Emotion, Argumentation and Informal Logic* MICHAEL A. GILBERT York Univers ity Abstract: Over the past 60 years there have been tremendous advances made in Argumentation Theory. One crucial advance has been the move from the investigation of static arguments to a concern with dialogic interactions in concrete contexts. This focus has entailed a slow shift toward involving both non-logical and non-discursive elements in the analysis of an argument. I argue that the traditional attitude Informal Logic has displayed toward emotion can be and ought be moderated. In particular, I examine the role of emotion in everyday argumentation, and how Informal Logic can encompass it alongside the more traditional logical mode of communication Resume: II y a eu un enorme avancement theorique dans l'argumentation durant les dernieres soixante annees. Au debut on investiguait seulement des arguments statiques, mais maintenant on prend en consideration les interactions dialogiques dans des contextes concrets. Cette approche importante a entralne un changement lent vers l'inclusion d'elements non logiques et non discursifs dans l'analyse des arguments. J'avance que l'attitude traditionnelle de la logique non formelle exprimee envers les emotions peut et do it etre moderee. En particulier, .i'examine Ie role des emotions dans l'argumentation de tous lesjours, et la fayon que la logique non formelle peut les inc lure avec les approches traditionnelles logiques de communication. Keywords: emotion, argument, informal logic, coalescent argumentation, critical reasoning 1. Introduction Over the past 60 years there have been tremendous advances made in the inter- disciplinary area known as Argumentation Theory. Building on the work done in Informal Logic, Discourse Analysis, Communication Theory, and Social Psychology and Sociology, the advances have led to new and dramatically improved models and constructs.! The central advance, from the standpoint of this essay, has been the move from the investigation of static arguments taken in isolation, to a concern with dialogic interactions situated in a concrete context. Within Argumentation Theory, for example, less attention is paid to the analysis of extracted bits of text, and more to ongoing interpersonal discussions. As a result, interest has grown in understanding a wider variety of conversational interactions than previously considered. This interest has entailed an achingly slow shift toward involving both non-logical and non-discursive elements of communication in the analysis of an argumentative interaction (see Willard 1981: l31). © Informal Logic Vol. 24, No.3 (2004): pp. 245-264 246 Michael A. Gilbert I will argue in what follows that the traditional attitude Informal Logic has displayed toward emotion can be and ought be moderated. There are, as I point out below, indicators of a softening approach that mark a difference from the earlier attitude which saw emotion essentially as a disruption unfortunately occurring in an otherwise logical world. But the tide has not changed dramatically. Critical Thinking and Informal Logic have always been dedicated to the promotion of rational argument. The idea that reasons must move us to the acceptance or rejection of a position is a central tenet of the field. But the term "rational" has generally been reserved to signify arguments that are linear, essentially discursive, and involve appeals to empirical statements that are verifiable or, at least, widely accepted by received authorities. So, Honderich describes a standard approach to rationality as follows: "Beliefs that are contrary to the dictates of reason are irrational. Rational beliefs have also been contrasted with beliefs arrived at through emotion, faith, authority, or by an arbitrary choice" (1995: 744). Arguments, consequently, composed of or relying upon non-rational reasons, themselves become at best weak arguments, and at worst irrational ones. The sorts of arguments favoured by traditional approaches to Informal Logic tend to be more linear and involve direct appeals to verifiable forms of argument. One way to think about this is to use the terminology of Perelman and Olbrechts-Tyteca (1969) who described certain arguments as "quasi-logical." These arguments, while not easily identifiable, have about them a structure that is similar to familiar forms of deductive reasoning such as modus ponens or disjunctive syllogism. They are quaSi-logical because, Perelman and Olbrechts-Tyteca maintain, no argument that is found in the natural language can be considered wholly formal or logical insofar as a degree of interpretation is always required to bring out its underlying structure (ibid., ! 93). In what follows, I will not use the term 'rational' but rather' logical' to signify arguments and reasons having this quality. The reason for my change of terminology is that one of my key points is that emotional "reasoning" or, if you will, the utilization of emotion in arguments in part or in whole, is perfectly rational. Not permitting emotion to be considered rational strikes me as begging the question, so here I stipulatively replace the term 'rational' with the term 'logical.' Moreover, the notion of "logical" is broader than deductive logic and covers inductive and abductive as well. The core point is that the kind of reasoning used is the sort that may elicit the expression, "That's logical." It should be noted that while being discursive is a typical hallmark oflogical arguments, it is neither necessary (since most arguments are enthymematic) nor sufficient. It is not sufficient because emotional argumentation may also be quite discursive, as when someone makes an assertion that she is, say, angry. However, emotional information is also often inferred from tone of voice, facial expression, body stance, or other contextual clues easily read by a native to the particular culture. 2 Concomitant with the idea oflogical and emotional arguments is the corollary that the components of such arguments will also be logical and/or emotional. That is to say, one may have emotional reasons or logical reasons or both in a given argument. Emotion, Argumentation & Informal Logic 247 To complicate matters even more, there is no suggestion that an argument or the premisses (reasons) of said argument may not be both logical and emotional. First, one may look at the logical and emotional aspects of a given statement or argument, and, secondly, there may be a variance between the discursive content of a reason and the emotional message expressed by it. The point of view I am defending is one that I have argued over the last ten years: all communication, and argumentation in particular, must be viewed as involving four modes, often highly intertwined and only separated with an analytical eye. These modes are the logical, the emotional, the visceral (or physical and contextual,) and the kisceral (the intuitive, mystical and religious). The Argumentation Theorist, I maintain is under an obligation to examine the rules and procedures for the non-logical modes so that the rationality inherent in them can be examined and normative rules developed. I have explicated and defended this view elsewhere (Gilbert 1994b; Gilbert 1997a) , and will here focus on issues involving the logical and emotional modes. In what follows I am not going