Foucault’s approach to writing

From a fascinating interview with his partner Daniel Defert:

Daniel Defert: Absolutely! He said once to me in a phrase which I remember well, “Intellectual work doesn’thave enough materiality. One has to construct that materiality by working to a strict schedule, one has to work the same hours every day, just like one would in a factory.”

AB: Starting early in the morning…

DD: Probably not before 9a.m, actually. Butin fact it is very difficult to talk about his work, since a very large part of it was done in the library, surrounded by other people. I couldn’t say whether all his time in the library was spent reading, or if he also wrote things there, apart from justtaking notes. He generally left the house around 8:30 so as to get the the library around 9, and he usually left the library in the late afternoon, around 5:30 or 6. From then on, he would be meeting with people, either here or in town. This was the time of the day for his social and political life, meeting people, and then dinner—generally with people he knew well, his closer friends—Pierre Cabat, Mathieu Lindon, HervéGuibert, Thierry Voeltzel for instance—usually three or four people. These evenings among friends seldom continued beyond 10pm, and they were usually followed by a hour of reading—not the things people tend to imagine, the latest publications, the literary avant-garde—no, he would read Chateaubriand’s Memoirs From Beyond the Grave, and everything by Thomas Mann, Gogol or Kafka, and then at 11, sleep.

Three days of holiday would be enough to set off a neurosis! Foucault could accept leaving his work, as long as it was to work somewhere else, to go and give some lectures somewhere and use that as the opportunity to rework something he had done in Paris. Stopping work for a holiday, that was just not thinkable.

Not exactly.He went to the theatre, the cinema, to concerts—but that would still be an activity: because he would talk about it, he could give an analysis, provide a critique, while he was still walking out of the theatre. One sensed he had been following everything in an active way—so it was leisure in the sense of otium rather than farniente: non-work activity, not idling.

For sure. Foucault wasn’t rigid, but in reality those occasions were not so frequent.On the whole things followed the habits which I mentioned. If he didn’t go to the Bibliothèque nationale, he got down to work at home, in a kimono. The table over there [pointing to a teak table laden with books and papers] is where he wrote History of Madness—because he kept the working furniture that he had in Uppsala, like that chair [AB: the one I am sitting in now]. Discipline and Punishwas written on that white table [pointing to it], and then rewritten on that other one over there [pointing to the first table]

In short, I watched Foucault working without trying to understand how he worked.It’s only in retrospect that I have beenstruck by the multiplicity of themes he covered and their secretcoherence over time: the return of the same questions at different moments, with the new displacement given to them by each phase of his work.

He did his writing at home. He refused invitations to stay in houses of friends that were considered to be quiet places for working. What struck me was that he could alwaysbe interrupted. It didn’t annoy him. Flaubert said about GeorgeSand that you could interrupt her at any point in her work and she would always start it again in complete continuity, that nothing had ever distracted her from that. Well, if I could risk a daring comparison, I would say it was like that with Foucault.At the same time, I would say he was quite single-minded: if he was following an idea, it would absorb him completely. If I interrupted him, he seemed available to talk about something else. And then, after a moment, I would expressly turn the talk to what I knew was his obsession of the moment, and he didn’t see that one had changed the subject, he would go straight back into it. During the whole conversation, in fact, he had been continuing towork on his idea… That was how he could seem so available.

He was certainly organized. He created some disorganization later on: for each book, his notes would range over a considerable number of sources that he used for it. But each time he accumulated a collection of notes and excerpts on a topic, the same documentation could later come to serve another aspect of his research. One set of papers could get taken out of one pile and put in another… Similarly with the lecture series:it wasn’t practicableto put the effort he did into his Collège de France courses and then writewholly different sets of lectures for theStates. So in this case it hasless to do with re-utilisations of the same documents thanwith different perspectives onthe same problem

I think that in the writing of the books there are indeed three main layers, but each part of the book could have been rewritten a great many times. In the first place, he didn’t like crossing out. I was struck to find on the reverse of the notes of the lecture courses there were manuscript pages that came from elsewhere, which he had started to cross through and then abandoned because he didn’t like to work on a manuscript with erasures. He rewrote everything, rather than have a draft with deletions. His manuscripts are very neat, very tidy. Even with the lectures, one has the impression that it has been written all at once, but very often there are two or even three versions of a lecture.

I thought this bit was particularly interesting:

Basically when you see someone working, you don’t know how they work. You see them read or write, you don’t see them think. So, looking at the reading notes that Foucault was taking when he was looking for the historical point of emergence of a new concept (for example, the modes of empirical description), I imagined he was plotting a kind of Gaussian curve showing the appearance and then disappearance of a concept across the history of several disciplines. I imaginedFoucault practising a very empirical mode of reading, a sort of statistical survey. But François Ewald convinced me that that wasn’t how it happened, that the development of the concept precedes the readings, that everything is constructed beforehand—even if we don’t have an explicit trace of this. When Foucault starts to take notes, copies out quotations, things are already constructed. What one is seeing is not a statistical sampling process: he already knows what he is looking for in his corpus,which he must by then have already thoroughly mastered…I can also remember the phrase he liked to use when he was heading off to the the Bibliothèque Nationale; “I’m going to check whether people at that date were saying what they should have been saying!” So there is a dimension herein the construction of his thought for which we have no visible trace. In reality, the fragments that we have called Foucault’s ‘intellectual journal’ comprisebeginnings of articles, draft plans for a piece of writing: the thought has already been elaborated. But how Foucault got to that point I have no idea. I can bear witness to the quantity of work he put in, talk about his regular habits, but as for saying how the work of thought was performed—on that, I can say nothing