This Critical Notice appeared in Dialogue (1991), Vol. 30, pp. 123-128.

A Guide for the Disputatious


Informal Logic: A Handbook

for Critical Argumentation

Cambridge: Cambridge University Press, 1989

Pp. xiii + 292.

NORMAN SWARTZ  Simon Fraser University

It is obvious that teaching and research in informal logic (critical thinking, rhetoric) is a growth industry, indeed one of burgeoning proportions. Judging by skyrocketing enrolments and the avalanche of books, journals, computer aids, conferences, and workshops during the last decade, it would seem to be no passing fad but the overdue recognition of a societal need.

     Among writers in this field, Douglas Walton is one of the most prolific: Informal Logic is his sixth book in this area since 1982.[Note 1] But exactly what genre of book this latest one may be, and what needs and audience it addresses, pose a considerable puzzle.

     Clearly Informal Logic is not a textbook: there are no exercises for students, no effective rules to be learned and applied, etc. Its subtitle, A Handbook for Critical Argumentation, suggests that this is a reference manual, a concise compendium that we might pull down from a shelf to look up a classification, rule, technique, or forgotten formula. The subtitle is misleading. I have never before come across a 'Handbook' which so little deserved its name. Unlike true handbooks, this book is obviously intended by its author to be read through, from cover to cover, not to be used as a mere reference tool.

     The most distinctive feature of this book is its insistence upon considering arguments as instances of dialectic — as exchanges of information, as objections, and as concessions and probings, engaged in by disputants. In this regard, Informal Logic departs widely from many familiar texts in that it presents few instances of 'static' arguments, e.g. passages drawn from the writings of philosophers, social scientists, economists and historians. It even shuns the stock-in-trade of many contemporary informal logic texts, viz. newspaper and television advertisements. Instead, Walton focusses on the cut-and-thrust of bona fide dialogue, and spends more time than is usual developing such themes as the norms and attitudes of argument, what constitutes a fair retort, and how to probe an opponent's position, both as a means to further the debate and as a means to draw out strengths and weaknesses. Argument, on this account, emerges more like a game of chess to be fought than as an artifact to be dispassionately picked to pieces. Even without his mentioning it explicitly, the analogy with chess is pursued to the point of identifying various 'stages' of argumentation: opening, confrontation, argumentation, and closing (10-11). The first two of these taken together correspond, it would appear, to the opening game in chess, and the latter two, to the middle, and end game, respectively. Plainly, Walton is trying to correct an imbalance he perceives in much of the traditional analysis of argument. But in his attempt to tilt the balance, he has written a book which is as lopsided, but in the opposite direction, as the books he would fault. There is virtually nothing in Informal Logic concerning techniques for assessing arguments outside the context of dialogue. The reader is given little help in grappling with written arguments (monologues) where there is no opportunity to interrogate the proponent. Debaters and courtroom lawyers will find more help in this book than will critical readers and writers.

     In places, Walton unmistakably aims his discussion at seasoned philosophers. His insights can be astute and instructive, as, for example, in his examinations of ad hominem argumentation and evaluating testimony of expert witnesses. To each of these studies he devotes a complete chapter – Chapters 6 and 7, respectively – far more than is usual in current textbooks. Walton takes great care not to condemn all ad hominem arguments as fallacies. In certain instances, on certain occasions, it is perfectly proper, he argues, to 'attack' one's opponents, if, for example, there is knowledge that one's opponents have a sufficient vested interest so as to make them less than totally honest in argument.

     Indeed, the chief worth of this book lies precisely in Walton's urging our being more subtle in judging our opponents' side than we are counseled in run-of-the-mill texts. Time and again, in a variety of contexts, Walton shows how a given manner of argumentation, historically condemned as being one sort or another of classical fallacy, may not really be fallacious at all. With examples such as Walton's in hand, one can begin to appreciate why many of the so-called fallacies are so seductive: these manners of arguing (e.g. ad ignorantiam 43-49) often do have appropriate occasions. True, they may be – and often are - used fallaciously; but they are not intrinsically fallacious. The fallacy arises only in misusing such arguments.

     His discussions of 'subtle equivocations' and the variability of strictness of standards (§9.8 and §9.9) are equally laudable and sophisticated.

     In other chapters, however, Walton seems to address utter novices in argumentation, and the quality of his exposition suffers badly. Although the nature of the subject matter does not lend itself to treatment by recipes, in their stead is offered a surplus of advice, much of which can be characterized only as avuncular, moralistic, hectoring, or insipid: e.g. "A participant should not try to opt out [of an argument] just because things do not seem to be going his[Note 2] way" (11), or, "the ignorant answerer should be able [i.e. permitted] to admit his[Note 3] ignorance" (43).

