Foucault's Point of Heresy: ‘Quasi-Transcendentals’ and the Transdisciplinary Function of the Episteme

‘Points of Heresy’: A Complex Disposition

It is widely recognized that the guiding category in Foucault’s archaeological inquiry is that of ‘episteme’, which is completed by such formulas as ‘epistemic field’ or ‘epistemic discontinuity’, and occasionally substituted with ‘system of positivities’ or simply ‘knowledge’ (as distinct from ‘science’). It performs a triple task: a periodization of discourses (and of knowledge inasmuch as knowledge is examined ‘from the outside’, through the exclusive analysis of its discursive realizations); a transdisciplinary unification (inasmuch as an episteme does not describe a system of concepts and postulates which are axiomatic for one single discipline, but rather the invariant ‘rationality’, or the system of questions and notions which, at a certain moment, or in a certain historical period, command the development of several disciplines – such as biology, economics, philology – and allow it to establish formal correspondences between them, provided they are ‘translated’ in each discipline’s language in order to be applied to its specific objects); finally a problematization, which, notwithstanding later uses of this term by Foucault, I will define as a double intellectual gesture of formulating questions which are meaningful, therefore possible within a certain organization of discourse, and, correlatively, identifying objects which are either visible or invisible for a certain ‘experience’ informed by this discourse.

This is a very powerful dispositive. It made it possible for Foucault to discover conceptual analogies between the disciplines that he was describing (by raising these descriptions from the status of empirical phenomenologies to that of rational constructions) and to propose a new philosophical interpretation of the transition from the ‘classical age’ of natural typologies to the ‘modern age’ of historical evolutions. In the specific field of the history of science, it allowed him to transform the formulation of such epistemological problems as the origins of evolutionist ideas and theories in biology and anthropology, by paradoxically rooting them in certain ‘fixist’ arguments. This is not to say that his theses on this point are unobjectionable, but they remain impossible to simply dismiss. ² We know that the notion of episteme is replete with difficulties of various kinds. Some have to do with the seemingly ad hoc selection of the disciplines that are included in Foucault’s construction of the two successive epistemes that he wants to compare: for example, why does he include ‘biology’ and ‘economics’ in the modern episteme of anthropological ‘finitude’, but never really consider sociology or psychology, which are two very different cases? The linguistic determinations, which are deeply rooted in national intellectual histories, are very important here: the tradition to which Foucault belongs is one in which the notions of ‘la science de l’homme’ (initially coined by Malebranche) and later ‘les sciences humaines’ (of Saint-Simonian and Comtean origin) enjoy a philosophical and institutional recognition uninterrupted since the 17th century, whereas ‘social science’, which is prevalent in the Anglo-Saxon academic world, is considered only as a recent and partial equivalent, and the ‘German’ notion of the Geisteswissenschaften is banned as incurably spiritualistic.

But the major difficulties (for philosophers) concern internal tensions between the historical function and the logical or transcendental construction of the episteme. On the one side, epistemic moments within which ‘knowledge’ is distributed and organized feature not only units of problematization, but also successive terms in a chronological line, modeled on classical histories of modernity: this produces the strange effect of a counter-historicist historicism. It seems to propose a ‘meta-narrative’ of the history of knowledge, where the idea of progress is abolished in favour of independent ‘mutations’, in the form of successive incompatible problematizations of the object. This does not amount to cancelling the notion of a ‘sense of history’ with its eschatological connotations – perhaps rather the contrary. The main demarcation proposed by Foucault separates the ‘classical’ episteme – where knowledge is commanded by the prerequisite of ‘order’ and the articulations of discourse provide a representation of the real that is always already adequate, since thought and being ‘belong’ to one another – from the ‘modern’ episteme, where the dominant category would be history or historicity, whose driving forces are supposed to lie beyond representation, opening something like a permanent line of escape. This mutation is said to take place in the revolutionary period during which philosophy (with Kant) becomes a ‘critique’. It marks a before and an after: the revolution, or the intellectual caesura. ³ And since, in order to locate a first discontinuity (whose idea he associates with the sense of hilarious absurdity produced by the reading of a passage in Borges), ⁴ it seems necessary to Foucault to evoke a prehistory, or a before the before, where the ‘classical’ categories are not yet in action (this would be the late Medieval and Renaissance age of ‘resemblance’); similarly our contemporary awareness of the limits of anthropology (which are also the limits of historicity) seems to indicate that ‘we’ (the collective bearers of western knowledge) are on the verge of entering an after the after: a post-humanist age of the ‘return of language’ as essential horizon of thinking. On the other side, the ‘invariants’ which cut across the boundaries of disciplines are presented at the same time, in Kantian fashion, as conditions of possibility for the formulation of scientific problems (such as: how to combine the articulation of types and the attribution of properties – a question that exists for classical grammar as it exists for natural history or the analysis of riches) and conditions of possibility for the recognition of objects in experience (such as ‘words’ and ‘languages’, ‘animals’ and ‘plants’, ‘commerce’ and ‘riches’, etc.). This induces the idea that the categories of knowledge are like a priori structures of rationality with respect to actual discourses.

But these two conceptualizations of episteme work philosophically in opposite directions. They are reconciled dialectically by means of a definition of the episteme as an ‘historical a priori’ – which owes something to neo-Kantian elaborations, but is ultimately a borrowing from Husserl turned against phenomenology. ⁵ This is, however, an extremely fragile notion, which, if it is not simply to name the difficulty in oxymoronic fashion, cannot but oscillate between ‘culturalist’ analogies (where epistemes are like systems of ‘cultural determinations’ subjecting knowledge to collective identities and social transformations) and an implicit reiteration of the Heideggerian ‘history of Being’ (where knowledge is not a ‘positivity’, much less a ‘rationality’, but a manifestation of the secret loss of truth in its successive figurations). Even a critic as sympathetic but also uncompromising as Canguilhem (1967) indicated this difficulty in his defence of Foucault’s book against the anathemas of what he called ironically the ‘League of Humanists of all Parties’.

My intention is not so much to resurrect these debates as to indicate why they remain distorted as long as we neglect an essential fact: the category of the episteme in Foucault’s book cannot be understood alone, in its own right, if we do not take into account its intrinsic correlation with another notion with which it forms a dialectical pair, the ‘point of heresy’, which signals the conflictual dimension of discursive knowledge. Admittedly, this correlation is somewhat concealed in Foucault’s book because the point of heresy, although it is explicitly mentioned in several passages, is invoked most of the time with other names and through paraphrases, and because it is never properly defined (as is formally the case for epistemes), except in the form of an additional determination in the course of their analyses – an overdetermination as it were. But this concealment works like a purloined letter, because as soon as the presence of points of heresy has been identified operating strategically in the text, we become aware that it is precisely the articulation of the two notions (heresy and episteme) that indicates what an ‘epistemic field’ exactly includes (or structures) and what it excludes, or what it exactly means to say that there are limits for any episteme, whose paradoxical ‘experience’ signals a possibility of transgression. Before anything else, I need therefore to retrieve in Foucault’s text the ‘traces’ of this notion, elusive but also unmistakable.

