In an earlier post in this series I discussed a misplaced reference in Foucault’s Birth of the Clinic, where Foucault’s reference pointed to a book by S.A.D. Tissot but the actual source of the quotation was a text by Archibald Pitcairn. In that instance, Foucault had either misnumbered his notes in his handwritten text, or mixed them up when typing the text.
As part of the work for the new translation and edition of this text, we are checking, completing and correcting all of Foucault’s references. As I mentioned in the previous post, Noah Delorme has done a first pass through the notes, checking whether they are in the places Foucault says. Many references are correct, but there are enough mistakes to make this necessary but very time-consuming work. Unless Noah could find the mistake fairly quickly – notes out of sequence, a slip in the title, a quote on a different page in the same work – he has marked them for further investigation.
Having access to so many old medical books online is, of course, incredibly helpful for this. Google books, HathiTrust, internet archive, Wellcome collection, Gallica, etc. Some can be resolved with a check in a library for a text which isn’t easily available online, so I made lists of things to check at the British Library, the Wellcome Library or the Bodleian. So far, fortunately, only a few look like they might need to be checked further afield. There have been a few other challenges in the notes, a couple of which I’ll discuss here.
One was a Latin quotation, with no translation, and with a reference which didn’t check out. Foucault references a letter and article in a work, not the page number. This should make it easier to find in any edition, if the reference was correct. I couldn’t easily find the passage by searching, so I went back to Foucault’s handwritten notes on the Fiches de lecture site.
Once I located the relevant note, I was able to work out what was going on. It seems he transcribed his own notes incorrectly, writing ‘faber’ instead of ‘febres’ (maker or craftsman vs. fever). I was then able to find the passage in the text he was quoting from, which was in the correct letter, but he had referenced the wrong article. I suppose I should be grateful he got the author and work correct. The French text has gone through four editions – the two editions in Foucault’s lifetime; the retyped version in the Quadrige series (after the earlier Quadrige version reused the second edition pages); and then the version in the Pléiade Œuvres – without this being spotted. Given it is in a chapter called “The Crisis of Fevers [La crise des fièvres]” the word slip really should have been noticed before. The passage is simply copied in Latin with the error and the incorrect reference in the existing English edition.
Some other reference errors are because Foucault uses ibid and loc. cit. but this is sometimes for authors where he references more than one work by them in the same chapter, and these can easily point to the wrong place. Successive editions/translations have just copied these references, repeating the errors, although François Delaporte corrects some in his notes to Œuvres.
If the work referenced seems at least plausible but the page doesn’t match, you can spend a lot of time looking in that work for something which is not there. Then you check another book and are more successful, though the page is sometimes wrong too (or is correct for a different edition). One of the challenges is that several of the books referenced have reset pages between different editions, and this means the pagination is different. So, to check a quotation you need to find the same edition Foucault used, which is not always the same as the one he referenced.
I am hardly the first to notice that Google search for quotations is much less effective than it used to be. A search for a short passage between quotation marks could often provide the source, or perhaps another place which quoted it, and then finding that and checking its notes might lead to the real source. But recently I’ve found that because Google often spits out some AI nonsense first, it will say something like “this is a famous quote by [Foucault’s source] in [incorrect reference]”, therefore replicating the error or mistaken reference which I’m trying to correct.
There are so many pirate versions of Foucault’s books online, and in translation, or quotations from them (with copied and unchecked references) that it seems to take repetition as confirmation. But this can mean that the error is reproduced rather than corrected.
I know why earlier publishers, translators and editors didn’t do this work of reference checking and correction – it is incredibly time-consuming, and largely thankless. As with much editorial work it isn’t noticed if done properly, but only if it is not done or done badly. Before digital search it would have been much harder, as I remember from my PhD and early-career. The notes for the first couple of Henri Lefebvre books I worked on were done before these kind of searches were easy to do. But it feels like things are now going backwards. I’m not completely anti-technology, and I can believe that some kind of mechanised process could, conceivably, make reference-checking much easier. I’m not sure it would be artificial ‘intelligence’, but perhaps valuable support. But at the moment it feels like the technology is inhibiting the work rather than helping it.
A final example was when Foucault quotes a law from the Revolutionary period, and the date is given is in the Republican calendar. A date in the Gregorian calendar is provided in a note. These two dates have been copied in all French editions and Alan Sheridan’s translation. In Delaporte’s note he gives a reference to a compendium of laws for this date. But the law Foucault is quoting from is not there, and some searching didn’t identify what it could be. I then realised the Gregorian date was wrong – the day given actually predates the introduction of the Republican calendar. Though the Republican date was applied retrospectively, at the time of the stated Gregorian date they might have called it Year II, but would not have used the Republican date. Once I realised this, and worked out what the real Gregorian date was, I was able to identify the correct law.
Generally, the political speeches and laws were the hardest to locate, since they were often referenced from obscure editions which are hard to find outside of France. The British Library has a lot of these, many bound together in a couple of volumes, but by no means all the ones Foucault references. Chapter Five presented the most problems here.
There are hundreds of references in the book, and some can be really difficult to track down. The number incorrect is relatively small, but these often present challenges. Delaporte’s notes have solved some, and having access to Foucault’s reading notes online has helped with a few more. We are down to only a handful as yet unresolved.
This is the 92nd post of a weekly series, posted every Sunday throughout 2025, and so far through 2026. The posts are short essays with indications of further reading and sources. The pieces are not as formal as something I’d try to publish more conventionally, but are hopefully worthwhile as short sketches of histories and ideas. They are usually tangential to my main writing focus, a home for spare parts, asides, dead-ends and possible futures. I hope there is some interest in them. They are provisional and suggestions are welcome. A few, usually shorter, pieces in a similar style have been posted mid-week.
I discuss what I am doing in this series in “Writing the ‘Sunday histories’ – an experiment in sharing short pieces on a blog”, my contribution to the Northern Ireland Legal Quarterly “Reflections on Writing” Series – all pieces are open access
The full chronological list of ‘Sunday histories’ is here, with a thematic ordering here.
