CAIRN-INT.INFO : International Edition
To Bernard Bourgeois

1 If we can agree on anything with regard to the notion of skepticism, we can at least agree on the polysemy of the term, or the indeterminacy of the concept. Ultimately, what does it mean to be a skeptic, what makes one a skeptic? In order to determine a unified meaning of skepticism, must we appeal to Sextus Empiricus, traditionally considered as the emblematic thinker of antique skepticism? But, as J. Brunschwig warns us, we will not find the word “skepticism” in his works—only the word “skepsis”. [1] Then should we seek the source earlier still, in the Pyrrhonian school? But wasn’t the latter divided into many tendencies without any organic link between them? From Carneades, who seems to profess a moderate probabilism, to Aenesidemus, who refuses every notion of truth, sensible, intelligible or mixed; from the practical skepticism that seeks to be a way of life similar to oriental spiritual exercises, to the ‘urban’ skepticism that boasts that it assents only to ordinary beliefs, we have no choice but to recognize that we find very little to conceptually unify the term. [2] Should we, then, turn toward modern skepticism which, since Descartes, has been distinguished from the ancient model in so far as it doubts, electively, our capacity to know alterity (the things of the world, or other minds)? But in fact, who, in this modern epoch, is a “skeptic”—that is to say, who consistently presents themselves as such? Hume, who declares that no-one can be a skeptic “sincerely and constantly,” who associated this stance with a “malady,” [3] and who appealed to Newton to construct a “mental geography” [4]—a model, finally clear and rigorous, for investigation into human nature? Bayle, who, while certainly defying all metaphysical essences, nevertheless sought a “scientia apparentalis” and concluded his celebrated arc as a philosopher—a fall from the heights of dogmatism into the abyss of skepticism—with a profession of faith that had an undeniably Pascalian accent: “Several others would be reduced to the like extremities, if they had not recourse to an assistant superior to reason”? [5] Nietzsche, for whom there are no facts, only interpretations, but who, just like Hume, associates skepticism with a wasting disease? Rorty, who, in spite of the claims of his readers (Putnam in particular), ceaselessly rejected this designation, just as he would refuse the “relativism” that, nonetheless, was implied by his positions? Kripke, whose “skeptical” interpretation of the very famous problem of the application of rules is often mentioned—an interpretation based on a reading of Wittgenstein, who himself, however, declared skepticism to be impossible and meaningless since it was not “sayable”. In short, if we conducted a thought experiment in which all philosophers were calmly gathered in a circle, and then we dared to ask them: “Is there a skeptic in the room?,” then, we believe, very few hands would be raised. [6] Doubtless the reader will protest here. Have we not seen, in the last thirty years, a “recrudescence,” a true “revival” [7] of skepticism, since B. Stroud demonstrated that “no one has been able to find the proof that has been sought, throughout the modern epoch, for the reality of the external world”? [8] Further still, those in the analytic world who refuse this radical skepticism nonetheless adopt certain aspects of classical skepticism, thus joining the ranks of what have been called the “new Humeans”. [9] In which case, doesn’t a good part of today’s philosophy, in one version or another, claim some form of skepticism? What will ultimately be at stake in our analysis will be to interrogate the meaning of this contemporary skeptical “revival,” but we shall do so by putting it into perspective, viewing it according to an unusual vanishing point: the point of view of those rare philosophers who, after Hume, and unlike him, did declare themselves to be “constant” skeptics. Let us therefore return to our circle of departed philosophers and dare to respond: Since Descartes, and before the famous analytic “revival,” only two philosophers have openly claimed the term of skepticism without denying, diminishing, or rejecting it at some other point in their work, or tempering it through a fideist recourse to some instance independent of reason. [10] These two philosophers are contemporaries: Schulze (who called himself Aenesidemus) and Maimon. What can they tell us about this “skepticism” that they assumed in a constant manner? And, once the true nature of their skepticism is elucidated, can it be fruitfully set against the new skepticism?

I. From Referential Skepticism to Self-Referential Skepticism

1. The Skeptical Context During the Years 1760–1790 in Germany: Referential Skepticism as Existential Solipsism and Epistemological Solipsism

2 To pinpoint precisely the nature of these two authors’ skepticism, we must first of all situate it within the theoretical context that was theirs: the Enlightenment Germany whose source is Thomasius, which culminates in the 1770s, and continues up until Kant. Recall firstly that it is traditional to oppose antique skepticism to modern skepticism by showing that the placing in doubt of the existence of the external world does not make much sense for antique skepticism, [11] which does not define knowledge as a relation between two distinct entities, the subject and the object. The question “how can our representations (the internal site of the subject) be related to an other referent (the external site of the object)?” only comes into focus within an epistemological model of a generally Cartesian type. Skepticism so defined is inscribed in this apparatus of representation, of which it is one possible consequence. This is the skepticism that Kant attacks in his Critique. This skepticism is a solipsism, which can take one of two forms: an existential solipsism that doubts whether we correctly perceive the material world or other people, and imagines a world that is unstable to such a point of variability, change, uncertainty and fluctuation, with neither reference points nor regularities, that madness lurks in wait for anyone who experiences it: will I not “see this stone vanish into air before me and become invisible”? Or again: this woman I love, and who thus gives me life, is she nothing but a mechanical puppet? [12] This type of solipsism, or existential skepticism, was not really at issue in the era that interests us here, and therefore did not give rise to the subtle counterarguments that seek to overcome it. The second figure is that of epistemological solipsism, which simply doubts the agreement between our rational principles (the principles of noncontradiction, of sufficient reason, of cause and effect, of lawful connection) and the course of nature (understood here scientifically as a set of regularities, and not in the ordinary everyday way as immediately perceptible sensible qualities). We say that the earth moves even if we do not perceive its movement; that a solar eclipse will soon take place in the Inca lands, even if this goes against their common beliefs; but how can we assure ourselves of such certainty? It is this form of skepticism that finds defenders (supporters of Hume, numerous in Germany at the time), involuntary propagators (the eclectics, advocates of “common sense”) and opponents (Kant, and Eberhard and the Leibnizians of the time).

3 This skepticism, as an epistemological solipsism, can be clearly defined as bearing upon the capacity of our statements to veritably refer to an external world, itself determined as the world treated by physics. It is what we might call a referential skepticism, or perhaps a semantic skepticism. This referential skepticism constituted the truly dominant current in Germany between 1750 and 1780 [13] and hence existed in multiple variations: from the radical skepticism that declares all our judgments in physics to be uncertain and even calls into question the status in re of a principle as general as that of noncontradiction, to a very watered-down version that we might just as well call “moderate skepticism” as “prudent rationalism” or “open positivism,” which, very simply, defends probabilism in physics, denies the interest (or even the rational possibility) of questioning the principle of noncontradiction, and pursues the progressive construction of a science of physical (but at the same time social) reality that is falsifiable and always a work in progress.

4 Now, Aenesidemus and Maimon were not even participants in this dominant skeptical current. If we have been able to distinguish two skepticisms, one antique and one modern, then we must admit a third form that we might very well call “contemporary” (respecting the customary periodization in history of philosophy), but which, more precisely, is distinct from modern skepticism in so far as it is no longer a referential skepticism, or an epistemological solipsism, but intends to definitively exit from the framework of a philosophy that conceives representation (the judgment or statement) as a schema or image standing between the external thing (or external referent) and the representing (judging or speaking) subject.

2. Schulze’s Deconstructive Skepticism, or the Strategy of Logical Incoherence

5 Thus, the Aenesidemus text [14] does not demonstrate, against Kant, that we cannot know external things, nor even that our concept of causality is not applicable to them; rather, it refuses to participate in the debate at all. [15] In fact, Schulze is not interested in the question of proof (or lack of proof) of the reality of the external world. His aim is instead to demonstrate the inanity of every theory of knowledge, whether it defines itself as empiricist, realist, rationalist, Leibnizian or Kantian. As far as theories of knowledge are concerned, anything goes—from the most trenchant criticism to its exact contrary. Schulze “shows” this not by demonstrating it positively and discursively, but through a subtle strategy of assumed incoherence. The skepticism of Aenesidemus becomes evident through the way in which it is expressed, rather than through what it says; it is defined on the basis of its argumentative strategy more than by its specific theses, which are, as per the reproaches of numerous later commentators, frankly incoherent. Thus, if one reconstructs the whole of the argument, Schulze’s strategy in this book against Kant is not at all to affirm the impossibility of proving the reality of the material world, nor even to focus on this question, but to deploy a series of multiple alternatives which he intends to show, according to his own terms, are neither “more absurd” nor “more unthinkable” than the Kantian theses. The purpose of the argument (which only the overall thrust of the whole text allows one to reconstruct) consists, then, in suggesting that such and such an hypothesis—that the author proposes without ever worrying about its coherence with other statements made a few lines later on—can claim validity on the same basis as the Kantian thesis. These hypotheses are, Aenesidemus tells us, “just as valid”. But “just as valid” means (if we reconstruct the chain of substitutions in the text) just as “arbitrary” and “indemonstrable”. [16] The entire construction of the text thus aims to reduce Kantianism to one opinion among others, or, in the expression of Aenesidemus itself, to reduce it to “a tissue of arbitrary opinions”.

6 This approach indicates to us that Schulze’s aim is to put into question the Kantian discourse’s claim to validity and, further, that of every philosophical discourse in general. This is why Schulze is obviously not bothered about being coherent, even less about proposing a consistent conception of knowledge, since, to his eyes, there is no point in trying to determine why and how science works. Hence incoherence is the only consequent way to express the intended purpose: the deliberate failure of every general statement, of every science, of every theory of knowledge and, beyond them, of all philosophical discourse. The general coherence of Aenesidemus is thus not given by the way its specific theses fit together, but by what they aim at through their very incoherence: challenging the claim to validity of Kant’s discourse.

7 This challenge to the claim to validity of Kantian discourse was to take a particular form which we should pause and examine, since it sums up the difference with the earlier form of skepticism, and allows us to understand how Schulze can call himself a “sincere and constant” skeptic. Aenesidemus claims that Kant’s philosophy, in the last instance, is self-refuting, without there being any need to oppose another, external thesis to it. For, as the text explains, Kant can only justify the truth of his own theses by presupposing principles that these very theses aim to recuse. So his notion of “validity” is contradictory. This is summed up in the alternative to which the end of Aenesidemus tries to consign the critical philosophy:


Either the Critique is true and, for a cognition to be true, there must be an intuition and a concept. Now, the enumeration of the conditions of possibility of knowledge do not satisfy this criterion of truth, since these conditions are not representable through intuitions and concepts. Therefore, if what the Critique of Pure Reason says is true, then the Critique of Pure Reason is false; or, there exists a world of truth apart from the linkage between a concept and an intuition. In this case, Kant is wrong to say that the only truth is that of mathematical and physical judgments. In which case, here again, the Critique of Pure Reason is false.

9 This alternative, expressed in multiple forms in Aenesidemus, was to be masterfully summed up by Schulze in his other work, Critique of Theoretical Reason:


So, if what the Critique claims to know of the foundations of experience is real knowledge, the affirmation of the same Critique, according to which all true knowledge of our mind is exclusively limited to objects of experience, is absolutely false. If, on the contrary, this latter affirmation must be true, all knowledge of the sources of experience must be taken as a whole to be an empty appearance. [17]

11 This type of challenge to Kant’s philosophy shows how much skepticism changed its meaning after Kant. Skepticism before the Critique bore upon the link between knowledge and being, or on the relation between subject and object. What Hume challenged was the passage from the causal relation in our mind to its objectivity in things. Now, after Kant, this question of the subject-object relation disappears from the skeptical refutations or readings (and the remaining idealistic ones) of the Critique. Post-Kantian skepticism does not reproach Kant stricto sensu with not having been able to prove the reality of material things, but abandons the classic theme of representation as a relation between two independent entities. Hence the “modern” skeptical stance (as a solipsist position or a referential skepticism) loses its leading role, with the emergence of a question of an entirely different nature. For the main question of Aenesidemus concerns Kant’s relation (and by extension, that of philosophy in general) to his own propositions. What is the status of the discourse of philosophy? What are the modes of knowledge (discourse) that it must bring into play in order to explain what it explains (discourse), and claim that what it says is more true or more likely than its opposite (criticism rather than skepticism, idealism rather than realism, finitude rather than its opposite, etc.)? Thus, to advocate skepticism, in the case of Schulze, is precisely not a matter of affirming that the reality of the material world cannot be “proved” nor that it is illusory, but instead means: to revoke the possibility of any philosophical discourse, and correlatively, to declare vain the pursuit of any “knowledge” whatsoever, be it physical, social, philosophical or other.

12 Here we attain a skepticism more radical than that of Hume, who, in affirming the possibility of the knowledge of human nature, could not advocate a total skepticism, and mitigates it with its opposite: confidence in a science of nature. Now, Schulze’s skepticism cannot be moderated by a position with a more positivist or naturalist allure, since it bears upon the very possibility of any enunciation with a philosophical or general aim. In so doing, it attacks the very root of the theoretical stance: the claim to truth (in Kantian discourse as much as in Newtonian physics, in Euclid’s mathematics [18] and in the possibility of a knowledge of man, of the social, of history) and refuses to “utter,” “say” or “demonstrate” anything whatsoever, other than through a succession of motley and incompatible theses, whose very incoherence is a sign of their inanity.

13 This skepticism is no longer a referential skepticism or a semantic skepticism but can be characterised, on one hand, as a self-referential skepticism (or pragmatic skepticism) [19] in so far as it bears upon the enunciation of philosophy; and on the other hand, as a “deconstructive” or “destructive” skepticism, in the sense that it incontestably needs some other philosophy, like Kant’s, to deconstruct. Negative skepticism can only survive on this deconstruction of philosophical discourse, as is attested by Schulze’s writings, entirely dedicated to refutations of Kant, Reinhold and classical metaphysics. This skeptic can never affirm anything other than “anything goes”; and he shows this by deconstructing, step by step, every edifice that claims to say otherwise. Which explains why Schulze opposes to the Critique of Pure Reason his Critique of Theoretical Reason, a critique of the “Greek” ethos (which cannot conceive of a good and fully human life without knowledge or truth-seeking), and ultimately one that can find coherence only through the abandonment of philosophy and science, in favour of an anti-intellectual or anti-theoretical stance which—as Hume had already envisaged, in the placid guise of a country gentleman—the pleasures of hunting and the joys of an “earthy” ordinary life suffice to fill. [20] Skepticism, thus strictly defined, if held constantly and sincerely, would thus consist simply in living so as to live out the distinction that Hume proposed between “living” and “philosophizing,” and taking it to breaking point. So, accordingly, this skepticism would have to finish in silence, with our gentleman going butterfly-hunting without a further care for deconstructing the presumably exorbitant claims of discourses that seek to be other than purely negative, claiming the truth of what they say and the possibility of proving it and sharing it with others.

14 With this first type of post-Kantian skepticism circumscribed, we should now go on to compare it with Maimon’s skepticism, which will be seen to be very close to it, in so far as it is also a self-referential skepticism that bears less upon the content of particular statements than on the status of enunciation itself (I.3); but at the same time is very distant from it, even frankly opposed to it, to the point of becoming its opposite, in so far as it seeks to be a skepticism that is constructive, rationalist, and even (as Maimon unhesitatingly declares) “transcendental” and “metaphysical” (II).

3. Maimon’s Reflexive Skepticism

15 Like Aenesidemus, and, strictly speaking, before him, [21] Maimon would begin to reflect upon the question: What procedures does Kant mobilize in order to write the Critique of Pure Reason? In the Critique, Kant claims to trace out the functioning of our knowledge of the object, but says nothing regarding the procedures by which he accedes to the knowledge of this functioning. Still, like Aenesidemus, Maimon’s “skeptical” strategy does not consist in refuting this or that thesis of the Critique; rather, it aims to think through the very possibility of the enunciation of these theses. Maimon would pose the question in the following terms: Kant divides the power of judgment into determinative judgments (mathematical judgments and, in part, those of physics, which operate through subsuming an intuition into a concept) and reflexive judgments (which set out from a singular given to move toward an absent universal, and which, as Kant will show more explicitly in the Critique of Judgment, bear upon the beautiful object or the organized object of nature). Determinant judgments are valid, whereas reflective judgments must be prefixed with the formula: “it is as if…”. Now, Maimon asks, to which type of judgment (determinant or reflexive) do Kant’s own philosophical enunciations belong? To answer this question, Maimon intends, in the Essay, to reconstruct the acts implicitly undertaken in the writing of the Critique. He will show that the text is written on the basis of concepts of reflection. If we hold to what Kant says in the chapter “The Amphiboly of the Concepts of Reflection,” this claim seems obvious; and yet Maimon was the first, and remains one of the only thinkers to have shown that transcendental reflection is the faculty that allows the Critique to be written at all. [22] Reflection is not only that which allows us to order the contingent (the reflective judgment, governed by an “as if,” that is at work in the appreciation of the beautiful and the living); it becomes the proper faculty of philosophy, in so far as the latter says something about representation, judgment, the faculties, language, affect, good, evil, etc. Inscribed in the Critique as its central kernel, the very condition of possibility of its enunciation, reflection, according to Maimon, gives us the essence of criticism. This reading, a literally correct reading, of Kantianism plunges it into skepticism. [23] For if every reflexive judgment must be prefixed, as Kant would have it, with an “as if,” and if the statements of the Critique are reflexive judgments, then we must in all rigor write: “it is as if the object turns around the subject (the Copernican revolution); it is as if the understanding and sensibility were the sources of (analytic and aesthetic) knowledge; it is as if the objects of metaphysics were (dialectical) illusions”; and finally—to the tune of a nursery rhyme—“it is as if what the Critique says were true”! The status of Kant’s discourse is, in fine, comparable to the status of our statements about the beautiful. Consequently, all of our judgments become, defined rigorously in Kantian terms, aesthetic judgments. This is why an accomplished and completed criticism is an authentic skepticism.

