Jump to ContentJump to Main Navigation
The EmotionsA Philosophical Exploration$

Peter Goldie

Print publication date: 2002

Print ISBN-13: 9780199253043

Published to Oxford Scholarship Online: November 2003

DOI: 10.1093/0199253048.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 20 August 2018

What Emotions Are, and Their Place in Psychological Explanation

What Emotions Are, and Their Place in Psychological Explanation

(p.11) 2 What Emotions Are, and Their Place in Psychological Explanation
The Emotions

Peter Goldie (Contributor Webpage)

Oxford University Press

Abstract and Keywords

Emotions and emotional episodes are distinguished. Emotions are intentional, but this intentionality cannot be fully captured in terms of feelingless beliefs and desires. The idea of feeling towards the object of the emotion is introduced as an antidote to this over‐intellectualization: emotions are intentional, but also essentially involve feelings. Emotions can be educated. They involve a conceptual tie between recognition and response: the recognition‐response tie. Commonsense psychological explanation of emotion and of action out of emotion are discussed.

Keywords:   action out of emotion, belief, commonsense psychology, desire, emotion, feeling, intentionality


All five of the themes which I introduced in Chapter 1 are brought to bear in this chapter.

I begin with an important distinction: between emotion and episodes of emotional experience; it is often not obvious from the way we speak (‘I'm angry with James’) which we mean. An emotion, I argue, is a complex state, relatively more enduring than an emotional episode, which itself includes various past episodes of emotional experience, as well as various sorts of disposition to think, feel, and act, all of which can dynamically interweave and interact. What holds these diverse elements together is their being part of a narrative.

I then turn to the intentionality of the emotions, and the thoughts which can be involved in an emotion. Here begins my attack on the over‐intellectualization of emotion. First, I consider and reject the idea that the intentionality of emotion can be fully captured by feelingless beliefs and desires. I introduce as an antidote the notion of feeling towards: the feeling one has towards the object of one's emotion is both essentially intentional and essentially involves feelings. Feeling towards is the subject of much more detailed discussion in Chapter 3. Next, I consider the education of the emotions and the relation between recognition and emotional response—what I call the recognition–response tie. I argue that there is a conceptual relation between recognition and response, because, for example, what is dangerous is what merits a particular sort of emotional response, namely fear; furthermore, we are typically brought up, as part of a single educational process, to recognize the dangerous (p.12) and to respond to it fearfully—in a way which is appropriate and proportionate to the circumstances. However, we need not be brought up in this way to have a grasp of the concept of the dangerous or of the concept of fear; and even if we are brought up in this way, recognition and response can occasionally come apart.

I then consider how we explain emotion and action out of emotion by appeal to the thoughts which are involved. A further idea which comes from over‐intellectualizing the emotions is the claim that beliefs and desires can show a person's emotion to be rational in the sense of being based on a sort of syllogistic reasoning. I put forward an alternative, arguing that the thoughts involved in an emotion can show it to be intelligible, intelligibility being a thinner notion than rationality. I also suggest that what lies behind the over‐intellectualizing of emotion is the idea that action out of emotion can be adequately explained by feelingless beliefs and desires, perhaps characterized impersonally, in the same way (so it is thought) that other actions can be explained which are not out of emotion. What we need, I argue, is a notion of action out of emotion which makes it out to be fundamentally different from action not out of emotion. Then, once this fundamental difference is established, any residual temptation to eliminate feelings—particularly feelings towards the object of the emotion—or simply to add them on as an afterthought once action has been explained, should subside.

In sum, then, the aim of this chapter is to show the importance of, and the right place for, the intentionality of emotion in understanding ourselves and others, but to insist that a proper emphasis on this intentionality should not force feelings out of the picture. Feelings are, as we all know, at the heart of emotion.

