Alan Sheridan’s translation of Foucault’s Surveiller et punir as Discipline and Punish is almost forty years old, and it is sometimes said that great works of literature need to be retranslated each generation. (For some examples of this for works of theory, see my post here). Foucault scholarship has advanced quite dramatically in the last forty years. The collected shorter writings, and especially the lecture courses, have given us a new sense of what Foucault was doing. The debates in the secondary literature have moved on too – Sheridan’s Michel Foucault: The Will to Truth was the first book on Foucault in English in 1980. Compare that book to more recent secondary studies and you’ll get a sense of how debates have changed.

Sheridan deserves enormous credit for the work he did, translating several of Foucault’s books and writing that first, important, study of his work. A good many of his translation choices are undoubtedly correct, and many of his phrasings felicitous. However there are several small errors, strange or unhelpful choices, missing words or phrases, and shifts in register that mar the reading. Additionally the frame within which he read and translated Foucault was the 1970s. Forty years on, we think with and appropriate thinkers in a very different context, arguably especially in the case of Foucault. In addition, Foucault is a much better writer in French than some of the translations would suggest, and some of his very deliberate stylistic choices are not respected. Here are some particular issues with the translation of Discipline and Punish… (page references are to French Gallimard Tel edition/English Penguin edition).

The title

Surveiller et punir does not translate as ‘discipline and punish’. Survey and Punish would be a closer title of the book. I know that Alan Sheridan explains his choice in the translator’s note, and discusses various possibilities for surveiller, but not, strangely, ‘survey’. Survey has a sense of both to oversee, and to catalogue.

The whole point is that discipline is made up of two elements – surveillance, of which the examination is a crucial element, and punishment. The danger of the current title is that it makes it look like discipline and punishment are discrete, when really one is contained within the other. Sheridan notes that “in the end Foucault himself suggested Discipline and Punish” which should be given due weight, but then, strangely, Sheridan adds that this choice “relates closely to the book’s structure”. In fact, the book’s four parts are entitled Supplice, Punition, Discipline, Prison.

Jeremy Gilbert has pointed out that the French surveillance and the English ‘surveillance’ are also not strict equivalents – the French having more of a sense of overseeing, or inspection, or supervision than covert, hidden surveillance, subterfuge or espionage. I think that’s broadly right, though it would be tricky to find a way to render this key term consistently without using ‘surveillance’, and using multiple English words would obscure as much as it would illuminate.

The Three Moments

The three moments of the change Foucault is discussing in the book gets somewhat lost in translation. The earlier model of spectacular public displays of punishment and torture—what Foucault calls supplice, which will be discussed below—to the model of discipline are the first and third of these moments. The second is underplayed and some of the translation choices reinforce this sense. The point is that the second is a direct opposite or contrast to the first, while the third retains elements of the first within a modified version of the second – “a torturous sediment [un fond «suppliciant»]” remains (23/16).

Foucault suggests “three ways of organizing the power to punish”. The first is the old, monarchical law; the other two are the alternatives – the reformers’ dream and what actually happened. He then proceeds to a set of alternative ways of conceptualizing the alternatives, each with a sequence of three, corresponding to the three ways, a rhetorical device somewhat masked by the English translation’s choice of punctuation and making this a single long sentence.

“The sovereign and his force, the social body, and the administrative apparatus [l’appareil]. Mark, sign, trace. Ceremony, representation, exercise. The vanquished enemy, the juridical subject in the process of requalification, the individual subjectified by immediate coercion. The tortured body [le corps qu’on supplicie], the soul with its manipulated representations, the trained body [le corps qu’on dresse]. We have here the three series of elements that characterize the three dispositifs that face one another in the second half of the eighteenth century” (155/131).

The problem, for Foucault, is why the third, and not the second, was adopted as the alternative to the first. “How did the coercive, corporal, solitary, secret model of the power to punish replace the representative, scenic, signifying, public, collective model? Why did the physical exercise of punishment [punition] (which is not supplice) replace, with the prison that is its institutional support, the social play of the signs of chastisement [châtiment] and the prolix [bavarde] festival that circulated them?” (155/131).