16 We see here how close Maimon’s skepticism is to that of Aenesidemus, which also highlights the philosopher’s enunciation. In this sense, we can here again speak of a self-referential skepticism, a pragmatic skepticism (which interrogates enunciation), or a reflexive skepticism, in contrast to a referential skepticism or solipsism. It is a question of asking not about representation or reference, but about what Bourdieu, not so long ago, did not hesitate to call “the science of science” or “reflexivity.” [24] It is this problem of reflexivity that Fichte and Hegel alike would fully confront—the problem that sets German idealism in motion and explains the unprecedented way of philosophizing that is deployed within it. Nevertheless, Maimon is, at the same time, antipodal to Schulze’s skepticism, as we can see not only in his critique of Schulze in his Letters from Philalethe to Aenesidemus, but in the spectacular turnaround that he will make in the understanding of the skeptical stance. For Maimon, far from deconstructing out of sight all existing statements, far from applying himself to a condemnation of philosophical activity, will make himself one of the greatest apologists for “theoretical” reason. [25] It is this reversal of deconstructive skepticism into a constructive skepticism that we must now try to understand.

II. Maimon and the Reversibility of the Term “Skepticism”: Unlimited Reason

1. The “Coalition of Systems” or the Strategy of Topological Coherence

17 Not only does Maimon not intend to destroy the Kantian system; he claims to go further still and to “coalesce” the entire set of philosophical systems in order to produce, in his turn, a new, authentically rationalist, system—even, he says straight-facedly, a “dogmatic” one. [26] What does this mean? That, like Hume, he recuses the word “skeptic” at certain moments in his work? That, like Bayle and others, he seeks, outside of reason, in a different instance (faith, common sense or everyday “life”) the “dogmatic” antidote to his skeptical “melancholy”? Not in the least. Maimon wants to be at once dogmatic and skeptical, empiricist and rationalist, realist and idealist, joyful and a philosopher. Let us explain this apparent extravagance starting from a precise example: the manner in which, according to his recurrent expression, he “coalesces” Kant and Hume, thus seeming, like Hegel, to integrate irreconcilables and thus surpass (aufheben) them.

18 In his Essay on Transcendental Philosophy, Maimon puts forward, on the question of causality (at the time a subject of debate between supporters of Hume and Kant) the following proof: Even admitting that every phenomenon must a priori have a cause, this cause could be anything whatsoever. Now, the physical sciences must exhibit the connection between specific events. Physics is involved with Wirklichkeit (effective experience), which Maimon distinguishes carefully from Realität. By virtue of this, to claim that Kant’s arguments effectively founded physics is to confuse possible

19 experience with effective experience. The scission between the universal and the particular is thus not, Maimon tells us, overcome by the Critique of Pure Reason. This is why Hume is right, and we must, on this point, be “skeptical empiricists,” by declaring that the statements of physics are at best probable and, in any case, endlessly revisable. Nevertheless, the Kantian analysis may be valid at another level, that of possible experience. Accordingly, for Maimon, Kant and Hume are both right, but at different levels of the construction of human knowledge. The Kantian subsumption, operative at a certain level, is not operative at a “lower” level, that which we address with Wirklichkeit. Accordingly, he undeniably rehabilitates the nodal concept of the Humean theory of knowledge: comparison. For all this, the Essay does not recuse the process of subsumption, but relativizes it, by assigning it a place, a site, a sphere of application. Subsumption does not exhaust the power of judgment—it is but one facet of it, one aspect, one of many possible variations.

20 These philosophies can coexist in the same space, that of reason, whose field of pertinence henceforth will be enlarged, will dilate, will grow ceaselessly, living on the different systems that are the possible crystallizations of it. Thus, subsumption may be valid for certain mathematical judgments—Maimon will show this, for example, in his texts dedicated to Euclid’s geometry. On the other hand, he rules it out for arithmetic, where he takes up Leibnizian logicism instead. And then, this Leibnizian logicism and this Kantian subsumption are not at all incompatible with the acceptance at another level of certain of Hume’s analyses. The latter, in this sense, does not so much contradict Kant as complete him and, by the same token, strengthen him. Maimon’s method takes up what we have elsewhere called a “spatializing approach” to the opposition of systems. [27] Maimon does not reject one of the theses that are present (according to the classical logic of the excluded middle); he does not deem them doubtful either (in this sense, his skepticism is not one that consists in recusing everything); neither does he resolve contradictions by reference to a future that would integrate them all (Hegel); in a word, he rejects nothing, he renounces nothing, nor does he temporalize; he makes systems into neighbors, coexistent and cohabiting, in one and the same widened space, which can only dilate infinitely in a ceaselessly resumed process of expansion. If, like the illuminators of medieval manuscripts, we were to arrange on one page of a book this “coalition” realised by Maimon, Kant’s text might, for example, appear in the middle of the page; that of Hume lower, like a necessary level demanded by the preceding text. Leibniz’s text would figure higher up than the level where Kant’s intervention was placed, at the level of algebra and of the more abstract operations of reason. This neighboring relation is without exclusion: none of them are false; and without inclusion: the three are not at the same level, and are therefore different, but “demand each other”. This neighboring is made possible by the existence of a common, infinitely expandable space, that of philosophy in action, which continually enlarges its field, integrating the analyses proposed by each thinker by situating them alongside one another, stepped, graduated, layered, like so many stratifications of one and the same rock. Hume’s analyses, like Kant’s and Leibniz’s as much as Aristotle’s, are so many aspects, pieces, fragments of one and the same logos that distends, “elasticizes,” with the production of each new philosophy.

21 Very well, the reader will say, but this does not at all resolve the initial problem, which is also no doubt the most important, since it is the most abstract: That of the status of philosophical discourse. Because, ultimately, if in the last instance all of our philosophical judgments are preceded by an “as if,” doesn’t this activity of filling the infinite page of reason become a coloring-book, a merely aesthetic assemblage that renders all systems compossible, in an aimless hermeneutic labour? Doesn’t topological coherence simply make for a co-errance of all philosophers, who are assuredly finally reunited in the same space, but a space of exile, far from the land promised by philosophy itself: Truth? [28] Is philosophy really nothing but a hermeneutical practice where, after Maimon, we must integrate, in a vast book in the form of a puzzle, or like Borges’s library (a common space, according to the topological hypothesis), Nietzsche, Husserl, Derrida, Henry, Rorty, Levinas, Wittgenstein, etc., with no other guiding light than the “as if” of aesthetic judgment and the pleasure that is attached to it? [29]

2. The Copernican Vortex

22 To respond to this question, we must firstly delve into the aporia. Maimon, undeniably, truly blows to smithereens the “sure and fixed point” much sought after by philosophers. To show this, let us take the example of his comparison (“coalition”) of Aristotle and Kant. Aristotle’s categories, he tells us, are enunciated from the point of view of the object. Aristotle wants to say what being is and to run through the different predicates that express it. Kant’s categories, on the other hand, are expressed from the point of view of the subject. The Copernican revolution consists in a transfer from the subject to the object, a transfer of “being” to the “I”; the “fixed point” is thus displaced from one to the other. Now, Maimon tells us, there is no more legitimacy in one of these fixed points than in the other (“as if”). Just as, to think of the movement of a body, we possess no fixed point other than that which we initially and arbitrarily determine—so to set out from the subject or the object is a matter of initial choice. It is on the basis of this optic that the metaphor of the Copernican revolution is reread. As M. Gueroult has already discussed in his commentary on Maimon, “in this phenomenon of relation between the subject and the object of knowledge, we are not obliged to attribute the movement to one or the other.” [30] The positing of an absolute (or the absolute of positing) on the basis of which one may then deduce the whole, is arbitrary. This is why philosophical statements (such as that of the Copernican revolution) are as relative as those of astronomy, where the fixed point chosen depends upon the movement that one attributes to other bodies, according to a conception of astronomical revolution maintained by Ossiander, in his preface to On the Revolution of Stars, and, moreover, by Descartes himself, who affirms that Ptolemy’s geocentrism is as correct as Copernicus’s heliocentrism, since the movement and rest of a body depend on the reference point one initially decides upon. [31] We always think on the basis of a fixed point, but any reference point will do, and there is no way to assert an absolute position. This is why neither criticism, which begins with the subject, nor Aristotelianism, which begins with the object, nor a philosophy that decides to find another name for a fixed point (God, history, society, language, drive, and other things of the same ilk) can be anything but reconstructions placed under the sign of “as if”. It would be difficult, it seems, to be any more skeptical in the straightforward sense of the word, since the Copernican revolution, thus conceived, ends in a veritable vortex of doctrines and foundations.