What Emotions Are

An emotion—for example, John's being angry or Jane's being in love—is typically complex, episodic, dynamic, and structured. An emotion is complex in that it will typically involve many different elements: it involves episodes of emotional experience, including perceptions, thoughts, and feelings of various kinds, and bodily changes of various kinds; and it involves dispositions, including dispositions to experience further emotional episodes, to have further (p.13) thoughts and feelings, and to behave in certain ways. Emotions are episodic and dynamic, in that, over time, the elements can come and go, and wax and wane, depending on all sorts of factors, including the way in which the episodes and dispositions interweave and interact with each other and with other aspects of the person's life. And an emotion is structured in that it constitutes part of a narrative—roughly, an unfolding sequence of actions and events, thoughts and feelings—in which the emotion itself is embedded. The different elements of the emotion are conceived of by us as all being part of the same emotion, in spite of its complex, episodic, and dynamic features. The actions which we do out of an emotion, and the various ways of expressing an emotion, are also seen as part of the same narrative, but not themselves as part of the emotion itself.

We should therefore not be misled into thinking that an emotion is unchanging, just because an emotion can be characterized as a sort of state: ‘He was in a state of anger’. (One particular feature of a state is that when we use a verb to speak of a state, this verb is not used in the continuous tenses; we do not say that he was being jealous or that she is being in love, any more than we say that he is being six feet tall; cf. Kenny 1963, Steward 1997, and Mourelatos 1978.) A state can be a relatively settled disposition of some sort, like a character trait, but an emotion is not at all like this. I will be discussing character traits in Chapter 6, but, briefly, the contrast with emotion is as follows. If you are a jealous sort of person, then you are a person who, in some sense, is disposed to be jealous in your relationships. And when you become jealous of Jane's relationship with John, then you are actually having the emotion. And this emotion—your jealousy of that relationship—can itself involve all sorts of episodes and dispositions regarding that specific relationship. Thus, both the character trait of being jealous and the emotion of being jealous involve dispositions, but the latter are more specific than the former.1 So, to say, for example, that James is jealous can, according to context, mean one of three things: it can mean that he is a jealous type (the character trait); it can mean that he is jealous of this relationship (the emotion); or it can mean that he is currently (p.14) experiencing a particular jealous thought or feeling (the emotional episode).2

An example can illustrate the typical features of an emotion. You are jealous because you think that she has run off with someone else. You cannot sleep: your heart and mind are racing all night. While you are getting dressed in the morning you cannot help imagining them together, talking and joking about you perhaps, and you are unable to keep your mind on anything else. On the way to work, you see another couple in the distance, one of whom looks just like her, and you practically faint, frozen to the spot in terror. Later in the day, you are preoccupied with work for a while, and then suddenly, like a blow to the body, you see on your desk something of hers which triggers your feelings again, and you think ‘If I'm not able to talk to her now then I don't know what I'll do’. The next minute your jealousy takes another turn, and you hope you never see her again; the telephone rings and the thought that it might be her fills you with dread. This complex of perceptions, thoughts, feelings, and bodily changes are dynamically related episodes of the same emotion—the same state of jealousy. And these elements fit in as part of a narrative of this part of your life, which will include not just these elements but also things which you do out of jealousy and your emotional expressions of jealousy: one minute you pick up the telephone to ring her, the next you are unplugging it so that she cannot get through to you; one minute you are biting your knuckles, the next you are hitting the steering wheel of your car. The narrative will also include all sorts of other aspects of your life on which your feelings of jealousy impinge—aspects which often cannot be fully understood apart from those feelings: for example, your way of seeing things in general, your mood, and your character traits.

Consider another example of an emotion, this time love, in this passage from Tolstoy's War and Peace. Pierre is just coming to realize that he is in love with Natasha: (p.15)

When she smiled, doubt was no longer possible, it was Natasha and he loved her.

At that moment Pierre involuntarily betrayed to her, to Princess Mary, and above all to himself, a secret of which he himself had been unaware. He flushed joyfully yet with painful distress. He tried to hide his agitation. But the more he tried to hide it the more clearly—clearer than any words could have done—did he betray to himself, to her, to Princess Mary, that he loved her.

