As far as I know, the first time Foucault introduces the concept of ‘regime of truth' is in chapter one of Discipline and Punish where, speaking of the formation (within the new penal system in the xviii and xix centuries) of a corpus of knowledge, techniques, ‘scientific' discourses that became entangled with the practice of the power to punish, he argues that a new "regime of the truth" emerged (Foucault 1975, p. 30; 23). Now, what makes this concept so interesting is the fact that, through this expression, Foucault links the notion of truth to the explicitly political notion of regime – as he does also in the February 18th, 1976 lecture of "Society Must Be Defended", where he speaks of "our regime of truth and error" and incidentally makes it clear that ‘regime' means here a certain power of separation between truth and error (Foucault 1975-76, p. 145; 164).
But the most interesting text, before 1980, with regard to Foucault's use of the concept of regime of truth – leaving aside a short passage in The Birth of Biopolitcs (Foucault 1978-79, p. 20; 18) –, is without a doubt the 1976 interview The political function of the intellectual, where Foucault argues, in contrast to a certain philosophical myth, that "truth isn't outside power, or deprived of power": on the contrary, truth "is produced by virtue of multiple constraints [a]nd it induces regulated effects of power". This is to say that "each society has its regime of truth", and by this expression Foucault means: (1) "the types of discourse [society] harbours and causes to function as true"; (2) "the mechanisms and instances which enable one to distinguish true from false statements" and (3) "the way in which each is sanctioned"; (4) "the techniques and procedures which are valorised for obtaining truth"; (5) "the status of those who are charged with saying what counts as true" (Foucault 1976, p. 112; 13). Therefore, "truth" is "a system of ordered procedures for the production, regulation, distribution, circulation and functioning of statements"; it is linked "by a circular relation to systems of power which produce it and sustain it, and to effects of power which it induces and which redirect it". And right at the end of the interview, Foucault adds that the essential political problem for us, today, is trying to change our "political, economic, institutional regime of the production of truth" (where truth is modeled on the form of scientific discourse), in order to constitute a new "politics of truth" (Foucault 1976, pp. 113-114; 14).
So, before 1980, the concept of regime of truth clearly refers to the well-known circularity Foucault establishes between power and knowledge: we should speak of a ‘regime' of truth because truth is produced, sustained, valorized and regulated by a series of mechanisms, techniques and procedures that are ‘political' – if we understand this term the way Foucault does: politics has to do not only with institutions, but with the complex and constitutive field of power relations within which we ordinarily live –, and at the same time truth itself reinforces and induces effects of power. A regime of truth is thus the strategic field within which truth is produced and becomes a tactical element in the functioning of a certain number of power relations.
Michel Foucault, fotografiert in Buffalo (NY) in den USA, 1971. (Foto: Bruce Jackson)
Posing this problem means, for Foucault, at least three things: (1) redefining the concept of regime of truth; (2) putting it at the heart of his historical study of Christianity; (3) putting it also at the core of his genealogical study of our contemporary regime of truth.
So, firstly, Foucault claims that a regime of truth "is that which determines the obligations of individuals with regard to procedures of manifestation of truth", thus stressing the role played by the subject within such procedures. But Foucault immediately objects to himself: "What does the addition of this notion of obligation mean in relation to the notion of manifestation of truth? How does the truth oblige, in addition to the fact that it is manifested?" (Foucault 1979-80, p. 91). This objection contests the legitimacy of the concept of regime of truth arguing precisely that these two notions – ‘regime' and ‘truth' – cannot go together: it is not possible to speak of a regime of truth like we speak of a political or a penal regime. Here, Foucault clearly assumes the point of view of the dominant conception of truth in the West, that is an ‘epistemological' point of view – the point of view of what he calls, in his lectures on Psychiatric Power, "truth-demonstration" as opposed to truth as an event (Foucault 1973-74, pp. 235-239; 235-239). According to this perspective, truth, if it is really true, does not need a supplement of force, an enforcement, a supplement of vigor and constraint to be accepted. It is the truth, and that's all: truth is sufficient unto itself for making its own law – its coercive force resides within truth itself. "Truth itself determines its regime, makes the law, and obliges me. It is true, and I submit to it". So, as Foucault argues, "for something like an obligation to be added to the intrinsic rules of manifestation of the truth", it must "involve precisely something that cannot be manifested or demonstrated by itself as true". In other words, there can be no genuine truth ‘obligation', no genuine ‘regime' of truth: there can only be the "coercion of the non-true or the coercion and constraint of the unverifiable" (Foucault 1979-80, pp. 92-93) – because truth, if it is really true, is rather on the side of freedom, it emancipates and redeems instead of subjugating.
