Conclusion

Walton (1989) has argued that a shift in the burden of proof can occur when one side in a debate is shown to be more trustworthy than previously thought or the other side is shown to be less trustworthy. Walton calls this "internal evidence" (156).

Although internal evidence derived from an arguer's position may always be weaker than external evidence, sometimes, when it is hard to decide on a controversial topic, weak evidence may be enough to shift the burden of proof (Walton, 1989:156).

Something like a shift in the burden of proof occurred, I believe, in the NAFTA debate, as a result of Perot's thoroughgoing delegitimation. That Gore "won" the debate seems incontrovertible, based on the polls. I have been interested in extending the analysis of the debate by drawing on the concept of rhetorical rationality. I have been maintaining, given the circumstances surrounding the debate, as well as the debate itself, that a reasonable viewer, having looked to the debate for guidance on a difficult matter that was soon to be decided by Congress, and having found Gore to be much the more credible speaker, would have been within reason in thus becoming pro-NAFTA or more pro-NAFTA on the basis of the debate. Admittedly, what Walton calls the "intern al" evidence is "weak" regarding NAFTA, but it is devastating regarding Perot. Moreover, the anti-NAFTA forces stood behind Perot, allowing him to serve as their spokesman in the debate. And Perot, by many estimates, was expected to be a formidable adversary. In a message-dense society, on a highly complex issue, where the pre-debate campaigning by each side was equally credible, debate viewers were acting rationally in inclining themselves against the anti-NAFTA position and towards NAFTA.

From a normative rhetorical perspective, then, not just Ross Perot, but the anti-NAFTA position, was "convicted" on Larry King Live. And this in turn suggests that the public grants potential relevance to character displays and ad hominem attacks in policy debates. Yet the very concept of rhetorical rationality, with its admittedly "weak" normative underpinnings, may seem to many a highly troublesome notion, perhaps an oxymoron. In defense of the concept, I would like to offer four points t o clarify its applications to policy disputes.

First, rhetorical rationality, as I see it, is not an implicit invitation to abandon critical message processing in favor of cognitive shorthands such as following the crowd or trusting to the judgments of one's opinion leaders. Much depends upon the cont ext--on whether under conditions of limited time, limited availability of information, message complexity, and so on, audience members are in a position to render competent judgments on their own. Much depends, also, on whether a given shorthand seems re liable in a particular case. And motivation must play a role as well. The evidence is overwhelming that under conditions of low involvement in an issue, people tend to rely on cognitive shorthands (Petty & Cacioppo, 1986). Perhaps they ought to be more reflective more of the time. But then they might not have as much time for the issues that are truly important to them.

A second and a related point is that endorsement of rhetorical rationality is not a license for indiscriminate reliance on judgments of source credibility in policy matters. Nor, on the other hand, does it entail dismissing source evaluations as irrelevant or unnecessary. Insofar as persuaders offer analyses, interpretations, speculative forecasts, and the like, we need to decide whether we can trust their judgments. The same holds true for the sources they cite. But these decisions need not be indiscriminate. Surely in the NAFTA debate, the two speakers gave evidence of their relative trustworthiness.

Third, one's commitment to rhetorical rationality does not involve countenancing personal attacks on an opponent in a policy debate, unless such attacks are relevant to audience assessments of the policy in question. But many ad hominem attacks are relevant when audience assessments of the policy in question are dependent to some degree on the judgments of the opponent. This takes the discussion back to my previous point, where I argued the potential relevance of source evaluations to policy considerations. Pertinent, too, is Walton's (1989) claim that perceptions of the relative trustworthiness of opponents can shift the burden of proof in a debate.

My position is apparently at odds with that of van Eemeren and Grootendorst (1992) on ad hominem attacks. After observing that the standard "logico-centric" approach permits too many exceptions to its criteria for relevance, van Eemeren and Grootendorst categorically insist that any denigration of the other party's expertise, intelligence, good faith, impartiality, or logical or behavioral consistency violates a pragma-dialactical rule against questioning another's right to speak (van Eemere n & Grootendorst, 1992:153). But denigrations of this sort, used by both sides in the NAFTA debate, did not call into question the opponents' right to speak; only their relative trustworthiness when they spoke. This was an issue of some relevance since each party made implicit character claims.

Finally, the idea of rhetorical rationality in no way entails a fixed commitment to a policy decision, or to judgments about a source, or to the decision criteria one has used in evaluating policy proposals or their advocates. Even cognitive shorthands, thought useful at one moment, may be reevaluated. Critical judgment may indeed be improved (or at least altered) over time as individuals and entire communities learn from experience, including rhetorical experience. Thus, one might conclude (though I would not) in light of subsequent revelations about Mexico's financial problems that Perot was right after all about NAFTA, and that even if Perot's behavior in the NAFTA debate had been eccentric or worse, his arguments should nevertheless have been heeded ( Kempton, 1995).

Having defended and hopefully clarified the concept of rhetorical rationality, I should like now to acknowledge that "rhetorical rationality" remains a troublesome concept, one reflective of the historically troublesome character of the field of rhetorical studies generally. Indeed, the term is troublesome in much the same way that "practical reason," "practical wisdom" (phronesis), "prudential reason" (prudentia) and "good reason-giving" are troublesome; all are vague and elusive, value-dependent, and context-sensitive (e.g., Farrell, 1993; Fisher, 1978). At the same time, they reflect an age-old insistence that these difficulties cannot be avoided on matters complex and contingent.

Understandably, rhetoricians have exhibited admiration for the certitudes of logics considered formally or geometrically. But their admiration has been tempered by the sense that the strictures are insufficiently flexible for dealing with matters of policy or value, and that logicians err especially on the side of condemning in discourse what might reasonably be condoned or even celebrated under a given set of circumstances (Farrell, 1993; Jonsen and Toulmin, 1988; Toulmin, 1958; Walton, 1992).

There are no quick solutions to the task of giving clearer meaning to concepts like "rhetorical rationality." If anything, the problems mount with postmodern critiques of foundationalist logics and objectivist presuppositions. Ironically, this has led to the so-called "rhetorical turn" in the human sciences. Or, as Allan Megill has proposed, the "postmodern turn appears as a 'weakening' of philosophy into rhetoric--as a turn from, say, the Aristotle of the Posterior Analytics and related works to t he Aristotle of the Ethics and the Rhetoric, where the philosopher admits an element of localism and uncertainty" (Megill, 1989: 142). Indeed, philosopher Joseph Margolis has argued that "the most comprehensive conceptual schemes we can [currently] imagine are all partial, fragmented, contingent. historically bounded, ultimately blind, and radically ungrounded..." (Margolis, 1990:310). (10) 

One promising approach to the development of norms of rational judgment, consistent with postmodern theorizing, is to think of them as practical accomplishments, rather than as indubitable postulates. This is an approach shared by a number of the theorist s cited in this essay (Farrell, 1993; Fish, 199 ; Margolis, 1995; McCloskey, 1985; Posner, 1990). Farrell (1993) has been especially persuasive in arguing that discourse norms can be abstracted from cultural experience only at great risk to their historic al embodiment in rhetorical practices and critical judgments. Along similar lines, Margolis (1995) has proposed that we think of logics as abstractions from rhetorics that audiences over time and across contexts have found convincing. Even Habermas's discourse postulates (e.g., truth, rightness, truthfulness, propriety, etc.), says Farrell, can best be appreciated as "powerful normative enticements, or goads, to the perfectibility of discourse practice. But this is not because these norms somehow hover outside history and spatio-temporal existence; such an a priori 'view from nowhere' is simply implausible" (Farrell, 1993:211). Maintains Farrell (1993: 228), in reference to Habermas's notion of critical judgment (and of communicative competence more generally) as a disembodied faculty,

If judgment is treated only as a faculty, it is unlikely that audiences will ever be lucky enough to share in the privilege. By contrast, if judgment is treated as an acquired competence, sophisticated through practice, the prospect for its democratization is considerably more promising.

Viewing norms of judgment as practical accomplishments has implications for cultures, discourse communities, and even audiences of one. All need to inquire self-reflexively into the utility of their norms of judgment, including their cognitive short hands, recognizing as Posner (1990) does the limitations of formalisms on matters complex and contingent. It becomes especially important to bring these norms to bear upon situations of considerable uncertainty for audiences (as in the NAFTA case) and to subject these decision criteria to critical scrutiny. This conclusion points us back to rhetorical rationality and to the need for case studies of the kind provided here. Rhetorical rationality can emerge as a clearer concept if we in rhetoric reflexively apply it to cases and genres at the same time that we attempt to learn from the cases what we mean by the term. In this admittedly circular mode of inquiry, we need to be asking what modes of reasoning were most appropriate for message recipients under the circumstances, inviting critical scrutiny by colleagues of our own judgments.

Thus I would tentatively derive from the Gore-Perot NAFTA debate the following normative principle of rhetorical rationality: To the extent that policy advocates seek audience adherence without providing full backing for their judgments, that audiences experience difficulty deciding policy issues, and that there is a disparity between opponents in source credibility, credibility itself becomes increasingly relevant.

With admittedly minimalist principles such as this one, we may ultimately conclude that in rhetoric's "weak" sense of rationality lies its enduring strength.