Chickens and Millet: Early Agriculture in China

Paper delivered at Society of Ethnobiology annual conference, Santa Barbara, CA, May 2015


Chickens and Millet:  The Significance of New Findings in Chinese Food Archaeology

Recent findings in archaeology have considerably pushed back the dates for domestication of chickens, millets, rice, pigs, and other domestic life forms of eastern Asia.  North China has taken a lead over south China, though this may change with further investigation.  Early evidence of milking and stockraising in central Asia is relevant. To a cultural anthropologist working with modern uses of plants and animals, the new findings confirm my models and suppositions about the origins and development of agriculture: it happened when environmental conditions improved and food got more abundant, not during periods of scarcity; it probably involved trade and certainly contact with other groups; it took place in favorable locations at probable trade crossroads.  Early items grown were those either storable or highly valued or both.  Uses of many items tended to shift over time as more efficient systems were discovered.  The development of food systems has to be understood in a context of induced biological development: changes were most likely when they removed bottlenecks that inhibited trade, contact, and efficiency.


A number of recent excavation projects in China have shown that agriculture there is much older than we once thought.  When I was a student, agriculture was known back to about 4000-5000 BCE.  It now appears to go back at least to 8000 BCE (for excellent recent reviews of Chinese archaeology, see Li 2013; Liu 2005; Liu and Chen 2012; see also Anderson 2014).

Pottery is even earlier, and those who still believes in the “Neolithic Revolution” will be delighted to learn that pottery goes back to 20,000 years ago or more in China (Wu Xiaohong et al. 2012).  It was soon quite widely spread, from the Pearl River at 15,000 BCE (Pearson 2006) to the Amur River on China before 11,000 BCE (Zhushchikovskaya 1997).  It is in Japan by 13,000 or earlier.  This pottery is probably ancestral to that of Europe, since one sees a slow spread of similar wares across Siberia.  Ground stone appears early in the form of milling stones (metates).

The first agriculture known in China involves two species of millet, foxtail (Setaria italica) and broomcorn or panic millet (Panicum miliaceum). Several sites report them around that date, but the most interesting currently at Nanzhuangtou, somewhat south of Beijing.  Here not only early millets but the earliest domestic chickens in the world are found, at 8000 BCE (Xiang et al. 2014, 2015).  The earliest dog in China is also there, and is even earlier, at 10,000 BCE (Liu and Chen 2012:64).  Very early pigs and dogs are found at nearby sites.

The Nanzhuangtou site got its domesticates during the rise of warm wet weather around 8000 BCE.  There and elsewhere, rise and spread of domestication tracks warming and wetting trends, with dramatic improvement of growing conditions.  Around Dadiwan, for instance, there were forests of oak, birch, maple, hazelnut, cherry, chestnut, hophornbeam, sorbus, persimmon, hornbeam, elm, Toxicodendron (I didn’t know China had that), locally even liquidambar, eucommia, and other warm-weather trees.  Spruce occurred locally, with sharp decline after 2600 BCE (Li et al. 2013).  Most of these must have been on the mountains above the site, not in the dry, desolate plains where the site is, but trees evidently moved down the valleys.  It is worth noting that the mountains support a forest today, though not such a subtropical one.  The high Qinling Mountains to the south also had a warm-temperate forest (Zhao et al. 2014).

Several sites in the Yellow River drainage report millets back to 7000 BCE.  Millet agriculture had reached Dadiwan, far out into west China and almost in the Central Asian desert, by 6000 BCE (Bettinger et al. 2010).  It had also reached Inner Mongolia by this time (Shelach et al. 2011).  Millets are C4 plants, almost everything else in China is C3 (including rice), so where C4 shows up in bone signatures one can be sure that millet is being devoured.  This allows us to find transitions to agriculture in the record, with C4 dominating by 6000 BCE.  Rice occurs at Jiahu, one of the millet-agricultural sites from around 5000 BCE (Zhang and Hung 2013).  Also at Jiahu were residues of millet beer brewed with honey, hawthorn fruit, and grapes (Liu and Chen 2012:120; McGovern 2009).  Patrick McGovern, who analyzed this residue, worked with Dogfish Head Brewery in Delaware to reconstruct it, and you can now buy “Chateau Jiahu” beer if you can find it.  It is possibly a bit less than the finest brew, and thus is rarely stocked by liquor stores.

