Possible sources of the appearance of obscurity:
1) Vocabulary: French being Latin based looks more complicated than it is when viewed by an English speaker. I remember my surprise when I first arrived in France and a fairly unintellectual friend said “I just want to deculpabilise him”. In English that would sound highbrow, in French not so much. Another lexical factor giving the appearance of abstraction is the tendency, contrary to English, to prefer nominal expressions to verbal ones. “Natation” means “swimming”, and is an ordinary word in French, but it has an abstract feel to an English speaker. Further, at the level of meaning the suffix “-ation” seems more reified and static compared to the processual “-ing”.
2) Conceptual density: Accusations of Foucault’s intentionally complicated style are perhaps missing an important distinction. There’s complex and there is complicated. Of course there are poseurs everywhere. Sometimes a book is difficult because it can be more ambitious than a conversation or even a lecture. Deleuze’s classes were very clear, but his books were denser. If we compare some of the classes here with the corresponding books, and we can see that Deleuze is in fact a good stylist, in his writings he packs a lot more than he can say in a lecture. Surely some people like Continental philosophy’s language for its difficulty, but I don’t think that is the main aim of its demanding style.
3) Logical grammar: The comparison with the clarity of Descartes’ writing is misleading. I think people who study Descartes deeply soon find that he is not as clear and simple as he seems. Borrowing from Wittgenstein, I would argue that if you are saying something new and different yet using ordinary words you are changing their “logical grammar”, and so the difficulty merely retreats to that grammatical dimension that is often less apparent. Foucault thought with Blanchot and Klossowski that all ordinary uses of language are based on the privileging of unity and identity, and so that it required a special creative use of language to overcome that presupposition. Foucault always kept a poetic and allusive aspect to his writing in a deliberate attempt to avoid univocal signification and to respect the deconstructive ambiguity, multiplicity and fluidity of language. Foucault also tried to fight the unnoticed but limiting presuppositions behind our abstractions by going to an even higher level of abstraction to make manifest and relativise entrenched concepts, and transform them. Clarity depends on recognising the subtended ontology and conceptual grammar of a text.
4) Singularity: Later Foucault no longer sought the solution inside language but in relation to other practices. This is the period where he said that the new role of the intellectual was to “speak from a singular practice”, and I think his writing cleared up then. Psychoanalysis is, or can be, such a singular practice. But so is getting outside the academy and talking to prisoners and sick people and exiles and trying to help organise with them. Yet this is not a return to naïve transparency of language. The same effort is made to foreground that which usually is presupposed invisibly in the background and which creates the illusions of transparent communication and of the universal scope of one’s pronouncements.