The Death of Philosophy: Reference and Self-reference in Contemporary Thought

Kant expresses himself in multiple ways—this is why current proposals to renew or “reappropriate”¹ the critical project do not all move in the same direction. Thus, if we undertake an analysis of these contemporary returns to Kant, we ought to distinguish two general ways of reading Kant—one starting from the Critique of Judgment and regulative judgment; the other from the notion of the a priori and from the transcendental, the fulcrum of the Critique of Pure Reason. But, as in previous chapters, we will see how these two ways of reading Kant embody the oscillation that I have stigmatized everywhere else—Kant, like Quine (in chapter 1) or Austin (in chapter 2), is the core that supports an oscillation between skepticism and positivism. On the other hand, we will see in this chapter that only one of these reformulations of Kantianism contains the logical pathology of pragmatic contradiction. The second, Karl-Otto Apel’s, contains a tension that I will elucidate but does not repeat stricto sensu the contradiction that I have analyzed up to this point. It will thus constitute an indirect proof of my thesis that the destruction (Rorty, Quine, Austin, Levinas) or even minimization of philosophy’s role (Habermas) leads to an insurmountable contradiction that annuls the denial or minimization. And indeed, among all the authors I’ve examined Apel is the only one to assert that philosophy is a first,² autonomous, and distinct discipline. Before I tackle Apel’s view, unique in an otherwise relatively unanimous context, I shall first analyze the reconstruction of Kantianism that surreptitiously, without ever saying as much or sometimes even meaning to, endorses the idea of an “end of philosophy” and thus finds itself “standing with” Rorty.

If Karl-Otto Apel is the source for a second reading of Kant,³ the first, which takes the Critique of Judgment as the nodal text for the critical project, is proposed by the ensemble of contemporary commentators on Kant. Indeed, beyond the difference between their propositions and the diversity of their sources, all the contemporary interpreters of Kant share a single presupposition: the Kantian enterprise as a whole should be reconstructed or reinterpreted starting with the notion of Urteilskraft, as it is particularly developed in the Critique of Judgment. In Germany, Manfred Riedel’s work⁴ best embodies this desire to take the principles of reflection proper to the faculty of judgment as the sole guiding thread. In France, Beatrice Longuenesse’s fine study, Kant and the Capacity to Judge,⁵ has systematically clarified the first Critique in light of the theory of reflection thematized in the third. The same perspective is shared by Claude Piché, whose Kant et ses épigones,⁶ moreover, underlines the undeniable preeminence of the third Critique today and traces it back to a 1982 colloquium held at Cerisy under the direction of Jean-François Lyotard, whose participants’ papers have been collected under the telling title La faculté de juger.⁷ Allow me to note, finally, that of the writers today who claim to be representatives of Kantianism, such as Luc Ferry and Alain Renaut, although their reading is apparently very far from Lyotard’s, they, too, propose nothing less than a promotion of the Critique of Judgment to the key position for the entirety of the critical project. Both as historians and as philosophers concerned to advance a “return to Kant,” Ferry and Renaut offer a reconstruction that begins with the Critique of Judgment—and it is for this reason that I have chosen to focus on their reading of Kant, among numerous other candidates.⁸ To give meaning again to the Kantian vision of the world is the goal that Ferry and Renaut gave themselves a few years ago, naming this project “critical humanism”—a goal whose implications are developed in Renaut’s book Kant aujourd’hui. By “the critical project,” Renaut means Kant’s oeuvre in its entirety and two years of Fichte’s philosophical activity: the 1794 Wissenschaftslehre and the doctrine of law. Critical humanism proposes to rethink our modernity starting with this entity (Kant-Fichte). So I must reconstruct their proposed reading.⁹

Relative to Kant, critical humanism accepts the Heideggerian thesis of radical finitude. It thus reads the Critique of Pure Reason in light of Kant’s aesthetics. However, the advocates of this rereading of Kant refuse to accept the corollary thesis that in the Critique of Practical Reason Kant cannot relativize the thesis of radical finitude with the idea of an absolute self-placement.¹⁰ Similarly, critical humanism accepts and takes up in its entirety Alexis Philonenko’s interpretation of Kant and the early Fichte; however, they do away with the basis for this interpretation, namely, his reference to Martin Luther and the idea of philosophy as practical theology.¹¹ This dual acceptance, accompanied by a dual refusal, explains their attempt to find in the Critique of Judgment the means to affirm the moral dimension of the Critique of Practical Reason without either renouncing radical finitude or founding it on a theology that would give it meaning and value. Thus, in The Era of the Individual, after having recalled the impasse at the Davos debate (either a radical finitude but no autonomy—Heidegger—or autonomy but a relativization of finitude—Cassirer), Alain Renaut proposes critical humanism’s solution: “The discourse on moral freedom (autonomy), as it is presented in Kant’s practical philosophy, exhibits the same distinctive features as the discourse in the Critique of Judgment on intuitive understanding—that is, on God as an infinite being.”¹² Moral statements should be read not as determinate but reflexive judgments: “This would involve conceiving of the supreme principles of practical reason as constituting principles of reflection.”¹³ The movement is clear: in order to preserve the thesis of radical finitude, moral judgments are understood as having the same status as judgments about art or life.

We can best bring out the simultaneously concrete and profoundly skeptical implications of critical humanism’s decision to transform [moral] judgments into reflective judgments by first examining another of these two writers’ positions, namely, on the question, which they consider as central for Kant and Fichte, of recognition of others.¹⁴

Let’s recall the problem as it has been posed in French Fichte studies of the last thirty years: for Fichte, if I do not manage to find a criterion that proves, in the phenomenal world, that I am dealing with a free being with whom I can communicate, and not with a machine stripped of all intentionality, then the Critique of Practical Reason loses all philosophical legitimacy. This question about the appearance of recognition of others as rational beings can be illustrated by the following example:¹⁵ a Nazi officer could conduct himself in a morally or juridically irreproachable manner with those whom he considers to be German, for they are, in his eyes, part of humanity. But we cannot say, from a strict Kantian point of view, that he transgresses the moral law in his attitude with respect to those whom he considers to be Jewish, because the latter, in his eyes, are not part of humanity. Put bluntly, if it is not possible to understand the recognition of every other as a free being, then it is illicit to condemn Nazism—what is called, from themes developed in the second Critique, the “moral vision of the world” thus has no philosophical validity. To be assured of consistency will thus mean to determine that status of the description of the empirical sign of freedom in Fichte.

