Monday, November 17, 2008

G. William Skinner

G. William Skinner was one of the most innovative social scientists to have turned his attention to China in the past thirty years. Bill passed away on October 25, 2008, but his influence on how we think about China will survive him for a long, long time. (See this list of Skinner's publications to get an impression of the depth and scope of his contributions.) Bill was a generous and open scholar, and many scholars working in the field today owe him a permanent debt of gratitude for his advice and support in the past several decades. (See this Chinese obituary documenting Skinner's significance for Chinese scholars.)

Bill was a particularly fertile thinker when it came to using analytical and spatial models to explicate social reality in China (and occasionally Japan or France). (His work on Japanese demographic behavior is a great example; he devised an analytical framework for permitting inferences about family planning choices within Japanese peasant families demonstrating very specific preferences about birth order depending on the age and wealth of the parents. See "Conjugal Power in Tokugawa Japanese Families" in Sex and Gender Hierarchies, edited by B. D. Miller (Cambridge University Press, 1989).) Skinner was insistent that social data need to be analyzed in spatial terms; he believed that many social patterns will be best understood when they are placed in their geographical context. And the reason for this is straightforward: human social activity itself is structured in space, through transport systems, habitation patterns, the circuits of merchants and necromancers, and the waterways that integrate social and economic systems in pre-modern societies. He also believed that identifying the right level of geographical aggregation is an important and substantive task; for example, he argued against the idea of making economic comparisons across provinces in China, on the basis that provincial boundaries emerged as a result of a series of administrative accidents rather than defining "natural" boundaries of human activity.

Several ideas that Skinner developed in detail have had particular impact. His analysis of the marketing systems of Sichuan using the conceptual tools of central place theory was very influential when it appeared in three parts in theJournal of Asian Studies (1964-65) (part Ipart IIpart III). This analysis was illustrative in several key ways. It gave an important empirical instance for the abstract geometry of cental place theory -- the nested hexagons that represent the optimal spatial distribution of towns, villages, and cities. But more important, the analysis creates an important shift of focus from the village to the larger social systems of interchange within which villages are located -- the patterns of social intercourse that are associated with periodic markets, the flow of ideas associated with the circuits of martial arts specialists, and the likelihood of intersections between economic, cultural, and political processes rooted in the geometry of social exchange.

SOURCE: Skinner, G. William. 1964-65. Marketing and Social Structure in Rural China. Journal of Asian Studies 24 (1-3), pp. 22-23

A second highly influential idea also falls within the intellectual precincts of economic geography. Skinner offered an analysis of the economic geography of late imperial China in terms of a set of eight (or nine) macroregions: physiographically bounded regions consisting of core and periphery, regarding which the bulk of trade occurs internally rather than externally. Skinner argues in this body of research that it is analytically faulty to treat China as a single national market system in this period of time; rather, economic activity was largely confined within the separate macroregions. He used empirical measures to establish the distinctions between core and periphery, as well as to draw boundaries between adjacent macroregions. As he points out, the economic geography of traditional Chinese economy was largely governed by transport cost, and this meant that China's river systems largely defined the shape and scope of intra- and inter-regional markets. And he demonstrated how human activity was structured by the patterns of social interaction defined by these macroregions -- for example, the transfer of soil fertility from periphery to core through the gathering of fuel wood in the periphery, which then is transferred to core soils after burning.

SOURCE: Skinner, G. William. 1977. Regional Urbanization in Nineteenth-Century China. In The City in Late Imperial China, edited by G. W. Skinner. Stanford: Stanford University Press, p. 215.

Another critical contribution that Skinner provided, through his own contribution to the highly important City in Late Imperial China volume (link), is the idea of a hierarchy of urban systems. Skinner argued that there was an orderly hierarchy of places, ranging from higher-level cities through lower-level cities, market towns, and villages. He distinguishes between two types of hierarchy: administrative-bureaucratic hierarchy of places and the economic-commercial hierarchy of places. These two systems create different characteristics and functions for the cities that fall within them. This body of formal analytical ideas is borrowed from urban geographers such as Walter Crystaller and Johann Heinrich von Th√ľnen. Skinner's genius was to recognize that these analytical approaches provided a lens through which to make sense of Chinese social activity across space and time that other approaches do not. In particular, Bill demonstrated the utility of a spatial and regional approach in contrast to both purely statistical analyses of China's economy and village-level ethnographic studies that ignored the urban and town relationships within which village society was situated.

SOURCE: Skinner, G. William, ed. 1977. The City in Late Imperial China, Studies in Chinese society. Stanford, Calif.: Stanford University Press, p. 289

SOURCE: Skinner, G. William, ed. 1977. The City in Late Imperial China, Studies in Chinese society. Stanford, Calif.: Stanford University Press, p. 294

A central and crucial aspect of Skinner's thinking is spatial; he was vastly ahead of the GIS revolution in the social sciences, in that he consistently tried to analyze China's social, economic, and cultural data in terms of the spatial patterns that it displayed decades before the corresponding desktop computation capability was available. I visited his research laboratory at UC-Davis sometime early in the 1990s, and was struck by a couple of vignettes. When I arrived he was poring over a Chinese census atlas in eight gradations with a magnifying glass; he was laboriously coding counties by the color representing a range of social estimates. And when he brought me to examine a wall-sized map he had produced mapping sex ratios across part of southeastern China, he was interested in pointing out how the values of sex ratios corresponded to the core-periphery framework mentioned above. I pointed out a small, bounded region in southwest China that appeared to be anomalous, in that it represented an island of normal sex ratios in a sea of high male-female ratios. He instantly replied: that's an ethnic minority population that doesn't discriminate against girls. Culture and space!

Another of Skinner's crucial contributions to the China field -- and to historical social science more generally -- was his devotion to the project of creating a public database of historical Chinese social, economic, and cultural data at the county level. This effort contributed to the eventual formation of the China Historical Geographical Information System (CHGIS). What is striking about this work is that it was begun at a period in which the desktop computing tools that would permit easy and flexible use of the data -- in producing historical statistical maps, for example -- did not yet exist.

G. William Skinner provided a genuinely unique contribution to our understanding of the social realities of China. His contributions were innovative in the deepest sense possible: he brought an appropriate set of tools to each topic of investigation he addressed, without presuming that existing analytical techniques would do the job.

No comments: