Book Reviews I Compes rendus 169 Critical Review Logic and Misery: Walton's Appeal to Pity HANs VILHELM HANSEN Appeal to Pity: Argumentum ad Misericordiam by Douglas Walton Albany: State University of New York Press, 1997.225 pp. ISBN 0-7914-3461-3 (hb) [US$53.50]; 0-7914-3462-1 (pb) [US$17.95]. Douglas Walton's book, Appeal to Pity: Argumentum ad Misericordiam,' is, in many ways, a paradigmatic essay on an issue in argumentation theory. It is a patient treatment of the complicated webs of concepts that entangle this kind of argument, both notions kindred to pity, such as sympathy, benevo- lence and mercy, and those associated with 'argument', like dialogue type, argument scheme and fallacy. The first chapter takes up a theme that has been dominant in Walton's work and shows his freely admitted debt to Hamblin: it is a review of the textbook treatments of the argumentum ad misericordiam. The next chapter seeks to take us further back in history than the twentieth century's slew of text books have done (as did Hamblin's second chapter), and it considers views on the use of pity and related concepts in Greek, Chris- tian and early modern thought. The third chapter is a study of the Jerry Lewis Telethons for Muscular Dystrophy and the issues they have inadvertently raised. In the fourth chapter is an exploration of what the structure, or schematic form, of the argumentum ad misericordiam might be. The next chapter is another case study: the Nayirah Case involves how the media and the viewing public were misled by a Kuwaiti citizen during the escalating tensions between United States and Iraq in 1990. Chapter 6 takes up a question anticipated by those who have followed Walton's prolific career: "When is the argumentum ad misericordiam a fallacy?" The concluding chapter introduces yet another case study-this time involving the patricide by the Menendez brothers in California-and gives general concluding remarks about Walton's method and its perspective on ad misericordiam arguments. Throughout the book the mo- tivating question is "to set out criteria that can be used to help a rational critic judge when an ad misericordiam argument is reasonable or not, in a given case" ( p. xiii). ©Informal Logic Vol. 20, No.2 (2000): pp.169-184. 170 Informal Logic We cannot pause here to repeat or add to Walton's survey of the text book literature on argumentum ad misericordiam. Suffice it to say that the popular texts are uneven, inconsistent, and not overly penetrating, and that Walton convincingly shows this. Appeals to sympathy and/or pity are older than logic (if logic is thought to have been fathered by Aristotle) and just as old as argument itself. In Plato's Apology, for example, Socrates-no stranger to paradox-arouses our sympathy by disdaining the use of pity in his own defence. After reviewing Aristotle, the Stoics and early Christian views of pity, Walton searches for the earliest signs of life in the natural history of the argumentum ad misericordiam as a distinct kind of argu- ment, as individual as Locke's argumenta ad verecundiam, ad ignorantiam, and ad hominem are. Walton locates the first mention of argumentum ad misericordiam in a 1929 logic text: J.E. Creighton's, An Introductory Logic (37); however, he traces the origin of the phrase ad misericordiam, as standing for a kind of argument, to the Edinburgh Re- view in 1824. We are also given a very vivid example of an ad misericor- diam from Wi1Iiam Thackeray's editorial correspondence in 1860 (38-40), a case we will review below. An occurrence of the phrase more likely to have been noticed by modem argumentation theorists, and which Walton might have mentioned, is in Mill's On Liberty of 1859. There the context is the discussion of the power of the public to suppress the free expression of certain opinions by stigmatising them. Mill observes that the only ones who are generally free from this public power are the moneyed classes. Those whose bread is already secured, and who desire no favours from men in power, or from bodies of men, or from the public, have nothing to fear from the open avowal of any opinions, but to be ill-thought of and ill-spoken of, and this it ought not to require a very heroic mould to enable them to bear. There is no room for any appeal ad misericordiam in behalf of such persons. 2 Here Mill seems to be using 'ad misericordiam' in a way that argu- mentation theorists have come to think of as the argumentum ad miseri- cordiam. Moreover, Mill appears to hold that ad misericordiam appeals are sometimes relevant appeals to an issue under discussion (although not in the present case). Such a view is in disagreement with the tradition which takes any appeal to emotion in arguments to invariably commit a fallacy of premise irrelevance. The second chapter, overall a mini-history of the concepts of pity and sympathy, ends with a review of the importance of sympathy as a virtue in the writings of eighteenth-century philosophers Hume, Smith and Schopenhauer. 3 Earlier, in the seventeenth century, Hobbes had important remarks about pity and compassion, as did Butler in the eighteenth century, but these contributions are not mentioned in this work. Book Reviews / Compes rendus 171 , One of the issues that perplexes the study of the argumentum ad mis- ericordiam is how we are to translate, and understand, the Latin. Many of the early logic texts understood it as "appeal to pity" (also the title that Walton has chosen for his book), but "pity is often seen as containing nega- tive connotations of condescension, and even contempt"(64). Hence, oth- ers have opted for "appeal to sympathy," since "sympathy is generally con- ceived as good or virtuous, from an ethical point ofview"(64). The objections to pity seem to be two. There is first what we might call the logical objection-Walton traces it back to the Stoics-that pity is a "sentiment or passion that is being exploited as a rhetorical device, an 'irra- tional impulse' that is substituted for and interferes with rational thinking" (68). On this ground, pity is an encumbrance to good judgement, the very kind of thing that a critical thinker must avoid. The second objection to pity is an ethical one. "No one should be pitied," goes the objection-or, at least, the ones most often pitied in the public eye should not be pitied-, "because it is morally wrong to pity someone." Why is it morally wrong to pity someone? This is one of the main ques- tions occupying the third chapter. A number of authors (Ben-Ze'ev and Mercer, among the modem commentators, for example) are quoted in sup- port of the view that "pitiers are putting themselves on a higher level than the pitied ... thereby robbing them of their human dignity" (71). It is also thought that one pities someone else only when one is unable to help them (69). I think there are some unnoticed consequences of this morally negative view of pity. First, if it is morally wrong for A to pity B, and B knows it, then it is morally wrong of B to ask A for pity, or for B to argue that A should do something for B (or someone else) on the grounds of pity. Thus, the person making an appeal "for pity's sake," and the one swayed by the resulting ad misericordiam argument, may be equally in the wrong: the petitioner for asking for something to be done from an immoral motive, the respondent by adopting an attitude of pity towards another person. Secondly, if"A pities B" implies "A cannot alleviate B's suffering," then an argument by B that A should, on the basis of pity, relieve B's suffering, would be a bad piece of practical reasoning on B's part (for the reason that B would be asking A to do what A cannot do). These observations are not considered by Walton. The ethical objection points to an ambivalence towards pity in our tradi- tion. We have been exhorted to pity the poor and the incurably ill, never thinking until recently that we might have been acting immorally. It must be, then, that 'pity' is not the right term to designate what it is that moves us when we do good acts for those who are in need and request our assist- ance. As a response to this problem, Walton takes us on a conceptual tour 172 Informal Logic of pity's cognate concepts: compassion, empathy, sympathy, and mercy. The details are many and I will not review them here. Suffice it to say that Walton ingeniously manages an analysis of pity which takes empathy as the elementary concept which can be augmented to become sympathy; sympa- thy in tum can be enlarged into compassion, and finally, given a certain twist -perhaps the inability of the argument's addressee to alleviate the other person's condition-the concept of compassion can be transformed into the concept of pity (73-4). Nevertheless, Walton continues to use 'appeal to pity' as the most convenient translation of ad misericordiam (101). Chapter 4 is titled, "The structure of the argument." Argumenta ad misericordiam are arguments that appeal to consequences, more particu- larly, negative consequences. Walton identifies four kinds of ad misericor- diam arguments, each kind an instance of one of four different argument schemes.4 An argument scheme is not an argument, and not a logical form of an argument, but a set of conditions (i) that an argument must meet in order to be an argument of a certain kind, (ii) that enables an argument to be used correctly in a context of dialogue to shift a burden of proof or pre- sumption, and (iii) that has a set of associated questions that should be answered whenever arguments satisfying the scheme are used (151-52). The four kinds of schemes that are used for ad misericordiam arguments are: the appeal for help scheme, the argument from distress scheme, the prac- tical reasoning scheme, and (later, in Chapter 6) the argument from plea- for-an-excuse scheme. In each of these schemes the appeal to sympathy is not "built in"; that is, there are no argument schemes definitive of the ad misericordiam. Appeals to sympathy are rather "grafted onto" (119) each kind of argument, thereby turning the argument into, e.g., an ad misericor- diam argument that appeals for help.There are, however, four different kinds of emotional appeals that fall under the general heading of ad miseri- cordiam: appeals to pity, sympathy, compassion and mercy. An argument that is of the kind defined by one of the four schemes and has one of the appeals to feeling added on to it, is an ad midericordiam argument. Looking ahead to Chapter 6, "When is it a fallacy?", Walton maintains that there are four basic sub-types of the fallacy: those that involve a twist- ing of the argument, those that exhibit a failure of relevance, those that exploit the timely impact of a pity-appeal, and those in which the pity-appeal is staged (161). Ifwe multiply all these possibilities together (the schemes, the appeals, the fallacious moves) we find that the ad misericordiam fal- lacy may exist in anyone of 64 different incantations. For example, it could be an appeal to help kind of argument with an appeal to pity grafted onto it, which becomes a fallacy by twisting the argument, or a plea-for-an excuse kind of argument which appeals to compassion which is a failure of rel- Book