Emotions and Argumentation- AARON BEN-ZE'EV University of Haifa Abstract: The relationship between emotions and argumentation is not always clear. I attempt to clarify this issue by referring to three basic questions: (1) Do emotions constitute a certain kind of argumentation?; (2) Do emotions constitute rational argumentation?; (3) Do emotions constitute efficient argumentation? I will claim that there are many circumstances in which the answer to these questions is positive. After describing such circumstances, the educational implications of the connection between emotions and argumentation will be indicated. 1. Do emotions constitute a certain kind of argumentation? A distinction can be made between argument and argumentation. An argument may be described as set of propositions, one of which is designated as the conclusion and the remainder as premises, whereby the conclusion is claimed to be based upon (e.g., derived from, supported by) the premises. Argumentation is a complex activity in which someone presents a thesis and substantiates it by reference to some premises. Such activity is a form of communication where, in the presence of what we perceive as incompatible positions, we convey our own position and somehow substantiate it (Goldman 1994; Willard 1989). Argumentations are then associated with the following features: (a) a communicative conkxt; (b) incompatible positions; (c) substantiation of one's position. In typical argumentation the communicative context is that of speech acts, the incompatible positions are explicitly presented, and the substantiation involved is that of a few-steps process of intellectual reasoning. Sometimes argumentation takes a less typical form: the communicative context is not that of speech acts, the incompatible positions and premises are not explicit, and the substantiation is not through a reasoning process. It is obvious that speech acts are not the only form of communication involving incompatible positions. Body language is another example. However, in order to be regarded as argumentation the element of substantiation should also be present- otherwise, the message will be merely a claim. "This person is dead wrong" is not an argumentation unless some form of substantiation is attached to it. The typical form of argumentation is an explicit process of reasoning. Sometimes the substantiation is implicit; consequently, messages which appear to be mere claims can actually be regarded as forms of argumentation. When an argumentation becomes familiar, it can be viewed in a single step, its substantiation implicit in the pattern commonly associated with it. In this way, a reasoning-why process becomes a one-step intuition (Margolis 1987: 134-136). Thus, if in a campaign speech a prominent politician claims "I am telling you that this view is dead wrong," this claim is delivered and received a~ an argumentation. It states not merely that the other view is wrong, but rather that "J am telling you that it is such and such." The implicit substantiation is the Informal Logic Vol. 17, No.2 (Spring 1995): 189-200 © Informal Logic 1995 190 Aaron Ben-Ze'ev assumed experience, knowledge and wisdom of this politician. Politicians do not merely wish to communicate certain claims; they want to deliver persuasive arguments. The form in which a certain claim is communicated or the way in which it was generated sometimes provides its substantiation. I will argue that an emotion can often be seen as a non-typical form of argumentation: it communicates a message incompatible with a certain state of affairs and its substantiation is implied in the profound nature of the emotional state and the way it was created. Emotions are complex attitudes involving the intentional components of cognition, evaluation, and motivation, and the feeling component. The cognitive component includes the information about the given circumstances; the evaluative component assesses the personal significance of this information; the motivational component addresses our desires, or readiness to act, in these circumstances. The feeling component is a primitive mode of consciousness which expresses our own state, but is not in itself directed at this state or at another object. These components are not separate entities or activities; they are distinct aspects of a typical emotional experience (Ben-Ze'ev, 1996). Ignoring the complex nature of emotions, reducing them to either feeling or cognition (as is often the case in theories of emotions), obviously makes it impossible to regard them as a kind of argumentation. Mere feelings, such as a toothache, or mere cognition, such as having certain information about someone, are not argumentation. Only when we take into consideration the complex nature of emotions and refer to all four components--and in particular to the evaluative component--does the connection between emotions and argumentation become evident. I have suggested that argumentation is associated with a communicative context, incompatible