Millet agriculture, complete with chickens, pigs, and dogs, spread from north-central China throughout what is now China by 4000 BCE.  Its expansion could very well have coincided with, and been responsible for, the spread of the Tibeto-Burman (a.k.a. Sino-Tibetan) language phylum.  The timing, location, and motivation all seem right.  This phylum may have started in high west China, judging from surviving origin myths, or from central north China.  Its more recent radiation, giving us the Tibetan, Burman, Qiang, and other branches and probably the Chinese too, is generally thought to have been in Sichuan; there are many good grounds for this (van Driem 1999, 2002).  If so, that was probably after the spread of millet agriculture into that mountainous region, which is perfect for differentiation and migration of groups.

Meanwhile, rice (Oryza sativa) was domesticated by 7000 BCE in the Yangzi area (Liu and Chen 2012:76)  The Peiligang culture, flourishing 7000-5000 BCE, had a lot of it, as well as millet (Liu 2012).  At Hemudu by 5000 BCE rice was common, with large containers of it having been found.  (Rice then probably yielded 500 kg/ha; it now yields over ten tons per ha, thanks to the masterful breeding programs of Yuan Longping [2002] and others.)

A great deal of controversy surrounds rice.  All evidence points to the Yangzi drainage, except for some recent genetics work that pinpoints the Pearl River drainage, far to the south (Huang et al. 2012).  But they sampled wild rice (Oryza rufipogon) largely from that area, and the plant is mostly gone in the Yangzi area, so this is probably an artifact.  Jeanmarie Molina et al. (2011) found that their genetic data pointed to a single origin in the Yangzi area for rice, but more recent work

There are two major divisions of rice, japonica (short grain) and indica (long grain).  These are very separate and hard to cross (Yang et al. 2012), indicating a very long period of divergence; they were very likely different in the wild long before humans came on the scene, and thus must have been domesticated separately.  They may both be native to the Yangzi area. Reports of indica in early Chinese sites are common but controversial.  However, recent work suggests that indica may have arisen in India from hybridization of introduced japonica with local Indian strains (Callaway 2014).  Rice is not found in south China till about 4000 BCE (Lu 2011).

It was in Taiwan, along with millets, in the Daben’geng culture, which represents Austronesians migrating from southern China to that island around 3000 BCE (Hung and Carson 2014).  The Austronesians, specifically the Malayo-Polynesian branch thereof, apparently radiated later from Taiwan throughout Oceania (Bellwood 2009; Bellwood and Renfrew 2002).  Many agricultural words reconstruct to proto-Austronesian or at least proto-Malayo-Polynesian, including words for grains, root crops, chickens, and pigs.

Rice may have been spread by the ancestors of the Thai-Kadai language phylum.  It in turn may be related to Austronesian and even other relevant groups (Sagart et al. 2005).  The Austroasiatic phylum is generally believed to have arisen in India, but now some think that it arose in China and spread rice there; there are many words associated with rice cultivation in its reconstructed original vocabulary (Sagart et al 2005).  In any case, it is hard to deny that the various phyla in south China—Thai-Kadai, Yao-Mian, Miao-Hmong, Austronesian, Austroasiatic—may all have been involved from a quite early time.  The Austronesian word for unhusked rice may even have invaded Tibeto-Burman: Bahasa Malaysia beras, Tibetan mbras (Sagart et al 2005), but the similarity of the words—if it is not purely accidental, which I think it is—would imply a very recent borrowing.

Returning to the chicken, a very interesting point is that the word for “chicken” all over east Asia and widely in the rest of the world derives from the Thai-Kadai root kai (Cantonese kai, surely the Thai word; Mandarin ji from *kai; and so on; Blench 2007).  This indicates to me that the ancestors of the Thai-Kadai, who were almost certainly in the Yangzi valley, domesticated the bird, in which case it spread north after domestication to Nanzhuangtou.

