Karl-Otto Apel, "The Problem of Philosophical Foundations in Light of a Transcendental Pragmatics of Language"

The Problem of Philosophical Foundations in Light of a Transcendental Pragmatics of Language

4. Philosophical Foundations via Transcendental-Pragmatic Reflection on the Conditions of Possibility of the Intersubjective Validity of Philosophical Argumentation

Before I make a final attempt to show the indubitability of certain paradigmatic evidence in the language game of philosophical argumentation, I would like to settle the question of whether, and to what extent, the principle of fallibilism is also to be employed in philosophical argumentation.

First, it should be noted that even logico-mathematical deductions are fallible, trivially, inasmuch as with regard to their pragmatic dimension they are the operations of finite men and thus can go wrong. In this respect I can gladly concede to critical rationalism that human reason in the psychological sense is always fallible. The philosopher too can never be certain that he isn't in error. But this does not mean that it makes sense to assume in either the enterprise of argumentation or that of criticism that one is always in error, or that the use of the concepts “argumentation” and “criticism” could be meaningful without indubitable presuppositions. More important than the empirical pragmatic concession that even logico-mathematical thought is always fallibilistic is the transcendental-pragmatic insight that the metalogical or metamathematical demonstrability of the absence of contradictions in axiomatic logico-mathematical systems remains in principle incomplete. Earlier, following Hans Lenk, I allowed this to be one aspect of the Münchhausen trilemma of deductively justified philosophical foundations. At the same time, however, I pointed out that this does not at all reduce the question of reflexive foundations to absurdity, but rather actually raises it. In the present context I want to appropriate the insight into the incompleteness of all demonstrations of an absence of contradictions as an aspect of reflexive transcendental-philosophical insight that is, of insight into both the conditions of the possibility and the limits of the objectification of arguments in axiomatized and formalized language systems.

It is difficult to conceive how this insight from the extended critique of reason could be revised in its transcendental-philosophical core. Nevertheless, one never knows definitively what belongs to the transcendental philosophical core and what belongs to the complex of results that are revisable through advances in metamathematics or metalogic. To that extent, the transcendental-pragmatic interpretation of the results of metalogic or metamathematics can give an indication of the general problem of transcendental philosophy, whose situation is here somewhat different than in Kant. The Kantian claim of the definitive completeness of the “system of pure reason” can no longer be sustained; our task is rather that of progressively opening up transcendental horizons, which grow wider with the expansion of the human knowledge that we are questioning as to conditions of possibility. It in no way follows from this, however, that the principle of fallibilism, and the principle of virtually universal criticism derived from it, could show the absurdity of the postulate of transcendental-philosophical foundations or function as its replacement.

That this is impossible is shown by the fact that the application of the principle of fallibilism itself leads to something similar to the “liar paradox.” If the principle of fallibilism is itself fallible, it is just to that extent not fallible, and vice versa. This application of the critical-rationalist principle of fallibilism to itself can hardly be rejected as meaningless by critical rationalists; for it is precisely they who absolutized the methodological principle of fallibilism beyond its original application to empirical science. In my opinion, it follows from this, with all the clarity one could wish, that pancritical rationalism represents an untenable standpoint, or at least an exaggeration. The principle of fallibilism and the principle of criticism derived from it are meaningful and valid only if they are restricted in their validity from the outset, so that at least some philosophical evidence is excluded from possible criticism namely the evidence on which these principles themselves are based. In this way the transcendental-pragmatic dimension of the noncriticizable conditions of possibility of intersubjectively valid philosophical criticism and self-criticism is brought out in a sufficiently radical way. What are these conditions? In this question, I believe, is posed the problem of philosophical foundations.