It obviously cannot be a matter of a purely empirical description, without which it could not function for even a moment as a proof. It is thus a matter of phenomenological description—hence the question becomes, What relationship does this phenomenology maintain with a theory of truth? What is the relation between philosophy as a rigorous science and lived experience, between scientific construction and phenomenal manifestation? What is the status of Fichte’s discourse when he claims to be able to have an apprehension of freedom in man? Put differently, what sort of judgment claims that the mechanism does not work for all phenomena?

To this question there are two possible responses:

French interpretations, prior to the critical humanist interpretation, tended to hold that Fichte must give a necessary status to his proposition; Fichte must demonstrate that the phenomenal world is not entirely ruled by the law of determinism, or else (and this is entirely Martial Guéroult’s view) the philosopher must either renounce “a certain number of his practical assertions,”¹⁶ or else rethink his speculative philosophy as a whole. The difference between the Wissenschaftslehre of 1794 (understood by Guéroult as a system of the finite subject) and the Wissenschaftslehre starting in 1801 (understood as a system of the absolute) could be understood by the necessity to ground in an absolutely true manner (in Kantian terms, from determinate judgments) the mechanism’s limitation in the world of phenomena.

Conversely, for Ferry and Renaut, Fichte’s discourse must be reflective. On this point, Fichte renews the Kantian analysis of the teleological antinomy of the Critique of Judgment.¹⁷ This last point can be summarized thus: certain beings are infinitely improbable, according to the laws of determinism alone. It follows that we can postulate purposiveness not as a determinate principle but as a principle of reflection (that is, as a maxim allowing the direction of thought).

In the Critique of Judgment, if the viewpoint of purposiveness must be considered as a reflective moment, it must likewise be so from the viewpoint of the mechanism [of nature]. If that were not the case, we would have absolute opposites, where the affirmation of a thesis would necessarily entail the negation of its antithesis, insofar as—unlike in the third antinomy—the subject of the two propositions is the phenomenal world. Hence, Kant tells us, [nature’s] mechanism and purposiveness can be maintained only if we refuse to think of them as constitutive principles—that is, as laws governing phenomenal reality—but rather make them principles governing the scholar’s investigations. The contradiction, as Schelling had already remarked, would be less a contradiction in reality than an irreducible opposition in the human mind.

This solution has the advantage of not transgressing the limits of finitude in the Heideggerian sense of the term, but it is not, however, exempt from paradoxical consequences. Indeed, nature’s mechanism cannot claim to limit purposiveness, but—and reciprocally—nothing can claim by means of argumentation to limit nature’s mechanism. Here again, we have two principles of reflection, both legitimate if the subjects will agree to consider them as simple principles of investigation. A scientist will choose to follow one rather than the other without having to understand why. To put this concretely, an imitator of La Mettrie today could well say that man is a machine and proceed to show its mechanisms if he underlines with his work that “everything happens as if”; reciprocally, a scientist could, if he wished, attempt to reconceive purposiveness provided that he claims to obey only a heuristic principle. A historian or a sociologist could maintain a thesis and accept its coexistence with its opposite, without claiming to dispute this, for in the final analysis, everything depends upon the principle of reflection that one adopts. This is a solution whose immediate adoption in the scientific community would have the advantage of giving scientific discussions an amiable, consensual, even ecumenical character but would have the disadvantage of delegitimizing any attempt at argumentation, or even of eliminating any discussion. Moreover, with this solution, Kant and Fichte strangely become precursors to Paul Feyerabend’s view that “anything goes in science,” the strongest purposiveness as well as the strictest determinism.¹⁸ Neither facts nor principles govern the scientist’s investigations. Here the skeptical appearance of this system becomes patently clear.

With such a lens, science is in line with aesthetic judgment; the evaluative structure that holds for works of art becomes the ultimate model to which every judgment must refer. If we accept this as Kant’s solution and if we apply it to moral questions, we risk being logically¹⁹ forced to the following assertion: moral questions are not subject to determinate judgments, rather they are placed, like our aesthetic judgments, under the sign of “everything happens as if.” We may not recognize another’s humanity, just like we might not recognize the beauty of a painting or the value of a poem, and we have nothing with which to oppose the Nazi apart from the opposite principle (the method of construction). To make moral questions depend upon reflective judgment risks leading us to a renunciation of any validity for ethical norms. In this domain, we could claim only universality, without necessarily, as in the case of a discussion of art, de facto condemning someone who claimed the opposite. Transforming truth into a problematical demand leads to defining meaning from only its aesthetic style. And yet a lack of taste (such as not recognizing Picasso as a great painter, or objecting to a given kind of musical harmony) does not seem, in its initially Kantian presentation, to be the object of the same kind of genuine condemnation as moral errors like, for example, exterminating an entire community. The critical humanist theory, however, does not seem to be able to philosophically produce a conceptual distinction. Aesthetic communication becomes the reference for all communication—moral, juridical, political, even scientific. On this point, Ferry and Renaut write, “Law and aesthetics are then reunited in the general theory of communication or direct intersubjectivty²⁰ whose core is surely to be investigated in the notion of reflective judgment.”²¹

On the basis of these analyses, we are able to perceive the structure of the entire system attributed to Kant and Fichte, and upon which critical humanism rests: the theory is founded on practice (repeating Philonenko’s reading of Fichte). This practice itself is founded on the juridical (Renaut’s contribution in Le système du droit). But the juridical has the same structure as aesthetic judgments (according to Renaut and Ferry). It follows that all theoretical, ethical, juridical, etc., judgments are aesthetic judgments placed under the sign of “everything happens as if?’ All the faces of human rationality are reduced to the single component analyzed in the Critique of Judgment. Truth must be replaced by meaning, whose model is found in art. “Valorization” (in the sense of Wertlehre) must not be understood in a conceptual or determinate sense but in an aesthetic sense.

To be sure, I could not dispute this philosophical reconstruction on the pretext that it cannot produce a conceptual difference between a moral judgment and an aesthetic judgment. Indeed, the demands of individuals, of the common conscience, of the citizen (who would evidently prefer to see Nazism condemned and to see every human being’s humanity founded in truth) cannot be taken into account here. The metatheoretical claim “Notions like the Good, Humanity, etc., have no absolute foundation” is philosophically acceptable. It is thus obviously imperative, to preserve the distinctness of philosophical questioning, to distinguish the demands of a given empirical conscience and the philosopher’s point of view. Philippe Lacoue-Labarthe, in reference to Heidegger’s membership in the Nazi Party, had to forcefully recall, in this respect, the distinctiveness of the philosopher’s question that asks, “‘From what position can we judge?’”²² Who, qua philosopher, has not “considered [it] an accepted fact, and beyond question, that being a Nazi was a crime[?] This is something one may argue politically. I do so myself.”²³ And indeed, we cannot mix different levels of argumentation and judge a philosophical discourse on the basis of criteria external to that discourse.