positions, and substantiation of one's position. There are at least some emotional attitudes that can be said to satisfy these features. When Bertrand Russell, during his long love affair with Lady Otto line Morrell, became jealous of the intimate relationship she maintained with her husband Philip, his jealousy may be regarded as a kind of implicit argumentation. The way he communicated his jealousy to Lady Ottoline was intended to make her realize he wanted her to change her attitude; their positions were incompatible (Le., while he demanded she cease allowing her husband access to her bed, she refused); and the care and great personal involvement he thus revealed, he believed substantiated his position. Let me discuss in more details the connection of the above features to emotions. Emotions are typically social in their nature. Accordingly, they characteristically have a communicative or expressive function. The unique social sensitivity typical of emotions ensures that the situation of others is taken into account by us and that our situation is taken into account by others. Our limited resources and multiple goals force us to make constant choices in our Emotions and Argumentation 191 daily life---we cannot do or attain everything we want. Such choices must consider the ability and needs of others and are fraught with uncertainty. Our emotional response must adequately communicate our attitude and must take into account the attitudes of others. Emotions express our profound attitudes better than words. People who do not know the language or customs of a foreign country nevertheless know how to flirt with the natives. They use emotional communication which is largely universal. When emotional communication fails, serious aversive consequences follow. It has been suggested, for example, that sexually aggressive men use a suspicion schema when interpreting the way women communicate their (lack of) sexual interest: such men assume that women do not tell the truth when it comes to sex. Misinterpreting women's emotional communication results in sexually aggressive behavior such as rape (Malamuth & Brown 1994). Not all emotions have the communicative function. When in the privacy of her room Martha grieves the death of her mother, she is not making an argumentation as the communicative context is absent. Nevertheless, it can be argued that even having an emotion in isolation may sometimes be described as an attempt to convince oneself of the sincerity of one's attitude. In this sense, some communicative context is retained. For the purpose of my argument, I do not need to claim that all emotions are communicative (I have some doubts concerning this claim), but merely that many of them are. The position expressed in emotions is incompatible with some prevailing position. Emotions typically occur when we perceive highly significant changes in our situation. Like burglar alarms going off when an intruder appears, emotions signal that something needs attention; when no attention is needed, the signaling system can be switched off. We respond to the unusual by paying attention to it (Ben-Ze'ev, 1996; Frijda 1988:353-4; Lazarus 1991; Lyons 1980; Oatley 1992:50; Spinoza 1677). In the social communicative context typical of many emotions, the emotional change is often constituted: by a perceived opposing position or a perceived possibility of such a position. In anger we are speaking about a perceived opposing position, and in jealousy about a possibility for such a position. The comparison to an opposing--or at least different--alternative is crucial for emotions. The substantiation feature seems to be the most serious obstacle for considering emotions as argumentation. Emotions do not typically involve the presentation of a verbal conclusion to be followed or preceded by a reasoning process substantiating it. Their assumed substantiation is to be found in the profound nature of the emotional communication. Emotions express our most profound values and stands; expressing a view in a sincere emotional manner indicates our seriousness and high personal stake in this matter. Since the claim implicit in the emotional state is very important to us, other people are expected to take account of it and attempt to act accordingly. An emotional claim may be characterized as argumentation having the following form: the claim implicit in the emotional state is valid since I sincerely believe in it and it has emerged 192 Aaron Ben-Ze 'ev through the activation of a cognitive schema whose validity has been supported during the course of personal and evolutionary development. The emotional substantiation is personal: it involves personal logic not always accepted in intellectual argumentations. Russell's emotional demand that Lady Ottoline will sleep only with her lover (himself) may sound extravagant, but it has its own substantiation. It may have been based on the importance Russell attaches to their relationship and the assumption that people should have sex only with those they