Point of heresy is a phrase that obviously has affinities with others in Foucault and some of his contemporaries. (A complete investigation of the topological metaphors of ‘points’, ‘lines’, ‘spaces’ or ‘fields’, ‘bodies’, ‘surfaces’ and ‘depth’, etc. in Foucault – and others – would be a fascinating study.) ⁶ Strictly speaking, however, it is uniquely linked to the problematic of The Order of Things, with the partial exception of a development in The Archaeology of Knowledge. Literally, it is used three times, in two separate contexts. First, in the course of the description of classical grammar as a general theory of ‘speaking’, when it is a question of the epistemic choice between decomposing words into meaningless units (sounds and letters) or into more primitive elements of signification, such as ‘roots’ (Foucault, 1966: 115–16). Second, when in the course of the description of the classical analysis of wealth as a theory of ‘exchanging’, it is a question of the epistemic conflict between theories of money as ‘sign’ and theories of money as ‘commodity’ (Foucault, 1966: 193–4). This is later repeated as an alternative to ‘psychological’ foundations of the value of goods in utility (advocated by the likes of Cantillon) and a foundation in the productivity of nature (advocated by the Physiocrats), which he calls a ‘familiar point of heresy’ (Foucault, 1966: 204). We must pay attention to the fact that the same idea of a choice between antithetical opinions (i.e. doxai: judgments, positions) relying on a bifurcation in the development of the same ‘disposition’ of knowledge (e.g. defining money as a ‘pledge’) – in other words, the idea of a choice whose result is arbitrary but derives from a necessary antithesis within the shared premises (‘though it is logically necessary, it is so on the basis of a single arrangement that simply creates, at a given point, the alternatives of an indispensable choice’ – Foucault, 2002: 181) – does not need a repetition of the name ‘heresy’ to be invoked. Quasi-literal equivalents such as ‘point of choice’, ‘fork of a choice’, ‘real choice of opinions’, ‘alternatives’, express exactly the same idea. ⁷ The distribution in the book becomes then much more general, but also more complex, and it is this that makes it a strategic object for our reflection.

The variations in the naming are easily explained by the fact that Foucault is using ‘heresy’ in its etymological sense (in Greek, hairesis means ‘choice’ or ‘decision’ between alternative possibilities), which in our culture is intimately linked to the religious institution and discourse. Early theologians defined ‘heresies’ as those doctrines whose opposition to the ‘orthodoxy’, or to the right opinion (judgment) sanctioned by the authority of the Church, resides not only in a ‘wrong choice’ or a ‘choice of the wrong’ but more precisely in the fact of choosing a one-sided interpretation of a dogma. ⁸ This involves the idea that the unity of opposites is precisely what forms the paradoxical condition of a theological utterance of truth, as in the two central – and related – cases of the dogma of the incarnation (the double nature of Christ: divine and human) and the doctrine of salvation as result of free actions that are nevertheless entirely determined by the grace/power of God. I submit that Foucault’s uses of ‘point of heresy’ and ‘point of choice’ are a reminiscence, if not a conscious transposition of this theological use, by means of the etymology, a transposition which ‘suspends’ any direct reference to religious doctrines, but keeps the idea that in matters of ‘dogmatic alternatives’ the truth-function (or, as he will later write, the mode of ‘veridiction’) does not depend on the fact that this or that ‘theory’ is separately verified or refuted (or is scientific or not), but on the fact that the system of antithetic propositions divides the possibilities of interpretation of a given principle or axiom.

Moreover, Foucault’s articulation of knowledge (episteme) and choice (heresy) in the archaeological discourse is obliquely commanded by a reference to the Pascalian reflection on this subject – as I suggested elsewhere is the case with his coining the expression ‘history of the truth’, which is in fact a rediscovery of a provocative proposition in Pascal (Balibar, 2002). In his polemical intervention in the controversy on the relationship between divine grace and redemption (linked to the great theological divide about predestination), Blaise Pascal, who was also a logician and a mathematician with a strong doctrine of axioms as ‘indemonstrable foundations of demonstrations’, proposed a radical definition of the relationship between orthodoxy and heresies, which is a variety of the theological via negativa (perhaps itself ‘heretical’): since the orthodoxy is about ‘mysteries’ (or revealed truths without any ‘rational’ possibility of explanation), there can be no such thing as an intelligible presentation of the positive truth, but only a permanent ‘negation’ of incompatible false representations which betray it, or an infinite effort to simultaneously avoid antithetic ‘errors’. In Foucault as in Pascal, we have ‘dialectical’ schemes for making the unity of opposites (coincidentia oppositorum) an intrinsic characteristic of thinking, which each in its own way distinguishes itself from the teleological procedure of ‘reconciliation’ or negation of the negation.

Under its several equivalent names, the concept of a point of heresy returns regularly in The Order of Things, with applications of different orders, which seem to accompany the progressive transformations of the concept of the episteme, and help to clarify its function.

In the classical age, each and every ‘field of positivity’, while isolating an object that also features a function in the general space of representation (‘speaking’ words, ‘classifying’ natural beings, ‘exchanging’ goods), exhibits one or several points of heresy. Their importance ranges from a simple difference between alternative interpretations of the same enigmatic notion (such as meaning or value) to a bifurcation separating ‘two ways of constituting [the discipline] as a language both possible and indispensable’ (as in the case, crucial for the history of ‘natural history’, of the choice between Linnaeus’ system and Buffon’s method) (Foucault, 2002: 139–40). In the case of Representation as the episteme of the Classical Age, there is clearly no understanding of the ‘epistemic field’ without identifying its points of heresy. More precisely:

Points of heresy, inasmuch as they are internal to an epistemic field which imposes rigorous conditions of possibility upon the perceptible and the sayable, are the exact opposite of choices ‘operated according to individuals, environments, social groups’, whose presentation in the form of personal and doctrinal controversies had provided the traditional history of ideas with guiding narratives to describe the opinions of certain authors (Foucault, 2002: 75). Foucault’s archaeology does not explain them as subjective decisions (however ‘situated’ in history), but as objective possibilities offered by a ‘disposition’ or structure of knowledge. As a consequence, points of heresy exhibit the kind of (limited) freedom that knowledge offers to its ‘subjects’, and explain what the function of the author consists in: namely, actualizing and individualizing antithetic positions that are inherent in the latent organization of knowledge, or virtually present as soon as a new principle of understanding things and experiences has emerged. ⁹ The ‘controversies’ or discursive conflicts as represented in these two incompatible ways appear thus as a major criterion for the discussion of the position of subjects in the history of knowledge.

