What is an Episteme?: An Explication via the Charge of Performative Contradiction

Episteme is a key concept in Foucault's first productive phase – considered commonly to extend until a shift in methodology announced in his 1971 essay Nietzsche, la généalogie, l'histoire – but it is still not very well understood. I think that a good way to illuminate what an episteme is to defend Foucault against the charge that he commits a performative contradiction when using the concept. The argument underlying this charge is often only hinted at in the literature, but there is a 1987 paper by Manfred Frank we can use as a foil that features the whole of it. It's an older one, but it checks out.

All of this mostly concerns Les mots et les choses (1966), in which Foucault is interested in a certain conception of "man" that underlies linguistics, economics and biology as representatives of the contemporary human sciences. Foucault ultimately reaches a historicist conclusion: "Man" was not always thought of in this way. Rather, the concept only appears in the late 1700s and is thus in a sense much younger than we as people influenced by the human sciences might think.

Les mots et les choses defines an episteme as a set of discursive rules governing propositional utterances in a given culture or society. Groups of these utterances are what constitutes knowledges, expressions of collectively held justified beliefs, be they alchemy or high-energy particle physics. epistemes can span disciplines: One upshot of les mots et les choses is that biology, economics and linguistics share their discursive rules.

Another is that epistemes appear and disappear at a given time. Chronological series of epistemes have sharp breaks between them. Les mots et les choses focuses especially on the break before modernity. For Foucault, this reveals anachronism in historiography: Biology, for instance, comes into being only in modernity – after the late 1700s break. Before, in early modernity (the age classique), we have natural history, which relies on a different set of rules to generate its propositions. Because of this, the difference between the two is quite fundamental and it would be a mistake to talk of the same discipline, although it might come easy to us to project the fundamentals of biology as we know it further back into the past.

Structuralism and the End of Modernity

One part of Frank's argument is that Foucault has preferences when it comes to certain epistemes. Firstly, in Frank's reading of les mots et les choses, Foucault (whom he understands to be a structuralist) predicts that the episteme of modernity is about to end and says that it is likely structuralism that will replace it.

Secondly, Frank thinks Foucault is playing favourites here. Foucault as a structuralist not only welcomes this new episteme, but, because the early modern order has some similarities to the "post-modern" one (I must emphasize that the term "postmodern" is not used anywhere in les mots et les choses) in regards to their consciously superficial approach to phenomena, he must also prefer early modernity to modernity. So in some way, Frank implies that Foucault advocates a return to a form of pre-modern "human sciences".

Without going into too much detail, I disagree with Frank on this side of the argument simply based on textual evidence. Indeed there is, for instance, an interview with Madeleine Chapsal (1966) in which Foucault declares the episteme of the tableau to be somewhat similar to structuralism, but predicting the return of an episteme is diametrically opposed to Foucault's deep-seated historicism. I therefore take this to be a superficial (pun intended) remark done for illustrative purposes vis-à-vis a broad audience. There are also some passages towards the end of les mots et les choses that make clear that Foucault sees some evidence that the episteme of modernity might come to an end soon – one of them being the literal end of the book where Foucault infamously stresses the possibility that "man" (in his technical sense) might disappear. He also thinks about about structuralism possibly being the harbinger of a new episteme, but these passages are clearly marked as speculative.

Episteme and the Historical Transcendental

More important for us is the second component of Frank's argument, which aims at the character of the modern episteme, which Foucault sometimes refers to as "history" or "man". A paradigmatic activity of natural history is the classification of plants as seen in Linné. Typically, it orders things on a tableau. We are talking about visible attributes. This is the aforementioned superficiality and transparency in early modern thought. In modernity, by contrast, animal physiology comes into focus. Life is now thought of in functional unities (organs and organisms) living in a milieu that shapes them. For whatever reason, all attributes, all representations need to be anchored now in man or animal and this anchor is conditioned by history. Think of how the discovery of the Indo-European language family shows that we have a historically contingent language. This is what Foucault means with "man" when he says that man only appears in Modernity and might disappear again. At the same time, history also appears, properly understood not just as the flow of time but something of impact that irreversibly shapes those that live within them. With man, an in-transparency takes hold because the forces that shape us – history, culture, language, material conditions, psychological makeup, even our faculties as conditions of possibility for all our knowledge – might influence what we believe, but they are not transparent to us.

But is this not what an episteme is? This is the second component of Frank's argument. If what we are able to say is constrained by given discursive rules, then an episteme is missing self-transparency in a subject engaged in knowledge-building. It sounds like a different gloss on other traditionally modern opacities, like the subconscious in Freud or ideology in a Marxist sense. episteme is a form of cultural relativism in the space of knowledge. Frank even traces the history of such relativisms back to the notion of the "historical transcendental" in Schleiermacher.

A student of Kant's writings, Schleiermacher of course is a thinker of the Modern Age just like Marx and Freud. Identifying epistemes is therefore considered a modern method. We can now see why Frank is downright confused by Foucault. How could someone that prefers a return to the transparency of the Classical Age use a modern concept? Foucault appears to him as a (non-apologetic) modernist that devalues Modernity.

Episteme as a Historiographical Concept

Foucault sometimes speaks of an episteme as a kind of subconscious of an entire discipline. This is confusing because a personal subconscious causally influences what someone does. The same holds for ideology or for Schleiermachers historical transcendental, but I don't believe this to be true of an episteme. Note that Foucault is uninterested in the people behind the discourse, their beliefs, their perceptions or even their circumstances. I think the text bears it out that Foucault uses the term "subconscious" as a simile, to explain what he means. This is, I think ,the crucial difference: Unlike the historical transcendental, an episteme does not cause the utterances it underlies. This is not the way they are primary to their discourses.

Ian Hacking has continuously stressed that the discoursive rules as Foucault conceives them can explicitly produce contradictory statements. Within the same episteme, Cuvier and Darwin disagree about whether organisms are changing or fixed. Epistemes are therefore not sufficient conditions for certain utterances and Foucault's disinterest in the circumstances of the speaker, who might combine with the episteme to cause certain things being said, suggests that he is not hot on them being necessary conditions either.

So what then is this thing that does nothing? The answer is, I believe, that epistemes are not in history at all, but purely historiographical concepts. "Man" does not reconstruct anything in the 19th century are, they are created by the historian examining a time, a methodical construct designed to help us explain or understand. epistemes come from a given archive, a set of documents containing utterances. The archive as a given is why they are also fundamentally unsuitable to discern actual epochs. All the assertions are taken and then we delineate the limits on what would count as a meaningful, reasonable or more specifically scientific statement at that point in time. Only in this sense does an episteme precede its discourses, is the former the construction rules of the latter. What we learn from them as historians is the limits on what can be said, not why someone said something.