The article reviews the philosophical importance of conversation and its attendant virtue of conviviality for the theory of knowledge. It argues that to appreciate the crisis Romanticism encountered in trying to maintain enlightened philosophical conversation in a colonial era can usefully inform discussions of secularism in the post-colonial age. This essay appears in _Romanticism, Secularism, and Cosmopolitanism_, a volume of _Romantic Circles Praxis Series_, prepared exclusively for Romantic Circles (http://www.rc.umd.edu/), University of Maryland.

1. The equation of secularization with demystification no longer seems to work. Baldly stated, the problem is this: the cultural insult in the assumption that one culture can enlighten another overrides the idea that enlightenment is a benefit blind to cultural difference. Classic 20th-century critiques of enlightenment, such as Adorno's, uncovered Enlightenment's own inability to escape the mythologizing it criticized. Its apparently disinterested quest for justice disguised a desire to equalise differences and reduce individual differences to a controllable uniformity. Easier legislation rather than the myth of justice for all was the real programme of Enlightenment thought. But now, any notion at all of being disabused of myth appears misguided. All attempts to find common ground between different peoples have their hidden, interested agendas.

2. Recently, the British Romantic mind-set that followed the Enlightenment has been plausibly thought of as facing a comparable dilemma. This has happened in the wake of scepticism, led by new historicist criticism, of the idea that Romanticism provides any substantial critique of Enlightenment uniformitarianism at all. Sceptically viewed, Romantic ideology took the practical solutions of the Enlightenment, problematised them, and then solved them on a higher plane of imaginary compensations. Its sublimations evaded rather than confronted the Enlightenment challenge to describe the common human nature on which a cross-cultural theory of justice might be based. To say, with the Romantics, that what we have in common is imagination appears precisely to avoid answering the Enlightenment question, since our uncommon fictions become our distinguishing characteristics. But what happens, scholars have wondered, if instead we start with the acknowledgement that British Romanticism's conscious inheritance from the Enlightenment was patriotic unanimity unparalleled in Britain's history? Unanimity, in a sense, is a misnomer, because the patriotism at work "forging the nation," in Linda Colley's words, between 1707 and 1837, was largely practical, not mental.[1] Actual successes in communication across classes and their different interests and cultures came to constitute a patriotic citizenry largely un-tempted by the example of the French Revolution. This consensus, though, was strikingly corroborated by imperialist successes and the imposition of English as a global language that followed. British colonial failure in America simply delegated the task of linguistic imperialism. The nationalist consciousness not only held together hearts and minds at home, but increasingly presumed to create a loyal citizenry across the world. Colonialism began to look like social communication by other means. Its hegemony took the form of an exportable, self-confirming patriotism which could morally justify the appropriation of the national goods of others.

3. But the link between this colonialism and its sources in a practical British Enlightenment became increasingly tenuous. The interesting question then is this: did British Romanticism go along with the transformation of Enlightenment into colonialism? Or did it not, rather, recover these Enlightenment sources in ways that typically resisted the contemporary colonial thinking into which they were dissolving? If it did, then could Romanticism be re-read so as to have already helpfully addressed our question of how to differ from someone in a non-coercive form of communication, instead of communicating so as coercively to efface difference? Does cosmopolitanism have to involve the colonization of one belief-system by another? Did Romanticism really engage with these paradoxes?

4. To start answering these admittedly broad questions, we have to look at Enlightenment theories of communication. While the questions are abstract, they need to be if, as I hope, we are to uncover the philosophical skeletons they rely on for their articulation. But in that case, we should evaluate the Enlightenment ideas of conversation, which the Romantics inherited, against models which have been subjected to stringent critique in our own time. We have to see if Romantic theory, no longer exclusively British, develops any workable symmetries between Enlightened and Modern explanations of communication in a manner revealing its need to handle an Enlightenment heritage collapsing into the colonial antagonisms it can hardly have intended. We do appear currently to be tangling with the post-colonial outcomes of this putative debate, and experiencing comparable threats of collapse or reversion to the original colonial problematic. Secularist positions are cast as sectarian; multi-culturalism is hard to defend from charges of ignoring the claims to singularity that seem necessary to the identity of its main players; language-games of different cultures multiply at the expense of their translatability into what is pejoratively described as a master-discourse. To find an historical parallel is not to find a solution, but it might help enlighten us about ourselves in a way which lets us decide if—ourselves suffering its power to secularize our own cherished myths—we really do want to incriminate enlightenment in revenge.