However, this explanation also means that epistemic conflicts become relativized, or they are presented as ‘superficial’ with respect to the ‘general system’ (or structure) where they take place. Not only do they not constitute the system, or involve a capacity to transform it, but they form only an ‘apparent contradiction’ if compared to the positivity itself, or to the common ‘deep’ determinations of experience and objects within a given episteme:

One must reconstitute the general system of thought whose network, in its positivity, renders an interplay of simultaneous and apparently contradictory opinions possible. It is this network that defines the conditions that make a controversy or problem possible, and that bears the historicity of knowledge. (Foucault, 2002: 75)

In turn, however, this relativization bears a strong counterpart, because it appears (or it is asserted by Foucault) that the points of heresy are the same, or they emerge at the same ‘place’, giving way to ‘isomorphic’ alternatives, when describing the various disciplines whose juxtaposition maps the classical episteme: grammarians, naturalists and analysts of wealth (who cannot yet be called ‘economists’) face the same choices and resolve them in analogous manners, albeit in completely different vocabularies and referring to heterogeneous empirical activities (speaking, classifying, exchanging). Points of heresy are essentially transdisciplinary. As such, what they reveal is the unitary construction of the episteme, or better said they provide an effective verification of the unity that could be hypothetically asserted when discovering (through formal analyses) that the disciplines had isomorphic conceptual matrixes. If the conflicts are the same at the surface where opinions are formulated, this must be because the latent problems (or the mode of problematization itself) are governed by the same structure. This also shows why Foucault’s concept of knowledge is not purely logical or ‘mathematical’: it reveals how knowledge works not at the level of formulating axioms, but of constructing experience discursively in their terms.

All these characters, which seem to apply in general to what Foucault calls an episteme are in fact presented in full detail with respect to one single episteme: the Classical field of ‘Representation’. This could mean that the very notion of ‘heresy’ (and the underlying scheme of the bifurcation in the implementation of axiomatic premises) has a privileged relationship to the formal structures of representation and order, which involve binary choices of definitions and taxonomies. But if this were the case, it would also mean that heresy and episteme are not absolute correlatives, as I claimed, since there exists at least another episteme. So we must take a closer look at the resurgence of this terminology in the framework of the modern episteme, characterized by Foucault as a knowledge of History, whose general object is Man (or the Human, qua living, speaking, and laboring being), therefore opening the possibility of new positivities called ‘the human sciences’. We discover that a transformation takes place, which indicates that the concept of the episteme is evolving: no two different epistemes embody ‘knowledge’ in exactly the same sense.

To be sure, there is one explicit return of the ‘point of heresy’ with exactly the same function that we had already seen. It takes place (in a manner that is far from innocent) when Foucault describes an internal conflict of opinions or a bifurcation within the new discipline of ‘economy’, between the Ricardian and the Marxian interpretations of ‘historical tendencies’ affecting the relationship between needs and capacities in the organization of productive systems, which he interprets as expressing a fundamental articulation of history and anthropological finitude:

But the alternatives offered by Ricardo’s ‘pessimism’ and Marx’s revolutionary promise are probably of little importance. Such a system of options represents nothing more than the two possible ways of examining the relations of anthropology and History as they are established by economics through the notions of scarcity and labour. For Ricardo, History fills the void produced by anthropological finitude and expressed in a perpetual scarcity, until the moment when a point of definitive stabilization is attained; according to the Marxist interpretation, History, by dispossessing man of his labour, causes the positive form of his finitude to spring into relief – his material truth is finally liberated. There is certainly no difficulty in understanding, on the level of opinion, how such real choices were distributed, and why some opted for the first type of analysis and others for the second. But these are merely derived differences which stem first and last from a doxological investigation and treatment. At the deepest level of western knowledge, Marxism introduced no real discontinuity (aucune coupure réelle); it found its place without difficulty, as a full, quiet, comfortable and, goodness knows, satisfying form for a time (its own), within an epistemological arrangement that welcomed it gladly (since it was this arrangement that was in fact making room for it) and that it, in return, had no intention of disturbing and, above all, no power to modify, even one jot, since it rested entirely upon it. Marxism exists in nineteenth-century thought like a fish in water: that is, it is unable to breathe anywhere else. Though it is in opposition to the ‘bourgeois’ theories of economics, and though this opposition leads it to use the project of a radical reversal of History as a weapon against them, that conflict and that project nevertheless have as their condition of possibility, not the reworking of all History, but an event that any archaeology can situate with precision, and that prescribed simultaneously, and according to the same mode, both nineteenth-century bourgeois economics and nineteenth-century revolutionary economics. Their controversies may have stirred up a few waves and caused a few surface ripples; but they are no more than storms in a children’s paddling pool. (Foucault, 2002: 261–2)

The clearest example of another point of heresy rooted in the history of ‘positive’ disciplines, but immediately endowed with a philosophical meaning, emerges with Foucault’s presentation of the great antithesis between formalistic and hermeneutic understandings of the function of language. It is illustrated there through an opposition between the contemporary ‘inventions’ of Bertrand Russell and Sigmund Freud: formal logic and the psychoanalytic interpretation of dreams and symptoms. Here is Foucault’s summary of the topological figure:

It is true that the division (le partage) between interpretation and formalization presses upon us and dominates us today. But it is not rigorous enough: the fork it forms has not been driven far enough down into our culture, its two branches are too contemporaneous for us to be able to say even that it is prescribing a simple option or that it is inviting us to choose between the past, which believed in meaning (croyait au sens), and the present (the future), which has discovered the significant (le signifiant). In fact, it is a matter of two correlative techniques whose common ground of possibility is formed by the being of language, as it was constituted on the threshold of the modern age … This certainly explains the nineteenth century’s double advance (la double marche), on the one hand towards formalism in thought and on the other towards the discovery of the unconscious – towards Russell and Freud. It also explains the tendency of one to move towards the other, and of these two directions to cross: the attempt, for example, to discover the pure forms that are imposed upon our unconscious before all content; or again, the endeavor to raise the ground of experience, the sense of being, the lived horizon of all our knowledge to the level of our discourse. It is here that structuralism and phenomenology find, together with the arrangements proper to them, the general space that defines their common ground. (Foucault, 2002: 299)

This indicates that identifying a point of heresy works in two directions. On the one hand, it leads us to asserting again and again that theoretical dilemmas that seem irreconcilable (such as a choice between the point of view of ‘meaning’ and a point of view of the ‘signifier’) are in fact determined by the general type of rationality that governs the historical emergence of some common objects of knowledge. On the other hand, it introduces a dynamic pattern of thinking where the possibility arises of ‘walking on two roads’: not a simple ‘antithesis of reason’, which remains static, but something like a dialogical adventure within rationality, whose details are retrospectively understandable, but never formed a simple repetition of the initial partage. It is because of this new dynamic, or in close relation to it, that Foucault now presents the ‘point of heresy’ not only as a ‘fork’, but as a place of interaction, retroaction, or double inscription for the extremes, where philosophy and positivity are practically exchanging their roles. Witness the reference to the way in which a formalistic conception of language as a game of signifiers can become the new ‘code’ for the interpretation of meaning, a clear allusion to Lacan’s structuralist concept of the unconscious and an indication, at least apparently, that a ‘dialectic’ of Russellian and Freudian conceptions of language is much more productive than a dialectic of Ricardian and Marxian conceptions of productive labor (except that Lacan’s preferred logical reference was Frege, rather than Russell).

