People who are quite fond of Foucault often tell me that Foucault does not understand himself as doing history and I should not read Foucault as a historian. With that said, how should I read Foucault?
I ask because it seems strange to me. When I read the first chapter of say Discipline and Punish it seems to me that Foucault is making what I would call a historical argument. He is talking about a change in practice from one time to another and then offering an interpretation of that change along with thoughts on what caused the change. That seems to me what historians do. (Albeit it is also what anthropologists and sociologists do.)
Am I misreading Foucault there? Am I being to general? Is there a particular way that I should Foucault rather than reading Foucault as a Historian?
(By the by, I like what I've read of Foucault.)
I am sure others have different views on this, but if you understand Foucault's work on his own terms, in Discipline and Punish he is doing the work of genealogy (italicized to make sure we are not confusing it with looking up one's family tree) and not history (or archaeology, which is what he said he was doing in Order of Things). For Foucault, both archaeology and genealogy are forms of philosophy that do use historical methodology (stories of change over time), but are less concerned with historical accuracy than uncovering unconscious shifts in discourse (the latter magic term being key to all-things-Foucault).
As a historian, I read this as something like a form of philosophical-intellectual history that does not mind taking very large flights of fancy and does not mind jettisoning (or just not discovering) inconvenient evidence when it comes to making a nice, complete-feeling argument. And this is of course historians' main problem with Foucault: the history doesn't totally add up.
So if you say, "don't read Foucault like a historian," what it means is, "don't cite Foucault for the actual transformation of 19th-century prisons as a historical subject; instead, see his work as a form of philosophy that while certainly informed by the past, is not at all exhaustive about the complexities of said past."
Now you might protest that this seems to let him get away with quite a lot. Which it does. It is a very generous reading of Foucault to say that his work is still useful even though there are probably lots of errors in its history. And saying "I don't understand myself as a historian" is a lame excuse for getting the history wrong (I don't understand myself as an accountant, but that doesn't stop the IRS from being all snippy with me when I make mistakes on my tax documents). One's perspective on this will entirely depend on how useful one finds Foucault. I find him useful, though again, I wouldn't cite him as the authoritative historical source on, say, prisons, mental illness, Victorian sexuality, and so on — if I engage with him, it's his broader ideas about power, sexuality, and so on, not the historical specifics.