From formal semantics to transcendental pragmatics: Karl-Otto Apel’s original insight

Karl-Otto Apel’s life’s work represents one of the groundbreaking philosophical achievements of the 20th century. Unfortunately, during the final decades of his career, Apel himself diverted the attention of his colleagues from the central theme in the philosophy of language and the path to discourse ethics to the secondary theme of what he called its ‘ultimate justification’ (Letztbegründung). I would like to justify my unconventional assessment of the relative importance of these themes by retracing the most important stages in Apel’s intellectual development from his habilitation in 1959 to the ‘Transformation of Philosophy’ presented in the second part of the second volume of his major work published in 1973 under that title. Apel’s central ideas were the result of learning processes he underwent during this period, even though the elaboration of his discourse ethics only followed later. ¹ It may be helpful to preface my discussion with some remarks on the academic situation in which German philosophy students of our generation found themselves after the Second World War and during the first half of the 1950s; for at traditional universities such as Bonn, as in the early Federal Republic generally, they encountered only those traditions that had persisted throughout the Nazi era.

Two traditions capable of satisfying a systematic interest played a major role in university teaching at that time. On the one hand, there was phenomenology, which was being developed in productive ways by Husserl’s students, foremost among them by Martin Heidegger with his hermeneutic analysis of Da-sein (being-there) in Being and Time; in his so-called late philosophy, Heidegger lent this a completely different turn charged with a tendentious reading of contemporary culture – a development that was received, however, only with a certain delay at the time. On the other hand, the philosophical anthropology founded by Max Scheler and Helmut Plessner enjoyed a certain vogue, although its main representative in post-war Germany was Arnold Gehlen (whose work was accessible via the politically sanitized second edition of Der Mensch) rather than Plessner, who did not return from emigration until 1952. Heidegger and Gehlen were the authors who posed the most stimulating challenges for philosophy students at German universities in the post-war era. Karl Jaspers had certain presence with his three-volume work Philosophie and his later writings. He triggered a controversy above all with his important work on The Question of German Guilt, which was published soon after the end of the war and clarified the relevant concepts of moral and historical responsibility, as distinct from moral and legal guilt. He deservedly enjoyed widespread public prominence until his death, but he did not inspire genuine philosophical passion among students. Meanwhile, the continent of neo-Kantianism had sunk without trace; and like ‘Lebensphilosophie’ in the wake of Dilthey, neo-Hegelianism, which was still represented in Bonn by a prominent figure such as Theodor Litt, was not able to acquire an independent profile in the face of the prevailing current of phenomenology and hermeneutics either.

That was more or less the state of play, not just in Bonn. But in Bonn, Oskar Becker even provided a personal link to Husserl and Heidegger, while Erich Rothacker in his lectures dispensed the fruits of the hermeneutic erudition of the Historical School from Ranke and Savigny to Wilhelm Dilthey and Karl Bühler. In his seminars in psychology, he dealt with the specialist literature in philosophical anthropology, while themes in the philosophy of culture and language were mainly discussed in the philosophy department (although ‘language’ was treated only in the very German tradition of Humboldt, which was represented at the time by the writings of Lohmann, Trier and Weißgerber). However, intelligent students do not stick to the curriculum but also discover the obvious lacunas in it. Sartre and Simone de Beauvoir were mentioned only in passing. What was lacking was an informed engagement with the two schools of thought of analytical philosophy then prevalent in the Anglo-Saxon countries, a task which Apel would go on to perform. Even more sorely lacking was a discussion of the tradition of Western Marxism that was especially vibrant at the time in France and was slowly regaining a foothold in Germany with the return of the émigrés of the Frankfurt School. It was impossible to overlook the isolation of the specifically German canon from the theories current in the West and the East, whether vital or, as in the case of ‘dialectical materialism’, wedded to the ideology of a political regime. We had to familiarize ourselves with them independently and on our own initiative.

Despite his moral shock at the eye-opening Nazi crimes, Apel was initially rather apolitical. It was only in the late 1950s, after the two of us had resumed the contact that had broken off following our shared time in Bonn, that he opened his mind to the philosophical impulses from critical social theory. As can be seen from the introductory chapter to his habilitation thesis, which was written post festum in the early 1960s, he was initially interested in connecting the philosophical universe of thought circumscribed by the theories of Heidegger and Gehlen with discussions in Anglo-American philosophy. ² For Apel, the tradition of the philosophy of language formed a bridge between these two continents – between the logical semantics of the Vienna School and Wittgenstein, on the one side, and the hermeneutic philosophy of language inspired by Wilhelm von Humboldt, on the other. And interestingly enough, Kant played the role of a connecting hinge. While Kant ultimately remained a subaltern link in the chain of the Occidental ‘history of being’ in the habilitation thesis submitted in 1959, after the Transformation der Philosophie (1973), he was granted the final word over Heidegger’s late philosophy, albeit in the transformed shape of a transcendental pragmatics of language. It was in pursuing this path that Apel initially appropriated the pragmatism of Charles Sanders Peirce. In my opinion, Apel’s genuinely original achievement consists in opening the door to the transition from logical semantics to formal pragmatics by translating the approach of transcendental philosophy from the paradigm of the philosophy of consciousness into that of the philosophy of language. By going beyond semantics, that is, the formal analysis of sentences, Apel also made the use of sentences in speech acts accessible to formal analysis. Of course, Wittgenstein had already focused attention on the deep structure of speech acts, although in different ways in the Tractatus and the Philosophical Investigations: in the former, the linguistic structure of ‘the world’ understood in transcendental terms was only supposed to ‘appear’ in nonlinguistic evidence, whereas in the Philosophical Investigations, the unity of reason dissolved in the description of the diversity of language games and their various views of the world contexts. In spite of his turn toward a pragmatic view of language, the late Wittgenstein had not subjected the operations of reason, to which empirical pragmatics necessarily remained completely blind, to any systematically generalizing analysis either.