are in love with rather than with those they happen to live under the same roof with. The characterization of emotions as a kind of argumentation is close to Aristotle's view. Nancy Sherman argues that for Aristotle, "To feel hostile or friendly, indeed to feel specific emotions, in general, is to have reason to feel one way or the other, and it is of this that the rhetorician must persuade his listeners." (Sherman 1994:10; Aristotle, Rhetoric, 1378a22-24; 1379al0-12). It may be the case that the issue as to whether emotions constitute a kind of argumentation is terminological rather than conceptual. If one insists on being a purist, and maintains that argumentations must always be propositional, then emotions cannot be regarded as argumentations. But if we characterize argumentation in a broader sense, talk of emotional argumentation becomes possible, provided we keep in mind that this sort of argumentation is different from the intellectual, propositional ones. I have indicated that in many circumstances emotions may be regarded as a kind of argumentation. In other situations emotions are part of an argumentation in the sense that they influence its content. As Douglas Walton suggests, appeals to emotions in argumentation may have a legitimate and important place when existing knowledge is insufficient to tilt the balance decisively to one side. In many situations the appeal to emotions in argumentation can be a good presumptive guide to a provisional action or conclusion. In light of the profound nature of emotional evaluations, heeding our emotions in argumentation can bring us in harmony with our deepest, fundamental commitments, which define our personal stance. Nevertheless, we should be cautious when appealing to emotions in argumentations since in many circumstances such an appeal may become fallacious (Walton 1992). 2. Do emotions constitnte rational argumentations? Granted that in some circumstances emotions may be characterized as a kind of argumentation, I turn to discuss further the nature of the emotional substantiation, in particular whether it is rational. I will argue that emotions are rational in the sense of being functional rather than intellectual. Two senses according to which emotions can be considered rational may be discerned: (a) a descriptive sense---the generation of emotions involves intellectual calculations; (b) a functional sense---emotions express a functional response in the given circumstances. The two senses are not interdependent; emotions can be rational or non-rational in each sense or in both. Emotions are Emotions and Argumentation 193 essentially non-rational in the descriptive sense since they are not the result of deliberate, intellectual calculations. Emotions are essentially rational in the functional sense: frequently, they are the most functional response. In many cases, emotions, rather than deliberate, intellectual calculations, offer the best means to achieve the optimal response. In such cases it is rational (in the functional sense) to behave non-rationally (in the descriptive sense). The failure to distinguish between these two senses of rationality underlies much of the heated dispute about the rationality of emotions. Emotions are non-rational in the descriptive sense, namely, they are typically not the result of intellectual reasoning processes. Unlike intellectual argumentation, typical emotional evaluations are not deliberate. They are spontaneous responses depending on a more elementary evaluative system than the intellectual one. The two types may clash. Thus, we sometimes persist in being afraid when our conscious and deliberate judgment reveals that we are no longer in any peril. We may explain such cases by assuming that certain non-deliberate evaluations become habitual to a degree where no deliberation can change them. This corresponds to situations in which intellectual knowledge fails to influence illusory perceptual contents. Spontaneous emotional evaluations are ready-made mechanisms of appraisal which have already set during evolution and personal development. Since the evaluative patterns are part of our psychological constitution, we do not need time to create them; we just need the right circumstances to activate them. We should distinguish between rule-following behavior and rule-described behavior, that is, behavior that follows rules and behavior that merely satisfies rules. The difference is between being guided by a known rule and simply being in accordance with a rule, or between intentional rule-following and non-intentional forms of mere lawful connection. Describable regularity need not imply actual calculation. Although a bird's flight and the spider's behavior in making webs can be described by complex, abstract mathematical formulae, neither birds nor spiders follow rules or make intellectual calculations (Ben-Ze'ev 1993: ch. 4). The same holds for the emotional system. The regularities typical of emotions should be described as assumptions structured into our