Printer Friendly


"Unwillingly willing science loves error"--Nietzsche, Beyond Good and Evil

In this paper, I divide the major late twentieth-century French philosophers typically classified as "poststructuralist"--namely Gilles Deleuze, Jacques Derrida, and Michel Foucault--based on what I diagnose as their fundamentally divergent epistemological attitudes. I suggest that Deleuze and Derrida, despite differences that in some ways amount to a diametrical opposition, are united by an assumption that it is possible--at least in principle --to produce a discourse that is in some sense continuous with reality. While both pointedly reject what they see as naive correspondence accounts of knowledge, they nevertheless hope to reconcile language and reality in various ways. On the other hand, I claim Foucault thinks the relationship of language and reality in terms of fundamental and irreparable discontinuity, in this respect having more in common with Jacques Lacan than the contemporary compatriots of his with whom he is more often associated.

This epistemological opposition could be read back across a wider swathe of the history of philosophy, but in this present article I will limit my inquiry to this highly specific geographical and historical locus. I would provisionally assert however that the significance of the division I am naming here is less parochial than this discussion might be taken to imply.


Inspiration for the distinction I draw comes from Foucault's (1978) influential depiction--in his introduction to the English translation of Georges Canguilhem's Normal and the Pathological--of twentieth century French philosophy as made up of two strands, distinct but not entirely separate. Foucault (x) does not give these two tendencies definite names, describing them rather in multiple ways as, on the one hand, "a philosophy of experience, of sense, and of subject" and on the other "a philosophy of knowledge, of rationality and of concept".

As Giuseppe Bianco (2011) has pointed out, Foucault's contribution here is not in itself particularly original, following rather in the footsteps of more senior members of the rational-conceptual camp--including Canguilhem, Claude Levi-Strauss, and Louis Althusser--in dividing things this way. Bianco shows how the marking of this historic division has its roots in the pointed drawing of battle lines between phenomenology and structuralism in the 1960s. Bianco also notes that this drawing of battle lines in French philosophy did not stop with Foucault, but has been continued in similar vein by Alain Badiou. I might humbly situate my own contribution here in this same partisan tendency, as a further extension of or redrawing of these lines in a continuing conflict.

Foucault in making his distinction divides up the generation of philosophers that came before him, putting Jean-Paul Sartre and Maurice Merleau-Ponty on the side of subjectivity, and Jean Cavailles, Bachelard and Canguilhem on the side of the concept. It is unclear whether his categorization is meant to be exhaustive in relation to French philosophy, but there are certainly some relatively major figures he does not name. Some French Hegelians are missing, in particular Foucault's sometime teacher Jean Hyppolite--though one such Hegelians, Alexandre Koyre, is included by Foucault in the conceptual faction, due presumably to his work in the history of science. Nietzschean literary figures such as Georges Bataille and Pierre Klossowski are also absent, which I think may be accounted for by the fact they can be considered writers rather than philosophers per se.

Foucault mentions multiple divisions prominent in French thought in the period under discussion in addition to the one he proposes: those between 'Marxists and non-Marxists, Freudians and non-Freudians, specialists in a single discipline and philosophers, academics and non-academics, theorists and politicians'. He proposes to posit a further bifurcation to be added to these, in no way denying their importance. It is in the same spirit that I now propose my own division. That said, the division I am proposing is posited primarily in relation to a slightly more recent period than Foucault's.

It is not immediately clear how the main division that Foucault proposes or indeed any of the others he mentions--might be said to play out in his own generation. There is no question that he situates himself in the tradition of Canguilhem, but he does have influences from both sides and in his generation there was more of a mixture of influences than in the previous one, whereas at least the canonical older figures he names belonged to one side of the other with relatively little admixture. Among Foucault's peers in the 1960s, there was a general inclination towards the philosophy of the concept in reaction to the reigning dominance of the philosophy of the subject, with Sartre at its head. This reaction encompassed not only Foucault but also the younger post-Althusserian philosophers who are still with us, as well as older French "structuralists," such as Althusser and Lacan.

However, Foucault's most prominent "poststructuralist" contemporaries, Derrida and Deleuze--two philosophers who like Foucault came to prominence during the 1960s (although slightly after him) and constituted the dominant position in French philosophy by the 1980s in particular--stand in profoundly ambiguous relations to his division. Deleuze's thought is certainly readily describable as a "philosophy of the concept," given his identification of philosophy with conceptual innovation (Deleuze 1994). However, he does not stand as closely in the lineage of either "structuralism" or French philosophy of science--unlike Foucault or the other figures Foucault names on the conceptual side--and Deleuze's emphasis on experience and "affect" is at least suggestive of a philosophy of the subject, indicating a certain debt to phenomenology, in particular via Merleau-Ponty. (1) Derrida is still closer to the side of the subject: though he was influenced very explicitly by structuralism, it is equally clear that he situates himself primarily in the broad (post)phenomenological tradition that was the major affiliation of those Foucault deems philosophers of the subject in the preceding generation. Notwithstanding Foucault's (1978, x) suggestion--following Cavailles (1970)--that both the strands he identifies take up the phenomenology of Edmund Husserl when it reaches France in the 1930s, reading it in different directions, phenomenology as such lends itself much more to a philosophy of the subject, and certainly the figures Foucault names as part of this tendency are the most aligned with phenomenology. That said, Derrida radically distances himself from all previously existing philosophy, much as Deleuze strikes a path that is highly idiosyncratic relative to any of his contemporaries, hence these various connections and affiliations are not immediately decisive.

It is precisely in light of this difficulty in tracing the continuation of the two tendencies Foucault had diagnosed in earlier French twentieth century philosophy in his own generation that I will suggest a novel one to divide it. While I speak of a 'generation' here, I will in fact focus on only three thinkers, Deleuze, Foucault, and Derrida, a sub-generation of major French philosophers born between 1925 and 1930. One could extend this range slightly to include other thinkers, such as Jean-Frangois Lyotard, born a year before Deleuze, or Jean Baudrillard, born a year before Derrida, but those thinkers attained somewhat less prominence than the big three, slightly late--and in any case tackling the three figures I have chosen is more than enough to occupy this article. Moreover, it is these three above all that are commonly thought of together, typically under the umbrella term of 'poststructuralism'. This term of course itself does indicate a certain lack of theoretical unity in this generation, indeed that it is united only by coming after 'structuralism'. Still, I think it is imagined commonly enough that Derrida, Deleuze, and Foucault can be treated as if they are of the same ilk that the corrective I am offering here might be useful. Moreover, when they are divided up, it is typically by placing Foucault and Deleuze together as differing from Derrida (e.g. Agamben 1999).


