Fallacies in Transition: An Assessment of the Pragma- Dialectical Perspective CHRISTOPHER W. TINDALE, Trent University Keywords: Argumentation, audience, fallacies, objective standards, pragma-dialectics; rhetoric. Abstract: The paper critically investigates the pragma-dialectics of van Eemeren and Grootendorst, particularly the treatment of falla- cies. While the pragma-dialectieians claim that dialectics combines the logical and rhetorical ap- proaches to argumentation, it is argued here that the perspective relies heavily on rhetorical fea- tures that have been suppressed in the account and that overlooking these features leads to sig- nificant problems in the pragma-dialectical per- spective. In light of these problems, the author advocates turning attention to a rhetorical account which subsumes the logical and dialectical. Resume: eet article examine d 'un oeil critique la pragma-dialectique de van Eemeren et de Grootendorst, mais tout particulierement leurs traitements des paralogismes. Bien qu'ils affirment que la dialectique reunit les approches logiques et rhetoriques, on maintient plutOt iei qu'elle s'appuie fortement sur des traits rhetoriques implicites et qu'en refusant de voir ceux-ci, la perspective pragma-dialectique connait d'importants problemes. On propose, ala lumiere de ces problemes, une approche rhetorique qui englobe la logique et la dialectique. Of the various perspectives on argument currently of interest, the three dominant ones may be characterized as the (geometrical) logical approach, the (anthropological) rhe- torical approach, and the (critical) dialectical approach. In this paper I will discuss a variety of the dialectical approach, namely the pragma-dialectical theory of Frans van Eemeren, Rob Grootendorst and the Amsterdam school, together with some of Doug- las Walton's North American modifications, and particularly the impact of this theory on work with fallacies. I do this for a number of reasons: the pragma-dialectical per- spective, partly through the successful proselytizing of its authors, is now receiving so much interest that no one studying argumentation theory can afford to continue to ignore it. But more than this, it is important because of the claim from members of the Amsterdam school that, "In the dialectical approach, the product-oriented [logic] and the process-oriented [rhetoric] approaches to argumentation are combined" (van Eemeren and Grootendorst, 1988, p. 281). Through generally reviewing the pragma- dialectical approach and offering more detailed comments on its treatment of fallacies, I intend to show that while there is recognition of rhetorical aspects in the approach, it is not sufficiently focused and credited, and that critical weaknesses in the approach can be traced to this shortcoming. This leaves us to ponder the merits of developing, not so much a dialectical approach which subsumes the logical and rhetorical, but a rhetorical approach which subsumes the other two. Informal Logic Vol. 18, No.1 (Winter 1996): 17-33 ©Informal Logic 1996 18 Christopher W Tindale The pragma-dialectical approach to argumentation is grounded in the speech act theories of Austin and Searle, and in Gricean ideas about conversational implicatures. Argumentation is viewed as a phenomenon of language usage with the focus of study being argumentative speech events. "Argumentation is a form of language usage that is part of an explicit or implicit discussion designed to resolve a difference of opinion between two or more language users" (van Eemeren, 1988, p. 46). This requires the use of dialectics and the associated idea of having a regulated critical discussion. The notion of a critical discussion and the rules of conduct for the same are fundamental to the pragma-dialecticians' project. In fact, fallacies will be understood within the con- text of the theory (and only within such a theory) as violations of such rules, as ma- noeuvres that disrupt and impede the resolution of disputes. The first developed account of the approach is given in Speech Acts in Argumen- tative Discussions (van Eemeren and Grootendorst, 1984), and refinements are made to the theory throughout the middle and late eighties, culminating in Argumentation, Communication, and Fallacies (van Eemeren and Grootendorst, 1992).1 Speech Acts in Argumentative Discussions analyzes argumentation as an ilIocutionary act complex which is performed by advancing a sequence of assertives and which is linked to an expressed opinion. In fact, much of the first part of the book involves investigating the relation between the illocutionary act of arguing and the perlocutionary effect of convincing (van Eemeren and Grootendorst, 1984, p. 47). In the tradition familiar to those steeped in Anglo-American philosophy, what they call the argumentation would be named premises, while what they would identify as a standpoint or expressed opinion would be named the conclusion. Argumentations are oriented toward the resolution of disputes and disputes only exist "if a language user has propounded a view and doubt is subsequently expressed about the acceptability of that view" (ibid., p. 79). The defender ofthe standpoint is the protagonist, while the party casting doubt on it is the antagonist (both may advance argumentation during the course of the discussion). The critical discussion surround- ing the dispute is presented by means of a normative model requiring four stages: the confrontation stage, the opening stage, the argumentation stage, and the concluding stage. Finally, in terms of this brief overview, the discussion is governed by a code of conduct which involves fixed rules. The rules are intended to enable language users to conduct themselves as rational discussants; they facilitate the resolution of disputes and the avoidance of fallacious moves. 2 Speech Acts in Argumentative Discussions detailed seventeen rules for conduct- ing discussions, some of them quiet unwieldy. These have been refined over time, and the 1992 book contains a "simplified version" of ten discussion rules. I will not repeat them all here, but they are distributed over the four stages of the discussion and cover things like burden of proof, the relevance of an arguer's contributions, the appropriate use of argumentation schemes,3 and the use of valid argument forms.4 Misunderstandings and Qualifications Several criticisms have been advanced against the pragma-dialectical perspective. But many of these are founded on misunderstandings or a failure to note qualifications Fallacies in Transition 19 made about the theory. The first point to stress is that the model for critical discussions put forward by van Eemeren and Grootendorst is normative rather than descriptive; in its essential features it is an ideal model. The extent to which a particular dialogue approximates the model is something decided at the time of reconstruction. They are not describing how disputes actually take place, but how they should be performed if the optimal chance for their resolution is to be achieved, or, rather-since they recognize that many disputes are too deep to be completely resolved-if disagreement between the parties is to be minimized (van Eemeren and Grootendorst, 1988, p. 286). Van Eemeren writes: "A normative reconstruction in the dialectical sense does not mean that every form of usage is automatically regarded in toto as a critical discussion, but that we look to see what happens if the analysis is carried out as ifit were a critical discussion" (van Eemeren, 1988, p. 47). A related misunderstanding is the belief that the pragma-dialectical perspective is concerned only with verbal dialogue. Among many North American and British argu- mentation theorists the practice has been to study written texts and the pragma-dialec- tical interest seems removed from this. Such a concern is expressed by Alan Hajek who, in a review of a book by Douglas Walton, makes the following comment about the pragma-dialectical approach: "It seems excessively restrictive to locate all slippery slope arguments in the context of a dialogue. Surely one can use such an argument in private thought. And even when the argument is presented publicly, need there be a respondent?" (Hajek, 1992, p. 23). In fact, the pragma-dialecticians believe that their approach easily accommodates such North American and British practice. While suggesting that the dialogue approach best reflects the behaviour of language users, they also submit that mono logical argu- ments can be treated as involving implicit dialogues (van Eemeren and Grootendorst, 1984, pp. 12-14). In "Some Fallacies about Fallacies," Grootendorst describes the or- dinary language situations for which the pragma-dialectical theory is appropriate as including "discursive text in a newspaper, a conversation in a cafe or a debate in some kind of meeting of a more formal nature" (Grootendorst, 1987, p. 336). In "Argumen- tation Analysis: A Dutch Counter-balance" van Eemeren, when talking of reconstruct- ing speech events, notes that this applies not only to ordinary conversation but also to . "more formal discussions, editorial comments, policy documents, scholarly polemics, and so on" (van Eemeren, 1988, p. 47). Finally, in Argumentation. Communication. and Fallacies. they insist that a discursive text can always be considered part of a discussion "real or imagined by the arguer, in which the arguer reacts to criticism that has been or might be levelled against his point of view" (van Eemeren and Grootendorst, 1992, p. 13 and fn.l). Argumentative discourse, even when presented in a monologue, is deemed essentially dialogical. This is demonstrated later in the text through an analysis of a letter to a Dutch newspaper. The standpoint of the author (protagonist) is identi- fied, along with the opposing standpoint of the "missing" discussant. The text of the discussion is then organized into the four identifiable stages of the dispute, from con- frontation to concluding stage (ibid., pp. 93-94). A final qualification to note is that the pragma-dialectical approach is restricted to the resolving of disputes. This is a limitation, and frequently acknowledged as such: 20 Christopher W. Tindale the norms provided by the rules for critical discussions apply only where the discourse is actually aimed at resolving a dispute (ibid., p. 105). Grootendorst, in allowing that the dialectical approach is neither perfect nor final, anticipates problems because the rules are applicable only ifthe discussion is aimed at resolving a dispute. Although it is often obvious or reasonable to assume that this is the case, it is also recognized that "sometimes that is not at all the purpose of a discussion, and sometimes this may not be clear" (Grootendorst, 1987, p. 340). In a later footnote he submits that "[t]his reason might be seen as an indication that besides a dialectical approach there is also a need for a theory to cover situations other than those in which the chief issue is the resolu- tion of a dispute" (ibid., p. 341, fn. 20). Such a furtherance of the pragma-dialectical programme is apparent in the work of Douglas Walton.5 Walton's Pragma-Dialectical Conversion Joseph Wenzel, in his evaluation of the early version of the pragma-dialectical theory, emphasized the fact that the variety of argumentative situations indicate there are pur- poses requiring argumentation beyond the resolution of disputes (1985, p. 150). Doug- las Walton addresses this in his identification of different types of dialogue in which argumentation is employed. An advocate of a general dialectical approach to argu- mentation, Walton has recently identified himself more solidly with the pragma-dia- lectical school,6 particularly in his work on plausible argument (1992a, p. 4). Yet he is convinced that in sticking to one model of dialogue only, that of the critical discussion, van Eemeren and Grootendorst's theory exhibits a narrowness which he cannot con- done (ibid.,p. 34). Walton has identified as many as ten types of argumentative dia- logue, although in his Handbook for Critical Argumentation he is satisfied with five (1989, pp. 3-6). This shorter list includes the personal quarrel, forensic debate, persua- sion dialogue (or critical discussion), inquiry, and negotiation dialogue.? In identifying a wider range of dialogues in which argumentation occurs, Walton likewise widens the scope for the occurrence and treatment of fallacies. Still, in central agreement with van Eemeren and Grootendorst, Walton promotes the persuasion dialogue (critical discussion) as "the single most significant type of dialogue" since it represents a normative model for good dialogue "because it has normative rules that, taken together, set a standard of how good persuasion dialogue should take place" (1989, p. 9). Accordingly, he introduces rules and the stages of discussion that correspond with those of the established pragma-dialectical account. Hamblin and the Standard Treatment A further point on which the pragma-dialecticians are in agreement is that their theory gives rise to an account of fallacies which remedies a number of problems associated with the so-called "Standard Treatment" made famous by c.L. Hamblin in his 1970 book Fallacies. s Taken at face value, the "Standard Treatment" has been criticized by recent com- mentators as inadequate and simplistic.9 Of particular weight here have been the charges of the pragma-dialecticians. They follow Hamblin in noting that central to the "Stand- ard Treatment" is the dual notion that a fallacy is an argument which is in someway not Fallacies in Transition 21 valid while appearing to be so. And yet, throughout the tradition, from Aristotle to Locke to Copi, the "Standard Treatment" contains fallacies which are not arguments at all and others which, while being arguments, are not invalid (Grootendorst, 1987, p. 331). A number of reactions have been formed to this. Grootendorst identifies three: one has been to deny the existence of fallacies altogether, that is, as argued by Finocchiaro (1980), a common body of such errors cannot be catalogued; a second response has been to use an array of logical systems to treat an array of fallacies 10; the third response is to disassociate fallacies from particular argument forms and argue that whether a certain argument form like, say, the ad hominem is used fallaciously will depend on the circumstances involved. I I None ofthese approaches is attractive to Grootendorst, but he is especially vexed by the third because, requiring as it does the assessment of individual cases, "it is difficult, not to say impossible, to give general criteria for the analysis and evaluation of fallacies" (1987, p. 334). And yet, it is exactly in the direction of this third option that Walton has taken the pragma-dialectical account. For Walton, although the "Standard Treatment" can be seen to have arisen natu- rally out of Aristotle's theory of sophistical refutation, it has quickly lost the force of the Aristotelian account, which recognized that the same argumentation strategies that could be employed legitimately in one case could be used in another to create the appearance of refutation. Hence, the focus on "seeming validity" made by the "Stand- ard Treatment. 12" Consequently, Walton advises that "the root concept of fallacy needs to be rethought by rejecting the standard treatment notion of 'seeming validity' and getting back to the basis of the Aristotelian notion of sophistical refutation" (1992a, p. 268). But before we go any further with Walton's account of fallacy we should first detail the parent account of the pragma-dialectical perspective. The Pragma-Dialectical Theory of Fallacies As noted above, the pragma-dialectical theory of argument effectively restricts falla- cies to the violation of rules for conducting a critical discussion. "Fallacies are speech acts intended by the speaker to make a contribution to the resolution ofthe dispute but often in fact obstructing the way to a resolution" (van Eemeren and Grootendorst, 1984, p. 151). Hence, van Eemeren and Grootendorst regard as afallacy "every viola- tion which may result in the resolving ofthe dispute being made more difficult or even impossible" (ibid., p. 182). While I have characterized this approach as restrictive, van Eemeren and Grootendorst see among its merits that it is a wider notion of fallacy than is usual, especially in the "Standard Treatment." It is wider because it does not link fallacies solely to the argumentative stage of the discussion. Fallacies of ambiguity and "Straw Man," for example, neither of which need be arguments, can both be identified at the first stage of the confrontation and speak to misunderstandin~s that can affect the dis- pute. This particular merit is further evidenced ina case like the ad hominem: it "is a good illustration of a fallacy whose analysis presents considerable difficulties if falla- cies are linked exclusively to the invalidity ofthe arguments expressed by the prot ago- 22 Christopher W Tindale nist in his argumentation at the argumentation stage of discussion" (ibid., p. 192). Other merits are claimed for the pragma-dialectical account offallacies. It is noted, for example, that fallacies may be committed by either the protagonist or antagonist. But of much greater interest to those who have followed fallacy theory since Hamblin is the claim that the approach is better able than others to account for the large and eclectic group of so-called informal fallacies that creates problems in the "Standard Treatment" (Grootendorst, 1985, p. 161). Such problems, as we have seen, are traced to traditional attempts to have fallacies cohere around notions of argument and invalid- ity. But fallacies like "Many Questions" or "Ambiguity" need not be arguments, and the petitio principii need not be invalid. A systematic account which treats these as violations of rules for rational discussion avoids such difficulties. The account in Argumentation, Communication, and Fallacies matches twenty- six traditional fallacies, informal and formal, to one or more of the ten rules of con- duct, to one or more of the four stages of the discussion and to either or both of the two parties involved in the discussion (1992, pp. 212-215). "Denying the Antecedent," for example, is a violation by the protagonist of rule #8 (a party may only use arguments that are logically valid) at the argumentation stage (seventeen of the twenty-six falla- cies occur at this stage, fifteen of them exclusively so). The argumentum ad misericor- diam is a violation by the protagonist of rule #1 (parties must not prevent each from advancing standpoints) at the confrontation stage. John Woods, in a 1988 paper, notes two tendencies in the pragma-dialectical fallacy theory which may be at odds with each other and difficult to reconcile: on the one hand, all violations of the rules are faIlacious; whereas on the other hand, these must be interpreted to correspond to the traditional fallacies (Woods, 1988, p. 7). In fact, van Eemeren and Grootendorst have not felt quite so constrained. In addition to the twenty-six traditional fallacies that their account accommodates, they also identify eight '''new' categories for resolving disputes ... that should be considered as falla- cies" (van Eemeren and Grootendorst, 1992, p. 216). These include "Declaring a stand- point sacrosanct" and H[flalsely presenting something as a common starting point." However, it still remains that the pragma-dialectical account can be viewed as doing two quite different things and we will return to Woods' concerns later. Walton's Modifications Before proceeding to evaluate the pragma-dialectical account of fallacies, a note on Walton's contribution to the discussion is in order. Van Eemeren and Grootendorst have explicitly restricted the usage of 'faI\acy': "Fallacies are only fallacies ifthe dis- cussion involved can be regarded as part of a critical discussion" (1989b, p. 107). Walton breaks from this restraint. As noted earlier, while his research also centres on the persuasion dialogue, he expands some of his remarks to the several other dialogue types that he has identified. "In the study of informal fallacies," he writes, "the persua- sion dialogue (critical discussion) clearly has a central, or even primary, place among the many contexts of dialogue. However, because of dialectical shifting, many other Fallacies in Transition 23 contexts of dialogue are important to study as well" (1991, p. 39). This remark leaves open the possibility that fallacies can arise in dialogues outside of the critical discus- sion. In "Types of Dialogue, Dialectical Shifts and Fallacies," he pursues this idea. Here the peripheral models of dialogue are needed to identify and treat fallacies, and he illustrates this by showing how the ad hominem and ad baculum can be used in dialogue shifts to quarrels or negotiation dialogues (1992b). Of greater interest, however, is Walton's disagreement with van Eemeren and Grootendorst on a more fundamental question: the nature offallacy. Quite simply, he does not wish to classify all violations of rules of dialogue as fallacies. Some viola- tions he prefers to call blunders "because they are not systematic or clever deceptions that attempt to prove a point, but are simply errors or lapses that damage or weaken the case" (1989, p. 116). Walton also tries to clarify things by pointing out that fallacies do not have to be committed intentionally, although this is often the case. He draws a distinction be- tween a sophism, where a participant in a dialogue uses "deliberate deception or trick tactics" to his or her advantage, and a wrong (erroneous) inference, which is simply an unintentional error in reasoning and may be no more than a blunder (1 992a, pp. 238- 240). Generally, Walton's qualification is not as clear as it might be: some incorrect inferences are fallacies (ibid., p. 246), but arguments that contain errors of logical reasoning are not always fallacious arguments (ibid., p. 269). What is clear is that Walton is using 'fallacy' to mean more than simply a viola- tion of a rule that prevents the resolution of a dispute. In its clearest formulations, it suggests the existence of objective standards that exist outside of particular dialogues. For example, "[a] fallacy (properly speaking) is a particularly strong kind offailure of reasoned argumentation exhibiting an underlying structural flaw" (ibid., p. 34); and "[a] fallacy is an argumentation technique that could be used rightly in one context of dialogue, but is used wrongly in the particular case in question" (ibid., p. 267). These remarks speak to concerns still to be addressed, that the criteria for deciding fallacies are too subjective in the pragma-dialectical theory-depending on the behaviour and decisions of the discussion participants. They make a further important point that an argument can be a fallacy or not depending on the context; or put another way, falla- cious arguments have as their counterparts good argument forms. 13 A Critical Evaluation I have already noted some concerns that have been raised about the pragma-dialectical approach. I want now to look specifically at some of the problems related to the treat- ment of fallacies. The first of these is what might be called the "same fallacy" concern raised by Woods. He notes conflicting tendencies in the pragma-dialectical account. A fallacy is a violation of one of the rules. Can we turn this around and say that a violation of each of the rules is a specific fallacy? From one perspective, it seems that we can. To violate rule #1, for example, means the prevention of either party from advancing standpoints or casting doubts on standpoints. Since this can be achieved by advancing an ad baculum, or an ad hominem, or even an ad misericordiam, then, Woods points 24 Christopher W. Tindale out, these fallacies are all "the same fallacy-infelicities all" (Woods, 1988, p. 11). The same is true for other groupings of fallacies. In Argumentation, Communication, and Fallacies van Eemeren and Grootendorst cross-list the fallacies under the ten rules, but they do not resolve the ambiguity which Woods has identified: that a traditional fallacy is a fallacy both by virtue of violating one of the new pragma-dialectical rules and by virtue of its own traditional conditions. A second complaint requires more attention. It concerns the apparent freedom given to participants in the discussion to decide what they will or will not allow toward the resolution of their dispute. In its most extreme statements, this attitude seems to eschew any kind of objective criteria for judging fallaciousness. Sometimes van Eemeren and Grootendorst speak this way; other times they refer to conditions exter- nal to the discussion that may override the parties' agreements. To a certain extent this is Woods' concern back under a different guise: can the arguers remain locked within the code of conduct rules for a critical discussion (rules which are supposed to ensure the rationality of the participants), or must they give foremost attention to other criteria transported in from without (the conditions of traditional fallacies, for example)? In "Some Fallacies about Fallacies," Rob Grootendorst, in discussing appeals to experts, stresses the intersubjective. Whereas in other approaches the expert's exper- tise is an objective verifiable fact, in the pragma-dialectical account "it is regarded as the intersubjective agreement of the discussants" (Grootendorst, 1987, p. 339). The discussants will decide who will count, who is acceptable, as an expert. To a certain degree this attitude can be extended to other apparently fallacious moves, because fallaciousness is conditional on the context of the argument. A number of people have drawn attention to the problem in this. Hajek writes: We can imagine situations in which even modus ponens ... is used fallaciously-its appearance in a "fallacious" slippery slope argument might be such a situation. Con- versely, an argument form that is fallacious in the traditional sense, such as affirming the consequent ... could be used non-fallaciously ... for example, by a discussant who sincerely believes that it is valid and makes no attempt to trick his opponent thereby. (Hajek, 1992, p. 23.) But if van Eemeren and Grootendorst (and Walton) are sufficiently tied to their "violation-of-the-rules" criteria for fallaciousness, Hajek's scenarios might simply be strange but acceptable results of a radical revision in fallacy theory. Biro and Siegel's (1992) observations deserve a less complacent response, however. Here the issue becomes at what cost the resolution of a dispute should be desired. Participants might abide by the code of conduct and resolve their dispute but do so in quite "unjustified" and irrational ways. Again, if justification and rationality are inter- nal to the code of conduct, the pragma-dialecticians may not be so concerned. But Biro and Siegel provide two cases that show the potential for problems. In the first, dispu- tants may agree at the start of their discussion that handsomeness is the key criterion on which a candidate for election to political office should be judged, but they disagree about which candidate is the handsomest. At the end of their discussion, following the rules, they agree that it is candidate C. In the second case, the discussants agree "in their argumentative practice that only arguments with an even number of premises are valid or that whoever argues the loudest has the best argument" (ibid., p. 189). In each Fallacies in Transition 25 case the resolution would seem unjustified or irrational and Biro and Siegel believe the problem lies in making good argumentation dependent on proceeding according to rules agreed to by all disputants. As Blair and Johnson have pointed out, saying that a piece of argumentation is counter-productive when the resolution of a dispute is the goal is not the same as saying that there is something wrong with the reasoning in the argumentation (Blair and Johnson, 1993, p. 189). Beneath the surface concern of dis- pute resolution is a quite traditional interest in what Walton has labelled the error of reasoning sense of fallacy. Woods does not think that the pragma-dialecticians would be completely suscep- tible to Biro and Siegel's criticisms. He believes that implicit in their account is "some kind of standard of objective correctness or adequacy" (Woods, 1988, p. 13). And, indeed, as much is implied. But what exactly such a standard would involve is far from clear. Some attention to discussions in Argumentation, Communication, and Fallacies will illustrate my point. In the chapter assessing "Fallacies in Utilizing Argumentation Schemes," van Eemeren and Grootendorst discuss the argumentation scheme "argument from anal- ogy". Here they make remarks likely to fuel the concerns of Hajek and Biro and Siegel. For example: Whether this argumentation scheme is allowed to be used in a discussion, depends on whether the protagonist and the antagonist can agree on the conditions for its use [my italics]. Ifthey cannot and the protagonist nevertheless goes ahead using it, or ifthese conditions have not been fulfilled, he is guilty of one ofthe variants of the fallacy wrongful comparison or false analogy. (van Eemeren and Grootendorst, 1992, p. 161.) Here, the existence of the fallacy depends partly on the agreement of the discussants. But the mention of "conditions" for the use of the argumentation scheme suggests an alternative (secondary?) means of evaluation: "The argumentation scheme has been used correctly only if certain correctness conditions have been fulfilled" (ibid., p. (62). These correctness conditions involve the asking of certain "critical ques- tions." For example, with the argument from analogy one of the questions to be asked is "whether the comparison is really justified or whether there are crucial differences. If the comparison is defective, the argument from analogy is used incorrectly and con- stitutes a fallacy of false analogy" (ibid., p. 162). Here, we seem to find traditional criteria for judging arguments from analogy transported into the pragma-dialectical account. And yet, the decision whether or not the conditions are adequately fulfilled still rests with the discussants. What if they do not recognize an argument as falla- cious? Or does its fallaciousness really depend on their recognition of it? Furthermore, whether the conditions are adequately fulfilled actually depends not just on the discussants individually, but on their joint agreement. Consider one of the examples that van Eemeren and Grootendorst provide. The case involves a dispute (imagined) between Neil Lyndon and Julie Burchill. It comes from one of Lyndon's contributions to The Sunday Times (December 9, 1990). Lyndon quotes Burchill from Time Out: A good part-and definitely the most fun part--ofbeing a feminist is about frighten- ing men. American and Australian feminists have always known this, and absorbed 26 Christopher W Tindale it cheerfully into their act; one thinks of Shere Hite julienning men on phone-in shows, or Dale Spender telling us that a good feminist is rude to a man at least three times a day on principle. Of course, there's a lot more to feminism ... but scaring the shit out of scum bags is an amusing and necessary part because, sadly, a good many men still respect nothing but strength. In order to show, not that feminism is Nazism, but that "the language of vulgar intolerance is readily transportable," Lyndon attempts what Trudy Govier might call a "Refutation by Logical Analogy" (Govier, 1992, pp. 269-271) by substituting certain key terms for those in Burchill's discourse. A good part-and definitely the most fun part-ofbeing a Nazi is about frightening Jews. German and Austrian Nazis have always known this, and absorbed it cheer- fully into their act; one thinks of Ernst Rohm julienning Jews in the ghettos, or Goebbels telling us that a good Nazi is rude to a Jew at least three times a day on principle. Of course, there's a lot more to Nazism ... but scaring the shit out ofthe scumbags is an amusing and necessary part because, sadly, a good many Jews still respect nothing but strength. (Cited in van Eemeren and Grootendorst, 1992, p. 163.) Now the question is, is this a good move in argument? Does it depend on Burchill's agreement in the imagined dialogue? Because, if it does, we can probably write it off as a false analogy for that reason alone; in the heat of a dispute, Burchill is unlikely to allow it. Nor should Lyndon expect her to. But surely this is not enough. Van Eemeren and Grootendorst recognize as much, I think, as is indicated by the following: "If disputes are to be resolved, false analogies must be avoided, but these examples illus- trate that it is not always clear when an analogy is false. For distinguishing false analogy from correct analogy, unequivocal criteria are badly needed" (ibid., p. 164).14 These unequivocal criteria must be objective, but what are they, and more importantly, can they override any agreement made (mistakenly) by the discussants (or force an agreement where it does not exist)? Once we answer yes to this, as we must, we have firmly placed our hopes in objective criteria for deciding fallacies and not in the intersubjective agreement of discussants. Yet toward the close of their discussion of fallacies and argumentation schemes, van Eemeren and Grootendorst are still reluctant to cede the point. The protagonist and antagonist can only try to find a common criterion for weighing the advantages and disadvantages against each other and then decide about whether or not to allow this type of argument in their discussion. If they cannot reach an agreement, the protagonist should not use it. (Ibid., p. 66.) One has to believe that the upshot of this will be that perfectly good arguments will be dismissed because they do not contribute to resolving disputes. The search for objective criteria in the pragma-dialectical account returns again and again to the agreement of the discussants and the rules which govern their behavior. It is these rules (or the observance of them) which guarantee the reasonableness of the proceedings. So perhaps all along we have only needed to recognize these rules as the necessary objective conditions. In fact, this is the best answer we get: "For dialecticians who maintain a critical outlook, reasonableness is not solely determined by the norm of intersubjective agree- ment but also depends on the 'external' norm that this agreement should be reached in a valid manner" (ibid., p. 6)-that is, according to the rules. I take the quotation marks Fallacies in Transition 27 around the "external" to be alerting us to some qualifying sense of the word. This norm is not completely from the outside because the discussants have some say in its appli- cation. Elsewhere, van Eemeren and Grootendorst appeal to a two-part criterion of valid- ity-problem solving and conventional: "This means that the discussion and argumen- tation rules which together form the procedure put forward in a dialectical argumenta- tion theory, should on the one hand be checked for their adequacy regarding the reso- lution of disputes, and on the other for their intersubjective acceptability for the discussants" (1988, p. 280). That the same explanation is repeated in subsequent texts indicates this to be a fundamental and continuing feature of the account (1995, p. 129). The rules should be adequate for resolving the dispute and be acceptable to the discussants. How do the two parts of this criterion work together? How is the agree- ment of discussants not, in fact, the resolution of the dispute? Or must the dispute only be resolved by the rules? This last question is answered in the affirmative. The rules should have priority over the agreement of the discussants . . . . the rules indicate when participants intending to resolve a dispute are entitled, or indeed obliged, to carry out a particular move. They must observe all the rules that are instrumental to resolving the dispute. Any infringement of a discussion rule, whichever party commits it and at whatever stage in the discussion, is a possible threat to the resolution of a dispute and must therefore be regarded as an incorrect discussion move. (Ibid., p. 131.) Several observations are appropriate here. (i) The rules govern the discussants. But, as we have seen, some rules depend upon the argument scheme (as with analogy) being acceptable to the discussants. In such instances the discussion rule does not so much govern the agreement of the discussants but cede authority to it (ii) Discussants are required to observe all the rules instrumental to resolving the dispute. Which rules are at stake are presumably recognized by the discussants. One might take from this that the failure to observe a rule is an "infringement of a discus- sion rule," which in tum is "an incorrect discussion move." As we have seen else- where, such an infringement (violation) constitutes, by definition, a fallacy. So one might take from this that there is a fallacy of omission: the fallacy of not observing a rule-any rule. But this would not be a fallacy like any other in the account since it is not, strictly speaking, a violation of a rule. (iii) Finally, the soundness of the argumentation is decided by its adequacy in resolving a dispute and its "intersubjective acceptability to the discussants" (van Eemeren and Grootendorst, 1988, p. 280, and 1995, p. 129). Thus, the acceptability of the argumentation to both of the discussants is a determining factor in deciding its validity. On this ground, it seems the discussants can reject good arguments. Is it enough for the argument to resolve the dispute without both discussants' agreement? Presum- ably not, since the dispute (between the discussants) can hardly be considered resolved ifthey cannot accept the reasoning instrumental in resolving it (unless that judgment is made by a third party). Now, in this normative model it might well be prescribed that the discussants should accept argumentation which resolves the dispute. But the prac- tical application of this begs to be clearly demonstrated. Besides which, it hardly makes 28 Christopher W. Tindale sense to place so much importance on the agreement of discussants if the rules con- strain them in how they can and cannot respond. I can begin to demonstrate what I take to be the underlying problem in all of this by analyzing the 1992 van Eemeren and Grootendorst text further. In the chapter on "Fallacies in Choosing the Means of Defense," the authors give some attention to playing on the audience's emotions and more particularly to the usage of the ad populum. Now the reader might well wonder: from where did the audience come? Van Eemeren and Grootendorst have been very clear to restrict their analysis to the resolving of disputes between protagonist and antagonist, to reach agreements or minimize disa- greements. And fallacies have been linked to discussion between the two parties in- volved only. "[M]anipulating the emotions of those present" (ibid, p. 134) does not fit happily into such a scenario. In fact, the entire text (and pragma-dialectical programme) has been set up to be resolution-oriented and not audience-oriented (dialectical and not rhetorical) (ibid, pp. 7-8). It is this bias which is starting to suggest difficulties in the account. In fact, van Eemeren and Grootendorst have an answer in this instance. "Gener- ally speaking, the argumentum ad populum thrives best in discussions on a broad scale, in which many people consider themselves involved" (ibid, p. p. 135). So the audi- ence in question is one which considers itself a participant in the dialogue, a party to be persuaded to accept a standpoint. Although certain difficulties are suggested by such a "pluralistic" discussant, I will not address them here but will proceed to a more substantial, and concluding, point which has been suggested by this turn in the discus- sion. Rhetorical Elements: Audiences, Readers, Third Parties The audience-as-participant has an overt place in the pragma-dialectical perspective and will require, one assumes, some consideration of rhetorical features. But the audi- ence-as-spectator is also involved in the perspective, although left in the background. Bringing it to the foreground will not only indicate its underlying importance (because what the audience-as-spectator brings with it is the demand for clear objective criteria) but also require considerations of further rhetorical features. The audience-as-spectatbr, onlooker, implicit judge, is hidden throughout the texts of the pragma-dialecticians as is evidenced by the mention of audiences, readers, and third parties. Consider some comments from two of the main texts. In Speech Acts in Argumentative Discussions we find "[the first two variants of the ad hominem] are not directly addressed to the other party (though they are of course directed against him), but to a third party consisting of spectators. These are rhetorical rather than dialecti- cal tricks" (1984, p. 191). In Argumentation, Communication, and Fallacies, in a dis- cussion of the "Straw Man" fallacy, we find: "the manoeuvre is more likely to succeed with an audience not entirely sure exactly what the other party has asserted than the opponent himself. In the case of a polemic in a newspaper, for example, the readers will rarely be able to lay hands on the article they need" (1992, p. 126). These quotes take us outside the dispute to a silent audience who is expected to play some role of adjudication. In the final quote that I will offer, that role is made explicit. Fallacies in Transition 29 In the chapter on "Fallacies in Concluding the Discussion" van Eemeren and Grootendorst write: . .. If they cannot agree on this, the dispute continues. As an unbiased third party will often be in a better position to decide who has won, to an outsider it may, mean- while, be perfectly clear what the outcome is, but this does not really solve the prob- lem. A joint discussion by protagonist and antagonist as to who has won the discussion is usually impossible if the discussion is implicit, as in the case with many written texts. Of necessity it is then up to the reader to decide for himself whether the pro- tagonist has provided a conclusive defense of his standpoint. (Ibid., pp. 184-185.) So this is how we would be expected to deal with the BurchilllLyndon dispute: thrown back on our own resources. The final arbiter for so many of the argumentative texts that we encounter and the dialogues in which the disputants cannot agree will be the audience/reader, and this audience/reader will of necessity use objective criteria to decide the rationality of the arguments concerned. One of the merits of the pragma-dialectical approach had been its combination of the logical and rhetorical perspectives. Hence the recognition paid to some "rhetorical tricks." But the dialectical approach was preferred because argumentation was deemed to be resolution-oriented and not audience-oriented. Now it seems it is far more audi- ence-oriented than was first allowed. Discussing the rhetorical approach of Perelman and Olbrechts-Tyteca, van Eemeren and Grootendorst note that its notion of success in argumentation depends on the audience accepting it. It offers "a rhetorical concept of reasonableness in which soundness is equated with effectiveness" (1988, p. 277) This is rejected by van Eemeren and Grootendorst because it is too relative a standard of reasonableness. But that standard now appears to underlie important aspects of the pragma-dialectical programme. In fact, the reader who must decide and the unbiased third party are reminiscent of the objective standard served by Perelman and Olbrechts- Tyteca's universal audience. Finally, we must ask whether all argumentation must be essentially dialogical rather than monological. Van Eemeren insists that when a protagonist anticipates a discussion, even though it may never be verbalized, "the moment he argues, the very fact of his argumentation testifies to the legitimacy of a dialectical approach to argu- mentation" (van Eemeren, 1986, p. 2). By contrast, in her assessment of the pragma- dialectical perspective, Sally Jackson suggests that: "S [speaker] may construct an argument not/or the sake oiL [listener] but/or the sake o/some audience, either real or imagined. S may make an argument with no expectation or intention of convincing L" (Jackson, 1985, p. 133). There is a distinction implied here which is worth noting. It may be a matter for debate as to whether I can make an argument with no intention of convincing, as Jackson proposes, but it is certainly a suggestion that I operate quite differently in relation to two different audiences: one who is a participant (real or imagined) in my argument, and the other who is a silent spectator. The silent audience is the one I invite in to read (or hear) my mono logical investigations, working through a problem for myself. On another occasion, I may argue the results of that mono logical inquiry and then arm my audience with all kinds of counter-argumentation tools to further the effectiveness of 30 Christopher W Tindale my appeal. The pragma-dialectician may insist that in each case a dialogue is involved, but, if so, it is not the same kind of dialogue or audience. The one who silently stands behind my private reasoning is not one to be convinced but one who represents objec- tive standards of rationality against which I judge the merits of my own thought. It is this representation of objective standards which is never given sufficient space in the pragma-dialectical account and which needs to be developed. But such a development must play down the emphasis on the dialectical and give more attention to the logical and, particularly, the rhetorical perspectives. What those objective standards should be awaits the outcome of such further studies, but one suspects that something of the traditional conditions ofthe fallacies (formal and informal) will still be around to play a part in them. Notes I In my judgement, excellent synopses of the approach may be obtained from the papers "Argumen- tation Analysis: A Dutch Counter-balance," (van Eemeren, 1988) and "Rationale for a Pragma- dialectical Perspective" (van Eemeren and Grootendorst, 1988). 2 So there is an underlying notion of "reasonableness" associated with this account. See my "Rea- sonableness and the Limits to Persuasion," (1993) for a critical discussion of this idea. l The argument from analogy is an example of an argumentation scheme. Generally, the introduction of this concept seems an innovation of the post-I 984 work, although it is influenr;ed by other au- thors. An argumentation scheme is defined as "a more or less conventionalized way of representing the relation between what is stated in the argument and what is stated in the standpoint" (Argumen- tation, Communication, and Fallacies, p. 96). 4 A summary of the rules can be found in Argumentation, Communication, and Fallacies, p. 209. l Here I am interested in the "pragma-dialectical" stage of Walton's work. As I discuss elsewhere, this is a stage he has moved beyond in developing a distinct notion of "fallacy" (see TindaJe, 1997). 6 Hence, fulfilling the promise seen in him by van Eemeren and Grootendorst: "What is needed for the development of the study of argumentation and fallacies is, in our view, a radical pragma-dialectical approach. Walton is on his way. He only needs a small push" (van Eemeren and Grootendorst, I 989b, p. 105). 7 In "What is Reasoning? What is an Argument?" (1990) Walton adds three more: the planning committee dialogue, the pedagogical dialogue, and the dialogue of expert consultation; and in Beg- ging the Question: Circular Reasoning as a Tactic of Argumentation (1991, p.43) we find a further two: the deliberation dialogue and the interview. S More recently, the very existence of such a tradition of a "Standard Treatment" has been challenged by Ralph Johnson who, while recognizing the significance of Hamblin's work in stimulating subse- quent study of fallacies, argues that Hamblin's list of texts is selective and even then offers no "uniform and homogeneous treatment of fallacies" (1990). 9 See remarks along these lines in David Goodwin (1992). 10 This is the approach favoured by John Woods and Douglas Walton in Argument: The Logic of the Fallacies (1982). See also their Fallacies: Selected Papers 1972-1982 (1989). II This approach is preferred by a number of people in one fashion or another, among them: Leo Groarke and Christopher Tindale, "Critical Thinking: How to Teach Good Reasoning" (1986). 12 See Walton, Plausible Argument in Everyday Conversation (I992a, pp.264-267) for a discussion of the Aristotelian account. 13The remainder of the paper focuses primarily on the Dutch account. For a detailed examination of how the problems discussed here affect Douglas Walton's treatment of , fallacy , see my ;'Walton and the Standard Treatment," in Proceedings of the Third lnternational Conference on Argumentation ed. Frans H. van Eemeren et at. (Amsterdam, Holland: SICSAT, 1995). 14 Certain "critical questions" for assessing analogies had been suggested earlier, see van Eemeren Fallacies in Transition 31 and Grootendorst (1992), p. 99 and p. 102. Bibliograpby Biro, John and Siegel, Siegel. 1992. "Normativity, Argumentation and an Epistemic Theory of Fallacies," in Argumentation Illuminated, eds.Frans H. van Eemeren, Rob Grootendorst, J. Anthony Blair, and Charles A. Willard, pp. 85-103. Amsterdam: SICSA T. Blair, J. Anthony and Johnson, Ralph H. 1993. "Dissent in Fallacyland, Part 1: Problems with van Eemeren and Grootendorst," in Argument and the Postmodern Challenge: Proceedings of the Eighth SCAIAFA Conference on Argumentation, ed. Raymie E. Mckerrow, pp. 188· 190. Annandale, VA: Speech Communication Association. van Eemeren, Frans H. 1985. "Response," in Argument and Social Practice: Proceedings of the Fourth SCAIAFA Coriference on Argumentation, eds. J. Robert Cox, Malcom O. Sillars, Greg B. Walker, pp. 154-159. Annandale, VA: Speech Communication Association. -. 1986. "Dialectical Analysis as a Normative Reconstruction of Argumentative Discourse," Text 6: pp. 1-16. -. 1987. "For Reason's Sake: Maximal Argumentative Analysis of Discourse," in Argumenta- tion: Across the Lines of Discipline, eds. Frans H. van Eemeren, Rob Grootendorst, 1. Anthony Blair and Charles A. Willard, pp. 201-215. Dordrecht: Foris Publications. -. 1988. "Argumentation Analysis: A Dutch Counter-balance," in Critical Thinking: Proceed- ings of the First British Coriference on Informal Logic and Critical Thinking, ed. Alec Fisher, pp. 39-53. Norwich: University of East Anglia. van Eemeren, Frans H. and Grootendorst, Rob. 1982. "The Speech Acts of Arguing and Con- vincing in Externalized Discussions," Journal of Pragmatics 6: pp. 1-24. -.1982. "Unexpressed Premises.: Part I,"Journal of the American Forensic Association 19: pp. 97-106. -. 1983. "Unexpressed Premises: Part II," Journal of the American Forensic Association 19: pp.215-225. -. 1984. Speech Acts in Argumentative Discussions. Dordrecht: Foris Publications. -. 1987. "Fallacies in a Pragma-Dialectical Perspective," Argumentation I: pp. 283-301. -.1988. "Rationale for a Pragma-Dialectical Perspective," Argumentation 2: pp. 271-291. -. 1989a. "A Pragma-Dialectical Perspective on Norms," in Norms in Argumentation, ed. Robert Maier, pp. 97-112. Dordrecht: Foris Publications. 1989b. "A Transition Stage in the Theory of Fallacies," Journal of Pragmatics 13: pp. 99- 109. -. 1989c. "Writing Argumentative Texts From Analysis to Presentation: A Pragma-Dialectical Approach," in Spheres of Argument: Proceedings of the Sixth SCAIAFA Conference on Argumentation, ed. Bruce E. Gronbeck, pp. 324-330. Annandale, VA: Speech Communica- tion Association. -. 1991. "The Relevance Problem in the Analysis of Argumentative Texts: A Pragma-Dialec- tical Reconstruction," Conference Paper: "Relevance in Argumentation," McMaster Uni- versity, Hamilton, Ontario: 15 pp. 1992. Argumentation. Communication, and Fallacies: A Pragma-Dialectical Perspective, Hillsdale, NJ: Lawrence Erlbaum Associates, Publishers. -. 1995. "Perelman and the Fallacies," Philosophy and Rhetoric 28: pp. 122-133. van Eemeren, Frans H., Grootendorst, Rob and Meuffels, Bert. 1989, "The Skill of Identifying Argumentation," Journal of the American Forensic Association, 25: pp. 239-245. 32 Christopher W. Tindale van Eemeren, Frans H., Grootendorst, Rob and Kruiger, Tjark. 1984. The Study of Argumenta- tion. New York: Irvington Publishers, Inc. [Reprinted 1987 as Handbook of Argumentation Theory: A Critical Survey of Classical Backgrounds and Modern Studies. Dordrecht: Foris Publications. ] Finocchiaro, Maurice A. 1980. Galileo and the Art of Reasoning: Rhetorical Foundations of Logic and Scientific Method. Dordrecht: D. Reidel Publishing Co. Grootendorst, Rob. 1985. "Response," in Argument and Social Practice: Proceedings of the Fourth SCAIAFA Conference on Argumentation, eds. J. Robert Cox, Malcom O. Sillars, Gregg B. Walker, pp. 159-161. Annandale, VA: Speech Communication Association. -. 1987. "Some Fallacies about Fallacies," in Argumentation: Across the Lines of Discipline, eds. Frans H. van Eemeren, Rob Grootendorst, J. Anthony Blair and Charles A. Willard, pp. 331-341. DordrechtIHolland: Foris Publications. Goodwin, David. 1992. "The Dialectic of Second-Order Distinctions: The Structure of Argu- ments about Fallacies," Informal Logic, 14: pp. 11-22. Govier, Trudy. 1992. A Practical Study of Argument, 3rd edition. Belmont, CA: Wadsworth. Groarke, Leo and Tindale, Christopher W. 1986. "Critical Thinking: How to Teach Good Rea- soning," Teaching Philosophy 9: pp. 301-318. Hajek, Alan. 1992. "Trick-arguing," The Times Literary Supplement, September 4, p. 23. Hamblin, c.L. 1970. Fallacies. London: Methuen & Co. Jackson, Sally. 1985. "What Can Speech Acts Do For Argumentation Theory?" in Argumenta- tion and Social Practice: Proceedings of the Fourth SCAIAFA Conference on Argumenta- tion eds. J. Robert Cox, Malcom O. SiIIars, et. al., pp. 127-138. Annandale, V A: Speech Communication Association. Johnson, Ralph H. 1990. "Hamblin on the Standard Treatment," Philosophy and Rhetoric, 23: pp.153-167. Johnson, Ralph H. and Blair, 1. Anthony. 1987. "Argumentation as Dialectical," Argumentation 1: pp. 41-56. Tindale, Christopher W. 1993. "Reasonableness and the Limits to Persuasion," The Canadian Journal of Rhetorical Studies, 3: pp. 133-148. -. 1995. "Walton and the Standard Treatment," in Proceedings of the Third International Con- ference on Argumentation, Vol. II, Analysis and Evaluation, eds. Frans H. van Eemeren, et al., pp. 274-285. Amsterdam: SICSAT. -. 1997. "Fallacies, Blunders, and Dialogue Shifts: Walton's Contributions to the Fallacy De- bate," Argumentation II: pp. 341-354. Walton, Douglas. 1987. Informal Fallacies: Towards a Theory of Argument Criticisms. Am- sterdam: John Benjamins. -. 1989. Informal Logic: A Handbookfor Critical Argumentation. Cambridge: Cambridge University Press. -.1990. "What is Reasoning? What is an Argument?" Journal of Philosophy 87: p. 399-419. -.1991. Begging the Question: Circular Reasoning as a Tactic of Argumentation. New York: Greenwood Press. 1992a. Plausible Argument in Everyday Conversation. New York: State University of New York Press. -. 1992b. "Types of Dialogue, Dialectical Shifts and Fallacies," in Argumentation Illumi- nated, eds. Frans H. van Eemeren, Rob Grootendorst, 1. Anthony Blair, and Charles A. Willard, pp. 133-147. Amsterdam: SICSAT. Fallacies in Transition 33 Wenzel, Joseph W. 1985. "Toward a Nonnative Theory of Argumentation: Van Eemeren and Grootendorst's Code of Conduct for Rational Discussions," in Argumentation and Social Practice: Proceedings of the Fourth SCAIAFA Conference on Argumentation eds. J. Robert Cox, Malcom O. Sillars, et, al., pp. 139-153. Annandale, V A: Speech Communication As- sociation. Woods, John. 1985. "Pragma-dialectics: A Radical Departure in Fallacy Theory," iSSA News- letter 4: pp 5-15. -. 1988. "Buttercups, GNP's and Quarks: Are Fallacies Theoretical Entities?" informal Logic, 10: pp. 67-76. Woods, John and Walton, Douglas. 1982. Argument: The Logic of the Fallacies. Toronto: McGraw-Hill Ryerson. -. 1989. Fallacies: Selected Papers 1972-i982. Dordrecht: Foris Publications. Wreen, Michael. "Look, Ma!No Frans!" Pragmatics and Cognition. (1994): 285-306. CHRlSTOPHER W TlNDALE, DEPARTMENT OF PHiLOSOPHY TRENT UNiVERSiTY PETERBOROUGH,ONTARlO K9J7B8 D