Common Knowledge 9.1 (2003) 119-131
[Access article in PDF]
Are There Rationally Undecidable Arguments?
Translated by Ruth Morris and Barry Allen
Great importance attaches to the question of whether conflicts that are beyond rational resolution can occur in negotiations designed to achieve agreement. It was on the basis of this question that Jean-François Lyotard hoped to assess the fate of the modern age. For if conflicts are more and more becoming fundamentally irresolvable—as he believed—then reason no longer counts as the supreme mediating authority: we are entering the condition postmoderne. Jürgen Habermas has challenged this argument, countering that the lack of an expansive conception of reason does not necessarily lead to the dead end defined by Lyotard. One can, Habermas says, accept a classical (say, Kantian) definition of rationality—in which assertions (and their grounds) must be universally true or false—but also feel no need for a final authority (God, tradition, transcendental subject, or absolute spirit) on which to rely: the universal truth or falsehood of propositions can in any case be guaranteed by the peaceableness of the participants in any discourse. All that they need do is recognize procedural principles such as nonaggression, fair allocation of opportunities to speak, and so forth. Observing principles whose violation would contradict the participants' declared [End Page 119] aim of a cooperative quest for truth is, according to Habermas, an uncomplicated condition, free of any suspicion of metaphysics.
In an essay ("Die Grenzen der Verständigung") that I wrote in 1988, I explained why Lyotard's plea for postmodernity—as a situation in which differences are irresolvable—is unpersuasive. I also showed how easy it would be for Habermas to dismiss that kind of objection. It has generally been supposed that, in reaching my conclusion, I wanted to subscribe to all the implications of Habermas's theory of consensus formation; but this is not the case. Rather, I believe that, in negotiations designed to achieve agreement, there can indeed be rationally undecidable conflicts, though it is my contention that Lyotard's "differend" and similar constructions provide a workable approach for resolving them.
Lyotard's Le differend (1983) advances the thesis that conflicts might be insoluble even under ideal conditions of discourse. 1 Lyotard's is a very strong claim, because it questions the mediating capacity of reason even when understood, in Habermasian postmetaphysical terms, as discursive intersubjectivity.
Lyotard's thesis is simple. In explicit accord with Wittgenstein, he denies the existence of a "superordination" of language, claiming that there exists a diversity of language games that are not reducible to each other and that cannot be derived from a single linguistic supraconcept. 2 Within this diversity of rules for language use, Lyotard distinguishes, further, between rules for the use of sentences and rules for speech-acts. The latter are rules for the use of sentences relative to "transphrastic" and "pragmatic" actions, such as narrating, giving a lecture, leading someone astray, and so on. Lyotard views both speech-acts and the use of sentences as governed by rules that prescribe linguistic practices only generally. The transition between speech-events of one type and another remains untheorized—there is, after all, no universal suprarule to regulate usages that violate the rules. Thus occurs the dearth of validation that Lyotard calls a "tort": a wrong or evil that no mediating authority can remedy. Since the (nonexistent) suprarules of discourse are what the modern age has meant by "reason," conflicts between pragmatic rules cannot be mediated rationally; by nature, discourses tend to be agonistic. Darwin's imagery of the struggle for survival finds its way into the postmodern view of discursive exchange in the simile of language at war with itself. Like all wars, this one is carried out by means of naked power and without an arbiter.
Lyotard's view of the nature of discourse is brutally realistic but also (fortunately) incorrect, because it is self-contradictory on a number of levels. First, [End Page 120] Lyotard's struggle against the fetishizing of langue as an...