The principle of the division I will posit here is derived from Foucault's Introduction, albeit obliquely. My division bears some relation to Foucault's, but is not meant as a redefinition it; it rather meant as distinct in its own right.

Foucault begins his introduction to The Normal and the Pathological by describing the cleavage in twentieth century French philosophy, but moves on to describe Canguilhem's thought in some detail. It is the point on which Foucault ends that particularly interests me. He suggests there that for Canguilhem, at base, the human is
   a living being who is never completely at home, a living being
   dedicated to "error" and destined, in the end, to "error". . . .
   Error is at the root of what makes human thought and its history.
   The opposition of true and false, the values we attribute to both,
   the effects of power that different societies and different
   institutions link to this division--even all this is perhaps only
   the latest response to this possibility of error, which is
   intrinsic to life. If the history of science is discontinuous, that
   is, if it can be analysed only as a series of "corrections," as a
   new distribution of true and false which never finally, once and
   for all, liberates the truth, it is because there, too, "error"
   constitutes not overlooking or delaying a truth but the dimension
   proper to the life of men and to the time of the species.

I am not concerned for the moment with the extent to which this is an accurate exegesis of Canguilhem and to what extent this is Foucault's own position, though I would argue that Canguilhem's vitalism points in a less relativist (and more progressivist) direction than Foucault's reading, and even on this reading Canguilhem remains more vitalist and anthropological than

Foucault himself is. What interests me is the general epistemological perspective that Foucault lays out here that error is the basis of knowledge. It is on this radical note that Foucault (xx) ends his Introduction.

This position is not synonymous with the "philosophy of the concept," though it is aligned with it in Foucault and Canguilhem's cases. It is rather whatever else it may be allied to or alloyed with--indicative of a distinctive epistemological orientation, which I will call discontinuism, the view that knowledge is fundamentally discontinuous with extra-discursive reality, and that all human life bears witness to the effects of this mismatch and the generation of error that goes along with any attempt at speaking the truth. It is this orientation that I want to posit as constituting a distinct and fundamental point of divergence between Foucault and his most prominent poststructuralist peers, whom I will by contrast characterise as continuist.

If human knowledge is always error then there is no entirely correct knowledge. Now, this premise might make discontinuism seem simply to be another word for fallibilism. There are any number of differences between these positions, however, coming as they do from quite different philosophical loci and traditions. At heart, though, I take it the difference is that while discontinuism claims that all knowledge actually is erroneous, fallibilism consists in the belief that any and all knowledge might be erroneous and is hence always defeasible. That is to say that discontinuists claim that all our beliefs actually are to some extent false, whereas fallibilism merely allows the universal possibility of error. Fallibilism and discontinuism are similar in that neither necessarily undermines scientific truth-claims, but rather instead both tend to support making such claims on the basis that they are not final statements of fact, but provisional and defeasible findings. The difference here is that while fallibilism allows a continual scientific progress through the improvement of findings, discontinuism tends to the more relativist conclusion that there is no clear arbiter of correctness and that since all positions are essentially erroneous, there is no superior position that can be moved to through falsification.

Discontinuism holds that knowledge will always be erroneous because there is an ineluctable scission between thought and its object. Though Foucault thematises the concept of "error" only here, in relation to Canguilhem, he reiterates this basic epistemological insight multiple times in the course of his career in other terms. As Foucault (1994b, 552) puts it in an earlier text, 'Truth and Juridical Forms', 'la connaissance est toujours une meconnaissance'. This French epigram is ultimately untranslatable, since there is no English term equivalent to meconnaissance, literally "misknowledge," but we may capture the point by saying that all knowing is a form of misknowing, or perhaps less outlandishly, though more liberally, that all understanding is a form of misunderstanding. One may refer also to his remark in his essay on Nietzsche and genealogy that "Knowledge is not made for understanding; it is made for cutting" (Foucault 1998, 380). It should be noted that both this line and his line from 'Truth and Juridical Forms' are exegeses of Nietzsche, and that Nietzsche's position is seminal to Foucault's here, though also that his exegeses of Nietzsche are tendentious and may to a large extent be taken as a ventriloquism of Foucault's own position. One can refer here also to a yet earlier injunction of Foucault's, not indexed to Nietzsche, from his 1970 inaugural lecture at the College de France, L'ordre du discours (which has appeared in English under multiple different titles, although never, as it did in French, as a book), that 'We must conceive discourse as a violence that we do to things'. (2) Foucault's infatuation with the metaphor of violence was of relatively brief duration, but the theme of a fundamental mismatch between words and things is consistent across his career, if instantiated during the late 1960s primarily by refusing to pay much attention to the latter. Nevertheless, this problematic is identified in the title of his Les mots et les choses ("Words and Things," published in English as The Order of Things), even though the book almost exclusively concerns itself with discourse. In the preface to this book, Foucault (2002, xxii) inveighs that 'there exists, below the level of [a culture's] spontaneous orders, things that are in themselves capable of being ordered', However, the discontinuist orientation is I think encapsulated most clearly not in Foucault's work, or in Canguilhem's, but in that of a contemporary who also may be said to belong to the conceptual camp in twentieth century French thought, but who does not figure in Foucault's counting of its members, and indeed who is not typically counted as a philosopher at all, and certainly doesn't consider himself one: Jacques Lacan. Lacan is also sometimes counted as a 'poststructuralist', though is clearly of an older generation than the figures I am treating under that label and is more commonly called a 'structuralist' (it is perhaps no coincidence that Foucault was himself the youngest French intellectual to find himself regularly called 'structuralist' during the 1960s). (3) Lacan conceptualises the scission between knowledge and its object schematically as the disjuncture between knowledge and truth--the former being symbolic, the latter real (Lacan 2006, 798). This conceptual division is quite foreign to Foucault, who uses the word 'truth' like 'knowledge' to refer to contingent ways of ordering statements. Still, as Amy Allen (2018, 179) notes, Foucault's Blanchotian notion of an 'outside' of discourse is closely equivalent to Lacan's concept of an extra-symbolic 'real'. Lacan is not alone in defining 'truth' as extra-discursive, however--Sartre (1992, 3) does the same thing, albeit without suggesting that the cleavage between truth and language is as basic or irreparable. The difference between Foucault and Lacan here is merely a semantic one, but it is apt to confuse, as is Lacan's conceptual division between reality (which for him is an imaginary image) and the real (which is what lies outside symbolisation). Indeed, as I argue elsewhere (author self-citation), such terminological differences lead Foucault himself to fail to understand how close his position is to Lacan's. I prefer to conceptualize the key distinction as the difference between knowledge on the one hand and the reality that knowledge tries but fails to describe on the other, (4) and to conceptualize the relation between the two things as always already riven. I would argue, after Foucault and Lacan, that there is not merely a gap here, but a tear, because this discontinuity appears for an animal which previously had not, as an infant, experienced it, and because language is an irruption into a reality that it does not leave undamaged. I regard this discontinuity between language and reality as necessary because of a complexity problem: reality is vastly more complex than language, hence the former cannot be adequately represented in the latter. This is true not only of the totality of the universe, but even of small pieces of it: even momentary human experiences contain a richness of elements that we cannot hope to be fully describe with language. There is in addition something akin to the qualia problem here: even if an experience and a language were of equal complexity, putting something in language would never fully convey what it is to have the experience.

