This book is supported by the French Ministry of Foreign Affairs, as part of the Burgess programme run by the Cultural Department of the French Embassy in London (www.frenchbooknews.com)

Ouvrage publié avec le concours du Ministère français de la Culture – Centre national du livre.

Published with the assistance of the French Ministry of Culture – National Centre for the Book.

Polity Press

65 Bridge Street

Cambridge CB2 1UR, UK

Polity Press

350 Main Street

Malden, MA 02148, USA

All rights reserved. Except for the quotation of short passages for the purpose of criticism and review, no part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the publisher.

ISBN-13: 978-0-7456-4641-1

ISBN-13: 978-0-7456-4642-8(pb)

ISBN-13: 978-0-7456-8380-5(epub)

ISBN-13: 978-0-7456-8379-9(mobi)

A catalogue record for this book is available from the British Library.

The publisher has used its best endeavours to ensure that the URLs for external websites referred to in this book are correct and active at the time of going to press. However, the publisher has no responsibility for the websites and can make no guarantee that a site will remain live or that the content is or will remain appropriate.

Every effort has been made to trace all copyright holders, but if any have been inadvertently overlooked the publisher will be pleased to include any necessary credits in any subsequent reprint or edition.

For further information on Polity, visit our website: www.politybooks.com

No, Foucault was not a structuralist thinker; nor was he the product of a certain line of ‘1968 thinking’. Nor was he a relativist or a historicist; nor was he bent on sniffing out ideology everywhere. He was something that, in this day and age, is rare, a sceptic thinker¹ who believed only in the truth of facts, the countless historical facts that fill the pages of his books, never in the truth of ideas. For he acknowledged no transcendent principles as the foundation of truth. Yet he was not a nihilist; he recognized the existence of human liberty (it is a word that frequently occurs in his texts) and he did not think that, even when set up as a doctrine of ‘disenchantment’, the loss of all metaphysical and religious bases ever discouraged that freedom from having beliefs, hopes, indignations and revolts (he himself was an example, for he was a militant of a kind, a new type of intellectual; and in politics, he was a reformer). However, he considered it to be false and pointless to argue about his battles, to wax loquacious about his indignations, or to generalize. ‘Do not use thought to confer the value of truth upon any political practice,’ he wrote.²

He was not the enemy of man and humanity that he was believed to be. He simply reckoned that humanity could not get any absolute truth to descend from heaven or to operate, in sovereign manner, in a heaven of truths. He believed that all he could do was react to the truths and realities of his time and perhaps respond to them in an innovative fashion. Like Montaigne, and in direct contrast to Heidegger,³ he reckoned that ‘we have no means of communication with Being’.⁴ Notwithstanding, his scepticism did not cause him to cry, ‘Ah, everything is doubtful!’ You could say that this supposed sixty-eightist was an empiricist and a philosopher of understanding, not of any presumptuous Reason. Without being too insistent about it, he ended up with a general conception of the human condition, the freedom with which it could react to things, and also its finite nature. Foucault's philosophy is, in truth, an empirical kind of anthropology with a coherence of its own, the originality of which is founded on a historical critique.

Now let me move on to details, but not before, in the interests of clarity, I first state the two principles by which I work. 1. What is ultimately at stake for human history, even above power, the economy and so forth, is the truth: what economic regime would ever confess to being false? But this problem of historical truth has nothing, absolutely nothing to do with any questioning of the innocence of Dreyfus or the reality of the gas chambers. 2. If historical understanding, for its part, desires to push its analyses of a given period as far as possible, it must move on from societies and their mind-sets, to the general truths in which minds in that period were, without realizing it, trapped like fish in a glass bowl.

A sceptic, meanwhile, is a dual being. So long as he keeps on thinking, he remains outside the fishbowl and can watch the fish swimming round and round inside it. But given that one has to live, he then finds himself within the bowl, a fish like all the rest, faced with deciding which candidate he will vote for in the forthcoming elections (even if he cannot claim his decision to be based on the truth). This sceptic is at once an observer, outside the fishbowl about which he has misgivings, yet also one of those very goldfish. However, there is nothing tragic about that duality.

The name of the observer who is the hero of this little book is Michel Foucault, a slender, elegant and decisive person whom nothing and nobody could force to back down and who, with his intellectual cut-and-thrust, handled a pen as though it were a sword: which is why I might well have entitled this book The Samurai and the Goldfish.

