Ethnic Engineering: Scientific Racism and Public Opinion Surveys in Midcentury Japan

The science of race and the scientific study of public opinion are not normally bracketed together. Academic efforts to measure and define the physical characteristics of races reached a zenith in the first four decades of the twentieth century and then fell into decline, as the events of World War II revealed the inhuman ends to which this branch of human science could be turned. The science of public opinion research, on the other hand, was still in the relatively early stages of development at the outbreak of the war and was to come into its own in the postwar decades, when sociologists, psephologists, statisticians, and advertising agencies developed increasingly powerful tools to grasp and quantify the shifting moods of mass society. Although many of the key techniques of opinion research were pioneered in the United States, it was perhaps in postwar Japan that the art of measuring every imaginable aspect of popular opinion was most highly refined.

Racial science was and remains profoundly controversial; public opinion research, on the contrary, is generally recognized as a necessary instrument for the management of modern societies, and its underlying philosophies have attracted little critical attention (though its techniques and political uses have been extensively debated). Yet these two very different sciences overlap at certain points. Both apply statistical techniques to the task of defining the essential characteristics of human groups and of determining the differences that separate one group from another. In racial science, this was commonly achieved by biometry: the measuring of physical characteristics, such as height and skull size. In the case of opinion research, what is measured is not biological similarity and difference but similarities and differences of the mind, whether transient reactions to current events or more lasting attitudes, such as religious beliefs. In either case, the processes of research require drawing a line around the group to be studied and then averaging the measured characteristics of individuals within the predetermined group. In some cases, the results, in turn, become the basis of policies to improve the group, either physically (through eugenics) or mentally (through education, opinion leadership, or propaganda).

Almost any attempt to make sense of the infinitely complex web of human relationships requires a definition of boundaries that cut through the living substance of reality, separating (more or less arbitrarily) individual from individual and group from group. My purpose here is not to suggest that statistical surveys of human groups are necessarily invalid. By pursuing connections and differences between racial science and public opinion research, however, it is, I think, possible to chart crucial shifts and continuities in mid-twentieth-century social thought. In the Japanese context, this study casts a somewhat troubling light on the tensions inherent in the transition to postwar democracy. It also raises questions about the ways in which, even today, groups are defined, averages are created and interpreted, and social science is used to engineer social change. Here I shall focus on the relationship between the sciences of race and public opinion through the rather narrow lens of a specific case study: the story of the career and ideas of the scholar and social administrator Koyama Eizo (1899–1983).

Democratizing Public Opinion

A few weeks after Japan's defeat in the Pacific War, sociologist Koyama Eizo made his way through the burned ruins of Tokyo to an interview with the Allied Occupation authorities (popularly known as GHQ). He went, not on his own initiative, but somewhat nervously in response to a summons from the GHQ's newly created Civil Information and Education Section (CIES). In the autumn of 1945 both the Allied authorities and the Japanese bureaucracy were developing projects for systematic surveys of public opinion in Japan, and Koyama had been identified as a potential participant in those projects. But first, “in fear and trembling” (osoru osoru), he faced the ordeal of an interview...