INFORMAL LOGIC XlVI, Winter 1992 The Dialectic of Second-Order Distinctions: The Structure of Arguments about Fallacies DAVID GOODWIN Key Words: Informal logic; argumentation history and theory; rhetorical theory; dialectic. Abstract: Arguments about fallacies generally at- tempt to distinguish real from apparent modes of argumentation and reasoning. To examine the structure of these arguments, this paper develops a theory of dialectical distinction. First, it explores the connection between Nicholas Rescher's con- cept of distinction as a "dialectical countermove" and Chaim Perelman and L. Olbrecht-Tyteca's "dissociation of ideas." Next, it applies a theory of distinction to Aristotle's extended arguments about fallacies in De Sophisticis Elenchis, prima- rily with a view to analyzing its underlying strate- gies of argumentation. Finally, it examines how second-order distinctions (those designed to chal- lenge previously formulated distinctions) under- pin current arguments against the Aristotelian or "Standard Treatment" of the fallacies. This paper is not about fallacies per se. By this I mean, it will not offer any claims about the nature of fallacies, nor speculate about whether they exist. It will not ad- vance a theory of fallacies, nor, equally, make judgments about their use (or abuse) in everyday reasoning. Indeed, it will not even attempt to differentiate the various types and functions of fallacies. Other pa- pers, monographs, and books have done so already, and many of these, I think, with great success. Rather than join in the various debates about fallacies-about what they are or how they work-this paper develops a meta- analysis of the debate: it examines the dia- lectical countermoves characteristic of ancient and contemporary arguments about fallacies. It discusses controvertible claims that philosophers have asserted about fallacies and the kind of evidence Unirersity of Waterloo and reasoning used to support these claims. Specifically, it analyzes the struc- ture and function of one countermove common to all such arguments about fallacies-distinction. To accuse an oppo- nent of fallacious reasoning is to argue that his or her argument is reasoning in appear- ance only, that it fails, in some important respect, to meet the criteria of "real" argu- mentation. Hence all arguments about fal- lacies proceed from, and frequently argue for, a primary distinction between appear- ance and reality as it pertains to the form, conduct, and function of argumentation. Moreover, all arguments about fallacies reveal a "dialectical reflexivity," as I call it, which may be their most intriguing, if not characteristic, feature. From Aristotle's De Sophisticis Elenchis to the works of con- temporary informal logicians, all argu- ments about fallacies are arguments about argumentation. How this reflexivity in- forms the dialectical strategies of such treatises is the main concern of this paper. Specifically, it asks: what roles do distinc- tions play in dialectical exchanges about fallacies? How do distinctions differentiate "real" from "apparent" modes of reason- ing? In what way do they determine the outcome of arguments about argumenta- tion? Are there different kinds of dialecti- cal distinctions? And finally, can these differences account for shifts in models of argumentation, and, in particular, shifts between "traditional" and "contemporary" arguments about fallacies? To answer these questions, the first section of this paper examines the role that distinction plays in dialectic generally. It develops a theory of distinction by outlining the relationship between Nicholas Rescher's concept of "dialectical countermove" and Chaim Perelman and L. OIbrechts- Tyteca's "dissociation of ideas." The sec- ond section applies this theory to Aristotle's Sophistici Elenchi in order to analyze its underlying strategies of argu- mentation. And finally, the third section examines how second-order distinctions (distinctions designed to undermine the probative force of previously formulated distinctions) play an instrumental role in the construction of current challenges to the Aristotelian, or "Standard Treatment," of the fallacies. I. Dialectic, Distinction, and the Dissociative Construction of the Real In Dialectics: A Controversy-Oriented Approach to the Theory of Knowledge, Nicholas Rescher defines "dialectic" as any form of "disputation, debate, and rational controversy" which "exhibits epistemological processes at work in a setting of socially conditioned interac- tions" (xii). These "epistemological processes," as he calls them, require partic- ipants to present evidence that is either plausible enough to shift the burden of proof away from. or garner presumption for, their respective assertions. Dialectic, therefore, is a process which shapes, and is shaped by, socially-constructed knowledge, knowledge which allows communities to establish and test plausi- bilities, and therefore to set standards and make judgements in human affairs. To achieve this goal, dialectic proceeds by successive motions of exchange: "assertions" (initial statements or claims), "denials" (challenges to such claims), "provisoed assertions and denials" (the presentation of grounds which support- that is, provide prima facie evidence for-an assertion or denial), and finally, "distinctions" (the presentation of an evidential ground that limits the probative force of a proviso) (5-17). As defined by Rescher, a distinction is a dialectical countermove which challenges a provisoed assertion or denial. Take. for example, the following provisoed asser- tion: "I know this is a human hand (P) because my senses tell me so directly" (Q) (16). Schematized by Rescher as P:Q and read as "P because of Q," this proviso can be challenged by the