However, Foucault spells out this objection just in order to reject it. In fact, the truth is index sui – that is to say: in every ‘game of truth' (considered from the point of view of its formal rules and not from the point of view of the individuals who are implicated in it) only the truth can legitimately establish the partage between true and false statements –, but it is not rex sui, lex sui or judex sui. On the contrary, the truth "is not creator and holder of the rights it exercises over human beings, of the obligations the latter have towards it, and of the effects they expect from these obligations when and insofar as they are fulfilled": it is not true "that the truth constrains only by truth". In other words, according to Foucault, under every argument, every reasoning and every ‘evidence', there is always a certain assertion that does not belong to the logical realm, but is rather a sort of commitment, a profession, and which has the following form: "if it is true, then I will submit; it is true, therefore I submit". This ‘therefore' that links the ‘it is true' to the ‘I submit' and gives the truth the right to say "you are forced to accept me because I am the truth", even if in some games of truth it is almost invisible, even if sometimes it goes so much without saying that we hardly notice its presence, well this ‘therefore' does not rest itself on any truth or evidence and does not arise from the truth itself in its structure and content. This "you have to" of the truth is, according to Foucault, a "historical-cultural problem" (Foucault 1979-80, pp. 94-95). Hence, while the rules of each game of truth define autonomously, within this specific game, the partage between true and false statements, Foucault makes it clear that these rules are not themselves autonomous: on the contrary, they are always the result of a historical, social, cultural and ultimately ‘political' production.
Three brief remarks before coming to the conclusion. (1) Foucault traces here, somehow implicitly, a distinction between ‘game of truth' and ‘regime of truth' – a distinction that becomes clear when (in a very Wittgensteinian way) he defines science as "a family of games of truth all of which submit to the same regime, although they are not subject to the same grammar, and this very specific, very particular regime of truth is a regime in which the power of the truth is organized in a way such that constraint is assured by truth itself" (Foucault 1979-80, p. 93). Nevertheless, science "is only one of the possible regimes of truth": there are "many other ways of binding the individual to the manifestation of truth" (Foucault 1979-80, p. 97). (2) Even if we can trace a distinction between ‘game' and ‘regime' of truth, there is no game of truth without or outside of a regime of truth. In other words, no game of truth has the privilege of being ‘pure': every game of truth is necessarily linked to a regime of truth that determines the obligations of the individuals who are implicated in it and thus accept – explicitly or not, consciously or not – the specific ‘therefore' that links the ‘it is true' and the ‘I submit'. (3) This acceptance takes the form of a subjection (assujettissement) or of a subjectivation (subjectivation), since every regime of truth asks to the individuals who are implicated in it a specific self-constitution. For instance, in the case of Descartes, the subject can say "I think, therefore I am" only if he or she is "qualified in a certain way", that is only if he or she has constituted him or herself and has been constituted by his or her society as someone who is not mad (Foucault 1979-80, p. 96).
To conclude, I should note that, in Du gouvernement des vivants, Foucault at the same time and by the same move puts the concept of regime of truth at the heart of his historical study of Christianity and of his genealogical study of our contemporary regime of truth indexed to subjectivity. Thus, we cannot consider the concept of regime of truth as a merely analytical or methodological tool: it carries in itself a critical force, and this is why it can still be useful for us, today. In fact, through this concept, Foucault shows us that we are not obliged to accept the scientific or epistemological regime of truth, and more importantly that we are not obliged to shape our subjectivity and our way of life on it. On the contrary, we should try to choose other values (that may have nothing to do with truth) on which to shape ourselves. The idea that truth gives us no choice, that truth necessarily forces us to accept it and build up our conduct in accordance to it, is an extremely dangerous ethico-political trap that Foucault can help us to unmask and overcome.
M. Foucault (1973-74), Le pouvoir psychiatrique. Cours au Collège de France. 1973-1974, ed. J. Lagrange, Paris, Seuil/Gallimard, 2003; trans. G. Burchell, Psychiatric Power. Lectures at the Collège de France. 1973-1974, New York, Palgrave Macmillan, 2006.
— (1975), Surveiller et punir. Naissance de la prison, Paris, Gallimard; trans. A. Sheridan, Discipline and Punish. The Birth of the Prison, New York, Vintage Books, 1977.
— (1975-76), " Il faut défendre la société ". Cours au Collège de France. 1975-1976, ed. M. Bertani and A. Fontana, Paris, Seuil/Gallimard, 1997; trans. D. Macey, "Society Must Be Defended". Lectures at the Collège de France. 1975-1976, New York, Picador, 2003.
— (1976), "La fonction politique de l'intellectuel", in Dits et écrits II, 1976-1988, ed. D. Defert and F. Ewald, Paris, Gallimard, 2001, pp. 109-114; trans. C. Gordon, "The political function of the intellectual", in Radical Philosophy, no. 17 (Summer 1977), pp. 12-14.
— (1978-79), Naissance de la biopolitique. Cours au Collège de France. 1978-1979, ed. M. Senellart, Paris, Seuil/Gallimard, 2004; trans. G. Burchell, The Birth of Biopolitics. Lectures at the Collège de France. 1978-1979, New York, Palgrave Macmillan, 2008.
— (1979-80), Du gouvernement des vivants. Cours au Collège de France. 1979-1980, ed. M. Senellart, Paris, Seuil/Gallimard, 2012.
 A first and longer version of this paper has been presented on October 22nd, 2013 at the University of Chicago, during the seminar "Michel Foucault: Self, Government, and Regimes of Truth" led by Arnold I. Davidson.