Humboldt had already distinguished between érgon and enérgeia as two main aspects of language, that is, between the grammatical structure of ‘language’ (Sprachaufbau) and the effective use of language in ‘conversation’ (Gespräch) – and it is the formal analysis of this language use that Apel wanted to approach in terms of the logic of reaching understanding in discourse. In tracing Apel’s path to what he would then call ‘transcendental hermeneutics’, I will begin by describing his starting position at the time of his habilitation. He was troubled from the outset by the universalistic claim of the Aristotelian logos and Kant’s transcendental apperception: the protest of these voices against the historical thinking of the hermeneutic tradition and its relativistic tendency became an increasingly insistent undertone in his thought, which was initially captivated by the all-encompassing perspective on the history of Western metaphysics that Heidegger interpreted in terms of the ‘history of being’ (1). Then, I will briefly recall the theory of ‘cognitive interests’ (Erkenntnisinteressen) that represents Apel’s appropriation of Gehlen’s legacy. This approach to the anthropology of knowledge, coupled with the simultaneous political critique of Gehlen, could already have liberated the universalistic claim of reason from the contextualism of the history of being, but Apel’s critique remained trapped in a peculiar ambiguity in this regard (2). Although Apel had paved a methodological path to the pragmatic thought of hermeneutics, it was only by means of a transcendental interpretation of the semiotics of Charles Sanders Peirce that he achieved the breakthrough to formal pragmatics – specifically by reading out of the discourse of the community of investigators an a priori of linguistic communication that he simultaneously interprets in ethical terms (3). With his next step of embedding the ‘unlimited interpretation community of investigators’ in the functionally unspecific context of the lifeworld, Apel shifts the discursive exchange of reasons, as the last forum of appeal, from the specialized practice of inquiry into everyday practice. At the same time, the ‘ethos of mutual understanding’ (Ethos der Verständigung), which in Apel’s view springs from the general orientation to reaching understanding already operative in communicative action and discourse, acquires a transcendental status (4). But this direct strategy of an ‘ultimate justification’ of moral demands fails because the strong obligatory meaning of the moral ‘ought’ (Sollgeltung) cannot be derived from the idealizing contents of what participants must conceptually presuppose when they enter rational discourse. Yet Apel was wary of the relativistic implications of an unacknowledged detranscendentalization of reason, although in fact he could not avoid it either. He resisted the pull of these sceptical implications by later supplementing his moral philosophy with a ‘Part B’, which is beyond the scope of the present essay (5). In my view, the question that becomes unavoidable under conditions of secular modernity – namely, why should we be moral at all? – cannot be dismissed as meaningless, but it cannot be answered within the conceptual framework of any deontological moral theory either. I will conclude with a brief outline of an alternative answer that a philosophy which transforms Kant as a whole could provide (6).

(1) The essay on the philosophical concept of truth published in 1959 ³ exhibits a close thematic relationship to the habilitation thesis; Apel probably wrote this text before completing the introduction to the latter work, which clearly was only written later on the occasion of its publication in 1963. At any rate, that essay reflects his continuing allegiance to Heidegger’s understanding of truth as ‘unconcealment’ or ‘disclosure’ (Unverborgenheit), which is supposed to have precedence over propositional truth (for which he uses Rothacker’s expression ‘correctness’ [Richtigkeit]): Vico, whose Scientia Nuova represents the mature idea of language rooted in Renaissance humanism, ‘falls just short of the general insight that “correctness” in the sense of the Aristotelian concept of truth, insofar as it is supposed to be verified in concrete, factual statements, presupposes ‘truth’ as ‘unconcealment’ (a-lētheia) of that which exists’ (p. 130). However, here, too, Apel is already preoccupied with the problem of how to do justice to the claim to universality of the logos that comes into play in propositional truth. For the corresponding notion of reason first makes it possible to translate expressions from one language into synonymous expressions in any other language. By transcending ‘all dogmatics of human standpoints’, every language is ‘rooted in the absolutely universally valid logos’ (p. 132). Apel does not want to deny the universality of the claim to truth that had been interpreted in different ways from Aristotle to Kant. Yet at the same time he insists that it is ultimately the task of the ‘dogmatic’ linguistic world view, which first ‘concretely opens up’ the objective world for each historical linguistic community, to integrate this world-disclosing a priori of the history of being with the ‘apriori of consciousness’ that insists on universal validity. He calls this paradoxical achievement ‘linguistic world integration’, by which he means the integration of the objects encountered in the world into the respective linguistic horizon. The universally valid logos that can only exercise its power within a linguistically ‘pre-given’ but unrevisable horizon of meaning is supposed to specify the conditions under which the interpreters can determine what is or is not true in each case. ⁴ The systematic priority of the world-disclosing ‘truth’ of linguistic world views over the ‘correctness’ of the statements that are made possible by this language explains why learning processes in the world lack a priori the power to revise the world view itself that is ‘sent’ in advance, which in turn is determined by a kind of mythical happening.

However, in the purduit of his project of the idea of language in the early modern humanist tradition, this premise creates difficulties for Apel, because it is obvious that the philosophical heritage is not continued and it does not develop along the lines of national languages and their communities but is instead determined by controversies that extend across national borders. The historical reception of philosophical theories reaches through national languages and their world views. In the later introduction to his habilitation thesis, Apel still maintains

that the truth of human discourse does not rest primarily on a logically correct semiotic representation of supposedly given empirical facts, but on the interpretation of the world as a meaningful situation of human beings that first discloses an order of facts. ⁵

But at the same time, the objective of confronting the hermeneutic with the ‘logos-centric’ conception of language makes the author aware of the right of the universality of reason; thus, he warns against turning prematurely from the anthropologically general analysis of the human condition in Being and Time to the ‘mythology of the fateful history of being’ (p. 54). Alluding to the title of my 1953 critique of Heidegger, ⁶ here Apel already emphasizes

that we must understand – if necessary thinking with Heidegger against Heidegger – both the so-called ‘turn’ from the analysis of Da-sein to the history of being and especially…Heidegger’s philosophy of language in terms of the basic approach of ‘transcendental hermeneutics. (p. 55) (emphasis added)

The turn to ‘transcendental hermeneutics’ opens the door for a deflationary reading of Heidegger’s history of being. The methodological insight of this hermeneutics is ‘transcendental’ in the sense that, through the explication of the hermeneutic circle, it uncovers the peculiar way in which understanding a text depends on the prior understanding of a widely shared historical context in which the interpreter and the author find themselves embedded. This ‘pre-structure’ of understanding, moreover, reflects a basic feature of human existence, namely ‘understanding’ as a mode of being-in-the-world that is constitutive for the possibility of empirically coping with the contingencies in the world.

In Apel’s view, the self-reflection of linguistic understanding (Verstehen) and communication (Verständigung) – where the latter is understood methodologically as an exploratory conversation between interpreter and author – had led to a transcendental deepening of hermeneutic insights. Reason itself seems to assume the shape of an existential mode of human existence in the guise of an everyday understanding that is constitutive for both experience and interpersonal communication. But in the introduction to his habilitation thesis, Apel was not yet able to advance to this specific interpretation, which is achieved through methodological reflection and is aimed at the rational core of Heidegger’s existential mode of ‘understanding’. Reading the second part of Gadamer’s Truth and Method (1960) would later inspire him to take this step. Initially, however, Apel continued to argue against the internal connection between ‘understanding’ and ‘rational communication’ in the following terms:

perhaps philosophy can and must be hermeneutics in a different sense, not insofar as it engages in formal reflection on the accomplishments of hermeneutics in the humanities, but insofar as it must reflect upon itself as emerging from the conversation between philosophers in the West since the Greeks and then introduce this Western philosophical conversation into the contemporary encounter of the major world cultures – or, more generally, insofar as it must shift the present-day human being into the situation of his linguistically mediated being-in-the-world. (p. 48)

