; so, for a long time, it was difficult for botanists to accept that one of the consequences of domestication and manipulation of barley has been to restore fertility in lateral spikelets, thereby increasing the grain production of each plant (Harlan 1968: 10).

A final important morphological change was the appearance of naked-grain barleys. In wild cereals, the modified seed leaves (glumes, lemmas, and paleas) typically tightly enclose the grain and form a protective husk. From a human perspective, one of the most attractive changes in domesticated cereals is the development of grains from which the glumes, lemmas, and paleas easily fall away. Because humans cannot digest the cellulose in the husks, such development considerably reduces processing effort. Naked-grain barleys (Hordeum vulgare subsp. distichum var. nudum and H. vulgare subsp. vulgare var. nudum) appeared shortly after the emergence of sixrow forms (Zohary and Hopf 1993: 63). Taxonomists have always recognized these as varieties rather than as distinct species of barley.

Genetics

Genetic relationships remain very close among all barleys, and modern taxonomic schemes collapse all cultivated barleys and the wild Hordeum spontaneum ancestor into a single species, Hordeum vulgare (Harlan and Zohary 1966: 1075; Briggs 1978: 78; Nilan and Ullrich 1993: 3). H. vulgare is a diploid with two sets of seven chromosomes (2 n = 14) and has proved an excellent subject for genetic and cytogenetic analysis (Nilan and Ullrich 1993: 8). Because the plant is self-fertile (that is, male pollen fertilizes its own or adjacent flowers on the same plant), mutations have a good chance of being copied and expressed in the genes of subsequent generations. This attribute was undoubtedly an important feature of barley domestication, for a very few mutations have caused major morphological changes that were easily favored by humans, both consciously and unconsciously (Harlan 1976; Hillman and Davies 1990).

A brittle rachis, for example, is controlled by a pair of tightly linked genes. A mutant recessive allele in either gene (Bt and Bt1) will produce a tough rachis in homozygous offspring (Harlan 1976: 94; Briggs 1978: 85). This condition will occur rarely in the wild but may be quickly selected for and fixed in a population under cultivation (Hillman and Davies 1990: 166-8). Experimental trials and computer simulations suggest that under specific selective conditions, the homozygous recessive genotype may become predominant in as few as 20 years (Hillman and Davies 1990: 189). Consequently, barley domestication depends on one mutation!

Furthermore, a single recessive mutation also is responsible for fertility in lateral florets and the conversion from two-row to six-row forms (Harlan 1976: 94). Another gene, also affected by a recessive mutant allele, controls the adherence of lemma and palea to the grain. Jack Harlan (1976: 95-6) has postulated a single domestication of wild barley followed by other recessive mutants for six-row and naked forms.

Objections to this parsimonious reconstruction revolve around the many brittle or semibrittle rachis variants of barley, some of which include six-row forms. Barley is rich in natural variants in the wild (Nevo et al. 1979; Nilan and Ullrich 1993: 9), and geneticists have long tried to incorporate six-row brittle forms (e.g., Hordeum agriocrithon Åberg) into an evolutionary taxonomy of the barleys. Most now agree that the minor genetic differences, wide genetic diversity, and ease of hybridization and introgression with domesticated forms account for the great number of varieties encountered in the “wild” (Zohary 1964; Nilan and Ullrich 1993: 3).

Ecological Evidence for Barley Domestication

 

Geographic Distribution

In the tradition of Nikolay Ivanovich Vavilov, geneticists and botanists have documented the distributions of wild and domesticated varieties of barley. The places where wild races ancestral to domesticated crops grow today may indicate the range within which a crop arose, because the earliest cultivators must have encountered the plant in its natural habitat. Patterns of genetic diversity in different areas may also offer clues about the history of a crop plant. Such patterns may suggest, for example, a long and intensive history of manipulation or an isolated strain. Harlan and Daniel Zohary (1966: 1075-7) have summarized the modern distribution of wild barley, Hordeum spontaneum, noting many populations outside the core area of the Fertile Crescent where the earliest agricultural villages lay.

Harlan and Zohary distinguish between truly wild barley (H. spontaneum) and weedy races—wild-type barleys derived from domesticated barley crops in areas to which barley farming spread after domestication. Modern distribution of truly wild progenitors is closely associated with the geography of semiarid Mediterranean climates and with an ecological relationship with deciduous oak open woodland that covers the lower slopes of the mountain arc of the Fertile Crescent.This landscape was made famous by Robert J. Braidwood’s expedition to the “Hilly Flanks,” where he and later archaeologists sought to uncover the first farming villages (Braidwood and Howe 1960: 3). A “small, slender, very grassy type” grows wild in steppic environments with somewhat greater temperature extremes and less annual rainfall. Another distinct truly wild race in the southern Levant has extremely large seeds (Harlan and Zohary 1966: 1078).

