8

Jürgen Habermas’s Kantian Pragmatism

Several years ago when Habermas was asked “What do you see as the most lasting contribution of pragmatism to the tradition of Western philosophy and social thought?” he replied:

Alongside Marx and Kierkegaard … pragmatism emerges as the only approach that embraces modernity in its most radical forms, and acknowledges its contingencies, without sacrificing the very purpose of Western philosophy – namely, to try out explanations of who we are and who we would like to be, as individuals, as members of communities, and as persons überhaupt – that is, as man. (Habermas 2002, p. 229)

This response echoes an interpretation of post-Hegelian philosophy that Habermas (together with Karl-Otto Apel) worked out in great detail. Hegel is the philosopher par excellence who deeply grasps the aporias of modernity – aporias concerning the many forms of rupture and fragmentation as well as their potential for healing and reconciliation. Marxism, Kierkegaardian existentialism, and pragmatism represent the three major attempts to rethink and resolve these aporias. The above passage also tells us something about Habermas’s philosophical trajectory – where he has come increasingly to think of his comprehensive philosophical investigations as developing what he calls a “Kantian pragmatism.” My aim in this chapter is to sketch the outlines of this Kantian pragmatism and raise some critical questions about it. (We will see just how important it is to explicate both terms: ‘Kantian’ and ‘pragmatism.’)

Kant Detranscendentalized

But first I would like to say a word about the influence of American pragmatism on Habermas. Habermas is one of the very few European philosophers who have taken classical American pragmatism seriously. He appropriates, reconstructs, and integrates pragmatic themes into his own thinking. “Influence” is perhaps too flat a word, because it suggests that Habermas read the pragmatic thinkers carefully and learned from them. This is certainly true. But what is more philosophically interesting is that Habermas – following his own distinctive dialectical journey – has come to realize how many of his central ideas were anticipated by the classical American pragmatists. This is, of course, compatible with the ways in which he has also been critical of the pragmatists (see Habermas 2002). Habermas has also critically engaged those contemporary American philosophers who (in different ways) strongly identify with the pragmatic tradition: Richard Rorty, Hilary Putnam, and Robert Brandom. In these exchanges he has refined and defended his Kantian pragmatism.

We must remember how deeply Peirce, the founder of pragmatism, was shaped by his encounter with Kant.¹ Peirce began his philosophical investigations with an attempt to rethink the Kantian categories – or to use Habermas’s expression – to “detranscendentalize” Kant. Habermas tells us: “In epistemology – and the theory of truth – Peirce had the strongest influence, from my Frankfurt inaugural lecture on Knowledge and Human Interests (1971) onwards up to Wahrheit und Rechtfertigung (1999).” Habermas (and Karl-Otto Apel) “perceived Peirce’s pragmatist approach as a promise to save Kantian insights in a detranscendentalized yet analytical vein” (Habermas 2002, p. 227). What does Habermas mean by a “detranscendentalized” Kant? I begin with an initial characterization, which I will refine as I proceed. Habermas has always felt that there is something profoundly right about the core idea of the Kant’s transcendental project.

Transcendental philosophy, as the famous phrase has it, “deals not so much with objects as rather with our way of cognizing objects in general insofar as that way of cognizing is to be possible a priori.” It takes itself to be reconstructing the universal and necessary conditions under which something can be an object of experience and cognition. The significance of this transcendental problematic can be generalized by divorcing it from the basic mentalistic concept of self-reflection as well as from a foundationalist understanding of the conceptual pair a priori – a posteriori. After the pragmatist deflation of Kantian conceptuality, “transcendental analysis” refers to the search for presumably universal but only de facto unavoidable conditions that must be fulfilled in order for fundamental practices or achievements to emerge. (Habermas 2003, pp. 10–11)

One may think initially that this “pragmatist deflation” is so radical that there is very little of the “historical” Kant left. Habermas doesn’t deny this; transcendental philosophy no longer deals with “consciousness überhaupt.” It now becomes the investigation that aims “to discover deep-seated structures of the background of the lifeworld. These structures are embodied in the practices and activities of subjects capable of speech and action” (Habermas 2003, p. 11). The linguistic turn in philosophy requires that we give up talk about “consciousness,” “self-consciousness,” and “apperception” as exclusively mental activities. There is no way of analyzing concepts and judgments without reference to language. And we cannot understand language and speech acts except in the context of social and communal practices. “Detranscendentalization leads, on the one hand, to the embedding of knowing subjects into the socializing context of a lifeworld and, on the other hand, to the entwinement of cognition with speech and action” (Habermas 2003, pp. 88–9). A great deal is packed into this brief statement. When Habermas speaks of “embedding of knowing subjects into the socializing context of a lifeworld,” he is alluding to his (and the Peircian) sustained critique of “the philosophy of consciousness” that has dominated so much of modern philosophy. He is underscoring the paradigm shift from a philosophy of subjectivity or consciousness to an intersubjective or social understanding of human beings that stands at the very heart of his communicative theory of action and rationality. When he speaks of the “entwinement” of cognition with speech and action, he is alluding to the linguistic turn that is characteristic of twentieth-century philosophy – especially the appreciation of the pragmatic dimension of speech acts, which can also be traced back to Peirce’s semiotics and Mead’s communicative analysis of speech acts.

Many of the classic Kantian distinctions – noumena/phenomena, appearance/thing-in-itself, understanding/sensibility – are no longer acceptable as they were originally developed by Kant. Habermas rejects Kant’s transcendental idealism (as well as Hegel’s absolute idealism) in favor of a post-Kantian and post-Hegelian epistemological realism. Nevertheless, even after the linguistic and pragmatic turns in philosophy, there is a proper role for a chastened transcendental philosophy: that is, the attempt to reconstruct the universal conditions that form the deep structures of everyday speech and action. This is the investigation that Habermas carries out in his theory of communicative action, a project that abandons any claims to finality, completeness, or epistemic certainty. Habermas is a thoroughgoing fallibilist – another basic conviction that he shares with Peirce and the classical American pragmatists. Furthermore, Habermas, as we shall see, follows Kant in maintaining a sharp distinction between theoretical and practical philosophy. In this respect, he differs from the classical American pragmatists (with the possible exception of Peirce). In sorting out the ways in which Habermas follows and departs from Kant, one should also mention his interest in naturalism. Habermas has always opposed a reductive form of naturalism: the type of naturalism closely allied to reductive materialism and scientific realism, which identifies nature with the entities ultimately countenanced by an “ideal” physical science. But in Truth and Justification he sketches a “weak naturalism.” The question he wants to answer is: “How can the normativity that is unavoidable from the perspective of the participants in this lifeworld be reconciled with the contingency of sociocultural forms of life that have evolved naturally?” (Habermas 2003, p. 2). Weak naturalism, as Habermas understands it, is “based on a single metatheoretical assumption: that ‘our’ learning processes, that are possible within the framework of sociocultural forms of life, are in a sense simply the continuation of ‘evolutionary learning processes’ that in turn gave rise to our forms of life. … [W]eak naturalism contents itself with the basic background assumption that the biological endowment and the cultural way of life of Homo sapiens have a ‘natural’ origin and can in principle be explained in terms of evolutionary theory” (Habermas 2003, pp. 27–8). In short, Habermas wants to combine a version of a detranscendentalized Kant with Darwin.²

When Habermas developed his theory of communicative action and rationality originally, he set aside theoretical issues of epistemology that preoccupied him in Knowledge and Human Interests. The linguistic turn did not acquire its significance for him in dealing with traditional epistemological and semantic issues such as our knowledge of objects and facts or the form of assertoric propositions. “Rather, the pragmatic approach to language [Sprachpragmatik] helped me to develop a theory of communicative action and of rationality. It was the foundation for a critical theory of society and paved the way for a discourse-theoretic conception of morality, law, and democracy” (Habermas 2003, p. 1). The theory of communicative action situates rationality in the everyday practices of the lifeworld and shows that the critical power of reason is rooted in ordinary language. “The expression ‘communicative action’ designates social interactions where language use aimed at reaching mutual understanding plays the role of action coordination. Through linguistic communication, idealizing presuppositions enter into action oriented toward reaching mutual understanding” (Habermas 2003, p. 110). But Habermas stresses that his communicative theory of action and his formal pragmatics presupposes an understanding of truth and objectivity.