He returns to this three-way contrast right at the end of the book. “We are now far way from the country of supplices, dotted with wheels, gibbets, gallows, pillories; we are far, too, from that dream of the reformers, less than fifty years before [i.e. around the time of the French Revolution]: the city of punishments in which a thousand small theatres would have provided an endless multicoloured representation of justice in which the chastisements, meticulously produced on decorative scaffolds [échafauds], would have constituted the permanent festival of the Code. The carceral city, with its imaginary ‘geopolitics’ is governed by quite different principles” (307/359).

The Prison and a History of the Present

The prison of the subtitle is both the prison in a literal sense and the soul, Foucault stating the ‘The soul, effect and instrument of a political anatomy; the soul, prison of the body’ (38/30). The translation of that phrase is okay in Sheridan, though slightly different to mine. But Foucault says the purpose of the study is the following:

“It is of this prison, with all the political investments of the body that it assembles within its closed architecture, that I would like to write the history. By a pure anachronism? [Par un pur anachronisme?] No, if one understands by that to write the history of the past in the terms of the present. Yes, if one understands by that the history of the present” (39-40/30-1).

Three points here. First, he wants to write the history of this prison, not the or that prison (39/30-1). Earlier he makes it clear his study is “a genealogy or an element in a genealogy of the modern ‘soul’” (38/29). Sheridan is fine on this, though most readers seem to miss it.

Second, Sheridan translates ‘Par un pur anachronisme?’ as ‘Why? Simply because I am interested in the past?’ – which is at best an extremely liberal translation. I think it distorts what Foucault is saying. Pur can be sheer, worthless, but it’s not clear to me that it does have that negative connotation here.

Third, Foucault says he does not want to write ‘the history of the past in the terms of the present’, not ‘the history of the past in terms of the present’ as Sheridan has it. The second obscures the way that terminology and vocabulary, supposedly a remnant of his earlier ‘archaeological’ work, remains a crucial concern in his later, ‘genealogical’, writings. Foucault wants us to avoid reading the past with the conceptual terms of the present, which would indeed be ‘pure anachronism’, importing those terms back into a past that worked with different categories.

L’éclat des supplices and the term ‘supplice’

The chapter entitled ‘L’éclat des supplices’ is translated as ‘The spectacle of the scaffold’. It’s hard to know where to begin with this, and I have sympathy with Sheridan. Le supplice is torment or ordeal, so a form of torture, potentially a public version. The word alone is the title of the first part of the book, rendered by Sheridan there simply as ‘Torture’. L’éclat is fragment – coming from éclater, to burst, explode or shatter – or perhaps the splendour, brightness or brilliance. I know there is a tension between le supplice and la torture, both of which might be translated by the same word in English, but translators have had to work around this with other terms in Foucault – la connaissance and le savoir, for example as the two forms of knowledge, though Sheridan does not always do that here (i.e. 238-9/204; 346-7/296). But for ‘L’éclat des supplices’ there must be a better solution to something that currently invites comparisons with Debord, especially given Foucault’s argument runs against that of Debord (on this, see Bernard Harcourt’s note 4 on pp. 40-41 of La société punitive).

The single word le supplice is then translated by Sheridan in different ways in different places, often as ‘torture’, ‘public execution’ or ‘scaffold’; sometimes as the phrase ‘public torture and execution’ (37/28) or even ‘the process of torture and execution’ (56/45). (Scaffold is also used for the literal French term, l’échafaud.) The one to whom supplice is applied is le supplicié, which Sheridan renders as the ‘tortured criminal’ (i.e. 296/254). The term supplice is absolutely crucial to the analysis because Foucault’s first claim of what the book is about focuses on this term. The contrast between Damiens and the house of prisoners opens the book and then Foucault says “We have a supplice and a time-table… Among so many changes, I shall consider one: the disappearance of supplice” (14/7; see also the contrast on 299/257, or 308/264, etc.).

Foucault provides a helpful discussion on 42-4/33-4, which makes it clear that supplice is a technique, which cannot be “assimilated with the extremity of lawless rage”, and that not all punishments are supplice. To be supplice punishment must, first, be “calculated, compared, hierarchized… a quantified art of suffering”; second, it must be regulated, set in correlation between the “corporeal effect, quality, intensity and duration of suffering with the gravity of the crime, the person of the criminal, the rank of his victims”; and third, be part of a ritual, “an element in the punitive liturgy” that marks the body of the victim or the memory of the observer. Supplice must be stunning, spectacular or explosive [éclatant], almost like a (Roman) triumph. Triumph is a term Foucault uses a few times and has a strong resonance with the analysis later (and see 220/188).