Along with rice came peaches.  Possibly domesticated peaches occur by 6000-5000 BCE in the lower Yangzi and are certainly domesticated by 3000 (Zheng et al. 2014).  It can be safely assumed that if people were domesticating peaches they must have domesticated a range of other fruits, as well as vegetables and other plants.  I also strongly suspect that China had long been managing wild trees, as Native Americans did (and locally still do), to maximize nut tree production; oaks and various nut trees abounded.   Elsewhere in China, buckwheat was being domesticated about this time (Ohnishi 1998).

By 4000 BCE, then, millets, rice, and the commoner domestic animals, as well as fruits and other foods, were all over what is now China, except for the remote mountain and desert areas.  (Tibet in particular remained long unsettled.)  Large villages developed, and beautiful, exquisitely made pottery was common.  Dairying was evidently beginning in central Asia; not long after, residues of kefir and other milk products appear on pots (Yang et al. 2014).

Before 3000 BCE, settlements grew large, implying rich chiefdoms (Drennan and Dai 2010).  Some settlements grew to near urban size, such as the mysterious towns of the Hongshan culture in far north China.  This culture goes back as far as 4500 BCE (Shelach et al. 2011; Zhang et al 2013).  By 3000 it was producing sizable towns that seem like capitals and have associated ceremonial and ritual items (Allan 2002).  It remained at a chiefdom level, with only about 1000 people in these large towns (Peterson et al 2010).  But then it declined and fell,  It had depended on C4 food entirely, but now regressed to getting 15% of its food from C3 plants (Liu and Chen 2012:177), indicating a return to foraging on wild foods or eating coarse grains and vegetables.


What lies behind this certainly includes climate.  China’s last and harshest ice age gave way, as elsewhere, betweem about 15,000 and11,000 years ago.  Then a warm and quite wet period came, between 10,000 and 5,000 years ago, with a maximum warmth around 8,000 to 5,000 years ago.  This is exactly the period when agriculture developed and flourished most, and when large settlements arose.  The period from 3000 to 2000 BCE was one of decline.  Cultures like Hongshan sank back to small village levels.  Significantly, the area at the great bend of the Yellow River, at the focus of the great logical trade routes of northern China, was countercyclic: it grew and flourished in population in the late 2000s BCE.

Around 2000 it broke over into full civilization, with bronze work, intensive agriculture, and massive architecture.  Soon after that, the city at what is now the village of Erlitou reached a population between 18,000 and 30,000 (Liu 2009; Liu and Chen 2003).  Nothing remotely close to that size existed elsewhere in China.  It has been awfully hard for people to resist equating this statelet with the legendary Xia Dynasty, China’s first dynasty, known only from reports in much later history works.  The only problem is that the Erlitou culture had no writing (though some marks on pots point toward it).  We can only guess.  Many, I think most, Chinese archaeologists, however, now simply assume that Erlitou was Xia.

There may have been some climatic improvement, but it seems more likely that Erlitou flourished because times were hard.  It was strategically placed to dominate trade, communication, and military adventuring in the Yellow River drainage.  Competition over scarce resources might well have driven a race to build bigger, more defensible settlements.  Ceramics and other stylistic markers show that Erlitou exerted at least cultural and possibly political dominance over a local area about the size of a typical early state or large chiefdom.  (See Liu and Chen 2003, 2012, esp. 2012:258-259.)


Significantly, through all of this, China lost almost no wild species.  Some megafauna went extinct at the end of the ice age, but China kept most of its megafauna.  Elephants, rhinoceri, and other large animals still existed well into historic times, and of course China still has pandas, tigers, bears, gibbons, and even a few elephants in the far south.  Hunting was a major source of food in the Neolithic, as was fishing.  China’s fantastic botanical diversity flourished.  During cold dry periods, animals and plants retreated southward.  In warm wet ones, they moved north again.


