公法评论

Contemporary Philosophy

Habermas on Virtue *

ABSTRACT: Although Habermas has never worked out a conception of virtue and indeed criticizes this notion whenever he uses it, his theory crucially depends on the virtuous attitude of participants in discourse — be it in the realm of democracy and law or that of morality. In this paper, in which I deal only with the ethical foundations of morality, I argue first that the norms of discourse which are gained from a presuppositional analysis of speech as such have to be complemented by the sensitive perception on the side of the recipients. Only when the claims are understood in their full significance for the speaker does the discourse live up to the ideal which is already anticipated in every speech act. This presuppositional analysis shows secondly that it is mainly the virtuous attitude that is morally relevant and not those capacities for acting morally that the agent already possesses. However, the virtuous attitude genuinely entails the obligation to strive to perfect all those capacities that enable us to sensitively understand the other's claim. A discussion of the (discursive) capacities that have to be promoted leads to the singling out of sensitivity as contrasted to immediate empathy. It is the reflexive transcendence of the agent's evaluative patterns that allows this emotion to sit well with a post-conventional morality. Finally, I discuss a possible caveat of Wellmer and Habermas who might claim that the proposed conception of virtue would only be valid for participants in discourse, but not for agents acting in the life-world. However, because the discursive virtue is of wide latitude, it does not fall under this objection.

The last several years have seen an impressive revival of interest in the concept of virtue. In contrast, the discourse ethics of Jürgen Habermas seems to be especially inadequate to account for the ethical significance of virtue, because by emphasizing intersubjective procedures it plays down the importance of individual qualities. In fact, Habermas uses the term 'virtue' rarely, and where he does use it, he criticizes it harshly for presupposing a shared conception of the good life. (1) However, as I will argue in this paper, there is a conception of deontological virtue which is not only compatible, but indispensable for Habermas's theory. Thus, although he attempts to externalize virtue as much as possible into procedures, discourse ethics has to presuppose that the participants possess a genuinely virtuous attitude. In practical discourses the willingness to reach a solution acceptable to all implies that one does as much as one can to sensitively understand the claims of the other. An intersubjectively transformed 'good will' is thus shown to entail a quasi-transcendental 'duty of virtue' (I). The presuppositional analysis also reveals that this 'wide duty' primarily demands to pursue the cultivation of the needed emotional patterns as much as one can. Although this reading stresses the intentional dimension of virtue, it surely implies a high esteem for those who have already managed to achieve the goal of a virtuous character (II). In a third step I argue that it is not immediate empathy, but rather reflexive sensitivity that is demanded by discourse ethics (III). Finally, I discuss a possible caveat of Habermas and Wellmer who might claim that the proposed conception of a quasi-transcendental 'duty of virtue' would only be valid for participants in discourse, but not for agents acting in the lifeworld. However, because the discussed duty is of wide latitude it does not fall under this objection (IV).

(I) Very roughly, in Kant virtue is the strength of the will to act only on maxims that accord with the moral law and to withstand counterstruggling inclinations. However, the obligatory actions, omissions and ends issued by the Categorical Imperative can be detected by the lonely subject herself. It is this allegedly monological conceptualization of the moral point of view (2) that has provoked the discourse ethical criticism. According to Habermas it is extremely likely that interests and needs will be distorted through individual perspectives or will be even completely neglected. Thus, an analysis of the necessary presuppositions of speech as such is not only supposed to show that the universalization test has (ideally) to be carried out by all affected, but to justify the universalization principle in the first place. Roughly the argument proceeds as follows: Already by delivering a speech-act we raise a 'validity claim' (be it one of propositional truth or normative rightness). We necessarily imply, lest our statement is intended as a nonsensical utterance for the fun of simply uttering, that it is backed with good reasons. Reasons, however, to be counted as good or valid would have to withstand detailed scrutiny. Thus, our validity claim can in principle not be restricted to the person(s) to whom it was raised, but implies that it must be acceptable to every reasonable person. If we honestly search for 'truth' or 'rightness' we have, in order to consider every possible (counter)argument, in principle to engage in an intersubjective dialogue that allows all others to criticize our proposals and to bring forward their own. Excluding others therefore amounts to a 'performative contradiction'. Habermas as well as Apel thus engage in a reconstruction of what it means to deliver speech-acts, be it in the unthematized contexts of the lifeworld or the reflective stage that follows a problematization, namely discourse.

However, Habermas follows Alexy in reconstructing the necessary presuppositions implied in speech acts as rules. And although he is at pains to make clear that they are not, such as chess rules, constitutive of the argumentative practice, the usage of 'rule' is ambiguous also in a second sense. Habermas's presentation might be seen as proposing rules of action. This interpretation of the rules of discourse as only demanding certain outward performances could draw on Habermas's remark that the counterfactual presuppositions have to be approximated by institutionalizing discourses which can be empirically studied in "authorizations, exemptions, and procedural rules". (3) However, Habermas reminds us not to confuse these measures with the rules of discourse itself. This is of utmost importance because the necessary presuppositions also entail demands of inner attitudes that cannot be captured by rules regulating action. Habermas himself calls for institutionalization because "participants are not Kant's intelligible characters but real human beings driven by other motives in addition to the one permitted motive of the search for truth.(emph. add.)" (4) In the following I want to argue that by raising a validity claim we already presuppose that we have to approach the others with a certain virtuous attitude.

The equal chance of everyone to articulate her own point of view is not sufficient to fully exhaust the idea of a universal role-taking. If it were, only morally righteous outward performance would be required. Yet, the chance to articulate oneself has still to be complemented by an adequate reception on the side of all others. In order to prove this as a necessary presupposition of argumentation and to thus situate it at the very core of the discourse theoretical enterprise I have to show that a performative contradiction is involved if a participant does not display such a virtuous attitude towards others. However, Habermas acknowledges a high standard for such type of argument: "Strictly speaking, arguments cannot be called transcendental unless they deal with discourses, or the corresponding competences, so general that it is impossible to replace them by functional equivalents (...)." (5) With the emphasis on functional irreplaceability we have to single out the demanded attitude as necessary for practical discourse in general. This, however, might involve the danger of a petitio principii. We might define practical discourse in such a way that it already entails the idea of virtue before we extract it (unsurprisingly) from an analysis of the presuppositions. Thus, "Apel tries to meet this objection by extending presuppositional analysis to the preconditions of argumentative speech as such, as opposed to restricting it to moral argumentation." (6) This "more radical version" (7) , however, seems to threaten our goal to anchor the described attitude in the very presuppositions of argumentation because theoretical discourses do not demand role-taking at all.

Though everyone engaged in theoretical argumentations accepts that only reasons can decide the debated question and must be attentive to any possible counterargument, she does not have to step into the other's shoes. If the discussion centres around the question whether the earth circles around the sun or vice versa it is simply irrelevant to anticipate the discussants' (e.g., the pope's) feelings about either position. Individual interests or feelings simply do not matter. In theoretical argumentations the main provisions only include that the argumentative process ought not to be influenced by anything except reasons mattering for the subject, all relevant information gathered and adequately processed. The usage of 'adequately', however, implies that the participants do not only have to understand with what kind of problem they are dealing, but also the content of the arguments made. Otherwise the entire discussion would miss its point. If someone presents a complicated equation in order to prove that the earth circles around the sun, the others do not only have to allow its utterance but have to try to follow the argument. Thus, a speaker would commit a performative contradiction if she would state that 'although we were neither able nor willing to grasp the real importance of A's claim that non-p would be the right solution we agreed legitimately on p.'

The refusal to adequately listen to A's claim results in the same actual exclusion of her argument as the suppression of its utterance. (8) Insofar as an agent is honestly interested in the victory of the better argument she does not only have to let everyone articulate her arguments, but try to understand their full meaning. This argument still covers both theoretical and practical discourses, but does so far not include any attempt at role-taking.

The critical move can be done by reflecting on what it means to adequately understand normative arguments. Because normative questions concern the legitimate reconciliation of conflicting interests (otherwise there would be no problematization and consequently no practical discourse at all), it is necessary to understand the interests which give rise to the claims made on their behalf. However, to fully understand the speaker's interests and needs, even when articulated explicitly, the listeners must do more than grasp the semantic content of her claim. They are only able to properly describe the normative problem faced if they discern the meaning the claim has for the particular speaker. This does not imply that the interpretation of individual needs could not be changed by challenging their authenticity and finally, their general acceptability. However, in order to argue against a particular (interpretation of an) interest one first has to adequately understand it. Such a discourse theoretical reading of the role-taking demands of the participants, at least to a certain degree, to situate the raised interests in the evaluative patterns that fuel them; perhaps even in the particular life-histories with all their personal strivings and the meaning-providing (sub)culture in which these claims are embedded. (9)

This process of proper understanding can be partly institutionalized in giving each participant the possibility to explain her background. However, the last step of ascribing adequate meaning to it can only be done by the listeners. Such an attempt at adequate discernment is especially important because "(s)ubstantial reasons can never 'compel' in the sense of logical inference or conclusive evidence", and allow for no "'natural' end". (10) Acceptability expresses an agreement on the side of participants who are moved by certain reasons, but could also not be moved. The more participants are enabled to understand the importance of the other's claims and to relativize their own ones in the face of the latter, the more adequate the reached solution will be. Adequate discernment is thus achieved only by a conscious attempt and can neither be enforced nor checked by others. Whereas it is relatively easy to detect violations of the rule not to hinder someone from entering into discourse, the requirement to step into the other's shoes is a 'duty of virtue' in Kant's sense: it has wide latitude. One never knows whether one's sensitivity was sufficient. One can only try as hard as possible. However, so far I have used formulations such as 'attempt' and 'try as hard as possible' in order to construct a performative contradiction.

(II) Indeed, I want to argue that it is the virtuous attitude which is primarily significant from a practical point of view; though this surely implies a high esteem for those who already have managed to achieve a character that displays the discursive capacities. As Kant put it in his Religion within the Limits of Reason alone, hereby trying to do justice to Aristotle's insistence on the importance of habituation, man has to be conceived of as an autonomous being that is always able to change from bad to good. Otherwise, morality would prove to be an impossibility. However, conceived of as empirical beings from an observer's perspective changes in character can only occur gradually and over time. Unfortunately, here I cannot discuss the problems of this distinction. I simply want to hold that the Kantian intuition is favoured by the presuppositional analysis undertaken in this paper. An agent who lacks the necessary moral capacities does not necessarily commit a performative contradiction because the way she is cannot be confronted with the pragmatic presupposition that adequate understanding is necessary. We can only confront the way she thinks about her state with the latter. If she would embrace the attitude that her incapability would be quite all right, that nothing remained to be changed, she would indeed commit a performative contradiction. Confronted with the meaning of intersubjective dialogue a person with such an attitude would implicitly deny that the others' interests count, that they have to be taken seriously. Yet, an agent who acknowledges her deficient ability to understand others and therefore strives as hard as possible to attain the previously lacking capacities does not fall short of pragmatic consistency. She only has to admit that the results of her encounters with others, though she tries to understand and bridge the difference as well as possible, are probably not as adequate as they would be if she was more sensitive to their demands.

The reconstruction thus shows that we have to acknowledge the following obligation: 'You have to try as hard as you can to understand the other's claim and to strive to improve the capacities necessary for it.' Because both parts of this obligation refer to the attitude of the participant, the presuppositions demanding certain discursive norms must be complemented by the requirement of a genuinely virtuous attitude: of wanting to step into the moral point of view as well as possible, to do justice to all the arguments brought forward and to reach a solution acceptable to all.

I will call the capacities that such a virtuous agent needs to foster 'discursive capacities'. However, so far I have not answered the question as to which capacities fall under this category. Which discursive capacities do we have to try to improve?

(III) The question what kind of discursive capacities are needed surely depends on the way we conceptualize role-taking more specifically. Therefore, we might conjecture that different kinds of discourses will demand different kinds of capacities, e.g., that we would further have to distinguish between discourses of justification and the more context-sensitive discourses of application. The latter structurally demand a concern for the situational details involved, and probably require more sensitivity than the rather abstract discourses of justification.

However, role-taking in discourse ethics can certainly not be conceptualized as empathic identification with the other. Such a picture of simply jumping into the other's mind has to presume that the other is not different at all. But such an assumption of a shared ethical background becomes especially unsuitable in a multicultural society where background values will considerably deviate. Because feelings are not mechanical phenomena but express deep-seated evaluations such as approval or disapproval, disgust, etc., different evaluations of an act will lead to different feelings about it. Any model of role-taking that stresses immediate identification therefore neglects the burdens of reason that shape our emotional responses into different patterns. A stranger who enters an otherwise ethically integrated group might even be depicted as insensitive, because she does not share the specific values of her social environment. To be sure, such identification is not normatively problematic as long as the participants do indeed share the same ethical convictions and it remains possible to challenge the consensus. However, such a challenge might come unexpected (and even undesired) once discourse is, as Habermas charges Kohlberg's account, overwhelmed by "empathy, that is, the intuitive understanding (...)". (11) The strong emotivistic tinge would threaten the role of discourse as a cognitive undertaking for the impartial evaluation of moral concerns, that is, of changing attitudes and interests in a reasonable way. If empathic understanding predominates, unanimity is secured without rendering explicit for what reasons. Consensus can only be achieved through a prestabilized harmony of moral feelings and does structurally not allow for deviance. This concern seems to me at work when Habermas claims that with the growing abstraction, necessary to scrutinize the universalizability of norms, mutual role-taking does no longer depend on emotional capacities such as empathy, but that it can properly reconstructed in purely cognitive terms. Yet, this cognition still entails "understanding for the claims of others that result in each case from particular interest positions (...)." (12) Empathic identification can structurally not become aware of the difference encountered in the other, and thus remains bound to a particular ethical background. It remains on the conventional level. We thus have to search for capacities that allow us to 'empathize' with otherness in a way compatible with a postconventional morality. Reflexivity seems to be paramount, emphasized especially in the hermeneutical tradition. According to Gadamer we can only understand someone else's utterance at all if we already possess a certain pre-understanding, but have to be aware that we already possess this 'prejudice'. The latter is only challenged by a tension or contradiction between our expectations and the way the other (re)acted. The more we are aware of our already working pre-understanding, the less likely it will be that we project our own evaluations inadequately into the other's situation. The desirable outcome of such a self-reflective encounter with different (e.g., cultural) points of view is what Gadamer calls the 'fusion of horizons'. This picture of understanding does not share the problem of an unreflective identification because it does not count with a shared background. The awarenes of one's own continual situatedness has, however, to be accompanied by an imaginative attempt to discover the different value background underlying the other's views. The never-finished fusion leads to an "enlarged mentality" (13) with which we do not only proceed self-reflectively and imaginatively, but attain more and more viewpoints that, in turn, may help us to encounter difference more adequately. Having included self-reflexivity and imagination (including learning about the other's background and views), we might ask if emotions do not figure at all in the proposed picture of discursive capacities.

Yet, if emotions express sedimented evaluations, these evaluations are already 'one step behind' when we encounter difference that requires us to reevaluate our previous judgements. Emotional perceptiveness does indeed depend on the soundnes of the underlying evaluations, often acquired in childhood. Sensitivity as the ability to adjust to the different emotions and views of others, I want to hold in contrast, entails from the start reflexivity. Emotions that sit well with such an approach have to be dynamic themselves. They will most likely be sedimented approvals of self-reflectiveness, imagination, openness, tolerance, etc. That is, those 'post-conventional emotions' (to use a somewhat paradoxical expression) are only post-conventional, because they support procedures that in their dynamic remain open for new voices. Wellmer expresses the same idea when he writes that the liberal virtues "are but nothing [else] than the expression of a habitualization of liberal and democratic behavioural patterns (Verhaltensweisen) and are in a way only to be determined 'procedurally' themselves." (14) Once we have come to deeply appreciate the necessity to take everyone's interests equally into account, acts that accord with this idea will be accompanied by positive feelings. In contrast, violations or neglect of justice will arise shame, contempt or anger. Thus, a virtuous person should try to cultivate the emotions that correspond to the idea of the moral point of view. Additionally, she has to master the difficult task of remaining open for new and contesting voices, although that implies a willingness to call into question those strong emotions that accompany one's present answers to substantive questions of morality and the good life.

(IV) So far I have argued for a (still formal) conception of virtue that transcends mere compliance with prefixed norms or rules. Rather, individual virtue in an intersubjective reading of Kant's 'good will' is essential for finding the right norms in the first place. Only after this discursive virtue, which already demands the perfection of some features of human nature, has issued substantial results in discourse (or its 'makeshift', the monologically performed role-taking) we can get a full picture of how the virtuous agent would have to act. And finally we may single out 'executive virtues' that help the agent to actually carry out her duties. However, leaving all substantial questions to the deliberations of the affected, discourse ethics remains silent on which substantive duties we actually have. Yet, the 'epistemic' virtue described in this paper does not depend on the results of those discourses because it renders them morally adequate in the first place.

However, according to a thesis first put forward by Albrecht Wellmer and since then also embraced by Habermas himself, the obligations of rationality within discourse are no obligations of actions, and thus no moral norms proper. This charge would equally apply to the conception of virtue proposed in this paper. "Thus, I claim that the obligation to suppress no arguments — an obligation that is founded in the validity-orientation of speech as such — has no direct consequences for the question when and with whom and about what issue I have to discuss." (15) This is surely correct. The obligation to engage in a discourse with anybody who demands it, seems extremely counter-intuitive. Just imagine a person on the street who addresses you with the following words: 'Dear Sir, the colour of your jacket does not fit that of your pants. I propose that you take the jacket off.' Neither would we feel obliged to follow this proposal nor would we be inclined to engage in a discourse with this person.

The virtue described in this paper, however, does not prescribe how to act, but how to deliberate. It is a way of approaching the other, of trying to take seriously the other's worries and concerns. In the situation just described the demand of the other falls too obviously in the class of 'unjustified complaints'. However, in the average case the capacities that enable us to step into the other's shoes in discourse will also enable us to detect more adequately whether we face a case in which a problematizing discourse is necessary or not. Thus, the conception of virtue described in this paper is not only valid for participants in discourse, but also for agents who act in a lifeworld where the transfer to discourse is permanently latent. It is the pervasiveness of virtue that blurs the clear cut distinction between deliberating and acting, between discourse and those situations in the lifeworld, where the normative rightness of an interaction has become contested. In fact, a virtuous person will probably not only enter more easily into a problematizing discourse, but also more readily relativize her own interests in the light of the sensitively perceived needs of the other. In the same fashion Wellmer might also argue against the proposed wide duty of perfecting the discursive capacities. Yet, by calling it a wide duty I have already acknowledged that no presuppositional analysis can determine how much we should strive for them in particular circumstances. However, even a wide duty remains, after all, a duty.

Notes

* For helpful comments on an earlier draft of this paper I am grateful to Thomas McCarthy and Michael Neblo.

(1) J. Habermas, Between Facts and Norms (Cambridge/Mass.: MIT, 1996), 277.

(2) I have to leave those criticisms aside that charge discourse ethics with fundamentally misinterpreting Kant's moral philosophy as a monological enterprise.

(3) Supra n.3, 92.

(4) Supra n.3, 92.

(5) Supra n.3, 83

(6) Supra n.3, 84f.

(7) Supra n.3, 85.

(8) This claim applies to the single individual. It surely makes a difference whether I do not (adequately) listen to A or whether I exclude her from raising her voice. In the first case at least the chance remains that others do listen, whereas in the second case nobody will be able to do so.

(9) See also T. McCarthy, "Practical Discourse: On the Relation of Morality to Politics", in: C. Calhoun (ed.), Habermas and the Public Sphere, (Cambridge/Mass.: MIT, 1992), 51-72; here 58-62 and A. Honneth, "The other of justice", in S. White (ed.): The Cambridge Companion to Habermas (Cambridge: Cambridge, 1995), 289-324; here 302-306.

(10) Supra n.1, 226.

(11) J. Habermas, "Justice and Solidarity: On the Discussion Concerning Stage 6", in T. Wren (ed.), The Moral Domain (Cambridge/Mass.: MIT, 1990), 224-251; here 234.

(12) Supra n.13, 234.

(13) S. Benhabib, Situating the Self (New York: Routledge, 1992), 121ff.

(14) A. Wellmer, "Bedingungen einer demokratischen Kultur. Zur Debatte zwischen Liberalen und Kommunitaristen", in M. Brumlik/H. Brunkhorst (ed.): Gemeinschaft und Gerechtigkeit, (Frankfurt/Main: Fischer, 1993), 173-196; here 187, my transl.

(15) A. Wellmer, Ethik und Dialog (Frankfurt/Main: Suhrkamp, 1986), 106, my transl.