1. First of all, there is the thesis of ethical relativism—which a consistent skeptic must fully adopt in accord with his philosophical principles. But the necessity of a moral vision of the world is sometimes still affirmed by the critical humanists as a true perspective and not simply as a judgment of taste (which in all logic, however, they ought to say). Indeed, they rebut other philosophers in the name of this moral vision. Thus, because Heidegger was not able to put a moral philosophy into practice, he is condemnable: “It has scarcely dawned on them [Heidegger’s disciples] that Heidegger risked being led astray [i.e., into joining the Nazi Party] at all only to the extent that, confronted by Nazism, he had no ethical point of view consistent enough to allow him to immediately condemn it.”²⁴ Similarly, in his article “Les subjectivités: Pour une histoire du concept de sujet,” Renaut denounces Michel Foucault’s statements in the name of morality.²⁵ Renaut’s only arguments against Foucault’s position are the absence of “morality” to which his position leads, and therefore the reader can have the impression, as in the sentence cited against Heidegger, that Ferry and Renaut claim to refute other philosophical viewpoints with these simple value judgments.

2. Next, Ferry and Renaut do not emphasize, as Maimon does, the question of the status of discourse in Kant’s Critique of Pure Reason. Kant’s statements in the Critique of Pure Reason are, in their eyes, true. Finitude is indeed radical; objective knowledge consists in the relationship between a category and an intuition, through the productive imagination of representable figures. But there is a paradox here, for if the Critique of Practical Reason is written from the viewpoint of reflective judgments, from what viewpoint must the Critique of Pure Reason be read? How is an affirmation of the eternal truth of the Critique of Pure Reason compatible with this other assertion that aesthetic judgment is the model for a finite and human rationality? Are we so finite that we can’t even find a legitimate viewpoint from which we could assert that “we are finite”? Don’t we find in this reading of Kant the very process of self-refutation that I have brought to light in the other trends? Here again, the question of the status of the discourse attributed to Kant is what is at stake. If we are told that not only statements about the beautiful and the living are reflective but also the statements in the Critique of Practical Reason, what then are we to say of Kant’s statements in the Critique of Pure Reason, statements that demand the division into determinate and reflective? If we are to reconstruct the entirety of the critical project from the Critique of Judgment, then what is the status of the statements in the Critique of Pure Reason? Why are its statements the only ones that aren’t reflective? But if we admit that they are just as reflective as practical principles, then how can we not see that we’re dealing here with a very powerful skepticism, of which Maimon is the most important representative? Why don’t Ferry and Renaut declare themselves to be “critical skeptics” like Maimon, unless it is because they deny a skepticism that all their analyses nonetheless combine to bring about?

Therefore, the reconstruction of the critical project from reflection as it is deployed in the Critique of Judgment leads to the advocacy of a skepticism that is not far from that of Feyerabend, or even Rorty.²⁶ Furthermore, we are led to an aestheticization of statements that critical humanism nonetheless denounces in thinkers like Jean-François Lyotard. In a word, the movement that promotes the Wertlehre into the place of the Wissenshaftslehre clearly leads to the legitimization of skepticism.

Our study of this reading of Kant thus shows the extent to which this reconstruction of the critical project is run through by skepticism—denied even though present, and thus this reconstruction is also run through by a tendency to self-refutation. There is, however, a contemporary attempt to stand up to skepticism. It is even more important that we examine it, since—in contrast to all the approaches presented so far—it is the first to have conceived of pragmatic contradiction as a symptom of the crisis into which philosophy has entered, and the first to have spoken of a “reflexive deficit” in contemporary thought. Of course, I am referring to the work of Karl-Otto Apel.

The project of transcendental pragmatics is thus simply expressed: against the radical deconstruction of reason (whether this deconstruction be Heideggerian, skeptical, or fallibilist), it is a matter of taking up the Kantian challenge of an ultimate foundation; but against Kant, and thus with his contemporary critics, he must challenge the idea of an “a priori of consciousness,” of a solipsistic and sovereign substance.³²

Having defined this project, his implementation is also rather easy to grasp, and Apel summarizes it admirably in “The A Priori of the Communication Community and the Foundations of Ethics.” It shares the view of ethics as simultaneously “necessary and impossible.” Ethics is necessary because contemporary science, by the nature of its own products, calls for a clear understanding of norms (consider, in this respect, our bioethical committees confronting scientific technologies like, for example, cloning). But although it is necessary, over the course of the centuries ethics has just as much seemed to be impossible. This impossibility is born, Apel tells us, in the very concept of science as it was established in the seventeenth century in the West. Why does modern science forbid ethics? To understand this, we should “attempt to present the most important of [these] propositions”³³ that demand the idea of an impossible ethics from the age of science. Three propositions, according to Apel, are covertly included in the current conception of science:

In a word, facts are subject to a rationally motivated agreement; on the other hand, they point to an arbitrary choice, irreducibly contingent and individual. This gap between an objective science and a subjective ethics is not only supported by the most traditional positivism but also by apparently opposed philosophical movements like the existentialism of a Kierkegaard or a Sartre. Paradoxical but objective allies, they indeed insist on the irreducibility of moral choices. Thus, Apel writes:

From this analysis, Apel undertakes to bring down the myth of a value-free science. He shows that in scientific practice there are “always already”³⁸ presuppositions of a prescriptive and not descriptive nature. “I should like to reconstruct the ethical preconditions for the possibility and validity of human argumentation and, consequently, of logic.”³⁹ The sciences depend upon a number of clearly normative conditions like, for example, the prohibition against falsifying the results of an experiment, the presupposition of equality among participants in an argument, and even the rejection of all physical violence, implicitly included in the very act of wanting to convince with a demonstration, etc. Science implies a number of conditions without which it could not be practiced, but these conditions turn out to be ethical norms, not natural facts. Apel condenses the set of these “always already there” conditions (which I have already detailed when I discussed Habermas’s universal pragmatics—itself, on this point, a clear copy of Apel’s analyses) to a single phrase: “the community of argumentation.” An extension of what C. S. Peirce called “the scientific community”⁴⁰ and an echo of Kant’s “community of saints,” this “community of argumentation” holds for any discussion whether its purpose is scientific, ethical, political, artistic, etc. This “community of argumentation” is the presupposition—always already made as soon as I begin to argue—that the truth will be attainable, without violence, in an ultimate and universal consensus. When I begin to argue in the factual world, I have always already presupposed the possibility, in an ideal world, of a community of argumentation with all its prerequisites (sincerity, universality, equality among participants, etc.). But, it will be asked, why should we accept rational discussion? If obeying logical norms presupposes an acceptance of ethical norms, why should we be logical? To put it in Kantian terms, aren’t Apel’s ethical norms just simple hypothetical imperatives and not a categorical imperative that demands of morality a foundation in reason? If logic and science presuppose morality, are we only moral if we want to be logical? And why would we want to be?