Of course, the formal pragmatics that I have developed since the early 1970s cannot do without the fundamental concepts of truth and objectivity, reality and reference, validity and rationality. The theory relies on a normatively charged concept of communication [Verständigung], operates with validity claims that can be redeemed discursively and with formal-pragmatic presuppositions about the world, and links understanding speech acts to the conditions of their rational acceptability. However, I have not dealt with these themes from the perspective of theoretical philosophy. (Habermas 2003, p. 1)

Truth and Normative Rightness

To see why it is so important to deal with these themes from “the perspective of theoretical philosophy,” we need to recall some of the key claims that Habermas makes in The Theory of Communicative Action. The heart of the theory concerns what Habermas calls “validity claims.” Communicative interaction is primarily a matter of raising validity claims. There are three basic types of validity claim that are raised by speakers in their speech acts: a claim to the truth of what is said or presupposed, a claim to normative rightness, and a claim to truthfulness. All three claims are raised simultaneously, although typically only one of them is thematized or made explicit, while the others may remain implicit. Habermas is, of course, aware that in actual empirical communication, any or all of these validity claims may be violated. We may deliberately lie and misrepresent something as true that we know is false with the intent of deceiving a hearer about what is normatively right. One of the most misunderstood features of Habermas’s theory is the way in which these “idealizing” counterfactual presuppositions of communicative action may be violated in any empirical speech act. These idealizations are not intended to delineate an “ideal” communication or an “ideal” community. On the contrary, they are intended to specify what is counterfactually presupposed in every communicative interaction. For example, we could not lie or deceive without presupposing that there are implicit standards for what counts as truth and truthfulness.³ I shall concentrate on the validity claims to truth and normative rightness – for this is where we will see the full force of his Kantian pragmatism. Each of the validity claims is distinctive, but they are also interdependent. When, in a specific communicative exchange, a question is raised about the truth of one’s claim or its normative rightness, then there is a call for argumentation (discourse) – the give-and-take of reasons. Consequently, disputes about truth and normative rightness are essentially cognitive. This sets Habermas apart from all those philosophers who claim that disputes about normative rightness are noncognitive. One of the key presuppositions of communicative action is that there is a common objective world about which we can make true and false assertions. If I claim that something is true and you challenge me, then at the very least we are both presupposing that there is a common objective world to which we can refer. Of course, this does not mean that in fact we will agree or be able to resolve our dispute. When Habermas initially sought to explain the validity claim of truth, there were some crucial ambiguities in his presentation. Unless Habermas can give an adequate account of the validity claim of truth, the very foundations of the theory of communicative action and rationality are seriously threatened.

The Epistemic Conception of Truth

Habermas originally approached the question of truth from an epistemic perspective. He calls this a “procedural” or “discursive” understanding of truth. The primary issue here is not how to characterize truth, but rather how to analyze what it means to know that a given claim is true. He originally sought to answer this question by appealing to the “ideal” conditions of justification. But many critics have argued that such an epistemic conception of truth is seriously flawed. Habermas now concedes that his critics are right.

For my part, I initially determined the meaning of truth procedurally, that is, as confirmation under the normatively rigorous conditions of the practice of argumentation. This practice is based on the idealizing presuppositions (a) of public debate and complete inclusion of all those affected; (b) of equal distribution of the right to communicate; (c) of a nonviolent context in which only the unforced force of the better argument holds sway; and (d) of the sincerity of how all those affected express themselves. The discursive concept of truth was on the one hand supposed to take account of the fact that a statement’s truth – absent the possibility of direct access to uninterpreted truth conditions – cannot be assessed in terms of “decisive evidence,” but only in terms of justificatory, albeit never definitively “compelling,” reasons. On the other hand, the idealization of certain features of the form and process of the practice of argumentation was to characterize a procedure that would do justice to the context-transcendence of the truth claim raised by a speaker in a statement by rationally taking into account all relevant voices, topics, and contributions. (Habermas 2003, pp. 36–7)⁴

This procedural conception of truth comes close to identifying truth with “ideal” justification.⁵ Habermas now realizes that this is inadequate. In his introduction to Truth and Justification, he writes:

I have given up an epistemic conception of truth and sought to distinguish more clearly between the truth of a proposition and its rational assertibility (even under approximately ideal conditions). In retrospect, I see that the discursive conception of truth is due to an overgeneralization of the special case of the validity of moral judgments and norms. A constructivist conception of the moral “ought” does require an epistemic understanding of normative rightness. But if we want to do justice to realist intuitions, the concept of propositional truth must not be assimilated to this sense of rational acceptability under approximately ideal conditions. (Habermas 2003, p. 8)

Habermas certainly does not want to deny that there is an “unavoidable epistemological connection between truth and justification,” but this does not amount to “a conceptual connection between truth and assertibility under ideal conditions” (Habermas 2003, pp. 37–8). “Truth,” he tells us, is a “property of propositions that they ‘cannot lose’ ” (Habermas 2003, p. 38). The proper characterization of truth is closely related to an understanding of objectivity and reference. Habermas now emphasizes that an epistemic understanding of truth is not sufficient, because no matter how strong our justification is – even if it were “ideal” – it may turn out that what is ideally justified is nevertheless false. Furthermore, serious problems arise when it comes to specifying what “ideal justification” really means. The epistemic account of truth does violence to our “realist intuitions.” It violates our sense that there is a world that is independent of our justificatory practices and that our claims to truth ultimately presuppose an independent objective world. Habermas’s realist intuitions are extremely strong. Although he emphasizes that philosophy is now post-metaphysical, he doesn’t hesitate to use ontological language.

What we want to express with true sentences is that a certain state of affairs “obtains” or is “given.” And these facts in turn refer to “the world” as the totality of things about which we may state facts. This ontological way of speaking establishes a connection between truth and reference, that is, between the truth of statements and the “objectivity” of that about which something is stated. The concept of the “objective world” encompasses everything that subjects capable of speech and action do not “make themselves” irrespective of their interventions and inventions. (Habermas 2003, p. 254)⁶

This is strong stuff. Except for the claim that the world is “the totality of things about which we may state facts” rather than the world as the totality of facts, it sounds like the Wittgenstein of the Tractatus.

One of the fundamental questions that Habermas seeks to answer in Truth and Justification is: “How can we reconcile the assumption that there is a world existing independently of our descriptions of it and that is the same for all observers with the linguistic insight that we have no direct, linguistically unmediated access to ‘brute’ reality?” (Habermas 2003, p. 2). As I hope to show, the viability of Habermas’s Kantian pragmatism depends on giving a satisfactory answer to this question.

Neither Contextualism nor Idealism

Let me first state what are Habermas’s chief worries and what it is that he wants to avoid. This may seem like a circuitous route, but it will take us to the heart of the matter. Briefly stated, Habermas wants to avoid what he takes to be two unsatisfactory extremes: the extreme of Rorty’s pragmatic contextualism and the extreme of Brandom’s pragmatic neo-Hegelianism. Once one makes the linguistic turn, then one is committed to holding that an adequate explication of the procedures of argumentative justification must be related to language – or what Rorty calls “vocabularies.” And if there is no “linguistically unmediated access to ‘brute reality’ ” (as Hegel, Peirce, Wittgenstein, Sellars, and many others have argued), then how can we know that there is a world existing independently of our descriptions of it? Rorty’s answer is straightforward: we can’t! And because there is a plurality of vocabularies, and no apriori limits on inventing new vocabularies, all we can do is play off our vocabularies – our descriptions and redescriptions – against each other. The world is well lost! There is simply no “it,” no world that stands independently of our descriptions. If we appeal to “realist intuitions” (as Habermas does), Rorty’s advice is to abandon these intuitions and develop new ones. Rorty is quite explicit in mocking the philosophical appeal to intuitions, especially realist intuitions.

What really needs debate between the pragmatist and the intuitive realist is not whether we have intuitions to the effect that “truth is more than assertibility”. … Of course we have such intuitions. How could we escape having them? We have been educated within an intellectual tradition built around such claims. … But it begs the question between pragmatist and realist to say that we must find a philosophical view which “captures” such intuitions. The pragmatist is urging that we do our best to stop having such intuitions, that we develop a new intellectual tradition. (Rorty 1982, pp. xxix–xxx)

From Rorty’s perspective, defenders of “realist intuitions” require philosophical therapy – a therapy that aims to relieve them of their anxieties and to abandon their “realist intuitions.” These “intuitions” don’t do any pragmatic work; they don’t help us to cope. Consequently, Rorty deflates the philosophical idea of truth and replaces it with the appeal to social (that is, sociological) justification: he urges us to replace objectivity with solidarity. All we need is a “cautionary sense of true,” which merely tells us that no matter how successfully we have justified a claim, it may turn out that we discover that the justification is no longer satisfactory; it no longer persuades. But if this happens, it is because we have come up with new forms of social justification: that is, procedures of justification that are more acceptable to our peers and the audience we are addressing – not because “realistic intuitions” have been vindicated. Rorty might claim that he is a more consistent “fallibilist” than Habermas, because he makes no pretence to speak about “ideal justification” or to anticipate what new forms of justification will be adopted in the future.

Habermas has raised a variety of objections against Rorty. He argues that this sort of contextualism cannot avoid relativism. Rorty doesn’t think that we can give good reasons (indeed, he doesn’t think he can give any context-transcending reasons) to justify one vocabulary over any other vocabulary. All we can do is try to persuade others to adopt the vocabulary we favor. Despite Rorty’s claims to the contrary, this leads to a self-defeating relativism and involves a “performative contradiction.” (Putnam has raised similar criticisms against Rorty. This is one reason why Habermas views Putnam as a fellow Kantian pragmatist who justifies a post-Kantian internal realism.) Habermas also argues that Rorty fails to make any sense of something that is implicit in our everyday linguistic practices: if I claim that something is true or just, I am not simply claiming that something is true or just for me or my group. I am making universal claims that “transcend” my context – “transcend” what I or my peers say or believe here and now. I may or may not be able to justify my claims, but I can’t avoid making universal validity claims. So, even on pragmatic grounds, Rorty fails, because he can’t make sense of the starting point for pragmatism – making sense of everyday practices.⁷

Habermas is far more sympathetic to Rorty’s student, Robert Brandom, precisely because – unlike his teacher – Brandom struggles to give an account of truth and objectivity that tries to do justice to our “realist intuitions.” Nevertheless, Habermas is skeptical of Brandom’s neo-Hegelian turn.