That the principle of pancritical rationalism does not belong to the uncritical conditions of possibility of philosophical criticism can be seen in the successful self-criticism of pancritical rationalism by its founder, W. W. Bartley. Bartley found that logic manifestly does “not also belong to that totality. . . which should be subject to proof,” since “the exercise of critical argumentation and logic are inseparably bound together.”⁵¹ In critical discussion with Bartley and Albert, Hans Lenk made Bartley's observation more precise. He stated that “at least some logical rules are fundamentally removed from rational revision.”⁵² Still more interesting to me is Lenk's remark that the stated rules of some minimal logic are removed a priori from criticism because they are analytically bound to the (idea of the) institution of criticism itself.⁵³ Thus, the rules of a minimal logic are seen as belonging to the paradigmatic evidence of that institution or language game, which can only be disclosed in transcendental pragmatic reflection upon the conditions of the possibility of criticism itself. I will call this institution the transcendental language game. Concerning this language game, the previously cited insight of the later Wittgenstein is emphatically valid: this language game as a “system” belongs “to the essence of what we call an argument”: it is, so to speak, “the element in which arguments live.”⁵⁴ Laying bare this system of argumentation in transcendental pragmatics provides philosophical foundations in a nondeductive manner, since its paradigmatic evidence is precisely of a kind that can neither be called into question by criticism without performative self-contradiction nor be justified deductively without presupposing itself. The contemporary discussion of the problem of foundations, which is usually oriented toward axiomatic systems of logic, would certainly interpret this situation differently: it would be seen as showing that some evidence—for example, the logic of justification itself—can neither be denied without self-contradiction nor be justified without petitio principii. Therefore, it is said, ultimate foundations must be superseded by some ultimate decision—somewhat in the sense of the self-confidence of reason as opposed to skepticism (Stegmüller)⁵⁵ or in the sense of belonging to the institution of critical discussion as opposed to obscurantism (Popper).⁵⁶ This “solution” to the problem of philosophical foundations clearly corresponds again to the purely logically derived Münchhausen trilemma in Albert's sense, if one disregards the fact that Stegmüller understands “appeal to evidence” not as “appeal to dogma” but rather as a necessity of all philosophizing that cannot be denied without self-contradiction, but that cannot be logically demonstrated without a petitio principii.

In light of our transcendental-pragmatic reflection, however, the problem as presented by these philosophers is the result of absolutizing the objectivization and externalization of argumentation that is presupposed by the axiomatic method—that is, the “estrangement” of arguments into syntactic-semantically interpreted sentences and systems of sentences. The analysis of such sentences abstracts from the transcendental-pragmatic dimension of the self-reflection of the arguing subject. Under this abstractive presupposition all paradigmatic evidence of the transcendental language game (such as, for instance, the validity of a minimal logic) must, of course, have the status of unprovable presuppositions of any proof. And the attempt to justify the necessity of such presuppositions now must look like some bad sophistic attempt at a proof by begging the question; for on the abstract level of an axiomatized sentence system there is no difference between arbitrarily chosen presuppositions and presuppositions that one must presuppose in all possible proofs because one cannot deny then without actual self-contradiction. Thus philosophy seems doomed to resignation in regard to the problem of foundations. As Bar-Hillel argues, the logical semantics of sentences and sentence systems can only indirectly clarify argumentation in ordinary language, which is pragmatically integrated in principle; this sort of clarification is based on an abstraction from the pragmatic dimension that has to be overcome if the significance of axiomatic systems is to be brought to bear in the context of argumentation.⁵⁷ Therefore, the reduction of the meaning of philosophical foundations to the deduction of sentences from sentences (or to the metalogical proof of an absence of contradictions in certain sentence systems) appears to me to be an illegitimate reduction resting on an “abstractive fallacy” that pervades recent, purely syntactic-semantically oriented logic of scientific inquiry. When this logic is absolutized into a philosophy of argumentation, it commits an abstractive fallacy; it banishes to the domain of empirical psychology the pragmatic dimension of argumentation that cannot be objectivized and formalized (for instance, the self-reflection of participants in argumentation as expressed in performative acts of asserting). The discussion of the impossibility of providing philosophical foundations rests on a confusion between argumentation as originally related to the dialogic situation of assertion and refutation—which was for Socrates the basis of philosophizing—and Aristotle's apodictic science, which, however, could only be an organon of argumentation purified of all possible pragmatic intrusions.⁵⁸

If, however, this abstractive fallacy is reversed by admitting transcendental-pragmatic reflection upon the subjective-intersubjective conditions of the possibility of intersubjectively valid argumentation, then the problem of philosophical foundations appears in a completely different light. The insight that certain evidence cannot be deductively grounded without presupposing itself (for example, the paradigmatic evidence of a minimal logic in the framework of an as yet unclarified transcendental language game) is no longer a proof of the impossibility in principle of philosophical foundations but rather a reflexive, transcendental-pragmatic insight into the uncriticizable foundation of argumentation itself. If, on the one hand, a presupposition cannot be challenged in argumentation without actual performative self-contradiction, and if, on the other hand, it cannot be deductively grounded without formal-logical petitio principii, then it belongs to those transcendental-pragmatic presuppositions of argumentation that one must always (already) have accepted, if the language game of argumentation is to be meaningful. One can, therefore, also call this a transcendental-pragmatic argument for foundations on the basis of critical arguments concerning the meaningfulness of certain practices.

