Mabiki: Infanticide and Population Growth in Eastern Japan, 1660–1950 by Fabian Drixler (review)

Some historical puzzles bear repeated examination, as new approaches can lead to new insights. The historical phenomenon of infanticide in Japan is just such a puzzle, and Fabian Drixler’s Mabiki: Infanticide and Population Growth in Eastern Japan, 1660–1950 is just such a creative reexamination. Although Japan’s demographic transition from the Edo period to modern times has needed—and has received—numerous reconsiderations, Drixler’s book marks a new starting point for any future studies of infanticide in the context of this transition. The book’s success results largely from the way in which it synthesizes the disciplines of historical demography, anthropology, and religious studies, approaches that had grown far apart during the previous fifty years.

In terms of breadth of perspective, Drixler follows in the footsteps of Irene B. Taeuber, the pioneering demographer and sociologist who wrote numerous works on Japanese population trends and provided the first substantial treatment of mabiki (infanticide) in the English language in her magisterial The Population of Japan (Princeton University Press, 1958). There she discussed several methods used to limit family size, including contraception and abortion. In addition, another option was available:

The procedure easily and cheaply available to the people was not abortion but mabiki (thinning). According to the ancient tales, when the woman in a poor family was delivered, the midwife asked the family whether to let the infant remain, okimasu-ka, or whether to return it, modoshi masuka. The midwife either cared for the infant who must be assisted to survive or managed the death of the infant for whom there was no room.

(Taeuber 1958, p. 29)

According to Taeuber, this practice was widespread: “Many accounts from the late Tokugawa period show that infanticide was practiced in all classes of the population in all regions of the country” (Taeuber 1958, p. 29). Peasants, she argued, would have chosen mabiki in view of values that gave greater weight to the existing family, implying that membership in a family was not simply a birthright. While Taeuber did not go on to explore the religious aspect of what “returning” the infant might have meant to families during this period, she made it clear that the way in which Japanese peasants understood life and death helped guide their decisions regarding infanticide.

Two years later, Taeuber reassessed the role of infanticide during the Edo period in an article published in the journal Population Studies.¹ The main conclusion of this later analysis was that “whatever infanticide existed was not sex-selective” (Taeuber 1960, p. 32). Since Taeuber had started with the assumption that infanticide would have resulted in the deaths of more female than male infants, this conclusion suggested that the practice might not have been as widespread as previously surmised. Taeuber’s earlier examination of the issue, based primarily on research by Honjo Eijirō and Takahashi Bonsen, had considered not only the structural role of infanticide in Japan’s demographic transition, but also its cultural dimensions. Her later study, however, focused on a problem that could be addressed through statistical methods. This approach placed the study of infanticide firmly in the realm of historical demography and its increasingly sophisticated analytical tools.

Subsequent studies sustained this epistemological trend toward emphasizing the “hard” numerical sciences over “soft” interpretive fields such as anthropology and religious studies. The groundbreaking work of scholars including Hayami Akira, Kitō Hiroshi, Saitō Osamu, Robert Eng, Thomas Smith, Susan Hanley, Kozo Yamamura, and G. William Skinner examined the place of infanticide through the epistemological lenses of numerical data, household structures, economic rationality, and geography. Most concluded that the practice was important to Japan’s demographic transition. Critics of those scholars’ conclusions, among whom were Carl Mosk and Laural L. Cornell, also based their work on similar numerically focused methods.

In the meantime, William R. LaFleur’s Liquid Life: Abortion and Buddhism in Japan (Princeton University Press, 1994) pushed back in the opposite direction by including an...