Sheridan has the following passage which doesn’t make a lot of sense

“the term penal torture does not cover any corporal punishment: it is a differentiated production of pain, an organized ritual for the marking of victims and an expression of the power which punishes; not the expression of a legal system driven to exasperation and forgetting its principles, losing all restraint. In the ‘excesses’ of torture, an entire economy of power is invested” (44/34-5).

There are lots of issues with this. The most important is the ‘does not cover any’ in the first clause. Foucault says “ne recouvre pas n’importe quelle”, which surely means something like ‘does not cover all’. In other words, not all kinds of corporal, bodily, punishment, but only some kinds, which he goes on and qualifies. There are a lot of other small issues. I’d suggest this as an alternative.

“Penal supplice does not cover all corporal punishment: it is a differentiated production of suffering, an organized ritual for the marking of victims and a manifestation of the power which punishes; and not the point of exasperation of a justice which, in forgetting its principles, loses all restraint. In the ‘excess’ of supplices, an entire economy of power is invested” (44/34-5).

The Search for Truth

The discussion of evidence and proof (44-5/36-7) trades on analyses in earlier Collège de France courses, especially Théories et institutions pénales. The subsequent discussion of proof and confession also trades on earlier courses, and by this time Foucault is able to make broad statements with some confidence, rather different from the more tentative claims in the courses. The important relation between épreuve as ordeal or test and preuve as proof is not marked.

Compare Sheridan’s translation of this passage

“The search for truth through judicial torture was certainly a way of obtaining evidence, the most serious of all – the confession of the guilty person; but it was also the battle, and this victory of one adversary over the other, that ‘produced’ truth according to a ritual. In torture employed to extract a confession, there was an element of the investigation; there was also an element of the duel” (41)

with a more literal version, marking key French terms

“The search for truth through interrogation [la «question», perhaps even ‘inquisition’, but not Sheridan’s ‘judicial torture’] was certainly a way of obtaining evidence [un indice, perhaps ‘a clue’], the most serious of all – the confession [la confession] of the guilty [coupable]; but it was also the battle, and this victory of one adversary over the other, that ritually ‘produced’ truth. In torture [la torture] employed to produce an avowal [pour faire avouer], there was an inquiry [l’enquête]; there was also a duel” (52/41).

The dual use of avowal and confession mirrors the two French terms l’aveu and le confession, though we are only recently beginning to understand how Foucault was working with these, in the light of the Mal faire, dire vrai and Du gouvernement des vivants courses. Sheridan does not mark this distinction, using ‘confession’ for l’aveu, i.e. 48/37-8ff. Inquiry is a crucial term in the Théories et institutions pénales course. Foucault contrasts ‘inquisitorial’ justice with ‘examinatory’ justice later in the book (356/305).

Sheridan almost always neglects the specific attention Foucault gives to the notion of ‘measure’, explored in detail in Lectures on the Will to Know. For example: “Penal punishment is therefore a generalized function, coextensive with the function of the social body and each of its elements. This gives rise to the problem of the degree of punishment, the economy of the power to punish” (107/90). I think this should read: “Penal punishment [or chastisement, le châtiment penal] is therefore a generalized function, coextensive with the social body and each of its elements. This gives rise to the problem of the ‘measure’, the economy of the power to punish” (107/90).

Generally it is worth stressing how important measure, inquiry and examination are to Foucault’s work of the first half of the 1970s. They are central themes, respectively, to his first three courses at the Collège; they are treated in some detail in the 1973 ‘Truth and Juridical Forms’ lectures in Rio, which use and develop material from those courses; and they are crucial to this book. The first term, ‘measure’ appears in relatively few instances; the second, ‘inquiry’ is discussed in some parts (especially the chapters ‘L’éclat des supplices’ and ‘Panopticism’); the third, ‘examination’ is treated in much more detail (especially in the third section of the second chapter of Part III, which has ‘Examination’ as its title). But the link between the courses, the 1973 Rio lectures and this book is not clear with current inconsistencies in translation.

Power, Force and Punishment

Sheridan also sometimes translates force de travail as ‘labour power’ instead of ‘labour force’, which is confusing given Foucault’s understanding of power, le pouvoir, generally. Sometimes he translates force as power and pouvoir as force in the same sentence (61/50).