5. This, then, is a discussion of the "theories" of Enlightenment and Romanticism, rather than the achievement of individual authors representative of the literature of the period. I hope, though, that the ideas of citizenship emerging while remaining necessarily abstract are nevertheless more recognizably secular and non-secular than in my own earlier Habermasian proposals for Romantic stand-ins for multiculturalism which perhaps remained unclear about this.[2] On the other hand, I do believe that reactions to Habermas tend as a rule to cast him as more of an abstract rationalist than he is and to neglect the affective basis of much of his thought about communication. His renewal of the Kantian heritage has something in common with the post-Kantian tactic of delegating to non-philosophical discourses an authority no longer sustainable within the monological meta-language traditionally attributed to philosophy. Inevitably this tends to involve affect in the theoretical project. Further, it updates and finds uses for pre-reflective forms of orientation which tend to call themselves "intuitive"—including, perhaps, forms of religious intuition described for example by Schleiermacher. Our apprehension of the lifeworld, we might prefer to say, must involve non-conceptual forms of situating and positioning which philosophy may invoke, but only to defer to them. Such deferral is, no doubt, calculated, but it remains a limit to disenchantment, a boundary of demythologizing. It is a secular deferral to the non-secular, where philosophical logic represents secular understanding and aesthetic, religious, or political discourse its non-secular stand-ins. For those who cannot make sense of any institutional religious affiliation or doctrinal orthodoxy, it is, honestly, about as far as they can go.

   The "exercise of self-converse"

6. Within a well-known Whig tradition deriving principally from Shaftesbury but extending to Adam Smith and beyond, sociability was a major player in sociological explanations. What is more, a ruling comparison furthered those accounts of what bound society together and made political agreement possible. If enthusiasm propagated people's incorrect relation to God, conversation crystallised people's correct relations with each other. Shaftesbury's attack on enthusiasm works through secularization: a down-to-earth redeployment that removes the privileged privacy from enthusiasm and makes it observe the logic of any other conversation.[3] In his Letter Concerning Enthusiasm, which became the first section of Characteristics (1711), it becomes apparent that enthusiasm secularized is still valuable. But, even more than a straightforward dismissal would have done, this rebirth requires the death of enthusiasm as a form of religious experience. Furthermore, Shaftesbury's taming of enthusiasm through its demystification is completed by his domestication of enthusiasm as a driver of the dynamics of his own philosophical dialogue. We are to see the defeat and secular recycling of enthusiasm at work in the conversational, persuasive force of Characteristics itself.

7. Lawrence Klein's recent Cambridge edition of Characteristics freely admits to underplaying the typographical resources with which Shaftesbury stressed that his ideas were produced by "different speaking parties" (xxxvi). Klein has explored as much as anyone the political implications of Shaftesbury's belief that the conversational attitude pursues us to our inmost self-communings. He is less interested in parallels with contemporary philosophical sources of the self and the logical priority of discourse to consciousness, although he does compare Shaftesbury with Gadamer and Habermas. But for Shaftesbury it seems that we cannot think except through the setting up of a debate between two parties. No knowledge is immediate, all is the product of intercession. The Romantic aporias of self-consciousness are still to come. In all likelihood such aporias would have appeared to him as either a questionable unwillingness to take his point or as an unteachable sublimity. Clearly there are political constraints on the view of Shaftesbury as a theorist of communication. From Habermas to Klein, the emergence of a public sphere of philosophical debate associated with the Whig ambitions of Shaftesbury and Addison is described as regulated by a normative politeness. Addison too wished to secularize, completing Socrates' achievement of bringing "Philosophy down from Heaven to inhabit among men" by further inviting it out of the "Closets and Libraries, Schools and Colleges, to dwell in Clubs and Assemblies, at Tea-Tables, and Coffee-Houses" (Addison I.44). The belletristic form to which The Spectator belongs is integral to implementing this future for philosophy. To be sure, studies of the Whig moment of philosophical dissemination have tended to divide between two approaches. Some investigate the specificities of its desired culture of conversation, noting its actual exclusiveness and the counter-public spheres which were strategically ignored as part of its cultural project. Others emphasize the genie let out of the bottle: the emancipatory dynamic in conversation that exceeds any polite self-regulation. I want to suggest that this happens as a matter of logic; and that one way of understanding Romanticism and its uses for us now is to see it as the consequence of this making of conversation: men speaking to men, but with their masculinist, Wordsworthian idiom de-gendered by becoming the necessary, quasi-transcendental framework for knowledge.