distinction, "you can't really support this claim CP) be- cause, although your senses provide evi- dence (Q), this evidence is quite often deceptive" (R). The distinction, schema- tized as -P:Q&R and read as "not P be- cause although Q yet R." limits the grounds established by the initial proviso (Q) by qualifying the status of these grounds. Put simply. it provides an over- riding exception to be considered: namely, that the senses are open to deception. Such a distinction adds a new set of evidential considerations (R) to the dispute, one which is meant to diminish the probative strength of the opponent's proviso. A distinction at- tempts to accomplish this by transferring the opponent's primafacie case to another, more hostile epistemological context, one that militates against, rather than provides support for, an assertion or denial. Extending Rescher's discussion, then, a distinction, I would argue, challenges an opponent's dialectical grounds by (1) pro- viding a new ground for consideration which (2) separates the old ground from the initial assertion by (3) splitting the old ground into two elements, one of which it concedes and the other it denies. In the ex- ample above, R introduces a new concept (that the senses can be deceived) in order to break the bond holding Q to P. the provi- so to its assertion. A distinction accom- plishes this by a strategy of "yes but," that is, by conceding that "yes, your senses pro- vide evidence" and then by denying "but this evidence cannot support the initial claim." The pivot, or "but," of the distinc- tion, then, is its most powerful feature. To function as a dialectical countermove, a distinction must grant more evidential weight to what is denied than to what is conceded. This, of course, is how it garners presumption for -P. For without the pivot, the distinction could not tip the scales of plausibility against the opponent's claim taken as a Whole, and instead, would only serve as a neutral division of the issue under discussion. In Rescher's account of dialectic, pre- sumption favours the plausible, and the plausible, in tum, favours evidence which conforms to the desiderata of cognitive systematization-simplicity, completeness, directness, regularity, uniformity, distinct- ness, normalcy, and the like--or what Rescher calls "thesis-warranting princi- pies" (39). Distinctions invoke a thesis- warranting principle to advance the cause of the newly-stated position, and to make it seem more epistemically desirable than antecedent positions. In the above exam- ple, the provisoed assertion appeals to "directness," that is, to an immediate sensory perception of the body. The dis- tinction, on the other hand, appeals to "per_ ceptual consistency"--or perhaps more accurately, "inconsistency"-by pointing out that the senses often generate false patterns of expectation, as, they do, for instance, when an oar in water appears to be, but is not in fact, broken. The disputant making this distinction, then, not only adds a new consideration (the deceptiveness of the senses) to the argument, but advances the cause of a new thesis-warranting prin- ciple (perceptual consistency), one that is more pertinent to, and true of, the issue under discussion than that advanced by the opponent. Distinctions tip the scales of plausibili- ty because they construct or derive from previously constructed concepts of the "real." The "real" both shapes and is shaped by dialectical processes. It stands for whatever a community designates to be the most true, meaningful, relevant, or nor- mal considerations touching upon the issue Dialectic of Second-Order Distinctions 13 under discussion. In The New Rhetoric: A Treatise on Argumentation, Chaim Perelman and Lucie Olbrechts-Tyteca discuss such distinctions under the heading of "dissociation" (411-50). Dissociation is the argumentative technique of reconfigur- ing previously constructed integrities. pri- marily by showing them to be unstable, illusionary, or in some way deficient. The real is created by dissociation, which not only breaks connecting links, but which involves a more profound change that is al- ways prompted by the desire to remove an incompatibility arising out of the confron- tation of one proposition with others, whether one is dealing with norms, facts. or truths, (413) Dissociations, then, postulate a coherent, self-consistent realm of being and a "rule of reality" against which conflicting claims can be judged. Dissociations resolve im- passes created by incompatible claims or principles, primarily by showing that one of the two claims only appears to be plau- sible, relevant, useful, and so on, "The ef- fect of determining reality," Perelman says, "is to dissociate those appearances that are deceptive from those that corre- spond to reality" (416), Dissociations generate what Perelman and Olbrechts-Tyteca call "philosophic pairs," ranked dichotomies that distinguish realities from appearances. Every philo- h' , Term I I' 'If' h h sop IC paIr, Term II ' a Igns ltse WIt t e primary dissociation appearance , Term I reality represents whatever is deemed by the dis- putant to be apparent, illusory, insubstan- tial, irrelevant, insignificant, false, erroneous. Term II, on the other hand, represents whatever is actual, substantial, relevant, coherent, true. The real is a construct: Term II provides a criterion, a norm which allows us to distinguish those aspects of term I which are of value from those which are not; it is not simply a datum, it is a con· struction which, during the dissociation of 14 David Goodwin term I, establishes a rule that makes it pos- sible to classify the multiple aspects of term I in a hierarchy. (416) The real, then, functions to distinguish and classify those aspects of appearance which conform to, or conflict with, previ- ously established guides for resolving incompatibilities. Applied to our example above, the distinction implicitly generates a series of philosophic pairs: 1 2 3 4 Q sensory directness concede appearance - = > > --,----- R non-sensory consistency deny reality As represented by 1, the distinction ad- vances a new ground, R, which serves as a measure of the real-at least pro tern-against which Q is found wanting. Specifically, at 2, R introduces the philo- sophic pair-sensory versus non-sensory evidence-aligning the first with Term I and the second with Term II. The ranking at 2 is derived from another distinction, 3, which ranks the principle underlying the appeal to the senses (namely, direct- ness) as inferior to the newly postulated principle of perceptual consistency. At 4, the distinction concedes the appearance of proof established by the opponent's provi- so (based on sensory evidence and the principle of directness), but then denies the reality of this proof in favour of the newly distinguished grounds it puts forward for consideration (based on non- sensory evidence and the principle of perceptual consistency). In many cases, distinctions employ pre-existing philo- sophic pairs created by longstanding or ongoing argumentation to generate disso- ciations (420-26). In other cases, the dialectical exchange itself introduces new pairs. If adopted by a discourse community, new pairs become part of the construction of the real, and in this way, may serve to reallocate presump- tion and burden of proof in future controversies. In summary, then, a distinction is a dia- lectical countermove that challenges a provisoed assertion or denial by introduc- ing a new, more plausible ground (R) which qualifies-and by qualifying, diminishes-the epistemic force of the op- ponent's claim. To accomplish this goal, the new ground introduces a dissociation which differentiates appearance from reali- ty, and which allows the disputant to con- cede the appearance of proof in the opponent's argument (the "yes") but then deny the underlying reality of such proof (the ·'but"). Finally, the dissociation of re- ality inherent in distinctions shapes the di- alectical process by generating and consolidating plausibilities. In this way, distinctions help determine which of two positions will garner presumption at any given point in a disputation. II. The Dialectic of First-Order Distinctions: Aristotle's Arguments about Fallacies Aristotle's "De Sophisticis Elenchis- or, indeed, any of the works inspired by it, such as the medieval treatises on sophis- mata and obUgationes-reveals a complex system of distinctions, all of which serve to delimit licit and illicit argumentative procedures. Fallacies, Aristotle tells us, frequently arise from confused thinking, and especially, the inability to make, or the unwillingness to grant, distinctions (1 69a23ff]. Verbal equivocation and am- biguous phrasing, for instance, can lead unsuspecting disputants into the abyss of self-refuting paradoxes if they fail to exer- cise their right to draw a distinction: If one does not distinguish (OWPWt:'tj1E VO£,) the meanings of a doubt- ful term, it is not clear whether he has been confuted or not, and since the right to draw distinctions (OzE AE tv) is conceded in ar- guments, it is obvious that to grant the question simply, without making distinc- tions (olOp(auvm), is a mistake .... It frequently happens, however, that, though people see the ambiguity, they hesitate to make the distinction (alalPEO faOal), be- cause of the numerous occasions on which people propose subjects of this kind, in or- der to avoid seeming to be acting perverse- ly all the time. Then, again, though people would never have thought that the argu- ment would hinge upon this point, they are often confronted with a paradox. So, since the right to draw a distinction (alalPEO fv) is conceded, we must not hesitate to use it, as was said before. (I75b 27-38) Disputants have "the right to draw distinc- tions" and thus to avoid falling into seem- ingly self-contradictory claims, but they frequently fail to exercise this right in or- der "to avoid seeming to be acting per- versely all the time." Derived from the Greek expression for conversation, "dia- lectic" requires disputants to follow the rules of "turn-taking" that regulate conver- sations generally, and which permit the ex- change of ideas and the equitable distribution of opportunities to speak and reply in argumentation. Those disputants who insist on making distinctions at every turn, and who admit little or nothing "for the sake of argument," may feel they are acting perversely, that is, against these rules of social interaction. Sophistical refu- tations. then, force disputants to sustain their ethos at the expense of logos, preying on misplaced goodwill and habits of socialization. Distinction allows disputants to differ- entiate relevant from irrelevant, true from false, real from apparent considerations in matters of reasoning. The fallacy of com- position and division, for instance, requires disputants to distinguish those complex ex- pressions which signify a compound from those which do not; the fallacy of accent, to distinguish among the "supra segmental phonemes" that signify different mean- ings; the fallacy of figure of speech, the different categories that a word may be- long to; of accident and consequent, the symmetrical and asymmetrical natures of relation; of secundum quid et simpliciter, Dialectic of Second-Order Distinctions IS the absolute and relative predication of a term; of petitio principii, the premises and conclusions of an argument; of "non-cause as cause," the relevant and irrelevant ele- ments of a proof; and of "several questions at once," the kinds of questions which can and cannot