Here, Apel still recaptures the claim to universality of a sheer methodological reflection on the ‘pre-structure of understanding’ with which reason seemed to transcend the fateful connection to a particular place in the history of being; he still reintegrates the Kantian ‘apriori of consciousness’ into the context of ‘Occidental metaphysics from Anaximander to Nietzsche’ that is an unavoidable fate for us (p. 57). As if to dispel any doubt about the historically captivating force of this eminent, but specifically Western understanding of the world, which is different from that of other world cultures, Apel adds: ‘In this way, Heidegger…performs the circle of the linguistic-hermeneutic situation of philosophy’ that is ‘allotted to us’ (p. 57) (emphasis added). At the same time, he tries to build a rather daring bridge from the Occidental ‘history of being’ to the ‘history of language’ by appeal to the – rather questionable – theory of the Indo-Germanic linguist Johannes Lohmann. ⁷

(2) During the decisive formative period of his thought in the 1960s, between his habilitation thesis and the essay ‘From Kant to Peirce’ originally published in 1970, Apel liberated himself from this dependence on the late Heidegger and laid the theoretical groundwork for the elaboration of his own theoretical approach. Of interest in this connection is an essay on Arnold Gehlen published in 1962 ⁸ that mainly deals with the theory of institutions developed in Urmensch und Spätkultur (1956). This interesting book can be read as a late substitute for the final chapter of Gehlen’s principal work Der Mensch (1940), which had dealt with ‘Führungssysteme’ and then had been purged from the first edition after the war. Apel specifically addresses Gehlen’s main thesis that the behaviour of human beings who are released from animal instincts, and are therefore no longer guided by nature and in this unshielded condition are ‘open to the world’, has to be stabilized through ‘strong’, that is, repressive, institutions. Against this thesis, Apel argues that, at the evolutionary level of the communicative socialization of acting subjects, language itself represents the functional equivalent of the instinctual behavioural programs tailored to species-specific environments that human beings lack. According to Apel, what takes the place of the instinctively secured connection between the behaviour of a species and its environment is the connection between the human mind and the correspondingly pre-interpreted pathways of a cultural environment that is provided by linguistic world views. Apel assumes furthermore that the contrast between this cultural connection, which is now assured through grammatical structures, and Gehlen’s rigid ‘institutional locking down’ of an amorphous subjectivity was rendered increasingly apparent by the progress of social modernization: here, he concludes, ‘the subjectivity of the modern spirit set free from institutions falls back on the meta-institution of language’.

This assertion is not surprising in itself, since as far as it goes it is still compatible with the fateful priority that Heidegger accords the world-disclosing force of language in the history of being. But what Apel highlights about the meta-institution of language in the following sentence is striking: he does not stress the prejudicing power of a pre-interpreting linguistic world view that would enable language to compensate for the loss of behavioural stabilization in the course of social modernization; rather, he emphasizes the anti-authoritarian feature of rational discourse, that is, of reaching mutual agreement, as opposed to a submissive hearing and understanding of the allotted word of being: ‘After all, the essence of parliamentary democracy, for example, is ultimately that the no longer taken-for-granted social institutions first emerge from the institution of rational discussion’. ⁹ With this, the transcendental role of the intersubjective process of reaching understanding shifts from the directing force of a linguistic pre-understanding of the world as a whole that members of a corresponding language community a fortiori share to the liberating force of rational discourse through which the participants themselves reach a justified agreement beyond dissent and objection. To put it in Humboldtian terms: Whereas Apel, in the course of an in-depth analysis of the hermeneutic pre-understanding, had previously ascribed a transcendental function only to the ‘linguistic worldview’, he now extends this function also implicitly to the unifying force operating in the logos of ‘rational conversation’. Thus, he now continues:

Without doubt, philosophy – since the time when it first discussed, in the dialogue between individual human beings, the foundations of the state and of civilization in general – is now the real ‘idée directrice’ of a meta-institution of language released from myth and the associated institutions, which as ‘logos’ is supposed to establish all other human institutions in the first place (p. 217) (emphasis added).

And to obviate elitist intellectualist misunderstandings of this statement, he continues:

Viewed in this light, parliamentary democracy is an institutional embodiment of the spirit of philosophy. Specifically, it is the embodiment of a mode of philosophical thought that is marked equally by sober skepticism and generous optimism, that does not despair of truth itself…but regards every single human being as capable of contributing to the discovery of truth under conditions of finitude. (pp. 217–18)

This wonderful dictum from 1962 expresses the impeccable Kantian credentials of one of the great 20th-century philosophers. It already points to what Apel would soon learn from Charles Sanders Peirce, namely the pragmatist grounding of the logos, that is, of the reflection on validity claims in the research praxis of a community of investigators engaging in the discursive exchange of experiences and reasons. But a further 6 years would pass before Apel would generalize the example of parliamentary discussion and draw the decisive conclusion from his study of Peirce for the introduction of what he would call transcendental hermeneutics. For it was still quite unclear how the transcendental dimension of the conditions of possibility of linguistic communication as such – that is, the justifying reason operating in rational discourse – was to be conceived as interacting with the interpretive cues that are, as it were, deposited in any linguistic world view. The clarification of this question was more hindered than facilitated by the elaboration of a philosophy of science based on the anthropology of knowledge that Apel undertook in the middle of the decade. ¹⁰ For, contrary to what might be expected from the interplay between hermeneutic understanding and reflection on validity – or to put it in Apel’s terminology, between the a priori of language and the a priori of consciousness – the natural sciences, whose empirical knowledge is supposed to depend on the ‘bodily apriori’ of targeted experimental intervention in natural processes, were not expected to bring universal reason to bear. Paradoxically, the ‘technical’ cognitive interest, on Apel’s account, shifts experimental knowledge to the side of the world-disclosing a priori of language. Apel’s guiding intuition here is the idea that the interest in making nature accessible to instrumental intervention – in a similar way to the perspectives of a linguistic world view – predetermines possible involvements in the world in a practically relevant respect that it prescribes once and for all:

In my opinion, the meaning of the problems of physics…cannot be made intelligible solely by recourse to ‘unifying’ (synthetic) functions of consciousness (categories). This meaning also presupposes a ‘unifying interpretation’ on the basis of a linguistic ‘agreement’ by the investigators of nature as well as the possibility of a realization of the question by an instrumental intervention in nature. This instrumental intervention in nature, which is presupposed a priori in every experiment, to some extent specifies that bodily engagement in the world through the sense organs which is already presupposed in pre-scientific experience: human beings’ ‘comparison’ [Sich-Messen] of themselves ‘with’ nature becomes the ‘measurement’ [Messen] of experimental science. (pp. 46–47; translation amended)

What is counterintuitive about this is not the pragmatist interpretation of experimentation as such, but the fact that Apel ignores the key difference between the ‘bodily apriori’ realized in the experimental intervention and the a priori of the linguistic world view; while the latter projects a particularistic view on the world, the bodily a priori shares with Kant’s ‘apriori of consciousness’ the property of grounding general, universalistic truth claims. ¹¹ In contrast, Apel insists on the supposed analogy between them: ‘like material-technical intervention…language refers back to a distinctive subjective a priori that was not considered in traditional Cartesian epistemology.… ‘knowledge through reflection’ and ‘knowledge through engagement’ stand as polar opposites’ (pp. 48–49). Only a closer study of Peirce would lead Apel to an awareness of the constitutive role of communication for the community of scientists who interpret their theoretical assumptions and the experimental verifications of their hypotheses. This would open his eyes to the fallibility of the linguistically channelled pre-understanding and to the revisionary power of the inner-worldly learning processes that sometimes extends even to the image of nature as a whole. At the same time, it would also lead to a shift in the relative importance of a hermeneutics that refers to the historically relative but fundamentally revisable a priori of a linguistic pre-understanding, on the one hand, and the results of the learning processes reached by the in principle unlimited discourse of investigators, on the other. The latter is based on the quasi-transcendental enabling conditions for criticism of claims to truth, which are rooted in the basic constitution of the communicative socialization of acting subjects. With this Peircean insight, Apel would finally overcome his dependence on the late Heidegger.

