Connecting China to Europe in the Bronze Age

by Edward Pegler on 1 November, 2013

How evidence of wheat, copper and broomcorn millet gives some clues to the first connections between West and East and the routes that were taken. A current best guess is for a steppe connection at the beginning of the third millennium BC and a ‘silk road’ connection at the end of the 3rd millennium. However, a much earlier connection (the sixth millennium BC) is still arguable.

Map showing Central Asia, centred on the Takla Makan and the Silk Road. The scale takes in Iraq and the Black Sea in the west to the Yellow River in the east. The Silk Road runs, from west to east, along the Kopet Dag, through the Hindu Kush, both to the south and north of the Takla Makan Desert and down the Gansu (Hexi) Corridor to join the lower reaches of the Yellow River (Huang He). Dark Green indicates forest, light green is open country or steppe, yellow is scrub and white is desert. Water is blue. Mountains are indicated in a suitably Tolkein-like way.

Map showing Central Asia, centred on the Takla Makan and the Silk Road. The scale takes in Iraq and the Black Sea in the west to the Yellow River in the east. The Silk Road runs, from west to east, along the Kopet Dag, through the Hindu Kush, both to the south and north of the Takla Makan Desert and down the Gansu (Hexi) Corridor to join the lower reaches of the Yellow River (Huang He). Dark Green indicates forest, light green is open country or steppe, yellow is scrub and white is desert. Water is blue. Mountains are indicated in a suitably Tolkein-like way.

A long discussed question for historians is how long ago China and the West (by which I mean the Middle East and Europe) developed contacts with each other. There are known to have been links by the first millennium BC, when Alexander the Great fought his way East to Afghanistan, and when China was having problems with tribes on its western borders. However, there is interesting evidence for prehistoric contacts between East and West well over a thousand years before this.

This evidence comes in the form of copper smelting, wheat, barley and oat agriculture, cattle-sheep pastoralism and broomcorn millet (‘birdseed’) agriculture. The idea that copper smelting was not invented in China is a relatively new one but is increasingly accepted by Chinese archaeologists. Wheat, barley and oat agriculture is almost universally accepted as having originated in the west. Domestic cattle and sheep show complex origins with earliest evidence in the west. The oldest evidence for broomcorn millet agriculture is in the east. However, all of these things had become relatively commonplace in both east and west by 2000 BC.

The problem for archaeologists is to decide whether pre 2000BC evidence of common features is due to independent invention in the east and west or due to earlier connections between them. If the latter then when were these connections made and by what routes?


Old World copper was first worked and smelted in the west (possibly the Balkans, perhaps Turkey), and is known to have been smelted in eastern Iran by the fifth millennium BC. Copper smelting and the production of bronze was carried as far east as the Altai Mountains by the late 4th or early 3rd millennium BC. This is associated with the Afanasievo culture, which shows links with other steppe cultures to the west (Anthony, 2006; Syvatko et al., 2009).

The first appearance of smelted copper in China (surprisingly, a relatively sophisticated tin bronze) is currently from the Majiayo culture at Linjia in Gansu, with an approximate date around 2800 BC. Another important early site is Taosi, from the second half of the third millennium BC, located west of the bend in the Yellow River (Huang He) just before the river is joined by the Wei River. This site includes items of both copper and arsenic bronze (Liu & Chen 2012).

(I’ve just found an article which puts copper smelting in China earlier, questionably around 4000 BC in the Banpo culture and, more confidently, around 3000 BC, in the Hongshan culture.)

Both of these sites are to the west of the great Central Plain, where many innovations of Chinese culture are often assumed to originate. Significantly, both are earlier than the great central plain bronze production culture of Erlitou (2000 BC and after).

Broomcorn Millet (Panicum miliaceum)

Also known as ‘proso’ millet, this always seems to be called  ‘broomcorn’ by archaeologists. It is an annual with no known wild relatives (Liu & Chen 2012), although the domesticated form, according to Zohary et al. (2012) at least, grows as a weed or is naturalised anywhere between the Caucasus and northern China/Mongolia in conditions best perhaps described as dry steppe. Broomcorn is cultivated as far south as Africa.

An early form of domesticated broomcorn was being harvested across northern China possibly as far back as the 8th millennium BC (Lu et al. 2009) and certainly by the late 7th millennium BC (Liu & Chen 2012). By the 4th millennium BC broomcorn was being farmed in highland Sichuan (SE China) (d’Alpoim Guedes 2011).

Domesticated broomcorn is first reported in the west around the early sixth millennium BC in places including Sacarovka, Moldavia and on the Dniester by mid 6th millennium (BC Anthony 2007). It is also found in slightly younger contexts at Tepe Yahya, Iran and Arukhlo, Georgia. It appears to have spread from the Balkans further west into northern parts of Europe , supposedly with the first farmers (Zohary et al. 2012).