to dwell on the question ofthe ultimate nature or metaphysical status of emotions. Needless to say, the question is vital and has an impact on the ways in which one expects emotions to play out in argumentation, and it may help to understand that my general orientation regarding both communication and emotion is social constructionist (see Jagger 1989, for example). However, I will leave these discussions aside in order that we can examine emotional communication within a dissensual framework as a process rather than focus on emotions as things. Whatever emotions are, we are all too aware oftheir presence in argumentative contexts, and whether they are sensations, feelings or some other psycho- philosophical entity is beside the point. The field of Argumentation Theory is a multi-disciplinary area that focuses on the dynamics, structure, and nature of arguments and argumentation. Considerations in this area focus on both descriptive and normative components, and, as in many fields, there are those whose priorities put normative considerations over descriptive, and vice versa. From the point of view of philosophy, the lead area has been the field ofInformal Logic which includes investigation of argument structure, premiss- claim relationships, argument evaluation, and the study offallacies. Scholars in the area can argue at length how an argument (almost exclusively a static mono logic argument) should be diagrammed, analyzed, and assessed. While there was a great deal of controversy about these issues, there was also, in certain instances, a good deal of concurrence. One such area of concurrence was the importance of rationality, and the irrelevance, or even fallaciousness, of emotional considerations put forth in an argument. These attitudes begin quite far back with Arnauld & Nicole's Port Royal Logic (1996 [1685]), and continue with such a standard work as Introduction to Logic by Copi, who writes, after Russell, "strongly felt emotions can prevent thinking about whether the reasons given for an action are good and sufficient for taking it" (1963: 131). More recently we have Govier (2001: 210), 248 Michael A. Gilbert who states, "even when emotions are legitimate and not manipulated, they are not in and by themselves good reasons for belief or action." There are two important caveats to my claim that Informal Logic holds a prejudice against emotional reasoning, i.e., reasoning with and by emotional arguments. The first is that most writers are happy to allow that emotions are normally present in many arguments, and that the presence of emotions or emotional premisses is acceptable so long as they do not become too powerful or overwhelming. In other words, a certain amount of attention is paid to emotional considerations, but these typically feel very much like add-ons where one finds it necessary to say, "Well, yes, emotions do have a role to play in argumentation, but one must be careful." The second caveat is that contemporary argl)mentation theorists are becoming more open to the role played by emotion in dissensual communicative interactions, and the inclusion of rhetorical considerations is more widely expected. More recently, there has been a softening, and a more general acceptance of the idea that if Argumentation Theory is to locate itself in actual marketplace argumentation, then the communicative tools actually used by arguers must not be ruled out of court because they do not meet ideal standards. This can be seen, for example, by examining the three separate editions of Johnson and Blair's Logical Self-Defense (Johnson and Blair 1977; 1983; 1993). In each later edition the discussion of emotion and rhetoric is more sympathetic and emphasizes its integral nature. This does not, however, mean that non-logical modes are being given equal billing, but that their presence does not instantly pollute the entire argument. In the Pragma-Dialectic arena, there is similarly a weakening of the idea that "ideal arguments" themselves must exclude affective components (van Eemeren and Houtlosser 2000). That is, rather than allowing that we must abandon the ideal, we redefine it, as I have been arguing, to include the simple human fact that affective and rhetorical components will occur in every dissensual interaction. And this brings us to my current project. Emotion, in all its forms, is an integral part of human communication, and, consequently, of human argumentation. As such, there are a number of choices to be made. We can view emotional arguments as intrinsically different and distinct from logical arguments and, consequently, expect them to be treated very differently. They will not have similar sorts of rules for proper usage and interpretation, and the limitations on emotional input will be much more restrictive than on logical. Some scholars believe no affective material ought appear in a rational discussion, while others believe that some may, but it must be limited and not permitted to influence the outcome. The agreement we reach as a result of our argumentation ought not be influenced by feelings, emotional appeals, desires, or other, non-logical reasons. The main reason for this attitude is that emotional information is unclear, fuzzy, subject to misinterpretation, and inclined to bias, while logical information is not. I am, once again, not going to address this issue directly, since I already have done so elsewhere (Gilbert 200 I). But, for my purposes here, I must claim that emotional messages are either as clear as logical messages or almost so. The defense of this claim will be Emotion, Argumentation & Informal Logic 249 found in (Gilbert 2002: 30ff.), but the essential point is that discursive logical communications and premisses require clarification and contain inherent and contextual ambiguity every bit as much as non-discursive non-logical communications and premisses. For Argumentation Theory this means that whenever there is disagreement, be it with an argument or reason in the logical mode or one in a non-logical mode, there is no great advantage to something's being communicated discursively. What I want to do here is ask the following question. Assuming that, I. emotional arguments are integral to human communication, and 2. the comprehension of the emotional content of emotional arguments is more or less on a rar with comprehending logical arguments, then what normative frame ought be developed in order to govern emotional argumentation? The core goal of Informal Logic is the creation of rules for the analysis and assessment of arguments. Are the arguments presented for a claim, position or standpoint good ones? Do they offer sufficient grounds for agreeing with the position or accepting the claim? Ifnot, then why not? Do the presented arguments contain fallacies or procedural errors that weaken the support the premisses provide for the claim? In short, does the argument stand up to scrutiny? Now, if emotion is an integral, even, as some say a necessary part of communication (Damasio 1994; Greenberg and Paivio 1997), then the rules of argument evaluation must be applicable to the emotional content of arguments. Certainly, there are minimal and maximal cases of emotional argumentation, and in the minimal cases there may not be a great need to bring special considerations to bear. However, many arguments contain