Aristotle and the Emotions

As published in Phronesis (1982) Aristotle and the Emotions* Stephen Leighton Since the object of rhetoric is judgment (Rhetoric 1377b2l) and since what appears does vary with the emotions (ibid. 1378al), a concern for rhetoric provided Aristotle with the opportunity to develop his most sustained thoughts on emotions; not only does he define, explicate, compare and contrast various emotions, but also he characterizes emotions themselves.1 His observation is quite striking. Emotions are the things on account of which the ones altered differ with respect to their judgments, and are accompanied by pleasure and pain: such are anger, pity, fear, and all similar emotions and their contraries. (ibid. 1378a20-23) Here a number of things provoke thought. First, how did Aristotle take the altering of judgments to occur? Second, what does Aristotle mean by speaking of the ‘accompaniment’ of pleasure and pain? Last, and resolvable only after the above questions are answered, is the conception of emotion like our own?2 These questions are worth answering not only for their value in understanding Aristotle, but also insofar as they shed light on our own understanding of emotion. Let us begin with the matter of altering judgment. I We would agree that emotions may alter our judgments. Love’s flame flaring, we view a beloved, and sometimes the whole world, through rose-colored glasses; our blood boiling, these same things are viewed rather differently. For it does not seem the same according as men love or hate, are wrathful or mild, but things appear altogether different, or different in degree; for when a man loves one on whom he is passing judgment, he either thinks the accused has committed * The copyright to this work is held by the author and publisher. I am grateful to Phronesis for the right to reprint this here. 1 As published in Phronesis (1982) no wrong at all or that his offense is trifling; but if he hates him, the reverse is the case. And if a man desires anything, and has good hopes of getting it, if what is to come is pleasant, it seems to him that it is sure to come to pass and will be good; but if a man is unemotional or in a bad humour, it is quite the reverse. (ibid 1377b30-78a4, based on a translation by Freese) From this and the previous quotation we can infer that emotions may move one to a particular judgment, may alter the severity of a judgment, or may change a judgment entirely. The field in which emotion operates is not restricted. Although the judgments altered which are foremost in Aristotle’s mind are formal verdicts given at the end of proceedings, there is no reason to doubt emotion’s effect on judgments on the way to a formal verdict, or for that matter on any other judgment. Thus the range of things to be included under affected judgment is quite general, forming, from our point of view, a rather heterogeneous group. For, so far, there is no requirement of belief in the judgment. Yet, as the passage immediately above suggests, belief may be present. Again, the group is not simply restricted to stated judgment, or even terminating judgment of any sort. Also, the sorts of changes Aristotle includes can be quite dramatic.3 Nevertheless, that such changes occur sits rather well with our own in tuitions on the matter. What, then, is (are) the explanation(s) of changes of judgment involving emotion? Aristotle nowhere explicitly reports on this matter. However, he does provide for a number of solutions. 1. An obvious place to begin is with the definitions of each of the emotions. The aim or end of an emotion could explain a change of judgments. In anger’s definition Aristotle speaks of seeking revenge (ibid, 1378a31-34). It is easy to see that one way of seeking revenge in a court room would be to return an unfavorable verdict. In more pedestrian settings one could achieve the same end by, say, slandering the person. Similarly, in love’s definition, Aristotle speaks of seeking the beloved’s good for his own sake (ibid, 1380b35-8la2). Again, it is easy to see that one way of seeking his good in a court room 2 As published in Phronesis (1982) would be by bringing down a favorable verdict. In more pedestrian settings one could sing the beloved’s praises (though not really deserved) and thereby alter one’s judgments. Similar considerations apply to other emotions Aristotle considers. Pity and indignation require a sense of justice. We are moved to pity because the misfortune suffered is undeserved (ibid, 1386b1 I); we are roused to indignation because the good fortune enjoyed is undeserved (ibid, 1386bl0). Thus we can suppose that our judgments concerning those we pity would become lenient and generous, while our judgments concerning those with whom we are indignant would be severe and mean-spirited. In both cases one would be compensating for the injustice that roused the emotion. This compensation takes the form of an alteration in judgment.4 In all likelihood one is well aware of what one is doing in such cases. We make certain judgments in public which are at odds with what we really believe.5 We are like the person who forms the right opinion, but through viciousness or lack of good will does not say what he really thinks (ibid, 1378a12-15). Just how this works is fairly transparent. One holds view A, but because one wants to do well or poorly by another, says B. This, then, is our first explanation of emotions altering judgments. It is most unlikely that Aristotle intends this sort of insincerity to bear much of the burden of explaining how emotions alter judgments. For though this can explain why one’s pronouncements vary and to that extent how judgments are altered, it is not helpful with Aristotle’s remark that things appear differently through emotion (ibid, 1377b30). Moreover, un less we suppose Aristotle to hold a rather eccentric, and suspiciously un acknowledged theory of self-deception, his theory has not yet begun to account for the interesting cases of emotions altering judgment: cases in which one is like the man who lacks good sense rather than like the vicious or those without good will (ibid, 1378a8cases in which the change of judgment has to be a matter of belief. Thus we must seek a further explanation of emotion’s effect. 2. A different sort of explanation is implicit in Aristotle’s remark about the speaker who 3 As published in Phronesis (1982) rouses the judges’ indignation towards those pleading for forgiveness. If then the speech puts the judges into such a frame of mind and proves that those who claim our pity (and the reasons why they do so) are unworthy to obtain it and deserve that it should be refused them, the pity will be impossible. (ibid, 1378b1721, based on Freese’s translation) The same sort of explanation is implicit in Aristotle’s remarks concerning one made envious. So that if the judges are brought into that frame of mind, and those who claim their pity or any other boon are such as we have stated, it is plain that they will not obtain pity from those with whom the decision rests. (ibid. 1388a26-29, based on Freese’s translation) The defendant vainly struggles to move the judges to one emotion while they are in the grips of another. The point seems to be that emotions have certain judgments connected with them such that certain other emotions, their judgments, and other judgments too are excluded. For example, John’s indignation with Mary involves John making a judgment of Mary’s unmerited good fortune (ibid, 1386b10) which thereby precludes John making the judgment of Mary’s undeserved misfortune (ibid. l385b14) that he would have made were he roused to pity her. Again, Mary’s envy of John involves, for example, Mary making a judgment of reproach concerning herself through John’s successes (ibid, l388a17) which thereby precludes Mary making a judgment of John’s undeserved misfortune that she would have made were she roused to pity him (ibid, l385b14). This is not a matter of insincerity on John or Mary’s part. Rather, being moved to one emotion with its judgments rules out being moved to another emotion with its judgments. Those judgments obtain which are connected with the emotion one is moved to. Thus, insofar as one moves to a given emotion one thereby alters one’s judgments; and this is underlined by the con sequent exclusion of other emotions and their judgments. We can see the same sort of alteration with other emotions as well. Should one be moved to anger, one thereby views the object of anger as having insulted one (ibid, 4 As published in Phronesis (1982) 1378a31-33). Becoming ashamed of a person involves being brought to view the person as involved in misdeeds that bring dishonor (ibid, 1383b15-16). Again, to the extent one is moved to these emotions one’s judgments are thereby altered. Similar points can be made for all the emotions Aristotle discusses. How this works is transparent also. To be moved to emotion A involves making judgments A; to be moved to emotion B involves making judgments B, etc. Thus, ‘things do not seem the same’ as one finds oneself in one emotional state as opposed to another, or none at all. Moreover, the judgments in any given complex may logically exclude those of another complex, or any other judgments. The most obvious contrast between the two cases considered is that while the former is a matter of insincerity, the latter is not. An equally striking contrast is that while the former is an example of emotion altering judgments, the latter is actually a matter of emotion itself being an alteration of judgments. In the second case, emotions are complexes involving judgments, each complex possibly excluding other emotion complexes, their judgments, and other judgments as well. It is not that envy brings about a change of judgments such that one does not show or feel pity; rather, to be moved to envy involves being moved to a particular set of judgments which excludes those of pity. Similarly, it is not that being angry makes us view the object of emotion as insulting, but being angry involves viewing the object as insulting. This sophisticated thesis sits very well with many modern analyses of emotion in which changes of emotion are, in part at least, changes in judgments. Although this is not what we began searching for, given the passages quoted above and Aristotle’s understanding of shame, anger, etc., and given how well this suits his claim that things do not seem the same when one is in different emotional states (ibid, 1377b30), we have no reason to doubt that the thesis is Aristotle’s. However, it does mean that we have not exhausted Aristotle’s thoughts on emotions and changes of judgments. For the characterization of emotion quoted at the outset of this paper speaks of emotions being that on account of which judgments change, not emotions themselves being changes of 5 As published in Phronesis (1982) judgments. Since, as we have seen, the matter of insincerity cannot be the whole of this explanation, there must be more to the account. Thus we must search for further explanations of emotions altering our judgments. So far, our study has revealed two species of the genus changes of judgment involving emotion: i) change of judgment as a consequence of emotion; ii) change of judgment as a constituent of emotion. The latter we have just considered. It will not be further subdivided; it works by means of emotions involving particular sets of judgments, judgments which may exclude the judgments constitutive of other emotions. The former will subdivide into four species, one of which is the matter of insincerity (method one). 3. Another way in which emotion might affect judgment is like our first explanation (I) insofar as it depends upon the aim or end of the emotion, but is like the second (2) insofar as it is not disingenuous, but rather a seduction by emotion. Consider again the angry person. We have seen that he seeks revenge. A change of judgment may result here only because one is disposed to give an unfavorable interpretation where the case is ambiguous. One never grants the benefit of the doubt, quite the opposite. In this way one would not only say that the person was worthless, but would have come to believe it, one’s anger having seduced one’s judgment. Again, in love’s seeking the benefit of a beloved, where circumstances are unclear, one would be inclined to give the beloved a favorable interpretation because one is ‘favorably disposed’. One thereby arrives at a far more charitable judgment than one would have had one been more rigorous and critical when considering the matter. As with the previous explanations, how this favor/disfavor method works is fairly transparent. Of a certain case, A, one is unsure how to evaluate it. But since the emotion disposes us and makes us desirous to favor or disfavor that to which the emotion relates, and since we do need to form some opinion of the case, we correspondingly judge the case harshly or favorably. Thus, emotion alters judgment.6 Should there be a number of related cases, one’s judgment not just of each particular case, but of the person will be 6 As published in Phronesis (1982) likewise swayed. For example, we shall tend to be charitable about the motives of a beloved, judging ambiguous cases in this light. Should we be faced with a number of such cases, this will strengthen our charitable interpretation not only of the cases, but also of him. This, then, is a third and twofold way in which emotion changes judgment. Although our ability to account for emotion changing judgments is increasing, further explanations are still needed. For the explanations appropriate do not seem at all helpful in the eases of fear, shame and shamelessness, given Aristotle’s definitions of these. For, on his view, these have no aim towards the realization of which our judgments might be bent. Of course, instead of searching for further explanations, this might lead some to think that the Rhetoric’s characterization of these emotions is inadequate. They too should have an aim in their definitions, or at least the general characterization of these emotions should include an aim. For example, it would be plausible to say that fear aims at flight. Now, if this sort of move could be made for all emotions, then we could say that all emotions could affect judgment in each of these three ways. However, this is not the position of the Rhetoric; and so it does not resolve the problems here. Nor is it plausible. What does sadness or shame aim at? If we disagree about Aristotle’s characterization of fear, I do not think we want to dispute that some emotions have no aim. That leaves us needing to search for further explanations. Moreover, where we do think emotions seduce our judgment, we do not always want to construe it in any of the above fashions. For example, consider Aristotle’s own example at 1378al. It is plausible that one could explain, in the favorable interpretation manner, the person who hopes for something good, and thus supposes it will come to pass. He desires its occurrence; and it is likely that he would give himself the benefit of the doubt concerning the many hurdles he has to face, but does not really know whether he can leap or not. Likewise, one without hope would be disposed to underestimate his prospects. However, if the case is unambiguous, if the hurdles to be faced are insurmountable, then any seduction of judgment seems inadequately explained in terms of a favorable interpretation. Because of this, and because certain emotions lack an aim, we must look for further explanations. 7 As published in Phronesis (1982) 4. The Nicomachean Ethics (1 149a24-31) is helpful.7 Anger seems to listen to argument to some extent, but to mishear it, as do hasty servants who run out before they have heard the whole of what one says, and then muddle the order, or as dogs bark if there is a knock at the door before looking to see if it is a friend. So anger by reason of the warmth and hastiness of its nature, when it hears, though not hearing an order, springs to revenge. (Ackrill’s revision of Ross’s translation) The position here is not that there is something particularly ambiguous; it is not just that one does not know how to take something, and so throws it in with one’s other judgments through the aim of one’s emotion. Neither is it a matter of conniving, nor of anger being a change in judgments. Rather, through the emotion one mishears, i.e., does not hear an order. Here the seduction arises from a mishearing, a misperception of what may be very clear evidence. This misperception in its turn results in faulty judgment. This is quite insightful, covering certain cases more plausibly. For emotion can alter perception and consequently the judgments based on these perceptions. Excited supporters of opposing tennis players often see rather different things. Their judgments based on these perceptions are accordingly influenced. Thus the seduction of the hopeful person facing nearly insurmountable hurdles can be explained by the misperception of those hurdles and the consequent misjudgment. However, even though we might like to agree that something along these lines is surely right, the account needs to be articulated more clearly. For the precise operation of emotion, especially upon perception, is not yet clear. Until it is clear, we do not know if Aristotle has adequately provided an additional way in which emotion alters judgment. To see how Aristotle can suppose this to work, we need to begin with a passage in De Somnis and then reflect upon the theory of objects of perception in De Anima. With regard to our original inquiry, one fact, which is clear from what we have said, may be laid down —that the percept still remains perceptible even after the 8 As published in Phronesis (1982) external object perceived is gone, and moreover that we are easily deceived about our perceptions when we are in emotional states, some in one state and others in another; e.g. the coward in his fear, the lover in his love; so that even from a very faint resemblance the coward expects to see his enemy, and the lover his loved one; and the mote one is under the influence of emotion, the less similarity is required to give these impressions. Similarly, in fits of anger and in all forms of desire all are easily deceived, and the more easily the more (hey are under the influence of emotions, So to those in a fever, animals sometimes appear on the wall from a slight resemblance of lines put together. Sometimes the illusion corresponds to the degree of emotion so that those who are not very ill are aware that the impression is false, but if the malady is more severe, they actually move in accordance with appearances. (460b based on Hett’s translation) Like the passage from the Nicomachean Ethics and unlike the first three explanations, this concerns the perceptual level of emotions affecting perceptions rather than the epistemic level of emotions affecting beliefs and knowledge. Aristotle claims that deception occurs readily when we are excited by the emotions — the coward by his fear, the lover by his love. With little basis the coward will see his foes; the lover, the beloved. Moreover, the more deeply the emotion is felt the more remote a resemblance may be which gives rise to illusory impressions. However, this is not to hold that we always get it wrong when in an emotional state: Aristotle suggests that we may recognize the illusion if the emotion is slight. In addition to concurring with the view of the Nicomachean Ethics, this provides part of the explanation we seek. Emotion is meant to alter perception through the expectation of emotion and the ‘putting together’ (suntiithemenōn) of things accordingly. If this occurs, then, having the wrong perceptions, we are likely to go on to make inadequate judgments. Still, how emotion can operate in this way on what we perceive is unclear, even if we grant, say, that the fearful would expect and put things together differently from the amorous. 9 As published in Phronesis (1982) If we recall the distinction between objects of perception per se and objects of perception per accidens in De Anima (book two, chapter six), we can make good sense of putting together through the expectation of emotion. An object of perception per se is a white thing; per accidens, Socrates. We are meant to perceive both sorts of objects.8 And what is noteworthy for us is that while the former object has little or no room for misperception or difference in perception, the latter object has a good deal of room. Thus, with this latter object error can occur; emotion can create illusions and alter perception. To explain exactly how this occurs, let me begin with the differences of perception that may occur without involving error or emotion. The object of perception that we all see (the object per se) is a black, circular, flat thing.9 If it is a record, a piece of plastic, and something different as well, then according to Aristotle, those things are perceived per accidens, even though the perceiver may not perceive it as those things and they may not, therefore, be ‘his object’. Although with my knowledge of records what I perceive it as is a record, and with another’s knowledge of the mysteries of Lil what he perceives it as is the sacred God, etc., still what is perceived per accidens is the record and the sacred God. While what is seen per se and even per accidens remains the same, the object per accidens that it is perceived as need not be the same. We can say ‘our objects’ of perception are different. As plausible as this may be, we need the case of misperception, and misperception through emotion.10 We have noted that the object of perception per se is not subject to misperception, while the object per accidens is. The sorts of error involved here includes misperception of what the object of perception per accidens is (De Anima 4l8a15), and illusion (De Somniis 460b 19). Thus, here, the object of perception per accidens that it is perceived as is not in fact the same as the object of perception per accidens. Given that emotion is held to be responsible for misperception, its means of influence is through the expectation and consequent putting together of a given object of perception. Let me illustrate these points. Suppose we have two people. George is swept by fear; Harry is exceedingly calm 10 As published in Phronesis (1982) and confident. Both hear a loud sound (object per se); George hears a gun firing (object per accidens); Harry hears the backfiring of a car (object per accidens). Suppose, in fact, a car did make the loud sound. Then we say that through his fear George is expecting (dokein, De Somniis 460b6) fearful events to occur. The object per se can be taken for the firing of a gun; and, through the expectation, the loud sound is heard, though misheard, by being put together (suntithemenōn, ibid, 460b13) as a gun firing. The emotion involves certain expectations by which what the object of perception per accidens is perceived as can be put together erroneously. Suppose now it was a gun firing. What we say of Harry is that he is not expecting anything untoward, that his confidence precludes any such thing. Hearing the loud sound (object per se), he puts it together differently — here misperceives — and this is what he hears. Turning to the example in De Somniis, we find that even where the resemblance is very faint, the coward is meant, through expectation, to put together an enemy that is not there as an object of perception per accidens. Turning to the examples of the Nicomachean Ethics, the servant hears a sound, is expecting something, and through the haste and warmth of the emotion puts it together accordingly, thereby misperceiving. Without further ado, he springs to action. Likewise, the dog hears a sound, is expecting some evil, and through the warmth and hastiness of the emotion puts together the knock of a friend as that of an enemy. Without further ado, it springs to action. Thus the distinction in De Anima helps to explain the suggestion of both De Somniis and the Nicomachean Ethics. Moreover, we can understand why Aristotle says in this latter work that the servant hears but does not hear an order (1l49a31). For he hears the object per se, the sound, but through his emotion he is expecting something, and does not hear (does not put together) the object per accidens, the order issued. Thus far, we can explain how Aristotle takes misperception to arise through emotion. That the emotion controlling us is seen to predispose us to see things in terms of it through expectation, is, I believe, a plausible suggestion on Aristotle’s part: emotion is supposed to be part of our way of viewing the world. Our way of viewing the world is the way we put things 11 As published in Phronesis (1982) together, and thus brings about an alteration of perception. That the emotion controlling someone affects his perception is a good part of the explanation of how a person is seduced to dissent from what is, for others, unambiguous. What it is perceived as (per accidens) differs for one moved by a particular emotion. The judgments based on this perception would be askew correspondingly. But this does not yet seem adequate to explain how, say, the lover gets all wrong what is plain to others, and never catches on. Of course, one should maintain that he constantly misconstrues, so long as the emotion is present. This must be part of the answer. But we should also recall the warmth and hastiness both of the dog and the servant mentioned in the Nicomachean Ethics. That warmth and hastiness helped to explain that and how he mishears what he hears. In addition, it helps to explain why neither the dog nor the servant takes in all the relevant information. Rather, they spring to action. Remember, the wary dog hears but does not look to see, so immediate is its reaction. Likewise, the amorous .person hears a little but through his emotion mishears, and springs to action (here, misjudges). But he has not listened to all the evidence, fastening on to some only, and that misheard and misjudged. All the rest he judges in terms of it. There can, therefore, be a variety of reasons why the lover seems to be able to misjudge even in light of what appears to be insurmountable evidence to the contrary. What he takes in, he misconstrues. To the extent he continues to take in, he continues to misconstrue. Through his warmth and hastiness, and the expectation of emotion, he stops considering further evidence, and instead views the entire matter in terms of what he has already taken in and determined. This completes the fourth explanation of emotion altering judgments — though one should add that it may well be augmented by any of those methods discussed so far, the favor/disfavor method seeming very likely to be involved here.11 What spurred us on to search for a fourth explanation has been found: we have an explanation that does not rely on emotion having an aim or goal, and one that can account for the seduction of judgment though the evidence to the contrary is clear enough. Rather 12 As published in Phronesis (1982) than rely on the end of the emotion, this explanation relies on the emotion having a certain expectation and a person’s putting together what something is perceived as per accidens in light of this. This may be plausibly said of all emotions; and thus this explanation applies most generally. 5. Although the need to search for additional explanations has been fulfilled, this does not exhaust all the answers implicit in Aristotle for the ways in which emotion alters judgments. In the characterization of emotion at 1378a20-23, we find that pleasure and pain play a key part. Pleasure and pain can provide us with a further, albeit very general, explanation of the effect of emotion upon judgment. According to De Anima (431a8-lO), as something is painful or pleasant it is avoided or pursued. Thus, the person experiencing a pleasant emotion (e.g. love) will be moved to focus on the matter more than he who is not in a state of pleasure. Contrariwise, the person experiencing a painful emotion (e.g. anger) will be moved to avoid the matter, unlike the person not in a state of pain. Hence, the lover is better able to understand the beloved insofar as the pleasure of his emotion moves him to more attention to the beloved; and the one angry is less able to understand the object of his anger insofar as the distress of his emotion moves him to shun that object. Through attention or its opposite, one’s judgments may be influenced. To this extent, things do not seem the same; and this is a fifth way in which emotion alters judgments. Clearly this is a very general explanation, relating to emotions only insofar as they are pleasant or painful. Although a change in judgment is brought on by the pleasure or the pain of the emotion, the operation of that change is a result of attention or its lack. Moreover, it should be added that other considerations may alter the effect of pleasure and pain. Indeed, in the case of love, insofar as the previous explanations are appropriate, they tend to counteract the force of this explanation. So we should retain a readiness to admit that it could be overridden by other factors. Nevertheless, insofar as one feels pleasure or pain, one has a better or worse opportunity to understand. Insofar as one is so 13 As published in Phronesis (1982) influenced, emotion alters judgment. These different methods answer our question as to how well equipped Aristotle is to explain emotion’s ability to alter judgment. He is very well equipped. Doubtless, there can be other answers consistent with the Aristotelian framework, and some of these we can anticipate. (For example, we might expect an explanation parallel with the fourth, but having expectation and ‘putting together’ alter judgment directly.) Nevertheless, it is these five that are implicit in Aristotle’s works. Their complexity varies from the simplicity of the fifth to the intricacies of the fourth. All, I think, provide plausible solutions to the problem addressed. To this, it should be added, what has been hinted at before, namely that when we come to account for instances of emotions affecting judgment, often we shall find that more than one explanation is involved. They need not be separate. Still, the principles remain distinct; indeed, as the examples illustrate, cases can be imagined in which only one need apply. Thus we end with the following two species of the genus changes of judgment involving emotion: i) change of judgment as a consequence of emotion; ii) change of judgment as a constituent of emotion. The former has the following species: a) connivance, b) seduction through favor and disfavor, c) seduction of perception, d) seduction through pleasure and pain. II Having dealt with the first of our tasks, let us now consider the second aspect of Aristotle’s characterization of emotion: pleasure and pain accompanying emotion. Some points maybe readily stated. The definition of anger holds that it is with pain; contemplating, dwelling upon and achieving its revenge is pleasant (Rhetoric. 1378b1-5, 1370b29); those disposed to be angry are those in pain (ibid, 1379a10-21, cf. De Anima 403al8-20). And one can go on to cite similar information regarding the different pathē. But to do so would not go very far to explain what Aristotle means by pleasure and pain accompanying ta pathē. What could Aristotle mean by saying that emotions are accompanied by pleasure 14 As published in Phronesis (1982) and pain?12 A number of interpretations are possible. This might have the status of an observation of a frequent concurrence, much like ‘mothers accompany their daughters to new schools’. There is no necessity here: it is just that the two often or always concur. Alternatively, it could be a conceptual point. If so, there are at least two ways this might go. ‘Accompanying’ might suggest a link between two separate concepts, ‘accompanying’ relating the two; or the point might be that the concept of emotion includes within it an accompanying pleasure or pain. in addition, we must enquire into the nature of that which accompanies. Is the pain, for example, that accompanies shame an instance of pain of the same kind that accompanies anger? Or is the pain peculiar to shame and of a different sort from that which accompanies anger? The thought that the accompaniment by pleasure and pain with emotion is like mothers and their daughters need not delay us very long. After all, the point is not stated in terms of ’it is often found that’ or ‘it is usually the case that’ or ‘it will be observed that’. Instead, the claim is stated as though a point were being made about the concept of emotion. Moreover, had the point been one of simple concurrence, then one would expect the language to reflect an analogous discussion in Plato’s Philebus, using ‘meta’ only and not ‘hepetai’ to make the point about ta pathē.13 Although we should understand Aristotle to be making a conceptual point, an oddity remains. When contrasting hatred and anger (Rhetoric, 1382a1 1-13), Aristotle goes out of his way to point out that hatred, unlike anger, has no pain. Since the implication is not that hatred is a pleasure, the point must be that hatred is without feeling, cold, accompanied by neither pleasure nor pain. This is definitely out of step with the thesis that the concept of emotion involves pleasures and pain. Mind you, with this oddity noted, we should reject any suggestion that emotion being accompanied by pleasure and pain is anything other than a conceptual point about emotion.14 If ‘accompanying’ introduces a conceptual link, what sort of link is it? Is it a link between two distinct concepts, one always attending the other (perhaps as do cause and effect), or is it that the concept of emotion includes within it pleasure and pain? Now, 15 As published in Phronesis (1982) simply speaking of ‘accompanying’ might suggest the former to us, so that, for example, when the definition of anger is given as a certain sort of longing, the pain could be understood as something necessarily accompanying this longing but itself distinct from the emotion. If true, this would mean that the pain was not part of the emotion, and would not be required in the definition of the emotion — though it would need to be noted as a necessary accompaniment of the emotion. Since we find that the accompanying pain is placed within anger’s definition, Aristotle means more than a necessary accompaniment; emotion includes the pleasure or pain. This conclusion is further confirmed when we observe that many of the emotions are defined as pains or disturbances (e.g. fear, shame). Thus Aristotle includes pleasure and pain within the concept of emotion when he speaks of ‘accompanying’.15 The link between emotion and accompanying pleasure and pain is conceptual. Further, pleasure and pain are part of the emotion. Now, we must ask whether the pain felt in, say fear, is unique to fear, or is it interchangeable with the pain of shame? Do the relevant pains or pleasures differ only in number and intensity? Before dealing with this question we should notice that even if the relevant pains or pleasures do not differ in kind, the absurdity would not follow that if, say, the judgment appropriate to fear was made and at the same time a pain arose (say, in the foot), one would then be afraid. This does not follow because the linking together of the elements in the definition is done in a way stronger than simple concurrence. This, I have argued, is part of the force of ‘accompany’. Moreover, if we look at the definitions of the various emotions, consider fear, we find that the pain is not just conjoined with a particular judgment, but caused by that judgment. ‘Let fear be defined as a pain or troubled feeling caused by the impression of an imminent evil that causes destruction or pain’ (ibid 1382a20-22). Thus, for a variety of reasons, there is no possibility that such an absurdity could follow within the Aristotelian framework. Having dismissed such a misunderstanding, let us turn to our alternatives: the pain or pleasure of an emotion being unique, or the pain or pleasure being different in number 16 As published in Phronesis (1982) and intensity but never in kind. The definitions given in the Rhetoric are plausibly interpreted either way; and within that work I see no reason for confidence that Aristotle holds that there are kinds of pleasures and pains.16 It is tempting, then, to restrict oneself to the modest conclusion: the pains and pleasures of different emotion types differ in number and intensity but not in kind. However, if we expand our horizons somewhat, I think we shall see the stronger position to be Aristotle’s. The Nicomachean Ethics provides reason to think that the pleasure or pain is specific to a given emotion and not shared with other emotions. For this reason, pleasures seem, too, to differ in kind. For things different in kind are, we think, completed by different things (we see this to be true both of natural objects and of things produced by art, e.g. animals, trees, a painting, a sculpture, a house, an implement); and, similarly, we think that activities differing in kind are completed by things differing in kind. Now the activities of thought differ from those of the senses, and both differ among themselves in kind; so therefore do the pleasures that complete them. (1175a22-28, Ross translator) Given that Aristotle goes on to talk about the pleasures of flute playing as opposed to those of argument, and given that the pleasures of the different senses vary, it is reasonable to conclude, concerning emotions, that the pleasure of love differs in kind from that of joy. Likewise, it is reasonable to conclude that the pain of anger differs in kind from that of shame. Thus the pain or pleasure of emotions differ from one to another in number, intensity, and kind. This means that the proper reading of the definitions, again taking fear as our example, is the following: ‘Let fear be defined as a painful feeling caused by…’ rather than ‘Let fear be defined as a painful or troubled feeling, caused by. . .’. The pleasure that accompanies completes the emotion, rather than supervenes upon it.17 We can say the following about the accompaniment of emotion by pleasure and pain. The pleasure or pain is part of the concept of the emotion; neither is separable from the emotion. For each emotion type there is a type of pleasure or pain peculiar to that 17 As published in Phronesis (1982) emotion. They complete the emotion.18 With this observed, it must be recalled that the role of pleasure and pain in emotion is not exhausted by the ‘accompanying’ relationship. As noted already, in addition to the pain or pleasure of the emotion, contemplating and achieving the aim of the emotion (where appropriate) is pleasant, the bodily precondition for the emotion may be pleasant or painful, and so on. III Concerning Aristotle’s characterization of emotions, we have seen how emotions alter judgment, are an alteration of judgment, and what it means to say that emotions are accompanied by pleasure and pain. We now come to our third, and perhaps most difficult task. Throughout, we have spoken of ta pathē as the emotions. That is surely the right translation, given the examples Aristotle offers us. But does his notion match our own? Implicit in Fortenbaugh (in his ‘Aristotle’s Rhetoric on Emotions’) is an answer to this question. Fortenbaugh takes the concern here to be clearly that of the emotions because he believes Aristotle’s characterization of ta pathē is implicitly qualified in terms of the Philebus’s ‘psychic attributes’. Because of this, ta pathē are held to be quite distinct from desires such as hunger and thirst; and all doubts that by ‘ta pathē’ Aristotle has grasped the emotions are dispelled. Although I find this conclusion agreeable, I do not think Fortenbaugh’s argument is adequate. First, there seems to be no reason to be confident that Aristotle’s characterization is so qualified: Aristotle never hints at this. Second, Aristotle does offer a list of what are emotions and takes them to involve the body (De Anima 403al6-19). Thus the suggestion that ta pathē are distinct from desires because Aristotle thinks that emotions are ‘psychic’ rather than ‘bodily’ is not an accurate portrayal of Aristotle’s position. Again, when it is recalled that Aristotle often does include bodily desire (epithumia) in with ta pathē (cf. note 2), Fortenbaugh’s proposal becomes more and more doubtful. Indeed, even if it were clear that the Philebus’s qualification was intended, and we did not have to worry about epithumia or Aristotle’s 18 As published in Phronesis (1982) claim of a bodily nature for emotion, it still would not be evident that ta pathē are the emotions. For included in the Philebus’s psychic attributes are (yearning) and erōs (sexual desire), 47e 1-2. As is especially clear concerning sexual desire, these are and are seen to be types of desire (cf. Republic 549c6-8, Rhetoric 1385a24, Nicomachean Ethics 1118b8ff.). Hence, even were Aristotle to be building upon Plato, it does not seem that this itself is to grasp the notion of emotion. Consequently, if Aristotle has grasped the notion, this is not to follow old blue prints, but to redraw the boundaries within the human soul. My own position is that Aristotle is redrawing boundaries. Yet that he is doing so requires justification; how and why he is doing so requires explanation. To resolve these matters, we should begin by turning our attention back to the Rhetoric’s characterization of ta pathē, examining how the ‘accompaniment’ of hedonē and lupē sharpen and refine this notion.19 This will lead us to consider other ways in which the Rhetoric hones ta pathē . Examining these matters should provide insight into what Aristotle understands by ‘ta pathē’ here and elsewhere, how well it matches our own notion of emotion, and how it contrasts with Aristotle’s notion of desire. Let us begin with our concepts. Were we to try to set forth all the elements of our ‘inner life’, we would wind up with an extensive list, including yearnings, moods, thoughts, wants, perceptions, pleasures, satisfactions, hankerings, and so forth. Obviously, a complete list would fill pages, but I think the following distinctions will serve here to mark off major areas: 1. sensations, 2. desires, 3. emotions, 4. thoughts, 5. perceptions, 6. attitudes, 7. pleasure and pain. We shall examine the first six in light of the seventh. The object is to see what work the accompaniment of pleasure and pain accomplishes. This, I think, will provide some insight into Aristotle’s notion of ta pathē with its accompaniment by hedonē and lupē. It seems to be the case that desires and emotions require pleasure and pain in a way sensations, attitudes, perceptions, and thoughts do not. Take thinking. As I think about how best to put my point, the process is neither pleasant nor painful; as I think 19 As published in Phronesis (1982) about a vacation to France, the thought is pleasant. Thus, while pleasure or pain may attend my thoughts, there is no necessity to it.20 Sensation too may be pleasant or painful: the warming of the sun is pleasant; its burning is painful. But sensation need not be either pleasant or painful, just as the sensation of a gurgling stomach or twitching eye is neither pleasant nor painful. Parallel considerations apply to perceptions and attitudes.21 But emotions and desires (and here we are thinking of their occurrent manifestation) do not seem like this. My desire for a drink is something disturbing to me: its ‘satisfaction’ is just that, a pleasure. And this would seem to be so for all desires. Of emotion, it seems that it must in some way involve pleasure or pain. Anger, shame, sadness are themselves painful or distressing; love, joy are pleasant. The pleasure or pain of these is not just coincident, but necessary to the emotion or desire. By speaking of what must involve pleasure and pain, we limit ourselves to emotions and desires. Turning to Aristotle, that is an interesting con sequence. For, likewise, by speaking of hedonē’s and lupē’s ‘accompaniment’, since that accompaniment is understood as a necessary and conceptual claim (cf. section II), Aristotle thereby limits himself to what we call emotions and desires. Moreover, all this seems to fit in with Aristotle’s theorizing. For that Aristotle does take pleasure and pain to do this work concerning emotions is clear both through his claim that ta pathē (which, at least, include the emotions) are accompanied by pleasure and pain, and through his definitions of various emotions as types of pleasure and pain. That he takes pleasure and pain to be central to desire (orexis), and its other face, aversion, we see in De Anima, 431a8-16. What is pleasant is pursued; what is painful is avoided. Again, in the Rhetoric, l385a23-25, we see that desire (orexis) is a discomfort seeking satisfaction. But how far does this get us? We are seeking to understand what sort of notion Aristotle develops with ta pathē in the Rhetoric’s characterization of them. We have looked at the necessary accompaniment of pleasure and pain, finding that through this accompaniment we understand how a realm exhausted by emotions and desires is delimited. Yet the 20 As published in Phronesis (1982) examples mentioned in the Rhetoric’s initial characterization of ta pathē, as well as the examples he goes on to discuss, concern emotions only and not desires. We are, then, puzzled on our own terms as well as on Aristotle’s by this exclusion. There is the possibility that desire’s exclusion from the Rhetoric is just an omission on Aristotle’s part. Here it is worth recalling that lists of ta pathē in other works do include desire (e.g. Nicomachean Ethics 1105b2l). Alternatively, the Rhetoric may develop in ta pathē a notion like that of emotion. But if this is so, the basis upon which Aristotle has excluded desire remains mysterious. These matters may be cleared up somewhat if we inspect Aristotle’s analysis of desire, and try to find its proper place in Aristotle’s psychology.22 I begin by observing that Aristotle’s notion of desire (orexis) is not some one, homogeneous, all encompassing entity, rather it includes: 1. spiritedness, thumos; 2. wish, boulēsis; and 3. appetite, epithumia (De Motu Animalium 700b22, De Anima 4l4b2, Eudemian Ethics l223a25-27).23 The differences between these must be noted; and the exclusion or inclusion of any one of these from the Rhetoric’s notion of to pat he must be considered and explained. The characterization given of thumos is very much like that of anger, orgē. It too seems painful, while the prospect and achievement of its aim is a pleasant thing, revenge (Eudemian Ethics 1229b31, Nicomachean Ethics 11 l6b23-l I 17a9). Indeed, thumos is sometimes offered as an example of a pathos in the Rhetoric. However, this inclusion does not expand the notion of ta pathē beyond that of emotion, since when ‘thumos’ is included, it is included as a synonymous expression for ‘orgē’ (Rhetoric 1378b4, 1379a4). Aristotle is not tempted to include boulēsis as a pathos; and this exclusion is quite appropriate. Pleasure and pain do not seem to characterize the desire. Moreover, boulēsis’s first aim is to on kalon (Metaphysics 1072a27, cf. Rhetoric 1369a2). If it relates to pleasure and pain at all in terms of its aim, it does so incidentally (cf. Topics l46b3, l47a1-4, Rhetoric 1381al-4). Thus boulēsis does not satisfy the pleasure/pain test 21 As published in Phronesis (1982) as emotion does, but, at most, as perception or thought does.24 This is to say that boulēsis does not satisfy Aristotle’s first test for ta pathē. Hence, this type of desire is not to be confused with emotion; and Aristotle has good reason to exclude boulēsis from his list of ta pathē in the Rhetoric and elsewhere (Nicomachean Ethics l105b21, Eudemian Ethics l220b12). One might object that this last argument argues too much. Shame, aischunē, like Boulēsis, does not aim at pleasure or pain, yet remains a pathos. Thus my argument that boulēsis does not aim at pleasure or pain is not a reason for its exclusion as a pathos. However, the cases are dramatically different. For while boulēsis only relates incidentally to pleasure or pain, shame is defined as a pain (Rhetoric, 1383b15). Thus this latter, but not the former, satisfies the pleasure/pain test; and boulēsis has rightfully been excluded as a pathos. So far, so good. Aristotle does seem to be developing a notion like that of emotion. We have seen why neither boulēsis nor thumos create difficulties; we have only epithumia to contend with. Since epithumia is not mentioned in the passage from the Rhetoric we are concerned with, this development seems quite likely. However, we still have to find justification for epithumia’s exclusion. In addition, we need an account of why epithumia is not counted as a pathos here when it has been elsewhere (Nicomachean Ethics I 105b21, Eudemian Ethics l220b 12).25 First, we need look at epithumia. Epithumia (appetite or sensual appetite) is a desire for the pleasant (Rhetoric I 37017, De Anima 4 14a5-6, Topics 147a2); like anger and other painful emotions it is characterized as painful (Nicomachean Ethics 11 19a4). Epithumiai include and are explained in terms of the desires for food, drink, sex (Nicomachean Ethics II l8bSff., cf. De Anima 414b13).26 As something itself unpleasant craving the pleasure of satisfying its lack, movement as a result of appetite is not very mysterious (De Anima 433a25, cf. Rhetoric 1369b1 5, 1379a10-I I). It is taken to be contrary to choice (Nicomachean Ethics 1111b16), a wild beast (Politics l287a31). Its operation occurs without involving reason: ‘But appetite leads without persuading, being 22 As published in Phronesis (1982) devoid of reason’ (Eudemian Ethics l224b2). We have here something well suited to causal analysis. From this characterization, we can see that epithumia’s exclusion (unlike boulēsis) cannot be accounted for through failing to satisfy the pleasure/ pain test. Moreover, there is no indication (as there was concerning thumos) that ‘epithumia’ is but another name for an emotion. What, then, accounts for its absence from the Rhetoric? I suggest that what justifies this desire’s exclusion is the Rhetoric’s other major characterization of ta pathē: emotions being the things on account of which the ones altered differ with respect to their judgments. If I am right in saying that epithumia is excluded because it does not meet this demand, then the Rhetoric delimits what we mean by emotions, offers a justification of this, and advances beyond Plato’s spirited realm. A fascinating and perceptive development. But one that requires some argument before we grant it. Although epithumiai do not seem to be emotions, and the Rhetoric’s exclusion of them as pat he seems to be a recognition of this, we cannot really be sure that it is this until we see that that part of the characterization of ta pathē which speaks of altering judgments does properly exclude epithumiai. I think that it does. There is no need for a difference in judgment between one thirsty and one hungry: they may hold all the same judgments, but the former seeks food while the latter seeks drink. Indeed, being hungry or thirsty does not require the holding of any particular judgments, or any judgments at all. Moreover, it is not itself a reasonable or un reasonable state. As Aristotle suggests, it is devoid of reason (Eudemian Ethics 1224b2). However, as Aristotle recognizes, emotions are rather different from this. Those in a different emotional state do differ with respect to judgment, e.g. whereas the envious man will view another’s good fortune as undeserved, the emulous will not. Being in an emotional state requires judgments, particular judgments. Moreover, it is itself reasonable (fear of a formidable enemy) or unreasonable (fear of a mouse). Thus the two, epithumetic desire and emotion, do seem importantly different with regard to the role of judgments. It would seem, then, that Aristotle has noted the 23 As published in Phronesis (1982) difference, distinguished the realms, and provided justification for this. Still, one might doubt that what Aristotle has set forth is really adequate, even though right-headed. After all, is there not a sense in which an epithumia might bring about a change of judgments? For example, hunger’s pang could make one so irritable that one comes to a very harsh view of someone who interferes with one’s attempt to acquire food. Again, the alcoholic’s thirst may be so strong that the person decides that the wood alcohol is not so bad. However, it is not the hunger or the thirst that alters judgment. For, as we have seen, what these desires do is seek out their own satisfaction. Rather, the difference in judgment that may arise in such situations will arise through one’s anger, irritation, despair, or reflections upon these matters. And Aristotle follows this up by noting that emotions often arise when desire is present. Men are angry when they are pained, because one who is pained aims at something; if then anyone directly opposes him in anything, as for instance, prevents him from drinking when thirsty, or not directly, but seems to be doing just the same; and if anyone goes against him or refuses to assist him, or troubles him in any other way when he is in this frame of mind, he is angry with all such persons. Wherefore the sick, the necessitous, the love-sick, the thirsty, in a word, all who desire something and cannot obtain it, are prone to anger and easily excited… (Rhetoric, 1379a10- translated by Freese) Any change of judgment here is only an incidental result of hunger or thirst, and quite remote from it. Epithumetic desire is not sufficiently complex to speak of it as altering judgments. The changes of judgments are to be explained by emotions or reflections upon these matters. Aristotle is right in thinking that emotions are quite different from these desires, epithumiai; and he is able to locate just what accounts for the difference. Like epithumetic desire emotions too have an object, involve pleasure and pain, and through this latter are involved in pursuit and avoidance. However, in addition, emotions have a much more wide-ranging aim. Through expectation, they alter the way we put things 24 As published in Phronesis (1982) together. Moreover, they require judgments. And because of this emotions are themselves alterations of judgments (anger views its object as having insulted one, Rhetoric 1378a31), and alter judgments (hope leads to a better view of one’s prospects, ibid 1378al-4). In contrast, because epithumia has only the satisfaction of eating, drinking, etc., what counts as satisfaction here is much more restricted; and it will not involve changes of judgment. That Aristotle excludes epithumia from the list in the Rhetoric is justified; epithuinia as an emotion does not belong.27 The thesis that Aristotle is delimiting a realm of emotions finds further confirmation, For not only does Aristotle recognize the difference between epithumiai and emotions, but also he utilizes this difference. Aristotle discusses the nature of epithumia, making the point that it is not subject to rational principle (is not reasonable or unreasonable). Moreover, when strong and violent it can expel the power of calculation (Nicomachean Ethics 1119b5-15). Thus the angry man may reason poorly in deciding to wreak a terrible vengeance; the man of unquenchable thirst does not reason at all, but simply seeks the object of his desire. This role for epithumia is utilized elsewhere. For example, in a discussion of incontinence Aristotle’s position is not that the desire for the sweet alters one’s universal opinion forbidding tasting. Rather, one follows one’s desire to taste and loses sight of the universal opinion. Once more, desire seems to expel rather than alter reasoning (Nicomachean Ethics 1l47a25-b 17). All this is rather different from the way, say, hope brings about a favorable interpretation of what is ambiguous or envy views the good fortune of another as undeserved. Yet another utilization of the difference between epithumia and emotion has to do with the obedience of emotion, but not epithumia, to reason. If emotions are the sorts of things that rationally alter our judgments, one can expect them to be open to reason. Similarly, if epithumia does not rationally alter judgments, one would not expect it to be open to reason. Aristotle appreciates this when he says: Therefore anger obeys the argument in a sense, but appetite does not. It is 25 As published in Phronesis (1982) therefore more disgraceful; for the man who is incontinent in respect of anger is in a sense conquered by argument, while the other is conquered by appetite and not argument. (Nicomachean Ethics, I 149bl-4, Ross’ translation, cf. 1119b7) This contrast is fairly drawn between epithumia and emotions in general. Thus while you might convince a person not to act on his epithumia, say, for food, you cannot talk him out of feeling hungry. Hence, we find Aristotle observing: ‘ . . . it is assumed that there is no gain in being persuaded not to be hot or in pain or hungry or the like, since we shall experience these feelings none the less’ (Nicomachean Ethics 11 13b27-30). In contrast, not only might you convince a person not to act on his emotion, say fear, but also you might talk him right out of it. This latter you might do by convincing him that one of his judgments whence his fear arose was wrong, or you might convince him that even though all is as he judged, the object he fears is not worth fearing. And by convincing him you also move him.28 The contrast between the two is that while we give grounds for emotions, we only give causes for thirst and other epithumiai. Thus the former, but not the latter, is, in this sense, conquered by argument. Thus it is the former, but not the latter, that Aristotle concerns himself with and explains the grounds upon which they are felt (Rhetoric 1378a28). In view of the interaction between the rational soul and desire, we must digress to notice that the contrast between epithumetic desire and emotion becomes more complicated in certain instances. Epithumetic desires, we have seen, are the sorts of things that get set in motion, halted, stemmed, suppressed, expelled, etc. The causal chains for any particular desire can be quite diverse. Consider sexual desire. Gestures, clothing, movement, glances, pictures — all these may serve to ‘turn one on or off’. A causal conception is in operation here in a clear-cut way. However, it may seem a little less clearcut when we consider that reading certain passages from novels may have the same effect. For here it seems as though epithumetic desire is available to reason. However, there are two objections to this conclusion. First, the case remains one of being ‘turned on’ or ‘off’, of causation. Although reading the novel may dampen or arouse 26 As published in Phronesis (1982) one’s ardour, still one has been turned off or on, shocked or titillated. One has not been reasoned into anything or persuaded, in the way one may be moved to anger by being persuaded that Fred has insulted you or by deliberating upon Fred’s character. So whereas emotion admits of rational persuasion, epithumetic desire still is not available to reason. The complication here has been that, as an animal capable of reason, the means of turning on and off epithumetic desire are that much richer, involving the rational soul, but still not in a way to be confused with emotion’s involvement with the rational soul. Second, to the extent one still wants to say: ‘No, my desire really has been rationally altered here’, that we can quite happily accommodate by the operation of boulēsis, not epithumia. That is to say, in the example above, we have not only epithumetic desire in operation, but also rational desire. For that the desires are distinct has no implication about forced separability or lack of interplay amongst them. And it is possible that deliberative desire could enter into the picture here.29 Thus the contrast between epithumiaa and emotion stands.30 Hence, not only is the exclusion of epithumia and all desire from the list in the Rhetoric reasonable, but also the implications are appreciated and utilized elsewhere. What I have just argued is that in the Rhetoric and elsewhere Aristotle shows a perceptive awareness of the differences of operation of epithumetic desire and emotion. Before that I argued that other sorts of desire (thumos and boulēsis) do not interfere with the suggestion that Aristotle is delimiting the realm of emotion in the Rhetoric. My conclusion is that the characterization of ta pathē in the Rhetoric distinguishes emotion from other elements of our inner life: the pleasure/pain test setting emotion and certain desires quite apart from the other elements, the alteration of judgment setting emotion quite apart from epithumia.31 That the Rhetoric does not mention or expand upon the pathos epithumia is not an oversight or error, but a recognition that epithumia is not an emotion.32 Setting forth the notion of emotion is a sophisticated advance within the realm of philosophical psychology. However, at least one problem lingers. Why is epithumia here 27 As published in Phronesis (1982) excluded from the list of ta pathē, while elsewhere included? Historical explanations are often employed in this sort of situation, arguing, for example, that here Aristotle abandons the Platonic psychology that mesmerized him elsewhere. However, we cannot be certain that the Rhetoric is Aristotle’s last word in this area of psychology; and since Aristotle utilizes these distinctions at some points in his ethical works (see above), but does not utilize them at other points (Nicomachean Ethics 11l05b21, Eudemian Ethics 1l20b12), an historical explanation cannot resolve this problem. We must search for some other sort of explanation. There are a number of possibilities. The most radical one suggests that the picture of ta pathē that has emerged is all wrong. In the Rhetoric Aristotle simply chose not to use epithumia as an example; and we have made a mountain from what is not even a mole hill. But too much has been gained; there is too much rigor, too much perceptiveness, too much following out of consequences on Aristotle’s part for this explanation to be seriously entertained. A different explanation urges that sometimes Aristotle wrongly includes epithumia (Nicomachean Ethics I 105b21, Eudemian Ethics l220b There are the differences noted between epithumia and ta pathē; he is aware of them; yet his inclusion of epithumia in the lists of ta pathē in the ethical works is a lapse, a failure to appreciate fully and mark out adequately what he does elsewhere. Alternatively, one can suggest that Aristotle is driving at a slightly different point than our analysis of the Rhetoric suggests. What he really wants to do is to note a group of things that a) relate to pleasure and pain, and b) ‘in one way or another’, however remotely, alter judgment. These all ta pathē do, including epithumia. Epithumia’s absence from the Rhetoric is just a failure to list fully. That we find very important differences between the ‘one way’ and the ‘other’ is interesting and important to us, but does not signify for Aristotle’s analysis of ta pathē. He may be dividing the cake differently from us, but not therefore mistakenly. These two approaches are not really that far apart. The latter tries to claim that Aristotle’s conceptions when brought forth on his own terms are somewhat different from our own — though it admits that at certain points he does draw the contrasts as we do. 28 As published in Phronesis (1982) The former views Aristotle in terms of distinctions we make (accusing him, in parallel and quite important passages in his ethical works, of failing to appreciate adequately what he at other times takes to be important). Neither of these ways of resolving the matter is as satisfactory as we might want. Both interpretations make Aristotle’s analysis of ta pathē broken backed and admit that he should have been aware of the broken nature. The former view’s contention that in the ethical works Aristotle is simply guilty of a glaring error through his inclusion of epithumia as a pathos is difficult to believe. Equally difficult to believe is the latter view’s contention that though Aristotle is aware that emotions as such alter judgments, while epithumetic desires do not, he nevertheless ignores this in the Rhetoric opting for an ‘in one way or another’ — especially since epithumia is not listed or discussed there. Also difficult to believe on the latter interpretation is Aristotle’s silence about his own thesis that emotions but not epithumiai are altered by reason. In addition, this interpretation by supposing epithumia as a legitimate candidate for a pathos in the Rhetoric fails to appreciate that whereas the pathē Aristotle does mention and discuss do have a ‘with whom’, epithumia does not. Thus neither of these explanations can be accepted. Moreover, when we look at the much greater scope given to ta pathē in De Anima 403a1-7, 17-19, and the Categories 9b910a10, we realize that something different might be occurring than has been suggested so far. The clue to a more satisfactory explanation (not without its own difficulties) is found in the last observation. Rather than trying to find a unified or developing (though broken backed) theory of ta pathē, let us look carefully at the different contexts in which Aristotle deals with ta pathē; and let us allow that Aristotle’s use of ‘ta pathē’ may vary in extension and intension with the purposes at hand. We shall focus on the relevant discussions in the Rhetoric, Eudemian Ethics, and Nicomachean Ethics. We have seen that the two criteria present in the Rhetoric distinguish in an insightful way the emotions from the other elements of one’s inner life. The examples Aristotle chooses, develops, and excludes bear this out. Turning to the lists of ta pathē in 29 As published in Phronesis (1982) the ethical works, we find similar lists, though epithumia is included. Hence these are not lists of the emotions. Moreover, we should notice that though we do find the pleasure/pain test, we do not find anything about altering judgments.33 Consequently, the lists with their inclusion of epithumia match perfectly with the single pleasure/ pain test. Viewed in this way, Aristotle seems to wield the two principles with great sensitivity to their implications both in the Rhetoric, the Eudemian Ethics, and the Nicomachean Ethics. Well and good, we might think, but still it remains puzzling in its way. Why does Aristotle speak of ta pathē in these similar, but importantly different ways? Why not stick with one, preferably the most subtle? These differences in intension and extension can be explained, I believe, by noticing the issues Aristotle is addressing at a given time. At the appropriate places in the ethical works, Aristotle is trying to discover where virtue lies. The alternatives he offers are: pathē, dunameis, hexeis. In light of these contrasts and the goals sought, it seems quite reasonable that ta pathē should include more than emotion. The distinction between epithumetic desire and emotion does not matter to his ongoing discussion. Whether ta pathē are subject to reason will not matter to the discovery that virtue is a hexis. Indeed, given that Aristotle wants to hold that virtue concerns the pathē, he means it to concern epithumetic desire as well as the emotions. For virtue concerns those occurrent rumblings which may lead us astray, whether they be rumblings subject to reason or not.34 Hence, Aristotle does not bother about the second criterion; and epithumia is rightly included. Here ‘ta pathē resemble what Hume and others call ‘the passions’. However, when Aristotle’s purposes are different when he is trying to offer a theory of those affections relevant to rhetorical purposes, when he is trying to avoid the Platonic tendency of seeing rhetoric as sophistical, and when, as in certain pans of the ethical works, he is trying to illustrate the differences between epithumia and anger — then the differences between things that do and do not influence judgment, are and are not influenced by judgment is crucial. The Rhetoric’s interest in ta pathē has to do with persuasion and as a result 30 As published in Phronesis (1982) Aristotle sharpens the notion to those things that do affect judgment. Thus Aristotle excludes epithumia which does not similarly affect judgment. Moreover, this explains the introduction of his second criterion, a criterion not introduced elsewhere.35 Aristotle does not hold a broken backed theory with all its awkwardnesses. Moreover, we appreciate how skilfully Aristotle uses the different senses of ‘ta pathē. Where he is concerned to speak of the role of judgments concerning affection, he adequately gives the notion of ta pathē as emotion. Where his interest is not so specific, he includes epithumia in with emotions, but there correctly excludes the judgment criterion. And where his concern, as in De Anima, is with any affection of the soul, he properly drops the pleasure/pain criterion. In all these cases the theory is adequate, skilful, and is not subject to the above complaints.36 This completes our third task. If the arguments are right, the consequences are impressive. In the Rhetoric Aristotle develops a notion of emotion to which he turns elsewhere. As well as coming to this notion, he isolates those features that set emotion apart from other elements of the human soul. We have come to see what it means to say that ta pathē are accompanied by hedonē and lupē, as well as how these help to refine the notion of emotion. We have come to see the ways in which ta pathē can alter judgment, as well as how this also helps to refine the notion of emotion. In addition, we have seen that Aristotle is quite able to call upon the notion of emotion when needed, and related notions when they are needed. By this ability to wield the features that distinguish these notions, by his sensitivity to the different notions and their place, we see an extremely subtle philosopher at work. 31 As published in Phronesis (1982) NOTES 1 I should like to take this opportunity to thank Professor J. L. Ackrill, J. Barnes, D. Browning, L. Judson, P. Mitchell, and the Euthyphrones Discussion Group for their criticisms and suggestions. Where I have not profited as I should the fault is mine alone. 2 One might think that the concern is so obviously that of emotion that this question hardly bears investigation. Given the examples he offers, given that ta pathē? are meant to be occurrent phenomena, given that ‘the emotions’ is a reasonable translation of ‘ta pathē’, Aristotle has surely grasped the notion of emotion here. However, we need to be a little more cautious before drawing this conclusion. For there are more occurrent phenomena than emotions: and Aristotle often includes as a pathos epithumia, a type of desire which includes hunger and thirst (Nicomachean Ethics 1105b21, Eudemian Ethics 1120b12). This plus doubts that Plato ever clearly distinguishes emotion and desire should lead us to take this question very seriously. Should it turn out that Aristotle does develop the notion of emotion, he has redrawn psychic boundaries in a very insightful way. 3 How dramatic is very apparent in the case of Ergophilus. Concerning the exhausting of anger and consequent growing mild, Aristotle observes: For although the Athenians were more indignant with him than with Callisthenes, they aquitted him, because they had condemned Callisthenes to death on the previous day. (Rhetoric 1380b11-14, Freese translator) 4 In his stated definitions of pity and indignation an aim is not explicitly announced. Rather, it is part of the larger concept of these emotions. We find the same thing in envy and emulation. Part of the concept of envy involves preventing one’s neighbour from possessing certain goods, while emulation strives to make oneself fit for such goods (Rhetoric l388a35-37). From these aims, which are part of the concept of emotion (but not part of their stated definitions), we still can explain certain changes of judgment. 5 It is plausible that there will be certain cases in which there is no such discrepancy. Someone might forget what he did believe and so come to be persuaded by his own pronouncement. 6 The alteration is a seduction unless one is simultaneously aware of the presence And workings of the disposition and desire. Such awareness is not typical, though it is certainly possible. 7 I assume that thumos is meant to be an emotion. This seems to accord with most translations and with the Rhetoric, in which Aristotle happily switches from thumos to orgē. This point, I comment on later. 8 If J. Cooper is right (in an unpublished paper “Aristotle on the Ontology of the Senses”), then contrary to Hamlyn’s translation of De Anima, ‘krinein’ means ‘distinguish’, not ‘judge’; and the perceiving of both types of objects is, properly 32 As published in Phronesis (1982) speaking, a matter of perception. 9 This is a variation of an example of Cooper’s. 10 Concerning the plausibility of the thesis, modern theorists would be inclined to reject different types of objects of perception, speaking instead of differences in perception. However one chooses to characterize the difference, there is here an additional, distinct way in which emotion alters judgment. 11 Under the favor/disfavor case I include what is objectively ambiguous. In the case of misperception I have spoken of what is not itself ambiguous. An interesting case is one in which something is not itself ambiguous, but seems so due to carelessness or inattention. This sounds very much like the case of the hasty servant. When the carelessness concerns perception, it is. Ho where something seems ambiguous through inattention in evaluation, then we have a second version of the favor/disfavor case: one version explicable by the ambiguity of the phenomena; the other explicable by ambiguity arising through inattention. 12 K. J. J. Hintikka, “On the Interpretation of’De Interpretatione’ Xll-XIll”, in his Time and Necessity (Oxford, 1973), pp. 53-5, speaks of the meaning of ‘hepesthai’. Unfortunately, his conclusions are meant to be restricted to that text, and will not help us here. 13 The problems that arise in the Philebus through using ‘meta’ to explain the place of pleasure and pain, and Aristotle’s appreciation of this in his Topics with respect to the emotions is nicely illustrated by W. W. Fortenbaugh, “Aristotle’s Rhetoric on Emotions”, AGP 52, 1970, 40-70, at pp. 55-6. 14 One attempt to resolve this apparent anomaly would be to observe that hatred Should take pleasure in the destruction of the hated. Even if true, it is to be explained by the fact that contemplating and achieving one’s aim is pleasant (Rhetoric l370b29, l378b It is no more a matter of hatred being a pleasure than the sweetness of anger’s revenge is a matter of anger being-a pleasure. I suspect that part of the reason for this anomaly is that the description of hatred in the Rhetoric is similar to what is elsewhere called a hexis. Since a hexis has more a dispositional than occurrent tone, the need to speak of pleasure or pain is that much weaker. But this is only partially satisfying. For hatred remains classed as a pathos. 15 If this is right, then although we find Aristotle using ‘meta’ in his definition of anger, the ‘hepetai’ controls the ‘meta’. The accompaniment of pleasure and pain does not suddenly become contingent here. 16 That being pained disposes one to emotion (Rhetoric, 1379al0- and that the point seems to be about pain in general, rather than a matching between a certain sort of pain predisposing one and a corresponding emotion disposed to might suggest that there are not kinds of pains and pleasures in emotion. However, that the pains that predispose one do not divide into kinds is no reason to doubt that the pain of the emotion does so divide. After all, the pain of anger is not the pain in one’s tooth that has disposed one to anger. 33 As published in Phronesis (1982) 17 Actually, Aristotle does not offer a full analysis of pain in the Nicomachean Ethics. But he often considers pleasure and pain in terms of health and disease. If pain is like disease, then it is a privation of pleasure; and as a divergence from a pleasant condition, separating pains into kinds becomes messy. More serious problems in applying an analysis of pain implicit in the Nicomachean Ethics to emotion occur insofar as a pleasure proper to each activity would imply that the activity of being angry or being ashamed would be a pleasure. This is both absurd and contrary to Aristotle’s analysis of these emotions. Hence, the analysis of pain in the Nicomachean Ethics is unsuited in some respects to account for painful emotions. The problem could be resolved by giving pain its own character (not simply a privation). and admitting that pains complete certain activities. Thus pain would as much complete anger as pleasure completes love. However, to the extent that the Nicomachean Ethics offers an account of pain, this is not it (but see note 20). 18 It is not the case that the completion in the case of flute playing or argument is just like the completion in the case of the emotions. For while flute playing can occur without being completed, the emotions do not. Anger is not anger unless it is painful. 19 To avoid confusion I will use Greek terms for Aristotle’s concepts, and English Terms for modern concepts. 20 One might object that by Aristotle’s analysis of hedonē and lupē in the Nicomachean Ethics, this claim could not be made. For any unhindered activity should be pleasant, including thinking. Thus an attempt to see what is behind the notion of ta pathē in the Rhetoric is doomed if one continues in this way. However, we have already seen that some thoughts on pleasure and pain in the Nicomachean Ethics are out of step with the analysis in the Rhetoric. Thus I am not assuming Aristotle to be bound in every detail to the theory in the Nicomachean Ethics: I am allowing that in thinking out a different problem Aristotle might not depend upon or be loyal to some of his conclusions elsewhere. This may be to skate on rather thin ice, but it is not unusual for Aristotle to forgo theoretical consistency for observations closer to the truth. Moreover, there is evidence to suggest that some of the thoughts on pleasure and pain are different in the Rhetoric. Many of the emotions are defined as types of pain. It seems implausible that by this each is meant to be a lack of something. A lack of what? Thus the disease model is inappropriate here. Pain seems to be understood as having a character of its own; and that is why it is sufficient for the definitions of the various emotions. Not every unhindered activity is pleasant. Elsewhere (Eudemian Ethics l220b 13), Aristotle speaks of perceptible pleasure and pain. Here too the disease model is unlikely. Thus we can expect some unhindered activities to be pleasant, others to be painful, others still to be neither pleasant nor painful. 21 At one point in De Anima, 413b23, Aristotle speaks of aisthēsis, including pleasure and pain (cf. De Sensu 436a8-11). This does not disturb my thesis, since I take his point there to be that where we speak of aisthēsis, the possibility of pleasure or pain is introduced, 34 As published in Phronesis (1982) and not that aisthēsis must be pleasant or painful, i.e. not that it must be accompanied by pleasure or pain. That this is the right way to interpret Aristotle is suggested by the erroneous nature of the alternative interpretation. It finds further confirmation in the fact that where Aristotle speaks of ta pathē and explicates this with ‘accompanying pleasure and pain’, he does not introduce aisthēsis as an example (Rhetoric 1378a20, Eudemian Ethics 1220b12, Nicomachean Ethics I 105b2l), whereas when ta pathē have been expanded and aisthēsis is included, the claim of an accompaniment by hedonē and lupē is dropped (De Anima 403a1-7). 22 That desire is not an emotion may need some argument. Its inclusion is counterintuitive: and I shall advance arguments shortly, one consequence of which is to distinguish emotion from desire. 23 For a somewhat different analysis of the desires than the one to follow see M. C. Nussbaum’s edition of De Motu Animalium, pp. 334-337. 24 Aristotle’s earlier claim that desire is a discomfort seeking satisfaction is, effectively, modified in the case of bouēhsis. This, in part, is an appreciation of the point that the intellectual desires do not run along the same lines of distress and pleasure in the way bodily desires do. 25 It is noteworthy that Fortenbaugh supposes that ‘ta pathē in the passages from the Ethics means ‘the emotions’ (cf. “Aristotle and the Questionable Mean-Dispositions”, TAPA 99, 1968, 203-31, at p. Here too the reference to the ‘psychic attributes’ from Plato’s Philebus is thought to be implicit. Earlier, I suggested, concerning the Rhetoric, that this reference to the Philebus was both questionable, and, if true, still does not provide us with the notion of emotion. Thus, I argued that care is needed when claiming that the concern of the Rhetoric was that of emotions. These considerations apply to the passages from the Ethics as well. More importantly, the inclusion of epithumia (which for Aristotle is to include desires such as hunger and thirst) in the Ethics precludes the idea that here Aristotle is implicitly referring to the Philebus’s ‘psychic attributes’ and bars the claim that by ‘ta pathē ’ in the relevant passage from the Ethics Aristotle means ‘the emotions’. What it does mean, we shall see shortly. It must be emphasized that this dispute about the Ethics (and the Rhetoric) cannot be dismissed as ‘quibbling’. For, as Fortenbaugh himself is keen to show, there is a world of difference for Aristotle between the operation and nature of shame or fear versus hunger or thirst. Where Aristotle includes or excludes these latter is significant. 26 I am not here concerned to compare and contrast each sort of desire. However, I would like to emphasize one point of contrast between epithumia and boulēsis. While epithumia aims at the pleasure of food or drink, boulēsis may take pleasure in achieving its aim (to on kalon) but does not act for the sake of such pleasure (cf. Eudemian Ethics 1235b1924). 27 A quite different and less central consideration for the distinction between epithumia 35 As published in Phronesis (1982) and ta pathē arises when we consider one of the headings under which ta pathē are analyzed, the person before whom one typically feels the pathos (Rhetoric 1378a24). Ta pathē seem to involve one with others: the person loved, hated, angry with, ashamed before, etc. Epithumia is not like this. Thirst is not bound up with others, but with the seeking of drink. So, similarly, hunger, and the desires of the senses. Erotic desire seems out of tune with this insofar as the object typically is another person. Short of withdrawing the point of contrast, one might urge that erotic desire can be satisfied without the existence of another, but philia cannot. Second, there need be no social involvement with another in the case of erōs while there is with philia and other emotions. Thus while ta pathē require others, epithumia does not. 28 This latter would be a matter of convincing a person to change his values. So By converting to Buddhism one might lose one’s fear of dying. It is important that emotions only ‘listen to argument to some extent’ (Nicomachean Ethics 1149a25, cf. De Memoria et Reminiscentia 453a25-30). The thesis is not so strong that the relevant change about the facts or values is or forces a change of emotion. Rather, emotion is available to reason. Aristotle leaves room for what we call irrational emotions, be they so from lack of foundation in the first place (fear of a mouse) or loss of a foundation. Hence, to convince is not necessarily to move. (Here we have a further contrast between boulēsis and ta pathē. In addition to failing to satisfy the pleasure/pain test, boulēsis does not just listen to some extent.) 29 That boulēsis is available to reason is not in question. Its exclusion from the realm of emotion has been accounted for on other grounds. It is also worth emphasizing here that the attempt has not been to say that Aristotle’s distinctions within desire match our own. Rather, the attempt has been to say that Aristotle’s characterization of ta pathē excludes desire; and ta pathē matches our notion of emotion. 30 But is not there still a sense in which one can and does speak of having ‘reasonable appetites’? Yes, but this sense is the following: one’s appetites are well brought up so that what they desire is in conformity with rational principle. Unlike rational principle or emotions, epithumia is not itself rational, but spoken of so only insofar as it happens to conform to logos. The truly virtuous have such epithumiai; the continent and incontinent do not. As a result, these latter have to control their epithumiai, though, as we have seen, sometimes epithumia will expel any reasoning present. 31 That part of thumos which is not to be seen as equivalent to orgē is excluded from the realm of emotion by this second test. Moreover, were one dissatisfied with the exclusion of aisthēsis (cf. note 21) its exclusion from ta pathē is supported by this test. 32 There are two spots in Aristotle that might present difficulties for this understanding of desire, and consequently the distinction between it and emotion. First, Rhetoric 1370al9-25 distinguishes desire into rational and irrational desire, instead of the typical triad. The rational desires seem to be more sophisticated. Such desires do not present serious problems for my analysis. For though this is a different way of examining desire, it can 36 As published in Phronesis (1982) be dealt with in much the same fashion I dealt with the sophistication within erotic desire, and the sense of ‘reasonable appetites’ spoken of in note 30. Second, a discussion of the soul in the Nicomachean Ethics (book one, chapter thirteen) may seem to present problems. There Aristotle talks of hormai, impulses. These seem to be available to reason, yet they do seem to be boulēsis. Still, this is not too troublesome. They are said to be reasonable and listen to reason as does the son to the father. Now, this seems to be a matter of a certain sort of habituation. If so, this is a ‘reasonable’ appetite of the sort mentioned in note 30. Moreover, the discussion of hormai is unique and very difficult to square with the earlier discussion of the soul in chapter nine! as well as with the discussions in De Anima. So, at worst, this passage can be dealt with as a matter of Aristotle wandering from his normal path. Most importantly, Aristotle makes it perfectly clear that this discussion lacks precision (1102a22-3). That Aristotle himself does not take this way of dividing the soul too seriously means that we need not be bothered if it conflicts with more serious attempts to understand distinctions within the human soul. It is these latter that are the important ones. 33 The absence of this criterion is further evidence that Aristotle is up to something very different in the Ethics than he is in the Rhetoric; and that ‘ta pathē’ in the Ethics cannot be ‘the emotions’. 34 This is another reason why Fortenbaugh cannot be right in his understanding of ‘ta pathē’’ from the relevant passages in the Ethics (cf. his “Aristotle, Virtue and Emotion”, Arethusa 2, 1969, 163-85, esp. note 24). Virtue is not just a preparation and control regarding emotions and actions, but also epithumiai. The inclusion of these desires is more in the Aristotelian spirit. For, as we observed in note 30, the man of perfect virtue has trained his desires so as to be moderate in them, while the continent man is not moderate in them, but has control over them. Indeed, Aristotle’s whole picture of moral education has to do with the training of the emotions and desires. Not only does this way of understanding what ta pathē are in the Ethics create a more Aristotelian view, but also a more accurate one. For the inclusion of the control of one’s desires seems to help create a better description of virtue’s place in our moral life. 35 I shall not here deal with the passages from the Categories or De Anima in detail. They are interesting, but a full analysis would take us far afield; and would not help us with the issues here. Let me only say that a similar approach to these passages will explain the use of ‘ta pathē’ there. The general direction would seem to be the following. In De Anima the concern is whether attributes of the soul involve the body. ‘Ta pathē is used to collect these attributes; and hence the list is much expanded from any so far examined (ibid, 403 A second list 403a17-19 more closely resembles that of the Rhetoric, because emotions are more obviously bodily than perception, thought and other such attributes mentioned in the first list. But it must be said of both lists that the remarks are problem initiating rather than problem-solving. Hence Aristotle uses a very non-technical and nonrefined sense of ‘ta pathē.’ Indeed, he provides no criterion for them. In the 37 As published in Phronesis (1982) Categories a general interest in ta pathē brings Aristotle to speak of those of the soul. These seem to be occurrent rather than dispositional features; and their temporary nature is featured. Again, a rather non-technical conception is in use. Aristotle is roughly mapping the area, rather than sharpening a philosophical tool with which to resolve a particular problem. As we have seen, matters are rather different in the Rhetoric and the Ethics. 36 Although this does solve our problems, I mentioned that even this proposal has a difficulty. In the Rhetoric (1388b33), having completed his analysis of ta pathē Aristotle reviews his progress. At this point he does include epithumia as a pathos. This runs contrary to Aristotle’s development here. Indeed, given Aristotle’s understanding of epithumia (see above), this inclusion must be seen either as an uncareful moment, or as the destruction of all that Aristotle has sought 10 achieve in his characterization and explanation of ta pathē in the Rhetoric. Thus I would suggest that it be seen as an uncareful moment. I should add that one alternative explanation of what Aristotle has done in the Rhetoric yet remains. Instead of seeing Aristotle as defining a notion of emotion which is distinct from desire, one might suggest that what Aristotle has done is to refine a subset within desire (orexis); the subset is emotion. Evidence for this view would include Aristotle’s inclusion of thumos as one of the key notions of orexis, yet his willingness to understand ‘thumos’ as synonymous with ‘orgē’. Evidence consistent with this includes both the exclusion of epithumia as a pathos, and the inclusion of orexis within the definition of anger. What seems to count against this is that Aristotle does not ever say that a type of desire is a pathos. More importantly, many of the emotions are defined without reference to desire, and without reinterpretation as a desire. It seems as though we must wait for another thinker within the tradition of Aristotle, Aquinas, to offer a motivational analysis of emotion. 38