So the record is one of agriculture expanding rapidly along with the improvement of the climate for plant growth, and the continuing flourishing of megafauna.  This goes totally against the received wisdom in studies of agricultural origins, which usually assume that agriculture was invented because people needed food.  Either they killed off the megafauna, or they grew rapidly in population or they just plain starved.  (For a summary of theories of agricultural origin, with full references, see Graeme Barker 2006.)  Yet, in China, it was not only the richest and most food-abundant areas that developed agriculture, but specifically the areas that were most rapidly getting richer still.

But it does dramatically confirm one theory:  Carl Sauer’s, from his book Agricultural Origins and Dispersals (1952).  Sauer understood that developing agriculture is a very long, slow process that does not pay off immediately.  It requires people who are playing around with plants, using them, letting them seed themselves.  He guessed that agriculture was invented by settled people who lived by fishing and gathering in southeast Asia.

The earliest agriculture in the world is in the Near East, and the earliest agriculture elsewhere is in northern China, but Sauer is close to right.  The communities in north China were relatively settled, plant-dependent, and above all on a climatic roll. Moreover, agriculture began in exactly the area that was turned from a cold, harsh steppe into a lush, warm-temperate paradise.  It spread first to other areas with that history, then south and out into other areas that were also improving, but less rapidly.

This being determined, what was the motive?  If wild food was rapidly increasing, why farm?  Sauer thought people might start with fibre crops instead of food, but this was not the case.  Brian Hayden and his associates (2001) hypothesized that the motive might be producing food for feasts.  Indeed, there is evidence of some feasting, but largely later in time.

For over 40 years, I have been arguing that agriculture developed because of trade.  People wanted trade goods to be around the settlement—both to have them on hand to trade and to be able to protect them from raid.  It so happens that the early Chinese sites are in good areas for trading.  However, there is not much evidence for trade on any scale until agriculture was well developed.  I suspect it was there, but it certainly was not overwhelmingly obvious.

Storage is another concern.  Grain can be stored easily.  Domestic animals are a form of storage: one controls them and their reproduction, in contrast to the situation with wild animals.  They are always around the house.  The larger and more settled a group is, the more useful storage is to them.  However, what we often find is a replacement of wild nut crops—acorns, chestnuts, walnuts—by grain (Liu 2012).  This happens quite dramatically in much of the north around 7000-5000 BCE.  Grain is much more controllable.  It grows fast, yields reliably, and can be spatially manipulated—you can plant it anywhere.  Tree crops, in contrast, bear erratically, cannot be moved around easily, and cannot regrow fast after a fire, flood or disease episode.  It would make a great deal of sense for people to take control of their destinies by growing their animals and quick-maturing plant foods, instead of depending on uncertain nature.

However, I do not believe that storage and control are adequate motives.  I still think the trade and protection theory is the only one that can explain existing patterns of early agriculture.  Only it, for instance, explains the persistent correlation of development and progress with areas that are central to trade.  The early sites are near trade routes, and the strategic location of Erlitou and other larger, later sites is unquestionable.  There is also the issue of those early chickens: they probably came from farther south—chickens are not native anywhere near Nanzhuangtou.  If they were traded up from a domestication farther south, as I think they were, we have good evidence of important trade in domesticated food items.  I await further research, in hopes it will provide more evidence. Meanwhile, at the very least, the theories that assume agriculture developed because people needed food do not fit the Chinese case, or any other case known to me.





Allan, Sarah (ed.).  2002.  The Formation of Chinese Civilization:  An Archaeological Perspective.  New Haven: Yale Unversity Press.


Anderson, E. N.  2014.  Food and Environment in Early and Medieval China.  Philadelphia: University of Pennsylvania Press.


Bellwood, Peter.  2009.  “The Dispersals of Established Food-Producing Populations.”  Current Anthropology 50:621-626.


Bellwood, Peter, and Colin Renfrew.  2002.  Examing the Farming / Language Dispersal Hypothesis. Cambridge:  McDonald Institute for Archaeology.


Bettinger, Robert; Loukas Barton; Christopher Morgan; Fahu Chen; Hui Wang; Thomas P. Guilderson; Duxue Ji; Dongju Zhang.  2010.  “The Transition to Agriculture at Dadiwan, People’s Republic of China.”  Current Anthropology 51:703-714.