personality, not as intellectual calculations carried out inside our heads. The emotional agent is not aware of premises and cannot therefore infer conclusions from them. Instead of assuming an intelligent agent who makes intellectual calculations, we should assume a well-designed and somewhat inflexible system, thus providing a more economical explanatory mechanism. Indeed, simple mechanisms often underlie what seems overwhelmingly complicated when described by formal idioms. Our emotional behavior is clearly not rule-following behavior. When we fall in love, or become angry, we do not calculate the emotional response; in most cases the relevant data and the general principles of calculation are simply unknown to us. Although we actually do not make intellectual calculations, the emotional response, being in accordance with such calculations, may be perceived as if it 194 Aaron Ben-Ze 'ev were the result of such calculations. When one is angry with the right person to the proper extent at the right time, one acts in accordance with what reason dictates, but not because of it. Here anger speaks with the same voice as reason, but this does not mean that we employ reason through deliberate, intellectual processes. 1 We do not need such processes here; we simply act in accordance with our character. If we are less conscientious people, acting according to character may not be in accordance with what moral reason dictates, but it can still be an immediate response. We should distinguish between the descriptive issue concerning the psychological nature of the emotional response, namely, whether it is an immediate or a mediate response, and the normative issue of whether this response accords with what reason prescribes. The emotional response can be immediate (or spontaneous), namely, not preceded by mediating intellectual processes, but it may still accord with such processes. Emotions are non-pure and direct: they are influenced by our personal make-up, but are not mediated by intellectual processes. The influence of personal characteristics is expressed by the responsiveness, or sensitivity of the system. Mediated calculations are not required for the emotional system to behave rationally in the functional sense: reason in emotions is not simply a matter of calculation but first of all a matter of sensibility (Solomon 1990:47). Emotions are rational in the functional sense, namely, they constitute a functional response in the given circumstances. We have seen already one important function of emotion, that it, the communicative function. Two other important functions of emotions are (a) an initial indication of the proper direction to respond, and (b) quick mobilization of resources. Emotions function within individuals to indicate and regulate priorities, and between individuals to communicate intentions. Since emotions are generated when we perceive a significant change in our situation, their functionality must be related to our ability to function in these circumstances. This is clearly expressed in the indicative and mobilizing functions. The indicative function is required for giving us an initial direction in the uncertain novel circumstances we are facing. The mobilizing function is to regulate the locus of investment, i.e., away from situations where resources would be wasted, and toward those urgent circumstances where investment will yield a significant payoff. The indicative function of emotions is that of telling us which is the positive or negative nature of the uncertain circumstances we face and of helping us choose the initial course of actions accordingly. Because of the uncertainty and urgency of the situation, a decision must be made quickly but it is not necessary that the decision will be detailed at the very first moment. Emotional responses indicate the general direction of our actions by presenting us with an immediate evaluation of the positive or negative nature of the situation. Indeed emotions often amplify, or even provide, the first indication that something has changed and in particular that something has gone wrong. The indicative function is fulfilled not by presenting a convincing argumentation, but by inducing certain feelings. Emotions and Argumentation 195 The neurologist Antonio Damasio provides some evidence for the neurological basis of the indicative function of emotions. He speaks about somatic markers which highlight the negative or positive nature of each option. For example, when a bad outcome connected with a given response option comes into mind, we experience an unpleasant gut feeling. The somatic marker forces attention on the negative outcome of a given option and functions as an automated alarm signal which warns us of the danger ahead (Damasio 1994). The rationalist tradition argues that in order to get the best result in a logical decision-making process, emotions should be kept out. In light of the indicative function of emotions, this contention is plainly incorrect. Indeed, Damasio describes pathological