But there is more, because in this new interactive pattern a reference to the past is involved as much as a projection towards the future (or the latter is possible only in relation to the former). This was clearly indicated when, a few paragraphs earlier, Foucault described the point of heresy inherent in the new ‘objectivity of language’ produced by modern philological techniques (after the invention of ‘flexion’ by Schlegel and Bopp) in terms of the opposition of formal logic and comparative grammar, which he presented as twin ‘products of the dissociation of general grammar’ – therefore as an active trace of the vanishing of the old (classical) episteme:

One might say in one sense that logical algebra and the Indo-European languages are two products of the dissociation of general grammar: the Indo-European languages expressing the shift of language in the direction of the known object, logical algebra the movement that makes it swing towards the act of knowing, stripping it in the process of all its already constituted form. But it would be inadequate to express the fact in this purely negative form: at the archaeological level, the conditions of possibility of a non-verbal logic and a historical grammar are the same. The ground of their positivity is identical. (Foucault, 2002: 297)

This third type is exhibited in the section where Foucault analyses the meaning and correspondences of discourses competing for a synthetic reconstruction of knowledge at the end of the 18th century: ‘Ideology’ (initiated by Destutt de Tracy and Cabanis), and ‘Critique’ (initiated by Kant). I shall leave aside discussing this selection from the point of view of the history of ideas. In particular, it seems to neglect the place occupied by British (and especially Scottish) empiricism in the ‘philosophical battlefield’ of the revolutionary age (although, in the same chapter, Foucault paid close attention to Adam Smith’s problematic, but only as an economist). A new interest in the doctrine of the French Idéologues was – temporarily – in the air in the 1960s: Foucault (already with Histoire de la folie and Naissance de la Clinique) contributed to it, and was praised for that by Canguilhem. But what interests me is the conceptual pattern of the opposition, as he constructs it. Quotations are again necessary here. Let’s recall how Foucault characterizes the imperceptible ‘event’ which, taking place simultaneously in the fields of general grammar, natural history and analyses of wealth, will show that representation is insufficient to express the being of objects whose essential attribute is an intrinsic plasticity or historicity – be it in the case of the flexion of words, the life of the organism as the antithesis of ‘inorganic matter’, or the productivity of labour as a measure of the value of commodities:

This somewhat enigmatic event, this event rising up from below which occurred towards the end of the eighteenth century in these three domains, subjecting them at one blow to one and the same break (rupture), can now be located within the unity that forms a foundation for its diverse forms. Quite obviously, it would be superficial to seek this unity in some progress made in rationality, or in the discovery of a new cultural theme […] In a more fundamental fashion, and at the level where acquired knowledge is rooted in its positivity, the event concerns, not the objects aimed at, analyzed, and explained in knowledge, not even the manner of knowing them or rationalizing them, but the relation of representation to that which is posited in it. What came into being with Adam Smith, with the first philologists, with Jussieu, Vicq d’Azyr, or Lamarck, is a minuscule but absolutely essential displacement (un décalage infime mais absolument essentiel), which toppled the whole of western thought: representation has lost the power to provide a foundation – with its own being, its own deployment and its power of doubling over on itself – for the links that can join its various elements together … The condition of these links resides henceforth outside representation, beyond its immediate visibility, in a sort of behind-the-scenes world (une sorte d’arrière-monde) even deeper and more dense than representation itself … The very being of that which is represented is now going to fall outside representation itself. (Foucault, 2002: 238–40)

What is described here in general terms is the emergence, out of transformations and inventions that took place within the positive disciplines themselves (which are ‘enigmatically’ correlated in a transdisciplinary unity), of a thing in itself of which any ‘representation’ can only express partial and disjunctive aspects. We could say: beyond the ‘objects’ of representation, it is the ‘function thing’ that emerges simultaneously in all disciplines, subjected to the same epochal change in the organization of knowledge. But it is not the same thing, because Life is not Language, and Language is not Labour, even if these three essences name what must be supposed to exist ‘behind-the-scenes’ of visibility, in an analogous manner.

No wonder that Foucault is led to explain that it is Kant’s criticism that most exactly expresses the new disposition of knowledge (because, in meta-disciplinary language, it is the manifestation of the same ‘event’ or ‘break’), although Kant’s critique was blind to the historical condition of its ‘rational’ correspondence with contemporary knowledge, for which it sought a foundation in nature – except that the ‘nature’ in question now is not Nature as an infinite order of beings, but the nature of ‘Man’ as an empirical-transcendental double. This may explain why Foucault is not busy here repeating the idea (variously illustrated in the neo-Kantian tradition) that a Kantian type of criticism provides the ‘foundation’ for a new science (or an extension of scientificity, beyond classical physics and mathematics, e.g. to psychology, or to culture and symbolic forms). He is busy explaining that the new transdisciplinary episteme that, in the end, generates the system of ‘human sciences’ will find a privileged expression of its philosophical contradictions in the inner tensions of Kantian criticism. And, in turn, this exquisite sensitivity of Kantianism to the tensions of an anthropological positivity (virtually: its heresies to come) will not be presented in terms of an internal reconstruction of the system, but rather in terms of its antithesis with another speculative anthropology, the ‘natural science of man’ proposed by Kant’s exact contemporaries, the ‘Idéologues’. The ‘deep unity’ underlying the formation of the new episteme is thus best understood in terms of a conflict between two philosophical discourses that were established during the same period, but are in fact ‘non-contemporaneous’. This is also a way to conceptualize an epistemic ‘event’, or more precisely it develops the idea of the event into the analysis of a complex ‘actuality’:

The coexistence of Ideology and critical philosophy at the end of the eighteenth century … divides, into two forms of thought, exterior to one another, yet simultaneous, what scientific forms of reflection, on the other hand, hold together in a unity doomed to imminent dissociation … We should note, however, that, in defining the thought of a relation by the sensation of that relation, or, in briefer terms, thought in general by sensation, Destutt is indeed covering, without emerging from it, the whole domain of representation; but he reaches the frontier where sensation as the primary, completely simple form of representation, as the minimum content of what can be given to thought, topples over into the domain of the physiological conditions that can provide an awareness of it … Analysis of representation, at the moment when it attains its greatest degree of extension, brushes with its very outermost edge a domain that is more or less – or rather, that will be more or less, for it does not exist as yet – that of a natural science of man. Different as they are in form, style, and aim, the Kantian question and the question of the ‘Idéologues’ have the same point of application: the relation of representations to each other. But Kant does not seek this relation … on the level of representation … he questions it as to what renders it possible in general … There is thus a definite correspondence between the Kantian critique and what in the same period was posited as the first almost complete form of ideological analysis. But Ideology, by extending its reflection over the whole field of knowledge … tried to resume in the form of representation precisely what was being formed and re-formed outside representation … In this sense, Ideology is the last of the Classical philosophies … Confronting Ideology, the Kantian critique, on the other hand, marks the threshold of our modernity; it questions representation, not in accordance with the endless movement that proceeds from the simple element to all its possible combinations, but on the basis of its rightful limits. Thus it sanctions for the first time that event in European culture which coincides with the end of the eighteenth century: the withdrawal of knowledge and thought outside the space of representation … In this sense, Criticism brings out the metaphysical dimension that eighteenth-century philosophy had attempted to reduce solely by means of the analysis of representation. But it opens up at the same time the possibility of another metaphysics; one whose purpose will be to question, apart from representation, all that is the source and origin of representation; it makes possible those philosophies of Life, of the Will, and of the Word, that the nineteenth century is to deploy in the wake of criticism. (Foucault, 2002: 240–43)