Reviews / Compes rendus 173 evance, etc. Perhaps there will be some overlap and redundancy in this taxonomy, so let us introduce a very generous allowance for a margin of error and halve the possibilities. We are then on much safer ground when we say that on Walton's view there are at least 32 distinguishable kinds of fallacious ad misericordiam argument. We haven't yet mentioned that there are at least six different kinds of dialogues (quarrel, critical discussion, de- liberation, negotiation, inquiry and information-seeking),5 and these, having different goals and standards may force us to extend the taxonomy even further. (A little math can be a frightening thing.) Let us pause to consider one of the schemes onto which an appeal to sympathy can be grafted, the argument from need-for-help (104). A 1. For all x andy,y ought to help x, if x is in a situation where x needs help, and y can help, and y's giving help would not be too costly for y. A2. x is in a situation where some action A by y would help x. A3. y can carry out A. A4. y's carrying out A would not be too costly for y-that is, the negative side effects would not be too great, as y sees it. AC. Therefore, y ought to carry out A.6 There are a number of things to notice about this argument scheme: (a) it indicates a deductively valid form of argument since the major premise (A 1) is a universally quantified conditional that connects premises A2-A4 to the con- clusion; 7 (b) the appeal to feeling or sympathy is not built into it at all; i.e., no one is asking for help; (c) the major premise is a moral principle, equivalent to "we ought to help those who need help, when we can, if it is not too much trouble" (is y to determine what will count as "too much trouble"?); (d) the 'ought' in the major premise and in the conclusion must have the same sense, on pain of equivocation. On the next page (105), Walton considers another of the schemes that may be used in connection with appeals to sympathy, the argument-from- distress scheme. B 1. Individual x is in distress (is suffering). B2. Ify brings about A, it will relieve or help relieve the distress. BC. Therefore, y ought to bring about A. Walton remarks (105) that this schema is a sub-type of the need-for- help scheme. We notice, however, that the sub-scheme differs from its parent-scheme in a number of ways. It doesn't talk about the cost to y, for example, and it doesn't have a principle containing an 'ought' among its premises. (Is there an is-ought problem here?) Perhaps, we are to under- stand that the elements of the second scheme are to be substituted for elements in the first scheme but we are not given any help to understand the relationship between the two schemes. At any rate, the' ought' in the con- 174 Informal Logic c1usion of the second scheme must surely be the moral one. We haven't room to reproduce the schemes for practical-reasoning arguments, or the plea-for-excuse argument. The former also has an 'ought' conclusion but it is normally taken to be the' ought' of rationality. The latter has an ought in its conclusion too, but it is not immediately clear what its sense is: one ought to be granted exception from a rule under certain circumstances in law, in morals, in education, in sports, etc. Even so, these are argument schemes to be conjoined with appeals to sympathy (or pity, or mercy, or compassion); that is, they are communications that one should do the thing indicated, or something like it, for the sake of someone else who is in an unhappy situa- tion. Will that not, in each case, transform the 'oughts' in the conclusions of arguments satisfying these schemes into moral oughts too, if they aren't so already? I think it will. I t surely does bear notice that a kind of argument widely recognised by argumentation theorists as stock-in-trade is essentially a moral argument. I can't think of any other kinds of argument (appeal to authority, analogy, modus ponens, slippery slope, straw man, etc.) of which this is true.s Walton doesn't remark upon this, as we might expect him to do. This is not to say that he is unaware of the moral aspects surrounding argumentum ad mis- ericordiam. In the opening chapter it is noticed that some of the logic books consider appeals to pity to be parts of moral arguments (8). In Chapter 2, sympathy is discussed as a moral concept. But these various allusions never go past being incidental remarks. In Chapter 6, under the sub-heading, "The logical leap from ethical premises," two defeasible conditionals are consid- ered: (P) If pity is (morally) a bad quality, then appeals to pity in argu- mentation should be generally condemned (logically) as fal- lacious. (S) If sympathy is (morally) a good quality, then appeals to sym- pathy in argumentation should generally be acceptable (logically) as reasonable kinds of arguments. Walton rejects these two conditionals on the basis that they are an over- simplification in the textbook literature (172, 197). But something stronger needs to be said. That moral sentiments give rise to logical properties is a thesis which runs counter not only to traditional moral theory, but more so to logical theory. Even when we take a view of logic which is much wider than that espoused by some formal logicians, as Walton and I both think we should, there is no moral quality that makes an argument logically good or bad, and there is no logical quality either that makes arguments good or bad as moral arguments. There is a related consideration, however, already mentioned. If arguments from pity are appeals to do what cannot be done, Book Reviews I Compes rend us 175 then any such argument is bound to be unsuccessful, and so a bad argu- ment. But what makes the argument bad in this case is not that it involves a moral quality, but that it is a futile argument (as are also many other argu- ments). Analogous considerations apply to (S), the other conditional: that sympathy is a good thing does not make an argument that turns on sympa- thy a logically good argument. 