Methods and Meanings: Reflections on Disciplinary Difference

Abstract

IN DEMOGRAPHY, a field long defined by its quantitative methods, a move toward qualitative methods at first blush seems revolutionary. Not surprisingly, then, there has been much interest in the qualitative turn in demographic methodology, accompanied by words of caution from some anthropological quarters. I do not intend to address the question of whether this shift is productive for population studies. Instead, my aim in this commentary is to relocate the discussion of qualitative methods, to remove it from the realm of methodology and place it in the arena of disciplinarity. By disciplinarity I mean the study of academic disciplines, more specifically, study of the control of knowledge organization and production by academic disciplines and the social practices by which that control is maintained (for more, see Messer-Davidow, Shumway, and Sylvan 1995). Although, institutionally, demography is interdisciplinary, with members based in different academic departments, as an intellectual practice demography operates as a relatively distinct discipline. Focusing on disciplinarity allows us to talk about disciplinary histories and disciplinary cultures, which in turn provide entree to fresh perspectives on methods. Let me begin by backtracking a bit. In a recent essay, titled The social construction of population science, I argued that demographic knowledges about fertility have been shaped by the discipline's distinctive history and culture (Greenhalgh 1996). More specifically, I argued that the late arrival of demography on the American social science scene and the close ties it was constrained to build with applied funding organizations had specific, identifiable effects on the types of thought that were formed. Various theories of fertility differ in many ways, but they all share three larger properties: a lack of interest in politics, a general neglect of history, and an assumption of a global distribution of values and power that is Eurocentric, in other words, that places the val-