What Foucault seems to be adding (as, in fact, Althusser would do a few years later in his own way) is the idea that the transition never ends, or the break with classicism giving rise to modernity, which is ‘radical’ at the level of epistemic principles (or dispositions), is also impure or impossible in the effectivity of discourse. ‘Under’ the deep level of the new positivity (or perhaps intertwined and conflicting with it) there remains a latent confrontation with the old positivity of representation, discourse itself: it is presented here as a principle of regression, but it could also become actively involved in a transgression. This is exactly what we observe when Foucault (who does not think in evolutionist terms of regression vs. progression, but rather in the polemical terms of regression vs. transgression) introduces ‘the return of discourse’ as a limit experience for modernity itself.

This leads me to a third comment: the point of heresy we are discussing here is not only a more complex one, it is also one that virtually reverses the articulations between surface and depth, epistemes and ‘choices’, or inclusion and exclusion. It is the point where a question can be asked about possible alternatives to the great alternative (or the great divide) itself. Therefore – in Pascalian terms – a possibility must be faced that epistemes are not simply divided internally by some local points of heresies or secondary conflicts, but they are caught in the bifurcation of several fundamental modalities of thought, which are not just successive, but permanently antithetic, even if some of them are still ‘to come’ or in a state of ‘imminence’ … It is this ‘transgressive’ figure of epistemological reflection according to Foucault that I want now to investigate. However, we cannot forget that its insistence in the book (particularly wherever Foucault invokes the allegoric figure of the ‘place of the king’, which he draws from his virtuoso analysis of Velazquez’s Las Meninas) has nothing to do with a pure formal definition of spaces and displacements between methodologies or even rationalities. It is closely related to the ‘object’ or ‘thing’ whose domination and fragility forms the general content of modern knowledge according to Foucault, namely ‘human finitude’ or ‘human historicity’. We must therefore now discuss his critique of anthropology in a more precise manner.

The ‘Quasi-Transcendental’ and the ‘Anthropological Sleep’

At the core of what, in a terminology borrowed from Althusser, we could call Foucault’s own version of ‘theoretical anti-humanism’, there lies a notion that represents a bold novelty and evokes deep resonances in Western philosophy and culture: the ‘empirical-transcendental doublet’. As we know, this doublet (which I am tempted to simply call ‘ET’) names a ubiquitous being or entity which should feature at the same time on both sides of the divide allowing us, in modern philosophy, to describe knowledge as an activity: it refers to an object among others which, on the other side, is also a faculty of representation confronting all objects (whether as a perception of the world, or its construction, or its appropriation by means of concepts and theories). While apparently breaking with an onto-theological tradition for which such unities of opposites exist only in God (or the Absolute), or derive from a privileged relationship to God (where Man as a ‘creature’ bears the resemblance of his ‘creator’), modern philosophy (especially in its Kantian, critical version) would explain that this double inscription characterizes ‘Man’ as a generic being, which is part of Nature while also being able to perceive or conceive it as a system of objects and an idea; which, therefore is also a ‘subject’, or, in a more complex formulation, a moral person bearing the universal subject of experience. This being is not absolute at all, on the contrary it is ‘finite’, in a double sense: because it exists only in relation to other beings in a given external environment, as a product of evolution, and because its capacity to appropriate the world is internally limited (in Kantian terms: through the restrictive condition that concepts must always be ‘imaginable’ in sensible forms).

Foucault’s innovation in the discussion of this central question in Western philosophy lies in the fact that he is not content with either accepting or rejecting it, but wants to provide a genealogy of its invention and content. (Incidentally, this whole discussion is one of the clearest illustrations of what Foucault means by ‘genealogical inquiry’, in the middle of the supposedly different ‘archaeological’ construction – which shows the absurdity of opposing abstractly the two ‘methods’.) And although it presupposes a long tradition in Western philosophical, moral and theological discourses, the immediate instrument of his genealogy is a specific epistemological category: that of the ‘quasi-transcendental(s)’, or more generally the quasi-transcendental function of certain concepts (typically: Life, Labour, Language) which serve to unify the regions or disciplines of modern knowledge, and to identify the attributes or conditions of human nature. It is this definition and critical use of the ‘hybrid’ notion of the quasi-transcendental that we need to clarify first.

I spoke of an epistemological category, but this is both tautological and equivocal. Borrowing from the Kantian terminology of which Foucault is making a free or ‘heretic’ use, we should rather say that the quasi-transcendentals are ideas: the ideas of Life (or organism), of Language (in its twofold modality, which the French more customarily distinguishes: langage as capacity and langue or ‘idiom’ as institution), Labour (or better productive labour). According to Foucault, however, these ideas are not only ‘regulative’ but constitutive of the new positivities of the modern age, because they are scientifically defined and empirically investigated. Foucault’s thesis is nevertheless that each of them indicates an excess of reality (albeit each time in a different way) with respect to representation, which for that reason must remain ‘uncovered’. As I indicated earlier, they name a thing (or a ‘power’) that is supposed or anticipated as a ‘real’ beyond representable reality itself. We may conclude that the three ideas are constitutive because they stand on the limit of the representable, from where they prescribe its organization in an analogous manner. But then the question inevitably arises: what do they have in common? Why do they contribute to the emergence of a single episteme, as if expressing some latent ‘event’ which would be the same for them all? Here we must be very careful, because Foucault’s answer is a complex one, ‘repeating’ something of the critical tradition only to be able, in the end, to criticize it from the inside. The answer is not that they form part or express aspects of a single superior idea, which would be the idea of ‘Man’ (or the Human), as if Foucault had reformulated the famous Kantian typology of ‘transcendental questions’, simply suggesting that the three questions: What is Life? What is Language? What is Labour? must be subsumed under a fourth one – What is Man? – which would be the ultimate question, bearing the capacity to endow the new episteme with its essential transdisciplinary function (a function without which, as we know, the very idea of episteme is deprived of epistemological interest).

