Conversation is, first of all, an event, something that happens. But the concept of conversation has also been appropriated by various thinkers as a model or metaphor of hermeneutical experience, of communication, political discourse, the acquisition of knowledge, and so forth. As an event it has been analyzed within the hermeneutical tradition, from Schleiermacher to Gadamer, and in this analysis it has been tied to Romantic conceptions such as the universality of language, "linguistic heritage" (Angeborenheit der Sprache), and what von Humboldt called "the common denominator of human nature," the universal and spiritual humanity that we all share.1 So conceived conversation has been made a model of hermeneutical experience. Gadamer, as is well known, characterizes reading as a conversation or dialogue with a text. This is more than metaphor, he tells us. Interpretation means bringing the text "into the living present of conversation" (TM 368), restoring it to an original condition.

As a metaphor, outside of the actual hermeneutical situation, it is quite easy to idealize conversation and to describe it as it ought to be. For Oakeshott, the "conversation of mankind" ideally conceived, is composed of a plurality of voices, each with its own idiom and each in relative emancipation from the others -- a conversation free of hegemony, lacking hierarchy, without the requirement of credentials for participation. Despite this ideal conception Oakeshott acknowledges the real imperfection of conversation -- its political nature.

It is the political reality, as well as the Romantic and metaphysical characteristics attributed to conversation, that motivate various critiques of conversation conducted from critical, radical, and feminist perspectives. In opposition to the champions of conversation, such as Gadamer, Oakeshott, and Rorty, it is not difficult to identify the critics of conversation, notably (and of course in variously different ways) Habermas, Derrida, and feminists such as Harding, Nicholson, and Comay.2 Within the context of this opposition, however, nothing stays clear cut for very long, and what I find most interesting and promising is a distinctive ambiguity introduced into the conversation about conversation by authors such as Lyotard and Caputo.

In this paper I want to explore some of this ambiguity and suggest why it is both interesting and promising. In specific I want to ask whether there is a particular model of conversation that will work within the context of postmodern thought and social practice. I formulate my answer in three parts, and with the help of three conversations. First, I explore some ways in which Gadamer's conception of conversation might be stripped of its Romantic and metaphysical conceptions and made more postmetaphysical, postmodern. For this I focus on the challenge that Derrida poses for Gadamer in their famous conversation in Paris in 1981. Second, I want to show that the conception of the conversation of mankind employed by Rorty and Caputo for postmodern purposes is not a good model for postmetaphysical/ postmodern thought. Here a conversation between Rorty and Lyotard at Johns Hopkins University in 1984 will be instructive. Finally, I want to take the stripped-down, transformed Gadamerian model of conversation and show in some detail how it could operate within the postmodern situation outlined by Lyotard. In effect, I want to develop and defend a Gadamerian-Lyotardian conception of conversation(s) as a model of postmodern hermeneutics. To my knowledge, no conversation (or at least no documented conversation) between Lyotard and Gadamer ever took place, so we can only explore a hypothetical conversation.

Part and parcel of Derrida's rejection of the hermeneutical project, "which postulates a true sense of the text," and which maintains itself within the "horizon of the meaning or truth of being . . . and the hermeneutic space of the question of the truth (of being)," is a rejection of a conception of conversation oriented toward an original experience of truth, being, or presence.3 One needs only to line up Gadamer's remarks about conversation in Truth and Method -- for instance, remarks about the originality and primacy of the living present conversation, about his own kinship with Plato's critique of the written word (TM 368-369), and about how in conversation we all come under the "influence of the truth" (TM 379) -- to see how Derrida can definitively shape his opposition to this concept. This involves the deconstruction of phono/logocentrism, the displacement of speech by arche-writing, his reading of Plato's Phaedrus, and so forth. It is no surprise, then, that in their 1981 encounter the differences between Gadamer and Derrida were much in evidence.

The contrast in their rhetorical styles reflected a fundamental disagreement in interpretive approaches. One difference clearly manifested itself in regard to language: for Gadamer, the essential nature of language is to be found in the event of conversation, and his style of discourse reflected an attempt to bring Derrida into a conversation about hermeneutics. Derrida, on the other hand, understands language to work as an uncontrollable play of signifiers with indeterminate meaning, and his style reflected the disruption from within of the very concept of conversation which Gadamer was theorizing about and practicing. Their minimal dialogue centered on the concept of trust or "good will" which Gadamer proposed as a prerequisite of conversation. Distrusting the very concept of "good will," Derrida attempted to show its metaphysical roots in Kant, and thus to show the metaphysical background for the notion of the primacy of conversation.4 If we could set things right, however, trust and good will are, for Gadamer, not only, as he acknowledges, tied to Platonic dialogue, but also concepts that depend more on the Romantic claims about the universality of language and the sharing of a common humanity, than on the Kantian will. Still, these concepts are no less metaphysical and no less open to Derrida's sort of deconstruction.