3. Co-errance and the Infinite

23 Now, it is on this precise point that skepticism can reverse into its apparent opposite. For in order to affirm this vortex, reason must have been capable of reflecting on itself, otherwise it could not say what it said. Here we need to understand that the term “reason” is not at all “substantialized,” since it is defined as the very discourse that, as a philosopher, I maintain when I say, for example, that physics is only probable, thus defining myself as a skeptic. Reason is rigorously understood as the very thing I am doing when I write books of philosophy. Now, says Maimon, we should concentrate on this act or capacity of reason to “reflect on itself” (to say itself, to auto-explicate itself), rather than on an external point, declared to be more firm and assured than others (God, nature, history, language, drive) and consequently proclaimed to be the yardstick by which everything else will be judged and organized. For to say what we say: “There is no fixed point,” “Our philosophical judgments are placed under the sign of ‘as if’,” is to say that our reason is finite. Through our claimed skepticism, we announce the finitude of reason. But this finitude of reason, on what basis do we declare it as such? Maimon, a little brutally to the ears of a standard skeptic, will respond: in comparison to its contrary—that is, in current language: on the basis of the infinite. This infinite on the basis of which we reflect on reason and declare it finite is not some external instance, a being that exists outside of our reason. As such, this infinite does indeed, in classical language, have the status of a “fiction,” the term Maimon uses to characterize it. Only the positing of this ideal fiction—infinite understanding, unscathed by any scission or finitude—authorizes us to reflect upon ourselves as we have just done—namely, as finite beings, possessing no fixed and assured point as we make judgments under the sign of the “as if”. Without this fiction, immanent to a reason that reflects and thus is reflected, these propositions themselves would be literally impossible (non-sayable); without it, no proposition, beginning with that of the skeptic, could claim to say what philosophy can or cannot claim, nor to determine the limits of sense and nonsense, nor even that man is finite (unless to say that man is so finite that he cannot discover any point of view on whose basis the validity of the proposition “man is finite” can be claimed!). This is, intrinsically, what is attested to by any philosophical enterprise that tries to unveil the limits of reason or knowledge, beginning with the Critique of Pure Reason itself. Certainly, today’s commentators on Kant, following Heidegger, have tended to accentuate the thesis of a radical and structural finitude through which Kant definitively dissociated himself from classical metaphysics, which, for its part, thought of finitude as a simple lack in relation to an actual infinite. It remains the case that, even if finitude in Kant is thematized only as a difference of degree in relation to an existing infinity, it is nevertheless—at the level of the writing of the Critique itself, and the possibility of its writing—posited on the basis of its contrary. For what gives Kant the right to say, in the transcendental aesthetic, that we are finite? By what right can he interpret sensible intuition as the mark of finitude? In relation to what does he posit the very notion of limits, the real subject of his text? [32] By analogy to the intuitus archetypus. This intuitus archetypus is determined by Maimon neither as being an external being or referent, nor as an external aim, but as the condition internal to reason, on whose basis reason can declare itself to be such and such (for example, finite and limited, in the critico-skeptical guise it takes on in Kant). This condition is a “fiction” (if one thinks according to the framework of external reference), but a fiction necessary to the exercise of thought itself; this fiction is thus a transcendental condition of possibility; without it we would not be able to say what we say, for example, of the finitude of man. This condition is generated in reason itself, by its effective implementation, its putting into practice or into action—that is: from the very fact of its own exercise. Every type of writing that claims to say what reason, language, history, man (etc.) is or is not, to decide on their limits, their legitimate and illegitimate usage, every philosophy that defines us as finitude, or again, every statement that tries to dissociate good from evil, the acceptable from the unacceptable, or which outlines some way of behaving more desirable than others, or which very simply determines what philosophy is and what it must not be, can only do so by analogy with what an infinite understanding or a perfect evidence (intuitus archetypus) would be. It is this internal infinite that organizes, orients and vectorizes the very activity of reason and, from the start, brings it to life.

24 As we see here, this reversibility is total, in so far as skepticism reverses itself, turns inside out like a glove, according to a metaphor of reversibility dear to Merleau-Ponty, in order to become the obverse of its apparent contrary. Consequently, the conjoint utilization of normally anthithetical conditions such as “skeptic” and “dogmatic” is justified: to be authentically and constantly skeptical, one must posit an infinity (one that is “dogmatic” in the sense that it is an insuperable and necessary condition of possibility), but to be “dogmatic” (to posit an infinity), one must be skeptical (enunciating finitude in a consequent manner). The positions are reversible and one cannot go without the other. It is because the author did not see this reversibility of positions that Aenesidemus, in spite of itself, imprisoned itself in the refusal of philosophy, the denunciation of reason and the appeal to its ultimate consequence: Hume’s country gent. Now, this deconstructive skepticism cannot claim to critique Maimon’s proposition, as Maimon demonstrates in his Letters from Philalethe to Aenesidemus—whereas Maimon can critique the position of the deconstructing philosopher. For anyone who, like Aenesidemus, renounces philosophy, cannot claim to make a counterargument against any philosophical position whatsoever, nor to emit any discourse or judgment of the type “truth does not exist,” “man is finite,” “we must live in such and such a way,” nor even “we should no longer do philosophy, which is a pernicious activity,” etc. He who goes hunting butterflies goes hunting butterflies (a tautology). The philosopher, who, inversely, accepts, takes in charge, assumes his claim to truth, cannot, according to Aristotle’s terms, enter into discussion with he who can no longer say anything in this field; he can therefore do no more than (in Fichte’s words) wish the newly converted countryman-hunter “a peaceable circulation of humours”.

25 It is permissible, at the end of this reconstitution of post-Kantian skepticism—once it is assumed—to propose a typology capable of reducing the initial polysemy of the term. We first encountered a referential skepticism, which divided into existential and epistemological solipsism (Hume’s supporters before Kant), and a self-referential skepticism which spoke either as a deconstructive skepticism (Schulze) or as a constructive skepticism (Maimon). Can this grasp of the skeptical field, this reconstruction of its perspectival depth, help us to understand the “revival of skepticism,” the ultimate aim of our analysis?

III. The “New Humeans” or the Two Contemporary Forms of “Skepticism” Compared with Self-Referential Skepticism

1. The Meaning of Actual Skepticism as Existential Solipsism

26 It is to B. Stroud and to his proof that “skepticism is unassailable,” along with the retinue of refutations that try to respond to it, that we can trace the current outbreak of skepticism in the analytic world. For Stroud, as we have said, skepticism consists in proving that one cannot exhibit any proofs of the reality of the external world. [33] In relation to our preceding analyses, it is easy to characterize this skepticism as a referential skepticism. Stroud supposes the dissociation between a subjective point of view (our mind which thinks that the sun will rise tomorrow) and a point of view that would be objective (the world as independent of us but which, in T. Nagel’s words, we can see “nowhere”). [34] This referential skepticism is, consequently, inscribed within a “representationalist” framework (in conformity with the meaning that Brandom gives to this generic term) [35] and which connects back to a problematic with a certain solipsist flavor. It is thus less a revival of skepticism than a resurgence of a very particular type of skepticism: solipsism. It even presents itself often as more of an existential than an epistemological skepticism, if we judge by the examples chosen both to support and to refute it. These examples, following Cartesian doubt, often have a theatrical flavor (along the lines of “the world is a dream [or a vat] told by an idiot”), and which are descended more from the poetic staging or the experience of a certain form of madness (“all of my representations are hallucinations”). This type of doubt, of course, has its legitimacy and its importance, in so far as it accepts, for example, the posing of the question of madness and makes it a philosophical experience; which may, at least, cathartically, through representation and distancing, cure us of its actual experience. Nevertheless, it raises a series of problems that our initial perspective allows us to make visible.

27 The first is as follows: Stroud tells us that no classical philosopher (Descartes, Kant, Carnap or Russell) was able to respond to skepticism, because “none of them were able to find the proof that was sought since the modern epoch of the reality of the external world”. [36] But who, in classical philosophy, said the inverse, namely: “There is a proof of the reality of the external world”? Descartes, of course; although we must note, in all rigor, that the sixth Meditation, whose theme is precisely this, limits itself prudently to the consideration of a “great propensity” (propensio magna) to believe in the reality of the sensible world and thus appeals to my natural tendency. Kant, who says rigorously that we can never know things as they are? The self-declared idealists such as Fichte and Hegel, who ceaselessly denounce the realist residue of the Kantian doctrine of the “thing in itself”? Russell, who declares that our belief in the existence of the external world is not of the order of a proof, but that we cannot disabuse ourselves of this belief, for if we could, we would not be able to live well, or indeed at all? Quite evidently, few philosophers inquired as to “the much sought-after proof”. If skepticism as defined by Stroud consists in saying that one cannot exhibit any proof of the reality of the external world, then Kant is skeptical, as are Fichte, Russell and Husserl. But then what can the phrase “revival of skepticism” mean if, from a certain point of view, everyone is a skeptic? [37]

28 What is more, even if we do not turn toward this existential solipsism, but instead toward an epistemological skepticism with regard to physics, it would be just as easy to show that judgments bearing on physics in Kant are so-called regulative judgments, and that, in this sense, they relate merely to verisimilitude. We could just as easily recall the very last paragraphs of Descartes’s Principles (on “moral certainty”), which resort to the metaphor of a ciphered message (nature), which the receiver (the physicist) tries to make sense of, by way of various hypotheses and experiments, without ever possessing the absolute proof that these correspond to the initial meaning of the message. In short, it is as if Stroud is telling us: “you cannot rationally refute this skeptical position so defined,” when no one has ever really dreamed of doing so. Which, it seems to us, poses the problem of the pertinence of this challenge which neither rationalists nor idealists have ever claimed to meet.