Apel proposes to answer this formidable question—in which the very notion of an ultimate foundation of morality and its corollaries of an a priori or a “categorical imperative” are at stake—by first working out the central notion of a transcendental condition, that is, of a necessary condition of possibility.

Performative contradictions are by definition always false—not because of their meaning (as in the analytic proposition “every bachelor is married”) but because of the status of their utterance. In addition, as in logic, the contrary of a statement that is always false is necessary. So statements whose contraries cannot be said without self-contradiction would therefore be necessary. Within the category of performative contradiction, defined as statements that cancel themselves out by the fact of being uttered, two types of contradiction can be found: (I) There are “pragmaticoempirical” contradictions, which relate to the empirical conditions of a discourse. In this case, the contents of a discourse and the empirical conditions that make that discourse possible are performatively contradictory, as in the statement “I was on the boat that was shipwrecked with no survivors.” (2) There are “pragmatico-transcendental” contradictions, which relate to the conditions of the actual argumentation itself, for example when I say, “The truth does not exist.” This is the classic self-refutation of radical skepticism, the necessary suicide of the consistent sophist. No discourse can, without contradiction, argue for a refusal of argumentation. In a word, there is only one alternative: accept these transcendental conditions or condemn oneself to silence; to speak is to enter into communication, signifying that one has already accepted these conditions. This refusal of performative contradiction is what gives the “a priori factum of argumentation” its meaning, transcendental philosophy its necessity, and reason its ultimate foundation. On this point, Apel does not hesitate to compare his pragmatico-transcendental argument to the fundamentum inconcussum:

Before I turn to look at recent objections to Apel’s philosophy, I should underline the extent to which we are dealing with a “strong” version of “pragmatics.” Apel simultaneously radicalizes the insights of standard pragmatics (for example, Searle’s, for whom, as we have seen, while a number of constraints are implied in discourse, they are never given as ultimate conditions of possibility)⁴² and of the “universal pragmatics” of Habermas’s “second” period. The difference between the two authors can be seen in three terms that echo one another and are implicated in Apel’s reasoning but rejected in Habermas’s: “transcendental,” “a priori,” and “ultimate foundations.” Apel emphasizes this difference in declaring:

A sharp distinction is drawn between the transcendental (universal and necessary) and the historical (particular and contingent). This refusal to index norms to the course of history, to the lived world, to contingent forms of life clearly allows Apel to avoid the various oscillations of Habermas’s philosophy—all caused, as I have shown, by an ultimate recourse to history and to social reality. But for all that, is this “strong” foundation exempt from aporias and difficulties? We must now turn to this question as we inventory the various objections to Apel’s philosophy.

The most immediately apparent objection to Apel’s “foundationalism”⁴⁵ bears on the notion of individual liberty with regard to an actualization of transcendental presuppositions. Recourse to performative necessity could seem to make morality,⁴⁶ in the particular sense of ethical norms underlying discussion, inescapable and necessary. The example of the devil (a frequent figure in Apel), forced regardless of his intentions to obey ethical norms if he will engage in discussion, only serves to fuel this simple question: where in all this is an individual’s free choice? To answer this first classic objection concerning the claimed necessity implied by “transcendental pragmatics,” we should distinguish, with Apel, between objectively necessary actions and their subjectively contingent realization. Indeed, I can always refuse discussion. I could hit someone instead of arguing. Recourse to violence is still obviously a historical and human possibility, without which good would already have been achieved for eternity. To understand this as something other than an ad hoc distinction, we must recall the structure of Apel’s “system.” I use the term “system” on purpose, for, in contrast to Habermas, Apel’s philosophy is presented as a systematic enterprise. Indeed, from 1973 to 2005, whether he is in dialogue with Heidegger, hermeneutics, Searle, Rorty, or Hans Albert’s fal libilism, whether he is addressing first philosophy or more immediately concrete problems (like the education of children), he has always maintained the same structural orientation, which he summarizes in “Limits of Discourse Ethics?”⁴⁷ It is important

All his books, articles, and lectures exemplify this clearly systematic organization, which takes its inspiration, without inhibition, from the great reconstructions of German idealism, like Fichte’s, for example, who in The System of Ethics⁴⁸ just as in the Foundations of Natural Right⁴⁹ first asks about the necessary conditions of possibility of the field under study (“deduction of the general principle” of ethics or of right), then asks about the applicability of these conditions to concrete communities or to real historical processes. In its very form, Apel’s philosophy revives the Kantian demand for systematic organization and runs counter to the entire contemporary doxa, for not only does he not hesitate to use the notion of an ultimate foundation but he also retains a systematic structure for philosophy in which diverse elements are integrated into a unified whole. And this systematic structure, organized in two large levels (A and B), allows us to respond to the initial question. Indeed, for Apel, there is and always has been a difference between grounding ethical norms in reason and understanding their application to history. The foundation (A) is necessary; the application (B) is aleatory. The individual’s freedom of choice remains whole and historically conclusive. Here, Apel is back in a typically Kantian structure that leaves space for individual action and does not understand moral achievement as a historical necessity. For all that, Apel protects himself from falling into certain Kantian pitfalls relative to the question of principles’ applicability. That a principle must be applicable to a real historical process does not mean that we must go back to the impasses of “On a Supposed Right to Lie.”⁵⁰ We can recall that, under the pretext that any universalization of a lie destroys it as such (demonstrating that to tell a lie is not moral), Kant moves without mediation to the application of this maxim to a real situation and deduces that he must never tell a lie regardless of the contingent circumstances. We know the possibly horrendous consequences of this intransigence from a now-classic example: on the pretext that I must never lie, must I tell a potential murderer where to find his intended victim? In Apel’s eyes, Kant’s error here is due to the still too “metaphysical” character of his doctrine, which remains imprisoned in the philosophy of subjectivity.⁵¹ In order that ethics can be an ethics of responsibility and not merely one of belief, some sort of limitation for the ideal principle must be found. The principle, while it may be universal, is applicable to a complex and multiform reality; it therefore must not be applied without mediation as Kant had done. We can say that it is a matter, for Apel, of finding a schema or a mediator between the condition (the universal principle) and the conditioned (the historical situation). Not only his Diskurs und Verantwortung (1988)⁵² but also all his articles written in the 1990s about concrete ethics are consecrated to developing and illustrating this principle of limitation. Let’s briefly summarize this aspect of his work that addresses a too-often-heard objection, for clearing out erroneous objections will allow us to concentrate on the real ones. If we consider the example of lying, it is clear, for both Apel and Kant, that a lie as a universal principle is a performative contradiction. Its contrary “Do not tell a lie” is thus a necessary statement. Nevertheless, we should not absolutize a single concept of reason, and we should recognize that sometimes the only reasonable way to realize the principle is to have recourse to a strategic action directly contrary to the rational principle. In a word, there are historical situations in which telling a lie is better in conformity with reason than refusing to do so. Thus, in the case of a Nazi who is searching for a victim, I must exercise a counterviolence in the face of this real violence. We can see, in this case, that the mediating principle consists in attempting to eliminate the obstacles to a future application of the universal principle. The obstacle here is the Nazi regime that denies universality. To eliminate this obstacle, I must implement a strategic action, in this case, I must lie. This movement from a morality of belief (“Do not lie.”) to a morality of responsibility (“Lie, in light of the given historical circumstances.”) can only be accomplished on the basis of a dual refusal: refusal of an unavoidable accession of morality in history (a rather Hegelian schema), and refusal of the terroristic application of principles in the mode of Fiat justitia, pereat mundus⁵³ (a rather Kantian-Fichtean schema).