9 Here, then, are two theses. The first is that ad misericordiam argu- ments are moral arguments, the second is that moral goodness and logical goodness are distinct. We are left with this: if the argument is a moral argument and it is logically bad, then it is a bad argument. But suppose the moral argument is not logically bad, then what? It is not thereby a good moral argument. Unless an ad misericordiam argument happens to be logi- cally bad, its evaluation as an argument must be based on criteria that lie beyond those that may be given by informal logic. That is, unlike other types of arguments, the ad misericordiam's acceptability may depend upon ex- tra-logical (and also extra-dialectical, or extra-procedural) considerations. Chapter 6 takes up the question of when argumenta ad misericordiam are fallacies. In Walton's pragmatic view, being a fallacy is a matter of degree: at one end of the spectrum are minor mistakes, at the other end are fallacies. This is parallel to a strain in the fallacies tradition, thought impor- tant by Mill, which has insisted on distinguishing slips in reasoning and trans- parent errors from fallacies, those kinds of mistakes worthy of notice. 1o A similar distinction is implied by the traditional conception offallacy: there are invalidities and good-looking invalidities, fallacies being of the latter kind. Walton, however, wishes to distance his theory from the traditional one which he finds wanting because it is deductivist and psychological (164).Thus, he restricts fallacies neither to the range of deductive arguments nor to the scope of those which falsely wear the face of rightness. Put more fully, it can be quite difficult and tricky to judge whether an argument is just weak-a blunder or an insufficiently supported argument that could be improved or corrected by further argument-or whether it is fallacious. A fallacy is a more serious error than a blunder. It is a systematic, under- lying kind of error in an argument, or deceptive tactic used to try to persuade another party to accept the argument in a conversational ex- change. (163)11 From this passage we see that there are two kinds of fallacies for Walton (as there were for Aristotle, although he based his distinction on something else): a serious error in an argument and a serious error in something external to an argument, the use of deceptive tactics with the purpose of getting the argu- ment accepted. More precisely, we may formulate Walton's theory of fallacy as fol- lows. A fallacy is a 176 Informal Logic (a) conversational "move" (b) which is a serious error (c) committed by the person attempting to persuade the other person (d) in a particular dialogue, D, with a particular goal, G. D belongs to a general kind of dialogue, Ll, and D inherits certain standards, S, from Ll, which must be abided by in D in order to reach G, and (e) the persuader misuses, or attempts to misuse, an argument accord- ing to the standards of the argument scheme of which type it is, thereby resulting in a violation of S, or (f) the persuader uses a deceptive tactic in conjunction with an argu- ment in order to attempt to get the argument accepted by the other party to the dialogue, thereby reSUlting in a violation of S.12 If this is on the right track, then it would be useful to have an explicit statement of the rules that go with each kind of dialogue in which the falla- des can occur. Next we must try to identify what the serious errors are, especially as the theory applies to ad misericordiam arguments. "Fallacies," writes Walton, "are associated with cases where the argument is put forward in such a way, in a context of dialogue, where the asking of critical questions is blocked by some tactic or other"( 156; see also 158-9). The ad misericordiam is attached to one or another of four kinds of schemes (mentioned above), and when it is fallacious it prevents the asking of the critical questions that go with the scheme, and should be asked. In other words, the ad misericor- diam fallacy is the kind of argument that seeks to prevent its own negative evaluation. Writes Walton: "the argumentum ad misericordiam is an unu- sual and distinctive type of fallacy in that it is more complex than other fallacies typically are" (153). We might add: it is a very clever fallacy. The four tactics found to have been used in the ad misericordiam fal- lacy are: (i) the twisting of the argument ... , (ii) the failure of relevance; (iii) the exploitation of the timely impact of the appeal to pity, where evidence in the context is ignored; and (iv) the use of appeal to pity as a staged effect (161). We can now amend the above general conditions of fallacies and par- ticularize for the argumentum ad misericordiam. (e') the persuader misuses, or attempts to misuse, an argument ac- cording to the standards of the argument scheme of which type it is (either appeal for help, argument from distress, practical reason- ing, or plea for an excuse), thereby resulting in a violation of S, or (f) the persuader uses a deceptive tactic (argument twisting, relevance, timely impact, staged appeal) in conjunction with an argument (which is an instance of one of the four schemes mentioned in e'), Book Reviews / Compes rendus 177 in order to attempt to get the argument accepted by the other party to the dialogue, thereby resulting in a violation of S. *** Argumentation theory means to be the theory of a practical art. Theory will be of interest only in so far as it gives shape to method and the method in tum pays off in practice. It is only in the last five pages of this book that Walton turns to a discussion of his method, a method he has followed in the analyses of the cases reviewed. It is a three-step method consisting of 'identification, analysis and evaluation' (194), urging that evaluation should follow only on careful analysis. But I think that Walton has presented us with the machinery for a potentially more sophisticated method which would be a more detailed aid to argumentative discourse analysis. These seven steps might be followed: MI. Identify a passage as containing an ad misericordiam appeal, and classify the appeal. M2. Identify the type of dialogue, a, in use. M3. Identify the kind of argument scheme onto which the appeal has been grafted, and determine what the standards are for it in a-type dialogues. M4. Determine whether the use of the ad misericordiam appeal has blocked, or was intended to block, the asking of any of the ques- tions belonging to the argumentation scheme which is in use. MS. If the answer to M4 is affirmative, then identify which question(s) it is that has/have been prevented from being answered, and deem the appeal an ad misericordiam fallacy. M6. If the answer to M4 is negative, then examine the argument in light of the questions that accompany its scheme. If it does not pass review of the critical questions, then there is a fallacy in the ad misericordiam argument. M7. If the argument does pass review of the critical questions in M6, then there is no fallacy present in the ad misericordiam argumenta- tion. This is a sketch of the method-Walton would undoubtedly express it somewhat differently-but I think it is fair to the intelIectual geography of Appeal to Pity: it ties the method to the theory. Of course, it may not always be possible to move through the method step-by-step, but it should guide the empirical investigation of cases under examination. An interesting question may emerge by the end of the process of evaluation, however: what are the implications if the ad misericordiam argument is acceptable? Let us now consider some of the cases in more detail. The Nayirah case, having a chapter of its own plus substantial discus- sion in the final chapter, is the one that occupies Walton the most. In its most 178 Informal Logic skeletal form, the Nayirah case involves the following elements. It took place at the time when the United States was considering what actions to take in response to the Iraqi invasion of Kuwait in 1990. Nayirah gave testimony that Iraqi soldiers had removed infant babies from life-sustaining incubators in a Kuwait hospital, leaving them to die. Later it was found out that Nayirah was really a member of Kuwait's royal family, that she was the daughter of the Kuwaiti ambassador to the United States, and that what she had said to a committee of the American Congress, and which was ultimately widely reported in the press, was not true (or so it is now be- lieved). It was found that her testimony, and the timing of it to sway the American government in the direction of taking military action against Iraq, had been orchestrated by a very influential Washington-based public rela- tions firm. The story of the inhuman behaviour by the Iraqi soldiers, accompanied by Nayirah's tears, constitute an appeal to sympathy and compassion which is grafted onto an argument described both as an appeal for help, and argu- ment from distress (144). This takes place in a context of dialogue which is deliberation (160).Walton thinks that the use of emotion on this occasion qualifies as three of the ways of bringing about the ad misericordiam fal- lacy (161). What is worthy of censure is that a false statement was made (argument twisting) at a critical time (exploitation of timely impact), and that it was calculated to bring about the eventual outcome (staged effect). These three aspects of the emotional appeal will have had the effect of blocking the asking of the critical questions associated with the argumenta- tion schemes. Walton studies this case in great detail, looking at it from one per- spective and then another. He is at first tempted to declare it a fallacy because it was intentionally misleading argumentation (142, 147), but he is concerned that this is tantamount to saying it is a fallacy because it has a false premise, which is not a structural mistake (141). If the alleged facts about the mistreatment of babies in Iraq had been true, then, goes the rea- soning, the effect would not have been staged, and the use of the argument for timely impact would have been excusable. Walton concludes that "ifwe are to say that the ad misericordiam argument in the Nayirah case is falla- cious" (notice the 'if), our finding must rest in part the claim that there was "a false or unproven (and doubtful) account" given (186). But I think there is less cause for worry here than Walton thinks there is. His theory is that there is a fallacy if there is an attempt to deceive (see (e') and (f') above) as there clearly was in the Nayrah case. To doubt that it constitutes a fallacy is to dpubt his own criteria for what makes for a fallacy. On Walton's view, I think, had everything Nayirah said been true, but she believed it false, and said it for effect, it would have been a fallacy. Book Reviews / C ompes rendus 179 The Menendez brothers' defence (1993-4 and 1996) in the charge of murder rested on the claim that they had been sexually abused as children by their parents, and because of this they lived in fear that their lives were in danger (175). In the first trial, this call for sympathy based on the alleged abuses led to a hung-jury (no decision), in the second trial the appeal to sympathy was disallowed, and the brothers were found guilty and sentenced to life imprisonment. Walton's focus is on the first trial, and the use of the imperfect self-defence, i.e., that the Menendez brothers, although there was no immediate threat from the parents, "honestly believed their lives were endangered" (175),13 It is not immediately clear which context of dialogue we are to under- stand this argumentation as occurring in but since the business of a court, at least in the pre-sentencing stage of its proceedings, is supposed to be con- cerned with questions offact, we should hazard the guess that Walton thinks that we are here in the context of an inquiry kind of dialogue, or perhaps a critical discussion. We may safely infer that it is the plea-for-excuse kind of argument that is in use, since the Menendez brothers' defence was that, although they did kill their parents, they did so because of a perceived dan- ger to themselves. If the brothers had been sexually abused as children and because of that were in fear of their lives, this counts as relevant testimony within the legal system.But an appeal to feeling was exhibited in another way too in that the brothers gave emotional testimony telling of their fears, and this might count as a staged tactic, another of the appeals that can block the asking of the relevant critical questions. It is believed that this testimony had an impact on the jury, leading to the no-decision. Walton's evaluation of the case is that there is no fallacy committed, because the use of ad misericordiam arguments was allowed by the judge, and so it was within the "legal framework of evidence" (189). Here, I think, Walton is sticking to his theory. When the brothers were retried, the new judge, on legal grounds, disallowed the same appeal to sympathy. The ques- tion here is not one of whether a particular ad misericordiam argument is allowable in one context of dialogue and not in another, since both trials would be typologically identical dialogical settings, with the same circum- stances, the same evidence, and the same actors, albeit different judges and juries. Either the argument is a fallacy in both cases or in neither case. 14 Hence, at the very least, one of the judges made a mistake, but we non- specialists don't know which one, and so we don't know whether the ruling in the first case was correct. Ergo, we cannot use that ruling as a basis for deciding whether the use of the emotional appeal was fallacious. In other words, the question of whether there is a fallacious use of appeal to sympa- thy in the first trial is not answered by what the judge actually allowed, but by what the judge should or should not have allowed. 180 Informal Logic The Thackery case is the very last case discussed in the book, and Walton's analysis of it is insightful and will give us a means of addressing the student's plea case as well, since the two cases are "closely related" (193). Thackeray was an editor of a journal. He received a submission from a poor woman who had several dependants. She pleaded that he should ac- cept her poems for publication on the basis that the money she would earn would help relieve her family's suffering. Thackeray, the editor, is made uncomfortable by the letter, and Walton observes: [T]he editor is also a private citizen, a person who will naturally feel an obligation, or at least a pull to respond to cries for help, from someone who appears to be in distress. This is a fundamental kind of human relationship, and an appeal based on it can be emotionally strong. So the editor is caught in a conflict between two forces. On the one hand, he is an editor, who has the job of judging submissions to his magazine by certain criteria. On the other hand, he is a private person who is responsive to cries for help from persons in distress (192). Walton treats this as an instance of the deliberation kind of dialogue employing a need-for-help argumentation scheme and an appeal that twists the argument. The poor woman has committed a fallacy. Thackeray feels an internal conflict between his role as an editor and his humanitarian in- stincts. Walton remarks that the pleader 'puts the editor on the spot' (192). Similarly, of the student's plea, Walton observes that "the inappropriate ap- peal to sympathy puts the professor on the spot, making him appear to be impolite or unhelpful" (191). This observation, that the person to whom the ad misericordiam argument is directed is caused to feel some discomfort, is an important piece of data. The student's plea typically involves a student who has not performed well in a course and now realises that untoward consequences will follow from his lack-lustre effort. The consequences can vary from having an upset grandmother to losing student status and being sent home to your native country where a dire fate awaits.The schemes used for these cases might be anyone of argument from distress (190), argument from conse- quences, or argument from need for help (191), augmented by a twisting of the argument (190). But in what kind of dialogue does the student's plea find itself? Walton reports on an earlierwork 15 in which the fallaciousness was found to lie in a shift from a dialogue about the merits of an exam to one about the consequences offailing to get into law school (191); perhaps a shift from a critical discussion to a practical reasoning dialogue, we might conjecture, or perhaps a new critical discussion. I think there is a duality in the typical student's plea case, but I'm not sure that it is best described as involving two kinds of dialogue. Book Reviews / Compes rendus 181 Let us develop the student's line of thought. He knows that his profes- sor has a life. With that life goes a number of roles: teacher, member of professional organisations, brother, father, husband, and also, plain old hu- man being. When my duties, defined by these various roles, conflict, I natu- rally sense an internal conflict. When the student asks me for a higher mark on humanitarian grounds, I feel put "on the spot." I must choose between what I ought to do qua professor and what I must do qua concerned hu- man being. Even something so slight as an unknown grandmother's pro- jected unhappiness is a relevant consideration for me as a human being. When it is brought to my attention, it creates a minor conflict, but one which is not sufficiently strong to override my feeling of professional obligation. If something truly awful awaits the student unless I raise his grade, my deci- sion about what to do becomes much tougher (especially if the student's performance is in a "grey area"). In principle, there could be a case where my obligations qua human being would out-weigh my obligations qua teacher. But the important point here is that the ad misericordiam is relevant to me in one of my guises, and not in another. Hence, our ambivalence with this kind of argument and our inability to dismiss its reasonable occurrences out of hand. Of course, to say that an ad misericordiam appeal is relevant is a far cry from saying that it is sufficient. Despite Walton's very reasonab Ie remonstrations to the contrary (152), I find it useful to view the argumentum ad misericordiam, as the traditional theory of fallacies might do, as having, in general, this structure. A is speak- ing to B, and possibly A and C are the same person: Unless B does X for C then Z (something deserving B's sympathy) will happen. For moral reasons, B does not want Z to happen to C. Bean doX Therefore, B ought to do X The second premise indicates the presence of a moral 'ought', and so we must take the 'ought' in the conclusion to be meant in the same sense. Hence, if our student is arguing validly, as he should, then, ifthe premises are accept- able, the student has made a sound moral argument. If, alternatively, the stu- dent thinks that the 'ought' in the conclusion is the 'all-things-considered ought' , he has committed a fallacy of equivocation. Having noticed this possi- ble ambiguity, and returning to the hypothesis that the professor has two roles at least, we now see that there may be four possible situations to consider: the student presents a valid moral argument (a) to the professor qua professor, or (b) to the professor qua human being; or he presents an invalid argument with an all-things-considered 'ought' conclusion (c) to the professor qua profes- sor, or (d) to the professor qua human being. The last two cases, the argu- ment being invalid, we may call fallacies (since they at least give us pause until 182 Informal Logic we see our way through them), and so discount them. As for the first argu- ment, it is valid but irrelevant: moral considerations don't bear on student's grades from the point of view of the course requirements. It is only the sec- ond possibility, (b), that deserves further comment; it is the interesting case for, although it may be an unwelcome argument, it is relevant to the professor and it is not a fallacy. A possible problem now presents itself: if a student's plea is relevant, is ad misericordiam, is not a fallacy, and is directed to me as a human being, am I not bound to yield to it? The answer is No. I have a stronger moral obligation to uphold the values of the university in which I teach than I do to thwart a granny's oncoming disappointment. So, in that case I resolve a moral conflict by deciding in favour of the more stringent obligation. A similar analysis may be given of the Thackeray case. But suppose another kind of case in which I am forced to choose between taking my dancing lesson and responding to an ad misericordiam plea to help at the local food bank. I do have a moral obligation to do charitable acts, and make charitable donations, but I am not morally obligated to respond to every charitable appeal directed at me. Therefore, even when an ad misericodiam is a good argument, val id and relevant to me in one of my roles, it may be one that, on balance, I need not heed. But at that point, logical evaluation of the argu- ment is at an end, and something else, my morals, have to take over. This is what I alluded to when, after reformulating Walton's method, I wondered what would happen when the ad misericordiam argument was found to be faultless. Walton says the student commits a fallacy; I think it need not be so if there is any merit to the dual-role analysis. The dual-role analysis has the advantage that it can explain why such ad misericordiam arguments make us uncomfortable, and it makes room for that most traditional hallmark of fallaciousness, ambiguity. Walton says of the Thackeray case that "it pro- vides a kind of model of what is fallacious in the ad misericordiam argu- ment (when it is fallacious)" (192), and the same will be true of the stu- dent's plea case since it is so similar. Both these clear cases of the ad misericordiam invite the dual role analysis.When there are two roles we actually do have, the analysis works best. Sometimes, however, we don't actually have the two roles that might give rise to conflicting responses, but we imagine, by an act of sympathy, other roles we might have. Thus, in the Menendez brothers case, members of the public imagine the additional role of being a member of the jury and they then feel a tension between what the law allows and what seems intuitively right to them as non-jury mem- bers. (Members of the jury may also be torn when they remember how they would have reacted in their pre-jury days.) Hence, the members ofthe Book Reviews! Compes rendus 183 public who can appreciate the law's "point of view" might feel an ambiva- lence about the case. Walton's book repays careful study. One cannot help but be attracted to the broad pragmatic vision, the patience for detail, and the desire for a unifying theory. Even those who are not in total agreement with Walton on every point are beneficiaries of his logical imagination and his persistent effort to make the study offallacies a cornerstone of argumentation theory. 16 Notes I All page references in the essay are to Walton's book. 2 On Liberty, Ch II, para 19. 3 Schopenhauer would have been twelve years old in the year 1800, thus making him one of the younger contributors to the eighteenth-century literature on sympathy. 4 Walton also uses the terms 'form' (104), and 'structure' (152), as synonyms for 'argument scheme.' I wish he wouldn't do that. 5 Actually, Walton only mentions five kinds of dialogue on p. 114, forgetting aboutdelibera- tions. But deliberation dialogues are the most frequently mentioned in the book and they are included in the inventory of dialogue kinds on pp. 163-4. 6 The associated critical questions for this scheme are given on page 155. Presumably when arguments fitting this scheme are turned into appeals to sympathy they are appropriately used only when the help being asked for does not fall under the normal requirements of duty. Thus, to answer a question raised by Adler, a student asking a teacher for help with an assignment is only asking her to do her duty, and an appeal to sympathy in such a situation would be misplaced. But when the student asks the teacher for extra help, beyond what could normally be expected of the teacher, the student may wish to employ an argument fitting this scheme, and embellish it with an appeal to sympathy. 7 Thanks to Jonathan Adler for pointing out the similarity of this claim to the key premise in Peter Singer's renowned essay, "Famine, affluence and morality": " ... if it is in our power to prevent something bad from happening, without thereby sacrificing anything of comparable moral importance, we ought, morally, to do it." Singer's essay originally appeared in the first volume of Philosophy of Public Affairs (1972) but has since been widely reprinted. The quotation is from the sixth paragraph of the essay. A difference between Walton's schematic premise and Singer'S moral principle is that Walton balances giving help against a cost to the agent whereas Singer weighs giving help against competing moral considerations. 8 The ad baculum is not a moral argument since it appeals to the addressee's self-interest, rather than his concern for others. Some think the ad baculum is typically used negatively and, of course, some of these uses may be immoral, but this is a different matter than the argument turning on moral considerations. Another perspective denies argument status to the ad baculum, considering it to be merely a threat. For an example of this, see John Deigh's critical study of Walton's The Place of Emotion in Reason, this journal, Vol. 17 (1995), p. 119. Deigh's review is well worth revisiting for those interested in Walton's work. 184 Informal Logic 9 Eugene Garver contends that "in matters of practical argument and judgment, ethical criteria apply to arguments, not only arguers." Philosophy and Rhetoric (1998) 31: 107-30, p. 108. Garver's claim, if it were true, would not touch the point made in the present paragraph: that is, that arguments are open to ethical evaluation as arguments does not imply that it is their dependence on particular moral concepts (pity or sympathy, for example) that makes them either logically good or bad arguments. 10 See J. S. Mill, Logic, V.l.iL II Similar statements are found on pages 152, 156, 163 and 164. 12 But now, to the old question, 'Are all fallacies arguments?'Waiton must answer No, thereby running against the grain ofthe fallacies tradition. Over 80% of the fallacy theo- rists up to 1970 thought that fallacies were arguments. See my "The straw thing of fallacy theory," forthcoming in Argumentation. Il Here I am quoting Walton quoting someone else. My knowledge of this case, and the Nayirah case too, by the way, is based wholly on Walton's presentation in this book. 14 Here I am assuming that there was no change in the law and no change in the charges. 15 Walton and E. C. W. Krabbe, Commitment in Dialogue (Albany: State University of New York Press, 1995). 16 I wish to thank Francisca Snoeck Henkemans and Jonathan Adler for their critical com- ments on an earlier version of this essay, and the Department of Communication, Argu- mentation and Rhetoric at the University of Amsterdam for supporting my research. Hans Vilhelm Hansen Department of History, Philosophy & Political Science University of Windsor Windsor. Ontario Canada N9B 3P4