cases in which a decline in rationality is accompanied by diminution or absence of feeling. Emotions assist "cool" reason by reducing the number of possible options to be considered and directing reason to the more advantageous options. Emotions may not only attribute an initial positive or negative value to each option, but also some preference order for choosing among them. Emotions make the cognitive landscape uneven, namely, with more or less salient and valuable contents. This may explain why the emotional defect in Damasio's patients is also connected to the way they consider future prospects: they appear to be insensitive to the future. The defect in these patients is not cognitive, but evaluative: they are aware of future options, but attach no value to them. Accordingly, they have no emotions toward these options. The mobilizing function of emotions is evident in light of the urgency of the situation: there is an urgent need to respond quickly and with all our resources to an event which can significantly change our situation. Since it is quick and intense, the emotional response is less accurate and more partial. By being partial, emotions focus our limited resources on those events that are of particular importance, thereby increasing the resources allocated for these events. Because of our limited resources and multiple goals we need a system which is able to switch our resources quickly from one event to another. Emotions constitute such a system. By switching our resources emotions discontinue a certain smooth operation, but this is done so as to jolt the system into a more helpful frame. Emotions enable the system to function in such instability. Once the initial switch is made, and the mental system has been shifted toward a more or less suitable frame, finer tuning of the system to its environment must follow. This process of fine tuning is no longer emotional; it expresses the eclipse of emotions. In some cases, however, emotions seem to be too slow from a practical point of view. Sometimes emotions surface after the significant change has already disappeared. For example, when we are driving a car and nearly miss having an accident, the intense fear that emerges comes too late to be effective for our behavior. In such situations the functional value of emotions lies not in providing an immediate response to a sudden change, but in realizing the significance of this change and appreciating its usefulness in preventing or facilitating in the future the circumstances that led to its happening. Incidentally, being already in an emotional state may either slow down or accelerate the next emotional 196 Aaron Ben-Ze 'ev response. Thus, sadness slows the agent down, while fear usually has the opposite effect. Emotions are the optimal functional response in the specific circumstances associated with their generation, namely, when we face a sudden significant change in our situation but have limited and imperfect resources to cope with it. That in these circumstances the emotional response is often the most functional, is because optimal conditions for the normal functioning of the intellectual system are absent. For example, we are not restless, a lot of the relevant data are missing, and speed may be more important than accuracy. In such circumstances our decision-making must be done in a more or less rigid form without having to think about what to do exactly. Fire drills are meant to help us acquire such forms of behavior. Emotions express these forms without undergoing such "drills"-those were done, so to speak, during our evolutionary and personal development. 2 The adaptive function of emotions is to be found in the way emotional patterns have evolved. The burden of explaining emotions should shift to developmental processes. Evaluative emotional patterns have emerged and have been modified throughout the evolution of the species and the personal development of the individual agent. In this sense, history is embodied in these patterns. Explaining emotional phenomena cannot be limited to (explaining) what happens in the fractions of seconds in which we are supposed to make the various intellectual calculations, but has to account for many evolutionary and personal factors. But we need not undergo the whole process of evolutionary and personal development each time we have an emotional encounter. This process has modified, or tuned, our emotional system in such a way that our surroundings immediately become emotionally significant. The evolutionary and personal background of emotions substantiates to a certain extent the claims implied in emotional responses, Le., these are not arbitrary, but often the most optimal given the circumstances. Others encountering us in an emotional state should consider these claims as expressing our sincere position based upon evolutionary and personal information and intended to have a certain effect--either on other people, in which case the emotions can be regarded as argumentations, or on the way we cope