(3) So far Apel had circumscribed rather than analysed ‘reflection on validity’ – as the opposite pole to hermeneutics – with allusions to a Kantian ‘apriori of consciousness’. But with the aid of Peirce’s semiotic transformation of Kantian epistemology, he now set about to convert ‘reason’, which until then had been conceived in terms of the philosophy of the subject, into the paradigm of language. For Apel, Peirce was the pioneer who had undertaken ‘to renew Kant’s question about the conditions of the possibility of objective judgments of experience as one concerning the possibility of intersubjective agreement about the meaning and truth of propositions and systems of propositions’. ¹² The place of the transcendental subject is now taking by the community of investigators who conduct experiments to test hypotheses, while the epistemologist’s attention is focused on the discursive dispute over the correct interpretation of experimental data in need of explanation. The investigators argue about the interpretation of the data they have collected in the light of hypotheses and about the hypotheses themselves in the light of competing theories. The dynamic of the cognitive process is located in rational discourse concerning controversial interpretations. Therefore, Peirce conceives of the community of investigators as a communication community whose participants take their orientation from the truth of fundamentally fallible statements, while they understand the truth-claim itself in universalistic terms. In the awareness of its fallibility, a claim to the ‘truth’ of a proposition can only mean the well-founded expectation that the propositions regarded as true for the time being will be included in a final consensus. This ultimate consensus will finally include the sum total of propositions about each and every thing that is the case. This regulative idea of a ‘final agreement’ goes hand in hand with corresponding idealizations: the research community must be conceived as unlimited and the process of inquiry itself as a chain of interpretations extending to infinity. What counts as ‘reality’ is then the totality of the facts on which the research community of interpreters would one day be able to agree as the final consensus. Although Kant’s concept of the ‘thing-in-itself’ is eliminated from this concept of reality, the internal relation of what is real to what in principle can be known by such an unlimited community of investigators reveals the quasi-transcendental character of this approach. For Peirce, philosophical reflection can no more reach behind the process of interpretation and argumentation within the community of investigators than for Kant reflection can reach behind the transcendental subject’s operations of synthesis. A transcendental philosophy thus transformed

is primarily concerned with reflection upon the meaning – and also upon the implications of the meaning – of the process of argumentation as such. This practice is what is obviously the ultimate and irreducible bottom line [das Nicht-Hintergehbare] for all who argue – no matter what their position. ¹³

For Apel, this conception had the advantage of bringing together, in two steps, the traditions in contemporary philosophy of language that continue Aristotle’s and Kant’s universalistic understanding of reason, on the one side, and the hermeneutic tradition, on the other. The first point of contact is the pragmatic aspect under which the use of language becomes a relevant topic both for the hermeneutic understanding of the human and social sciences and now for the process of inquiry in the natural sciences. One recognizes similar initial starting points when one extends the hermeneutic question-and-answer situation in the human sciences between the interpreter and the author of a text up for interpretation to the discursive dispute between experimenting investigators about a corresponding interpretandum. On the side of the investigators, however, this commonality is specified by hypothesis testing and theory formation based on physical measurement and experiment. While hermeneutic understanding seeks to explain the meaning of symbolic expressions, scientific investigators search for the true causal explanation of an empirical phenomenon and thus also for its correct description. This difference between the explanation of meaning and the search for truth points the corresponding hermeneutic and analytical theories of language in different directions. The explanation of the meaning of a text in the light of a pre-understanding that is simultaneously rendered explicit directs the attention of philosophers of language, as in the case of Humboldt, to the structured whole of the semantic network of particular languages, whereas the controversy over the theoretical explanation of empirical phenomena directs their attention, as with Frege, to the composition and logical structure of individual sentences – and also to the pragmatic aspects of the discursive justification of their truth. At any rate, Peirce conceives of the philosophy of language from the outset as semiotics, that is, as a theory of the use of signs that takes as its starting point the triadic relation of the sign to the designated object and to the interpreter. This relation already holds for ‘indices’ (Anzeichen) and ‘icons’, but it is only ‘symbols’ or sentences that have a complete or self-supporting meaning. An assertoric sentence represents a fact for an interpreter in that it can tell him or her something about an (‘indicated’) referent by attributing a predicate (or a quasi-iconic content) to it.

The second point of contact of hermeneutics with Peirce’s pragmatism concerns the kind of relationship between truth and meaning that the two approaches explore in complementary directions. Hermeneutics uncovers this internal relationship by making the possibility of taking an utterance to be true the yardstick of its interpretation in a certain way. An interpreter has understood a text only when she can in principle specify the reasons that could have rationally motivated the author to consider his statements to be true. Insofar as it is a question of the hermeneutic appropriation of classical texts that claim canonical validity, as with Gadamer in Truth and Method, the main hermeneutic challenge facing later generations of interpreters is to ‘bridge’ the ‘time gap’ separating them from the classical author. By advancing to the author’s background understanding in an attempt to understand what reasons he could have had for his statements, the interpreter becomes increasingly aware of her own pre-understanding in a dialectical process – the one becomes explicit in the light of the other. From the vantage point of later generations, the historical appropriation of a text that – for example, in the case of theological or legal exegesis – possesses canonical authority can in this way call into question and relativize the correctness of earlier interpretations but without impairing the claim to the truth of the text that is now ‘better understood’. This is what Gadamer is getting at in his hermeneutics, which takes express aim at Nietzsche’s criticism that historicization neutralizes the power of binding traditions. Gadamer’s hermeneutics seeks to explain why still vital traditions do not necessarily lose their persuasive power when they are exposed to hermeneutic analysis in the human and social sciences. Of course, this narrowing of hermeneutics to the continuation of culturally influential traditions can avoid arbitrarily restricting hermeneutics only if we assume that every interpretation is obliged to follow the uncorrectable grammatical pathways of a particular world view or a corresponding ontological pre-understanding. However, if we reject the assumption that the structure of language founds a prior connection between truth and meaning, then the theory of meaning can only establish a hypothetical connection between these two elements. According to truth semantics, we understand the meaning of a proposition in this sense when we know the conditions under which the corresponding sentence is true: understanding a proposition requires knowledge of its truth conditions. The principle of formal pragmatics postulates a similar hypothetical connection between truth and meaning but generalizes Apel’s reflection on validity (Geltungsreflexion) from ‘truth’ to a variety of rationally redeemable validity claims: We understand a speech act when we know the reasons that make its claim to validity rationally acceptable. ¹⁴