When The History of Madness was published, some of the most well-disposed French historians (including myself) failed at first to appreciate the scale and significance of the book. I thought Foucault was simply showing that our conception of madness has varied greatly in the course of the centuries. But that told us nothing new; we already knew that human realities betray a radical contingency (well known in the form of the ‘arbitrariness of cultures’) or, at the very least, are diverse and variable. There are no historical invariants, no essences, no natural objects. Our ancestors developed strange ideas about madness, sexuality, punishment and power. But it was as though we ourselves had silently recognized that those days of error were over, and believed that we were doing better than our ancestors and had discovered the truth around which they had stumbled. ‘That Greek text speaks of love in accordance with the ideas of the time,’ we would say to ourselves. But was our modern idea of love any better than theirs? We should not have presumed so to claim had that apparently trifling and unimportant question been put to us. But do we, even now, think about it seriously and philosophically? Foucault did.

I had not realized that, without claiming to, Foucault was taking part in one of the great debates in modern thinking: does truth, or does it not, correspond to its object; does it or does it not resemble what it states, as common sense supposes? The fact is that it is hard to see how we could possibly know if it does resemble what it states, since we have no other source of information that might offer confirmation. But let that pass. For Foucault, as for Nietzsche, William James, Austin, Wittgenstein, Ian Hacking and many others, each of them with views of their own, knowledge cannot be a faithful mirror to reality. No more than Richard Rorty¹ does Foucault believe in that mirror, or in that ‘specular’ concept of knowledge. According to him, the object, in all its materiality, cannot be separated from the formal frameworks through which we come to know it, frameworks that Foucault, settling upon an ill-chosen word, calls ‘discourse’. That, in a nutshell, says it all.

Misunderstood, this concept of truth not corresponding to reality has made some people believe² that, according to Foucault, mad people were not mad and that to speak of madness was nothing but ideology. Even a man such as Raymond Aron believed this to be the line taken by A History of Madness, as he told me, without beating about the bush; he protested that madness is all too real: you only need to see a madman to be sure of that. And he was quite right. Foucault himself held that, even though madness was not what its ‘discourse’ claimed it to be, that did not mean that there was no such thing.³

So what does Foucault mean by ‘discourse’? Something very simple: a most precise and close description of a historical formation, stripped bare, a revelation of its ultimate individual difference.⁴ Reaching the differentia ultima of a dated singularity requires an intellectual effort of perception: it is necessary to strip the event of the excessive draperies that make it unexceptional and rationalize it. The consequences of doing so are far-reaching, as we shall see.

In his first book, Foucault's heuristic starting point was a classification of the ‘discourse’ on what we call madness (or insanity, as earlier ‘discourse’ put it). The books that followed offered other subjects to exemplify the sceptic philosophy that he had developed from his attention to details. However, he himself never fully expounded his doctrine, but left that redoubtable task to his commentators.⁵

In the present work, I shall be endeavouring to explain to myself the thought of this man who was a great friend and, it seems to me, a great mind. I shall be citing repeatedly from his Dits et Ecrits (Sayings and Writings), as he refers there to the bases of his doctrine more often than he does in his major works.

Before I take the plunge, though, let us consider an example. Suppose we were planning to write a history of love or of sexuality through the ages. We might feel satisfied with our work when we had reached the point at which a reader could learn about the variations that pagans or Christians had elaborated, in their ideas and practices, on the well-known theme of sex. But suppose, having reached that point, there was still something that bothered us and we thought that we should press on further with our analysis. We might, for example, have felt that in one way or another a Greek or medieval author had expressed himself using particular words or turns of phrase which, despite our analysis, left a residue, a hint that suggested that, instead of taking no notice of that residue, as if it were just a clumsy expression, an approximation, a dead passage in the text, we should make an extra effort to make explicit what it appeared to imply. And suppose we were successful.

Then the scales would fall from our eyes: once a variation is made thoroughly explicit, the eternal theme is wiped out and all that remain in its place are successive variations, all different from one another, which we may call the ‘pleasures’ of Antiquity, the ‘flesh’ of the Middle Ages and the ‘sexuality’ of the modern age. Those are three general ideas that some people have successively elaborated around a kernel that is incontestably real and probably trans-historical, but that remains inaccessible, lurking behind them. Inaccessible or, rather, impossible to extricate: we would inevitably turn them into ‘discourse’.