‘Man’ is not the deep source of the new epistemic figure, it is rather the ‘surface effect’, not to say the mirage of the epistemic transformation, and in particular it is not what really unifies the ‘human sciences’, but only what the human sciences project as their common ideal object. One is tempted to say their common opinion. However, this only displaces the question, because there must exist nevertheless a transdisciplinary notion that traverses or ‘intersects with’ all the sciences (of Life, Language, Labour), in order to induce the anthropological question and install the idea of Man as a common horizon. I believe that, from Foucault’s point of view, this transdisciplinary notion is historicity, or even better, historicity articulated within the horizon of finitude (since Foucault never claimed that knowledge had been unaware of history or historicity before the modern age). We may therefore summarize this first understanding of the function of quasi-transcendentals in the following manner:

they unify each discipline around its constitutive problem, which is to investigate the phenomena of Life, or Language, or Labour as manifestations of a finite (or conditioned, contingent) ‘historical’ power (of permanent transformation);

At this point, however, Foucault exhibits a possible bifurcation (therefore a latent point of heresy, which is exemplified in the history of philosophical systems): the constitutive circle can be elaborated in two different ways. Quasi-transcendental ideas could become foundations for new metaphysical systems, each in its own right: metaphysics of Life, metaphysics of Language, or even metaphysics of Labour (or production: Marxist ‘dialectical materialism’). This would mean that each of the quasi-transcendentals emerges in turn as a Name of Being as such, for which the others represent only secondary or partial expressions. We know that this actually exists, but Foucault submits that it is not the dominant or characteristic tendency within the modern episteme: both because it tends to obliterate the real difference of positivities (or negate the empiricity of the sciences, which goes along with their pluralism), therefore nullifying the very question of transdisciplinarity; and because it amounts, in fact, to transmuting the historical finitude which they have in common, or which they project as a common horizon, into renewed figures of the Absolute. For this reason, the ‘choice’ operated by the modern episteme (or its dominant, most typical orientation) must be the opposite one: not a regression towards metaphysics, but a redoubling of the ‘quasi-transcendental’ itself, or a reformulation of the ‘circle’ in the form of a relation in a mirror, whereby the historicity of the empirical sciences is ‘founded’ in the finitude of a transcendental subject (‘Man’ as a limited power of knowledge), while the finitude on the side of the objects (their material limitations and contingency) imposes a ‘constitutive historicity’ on the (human) subject itself. To describe this circularity is precisely the function of the ‘analytic of finitude’, and it is at this point that the definition of the quasi-transcendentals becomes the elucidation of the ‘anthropological’ function of the empirical-transcendental doublet, which not only presents itself as a crystallization of an epistemological circularity (being at the same time founding and founded, originary and derived, etc.), but as ‘ontologically’ consisting in a circle, since it is at the same time subject and object. This is a paradox from the logical point of view, or perhaps an absurdity from a rigorous ‘critical’ perspective which wants to distinguish the two functions, but it is the very core of the promotion of ‘Man’ as the founding idea of its own ‘sciences’: at the same time their general ‘object’ of investigation and the ultimate ‘subject’ of their development.

Quasi-transcendentals all involve the same identification of historicity with a finitude which cannot be reduced to the lack or the retreat of the infinite (as it used to be thought in the ‘classical’ philosophical systems), ¹⁰ but must be conceived as self-limitation, or a relationship of the finite being with itself which remains affected with some unthinkable intrinsic otherness. Nevertheless, it is not the case that they equally contribute to a critique of the critique. In fact, there emerges an extraordinary dissymmetry and difference of interest for Foucault between the cases of Life, Language, and Labour; and therefore the corresponding disciplines of biology, philology, and economy. Only language radically exhibits the relationship of the empirical-transcendental doublet to its intrinsic otherness (or the necessary relation of any cogito to the unthought) in a manner that breaks the anthropological circle, or disturbs the anthropological ‘sleep’. In other terms, language has an intrinsic relationship to the unthought and a capacity to manifest the instability of finitude that neither life nor labour possess, because language has an intrinsic relationship at the same time with the sciences (‘knowledge’) and a certain literature. This is a position proper to Foucault at the time of The Order of Things, also strongly asserted in such contemporary essays as ‘The Thought of (from) the Outside’ (1966 – a tribute to Blanchot and a commentary of the capacity of literature to destroy the mastery of the subject). This will not be maintained in the same terms throughout Foucault’s career, and in this sense there will be a shift in the genealogical project itself. This is an idea that seems to me more interesting than the more simple one that Foucault exchanged an ‘archaeology’ for a ‘genealogy’. In The Order of Things it is implicitly based on the supposition that the Kantian philosophy was prevented from uncovering within its own project a more radical possibility of critique by the fact that it stubbornly maintained the primacy of the ‘I think’, or did not substitute the ‘I think’ with an ‘I speak’. This should return it to the limitations of the anthropological discourse even before it could begin to question them – an idea that Foucault had elaborated since his early work on Kant’s Anthropology from a Pragmatic Point of View, which remained unpublished at the time, but is now available (Foucault, 2008). We can comment on this critical function of language as follows.

Foucault calls the ‘human sciences’ that would ultimately contribute to unsettling the idea of the human, ‘counter-sciences’ (Foucault, 2002: 378–80). They represent a point of heresy of the second order within the field of the human sciences. Two of them are explicitly identified and discussed: first, ‘psychoanalysis’ (in terms which strangely combine Foucault’s own reflections on the ‘limit-experiences’, where death and madness are destroying the communicative function of language, with quasi-Lacanian considerations on the articulation of the Law, desire and language, which forms the structure of the unconscious and commands the possibility of the cure (see Foucault, 2002: 375); and second, ‘ethnology’ (in terms which directly evoke a structuralist problematic à la Lévi-Strauss ¹¹ and connect it to a critique of the ‘colonial’ dimension of western culture – one of the very few ‘political’ indications in The Order of Things). Of course, they share a capacity to push ‘finitude’ to the recognition of its own constitutive limits, inasmuch as their ‘object’ becomes the unthought itself, whose indeterminate alterity is also incompatible with any rational or empirical construction of the human as a ‘subject’. But clearly this capacity does not result (or only virtually) from the traditional definitions of psychoanalysis as a psychological and medical technique, or ethnology as a form of sociology applied to the ‘primitive’ societies. It is entirely linked to the ‘structuralist’ recasting operated by contemporary theorists such as Lacan and Lévi-Strauss, who have recognized the primacy of language within their own fields, and adopted a scientific model provided by structural linguistics. This takes us to the most difficult question – the one that also involves Foucault’s extremely strange oblique (not to say perverse) relationship to structuralism as an epistemological program in these years, which was largely perceived as an ‘extension’ of linguistic structuralism towards anthropological disciplines, in the broad sense. ¹² Should we say that, for Foucault, linguistics is the ‘counter-science’ par excellence? And does it, from that point of view, provide something like a principle of ‘transgressive transdisciplinarity’, i.e. a critical idea that crosses the boundaries of disciplines, as they progressively adopt a model that was first developed by linguists, refuting historicism and manifesting a new, subversive, understanding of finitude itself?

The answer seems to be explicitly yes, in the first instance, witness the development, immediately preceding the final considerations on the ‘near end’ of the ‘figure of man’:

Whereupon there is formed the theme of a pure theory of language which would provide the ethnology and the psychoanalysis thus conceived with their formal model. There would thus be a discipline that could cover in a single movement both the dimension of ethnology that relates the human sciences to the positivities in which they are framed and the dimension of psychoanalysis that relates the knowledge of man to the finitude that gives it its foundation. In linguistics, one would have a science perfectly founded in the order of positivities exterior to man … which, after traversing the whole space of the human sciences, would encounter the question of finitude (since it is through language, and within it, that thought is able to think: so that it is in itself a positivity with the value of a fundamental). Above ethnology and psychoanalysis, or, more exactly, interwoven with them, a third ‘counter-science’ would appear to traverse, animate, and disturb the whole constituted field of the human sciences … it would form the most general contestation of that field. Like the two other counter-sciences, it would make visible, in a discursive mode, the frontier-forms of the human sciences; like them, it would situate its experience in those enlightened and dangerous regions where the knowledge of man acts out, in the form of the unconscious and of historicity, its relation with what renders them possible. In ‘exposing’ it, these three counter-sciences threaten the very thing that made it possible for man to be known. Thus we see the destiny of man being spun before our very eyes, but being spun backwards … And is that not one way of bringing about its end? For linguistics no more speak of man himself than do psychoanalysis and ethnology. (Foucault, 2002: 381)

But ultimately, this is not the case – and the reason why Foucault writes in the conditional tense. Or it is the case only on the condition that structural (or formal) linguistics ‘communicates’ with another way of ‘de-subjectivizing’ the use and the meaning of language, becoming itself subverted by it: not the science that masters language, but the knowledge that lets itself be subjected by the ‘sovereignty’ of language. That other de-subjectivation is provided by literature in its modernist, anti-humanist understanding, whose theory and practice is understood to have been initiated by Mallarmé and accomplished by Blanchot. ¹³ Otherwise, the ‘return of language’ in contemporary discourse would be nothing other than a return to the ‘order of discourse’ (which was also the order of representation), which dominated the episteme of the classical age – even if in an abstract or formal manner. (This is something that many ‘structuralists’ or formal linguists gladly admitted at the time, when they discovered the affinities between their semantic and syntactic structures and the models of the Grammaire générale or ‘Cartesian Linguistics’.) This is a crucial indication, because it shows that – in the same place, at the same ‘point’ – not two, but three possibilities are offered: (i) a continuation of the ‘anthropological sleep’ (or a development of the human sciences on the continuous assumption that their ‘subject-object’ is Man, the empirical-transcendental being); (ii) an ‘anachronistic’ reconstitution of the Order of Discourse in the guise of formalized anthropological discipline (which could perhaps appear, with respect to the classical models of representation, as a ‘fusion’ of mathesis and taxonomia, in the restored horizon of universality); (iii) an enigmatic (and perhaps aporetic) critique of criticism, or a new ‘analytic of finitude’, where it is not the normality of the human that is considered the rational object of science, but the ‘excessive powers’ or extremities announced by Nietzsche which erase the subject (or its foundational, ‘sovereign’ place), with the help of a new coincidentia oppositorum, the contemporary encounter of formalization and literature. ‘Structuralism’ is not the answer, rather it is the stake, because its concept is torn between the last two possibilities, and in fact indicates their antagonism. But in any case, it rules out historicism – which seems to mean also history.

Foucault’s section on ‘History’ in the last chapter of the book is remarkably convoluted because it is aiming at two targets at the same time. Given that ‘historicity’ was this transdisciplinary idea which could be retrieved (in specific, but analogous manners) in every modern science (as historicity of Life, historicity of Language, historicity of Labour), thus making it possible to constitute the anthropological ‘doublet’ by defining the subject itself as ‘historical being’, there is a necessity to demonstrate that finitude can be thought beyond historicity, or rather beneath historicity, in the form of its excesses or extremities. And, in fact, what the unconscious, or the violence of madness, or the limit-experience of death formally have in common is their ‘a-historical’ character, at least from the point of view of the ‘laws of history’ (which are the laws of progress, or evolution, of organization). Better said, it is the uncovering of another historicity (later, as we know, called that of the ‘event’, but the term already plays a crucial role in The Order of Things in ‘disposing’ of the figures of knowledge), which is neither objective nor subjective. But there is also a necessity to demonstrate that history as such (as a discipline) cannot become a ‘counter-science’ in its own right, a thesis which seems in sharp contrast with what Foucault will propose later (rather soon, in fact, but in a completely different intellectual and political conjuncture), when he will retrieve in the archives and appropriate for himself the notion of a ‘counter-history’, which bears at least a formal resemblance to the ‘counter-sciences’ (Foucault, 2003). In this moment, I submit, the demonstration takes the double form of showing that there can exist no such thing as a ‘structural history’, and that the internal critique of anthropology takes the form of a reconstruction, involving a preliminary dissociation of the motives of historicity and finitude, which made it possible to construct the empirical-transcendental doublet. The latter is an implicit rejoinder to a Heideggerian notion of Dasein that would avoid any anthropological recuperation (as in Heidegger’s own mixtures of Ek-sistenz and Sorge with community or authenticity). The former obviously, once again, includes a refutation of the Althusserian program of a structuralist-Marxist ‘science of history’ (or structuralist theory of social formations) – another very clear indication of polemical intentions. They do not define which ‘place’ Foucault wanted to institute for the epistemological ‘choice’ (or choices) that he allegorically covered with the repetition of Nietzsche’s prophecy about the ‘death of man’ (following the ‘death of God’, or even identical with it). But they indicate in which strategic field the construction of his own ‘point of heresy’ must be located.

The Place of the King

At the beginning of Chapter IX (‘Man and His Doubles’), Foucault installed a section called ‘The Place of the King’, which begins to explain why the idea/figure of ‘Man’ should be absent from the construction of the classical episteme (qua Order of Representation, where the adequacy of Thinking and Being is assumed as an effect of the transparency of Discourse), but necessarily present (or even omnipresent) in the construction of the modern episteme (qua generalized historicization of knowledge, where the human is at the same time the effect of all objective transformations and the source of a reflexive idea of history). This involves in particular a discussion of the difference between a ‘natural concept of man’ and a ‘concept of human nature’, and a reiteration of the Cartesian cogito, where ‘I think’ and ‘I am’ are immediately identified, without any anthropological mediation. But the same section – hence its title – includes a return to the theme of the elusive sovereign of representation, as it had been allegorically presented in the opening chapter of the book, through the interpretation of Velazquez’s picture, Las Meninas. This return plays a very disturbing role. Although Foucault proposes it in a strangely detached tone – not deprived of some coquetterie or intellectual dandyism – what it suggests is that the two epistemes must delineate (as their internal limit or perspective) the same place (the ‘place of the king’, or the sovereign place of the sovereign) – except that, in the first case, the place remains empty (or is a pure void), whereas, in the second, it is occupied or filled (even by a paradoxical ‘double’ figure). How can the two epistemes be radically distinct, if they contain the same ‘place’, or can be subjected to the same perspective? How can the two ‘places’ be ‘the same’, or isomorphic, involving the possibility of such a binary alternative (empty vs. filled, or negatively vs. positively defined), if a radical mutation had taken place, whereby the very construction of knowledge changed, so that each knowledge offered to the other only the ‘absurd’ meanings of a Borgesian ‘Chinese encyclopaedia’?