Blench, Roger.  2007.  “Using Linguistics to Reconstruct African Subsistence Systems:  Comparing Crop Names to Trees and Livestock.”  In:  Denham, Tim; José Iriarte; Luc Vrydaghs (eds.).  2007.  Rethinking Agriculture:  Archaeological and Ethnoarchaeological Perspectives. Walnut Creek: Left Coast Press.  Pp. 408-438.


Callaway, Ewen.  2014.  “The Birth of Rice.”  Nature 514:S58-S59.


Drennan, Robert, and Xiangming Dai.  2010.  “Chiefdoms and States in the Yuncheng Basin and the Chifeng Region: A Comparative Analysis of Settlement Systems in North China.”  Journal of Anthropological Archaeology 29:455-468.


Hayden, Brian (ed.).  2001.  Feasts:  Archaeological and Ethnological Perspectives on Food, Politics and Power.  Washington, DC:  Smithsonian Insitution Press.


Huang, Xuehui; Nori Kurata; Xinghua Wei; Zi-Xuan Wang; Ahong Wang; Qiang Zhao; Yan Zhao; Kunyan Liu; Hengyun Lu; Wenjun Li; Yunli Guo; Yiqi Lu; Congcong Zhou; Danlin Fan; Qijun Weng; Chanrang Zhu; Tao Huang; Lei Zhang; Yongchun Wang; Lei Feng; Hiroyasu Furuumi; Takahiko Kubo; Toshie Miyabayashi; Xiaoping Yuan; Qun Xu; Guojun Dong; Qilin Zhan; Canyang Li; Asao Fujiyama; Atsushi Toyoda; Tingting Lu; Qi Feng; Qian Qian; Hayang Li; Bin Han.  2012.  “A Map of Rice Genome Variation Reveals the Origin of Cultivated Rice.”  Nature 490:497-501.


Hung, Hsiao-chun, and Mike T. Carson.  2014.  “Foragers, Fishers and Farmers: Origins of the Taiwanese Neolithic.”  Antiquity 88:342:1115-1131.


Li Feng.  2013.  Early China: A Social and Cultural History.  Cambridge: Cambridge University Press.


Li Linna (chief ed.).  2007.  Treasure from the Museum of the Nanyue King.  Guangzhou: compiled by Museum of the Nanyue King, published by Cultural Relics Press.


Li, Xiaoqiang; Nan Sun; Xinying Zhou; Keliang Zhao.  2013.  “Vegetation Characteristics in the Western Loess Plateau between 5200 and 4300 cal. B.P. Based on Fossil Charcoal Records.”  Vegetation History and Archaeobotany 22:61-70.


Liu Li.  2005.  The Chinese Neolithic:  Trajectories to Early States.  Cambridge: Cambridge University Press.


—   2009.  “State Emergence in Early China.”  Review of Anthropology 38:217-232.


—  2012.  “Archaeology Under a Microscope.”  Talk, Fowler Museum of UCLA, Jan. 6.


Liu Li and Xingcan Chen.  2003.  State Formation in Early China.  London:  Duckworth.


Liu Li and Xingcan Chen. 2012.  The Archaeology of China: From the Paleolithic to the Early Bronze Age.  Cambridge: Cambridge University Press.


Lu, Tracey L-D.  2011.  “Prehistoric Migration, Historic War, and the Occurrence of Rice Farming and Consumption in South China.”  Proceedings, 12th Conference of the Foundation for Chinese Dietary Culture, Okinawa.  Pp.  239-251.


Molina, Jeanmarie; Martin Sikora; Nandita Garud; Jonathan M. Flowers; Samara Rubinstein; Andy Reynolds; Pu Huang; Scott Jackson; Barbara A. Schaal; Carlos D. Bustamante; Adam R. Boyko; Michael D. Purugganana.  2011.  Molecular Evidence for a Single Evolutionary Origin of Domesticated Rice.  Proceedings of the National Academy of Sciences 1104686108.  Online, May 2.


Ohnishi, Ohmi.  1998.  “Search for the Wild Ancestor of Buckwheat III.  THe Wild Ancestor of Cultivated Common Buckwheat, and of Tatary Buckwheat.”  Economic Botany 52:123-133.