be adequately answered by single affirmation or negation. Distinctions block a sophistical refuta- tion by stipulating the conditions under which one disputant will assent to a claim formulated by another. That is, distinctions clarify what can be accepted and what re- jected in the argument of the other dispu- tant. For example, if someone should argue for the following paradox-"the same man can be seated and walking" (166a2Sff)- Aristotle advises that we distinguish the various meanings of "can." Does it mean "has the power, potential, or desire to walk" (while sitting)? Or does it mean "has the ability to do both activities (walking and sitting) at the same time"? Similarly, if someone should argue that "something can be white and not-white because an Indian is black and has white teeth," we should distinguish between the absolute and rela- tive use of the predicates "black" and "white" (l67a8ff). Does "black," for in- stance, apply "in every respect" or just "in some respects"? And finally, if someone should argue "I know that it has rained be- cause the ground is wet," we must distin- guish the cause from the effect, ensuring that the first has not been inferred from the second (167b6ff). Is the ground wet for other reasons? Was someone watering the garden, for instance? Using Rescher's notation system, we can schematize Aristotle's examples of fallacious arguments as the following provisoed assertions: I. "the same man can both walk and not walk at the same time (P) because he can walk while sitting" (Q). Hence P:Q, read as "P because of Q." 2. "something can be white and not-white (P), since an Indian is black and has white teeth (Q)"; hence P:Q. 16 David Goodwin 3. "I know that it has rained (P) because drenched earth accompanies rain" (Q); hence P:Q. The first is an instance of the fallacy of equivocation; the second, the fallacy se- cundum quid et simpliciter; the third, the fallacy of affirming the consequent. To counter these fallacious provisoed asser- tions, Aristotle recommends introducing the following distinctions to the argument: I. "you can't support this claim CP) be- cause, although you argue that a man can both walk and not walk (Q), you have not distinguished between two conflicting senses of 'can' (R)"; hence -P:Q&R, read as "not P because, although Q, yet R" 2. "you can't support this claim (-P) be- cause, although the Indian is black and white (Q), the predicate, 'black: is being used both absolutely (black in every re- spect) and relatively (black in one respect only) at the same time (R)"; hence -P:(Q&R). 3. "you can't support this claim CP) be- cause, although drenched earth accompa- nies rain (Q), the relationship between an antecedent and a consequent is not the same as that between a consequent and its antecedent. i.e., the relationship is not con- vertible (RY'; hence -P:Q&R Each distinction, then, adds evidential grounds (R) to a provisoed assertion which limits or diminishes the probative strength of the opponent's prima facie case. And each distinction limits the initial proviso (Q) by invoking an overriding exception to be considered: in the first case, that "can" denotes both potentiality and actuality, namely, both the desire to do, and the per- formance of, two contrary activities; in the second, that a word cannot be predicated of a term both absolutely and relatively at the same time; and in the third case, that the relationship between antecedent and consequent is neither symmetrical nor transitive. In the three examples above, (R) divides the grounding proviso (Q) into two aspects, one of which it denies, the other it concedes. In the first example the dispu- tant concedes: "that a man may desire to walk while sitting." But denies: "that the man can perform both tasks at the same time." In the second example, the disputant coneedes: "that the Indian is black and white." But denies: "that the Indian is black in every respect and is not black in every respect." And in the third example, the disputant concedes: "that drenched earth accompanies rain." But denies: "that the relationship between antecedents and consequents is necessarily convertible." By employing distinctions, disputants concede that many features of argumentation-the ambiguity of lan- guage, the predication of terms, the princi- ples of causality and validity, and so on-are easily overlooked or confused or omitted from consideration, but deny that such oversights or confusions or omissions reflect the "reality" of argumentation. That is, by differentiating antecedent from con- sequent or relative from absolute. dispu- tants accomplish two tasks: first, they introduce a qualification, R, which under- mines the ability of the fallacious ground, Q, to support the initial claim, P; second, they demonstrate their expertise in matters of reasoning, since, according to Aristotle, the ability to distinguish real from appar- ent modes of reasoning is the sign of a true dialectician. Aristotle's motives for writing this treatise are clear. During the fifth and fourth centuries B.C., students could learn from certain teachers, or sophists (croqJLcrrm), how to use paradoxes or sophisms (croqJfcrJ.uxra) as refutations (e-AEJXOl) of established doctrines or be- liefs. Although Aristotle discusses dialec- tical strategies in Topics book VIII, erroneous reasoning in Analytics II, and fallacies in Rhetoric book II chapter 24, De Sophisticis Elenchi is the only extant trea- tise which attempts, on a large scale, to outline procedures for challenging, in turn, the sophistic refutations of philosophic thought. The treatise differentiates, then, between two kinds of disputants: those who accept the force of an argument such as "a man can be black and not black" and those who do not. The first group fail to distinguish