In areas outside the semiarid Mediterranean woodlands, wild-type barleys survive in a more continental climate (hotter summers, colder winters, year-round rainfall). Because all the collections in such areas have proved to be weedy races (including brittle-rachis, sixrow Hordeum agriocrithon Åberg from Tibet), their range provides better information on the spread of barley farming than on the original domestication of the plant.

Ecological Factors

Ecologically, wild barley shares some of the preferences of wheat, but wild barley has not only a much more extensive geographical range but also a wider ecological tolerance. Barleys thrive on nitrogen-poor soils, and their initial cultivation must have excluded dump-heap areas enriched by human and animal fertilizers (Hillman and Davies 1990: 159). But the wild barley progenitors do thrive in a variety of disturbed habitats (Zohary 1964). In prime locales, wild barley flourishes on scree slopes of rolling park-woodlands. It likes disturbed ground—in fields and along roadsides—and is a moderately aggressive fire follower (Naveh 1974, 1984).

Ecological and botanical attributes of wild barley have convinced some that it was the first domesticate (Bar-Yosef and Kislev 1989: 640). Archaeological evidence, however, indicates that despite possible cultivation in the PPNA (Hillman and Davies 1990: 200; Bar-Yosef et al. 1991), barley was not domesticated as early as wheat and some legumes. From the perspective of cultivators, wild wheats have several advantages over wild barley, including greater yield for harvesting time (Ladizinsky 1975) and easy detachment of lemma and palea. Modern wild barleys demonstrate a number of features potentially attractive to foraging peoples, including large grain size, ease of mutant fixation, local abundance, and wide soil and climate tolerance (Bar-Yosef and Kislev 1989: 640). Nevertheless, the archaeological record holds earlier evidence of domesticated wheat than of domesticated barley.

Was wheat domestication fast and that of barley slow? Or were these cereals domesticated at the same rate but at different times? Experimental studies offer significant insights into the archaeological record of cereal domestication, including probable causes of ambiguity where wild forms may or may not have been cultivated (Hillman and Davies 1990; Anderson-Gerfaud, Deraprahamian, and Willcox 1991). Although barley domestication can happen very quickly, the rate of domestication would have varied according to planting and harvesting conditions. It would be nearly impossible to discriminate between collection and cultivation (reseeding) of wild barley in the archaeo-logical record; therefore, it is difficult to know whether barley was cultivated for a long time before becoming a recognizable domesticate. The excellent work of Gordon Hillman and Stuart Davies (1990) with mutation rates, cultivation variables, and harvest strategies suggests little inherent difference between wheat and barley plants for their domesticability. Perhaps the very different archaeological record, with a PPNA emergence of domesticated wheat and a later PPNB emergence of domesticated barley, implies that we should look beyond the genetics and ecology of the plants for other variables in the early history of agriculture.

Uses of Barley and Barley Domestication

Today, barley is primarily important for animal feed, secondarily for brewing beer, and only marginally important as a human food. Although researchers typically assume that cereals were first domesticated as foodstuffs, we do not know what prominence, if any, barley had in early agriculturalists’ diets. In its earliest, most primitive, hulled form, barley required extra processing to remove lemma and palea. Once naked barleys appeared, one suspects that they were preferred as human food, but one cannot conclude that hulled barleys were domesticated for animal feed. Although the domestication of barley coincided with the appearance of domesticated animals in the Levant, the Natufian and PPNA evidence clearly indicates harvest and, perhaps, cultivation of wild barley before animal domestication. Some quality of barley attracted cultivators before they needed animal feed.

Solomon Katz and Mary Voigt (1986) have hypothesized that barley domestication was a consequence of early beer brewing. They suspect that epi-Paleolithic peoples intensively cultivated wild barley because they had come to understand its use in fermentation and the production of alcohol, and it was this use that prompted the advent of Neolithic farming. Their theory implies that epi-Paleolithic peoples were sedentary, as Natufians apparently were (Tchernov 1991). Beer brewers must also have possessed pottery or other suitable containers, yet the invention of pottery took place long after cereal domestication in the Near East. And if cereal (barley) domestication was brought about by demand for beer, then domestication probably was impelled by social relationships cemented by alcohol rather than by subsistence values of cereal grains. The social context of drinking has been explored recently by a number of anthropologists (e.g., Moore 1989; Dietler 1990) who have emphasized the important roles that alcohol and other costly perishables play in social relationships, especially in matters of reciprocity and obligation.