8. Romanticism, though, is mediated by Enlightenment. For once, rough periodization helps make clear the philosophical developments and issues in play. In the Letter Concerning Enthusiasm, it is philosophical conversation, inspirational or abrasive, that absorbs enthusiastic energies, and, in redeploying them argumentatively leaves enthusiasm "in some measure justified" (28). Shaftesbury begins to impose his secular turn on enthusiasm by considering the function of a Muse. The exaltation the writer gained from imagining such a divine figure can be gauged, suggests Shaftesbury, when we consider how important for the quality of communication—whether witty, dramatic, or philosophical—is the intended addressee. The quality of recognition of our purpose which we can reasonably expect affects the quality of what we have to say. The flattery intended by Shaftesbury's addressing of the present Letter to the Whig grandee Somers will therefore only work if his philosophy works. The quality of matter and interlocutor are mutually implicated. That Shaftesbury does not want the conceit involved in this equation to appear pompous or self-serving is evident from the virtues he ascribes to the conversation that a good conversationalist facilitates. "Gravity," we are told, "is the very soul of imposture" (8), and the conversationalist should submit his ideas to an opposite raillery, satire and critical inspection. To do this cheerfully requires "good humour" not the "melancholy" belonging to the religious enthusiast (8). A "divine" temper in Shaftesbury's modern, secular sense extends this tolerance to a public who can "partake with us" in our conception of its best interests (20). Contrary to French gloire (an ever-present anxiety for Shaftesbury), this good-humoured exchange observes a fundamental truth about ethical judgement: "we can have no tolerable notion of goodness, without being tolerably good." Shaftesbury's optimism here (what of the villain who knowingly delights in his evil deeds?) is again based on a broad logic of conversation. That is, Shaftesbury assumes a good faith in the critical challenge of one's interlocutor, which, if genuine, means s/he wants to advance the process in which we are engaged. His further assumption is that such progress, logically speaking, is the most critically effective intervention she could make. Knowledge emerges from the recognition of an intention, and from a judgement on that intention's efficacy. Judgements of the goodness or badness of an action similarly only make sense as they lay claim to an integrity of their own, recognizably on the side of goodness.

9. This "conversational implicature," as we will hear it later called by Paul Grice, runs deep in Shaftesbury's thinking, connecting it up with more modern philosophical idioms. Shaftesbury links "soliloquy" with "advice," for instance. In "Soliloquy, or Advice to an Author," also collected in Characteristics, he in effect wonders how to revive the Renaissance role of the consigliere or adviser in a modern non-courtly society.[4] It turns out that "the best way and manner of advising" is a practice best learned through self-experiment (71). By now writing in a very non-empiricist, un-Lockean epistemological vein, Shaftesbury dispenses with the notion of a private language of sensations. "In reality, how specious a study, how solemn an amusement is raised from what we call "philosophical speculations," "the formation of ideas, their compositions, comparisons, agreement and disagreement" (134). Instead, Shaftesbury assumes that the language in which we conduct our introspection is always public. Already, in other words, we grasp ourselves in debate, split into two parties, one half trying suggestions on the critical responses of the other, engaged in our typical "exercise of self-converse" (75). Far from being eccentric or pathological, this "doctrine of two persons in one individual self" only articulates the logic of the Platonic and Stoic tradition of the examined life which Shaftesbury admires (83). To "recognize yourself" is "as much as to say, 'Divide yourself!' or 'Be two!'" (77). To conjure your "daimon" or consult your "genius" is comparably to enter into that dialogue which constitutes self-consciousness for Shaftesbury. Again, in contrast to enthusiasm, the model correspondent here seems to be not an inner voice nor immediate spiritual assurance, but a public body settling its differences. The result is to "make us agree with ourselves and be of a piece within" (77). A dialogue has taken place, has been resolved, and the skills and appetite for further such dialogues have been stimulated. Sociability has been inculcated at the site notionally furthest from its centre, the self-communing individual.

10. Conversation of Shaftesbury's kind creates meaning through disagreement as much as through agreement. The fact that dialogue is taking place authenticates the identities of both participants; in fact they have no other model for self-consciousness. The conversational contract, though, insists that a disagreement is only recognizable as such where it solicits an agreement to come, even if finally we agree to disagree. But this logical propriety exceeds the restrictions of propriety in its more ordinary, polite sense of what is fitting. More interestingly, it harnesses all improprieties in their usual unmannerly sense to this same logic of dialogue. No performance can be so outrageous that it avoids affirming this exchange, even while opting out of it. Something notionally escaping reciprocity and exchange, such as a gift, is, writes Shaftesbury, hard to imagine in the context of conversation. He is thinking of our natural resentment of the advisor whom we suspect of using the conversational occasion for "raising himself a character from our defects" (70). But Shaftesbury's overall argument suggests that free advice will always serve the advisor well too, because it shows conversation working. She or he has a fundamental interest in its practicability, in fact his or her own identity and degree of self-awareness depends upon it: the conversation of soliloquy is "our sovereign remedy and gymnastic method" (84). Conversation for Shaftesbury is the technique and care of the self. To see advice as the conversational outcome is another way of seeing that we are implicated in conversation as a fundamental good, not as a gift but as ethically given.

11. It is obvious that, depending on your point of view, religious presumption might appear heretical, and thus socially divisive. But can't conversation still be improper in comparable ways? Moving on from my basis in Shaftesbury now, we can find arguments for replying: yes, conversation can be improper, but it is in conversation properly so called, conversation that works, that people take just stock of each other. An improper conversation offends against the logic of conversation; it doesn't offend merely by entertaining unsuitable, unmannerly, impolite topics. It is self-defeating. My question is, firstly, whether or not this proposition about conversation as such is sustainable? Secondly, if the answer is yes, can conversational success model, in a more than trivial sense, a kind of communication we need now, one which achieves an equality of exchange without compromising cultural difference?