with the current situation. The functional value of emotions does not imply that emotions are beneficial in all circumstances. The popular advice to count to ten before expressing our anger--and to count to an hundred when very angry--i'eflects an awareness of the risks of an immediate emotional response. Such advice, however, does not completely dismiss the functional value of emotional responses: it does not suggest to count to the thousand. It may also be the case that in many circumstances the functionality of the emotional response will be enhanced if connected to the intellectual system. In any case, emotions are quite often rational in the sense of being the optimal response. This rationality may be characterized as a localized rationality since it does not take into account all possible implications but only those limited to the local present situation. Emotions and Argumentation 197 3. Do emotions constitnte efficient argumentations? After arguing that in some circumstances emotions can be regarded as a kind of argumentation and that since being functional they are rational, I turn to discuss the efficiency of such argumentation. I believe that emotions are quite common and efficient types of argumentation; their persuasive power is enormous and accordingly they are used quite often in everyday life. Emotions are not theoretical states; they involve a practical concern, associated with readiness to act. Aristotle argues that people are persuaded to do something when what is said stirs their emotions; accordingly, his main discussion of emotions can be found in his book on Rhetoric. Indeed, we often explain and justify our actions by reference to emotions. No wonder that politicians persuade us to vote for them by mainly referring to the emotional domain. Pictures and emotional slogans rather than intellectual discussions are at the center of election campaigns. Commercial ads are also directed essentially at the emotional domain. Their primary goal is not necessarily direct persuasion, but to make the viewers aware of the product and to have them associate it with something desirable. The aim is to trigger positive emotions that are then associated with the product. Accordingly, many more viewers gain pleasure from advertisement than buy the products being promoted. Rather than conveying information about product characteristics, advertising offers information about lifestyles and about viewers; the latter has greater emotional impact. It has been found, for example, that if you can induce a happy mood and at the same time not give people too much time to think, you can incline a person favorably towards your product, independently of its merit or usefulness. Indeed, commercial ads often lack "hard" product information such as material, price, or performance, and rely on techniques that emphasize stylistic features such as action, appearance, fun, or newness. For example, background music and the character of the person delivering the commercial message can significantly affect the viewer's evaluation of the product. The major function of such forms of "argumentation" is not to explain the phenomena--although they pretend to do that as well---but to make us experience certain emotions intended to induce certain attitudes and activities (Fiske 1988: 104; Oatley & Jenkins 1996: ch. 9; Van Evra 1990: ch. 7). Generating emotions consists mainly of activating certain evaluative patterns rather than persuading us to become emotional. So successful in inducing emotions, poetry does not present long intellectual descriptions, but rather touches upon sensitive points which activate the emotional system. Similar considerations explain why from an emotional point of view one picture speaks louder than a thousand words. The emotional system is more easily activated by visual than by verbal stimuli; the intellect is more related to verbal stimuli. It is not that emotional persuasion is devoid of information in whatever form but that in accordance with the nature of the cognitive component in emotions, this information is often partial and superficial. 3 198 Aaron Ben-Ze 'ev Emotions express our most profound values and attitudes: holding a certain value emotionally is necessary if that value is to become central to ourselves. But then also, it is easy to evoke them. We do not need a profound argumentation to generate emotions; on the contrary, very superficial matters easily induce emotional reactions. Because of their depth, emotional values are comprehensive and relate to many events in our life. 4. Educatioual implications The fact that in many circumstances emotions are more efficient persuasive tools than intellectual thinking has significant educational implications. There is a long tradition that criticizes the role of emotions in moral behavior and hence considers controlling emotions------even to the extent of almost abolishing them----as an important educational task. Since I believe that emotions have an important functional role in our life, I must oppose this tradition. Emotions should not be overlooked, but their weight should often be limited. Virtuous people are neither calm and unfeeling people nor ones led by passion. Their behavior is in accordance with the dictates of reason but it is not generated by intellectual deliberations; it is rule-described behavior rather than rule-following behavior. The role of emotions in such behavior is crucial. As Plato suggests, a sound education consists in training people to find pleasure and pain in th~ right objects. Similarly, for Aristotle a virtuous person is one who not only acts virtuously, but also has the appropriate emotional dispositions and character traits when acting this way. Not having the proper emotion is as significant as not acting in accordance with it. The virtuous, good-tempered person is not only the one who acts angrily in the appropriate manner on the appropriate occasion, but the person who also feels anger in these circumstances (Aristotle NE: II 05a28-b4. See also Frank 1990; Sherman 1994). The spontaneous, emotional system and the deliberate, intellectual system are both important for everyday argumentations. Each has its own function and is more adequate than the other depending on the circumstances. The presence of several systems of argumentation is as valuable as the presence of several powers in the political domain, often expressing opposing tendencies and competing interests yet each retaining a somewhat independent voice and influence. If our moral decisions were reached only through intellectual deliberations, those decisions would often be distorted since they would be one-sided-they would neglect important aspects of our life. The presence of conflict between the intellectual and emotional systems is frequently useful from a moral viewpoint since it indicates moral predicament to which we should pay attention. Emotions and Argumentation 199 • This paper was written while 1 was on sabbatical at the Centre for Applied Cognitive Science in the Ontario Institute for Studies in Education. I am grateful for the warm hospitality. 1 also thank J.A. Blair and an anonymous reviewer for helpful comments. I See also Aristotle, Nichomachean Ethics, 1102b28. 20adey (1992) claims that emotions may be viewed as biological adaptations to situations that "have no fully rational solutions." He argues that "Mechanisms that cope with limited and imperfect resources are not to be regarded as failures of rationality. They are among our most highly sophisticated cognitive features" (165,175). See also De Sousa 1987. 3 See also Ben-Ze'ev 1992; 1996. References Aristotle, The complete works of Aristotle: The revised Oxford translation. (1. Barnes, ed.). Princeton: Princeton University Press, 1984. Ben-Ze'ev, A. (1992). Envy and inequality. Journal of Philosophy, 89: 551-581. Ben-Ze'ev, A. (1993). The perceptual system: A philosophical and psychological perspective. New York: Peter Lang. Ben-Ze'ev, A. (1996). Typical emotions. In W. O'Donohue and R. Kitchener (eds.), Psychology and Philosophy. London: Sage. Damsio, A.R. (1994). Descartes' error: Emotion, reason, and the human brain. New York: Putnam. De Sousa, R. (1987). The rationality of emotions. Cambridge: MIT Press. Fiske, J. (1988). Television culture: Popular pleasures and politics. London: Methuen. Frank, D.H. (1990). Anger as a vice: A Maimonidean critique of Aristotle's ethics. History of Philosophy Quarterly, 7: 269-281. Frijda, N.H. (1988). The laws of emotion. American Psychologist, 43: 349-358. Goldman, A. (1994). Argumentation and social epistemology. The Journal of Philosophy, 91: 27-49. Lazarus, R.s. (1991). Emotion and adaptation. New York: Oxford University Press. Lyons, W. (1980). Emotion. Cambridge: Cambridge University Press. Malamuth, N.M. & Brown, L.M. (1994). Sexually aggressive men's perceptions of women's communications: Testing three explanations. Journal of Personality and Social Psychology, 67: 669-712. Margolis, H. (1987). Patterns, thinking, and cognition: A theory of judgment. Chicago: The University of Chicago Press. Oatley, K. (1992). Best laid schemes: The psychology of emotions. Cambridge: Cambridge University Press. Oatley, K. & Jenkins, J.M. (1996). Understanding emotions. Oxford: Blackwell. Sherman, N. (1994). The role of emotions in Aristotelian virtue. Proceedings of the Boston Area Colloquium in Ancient Philosophy, 9: 1-33. Solomon, R.C. (1990). A passion for justice: Emotions and the origins of the social contract. Reading: Addison-Wesley. Spinoza, B. (1677). Ethics. In E. Curley (ed.), The collected works of Spinoza. Princeton: Princeton University Press (1985). 200 Aaron Ben-Ze 'ev Van Evra, J. (1990). Television and child development. Hillsdale: Erlbaum. Walton, D. (1992). The place of emotion in argument. University Park: The Pennsylvania State University Press. Willard, C.A. (1989). A theory of argumentation. Tuscaloosa: The University of Alabama Press. AARON BEN-ZE'EV DEPARTMENT OF PHILOSOPHY UNIVERSITY OF HAIFA HAIFA 31905, ISRAEL CI