One might argue that this implies only that we simply partially understand rather than always misunderstand a world that we will never understand entirely. (5) The problem is that partial understanding does always imply misunderstanding, because facts are contextual. If I understand some fact about a thing, but miss others, I do not understand the thing, and therefore do not understand the fact. Indeed, this way of speaking is flawed in itself: there are no 'facts' which exist as discrete bits of information that we can correctly understand without understanding others, but rather we understand things according to an imperfect total picture that is ultimately always erroneous both in its details and as a whole.

I must distinguish what I am calling "discontinuism" from a pre-existing use of the discontinuisme in French, which indeed is invoked by Foucault (1978 xv) adjectivally in scare quotes in his Introduction to Canguilhem. While the word can refer to any form of emphasis on discontinuity, it primarily--as it does in Foucault's usage here--refers to the emphasis on discontinuity in the history of knowledge. This form of discontinuism, which I will qualify after Foucault (xvi) as "historico-epistemological," is also one to which his thought clearly cleaves, following Canguilhem, and--though Foucault does not mention him here--Gaston Bachelard, whose concept of rupture Foucault (2002) adopts elsewhere. The "discontinuity" Foucault invokes here is in what is considered true between different scientific theories or time periods. Positing such discontinuity implies a form of relativism about scientific truth for Foucault (1978 xv), but discontinism is not merely historical relativism: it implies that historical periods are divided by discrete breaks that do not (and cannot) involve a total lack of continuity, but nonetheless constitute discontinuities in the history of knowledge; this could be compatible with a certain historico-epistemological progressivism, though it is not in Foucault's case.

The epistemological orientation I am calling discontinuism is only contingently related to historico-epistemological discontinuism: it is possible to be discontinuist in this sense while being historico-epistemologically continuist or vice versa, since the variation of knowledge throughout history does not logically require an explanation in terms of a scission between thought and reality, and the scission between thought and reality does not logically require periodical revolutionary changes in knowledge. That said, there is a tendency for both forms of discontinuism to correlate to one another, and this is not accidental, inasmuch as a basic mismatch between knowledge and reality would tend to imply that knowledge will tend to lack any secure basis that prevents it changing, that it will never reach a final equilibrium at which it stops changing, and that since it has to find its principle of stability in itself rather than in reality that it will lurch from self-sustaining formation to self-sustaining formation rather than changing continuously. These historicoepistemological shifts also imply importantly though I think that knowledge is nonetheless trying to grasp reality, and it is in this attempt that it is made to continuously change, in part in response to changes in the real itself, in part simply by the weight of its inevitable errors.

One might also confuse the position I am calling discontinuism with Quentin Meillassoux's (2008) critique of "correlationism." What I call "continuism" is not the same thing that Meillassoux, and after him the other "speculative realists," call "correlationism." For Meillassoux (5), "correlationism" refers to "the idea according to which we only ever have access to the correlation between thinking and being, and never to either term considered apart from the other." For Meillassoux, all late modern philosophy is correlationist, including all the thinkers considered here. From my perspective, Meillassoux's position is itself continuist: he attacks those who assume a correlation of thinking and being (which is certainly one form of continuism), but does so not in the name of an ultimate impossibility of describing reality (which is what the perspective I call discontinuism implies by my lights), but rather in favour of arguing for a positive ontology of what is outside thinking (itself a continuist aim, insofar as it aims to reconcile thought to being at a meta level).

This discontinuist perspective can be counterposed most obviously to a progressive vision of scientific knowledge allied to a correspondence theory of truth, according to which the history of science can be characterised as a gradual progress towards a perfectly apt description of reality. This view can be characterized as continuist in its view both of history and of knowledge. I want to suggest that such bald continuism is rare, however; few if any thinkers would explicitly take up such a naive position, particularly in twentieth century French philosophy. Yet, epistemological continuism nonetheless infects it widely, even where there is historical-epistemological dicontinuism. Assertions of discontinuities in the history of knowledge can be reconciled with a continuist epistemology via a broadly Hegelian solution, a dream that ultimately this discontinuity will end and things will coincide once again.

Now, I should perhaps clarify the implication of my invocations "continuity" and "discontinuity". Discontinuists, in either sense, do not believe in an absolute discontinuity--they are not substance dualists with an interaction problem--any more than continuists believe that there is literally no difference at all between words and things, or that nothing ever changes. Foucault (1989) considers concepts to be part of reality, to be material, but nevertheless as a different type of thing to concrete objects (Author citation redacted--Tom Eyers [2011] has similarly claimed Lacan's position regarding language is materialist). They however oppose a conception of the interaction between historical periods, or between words and things, which has them fitting together easily, harmoniously, and without remainder. Indeed, discontinuism should be understood as having the recursive implication that knowledge, being material, is also discontinuous with itself, in the sense that we not only have an erroneous conception of our reality, we also have an erroneous conception of our own knowledge (a problem then that would necessarily affect all of my attempts to understand the thoughts of the thinkers under discussion in this paper). Epistemological discontinuity is not so much a hard break between knowledge and reality, as one that infects knowledge itself as such. We may indeed suspect that discontinuity infects reality as such--and clearly does insofar as knowledge itself is real and exhibits this discontinuity. Continuism differs in thinking that words and things, or even words and other words, are ultimately reconcilable.

Foucault (2006, 105) says that 'continuity is actually a phenomenon of discontinuity'. He glosses this by saying that even things that apparently stay the same are changed by the changing context. Continuists by contrast can be characterised as holding that discontinuity is actually a phenomenon of continuity, that beneath and across all the changes something remains the same, even--I will claim--if it is difference itself as such.


If Foucault and Lacan are exemplary discontinuists, I will argue that Derrida and Deleuze are united in epistemological continuism, albeit of an obscure kind in each case. Given the enormous difference between these two philosophers' thoughts, their continuism manifests itself in quite different ways, although not without some similarities.

There is much that can be said about the triangular relations between Derrida, Deleuze, and Foucault (bracketing Lacan, who was not of their generation and had complicated relationships with each of those three). Their personal relations varied: Foucault taught Derrida but they spent most of their intellectual career at daggers drawn with little contact; Foucault and Deleuze were close personal friends for some years before they fell out; Deleuze and Derrida knew each other only relatively slightly, but according to Derrida (2001) were always on friendly terms. However, I'm interested here solely in the relation between their ideas. While the occasional sniping and general coolness between Foucault and Derrida does reflect the divergence of their positions, it doesn't really quite represent its depth. Conversely, I think that the personal closeness of Deleuze and Foucault did much to conceal--even from the men themselves--a trench between their positions.

Philosophically, there are some noteworthy commonalities between all these three poststructuralists. The decisive influence of Nietzsche is one, though they each took different things from him. They also share a certain ambivalence towards Freudian psychoanalysis, at once influenced profoundly by it but keen to challenge it (although Derrida less than the other two). Beyond that, we find differences: Deleuze considers himself a metaphysician, Derrida rails against metaphysics, whereas Foucault eschews discussing it; Deleuze considers himself a Marxist, while Derrida and Foucault each distance themselves from Marxism (while remaining also quite explicitly influenced by it); perhaps most decisively for our purposes, Derrida and Deleuze both make much of the notion of 'difference', where Foucault does not. Lastly, we should mention the relation to structuralism: Foucault was considered a structuralist in the 1960s, and, while he did not accept the label, he nonetheless closely identified his project with others who were also considered part of this movement, most explicitly Levi-Strauss (Foucault 2002, 413), and also sometimes Lacan and Althusser (Foucault 1994a, 653). Derrida (1997), engaged in an explicit reaction against Saussure's structuralism, turning it against itself, while pillorying Levi-Strauss, using the Saussure's model of language against its inventor by making the relation between words a model for the relation between words and things. Deleuze, for his part, had less overt relation to structuralism than the other two.


Derrida, it seems to me, inherits continuism from phenomenology, despite his significant divergence from its classical mode. Husserl's war cry, hin zu den Sachen, explicitly expresses a desire to base knowledge directly on things, in the sense of phenomena, to deal with the failure of the conventional sciences. French phenomenology in its first generation, in the principal shapes of Merleau-Ponty and Sartre, is resolutely continuist. While ontologically there is a serious divergence between Merleau-Ponty's (1968) conceptualisation of subjectivity as a fold in being--clearly indicating a deep continuity between material reality and consciousness--and Sartre's theory of subjectivity as negativity, and while the latter would seem to amount to a form of ontological discontinuity, it is not at all clear that this is carried over by Sartre into a discontinuist epistemology; rather, he continues to hope for redemption through authenticity and historical progress. Even the late Heidegger's ruminations on language and thinking, which are so influential on Derrida, don't entirely break with the idea that language and reality can and must be brought together.

Still, the general tendency of (post-)phenomenology after Husserl's initial formulation of the project is to depart from mere correspondence in a direction of a problematisation of the relationship between language and reality. Philosophers in this tradition tend to think that language has done violence, but unlike discontinuists to believe that this problem is fixable. Many figures in this tradition are concerned to criticise 'dichotomous' thinking, particularly the positing of splits between subject and object, and hoping to heal these rifts by conceptualising things differently. Hence Emmanuel Levinas for example wants to oppose the dichotomy between self and other. A typical move--again not made by all thinkers of this tradition, but made by Heidegger as an essential part of his philosophical perspective, and then taken up by several thinkers influenced by him--is to posit an historical orthodoxy (which may be called 'Cartesianism' or 'metaphysics') that propagates dichotomies, from which we now need to rescue ourselves. It takes dichotomies that are traditionally regarded as ontologically basic and instead understands them as a matter of an historically contingent ontological view.

This is not to suggest that discontinuists accept these traditional dichotomies. Rather, for it the language-reality dichotomy displaces older ones. While it might appear to replicate them, in particular the Kantian break between the subject and the object, discontinuism differs for one thing in implying discontinuity within the subject itself, insofar as 'subject' is a concept imposed on and deeply altering the non-linguistic reality of subjective experience. Indeed, we can go further and note that words cannot even adequately describe their own meaning exhaustively: language itself is a reality that language cannot accurately describe, that the word 'language' is inadequate to conceptualise. This might seem to suggest that Derrida is right, and that a hard distinction between language and reality is untenable, but I am suggesting that on the contrary it merely indicates how ubiquitous and far-reaching this divide is.

Derrida's project, following Heidegger, is a critique of metaphysics, understood as requiring a wholesale revision of language that will reconfigure what I am calling the reality-language relation. Indeed, I take it that this is the principal target of Derrida's early critique of structuralism. Derrida's (1997) critique of "logocentrism" is explicitly primarily a critique of correspondence theories of truth, of the idea that words are supposed to operate via a one-to-one correlation to things. He takes up Ferdinand de Saussure's insight that words derive their meaning through the interplay of differences from other words within language, and contends against Saussure that this interplay is also how words refer to non-linguistic things, through a play of differences that encompasses both all words and all things. This argument is based in the Heideggerian insight that for humans all objects are 'sign-like'. For Derrida, language and reality are both ultimately a matter of the same thing, what Derrida calls 'writing'. Foucault, by contrast, while allowing certainly that words are something that exist in the same world as things and which therefore can affect them, and to that extent that language is part of reality, wants nonetheless to insist that there is an extent to which language will never be like non-linguistic reality.