an appreciable degree of emotional content, and, so, it is important to develop the normative frame necessary for evaluation. Obviously, this is a huge project, and far beyond the scope of a single essay. But, as a component of the larger programme, I want to map out a core model ofInformal Logic presently used for framing logical argumentation, and suggest that there is no reason for using it to exclude non-logical argumentation. More, I argue that the most popular Infonnal Logic model can be, with a minimal amount of tailoring, applied to emotional argumentation as well. There are, of course, other models, regarding one of which, the Pragma-Dialectic model, I have begun a discussion elsewhere (Gilbert 2005). Another, the Gricean Cooperative Model, can, I believe, offer a rich interpretation ofthe usage of emotional infonnation and communication. Moreover, there are important models to be taken out of the therapeutic arena which also require investigation and where interesting work has been done. In this essay, I want to focus on that model most analogous to the one used in logical communication, i.e., communication utilizing logical argumentation, in the sense that I defined the term above. This model, introduced by Johnson and Blair, became used in a variety of ways by various informal logicians. For Johnson and Blair, it was the Relevance- Sufficiency-Acceptability [RSA] model (1993), while 250 Michael A. Gilbert for Govier it is the Acceptability-Relevance-Grounds [ARG] model (2001), and for Groarke and Tindale Acceptability-Relevance-Sufficiency (2003). Each author, and there are many others, changes the details, but the overall picture is the same. There are three main aspects of an argument that must be examined in order to determine ifit provides adequate grounds for accepting its conclusion. My claim is that using this model, we can establish a frame for emotional argumentation that is at least as reliable and normatively governed as logical argumentation. This effort can be viewed as a plank in my programme of destigmatizing emotional argumentative communication within the frame of argumentative communication. 2. The Informal Logic Approach - Acceptability A non-emotional reason or premiss is acceptable when it is either true or meets the standards of acceptability ofthe particular field in which the discussion is taking place. This characterization of acceptability is geared to argumentation insofar as the emphasis is on either truth, in some sense of the term, or standards which are accepted by the members of a particular field. Hambl in, for example, begins his examination of argument acceptability by looking at truth as the standard for acceptability, but quickly moves to a dialectical notion that is based on acceptance within a field (Hamblin 1970). He begins (234) by stating that the premisses of an argument must be true, but quickly moves to an epistemic foundation and suggests that, "The premisses must be known to be true" (236). Hamblin's final resting place, however, drops the notion of "true" altogether in favour of acceptability: "The premisses must be accepted' (245). The point to be made here is that trying to tie the acceptability of a premiss to a strong metaphysical notion of truth is by no means straightforward. In what fo Ilows, by "true" and "false" I shan be referring to Hamblin's notion of acceptability and unacceptability in a dialogue. Emotions, of course, are neither true not false; what are true and false are assertions concerning emotions. Emotions themselves are not usually statements of fact, and just as definitions are not true or false, but accurate or useful, so emotions require their own standard of evaluation. When we move into the emotional realms we find evaluation to be not so much with truth as with whether or not an emotional expression is genuine or counterfeit. That is, when I describe someone as, say, angry, then that assertion is acceptable or unacceptable just when the standards of the relevant field are met, but when I say that I am angry, then the assertion is acceptable if I am being genuine or unacceptable if! am being counterfeit, i.e., lying. In the third person context, I am, of course, saying that I believe that S had a genuine emotional experience. Human beings frequently, while communicating, express emotions. These expressions of emotion often guide us in how to proceed during a dissensual interaction as well as giving us information that mayor may not be discursively expressed. Since emotional expressions, that is to say, the physical demonstration of an emotion, may be discursively declared or not, it is necessary to distinguish Emotion, Argumentation & Informal Logic 251 between emotional expressions that are explicitly declared and those that are not. I shall say that an emotional expression is explicit when the presence of the emotion is discursively announced. As for example, ifS says, "I wish you hadn't done that, as it has made me very sad." Also, I shall say that an emotional expression is inferred when the hearer concludes that an emotion is being expressed, but the speaker has not done so discursively. An example of an inferred emotional expressions would be deciding that T is upset with me because she is glaring at me. Given this language, there are a number of issues that arise regarding the differences between logical argumentation, i.e., "rational" argumentation, and emotional argumentation. These difference are just those often referred to as the grounds for believing that the best arguments, the finest critical discussions, involve little or no emotional content. First of all, there is the difficulty in distinguishing between genuine and faked emotions, and secondly, the friability of conclusions concerning inferred emotions. Neither of these, I have argued elsewhere, is telling (Gilbert 2001), for the simple reason that the same conditions afflict logical material. In other words, it is not so much that emotional argumentation is clear and emotional inferences reliable, as it is that logical argumentation is unclear and logical inferences unreliable. Ben Ze' ev writes, "If one insists on being a purist, and maintains that argumentation must always be propositional, then emotions cannot be regarded as argumentations" (1995: 192). But the position that arguments must be propositional, it seems to me, is a secret way of saying they must be discursive, for, if not, they are then metaphysical objects just like emotions. If it is not, in other words, that propositions are discursive while emotions are not, then what is left is the information role of propositions. And emotional expressions have an information role as well. So the issue concerning the comprehension of emotions as opposed to propositions falls away. In other words, there is nothing more inherently reliable in propositional argumentation, discursive or otherwise, than there is in emotional argumentation. In both cases the message being sent is subject to the same sorts oflimitations and difficulties: (I) is it being sent correctly? (2) is it being received correctly? (3) is it true (genuine)? (4) is the inference of suppressed premisses or unexpressed emotions justified? The analogies between logical and emotional argumentative communications are striking, and it is only the misguided reliance we