     In Chapter 5, again seemingly for the benefit of the inexperienced, Walton makes a half-hearted attempt to elucidate the concept of deductive validity. But in explicating this concept, Walton exhibits the same ambivalence characteristic of too many other writers in his vacillating between a semantic conception (114) and various syntactic accounts (e.g., 131). And he can be terribly careless. Eschewing symbolism, he stumbles into an egregious ambiguity: "We could interpret the conditional [If A, then B] as meaning that it is not true that A is true and B is not true" (110). Persons familiar with logic will know that Walton means "~(A & ~B)". But the very fact that Walton is even making this point, and deliberately avoiding symbolism, shows he is trying to explain this concept for the neophyte. Unfortunately, the neophyte is almost certain to interpret Walton's English sentence as saying "~A &~B". There is something to be said for well-chosen symbols, and Walton ought not to be so diffident about using them.

     It is not just the avoidance of formal resources, nor the two-mindedness of his explication, which spoils his account of deductive validity. The order of presentation is the reverse of what pedagogy demands: the concept of deductive validity is appealed to (47, 61, 89) long before Walton attempts (108-133) to explicate it. And the flaws continue to mount up. He invokes the technical concepts of most specific form (122) and implication (126), but nowhere defines these notions. His account of inconsistency sounds almost Hegelian in allowing for the actual instantiation of inconsistency: "Inconsistency can be very difficult to deal with. Psychologists know that, when laboratory animals are subjected to inconsistent treatment, they can start to exhibit frustration and loss of interest in activity" (126). Although such a concept is sanctioned and perfectly familiar in ordinary speech, it cannot be used coherently, in spite of Walton's attempt, to explicate the modal dimension of the concept of deductive validity. For the latter purpose, a concept of logical inconsistency is required which cannot, under any possible circumstances, be instantiated in a state of affairs. And, lastly, although he reviews a few principles of valid inference, e.g. Modus Ponens and Hypothetical Syllogism, his choices appear arbitrary. He does not, for example, mention the common principle of Simplification (And-elimination) or the common use of reductio in ordinary argument. To be sure, Walton makes no claim that his set of rules is complete; but his very silence may mislead the unsophisticated student into believing that this book delivers more than it does in fact.

     A book in informal logic ought to exemplify the virtues it extols: clarity of thought, fairness to one's opponents, etc. Yet the very first example (2), which allegedly illustrates the difference between the semantic and pragmatic aspects of the context of argument, seems not to concern an argument at all. The example is of a ship captain who, out of uniform, finding a cigarette butt on the deck, asks a new sailor: "… who the hell owns this damned thing?" The reply: "I'd say you do, mate. You found it." While one can agree with Walton's diagnosis that this latter is an inappropriate response given the pragmatics of the context, one must, nonetheless, if one wants to preserve one's pre-analytic intuitions, find it bizarre to characterize such an exchange as an 'argument.' For illustrating an argument, Walton could hardly have chosen a worse example. Or, again, in his discussion of the complex question (e.g. "Has Beethoven's Twelfth Symphony been found?"), Walton commends the obvious technique of dividing the question. But he then falls victim to an equivocation[Note 4], invoking the parliamentary rule of 'dividing the question', saying that dividing the question is not always permitted in every legislature (34-35). But what is called 'dividing the question' in a legislature is not the making explicit of a tacit presupposition of a question, but the splitting of a conjunctive motion, e.g. of the sort, "Be it moved (1) that parcel post rates be increased by 5%, and (2) that first class rates be increased by 8%." Walton has wandered far from the target.

     One case study, which is examined twice, is of a father, who himself smokes, telling his son that he must not smoke[Note 5] (127, 141-143). Walton leans strongly toward describing this as a situation where the parent's words are inconsistent with what his actions express, and thus those actions undercut the cogency of the argument. But in fairness, before one leaps to such a verdict, one ought at least to consider another analysis. In such cases, it is not unreasonable to ask whether the father might not genuinely believe his own advice, even wishing that he himself could follow it, but finds himself quite unable to do so. It is not, after all, unusual for persons in perfect sincerity to offer good advice which they, for one reason or another (e.g. perhaps because of chemical addiction) are themselves incapable of following. Of this latter, reasonable, alternative analysis, Walton does not breathe a word.

     Analyzing the fallacy of false cause is de rigueur in every book of informal logic. But one reads the relevant sections (§§8.5-8.10) in this book at one's peril: they are riddled with errors. [1a] Walton obliterates the distinction between individual events and events of the same type. He begins §8.5 by writing, "The traditional post hoc fallacy is said to be the unjustified argument which concludes that one event causes another event simply because there is a positive correlation between the two events" (212). [1b] The error is compounded when he further blurs the distinction between events and types of events (236), i.e. he treats individual events, events of the same type, and the same type of events as if they were all of the same ontological category. [2] As we have just seen, Walton uses the concept of correlation in an idiosyncratic, unexplicated fashion, at wide variance from the similarly-named (and well-defined) concept in statistics. For Walton allows that 'correlation' can obtain between two individual events (212, 213, 216, 229, etc.).[Note 6] [3] Besides "correlation," he uses other standard mathematical vocabulary in eccentric ways: "recursive" (219) and "nonlinear" (222). [4] He argues that 'positive correlation' (in his peculiar sense) may be indicative of causal connectedness (213, 230), overlooking that a 'negative correlation' (in the standard mathematical sense) can be equally indicative of causal connectedness. [5] And although Walton states explicitly that he "will not try to offer an analysis of the causal relation" (199), he certainly does offer one particular analysis, the Manipulability thesis, without, however, admitting to doing so (199, 218, 228 [twice], and 236). Naive readers are here having a particular philosophical thesis foisted on them unawares. The Manipulability Thesis is, of course, a highly controversial one, and a book such as this aspires to be, ought not to be promoting hotly disputed theses as if they were settled matters.[Note 7]

     One frequent theme in the book (despite the paucity of references in the Index) is the contrast between plausible and probable reasoning (e.g. 15-16, 192, 196, 228). On this topic, Walton relies on the earlier work of Rescher.[Note 8] But while Rescher explicitly acknowledges that plausible reasoning "represents a more basic — or, if you insist, more primitive — level of analysis [than probability theory]" (Rescher, 17), that admission is a far cry from, actually in distinct opposition to, Walton's characterization of plausible reasoning as "intrinsically weaker [than probabilistic reasoning]" (15; see also 192.) The relative efficacies of the two modes of reasoning are not fixed. Rescher has tried to show how, in some instances, plausibilistic arguments are superior to probabilistic ones (Rescher, Chapter 4). To catch the authentic flavor of the theory of plausible reasoning, its nuances and its prospects, one will have to turn to Walton's original source, viz. to Rescher's book.

     By the time I had finished reading the first sentence of Chapter 1, I was experiencing déjà vu. This sense of familiarity did not grow out of having read the same material in other books, but from the annoying recognition that the sentence was a virtual restatement of the opening sentence of the Preface, four pages earlier. This inauspicious beginning was to presage one of the most considerable defects of this book, viz. its inordinate repetitiveness, much of it irritatingly banal: "judgment is needed to evaluate whether an answer is relevant, if it seems to be heading in the right direction" (42) and "great care and judgment are needed in determining the important presuppositions of some tricky or objectionable questions" (55).

     In short, this book swings dizzyingly from one extreme of merit to the opposite. There is little in it that can be called mediocre. Parodying Longfellow, we might say of Informal Logic, "where it is good, it is very, very good; where it is bad it is horrid." All told, there is enough solid material in Walton's latest addition to his ever-expanding corpus to constitute a good book of half its present length.


  1. His previous books are Argument: The Logic of the Fallacies, co-authored with John Woods, (McGraw-Hill Ryerson, 1982); Topical Relevance in Argumentation (Benjamins, 1982); Logical Dialogue-Games and Fallacies (University Press of America, 1984); Arguer's Position: A Pragmatic Study of Ad hominem Attack, Criticism, Refutation, and Fallacy (Greenwood, 1985); and Informal Fallacies (Benjamins, 1987).[ Resume ]
  2. Neither Walton nor the editor at Cambridge University Press seems to have cared that the masculine pronoun is used virtually throughout.[ Resume ]
  3. See footnote 2.[ Resume ]
  4. Not unlike the one he himself warns against in the case of "begging the question" (245-250).[ Resume ]
  5. Again see footnote 2. Surely it would have compromised no stylistic principles to have made this a case of mixed-sex or of mother and daughter.[ Resume ]
  6. On the standard definition of "correlation" (i.e. of "correlation coefficient"), correlation is undefined for a single pair of numerical values (e.g. ⟨2, 7⟩) and must have a value of either 1 or -1 for any two pairs of values (e.g. ⟨2, 7⟩ and ⟨4, -6⟩). Correlation coefficients are significant only for n pairs of numbers, where n > 2.[ Resume ]
  7. A minor, sixth, point: one sentence is turned inside out. "If B [sic] tends to be followed by A [sic], even in different circumstances after many trials, the claim that A causes B is made stronger" (232).[ Resume ]
  8. Nicholas Rescher, Plausible Reasoning, Assen, Netherlands: Van Gorcum, 1976.Resume ]

Return/transfer to Norman Swartz's Philosophical Notes

Return/transfer to Norman Swartz's Home Page