Brandom combines Wilfrid Sellars’s inferential semantics step by step with a pragmatics of discourse in order to explain the objectivity of conceptual norms from the perspective of the intersubjectively shared “practice of giving and asking for reasons.” In the end, Brandom is able to do justice to the intuitions underlying epistemological realism only at the price of a conceptual realism that obliterates the distinction between the intersubjectively shared lifeworld and the objective world. This assimilation of the objectivity of experience to intersubjectivity of communication is reminiscent of an infamous Hegelian move. (Habermas 2003, p. 8)⁸

So Rorty’s pragmatism is too contextual, and Brandom’s is too Hegelian. Neither fully appreciates the virtues of Kantian pragmatism. Habermas sees himself as closer to the Kantian pragmatism of Hilary Putnam. He draws upon Putnam’s internal realism and his theory of direct reference to vindicate his own thesis that there is an objective world independent of us about which we can make true claims. “With regard to the objective world, the proposition’s truth signifies a fact – the obtaining of a state of affairs.” We will soon see that, according to Habermas, even Putnam is not a consistent Kantian pragmatist, because he fails – so Habermas claims – to make the important distinction between values and norms. Putnam fails to appreciate the important Kantian distinction between theoretical and practical philosophy and reason.⁹

The question now becomes, How does Habermas account for truth? How does he avoid identifying truth with ideal justification? How does he avoid the aporias of contextualism and neo-Hegelian idealism? Before answering these questions, we must deal with one further topic in order to appreciate fully the contours of Habermas’s Kantian pragmatism.

Moral Rightness

The theory of communicative action and rationality is not yet an explicitly moral theory, even though normative rightness is one of the primary validity claims. Normative rightness is a much more general concept than moral rightness. Not all norms are moral norms. Nevertheless, the theory of communicative action and the formal pragmatic specification of the unavoidable conditions of communication serve as the basis for developing his discourse theory of moral discourse, law, and democracy. Barbara Fultner gives a succinct statement of the core of Habermas’s cognitivist moral theory in the form of a discourse ethics.¹⁰

The core of this theory is the so-called Principle of Universalization, according to which a moral norm is justified if all those affected would assent to it under conditions of an ideal speech situation. Moral norms, unlike ethical values, have a universal and unconditional validity. At the same time, moral rightness is an epistemic notion. That is, it is defined in terms of what rational agents would agree on under (approximately) ideal conditions. (Fultner 2003, pp. viii–ix)

There are several features of the discourse theory of morality that call for comment. Although there is practical (moral) knowledge, it must not be confused or identified with theoretical knowledge about the objective world. “It seems that a cognitivist interpretation of the normative validity of binding norms, which takes account of the inescapable sense of the ‘respect of the law’ as a ‘fact of reason,’ is possible only if we conceive morality on analogy with cognition” (Habermas 2003, p. 239). And when a moral norm is contested, the only proper way to resolve the dispute is by discourse, by argumentation. In this argumentation we seek to justify moral norms. But what does justification mean here? It means that “all those affected would assent to it under conditions of an ideal speech situation.” Once again, it is important to note that Habermas is not making a prediction about what will happen or suggesting that there will not always be moral disagreements; rather, he is specifying a demand and a condition for what counts as the justification of a moral norm. It begins to look as if much of what Habermas claimed with regard to the justification of truth claims is just as relevant to what he has to say about the justification of universal moral norms. Indeed, as we noted earlier, it was primarily Habermas’s concern to develop an epistemic account of “normative rightness” that inspired his epistemic account of “truth.” And it is this analogy between justifying normative validity claims and justifying truth claims about the objective world that initially led Habermas to claim that an epistemic account of truth was sufficient to explicate truth – the claim that he now rejects. This is why we need an account of truth that explains how truth involves more than justification – more than even “ideal” justification. Unless he can satisfy this need, he thinks that we will not be able to escape from the aporias of contextualism and linguistic idealism. Consequently, Habermas tries to give an account of truth that will explain our “ontological way of speaking [that] establishes a connection between truth and reference, that is, between the truth of statements and the ‘objectivity’ of that about which something is stated” (Habermas 2003, p. 254).

The substance of Habermas’s “Kantian pragmatism” now emerges with greater clarity. Like Kant, Habermas claims that there is a sharp distinction between theoretical and practical philosophy and reason (although, of course, they are interdependent). The task of theoretical philosophy is to provide an adequate account of truth, and the task of practical philosophy, an adequate account of normative rightness and, especially, moral norms.

Moral norms are universal; these norms are binding on all human beings.¹¹ Ethics, as distinguished from morality, pertains to my personal life or the life of my group – whether this is an ethnic group, a religious group, or a national group. “Ethics” as Habermas uses this term is always oriented to the first-person singular or plural perspective. There is a plurality of ethical orientations, but there is only a single universal morality. In making and emphasizing the distinction between morality and ethics, Habermas is clearly identifying himself with the Kantian deontological conception of morality and justice that is applicable to all human beings. Theories of the good life – that is, theories about what is good for me or my group – are particularistic; they are not truly universal. Habermas gives the following illustration to indicate the difference between “a universalistic morality of justice and a particularistic ethics of the good life.”

We call the torture of human beings “cruel” not only here for us, but everywhere and for everyone. Yet we feel by no means justified to object against strange child-raising practices or marriage ceremonies, that is, against core components of the ethos of a foreign culture, as long as they do not contradict our moral standards. The latter are those central values that differ from other values in virtues of their universal claim to validity. (Habermas 2003, pp. 228–9)

Concerning the analogy between moral judgments and true descriptive statements, Habermas writes:

Those moral judgments that merit universal recognition are “right,” and that means that in a rational discourse under approximately ideal conditions they could be agreed to by anyone concerned. The analogy to the claim to truth consists in the demand for rational acceptability; the truth of descriptive statements can also come out and be confirmed only in rational discourses that are as comprehensive and persistent as possible. (Habermas 2003, p. 229)

But Habermas warns against blurring the distinction between these validity claims – between rightness and truth. To say that there is an analogy between moral judgments and true descriptive statements because both can be justified by argumentative discourse is not to identify them; there are crucial differences.

The validity concept of moral rightness has lost the ontological connotation of the justification-transcendent concept of truth. Whereas “rightness” is an epistemic concept and means nothing but worthiness of universal recognition, the meaning of the truth of statements cannot be reduced to epistemic conditions of confirmation, no matter how rigorous they might be: truth goes beyond idealized justification. This difference between “truth” and “moral rightness” mirrors the distinction between theoretical and practical reason. (Habermas 2003, pp. 229–30)

Despite Habermas’s many criticisms of the “historical” Kant, we are now in a position to see just how Kantian he is – what he wants to appropriate from the Kantian legacy. First, there is the reconstruction of what he takes to be the viable core of the transcendental project – the formal-pragmatic statement of the unavoidable conditions of speech and action. Secondly, Habermas is Kantian in insisting on a sharp (categorical) distinction between the right and the good. Moral rightness is universal and can be justified only “in a type of discourse that (approximately) meets the ideal conditions that could be agreed to by anyone concerned.” We might say that Habermas is a “communicative Kantian.” Thirdly, he is Kantian in maintaining a strict distinction between theoretical and practical reason, and consequently between theoretical and practical philosophy. But here Habermas significantly departs from the classical pragmatists or, at least, from James and Dewey, who deny any sharp distinction between theoretical and practical philosophy. Most pragmatists – old and new – do not accept this Kantian distinction between theoretical and practical reason. And here Habermas also departs from Hilary Putnam, the contemporary Kantian pragmatist with whom he feels the greatest affinity, because Putnam questions the sharp distinction between values and norms.¹²

Action and Discourse

We are finally prepared to see how Habermas distinguishes rightness from truth. Here is a succinct statement of his guiding intuition:

My guiding intuition can be characterized as follows. On the one hand, we discover the rightness of moral judgments in the same way as the truth of descriptions: through argumentation. We no more have direct access unfiltered by reasons to truth conditions than we do to the conditions under which moral norms merit universal recognition. In either case, the validity of statements can be established only through discursive engagement using available reasons. On the other hand, moral validity claims do not refer to the world in the way that is characteristic of truth claims. “Truth” is a justification-transcendent concept that cannot be made to coincide even with the concept of ideal warranted assertibility. Rather it refers to the truth conditions that must, as it were, be met by reality itself. In contrast, the meaning of “rightness” consists entirely in ideal warranted acceptability. (Habermas 2003, pp. 247–8; emphasis added)

So the primary question now is, How does Habermas account for truth? How does he complement an epistemic discursive concept of truth with a non-epistemic account of truth that satisfies “realist intuitions”? How does he account for the presupposition that there is an objective world that is the same for everyone, and about which we can make true assertions?

