OrgTheory has an interesting recent thread about the frustrating task of reading and teaching Foucault. Various commenters note that Foucault’s lack of citations reflects a broader pattern of French scholarship, which uses different criteria for citations (i.e. tending not to cite living rivals). Omar suggests that Foucault’s style opens up space for a productive secondary literature that elaborates the connections between authors in useful ways, and linked to a fascinating piece:

A Discourse on Discourse: An Archeological History of an Intellectual Concept (Cultural Studies, 2002) by Sawyer. In this article, Sawyer traces the history of the modern concept of discourse. Sawyer argues that English-language scholarship, starting in the 1980s-1990s, began to attribute the concept of discourse to Foucault, even though Foucault himself abandoned the term, and had never used it in the broad sense most cultural studies scholars employed it in. Rather, Foucault used discourse to refer narrowly to a collection of statements. Foucault distinguished in his work between discursive and non-discursive practices, and even in his earlier work rejected many broader notions of discourse (see especially his treatment of the term in Archaeology of Knowledge, described in detail by Sawyer).

To make matters worse, Foucault actually stood against the three figures of Lacan, Althusser and Saussure (the “Triple Alliance” of French thought) who inspired British cultural studies scholars in the 1970s-1980s to begin speaking in terms of this broad concept of discourse. Sawyer theorizes that Foucault was brought into the mix to fill certain gaps, particularly the desire to add history to the concepts of the “Triple Alliance.” Later, Foucault gained a much wider readership, and thus the concept was eventually attributed to him.

Perhaps the most useful thing Sawyer does is lay out a nice description of the broader usage of discourse prevalent in cultural studies, sociology, and so on. Sawyer argues that discourse replaced “culture,” “ideology,” and “language,” in the vocabulary of cultural studies because it solved a variety of problems:

‘Discourse’ uniquely satisfies these multiple requirements. ‘Discourse’ has captured the totalizing and semiotic connotations of ‘culture’, combined it with the Gramscian and Althusserian notions of ‘hegemony’ and ‘ideology’, blended it with Lacanian psychoanalytic concepts, tapped into the linguistic turn in literary theory, and then introduced Foucault’s historical perspective on power/knowledge relations. ‘Discourse’ thus retains many connotations of 1970s Marxist and Lacanian theory, but in a way that allows the incorporation of history, culture and both structuralist and post-structuralist insights. It is not surprising that such an all-encompassing term is now associated with a wide range of conflicting and confusing meanings, such as those quoted at the beginning of this article; perhaps this is simply too much weight for a theory to place on one word. (449-450)

That sounds right to me, and helps to explain why discourse seems to mean so many things to so many people. Though not exactly intended as a definition of discourse, it works brilliantly by describing the problems that ‘discourse’ solved and continues to solve. That being said, the article ends with the reasonable argument that we might want to narrow our usage of the term, perhaps back to Foucault’s usage, in order to regain some clarity.

I wonder though if Sawyer overly privileges rigorous terms: given the constant (re)emergence of messy concepts that blend conflicting theoretical traditions and lump together distinct modes of analysis and objects of study, it seems hard to argue that they always represent an impediment to understanding.* Perhaps we need some way of asking ourselves when a concept has grown too big, unwieldy and all-encompassing, and it becomes worthwhile to break it back apart into littler pieces.

Returning to Foucault’s role in all of this, I have one one dissatisfaction with the article. I was hoping that this archaeological voyage into the history of “discourse” would help me understand Foucault better, would expose some misconceptions, something like that. Sawyer opens the door to such arguments by disentangling Foucault’s rich vocabulary in his early work, and his decided break with discursive arguments in his later work, from the messy concept of discourse we rely on today. But the article stops there, without telling us how, if at all, this should change our reading of Foucault. Still, fascinating grist for the theoretical mill and a really helpful jumping off point for that sort of analysis.

* The STS literature on “boundary objects” (Star and Griesemer) and “trading zones” (Galison) seems relevant here. A precise definition used by all individuals is not the hallmark of a good “working object” of knowledge (cf. Bowker and Star), though they do require some coherency.