place on discursive communication that makes the logical appear more reasonable than the emotional. In fact, we cannot interpret logical communications correctly without using, at the same time, emotional markers. In Gilbert (2002) I argue that, "Communication takes places holistically. The recipient of a message uses a wide variety of cues and considerable range of information in order to interpret what is being communicated. The fewer cues available, the less reliance one puts on the interpretation. These cues range from the very words being uttered, to the context in which the communication is taking place, to the emotional and intuitive feel of the message" (25). Words rarely communicate their message clearly without at 252 Michael A. Gilbert least some reliance on the non-logical aspects; supposing otherwise is an instance of what I have called the logocentric fallacy: "The Logocentric Fallacy is just the assumption that verbal pronouncements take precedence over other forms and modes of communication, and it is a fallacy because relying on it can often lead us to accept falsehoods rather than truth" (27). Consider the following examples. Example 1. Logical & Emotional Assertions A. I am six feet tall. B. Colette is six feet tall. C. I am angry. D. Andrew is angry. A is a straightforward assertion of fact, and C is an explicit emotional message. Both can be uttered insincerely, in the A case, say, to impress, in the C to intimidate. Band D can also be uttered insincerely, and, as well, may both be false though sincerely uttered. A can be uttered sincerely but still be false, while some would say C cannot. I think C can be sincerely uttered while not being true, but it is not crucial to the discussion. 3 Now it may be quickly pointed out that determining the truth or factual status of A and B is simple, while doing the analogous for C and D is not. I do not, however, really think this is the case, and certainly not always the case. If I assert that I am six feet tall, it is unlikely that someone who doubts it will whip out a tape measure and hold it alongside me. Even less likely is applying the measure to Colette because I said she is six feet tall. Moreover, there are numerous errors that occur in measuring. Must I remove my shoes? Will you measure to the height of my hair? And so on. Even so, our intuitions tell us that one is easier to deal with than the other. Perhaps we should change the example from being six feet tall to earning $250,000 a year. Suddenly, the ability to determine the answer disappears like mist in the breeze. The number of issues involved in deciding the answer to our new assertions can range from looking at a tax return to requiring a forensic accountant. But isn't that like requiring a psychologist to know if Andrew is angry? The verification issue is simply not strong enough to put emotions in a separate, isolated, category. Emotions expressed within an argument provide information that can playa crucial role in determining the acceptability of a premiss. Someone exhibiting anger or sadness when uttering a premiss indicates the degree to which that premiss is important, the role it plays in the argument, or a reaction to a received message. Moreover, we rarely believe a statement that we expect to be accompanied by an emotional expression, when it is not. Indeed, this often provides us ground for a further investigation of the authenticity of the stated premiss (Gilbert 1997b). But with respect to deciding acceptability, emotional expressions are no worse off than any other sort. It is certainly the case that displayed emotions may not be Emotion, Argumentation & Informal Logic 253 genuine. There can be many times and many reasons when one exaggerates or emphasizes emotional expressions. (When asked if they ever produce tears on demand, my students, primarily the female students, look at me as if I were asking if they ever washed their hair.) But this does not mean that the underlying feeling is not present, just that its demonstration is called forth intentionally and, perhaps, exaggerated, when it is argumentatively advantageous. Similarly, someone might put on certain trappings to make them look rich, or even lie about previous education or work experience. The point, in other words, is not that emotions and our readings of them are always reliable, but rather that ascertaining their status has the same difficulties and issues facing other more traditionally embraced modes of information. So I hypothesize that premiss acceptability, i.e., the determination that an expressed emotion is genuine or an explicitly declared emotion does in fact exist, is not particularly more complex than determining that someone is uttering the truth as opposed to exaggerating or outright lying. Exaggeration, logical or emotional, can be a more or less innocent rhetorical move, or it can undermine the value of an argument altogether-this depends on the context. Lying or faking is wrong and ought be eschewed in argumentation, but we well know that, alas, it occurs. The point is that when we suspect faking, just as when we suspect lying, further inquiry must be undertaken, and that is a natural part of the argumentation process. Emotional communications cannot be perforce excluded from rational communication merely because they may be faked or fail to be genuine. 3. The Informal Logic Approach - Relevance Within the realm of Informal Logic, relevance finds its roots in quite classical definitions. The core, as defined in two exemplar texts (Govier 200 I; Johnson and Blair 1983), uses the probabilistic approach laid out in Carnap (1962). The basic idea is that, Definition 1. A. (P*Q) B. Pr(QIP) C. Pr (Q) Then, (P*Q) Probabilistic Definition of Relevance df P is relevant to Q -df the probability of Q given P df the probability of Q .=, Pr (QIP) *- Pr (Q) One difference between the classic definition and the Informal Logic usage is that there is discussion of truth values rather than probability. So, Johnson & Blair ask, "does the truth of P dictate a truth value for C (the conclusion)?" (I 993: 54- 5). Govier explains that there is positive and negative relevance, depending on whether or not the truth of a statement P counts in favour or against the truth of a statement Q (163). The expression "in favour" has a more probabilistic sense than the idea of determining a truth value, so is even closer to the Carnapian definition. 