That said, having brought objective necessity and subjective action into agreement, without for all that reiterating Kant’s mistakes on the question of application (one does not move from the principle to its application without mediation nor without taking a given historical situation into account), Apel seems to be confronted by a much more serious objection, posed by a certain form of decisionism.

I have said that if we actualize a transcendental presupposition in argument, we still have the empirical possibility to refuse to argue. But what is the status of this refusal? Can’t we say, with Karl Popper, that the first commitment for or against reason would be, in the final analysis, irrational? Hans Albert would thus be correct in asserting, in his Treatise on Critical Reason,⁵⁴ that Apel’s project for an ultimate deduction of reason cannot exclude a decisionist moment. Reason can no more found itself on itself than the “hilarious baron” could get himself out of a swamp by pulling himself by the hair. The first, initial, original commitment in favor of reason would thus be itself without reason; it follows that the project of ultimate foundation would be condemned to defeat, the actualization of a categorical imperative would be null and void, and the very notion of transcendental necessity would be vain. Apel directly answers this strong argument by analyzing the presuppositions implied in the objection.⁵⁵ The principal presupposition is that the objector forgets that the first fact is the discussion and covertly supposes that a decision can be made before any discussion. But the factum of which Apel speaks, which he means to account for, is discussion and nothing other than discussion. But to say that a choice for reason would be preceded by an irrational decision, which would be its condition of possibility, has no meaning unless one first abstracts from the discussion. This surreptitious abstraction shows that Popperians sink into the most banal solipsism in implicitly supposing a subject prior to the discussion, before the intersubjective exchange, a sovereign and voluntary subject, deciding for or against reason. A pure metaphysical myth, this subject opting for reason or not, is in fact the answer to a question devoid of meaning after Wittgenstein and the “linguistic turn,”⁵⁶ namely, “What is there before discussion?” This question, the implicit presupposition of the objection, clearly does not enter into Apel’s philosophical horizon, as he takes care to specify that a decision in favor of reason is “the only decision possible” for “anyone who does not make this choice but instead chooses obscurantism, for instance, terminates the discussion itself and his decision is, therefore, irrelevant for the discussion.”⁵⁷ That Apel takes care to italicize the expression “for the discussion” shows the extent to which he is speaking only of an ultimate foundation relative to the project of accounting for the phenomenon of discussion, and shows how he is not at all invoking a foundation in itself, an original act, first decision, irreducible choice, the cause of itself and everything else. There is no choice stricto sensu, and if, to be sure, the devil’s choice is empirically possible (once again, I can hit someone instead of speaking with him), this choice is without meaning in accounting for the factum of discussion and for understanding its conditions of possibility.

With this answer, Apel shows the true nature of his project: when he speaks of an ultimate foundation, he remains within the order of discourse, because the framework of discussion is always presupposed. Too many objections are born from the abstraction from or the forgetting of this key point. As another illustration of this type of mistake, I can cite the objection made by both Sylvie Mesure⁵⁸ and Jean-Marc Ferry. They suspect Apel of moving unduly from the müssen of transcendental argumentation (I am constrained to accept such and such rules in order to argue) to the sollen of moral duty (I ought to will these rules as norms of my action). For Mesure and Ferry, the theoretical necessity of argumentation cannot under any circumstances ground the practical obligation, the duty to argue. But, as in the preceding objection, here again we should inquire into what Apel’s adversaries presuppose to be able to say what they say. From what point of view does the notion of an authentically moral obligation, the notion of sollen—in contrast to the only theoretical constraint of müssen—obtain meaning? It can only be from the point of view of a duty conceived of as separate from discourse, that is, from the point of view of an “obligation” in itself. Consequently, the objection against Apel is, again, formulated from the hypostasis of a sovereign and first subject who, beyond being constrained to respect the rules of argumentation, should, before any argumentation, will them as a duty. This objection has its origins in an absolutization of the moral subject, and of practical duty, that Apel unambiguously rejects when he writes, “I think that transcendental philosophy in general and ethics in particular may be grounded in a radical way by avoiding all metaphysical implications of the Kantian system—as, e.g., Kant’s unsolvable problem of a ‘transcendental deduction’ of the reality of the free will and hence of autonomous reason.”⁵⁹ Although he is always worried about an ultimate foundation, Apel’s philosophy is much less “absolutist” than are his Kantian adversaries. For Apel, the notion of a right intention has no meaning for the discussion. The good will is not only unverifiable but, moreover, it doesn’t really matter. This is why Apel reproaches Kant for having grounded an ethics of belief that still always secretly supposes a man’s good will would be recognized in its true value (by a God, somehow assuming responsibility for history). And indeed, if we analyze, for example, the demand for sincerity implicit in a given instance of speaking, we realize that the speaker’s real intention is not a pertinent parameter. From the fact that I speak, I am not at all committed to being truly sincere, but the act of discussion commits myself relative to the intentions that I have expressed. In this sense, the rule of sincerity, if it says nothing about the psychological state that the speaker ought to have, nevertheless constrains him to answer for his intentions publicized through language. This allows Apel to write that “it is not the good will that matters but rather that the good be realized.”⁶⁰ In this sense, from a strictly Kantian point of view, Apel’s undertaking appears more like a foundation for the law than a foundation for morality, inasmuch as he leaves to the side the notions of a free subject, a right intention, a good will, and even of wrongdoing (every breach of argumentation being interpreted as nonsense, as madness). In a word, it is only a matter, for Apel, of demonstrating that once I enter into a discussion, I have always already recognized that there are a certain number of rules immanent in that discussion. That is all that is meant by the project of a “transcendental” and “ultimate foundation” of “the ethics of discussion.”

