Book Reviews Scare Tactics: Arguments that Appeal to Fear and Threats. Douglas Walton DordrechtiBostoniLondon: Kluwer Academic Publishers, 2000. pp. xv, 1-218. ISBN 0-7923-6543-7. Hardbound 91.00 EURJ 105.00 USDI 66.00GBP. Reviewed by A. Francisca Snoeck Henkemans In Scare Tactics, Douglas Walton aims to explain how arguments that use scare tactics, such as appeals to fear, threats and force, can be identified, analyzed and evaluated. These arguments are quite commonly used in adver- tisements and negotiations, but are generally considered to be fallacious by authors oflogic textbooks. Walton thinks this judgment is too simplistic, and attempts to develop criteria for establishing when fear and threat appeal argu- ments are fallacious. In this endeavor, Walton first discusses numerous cases of fear appeal and ad baculum arguments, and presents a historical survey of the treatment of ad baculum arguments in the logic textbooks and in argumen- tation theory in general. From this overview of treatments it emerges that a dual approach to the analysis of ad baculum is required in order to evaluate cases of ad baculum as fallacious or non fallacious: (1) a logical approach in which the structure of inference of ad baculum arguments is defined, and (2) a dialectical approach in which it is established how the ad baculum argument is used in a context of dialogue. According to Walton, the distinction between fear appeal arguments and threat appeal arguments as they are defined in the literature is not clear enough. In order to clarify this distinction, a pragmatic analysis is required in which the making of a threat is defined as a specific kind of communicative action. The book ends with a classification system of the various types and subtypes of scare tactics and a proposal for a method of evaluation of ad baculum arguments. In Chapter 1 a characterization is given of fear appeal arguments. A fear appeal argument has three central characteristics: !I(i) it cites some possible outcome that is fearful to the target audience, (ii) in order to get that audience to take a recommended course of action, (iii) by arguing that in order to avoid the fearful outcome, the audience should take the recommended course of action." (20). Walton makes use of the work of social scientists in order to explain how fear appeal arguments work to alter behavior. According to one ©In/orma/ Logic Vol. 21, No.3 (2001): pp. 267-274. 268 Francisca Snoeck Henkemans influential social science model, fear appeals are only effective ifthe respond- ent takes the fearful outcome as a real danger, and if the action recommended to deter the threat is perceived by the respondent as feasible and easy to carry out. Only if the feasibility and ease of the recommended action outweigh the seriousness of the danger, will the respondent react cognitively and will he be persuaded to take the action. Otherwise, the respondent will choose to deal with the threat emotionally, by trying to cope with the fear instead of the danger. Because of their cognitive component, Walton argues that fear appeal arguments can be seen as a species of practical or prudential reasoning, and therefore cannot automatically be regarded as fallacious. Chapter 2 presents a survey of the standard treatment of the ad baculum fallacy in logic textbooks and manuals in English from 1906 to 1995. In the early textbooks, ad baculum is defined as an appeal to force, sometimes in- cluding the threat of force. It is typically regarded as a fallacy of irrelevance, either because it is not an argument at all, but just a threat or appeal to physical force, or because threats and the use of force cannot establish the truth or falsity of a proposition. In the more recent textbooks (after 1956), ad bacu- lum is often defined as an appeal to fear, but there are also authors who make a distinction between appeals or threats offorce (ad baculum arguments) and appeals to fear. Walton concludes that there is no agreement on fundamental matters of definition, analysis and evaluation in the standard treatment of the ad baculum. Chapter 3 discusses the scholarly literature that has begun to develop since 1975. In Walton's earlier work together with Woods, he considered the typical example given of ad baculum in the logic textbooks as a prudentially reason- able and thus non-fallacious argument. In retrospect, Walton thinks that "the fallaciousness of the ad baculum [ ... ] is not to be sought just in its internal logical structure as a type ofinference but, somehow, it is to be sought in how that inference is used in a broader context of dialogue" (73). Other authors claim that the fallaciousness depends on the type of conclusion that is being defended by means of an ad baculum. If the conclusion is descriptive, then pointing out negative consequences of not accepting the conclusion is falla- cious. If the conclusion is prescriptive and regards an action that should be carried out by the respondent, then the ad baculum argument is non-falla- cious. Walton quotes an example given by Kielkopf of a case where both a prescriptive and a descriptive reading are possible: Liberia warns the United states and Britain that lifting the economic sanctions against Zimbabwe-Rho- desia would be considered a hostile act. The (implicit) conclusion aimed for with this threat is: The U.S and Britain should not lift their sanctions against Zimbabwe-Rhodesia. This conclusion could be read descriptively "telling us that as a matter of sociological fact our standards or conditions for justifiable lifting of the sanctions have not been met" (75) or prescriptively, i.e. that it is wise not to lift the sanctions if we want to avoid certain negative conse- Review of Walton 269 quences of that course of action. In the latter case the argument is a non- fallacious ad baculum argument. There are also authors, however, who think that even if an ad baculum argument is used to persuade the opponent to perform some action, the argu- ment can be called fallacious, be causes it uses scare tactics, and therefore seeks to manipulate or coerce rather than persuade. Another matter of contro- versy concerns the question of whether threatening is essential to the ad baculum or not. Some argue that if a warning is issued instead of a threat, the argument is a non-fallacious argument from consequences and should not be regarded as a case of an ad baculum argument. Walton concludes from his discussion of the various analyses of ad baculum that a combination of a logical and a dialectical approach is required, so that attention can be paid both to the rea- soning used in the ad baculum argument and to the context of dialogue in which the argument is put forward. In Chapter 4 Walton attempts to give a pragmatic analysis of threats, by discussing various proposals made in the pragmatic literature. His conclusion is that a threat is a kind of speech act, but not an iIlocutionary act. Threats can be conditional ("If you don't do A, I'll do B") or unconditional, but Walton focusses on conditional threats, since it is only the conditional threat that can be seen as an argument. Walton formulates the following essential condition for a threat: "The speaker is undertaking to see to it that the event will occur unless the hearer carries out the particular action designated by the speaker" (128). In Chapter 5 ("The New Theory"), Walton presents his own theory of fear and threat appeal arguments. The first part of this theory is that fear and threat appeal arguments share the same kind of underlying structure, that of the practical inference, more specifically, that of an argument from (negative) consequences, and that therefore, these arguments can be seen as reasonable prudential arguments when used in an appropriate context of dialogue. One difference is that in the case of the threat argument it is the proponent himself that will bring about the negative consequences whereas this is not the case in the fear appeal argument. Another difference is that in a fear appeal argument the negative outcome that the respondent is incited to avoid, should, accord- ing to Walton, be "such a bad outcome that it is likely to evoke fear in [ ... ] the respondent" (143). Threat appeal arguments, on the other hand, are not nec- essarily intended to exploit the respondent's emotion offear. The second part of Walton's theory is dialogical in nature: both fear appeal and threat appeal arguments involve multi-agent reasoning: there has to be a sender and a re- ceiver of the argument. But Walton remarks that these two agents could be the same person, for example in a case of self-deliberation. In the dialogue ex- change that is typical of practical reasoning, a number of critical questions (concerning among other things alternative ways of avoiding the negative consequences and possible negative side effects of performing the requested 270 Francisca Snoeck Henkemans action) can be raised by the respondent, which the proponent should attempt to answer satisfactorily. Chapter 6 aims to resolve questions of a terminological and classificatory nature. Walton presents a classification system of the different types of scare tactics, by means of which he makes it clear how the different argument types distinguished in the literature are related to each other. In order for an argument to be classified as an argumentum ad baculum, it should involve the making of a threat or of an appeal to force. An appeal to fear is therefore not considered an ad baculum, and both the threat appeal and the fear appeal argument are subsumed under the category of danger appeal arguments. These are the main categories in Walton's classification system: Danger Appeal Arguments / ~ Ad Baculum Arguments Scare Tactics / ~ / ~ Use of Force Threat Appeal Arguments Fear Appeal Arguments / ~ Simple Threat Complex Threat Ad Baculum Ad Baculum Finally, in Chapter 7, Walton deals with the question of how ad baculum arguments and scare tactics should be evaluated as arguments. Walton pro- poses a dialectical method of evaluation: "in order to evaluate a case, you need to look not just at the premises and conclusions of the argument, but also at how that reasoning is used as part of a goal-directed conversational exchange" (177). As in his earlier work, Walton distinguishes between six types of dia- logue, and explains that the question of whether the use of scare tactics is fallacious or not depends on the type of dialogue the participants are engaged in. In a critical discussion, for example, threats and fear appeals are out of place. In negotiation, the making ofa threat can be an acceptable tactic. Walton next makes a distinction between weak and fallacious arguments: an ad bacu- lum argument may be prudentially weak in the sense that it does not provide adequate answers to the appropriate critical questions applying to an argument from consequences, but that does not necessarily mean that it is fallacious: "To make the allegation that the given argument is fallacious is to claim more than just that the argument is weak, or insufficiently supported, in the given case. What needs to be shown is that the argument [ ... J is such a severe systematic error or tactic of deception that it disrupts the dialogue the partici- pants are supposed to be engaged in, blocking the goal of the dialogue" (203). Contextual clues should be used to determine what type of dialogue the par- Review o/Walton 271 ticipants are supposed to be engaged in. If such clues are not present, then the critic should give a conditional evaluation of the form: 'if this argument was supposed to be part of a dialogue of type x, then it is fallacious (or nonfallacious ).' In Scare Tactics, Walton discusses an abundance of examples, makes nu- merous interesting observations and many subtle distinctions that are often lacking in other treatments of ad baculum arguments. Nonetheless, there are a number of problematic aspects in his treatment. My main criticisms concern Walton's pragmatic analysis of threat appeal arguments and his classification system of ad baculum arguments and scare tactics. Pragmatic analysis of threats By giving an analysis of the speech act of making a threat, Walton hopes to be able to give a clearer definition of what an ad baculum is, and to develop guidelines for determining in a particular case whether a threat has been made or not. The latter is necessary, since threats are often issued in an indirect or covert way. After having discussed some approaches to threats, Walton concludes that a threat is a speech act, but not an illocutionary act, since one cannot make a threat by uttering a sentence like "I hereby threaten to ... ". But this criterion is not sufficient reason for not considering threats as illocutionary acts. As van Eemeren and Grootendorst have pointed out in their book Speech acts in argumentative discussions, there are other illocutionary acts that cannot be introduced with the phrase "I hereby X", such as for instance the speech act of argumentation. Also, there are other ways of making clear the illocutionary force of a threat, for instance by saying: If you don't clean up your room today, I'll do it for you, and that's a threat! Walton also hesitates over whether threatening should be classified as a perlocutionary or illocutionary act. But calling a threat a speech act implies that it is an illocutionary act. According to Walton, "if threats include fear appeals, or if the notion of frightening is in- cluded within the definition of the concept of making a threat, then there are grounds for seeing threatening more as a perlocutionary type of speech act" (126). But why would this be the case? The illocutionary purpose of a threat (and thus the perlocutionary effect the speaker aims for with his ilIocutionary act) could be to frighten the addressee, without it being therefore necessary to see threatening as a perlocutionary act. Another inaccuracy concerns Walton's formulation of the sincerity condition for threatening (by the way, why for- mulate felicity conditions for threatening if you don't think making a threat is an illocutionary act?): "Both the speaker and the hearer presume that the oc- currence of the event will not be in the hearer's interests [ ... ]" (114). It is the speaker who should believe that the event will not be in the hearer's interests, but if it turns out that the hearer does not think that the event in question is bad 272 Francisca Snoeck Henkemans or undesirable, one cannot for that reason accuse the speaker of being insin- cere. Apart from these inaccuracies in Walton's analysis of the speech act of making a threat, my main problem is that he does not make use of this analysis in the rest of the book. The only place where his speech act definition turns up again is in Chapter 7 where the evaluation process is summarized. There the speech act definition precedes the representation of the inference structure of the threat appeal argument. But why is it only in the case of the threat appeal argument that a speech act definition is given? Couldn't the distinction be- tween a warning and a threat have been helpful in clarifying the distinction between a fear appeal argument and a threat appeal argument? Couldn't the correctness conditions for making a threat have provided a basis for formulat- ing more specific critical questions for the evaluation of threat appeal argu- ments? The classification system Walton's aim in developing a classification system of scare tactics is twofold. In the first place the classification system should provide clearer definitions of the main concepts involved in ad baculum and fear appeal arguments, and make it clear how the different types of scare tactics are related to each other. In the second place, the classification system should provide a basis for the evaluation: "until we have a clearly defined structure of reasoning underlying the ad baculum and fear appeal types of argumentation, no further useful advance in attempting to evaluate such arguments is possible" (171). It is, however, doubtful whether Walton succeeds in achieving these aims. The main problems with Walton's classification are that some types of arguments he distinguishes in his classification are not arguments, that other distinctions seem rather arbitrary, and that some distinctions are even coun- ter-intuitive. The direct use of force is included under the heading of argumentum ad baculum, even though Walton admits that using force is not an argument. The same goes for the unconditional (or simple) threat: it is considered to be a nonargument, but it is nonetheless defined as a subspecies of ad baculum arguments. In a classification system of "argument types" these tactics should not have been included. An example of the arbitrariness of the distinctions is the way Walton dis- tinguishes threat appeals from fear appeals. His main reason for considering a threat appeal a different type of argument from a fear appeal is based on an observation of Kevin Donaghy about threats, i.e. that "an utterance may be accurately characterized as a threat even though it was not intended as a threat" (147). The example given by Donaghy is a case where a speaker says to a listener "If you can't pay us something, we will have to repossess the Review o/Walton 273 car". Even if this remark were just intended as nothing more than information, the utterance may, according to Donaghy, be accurately characterized as a threat by the respondent. According to Donaghy, this shows that threats are not necessarily "attempts to frighten the hearer, to overpower the hearer, or to coerce the hearer into performing some action" (148). Walton uses Donaghy's analysis to justify his own classification ofthreat and fear appeal arguments in the following way: Donaghy's analysis points to the right way to classifY threat appeal and fear appeal arguments, by indicating that the threat appeal is not simply a subspecies of fear appeal argument [ ... ] The fear appeal has an addi- tional element. That element is one of the speaker's using a calculated tactic of appealing to the fear of the hearer. In the new dialectical theory, this element does not necessarily require that the speaker's act of ap- pealing to the hearer's fear be intentional. It does imply [ ... ] that the appeal to the hearer's fear [ ... ] is an integral part ot the argument. (148) In my opinion, Walton takes over Donaghy's analysis of threats too uncritically, whereas it is in fact rather problematic. To begin with, what does "accurately characterized" mean? Clearly not that the hearer's characterization is an accurate description of the speaker's intentions. It may of course be an accurate description of the way the listener understood the speaker's utter- ance. But then this fact is not typical of threats. A speaker may for instance inform a listener about his plans by saying: "I'll be at the office tomorrow". The listener may report this utterance to another person by saying: "He prom- ised to be at the office tomorrow". Then the listener perceives the original utterance as an attempt by the speaker to commit himself to a future course of action that is in the listener's interest. This erroneous uptake by the listener is in itself no proof of the fact that a promise does not necessarily involve the intention to commit oneself to a future action. The ill-motivated distinction between fear appeal and threat appeal argu- ments leads to a number of rather counterintuitive (and sometimes even in- consistent) definitions in Walton's classication system. Threat appeal argu- ments fall under the more general heading of 'scare tactics'. This means that Walton is now faced with the problem that he has characterized threat appeal arguments as arguments that are not necessarily aimed at frightening the op- ponent. He is therefore forced to give an extremely counterintuitive definition of 'scare tactics': "the term 'scare tactic' as used in the new system, does not necessarily imply an attempt to frighten, or an intent to arouse fear in the respondent" (173). Since fear appeal arguments are the other subspecies of scare tactics, Walton is also forced to say that fear appeal arguments can be made without the intent to frighten the hearer. How can this be consistent with the description offear appeals as "calculated tactics of appealing to the fear of the hearer" (148)? And how can the distinction with threats be upheld? 274 Francisca Snoeck Henkemans It is also doubtful whether the classification system provides a good basis for the evaluation of fear appeal and threat appeal arguments (Walton's second aim). Both fear appeal and threat appeal arguments are based on the same underlying form of inference, the argument from negative consequences. There are some differences in the types of premises in a fear appeal argument and a threat appeal argument, but Walton does not formulate different critical ques- tions for these two types of argument. So if there are any consequences for the evaluation, they are not mentioned by Walton. And for the question of whether the fear appeal or the threat appeal argument is fallacious or not, the evaluation of the argument is not decisive. That is to be decided byestablish- ing whether the argument is an obstacle to the goal of the dialogue. In his survey of recent approaches to scare tactics, Walton does not de- vote any systematic attention to other dialectical approaches, such as the pragma-dialectical approach to fallacies. The fact that scare tactics may be seen as violations of three different types of rules of critical discussion, is therefore not noticed by Walton. A threat appeal argument can be seen as a violation of the rule that parties must not prevent each other from advancing standpoints or casting doubt on standpoints, a fear appeal argument can be classified as a pathetic fallacy which violates the rule that parties may not defend their standpoint with nonargumentative means of persuasion or with irrelevant argumentation: the arguer then uses nonargumentative (or irrelevant) means of persuasion by exploiting his opponent's fear. A fallacious argument from consequences (the argumentum ad consequentiam) can be seen as a violation of the rule that the defense should take place by means of an appro- priate argumentation scheme. In order to test an assertion (descriptive propo- sition), the arguer then points out undesirable consequences of the assertion (evaluative proposition), thereby confusing facts and values. Seeing the dif- ferent types of scare tactics as violations of different rules for critical discus- sion is thus an alternative way of making a distinction between these types. It would have been interesting to see how this approach would have fitted in with Walton's own approach. A. Francisca Snoeck Henkemans Department of Speech Communication, Argumentation Theory and Rhetoric University of Amsterdam Spuistraat 134 1012 VB Amsterdam The Netherlands afsnoeckhenkemans@hum.lIva.nl