and its rulers project their authority. Not unreasonably, then, the same was taken to be true of the intervening Neolithic and Chalcolithic periods.
It was only in the early 1980s, and then not without misgivings, that a Chinese scholar first publicly questioned this accepted view. He suggested it was ‘incomplete’, though one might now call it downright mistaken. Examples of dozens of distinct Neolithic cultures, like the ‘Hemudu Relics’, have been excavated at sites ranging from Manchuria in the extreme north-east to Sichuan in the west and Guangdong and Fujian in the deep south. None is significantly more ‘advanced’ than the others; and many more sites undoubtedly remain to be discovered. Indeed, later references to this period as being that of ‘Ten Thousand States’ (or ‘Chiefdoms’) may not be too wide of the mark.
As usual with Neolithic peoples, pottery provides a ready means of classification and so is used to distinguish them. Burial sites can also be revealing. But graveyards and ceramic workshops presume the existence of a settled population. The first conclusion to be drawn from the new discoveries is that settlement based on growing crops and husbanding domesticated animals was a development common to many regions of China and not just the north’s Central Plain. If millet was grown in the Yellow River region from perhaps 8000 BC, so was rice grown in the Yangzi region from about the same time. Silk production based on silkworm rearing, a form of animal husbandry unique to China, also has a remote provenance and is now known to have been practised in the Yangzi valley from at least the third millennium BC.
The links, if any, between these Neolithic cultures are as yet unclear. For the Indian subcontinent and for inner Asia, trails of diffusion have been proposed to fit the distribution patterns of pottery types and other distinctive artefacts; population movement in the form of migration, colonisation or conquest has often been inferred from them. But such theorising may owe something to retrospective assumptions. In both cases the incidence in later times of migrations, mostly inward in India, both inward and outward in inner Asia and Siberia, may have been projected back into prehistory. Consequently early settlement in these regions is supposedly fluid, with levels of technology uneven and population shifts frequent.
The more static model preferred in China may likewise reflect later historical orthodoxy. Neolithic cultures are grouped into regional ‘spheres of interaction’ rather than into peripatetic societies tracking across the face of the country; and attention is directed to those cultures and sites exhibiting the most in the way of continuity and internal development. Perhaps because so much archaeological effort was initially expended on the Yellow River basin in the north’s Central Plain, the key locations in this context are indeed concentrated in the north. Here, notable for their red pottery, often with painted designs, the so-called ‘Yangshao’ settlements (c. 5000–3000 BC and so contemporary with Hemudu), were succeeded by larger concentrations of the black-pottery ‘Longshan’ culture from about 3000 BC. Some ‘Longshan’ sites have urban proportions. Though centred in Shandong they are scattered over a much greater area than the Yangshao settlements. They introduce a building material called hangtu that was produced by pounding the friable loess soil into a concrete consistency; it would remain in use for the construction of foundations and walls until replaced by concrete itself in the twentieth century. And to the delight of archaeologists the ‘Longshan’ people honoured their dead with lavishly furnished tombs.
The size of some ‘Longshan’ tombs and the wealth and nature of their grave goods betray a highly stratified society. Privileged clans (or ‘lineages’) evidently exalted their ancestors in order to legitimise their own position, and through the mediation of this ancestry enjoyed a monopoly on contact with the gods. In this context they lavished on their dead both exotica, such as carved ivories, and a great variety of ritual objects ranging from vessels for food and drink to musical instruments and jade objects. Many such items incorporate pictorial devices known to have been used in shamanic intercourse with the supernatural world of ancestors and gods.
It all sounds mildly familiar. ‘Longshan’ society, or some part of it, could well have been that over which the Xia kings ruled. Erlitou, a Longshan type-site near Luoyang on the south side of the Yellow River in Henan province, has been confidently dated to c. 1900–c. 1350 BC, which roughly synchronises with the revised dates deduced for the Xia dynasty from later textual sources. Erlitou has therefore been tentatively assigned to the Xia. Moreover the site has yielded two types of material evidence, one apparently primitive, the other highly sophisticated, that connect its culture unmistakably to that of the later (or more probably overlapping) Shang and Zhou dynasties. In fact these material finds constitute prime sources for the social, cultural and political history of the second and early first millennia BC.
The first of them is burnt bones, mostly the shoulder blades of various animals that have been subjected to fire so as to produce a cracking. The cracking was ‘read’, much like entrails by the Greeks, to discover supernatural responses to human predicaments. More will be said of the practice, for it led to the earliest extant form of documentation and the first certain appearance of a written script in China. The other source material encountered at Erlitou, however, is even more sensational. For here were discovered some of the earliest examples of bronze-casting, a technology that more than any other defines ancient China’s culture and whose hefty products – urns, tureens, jugs – age-blackened or verdigris-tinged but otherwise deceptively pristine, still grace the galleries of the world’s museums.
Robert Bagley puts it better in the Cambridge History of Ancient China: ‘Artifacts of cast bronze are technologically and typologically the most distinctive traits of material culture in second millennium [BC] China…[and furnish] a revealing index of cultural development.’7 Indeed, bronze came to occupy much the same position in ancient China as stone in the contemporary civilisation of Egypt or, later, those of Iran (Persia) and Greece. Enormous effort was devoted to producing bronze-ware, highly sophisticated ideas were expressed through it, some of the earliest inscriptions are found on it, and its durability has ensured that plentiful examples have survived. Bronze production in China, though inferior in its labour requirement to, say, the great megalithic constructions of pharaonic Egypt, was yet on a sufficiently large scale to be rated an ‘industry’. Single vessels weighing close to three-quarters of a tonne have been excavated at Anyang in Henan province; elsewhere the total bronze component in one fifth-century BC tomb (at Suizhou) was found to amount to 10 tonnes. ‘Nothing remotely comparable is known elsewhere in the ancient world.’8
Compared to quarrying and carving stone, the technology involved in casting bronze was infinitely more demanding. Earlier small-scale production in Gansu province suggests that China’s metallurgical skills may have actually originated in China; certainly the abundance of suitable ores argues for an indigenous development, as do the advanced ceramic skills needed to create the moulds and achieve the high furnace temperatures for bronze-casting. The most impressive products were large vessels, often incorporating an udder-like tripod base but taking a variety of different shapes – known as ding, gue, jia, etc. – depending on their function as food containers, cooking pots, ale jugs, etc.
All at first replicated pottery designs but were then subjected to increasing elaboration in both shape and decoration. The ceramic moulds in which they were cast were themselves considerable achievements, with the decoration being incised on the inner side of the outer mould so that it emerged as raised on the finished product. (Engraving of the finished surface came later.) The moulds, both inner and outer, were cut into sections for the first pourings, typically three sections for the three-legged urn known as a ding but many more for more complex shapes. The vertical joins of the moulds ran up through the legs to the top of the vessel. Each section, including devices like spouts and handles, was cast separately but was recast as part of the whole in the final pouring. This eliminated any need for soldering or jointing while encouraging decorative designs, patterns and inlays, often with an animal motif, that were repeated within the subdivisions which resulted from the use of sectional moulds.
Ingenious as well as skilful, the technique underwent rapid development; so did the vessel shapes and the often fantastic ornamentation given them. Studying these variations, art historians