If hermeneutics is conceived to be seeking meaning, truth, or consensus through interpretation modeled on conversation, it reflects a Romantic optimism or trust that in some sense truth will be found on the basis of a shared humanity. On the Derridian reading, Gadamer is placed clearly in the camp of the "hermeneutics of trust." If, as Gadamer contends, in conversation "both partners must have the good will to try to understand one another," Derrida wants to know what happens to trust and good will in cases of distorted communication, which demand suspicion, as in the psychoanalytic situation. But even more than this, Derrida embraces a (non)hermeneutics of radical suspicion, which is neither a Habermasian-style critical suspicion eventuating in consensus and emancipatory truth, nor a psychoanalytic cure that would somehow restore trust, but a Nietzschean suspicion that is never resolved. In opposition to hermeneutics in any traditional sense, Derrida proposes an "active interpretation, which substitutes an incessant deciphering for the disclosure of truth as a presentation of the thing itself in its presence."5

Gadamer's defense is, in some respects, a denial, in some respects an apology, and in some respects a confession. He holds fast to hermeneutics, but refuses to understand this as holding fast to the metaphysics of presence (DD 94-95), and he would associate his own work with certain aspects of Derrida's mode of interpretation (DD 96, 118). Still, with so much of Plato and Hegel in his background, Gadamer understands how his philosophical hermeneutics might be construed as belonging to the framework of the metaphysics of presence and the language of metaphysics (DD 98). He maintains, as he has from the very beginning of his philosophical career, the connection between hermeneutical conversation and Platonic dialectic. He also suggests, however, that the roots of hermeneutical conversation go back even further than the Socratic dialogue to the very rationality (the pre-metaphysical rationality) of anamnesis, which "is not only that of the individual soul but always that of 'the spirit that would like to unite us' -- we who are a conversation" (DD 110). This idea, which is an ancient one, but nonetheless Romantic and of the German 19th century, stands also behind Gadamer's contention that "to be in a conversation, means to be beyond oneself." This is the Romantic trust once again, the shared spirituality of humanity which makes "to be beyond oneself" mean "to think with the other" (DD 110). Gadamer admits this -- he confesses to being "in the neighborhood of the early Romantics" (DD 118; see 120-122). He remains "deeply rooted . . . in the Romantic tradition of the humanities and its humanistic heritage" (DD 24). On this conception conversation not only encompasses the "multiplicity of languages and linguistic families" represented in the biblical story of Babel, but holds out the promise of overcoming such differences (DD 106).

Now, whatever progress Gadamer can make in moving away from Romantic metaphysics, away from a presupposed trust, the closer he would come to a postmetaphysical conception of conversation. To be fair to Gadamer, he does learn from his conversations. Derrida's questions about good will remind him of his debate with Habermas. "This objection is one that Habermas had earlier raised against me when he argued that distorted understanding makes what I call mutual agreement mere appearance and even a form of manipulation" (DD 97). Gadamer goes on to acknowledge the importance of Levinas in this regard: "One can learn from Levinas how serious this objection is . . ." (DD 97). Further on he remarks: "Now Derrida would object by saying that understanding always turns into appropriation and so involves a covering up of otherness [by a romantic oneness]. Levinas, too, values this argument highly, so it is definitely an observation that one cannot dismiss" (DD 119). Perhaps Gadamer was thinking of passages like this one in Levinas: "The face, whose ethical epiphany consists in soliciting a response . . . is not satisfied with a 'good intention' and a benevolence wholly Platonic."6

For Levinas good will and benevolence are attitudes assumed under the community-form of commercial civil society and can be elements of history and politics. But this reduces the ethical face-to-face, an asymmetrical relation, to an interchangeable, economic sameness. So in the move from conversational communication to community, a move made quite often in Gadamer's theory, and reminiscent of Platonic harmony and Hegelian civil society, Gadamer would efface the differences between the self and the other. Gadamer takes this as a serious objection, and tries to be more careful. If he were to give up Romantic trust, good will, and benevolence in conversation, however, he would also need to give up the notion of a fusion of horizons, which reduces the other to the same. Gadamer is reticent to give this up, but he does attempt to formulate it so as to avoid the objections of Levinas and Derrida, and in doing so, we might say, he moves, slightly, in the direction of postmodernism. "When I speak in my own work of the necessity for the horizon of one person and the horizon of another to merge into one for any understanding between them to take place, I am not referring to an abiding and identifiable "oneness" [Eines], but just to what takes place in conversation as it goes along" (DD 119, trans. rev).

In other terms Gadamer might say that one is able to be suspicious of X only by trusting in Y, where Y might be simply our ability to ask the right questions. A less Romantic trust, a trust that is something more than naive or blind faith, consists of the informed outcome of previous suspicions. Too much trust, as Derrida points out, would lead us unwittingly into traditions and power structures of domination; too much suspicion, as Gadamer points out, would cause the conversation to self-destruct. Conversation has to occur somewhere between the extremes of absolute trust and absolute suspicion.

How far Gadamer moves toward a postmodern conception of conversation in this respect is open to interpretation. More interesting and promising is the aporia that Gadamer tries to make clear to Derrida, an aporia in which Derrida is caught up. Simply put, this aporia involves the inevitability of conversation.7 In effect, Derrida is already in a conversation with Gadamer: "Derrida directs questions to me and therefore he must assume that I am willing to understand them" (DD 55). He continues: "Surely this is not at all a kind of metaphysics, but the presupposition that any partner in a dialogue must assume, including Derrida, if he wants to pose questions to me" (DD 56). Now whether this is "a kind of metaphysics" or not, it is clear that what Gadamer points to is the inevitability of conversation, even for the critique of conversation. Even if Derrida provides some cause to suspect and move away from the concept of conversation, he can never escape the hermeneutics of conversation, he cannot help but work out his never-ending suspicions within a never-ending conversation.

Thus, deconstruction is not something that falls outside of this hermeneutical model. Even agonistic discourse follows hermeneutical principles, insofar as it produces discursive meaning and requires a tradition to react against. Derrida's deconstructive activity, no matter how radical or how postmetaphysical, or how anti-hermeneutical, is always in a conversational/hermeneutical situation. As Gadamer puts it, "the unavoidability of the hermeneutic standpoint is the theme of my confrontation with Derrida" (DD 125).

The inevitability or unavoidability of conversation involves in some fashion the universality of hermeneutical experience. Here we have a difficult problem. Is it possible to think of this universality in a postmodern context? Gadamer offers us the following text, which for some interpreters8 seems too ambiguous and even impossible. I think it is a promising one.

Universality pertains to conversation as an event. Conversation is something that happens to us, as Gadamer would say, beyond our wanting or willing.

One can immediately think of objections to introducing the notion of universality into postmodern contexts. The emphasis in postmodern hermeneutics is on the local, the particular. Postmodernism challenges universality wherever it finds it -- from metanarratives to categorical imperatives, from performativity principles to Enlightenment politics. The universality of hermeneutics is only one more metanarrative; another impossible traditional prejudice, Romantic and metaphysical. Despite this protest against pretentious universality, a conception of conversation that retains an aspect of universality is sometimes put forward in postmodern discussions -- namely, the notion of the "conversation of mankind." Of course it does seem odd that Richard Rorty, the pragmatic liberal, whether or not we would assign him the title of postmodern, retrieves this conception of conversation from the politically conservative Michael Oakeshott. And isn't it also odd that in the wake of Derrida's critique of conversation, John Caputo should suggest that the conversation of mankind is precisely the medium necessary to translate Derrida's radical deconstruction into the possibility of real political and moral liberation?

But first, we should try to avoid some confusion. The conversation of mankind is not the same as Gadamer's hermeneutical conception of conversation. Despite some confusion on this issue we do have some consensus (temporary and tentative though it be). Jeff Mitscherling, on the Gadamerian side, guards against this confusion by pointing out that for Rorty, the conversation is something that we control -- we make the effort to keep the conversation of the West going -- which seems to imply that it could stop at any moment. For Gadamer, on the other hand, conversation is unstoppable -- we do not control it but rather it is the condition of possibility for continuing or discontinuing personal or social practices and discourses.9 Caputo, on the more radical side of things, also recognizes the difference between Gadamer's conversation and Rorty's, favoring, of course, Rorty's. In Radical Hermeneutics he calls Gadamer's conversation inadequate to the postmodern situation at the same time that he champions the conversation of mankind. For Caputo, and he claims, for Derrida, "things get worked out in a way which is very much like what Rorty (following Oakeshott) calls the conversation of mankind (but with no Rortian illusions about the charms of bourgeois liberalism) . . ."10 Caputo explains that this "is not merely a hermeneutic problem in the Gadamerian sense . . . . It is rather a deconstructive problem which requires vigilance about the subversion of discourse by a priori metaphysical schemes, by exclusionary practices, by a rhetoric systematically bent on sustaining the prevailing order" (RH 261).

It is Caputo's contention that the conversation of mankind operates as a better model for postmodern hermeneutics than does Gadamer's conservative notion of conversation. Certainly the descriptions of the conversation of mankind which one finds in Caputo and Rorty, and even Oakeshott, paint the picture of a progressive, liberal, and open-ended scheme. Ideally, for Oakeshott, the conversation is composed of a plurality of voices each with its own idiom, and each being in relative emancipation from the others. This ideal conversation is prescribed, by all three theorists, in the face of exclusionary social practices. Oakeshott recognizes, however, the political reality: "An established monopoly will not only make it difficult for another voice to be heard, but it will also make it seem proper that it should not be heard: it is convicted in advance as irrelevant."11 The real event of conversation, we might say, is hegemonic, and this is a constant. Political practice, in its essence and regardless of which particular ideology it follows, operates as exclusionary. Granted this real situation, however, on the model of the conversation of mankind one argues that there is always room in the conversation for the excluded voices, or at least there is always room for a discourse which speaks about the exclusion, an agonistic discourse which champions the excluded. This is what attracts Caputo to the concept of conversation. With the conversation of mankind his "ethics of dissemination" takes its stand "with those for whom the system was not designed -- women, children, the mad, the ill, the poor, blacks, the religious and moral minorities -- those who are being excluded by the system" (RH 264). Likewise, Rorty protests against "attempts to close off conversation by proposals for universal commensuration through the hypostatization of some privileged set of descriptions."12

All of this looks pretty good. At the very least, it looks very liberal, very progressive, with the greatest intentions of the greatest inclusion. But is it genuinely radical, postmetaphysical, postmodern? I want to suggest that in a certain formal but still pragmatic way, the conversation of mankind is not radical enough, and indeed may be simply another apriori metaphysical scheme which, despite the rhetoric of inclusion, operates in an exclusionary way. Let me put this first in Lyotard's terms.

The conversation of mankind, or to be precise, the theory of the conversation of mankind (I don't think the conversation exists outside of theory anyway) is a metaphysical scheme in the sense that it operates in the same way as a metanarrative. For a discourse to be a metanarrative it must meet (or claim to meet) two conditions. First, it must 'cut across' or encompass all local narratives (conversations, interpretations, language games, and so on). If the discourse defines principles they must apply to all cases in its domain; it must have a universal scope. Second, it must be capable of adjudicating all differences; that is, it must be able to prescribe resolutions to conflicts (misunderstandings, misinterpretations). Lyotard suggests that Plato's theory of justice is a metanarrative because it claims both universality and adjudicative power.13 I would argue that the conversation of mankind is a metanarrative in the same way. It signifies the largest and most all-embracing of conversations, and it prescribes that attainable solutions must be sought within the ongoing universal conversation. It prescribes against nonparticipation. The universality and adjudicative power of the conversation of mankind is clearly discernible in Rorty. First, he understands conversation "as the ultimate context within which knowledge is to be understood" (PMN 389) -- this is its universal aspect, it is all-inclusive. Second, he regards the continuation of the conversation as a moral imperative (PMN 389) since, as Caputo puts it, this is where "things get worked out," that is, this is where things get adjudicated, and the hope for adjudication is nonexistent outside of the conversation.

Of course, even within the postmodern condition we may have to coexist with metanarratives and metaphysical schemas -- after all, we need something to be suspicious of and to find incredulous -- and besides, such things are inescapable. But, is it right to claim that the conversation of mankind is exclusionary. After all, for Oakeshott, the ideal is to have a conversation that is not hegemonic or hierarchical or exclusionary; and Rorty and Caputo clearly set up the model to include the incommensurability of abnormal discourse. Despite this rhetoric of inclusion, this inclusion in a curious way is predicated on the basis of a prior exclusion. In different terms, incommensurability is possible only on the basis of a more fundamental commensurability. There is a nice phrase in English that summarizes this: we can "agree to disagree." In the conversation of mankind, we have to agree to converse, we have to agree to be included; nonparticipation is not tolerated. Silence is always treated as a moment within the conversation rather than as a refusal to participate in it. Any protest to the conversation itself must be made within the conversation, which means objections to the conversation which do not take the shape of the conversation itself are not included. Abnormal discourse occurs only when it is voiced within the conversation. Abnormal discourse is possible only on the basis of normal discourse; abnormal, paralogical, agonistic discourse, Rorty maintains, "is always parasitic upon normal discourse" (PMN 365). One can discover a new paradigm only by first knowing and conforming to the existing paradigm. "To attempt abnormal discourse de novo, without being able to recognize our own abnormality, is madness in the most literal and terrible sense" (PMN 366).

This requirement can be seen very clearly in Rorty's support for E. D. Hirsch's educational ideal of cultural literacy. For Rorty and Hirsch, education -- "even the education of the revolutionary or the prophet -- needs to begin with acculturation and conformity . . ." (PMN 365). Rorty argues that the program of cultural literacy supports a pluralistic democracy precisely because of the neutrality of the cultural knowledge required for democratic communication.14 Neutrality, Hirsch claims, means value-neutrality, ideologically-free knowledge which therefore signifies an inclusiveness and accessibility for anyone as long as they adopt the basic non-ideological premises which form the vocabulary or idiom of the conversation. But this solution to cultural and political exclusion is not as clear-cut as either Hirsch or Rorty would like. Either, (a) language is value neutral and excluded groups need only become culturally literate to become included, while still retaining their own values, worldviews, and so on, or (b) language is not neutral but conserves established values and traditions, so that excluded groups, in becoming literate, must give up their own "un-common" values, worldviews, and so on, and adopt the established common ones. In the first case, (a), Hirsch and Rorty would have us believe that the original exclusion (which cultural literacy tries to correct) was based only upon the lack of literacy and that prejudices or differences in values and worldviews were not relevant to exclusion. If this solution would still maintain pluralism it would be based on a naively simple conception of political exclusion and communicative competence, not to mention a modern and simplistic Enlightenment conception of language as something value neutral -- certainly not postmodern, and not even hermeneutical. In the second case, (b), the excluded group would become included only by becoming the same as everyone else. As Hirsch indicates, "the specific contents of the different national vocabularies are far less important than the fact of their being shared."15 In this case, however, there is no pluralism, but only a nonpluralistic cultural leveling in which everyone would share and reproduce the same unchanging or slow-changing cultural meanings -- hardly postmodern, or even hermeneutical.

In terms of the conversation of mankind this means that incommensurability is defined only within the conversation, in the vocabulary of the metadiscourse, and if a group refuses to take up, or is incapable of taking up that vocabulary they are not agreeing to disagree. One can find numerous examples of such groups. The Sinn Fein (the political wing of the IRA) in Northern Ireland had at one time refused to take their seats in the British parliament when elected. They attempted to remain within their own conversation and did not accept the vocabulary of the other official one; an overarching conversation was refused. A group of aborigine coal miners in New Zealand refused to take up the terms of collective bargaining because in their view the vocabulary and legal processes of the industrialists -- which in the West would represent civility (agreeing not to disagree) and participation in conversation -- was immoral. This group, to join in the conversation, would have to adopt a vocabulary and procedure to which they object and find abhorrent. So also, native American tribes who did not have within their vocabulary the conceptions of property, legal ownership, purchase or sale, were not genuine participants in the negotiations which resulted in the white man's ownership of the forests and rivers.

David Ingram suggests other examples:

Nancy Fraser, within the context of discussing the dynamics of conversations between the sexes, not only cites empirical evidence to demonstrate the inequality of participants, but also indicates, in the case of legal considerations of marital rape, an instance, not of refusal to enter into recognized discourse, but of the inability of the established terms of discourse to recognize a claim made by the victim. (Nancy Fraser, "What's Critical about Critical Theory?" in Critical Theory: The Essential Readings , ed. David Ingram and Julia Simon-Ingram (New York: Paragon House, 1992), p. 369). Carole Pateman indicates that "women find their speech ... persistently and systematically invalidated in the crucial matter of consent, a matter that is fundamental to democracy. [But] if women's words about consent are consistently reinterpreted, how can they participate in the debate ..." (Carole Pateman, "The Personal and the Political: Can Citizenship be Democratic?" Lecture 3 of The Jefferson Memorial Lectures, delivered at University of California, Berkeley (February 1985), unpublished, cited by Fraser, p. 369). Fraser, following Pateman, suggests that "there is conceptual dissonance between femininity and the dialogical capacity" that seems central to the notion of a universal conversation (369).

What we have in these instances are what Lyotard calls differends, and it is precisely differends that are excluded from the conversation of mankind which operates on the basis of shared vocabulary and "civility" (Oakeshott, Rorty, and Caputo all use this word). The conversation of mankind reduces deprivations to negations. As Lyotard puts it, "to be able not to speak [= a negation] is not the same as not to be able to speak [= a deprivation].16

The conversation of mankind fails as a model of postmodern hermeneutics not only because it is a metadiscourse and worthy of our incredulity, but because it hides exclusionary rules beneath a rhetoric of inclusion. The overarching conversation of mankind aspires to resolve all differends. But by requiring what is genuinely incommensurable (i.e., incommensurable with the conversation itself) to be voiced within the conversation, it denies it expression and helps to constitute it as a differend at the same time that it disguises it as a litigation. The very attempt to include something which cannot be included makes the conversation of mankind a terrorist conversation.

This is one of the issues between Lyotard and Rorty in their own conversation in 1984. For Lyotard, the conversation of mankind forms part of the modern Enlightenment tradition.

In this scheme the particular I or you is subsumed into a universal we. Rorty responds that as long as the subsumption is accomplished by persuasion rather than force, there is no terrorism involved.18 The conversation of mankind would avoid terrorism by giving all participants a "respectful hearing." For Rorty, incommensurability is simply a "temporary inconvenience" that can be overcome by honing our translation abilities (CWE 216). He sees nothing wrong with trying to replace differends with litigations (CWE 217), a conversational process that would sometimes work and sometimes not. Lyotard's response to this is that a differend exists between Rorty and himself, a differend that Rorty would reduce to a mere litigation. Lyotard claims "My genre of discourse is tragic. His is conversational."19 He also points out that the persuasion which Rorty advocates is not free of violence, is not the same as convincing someone. "Persuasion is a rhetorical operation, and the Greeks knew that this operation utilizes ruse, mental violence" (Discussion, 582). It is a pragmatic activity which seeks to gain the trust of the interlocutor and which operates in an imperialist manner. Lyotard calls this "conversational imperialism."20

Does this critique of the conversation of mankind, taken together with the postmetaphysical critique of Gadamer's model of hermeneutical conversation, leave us without a conception of conversation that could work in postmodern theory? I want to make the following case: despite and even granting Derrida's deconstruction of Gadamer's concept of conversation, and despite and even granting the claim made for the hermeneutical universality of conversation, Gadamer's hermeneutical model, with some modifications, can serve postmodern hermeneutics and is not incommensurable with the incommensurability entailed in Lyotardian paralogy and the concept of the differend. The postmodern situation is nothing other than a paralogical multiplicity of conversations.

The plural here is important. It serves to qualify the type of universality that we can claim for conversation. The kind of universality claimed by the conversation of mankind is quite different. The conversation of mankind would be a universal conversation based on a presupposed (meta)consensus -- a contract, expressible in prescriptive terms, an agreement to disagree. In contrast, the universality claimed for the hermeneutical model of conversation involves neither a metaconsensus nor a method of adjudication. This model does not entail a metanarrative. Although it lays claim to universality, it does not claim adjudicative power. It is not prescribed as a solution to problems; it is not that we ought to converse. The claim is rather that we cannot avoid conversing. It is not a matter of agreeing to participate in a particular conversation, but rather a matter of finding ourselves already cast (sometimes as unwilling participants) in one or many conversations which are organized (or disorganized) in paralogical fashion. This means that wherever we find ourselves we are always in a hermeneutical situation, in a conversation, and more precisely, in one conversation among others. This universality has nothing to do with a universal conversation. As Gadamer indicates, the universality of hermeneutics is in no way inconsistent with the fact that a particular conversation contains its own limits within itself, but "fits perfectly well with the factual limitedness of all human experience and with the limits governing our linguistic communication and possibility for expression" (DD 95).

The postmodern idea is not that there is one overarching conversation, but that there is a plurality of conversations, some constituting relative differends in relation to others. It is still possible that fusions can happen between conversations, not in the sense of unifying or reducing different conversations, but in the sense of creating new and different conversations by linking one to another; or again, not in the sense of a fusion of horizons, but in the sense of a creation of new horizons.

Paralogy is not prescribed, it already is the case, over and above our prescribing, as Gadamer might say. This model, like Lyotard's own work in The Differend, goes beyond his considerations, in The Postmodern Condition, about language games and his prescription for paralogy. In the latter work Lyotard makes paralogy an issue of justice, a program to be opposed to performativity. In that way Lyotard himself had been involved in a metanarrative, not only telling "the great narrative of the end of great narratives" (D # 182), but prescribing paralogy as a means of adjudication and legitimation. On the postmetaphysical, Gadamerian model, however, the case is already paralogy, that is, already a plurality of conversations. We find ourselves, over and above our wanting and willing, in a postmodern paralogical situation (the situation of Babel always already in effect) -- what Lyotard calls the "postmodern Babel" ("The Sign of History," LR 410). Lyotard clearly goes beyond Gadamer's romantic conception of bridging, by conversation, the multiplicity of languages represented by the tower of Babel (DD 106)].

In The Differend -- which, you will note, conducts itself in the genre of a dialogue, even if it is Lyotard conversing with himself -- Lyotard provides, among other things, a philosophical hermeneutics of "phrasing" which is commensurable to a philosophical hermeneutics of conversation. According to this postmodern hermeneutics, a conversation is itself a particular genre of discourse. Still, some explanatory power is provided by taking conversation as a paradigmatic model of all genres. A phrase presents a universe, i.e., a situation in which an addressor, an addressee, a sense, and a referent are located. A conversation happens as a set of phrases come together (are linked). Dialogue, Lyotard says, "links an ostension (showing) or a definition (describing) onto a question; at stake in it is the two parties coming to an agreement about the sense of a referent" (D, p. xii). A differend exists where one is preempted from participating in a conversation by virtue of the fact that a would-be participant lacks or is divested of the means (the vocabulary, the referent) to converse. A differend exists when the conversation in question is conducted "in the idiom of one of the parties while the wrong suffered by the other is not signified in that idiom" (D #12). A differend exists between two conversations because they are heterogeneous. Differends notwithstanding, conversation is inevitable because "there is no last phrase." This means that injustice is inevitable because there is no conversation that has a universal authority to adjudicate; the inevitability of conversational linkages causes necessary wrongs. Given "the impossibility of avoiding conflicts," and the absence of a universal conversation that could adjudicate them (i.e., given the inevitable partiality of the participants in any particular conversation, consensus always being a temporary thing), the problem Lyotard sets out to solve involves finding, not a way to get rid of the inevitable (so Lyotard is not an idealist like Habermas), but how to participate in conversations honorably, that is, an "honorable postmodernity" (D p. xiii).

In attempting to link the conversations of Gadamer and Lyotard one finds (a) points of agreement, (b) points of disagreement, (c) places where Lyotard rightly goes beyond the Gadamerian model, and (d) places where Gadamer could operate as a corrective to Lyotard.

(a) Agreements: Both agree to disagree with the traditional idea that man controls language and simply makes use of it for his own ends.21 Lyotard's contention that addressor and addressee are instances presented by a phrase, is quite consistent with Gadamer's notion that the participants are carried along, conducted by the conversation. The participants are not independent of the conversation; "they are situated in the universe the phrase presents" (D, # 18).

Hermeneutical principles that apply universally to conversations also apply to phrasing. Gadamer and Lyotard would agree that

(1) All conversation/phrasing is linguistic;

(2) All conversation/phrasing is historically situated or local;

(3) All conversation/phrasing is productive (of further phrases).

These principles describe how phrasing/conversation happens, and they define the conditions of possibility and the limitations which one must take into account if and when one goes further and tries to say how any particular phrasing/conversation ought to take place.

(b) Disagreements: It is clear that Lyotard would leave behind the metaphysical, Romantic concept of trust defended by Gadamer. In number 32 of The Differend we find a reenactment of the Gadamer-Derrida encounter.

A postmodern hermeneutics would be one that is free from the Romantic conceptions of humanity and trust.

(c) Lyotard's Advance: Lyotard's distrust of metalepsis indicates that in postmodern hermeneutics the fusion of horizons which would efface the differences between the self and the other must be displaced by a conception of linking that includes the impossibility of complete fusion, along with the possibility of an agonistic refusal to be fused, as well as the possibility of progressive dialogue. Metalepsis is the transformation of an observer left outside the conversation into a participant through his judgment about the conversation. When Socrates speaks to Thrasymachus, Plato intends for the reader to enter into the same conversation. We enter into the conversation, Gadamer would contend, through this metalepsis in which we judge whether Socrates or Thrasymachus is right. For Gadamer, every time one reads Plato one enters into a conversation that is fused with the Socratic dialogue. Lyotard, in contrast, equates metalepsis with an absorption of the difference that exists between agonistics (debate) and dialogue, two incommensurable genres. For Lyotard, it is "never certain nor even probable that partners in a debate, even those taken as witness to a dialogue, convert themselves into partners in dialogue" (D, p. 26). Rather, what is certain here is that we end up with more than one conversation, each structured in its own genre, with different participants, and different senses. Despite Gadamer's addiction to metalepsis, this paralogical result is not inconsistent with Gadamer's own principle that we always understand differently. In doing so, however, we do not enter into the original conversation, but create a new one for ourselves. I think Manfred Frank is correct in this regard: "Conversations are always transformations of other and earlier conversations, just as signs are always new articulations of other and earlier signs" (DD 161).

(d) Gadamer's Corrective: It is within this model of postmodern, paralogical conversations that one must try to work out the notion of honorable participation, an ethics of participation. In this situation one has no universals; one has only conversational experience to fall back on. Of course, this relativism is what everyone worries about when they confront postmodernism. Richard Bernstein, for example, worries (just as Caputo used to worry) about the lack of critical justification-criteria in Rorty's conversation of mankind.22 Caputo used to worry about the lack of any anchor within the conversation of mankind and complained that Rorty denied the hermeneutical situation,23 a complaint that could now be turned against his own radical conception of conversation. Rorty himself now uses this same complaint against Lyotard,24 and he follows Habermas in worrying about Lyotard's paralogy and its devaluation of consensus and communication.25

Here, I think, Gadamer can offer something to postmodernism that I find lacking in Lyotard. Nonetheless, there is a weak linkage between them on this point. Their two conversations touch on the concept of Aristotelian phronesis. They understand this concept differently, however. The difference is that for Gadamer, as for Aristotle, phronesis requires the educational backdrop of experience, whereas for Lyotard, phronesis operates in an entirely ad hoc manner and without background knowledge. In this sense, Lyotard denies that the paralogical situation is still a hermeneutical situation (see Gallagher, 1993).

In Just Gaming, Lyotard contends that for phronesis to be postmodern it must rely on nothing prior. He denies that phronesis depends upon education or on hexis, a habit of virtue, an ethos (see JG 26). Although both Lyotard and Aristotle would agree that there is no theory or method that explicitly defines phronesis, and that phronesis is not a trained ability, they would disagree about the proper educational background for phronesis. Phronesis, according to Aristotle, requires an education, a knowledge of particulars which comes from experience (NE 1141b17; 1142a14). Phronesis also requires virtue; the person who wants phronesis must develop the right habits, an ethos developed over time. This formation of the right ethos in experience -- I call this the 'educational backdrop of phronesis' -- is precisely what Lyotard denies. Clearly, without the ethos, without a backdrop of educational experience, what Lyotard calls phronesis is nothing more than what Aristotle would call cleverness. He ignores Aristotle's clear contrast between these two concepts (see NE: 1144a25ff), as well as Aristotle's denial that "quickness of mind" is sufficient for the excellence in deliberation which is phronesis (1142b5-16). Lyotard equates phronesis with speedy imagination, "the capacity to actualize the relevant data for solving a problem 'here and now', and to organize that data into an efficient strategy" (PMC 51), the ability to play the game with inventiveness, to play "master strokes" (JG 61).

In contrast, on the Gadamerian view, every situation, no matter how paralogical, is always a hermeneutical situation, informed (like it or not) by particular traditions. No matter which of the many conversations I find myself participating in, I never start ex nihilo. I always have something, some tradition, some background knowledge to fall back on in making my judgments. Yet this is different from Rorty's claim about the requirement of a common vocabulary or idiom. This educational backdrop does not constitute an already predetermined universal under which I simply locate the particular situation -- phronesis works, not from universals, but from particulars. Phronesis works best in those cases where there are no rules, on the paralogical edge where new conversations are started. A sufficiently developed, hermeneutically-informed conception of phronesis promises a good postmodern response to the questions of ethics and community.

It is possible that a conversation between Gadamer and Lyotard might start here, with this difference.

Gadamer: That I, as an individual, find myself always within a hermeneutical situation, a conversation, signifies that I am not alone. Even if I am only talking with myself, my language is something that I have inherited from others, and their words interrupt and make possible my conversation. Even if there is no universally shared human nature as a basis for Romantic trust, within the hermeneutical situation there is still some shared aspects, a certain range of background knowledge, some limited common ground which enables the particular conversation to happen. Otherwise communication would be impossible. Neither the common ground, nor the communication it makes possible, will necessarily guarantee community, consensus, or a resolution of differends. We are not focused here on outcomes, a particular consensus to be reached, but on what is anterior to (as a condition of possibility for) conversation. This anterior common ground may only be the battlefield on which our conflicts can be fought. Isn't the principle something like, no differends without a battlefield?

Lyotard: You know yourself how even "the battlefield" is open to conflicting interpretations. This was a favorite example used by Chladenius in his Enlightenment hermeneutics. Differends are not fought out on the battlefield; they remain outside the circumference of the battlefield, unable to enter the conflict within. So we must define many small battlefields, each of which might be called a community of difference, which is not presupposed but accomplished in and through conversation which remains dialogue without ultimate synthesis. Conversations, in such cases, always remain incomplete, imperfect. In them the we is always in question, always at stake, the consensus always local and temporary, community always deferred. Perhaps, within these conversations, a trust which is not good will is required; a trust that we are different and for that very reason require conversation to create a we. This is not a trust in a preexisting we, but a trust in the promise of a we, a not yet we which will always remain not yet, defined by our differences. I have stated elsewhere, "the true we is never we, never stabilized in a name for we, always undone before being constituted, only identified in the non-identity between you ... and me ..."26 A conversation "could do no more than put the we back into question" (D # 210).