Food and Environment in Early and Medieval China. E. N. Anderson

to Uralic and Altaic peoples and to some settled Iranic-speakers. The cult is clearly continuous with Chinese reverence for trees; the same ideas and behaviors are visible.

      Agriculture flourishes in Ukraine and in river valleys and montane out-wash fans throughout inner Asia, but full steppe conditions are impossible for agriculture. They are, however, ideal for herding the hardier kinds of stock: sheep, goats, and horses. The riverine zones along the Tarim, Syr Darya, Amu Darya, and other rivers were once among the most agriculturally productive tracts of land on earth—grain, forage crops, fruit trees, vegetables, and other crops (including early cotton) flourished. In recent decades, however, pollution, salt buildup in the soil, monocropping (especially cotton), urban sprawl, and other features of extremely bad land management have ruined much of the land.

      The existence of extremely rich zones near vast tracts of nomadic herding country was an invitation for trouble. The steppe nomads could raise huge mobile forces and descend on the cities and farms, especially when warm and moist climatic periods allowed the nomads to increase both human and animal populations. Then the nomad leaders settle in the cities, succumb to luxury, lose their martial ability, and the whole cycle starts over again—as pointed out by the great Arab social scientist Ibn Khaldun in the fourteenth century.

      This sequence is complicated by the fact that steppe nomads were never independent of settled people (Barfield 1989; Khazanov 1984). They required grain to supplement the products of their herds. They produced felt and wool goods but depended on settled people for other fabrics. They needed more metal than they could produce themselves. Metal goods—especially gold—became major wealth and show items. (Stock-herders who could produce all their own food and everyday goods existed in Arabia and Africa, but could not do so in Central Asia, where at least some grain, clothing, metal, and the like had to be bought.)

      At the margins of the steppes, farming people encroached during warmer, moister periods. Since these are also the periods when steppe populations were increasing, tensions naturally arose. The infamous “barbarians” that harassed the Roman Empire rode out during such a time: the favorable climatic period in the early centuries of the Common Era. So did Mongol hordes during the Medieval Warm Period a few centuries later. Cold periods, by contrast, were deadly. Late winter and early spring storms dropped deep snow or, worse, ice over the young grass, starving the herds just when they needed feed the most. The old myth that “droughts” forced the nomads out on raids is long dead; droughts kept the nomads at home, scrabbling hard to survive, with no strength to raid. It was good times that made them raiders.

      The steppe world began to take shape around 4000 BCE with the coming of livestock to Central Asia. Sheep and goats slowly spread from their homes in the hospitable, pleasant Near East out onto the desolate, cold steppe and desert lands. The real dawn of steppe power, however, was the domestication of the horse. It apparently took place around 3500 BCE in what is now Ukraine and Kazakhstan. Horses are first known as tamed livestock from the Botai culture of the Ukraine area, around 3500–3000 (not 4000, as previously reported). Horses were apparently domesticated only once, though herds recruited mares from local wild populations all over Eurasia (Achilli et al. 2012). The first secure evidence comes from the Botai culture in Kazakhstan (Anthony 2007; Frachetti 2008, 2012; Levine et al. 2003). The Botai people depended almost exclusively on horses for animal protein—not just the meat (of which they ate an enormous amount) but also the milk, as shown by residues in pots. Milking horses implies domestication.

      No one knows when riding started—claims of bit wear on ancient horse teeth have not held up—but presumably it entered the picture about this time. By 3000 BCE, mounted riders seem to have been ranging widely over the steppes, and by 2000 the war chariot was a major part of warfare as far afield as the Near East. Horses and war chariots reached China around 1500. The Indo-Europeans were among those who took advantage of the horse and of livestock nomadism in general to radiate in all directions and build up large populations. It is tempting to associate the Botai with them, but the Botai are farther east than the presumed IE center in Ukraine. Perhaps the Indo-Europeans were already in the east, or perhaps the Botai people were ancestral Uralic or Altaic groups.