There are no reports of broomcorn millet of this date in the lands between these two extremes. In fact, the oldest dated reports of broomcorn between the Caucasus and China are three thousand years later. For example, in Tahirbaj Tepe in Turkmenistan from the mid 2nd millennium BC (Herrmann & Kurbansakhatov 1994), in Begash, Khazakstan (dated 2400-2100BC) (Frachetti et al. 2010) and in Shortugai, Afghanistan (late third millennium BC). The latter are linked to the Harappan culture of NW India and Pakistan (Zohary et al. 2012).

Foxtail millet

Despite being another early Chinese domesticate (maybe late 7th millennium) found in much the same region as broomcorn (Liu & Chen 2012), foxtail millet does not appear in Europe until the 2nd millennium BC (Zohary et al. 2012).


Although first cultivated between 7000-5000 BC in southern and central China, the earliest examples are reported in India at the end of the 3rd millennium BC in the central and eastern Ganga valley.

Wheat, Barley and Oats

These are all western crops, originally from the Middle East where they were cultivated by the eighth millennium BC.

In the East, wheat is first recorded in the Yellow River region, east of the Wei River, and on the eastern coast of China by the late third millennium BC (Liu & Chen 2012) (an earlier date of 2600 BC further west is disputed due to context – Jia, X. et al. 2012). It also appears in Begash around the same time (2400-2100 BC) (Frachetti et al 2010). However, not until the second millennium BC does it appear in the intermediate territories of the later Silk Road, such as in Gansu and at the edges of the Takla Makan (Liu & Chen 2012).

Barley appears slightly earlier, around 2800-2700 BC, further west at the higher bend of the Yellow River. However, it is rare and may not be an indicator of agriculture but simply of some limited east-west trade at this time, as wheat and barley agriculture does not really start until the beginning of the second millennium BC in this area (Jia, X. et al. 2012).

Cattle and Sheep

Both of these animals were first domesticated in the Middle East or Anatolia by the 7th millennium BC. However, dating of their exact origins is complicated by the presence of almost indistinguishable wild forms.

Domesticated sheep (Ovis sp.) in ancient China have a poorly understood origin, partly due to the presence of wild sheep in China before domesticated ones appeared. Modern Chinese sheep are largely of a Mitochondrial DNA lineage known as lineage A and may either be derived from northern China or from a small population introduced from the west (Cai et a. 2007). Either way, it appears that they started to show signs of domestication in China in China’s central plain and to the west and the north during the fourth millennium BC (Liu et al. 2012).

The genetic situation with cattle is slightly clearer, suggesting an introduction of western (‘taurine’) cattle to northern China. Earliest domesticated cattle in northern and central China appear to date to the late third millennium BC (Liu et al. 2012).


Looking at the evidence, it’s tempting, as I did before, to see a very early link along the Silk Road,via the Hindu Kush, between the cultures of the west and east, allowing two-way movement of various products. However, current data implies that this route didn’t open until the end of the third millennium BC or later.

Generally, this period appears to be a time of making connections across the Old World, with routes opening up simultaneously between China and India (e.g. rice and chinese-style pottery in northern India), India and Africa, and the Middle East and India. Significantly, there’s no evidence of wheat agriculture in the Tarim Basin (around the Takla Makan desert) until this late date. This is also the time when the Bactria Margiana ‘civilisation’ (or BMAC) started to emerge to the west of the Hindu Kush in Turkmenistan.

In fact, if current evidence is anything to go by (and it’s not that good), it seems that the earliest appearance of western grains in China may have been considerably earlier than this, around the beginning of the third millennium BC, and via a northerly ‘steppe’ route, between the northern forests and the desert, and then down the northern bend of the Yellow River from Mongolia. This may have been made possible by cultures postdating the Afanasievo culture. Corroborating this, there are interesting connections between the bronzes of eastern China and those of the steppe cultures (Anthony 2007? and others).

If this is so, then cultures such as Afanasievo and Okunievo may, in fact, have been more agriculturally based than current evidence suggests, growing wheat and barley (and maybe buckwheat). It’s certainly interesting to note that the Altai Mountains’ location of the Afanasievo culture is now a rare patch of wheat growing in the eastern Steppe. Whatever, as Anthony (2007) comments, “a true understanding of the importance of plant foods in the steppes will come only after flotation methods are widely used in excavations”.

As for broomcorn millet, despite some indications to the contrary (Hunt et al. 2011) it appears to be a rare case of two domestication events, one in China, one around the Black Sea. If it is not, then something very important appears to be missing from our archaeological evidence.

In fact, this whole story is liable to change in the future as so much data is missing. However, what I’ve put down above is a best guess based on current evidence.


Anthony, D.W. 2007 The Horse, the Wheel and Language: How Bronze-age rivers from the Eurasian steppes shaped the modern world, Princeton, pp553.

