Craft Production and Social Change in Northern China (review)

Following on her innovative studies of ceramic production in Late Neolithic China, Anne Underhill of The Field Museum in Chicago has, in this book, undertaken a massive investigation of the textual and archaeological evidence for changes in craft production in relation to the development of complex society between 4000 and 1200 B.C. in northern China. She adopts an unabashedly processualist approach in focusing on social evolution and structuring her work by hypothesis testing. This results in probably the most meticulously systematic argument in the Chinese archaeological literature, where a rich array of Chinese historical and archaeological as well as worldwide ethnographical data are marshaled in its support. From a survey of the Western theoretical literature on complex social development, Underhill develops expectations about the nature of craft production at different levels of society. However, recognizing that food vessels are a dominant element in the grave goods during her investigative time slot, she uses early Chinese documentary sources together with ethnographic data to develop expectations about the use of food in funerary rituals and the ramifications of doing so for ceramic and bronze craft production. Her resulting model is perhaps most succinctly stated retrospectively in her final chapter, where she equates (pp. 242-243), where + means "plus," = "results in," and > "leading to":

little social inequality + widespread access to food surplus > small-scale feasting in society in general

marked social competition + considerable social mobility > large-scale feasting by a minority to negotiate economic power

high social inequality + restricted access to food surplus = low social mobility > few households host large-scale feasts

From models in social evolutionary theory, we should expect to see a gradual increase and intensification of both craft production and feasting in the archaeological record. However, Underhill reveals that the Chinese sequence is different in several ways-not least in the already mentioned prominent use of food and drink to negotiate status in funerary rituals. A surprising result of her analyses is that the Late Longshan, where we would expect to find more "attachment" of specialists to chiefly figures for production of prestige ceramics such as the known eggshell-thin blackwares, is actually a period of considerable regional variation without clear centers and with general access to rare goods despite their technical perfection. Even in the Shang period, when the production of bronze vessels is controlled by elites, central workshops are lacking, and craft intensification is evidenced by an increase in the spatial scale and increasingly fine division of labor. At Anyang, bronze production is still carried out with the household but with different households responsible for different parts of the production process or for different products, indicating not a change in the "mode of production" from earlier periods but an "increase in structural heterogeneity" (pp. 234-235). Nevertheless, Underhill includes this mode of production under the rubric of "attached specialization" and notes that such specialists produced not only prestige goods but also utilitarian wares for the elites. And finally, despite some crafts (prestige or utilitarian) being organized under attached specialists, there were also "independent specialists" who made goods of various kinds for all levels of people, including elites.

Such variability within one phase of the Late Shang capital's existence is amplified on a regional and temporal basis. Underhill's work makes it very clear that it is difficult to generalize about "China" or even about "northern China" in terms of the nature of funerary rituals and productive activities. She urges more regional analyses to be undertaken, especially paying attention to areas other than centers of power and especially looking at ordinary villages and cemeteries. Without these controls on regional and hierarchical variation, it is difficult to come to conclusions about the rise of power-holders in the Chinese sequence.

A couple of notes on the language used in Underhill's text: in reading it one will soon become aware of the presence of the triumvirate "should be," "would be," "could be." The first should...