    The logic of conversation

12. What do I mean by the logic of conversation? In conversation, men and women reciprocally communicate their acknowledgement of each other's access to the subject under debate. A one-way conversation is a description of a conversation that does not take place. Like other non-conversational speech acts, it fails to grant the right of reply, or never takes off because that right is never exerted. Conversational exchange takes place in addition to the information one conversationalist imparts to another. But a persuasion of another's knowledge of the subject is what generates an equal estimate of that person with oneself, at least in respect of the particular topic under discussion. One may know more than another, but an assumption of equality lies in the idea of the conversation in which such disparities of knowledge may be overcome. Otherwise what happens is not a conversation but a lecture ("Meaning" 45).

13. Partly this describes a theory of meaning, a Gricean one perhaps, in which, as he describes it in a famous paper of 1957, natural and non-natural senses of propositions co-exist. Some things can only be meant if they are true ("natural"), others do not require this condition. Those spots only meant measles if the child actually had them. On the other hand, I may say one thing but can choose to mean a number of things by it. In both cases, however, if someone attempts to mean something, they are "inducing a belief by means of the recognition of this intention." Can't conversations be meaningless? Yes, of course, but presumably only through a failure happening within this framework, in which the belief is not induced or the intention to produce it not recognised. Otherwise, whatever is meaningless in the words exchanged does not amount to a conversation. If you didn't understand me, but recognised my failed intention to get my meaning across, this would be a conversational failure rather than a failure of some other kind. It would be a comparable conversational failure if you understood me but could not recognise my attempt to have you as my responsive audience. The first case is a failure to induce belief, the second fails rhetorically. Both cases are covered by being unsure "about how what [I] said is to be taken. ("Meaning" 48). Either you do not know what I am talking about, or else you cannot recognise my intention—literal, ironic, histrionic, expressive, performative in any number of ways. Both cases tie meaning to the possibility of conversation, the possibility of your continuing from where I left off.

14. Grice refines on the possible variations here. His 1957 article was "only intended as a model," he remarks somewhat disingenuously ("Utterer's" 59). Interestingly, his theory is not disqualified by the Derridean objection that we can never know the natural meaning because it is always framed by a non-natural plethora of possible intentions.[5] For Grice, the admission that "intensionality seems to be imbedded in the very foundations of language" doesn't preclude "extensionality." Performances themselves, we might gloss him as saying here, can be literal as well as metaphorical. The literal ones defer to the authority of facts they are trying to get us to believe in. But they are still synthetic performances, not illocutionary acts analytically dependent (as Searle argues against Grice and Strawson) on the meaning of the words they use.[6] We are not, Grice sees, anchored in a shared world by the gravitational force of linguistic presuppositions. The perlocutionary success of performance hinges, though, on the possibility of "recognition," a word undoubtedly redolent of continental philosophical contexts not usually containable by this Anglo/American approach. Setting aside these tempting ancestral voices of Hegelian struggle and glory, we can less spectacularly still assert that this way of putting a theory of meaning hangs on the possibility of conversation. And we saw Shaftesbury earlier argue that my function as conversationalist is affected by the quality of recognition my efforts receive.

15. Grice went on, in his influential William James lectures at Harvard on "The Logic of Conversation," to talk of "conversational implicature"—a non-conventional principle of cooperation necessary for communication to work. Again this specifies an intended recognition by the audience of the utterer's intention, a process that locates meaning within the pragmatics of conversational performance.[7] Grice would, I believe, have agreed with Shaftesbury that we are here directed by a good faith integrated with the surrounding project of a well-directed life. Shaftesbury's tendency to make optimism and progressiveness a matter of logic is conspicuously cast in the Enlightenment idiom of his time. It also translates into that "moral background" that Aristotelian colleagues of Grice, like Philippa Foot, assumed to be logically required for the virtues to make sense. Grice, though, again treats such connections as synthetic rather than analytic; or at least he allows for the less "instrumental" Aristotelianism which (existentially) modulates into a more generous "description of the appropriate forms of presentation of self to others, and of the appropriate ways of responding to and recognizing others."[8] Now, "recognition," in Christopher Cordner's remark here, is again more richly freighted with significance than can be contained within one tradition, period or ethical discourse. It does, though, appear a less anachronistic option than some to think of Grice in connection with Shaftesbury's dialectic.