Does this dissolution of the most current objections against the Apelian project suffice to give credence to it as a whole? Probably not, for a serious difficulty remains, an important problem, an apparently unsolvable aporia that is liable to mortgage this second great reconstruction of the “transcendental.”

The problem takes its origin in the following question: what status should be accorded to the notion of “being constrained to argue”? Unlike the preceding objections, this question is not asked with respect to an empirical choice—which Apel has always accepted—but with respect to the transcendental possibility to reflect upon the rules of discourse. Apel identifies this problem as the problem of “transcendental reflection.” When I find myself in the language game of discussion, I have the capacity “to go back over” this language game and, through analysis of it, to illuminate the rules that govern it. But how, precisely, are we to understand this expression “to go back over”? It cannot be a matter of reviving the Cartesian model of a reflection guided by a subject exterior to this discussion, since Apel rejects the solipsistic temptation inscribed in nucleo, in his eyes, in Cartesian reflection. To avoid any surreptitious revival of a reflection conceived as the movement of a sovereign subject that is outside and above the phenomenon to be analyzed (in this case, language and its rules), Apel hypothesizes that, in fine, language itself is what allows reflection on its own rules. To the question, “How is reflection on language possible in language?” Apel will answer, in Transformation der Philosophie, that language is simultaneously “the subject and the medium of transcendental reflection.”⁶¹ Furthermore, he will go so far as to show that self-reflection on the use of language is the condition of possibility for participation in a discussion.⁶² Thus, language produces in and of itself the demonstration of its functioning. Language would be the object, the means, and even the authority for reflection. To put it in yet another way, “the principle of discussion [is] established through reflection,”⁶³ and this reflection itself is allowed by language.

To better grasp this definition of a transcendental reflection automatically produced by language itself, we should go back to Paul Ricoeur’s clarifying distinctions. In Oneself as Another, Ricoeur takes care to distinguish between “sui-reference” (unique to language) and self-reference, the act if not of a subject in the Cartesian sense at the very least of an enunciative authority in Emile Benveniste’s or G. G. Granger’s sense. In his analysis, Ricoeur notes first of all that if semantic inquiry (represented by the first wave of analytic philosophy) was interested only in an utterance’s referential dimension, pragmatic analysis on the other hand (encouraged by the second wave) was able, by means of the speech act, to become interested in its reflexive or self-referential dimension. Initially only a simple “complication along the path of the reference”⁶⁴ (or even an “obstacle”⁶⁵ to its veritable access), self-reference is later considered to be part of every utterance that refers to something other than itself (in the statement “Snow is white,” the statement refers to a fact that is not of language) and simultaneously refers to itself (in the sense that a statement marks itself as a statement and not as the thing itself). A sign says something (reference ad extra) and proclaims itself as a sign (self-reference). Ricoeur insists on the difference between this self-reference, a very fact of language, and another, produced by the act of an “I.” This “I” must be understood not in the metaphysical sense of a material subject but as an act of utterance, which Benveniste has shown to be the first “indicator”⁶⁶ of any utterance. Contrary to Benveniste, pragmatics as François Récanati promotes it has more than a tendency to want to eliminate the speaker and to make the speech act make due with the utterance itself. Ricoeur underlines this point, “The reflexivity in question up to now has been constantly attributed, not to the subject of utterance, but to the utterance itself … Récanati … relates reflexivity to the utterance considered as a fact, that is, as an event produced in the world.”⁶⁷ In this framework, the utterance itself is in fact what, reflecting itself, produces the demonstration of its functioning as of its identity. This conception of “sui-reference,” distinct from a “reflexivity” conceived as an act of utterance, is what Karl-Otto Apel develops on the whole. Indeed, reflection, immanent in linguistic phenomena, is a fact of language itself. Language or speech (die Sprache) is the “theme” and the “medium” but also the vehicle and the condition of possibility for reflection.

And yet this conception of reflection cannot avoid giving rise to several paradoxes:

First of all, the question arises as to how, if language itself produces the possibility of reflection, philosophical mistakes about language have been and are still possible? Why, since the authority for reflection is a process automatically generated by use, can we be mistaken in bringing to light the rules that govern communication? To put it differently, reflection on language must allow, Apel tells us, the construction of a critical theory of meaning, a theory that must supplant the classic but obsolete theories of knowledge from before the “linguistic turn.”⁶⁸ But if this reflection is done by language itself, why would there have been anything other than this contemporary critique of language? How can language reflect upon itself today while yesterday it was misled? We can see the wide gap between these temporal markers (“today, the truth,” “yesterday, errors”) and a thesis that implies that it should hold for all eternity.

Furthermore, this conception of reflection immanent in language implies a certain form of essentialism that makes it difficult, or even impossible, to explain philosophy’s role in this system. What is Apel doing when he writes these books that reveal, bring out, and show the rules that govern our utterances? Is it a simple echo, reflection, mirror, or container for a language that speaks itself? Apel has never maintained such a thesis, though his choices about reflection seem to call for it. In a word, his “immanentist” conception is cornered by questions of a Spinozist, Hegelian, or even Heideggerian nature, but he nevertheless refuses to ask these questions that his system provokes. If the authority—whether this authority be God or nature (Spinoza), Spirit (Hegel), Being (Heidegger), or language (Apel)—reveals its own identity, speaks itself through philosophy, then not only should the possibility of error, of negativity, be explained (objection 1) but also how Being, substance, Spirit, etc., speaks itself through philosophy (thus for Spinoza substance speaks itself through the modes and then the accidents that are thinking individuals). And yet Apel clearly refuses to close off his system with this type of consideration, to which any form of immanentism leads. Stricto sensu, language speaks itself and thus reflects upon itself without an external intervention from the philosopher. Nevertheless, there are many passages in which Apel denies this implication of his system and falls back upon a classical conception of the philosopher’s critical reflection. For example, he explains that participation in the use of language can be interiorized and distanced by reflection. If language speaks itself, then talk, with respect to reflection, of a process of distancing oneself from language has no meaning, for such a proposition implicitly postulates an inquiring philosopher who oversees language and its rules. Rüdiger Bubner has noted this ambiguous, even impossible, status of philosophical reflection in Apel, in a sense different from but analogous to mine:

The second objection is directed at the lack of a place for the reflection called transcendental itself. The interaction community of the partners participating in the language-game forms the point of departure, at which the character of the dialogue, as something which encroaches upon the individual subjectivity, is emphatically set up as a corrective against Kant’s alleged narrowing of vision to the isolated ego. The ideal norm of mutual recognition of subjects forms the vanishing-point towards which the transcendental reflection is to be orientated. The activity of transcendental reflection for its part, however, which is supposed to mediate between the given point of departure and the counterfactual assumption, does not fit at all into the assumed picture. The communication community does not reflect on itself consistently, as a kind of collective subject, with a view to the unalterable presuppositions for its existence, but a philosopher approaching from the outside points to certain normatively characterized premises on the basis of the various language-games, only loosely connected by family resemblances. This act of external elucidation and criticism cannot however be called “transcendental” even in the most generous interpretation of the term. The philosopher, as privileged subject, is not by any means associated by virtue of reflection with what is reflected upon but speaks externally about it from a special meta-position … We are faced with the problem of the non-identity of the reflecter with the reflected.⁶⁹

In a word, the question of reflection as self-reference leads Apel’s system to a quite ruinous alternative: Either (1) there is a deficit of explanation and justification, for if language speaks itself, if the philosopher’s analyses are the reflection, mirror, or echo of linguistic mechanisms, then Apel must explain this possibility, as Hegel, Spinoza, and even Heidegger have done. Or else (2) Apel accepts the idea of a critical philosopher who uncovers, through his analyses, the rules that govern language, but then his system, in fine, contradicts itself.

Having developed this objection, I ought next to better specify its status. It is not a matter of demanding, through this objection, some sort of return to the subject of reflection, in the sense of a metaphysical substance or an authority that is prior, exterior, and superior to language. It is quite simply a matter of asking about the status of the philosopher’s act of speaking. With his critical theory of meaning, Apel presupposes that the philosopher has a certain position (in the example given above, he presupposes the possibility of putting distance between oneself and the object, “language”); but this “philosopher’s position” comes into conflict with the thesis that language speaks itself, reveals itself, and shows itself. In contrast to the objections that I’ve just discussed, I am thus not calling for a substrate of reflection but posing a question about the status, within Apel’s discourse, of the philosopher’s utterances. Likewise, I do not mean to reproach pragmatics, as Ricoeur, for example, does, for its “depsychologized”⁷⁰ conception of reflection. Nor do I mean to claim, in a way external⁷¹ to Apel’s system, that it is not language but concrete subjects, subjects of flesh, individuated and anchored in eternally unique places in time and space, that reflect. My problem is not to determine what the “subject” of reflection is (whether this subject be conceived as a universal authority or as a finite subject, a “self” forever shattered by theoretical finitude or practical wisdom’s own body) but to understand how Apel can consistently account for the status of his own philosophical analyses. There is indeed an ambiguity, internal to Apel’s philosophy, in the term “reflection,” which sometimes designates linguistic self-reference and sometimes designates the distancing act of the critical philosopher who brings to light the rules of discussion.

The aporia of metalanguage that burdens Apel’s system prevents me from following his reconstruction, from leaning on his philosophy. We have seen a characteristic movement in the two extreme styles of criticism today—a pendular oscillation between a skeptical reading of an author (the reconstruction of Kant in light of the Critique of Judgment) and a reading that is, if not positivist,⁷² at the very least more foundationalist and essentialist, in which a unique and first authority (here, language) becomes the theme, the medium, the vehicle, and the foundation of all things. Having shown this pendular movement, I could of course take up the second reading, since it accepts philosophy as a first, autonomous, and distinct science. But Apel’s system contains a tension between the general thesis about language and the philosopher’s presupposed position, a tension between what the philosopher “says” (language reveals itself) and what he presupposes in order to say it (the philosopher puts himself at a distance from language to be able to extract its rules).

Nevertheless, I should underline that if there is a tension in Apel, it does not immediately refute⁷³ the whole of his project, as did the contradictions of radical skepticism or a phenomenology inebriated by the singular. In addition, we must remember that Apel was the one who demonstrated that the notion of performative contradiction was the necessary fulcrum of logicophilosophical reflection. And yet I have shown that this contradiction runs through all the contemporary strains of thought that seek to be done with the classic stance of philosophy as a first and distinct science. It follows that Apel’s philosophy can undeniably easily spot this contradiction in different discourses. Finally, it was Apel who showed the importance that consideration of self-reference has taken on today, in contrast to a polarization only in terms of reference. Apel has contributed much to open the way for pragmatics to be able to alter semantics’ weaknesses and dictums, and for it to revitalize philosophical analysis from top to bottom.

However, with this legacy Apel also bequeaths to us some formidable difficulties. First of all, there is an ambiguous concept of self-reference, because he juxtaposes two concepts: the “sui-reference” of language and a “speaker’s reflexivity” (here, the philosopher’s).⁷⁴ To avoid confusion, it seems to me that we must carefully distinguish three senses of self-reference, which each belong to different conceptual fields: (1) “Reflection” (and even self-reflection) as in the classic Cartesian model, which always defines the movement by which a subject (prior to this movement) looks into himself as if looking into a preexisting x. This model of self-reference as reflection revives what we can call, following many others, the philosophy of the subject. (2) “Sui-reference,” a property of signs, by which a sign says or indicates that it is a sign and not a thing. The Port-Royal logic defined this type of self-reference. According to that analysis, a sign refers to a thing (“it represents something else”) but simultaneously marks itself as not being that thing (“it represents itself”). In the final analysis, this “sui-reference” of the sign, which Récanati explains, is what defines self-reference in Apel. (3) Finally, there is what we can call, with Ricoeur, Benveniste, and Granger,⁷⁵ “reflexivity.” This kind of self-reference refers to an act (initially understood in the pragmatic expression “speech acts”)⁷⁶ and introduces the problem of the speaker. In a word, we have a generic term, “self-reference,” that can be understood in at least three ways: the “self-reflection” of a subject, the “sui-reference” of language, and the “reflexivity” of an utterance.

But, if we consider what my objections to Apel have established, the last form of self-reference, reflexivity, must be understood as the philosopher’s utterance. Indeed, Apel’s problem emerges from the philosopher’s position or status. In the same way, the contradictions that I have already pointed out in other philosophical currents also reflect this problem of the status of the philosopher’s discourse. This is why I do not at all mean, in this provisional determination of self-reference as a possible reflexivity, to refer to an ailing and concrete subject, such as Ricoeur’s “shattered cogito.” My question is not to ask what it means to say “I”⁷⁷ in a world that is always already there, at the heart of an eternally finite existence, but rather to ask, What is a philosopher? What is presupposed by this type of utterance? What are we saying when we speak of “philosophy”? Is the philosopher’s position tenable? If not, why not; and if yes, under what conditions?