29 Further still, what we call the “new skeptics,” are they constant skeptics, in the restrictive sense in which they themselves define this term? When Nagel writes, in regard to the impossibility of philosophy’s ever proving the reality of the material world: “Our problem has no solution […] but to recognize this fact is to come as close as we can to the possibility of living in the light of truth,” doesn’t he make himself the defender of truth and, like Kant (and, to our eyes, Hegel also), a “prudent rationalist” in so far as judgments on Wirklichkeit are concerned? Likewise, does not Stroud tell us that skepticism is “something deep in our nature,” [38] adopting a naturalist tone (there is a “nature” of man which I know) that has a problematic status within his argument? Here we rediscover one of the features that we have only lightly touched on, as a constant undercurrent, in Hume—namely, a naturalism whose relation to skepticism is not immediately evident. In fact here we find once again an illustration of a certain reversibility: either there are no skeptics so defined, or there are very many of them. We should not misunderstand the nature of our perspective here: we do not accuse (far from it) current skeptics of “reviving” a problem that has never existed, and of thus being to rationalist and idealist philosophy what Don Quixote is to chivalry. On the contrary, indeed, we maintain that the problem posed by Stroud is a real problem. It is the problem of madness (itself defined within the framework of solipsism and representationalism). This problem, like others, can and must be interrogated. But it cannot be posed in the terms in which current skeptics pose it, when they accuse previous (metaphysicist, idealist, non-empiricist, non-Humean or non-analytic) philosophers of having wished to resolve it by way of a proof of the existence of the material things outside of us. And it is this point that should be retained: what is there to their refutations of classical and traditional metaphysical philosophy if the latter, thus restrictively defined, never really existed?

30 This referential skepticism has been criticized by numerous contemporary analytic philosophers who are also defined as “new Humeans”: either by the naturalists, who are numerous, as S. Laugier shows in the preface to the collection Naturalisme, héritages contemporains de Hume; [39] or else by the advocates of ordinary language philosophy. It is the refutations of skepticism produced by the latter that we must now approach, within the framework of our putting into perspective of contemporary skepticism on the basis of another skeptical vanishing point, that of Maimon. For an avowed naturalism (according to which there is a nature knowable by scientific means) no longer has anything to do with any conception of “skepticism,” since it is related to positivism; so it does not concern us here. On the other hand, it seems to us that ordinary language philosophy, although it refutes solipsist skepticism, also presents a figure of “skepticism” that can also be seen in the light of the self-referential skepticism proposed by Maimon.

2. The Meaning of Current Refutations of “Referential Skepticism”: The Relativism of Usages

31 This ordinary language philosophy, [40] beyond the divergence of its representatives, professes, minimally, a certain “realism,” which is not that of the Affiche Rouge of my everyday perception (the reality of the material world), nor the realism of the figures and laws of physical science, but the realism of collective institutions and customs, and primarily those of language and the set of practices attached to it. The plane of the object, from which every realist claims to set out, “hard facts” (in Bouveresse’s expression) are practices of the human community. It is from the point of view of the irrevocability of these practices that we will judge, on one hand, the so-called exact sciences (whose idealism is denounced by Hacker and Bouveresse alike), [41] and, on the other, traditional philosophy (equally idealist), [42] which will have bewitched the sciences, so as to lead them astray from the practices of our ordinary languages.

32 This “realist” philosophy obviously comprises an important component: the “refutation of skepticism,” as the approaches of Wittgenstein, Bouveresse and Cavell attest. Now, it is troubling to note that this refutation of skepticism is most often effected by way of stating the inanity of philosophical discourse, of which skeptics are, in fact, the unconscious symptom. [43] In a word, for the proponents of ordinary philosophy, it is traditional metaphysics—and its meaningless answers to questions that are not even posed—that are at the origin of Stroud’s skepticism and other “referential skepticisms”. From this they conclude therefore that it is philosophy, and its theoretical claims, that must be recused, in favor, if not of life in the woods, at the very least of a non-theoretical, non-speculative or non-“knowing” life, according to an “anti-intellectualist” position expressed by Cavell among others. [44]

33 Confronted with this strategy for the refutation of skepticism, a series of questions cannot fail to pose themselves to us; all of which converge once more on the idea of the reversibility of the terms “skepticism” and “anti-skepticism” that Maimon, from a completely different angle, revealed to us.

34 First of all, if traditional philosophy, or all metaphysics up until now, has never said what “current skepticism” determines as its content, its refutation by way of transitivity or, more rigorously, by contagion (current skepticism = metaphysics; current skepticism = nonsense; therefore…etc.) no longer has any basis. If we accept the restoration, which we have carried out, of the true content of this “current skepticism,” a referential skepticism, those who task themselves with refuting it cannot claim to refute, in the same breath, metaphysics or idealism. In refuting Stroud’s skepticism, the proponents of ordinary language affirm only the following: madness—not philosophy—is a disease, a potential malaise or a discomfort of the human being; consequently, we must seek therapies for it.

35 Correlatively, the argument according to which the theoretical stance must be abandoned because it generates the false problems of skepticism, which are its symptoms, loses all credibility. Cavell’s position, according to which “the basis of human being in the world in its entirety, its relation to the world, is not that of knowledge,” [45] becomes an affirmation that should be envisaged as a candidate for validation or contestation just like the affirmation that “all men naturally desire naturally to know”. Which leads us toward a third, more general problem concerning the skepticism of one (Stroud) and the relativism of the other (ordinary language philosophy), both declared “Humeans”.

36 Cavell asserts that our authentic relation to the world is not that of knowledge. How can he advocate this position (that is to say, propose it, while separating it from others)? By writing a theoretical treatise about what knowledge is? Assuredly not, for then we would fall back into the performative contradiction that the self-referential paradoxes formerly brought to bear against Kant. By invoking as a criterion of rectitude or normativity the way in which the uses of a language function? But which language? Language in general (which, moreover, seems like one of those generalities or generic entities that metaphysicians are criticized for), or a particular language, particular uses or practices? And therefore what is ordinary language, the mother of all norms and all virtues? Must one decide this by way of philosophical reflection? Obviously not, since the observable empirical facts are the basis of “realism,” ever advocated by the philosophy of the ordinary. In which case, will it be a question of a series of verbal customs that are empirically classifiable and belong to a given civilization? It is this solution that seems most often to be implicated by the discourse of the advocates of ordinary language. But then, as Dennett objects to Hacker, philosophy becomes the anthropology of civilizations, or the sociology of factual groups. This position is, assuredly, a hypothesis that is entirely philosophically envisageable: philosophy can no longer distinguish itself from empirical anthropology, and becomes the study of such and such practices of particular, defined societies. Nevertheless, particular rules of a particular civilization (for example, authentically American civilization) cannot claim to provide the norm for all others. [46] Cavell tells us that knowledge is not our authentic relation to the world, nor an ethically desirable relation. But Plato and Aristotle—and along with them, many ancient Greek and medieval Arab philosophers—were convinced of the contrary, as is Maimon, obviously. Let us pause to consider this situation of Maimon, which can be made a counterexample to an assertion that tries to authorize itself, through a “realism of facts,” themselves referred to uses and customs, to tell us that our relation to the world is “thus” rather than “so”.

3. The Other Paths of Reason?

37 Fichte and Hegel often argued vehemently against the partisans of common sense of their epoch. The latter are often close to the ordinary language philosophy of today, in so far as they always advocate common sense, common consciousness, and everyday life, against the culpable pretentions of philosophical speculations. Against them, Hegel writes:


The Germans always protect common sense against the supposed arrogance of philosophy. A vain effort, for if philosophy conceded everything to them, this would not be of any use whatsoever, for they have no good sense. Good sense does not reside in rough peasantry but treats with violence and freedom the determinations of culture, and does so according to the truth. [47]

39 Now Maimon, in a very different manner, approaches the positions of advocates of common sense as simply lacking in meaning, and even seems to have trouble considering their propositions otherwise than as huge pranks. Certain of his incidental remarks show that he finds himself confronted by their doctrines as today an individual might find himself coming out of his house and seeing everyone on the street walking on their hands. This individual would probably imagine a circus parade was passing, publicizing its acrobats; but, however this might be, his immediate reaction would be neither to want to convince them not to proceed in this way, nor even less to say to himself that he must, from now on, also walk on his hands. Now, if we seek a sociological or factual explanation for this large difference in reaction between the three philosophers in the same epoch, it may be very simple: Hegel and Fichte, because of their education and their social world, developed their thought in a Protestant context where the denunciation of reason and knowledge, the glorification of humble “unknowing,” could be recommended as a possibility and sometimes erected into an ethical necessity. We might recall, in relation to this, Luther’s discourse against Saint Thomas’s philosophical speculations: “Pure vanities […] Believe in Christ and do what you ought.” Fichte and Hegel were thus prepared to understand this doctrine, symbolically and intuitively, even if they did not at all share it. In fact, their “habitus” (to invoke Bourdieu again) had taught them to envisage this position as possible and often even as being incarnated in their social world. They were thus able to situate themselves “against,” where others declared themselves “for”. Maimon, on the other hand, who was initially set to become a rabbi, [48] began his development in a world where the separation of life, knowledge, and faith was literally inconceivable (even as an exotic hypothesis or crazy fiction). In this world, every believer had to study (according to his own capacities) to be able to render grace to God and to perfect himself morally. Life cannot be dissociated from this study—it is this very study. Having absorbed this habitus when very young, Maimon, in the course of his many travels without return, cultivated and reinforced it, as his one homeland and succour. In an entirely different context, P. Levi has shown that knowledge (like studying the Divine Comedy with a companion who knows it by heart, when one does not speak Italian oneself and our chances of being alive next week are almost nil) is a manner of living just as ordinary, or in any case just as necessary for the survival of “man,” as simple concrete and utilitarian exchange, apparently better suited to the context and circumstances (such as “pass me the brick and the trowel”). What can be said, also, of the numerous testimonies of mathematicians, who express their intense joy when faced with the truth given by the law that they have discovered? What can we say of the recognition of the physicist toward nature, often comparable to the amazed gratitude of the child who, in the chaos of the world, suddenly becomes conscious of the regularity of the seasons and their assured return? Should we say, like Luther, “Pure vanities!”?