between relative and absolute predication, and, equally, between real and apparent dialectical exchanges. By accept- ing the system of distinctions constituting Aristotle's treatise, however, the second group demonstrates a real knowledge of argumentation. For example, the ability to distinguish between absolute and relative predication generates the dissociation absolute/relative not distinguished h' h . absolute/relative distinguished ,w IC , m turn, aligns itself with the dissociation sophi stical reasoning Aristotelian dialectic . In other words, true dialecticians know and can make use of distinctions proper to the conduct and study of argumentation. As Sten Ebbesen points out, dialectical exchanges about dialectic determine the very basis of reasoning, and thus determine the paradigms or general theories govern- ing the conduct and outcomes of intellec- tual debate and problem solving generally: A standard procedure for proving the supe- riority of a new theory to an old one is to show that 'unreasonable arguments' known to be constructible and not soluble within the framework of the old theory present an obstacle no longer; either because they cannot be constructed at all within the new framework, some general principle pre- venting this; or because they need not, in the new theory, be considered unreasona- ble. And a standard procedure for attacking the new theory consists in demonstrating either that it does not neutralize all the old paralogisms or that it gives rise to some new ones, (1) In other words, the distinctions attending an argument about argumentation generate dissociations which serve to regulate what constitutes the structures, purposes, and outcome of reasoning itself. One of the pri- mary dialectical functions of Aristotle's treatise, I would argue, is to align the "un- reasonable" (i.e., paradoxical) arguments of the sophists with mere appearance, Dialectic of Second-Order Distinctions 17 showing them to be solvable within his new philosophic framework of distinction, so that they "present an obstacle no long- er." In short, Aristotle's treatise outlines a system of real distinctions: real, because they reflect the truth of reasoning, namely, truths such as "the antecedent to conse- quent relationship is not symmetrical"; real, too, because they serve as plausibili- ties against which the sophistical grounds of reasoning can be found wanting. In short, his treatise is an extended argument in defense of his concepts of reasoning. As such, it seeks at every point to dissociate sophistical reasoning from his own, more "real" reasoning, and equally, to dissociate sophistical refutations from his own refuta- tions of them, III. The Dialectic of Second-Order Distinctions: Current Arguments against "The Standard Treatment" So far I have discussed only the dialec- tic of first -order distinctions, namely, those which challenge a provisoed assertion or denial. As Rescher points out, however, distinctions can also challenge other dis- tinctions (15). These complex distinctions I call "second-order." An example of a sec- ond-order distinction in action might be: opp. CP) "You can't really support your claim (that you know that 'this is a hand'), because, although your senses provide evi- dence (Q), this evidence can be deceptive" (R), Hence -P:Q&R (first-order distinction), prop. (P) I can support my original claim because, although my senses have provided evidence (Q) that has misled me in the past (R), those past cases differ from this present case in some significant respect (S). Hence, P:(Q&R)&S (second-order distinction). The proponent uses (S) to restrict the op- ponent's ground of distinction (R), by showing that claims about the deceptive- ness of the senses may be irrelevent to the issue at hand. That is, the proponent uses a 18 David Goodwin second-order distinction to concede that the first-order distinction (R) appears to support its claim (that the senses can de- ceive) but denies its reality on the strength of a new claim (that past cases differ from the present in some significant respect). Equally, S introduces a new thesis-war- ranting principle-the principle of con- formity (that like cases should be treated alike, different cases should be treated differently)-which displaces the principle of perceptual consistency introduced by R. A first-order distinction introduces a qualifier to restrict an opponent's proviso, and thus signals a greater complexity of re- lation than that found in the provisoed assertion/denial. Similarly, the presence of a second-order distinction signals yet greater complexity, since S qualifies and restricts R by introducing new considera- tions to the debate. Put simply, a second- order distinction "yes buts" a "yes but," conceding the appearance but denying the plausiblity of the evidence introduced by the opponent's first-order distinction. This complexity, in tum, signals progressively higher levels of expertise, primarily be- cause, by definition, experts know the divi- sions constituting their discipline and are capable of making precise discriminations. In the above example, disputants use dis- tinctions to demonstrate their familiarity with the concepts, information, and argu- ments relevant to the issue being dis- cussed: epistemology. In the first instance, the opponent employs a first-order distinc- tion to demonstrate knowledge about the unreliability of the senses and the need for perceptual consistency. In the second in- stance, the proponent attempts to regain ground by introducing a sophisticated second-order distinction, one that demon- strates an even greater epistemic knowledge by acknowledging that the reliability of the senses must be judged on the strengths of the evidence surrounding each case. The expertise that Aristotle claims for dialecticians in general, and for himself in particular, derives from his ability to draw first-order distinctions. According to Aristotle, the dialectician knows "the various sources of apparent refutations- apparent, that is, not to everyone but only to a certain kind of mind" (170b5ff). This "certain kind of mind" is the mind expert in argumentation, the one that can distin- guish "the various ways in which, on the basis of common principles, a real or apparent refutation ... is brought about." Certainly, Aristotle believed himself to be the first expert in dialectical reasoning, and his "system" of dialectic to be the most "adequate" and "complete" account of rea- soning yet postulated. For whereas his predecessors, "paid teachers of contentious argument," trained students in a "rapid but unsystematic" manner, he claims to impart an art-that is, true expertise-to a student (l83b35). Aristotle's self-professed exper- tise in reasoning rests, as he himself says, on his ability to distinguish systematically between true and false reasoning. And it is on the strength of his carefully worked out system of distinctions about argumenta- tion that he successfully argues against sophistic reasening, showing it to be reasoning in appearance only. Contemporary informal logicians and dialecticians, however, make claims to ex- pertise based on their ability to draw second-order distinctions. Where Aristotle introduces first-order distinctions to challenge the paradoxical provisoed assertions/denials of the Sophists, contem- porary scholarship introduces second-order distinctions to challenge the Aristotelian formula, or what C. L. Hamblin calls the "Standard Treatment" of the fallacies. By discriminating among the complex circum- stances and factors attending claims of fal- lacious reasoning, this scholarship attempts to reclaim many forms of argument. It at- tempts to show, therefore, that modes of reasoning once considered fallacious by traditional standards, only appear to be so, and that, in reality, they may be relevant, informative, probable, valid, epistemically interesting, predictive, or useful. For example, we find J. E. Broyles re- claiming the fallacies of composition; D. Gerber, the argumentum ad hominem; T. Govier, the slippery slope argument; L. Groarke, the argument that two wrongs make a right; J. Woods and D. Walton, the ad verecundiam; J. Barker, the fallacy of begging the question; D. Van de Vate, Jr., the appeal to force, just to name a few. Each of these scholars engages in argu- ment about argumentation. And each argu- ment, in turn, attempts to garner presumption by introducing second-order distinctions. By virtue of these new dis- tinctions, informal logicians claim to posit more comprehensive, complex, representa- tive, or "real" grounds for understanding and evaluating the argumentative process than those of their opponents. Conversely, they attempt to shift the burden of proof against positions that are derived from, or that provide support for, earlier and more "simplistic" treatments of fallacies. Since the publication ofe. L. Hamblin's Fallacies, informal logicians have com- plained that the Standard Treatment of the Fallacies, derived in part from Aristotle's Sophistici Elenchi and other related hand- books of argumentation, is reductive, inad- equate, and extremely simplistic. As Walter Ulrich puts it, we should stop "arbi- trarily rejecting all arguments that have been traditionally labeled fallacies," and "develop standards for determining how much weight should be given to these ar- guments. Factors such as the nature of the decision being made, the nature of the con- flicting arguments, and the presence of other supporting arguments would be rele- vant to such calculations" (111). Similarly, S. Toulmin, R. Rieke, and A. Janik argue that "most disturbingly to some people, ar- guments that are fallacious in one context may turn out to be sound in another con- text. Therefore, we shall not be able to identify any intrinsically fallacious forms of arguing" (132). And finally, as J. Woods and D. Walton point out, "on closer inspec- tion, many of the examples [of fallacies] Dialectic of Second-Order Distinctions 19 turn out either to be arguments that are not fallacious at all, or arguments in which guidelines are so lacking that a non- arbitrary sorting of the correct from the fallacious cases seems highly problematic or impossible" (v.). Though recent reappraisals of fallacies are varied in their methods and aims, one aspect of their work remains constant: the use of second-order distinctions both to demonstrate the complexity surrounding claims about fallacies and to underscore the expertise of authors/disputants who make such distinctions. Because of the ltmitations of space, I plan to follow Aristotle's advice to rhetoricians. Rather than attempt a full induction, I will choose representative examples, ones that I hope will demonstrate the "distinctive" structure of arguments about fallacies. My examples are taken from two recent sources: D. Walton's Informal Logic: A Handbookfor Critical Argumentation and R. Grooten- dorst's "Some Fallacies about Fallacies." Both authors attempt to develop criteria for discriminating between fallacious and non-fallacious forms of argument from au- thority. The first author advances "six criti- cal questions" and the second, three rules of argumentative discourse to differentiate between real and apparent argumentation as well as between real and apparent accu- sations of fallacious reasoning. Specifical- ly, I will focus on their discussion of one fallacy-the ad verecundiam-in order to show how their discriminations function as an inventory of both first- and second- order distinctions about argument. Both Walton and Grootendorst believe that expert opinion can be a legitimate and helpful strategy for introducing evidence to a discussion, but that such arguments from authority must meet certain condi- tions lest they lapse into the ad verecundi- am fallacy. Walton sets out these conditions as "six critical questions [which] must be kept in mind when evalu- ating any successful appeal to authority" (194). According to Walton, arguments 20 David Goodwin from authority require the disputant to reflect on the following questions. Does the "judgment put forward by the authority actually fall within the field of competence in which that individual is an expert"? Is "the cited expert actually an expert, and not merely someone quoted because of his prestige, popularity, or celebrity status"? Is the expert's opinion authoritative? Is there "disagreement among several qualified authorities who have been consulted?" Is "objective evidence on the cited opinion presently available ... and is the expert's opinion consistent with it?" And finally, has the "expert's say-so been cor- rectly interpreted" (194-97)? Similarly, Grootendorst argues that argumentum ad verecundiam is a fallacy only insofar as a disputant violates at least one of three per- tinent rules of argumentative discourse. These rules include, first, "whoever ad- vances a standpoint is obliged to defend it if asked to do so"; second, "a standpoint can only be defended by advancing evi- dence which relates to this standpoint"; third, "a standpoint must be considered to be defended conclusively if arguments are used in which a commonly accepted argu- mentation schema is being applied in a correct way" (138-39). Any argument that fails to satisify the requirements posed by the questions or in- scribed in the rules can be deemed falla- cious. Conversely, any argument that does meet these conditions can defend itself, if need be, against such accusations. In this way, Walton's questions and Grootendorst's rules serve as inventories for both first- and second-order distinctions. Insofar as they can be invoked to challenge an opponent's provisoed assertion, the rules and ques- tions serve as inventories for first-order distinctions. For example, a disputant might offer the following provisoed asser- tion: "I don't need to argue this further (P) because I can assure you that it really is as I say (Q)." To which a respondent, in turn, might argue that this claim is fallacious us- ing Grootendorst's first rule: namely, your statement is dubious CP) because, al- though you have offered your own person- al guarantee of truth (Q), yet you have "advanced a standpoint without defending it" (R). In the same way, the two other rules and all six of Walton's questions can serve as R in a first-order distinction chal- lenging an assertion grounded in an appeal to authority. Insofar as a disputant can invoke Grootendorst's rules or Walton's questions to challenge an opponent's first-order distinction-that is, to challenge a previ- ously formulated accusation of fallacious reasoning-they serve as inventories for second-order distinctions. Again, a dispu- tant could respond to the argument above by using Grootendorst's first rule of dis- course: "my claim is not dubious (P) be- causc. although I support the truth of my claim based on my own authority (Q&R), you have not, until now, asked me to de- fend it on any other ground (S)." In other words, because the first rule uses the for- mulation "is obliged to defend it if asked to do so," the disputant can distinguish be- tween refusing to defend a standpoint and not being asked to do so. As Grootendorst points out, the importance of the first rule is to keep the discussion from ending be- fore it has a real opportunity to develop. In this case, however, the respondent argues that his or her argument is not fallacious since he or she is willing to provide further grounds for making his claim. To the ex- tent that these new grounds meet the re- quirements of rule three, and thus fit into a schema acceptable to the opponent, the ar- gument will not be fallacious. Of course, Grootendorst's other two rules or any of Walton's six questions could be used as the ground (S) of a sec- ond-order distinction defending an argu- ment from a charge of fallaciousness. For example, a disputant could argue from Walton's rules that "my argument is not fallacious (P) because, although I base my evidence on the authority of an expert (Q) and that authority disagrees with other authorities (R), yet this expert's assertions accord with objective evidence" (S), or P:(Q&R)&S. In this case the disputant uses question five (objective corroboration) to undermine an objection based on rule four (expert agreement). The division between one rule or question and another provides the grounds of second-order distinction. "Yes, my opponent seems to have grounds for a first-order distinction (one that char- acterizes my argument as fallacious) but, in reality, these grounds fail to take into ac- count some relevant, important, pertinent, and thus real aspect of argumentation (in- scribed in another rule/question governing discursive reasoning). By arguing in this manner, disputants employ a second-order distinction to show that they understand the divisions consti- tuting the study and practice of argumenta- tion. In short, they introduce more complex, and more finely discriminating grounds to the debate, showing their oppo- nent's argument, by contrast, to be more simplistic and less plausible. By doing so, of course, the disputants accomplish two things. First, they attempt to determine the real as it pertains to argumentation, and therefore, to shift presumption in favour of their own position and against their oppo- nent's. And second, they establish their own expertise in the very area of endeav- our they are presently engaged in: argu- mentation. The issue of complexity and expertise becomes more pressing as an ar- gument about fallacies moves through suc- cessive orders of distinction. The higher the order, the more reflexive the grounds are for disputation. For while first-order distinctions discriminate between the real and apparent in matters of reasoning, second-order distinctions discriminate between real and apparent discriminations in matters of reasoning, and so on. At each point, the introduction of new divisions-based on rules, questions, spec- ified conditions, etc.-provides progres- sively more reflexive grounds for disputants to hold their positions, since Dialectic of Second-Order Distinctions 21 each division engages disputants in mak- ing discriminations about argument in order to argue effectively. IV. Conclusion: The Future of Argumentation Clearly, Walton's book and Grooten- dorst's article are, first and foremost, argu- ments about argumentation, about what it is and how it functions. Specifically, these works are extended arguments which at- tempt to distinguish real from apparent ac- cusations of fallacious reasoning. As arguments, these works attempt to garner presumption and levy burden of proof by introducing distinctions whose divisions of the subject matter-argumentation-are more complex, comprehensive, plausible, and thus more real than those grounding the arguments of their opponents. Both intro- duce rules or questions to generate divisions in order to challenge the simplistic and over- ly reductive nature of arguments about fal- lacies based on the "Standard Treatment." Not surprisingly, ongoing debates among contemporary informal logicians provide an impetus for fostering a new, third-order level of distinctions. For exam- ple, Grootendorst argues that fallacies are neither "buttercups" (the ontological enti- ties of the Standard Treatment) nor "irreg- ular verbs" (the relativistic entities of Woods and Walton), pointing out that "in the Standard Treatment or its liberal vari- ants no distinction is drawn between the different consequences [of breaking the rules]." In other words, he distinguishes among the various consequences of break- ing the rules, and, by doing so, seeks to distinguish his own "dialectical approach" from those adopted by his predecessors. "Distinction," according to Rescher, "is the most characterstic and creative of dia- lectical moves" (12). Rescher's point aptly describes the nature of dialectical exchanges about fallacies. In such exchanges, distinc- tion is the most characteristic and creative 22 David Goodwin move: the most characteristic, because all such arguments use distinctions to distin- guish the real from the apparent as it per- tains to argumentation; the most creative, because, as a discipline, informal logic gen- erates complex discriminations about rea- soning, and thus introduces more expert and higher orders of distinction to gamer presumption for an ongoing succession of conflicting positions about argumentation. Given the progression from first to second, and perhaps second- to third-order distinc- tions, the history of argumentation may well be the history of the distinctions made by its various disputants. I suspect that the trend will continue, with distinction re- maining the engine of reflexivity driving informal logic. I suspect, too, that Nicholas Rescher's concept of distinction as a dialectical move, when conjoined with Chaim Perelman's and Lucie Olbrechts-Tyteca's concept of argumentative dissociation, can provide an important tool for examining, if not fine-tuning, this "engine." For as a theory of distinction emerges, it will re- mind informal logicians of the dialectical nature of their own discipline: that they always reason about reasoning and argue about argumentation. Bibliography Aristotle, On Sophistical Refutations. Trans. E. S. Forster. (Cambridge: Harvard University Press) 1965. Barker, John A. "The Fallacy of Begging the Question." Dialogue, 15 (April 1976) 241-55. Broyles, James E. "The Fallacies of Composition and Division." Philosophy and Rhetoric, 8 (Spring 1975) 108-13. Ebbesen, Sten. Commentators and Commentaries on Aristotle's Sophistici Elenchi: A Study of Post-Aristotelian Ancient and Medieval Writ- ings on Fallacies. (Leiden: E. J. Brill) 1981. Gerber, D. "On Argumentum ad Hominem." Personalist, 55 (1975) 23-29. Govier, Trudy. "What's Wrong with Slippery Slope Arguments?" Canadian Journal of Philosophy, 12 (June 1982) 303-16. Grootendorst, Rob. "Some Fallacies about Falla- cies," in Argumentation: Across the Lines of Discipline. (Dordrecht: Foris Publications) 1987.331-42. Hamblin, C. L. Fallacies. (London: Methuen & Co., Ltd.) 1970. Perelman, Chaim and Olbrechts-Tyteca. L. The New Rhetoric: A Treatise on Argumentation. Trans. John Wilkinson and Purcell Weaver. (Notre Dame: Notre Dame University Press) 1971. Rescher, Nicholas. Dialectics: A Controversy- Oriented Approach to the Theory of Knowledge. (Albany: State University of New York Press) 1977. Toulmin, Stephen, Rieke, Richard and Janik, Allan. An Introduction to Reasoning. 2nd. ed. (New York: ~acmillan Publishing Co.) 1984. Ulrich, Walter. "In Defense of the Fallacy," in Argument and Social Practice: Proceedings of the Fourth SCAIAFA Conference on Argumentation. (Annandale: Speech Commu- nication Association) 1985. 110-26. Van der Vate, Jr. Dwight. "The Appeal to Force." Philosophy and Rhetoric, 8 (Winter 1975) 43-60. Walton, Douglas N. Informal Logic: A Handbook for Critical Argumentation. (Cambridge: Cambridge University Press) 1989. Woods, John and Walton, Douglas. Argument: The Logic of the Fallacies. (New York: Mac- millan Publishing Co.) 1980. DAVID GOODWIN DEPARTMENT OF ENGLISH UNIVERSITY OF WATERLOO WATERLOO, ONTARIO N2L 3Gl o