Perhaps one of the most significant insights into the theory that a fermented beverage (rather than a nutritive grain) impelled domestication lies in the methods by which early beer was made. Recipes on Mesopotamian clay tablets and iconographic documentation clearly indicate that early beer was made from bread (Katz and Maytag 1991) rather than from malt (sprouted grain), as in modern practice. Both bread making and malting produce a fermentation material in which the cereal-grain endosperm has already been partially broken down mechanically and chemically (Hough 1985: 4-5). Archaeologists have detected residues consistent with beer making on ceramics from as early as the late fourth millennium B.C. (Michel, McGovern, and Badler 1992). If this Sumerian beer-making tradition developed from early antiquity, the most parsimonious theory is that beer-making developed from fermented bread.

But even if barley was first cultivated for its food value (perhaps as grits or gruel), it clearly offered an important array of products to farmers. In addition to grain feed, beer, and bread, barley also yielded straw for fodder, thatch, basketry, mudbrick, and pottery temper. Many of these products must have been essential as barley farming spread to arid regions all but devoid of trees, wood, and lush, wild vegetation.

The Spread of Barley Farming

Perhaps the most significant expansion of domesticated barley was into the truly arid steppes and deserts of the Near East, where irrigation was critical to its survival. Hans Helbaek (1964a: 47) has argued that barley was irrigated by the occupants of Tell es Sawwan in the early fifth millennium B.C. Six-row barley was evident among the site’s archaeological plant remains, yet with the available rainfall, plants producing even a third as much seed (such as the two-row forms) would have been hard-pressed to survive; the six-row form could have thrived only under irrigation. Six-row barley was one of the principal crops in ancient southern Mesopotamia, and some have even suggested that barley production lay at the heart of the rise of the earliest truly complex societies in a river-watered desert.

Barley farming has also expanded into temperate regions of China, and into the tropics, where dry and cool highland regions (such as in Ethiopia, Yemen, and Peru) offer appropriate locales for cultivation. Christopher Columbus’s voyages first brought the crop to the New World (Wiebe 1968), where it spread most successfully in North America (Harlan 1976: 96).

Conclusion

Domesticated barley was an important crop in human prehistory and provided many products to a wide range of settled peoples. Barley has long been considered one of the initial domesticates in the Southwest Asian Neolithic package of cereals and legumes, and in a broad chronological sweep, this conclusion remains true. But recent archaeological and botanical research indicates that domesticated barley did not appear among the very first domesticated plants. With a growing corpus of plant remains, more attention can be paid to regional variation in the Southwest Asian Neolithic, and archaeologists can now develop a more complex understanding of early plant domestication. Just as there seems to be no barley in PPNA agriculture, barley also seems to be absent from early PPNB farming communities in the Taurus Mountains. Archaeologists have long recognized other late domesticates, such as grapes, olives, and other fruits. The progress of barley domestication, along with its origins, offer potentially interesting insights into the development of domestic lifestyles and the expansion and adoption of farming in new ecological zones.

One explanation for different timing of wheat and barley domestication might be found in the possibility that differences in cultivation practices led to differences in cereal domestication. Modern studies suggest that domestication rates should be similar for wheat and barley, but they also demonstrate that different practices—tending cereals on the same or on new plots each year, for example—can affect domestication rates. An archaeological record implying a long period of cultivation for barley prompts us to wonder if cultivators treated wheats and barleys differently. Did they value one cereal more than another? Can we use such evidence from different regions and different crops to understand the significance of different plants in the diets and lives of the earliest cultivators and farmers?

Increasing botanical and ecological knowledge of cereals will help us address such questions. It may be that differences between wheat and barley domestication are related to the ease with which backcrossing between domesticated and wild cereal plants reintroduces wild traits in generations of barley crops. Ongoing experiments in cereal domestication will provide important information.

Ultimately, our reconstruction of barley domestication, and of its prehistoric importance in human diet and nutrition, depends on interdisciplinary research—the combination of archaeological evidence with botanical, ecological, and experimental evidence. There will always be uncertainties. Archaeological sites have been excavated and analyzed by different researchers practicing different methods and taking plant samples of differing quantity and quality. Modern distributions and plant ecology are the result of historical, environmental, and climatic changes, and ecologists and botanists can only guess in what ways these have affected plant geography. Nevertheless, as more archaeological and botanical evidence emerges, some of these uncertainties may be more conclusively addressed and the process of barley domestication more fully understood.