The Origins of Food Production in North China: A Different Kind of Agricultural Revolution ROBERT L. BETTINGER, LOUKAS BARTON, AND CHRISTOPHER MORGAN By roughly 8,000 calendar years before the present (calBP), hunter-gatherers across a broad swath of north China had begun small-scale farming of broomcorn millet (Pan- icum miliaceum) and foxtail millet ( Setaria italica). 1–6 According to traditional wisdom, this early millet farming evolved from the intensive hunter-gatherer adaptation repre- sented by the late Pleistocene microblade tradition of northern China, 2,7 termed here the North China Microlithic. The archeological record of this hunter-gatherer connection is poorly documented, however, and as a result the early agricultural revolution in north China is not as well understood as those that occurred in other parts of the world. The Laoguantai site of Dadiwan, in the western Loess Plateau, Gansu Province, PRC, fur- nishes the first complete record of this transition, which unfolded quite differently from other, better known, agricultural revolutions. Early millet farming is represented by at least five geographically separate but roughly contemporaneous cul- tural complexes (Xinglongwa, Houli, Cishan, Peiligang, and Laoguantai) distributed over an environmentally diverse area stretching 1,500 km from the northeast China Plain to the western Loess Plateau (Fig. 1). That these complexes appear so close in time but far apart in space and vary so much in cultivar choice and cul- tural details makes it unlikely that the millet farming revolution of north China was the spread of an idea, crop, or people from a single source. This phenomenon is better understood as an evolutionary contest among many varied regional adaptations, some of them (at least the five above) involv- ing food production, others not. However, all these adaptations are rooted in a trajectory of hunter-gath- erer intensification running well back into the Pleistocene. Unfortunately, the published litera- ture does not present even one continu- ous stratigraphic sequence or statisti- cally convincing seriation connecting any of the cultural complexes represent- ing north China’s preagricultural hunter-gatherers to any of the five repre- senting its early millet farmers. The tra- ditional view that millet farming evolved from the North China Microlithic is entirely plausible and arguably the most viable scenario, except that its artifact assemblages do not match. Microblades and microblade cores which are com- mon in North China Microlithic sites (such as Xiachuan, Xueguan, Shizitan, Shayuan, and Hutouliang), are rare or absent in nearby early millet farming sites such as Cishan, Peiligang, and Dadiwan. 2,8,9 To date, not one stratified assemblage or series of related assemb- lages bridges this technological gulf. In this and other key respects, the agricultural transition in north China is unlike the better-known transitions of the Near East 10–19 and Mesoamer- ica, 20–27 where archeological continu- ities make the forager to farmer con- nection much easier to trace. We argue here that this difference is not an artifact of archeological bias, pres- ervation, or sampling, although those have muddied the water. Rather, using a combination of optimal foraging and evolutionary game theory, we con- struct a model that explains the dis- connect between forager and millet farmer in north China as the signature of an agricultural transition quite dif- ferent from those more richly and con- tinuously attested to elsewhere around the globe. To support that argument, we report recent work at the Dadiwan site, which provides the first complete archeological sequence from north China to record behavioral variation across the transition to agriculture. While controversial and certainly far from conclusive, our model matches the Dadiwan record on key points much more closely than does any al- ternative account of the transition to millet agriculture in North China. The Dadiwan record begins at about 60,000 calBP, revealing a history of use by human hunters throughout the ARTICLES Robert L. Bettinger is Professor in the Depart- ment of Anthropology, University of California, Davis 95616. E-mail: rlbettinger@ucdavis.edu Loukas Barton is a research affiliate of the Center for Arid Environment and Paleocli- mate Research at Lanzhou University in China and archeologist for Katmai National Park and Preserve, the Alagnak Wild River, and the Aniakchak National Monument and Preserve, P.O. Box 7, King Salmon, AK 99613. E-mail: loukas_barton@nps.gov Christopher Morgan is Assistant Professor in the Department of Sociology, Social Work, and Anthropology, Utah State Univer- sity, 0730 Old Main Hill, Logan, UT 84322- 0730. E-mail: chris.morgan@usu.edu Part of a US-PRC research team that has been conducting research in China for more than twenty years, Bettinger, Barton, and Morgan are interested in human behavioral change over the last eighty thousand years in the western Loess Plateau, specifically in relation to the appearance of anatomically modern humans, rapid climatic change, environmental deterioration during the Last Glacial Maximum and Younger Dryas, the shift from Pleistocene to Holocene climatic regimes, and the origin of millet agriculture. Key words: origins of agriculture, China, Neolithic, millet domestication, hunter-gatherers, microlithic V V C 2010 Wiley-Liss, Inc. DOI 10.1002/evan.20236 Published online in Wiley InterScience (www. interscience.wiley.com). Evolutionary Anthropology 19:9–21 (2010)