254 Michael A. Gilbert All the authors emphasize that relevance is highly contextual, and that two statements deemed irrelevant to each other in one context may be deemed relevant to each other in another. That is, relevance does not inhere in statements, but in the relationship between them, given a set of contextual parameters. A discussion of the relevance of two statements to each other occurs during the course of an argument when two contemporaneous events happen. First; one of the discussants must question whether or not P is relevant to Q, and secondly, the respondent decides to argue the case that P is relevant to Q. Once that happens a case may be made. In other words, when we are actually arguing a question of relevance may arise and need to be addressed, or, on the other hand, when we are analyzing an argument the same may happen. But, like most questions concerning the nature and reliability of an argument, issues of relevance arise when one of the participants raises a question of relevance or an Argumentation Theorist is undertaking an analysis. Arguments about relevance are often subtle and can frequently be the focus of an argumentation. Examples of irrelevance frequently cloak this by being ridiculous. One asks, Why should I give you $1 OO? The answer is, Because there's a storm in Malaysia, and there is no further connection made. In reality, relevance is generally obvious from the context unless it becomes a core point at issue. Consider: Example 2. The Appointment Carlos: Our next faculty appointment should be someone in meta-ethics. Denise: Oh, I was thinking Eastern philosophy. There's a lot of student demand for it. Carlos: Student demand's irrelevant. We're a research department and we need to go with our strength. Denise: Student interest is hardly irrelevant! If we don't pay attention to what our students want, then ... The argument that follows will be about the relevance ofP to Q, but wiII clearly be tantamount to arguing about the claim at issue. Within the posits of the current discussion, the idea that emotional information is intrinsically irrelevant to an argument is ruled out of court; we have previously posited that emotion is integral to argumentation, and, ipso jacto, certainly can be relevant to the argument at hand. Besides, the suggestion is absurd. Whether or not emotional information is crucial or central or decisive in an argument may be open to question and investigation within the context of a given argument, but in reality, and especially in actual marketplace argumentation, emotional information plays an important role. However, the question of when emotional considerations are relevant to the claim in contention is a perfectly acceptable one. Not surprisingly, conditions similar to those for logical components, i.e., premisses and/or claims, apply to the emotional mode as well. What does this mean? It means that we can find emotional considerations that are obviously irrelevant, those that are obviously relevant, and Emotion, Argumentation & Informal Logic 255 those that require further consideration. Remembering that anything can be made relevant to anything else, someone being sad in Malaysia is, ipso facto, no reason for me to fork over $100, any more than the aforementioned storm was. But, on the other hand, that is not to say that a story cannot be constructed in which Malaysian malaise may rightfully cost me money. When we begin to deal with situations where emotional content mayor may not be relevant, where there is dissensus, we can see that the core idea of one statement being relevant to another when it makes a difference to the statement's degree of acceptability is, more or less, the same. In the following example there is a real question as to the relevance of the emotional premiss brought forth. It is worth noting that the examples I use are largely those involving should/ought or value situations. The reason for this is that my concern here is with argumentative contexts, and not with relevance simpliciter. Ifwe are concerned, for example, to know if having committed plagiarism once is relevant to committing it again, we can determine the answer via a study. But when we are trying to decide if a first time plagiarizer should have a sealed memo placed in her record, the answer will more likely arise through a dialogue than a study, though any number of studies may be relevant. So, Perelman and Olbrechts-Tyteca refer to Plato's dictum in the Gorgias that we do not settle issues concerning facts through dialogue, but simply through inspection. "When agreement," Perelman writes in The Realm of Rhetoric, "can easily be reached by means of calculation, measuring, or weighing, when a result can be either demonstrated or verified, nobody would think of resorting to dialectical discussion" (I979 13). Example 3. The Raise - 1 Mohit: So we're agreed, Catherine should receive the promotion. Elaine: Yes, but then we oUght to give Harold a merit increase. Mohit: Well, there are more deserving than him. Elaine: Perhaps, but he's going to feel bad about not getting the promotion, so this will at least be some recognition. Mohit: Hmm, I see your point, but I'm not sure that's relevant. I mean, '" In this case there is a question as to whether or not Harold's feelings are relevant to his receiving a raise. Elaine will argue that it is relevant because Harold had a reasonable expectation of the promotion, and giving him a raise will maintain his morale. Regardless of the outcome, we see clearly that Harold's emotional state is an issue whose relevance can be debated. This example can go even further. Elaine might confess to Mohit that she, Elaine, feels bad that Harold is not getting the promotion even though she agrees that their choice is the correct one. So, she might put forward her own feelings as well as Harold's for giving Harold the merit pay increase. Given the inexactitude of such decisions, why would her feelings be irrelevant? Well, for one thing, Mohit might respond that the department's guidelines 256 Michael A. Gilbert talk about such things as publications a''J teaching and administration. Therefore, assuaging Harold's feelings is not relevant according to the guidelines. Elaine might then try to bring her feelings and Harold's under the umbrella of one of these three categories, or, perhaps, might argue that there are overweening factors that warrant an exception. The point is not that Elaine's feelings or Harold's feelings are or are not clearly relevant. Rather the point is that we can argue about them in a way similar to arguing about logical P1.atters. If, for example, Wai-Chen had published a book of poems, would she receive merit in her home department of philosophy? Is her book of poems relevant to her philosophical pUblications? This presumably logical issue would, I suggest, generate a great deal of discussion. Consider another situation. Carl has perpetrated a crime against Ahmed. During the trial, Carl's remorse is quite properly deemed irrelevant to his conviction, but at sentencing, Ahmed, the victim, may express forgiveness, and that might be deemed relevant, as might Carl's remorse. On the other hand, saying, "He did it!" would be relevant at Carl's trial, but not at a sentencing since at that time his guilt is already established. 