40 What can we learn from our apparently untimely digression into the sociology of habitus, usages and practices? The following: that if ordinary language means the set of verbal customs proper to a given civilization or factual social group, it cannot avail itself of any prerogative to dictate what is the case, what is the norm. Ordinary language philosophy denounces theoretical knowledge and most often sets against it practice, the ethical (such as the “ethics of virtue” and “perfectionism” in Cavell), and everyday life. Quid juris? Whence proceeds the distinction between theory and practice? From philosophy, and more remotely, from a certain type of theology. In philosophy, it comes from Hume and then from Kant, who endorses it. Neither Fichte, who dissolves it, [49] nor Maimon, nor Hegel, nor Husserl take it up again; they try on the contrary to show the total reciprocity of these two dimensions. Whence the distinction between “life” and “knowledge,” the scission between man and the desire to know? It comes from Hume, among others, as we have said, but Aristotle never subscribed to it. How to establish the legitimacy of one position over another (Hume rather than Aristotle, Kant rather than Fichte)? By saying that neither Fichte nor Aristotle had common (American) sense, [50] and by reasserting without any further distinction that every true man does not naturally desire to know? Or else by arguing, in a rather “moralizing” way, [51] that he is culpable for wanting to know, that this betrays a suspect desire for “mastery” of the world and of others, or that it is the symptom of a frightful lack of “modesty”? Luther authorizes himself (rightly or wrongly), through a certain idea of God and of his creation, to pass such judgments on humans and their “sickly” symptom or care for theoretical knowledge. But what authorizes ordinary language philosophy to do so if, in other matters, it professes a relativism of usages and customs?

41 In truth, the only way to establish the legitimacy of this position according to which theoretical knowledge is neither a natural nor a desirable attitude, would be to demonstrate it theoretically. But, in order to do so, would it not be useful to tell us first of all what this “nature,” this “man” is “in general,” and to show us how it can be attained? And once this demonstration has taken place, would it not be necessary, then, to show us the legitimacy of passing from the proposition “it is in the nature of man not to have knowledge as his primary relation to the world” (fact) to this other proposition: “This therefore must be the desirable norm” (value)? Something which, obviously, ordinary language philosophers would not want to do for multiple reasons, of which we will mention here only one: it would mean entering once again into the theoretical process, into considerations that are not “concrete” (“man,” “nature”), and would be out of context. For, as Cavell tells us, claims to “knowledge” (which, he elaborates, “is not what we think under the name of knowledge”) must be “concrete,” “inscribed in contexts” and do not bear upon “generic objects”. [52] At the end of this putting into perspective of “skeptical realism” by skeptical idealism, it seems that the vanishing points chosen (Maimon, self-referential skepticism) make more visible the curious problems raised by the widely-held “New Humean” position today.

Conclusion: The Limit of Skepticism and the Skepticism of the Limit

42 Ordinary language philosophy opposes “realism” to “referential skepticism”. This realism, once elucidated, turns out to be a relativism. The starting point has reversed into its opposite: skepticism has been combatted by opposing to it another form of skepticism (another level on the page, as Maimon would say). This explains why certain fervent disciples of Wittgenstein have been able to characterize his position as “skeptical,” when part of his oeuvre is dedicated to showing that “skepticism” is nonsensical. This skeptical stance, we note, has nothing in common with Maimon’s, which, as we have shown, ends up in an apology for theoretical reason. Faced with this “option” of Maimon’s, skeptical realism cannot, as we have seen, justify its antitheoretical claim. In this way, we suggest, it indicates the limits of a skepticism so defined.

43 More generally speaking, our analyses bring to light the following result: the “revival of skepticism,” borne by the “New Humeans,” is divided into two distinct branches, namely referential solipsism, and the relativism of practices or usages. Neither of these branches is connected with self-referential skepticism. The question of self-reference is absent from the “revival of skepticism,” whereas it profoundly marked the skepticism of the only two philosophers of the modern epoch who assumed the position in a constant manner. Both Rorty and Apel have, with other vanishing points in sight, raised this absence of the problematic of self-reference within analytic philosophy: Apel, to deplore it and to show that this problem traverses this whole movement and weighs against its claims to validity, given a constant resurgence of performative contradictions or pragmatic paradoxes. [53] Rorty simply indicates it, writing that:


The current caesura between analytic and non-analytic philosophy coincides well with the division of philosophers who are not interested in historico-metaphilosophical reflections on their own activity and philosophers who are. [54]

45 In fact, the problem that constantly crops up again, throughout our analyses, is the problem of the status of the philosopher’s discourse, and the question of the very possibility of its enunciation. Maimon brought this problem to light, but he also assumed it for his own part. He unfolded all of its consequences, to the point of proposing what we might call, in the end, a skepticism of the limit. This skepticism can take account of the status of its own discourse, indicating the point from which it is spoken. It can do so, without recourse to indirect strategies (style) or destructive strategies (resolution through the “country gentleman”) proposed by Aenesidemus, and surreptitiously implied in many other more contemporary philosophical practices. He outlines a constructive skepticism that tries to establish a relation to the philosophical tradition other than that of the incessant deconstruction or denunciation of past metaphysics. The topology of systems, to which his work directs us, allows us to define realism otherwise than as a “systematic suspicion toward idealism,” and to thus make visible the true positions of one and the other alike. An idealist, Maimon is also a “Humean”. If he demonstrates how the coherence of criticism is skepticism, at the same time he demonstrates that the coherence of skepticism is transcendental philosophy. Making of skepticism an ode to reason, a witness to truth, and at the same time a trace of the desire for universality, Maimon invites us to conclude on this entreaty: “Skeptics of the world, unite.”


