book review Alec Fisher THE LOGIC OF REAL ARGUMENTS Cambridge: Cambridge University Press, 1988. 190pp. ISBN 0-521-30849-6 (cloth), 0-521-31341-4 (paper). Review by MICHAEL MALONE and DA VID SHERRY, Northern Arizona University There is a certain artificiality which in- fects logic texts, formal and informal alike. Examples and problems are either created, pruned, or carefully selected to give stud- ents arguments which can be straightfor- wardly treated by methods proposed within the text. We have in mind cases like: or: or: All butlers and valets are both obsequious and dignified. Therefore all butlers are dignified. 1 Tax reform involves two different measures: it reduces tax rates, and it closes tax loopholes. It seems to me that tax reform is politically possible only if the reduction in rates comes first, and the loopholes are closed later. The reason is this. Until rates are reduced, the loopholes are valuable to interest groups, and many of these groups are powerful enough to block reform. 2 Thc process of uniting two or more adja- cent school districts ought to be commenc- ed at once. The number of teachers in de- mand would thus be reduced, while the rate of compensation might be increased without adding to the burdens of the people and thus the facilities for obtaining good instructors would be multiplied in a two-fold ratio.3 Our experience has been that facility with arguments like these pays few dividends in the face of extended, theoretical arguments encountered in course work. Informal Logic X.2, Spring 1988 We regularly require students to find arguments in various contexts. Far too often they have returned empty handed when they have gone to their own textbooks. One better-than-average student complained that his " Ancient Life" text contained no reasoning, though examination yielded a nice argument to the conclusion that the dinosaurs were warmblooded. There are a couple of explanations for these blinders. Students labor under the misconception that whatever they are learning in classes is either fact or opinion, neither of which needs argument. They are also most com- fortable finding arguments that look, smell, and feel like those in their logic text, which tend to be (student) newspaper editorials, excerpts from informative articles, and in- ventions of logic teachers. None of these give students a feel for sustained theoretical discourse, even though logic must do this to function as a foundation for education. Of course, teachers and texts continually re- mind students that logic has wide applica- tion; but to say it is not to inculcate it. Alec Fisher's The Logic of Real Arguments is the first text that concentrates on the sort of ex- tended passages students encounter in their courses. In our experience, it is the best at- tempt yet to bridge the gap between logic and the rest of education. Fisher's book features two things. First, there are long, difficult, frequently classical arguments. Second, there is a rigorous method for extraction and evaluation, viz., a standard diagramming technique, sup- plemented by the "assertibility question" and a new twist on suppositional reasoning. Students had much better success at identi- fying real arguments in texts, indicating that they were becoming aware of the depth and importance of argument in academic discourse. Late in the semester, students found interesting and substantive arguments 106 Malone & Sherry in, e.g., mathematics, economics, politics and jurisprudence; there was even an ex- planation why water behaves the way it does at various temperatures. The passages were generally long and challenging, although sometimes a little too challenging. Our ex- perience has been that students acclimate quickly to the rarefied air of theoretical reasoning found in Real Arguments. Merely dwelling at this level for several weeks strengthens confidence in the face of reason- ing better than anything else we have tried. The book also has its share of problems. Fisher seems to restrict the term' 'reason- ing" to persuasion (16), and to exclude ex- planation from the domain (18). This is too narrow. For one thing, often the same set of sentences could be an attempt to persuade or an explanation, depending on whether the audience knew in advance that the con- clusion was true. For another, there are cases which seem to be both explanations and attempts to persuade. At least it is not obvious that Caspar Weinberger's letter to the U.S.'s NATO allies, the centerpiece of Chapter 4, is an attempt to persuade the allies that we must modernize rather than an explanation of why we are modernizing. Since Fisher's diagramming and evaluation techniques work equally well for explana- tions that occur in theoretical discourse, it is puzzling that he restricts his domain so narrowly. Real Arguments is not a good book for the beginning teacher of informal logic. Students need to be comfortable with the diagramming technique before tackling Fisher's first real argument, an excerpt from Malthus's "Essay on Population." But Fisher presents the technique in less than ten pages, providing few examples. Our students required lots of illustrations and practice to catch on to diagramming. For a two week interim, we did not use the text; instead we presented our own explanatory notes and problem sheets. What would be only an inconvenience for someone familiar with a book like Thomas's Practical Reasoning would be a real problem for one who is learning the method while teaching it. In short, it would greatly improve Chapter 2 to add some exercises of the 3-5 sentence variety to ease the transition. There are additional problems, which become acute as one moves from artificial to real arguments: how does one sort out the logically relevant material and what does one make of the rest? At the outset students presume that the task of logic is to rearrange the sentences of a given passage into the scheme insisted upon by the textbook and instructor. The standard, 3-5 sentence argu- ment encourages this disposition, for students can generally work such problems without having to discriminate the relevant from the irrelevant. By contrast, Fisher's treatment of a real argument excludes most of the original text from the final analysis. (A rough count of the Weinberger passage, pp. 58-61, shows 17 of 47 original sentences appearing in the final analysis.) The capacity for recognizing the logically relevant strikes the beginner as an occult power, and Fisher owes some account of extra-logical devices in which argument is embedded, including, at least, discounts, parentheticals, qualifiers and setting the stage. 4 It would also be appropriate to discuss the technique of paraphrase in order to deal with stylistic devices, like rhetorical questions, figurative language, repetition, and summary. The need for such an account is directly proportional to the length of passages being considered, and Real Arguments is concomitantly less effective for not including such remarks under' 'The language of reasoning" (15 ff.). Real Arguments is typical in failing to discuss arguments about arguments. Although routine for philosophers, it is very difficult for students to recognize when an argument is not about someone's conclu- sion, but about the argument by which his conclusion is reached. Fisher does not discuss this distinction, yet several of his problems and examples require that one notice it readily. The appendix on elementary formal logic (140-155) includes some basic argu- ment patterns. Fisher is not opposed to for- malization, but in keeping with the spirit of informal logic, he analyzes without resor- ting to formal logic. Unfortunately, there are cases (6-7) in which he presumes that students already have a grasp of basic argu- ment forms, and others in which the analysis would proceed much more smoothly if they did. Exercise 3 (158-60) illustrates this criticism and the previous one. It concludes with' 'That argument either settles the ques- tion [of the morality of deterrence] very quickly or it does not settle it at all. I think it doesn't settle it. " The argument ends on a disjunctive syllogism with a suppressed premise that is the conclusion of the bulk of the argument. But most students could not see this elementary move without help. We recommend using a chapter to discuss arguments about arguments and an enrich- ment of "the language of reasoning" with remarks on basic forms. As Fisher notes, suppositional reason- ing is not discussed by most informal logic texts, although it is "elegant, powerful, and extremely common" (82). His treatment is welcome. It is neater and richer than Thomas's in Practical Reasoning, the on- ly other account of which we are aware, and its exercises are, unlike Thomas's, not ar- tificial (cf. Fisher, 1-2 and 4).5 There are other differences between Thomas's and Fisher's approaches to sup- positional reasoning. Thomas requires that suppositional reasoning be deductively valid, thereby holding it to a standard higher than that for other kinds of reasoning. Fisher treats suppositional reasoning on a par with its more assertive relatives, and this is clearly correct, given arguments like Mill's case "in favour of restricting to the narrowest compass the intervention of a public authority in the business of the com- munity. " quoted from Principles of Political Economy (71-3). Thomas encloses suppositional reasoning within a box to em- phasize its hypothetical nature. This creates confusion about suppositional arguments Review of Fisher 107 which also involve asserted reasons. 6 Fisher succeeds at integrating assertions into sup- positional reasoning by superscripting 'u' (for unasserted) and observing that a superscripted 'u' infects everything which follows from the statement which is superscripted. Not only is this strategy more convenient than Thomas's, it also connects suppositional reasoning with the distinction between asserted and merely proposed clauses (cf. 23-4). This distinction is crucial to determining which statements make up an argument. Fisher's use of the superscripted 'u' allows him to distinguish results of suppositional reasoning which are unconditionally asserted from those which are contingent upon other assumptions (89-90). Finally, Fisher encourages students to consider convenience in choosing be- tween analyzing a passage as a piece of sup- positional reasoning, on the one hand, or as a conditional, on the other. This attitude promotes understanding of conditional statements, the staple of argumentation. We have some complaints about Fisher's treatment of suppositional reason- ing, though they are mostly aimed at his decision to "develop technicalities ... no further here" (90). His technique is insuf- ficient for analyzing the very first illustra- tion in the book (1). The argument begins with several suppositions that lead, separate- ly, to a pair of conclusions which are con- tradictories. 7 These must be conjoined prior to conditionalization. 8 It's a trivial matter, but something students ought to see in the text, if only to preserve the instructor's credibility. In addition, to reach the con- clusion that the initial supposition is false, something needs to be done with the remain- ing suppositions, all of which are necessary to reach the contradiction. It's not clear what Fisher would have us do here. Are we forced, now, to assert what was originally supposed? All of our criticisms to this point could be remedied in an expanded revision, and, in any case, an experienced teacher will overcome them with supplementary 108 Malone & Sherry materials. Another problem is not so easi- ly assuaged. Fisher alleges his approach is distinctive in employing the Assertibility Question (AQ), which asks: What argument or evidence would justify me in asserting the conclusion C? (27) It is used at two stages in analyzing an argu- ment. First, when inference indicators and context fail to disclose what the author's argument is, we use AQ to construct the best argument we can, and then "see if the author asserts or clearly assumes these same claims" (129). Second, Fisher relativizes the test for validity to the standards of evidence appropriate to the particular argu- ment. AQ determines the appropriate standards. We question whether the first use of AQ is as helpful as Fisher claims. It is a third line of attack, to be employed after inference indicators and context (i.e., nearby statements) have been exhausted. In his analysis of Weinberger's letter to NATO (49 ff.), Fisher uses AQ immediately after the inference indicators give out. The analysis includes reasons which are more Fisher's than Weinberger's, whereas the se- cond line of attack yields an analysis more faithful and charitable to Weinberger's text. A difficult but important lesson of logic is fidelity: We continually force our students back to the text to keep them from putting words in an author's mouth-words which are generally less than charitable. AQ en- courages fast, loose and uncharitable analysis, especially in the hands of students not inclined to keep their noses to the text. Fisher's presentation of AQ is troublesome pedagogically. There are no exercises in applying AQ, and we found it difficult to create them. Our own difficul- ty may subside with practice, but students weren't readily able to answer AQ in a non- trivial way. They were much more creative in showing arguments invalid by counterex- amples, which, conveniently, point the way to suppressed premises. And the task of fin- ding explicit premises seemed better served by trying 'because' and 'therefore' be- tween candidates for reason and conclusion and then following common sense. These strategies also decrease the likelihood of creating an argument instead of extracting it. The relativized test for validity is perhaps as old as Aristotle, who reminds us that "precison is not to be sought for alike in all discussions" (Nicomachean Ethics, 1094b 13-4). Fisher asks us to reflect on the appropriate standards and de- mand only the appropriate evidence, whereas Thomas asks us to determine how likely a given counterexample would be and to ignore unlikely ones. Both have the ef- fect of substituting 'beyond a reasonable doubt' for 'deductively valid.' Again, our complaints about AQ may indicate nothing but lack of practice; they do not outweigh the virtue of the text. As an introduction to logic which prepares students to use the discipline in further studies, Real Arguments has no serious competitors. Notes I Irving Copi, Symbolic Logic, 5th ed. (New York: Macmillan & Co., 1979), p.77. 2 David Kelly, The Art of Reasoning (New York: W.W. Norton & Co., 1988), p. 94. The exercises and examples of ex- tended argument analysis in this text and the next are not nearly the caliber of those we will discuss below. 3 Stephen N. Thomas, Practical Reason- ing in Natural Language, 3rd ed. (Englewood Cliffs, Nl: Prentice Hall, 1986), p. 33. 4 We are aware of two texts which try to deal with this problem, but each is in- complete. (1) Robert Fogelin, Understanding Arguments, 2nd ed. (New York: Harcourt, Brace, 10vanich, 1982), 58 ff., does a "close analysis" of a long passage, but only one, and does not com- plete the analysis. (2) R. Rubin and C. Young, Formal Logic: A Model of English, (Mountain View, CA: Mayfield Publishing, 1989), ch. 2, do not discuss any long passages. 5 A notable exception is the suppositional argument borrowed from Thomas (Fisher, 89). This piece of reasoning takes four tortuous steps to make what the medievals termed an "immediate in- ference" from 'Only A's are B's' to 'Every B is an A'. 6 Such reasoning is very common in for- mal science. For instance, the proofthat there is no rational number whose square is two begins by supposing that there is a rational number whose square is two; Review of Fisher 109 it draws consequences of this supposi- tion in conjunction with basic laws of arithmetic (e.g., if p is even, then p = 2r, for some r) which are asserted, not supposed. 7 This strategy is also used in the proof that the square root of two is not rational. The renowned mathematician, Paul Erdos, observed that with the discovery of this proof in the 6th century, we became human. We think there is a great deal of insight in Erdos's whimsical comment and so place some importance on students' being able to grasp the stategy. 8 Also, the schematic diagram of RAA (94) is incomplete. It needs to indicate somehow that 'C' is absurd. 0