From a Foucauldian perspective, Derrida's position is a case of what Foucault (1971b, 21) calls "logophobia," either the mistrust of language as such because of its difficulties in dealing with reality, or the mistrust of the specific language we currently speak. Derrida clearly exhibits the latter tendency, and evinces utterly explicit logophobia, inasmuch as he takes as a chief critical target the logos itself, qua model of the operation of language.

Derrida's elision of the distinction between words and things is epistemological continuism par excellence. Derrida dreams of the emergence of a language which is not logocentric, which does not think of the word as referring the things in a linear way, distinct from other words, but which rather acknowledges the endless play of difference that is found in reality, both linguistic and concrete. He proposes in the interim using old metaphysical concepts 'under erasure', but hopes to get beyond this (Derrida 1997, 61). Rather than hope for correspondence, Derrida hopes that by giving up the hope of referentiality, we can allow words and things to coexist, through an awareness of the ineluctably poetic relations between them.

Heideggerian emphasis on epochal shifts is historico-epistemologically discontinuist, and indeed was influential on both Foucault and Lacan. The difference with this acknowledgement of historical discontinuity by Heidegger and Derrida is that they invoke it as a way of explaining the existence of discontinuity between language and experience, rather than as its result, seeing this as a kind of regrettable perversion of the proper relationship. Heideggerians seek, unlike Foucault (or Lacan), to set things aright (again).

Now, Derrida does diverge from Heidegger by taking up the latter's most radical insights. Derrida (1982, 17) thus attempts to overcome any binary between continuity and discontinuity: for him, the terms of every binary opposition are united in the sameness of their difference. And indeed, it is true that there can be no continuity without discontinuity or vice versa. In a broad sense, then, for Derrida, discontinuity is just as ubiquitous as continuity. However, the radical form of discontinuity as irreparable cleavage between language and reality is not possible for Derrida, precisely because it means privileging discontinuity above continuity. Even if he can allow that there is a generic difference between words and things, deconstruction can at most acknowledge the general difference here as an aporia that cannot be resolved.

From a discontinuist position, however, mere acknowledgment of difference is insufficient to capture the radicality of the phenomenon of discontinuity, on those exceptional occasions when it does occur radically. As Badiou (2001, 26) so incisively instructs us, since everything is different from everything else, difference is so ubiquitous it tells us approximately nothing. To say, as Derrida effectively does, that everything is different from everything else, ends up having more or less the same implication as the arch continuist claim that everything is the same. His valorisation of difference ends up reducing any special difference to banality. Thus, the division between language and reality is elided and made simply another instance of difference.

What can the discontinuist critique of such a position be? Not, surely, that it is erroneous, because just as Derrida's position renders difference banal, discontinuism renders error omnipresent. However, from a discontinuist position it is imperative to acknowledge the existence and importance of error as such, just as it is for Derrida to acknowledge the importance of difference. From Derrida's point of view this specialness of error disappears, since error is reduced to difference. (6) Indeed, error might be said to figure in Derrida's worldview not as ubiquitous to human knowledge, but as what happens when the ubiquity of difference is not acknowledged as such, even if this is not how Derrida conceptualises things.

Discontinuism does not however commit the converse error to Derrida's reduction of error to difference, does not reduce difference to error. Rather, error is peculiar kind of difference specific to human knowledge. This is not to say that discontinuism asserts anthropocentrically that error is a form of difference that is unique in the universe: it cannot pretend to such knowledge. But by the same token we cannot know that error is fundamentally similar to any other kind of difference. While certainly it seems true that this cleavage between knowledge and reality could not exist if there were no difference in the universe, this is a trivial prerequisite.

Discontinuism does mirror Derrideanism however inasmuch as discontinuism implies a special criticism of the meta-error which consists in not acknowledging the significance of error as such. Of course, strictly speaking such a position cannot be conceived as peculiarly erroneous, but we can say it is the one error to which discontinuism as such is a priori opposed. Of course, one could certainly accuse discontinuism of a viscious circularity, since by its own lights even its core thesis must be erroneous. Indeed, it must allow that even it itself is in error, but since error is ubiquitous for discontinuists, the inevitable presence of error is not enough to invalidate a position. Rather, the discontinuist speaks regardless.

Derrida's epistemology is correct as far as it goes: the meanings of words are always ambiguous and everything refers to everything else. The discontinuist disagrees with Derrida about the implications of this, however. Discontinuists take it that what Derrida sees as the failings of "logocentrism" are really ineluctable problems of language use as such that cannot be superseded. It ultimately implies (at least in the extreme form I am imputing to Foucault and Lacan) rather taking a radical step that Derrida refuses, namely entirely sidestepping ontology and metaphysics, rather than Derrida's path of criticising them, placing them under erasure and converting Being into a still more abstract differance.

The discontinuist position is effectively that when we use language we are trying to describe reality even though we are always going to get it wrong. Historically, philosophers have tended to say that we should be getting it entirely right, and that we will get it entirely right one day. Correspondence theory is one mode of saying this, but other continuists--such, I am arguing, as

Derrida--hope effectively to achieve the same result by less direct routes, specifically in Derrida's case the acknowledgement of ambiguity. As Foucault states in the above-quoted exegesis of Canguilhem, however, while the errors in our current description can be pointed out in order to transform our knowledge, this will not generate a more accurate total description--since there is no neutral standpoint from which to judge which knowledge is more accurate --or an asymptotic approach to a final description. Derrida by contrast proposes a form of radical progress by stopping thinking of words in terms of univocal description, and instead approaching them via their inherent metaphorical play, the aim of which is not a more accurate description in a conventional sense, but rather a way of honouring the differences in reality. The discontinuist's preference is to acknowledge knowledge's ineliminable erroneousness at a meta-level.

Derrida's aporetic logic means gesticulating/genuflecting towards complexity rather than proceeding decisively in awareness of it. He opposes the standard continuist aim of having language reach a final, perfect hold on reality, but in doing so gives up attempting to hold onto it per se, in favour one might say of a caress of its surface.