4 There are any number of circumstances where emotions are relevant to the result of an argument, and they are relevant in the same way as other sorts of information. My wife and I may be arguing over whether I should attend a conference on our anniversary, necessitating our separation on that date. Considerations of professional advancement, seized opportunities, and so on will come into play, but so wiJI our sadness at being apart on that special day. The degree of my wife's upset will quite rightly, it seems to me, be a relevant factor in the determination. These kinds of considerations are very common. On the other hand, when a student comes to me to plead for a passing grade when her course average is below pass, I may rule that her degree of distress is irrelevant. I may do so because to do otherwise is to import a valuation that I believe cannot reasonably be factored in. Thus, I might explain that everyone heading for failure feels bad, but I can't pass them all. As a result, the feeling itself does not come into my computations. This is not, however, to deny that I might allow that, in this case, the emotional factors at play are indeed relevant. What would make my conclusion reasonable is just what normaJly would: am I applying a measure that is available to all students? It would seem, then, that discussions of the relevance of the premisses of an emotional argument are not that different from discussions of the premisses of a logical argument. Both must be seen in context and cases made where there is disa,greement; relevance is important for both logical and emotional reasons, but one is no less easy or difficult to establish than the other. The point is that simply saying that S's feelings, or your own feelings, are irrelevantto making a claim more acceptable does not, ipso facto, make them so. The determination of their relevance or irrelevance to a claim's acceptability requires as much defense as does the same determination for a logical premiss. Emotion, Argumentation & Informal Logic 257 4. The Informal Logic Approach - Adequacy The final component of assessment is what is called by Johnson and Blair, sufficiency, and by Govier, adequacy or adequacy of grounds. In other words, given that one has been presented reasons that are acceptable and relevant, does the claim follow from them? Not surprisingly, this is the fuzziest and most difficult of the three classic criteria. Once one leaves deductive entailment behind, the precise ways in which conclusions are derived from premisses becomes considerably less precise and more complex. Johnson (2000) suggests that very little work has been done on sufficiency, but there is another way to look at the issue. That is to realize that most of the work done in Informal Logic has focused to one degree or another on the question of just when the premisses of an argument are strong enough to support the conclusion. This includes discussions offallacies, induction, abduction, and a myriad of other conventions and tools. The issue now becomes one ofthe relationship ofthe premisses of an argument to its claim: does the presence of an emotional premiss or the expression of emotion alter judgments of sufficiency in some essential or radical way? We have previously decided that we can detennine ifan emotional premiss is acceptable or unacceptable, and that emotional premisses can be relevant to the claim in question. So, given acceptable and relevant premisses, can emotional information, i.e., the feelings surrounding the issue as expressed directly or indirectly, assist in the formation of an argument sufficient to warrant a claim. Let us return to Example 3 The Raise - I. Mohit and Elaine are continuing their discussion. Mohit is not yet convinced that Harold ought receive a share of merit pay, and Elaine is continuing to make her case. Example 4. The Raise - 2 Mohit: Hmm, I see your point, but I'm not sure that's relevant. I mean, merit is supposed to go to those who are producing the most. Elaine: Maybe, but just stacking up publications is only one way of looking at it. Merit can also be used as an incentive, as encouragement, or, as I'm suggesting, to compensate for going along with a decision you may not like. Mohit: Well, Harold certainly won't like the decision. Elaine: Besides that, I feel bad that we're going to make Harold unhappy, and if we give him some merit, at least he'll know that we do feel bad. Mohit: I suppose, ... Elaine: And we don't have to give him a lot. Just something to salve his pride. Think of it as good for departmental morale. Mohit: What about the others? 258 Michael A. Gilbert Elaine: They'll all understand why we did it, and I think they'd agree. It's not like no one else will get any. Mohit: I suppose you're right. OK, let's do it. Mohit is now convinced that all things considered Harold can rightly receive some merit pay. The simple truth is that when we do consider all things, we take emotional reasons into account every bit as much as, if not more than, so-called rational, i.e., logical, reasons. What is considered in this discussion? The following factors are taken into account. Example 5. The Raise - 3 1. Harold is not getting the promotion. 2. Harold will feel bad. 3. Elaine (and let's say Mohit) will feel bad that Harold feels bad. 4. Merit pay can be used for a variety of purposes. 5. Getting some merit pay will make Harold feel less bad. 6. Harold's feeling less bad will make Elaine and Mohit feel better. 7. Harold's feeling less bad is good for morale in the department. 8. Objections from others are not anticipated. The dynamics of this argument involve two logical facts and six emotional "facts," or, three and five if (8) is considered logical. Each of (2)-(8) can be true or false, and their assessment-their acceptability-depends, in part, on the arguers' ability to assess and anticipate their own emotions, Harold's emotions, and the emotions of the remainder of their colleagues. Note that Mohit's acquiescence to (4), what might be considered a non-emotional premiss, allows the emotional premisses to enter the fray. I want to suggest, then, that this argument is quite typical in the way in which it mixes logical and emotional information. Some discussions might emphasize logical more than emotional, and some might even leave it out entirely, though that would be rare. Consider a brief argument that might be considered highly logical. Example 6. The Coupler Gordon is in a hardware store to purchase parts to connect his natural gas barbecue to his main gas line. Anandi, the clerk, is assisting him. Gordon: OK, so I think I've got everything I need. Anandi: Well, what about the coupler itself? Gordon: Oh, I've got one from my previous barbecue. I'll just use that. Anandi: I wouldn't recommend it. If the threads are worn at all, you could end up with a gas leak. Gordon: Is that likely? Anandi: Can't say, but do you want to risk it to save $10? Emotion, Argumentation & Informal Logic 259 Interestingly enough, the final decision Gordon makes is one of risk assessment. Given the facts, how does he feel about Anandi's suggestion? She is not telling him what to do, but offering what is, in fact, an emotional consideration. She is suggesting that, given the price, assuaging any fears of a gas leak are worth the cost. Given that most political decisions, social decisions, and the arguments from which they result, have a similar texture, it's clear that emotion is inherent in our deliberations. Going back to Example 4, The Raise - 2, the question remains: was the reasoning in the argument sufficient to warrant Elaine and Mohit's agreeing to give Harold merit pay? I think the answer is, yes, given their context and beliefs. In other words, I am not suggesting that if we replaced Mohit with another player, the result would be the same. A different person might insist that Harold and Elaine's feelings were irrelevant and that (4) is not acceptable. Nor does that mean the he would win the argument. The suggestion, however, that there is only one right answer to the basic issues, and that merit pay should be disbursed one and only one way, is indefensible-as anyone who has attended a department meeting on the subject can attest. Such a view fails to take into consideration the simple fact that