This remark anticipates developments that were first made possible by Apel’s reception and specific interpretation of Peirce’s logic of inquiry. Like philosophical hermeneutics, Peirce examines the internal relationship between truth and meaning in methodological terms. Of course, Peirce’s concern is not with the hermeneutic analysis of how to understand the meaning of a text but in an analogous way with how to determine the truth of a belief – with what he calls the fixation of belief. Specifically, Peirce bases his pragmatic logic of inquiry on an analysis of those three patterns of inference involved in the construction and verification of empirical hypotheses. The starting point is the problematic situation of actors when the implicit assumptions that guide their actions are frustrated by experience. For, in a quite similar way to interpretation, inquiry turns on problematizing well-established assumptions. Inquiry also starts from the problem of a phenomenon in need of explanation that prompts the hypothetical imagination to ‘find out about’ or ‘abduct’ a potentially more suitable assumption about the perplexing phenomenon. This speculative hypothesis is then inductively tested in the light of the results of appropriate experiments, and if it is confirmed, the phenomenon in need of explanation can be deduced from it, and thus correctly explained, as one of its possible cases. Finally, insofar as the problematic explanation of a phenomenon is replaced by a true one in the manner outlined, the description and thus also (as Hilary Putnam has shown using the example of natural terms) the meaning of the phenomenon changes accordingly. While the successful hermeneutic understanding of an expression is measured by the conditions under which it can be used correctly, the logic of inquiry is oriented to a complementary goal: insofar as this procedure leads to the revision of the false description of a phenomenon in need of explanation, the truth conditions for the correct use of the expression itself change.

Decisive for the development of Apel’s thought was Peirce’s belief that the logic and dynamics of the research process give rise to a cumulative improvement in the description of reality. For if the research processes of different communities of investigators, whatever contexts they proceed from, must be assumed to converge on a shared description of the world, the established pre-understanding of reality from which each of them starts must also change in step with the progress of science. In view of this idea of a convergence on a single description of reality from different angles, Apel recognizes in Peirce both a certain hermeneutic sensibility and a rationalist opponent of all forms of historicism. Peirce can only be confident that the imagination of investigators will lead, sooner or later, to the abduction of a hypothesis that fits the facts because he assumes that they do not just speculate at random but take as their guide an intuitively known ‘pre-structure’. Abduction is more than blind guessing, since it is guided by a certain hermeneutic pre-understanding of the larger context in which the problematic case arises. On the other hand, the learning processes can converge on the same goal of a proper grasp of reality notwithstanding their different starting points only because the same logic of inquiry drives them in the same direction of revising their respective pre-understanding in the light of the truth of their findings. Peirce thus relies on the merits of hermeneutics but objects to the priority of the world-disclosing function of linguistic pre-understanding over the propositional truth of the statements that, according to Heidegger, it first makes possible. ¹⁵ Abduction lacks the conclusiveness of deduction and even the probable character of inductive inferences; it merely puts us on the trail of a hypothetical view that needs to be confirmed by empirical evidence. What has the final say, therefore, is the consensus achieved in the process of inquiry for the time being (even if this can subsequently be placed in question by new evidence). In the light of this Peircean conception, the Apel of 1969 poses the following pointed question to the Apel of 1959: ‘What prevents hermeneutics from subjecting itself to the regulative principle of the Peircean explanation of meaning [i.e. scientific progress, J.H.]?’ Hermeneutics is ‘transcendental’, Apel now argues, only in the sense of an interaction between the intuitive pre-understanding of the world laid down in advance by the inferential network of the vocabulary of a language as a whole and inner-worldly learning processes, but these empirical learning processes exercise a revisionary force and can well lead to an improved knowledge of the world, and, through the accumulation of knowledge, in the long run to a revision of the holistic pre-understanding as well. Universal reason ultimately operates in learning processes that are measured by the ‘regulative principle of an absolute truth of understanding in an unlimited community of interpretation and interaction’ (Towards a Transformation of Philosophy, 126), that is, by the irreducible effort to reach a universal consensus in rational discourse.

(4) In the last quotation, Apel already tacitly extends the community of interpreters and investigators to the community of actors in their everyday communication. That coheres with the tenor of those three famous lectures delivered by the young Peirce in 1868 ¹⁶ that marked the foundation of the pragmatic tradition:

Here the process of inference and interpretation of sign-mediated cognition is embedded in the life-process of feedback-monitored behaviour. The aim of this process no longer seems to be the establishment of a consensus concerning truth in the unlimited community of investigators but merely that ‘fixation of belief’ which restores the security of behavior that has been disturbed by doubt, through establishing a new habit of action, which is tested practically (experimentally). ¹⁷

Apel was the first to work out an important implication of this transition from research to everyday practice through the methodological clarification of the difference between causal explanation and hermeneutic understanding. In their everyday contexts, actors not only have to cope with an environment in which their instrumental and purposive-rational actions can fail; in cooperation, they have to act in concert with each other. What is at stake here is not only the truth of empirical assumptions that, when they are problematized and cannot be improved in an ad hoc manner, have to be further processed and clarified in the interpretation community of investigators. Rather, in their social interactions, the actors also confront normative questions in the broadest sense – questions of the controversial evaluation of controversial objects, situations and experiences, of the ethical self-understanding of individuals and communities, of the interpretation of aesthetic and religious experiences, but above all of how to resolve pressing moral, legal and political conflicts. These normative questions cannot be addressed and resolved in accordance with the model of the logic of inquiry. Although there are also experts for all of these questions, reflection on them remains more closely linked to everyday practice and more deeply embedded in the ‘historical interaction communities’ than is true of the theoretical treatment of scientific questions – even if they, too, develop into independent research areas of the humanities and the social sciences, of jurisprudence, art criticism and so forth.

On the other hand, the same general structures of ordinary language are also operative in the communicative practice of the community of investigators programmed to achieve progress in knowledge. This leads Apel to displace into the functionally unspecified lifeworld the a priori of rational discourse which he had first identified, following Peirce, as the practice of last appeal in the field of research. Therefore, Apel now defends

the thesis that the life-world has always been interpreted linguistically and that the a priori of communication in everyday language [sic!] within the context of the life-world is, in a precisely definable sense, the irreducible precondition for the possibility and intersubjective validity of all conceivable philosophical or scientific theory-formation and even of the ‘reconstruction’…of language itself. ¹⁸

From this shift of the a priori into the communicative practice of the lifeworld, Apel now seems to infer that normative beliefs are situated at a level of reflection prior to that of the functionally differentiated practice of inquiry. At any rate, in this text originally published in 1970, he lays the groundwork for his own variant of discourse ethics with its claim to ‘ultimate justification’.