Let us suppose that, thanks to the ‘programme’ of some branch of science, we learn something true, something scientific about homosexuality (for Foucault, the sciences amounted to more than empty words): for example (a gratuitous supposition on my part), that homosexual tastes are genetic in origin. Well, and then what? What, actually, is homosexuality? What should we do with that nugget, be it small or large, of the truth? Foucault wanted there to be some kind of ‘discourse’ on the subject of even an insignificant detail that related solely to anatomy or physiology, and not to the identity of individuals: in short, some detail that one would only discuss either in bed or with one's doctor.

Love in Antiquity gave rise to ‘discourse’ about the ‘pleasures’ of Aphrodite that were in no way suspect, and about how they should be controlled ethically and within the city. It addressed amorous gestures of the period, as timid as it was sinless, a period in which, at night, only a libertine would make love, not in the dark, but by lamplight, and in which civic morality distinguished not so much between the sexes as between the active and passive roles of the lovers – a period in which the ideal of self-control meant that a Don Juan would be considered effeminate and in which the obsessional reprobation of cunnilingus (which, nonetheless, was practised) implied a reversal of a hierarchy of the sexes; a period in which a pederast became a figure of fun because he carried his pleasure so far that his heart was left as bereft of feeling as an artichoke.

Now let us take an example less agreeable than love: the penal code through the ages. It is not enough to say that, under the ancien régime, punishments were atrocious, reflecting the brutality of the mores of the age. The royal sovereign ‘came down with all his might’, inflicting the horrific tortures of the period upon a rebellious subject, so as to demonstrate to all and sundry the enormity of both the punishment and the disproportion between the powers of the rebel and those of his king, whom the ceremonial torture avenged. With the advent of the Age of Enlightenment, punishment inflicted in private by a specialized administrative apparatus became preventative and corrective. Now imprisonment was to be a coercive technique of training, designed to set in place new habits in any citizen who had no respect for the law.⁷ This was assuredly an instance of humanitarian progress, but we need furthermore to recognize that it was not just an improvement: it was a total change.

Fifteen centuries earlier, in the arenas of the Roman Empire, the deaths of those condemned were prepared in mythological settings. A condemned man was dressed up in the costume of a Heracles committing suicide by fire, and was burnt alive; Christian women were disguised as Danaids and were accordingly raped before they died, or else as Dirce, strapped to the horns of a bull. These were sarcastic staged events, each one a ludibrium. The civic body, which the offender had presumed to rival, sneered at him, laughing in his face, to show him that he was not the stronger. Each of those successive ‘discourses’ was implied in the penal law, actions, institutions, powers, customs and even buildings, all of which reflected it and formed what Foucault calls the ‘set-up’ (dispositif). (Translator's note: dispositif is usually a hard word to translate, as it can mean so many things, depending on the context. In the present book, however, it consistently means what I have called the ‘set-up’.)

As you can see, we started off without any preconceptions, with detailed ‘concrete facts’;⁸ and we then discovered variations so original that each constituted a theme on its own. I have been using words such as ‘theme’ and ‘variations’, but Foucault had a better way of putting things. In 1979, he observed in his notebook: ‘Do not pass universals through the sieve of history; rather, strain history through a line of thought that rejects universals.’⁹ Ontologically speaking, variations are all that exist and the expression ‘a trans-historical theme’ is meaningless. Foucault, like Weber and all good historians, is a nominalist. Heuristically, it is better to start off with detailed practices, details of what was done and what was said, and then make the intellectual effort to make explicit the ‘discourse’ surrounding them. This is more fruitful (but more difficult both for the historian and for his readers)¹⁰ than starting off from a general, well-known idea, for if that is what you do, you are in danger of looking no further than that idea and failing to notice the ultimate, decisive differences that would reduce it to nothing.

But let us now forget those tortures and return to pleasures. It has been easy to distinguish pagan pleasures from the Christian concept of the ‘flesh’ (that ‘discourse’ on the sinful flesh and also on nature, which should be followed because it is a divine creation). That was followed by other ‘discourses’, for instance, the modern one about ‘sex’,¹¹ to which contributions have been made by physiology, medicine and psychiatry; and possibly also the discourse of postmodern ‘gender’ studies, along with feminism and permissiveness, or rather the subjective right to be oneself and say so (Didier Eribon would at this point remark that psychoanalysis would not survive this). One senses that every ‘discourse’ brings into play, around love, a whole host of associated elements: customs, words, bodies of knowledge, norms, laws and institutions; in fact, we should do better to speak of discursive practices or even, using a term loaded with meaning, to which we shall be returning, the whole ‘set-up’.¹²

But where were we? Instead of the commonplace notion of love, we thus discovered many bizarre little objects peculiar to the particular period, details that had never before been noticed. What we did was bring to light the submerged part of love in the period under consideration. The visible part, which was all that had been seen, was on the whole familiar. In contrast, once we had managed to make explicit the part that was not visible, not consciously recognized, what we were faced with was an object that was ‘incomplete and fragmented’,¹³ with jagged contours that corresponded to nothing sensible and by no means filled the capacious and imposing draperies that had previously covered it. Those contours put one in mind of the historical frontiers of nations that are traced in zigzag lines by the hazards of history, rather than by natural borders.

To be sure, our idea of sexuality or of madness (an idea that the subconscious, implicit ‘discourse’ follows closely, recording most precisely the singularity and strangeness that we cannot see) – that idea, together with its ‘discourse’, assuredly does relate to ‘the thing in itself’ (if I may take advantage of Kantian vocabulary), namely, the reality that it claims to represent. Sexuality and madness are things that certainly exist; they are not ideological inventions. However much one speculates, the fact remains that a human being is a sexual animal, as physiology and sexual instinct prove. All that has been thought about love and madness, down through the centuries, signals the existence and, as it were, emplacement of things in themselves. However, we are not in possession of a truth that corresponds to things, since we can only reach a ‘thing in itself’ by way of the idea that we have constructed of it in each different epoch (an idea of which its ‘discourse’ is the ultimate formulation, the differentia ultima). So we can only reach it as a ‘phenomenon’, for we cannot separate the thing in itself from the ‘discourse’ in which it is bound up for us or ‘buried in the sand’, as Foucault put it. We could know nothing without these kinds of presuppositions: had there been no ‘discourses’, object x that has successively been seen as divine possession, madness, insanity or dementia, and so forth would nonetheless exist, although, in our minds, we would be unable to place it.

The point is that all phenomena are singular, every historical or sociological fact is a singularity. Foucault thinks that general, trans-historical Truths do not exist, since human facts, acts and words do not come about naturally from a cause that is their origin; nor do they faithfully reflect the object to which they refer. Over and above their misleading generality or their supposed functionality, their singularity stems from their bizarre ‘discourse’. And that singularity of theirs in every case stems from chance developments and a complicated concatenation of the causalities at work; for the history of humanity is not upheld by reality, rationality, functionality or any dialectic: we must ‘identify the singularity of events, stripped of any uniform purposiveness’¹⁴ or any functionalism. The tacit suggestion that Foucault puts to sociologists and historians (a suggestion that some were independently putting into practice)¹⁵ is to push the analysis of historical and sociological formations as far as possible, in order to strip bare their singular strangeness.

To sum up: big words cover thoughts and realities (‘discourses’ and ‘discursive practices’) that are far narrower and have quirky edges. Here is another example of the gap that separates general and trans-historical ideas, which are always false, from little facts, the truth of which can be verified. These days, a Catholic with generous political opinions will be inclined to ascribe to Christian charity such of his opinions that favour socialism, greater economic equality and a redistribution of wealth. But how is it that his religion never conceived of such ideas until the workers’ movement of the nineteenth century? Catholicism never thought of abolishing slavery.

Around ad 300, the Christian Lactantius noted that some were rich while others were poor and some were masters while others were slaves. He went on to say that wherever all are not equal, there is no equality; and inequality excludes justice, which rests upon the fact that all men are born equal. He did acknowledge that among the Christians there were likewise both rich and poor and masters and slaves. However, he explained serenely, Christians consider them all as equal and brothers, for what matters is the spirit, not the body; the slaves are only slaves physically. Spiritually they are our brothers.

Charity was a big word that only covered small gestures, such as alms-giving and the notion of a religious fraternity in Christ. So in the Old South of the USA, prior to 1865, large landowners took care to baptize the black slaves that they acquired through the slave trade.