- Rhetoric and Power: An Inquiry into Foucault’s Critique of Confession
On October 10, 1979, Michel Foucault revised his thesis on confession. On that day, some three years after the publication of his magisterial treatment of confession in the first volume of The History of Sexuality, Foucault argued that the Pythagoreans, Stoics, and Epicureans had, before the advent of Christianity, their own practices of confession. Yet these practices, unlike their Christian variants, were not complicit with practices of domination (OS 310, 312). In the context of Foucault’s thought theretofore, in which the confession consistently figured as an essential component in the exercise of modern power, this is a radical shift. In lectures given at both Berkeley and Dartmouth in the fall of 1980, Foucault made similar arguments. On these occasions he argued that Seneca’s De ira constituted a “stoic confession,” the object of which was to enable the subject to “live differently, better, more happily, than other people” (AB 210, 205). Judith Butler captures the gravity of the shift: “In the last years of his life, Foucault returned to the question of confession, reversing his earlier critique in the first volume of the History of Sexuality, where he indicts confession as a forcible extraction of sexual truth, a practice in the service of a regulatory power” (2005, 112).
This reversal was short lived. Although Foucault would celebrate ancient Greek techniques of the self for the rest of his life, he would not [End Page 95] long refer to these techniques in terms of confession. Thus, in another reading of De ira, this one in March 1982, Foucault again noted that the Stoics have “confessional practices” that are “very close to what is found in Christianity” (Herm 482). This time, however, he insisted that the Stoic practices were fundamentally different: “However, all these elements seem to me to be profoundly different from what we should call ‘confession’ in the strict, or anyway, spiritual sense of the word” (Herm 365). In The Care of the Self, Foucault again conceded that De ira—what he now called “the most detailed description” of antique practices of the self—looked “at first glance” like a Christian confession, but he again maintained that it is something quite different: “The purpose of the examination is not therefore to discover one’s own guilt, down to its most trifling forms” (CS 61–62). Foucault concluded that the various “discourses of the self”—the confession and the Greek techniques of the self—could not be placed into a single category (see GE 250). Indeed, the “profound difference” (Herm 299) between these two discourses of the self was perhaps Foucault’s most repeated claim of the 1980s (see Herm 222, 237–44, 299, 308; GE 247–48; WS 237; AB 207–8; RFT 27; UP 89; CS 239; SS 5; FI 5, 12; FS 30, 35, 44, 87, 96, 97, 143–45, 147).
It is worth asking why Foucault returned so consistently to the distinction between these two discursive forms. This question gains even more relevance for readers of this journal when we remember that, on at least two occasions, Foucault defined one of these forms, the ancient techniques of the self, in terms of its opposition to rhetoric: the techniques of the self, he claimed, are “nonrhetorical” (Herm 368; FS 20). Thus the question that drives this essay is not so much why Foucault attached such importance to the distinction but rather how rhetoric itself might differentiate the confession from the techniques of the self. Stated differently, Foucault’s work on the confession vis-à-vis the technologies of the self provides a valuable opportunity to study the role of rhetoric in Foucault’s thought—not because they are both rhetorical forms but, curiously, because Foucault insists that one is not.
In order to foreground the work of rhetoric in Foucault’s thought, I situate his discourses of the self in the context of Nietzsche’s critique of metonymy. Although this may seem arbitrary, it gains initial plausibility from three facts. First, we know that Foucault understood rhetoric in terms of tropes (OT 84; Cooper 1988, 4). Second, we know too that Foucault had a deep...