Pearson, Richard.  2006.  “The Social Context of Early Pottery in the Lingnan Region of South China.”  Antiquity 79:819-828.


Peterson, Christian; Xueming Lu; Robert D. Drennan; Da Zhu.  2010.  Hongshan Chiefly Communities in Neolithic Northeastern China.  Proceedings of the National Academy of Sciences 107:5756-5761.


Sagart, Laurent; Roger Blench; Alicia Sanchez-Mazas.  2005.  The Peopling of East Asia:  Putting Together Archaeology, Linguistics and Genetics.  London:  RoutledgeCurzon.


Shelach, Gideon.  2012.  “On the Invention of Pottery.”  Science 336:1644-1645.


Van Driem, George.  1999.  “Neolithic Correlates of Ancient Tibeto-Burman Migrations.”  In Archaeology and Language II, ed. by Roger Blench and M. Spriggs. London:  Routledge.  Pp. 67-102.


—  2002.  “Tibeto-Burman Phylogeny and Prehistory:  Languages, Material Culture and Genes.”  In Examing the Farming / Language Dispersal Hypothesis, ed. by Peter Bellwood and Colin Renfrew. Cambridge:  McDonald Institute for Archaeology.  Pp. 233-249.


Von Falkenhausen, Lothar.  2006.  Chinese Society in the Age of Confucius (1000-250 BC).  Los Angeles:  Cotsen Institute of Archaeology, UCLA.


Wu, Xiaohong; Chi Zhang; Paul Goldberg; David Cohen; Yan Pan; Trina Arpin; Ofer Bar-Yosef.  2012.  “Early Pottery at 20,000 Years Ago in Xianrendong Cave, China.”  Science 336:1696-1700.


Xiang, Hai; Jianqiang Gao; Baoquan Yu; Hui Zhou; Dawei Cai; Youwen Zhang; Xiaoyong Chen; Xi Wang; Michael Hofreiter; Xingbo Zhao.  2014.  “Early Holocene Chicken Domestication in Northern China.”  Proceedings of the National Academy of Sciences 111:17564-17569.


Xiang, Hai; Michael Hofreiter; Xingbo Zhao.  2015.  “Reply to Peng et al.: Archaeological Contexts Should Not Be Ignored for Early Chicken Domestication.”  Proceedings of the National Academy of Sciences 112:E1972-1973.


Yang, Jiangyi; Xiaobo Zhao; Ke Cheng; Hongyi Du; Yidan Ouyang; Jiongjiong Chen; Shuqing Qiu; Jianyan Huang; Yunhe Jiang; Liwen Jiang; Jihua Ding; Jia Wang; Caiguo Xu; Xianghua Li; Qifa Zhang.  2012.  “A Killer-Protector System Regulates Both Hybrid Sterility and Segregation Distortion in Rice.”  Science 337:1336-1340.


Yang Yimin; Anna Shevchenko; Andrea Knaust; Idelisi Abuduresule; Wenying Li; Xingjun Hu; Changsui Wang; Andrej Shevchenko.  2014.  “Proteomics Evidence for Kefir Dairy in Early Bronze Age China.”  Journal of Archaeological Science, online, March 2014.


Yuan Longping.  2002.  “The Second Generation of Hybrid Rice in China.”  20th International Rice Commission meeting, Bangkok.  Retrieved 2013 from FAO website.


Zhang Chi and Hsiao-chun Hung.  2013.  “Jiahu 1: Earliest Farmers beyond the Yangtze River.”  Antiquity 87:46-63.


Zhao, Xiangui; Caihong Ma; Ling Xiao.  2014.  “The Vegetation History of Qinling Mountains, China.”  Quarternary International 325:55-62.


Zheng, Yunfei; Gary Crawford; Xugao Chen. 2014. “Archaeological Evidence for Peach Cultivation and Domestication in China.” PloS One DOI:10.1371/journal.pone.0106595.


Zhushchikhovskaya, Irina.  1997.  “On Early Pottery-Making in the Russian Far East.”  Asian Perspectives 36:159-174.

Leave a Reply