What then of Deleuze? The thrust of his philosophy appears diametrically opposed to that of Derrida's, viz. the deliberate formulation of a metaphysics, whereas Derrida, following Heidegger, makes destroying metaphysics his project. My thesis however is that--rather like in the horseshoe theory of political ideology--the apparently opposite Derridean and Deleuzean extremes in epistemology meet in continuism. Derrida (2001, 192-193) himself indeed claimed on the occasion of Deleuze's death that their two oeuvres shared a 'nearly total affinity', and that 'Deleuze undoubtably still remains, despite so many dissimilarities, the one among all those of my "generation" to whom I have always considered myself closest'.

Deleuze straddles the divide between Foucault's two categories of philosophies of concept and of the subject, perhaps to a greater extent than any other French thinker of twentieth century philosophy: simultaneously a philosopher of experience and of conceptual creation, he recovers the deeply unfashionable project of metaphysics, harking back to pre-Kantian philosophy. As Badiou (2009, 114) says, 'It is not easy to insert Deleuze into the usual genealogies." One might expect me then to place Deleuze in an intermediate position between continuism and discontinuism. However, while a philosophy of experience lends itself to continuism by putting faith in experience as an inerrant contact with reality, there is nothing about a philosophy of the concept that implies seeing the relationship with reality as discontinuous, even if it is prima facie more amenable to such a viewpoint than a philosophy of experience is. If there is a single obvious point of agreement between Deleuze and Derrida, it is the thematisation of difference: it is the key concept of Derrida's early work (albeit in the altered form of "differance"), and first word invoked in the title of Deleuze's first major statement of his own philosophy; both thinkers make difference itself into something like their ultimate ontological principle. To invoke Peter Hallward's (2006, 13) not entirely flattering--but I think accurate--characterization, Deleuze produces a "philosophy of creation" which valorises novelty, hence difference. However, as with Derrida's philosophy of difference, Deleuze's ontology of difference is paradoxically simplistic, ultimately revolving around a simple binary between good deterritorialisation and bad forces of stagnation. Deleuze describes a world where everything should move freely (which is to say that a normativity is built into his metaphysics), and where problems are due to the arresting of flows, the perversion of nature by malign forces (see in particular Deleuze and Guattari 1983 passim). I am tempted to deem this vision 'Manichaean', but this would be inaccurate, because it is precisely not for Deleuze an eternal battle, so much as an historical one that can be won. As with the Heideggerian view of the history of metaphysics, the Fall is viewed as a contingent mistake that can be overcome. This position is thus much more akin to the Christian one because it is somewhat unclear what the origin of evil is, beyond the fact that it has contingently appeared in the world and will in the fullness of time be defeated. Hallward deems Deleuze's thought theophanic, indicating the existence of a God, that the creative force in Deleuze is singular, hence identifiable with God. Of course, this is certainly not how Deleuze himself understands his thought.

Deleuze like Foucault thinks that truth is not universal or transhistorical, but rather evental (Bartlett, Clemens, Roffe 2014, 182). Having declared this, however, he proceeds to produce a metaphysics that tells us how truth comes about which is itself universal and transhistorical. Now, one could argue that discontinuism also does precisely this, but there is an important difference between the very minimal claims made by discontinuism--viz. that language and reality as such exist and that there is a mismatch between them--and the developed metaphysics of Deleuze that pretends to provide an account of underlying reality in language, but somehow wishes to bypass truth in doing this.

How is it possible to bypass truth? Deleuze's conception of truth, not unlike Foucault's, situates truth at the level of imaginary representations of reality. Deleuze thinks we can have a more immediate contact with reality than this, without the mediation of 'truth'. We can see this as a kind of natural outgrowth of Nietzsche's ontology, which reduces truth to the basic 'will to power'. Foucault, though he focuses on Nietzsche's notion of 'will to knowledge' (eschewing any mention of the will to power'), ultimately does not engage in any ontological reduction of knowledge. Foucault thus retreats from the metaphysical aspect of Nietzsche's thought, while Deleuze seeks positively to extend it.

Deleuze (1994, 148) criticises the concept of error as part and parcel of the conception of truth he wants to reject. Specifically, he condemns the determination of 'error' as 'the sole "negative" of thought'. Now, this contradicts Foucauldian discontinuism directly, since this is exactly what it also does, albeit that it conceives 'error' not merely as a 'negative' but also as a constitutive facet of knowledge. While Deleuze's formulation of his critique here seems to allow open the possibility that error can still be a useful concept if the exclusive focus on it is overcome, the thrust of his invective is to oppose the concept of error altogether. For Deleuze, error is the reverse of truth, and he wants to dispose of them both.

For Deleuze (1994, 137) as for Foucault--both of them self-consciously following Nietzsche--thought figures as a kind of violence. It is not for Deleuze however the violence of words on things. Rather, for Deleuze thought figures as the violence of things on words: we are made to think by forces outside of thought. Now, discontinuists do not have to deny that there is such an effect.

Lacan for his part says more or less exactly the same thing as Deleuze about the way the real disturbs knowledge. Foucault does not ask such a question as to the origin of thought as such, but again clearly does think knowledge in large part arises from extra-epistemic forces--indeed, his writings on epistemology certainly imply precisely that, namely that power is a major motive force of knowledge (though also vice versa). What both Lacan and Foucault both posit a specific power of language (that is of 'the symbolic', or 'discourse', or 'knowledge') as such that cannot be reduced. Deleuze is indeed simply inattentive to the problem of language: as Jean-Jacques Lecercle (2002, 2) notes, Deleuze 'resists the linguistic turn' in French philosophy, treating language 'as a secondary phenomenon'.

