Published in Fallacies: Classical and Contemporary Readings, edited by Hans V. Hansen and Robert C. Pinto (1995).
The Pragma-Dialectical Approach to Fallacies
Frans H. Van Eemeren and Rob Grootendorst
Thanks to Hamblin's book Fallacies (1970), it is now common knowledge that the Standard Treatment of fallacies suffers from serious theoretical and practical defects. Many generally recognized fallacies clearly fall outside the scope of the standard definition of a fallacy as an argument that seems valid but is not valid: in some cases because there is not the slightest question of there being an argument (many questions, ad baculum); in other cases because, logically speaking, the argument in question is not invalid at all (circular reasoning); in still other cases (ad verecundiam, ad populum) because it would be missing the point completely to identify the error as one of invalidity.1
In our own efforts to offer an alternative to the Standard Treatment we started from the consideration that there is no reason to assume from the outset that all the fallacies are essentially logical errors. We were convinced that the single-minded preoccupation with the logical aspects of arguments should be rigorously abandoned. For the informal fallacies it had, after all, only led to largely unsatisfactory and unsystematic ad hoc analyses. In our opinion, the fallacies could be better understood if they were treated as faux pas of communication -- as wrong moves in argumentative discourse. Viewed from this perspective, a fallacy is a hindrance or impediment to the resolution of a disagreement, and the specific nature of each of the fallacies depends on the exact manner in which it interferes with the resolution process. This was our starting point in setting about to develop a general and comprehensive approach to argumentation that covers the whole domain of the fallacies.
Some Basic Characteristics of the Pragma-Dialectical Approach
Argumentation, being a phenomenon of verbal communication, should be studied as an integral part of the conduct of argumentative discourse. Its quality and possible flaws are to be measured against criteria that are appropriate for determining the reasonableness of such discourse. The study of argumentation should therefore be construed as a special branch of linguistic pragmatics in which descriptive and normative perspectives on argumentative discourse are methodically integrated. Contrary to what some formal and informal logicians seem to think, this study cannot be based, unilaterally, on mere intellectual idealization; and contrary to what some discourse and conversation analysts seem to think, neither can it be based on mere empirical observation.2 Both the limitations of the nonempirical regimentation exemplified in normative modern logic and the limitations of the noncritical explanation exemplified in contemporary descriptive linguistics are to be systematically overcome. This calls for the development of an integrating and coherent research program in which normative and descriptive insights are to be closely interwoven.
The pragma-dialectical research program is based on the assumption that, on the one hand, a philosophical ideal of reasonableness must be developed and, starting from this ideal, a theoretical model for acceptable argumentative discourse. On the other hand, argumentative reality must be investigated empirically, so that it becomes clear how argumentative discourse is in fact conducted. Then the normative and the descriptive dimensions must be linked together by developing instruments that make it possible to analyze argumentative practice from the perspective of the projected ideal of reasonable argumentative discourse. Finally, the problems that occur in practice nust be systematically diagnosed and it must be determined how they can be tackled methodically. The pragma-dialectical research program therefore includes a philosophical, a theoretical, an empirical, an analytical, and a practical component.
In the implementation of the pragma-dialectical research program, the study of argumentation is approached with four basic metatheoretical premises, each of which represents a point of departure from other contemporary perspectives. It is characteristic of the pragma-dialectical approach that the subject matter under investigation is being externalized, functionalized, socialized and dialectified. What do these meta-theoretical labels mean?
Making an argument pragmatically presupposes a standpoint and at least the potential for opposition to that standpoint. In order to find out whether or not a person's opinions make sense and whether or not his reasoning holds water, he must submit them to public scrutiny. While beliefs, inferences, interpretation, and so on certainly underlie argumentation (and other discursive activities), the way in which the argument is expressed and proceeds is channeled by a system of public commitment and accountability. The motives people have for holding a position may sometimes be different from the grounds they will offer and accept in its defense. What is at stake in the study of argumentation is not so much the psychological dispositions of the arguers, but the positions to which the parties can be held committed in the discourse, whether these positions have been expressed directly or indirectly. This is why the pragma-dialectical approach to argumentation concentrates on externalized commitments.
In approaches that are merely concerned with argument "as a product" arguments are typically seen as an extemalization of an individual thought process, abstracting elements of reasoning such as "major premise," "minor premise," and "conclusion" from the communicative process in which they occur. Then, the central question becomes one of assessing whether and how these elements hold together to validate the arguer's position. But argumentation does not consist in a single individual privately drawing a conclusion: it is part of a discourse procedure whereby two or more individuals try to arrive at an agreement. The collaborative way in which the protagonist of a standpoint in the fundamentally dialogical interaction responds to the -- real or projected -- questions, doubts, objections, and counterclaims of an antagonist is reflected in the argumentation. This is why in the pragma-dialectical approach argumentation is put in the social context of a problem-solving process.
Not only in formal and informal logical approaches to arguments, but also in studies of fallacies and practical argumentation, argumentation is often described in purely structural terms. Although structural descriptions have much to recommend them, they tend to ignore the functional rationale of the structural design of the discourse. Argumentation arises in response to, or anticipation of, disagreement, and particular lines of justification are fitted to realize this purpose in a particular case. The need for argumentation, its structure, and the requirements of justification are all adapted to the opposition, doubts, objections, and counterclaims that have to be met. An adequate description and evaluation of argumentation can only be given if the purpose for which the argumentation is put forward in the interaction is duly taken into account. This is why the pragma-dialectical approach to argumentation concentrates on its function in managing the resolution of disagreements.
Discourse and conversation analysts generally restrict themselves to describing argumentation as it occurs, without regard for how it ought to occur if it is to be appropriate for resolving a difference of opinion. The argumentation is appropriate only if it is capable of accommodating the critical reactions of a rational judge. In order to determine whether this is the case, the argumentation is to be viewed as part of a critical discussion conducted in accordance with a problem-valid and conventionally valid discussion procedure.3 The problem-validity of a discussion procedure depends on its efficiency in achieving a resolution to the disagreement and its efficacy in furthering the resolution process while avoiding "false" resolutions. The conventional validity depends on the intersubjective acceptability of the procedure. The procedure provides a set of standards for rational judgment. This is why in the pragma-dialectical approach argumentation is put in the dialectical perspective of a critical discussion.
In the pragma-dialectical approach, externalization, socialization, functionalization, and dialectification of the notion of argumentation are realized by making use of pragmatic insight from discourse and conversation analysis and dialectical insight from critical rationalist philosophy and dialogical logic.4 Dialectification is achieved by viewing argumentation in the perspective of a regulated critical discussion aimed at resolving a difference of opinion. Functionalization is realized by defining argumentation as a complex speech act that can only serve its purpose adequately if certain identity and correctness conditions are fulfilled. Socialization is a result of putting argumentation in the collaborative context of an interaction between two or more discussants. Finally, externalization is accomplished by identifying argumentation with the specific commitments created by the performance of the speech act of arguing in a certain context of disagreement.5 As a matter of course, notions referring to other components of argumentative discourse, such as standpoints, are treated similarly.
In the pragma-dialectical approach, argumentation is treated as part of a reasonable argumentative discourse aimed at resolving a difference of opinion. In implementing the research program, it should therefore be clarified what this involves. For this purpose, in the theoretical component, the critical-rationalist philosophy of reasonableness is given shape in an ideal model of a critical discussion.6 The model specifies the various stages that are to be distinguished in the resolution process and the verbal moves that are integral parts of each of these stages. The principles authorizing the distribution of the verbal moves over the consecutive stages are accounted for in a set of rules for the performance of speech acts. Taken together, these rules constitute a theoretical definition of a critical discussion.
Fallacies and the Rules for Critical Discussion
The four stages that are analytically distinguished in the resolution process are the confrontation stage, the opening stage, the argumentation stage, and the concluding stage. In the confrontation stage a difference of opinion presents itself through the confrontation between a standpoint and (real or imagined) nonacceptance of this standpoint. In the opening stage the parties to the dispute are identified as well as their apparent premises -- procedural or otherwise. In the argumentation stage the party that acts as the protagonist defends his standpoint methodically against the critical reactions of the antagonist. In the concluding stage the parties establish what the result of the discussion is. By specifying a critical discussion in this way, a heuristic, analytical, and critical framework is created for dealing with argumentative discourse.
In a critical discussion, the protagonist and the antagonist of the standpoint at issue must in all stages of the discussion observe all the rules that are instrumental to resolving the dispute. For the elucidation of the pragma-dialectical analysis of fallacies aimed for in this article we offer the simplified and nontechnical version of the rules introduced in van Eemeren and Grootendorst (1992a), where the rules are reduced to a succinct recapitulation often basic rules:7
- Parties must not prevent each other from advancing standpoints or casting doubt on standpoints.
- A party that advances a standpoint is obliged to defend it if the other party asks him to do so.
- A party's attack on a standpoint must relate to the standpoint that has indeed been advanced by the other party.
- A party may defend his standpoint only by advancing argumentation relating to that standpoint.
- A party may not falsely present something as a premise that has been left unexpressed by the other party or deny a premise that he himself has left implicit.
- A party may not falsely present a premise as an accepted starting point nor deny a premise representing an accepted starting point.
- A party may not regard a standpoint as conclusively defended if the defense does not take place by means of an appropriate argumentation scheme that is correctly applied.
- In his argumentation, a party may only use arguments that are logically valid or capable of being validated by making explicit one or more unexpressed premises.
- A failed defense of a standpoint must result in the party that put forward the standpoint retracting it, and a conclusive defense of the standpoint must result in the other party retracting his doubt about the standpoint.
- A party must not use formulations that are insufficiently clear or confusingly ambiguous and he must interpret the other party's formulations as carefully and accurately as possible.
In principle, each of these ten discussion rules constitutes a separate and different standard or norm for critical discussion. Any infringement of one or more of the rules, whichever party commits it and at whatever stage in the discussion, is a possible threat to the resolution of a difference of opinion and must therefore be regarded as an incorrect discussion move. In the pragma-dialectic approach, fallacies are analyzed as such incorrect discussion moves in which a discussion rule has been violated. A fallacy is then defined as a speech act that prejudices or frustrates efforts to resolve a difference of opinion and the use of the term "fallacy" is thus systematically connected with the rules for critical discussion.
In this approach, as soon as fallacies in argumentative discourse are discussed, the discourse is treated as if it were aimed at resolving a difference oi opinion. In practice, a discourse will hardly ever be completely resolution-oriented -- nor completely non-resolution-oriented, for that matter. For a realistic appreciation of the scope of the pragma-dialectical approach to fallacies, it is important to note that the norms provided by the rules for critical discussion apply only where and insofar as the discourse concerned is indeed aimed at resolving a difference of opinion.8 Although it is often clear, or can he reasonably assumed, that this is -- or is not -- the case, it is not always obvious. This is one of the reasons why a discourse can only be fully and methodically screened for fallacies if it is first adequately analyzed.
As it occurs in practice, even a discourse that is clearly argumentative will in many respects not correspond to the ideal model of a critical discussion -- or at least not explicitly, completely, and immediately. In many cases, the how's and why's of the divergent forms of argumentative reality can be easily explained with the help of some empirical insight and in a great many cases the differences can even be explained away. At any rate, it would certainly not do if all verbal behavior that does not seem to agree with the model was automatically declared defective just like that. The discourse as it has been brought to the fore can only be evaluated adequately if one has first accurately determined what it actually conveys.
An analysis undertaken from a pragma-dialectical perspective is aimed at reconstructing all those and only those elements in the discourse that are pertinent to the resolution process. Such a resolution-oriented reconstruction concentrates on identifying the speech acts that play a potential part in bringing a difference of opinion to an adequate conclusion. In this endeavor, speech acts that are immaterial to the resolution process are ignored, implicit elements are made explicit, indirect speech acts are restated as direct speech acts, and unintentional swerves from the resolution path are rearranged.9 By pointing out which speech acts are relevant in the consecutive stages of the resolution process, the ideal model of a critical discussion gives specific heuristic direction as to which speech acts are to be considered in the reconstruction process. Thus, the model is a valuable tool for accomplishing a systematic analysis.
It is, of course, crucial that the proposed analysis of a discourse in terms of a critical discussion is indeed justified. The reconstruction should be faithful to the commitments that may rightly be ascribed to the actors on the basis of their contributions to the discourse. In order not to "overinterpret" the speech act potential of the discourse, a sensitivity must be maintained to the details of the presentation, the general rules for communication, and the specific contextual constraints inherent in the speech event concerned.10 In those exceptional cases where there is really no clue whatsoever as to whether a speech act is intended to contribute to the resolution process, it is the most charitable solution to opt for a "maximally dialectical analysis" and interpret the utterance as a constructive move in a critical discussion, thus deciding for an analysis "for reason's sake."11
When it comes to the detection of fallacies, the pragma-dialectical analysis proceeds in a number of steps. An utterance must first be interpreted as a particular kind of speech act. After it has been established that the speech act concerned has indeed been performed in a context of discourse aimed at resolving a difference of opinion, it must be determined whether the performance of this speech act agrees with the rules for critical discussion. If the speech act proves to be a violation of one of the norms pertaining to a particular stage of the resolution process, it must be determined what kind of violation it entails. Which specific criterion for satisfying the norm has not been met? Only after this question has been answered can it be determined which fallacy has been committed.
There are many things that can go wrong in resolving a difference of opinion by argumentative discourse. We shall now turn our attention to a discussion of the most important violations of the rules for a critical discussion. The list of violations we will explain is by no means complete, but it gives one a fair impression of the great variety of fallacious moves that might be detected in the various stages of an argumentative discourse that has been analyzed as a critical discussion.
Overview of Violations of Rules for Critical Discussion
Rule 1 can be violated -- at the confrontation stage -- in various ways, both by the protagonist and the antagonist. A discussant can impose certain restrictions on the standpoints that may be advanced or called into question; he can also deny a certain opponent the right to advance the standpoint he likes to advance or to criticize the standpoint he likes to criticize. Violations of the first kind mean that certain standpoints are in fact excluded from the discussion or that particular standpoints are declared sacrosanct, so that the opponent is prohibited from casting doubt on them and they are rendered immune to criticism. Violations of the second kind are directed to the opponent personally and aim at eliminating him as a serious partner in the discussion. This may be done by putting pressure on him by threatening him with sanctions (argumentum ad baculum) or by playing on his feelings of compassion (argumentum ad misericordiam), but also by discrediting his expertise, impartiality, integrity, or credibility (argumentum ad hominem).
Rule 2 can be violated -- at the opening stage -- by the protagonist by evading or shifting the burden of proof. In the first case, he attempts to create the impression that there is no need to defend his standpoint and no point in calling it into question by presenting it as self-evident, giving a personal guarantee of the rightness of the standpoint (special variant of argumentum ad verecundiam) or immunizing the standpoint against criticism. In the second case, the protagonist challenges the opponent to show that the protagonist's standpoint is wrong (special variant of argumentum ad ignorantiam) or that the opposite standpoint is right.
Rule 3 can be violated -- at all stages -- by the protagonist or the antagonist in a "mixed" discussion (in which both parties have a standpoint to defend) by imputing a fictitious standpoint to the other party or distorting the other party's standpoint (straw man). The first effect can be achieved by emphatically advancing the opposite as one's own standpoint or by creating an imaginary opponent; the second by taking utterances out of context, oversimplification (ignoring nuances or qualifications), or exaggeration (absolutization or generalization).
Rule 4 can be violated -- at the argumentation stage -- by the protagonist in two ways: first, by putting forward argumentation that does not refer to the standpoint under discussion as advanced at the confrontation stage (irrelevant argumentation or ignoratio elenchi); second, by defending the standpoint using nonargumentative means of persuasion. Among the latter are playing on the emotions of the audience (special variant of argumentum ad populum) and parading one's own qualities (special variant of argumentum ad verecundiam). If the audience's positive or negative emotions (such as prejudice) are exploited, pathos replaces logos; for this reason, such violations of Rule 4 are sometimes called pathetic fallacies. If the protagonist attempts to get his standpoint accepted by the opponent just because of the authority he has in tne eyes of the audience because of his expertise, credibility, integrity, or other qualities, ethos replaces logos; for this reason, such violations of Rule 4 are sometimes called ethical fallacies.
Rule 5 can be violated -- at the argumentation stage -- by the protagonist by denying an unexpressed premise or by the antagonist by distorting an unexpressed premise. By denying an unexpressed premise ("I never said that"), the protagonist in effect tries to evade the responsibility assumed in his argumentation by denying his commitment to an unexpressed premise that is correctly reconstructed as such. The antagonist is guilty of the fallacy of distorting an unexpressed premise if he has produced a reconstruction of a protagonist's unexpressed premise that goes beyond the pragmatic optimum to which the protagonist can actually be held, given the verbal and nonverbal context of the discussion.
Rule 6 can be violated -- at the argumentation stage -- by the protagonist by falsely presenting something as a common starting point or by the antagonist by denying a premise representing a common starting point. By falsely presenting something as a common starting point, the protagonist tries to evade the burden of proof; the techniques used for this purpose include falsely presenting a premise as self-evident, wrapping up a proposition in a presupposition of a question (many questions), hiding away a premise in an unexpressed premise, and advancing argumentation that amounts to the same thing as the standpoint (petitio principii, also called begging the question or circular reasoning). By denying a premise representing a common starting point, the antagonist in fact denies the protagonist the opportunity of defending his standpoint ex concessis, which is a denial of a conditio sine qua non for all successful argumentation.
Rule 7 can be violated -- at the argumentation stage -- by the protagonist by relying on an inappropriate argumentation scheme or using an appropriate argumentation scheme incorrectly. The various violations can be classified according to the three main categories of argumentation schemes: symptomatic, similarity, and instrumental argumentation. Symptomatic argumentation is being used incorrectly if, for instance, a standpoint is presented as right because an authority says it is right (special variant of argumentum ad verecundiam) or because everybody thinks it is right (populistic variant of argumentum ad populum and, as such, also a special variant of argumentum ad verecundiam), or if a standpoint is a generalization based upon observations that are not representative or not sufficient (hasty generalization or secundum quid). Similarity argumentation is being used incorrectly, if, for instance, in making an analogy, the conditions for a correct comparison are not fulfilled (false analogy). And, finally, instrumental (or causal) argumentation is being used incorrectly if, for instance, a descriptive standpoint is being rejected because of its undesired consequences (argumentum ad consequentiam); a cause-effect relation is inferred from the mere observation that two events take place one after the other (post hoc ergo propter hoc); or it is unjustifiably suggested that by taking a proposed course of action one will be going from bad to worse (slippery slope).12
Rule 8 can be violated -- at the argumentation stage -- by the protagonist in a variety of ways. Some logical invalidities occur with a certain regularity and are often not immediately recognized. Among them are violations that have to do with confusing a necessary condition with a sufficient condition (or vice versa) in arguments with an "If . . . , then . . ." premise (affirming the consequent, denying the antecedent); other violations amount to erroneously attributing a (relative or structure-dependent) property of a whole to its constituent parts or vice versa (fallacies of division and composition).
Rule 9 can be violated -- at the closing stage -- by the protagonist by concluding that a standpoint is true just because it has been successfully defended (making an absolute of the success of the defense) or by the antagonist by concluding from the fact that it has not been proved that something is the case, that it is not the case, or from the fact that something has not been proved not to be the case, that it is the case (making an absolute of the failure of the defense or special variant of argumentum ad ignorantiam). In making an absolute of the success of the defense, the protagonist makes a double mistake: first, he ascribes to the common starting points the unjustified status of established facts whose truth is beyond discussion; and, second, in doing so, he erroneously invests his successful defense with an objective rather than (inter)subjective status. In making an absolute of the failure of the defense, the antagonist makes a double mistake as well: first, he confuses the roles of antagonist and protagonist; and, second, he erroneously assumes that a discussion must always end in a victory for either a positive or a negative standpoint, so that not having the positive standpoint automatically means adopting the negative standpoint, and vice versa, thereby ignoring the possibility of entertaining a zero standpoint.
Rule 10 can be violated -- at all stages -- by the protagonist or the antagonist by misusing unclarity (fallacy of unclarity) or misusing ambiguity (fallacy of ambiguity, equivocation, amphiboly). Various sorts of unclarity can occur: unclarity resulting from the structuring of the text, from implicitness, from indefiniteness, from unfamiliarity, from vagueness, and so on. Again, there are various sorts of ambiguity: referential ambiguity, syntactic ambiguity, semantic ambiguity, and so on. The fallacy of ambiguity is closely related to the fallacy of unclarity; it can occur on its own but also in combination with other fallacies (such as the fallacies of composition and division).
Comparison with Other Approaches
This brief overview may suffice to show that the pragma-dialectical analysis of the traditional fallacies as violations of the rules for critical discussion is more systematic than the Standard Treatment. Instead of being given ad hoc explanations, all the fallacies fall under one or more of the rules for critical discussion. A comparison between the violations of the pragma-dialectical rules and the traditional categories also shows that the pragma-dialectical analysis is more refined. Fallacies that were only nominally lumped together are now either shown to have something in common or they are clearly distinguished, and genuinely related fallacies that were separated are brought together. Distinguishing two variants of the argumentum ad populum as a violation of Rule 4 or 7 makes, for instance, clear that these variants are, in fact, not of the same kind. Analyzing one particular variant of the argumentum ad verecundiam and one particular variant of the argumentum ad populum as a violation of the same Rule 7 makes clear that seen from the perspective of resolving a difference of opinion these variants are of the same kind.
The overview also reveals that the pragma-dialectical approach makes it possible to analyze thus far unrecognized and unnamed "new" obstacles to resolving a difference of opinion: declaring a standpoint sacrosanct (violation of Rule 1), evading the burden of proof by immunizing a standpoint against criticism (violation of Rule 2) or falsely presenting a premise as self-evident (violation of Rule 6), denying an unexpressed premise (violation of Rule 5), denying an accepted starting point (violation of Rule 6), falsely presenting something as a common starting point (violation of Rule 6), making an absolute of the success of the defense (violation of Rule 9), and so on.
Rather than considering the fallacies as belonging to an unstructured list of nominal categories, which happen to have come down to us from the past, as in the Standard Treatment, or considering all fallacies to be violations of one and the same (validity) norm, as in the logico-centric approaches, the pragma-dialectical approach differentiates a functional variety of norms. A whole series of norms other than just logical validity are taken into account, depending on the rule that has been violated. The logical norm of validity gets its proper (and limited) place in the argumentation stage.
The pragma-dialectic approach is both broader and at the same time more specific than the traditional logico-centric approaches, including the Standard Treatment. The scope is broader because at any stage of a critical discussion all possible violations of a discussion rule are treated as fallacies -- not just the "logical" errors. The greater specificity ensues from the fact that fallacies are systematically linked to the (non)resolution of a difference of opinion, which at the same time makes it possible to explain why something is a fallacy.
Analyzing fallacies systematically from the perspective of a well-defined theory of argumentation, set up as a set of rules, is not a unique feature of the pragma-dialectical approach but it is also promoted in the formal dialectics of Barth and Krabbe (1982).13 The first attempt to analyze a fallacy, ad hominem, within a formal-dialectical framework is reported in Barth and Martens (1977).14 According to Barth and Martens, a theory of rational argumentation must be consequently envisaged as a finite set of production rules for generating rational arguments. Each individual rule states a sufficient condition for the rationality of a generated argument: all (and only) arguments that can be generated by one or more of these rules are rational arguments. Fallacies can be analyzed -- -"unmasked" in the words of Barth and Martens -- as argumentative moves that can not be generated by the production rules (Barth and Martens 1977, 96).15
A first difference with the pragma-dialectical rules is that the latter are not formulated as production rules for generating rational arguments, but as conditions for resolving differences of opinion. Each pragma-dialectical rule refers to a necessary condition for resolving a difference of opinion. Only observing all the rules constitutes a sufficient condition. A second difference is that the pragma-dialectical rules are not primarily about the use of logical constants in a formal dialogue; they are rather about the performance of speech acts in the various stages of a critical discussion aimed at resolving a difference of opinion., All aspects of a critical discussion fall within the scope of the pragma-dialectical theory, not only the logical aspects. Owing to their more encompassing pragmatic orientation toward the performance of speech acts, the pragma-dialectical rules link up better with real-life argumentative discourse and are also better equipped for dealing adequately with the awkward informal fallacies.16
Where the peculiarities of ordinary argumentative discourse are, in the logico-approaches, either completely ignored or treated as "infirmities" of natural language, the pragma-dialectical approach takes all these phenomena, notably implicitness and indirectness, systematically into account.17 In principle, the rules for critical discussion provide all the norms that play a role in resolving a difference of opinion. The logico-centric approaches make use of just one norm: formal validity, in one sense or another. The pragma-dialectical rules amount to ten different norms that not only cover formal (invalidity but also many other things that can go wrong in argumentative discourse.18
[References below are to the Bibliography]
1. One of the most constructive reactions to Hamblin's devastating criticism of the Standard Treatment is provided by the works of Woods and Walton (1982a, 1989). Their remedy is to call o more sophisticated modern logics than just syllogistic, propositional, and predicate logic; every fallacy gets, so to speak, its own logic. Among the other constructive reactions are Finocchiaro (1981, 1987) and Hintikka (1987).
2. Biro and Siegel (1992) are among the most outstanding protagonists of a purely normative approach whereas C. A. Willard, Argumentation and the Social Grounds of Knowledge (Tuscaloosa: University of Alabama Press, 1983), and A Theory of Argumentation (Tuscaloosa: University of Alabama Press, 1989), is a prominent advocate of a purely descriptive approach.
3. For the notions of problem-validity and conventional validity, see E. M. Barth and E. C. W. Krabbe, From Axiom to Dialogue: A Philosophical Study of Logics and Argumentation (Berlin: De Gruyter, 1982), 21-22; and for a similar distinction R. Crawshay-Williams, Methods and Criteria of Reasoning: An Inquiry into the Structure of Controversy (London: Routledge and Kegan Paul, 1957), 175.
4. The pragmatic insight we are referring to is primarily gained from J. L. Austin, How to Do Things with Words (Cambridge: Harvard University Press, 1962); and J. R. Searle, Speech Acts: An Essay in the Philosophy of Language (Cambridge: Cambridge University Press, 1969) and Expression and Meaning: Studies in the Theory of Speech Acts (Cambridge: Cambridge University Press, 1979); and H. P. Grice, "Logic and conversation," in Syntax and Semantics, vol. 3, Speech Acts, edited by P. Cole and J. L. Morgan (New York: Academic Press, 1975), 43-58. Among the providers of fundamental dialectical insight are Crawshay-Williams, Methods and Criteria, and Barth and Krabbe, From Axiom to Dialogue; also K. R. Popper, Objective Knowledge: An Evolutionary Approach (Oxford: Clarendon Press, 1972), and Conjectures and Refutations: The Growth of Scientific Knowledge (London: Routlcdge and Kegan Paul, 1974).
5. For a more elaborate exposition, see van Eemeren and Grootendorst (1984, 7-18).
6. For a critical-rationalist philosophy of reasonableness, see both works of Popper referred to in note 4.
7. In van Eemeren and Grootendorst (1984), a full exposition has been given of the pragma-dialectical version of these rules. There, the rules that are constitutive for a critical discussion are stated in terms of the speech acts to be performed by the parties who are engaged in the resolution process. It is worth noting that observance of the rules can only constitute a sufficient condition for resolving a difference of opinion in conjunction with the fulfillment of the appropriate "higher-order conditions" pertaining to the attitudes and dispositions of the discussants and the circumstances of discussion. See F. H. van Eemeren, R. Grootendorst, S. Jacobs, and S. Jackson, Reconstructing Argumentative Discourse (Tuscaloosa: University of Alabama Press, 1993).
8. The pragma-dialectical identification of fallacies is in fact always conditional: only if it is a correct interpretation that the discourse is aimed at resolving a difference of opinion can it be maintained that a fallacy has occurred.
9. For the transformations of deletion, addition, substitution, and permutation carried out in a pragma-dialectical reconstruction, see van Eemeren, Grootendorst, Jackson and Jacobs, Recomtructing Argumentative Discourse, chap. 4.
10. At this juncture, insight from conversation and discourse analysis can be beneficial. The empirical research augmenting the analyst's intuitions so as to go beyond a naive reading of the discourse may vary from quantitative measuring to qualitative studies. See van Eemeren, Grootendorst, Jackson and Jacobs, Reconstructing Argumentative Discourse.
11. See F. H. van Eemeren, "For Reason's Sake: Maximal Argumentative Analysis of Discourse," in F. H. van Eemeren, R. Grootendorst, J. A. Blair, and C. A. Willard (1987a, 201-16).
12. For a discussion of the three main categories of argumentation schemes, see van Eemeren and Grootendorst (1992a, 94-102).
13. In several aspects, the pragma-dialectical approach links up with formal dialectics. Maintaining the term dialectics points to the agreement in general objectives, replacing formal by pragma (from pragmatic) to the differences in orientation. See note 3 above.
H. In fact, in Barth and Martens (1977), ad hominem is not treated as a fallacy but as an admissible discussion move "ex concessis."
15. Barth and Martens do not undertake any attempt to analyze the various fallacious variants of ad hominem. This is no wonder: Lorenzen's dialogue rules on which their analysis is based relate, after all, only to the use of logical constants such as "if . . . then . . . ," "and," "or," "not," "all," and "none."
16. The crucial difference between the opponent's concessions in a format dialogue and the protagonist's arguments in a critical discussion shows how much more distant formal dialectics is from argumentation in ordinary discourse. See van Eemeren and Grootendorst (1984, 15-15).
17. Interpretation problems in the literature on fallacies are generally ignored or trivialized. Furthermore, no distinction is made between norms defining the various types of fallacy and criteria for deciding whether a certain verbal move is indeed to be regarded as a perpetration of a particular type of fallacy.
18. For a more detailed illustration of our position, see our treatment of the argumentutn ad hominem in Part III of this volume.