As far as I can see, this reflexive transcendental-pragmatic argument for philosophical foundations is confirmed by a critical yet affirmative reconstruction of the argument of Cartesian doubt. In this way it can be shown, for example, that Descartes unreflectively undermines the possible significance of the language game he uses when he grants, in the course of his methodological doubt, that in the end all that is supposed to be real might be merely his dream, viz., merely in consciousness. If all that is supposed to be real is merely a dream, then precisely the critical significance of the expression “merely a dream” (or “merely in consciousness”) cannot be maintained, since it presupposes as paradigmatic evidence that all is not merely a dream (or merely in consciousness). However, this pseudo-argument, which manifestly rests upon Descartes's illegitimate abstraction of the methodical-solipsistic search for evidence from the a priori of the language game of argumentation, can be revised, as it was by Peirce and Popper, into a virtually universal doubt (viz., the principle of fallibilism). If one undertakes this correction, then the proper significance of the Cartesian doubt is revealed, in that the certainty of the “dubito, cogito, ergo sum” can also not be doubted in the sense of the virtually universal doubt of all that is supposed to be real. What is the basis for the certainty of the “cogito, ergo sum”?

It cannot rest upon the fact that (in the sense of logical semantics) a syllogistic inference is made from thinking to the existence of that which thinks, as Hintikka showed in 1963, using the conceptual apparatus of Austin's speech act theory.⁵⁹ Descartes himself repeatedly denied that the cogito was based on such an inference. Hintikka, however, explicitly states the reason why such an interpretation is inadmissible: in the use of a syllogistic inference from thinking to the existence of that which thinks, the existence of the thinking being must be tacitly presupposed in order to reject the thinking of a fictitious person (say, Hamlet) as irrelevant. In other words, the certainty of the “cogito, ergo sum” cannot be logically demonstrated in any direct way. In this sense Descartes does not supply any philosophical foundations that could be affirmatively reconstructed. That the same person who thinks also exists is, from the viewpoint of formal logic, a claim that, in the sense of the Stegmüllerian dilemma, can be neither denied without self-contradiction nor demonstrated without petitio principii; for it cannot be made in the case of a fictitious person such as Hamlet, but rather only in the case of an existing thinker. For just that reason, however, the certainty of the “ego cogito, ergo sum” is a transcendental-pragmatic condition of the possibility of the language game of argumentation in our sense. How can this be shown? As Hintikka demonstrates, that my doubting or thinking guarantees my existence rests upon the fact that when I perform the act of doubting my existence—an act that is explicitly expressed in the sentence “I doubt herewith, now, that I exist”—I refute the sense of that very sentence for myself and, virtually, for every dialogue partner.⁶⁰ In other words, the propositional component contradicts the performative component of the speech act expressed by that selfreferential sentence. The irrefutable certainty of the “cogito, ergo sum” thus rests not on an axiomatically objectifiable deductive relation between sentences, but rather on a transcendental-pragmatic reflexive insight mediated by the actual self-reflexivity of the act of thinking or speaking.

Hintikka remarks in addition that not only is the assertion “I do not exist” refuted by the thought or speech act that is its performance, but this is also true of the assertion “You do not exist.” I would explain this as follows: Someone who used such an expression in, say, an exorcism aimed at a ghostly apparition would in truth not be denying existence to an object by means of an act of predication; rather, he would be canceling the expression of address, that is, he would be reflexively designating his communicative act as failed. I prefer to see in this an indication that the irrefutable certainty of the “ego cogito, ergo sum” does not rest upon the primacy of “inner experience,” the “introspection” of an in principle solitary consciousness, as is assumed in the Cartesian theory of “evidence,” right up to Franz Brentano; rather, it rests upon the primacy of an experience of the situation that is simultaneously communicative and reflexive, an experience in which actual self-understanding (and with it ego-consciousness) and understanding the existence of the other are equiprimordial—as is convincingly maintained by Mead and Heidegger. Confirmation of personal existence in the performatively understood “ego cogito, ergo sum” is only possible as an understanding with oneself about oneself, and that is to say, as part of a virtually public discussion—more precisely, as the deficient mode of such a discussion in which I am the other for myself. It is precisely this virtual publicity that is attested to in the fact that reflexive self-certainty can be made explicit with the help of a performative speech act.

Therefore, the certainty of the “cogito, sum” is not, as Husserl would have it in Cartesian Meditations, to be understood in such a way that it cannot be formulated in the “communicative plural.”⁶¹ In such an epoche of “methodological solipsism,” in which the existence of other subjects is bracketed along with the real world, the evidence of Cartesian insight could in principle not be formulated in the sense of an intersubjectively valid philosophical judgment. Everyone of us can see, with subjective evidence and with an a priori intersubjective claim to validity, that he cannot doubt the existence of his own ego without actual self-contradiction.⁶² Unless Husserl could somehow formulate this statement in the “communicative plural,” he could not bring to our knowledge the results of his transcendental reduction or epoche—that is, the insight, to which he gives the status of certainty, into the irreducibility of the sphere of the pure noetic-intentional, meaning-constitutive ego and its noematic correlates. This can be applied even more radically: Like Descartes, Husserl could not even bring to his own consciousness the indubitability of his ego-consciousness, in a form both intelligible and valid for him, unless he could formulate this insight as an argument in the framework of a transcendental language game of an ideal communication community. To sum up: Along with ego-consciousness, a language game is presupposed as the fundamentum inconcussum in the sense of the critically reconstructed and transformed Cartesian tradition of philosophical foundations. In this language game the existence of a real lifeworld and the existence of a communication community are presupposed along with the actual evidence of thinking myself as existing in the sense of paradigmatic language-game evidence. For it is of prime importance that the Cartesian insight (solitary as it actually is) must be capable of being reexamined and, in this case, also capable of being confirmed by a communication community that is in principle indefinite. This transcendental-pragmatic version of the Cartesian insight could be valid, in principle, in the form of an a priori certain and at the same time a priori intersubjectively valid judgment even for a man who happened to be the last representative of the communication community and thus was alone in an empirical sense. Even this man would have to presuppose (1) that there must have been a real communication community and (2) that there might be an unlimited ideal communication community, both capable in principle of confirming his certain insight.⁶³

From this I conclude that the “vital element” of philosophical arguments is a transcendental language game in which, along with some rules of logic and the existence of a real world, something like the transcendental-pragmatic rules or norms of ideal communication are presupposed. The individual can secure a priori certainty in the solitary thought of his existence only by appealing to this transcendental language game and its rules. This means, however, that the individual cannot step into or out of the “institution” of this transcendental language game of critical argumentation in the same way we suppose he can in the case of empirical “language games” and “institutions” as “forms of life” (Wittgenstein).⁶⁴ Rather, as a successfully socialized homo sapiens with “communicative competence,”⁶⁵ he is necessarily constituted as a being who has identified himself with the ideal communication community in the indicated sense and who has also implicitly accepted the transcendental-pragmatic rules of communication as ethically relevant norms. This is not contradicted by our capacity to bring to consciousness the discrepancy between the normative ideal of the ideal communication community and real situations of discussion. It seems to me that this suggests instead the possibility of finding the presuppositions for a transcendental-pragmatic grounding of ethics in the a priori of communication presupposed by rational argumentation—more precisely, in the contradiction (which cannot be resolved by formal logical means) between the presupposition of a real communication community (including our real selves) and the situation of an ideal communication community that is necessarily “counterfactually anticipated” in that presupposition.⁶⁶ To that extent, the “institution” of the transcendental language game turns out to be rather different from the conventionally based institutions of empirically describable “language games” or “forms of life,” in Wittgenstein's sense.⁶⁷ More accurately, the former institution could be characterized as the meta-institution of all possible human institutions,⁶⁸ since it involves the conditions of the possibility of transparent and rational conventions (“agreements”). Man can withdraw from this institution only at the price of losing the possibility of identifying himself as a meaningfully acting being—for instance, in suicide from existential despair or in the pathological process of paranoid-autistic loss of self.

Therefore—to draw the final conclusion—one cannot choose this rational form of life in an “irrational choice,” as Popper would have it,⁶⁹ since any choice that could be understood as meaningful already presupposes the transcendental language game as its condition of possibility. Only under the rational presupposition of intersubjective rules can deciding in the presence of alternatives be understood as meaningful behavior. From this it does not follow that every decision is rational, but only that a decision in favor of the principle of rational legitimation of the criticism of behavior according to rules is rational a priori. In that respect the decision in favor of the “framework” of critical argumentation or discussion demanded by Popper can only be understood as an a priori rational and deliberate affirmation of the transcendental language-game rules that are always already implicitly accepted as valid. Such a decision—which is even to be repeated again and again, particularly in “existential boundary situations” is indeed required in the interest of the realization of reason.⁷⁰ However, reason in no way needs to replace, through a decision, its rational justification, as is demanded by decisionism.

For it can always confirm its own legitimation through reflection on the fact that it presupposes its own self-understanding of the very rules it opts for. Popper's assertion that irrationalism can be defended without self-contradiction because one can refuse to accept the argument⁷¹is simply false, since the defense of irrationalism actually refutes the attempt to refuse to engage in argumentation—it refutes it, that is, through the accompanying performative act. The effective refusal to engage in rational argumentation (or a corresponding self-understanding) is on the other hand a very much more serious matter than Popper seems to assume; it is an act of self-negation and, moreover, of self-destruction, as I have already indicated.⁷²Even in such a case, however, the person making the decision must himself presuppose the denied principle so long as he understands his own decision as such. Otherwise, philosophical decisionism (upon which, in the final analysis, Popper's arguments for critical rationalism rest) could not treat the act of denying reason as an intelligible possibility of human choice.

With that I can summarize the issue at stake in this attempt at a metacriticism of critical rationalism. Critical rationalism cannot, it seems to me, succeed in putting the principle of criticism as such in the place of the principle of philosophical foundations, because its criticism of this principle—like every meaningful criticism—itself needs justification. Justification of the principle of criticism is, however, possible if and only if the principle is not absolute—if and only if it is restricted by the principle of the self-justification of critical reason through transcendental reflection upon the conditions of its own possibility.⁷³ The point of philosophical foundations lies, then, in the reflexive—transcendental-pragmatic and not deductive argument that one can discursively or practically decide neither for nor against the rules of the transcendental language game without these rules being presupposed.

55. W. Stegmüller, Metaphysik, Skepsis, Wissenschaft, 2nd ed. (Heidelberg, New York, 1969), p. 169.

56. K. Popper, The Open Society and Its Enemies (Princeton, 1950), p. 413ff.

58. An interesting example of the early anticipation of this confusion and of the modern reduction of philosophy to logical semantics is the following text ascribed by the commentator on Aristotle, Ammonius, to Theophrastus:

Since discourse (logos) has a twofold relation. . . one to the listener for whom it has a meaning, the other to the things concerning which the speaker wishes to produce in the listener a conviction, poetics and rhetoric exist with regard to the relation to the listener. . . however, the philosopher will be particularly concerned with the relation of discourse to things, by refuting the false and demonstrating the true.” See Ammonius, Aristotelis De lnterpretatione Commentarium, ed. Busse (Berlin, ] 887), p. 365ff. The logic of language in logical empiricism has revived this division by putting empirical pragmatics in the place of poetics and rhetoric. Since, however, modern linguistic analysis was preceded by the transcendental philosophy of the knowing subject, today we should be in a position to see that this division is incomplete with regard to the interpreting subject. The completion can certainly not be undertaken by a transcendental philosophy of consciousness, which—like Kant—expels linguistic discourse in general into “anthropology from a pragmatic point of view.

59. J. Hintikka, “Cogito, Ergo Sum: Inference or Performance,” Philosophical Review 71 (1962), pp. 3-32.

60. Analogously, Stegmüller shows, by the very performative act through which he claims validity for his thesis “that the problem of evidence is absolutely insoluble” (Stegmüller, p. 168), that the existence of evidence is a necessary condition of the possibility of meaningful argumentation. Naturally this does not contradict his observation that the existence of evidence cannot be demonstrated (that is, logically deduced) without logical circularity. But it indicates that the reduction of the problem of justification to the possibility of logical demonstration in the framework of an objectivized syntactic-semantic sentence system amounts to an “abstractive fallacy” when it comes to the problem of philosophical foundations. For Stegmüller himself, after all, cannot avoid entering the sphere of (transcendental) pragmatics. He does this when he concludes that the arguing subject, faced with the dilemma that the existence of evidence can be neither denied without self-contradiction nor demonstrated without petitio principii, is compelled to a “prerational decision concerning certainty.” However, this way of entering into a pragmatic dimension without the aid of transcendental-pragmatic reflection leads him to miss the point that the reflective insight that the existence of evidence is a condition of the possibility of argumentation (which can be neither denied without self-contradiction nor logically demonstrated without petitio principii)—as an insight into the pragmatic situation of argumentation—renders a pre-rational decision in favor of the supposition of evidence completely superfluous. For, as an insight of transcendental-pragmatic reflection, it is not about some formal-logical dilemma but about an indispensable condition of the possibility of performing the act of arguing.

61. E. Husserl, Cartesian Meditations, trans. D. Cairns (The Hague, 1973), pp. 18-19.

62. One has to observe Husserl's uncertainty in the following formulation (Husserl, pp. 20-2]): “. . . this 'phenomenological epoche' or 'bracketing' of the objective world . . . therefore does not leave us confronting nothing. On the contrary we gain possession of something by it; and what we (or, to speak more precisely, what I, the one who is meditating) acquire by it is my pure living with all the pure subjective processes making this up, and everything meant in them: the universe of 'phenomena' in the phenomenological sense” (italics added, K.-O. A.).

63. When Husserl declares, “By my living, by my experiencing, thinking, valuing, and acting, I can enter no world other than the one that gets its sense and acceptance or status in and from me, myself” (p. 21), he looks through the language (game) presupposed a priori by his thought as through a glass—no differently than did Descartes at the beginning of the epoch of philosophy justifying itself through the evidence of self-consciousness. Certainly if this whole epoch is rejected as in error because of its reflection upon the subjective conditions of the possibility of epistemic evidence—as has recently been done by Werner Becker, who provides a destruction of the history of transcendental philosophy from the perspective of “critical rationalism” (W. Becker, Selbstbewusstsein und Spekulation. Zur Kritik der Transzendentalphilosophie (Freiburg, 1972)—then, in my opinion, the baby is thrown out with the bath water. For it is not the will to evidence or the “reflection model” (Becker) that is to be rejected from the standpoint of critical discussion. What is to be rejected, rather, is the confusion of reflection on validity with genuine knowledge of a special sphere of being (as in both Descartes and Husserl) or with substantive knowledge in general (partially in German Idealism) and the confusion of actual evidence (for my consciousness) with the intersubjective validity of knowledge. It seems to me, however, that these confusions can be unraveled and avoided through a transcendental pragmatics of language. For a convincing treatment of the Husserlian aporetic, cf. also H. Rouges, “Evidenz und Solipsismus in Husserls ‘Cartesianischen Meditationen,’” in Philosophische Beziehungswissenschaft, Festschrift für J. Schaaf, eds. W. F. Niebel, D. Leisegang (Frankfurt, 1971).

64. In this regard I have not only to add to Hans Lenk's characterization of the noncriticizable rules of the “institution of rational criticism” but also to “dramatize” them in a transcendental philosophy, to use an expression of H. Albert. “The rules and the very idea (or institutions)” of rational criticism are, in my opinion, not only “bound together by linguistic convention” (Lenk, p. 108), but linguistic convention is in this case only the “conventional realization” of rules that originally make explicit conventions (“agreements”) possible. More clearly, the idea and institution of rational criticism is not just a historical form of life among other possible forms of life although in the form familiar to us it was grounded, that is, conventionally realized, for the first time by the Greek philosophers. It may be that the institution of rational discussion has contributed to the realization of homo sapiens, but obviously it could do this only because it made explicit fundamental conditions of meaningful interaction between men and between forms of life. In any case, today the situation is such that not only “can the notion of rational criticism not renounce itself” (Lenk, p. 109), but also we cannot renounce it without renouncing ourselves as men in a nonpathological sense. Naturally, this does not mean that all men must be philosophers (in the academic sense) or even disciples of critical rationalism.