      Horses were in China by the middle Shang Dynasty, 1400–1500 (Harris 2010: 82; Lawler 2009), but, so far, are not reported earlier. China is not good horse country; there is little good grazing, and, in historic times, there was little room to grow fodder. Much of China’s lands are deficient in selenium, which horses need (May 2012). China always obtained its best horses from the steppes.

      Local conditions—ecological and cultural—led to different emphases in different areas of China and Central Asia: sheep and goats dominated widely, and there were even cattle specialists in some relatively favorable areas, but the all-importance of the horse in Kazakhstan was slow to change (Frachetti 2012). The western, central, eastern, montane, and far eastern steppes all had different histories, political as well as ecological; nearness to great civilizations, isolation by mountain ranges, and ease of mobility all mattered.

      When the steppe peoples entered Chinese history, their way of life was already ancient. It was, however, far more than nomadic herding. Central Asia, especially at the western and eastern ends, was a complex intermingling of steppe nomads, seminomadic groups with varying degrees of agriculture, settled riverine farmers using intensive irrigation, and dry-farmers taking advantage of every wet period to extend farming far out into dry lands—as pointed out by scholars such as Owen Lattimore (1940) long ago and many others since (e.g., Barfield 1989, 1993; Barthold 1968).

      By 1500 BCE there were substantial farming settlements in the Zhunge’er (Junggar, Dzungarian) Basin, in what is now far northwest China. The people dry-farmed wheat, barley (naked barley was prominent), and foxtail millet. They had sophisticated pottery, similar to that from other parts of eastern Central Asia at the time (P. Jia et al. 2011) but quite different from the wares of China—at that time just entering the Shang Dynasty. No hints of their ethnic affiliation exist. The area is traditionally a haunt of “nomads,” but these people were not nomadic. The widespread occurrence of early intensive farming in Central Asia, now established, has changed some historical speculation.

      Other high cultures with distinctive art and architecture have been discovered in Central Asia (see Lawler 2009 for a quick overview). They share many broad patterns with the better-known early cultures of the Yellow River plain but are still distinctive. Data on these societies are only beginning to appear, and the instability of the region makes excavation difficult at best.

      The early Chinese and Roman historians shared a tendency to overstate the nomadism and the dependence on stock as a way of differentiating the “Huns” and “Xiongnu” and other “barbarians” from “civilized” folk. In fact, every major stable Central Asian state or conquering horde had to depend on agriculture for a great deal of its food, clothing, and wealth (cf. Honeychurch and Amartuvshin 2006). The Xiongnu, for instance, held vast areas that were very dry but that were and are farmed, as well as several major riverine oasis-strips.

      The Central Asian cultures have produced many mummies, preserved by the dry, cold climate. They show that most of the people there were of West Asian (some perhaps even European) background. Current genetic theory holds that the East Asian peoples are derived largely from groups that moved up very slowly from Southeast Asia. So their late radiation into Central Asia led to a meeting of quite different stocks when they encountered Caucasians spreading through Central Asia from the west. Many of the Central Asian mummy-wrapping textiles are wool woven in patterns similar to European ones; some are strikingly similar to Scottish plaids (Barber 1999; J. Mallory and Mair 2000). The earliest mummies date to 1800–1500 BCE. These people certainly include the ancestors of the Tocharians. (The Tocharoi of Greek history were in northern Afghanistan, whereas the people discussed here, the Twghry, occupied what is now Xinjiang. Tocharoi is a very reasonable Greek spelling of Twghry, so the mistake may simply be a minor misplacement by the Greek writers. See Hansen 2012: 73.) At least three Tocharian languages were spoken in this area in early historic times. The better known ones are usually called Tocharian A and B, but the more useful names Kuchean and Agnean are coming into use (Hansen 2012: 74). They are Indo-European, close enough to eastern European languages that their word for “fish” was “lox”! (Phonetically laks, lakse, or laksi.) And a modern Uyghur bread resembles