In order to give a more adequate theory of truth, Habermas returns to the starting point for pragmatism – how truth claims function within the lifeworld, that is, within the world of everyday actions and practices.

Pragmatism makes us aware that everyday practice rules out suspending claims to truth in principle. The network of routine practices relies on more or less implicit beliefs that we take to be true against a broad background of intersubjectively shared or sufficiently overlapping beliefs. Everyday routines and habituated communication work on the basis of certainties that guide our actions. … As soon as such certainties are dislodged from the framework of what we take for granted in the lifeworld and are no longer naively accepted, they become just so many questionable assumptions. In the transition from action to discourse, what is taken to be true is the first thing to shed its mode of practical certainty and to take on instead the form of a hypothetical statement whose validity remains undetermined until it passes or fails the test of argumentation. Looking beyond the level of argumentation, we can comprehend the pragmatic role of a Janus-faced truth that established the desired internal connection between performative certainty and warranted assertibility. (Habermas 2003, pp. 252–3)

Consider the following mundane example. If I take a subway to my office as I routinely do, I know how many blocks I have to walk to get to the subway, and that the train I take will stop at the same stations that it normally does. All of these are “more or less implicit beliefs” that I take for granted. They function as practical certainties and do not call for any further inquiry or argumentation. But suppose that, when I get to the subway station, I discover that, due to a breakdown, the train is not running. The situation has become problematic for me, and I may initially be somewhat uncertain about what to do. I may start a conversation with strangers about the best alternative way to get to my office. There may even be some dispute (an argument) about the best solution: for example, which is the best bus to take. And after listening to the pros and cons, I decide that it is best to take the Fifth Avenue bus. I walk to the bus stop and proceed with my journey.¹³

This example illustrates the way in which practical uncertainties arise in everyday life and how they may be resolved by discourse (my conversation with strangers about the best way to get to my office). It illustrates the movement from action to discourse and back again to action (taking the bus). But how do the appeal to action in the lifeworld and the move to argumentative justificatory discourse further our understanding of pragmatic truth?

In the lifeworld actors depend on behavioral certainties. They have to cope with a world presumed to be objective and, for this reason, operate with a distinction between believing and knowing. There is a practical necessity to rely intuitively on what is unconditionally held-to-be-true. This mode of unconditionally holding-to-be-true is reflected on the discursive level in the connotations of truth claims that point beyond the given context of justification and require the supposition of ideal justificatory conditions – with a resulting decentering of the justification community. For this reason, the process of justification can be guided by a notion of truth that transcends justification although it is always already operatively effective in the realm of action. The function of the validity of statements in everyday practices explains why the distinctive vindication of validity claims may at the same time be interpreted as the satisfaction of a pragmatic need for justification. The need for justification, which sets in train the transformation of shaken-up behavioral certainties into problematized validity claims, can be satisfied only by a translation of justified beliefs back into behavioral truths. (Habermas 2000b, p. 49)

Let me elucidate and simplify what Habermas is saying in this rather complicated passage. Habermas wants to explicate the relation of justification and truth. He realizes that an epistemic conception of truth – a conception that restricts itself to the discursive justification of truth claims – cannot by itself provide an adequate theory of truth, because, no matter how “ideal” our justification may be, it may nevertheless turn out to be false. So we have to complement this epistemic concept of truth with a non-epistemic concept – or, more accurately, we have to grasp the epistemic and non-epistemic aspects of the concept of truth. Taking his cue from the classical pragmatists, Habermas suggests that we return to the context of everyday practices and actions. In the “realm of action” we are always operating with behavioral certainties. We act as if we know what is true. But sometimes we encounter frustrations and resistances that compel us to question what we have taken as unproblematic. When this happens, we can react in many different ways; but one of our options is to engage in a discourse in which we problematize what we initially took for granted, and seek to resolve the problem by moving to a level of discourse in which we evaluate and justify our truth claims. We cannot account for truth by appealing solely to justificatory discourse; nor can we account for truth simply by appealing solely to the “behavioral certainties of everyday life.” But if we grasp that truth is Janus-faced because it stands between action and discourse, then we can elucidate a notion of truth that transcends justification. By relating action to discourse in this manner, we can understand both why there is a “pragmatic need for justification” and how this need is satisfied by translating discursively established “beliefs back into behavioral truths.”

On one interpretation, what Habermas says seems thoroughly noncontroversial, although one might object to his complicated form of expression. In our everyday lives we implicitly act on all sorts of beliefs that we take for granted – “practical certainties.” Sometimes we are thwarted and face a practical uncertainty. This uncertainty may initiate a discourse in which participants offer reasons for the best solution about what is to be done. If the situation is resolved satisfactorily, I proceed on my way. If this is what it means to vindicate “realistic intuitions,” then it is hard to see what all the fuss is about. Not even Rorty would deny that in our everyday lives we are frustrated and meet all sorts of obstacles, and sometimes we need to engage in inquiry and discussion to work out how to cope. (Rorty, of course, would question whether it is helpful to describe this as commonsense “realism,” or whether any of this requires us to speak about a truth that transcends justification.)

Moral Constructivism and Epistemological Realism

Before probing further what Habermas means by his “Janus-faced concept of truth” and in what sense, if any, it helps us to “vindicate” our “realist intuitions,” I want to consider why it is so important for him to come up with an adequate pragmatic theory of truth. In advancing a pragmatic theory of truth, Habermas is not concerned exclusively with vindicating our “realist intuitions.” He wants sharply to distinguish practical moral discourse, which is concerned with justifying universal moral norms, from theoretical discourse, which is concerned with justifying truth claims. This distinction between these two forms of discourse (and argumentation) is a fundamental doctrine of his Kantian pragmatism. In practical moral discourse, there is nothing in the “objective world” to which we can appeal in order to resolve our disagreements, whereas in theoretical discourse we do appeal to the objective world (in a linguistically mediated manner) to resolve disagreements.

Moral validity claims lack reference to the objective world that is characteristic of claims to truth. This means they are robbed of a justification-transcendent point of reference. The reference to the world is replaced by an orientation toward extending the borders of the social community and its consensus about values. (Habermas 2003, pp. 256–7)

With moral norms, the only relevant consideration is the recognition of the worthiness of norms by those affected by these norms.

For the worthiness of norms to be recognized is based not on an objectively determined agreement of interests that are given, but on participants interpreting and evaluating interests from a first-person plural perspective. The participants can develop norms embodying shared interests only from a We-perspective. This perspective has to be constructed out of a reversible exchange of perspectives of all those affected. (Habermas 2003, p. 268)

So we must not identify or assimilate the type of “objectivity” we can achieve in moral discourse (intersubjective agreement by all those affected) with the strong realist objectivity indicated by an unanticipated reality – by a world that is independent of us. “The ‘objectivity’ of another mind is made of different stuff than the objectivity of an unanticipated reality” (Habermas 2003, p. 256). Theoretical reason is concerned with truth, and it deals with an objective world that is independent of us. Practical reason (in morals) is concerned with constructing and validating universal moral norms. Argumentation is essential for both forms of reason, but we corroborate moral beliefs “by consensually resolving interactions,” and not by “successfully manipulating otherwise independently occurring processes” (ibid.). Habermas’s Kantian pragmatism combines epistemological realism with moral constructivism.

Janus-Faced Truth?

Thus far, I have presented a sympathetic account of Habermas’s Kantian pragmatism, explaining in what sense he is “Kantian” and “pragmatic.” I want to raise some critical issues – primarily about his account of truth, his distinction between action and discourse, and his Kantian distinction between theoretical and practical reason.

We have seen that the core of his pragmatic conception of truth depends on the distinction between action and discourse (argumentation). This distinction is vital for his “Janus-faced” concept of truth. Habermas originally introduced this distinction and his pragmatic theory of truth in developing a critique of Rorty’s contextualism. But here I agree with Rorty’s response about this distinction.

[W]hen Habermas makes a distinction between “two pragmatic roles … played by the Janus-faced concept of truth in action-contexts and in rational discourses respectively,” and when he goes on to say that “the concept of truth allows translation of shaken-up behavioral certainties into problematized propositions,” I would rejoin that he is ignoring Peirce’s point that beliefs are just habits of action. A rational discourse is just one more action-context in which a behavioral certainty evinces itself. There is no Janus-like role to be played, and no translation to be performed. (Rorty 2000a, p. 57)

Unlike Rorty, I do not want to deny that in some circumstances it is helpful to draw a distinction between action and discourse, but we should realize how misleading this distinction can be: how much “discourse” is embedded in action-contexts and how much action is embedded in discourse. Habermas’s pragmatic theory of truth depends on drawing a clear distinction between action and discourse. But if we begin to question the rigidity of this distinction, we question the very idea of a Janus-faced concept of truth. When acting, I do take all sorts of things for granted without questioning them, but action is not dumb. It is – as Dewey says – funded with meaning and what I have learnt from the past. And as I act in the world, I am frequently adjusting my actions as I encounter unanticipated obstacles – as practical certainties turn into uncertainties. When I stop and reflect about how to resolve an uncertainty, I do not necessarily “move into” discourse. When I discover that the subway isn’t running – being acquainted with New York’s bus system – I, after some hesitation, decide to take the Fifth Avenue bus, without engaging in any discourse. But this feature of acting in the lifeworld is just as true for discourse. In any discourse, including sophisticated scientific discourse, I take all sorts of things for granted as “practical certainties,” and I may well discover that what I have taken for granted is in fact problematic. Even in the “lifeworld,” when my anticipations are frustrated, there are many different ways that I can adapt to the situation – and these adjustments are integral to acting in the world. Acting in the world is rarely simply routinized; it involves reflection and deliberation – and occasionally consulting with others about what is to be done. One of the primary aims of the classical pragmatists was to encourage the development of intelligent habits of action that would enable us better to cope with the world; to make our actions more intelligent and imaginative in responding to new situations. But Habermas makes it sound as if action is completely unreflective when he speaks of action and discourse as “two domains” and says: “Only once they make the transition from action to discourse do participants take a reflective attitude and dispute the now thematized truth of controversial propositions in the light of reasons for and against it” (Habermas 2003, p. 39; emphasis added). Furthermore, discourse frequently involves all sorts of actions, experiments, and interventions – as the classical pragmatists also emphasized. Experimentation and intervention are just as important and relevant to “discourse” as they are to “action.” This is why I find Dewey’s description of the pattern of inquiry – where one begins with a “felt difficulty,” moves to specification of a problem, advances hypotheses, then tests these hypotheses in order to resolve an “indeterminate situation” – more illuminating than Habermas’s analysis of the move from action to discourse and back to action. As Dewey stresses, this pattern of inquiry is exemplified in ordinary everyday actions and sophisticated theoretical inquiry (as well as in what Habermas calls “ethical” and “moral” inquiry).¹⁴ But all of this can be described without any reference to the “Janus-faced concept of truth.”

Habermas recognizes the “embeddedness of discourses in the lifeworld,” but nevertheless he claims:

Convictions play a different role in action than in discourse and “prove their truth” in a different way in the former than in the latter. In everyday practices, a prereflexive “coping with the world” decides whether convictions “function” or are drawn into the maelstrom of problematization, whereas in argumentation it depends solely on reasons whether controversial validity claims deserve rationally motivated recognition. (Habermas 2000b, p. 48)

Do “convictions play a different role in action and discourse”? I don’t think so. Whether it is action or discourse, I take all sorts of things for granted – as practical certainties. Talk about a “prereflexive ‘coping with the world’ ” makes it sound as if in such coping there is no reflection and deliberation. But most of our acting in the lifeworld involves some sort of reflection and deliberation – no matter how fleeting and incidental it may be. To speak about the “maelstrom of problematization” exaggerates the difference between discourse and action. “Problematization” is just as characteristic of everyday actions and practices as it is of cultivated argumentation. So what we have here is a difference of degree, ranging from routinized behavior that doesn’t involve reflection (but is typically funded by previous reflection) to abstract forms of discursive argumentation. But this is a continuum – not a movement from action to something that is categorically different, discourse. If we give up this (idealized) distinction between action and discourse, this puts in question the relevance of speaking of the “Janus-faced concept of truth.” I am not denying that, in some contexts, it is useful to introduce a distinction between action and discourse. But if we do so, it is we who make this distinction, and we do so for specific reasons. Consequently, it is misleading to speak of “the different roles played by the concept of truth in the two domains” (Habermas 2003, p. 39).

Habermas exaggerates the difference between action and discourse in order to make his “Janus-faced concept of truth” seem plausible. Consider the following passage:

Although when we adopt a reflexive attitude we know that all knowledge is fallible, in everyday life we cannot survive with hypotheses alone, that is, in a persistently fallibilist way. The organized fallibilism of scientific inquiry can deal hypothetically with controversial validity claims indefinitely because it serves to bring about agreements that are uncoupled from action. This model is not suitable for the lifeworld. … We would step on no bridge, use no car, undergo no operation, not even eat an exquisitely prepared meal if we did not hold assumptions employed in the production and execution of our actions to be true. At any rate, the performative need for behavioral certainty rules out a reservation in principle with regard to truth, even though we know, as soon as the naïve performance of actions is interrupted, that truth claims can be vindicated only discursively – that is, only within the relevant context of justification. Truth may be assimilated neither to behavioral certainty nor to justified assertibility. (Habermas 2000b, p. 44)

Of course, in everyday life we cannot survive on hypotheses alone; but neither is this the case in scientific inquiry. We could not conduct any inquiry – commonsense, scientific, ethical, or moral – unless we took all sorts of background knowledge and implicit convictions for granted – or, in Habermas’s language, acted as if there were “behavioral certainties.” (This is a lesson that Peirce taught us long ago.) It is misleading to say that the “organized fallibilism of scientific inquiry can deal hypothetically with controversial validity claims indefinitely because it serves to bring about agreements that are uncoupled from action.” Scientific inquiry is rarely, if ever, uncoupled from action, because performing experiments and testing our hypotheses are actions integral to scientific inquiry. I would not step on a bridge unless I assumed it were relatively solid, but I would not conduct a scientific experiment unless I assumed that the instrument I am about to use is functioning properly. In both cases I may well discover that my “practical certainties” have been problematized.

But, even more troubling, it isn’t clear what Habermas really explains with his non-epistemic conception of pragmatic truth. Suppose, for the moment, we accept Habermas’s characterization of practical certainties in the lifeworld – that we act as if many of our implicit beliefs are unquestionably true; that we bump up against a world that frustrates our expectations and compels us to question what we took to be certain; and that this questioning gives rise to a discourse in which we seek to evaluate and justify claims to resolve the problematic situation. In what sense does this contribute to an explanation that accounts for the philosophical understanding of realism and truth? Habermas says that, although justification (even if it were ideal) is not to be identified with truth, nevertheless there is an “internal connection between justification and truth,” which explains “why we may, in light of the evidence available to us, raise an unconditional truth claim that aims beyond what is justified.” But why not drop this talk of “unconditional truth” and simply say that no matter how rigorous and comprehensive our justifications, we may discover that they are not true? This is what Rorty calls the “cautionary” use of ‘true.’ It is hard to see what work the appeal to “unconditional truth” does. If we do discover that something we take to be thoroughly justified is false, it will be because we have reasons to question our justification, not because we presuppose that there is an “unconditional truth.” For example, before the United States invaded Iraq in 2003, it seemed to be true that Saddam Hussein had some weapons – at least some chemical weapons – of mass destruction. After all, he had already used chemical weapons against the Kurds. But it turned out that this (justified) belief was false. We falsely believed that he still had these chemical weapons at the time of the Iraq invasion. Consequently, we now have good reasons to reject our earlier belief. We can now say that we have good reasons to say it was false, because no such weapons were discovered. And, of course, further investigation (the discovery of a hidden cache of such weapons) may, in the future, necessitate revising our current beliefs. We don’t need to appeal to an “unconditional truth” or “justification-transcendent truth” to make sense of why we reject a former justification. Not only do we reject former “justifications,” we sometimes also reject former criteria or standards of justification – as we learn from the history of science.¹⁵

We can put the difficulty in a slightly different way. Habermas tells us: “It is the goal of justifications to discover a truth that exceeds all justification” (Habermas 2003, pp. 33–4). But the phrase “exceeds all justification” is crucially ambiguous. It may simply mean that it is the goal of justification to discover a truth (full stop). But, of course, our justifications may fail to achieve this goal. This is why we need a cautionary use of ‘true.’ But Habermas thinks that something more is needed – “unconditional truth.” He thinks that it is a deep and troubling question to ask: how is it that our justifications can tell us the truth about the “real world,” especially since we know that no justification is sufficient to guarantee this truth? Although I have serious reservations about the way in which Rorty seeks to dismiss and/or trivialize the very idea of truth, I do think his critique of Habermas with regard to this issue on target.

There is, to be sure, something unconditional about of truth. The unconditionality is expressed by the fact that once true, always true: we regard people who use the word in such expressions as ‘true then, but not now’ as using it incorrectly. Since “once justified, always justified” is obviously false, one can indeed express the contrast between truth and justification as a contrast between the unconditional and the conditional. But the unconditionality in question does not provide a reason for the fact that the cautionary use of ‘true’ is always apropos. … The whole pragmatic force of the claim that truth is not conditional is to express willingness to change one’s mind if circumstances alter, not to explain or justify this willingness. We are not contritely fallible because we are in awe of the unconditionality of truth. Rather, to speak of truth as being unconditional is just one more way of expressing our sense of contrite fallibility. (Rorty 2000a, p. 57; emphasis added)

Although we can speak of “unconditional truth” or “true” as a predicate that a proposition “cannot lose,” it is difficult to see that appeal to such notions helps to explain anything. Habermas also declares that the “nonepistemic concept of truth, which manifests itself only operatively, that is, unthematically, in action, provides a justification-transcendent point of reference for discursively thematized truth claims” (ibid.). But in what sense do “behavioral certainties” – the facts we implicitly assume to be true – serve as a point of reference for a justification-transcendent concept of truth? Why not simply say that when something that we take to be practically certain in the lifeworld becomes problematic, we try to figure out what to do and say? And when Habermas goes on to say that “it is the goal of justifications to discover a truth that exceeds all justification” (ibid.), our perplexity is only increased. Does this mean anything more than that we expect (or hope) “the truth” that we have presumably discovered and sought to justify really is the truth – that our justification will withstand criticism and refutation?

Habermas is right when he claims that in the everyday lifeworld of actions, coping requires that we take all sorts of things for granted as practical certainties, and that reflection is provoked when these certainties become problematic. This was the starting point for the classical pragmatic conception of inquiry – when real doubt arises. (Not even Rorty would deny this. If something didn’t resist and frustrate us, we could not even speak about coping.) But the basic question is whether the appeal to “unconditional truth” or “justification-transcendent truth” does any real work besides expressing “our willingness to change our minds if circumstances change.” And if we are asked why we are willing to change our minds, why we are willing to reject justifications that we once accepted, we can answer: because we have learned over and over again that what we thought was thoroughly justified has turned out to be mistaken. It seems that – pragmatically speaking – Habermas’s “realistic intuitions” come down to little more than what Peirce, in his categorical scheme, called Secondness – the non-epistemic sense of resistance, brute force, and compulsion, “the absolute constraint upon us to think otherwise than we have been thinking that constitutes experience” (1.336).¹⁶

Moral Constructivism Again

I have raised a number of questions about Habermas’s pragmatic “Janus-faced” concept of truth, which is intended to clarify the validity claim of truth and help us to distinguish theoretical reason from practical reason. I now want to turn to the other pillar of his Kantian pragmatism – his understanding of normative rightness as a validity claim and his moral constructivism – in order to consider some of his key claims about practical (moral) reason. There are many issues that can (and) have been raised by critics. Do the Habermasian distinctions between norms and values, between rightness and goodness, between morality and ethics, really hold up? Many of Habermas’s critics – including Rorty, Putnam, Brandom, McCarthy and I – have criticized Habermas on some or all of these issues.¹⁷

I will limit my discussion to Habermas’s claim that truth and objectivity play a different role in theoretical and practical (moral) discourse, in order to challenge the thesis that is so fundamental for his version of Kantian pragmatism – that there is a fundamental distinction between theoretical and practical reason. I have already suggested that Habermas proposes his version of a pragmatic concept of truth not only to avoid Rorty’s contextualism and Brandom’s neo-Hegelianism, but also because of his concern to call attention to the analogy – yet fundamental difference – between truth and rightness. He thinks that we are tempted to confuse the two if we assume – as many analytic philosophers have – that the paradigm of a validity claim is the assertion of truth. “Realist intuitions” can be vindicated when it comes to matters of truth, but they have no place in relation to moral norms. Now why is this distinction so important for Habermas? Why does he have a “horror” of moral realism? He believes that moving in this direction, obliterating or softening the distinction between truth and rightness, would undermine his communicative theory of action (with its different types of validity claims) and his discourse theory of ethics. “The concept of ‘normative rightness’ can be reduced without remainder to rational justification under ideal conditions. It lacks the ontological connotation of reference to things about which we state facts” (Habermas 2003, p. 42).

Habermas concedes that there is a proper role for speaking about “moral truth” and “moral objectivity.” But he insists that the meaning of “truth” and “objectivity” when we are speaking from a theoretical perspective is radically different from when we speak from a practical perspective. So now we must ask whether the role of the concepts of ‘objectivity’ and ‘truth’ are as different in theoretical discourse and practical (moral) discourse as Habermas claims.

Why Habermas Rejects Moral Realism

Before turning directly to these issues, I want to consider why the expression “moral realism” is loaded and misleading. (To the best of my knowledge, none of the classical pragmatists ever used this expression.) We might say that there is a picture here that holds one captive. And it certainly seems that this is the picture that Habermas has in mind when he categorically rejects moral realism and defends moral constructivism. Earlier I cited the passage where Habermas says: “What we want to express with true sentences is that a certain state of affairs ‘obtains’ or is ‘given’. And these facts in turn refer to ‘the world’ as the totality of things about which we state facts.” If we take this ontological way of speaking as presenting the paradigm of realism, then we can certainly understand why Habermas rejects moral realism. There don’t seem to be any states of affairs that “obtain” in “the world” that correspond to our moral norms. When, for example, Habermas critiques Putnam’s notion of “ought-implying facts,” he declares that “to extend cognitive realism to values is to postulate facts that are ‘queer’ (in Mackie’s sense) inasmuch as they run counter to our grammatical intuitions” (Habermas 2003, p. 224).¹⁸

But to appreciate fully why Habermas finds moral realism so objectionable and defends a version of Kantian constructivism, we need to grasp a deeper Habermasian theme: his understanding of how modernity has transformed the issue of moral autonomy and authority. Kant’s great insight regarding moral norms is not only that they are universal, but that there is no higher moral authority than the appeal to our own practical rationality. The key to understanding moral autonomy is self-legislation. Any appeal to an authority other than our own practical reason is a form of heteronomy. Habermas accepts this claim, and he thinks of his own discourse theory as a continuation of this Kantian project. Moral constructivism shows that we, and we alone, have the responsibility for justifying universal moral norms. “If moral claims to validity nonetheless owe their binding force to something unconditional and analogous to truth, then the orientation toward ever increasing inclusiveness of other claims and persons must somehow compensate for the missing reference to the objective world” (Habermas 2003, p. 43). Because the expression “moral realism” is so loaded with misleading connotations and calls to mind a world of “queer” facts, I think it best to drop the expression altogether and concentrate on the role of truth and objectivity in moral argumentation.

Let us consider again the example that Habermas uses to illustrate the universality of moral norms, which he distinguishes from the particularity of ethical values:

We call the torture of human beings “cruel” not only here for us, but everywhere and for everyone. Yet we feel by no means justified to object against strange child-raising practices or marriage ceremonies, that is, against core components of the ethos of a foreign culture, as long as they do not contradict our moral standards. The latter are those central values that differ from other values in virtue of their universal claim to validity. (Habermas 2003, pp. 228–9)

I do not want to deny that there are some “central values that differ from other values in virtue of their universal claim to validity,” where this is understood as being applicable to all human beings rather than to some specific group, people, or nation. The primary issue concerns how Habermas interprets this difference. He thinks that there is a sharp distinction between ethics and morals. The “domain” of ethics is that of values, and implicitly or explicitly they are limited to first-person singular or plural questions. And the “domain” of morals – or the moral point of view – concerns binding moral norms for which we claim universal validity.

Ethical-existential questions – what is best for me overall? who am I and who do I want to be? – arise from the first-person perspective just as ethical-political questions about the collective identity and way of life do. Casting the issue in terms of ethics already means selecting the context of one’s life history or of our collective life as the point of reference for this kind of hermeneutic self-reflection. This explains why practical wisdom is intuitive as well as context-dependent. … And since such reflection guides what we do or don’t do within the horizon of our own lifeworld, there is no universal validity claim connected with ethical wisdom. (Habermas 2003, p. 228)

I find this sharp distinction between ethics and morality unpersuasive. And this becomes evident if we turn to “real” ethical perspectives, especially those embedded in religious orientations. Ethical orientations embedded in religious world views have both particularistic and universal aspects (and sometimes it is difficult to draw a sharp line between these). As Jews, Christians, or Muslims, there are particular values that we share with our co-religionists, and we may also have binding obligations to them. But every one of the great religions also incorporates universal binding obligations to one’s fellow human beings. And if I take my ethical responsibilities seriously as a member of a religious community, then I am committed to norms that apply to all human beings – whether or not the person is a member of my religious community. Habermas is right to point out that there are more particularistic and moral universal validity claims in every living ethical orientation, but it is misleading to speak of “ethical world views” as particularistic and “the moral point of view” as universalistic, and to contrast “a universalistic morality of justice” with “a particularist ethics of the good life” (ibid.).¹⁹ Here one would like Habermas to be more pragmatic and recognize that our values and norms form a dynamic shifting continuum such that some are more particularistic and some more universalistic. Furthermore, we need to recognize just how open-ended “universalistic” moral claims are. It is not just that there are learning processes about what count as universal moral norms – a point that Habermas emphasizes. But in the course of history, unanticipated consequences and structural transformations require radical rethinking of the very meaning of morality, responsibility, and justice. Adorno argued that “after Auschwitz” we need a new categorical imperative, and Hans Jonas argued that, as a result of modern technology, we need a new imperative of responsibility. At its best, the universalistic characteristic of what Habermas calls the moral point of view is an aspiration that can take on radically different, and even incompatible, concrete meanings in novel historical circumstances. At its worst, it is merely formal and empty.²⁰

Let us consider the role of truth and objectivity in practical discourse. Assume – as Habermas does – that the prohibition engaging in the practice of torture is a good candidate for a universal binding moral norm.²¹ Consider some of the heated arguments that have been taking place recently in the United States about “waterboarding”: whether it is a form of torture and whether or not it is a practice that can be morally justified – or at least be considered morally permissible – when used on suspected “terrorists.” In these practical debates, issues arise about the practice of “waterboarding,” its history and how it is used, and what are its effects on victims, what sorts of physical and psychological pain it inflicts, etc. Answering these questions in a responsible way requires making validity claims about what is true about objective practices. Now, of course, the type of facts that I will take to be relevant in arguing about the propriety of waterboarding will be quite different from the type of facts that are relevant in deciding the best way to get to my office; but I fail to see that Habermas has given us any reason to think that the meaning of ‘true,’ ‘justify,’ and ‘objective’ differs when we engage in these inquiries. There are factual issues concerning the practice and consequences of waterboarding.

Habermas sometimes suggests that the tasks of justifying and applying must be carefully distinguished. But this distinction, which is sometimes useful, can obscure the difficult practical issues that we have to face. It is not as if we have clear or fixed criteria for determining what constitutes torture, and can then decide whether they are applicable to the practice of waterboarding. The hard – and contestable – practical issue is to decide whether waterboarding should be judged as torture. We can’t have a serious discussion about what moral norms ought to be binding unless we know what we mean by these norms and what are their consequences – the very meaning of such norms is always, in principle (and frequently, in fact), open to further determination and argumentation.²² Habermas stresses the importance of “learning processes” in determining moral norms; but these learning processes frequently involve coming to a better understanding of the factual consequences of moral norms.

But still Habermas might object. There clearly is a difference between arguing about the truth of facts (even if these facts are relevant to practical disagreements) and arguing about universal moral norms. Again, the point is not to deny this difference, but rather to question how to interpret it. When we have to decide “moral,” “ethical,” or “merely pragmatic” issues, do we really engage in different types of practical deliberation? One wishes that Habermas had been less “Kantian” and more “pragmatic.” Dewey, for example, characterized practical judgments as “judgments of a situation demanding action. There are, for example, propositions of the form: M.N. should do thus and so; it is better, wiser, more prudent, right, advisable, opportune, expedient, etc. to act thus and so. And this is the type of judgment I denote practical” (Dewey 1998, p. 234). This doesn’t mean that judging what is right to do is the same as judging what is expedient. But insofar as they are both practical judgments, they form a continuum and differ only in degree. Furthermore, there is no dichotomy between theory and practice.

The depersonalization of the things of everyday practice becomes the chief agency of their repersonalizing in new and more fruitful modes of practice. The paradox of theory and practice is that theory is with respect to all other modes of practice the most practical of all things, and the more impartial and impersonal it is, the more truly practical it is. And this is the sole paradox. (Dewey 1998, p. 268)

Although I have been critical of the way in which Habermas vindicates his “realistic intuitions,” I certainly endorse his shift to the pragmatic context of action in explaining what he means by ‘truth.’ But I think he also needs to shift more radically to the context of action and practice in his analysis of the justification of moral norms. In the context of action we appeal to (tentative) norms to guide our action. When we confront new problems, we have to clarify and defend moral norms. We do not first discursively justify moral norms and then apply them to new contexts. There is a more dynamic and dialectical relation between confronting practical choices, clarifying the precise meaning of the norms that guide us, and seeking to justify them.

At times, especially when he describes his “epistemological realism,” Habermas sounds as if he is still wedded to what Dewey called “the spectator theory of knowledge.” This becomes evident in the way in which he speaks about the “objective world.” Typically, he limits the expression “objective world” to what empirical descriptive judgments describe. “Empirical judgments say how things are in the objective world whereas evaluative judgments enjoin us to value or treat something in our lifeworld in the same way or other” (Habermas 2003, p. 224). If this is taken to be the primary or privileged use of ‘objective,’ then, of course, we would have to conclude that value and normative judgments do not describe the objective world – or if we speak of them as “objective,” we are using ‘objective’ in a different sense. But we can challenge this limited or privileged use of the expression ‘objective.’²³ Whether we are speaking about empirical judgments or normative judgments, we can speak of them as ‘objective’ if they can properly be justified – insisting, of course, that any specific justification may turn out to be unwarranted.

There is a tendency in Habermas, despite his disclaimers, to reify fact talk, to think that facts are simply the way in which we (correctly) describe the world. “What we want to express with true sentences is that a certain state of affairs ‘obtains’ or is ‘given’. And these facts in turn refer to ‘the world’ as the totality of things about which we may state facts” (Habermas 2003, p. 254). But Habermas might have developed a more pragmatic understanding of facts viewed from the perspective of agents (inquirers) solving problems if he had adopted Dewey’s “operational” perspective. Dewey writes:

What is meant by calling facts operational? Upon the negative side what is meant is that they are not self-sufficient and complete in themselves. They are selected and described … for a purpose, namely statement of the problem involved in such a way that its material both indicates a meaning relevant to resolution of the difficulty and serves to test its worth and validity. … They are not merely results of operations of observation which are executed with the aid of bodily organs and auxiliary instruments of art, but they are particular facts and kinds of facts that will link up with one another in the definite ways that are required to produce a definite end. (Dewey 1981, p. 234)

When facts are viewed in light of their pragmatic role in problem-solving, then they are just as relevant and essential for empirical description as they are for clarifying and justifying moral norms.

I have argued that when Habermas’s “epistemological realism” is analyzed, it amounts to little more than the recognition of what Peirce called Secondness – that we encounter resistances in our “theoretical” and “practical” dealings with the world, resistances that limit our “theoretical” and “practical” justifications. “Truth” and “objectivity” are just as relevant to arguments about the empirical character of the world as they are to serious practical disputes about what ought to be morally binding. There are, of course, differences between different types of inquiry, but these differences have more to do with the context of inquiry than with the meanings of ‘true’ and ‘objective.’ Habermas’s “Kantian pragmatism” is an unstable stopping stage. Habermas speaks about moving from Kant to Hegel and back again to Kant. Many of my criticisms of Habermas’s Kantian dichotomies are in the spirit of Hegel. But the main point is not to move back again to Kant, but to move forward to a more dynamic, flexible pragmatism in the spirit of Dewey and Mead. Dewey and Mead clearly discerned – despite their many limitations – that although distinctions play a crucial role in philosophy, we must understand their functional changing roles in human experience, inquiry, and discourse.²⁴ They – and I fully agree with them – affirm the continuity between theoretical and practical reason.

Notes

1 Writing about his college years, Peirce says: “The C.d.r.V. [Critique of Pure Reason] chiefly occupied my mind for three years; and it was several years more before I began to see that it was mistaken. Before I left college, I almost knew the book by heart; and certainly no day passed without my spending two or three hours in hard thinking about it” (cited in Ketner 1998, p. 139).

2 Habermas’s remarks about his “weak naturalism” are extremely sketchy. He never spells out in detail the precise meaning of this “background assumption”; nor does he provide arguments to show that his “weak naturalism” can be defended against the claims of “strong (reductive) naturalists.” But his naturalism is similar to the naturalism of the American pragmatists. He claims that distinguishing the methodological distinction between a participant’s perspective (second person) and an observational perspective (third person) is “ontologically neutral.”

This blanket assumption of an evolutionary continuity that permeates culture, as it were, refrains from making any philosophical assumptions about the relationship of mind and body (in the sense of eliminative or reductive materialism, for example); on the contrary, it keeps us from reifying a difference between methodological approaches that are themselves ontologically neutral. As long as we cast the issue in transcendental terms, we have to distinguish sharply between the hermeneutic approach of a rational reconstruction of the structures of the lifeworld, which we undertake from the perspective of participants, and the observation-based causal analysis of how these structures naturally evolve. (Habermas 2003, p. 28)

3 In his article “From Kant’s ‘Ideas’ of Pure Reason to ‘Idealizing’ Presuppositions of Communicative Action: Reflections on the Detranscendentalized ‘Use of Reason’ ” (in Habermas 2003, pp. 83–130) Habermas explains in detail what he means by the “idealizing presuppositions of communicative action” and develops a genealogical account of them – showing how they can be related to Kant’s Ideas of Reason. His reconstruction indicates how indebted he is to a pragmatically “detranscendentalized” Kant.

4 Habermas has always engaged in discussion with his critics and has modified his views in response to what he takes to be valid in these criticisms. His revision of his epistemic theory of truth exemplifies his deep fallibilistic commitment.

5 Much of the confusion and criticism of Habermas’s understanding of truth has its source in Habermas 1973. Although he rejected both correspondence and coherence theories of truth, he nevertheless coined the unfortunate phrase a “consensus theory of truth.” He now acknowledges that his primary concern had been an epistemic understanding of truth – justifying true claims about the world. The objections raised against his earlier discussion have prompted him “to revise the discursive conception of rational acceptability by relating it to a pragmatically conceived, nonepistemic concept of truth, but without thereby assimilating ‘truth’ to ‘ideal assertibility’.”

Despite this revision, the concept of rational discourse retains its status as a privileged form of communication that forces those participating in it to continue decentering their cognitive perspectives. The normatively exacting and unavoidable communicative presuppositions of the practice of argumentation now as then imply that impartial judgment formation is structurally necessary. Argumentation remains the only available medium of ascertaining truth since truth claims that have been problematized cannot be tested in any other way. There is no unmediated, discursively unfiltered access to the truth conditions of empirical beliefs. (Habermas 2003, p. 38)

It is frequently claimed (by Habermas, Putnam, McDowell, and many others) that the pragmatic “consensus theory of truth” originated with Peirce. The passage frequently cited is Peirce’s early (1878) statement from “How to Make Our Ideas Clear”: “The opinion which is fated to be agreed to by all who investigate, is what we mean by the truth, and the object represented in this opinion is the real” (Peirce 1992, p. 139). But Cheryl Misak has shown that this early statement does not do justice to Peirce’s reflections on ‘truth’:

C. S. Peirce, the founder of pragmatism, argued that a belief is true if it would be ‘indefeasible’, or would not be improved upon, or would never lead to disappointment, no matter how far we were to pursue our inquiries. Although he occasionally articulated this view of truth in terms of a belief’s being fated to be believed at the end of inquiry, on the whole he tried to stay away from unhelpful ideas such as the final end of inquiry, perfect evidence, and the like. (Misak 2007, p. 68)

See also Misak 1991.

6 For the reasons why Habermas speaks of the world as the totality of things, not facts, see Habermas 2003, pp. 30–6. He wants to avoid the type of representationalism that reifies facts. He wants to defend epistemic realism, but not metaphysical realism.

7 For Habermas’s critique of Rorty and Rorty’s response, see their exchanges in Brandom (ed.) 2000.

8 Brandom has defended himself against the criticism that he “obliterates the distinction between the intersubjectively shared lifeworld and the objective world.” See the exchange between Habermas and Brandom in the European Journal of Philosophy (Habermas 2000a and Brandom 2000b).

9 Doubts may be raised about whether Habermas’s understanding of realism and the objectivity of the world is really compatible with Putnam’s understanding of conceptual relativity and conceptual pluralism. For an explication and defense of these views, as well as Putnam’s “obituary” for ontology, see Putnam 2004.

10 When Habermas initially developed his discourse theory of ethics, he had not yet fully articulated his important distinction between ethics and morality. He worked out this distinction in his essay “On the Pragmatic, the Ethical, and the Moral Employments of Practical Reason.” Habermas explains how practical reason may be employed in three very different ways – and why it is essential to distinguish these three ways. In his preface to Justification and Application he writes: “Since then it would be more accurate to speak of a ‘discourse theory of morality,’ but I retain the term ‘discourse ethics,’ which has become established usage” (Habermas 1993, p. vii).

11 Habermas makes it clear that “discourse ethics situates itself squarely in the Kantian tradition. … Admittedly, it adopts a narrowly circumscribed conception of morality that focuses on questions of justice” (Habermas 1993, pp. 1–2).

12 In his critique of Putnam, Habermas writes: “Whereas in metaphysics and epistemology [Putnam] proceeds along the lines of a linguistic Kantianism, in practical philosophy, he takes a pragmatist reading of Aristotle as his point of reference. Here, eudaimonia – human flourishing – has the last word. Putnam understands autonomy in the classical sense of leading a reflective life rather than in the Kantian sense of rational moral self-legislation. By putting my description of Putnam’s philosophy in these terms, I am implicitly raising the question of how high the price of this split loyalty is. Would Putnam the pragmatist not be better off if he remained a Kantian all the way?” (Habermas 2003, p. 214). For the dispute between Putnam and Habermas concerning values and norms, see Putnam’s “Values and Norms,” in Putnam 2002a, pp. 111–34, and Habermas’s reply, “Norms and Values: On Hilary Putnam’s Kantian Pragmatism,” in Habermas 2003, pp. 213–36.

13 According to the three different employments of practical reason that Habermas discusses in “On the Pragmatic, the Ethical, and the Moral Employments of Practical Reason,” the reasoning involved in figuring out how to get to my office is “pragmatic.” Pragmatic reasoning is required in those cases where “we look for reasons for a rational choice between different available courses of action in the light of a task that we must accomplish if we want to achieve a certain goal” (Habermas 1993, p. 2).

14 Habermas prefers to speak about “discourse” rather than “inquiry,” because he thinks that “inquiry” can be monological and doesn’t fully bring out the participants’ second-person perspective in communication and discourse. But pragmatists, from the time of Peirce, have always emphasized that inquiry is essentially a social process involving a “community of inquirers,” and consequently encompasses a second-person perspective.

15 Although I am drawing on Rorty for some of my critiques of Habermas, I want to make it clear that I do not accept his extreme contextualism. I do not agree with his claim that the appeal to “good reasons” amounts to little more than an appeal to what some sociological group contingently accepts as good reasons.

16 See pp. 131–6 for my discussion of Secondness in Peirce. See also Steven Levine’s perceptive criticism of Habermas’s “Janus-faced” theory of truth (Levine, forthcoming).

17 Following Habermas, I have characterized his view of morality as constructivist. But, of course, there are many varieties of moral constructivism, including John Rawls’s version. Maeve Cooke notes that Habermas’s moral construction “defines justice as an argumentatively achieved consensus.” “[A] consensus achieved under ideal justificatory conditions guarantees validity.” She argues that this strong version leads to all sorts of problems, and that Habermas should adopt a weaker version – one where intersubjective agreement is a necessary but not a sufficient condition to justify the validity of moral norms (Cooke 2003, p. 288).

18 In fairness to Putnam, it should be noted that he also rejects this picture (caricature) of moral realism. See his “Values and Norms,” in Putnam 2002a, pp. 111–34.

19 I have developed this criticism of Habermas’s dichotomy between ethics and morality in Bernstein 1996.

20 One may seriously wonder – given the complex global world in which we live, with its conflicting claims about how to even frame moral issues and questions of justice – whether a formulation such as the following has any real normative force: “Only a maxim that can be generalized from the perspective of all affected counts as a norm that can command general assent and to that extent is worthy of recognition, or, in other words, is morally binding” (Habermas 1993, p. 8).

21 There are some persons who think that torture is ‘justified’ or, at least, permissible in extreme situations where we have good reasons to believe that it will extract information that will help save lives.

22 Habermas is certainly aware of the distinction between justifying a moral norm and applying it. But he thinks that these call for two different types of argumentation.

[E]very justification of a norm is necessarily subject to the normal limitations of a finite, historically situated outlook that is provincial in regard to the future. Hence a forteriori it cannot already explicitly allow for all of the salient features that at some time in the future will characterize the constellations of unforeseen individual cases. For this reason, the application of norms calls for argumentative clarification in its own right. In this case, the impartiality of judgment cannot again be secured through a principle of universalization; rather, in addressing questions of context-sensitive application, practical reason must be informed by a principle of appropriateness [Angemessenheit]. What must be determined here is which of the norms already accepted as valid is appropriate in a given case in the light of all the relevant features of the situation conceived as exhaustively as possible. (Habermas 2003, pp. 13–14)

But Habermas fails to emphasize the intricate dialectical relationship between justification and application. The very meaning of a moral norm is determined by what we take to be its range of application. We cannot fully anticipate unexpected applications and their consequences. Since the time when Peirce first proposed his “pragmatic maxim,” pragmatists have always emphasized how consequences affect the meaning of a concept.

23 Putnam also questions Habermas’s limited conception of objectivity. See my discussion of Putnam, pp. 161–7.

24 I mention Mead here for two reasons. First, Mead plays a fundamental role in the development of Habermas’s Theory of Communicative Action. Second, in his critique of Putnam, Habermas cites Mead as a pragmatist who gives a constructivist account of the moral point of view.

Mead’s intersubjectivist reading of the categorical imperative emphasizes the necessity of decentering one’s given ego or ethnocentrically limited interpreted perspectives. Reciprocal perspective-taking makes one’s own position dependent on the consideration of the polycentric structure of how all other parties understand themselves in the world. In this process, the normative validity of binding norms is understood in the sense of worthiness of universal recognition. … For the gentle force of inclusive reciprocal perspective-taking is embedded in the pragmatic presuppositions of discursive practice, on which justification of all beliefs – be they empirical or mathematical, evaluative or moral – depends. Of course the need to decentralize becomes particularly significant with regard to questions of justice. (Habermas 2003, pp. 234–5)

I agree with what Habermas says about Mead, but I also think it important to call attention to what Habermas does not say. Mead’s analysis of the “generalized other” begins with situated local communities. In Mead (as in Dewey) there is no sharp distinction between local “ethical” communities and a more universal “moral” community. By extending the boundaries of local communities, we strive to achieve greater inclusiveness and universality. Neither Dewey nor Mead would accept a sharp distinction between ethics and morality. Both seek to do justice to the local (particular) and universal dimensions of communal ethical life.