people have goals, and different people have different goals. Those goals go toward determining values and beliefs, and the way in which those values and beliefs are prioritized. An argument, then, is sufficient when the reasons put forward for the claim are enough to convince the participants that, given their beliefs, values, goals, and criteria for argumentation, that the claim has been warranted. This is a clearly circular statement, but there is little way around that. It is the cultural context, the socially held matrix of a given field, from which we derive the standards of argument (Toulmin 1958; Weimer 1983; Willard 1989). Different warrants will be acceptable in different realms. I often hear conversations that depend on at least a minimal beliefin astrology, something I find ludicrous. Nonetheless, I can understand it when someone argues that Larry and Steve will never last as a couple because Larry is a water sign and Steve an earth sign. For some people, this argument is sufficient grounds for them to stop dating, while to me it is irrelevant as it is based on wrong beliefs. It is the frame, the conceptual social field of the arguers in which a discussion is taking place that defines the criteria that will be used. One can object, and even sometimes make headway, but to suppose that one system external to the others has the right of censure is to suppose an ability to bootstrap that is at least suspicious and at most epistemologically fascist. It is important to remember that when we are talking about sufficiency, we are talking about the grounds for which acceptable and relevant premisses may warrant a claim. This has to do with the logical relationships between them as well as the emotional weight given to them. Since the presumption is that the acceptability and relevance criteria have already been met, adequacy must be considered with that frame. In other words, given that the premisses are acceptable and relevant, do 260 Michael A. Gilbert they provide sufficient grounds for the claim? Within that context, the distinction between logical and non-logical premisses fades dramatically. 5. Conclusion The driving force of Informal Logic has been instilling a respect for careful procedures of judgment and analysis that will foster intelligent, considered, and unbiased evaluations of claims and the reasons and reasoning that support them. It is not surprising, therefore, that these extremely valuable goals have led to systems that prize precision and regulation. The result has often been an idealization of argument and a limitation to contexts (e.g., written materials, rarefied and/or sterilized dialogues,) in which analysis is possible. This is a consequence of the general attitudes within critical reasoning and the goal of ideal argumentation. Examiningjust what scholars hope to have as standards in theories of argumentation one quickly realizes that the criteria for "good" argumentation being bruited about are quite "high," so high, in fact, that they may only exist in abstracto. On the Pragma-Dialectic model in a "critical discussion," one is not supposed to have any attachment whatsoever to the outcome (van Eemeren et al. 1993: 32). But they also say,justtwo pages later: " ... actual human interaction is not 'naturally' resolution oriented. People involved in disagreement are not normally disinterested in the outcome but have a heavy interest in one outcome or another" (van Eemeren et at. 1993). "Critical doubt," writes Walton (1992: 271), "requires a neutral point of view, neither positive nor negative." Bias is what interferes with this, and bias, Walton allows, is a fact of life. Others are not so sure that ideal practices are what one should aim at. Issues abound regarding the role of emotion as a factor in maintaining a socio-political structure both with respect to gender (Jagger 1989, Gilbert 1994a), and also regarding political dissent. Reygadas (2001) makes a compelling case that the denial of a ro Ie for emotion in argumentation is central to the silencing of gender, class and racial minorities. So, on the one hand, we have a view that precludes emotion having a role in argumentation because it is less than ideal; and, on the other, a view that such an exclusion has negative socio-political consequences for the disadvantaged. The core issue is this: Emotion is seen as having to do primarily with how something is expressed, and not with what is expressed. This is why emotion has always been relegated to the rhetorical realm as opposed to the dialectical. The rhetorical, on the traditional model, concerns the ways in which something is expressed that make it more or less persuasive, and relies on the convince/persuade dichotomy. This view maintains that we can come to a conclusion either by using our rational resources, weighing evidence, being open-minded, following more or less logical reasoning, avoiding fallacies and bias, or by allowing ourselves to be led by our feelings, and accepting or rejecting something based on how we feel about it rather than its intrinsic merits. Clearly, for most philosophers there is no contest. Emotion, Argumentation & Informal Logic 261 Persuasion and its handmaiden--emotion-aJ'e bad, and convincing with its valet, reason, are good. 5 The difficulty is that this is just wrong. It is wrong on all counts, since, first, emotional messages do carry information rather than merely colour the dialectical information contained therein, and, secondly, because even if we do assume that the convince/persuade distinction can be made, emotional information is used whenever we are convincing as well as persuading. It is emotions and feelings that permit us to perform such basic argumentative tasks as selecting data, choosing examples, weighing alternatives, and deciding whether or not we agree or disagree. To suppose otherwise would embroil us in an infinite number of decisions that would have to be made prior to any communication at all (Damasio 1994). This crucial role of emotions notwithstanding, the prejudice still does not address the situation where emotional information per se is being used in the first or second person, or as complete or collateral warrant for a claim. In such instances, as I have argued above and at length elsewhere, (Gilbert 1997b; Gilbert 2001; Gilbert 2002), the criteria we use for evaluating emotional premisses and/or emotional expressions are just not that different from those we use with other data. In the preceding I have not attempted to show how one decides in any particular case when an argument is a good one and when it is not. What I have argued is that given that we distinguish between good and bad arguments, then there are no grounds, ipso facto, for supposing that the presence of emotion or of emotional content is sufficient for labeling an argument inadequate. A good argument can be considered one in which the premisses are relevant and acceptable, and where collectively they provide adequate grounds for accepting the conclusion. There certainly are rules for emotional argument. They are similar to the rules for logical argument, and include such factors as veracity, non-exaggeration,justification of evidence, avoidance of bias, consideration of alternatives, and so on. Emotional intimidation, for example, is wrong not because it is emotional, but because intimidation in an argument is often wrong; i.e., an argumentum ad baculum can be a fallacy whether based on logical or emotional threats.6 Emotion, when improperly used, can lead to manipulation as in, for example, emotional blackmail, when one improperly claims that the actions of one person will affect the feelings of another. But this is a far cry from maintaining that emotion has no legitimate role in argumentation. To the contrary, one form of an emotional fallacy is dismissing someone because she or he is being emotional or sentimental (Campbell 1994); ignoring emotion or dismissing it can be as argumentatively fallacious as its overuse. Human beings can and do argue about everything. All that is required is that I utter or otherwise communicate something, and you disagree, object, or query that something. It might concern the ontological status of emotions as opposed to feelings, it might be about what time we should leave for the dinner party, if young Gerald is a suitable partner for our Emma, whether 1 + 1 = 2, or what time the 1 :04 262 Michael A. Gilbert from Manchester arrives in Edinburgh. When we discuss these things we like to think that we can distinguish between facts, feelings, emotions, goals, and values. While I do not want to say that in some abstract theoretical way any such distinctions are without merit or completely useless, I do want to say that in the vast majority of cases alI these categories combine and intermix in such a way as to make their separation impossible. So, I have argued that emotion does have a legitimate role in argumentation, and that role is not so far from the role of so-called "rational," or as I prefer, logical, reasons and premisses. This said, it behooves us to examine the role of feelings and emotions rather try to pretend that they do not exist. Notes * The author would like to acknowledge the support of the Canadian Social Sciences and Humanities Research Council. Grant # 4 I 0-2000- I 340. An earlier, shorter version of this paper was presented to the Royal Society of Philosophy Conference on Emotions, Manchester, GB, July 2001. Finally, my thanks go to the reviewers and editors for helpful suggestions and improvements. I Good surveys are available in van Eemaen, et (II. (1996) and in Gilbert (1997, Ch. I). 2 In fact. one who is incapable of reading basic emotions is considered ill as, for example, autistic children. 3 The sort of situation I have in mind is when someone realizes he is not really angry, but very hungry. The feeling may be very similar to the feeling of being angry, but when the cause is recognized, the feeling is re-identified from anger to low blood sugar. , This is not to say that emotional information is never relevant at a trial, just that in this example the particular information is not. 5 Gender choices arc used intentionally; there is a long trad ition of emotion being considered feminine and reason masculine. (, Not every occurrence of ad bacll/llm is fallacious. Threatening to report harassment, or leave an abusive partner, and so on is not fallacious. References Arnauld, Antoine and Pierre Nicole, (1996) [1685]. Logic, or, The art of thinking .. containing, besides common rules, several new observations appropriate for forming judgment. Jill Vance Buroker (trans.). CambridgelNew York: Cambridge University Press. Ben-Ze'ev, Aaron. (1995). "Emotions and argumentation," Informal logic 17: 189-200. Campbell, Sue. (1994). "Being Dismissed: The Politics of Emotional Expression," Hypatia 9: 46-65. Carnap, Rudolf. (1962). Logical foundations of probability. Chicago: University of Chicago Press. Copi, Irving M. (1963). Introduction to logic (2 nd Ed.). New York: Macmillan. Damasio, Antonio R. (1994). Descartes' error: Emotion, reason, and the human brain. New York: G. P. Putnam. .... r Emotion, Argumentation & Informal Logic 263 Eemeren, Frans H. van, Rob Grootendorst, Sally Jackson and Scott Jacobs. (1993). Reconstructing argumentative discourse. Tuscaloosa: University of Alabama Press. Eemeren, Frans. H. van and Peter HoutIosser. (2000). "Rhetoric in pragma-dialectics," Argumentation. Interpretation. Rhetoric I. http://argumentation.ru/2001_I1 Gilbert, Michael A. (I 994a). "Feminism, argumentation and coalescence," Informal logic 16:95-113. Gilbert, Michael A. (1994b). "Multi-modal argumentation," Philosophy of the social sciences 24: 159-177. Gilbert, Michael A. (1997a). Coalescent argumentation. Mahwah, N.J.: Lawrence Erlbaum Associates. Gilbert, Michael A. (1997b). "Prolegomenon to a pragmatics of emotion," in Argumentation & Rhetoric [CD-ROM], edited by H. V. Hansen, C. W. Tindale and A. V. Colman, Eds., St. Catherines: OSSA. Gilbert, Michael A. (2001). "Emotional messages," Argumentation 15: 239-249. Gilbert, Michael A. (2002). "Effing the Ineffable: The Logocentric Fallacy in Argumentation," Argumentation 16: 21-32. Gilbert, Michael A. (2005). "Let's talk: Emotion and the pragma-dialectic model." in Frans. H. van Eemeren and Peter Houstlosser (Eds.), Argumentation in practice. Amsterdam/Philadelphia: John Benjamins, 43-52. Govier, Trudy. (200 I). A practical study of argument (5 th Ed.). Belmont, CA: Wadsworth Thomson Learning. Greenberg, Leslie S. and Sandra C. Paivio. (1997). Working with emotions in psychotherapy. New York: Guilford Press. Groarke, Leo and Christopher W. Tindale. (2003). Good reasoning matters' : A constructive approach to critical thinking. TorontolNew York: Oxford University Press. Hamblin, C. L. (1970). Fallacies. London: Methuen. Honderich, Ted. (1995). The Oxford companion to philosophy. OxfordlNew York: Oxford University Press. Jagger, Alison M. (1989). "Love and knowledge: Emotion in feminist epistemology," Inquiry 32: 151-176. Johnson, Ralph H. (2000). Manifest rationality: A pragmatic theory of argument. Mahwah, N.J.lLondon: Lawrence Erlbaum Associates. Johnson, Ralph H. and J. Anthony Blair. (1977). Logical self-defense: The craft of evaluating persuasion in everyday life. Toronto: McGraw-Hili Ryerson. Johnson, Ralph H. and J. Anthony Blair. (1983). Logical self-defense, 2nd Ed. Toronto: McGraw-Hill Ryerson. Johnson, Ralph H. and J. Anthony Blair. (1993). Logical self-defense, 3rd Ed. Toronto: McGraw-Hill Ryerson. Perelman, Charm. (1979). The new rhetoric and the humanities: Essays on rhetoric and its applications. Dordrecht/Boston: D. Reidel. Perelman, Charm, and Lucie Olbrechts-Tyteca. (1969). The new rhetoric. A treatise on argumentation. Notre Dame, Ind.: University of Notre Dame Press. 264 Michael A. Gilbert Reygadas, Pedro. (2001). "In praise of eristics," in Frans H. van Eemeren, Rob Grootendorst, J. Anthony Blair and Charles Arthur Willard (Eds.), 4th ISSA Conference on Argumentation. Amsterdam: Sic Sat. Toulmin, Stephen E. (1958). The uses of argument. Cambridge: Cambridge University Press. Walton, Douglas. (1992). The place of emotion in argument. University Park: The Pennsylvania State University Press. Weimer, Walter B. (1983). "Why all knowing is rhetorical," Argumentation and advocacy: 20: 63-71. Willard, Charles Arthur. (1981). "The status of the non-discursiveness thesis," Argumentation and advocacy: 17: 190-214. Willard, Charles Arthur. (1989). A theory of argumentation. Tuscaloosa: University of Alabama Press. Michael A. Gilbert Deparment of Philosophy York University Toronto, Ontario M3J I P 3 Gilbert@YorkUca