In my view, the reason for this decision lies in a certain fixation of Apel’s early work on the problem and perspective of epistemology. At this important crossroads for the strategy of introducing moral theory, I would like to briefly recall an alternative theoretical path that I myself preferred at that time. I can only try do justice to the fascinating honesty of a great philosopher, and even more so of a friend, such as Karl-Otto who in his life subordinated everything to the consistency of his thought, by clarifying for myself, who owe so much to him, the complicated reasons for our disagreement – not over the substance of discourse ethics but over its justification. Although I am taking advantage of the undeserved opportunity of the younger one who has survived, I do so in the awareness that our friendship always assumed the form of a discourse between partners who were ready to learn from each other.

In the early 1970s, it seemed natural to me to further elaborate in the spirit of JL Austin and the early John Searle the innovative path that Apel had taken – in connection with Peirce’s semiotics – from logical semantics to formal pragmatics. The speech act theory developed by Austin and Searle seemed to be an obvious choice for the design of a formal pragmatics that directs attention to linguistic communication in general, that is, one which, as a theory of meaning, had left behind the Kantian path of epistemology and the methodological perspective of hermeneutics. It deals, instead, with the illocutionary-propositional double structure of speech acts and, moreover, with those universals that are performatively mastered by a competent speaker, like the system of referential expressions and especially the system of personal pronouns, and are constitutive for the use of sentences in speech acts. This formal pragmatic framework also takes into account the phenomenon on which Apel had focused from his epistemological perspective – namely, the distinction between communicative action and rational discourse, now analysed from a perspective that is no longer focused on the community of investigators. It is a general fact in everyday communication that the validity claims raised for propositions when performing speech acts can be thematized and examined as such only under the demanding pragmatic presuppositions of rational discourse, that is, under the presuppositions of a regular exchange of arguments pro and contra. At the same time, the method of uncovering the idealized content of necessary and general, and hence ‘irreducible’ or unavoidable, presuppositions draws on Apel’s own analytic strategy of indirect confirmation by identifying performative contradictions.

From such a formal pragmatic point of view, communication between investigators represents an outstanding case but nevertheless a special one. Just as the truth claims raised and accepted naïvely in everyday life are problematized and made into objects of discourse – whether temporarily or in the methodologically perpetuated form of an institutionalized research process – so, too, are practical validity claims. In both cases, the discursive examination of the validity of a statement has, if not the same, nevertheless the epistemic meaning of a justification based on reasons. The meaning of the respective validity claims, however, differs depending on whether the statement in question refers to the existence of states of affairs, to the worthiness of intersubjective recognition of moral or legal norms, to the preferability of particular values selected from an established, intersubjectively shared and transitively ordered hierarchy of values or, finally, to the ‘beauty’ or, more generally, the evocative potential of nature or art to affect our senses in extraordinary ways. And while the validity claims raised for descriptive, moral, ethical or aesthetic statements refer either to something in the world or – under those three further aspects just mentioned – to something in the symbolically structured lifeworld, the respective patterns of justification vary with the specified meaning of the validity claims.

Depending on the validity claim, however, the discriminatory power of the corresponding type of reasons also varies according to the nature of such patterns of justification. Thus, neither ethical recommendations nor especially aesthetic judgments can claim universal validity. Only moral statements (and mutatis mutandis legal statements) claim a form of universality analogous to truth, so that, especially since Hume’s sharp differentiation between moral and descriptive judgments, a comparison between the two corresponding patterns of justification suggests itself. With his logic of inquiry, Peirce analysed a method of justification for the truth of empirical or theoretical propositions. In a corresponding way, a pragmatic reformulation of the Kantian principle of universalization suggests itself as a similar method for justifying moral judgments: only those norms are valid that would meet with the reasoned approval of all affected in practical discourses. This discourse-ethical procedure calls for an operation of justified generalization of interests in which all those potentially affected participate in principle, so that only those maxims can be recognized as laws that everyone – each from his or her point of view and taking into consideration each other’s points of view – could want for good reasons.

Although Apel did not reject this conclusion, interestingly enough he did not explicitly justify a corresponding moral principle himself, because on his premises no such justification was required. He proposed to arrive at an immediate justification of moral norms via an epistemological path that adopts a transcendental approach from the outset without having to make the supposed detour through a formal pragmatic theory of language. Apel does not begin with the conventional reflection on the question ‘What should I do?’ but starts his ethical reasoning, prior to any differentiation between questions of theoretical and practical validity, from a deeper, more radical basis of epistemological reflection: he insists on the fact that ‘the objectivity of value-free science itself presupposes the intersubjective validity of moral norms’ (p. 256; emphasis in original; transition amended). In other words, practical reason is the core of theoretical reason, because certain basic moral norms are already implicit in the unavoidable pragmatic presuppositions of any rational discourse – and thus in the conditions that first enable the discursive examination of empirical hypotheses. In other words, the assumed moral implications of what we have to presuppose when we enter any kind of rational discourse provide the link between the epistemological interest in the pragmatics of inquiry, on the one side, and the interest of discourse ethics in fundamental moral norms, on the other. Discourse ethics can be understood from Apel’s perspective as the result of a linguistic transformation of Kant’s First Critique. In pursuing the basic epistemological question, transcendental hermeneutics encounters not only the argumentative practice of investigators but also certain moral implications of their willingness to communicate and to cooperate. The investigators’ participation in discourse presupposes a willingness to share responsibility for the successes (or failures) to be achieved collectively, and above all it presupposes the truthfulness of all participants in communication and that ‘all the members mutually recognize each other as participants with equal rights in the discussion’ (p. 259) (emphasis added). In response to the objection that these normative requirements are functional for the specific purpose of the enterprise of ‘research’ and are therefore in effect merely hypothetical imperatives, ¹⁹ Apel argues

that the life-world has always been interpreted linguistically and that the a priori of communication in everyday language within the context of the life-world is, in a precisely definable sense, the irreducible precondition for the possibility and intersubjective validity of all conceivable philosophical or scientific theory-formation and even of the ‘reconstruction’…of language itself. (p. 251; translation amended)

The unlimited community of investigators is rooted in the lifeworld’s ‘interpretation community of an unlimited community of interaction’, so that the reflection on the epistemological question is deepened by encountering the very source of a justification of last resort only in the everyday practice of joint reasoning, that is, in the exchange of arguments in discourse. With this move, the idealizing presuppositions of the community of investigators acquire a place in everyday life. And, as can then be shown with reference to the performative contradictions in which any sceptic would become entangled if he explicitly contradicted the unavoidable pragmatic presuppositions that he cannot avoid making in performing this very argument, this a priori of everyday communication has a normative content in its own right.

Apel defends the foundationalist thesis that rational discourses in the which claims to truth antecedently taken for granted in communicative action are exposed to reflection presuppose a strong ethos of communication (or Verständigung) and – in an analogous way to the accumulation of scientific knowledge – even point to moral progress. This is the mode in which Apel himself presents the argument:

If one considers that the real communication community that is presupposed by each person who engages in critical discussion in the finite situation never corresponds to the ideal of the unlimited community of interpretation, but is instead subject to the restrictions of consciousness and interest that are manifested by the human species in its various nations, classes, language-games and life-forms, then from this contrast between the ideal and the reality of the interpreting community there arises the regulative principle of practical progress, with which the progress of interpretation could, and ought, to be entwined. ²⁰

According to Apel, the counterfactual anticipation of a final state aims at a consensus omnium also in the case of moral discourse – not at an agreement among investigators on the totality of the facts but at an agreement among the communicatively acting members of a globally expanded social lifeworld about the norms that regulate their social coexistence.

(5) In my view, the program of discourse ethics does not depend on the persuasiveness of this transcendental justification strategy any more than does Apel’s innovative extension of hermeneutics to an examination of the formal properties of discursive processes of reaching understanding. As mentioned above, I regard his critical step from logical semantics to formal pragmatics (and the paradigm shift from the philosophy of consciousness to the philosophy of language built upon it) in any case as his groundbreaking achievement. But at the interface of the transition from theoretical to practical reason, Apel holds fast to the epistemological question even after he has accomplished the paradigm shift. ²¹ Theoretical reason – and this is where I cannot follow him – is supposed to become aware of its own limits as soon as it reflects on the roots of the process of inquiry in the lifeworld. This gives rise to a hierarchy of stages of a progressively deepened self-reflection of knowledge and cognition that leads back to a shared origin of theoretical and practical reason. Then, by extending Kant’s transcendental reflection, which is limited to the conditions of the objectivity of possible experience (i.e. to the theoretical use of reason), into an analysis of the necessary and general presuppositions of the use of language oriented to reaching understanding in general, Apel wants to uncover an ethos that is inherent in communicative socialization as such. In doing so, he uses the identification of performative contradictions in quite an original way as a method of indirect justification: The moral content implicit in those unavoidable pragmatic presuppositions of rational discourse can be demonstrated to any sceptic by making him aware of the fact that, insofar as he denies the position of discourse ethics, he becomes trapped in performative self-contradictions by his very participation in the exchange of reasons. In this indirect way, Apel can show, for example, that every serious participant in discourse is actually expected to be truthful and to treat all other participants equally regarding their arguments. The primordial anchoring of this ethical substance in the discursive practice of mutually convincing each other – even prior to the separation into theoretical and practical discourses – is supposed to lend the proof of the entanglement in performative contradictions the foundational status of an ‘ultimate justification’ of that substance. On closer inspection, however, this construction is compromised by its narrow, epistemological conception of the justification strategy and its failure to explicitly de-transcendentalize reason. As we shall see, Apel’s own investigations, notwithstanding all assertions to the contrary, will suggest such a de-transcendentalization of reason.

I do agree that Apel’s method of involving the sceptic in performative self-contradictions allows us to uncover in the general and necessary pragmatic presuppositions of rational discourse a normative layer comprising a few strong idealizations, such as the mutual expectation of equal treatment, comprehensive inclusion, as well as equal and unforced treatment of everybody possibly affected. But these pragmatic presuppositions are tailored to the epistemic goal of testing the validity of statements of whatever kind, including the rightness of moral norms and commands. However, the meaning of the moral ‘ought’ as such cannot be extracted from the presuppositions of discourse. ²² The presuppositions that competent participants in argumentation tacitly accept as soon as they enter the game are designed for checking hypothetical validity claims in the light of a discursive exchange of topics, reasons and information. This reflexive mode of communication must grant (a) that everybody concerned and anyone who could make a relevant contribution is included; (b) that all participants have equal opportunities to make contributions; (c) that all participants are obliged to be truthful in the sense that they must mean what they say; and (d) that the flow of communication is free from external and internal (!) constraints, so that the yes or no stances of participants on criticizable validity claims can be determined solely by the unforced force of the better argument. In these presuppositions, without which the communicative conditions for a rational discourse are not fulfilled, we recognize, among other things, the norms of truthfulness and equal treatment emphasized by Apel; they even imply ‘co-responsibility’, if by this we only mean the willingness to mobilize both the relevant and best possible information and the reasons that can facilitate the search for truth and, in addition the willingness to recognize the fallible result of argumentation for the time being. However, these specifying restrictions remind us of the exclusively epistemic purpose of rational discourses. Rational discourse, as the reflexive mode of communicative action, inherits from this basis the ‘orientation to reaching understanding’ but not the binding force of the intersubjective recognition of criticizable validity claims on which the action-coordinating role of linguistic communication rest in the communicative practice of everyday life. In rational discourse, validity claims are turned from a source of coordinating interactions into a topic to be examined as to whether the claims can be justified. And that issue is solely a matter of the persuasive force of the better reasons that rationally motivate participants to agree on the content of a statement. But an epistemic agreement on a validity claim does not generate any compelling motivation for shared action outside of discourse. It is up to the actor whether she maintains the communicative attitude or shifts from communicative to strategic or violent action.

This is manifest in the case of the truth claims that are examined in theoretical or empirical discourses. Within discourses, the ‘obligations’ to truthfulness and equal treatment that follow from the unavoidable pragmatic presuppositions of discourse apply only to the discursive practice as such – participants must respect these obligations within the discourse itself if this is not to become pointless. It is clear that violating these presuppositions of discourse is dysfunctional, but it is not per se morally wrong, because it does not immediately intrude on contexts of action and does not directly affect interests that could be violated in a normative sense. By its very nature, the orientation to reaching understanding is bound up with the discursive examination of a truth claim and therefore has a purely epistemic meaning. This is measured by whether the reason on which the argument turns speaks for or against the existence of the state of affairs affirmed. Something similar applies in the case of a moral validity claim for a general mode of conduct or a particular behaviour (under given circumstances and in the light of recognized moral norms). The claim to rightness of a controversial moral norm depends on whether there are reasons to believe that it is in the equal interest of all those possibly affected (and the appropriateness of a morally debatable behaviour depends on whether it is in agreement with a morally recognized norm under the present circumstances). It is true that practical discourses, because they deal with practical norms that affect acting subjects in their various life situations, also require the participants to be willing to adopt each other’s perspectives. But in these cases, too, what is at stake is mainly the epistemic meaning of the moral-cognitive question of whether a mode of conduct or a behaviour should or should not be recognized as morally binding.

However, there is a subtle but momentous difference between theoretical and practical discourses. It seems that the assertoric validity claim raised for descriptive statements does not first have to be introduced into the discourse from outside, because the semantic form of a descriptive statement per se refers to the meaning of the claim to truth, namely, the existence of a corresponding states of affairs. The meaning of truth claims is exclusively a matter of the logical structure of the statements for which they are raised; they form a component immanent in discourse, as it were, even though one must refer to something ‘existing’ in the objective world to explain the meaning of the claim to truth in greater detail. In contrast, the meaning of the claim to normative validity (Sollgeltungsanspruch) raised for moral statements can only be explained with reference to validity claims that already exist or are still to be realized outside of discourse. Before a normative validity claim can be thematized as such in a discursive exchange of reasons, it is attached to a norm that exists or can come into existence only in the mode of socially recognized normative validity. Moral norms can only exist insofar as they are recognized as valid by social subjects and, in virtue of this intersubjective recognition, can actually expect to meet with compliance on the part of their addressees. In contrast to the truth claim raised for propositions, the claim to normative validity of a moral norm or of a corresponding action that has become problematic can only be thematized in discourse after it has been imported into discourse from the lifeworld.

Therefore, the mode of the normative validity of moral statements cannot be derived from the idealized content of any presupposition of discourse but only from the obligatory meaning and the socially integrating force of social norms that demand recognition and actual respect from acting subjects. ²³ The validity claim to moral rightness that is subject to discursive examination does not derive its meaning from discourse itself but from the claim raised by the moral principles and norms that exist or shall be realized in society; it is only whether they are worthy of recognition that is up for discussion in discourse. But if, contrary to Apel’s assumption, a morally binding ethos of mutual understanding cannot be derived from what communicatively acting subjects must presuppose when they engage in a rational discourse, then the participants in discourse must introduce the notion of normative validity from their social life into discourses of justification and application in the first place. And instead of a transcendental justification developed by moral theory directly from the ‘apriori’ of communication rooted in the lifeworld, a justification of moral statements, as we have seen, ²⁴ calls for a procedure analogous to the justification of empirical–theoretical propositions. ²⁵

(6) In this place, there is no need to revisit my own attempt, originally published in 1983, to justify the principle of universalization. ²⁶ This strategy makes use of Apel’s procedure of reconstructing the content of the universal and necessary pragmatic presuppositions of rational discourse, but without claiming to derive from these presuppositions the obligatory meaning of the mode of normative validity through which morally binding moral norms actually operate as mechanism for coordinating interactions in the lifeworld. The quasi-transcendental necessitation of unavoidable presuppositions of argumentation alone cannot generate morally binding force. This force belongs instead to the norms that are introduced into moral–practical discourses because they have become problematic, and it is to these norms that the pro and contra reasons mobilized in deliberation refer. Thus, the presuppositions of argumentation cannot shoulder the burden of a justification of moral norms and commands – not to speak of a ‘transcendental’ justification in the strong meaning of a transcendental a priori. The discursive mode of the examination of criticizable validity claims that we raise for propositions is only ‘ineluctable’ or ‘unavoidable’ in the weaker sense that there is no functional equivalent for this practice of giving and asking for reasons in any of the sociocultural forms of life that we know of. In my opinion, such a detranscendentalized conception of de facto unavoidability for our species also follows from Apel’s own anthropological grounding of epistemology. Perhaps it is this unresolved problem in the background of his theory that led Apel later on to supplement his discourse ethics with a ‘Part B’ that is supposed to answer the question ‘Why should we be moral at all?’ ²⁷ However, this question calls for a teleological answer; it cannot be asked, let alone answered, within the framework of a deontological approach to moral theory. This problem challenged Apel to introduce a super-norm that paradoxically puts us, in addition to the sum total of moral norms, under the further obligation ‘to be moral’ in the first place. I cannot address this part of Apel’s theory here. But it is worth speculating in conclusion about why Apel stubbornly clung to the transcendental meaning of an ‘ultimate’ justification of discourse ethics. I suspect that the answer lies in the problematic tension between an ‘absolute’ or unconditional meaning of ought-sentences and the motivation of socialized subjects who exist in space and time to follow them.

If the Kantian question of what we ought to do loses its transcendental background and depends on the occurrence of binding norms whose normative validity claim we can critically examine from the moral point of view, we must confront the unsettling question of whether such categorically obligatory norms in fact still exist today – and whether they should continue to exist at all. Can norms that require their addressees to act in a certain way under given circumstances unconditionally – that is, without regard to the consequences – still claim ‘to exist’ in the secular societies of Western modernity in the sense that, when they are regularly problematized as objects of criticism and justification, they can still impose themselves in the light of good reasons alone? To be sure, in the role of social scientists, we can observe that, in the largely secularized spheres of action of modern societies – above all in patterns of socialization and systems of formal education, and certainly in informal interaction routines of almost all areas of life, and in the legal systems and constitutional states – conceptions of justice are still embodied that in many cases possess a deontological core of unconditional obligation. A prominent example is the decision of the German Federal Constitutional Court of 15 February 2006 that § 14, para 3 of the Aviation Security Act that leaves it up to the discretion of the state authorities to weigh the lives of a large number of threatened people against the lives of a limited number of others by ordering an airplane to be shot down to protect lives on the ground, violates human dignity (Article 1, Basic Law) and the basic right to life (Article 2, para 2, Basic Law). However, the very uncertainty over whether such decisions that used to be justified in deontological terms will continue to be so justified in the future demonstrates that philosophy can no longer rely on this ‘self-evidence’ in the way Kant did (when he treated the feeling of duty as a ‘transcendental fact’). Certainly, since the 17th century, philosophical attempts to reconstruct this core of religious moral conceptions in the form of rational law and rational morality have prevailed, but nobody knows whether competing noncognitivist approaches that appeal to empiricist or naturalistic arguments will not ultimately prevail in the future and dismantle the strict moral self-understanding of individuals who act responsibly as an illusion. From this noncognitivist perspective, the categorical ‘ought’ associated with deontological notions of justice is an intuition based on self-deception. Although it is hard to imagine that children growing up in families or corresponding social environments without the categorical ‘ought’ of obligatory moral commands would not develop into clinical monsters, it is difficult to gauge to what extent, as capitalist modernization progresses, those ‘enlightened’ interpretations will trump Kantian intuitions and undermine deontological beliefs concerning justice by replacing them with utility calculations based on rational choice or game theory.

These remarks are only meant to draw attention to the burdens of argument that may not even exist from a strict transcendental philosophical perspective. However, if, viewed from a postmetaphysical point of view, strict moral norms, as historical components of our sociocultural forms of life, can ultimately only have been generated by social subjects in the course of social evolution, it is plausible that, as morality becomes secularized, the immediate question ‘What should I do?’ will be trumped by the reflexive question ‘Why should I be moral at all?’ Since the latter question can neither be dismissed out of hand nor be answered within the conceptual frame of deontology, I have tried to explain within the broader context of a history of philosophy why, in the constellation shaped by Hume and Kant, postmetaphysical thinking has split into opposed traditions specifically in response to this question – and why a justification of rational morality can still be defended with reasons against the defeatism that this kind of postmetaphysical thought itself nourishes. Because Kant justifies the moral principle in terms of a concept of autonomy that binds the free will of the individual to precisely those laws that all possibly affected by them could want for good reasons, everyone can answer the question ‘Why be moral at all?’ by realizing that only an affirmative answer enables them to understand themselves as ‘autonomously acting’ subjects. To be sure, morality and freedom relate to and mutually depend on each other in the very concept of autonomy; but if, like Kant, our primary concern is the self-understanding of the person as an autonomous rational being, this alone provides a good reason, even without transcendental certainties, to ‘be moral’ in the deontological sense – for only then can we understand ourselves as autonomous. ²⁸

Translated by Ciaran Cronin