Mid 6th millennium BC broomcorn on Dniester. 4000 BC broomcorn east to Dnieper. Anthony quotes Vainshtein and DiCosmo saying more recent Nomads raising barley and millet (not wheat).

Bellwood, P. S. 2008 First Farmers: The Origins of Agricultural Societies, Blackwell, pp360.

Cai, D. et al. 2007, DNA analysis of archaeological sheep remains from China, J. Archaeological Sci. 34, p1347–1355.

Chang, K. et al. 2005 The Formation of Chinese Civilization: An Archaeological Perspective, Yale pp363.

d’Alpoim Guedes, J. 2011 Millets, Rice, Social Complexity, and the Spread of Agriculture to the Chengdu Plain and Southwest China, Rice 4, 104-113.

Broomcorn millet in Sichuan 4000 BC.

Flad, R. K. 2010 Early wheat in China: Results from new studies at Donghuishan in the Hexi Corridor, ???????

The earliest direct dates of wheat in East Asia come from Donghuishan in Gansu Province, China. Few other dates of wheat in East Asia are direct dates. The previous direct dates at Donghuishan were obtained from wheat without secure context. New samples were taken from a stratigraphic profile at Donghuishan and directly dated. The wheat remains are earlier than any other directly dated wheat east of the Xinjiang Autonomous Region, but considerably later than the previously dated specimen from the same site. These new dates, from the early second millennium BC, are the earliest evidence of significant wheat and barley production and show that the Hexi Corridor played a critical role in the introduction of wheat to China.

Frachetti, M.D. et al. 2010 Earliest direct evidence for broomcorn millet and wheat in the central Eurasian steppe region, Antiquity 84, p993-1010.

Fuller, D.Q. & Boivin, N. 2009 Crops, Cattle and Commensals across the Indian Ocean: Current and Potential Archaeobiological Evidence, Etudes Océan. Indien 42, p13-46.

evidence for dispersal of millet around late 3rd millennium BC from perhaps central Asia south to India and Iran.

Hunt, H.V. et al. 2008 Millets across Eurasia: chronology and context of early records of the genera Panicum and Setaria from archaeological sites in the Old World, Veg Hist Archaeobot. 18, p5-18.

Hunt, H.V. et al. 2011 Genetic diversity and phylogeography of broomcorn millet (Panicum miliaceum L.) across Eurasia, Molecular Ecology 20, p4756–4771.

argues for a genetic split between eastern and western forms of Broomcorn millet, although not arguing whether both were from China or one was from central Asia/Caucasus. Problem for western origin is lack of genetic diversity in the western form which would be expected if so.

Jia, X. et al. 2012 The development of agriculture and its impact on cultural expansion during the late Neolithic in the Western Loess Plateau, China. The Holocene 23, p85-92.

1 barley grain from around 2900-2700 BC at Majaiao sites shows not major feature of agriculture, but does suggest some connection to west.

Li, X. et al. 2009 Holocene agriculture in the Guanzhong Basin in NW China indicated by pollen and charcoal evidence, Holocene 19, p1213-1220.

First appearance of buckwheat in China just east of Wei junction around 3500-3000BC.

Liu, L. & Chen, X. 2012 The Archaeology of China: From the Late Palaeolithic to the Early Bronze Age, Cambridge, pp475.

Lu, H. et al. 2009 Earliest domestication of common millet (Panicum miliaceum) in East Asia extended to 10,000 years ago, PNAS 106, p 7367–7372.

Syvatko, S. et al. 2009 New Radiocarbon Dates and a review of the chronology of prehistoric populations from the Minusinsk Basin, Southern Siberia, Russia. Radiocarbon 51, p243-273.

Sets the date of the Afanas(i)evo culture between about 2900 BC and 2500BC, succeeded by Okunevo culture which continued until 1900BC (no overlap with Andronovo)

Zohary, D. et. al. 2012 Domestication of Plants in the Old World: The origin and spread of domesticated plants in Southwest Asia, Europe, and the Mediterranean Basin, (4th ed.), pp251.








{ 1 comment… read it below or add one }

John Rudmin January 15, 2014 at 6:42 pm

I’ve wondered if there was a maritime connection ca. 6500 B.C., to explain the appearance of pottery (first invented in Japan?) in the Mehrgarrh culture of India and the Persian Gulf, and the appearance of cotton fishing nets (first invented in Mehrgarrh?) in the Norte Chico culture of Peru. Perhaps the seagoing intermediaries originated out of Okinawa (Sobata culture) or the Sunda region?

Also, across the far north, there have recently been a lot of ancient genetic haplotype results in archaeology, indicating migrations and intermixture between the trans-Baikal region to the east and greater Finland to the west, and in a few cases all the way to the NW European megalithic cultures (e.g. an mtDNA lineage D4 turning up).


Leave a Comment

Previous post:

Next post: