Search results

The use of the concept of “agency,” in the sense of action that is to some extent free of “structural” constraints, has enjoyed enormous and growing popularity in the sociological literature over the past several decades. In a previous paper, we examined the range of theoretical rationales offered by sociologists for the inclusion of the notion of “agency” in sociological explanations. Having found these rationales seriously wanting, in this paper we attempt to determine empirically what role “agency” actually plays in the recent sociological literature. We examine a random sample of 147 articles in sociology journals that use the concept of “agency” with the aim of identifying the ways in which the term is used and what function the concept serves in the sociological explanations offered. We identify four principal (often overlapping) uses of “agency”: (1) purely descriptive; (2) as a synonym for “power”; (3) as a way to identify resistance to “structural” pressures; and (4) as a way to describe intelligible human actions. We find that in none of these cases the notion of “agency” adds anything of analytical or explanatory value. These different uses have one thing in common, however: they all tend to use the term “agency” in a strongly normative sense to mark the actions the authors approve of. We conclude that “agency” seems to serve the purpose of registering the authors' moral or political preferences under the guise of a seemingly analytical concept.