There are different ways indeed to reflect on this difficulty. For the sake of brevity, I will leave aside considerations that resort either to the ‘history of ideas’ or the ‘history of philosophy’, because they do not suffice to resolve the formal problem, which articulates a strategic with an epistemological dimension. We could draw the attention to the fact that the ‘absent sovereign’ in the first presentation of the picture is a king (immediately redoubled, it is true, with an artist, the painter himself), whereas the ‘present subject’ in the new age of knowledge announced by this allegoric repetition is an ordinary (or generic, collective) man. This consideration cannot be neglected if one believes that there is no innocent signifier, but it might be misleading in interpreting Foucault’s dispositive, because it might bifurcate towards the idea that, after all, the succession of the classical and modern epistemes expresses a social and political transformation which leads from an implicitly monarchic order towards a republican humanism. This is certainly not what Foucault wanted to explain, even at a symbolic level. But it could take us in the direction of another interpretation, which would explain that an ‘absent sovereign’ is in fact a name for God, therefore what the allegory suggests is that the classical order of representation (or positive infinity displayed in the rationality of discourse) is ‘haunted’ by the absence of God (a Pascalian theme). And it is only when this absence becomes visible as such that God can be ‘incarnated’ again, albeit not as Himself, but as a purely finite being. This would not be uninteresting, if we want to understand why Foucault in the end chose to replicate Nietzsche’s ‘death of God’ in the form of the ‘death of Man’, rather than, for example, following Feuerbachian and Marxian schemes of the ‘secularization’ of theology in modern culture. But it is an allegory to interpret another allegory, whereas what we need to explain is the play of visibility and invisibility itself.

I draw a first indication from an essay by John Rajchman on ‘Foucault’s Art of Seeing’ that discusses the relationship between the archaeology of epistemes and the question of visibilities (in the wake of Deleuze’s idea that Foucault should be considered ‘a seer’ or voyant):

The visual thinking of a period would thus have a positive organization. But the organization is not rooted in keeping something concealed. As Foucault came to realize, the ‘classical’ way of making madness visible was not based on the repression or concealment of the true way of seeing it. The conceptual scheme that determines what can be seen is, in the phrase of the Archaeology, ‘invisible but not hidden’. The visibility of a period may be invisible to it, but not as something hidden or kept from sight. What is invisible is just the light which illuminates things or makes them visible. (Rajchman, 1988: 92–3)

In order to jump from there to a question of the second order, which would be the question of the ‘invisible’ relationship between antithetic regimes of visibility themselves, we may complement these indications with Vinciguerra’s recent work, La representation excessive (2013). This is a book that with great precision and novelty deals with the relationship between the construction of objectivity and the developments of geometry in the works of Descartes, Leibniz, Locke, and Pascal – moving back, therefore, from the field of taxonomia to that of mathesis universalis. In his concluding remarks, Vinciguerra addresses the same puzzle about the epistemological ‘ubiquity’ of the ‘place of the king’, which Foucault’s analysis of the Velazquez painting had first identified as a representation of the conditions of representation themselves:

In this place, at the same time empty and attractive, Foucault then wanted … to find, as in a palimpsest, the image of 19th century Man … that a new episteme will make the core of its knowledge. How come that this subject-object, which breaks with the transparency of Discourse … could be offered to visibility by classical philosophy, in a picture that would appear as a miraculous tearing apart of its own thinking? (Vinciguerra, 2013: 164)

I think that ‘points of heresy’ of different orders must be distinguished. Foucault is not discussing the same differences or discursive conflicts when he distinguishes bifurcations or oppositions which are meaningful within a given rationality, but not outside it, and when he identifies the latent conflict that makes visible for a certain regime of knowledge what is invisible in another one – even if, as we have seen, there are ‘choices’ which move from the centre of the episteme towards its constitutive limits, thus approaching a reversal of its perspective, or a subversion of its rationality, creating for it a danger of disruption. Further, the ‘place of the king’ is indeed an allegory, through which Foucault attempts to think, not the ‘unthinkable’, but the impossible: a representation of certain objects or a formulation of certain epistemic questions in a rationality for which they cannot ‘exist’. But, pace Vinciguerra, in order to ‘problematize’ this thinking of the impossible, there is no need to create an equivalent of the absolute knowledge or to install in the position of the philosophical synoptikos a ‘theory of theories’, an epistemic hyperspace. It is only a question of reaching the limit and, as it were, staying at the limit to analyse or describe it ‘as an outside’, by turning outwards its inside. This is, in Foucault’s perspective, a matter of choice or decision, rather than theory, even if it is based on careful description and archival work.

But there is more. An additional dialectical twist can be brought to this configuration of choice, as our discussion of the ‘counter-sciences’ and their problematic relationship to the unthought should have indicated. To reach this limit is not to find oneself oscillating indefinitely between two epistemic possibilities, to remain enclosed in the symmetry of the existing closures (representation, anthropology). On the contrary, it is to discover a ‘point of heresy’ that is in fact a triple point, where the third, the ‘neuter’ that is neither one nor the other can be, at least, anticipated, or even ‘read’ in the transgressive moments of existing knowledge. Allegorically, this triple point is one where it can be a question not only of ‘choosing’ between God (or Nature) and Man, but of anticipating the Overman, as Nietzsche called the neuter. Philosophically, or epistemologically, this triple point is one where a ‘return of language’ that was carried on by the development of philology and linguistics themselves, but above all by the literary ‘externalization of thought’, confers upon the human sciences the possibility of breaking the circle of their own constitutive paralogism. Strategically, this triple point is one where the vacillation of pre-existing epistemes or models of rationality does not open so much the perspective of a new age for knowledge (as some ‘structuralists’ and particularly some ‘Marxists’ seem to have believed), as the possibility of thinking otherwise, in the light of their own alterity, that which historical knowledge had subjected to quasi-transcendental ideas. This is something that, strangely, seems closer to a notion of ‘deconstruction’ than to a notion of ‘epistemological break’. Foucault’s ambitious ‘book of philosophy’, The Order of Things, was probably not purely in his opinion a way to ‘discover’ this point of heresy. It was rather an attempt at constructing it, in order to be able to ‘occupy’ it and speak from there. Whether this point – the ‘place of the king’, where not only the king himself, or his painter, but also an equally absent but even more insistent spectator, ‘seer’, could come – would appear to him as a place to stay or to work from for long is another matter. The triple point thus acquires a decisive function at two different levels (but its allegorization as ‘the place of the king’ also shows why the two levels cannot remain epistemologically separated): it is the point from which, ultimately, the transdisciplinary effect of scientific discoveries concerning Life, Language and Labour can become intelligible. It is also the point that – when reached or anticipated – makes it possible to question the ‘disciplinary order’ itself. Coincidentia oppositorum, once again.