Deleuze (1994, 154) envisions a productive truth that is not about being adequate to its object. This seeks to eliminate the negativity in discontinuism: where thought has been haunted by negativity, Deleuze would make it positive. Deleuze in his late work with Guattari thus naturalises the violent irruption of language through a reduction of truth to describabale material elements, to events and bodies, in an opposite direction to what Derrida does, which might by contrast be characterised as a denaturalisation of the material through its assimilation to writing. For Deleuze, we create concepts, which attach in various ways to bodies (Deleuze and Guattari 1994, 21). As Jon Roffe puts it, "Deleuze's move here is to resituate truth as an ontological category rather than an epistemological one" (Bartlett, Clemens, Roffe 2014, 187). Truth for Deleuze is relative and historically discontinuous, constructed, but as such truth is simply part of a reality which we in principle can exist in and understand relatively unproblematically: the problem with this reduction of truth to materiality is that it begs the question of what the basis can be for these claims about the being of truth. He criticises traditional conceptions of knowledge, but the question of the mechanism of knowledge is in effect a yawning gap in Deleuze's thought: he expounds a metaphysics without giving it a formal epistemological basis. Deleuze (1994, 132) thinks in terms of 'liberation' from a 'distorting image of truth'. That is, there is an error we have fallen into that we have to escape from. Far from residing in thinking truth too simply, this is actually for Deleuze a matter of thinking truth in too complicated a way. That is, Deleuze identifies truth in the final sense, as both Sartre and Lacan do, with the real itself, but also believes in a kind of unmediated contact with that truth that knowledge can be based around. Deleuze thinks however that by maintaining an openness to the immanent forces that bring about thought we can overcome the dehiscence of thought and nature, by welcoming a thought that is much more like nature in its chaoticness than conventional philosophy. For Deleuze, in his immanentism, we encounter bodies directly and thus know them, and questions of truth are merely a relation between our body and others. That is to say that epistemological continuism is implicit and unchallenged in Deleuze's thought, even if his metaphysics deliberately accommodates a discontinuous history of knowledge. We see here a series of lurking influences, I think, overdetermining Deleuze's epistemological hubris: Bergson's intuitionism about truth; Nietzsche's inheritance from Schopenhauerian monism that allows us direct contact with a reality that, like us, partakes of the element of will of an unbridled confidence that the world is as we experience it; Merleau-Ponty's body-philosophy; and perhaps most deeply Spinoza's monism, which Hallward (12) credits with giving Deleuze the idea of the "total intelligibility of things." (7


One could argue that Deleuze cashes out a metaphysics of forces and bodies that Foucault himself effectively works with, and that Foucault simply hides from the necessity of making these ontological commitments explicit. Deleuze (1988) certainly saw Foucault's project as something that could be ontologised within his framework. I think Foucault's stance is different, however. Pace many readings of Foucault--for example that of Johanna Oksala (2010, 449) who pointedly asserts that it is impossible to avoid ontology to some extent--he posits such things only qua the outside of discourse which disturbs discourse. Foucault's references to bodies and forces and even to power are precisely a reference to the outside of discourse; pace Butler (2002), it is not meant to problematise the concepts he invokes. They are not meant either to put these concepts under erasure nor build a metaphysics out of them.

It is true that ontology is unavoidable iff we understand "ontology" to refer to any kind of use of language to refer to mind-independent reality. However, I take it that this is not what "ontology" means in any accepted usage, but rather that this is just how language itself is generally understood to operate, and that "ontology" occurs only when one sets out systematically describe either being itself (in Heidegger's idiosyncratic but influential definition) or, more commonly, what kinds of beings and relations exist, as Deleuze does. Foucault and Lacan do not go in for any such systematic description of being or beings: Lacan instead gives us at most a theoretical metapsychology, whereas Foucault does something even more ontologically parsimonious, trying to use language to refer to "visible" "things"--statements, forces, bodies--linking them up to "invisible" hypothetical forms like epistemes or strategies of power. (8) This division might be said to invoke some form of ontology, but if it does, it is absolutely minimal. The distinction between words and things at work here is essentially epistemological--for discontinuism, the only ontology that is endorsed is the one that forbids further systematic ontology by asserting that words are split off from things--and while in a sense Foucault relies minimally on what might be described as the folk ontology embedded in our linguistic concepts, he does this only with the utmost caution, stipulating that his formulations are provisional rather than theoretical. This is ontology if, like Oksala (2010, 457), one takes it that "ontology consists of mere analytical grids." At this point any disagreement between us may be merely semantic, but in this case we may say that if Foucault's "analytical grids" constitute an ontology, it is a profoundly anti-realist one. The Deleuze-Foucault difference here can be schematised by designating Deleuze a realist about analytical grids, and Foucault an anti-realist about them, even though their analytical grids are not dissimilar.

One might perhaps argue that since some minimal engagement with ontology is necessary, we ought to engage in it deliberately and maximally, as Deleuze does. I do not agree this conclusion follow, however. One could argue that to allow our ontological bases to remain unexamined means relying on a kind of spontaneous ideology that might be problematic, but Foucault obviates this by engaging in the critique of received opinion, without engaging in a systematic attempt to replace it, even though he is apt to invoke new concepts piecemeal to describe things where he finds the received vocabulary inadequate. Foucault and Lacan do not go the Derridean route of a sceptical deconstruction of all language, and rather are happy to posit all sorts of things, but fight shy of producing a metaphysics, even if Lacan, unlike Foucault, does deliberately propound a theoretical framework, albeit, like Freud's, one that is under constant revision. The one place in his corpus where Foucault (1997) does positively invoke "ontology" as such, viz. in his late essays on Kant's "What is Enlightenment?", the invocation is always of a "critical" or "historical" ontology "of ourselves"--like Lacan, or indeed even Kant in his way, the only domain he is ultimately willing to do the ontology of is the human subject, not external reality.

That said, Foucault does not prescribe abstaining from ontology--he rather only abstains from it on his own account. Discontinuism does not as a position imply that no one should engage in positive ontology, only that we should understand ontological reflection as an historically contingent form of fictive (to invoke indeed Foucault's description of his own works as "fictions") speculation, rather than a matter of finally uncovering the way things really are, which is to say, we should not understand ontology as really doing what it typically understands itself to be doing qua ontology, viz. describing accurately what there is.

We can schematically say that all three of our 'poststructuralists' reject all earlier conceptions of truth. For Derrida, we have understood it as the correspondence of the signifier with the object (a definition incidentally which all three thinkers pointedly reject). From Foucault's NietzscheanCanguilhemian point of view, we have understood truth to mean non-error, when we really need to grasp its erroneousness. For Deleuze, we have wrongly understood thought as adequation to its object, and the very idea of error is bound up with this. The discontinuist case against Deleuze and Derrida is not that adequation of words to things is achievable but that we can't simply sidestep the problem in relation to adequation, inasmuch as the referential function of language is ineluctable.