  • [1]
    J. Brunschwig, Études sur les philosophies hellénistiques: épicureanisme, stoïcisme, skepticisme (Paris: PUF, 1995), 341.
  • [2]
    On the definition of “urban” (as opposed to “rustic”) skepticism, see J. Barnes. “The Beliefs of a Pyrrhonist,” Proceedings of the Cambridge Philosophical Society, ed. Kenny and McKenzie, 28 (1982): 2–3. Reprinted in The Original Sceptics: A Controversy, ed., M. Burnyeat and M. Frede (Indianapolis: Hackett, 1997), 58–91: “he holds that he is content to believe only in things to which ordinary people assent in the ordinary course of life”; this is why the urban skeptic “directs the epoche only toward a precise target, substantially: philosophically and scientific questions”.
  • [3]
    D. Hume, A Treatise of Human Nature (Oxford: Oxford University Press, 2000), 121, 142 [Traité de la nature humaine, tr. Baranger and Salter (Paris: GF, 1995), 264, 303]. In the Inquiry Hume also speaks of the malady of “philosophical melancholia”.
  • [4]
    D. Hume. An Enquiry concerning Human Understanding (Oxford: Oxford University Press, 2007), 9 [Enquête sur l’entendement humain (Paris: GF, 1983), 55]. We do not intend to suggest, with this overhasty enumeration, that the authors in question never said, at some moment in their oeuvre, that they were indeed skeptics. The tension between Hume’s skepticism and his rather positivist naturalism is widely debated and is not our concern here—we are concerned solely to point out that in each of these “skeptical” authors we also find a denunciation of “skepticism”.
  • [5]
    Article “Xenophanes” in P. Bayle, A Historical and Critical Dictionary in 4 Volumes, Selected and Abridged (London: Hunt and Clarke, 1826), vol 4, 30 [Dictionnaire historique et critique (Paris, 1697 [second edition 1702])].
  • [6]
    Of course, we must provisionally distinguish skeptical strategies—devices devised so as to attain more reason or truth—from a permanent skeptical position. The first stance is incarnated first and foremost by Descartes and, as a general rule, by all rationalists—such as Diderot, who, inquiring into the definition of “skeptic,” responds: “It is basically a philosopher who has doubted everything he believes, and who believes only that which a legitimate usage of reason and of his senses has demonstrated to be true”. (D. Diderot, Oeuvres philosophiques [Paris: Garnier, 1964], 27–8); the second, as we suggest with our enumeration, is far more difficult to identify since, among all “modern” philosophers who subscribe to the term “skeptic,” there is not one who did not vehemently denounce skepticism—which renders problematic their “constant” attachment to this position.
  • [7]
    All of these terms are taken from the preface to C. Tiercelin’s Le Doute en question (Paris: Éclat, 2005).
  • [8]
    Ibid., 22.
  • [9]
  • [10]
    Of course, we take the risk here of inviting dozens of counterexamples: Bayle, quite obviously, Gassendi probably, as well as innumerable libertines. Given our aims here we cannot discuss them. Let us say simply that this skepticism is always an attenuated one, presented as “provisional” or “mitigated” (on this last adjective, see its reprise in Tiercelin, who takes it from Hume—Tiercelin, Le Doute, 11; or the way G. Paganini applies it to Gassendi and Mersenne, in Skepsis [Paris: Vrin, 2007], 9).
  • [11]
    On this widely-accepted point, see R. Popkin’s major history of skepticism, The History of Skepticism from Savonarola to Bayle (Oxford: Oxford University Press, Third edition 2003).
  • [12]
    The first example is borrowed from Kandinsky (Regards sur le passé [Paris: Hermann, 1974], 91), the second from Hoffmann (Olympia in the Tales).
  • [13]
    We have shown, in our book Fichte, reflexion et argumentation (Paris: Vrin, 2004) (chapter 1: “Le crise de la métaphysique”) as well as, more recently, in our article “Leibniz or Thomasius? An Enquiry Concerning German Modern Philosophy,” Idealistic Studies, 37:2 (Summer 2007), that this period, in Germany, far from having been dominated by what is known as leibniziano-wolffianism, was dominated by skepticism in all its forms.
  • [14]
    Initially published without an author’s name, Aenesidemus (the work of G. E. Schulze, whose contemporaries, for convenience, are referred to by the name of the work’s principal character) debates with a Kantian character by the name of Hermias. The full title of the text is Aenesidemus oder über die Fundamente der von dem Herrn Prof. Reinhold in Jena gelieferten Elementarphilosophie, and it was published in 1792 (republication Hamburg: Meiner, 1996). An excerpt in English, translated by G. de Giovanni, can be found in Between Kant and Hegel: Texts in the Development of Post-Kantian Idealism, ed. G. di Giovanni and H. S. Harris (Indianapolis: Hackett, 2000), 104–35.
  • [15]
    We have commented at length on the position of Aenesidemus: in the preface to Méditations personnelles (Paris: Vrin, 1999), but above all in our Fichte, which contains a whole chapter dedicated to him (chapter 3: “L’effondrement de la théorie kantienne de la validité”). In this chapter, on one hand, we describe the philosophical shockwave that this text would have provoked in 1792, and on the other, we identify and set in order the arguments against it that were expounded in the commentaries that followed—for example, its pre-Kantian (since dogmatico-realist) character, its incoherence, and its supposed confusion between transcendental and metaphysical sources. We sought to show how Aenesidemus’s text escapes these criticisms because it professes an authentic skepticism (the point of view of its enunciation) and not a dogmatic realism (the point of view of the content of what it says). Thus, in the present article we suppose these prior demonstrations to be established, and so as not to overload the footnotes, we refer the reader to these more exhaustive arguments on the text and its reception.
  • [16]
    Schulze, Aenesidemus, 142–3, among other occurrences.
  • [17]
    Aetas Kantiana, vol. 244: II (Brussels, 1968), 578.
  • [18]
    Aenesidemus even takes on the basis of logic and mathematics, namely the intangibility of the principle of noncontradiction; on this point, we refer the reader again to our chapter on Schulze cited above.
  • [19]
    Armengaud defines the term “pragmatic,” in the least vague way it is currently employed, as an attempt to respond to the question “What do we do when we speak?” (F. Armengaud, La Pragmatique [Paris: PUF, 1985], 3). On the reconciliation of pragmatics and self-referentiality in German idealism, see, among others, our article “Fichte and Austin,” Fichte-Studien, supplementa series (2006 Halle Colloquium) (Amsterdam: Rodopi).
  • [20]
    Hume, Treatise, 175 [Traité, 560]. Hume also admits that it concerns an individual who would have neither “curiosity nor love of truth”. On this relation between life and philosophy Hume thus seems to oscillate, as he also does between the affirmation of a possible science of human nature on the model of physics and skepticism with regard to knowledge. Nevertheless, he seems to have been one of the first to have dissociated life and philosophy in so decisive a manner, whereas the ancients held them to be indissolubly linked. Schulze seems less hesitant to take up his position as a skeptic. His critique against philosophical speculations (which are all held to be vain) implies theoretical silence, and the appeal to a life from which philosophy would have disappeared. In this sense, he is indeed a “sincere and constant” skeptic.
  • [21]
    The Essay on Transcendental Philosophy appeared in 1790, taking its impetus from the Kantian distinction between reflective and determinant judgments, already present in the first Critique, and in this way anticipating the third Critique. Kant recognized in Maimon his most acute reader and Fichte saluted him as one of the greatest philosophers of his age. English translation, trans. N. Midgley et al (London and New York: Continuum, 2010).
  • [22]
    Again, so as not to burden our notes with too many citations and proofs that have already been undertaken elsewhere, we refer the reader here to the long chapter we dedicate to Maimon in our Critique de la représentation (Paris: Vrin, 2000), in chapter 2, “La notion de réflexion et la question de la référence,” along with our prefaces to Fichte’s works, notably Doctrine de la science [1805] (Paris: Cerf, 2005), and also Doctrine de la science nova methodo (Paris: Le Livre de Poche, 2000). Our aim here is not so much to analyze the whole of Maimon’s thematic as to put his skepticism into perspective with that of Aenesidemus, so as then to compare them with other more contemporary acceptations of the term.
  • [23]
    In our article “Coalition des systèmes et topologie des contradictions : la pratique herméneutique de Maimon,” we have shown how Maimon makes a philologically correct and literal reading of the Critique even as he totally subverts its final meaning. In Haskala et Aufklärung, philosophes juifs des Lumières allemandes, ed. W. Adam et al. (Paris: CNRS, 2009).
  • [24]
    The title of the last book written in his lifetime: Science de la science et réflexivité (Paris: Raisons d’agir, 2001).
  • [25]
    A theoretical reason which, in spirit, is not dissociable from (nor opposable to) practical reason, or life. We will come back to this most important point.
  • [26]
    On this point, see, among others, the collection edited by G. Freudenthal, Salomon Maimon: Rational Dogmatist, Empirical Skeptic (Dordrecht: Kluwer, 2003), which takes up the terms in which Maimon defines himself. Let us nevertheless recall, so as not to frighten the contemporary reader too much with this now-infamous adjective “dogmatic,” that at the time (and according to the acceptation that Kant himself employs in certain contexts), “dogmatic” simply meant someone who says something positively; it thus has an apophantic status (such as: “it is sunny; physics is only probabilistic, man is finite,” etc.).
  • [27]
    See our article : “Coalition des systèmes et topologie des contradictions: la pratique herméneutique de Salomon Maimon,” in Haskala et Aufklärung, philosophes juifs des Lumières Allemandes.
  • [28]
    Maimon makes obvious use of this biblical metaphor. Citing Deut. 34:4 (for example in Gesammelte Werke [Hildesheim: Olms, 1965-1976], vol VII, 554), he shows that the path is the land and the eland, the path (taking up a classic refrain from his own cultural space of Eastern European Judaism) which is epitomized in the well-known joke about the Russian who, arriving at the “study house” of the “shtetl,” asks the rabbi: “Are all these people studying so that they will go to paradise?”—to which the response comes: “No, they are in paradise because they are studying”.
  • [29]
    If one seeks a possible illustration of this attempt at a topology of contemporary philosophies, one might, it seems (and with the caveat that the author himself will perhaps not recognize himself in this context) turn to Hugues Choplin’s book L’Espace de pensée contemporaine (Paris: L’Harmattan, 2007) (especially the introduction, “une démarche topologique”). Derrida, Henry, Levinas, Deleuze, Marion, and Laruelle are united in what Choplin calls a “topological framing [cadrage topologique—chapter 1]” of contemporary philosophy.
  • [30]
    M. Gueroult, la philosophie transcendantale de Maimon, Alcan, 1929, 17; see also the whole of the paragraph entitled “La fiction copernicienne de la Critique,” 14sq.
  • [31]
    R. Descartes, “Principles of Philosophy,” in Philosophical Essays and Correspondence, ed. R. Ariew (Indianapolis: Hackett, 2000), 222–72, §29–30 [A. T. (Paris: Vrin, 1964), vol. VIII, 92–3].
  • [32]
    It seems that this proof would be possible on all texts of philosophy, starting with the most skeptical. Nevertheless, to shed some light on this question that the thinkers of autoreference always pose—namely “in relation to what?”— one might also, more pleasantly, recall that passage in In Search of Lost Time where Proust reports the sighing of a melancholic marquise who ceaselessly exclaims “How boring life is!” without (the narrator tells us) one ever knowing quite what she is comparing it to.
  • [33]
    Stroud intervened as a defender of skepticism in 1968, in an article written against Strawson. But it is his 1974 work The Significance of Philosophical Scepticism (Oxford: Oxford University Press) that is the most complete.
  • [34]
    T. Nagel. The View from Nowhere (Oxford and New York: Oxford University Press US, 1986) [Le Point de vue de nulle part, tr. S. Kronlund (Combas: L’Éclat, 1993)].
  • [35]
    For the meaning of this term, see R. Brandom, Articulating Reasons: An Introduction to Inferentialism (Cambridge, MA: Harvard University Press, 2001) [L’Articulation des raisons, tr. J.-P Cometti and C. Tiercelin (Paris: Cerf, 2010)] and Making It Explicit (Cambridge, MA: Harvard University Press, 1998) [Rendre explicite, tr. ed. I. Thomas-Fogiel (Paris: Cerf, 2010)].
  • [36]
    Stroud, Significance of Scepticism, 22. In complete agreement with Tiercelin, who shows how Stroud identifies “skepticism” with “the impossibility of proving the reality of the external world,” we have noted this restrictive definition of skepticism by Stroud in our books Fichte, reflexion et argumentation (Paris: Vrin, 2004) and Référence et autoréférence (Paris: Vrin, 2006; American Translation : Columbia, New-York, 2011), and in our article “Fichte et l’actuelle querelle des arguments transcendentaux,” Revue de métaphysique et de morale 4 (2003).
  • [37]
    Obviously, we are not suggesting that all philosophers say the same thing, but simply that the difference between them does not pass by way of this question. On the question of the reality of material things, Descartes and Russell appeal to the “great propensity” to believe in the reality of material things. Kant, Fichte, Hegel and Maimon consider themselves to be situated in a philosophical context where the question no longer makes sense. It is therefore not their aim to respond to it.
  • [38]
    The Significance of Philosophical Skepticism, 39.
  • [39]
    Revue de Métaphysique et de Morale 2-2003 (June). Let us note, in passing, that the question of the compatibility between Hume’s naturalism and his skepticism, which we have only evoked as a problematic horizon of the clear assumption of the term “skeptic,” is discussed by J. Benoist, “Le naturalisme avec ou sans le skepticism,” Ibid., 127sq.
  • [40]
    It seems admissible to say that the second Wittgenstein, Austin and then Hacker, Bouveresse and Cavell, belong to this current. Certainly the delimitation between those who belong to it and those who do not can sometimes seem rather fuzzy. Thus, having been refuted at length by Bennett and Hacker in Philosophical Foundations of Neuroscience (which aims at a critique of the neurosciences from an ordinary language point of view), Daniel Dennett, in a 2005 paper entitled “Philosophy as Naive Anthropology: Comment on Bennett and Hacker” (, accessed April 19, 2013), confesses to being surprised that his detractors had not perceived more points in common with him, in so far as (he says) he is himself the fruit of that epoch of the “early 1960s, when Ordinary Language Philosophy ruled,” and that “Ryle and Wittgenstein were the authorities on the meanings of our everyday mentalistic or psychological terms.”
  • [41]
    Hacker in Philosophical Foundations, and Bouveresse in Langage, Perception et Réalité, (Nîmes: Chambon, 2004), vol. 2, which observes that quantum physics revives or endorses the constructivist and idealist paradigm (see, in particular, 20sq) and thus undertakes a critique of this “idealism” of physics in the name of ordinary language.
  • [42]
    Bouveresse’s obsession with idealism is such that he ends up defining “realism” only by way of a negation of idealism. Interviewed by Rosat on the “realist exigency,” which he puts forward in all of his works, he responds that this realism consists in “systematic distrust of idealism”. Rosat, Le Philosophe et le Réel (Paris: Hachette Littératures, 1998), 35.
  • [43]
    C. Tiercelin has emphasized this symptomal approach of Cavell. She writes that he intends to “arrive at a symptomal (Nietzschean) analysis: The basis of human being in the world in its entirety is not that of knowledge.” Tiercelin, Le Doute, 23.
  • [44]
    S. Cavell, The Claim of Reason (Oxford and New York: Oxford University Press, 1979) [Les Voix de la raison, tr. S. Laugier and N. Balso (Paris: Seuil, 1996)—especially 358 and also 329]. Cavell denounces the knowledge claims as not sufficiently “concrete,” too theoretical, far from life. This is why we allow ourselves to use the term “anti-intellectual” which Brandom also employs, along with that of “quietism,” in reference to Wittgenstein (on his relation to Wittgenstein, see our preface to the French translation of Making Things Explicit). The use of the term “anti-intellectual” on our part should not be understood as a value judgment; one can be effectively “anti-intellectual,” and this stance, like any other, can be discussed. The term simply expresses the stance of those who denounce theoretical philosophy (or the philosophical drive).
  • [45]
    Cavell, Claim of Reason, 45.
  • [46]
    Above all, it is important not to misunderstand our inquisitive objections here. Our aim is not to “violently” attack Cavell, something we have been reproached for in relation to a chapter of a book discussing a stance of Wittgenstein’s. Cavell’s analyses on cinema, literature, etc. are enchanting for any intelligent reader. All the same, in France, Laugier’s analyses of Austin and Cavell, and her own arguments relative to questions of ethics, are models not only of hermeneutic understanding but of human understanding, as are Hume’s. The generosity toward the human evidenced in these analyses shows that, sometimes, the belief that man is only matter proves more comprehensive, tolerant and benevolent to him than certain beliefs that think him the image of God. So it is not a matter of attacking ordinary language philosophy but of expressing certain concerns with it. And this for two reasons: (1) If we do not want to fall into the gap already denounced by Dennett, when he ironizes on Hacker’s “Saint Wittgenstein”—that is to say if we want to respect the idea of an argued exchange, then we must do so; (2) Ordinary language philosophy uses the ordinary to authorize it to recuse metaphysical (and in particular idealist) speculation. Thereby it intends to situate itself on the terrain of philosophy, not that of sociology (the level of factual enquiries to which we could adhere like, here, Bourdieu). So we must, says Bouveresse, “systematically distrust idealism”. This is why we insist on asking: Isn’t the term “realism” so defined, as a relativism of usages and customs, transformable without rupture or opposition into this other: “skepticism”; and if this is the case, what is the philosophical status of these attacks against the philosophical tradition and idealism?
  • [47]
    Cited by Rosenkranz in Hegels Leben,translated into French in the journal Philosophie, 13. We will abstain from citing Fichte, whose vehement refutations of this current would run the risk of shocking sensitive souls.
  • [48]
    On Maimon’s life see, among others, the recent article by N. Weill, “L’expérience autobiographique de Maimon,” in W. Adam et al. (eds.) Haskala et Aufklärung, 23–35. Recall the absolutely watertight separation between his initial milieu and the German world (no exchanges, no common language, etc.). The knowledge Maimon was able to acquire when he established himself in Berlin could therefore not have been purely intellectual (he learned German from reading Mendelssohn, as he learned English by reading Hume and Latin through Spinoza); this knowledge was not anchored in any habitus nor any practice. Which explains his sometimes “off-kilter” statements, like those evoked here.
  • [49]
    On this point, see our article “Fichte Against Kant in the System of Ethics” in Philosophy Today, 52 (Chicago, Fall-Winter 2008), 3-4.
  • [50]
    We allow ourselves this qualification in so far as it is Cavell himself who intends to return to a properly American heritage (Thoreau, Emerson), to an “original voice” to be rediscovered outside the way in which American philosophy would later define itself, on the basis of the contribution of European analytic philosophy (particularly Carnap, who would mark it so profoundly with his logicist imprint). On this point, see the collection Cavell, Qu’est-ce que la philosophie américaine?, tr. C. Fournier, S. Laugier (Paris: Folio-Essais, 2009). We should note the link between a certain form of anthropological relativism (usages are contextual usages, in a given civilization) and the search for roots, henceforth condemned to be specific. Which is consequential, if not necessarily desirable for all, in so far as what is placed in question in both cases is universalism.
  • [51]
    C. Chauviré has already noted this tendency toward moralization of a certain strain of analytic philosophy in his fine article “Pourquoi moraliser les normes cognitives?,” Cités 1-2001.
  • [52]
    Cavell, Claim of Reason, 358, 329.
  • [53]
    See, for example, his article in Critique (1983), as well as our commentary on his position in Référence et autoréférence.
  • [54]
    R. Rorty, Essays on Heidegger and Others, in Philosophical Papers (Cambridge: Cambridge University Press, 1991), Vol. 2, 21.


Paradoxically, there are very few skeptical philosophers in the modern epoch (since Descartes) who have laid claim to skepticism “consistently,” that is to say, without denying, diminishing or rejecting it at some or other period of their working life. Surprising as it may at first seem, only two philosophers—before the “skeptical revival” seen in analytic philosophy in the last thirty years—truly assumed skepticism without qualification of any kind: these two philosophers, who were contemporaries, are the post-Kantians Schulze (also known as Aenesidemus) and Maimon. This article defines the skepticism that each of them assumed “consistently”. Once the meaning of their positions is elucidated, the article goes on to compare them with a more contemporary form of skepticism. This putting into perspective ultimately reveals the coherency and consistency of Maimon’s transcendental skepticism compared with the position of those who are today known as the “New Humeans.”

Isabelle Thomas-Fogiel
Université de Paris-1 (France), University of Ottowa (Canada)
Uploaded on on 03/07/2014
Distribution électronique pour P.U.F. © P.U.F.. Tous droits réservés pour tous pays. Il est interdit, sauf accord préalable et écrit de l’éditeur, de reproduire (notamment par photocopie) partiellement ou totalement le présent article, de le stocker dans une banque de données ou de le communiquer au public sous quelque forme et de quelque manière que ce soit.
Loading... Please wait