16. There is something of the idiom of Shaftesbury's version of Enlightenment in the fact that Grice saw himself as a serial philosophical collaborator, someone for whom "the unity of conviviality in philosophy" was not altogether trivial. "Philosophy," as he said, "like virtue is entire."[9] (Dante's Convivio, after all, took on the task of setting his philosophy in a total intellectual context, one primarily challenged by the reconciliation of secular Aristotelian with Christian doctrine). While Grice's colleagues and opponents concentrated on the extent to which the pragmatics of conversation might escape or fall under a formal logical theory, I am concerned with this ethical dimension and its political implications, something shared by the Whig tradition I began by sketching. Indeed, Grice's background was one of strikingly non-conformist "dissenting rationalism" on his culturally influential father's side (Grandy and Warner 46). In this tradition, it is thought that communication partly depends on the mutual building of a common culture of recognition rather than on taking it for granted, and that when this activity is in the service of delivering natural meanings, it can owe nothing to existing social and political maps of people's relations with each other. Conversation, in other words, requires conventional behaviour, but in order to institute its own convention or meeting place; and this effort takes shape by neglecting what is colloquially known as conventional behaviour, all the multifarious speech acts which need not at all evidence exuberant escapes from the tyranny of the "natural" but may more likely embody repressions: institutional esotericisms requiring us to speak the language of some court or other, snobberies, social exclusion orders and so on which as much as playful liberties can so complicate any claim to get to grips with natural meaning. To insist that conventional conventions always win here, that they muddy the epistemological issue, make interpretations undecideable and conversational exchange (meaning) impossible, is to be very conservative. It is to invoke that emptiness of cultural sufficiency, which Hegel saw as the other side of the coin of emancipation from the State.[10] To be conversible, on the other hand, is, as the etymology almost suggests, to relate to others, but through conversational conventions whose meaningful possibilities, rather than their subversions of meaning, make conversation potentially critical of the orthodox social prescriptions it has converted in its quest for information.

17. This, at any rate, is the dynamic of conversation which takes such attractive eighteenth-century shapes as the republic of letters, the public sphere, new forms of sociability in scientific clubs, in debating societies and in other opinion forming forces for change—everything, in short, that tends towards the highly radical historical outcomes of Dissenting rationalism in such bodies as the Society for the Exchange of Constitutional Information and the London Corresponding Society. Let us investigate the formal rather than historical characteristics of conversational logic a bit further, though. The acceptance of someone's right to speak with authority on a particular topic displaces conventional assumptions of superiority. For conversation is potentially ubiquitous, as ubiquitous as nature, beliefs about which heuristic conversation (if we can call it that—conversation that wants to find something out) typically wishes us to recognise as its intention. Conversation can be had on whichever subject comes to hand. Described in this way, it sounds like the radical force of Enlightenment—democratic provided one has access to the skills required to participate. And from being a courtly attribute or bourgeois accomplishment, conversation becomes a name for the rational claim one person has on another. Institutionalised, the obligations of 18th-century conversation extend from the learned club to the university, bypassing the exclusiveness of legal formality or any of the proprieties restricting the swapping of information, but bypassing them in guises that never explicitly threaten them. Nevertheless, the rational universal in which conversation participates is so little a respecter of persons that the implicit threat to the conventions surrounding its own convention remains.

18. Paradoxically, in describing this openness to relevant evidence and argumentation, commentators on the phenomenon of conversation have uncovered an affective template. Habermas, in particular, found a crucial transition from opinion to politics via the intimate sphere of the (emergent bourgeois) family commemorated in epistolary novels of sentiment. This doesn't work, though, does it? What more striking scene of repression and misinformation can be imagined than the post-Freudian family? But the point remains that, in order to imagine communication free of the conventional pressures that restrict conversation at most times, theorists feel forced to transpose the scene entirely. In the process they turn society into the family and knowledge into love. Only in such a forum can that attitude to conversation be figured which does not require qualifications over and above that of being a member of the group by virtue of being able to participate in the conversation. This point, the relevant one, remains valid whatever that family structure was really like. Gender discrimination and all the other containments were, up to Burke's use of the family in his conservative polemic, constituent parts of this membership, not constraints upon it. That was its use for him: you could not sensibly want to be the inheriting son if you were actually the youngest daughter; hierarchy was for once unarguably natural. Yet this group could also model the heuristic conversation, uninhibited by the need to qualify for membership, unconstrained by conventions other than those facilitating recognition of the intention to communicate.

19. The other anti-hierarchical component built into conversation is the heuristic concession it makes to dialogue. To find out the truth conversationally one person must learn from another, or both must learn from each other. As soon as conversation acknowledges its own dramatic character, in which different people contribute to a story larger than any of them, the truth sought in this way has itself to be re-characterised as well. No longer a single object to be mastered from one point of view, truth might vary depending on the perspective from which it is viewed. Truth proper, then, would have to be an aggregate, the product of a composite inquiry. It might also be only fully graspable as the sum of its historical stages; part of its essential character might belong to its own genesis, its actual resistance to instantaneous access. In that case, post-Enlightenment theories of irony and phenomenology beckon. Maybe the final settlement of what a truth is would be political, something achieved when its different interlocutors compromised and negotiated a final agreement, a settlement depending on how far each was prepared to concede his or her own points and accept others? The point about heuristic conversation, though, holds true: the political winner would always be the one who got others to concede that from positions of equal access things looked the same to them. This politics seeks not ascendancy but results within a common franchise.

20. This franchise can be almost entirely one of technique and competence, or it can be one that is inherently negotiable, as in the case of subscribers to aesthetic or political agreement. The fact that the former kind of franchise are often difficult to qualify for—in mathematical theory for instance—does not invalidate the point being made here. Would a mathematical theorem be false if we couldn't persuade anyone that it was true? Clearly not, but one can also define the truth of such a theorem as entailing that only mathematical incompetence would foil its universal acceptance. Conversations about mathematics, in the decisive sense we have been tracing, will therefore be fewer, but not substantially different from other conversations.

    Romantic dialogism

21. The truth emerging from Enlightenment conversation qualified each effective participant as an equal inquirer in the enterprise. This levelling worked through the power of conversation to establish its own conventions irrespective of any already existing ones that might socially position its members. Also, the dramatic method driving this socially unrestricted play to establish the truth privileged neither side in the debate. The negotiated outcome resulting from the dialectic was the dominant interest of both parties. Shaftesbury's theory of conversation stressed the logical presupposition of its viability, and an accompanying ethical stance of some benevolence.

22. Romantic dialogue perhaps has a different emphasis, stressing the power to accommodate difference within a conversational economy. It looks for discursive forms and genres with the elasticity required for this kind of tolerance. Taken to its logical conclusion in Romantic irony, and then in Nietzsche's perspectivism, agreements about truth did not overcome but preserved their perspectival origins. In the case of Romantic irony such as Friedrich Schlegel's, this was in order to show that agreement was possible, and that a republic could be modelled on difference. In the case of Nietzsche, the palimpsest of difference was rather intended to show what a fraud consensus was, and how it masked an inherently competitive will to power. The conversational discovery continued in Romantic dialogue, though, was that the agreement to differ signified not failure to establish the truth but the accurate recognition of its essentially composite character. Incommensurability need not entail incompatibility. An obvious need then arose for a form of writing not disabled by the need to hold competing views in focus simultaneously. Many versions of this opera aperta arose: from the fragments of Schlegel to the encyclopaedic Allgemeine Brouillon of Novalis, and the collaborative forms of the Frühromantiker generally: Mischgedicht (medley), Symphilosophie,Sympoesie and Gespräch (dialogue). All were outcomes of the continuous symposium of brilliant men and women informally and intermittently convened in Jena at the end of the 18th-century. Contradiction would bind rather than unravel, and enhance rather than inhibit, such discursive initiatives. Key British contributions to this post-Kantian dramatic alternative to introspection could be Coleridge's Biographia Literaria, with its abandonment of an introspective transcendental deduction for the discursive histrionics and critical warfare of the second volume. But the entire second generation of British Romantics are now usually read as critiquing the inwardness of their immediate predecessors, and as resisting the spread of inwardness that Hegel characterised as fatefully Romantic with that synthetic play of genres that the Jena group had championed.

23. A possible consequence of this, though, links rather than estranges Nietzsche from the conversational tradition. Truth is just that movement outdistancing individual perspectives upon it. Truth, then, does not simply enrich the individual's version of it: it embarrasses, confounds and disturbs. The dramatic containment or management of its energies is never conclusive, but always a stop-gap round which more veridical energy can pour. The dramatic character that a necessarily dialogic approach to truth requires is a needful euphemism. Conversation, dialogue, the aesthetic itself become the new masks allowing us to contemplate the truth without becoming hopelessly confounded by it. The crucial transition from Aeschylean tragedy to Platonic dialogue, via Euripides, is described by Nietzsche both as decline from a supreme moment of human expression, and as the rescuing of that moment for modern discourses supposed to be non-dramatic. In Benjamin's re-articulation of Nietzsche's idea in The Origin of German Tragic Drama, the Trauerspiel replaces the tragic agon, in which the hero is symbolically clenched in a silence proleptic of a language still to come. The Trauerspiel is, by contrast, aligned with an art-form which is garrulous, allegorical, and indiscriminate in its choice of allegorical objects, thus offering the maximum number of points of entry for the reader. Benjamin's Trauerspiel, unlike tragedy, is hospitable to difference. It re-works Schlegel's Romantic idea of the mixed form, the philosophical conversation and the hybrid poem: "the dialogue," writes Benjamin, "contains pure dramatic language, unfragmented by its dialectic of tragic and comic" (113-118).

24. Benjamin encourages us to think of what happens to tragedy in the Nietzschean story as a parable of what happens to philosophy in ours, or the philosophy for which he failed to secure academic recognition in his own lifetime. The application of his tale to subsequent philosophical developments is easy to see, though. As the death of tragedy is the birth of dialogue, so the death of philosophy is the birth of communicative action, one version of 20th-century philosophy's famous "linguistic turn." Or, in Habermasian terms, philosophy abandons its privileged status as a meta-language pronouncing on the conditions necessary for legitimate intellectual inquiry, and becomes the convener of a dialogue between other disciplines and discourses. This post-metaphysical philosophy, in line with Wittgenstein's Philosophical Investigations, becomes a therapy for its own ambition to ground everything else, an ambition which it dissolves in a healthy proficiency in the logic of different language games. Philosophy thus secularizes its theological ambitions and becomes discursively cosmopolitan in the process. More accurately, philosophy suffers a displacement of its authority comparable to the submission of a religion to secular authority. Or, perhaps even more basically, there is no other language for describing this modulation of philosophical expertise than to describe it as a version of such a secularizing process, but this time imposed on the secularizer. Standing on this common ground, perhaps there is something both secular and non-secular sides can say to each other?

25. What might these non-secular philosophical surrogates be? Whatever they are, their devotees would not be pleased to hear them described as the stand-ins for an eventual philosophical competence, as Habermas sometimes does. To satisfy the non-secularists, the handing-over of philosophical authority would have to be a bit more whole-hearted or sincere than that. There is, however, nothing unusual about such interplay. It is as easy as, say, friendship, and as mysterious. Affinities without strict analogies, as studied in the recent literature on friendship by Nancy, Blanchot, and then Derrida, testify precisely to this ability to find a common cause with someone without the resemblance being, as it were, motivated or emulative. Friendship reaches over and crosses boundaries, and to say someone is my friend, as Montaigne saw, is sufficient explanation of this breach. Equally, following Aristotle, friendship can be virtuous: it can be based on attraction to my friend's exemplary exposition of a principle of flourishing, or the aspiration to human excellence common to us both. So I am not pushing a kind of irrationalism here, because just as a society of friendship would, for Aristotle, make the institution of justice unnecessary, so the kind of transfer envisaged for philosophy above is validated by an analogous kind of displacement of one authority by another—by the power of the discourse endowed with new philosophical responsibilities to make the expected disciplinary procedures inessential.[11] Who needs Fichte or Novalis's Fichte Studien when reading Novalis's novels? Who needs Bergson when reading Proust, or William James when reading Henry James? Yet, undeniably, these different writings are "friendly" towards each other. Not how one text comments on the other but the fact of their conversational relationship is what is critically informative and important: the fact that it continues something to talk about one after talking about the other or when talking about the other.

26. Yet apart from talking about the surprisingness or not of their community, celebrating the illustrative likenesses and un-likenesses of its members, its differences in relationship, there is little else we can say about this connection. Precisely the way neither side could have prescribed their relationship is what makes it friendship. The relationship itself is what matters; the fact that different speakers in a different discourse can take over the original story and carry it on in their own terms; the fact that we make a "natural" transition from one to the other in the course of interpretation, exposition or whatever. And this discursive friendship seems to replicate what we want to happen in societal cooperation. But let us retrace this journey to informality one last time, eventually addressing the specific relationship between logic and religion.

27. Arguably, then, dialogue is the forgotten destination of philosophy that is recovered in the post-Nietzschean story. The modern reading makes the delivery of Platonic theory its real end, rather than its world of ideal forms, its defence of justice, its scepticism of art or its political utopianism. These Platonic theories are not ignored or marginalised as a consequence. But they are regarded as plausible only as they are susceptible of convincing delivery, only as they are topics proper for conversation and dialogue. In Grice's strong sense, they must enhance the way in which philosophy is "convivial," and the drive of the line from Shaftesbury to Grice is to make the inherent feasibility of conversation a matter of logic. The affect here sticks as closely to the reasoning as any Kantian feeling of "respect" clings to the rationality of "law." Can this definitive turn in philosophical thought impinge on secularism and its discontents—or secularism's inability to sustain any longer a position outside the myths to which it supposedly was an alternative. Habermas, characteristically perhaps, sees a kind of philosophical modelling in play here, one standing in for sociological solutions. He writes that "the secular awareness that one is living in a post-secular society takes the shape of post-metaphysical thought at the philosophical level" (4). In other words, philosophy's delegation of its authority to other discourses, to dialogue, is like the translatability of secular values into non-secular discourses. One is aware that a translation is taking place, so one remains secular, but secularism has handed over its original authority as a kind of prima philosophia in such matters, and pragmatically accepts the consequences. For this to work, though, it must be corroborated by a willingness on the religious side to acknowledge the separate coincidence of secular values with the ones supported for religious reasons. The key word is "reasons," whose exchangeability Habermas understands as an assumption inalienable from democratic society as he understands it. This is perhaps the strength and the limitation of his argument. He makes a logical necessity out of the viability of communication; but, staying in this (neo-)neo-Kantian posture, he seems unwilling Romantically to credit communicative successes in other discourses for keeping his philosophy of a logically necessary communication viable. Yet he concedes that this is what philosophy in its interpretative function in fact shows other discourses capable of doing!

28. Habermas argues that the liberal state cannot "expect of all citizens that they also justify their political statements independently of their religious convictions or world views" (8). He concludes that they should "therefore be allowed to express and justify their convictions in a religious language if they cannot find secular 'translations' for them" (10). We might wonder how this could work, and search for those "reasons" common to religious and secular outlooks, by looking back into Habermas's German heritage rather than at the exigencies of the post 9/11 scene apparent around him when he wrote, and even further back than the reactions to Kant with which he usually begins.

29. Let us consider from this perspective two texts strategically anthologised in one volume of the putatively canonical Cambridge Texts in the History of Political Thought.[12] When Luther argues that Christians have a duty to obey the law, he manages to do so from both within and outside his faith. Christians themselves do not need laws, he thinks, since they behave morally anyway. Those who are not Christian cannot be similarly relied upon to behave properly, and need laws backed by the power of the "sword" to make them do so. However, in addition to this separation of the sheep from the goats, Luther maintains that it is the duty of his Christian to help others, and that this service can only be rendered under the rule of law. For that to be possible, evil and justice must exist out there, publicly visible and not peculiar to a private religious insight. Luther's Christian uncomplainingly suffers injustice against himself, but does not tolerate injustice visited on another. So it is because some people are not Christian that there has to be a secular authority; and only by maintaining a secular authority can Christians be empowered to behave in a properly Christian manner towards non-Christians. To be a Christian demands faith, not works. Nevertheless it is the duty of a Christian freely to carry out good works. The distinction between being a Christian or a heathen is not reducible to a private / public distinction in part because the privately assured Christian grace can express itself in support for the public realm. Secular authority certainly exceeds its remit, for Luther, when it pronounces on matters of faith. But the distinction between faith and works only apparently does the work of the private / public distinction. Significantly, Luther argues that not the exertion of power but a "different sort of skill" is needed to negotiate the relation between the subject and secular authority (25, 30). Ultimately, Luther claims that it is "unfettered reason" through which the just person makes the right judgement ("in accordance with love") and can "find written in his heart that it is right" (43). It is a matter of logic ("reason") supported by affect ("love") that common values "out there" exist. But for Luther, reason is the Word, and, again, his religion articulates both Christian and secular spheres and requires the interactive separation of both.

30. By the time Calvin writes his chapter on civil government in Book IV of his Institutio Christianae Religionis, this Christian versatility in making secular justice a matter of Christian principle is compromised by the development of the public / private distinction. "It would be utterly pointless," we are told, "for private men, who have no right to decide how any commonwealth whatever is to be ordered, to debate what would be the best state of the commonwealth in the place where they live" (56). For privacy to be such a disqualification, it must have ceased to stand solely for the realm of Christian assurance which Calvin, like Luther, believed underwrote the political order. Privacy was now developing an interiority of its own, eventually capable of stimulating that great Romantic satire on Calvinist inwardness, James Hogg's The Private Memoirs and Confessions of a Justified Sinner (1824). When the private man presumes upon the public interest then he does so for reasons different from the common religious support that makes the public realm justifiable on its own account—the Lutheran logic. In Luther we encountered a perspectivism enjoined by virtue of being religious. Religion gives the lead to secularism in matters of "unfettered reason." Calvin, a trained lawyer, carefully argues that "rules of justice and equity," originally learned in the (public) service of God could subsequently be distinguished from that service (67-8). This separation allows him to explain how different systems of laws and penalties might be equally just and equally demanding of observance. They are individualised without being dictated by private interest. But this casuistry, as well as lying open to the distortion Hogg attacks (for private interest can be a kind of casuistry too), forfeits Luther's clear point about religious support for secularism. In the process Calvin sounds more liberal and pluralistic. But he loses the idea of the godly living among the un-godly for their benefit in the literally "convivial" way that Luther describes.

31. The salient point emerging from a consideration of these texts at this stage of a discussion like this is perhaps the reciprocal goods that both secularism and non-secularism gain from mutual translation of each other. The Lutheran is more of a Christian for his or her care for the maintenance of civil order. And previously we saw that it was possible to argue that a secular philosophy reaching its own limits could only develop further through its self-effacing, collaborative enlightenment of different discourses or interpretation of them to each other. But the model anterior to these pleasing correspondences, if we can entertain them for a minute, is one of living together in convivial conversation. The possibility certainly remains highly abstract, or what I have called a matter of logic. Practically speaking, a coach-ride in the park with Shaftesbury or a banquet (convivio) in Oxford with Grice could no doubt have been enjoyable, but not particularly important theoretically. But both philosophers do think it important to redeploy enthusiasm and conviviality in philosophically affective ways. In this they can be seen to try to match the non-secular, often religious care for the communal which is so much less embarrassed or logically nervous about its emotional commitment. But the new conversational possibilities are mainly facilitated by the mutual tempering of the two sides' pretensions to absolute theological or quasi-theological authority. And, additionally, the fact that the secular is thus itself subject to a process for which "secularization" is, paradoxically, the best description, maybe shows that slightly comic difference from itself needed to begin the task of unclenching some oppositions currently hell-bent on remaining tragic.[13]