Western Sydney University



Agamben, Giorgio, Absolute Immanence, Potentialities, Stanford, CA, Stanford University Press, 1999.

Allen, Amy, 'Foucault, Psychoanalysis, and Critique', Angelaki 23:2, 2018.

Alain Badiou. 2001. Ethics. London: Verso.

Alain Badiou. 2009. Pocket Pantheon. London: Verso.

Alain Badiou. 2012. The Adventure of French Philosophy. London: Verso.

Bartlett, Adam, Justin Clemens and Jon Roffe. 2014. Lacan Deleuze Badiou. Edinburgh: Edinburgh University Press.

Giuseppe Bianco (2011) Experience vs. Concept? The Role of Bergson in TwentiethCentury French Philosophy, The European Legacy, 16:7, 855-872.

Butler, Judith. 2002. Bodies and Power, Revisited. Ra.dical Philosophy, no. 114:13-19.

Cavailles, Jean, 'On Logic and the Theory of Science', Phenomenology and the Natural Sciences, eds. Joseph J. Kockelmans and Theodore J. Kisiel, Evanston, IL, Northwestern University Press, 1970.

Deleuze, Gilles. 1988. Foucault. Minnesota: University of Minnesota Press.

Deleuze, Gilles, Difference and Repetition, New York, Columbia University Press, 1994.

Deleuze, Gilles and Felix Guattari. 1983. Anti-Oedipus: Capitalism and Schizophrenia. Minneapolis: University of Minnesota Press.

Deleuze, Gilles and Felix Guattari. 1994. What is Philosophy? New York: Columbia University Press.

Derrida, Jacques. 1982. Margins of Philosophy, Alan Bass (trans.), Chicago: University of Chicago Press.

Derrida, Jacques. 1997. Of Grammatology. Baltimore: Johns Hopkins University Press.

Derrida, Jacques. 2001. The Work of Mourning. Chicago: University of Chicago Press.

Eyers, Tom, 'Lacanian Materialism and the Question of the Real', Cosmos and History 7:1, 2011.

Foucault, Michel. 1971a. L'ordre du discours. Paris: Gallimard.

Foucault, Michel. 1971b. Orders of discourse. Social Science Information. 10: 7.

Foucault, Michel. 1973. The Birth of the Clinic. London: Tavistock.

Foucault, Michel. 1978. Introduction in Georges Canguilhem, The Normal and the Pathological. Dordrecht: D. Reidel.

Foucault, Michel. 1989. The Archaeology of Knowledge. London: Routledge.

Foucault, Michel. 1994a. "Interview avec Michel Foucault." Dits et ecrits I/4. Paris: Gallimard.

Foucault, Michel. 1994b. "La verite et les forms juridiques." Dits et ecrits II/4. Paris. Gallimard.

Foucault Michel. 1997. The Politics of Truth. Los Angeles: Semiotext(e).

Foucault, Michel. 1998. "Nietzsche, Genealogy, History." Aesthetics, Method, and Epistemology. New York: The New Press.

Foucault, Michel. 2002. The Order of Things. London: Routledge.

Foucault, Michel, The History of Madness, Abingdon: Routledge, 2006.

Hallward, Peter. 2006. Out of This World: Deleuze and the Philosophy of Creation. London: Verso.

Lacan, Jacques. 2006. The Subversion of the Subject and the Dialectic of Desire in the Freudian Unconscious. Ecrits. New York: Norton.

Lecercle, JeanJacques. 2002. Basingstoke: Palgrave Macmillan, 2002.

Meillassoux, Quentin. 2008. After Finitude. London: Continuum.

Merleau-Ponty, Maurice. 1968. The Visible and the Invisible. Chicago: Northwestern University Press.

Norris, Christopher, 'Truth in Derrida', A Companion to Derrida, eds. Zeynap Direk and Leonard Lawlor, Oxford, Blackwell, 2014.

Oksala, Johanna. 2010. "Foucault's politicization of ontology." Continental Philosophy Review 43:445-466.

Sartre, Jean-Paul. 1992. Truth and Existence. Chicago: University of Chicago Press.

Shain, Ralph, 'Derrida on Truth', The Philosophical Forum, 49: 193-213, 2018.

(1) One might defer here to Badiou's (2012) rediscription of the history of French philosophy as a division between a philosophy of the concept and a "philosophy of life," since this places Bergson as the origin of the latter camp and hence allows us to place Deleuze there more decisively due to the influence Bergson had on him. The most obvious problem with such a division, however, is that it ends up seeing Georges Canguilhem, whose thought is effectively the lynchpin of the philosophy of the concept for Foucault, being placed now in the other camp, because of his vitalism.

(2) Foucault 1971, 22.

(3) Lacan was fully 29 years older than Derrida--but due to the fact Lacan was a practising psychiatrist rather than an academic, his entry into theory didn't begin till he was in his 30s, and he published very little, with his first book (the anthology Ecrits) appearing only in 1966, contemporaneously with Derrida's earliest work.

(4) Lacan does not like to use the word 'reality' in this way, because for him this word refers to our imaginary fantasy of what the world is like.

(5) I am grateful to Alexei Procyshyn for this objection. Procyshyn suggested that this is Deleuze and Guattari's position, one shared by Walter Benjamin.

(6) My reading of Derrida here is offered pace that of Christopher Norris (2014) in particular, who reads Derrida as an epistemological realist with a relatively conventional conception of truth, but I take it that this is an extreme outlier in terms of readings of Derrida on this question, and indeed is a reading that lacks any real support to be found in Derrida's explicit claims (see Shain 2018).

(7) While I think Hallward is right here, his quotations from Deleuze are quite misleading--this phrase is cited as if from Deleuze's own mouth, but it's actually from Martial Gueroult's Spinoza, from an article about Gueroult's interpretation of Spinoza. One comes here across the general problem of trying to discern what Deleuze thinks because so much of what he says is ventriloquised through other thinkers.

(8) This distinction is made by Foucault (xi) in the preface to his Birth of the Clinic.
COPYRIGHT 2019 Ashton and Rafferty
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2019 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Kelly, Mark G.E.
Publication:Cosmos and History: The Journal of Natural and Social Philosophy
Geographic Code:4EUUK
Date:Jan 1, 2019

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters