Foucault’s main insight (in my view), is that power and knowledge are mutually constitutive, and that we can see this through “discourses.” Discourses are the conversations that societies have about topics, the assumptions that people have about those things, and therefore the things that are possible to say and even to think about something. I like to think of discourses as boundaries on what it is possible to “know” about something. We’re not really concerned with whether or not what people “know” about something is objectively “correct,” since any standard we would deploy to determine if that knowledge is correct would itself be part of a discourse. Further, in terms of explaining human behavior, it doesn’t really matter if what people know is “correct” or not, since they will act on what they believe to be true.
Discourses reflect power relationships, as some groups in society have the ability to shape discourses, and to in effect shape what is “true” for that society. The implication of this is that knowledge is not objective, because all knowledge exists within a discourse that is shaped by power. For example, the idea of the “nation” is a discourse, something that people can know which exerts power over us. Ideas of the nation hold that certain people are part of certain nationalities, that those people should be part of the same political unit, and that those people have certain characteristics and therefore should be governed in certain ways by certain individuals. The members of these nationalities “know” these things because experts construct knowledge about that nation’s past and its characteristics, this knowledge is preserved in archives and museums, taught in schools, and discussed by politicians.
Discourses serve to decenter power in history. In social history, we generally assume that change happens as a result of experiences people share. Groups of people experience something pretty much objectively, and react in pretty much the same way; power in history therefore resides in individual subjects reacting roughly objectively to transparent stimuli. However, Foucault’s ideas of discourses shaping what it is possible to know suggest that people do not experience things objectively. Rather, discourses make certain kinds of knowledge possible, and others impossible (or at least much more difficult), thereby governing—or at least putting boundaries around—people’s behavior. In this way of looking at history, power is not concentrated in individual subjects, but is diffuse, in the discourse itself, everywhere and nowhere at once. Instead of change in history being driven by the experience people share, cultural history argues that change in history is driven by the stories that we tell ourselves. These stories include things like “the nation,” “gender,” “medicine,” and one of my personal favorites, “nature.”
Thus, the idea of discourse opens up the historical study of categories that were previously understood to be pretty much stable, unchangable, and ahistorical: ideas of the nation, gender, sexuality, race, etc. Older histories assumed that categories of nationality were natural and objective, while now historians understand that the very idea of “the nation” has a history that reflects power relationships. In this way, Foucault has helped open up the study of history to include critical examinations of the very categories through which we understand the world.