This analysis also allows me to situate my questions with respect to Apel’s. I have said that I cannot follow Apel because of the tension in his conception of self-reference, but moreover, his initial project is not what I intend to accomplish here. Indeed, Apel begins with the problem of grounding ethical norms in reason. My question is about the foundations of philosophical practice, about the possibility of philosophy (always with another question on the horizon: if philosophy is not possible, why not; and if it is, under what conditions?). Our projects’ different orientations will allow me to conclude my discussion of Apel by bringing other aporias of his system to light, aporias that I was not able to discuss in the course of my analysis but that are nevertheless important for my questioning with respect to his.

Apel wants to ground ethical norms in reason—but why this project? Of course, one might retort that this question can be asked of any project at all. Nevertheless, there are cases in which a failure to answer the “why?” question does not entail the disappearance of what one wants to understand, but sometimes the impossibility to answer why requires an abandonment of what one wanted to understand. For example, an inability to ground ethical norms in reason does not imply the disappearance of what is considered ethical behavior—the “peasant of Savoy,” so dear to philosophers, can continue to behave morally even if Hume is correct. Furthermore, the possibility that ethical norms are, in fact, simple customs or pure illusions is not in itself self-contradictory. On the other hand, to say that philosophy’s relevance and uniqueness cannot be established entails the disappearance of the discipline. Likewise, to do philosophy for the sole purpose of denouncing the inanity of such a practice is an immediately self-refuting position. This is why we can reproach Apel for not having sufficiently elucidated the necessity for a project according to which “ethical norms should be grounded.” To be sure, he shows us that the current situation of the world demands a morality, but notwithstanding that a given empirical and historical situation could demand many kinds of grievous things (for example, the creation of a thousand-year Reich), this empirico-historical situation cannot have the force of a proof within Apel’s epistemology. It follows that Apel cannot legitimate his project to “ground ethical norms” in reason, whereas I can legitimate my project because if it is impossible to answer my question, then the discipline that I practice is invalidated.

Moreover, we are entitled to ask if it is really “morality” that Apel grounds. I have shown in the course of this analysis that his system as a whole leads rather to a foundation in the legal sphere because, in a discussion, the devil is constrained to behave “as if” he were moral—that is, to take up the Kantian categories, “in conformity with duty” (law) and not “from duty” (morality). Why then does Apel insist on speaking of ethical foundations, since by this term he means what Kant termed law? ⁷⁸ In reading his challenge to notions of the will, right intention, and autonomy, a reader can get the impression that Apel is one of the fiercest destroyers of Kantian morality. If his reference to Kant can be justified from the viewpoint of transcendental argumentation (as a regressive investigation into the conditions of possibility of a factum), on the other hand it is hard to see what is left of the substantive content of the Critique of Practical Reason (which still defines what should be understood by Kantian morality). In a word, why not do without reference to Kantian morality?

Finally, concerning the factum rationis itself, we are entitled to wonder whether Apel does not effect an absolutization of the language game of argumentation. Indeed, argumentation becomes a synonym for all discussion, all communication. It is on this basis that Apel can establish that we are always already moral (at least following ethical norms such as not to “falsify your results,” or “claim universality,” etc.). And yet we can wonder about the legitimacy of this reduction of all communication to the rules of argumentation. Without speaking of strategic actions, one need only mention other discursive registers, such as poetry, irony, or artistic expression in general, which immediately seem to escape the constraints of argumentation.⁷⁹ The question arises whether, in the final analysis, Apel has unduly stretched Peirce’s idea of the “scientific community” to encompass all linguistic situations. A good number of his proofs seem to hold for only those who are engaged in discourses meant or claimed to be scientific or philosophical but don’t seem to hold for other discourses, like poetry or irony. It is important to note this point insofar as the notion of “claiming to” is what allows a performative contradiction to take effect. Also, we should immediately note, pace Apel, that certain constraints hold only for certain kinds of discourse, like the philosophical discourse, which is all I’m concerned with in this book. Having noted these final details about Apel’s philosophy, I should now sum up what we have learned.

1. Any series of propositions that advance in this direction necessarily produces self-refuting arguments. I have shown this in every position we’ve examined. A good portion of philosophy today is marked by a specific kind of paradox, performative contradiction, in which a statement is destroyed by the very fact of being uttered.

2. We’ve also seen an oscillation between skepticism and positivism, relativism and scientism, within the approaches we’ve examined. This oscillation is not a movement in time from one extreme to the other but is often the movement from a quasi premise to its consequences within a single philosophy.

3. The propensity to self-refutation or self-contradiction is most often born from a lack of reflection upon the status of the maintained discourse (what must be presupposed in order to say that there is no truth or that philosophy should disappear or that it is only therapy or even that there is no other use of language than ordinary usage, etc.?). What seems to be lacking is this “reference to self,” this questioning of one’s own status—or, to put it in phenomenological terms, as Jocelyn Benoist has done, a look is not only for seeing but also for telling itself that it is a look and that “in the fact that a look has something to say to itself and in the way that it can do so [there is] something not at all obvious, but extremely problematic.”⁸⁰ We have seen that this concealment of the question of the status of one’s own discourse, or “reflexive deficit” in Apel’s sense, holds for a good number of the cases we’ve considered—all taking place as if contemporary philosophy, in the multiplicity of its manifestations, shared a single presupposition, a refusal of the question of self-reference.

Calling the distinctness and autonomy of a discipline into question—this is the general picture with which we began. I have been able to show the recurrence of a particular kind of contradiction (self-refutation) at the heart of this picture, born from neglect of the notion of reference to self as reflection on one’s own status—this is the mark that specifies the landscape I’ve analyzed, this is the figure or design that accompanies and defines, at the same time as it deconstructs this landscape.⁸¹

Consequently, the choice is simple and the challenge is clear: Either we rally to this questioning of philosophy while trying not to commit a pragmatic contradiction, that is, while refusing this paradoxical stance that consists of spending one’s life practicing a discipline for the sole purpose of saying that it is dangerous, harmful, and vain. In that case, the only and unique consistent position is an abandonment of philosophy, relegating it to the dustheap of obsolete disciplines, like astrology (which Kepler still practiced but in the course of time lost any scientific connotation). Or else we must understand the possibility of philosophy as philosophy, that is, as a discipline that is neither philology, nor literature, nor mathematics, nor “the reserve angel of jurisprudence.” To take up this challenge, we must undertake a thought experiment and assert, as a hypothesis, the opposite of what the evidence of our time seems to dictate, namely that