65. Cf. J. Habermas, “Vorbereitende Bemerkungen zu einer Theorie der kommunikativen Kompetenz,” in J. Habermas and N. Luhmann, Theorie der Gesellschaft oder Sozialtechnologie (Frankfurt, 1971), pp. 101-141.

66. For an attempt to carry out this program, cf. my essay “Das Apriori der Kommunikationsgemeinschaft und die Grundlagen der Ethik,” in Transformation der Philosophie, vol. 2, pp. 358-435. Also see there (p. 397ff.) my objections to Albert's proposal to consider competing systems of morals as empirically falsifiable theories of science. Such a treatment already presupposes, in fact, an ethical norm.

67. Cf. my essay, “Die Kommunikationsgemeinschaft als transzendentale Voraussetzung der Sozialwissenschaften,” in Transformation der Philosophie, vol. 2, pp. 220-263.

68. Cf. my essay, “Arnold Gehlens ‘Philosophie der Institutionen’ und die Metainstitution der Sprache,” in Transformation der Philosophie, vol. I, pp. 197-221.

70. To that extent Popper's commitment to the voluntaristic tradition from Duns Scotus to Kant (The Open Society and Its Enemies [Princeton, 1950], p. 780) is justified, but only because the engagement of the will in favor of the realization of reason is not directly synonymous with establishing its reflexive self-justification by means of a decision is tic “Sic volo, sic jubeo; stet pro ratione voluntas.” This viewpoint, however, must, it seems to me, be brought to bear not only against Popper's decisionism but also against Habermas's argument in Legitimation Crisis (Boston, 1975), pp. 158-159. Indeed, I concur—as I scarcely need to emphasize—wholly and completely with Habermas's theory that we human beings (not only as arguing beings but also as acting beings) have always implicitly recognized the validity of norms of ideal communication through the counterfactual anticipation of an ideal communication situation. Nevertheless, it seems to me necessary that transcendental reflection on this “fact of reason” be mediated by the reflection of those arguing upon the conditions of the possibility of their practice. It is only in connection with argumentative discourse that the conditions of possibility of all meaningful action within the framework of language games can be made explicit and distinguished from mere convention. More importantly, reflection upon those ethical principles we have always necessarily recognized does not remove the necessity of a deliberate affirmation (renewed again and again) of this recognition in the sense of a commitment to the realization of reason. This demand amounts, in my opinion, not to a “residual decisionism,” as Habermas asserts, but rather to the validation of the indispensable function of good will in the sense of an ethical unity of knowledge and interest.

72. Decisions against realizing reason do not signify as a rule a denial in principle of the validity of transcendental-pragmatic rules of rational discourse. On the contrary, one claims only to be an exemption—the Devil lives on such things, as it were.

73. That this depends decidedly on following the path of transcendental reflection is indicated, in a very interesting way, by the dilemma of the pure constructivism of the Erlangen School. Although Paul Lorenzen would like to solve the problem of philosophical foundations by reconstructing Kantian transcendental philosophy, he thinks it necessary to grant that an “act of faith” is its starting point, since “the term ‘justification’ makes sense only after one has accepted. . . principles” (Normative Logics and Ethics [Mannheim/Zurich, 1969], p. 74). This problematic, however (which bears obvious analogies to that of Popper), occurs, in my opinion, only if one either no longer recognizes transcendental reflection (upon principles that one must necessarily have always accepted) as a legitimate move in the philosophical argumentation game, or simply overlooks this possibility. This appears to me to be a typical modern conceptual compulsion: One wants to practice Kant's Copernican revolution and, hence, begins immediately with an act of construction. However, in order to be able to present a logical construction as the reconstruction of our competences, we must first reflect on that which is not capable of being meaningfully questioned, the conditions of the possibility of valid criticism that are implicit in the transcendental language game. Only this act of transcendental-philosophical reflection saves us from a “framework” relativism grounded in decisionism, on the one hand, and from a naturalistic absolutization of the empirical critique of ideology into a self-reflection as self-unmasking (in the sense of the “nothing but” reductionism of the nineteenth century), on the other hand. On the distinction between transcendental reflection and critical self-reflection, cf. J. Habermas, “Nachwort” to the paperback edition of Erkenntnis und Interesse (Frankfurt, 1973), p. 411ff.

Return