Statistical practices have often been assumed as routine and prosaic, if not entirely value-neutral, activities of the modern state, and as more or less universal in nature across European and non-European countries. Yet, the recent STS literature reveals that the very content and exercise of these practices are deeply political, and that their evolution has taken different paths in different historical, political, and cultural settings. Colonial Authority and Statistics, a painstaking study of the statistical systems and censuses of the Government-General of Chosŏn (Korea) during the Japanese colonial period (1910–1945), is most welcome since it represents one of the first and few empirical studies touching upon this interesting nexus between statistical practices and politics/power in a Korean context.
After an introduction, the book begins with a chapter reviewing the development of modern statistical systems and practices in Japan since the Meiji Restoration. Ever since the terms and concepts of statistics and census were introduced from Europe in the 1860s, the use of statistical practices was perceived as an essential part of Japan's civilizing project, the most important example being the national census that began to be systematically conducted in 1920. But while the term statistics was by and large translated into tokei (統計)—meaning total sum—how to translate the term census turned out to be more controversial. Initially, behind the drive to replace the method of “inquiries per household (戶口調査)” by that of census must have been the need to establish a realistic estimate of population, whose mobility was being significantly affected by industrialization and urbanization. According to the authors Pak and Seo, however, there was also a political dimension. As the eventually accepted translation of census—kokuseichōsa (國勢調査), meaning a survey of national wealth or national prosperity (國勢)—implies, the shift to census reflected not only Japanese pride as a prosperous empire but also militaristic nationalism of the day (p. 30).
The next chapter provides a general discussion of modern, centralized statistical systems adopted by the Government-General of Chosŏn. While some scholars trace Korea's modern statistical systems to the second Kabo Reform period in 1895, Pak and Seo point out that such an effort would be futile since pre-colonial systems were already modeled after those of Meiji Japan (p. 42). The authors argue that the Meiji style of statistical practices, which gave precedence to the compilation of statistics for administrative use over field survey and research, persisted in colonial Chosŏn. For instance, the Government-General tended to gather data mainly from police or other official reportings, and did not establish a separate apparatus devoted to statistical data collection (pp. 45–46). Although the year-end count (每年末戶口) of household surveys was published annually, possibly as a consequence of this pre-kokuseichōsa style, a large portion of the collected data remained classified. It presents a very different picture from what Hacking has called “an avalanche of printed numbers,” referring to social statistics in nineteenth-century Europe.
Finally, in Chapter IV, Park and Seo turn to the implementation of kokuseichōsa in colonial Chosŏn, the main focus of their study. Japan's first attempt to introduce kokuseichōsa to Chosŏn in 1905 was thwarted by the Russo-Japanese War. In 1920, Japan once again planned an empire-wide census. Through this initiative, by placing each member of population within the grids of abstract modern time and space, Japan sought to demonstrate the power of its modernity and to achieve the integration of colonial subjects into the empire (pp. 63–65). Thus, the authors claim, censuses in colonial Chosŏn were not just neutral techniques to produce objective information to be used for administrative purposes, but an inherently political instrument of colonial power. Yet, their efficacy was at best modest. With the outbreak of the March First Independence Movement in 1919, the 1920 census could not be conducted. Though subsequent censuses were successfully executed every 5 years from 1925 till the end of Japanese rule, the Government-General of Chosŏn had to face difficulties in hiring enough surveyors and in gaining cooperation from Koreans. Even the Government-General itself relied more on the data from household surveys collected by the police than on those from kokuseichōsa (p. 119).
This relatively short, but meticulously researched, volume offers an excellent starting point for a contextualized, sociopolitical reading of statistical systems and practices in colonial Chosŏn. Unfortunately, the lack of theoretical perspectives makes it difficult for the book to purse that direction more thoroughly. Pak and Seo do not seem to engage with the recent STS and adjacent social sciences/humanities literature on statistics and governmentality. Some of the earlier works, including that of Foucault, are briefly mentioned and footnoted, and yet theoretical insights from these studies are not adequately used in the remaining of the book.
For example, the authors note that political arithmetic in seventeenth- and eighteenth-century Europe was closely connected with the state's desire to calculate the intrinsic value of its population, but assume that it disappeared from the scene, replaced by new, modern statistical practices. It is true that, in the mid nineteenth century, Europe witnessed a turning point in the history of statistics. In tandem with the emergence of a new idea of the social body, statistical sciences shifted from sporadically collecting numbers to uncovering general patterns of mass phenomena. However, one may approach political arithmetic not merely as a product of a specific historical moment but as a particular form of “the science of the state,” with an observing gaze at individuals in quantitative context rather than at an entire population as a demographic phenomenon. Seen that way, the statistical practices of the colonial state in Chosŏn were close to political arithmetic. Perhaps this partly explains why the Government-General, while celebrating kokuseichōsa as a superior and advanced technique appropriate for the modern Japanese empire, actually preferred to use household survey data. The political arithmetic aspect was further illustrated by the fact that, unlike in European countries at the time, demographic indexes such as longevity, fertility rate, or mortality rate never made their way into the Government-General's official statistics and census. My point here is not to privilege the notion of political arithmetic for its usefulness, but to emphasize that, contrary to what Pak and Seo suggest, statistics and censuses have never been homogenous entities invariably tied to monolithic power. The case of colonial Chosŏn was no exception. Therefore, we need a more theoretically sound approach that can account for the heterogeneous and versatile nature of statistical practices, the relationship of which to power and politics would be quite complex and multi-layered.
Another related drawback of this book is that its treatment of subject formation is somewhat too simplistic. While the authors conclude that kokuseichōsa did not succeed in integrating the colonial population into the Japanese empire, they seem to implicitly assume that, in principle, subjects can be formed directly through the techniques of power such as kokuseichōsa. As the book well illustrates, however, the Government-General of Chosŏn's collection of statistical data often met with resistance and protest. Censuses were also equipped with devices to distinguish and discriminate against Koreans. Given these findings, it is regrettable that Pak and Seo do not pay close attention to the ways in which kokuseichōsa and other statistical practices were intertwined with the identity formation of residents in colonial Chosŏn. Readers are left puzzled: did these practices prompt Koreans to imagine themselves as colonial subjects or as imperial subjects?
Despite the shortcomings, Colonial Authority and Statistics is a valuable book that deserves the attention of STS scholars or anyone concerned with the interweaving of knowledge practices, politics, and power. This book is also useful in highlighting the need for a dialogue between STS and other social sciences/humanities fields as well as for broadening the breadth and scope of future STS research. One hopes that there will be a follow-up volume, possibly on a more detailed study of the aborted kokuseichōsa of 1920 in colonial Chosŏnorona comparative study of different styles of conducting kokuseichōsa across the various Japanese colonies. In the mean time, interested readers will await the forthcoming book by the same authors, featuring a compilation of primary sources on the statistical systems of colonial Chosŏn.