Foucault uses two verbs – and ones derived from them – that are translated by Sheridan as ‘punish’ without any explanation. These are châtier and punir – to punish, castigate, chastise (or perhaps to polish, refine), and more straight-forwardly to punish. Two common nouns that come from them are châtiment and punition.

Foucault frequently says le pouvoir de punir, the power to punish, but not always, and at one point his phrase la puissance du châtiment is translated the same way by Sheridan (108/90). It’s not clear to me that Foucault uses châtier/châtiment and punir/punition in clearly distinct, consistent ways – though there are moments when he comes close – but I think it needs at least a note. Perhaps chastise/chastisement could be used, even if only as an alternative word in the same way Foucault uses châtier/châtiment. But puissance should definitely not be confused with pouvoir.

Other Issues

Assujettisement is now frequently translated as ‘subjectification’; Sheridan prefers ‘subjection’ (i.e. 32/24; 34/26).

Compare “Judicial supplice is to be understood also as a political ritual” (SE) to Sheridan’s “The public execution is to be understood not only as a judicial, but also a political ritual” for “Le supplice judiciaire est à comprendre aussi comme un ritual politique” (58/47).

This passage clearly links to the last chapter of the first volume of the History of Sexuality (and the last lecture of ‘Society Must be Defended’), though Sheridan omits part of the quote: “the right to punish, therefore, is an aspect of the sovereign’s right to make war on his enemies: to punish belongs to ‘the right of the sword [droit de glaive – the phrase omitted], of that absolute power of life and death which Roman law calls merum imperium…” (59/48 quoting Muyart de Vouglans, Les lois criminelles de France, 1780, p. xxxiv).

Sheridan doesn’t capitalize le Code, as the Code, as Foucault does (i.e. 129/110). Foucault’s point is specific, to a moment in French history, usually to the 1810 Napoleonic penal code or the 1808 Code d’instruction criminelle. Foucault’s capitalization of State is generally not respected either.

Sheridan usually translates the key term dispositif as mechanism (i.e. 155/131). Apparatus is a common alternative in other translations and commentaries, even though that word also translates l’appareil (Sheridan sometimes uses  ‘apparatus’ for l’appareil, but also he sometimes renders that as ‘machine’ or ‘mechanism’ instead). It appears in several instances in this work – the Panopticon is described as a dispositif as is the army, and as are various institutions; and there are the three dispositifs contrasted throughout the book (see comment above). Whatever choice is made, it would be as well to signal the crucial word dispositif in Foucault’s work, even though it does not take on its major (and, I think, consistent) significance until the first volume of the History of Sexuality where, for reasons unknown, Robert Hurley translated it as ‘deployment’.

L’homme-machine is a book by Julien Offray de la Mettrie, usually translated as Machine-Man or Man, A Machine, not as Man-the-Machine. Foucault references la Mettrie and the book on the same page, but Sheridan doesn’t translate the title at that point, so the link is not clear to English readers (160/136).

Sheridan generally translates schémas as ‘projects’, but it is surely closer to ‘diagrams’, or even ‘schema’ (which Sheridan does use at least once), especially given the role Foucault gives to spatial organization. See, for example, the claim that leprosy gave rise to “rituals of exclusion”, the plague to “disciplinary diagrams [schémas]” (231/198). (N.B. while ‘diagram’ is an important theme in Deleuze’s book on Foucault the word there is diagramme, which can be found in this text – i.e. the description of the military camp as “the diagram of a power that acts by means of general visibility” [202/171].)

Sheridan also sometimes neglects the keyword le regard, the gaze – rather strange given he was the translator of the Birth of the Clinic, although even there he can underplay its importance, translating its subtitle Une archéologie du regard médical as ‘An archaeology of medical perception’.

Sheridan translates dresser, dressement as to train, training… and the title of the chapter as ‘The means of correct training’. I think it’s a shame that the English reader doesn’t realize that the other word Foucault uses for training is dressage, which in English is used for a horse event involving careful training of the horse (and rider’s) body. Foucault says, for instance, that correction “involves only incidentally expiation and repentance; it is obtained directing through the mechanics of training [dressage]. To punish is to exercise [Châtier, c’est exercer]” (211/180; see 155/131). I don’t necessarily think that dressage should be translated as ‘dressage’, especially because this would make the link to the other words difficult, but it’s worth noting, especially because I think the French dresser is used largely in relation to animal training, whereas there are other words for training of humans in the sense of sport, armies, police, education – former, entraîner, even suivre.

I’m not at all sure how Sheridan gets “The prison is the clearest, simplest, most equitable of penalties” from ‘Clarté en quelque sorte juridique de la prison” (269/232).

Foucault writes the book largely in the passive voice, with few active subjects of verbs. The English translation does not always respect this. Additionally, some of Foucault’s use of the impersonal pronoun ‘on’ and locutions such as ‘il faut’ become ‘we’ – such as “Il faut entrende le grondement de la bataille”, which Sheridan renders as “we can hear the distant roar of battle”, but probably should be “the rumble of battle must be heard” (360/308). Some, though not all, instances of Foucault’s first-person voice appearing in the text are the choice of Sheridan: Foucault tends to write in a more detached way. Given the argument of the book, of how subjects are made into objects, personalization and activation are quite fundamental problems.

Foucault is not good at avoiding gendered language, with les hommes appearing when the corresponding adjective is humain. But a lot of the ‘he/him/his’ language of Discipline and Punish is Sheridan’s, as a not always terribly elegant solution to the way the French language gives a gender to all nouns.

The English version only includes some of the images from the French. Most of the missing images used to be found at Jeremy Crampton’s site, but I don’t understand why they were missing in the first place. Tellingly, the English edition tends to have the ones most directly related to the prison, whereas the French edition, and of course the text itself, brings in a much wider range of examples.

Foucault uses footnotes with full references, and no bibliography. The English version provides a bibliography, turns most notes into short-form in-text references, takes other notes into the text, retains some but as endnotes, and occasionally mixes this up. Some examples from the first chapter of the English: on p. 3 the reference to source of the Damiens story is abbreviated in the text and the date of the edition of the Gazette d’Amsterdam is taken into the text; sometimes instead of a standard reference we are told in the text to ‘cf. Bibliography’ (i.e pp. 8, 23); and on p. 14 the whole of a note is taken in the text, in parentheses, with a reordering of sentences (this also happens on 283/244). Sometimes just the name of the author, but not the date, is taken into the text, so it’s not clear which text is being referred to (i.e. 90/76). At least one internal reference (p. 320 of the French edition) does not appear in the English version. Additionally, the final lines of the book, in the original, are “…the rumble of battle must be heard” (360/308) – discussed above. The final paragraph in the English was originally a note in the French (360 n. 1/308).

There are some additional errors noted on Clare O’Farrell’s michel-foucault.com site, most reproduced from Meaghan Morris’s notes in Power, Truth, Strategy.

Conclusion

A new translation might allow for a few more editor notes that point to things that are obscure today but would have been well-known at the time. For example, the reference to “the execution of Buffet and Bontemps at the Santé in 1972” (22/15) is to the two prisoners executed for killing a guard and a nurse in the Clairvaux prison; a case Foucault wrote about in his piece “Pompidou’s two deaths”. There is already a translator note on p. 268 of the English about recent prison uprisings. There could be a few more on links to lecture courses that expand discussions, or to explain difficult translation choices. If Foucault’s own notes were restored to the foot of the page, then editorial comments could be endnotes, as in the lecture courses.

Graham Burchell has given us Foucault’s voice in English in the translation of the lecture courses. Perhaps when he has finished the last of these – there are only three Collège de France courses to come in English – he should be commissioned to retranslate Foucault’s books. Discipline and Punish is probably the most important of these, given it is widely taught and still frequently the first book of Foucault’s to be read. The first volume of the History of Sexuality has some other major translation issues, which I hope to explore in a subsequent post, and I’m sure Birth of the Clinic and other books of the 1960s could be improved. The most pressing translation project is a complete, chronological translation of Dits et écrits, but given that many of the key texts have been cherry-picked for the Essential Works volumes, that may be logistically and financially unfeasible. But a new translation of Surveiller et punir, perhaps as Survey and Punish, in a related way to how History of Madness replaced the abridged Madness and Civilization a few years ago, might be a more realistic and viable undertaking. It should be stressed that such a retranslation would likely remain heavily indebted to Sheridan’s version, which on many points I find hard to fault, even if it would go beyond it in many other respects.

In the absence of such a retranslation of Surveiller et punir, I hope these indications of many of the major and some of the minor problems of the text are helpful as a companion to Sheridan’s version Discipline and Punish. I’d welcome comments, additions or alternative suggestions.