Abstract

In this paper, I argue against Mark Bevir’s contention that Foucault is committed to reject autonomy but affirms agency. I argue that Bevir’s claim is extravagant, that (a) Foucault need not reject autonomy, and (b) the rejection of autonomy flies in the face of the emphasis Foucault placed on autonomy in his late work. Central to my argument is the distinction between autonomy as self-sufficiency and autonomy as self-rule. By deploying Harry Frankfurt’s structural account of identification, I show that Foucault’s discussion of autonomy in his late work is best understood as self-rule. Such an interpretation has the virtue of not adopting the counter-intuitive view that Foucault is committed to reject autonomy. By way of conclusion, I explore how such a concept of autonomy as self-rule figures in Foucault’s project of self-formation in his late work.