‘No, it's not the unexpectedness of it,’ thought Pierre. But as soon as he tried to continue the conversation he had begun with Princess Mary he again glanced at Natasha, and a still deeper flush suffused his face and a still stronger agitation of mingled joy and fear seized his soul. He became confused in his speech and stopped in the middle of what he was saying. (iii. 398)

Later, as Pierre goes through the routine of the day, his love for Natasha affects his mood, and colours the way things look to him:

‘And this man too,’ thought Pierre, looking into the face of the Chief of Police. ‘What a fine, good‐looking officer, and how kind! Fancy bothering about such trifles now! And they actually say he is not honest and takes bribes? That's the way he was brought up and everybody does it. But what a kind, pleasant face and how he smiles as he looks at me.’

Pierre went to Princess Mary's to dinner.

As he drove through the streets past the houses that had been burnt down [during Napoleon's occupation of Moscow] he was surprised by the beauty of those ruins. The picturesqueness of the chimney‐stacks and tumble‐down walls of the burnt‐out quarters of the town, stretching out and concealing one another, reminded him of the Rhine and the Colosseum. The cabmen he met and their passengers, the carpenters cutting the timber for new houses with their axes, the women hawkers, and the shopkeepers, all looked at him with cheerful beaming eyes that seemed to say: ‘Ah, there he is! Let's see what will come of it!’ (iii. 409)

What we have is a narrative of this part of Pierre's life, involving his love of Natasha, a love which has existed unrecognized for some years, which has just been recognized by Pierre and others as being what it is, and which will, we later find out, continue and mature over many years. That narrative includes his joyful flushing, his painful distress, his efforts to hide his agitation and to continue the conversation, his thoughts about the unexpectedness of his feelings, his deeper flushing, his mingled joy and fear, his confused speech; later, his mood, his way of thinking of the Chief of Police and the ruined buildings of Moscow, and his thinking of the people he passes in (p.16) the street as directing kindly thoughts towards him; and his dispositions to have further loving thoughts, feelings, and responses towards Natasha. By considering all these things, so wonderfully human and utterly intelligible, as part of a structured narrative, we can begin to understand this part of Pierre's life.

So an emotion, such as your being jealous, or Pierre's being in love with Natasha, is typically complex, episodic, dynamic, and structured. Because of these features, it is, as I say, a mistake to think of the emotions as simple and unchanging. Of course, from a sufficiently distant perspective, seen through the wrong end of a telescope so to speak, an emotional experience can be described in this way; you might say that you had been in love with Erika from 1963 to 1965. But if the telescope is turned the right way round, and then the telescope exchanged for a microscope, more and more teeming detail and dynamic structure will emerge.

Perhaps all this is rather obvious. If so, good. For much of what I have just been saying tends to get forgotten in philosophical discussion of the emotions. In particular, it tends to get forgotten in discussion of how one can understand and explain a person's emotions, as if an emotional experience or a disposition can be understood and explained by detaching it from the narrative in which it is embedded, and holding it up for examination as one might hold up a section of bone tissue in a laboratory. On the contrary, understanding and explanation can only adequately be achieved from the personal perspective, seeing the emotion as evolving over time, embedded in and interweaving with the rest of the person's emotional life and other aspects of his mood and character.

The Intentionality of the Emotions

The emotions are intentional. By this I mean that the thoughts and feelings involved in an emotion have a directedness towards an object. (This way of characterizing intentionality, in preference to involving aboutness or ofness, is very effectively argued for in Crane 1998. He also argues, as I will, that emotions can be directed onto objects which are not states of affairs.) So if John is angry or Jane is in love, there will be someone or something with which John is angry and someone who is loved by Jane. ‘Object’ is to be understood in (p.17) the sense of being the object of the transitive verb which refers to the emotional state, as in ‘A φs X’, where A is the person having the emotion picked out by the emotion‐verb φ, and X is the object of the emotion. So when you are asked, for example, ‘What are you afraid of?’, you are being asked to state the object of your fear; the reply will specify something—the object—or, more specifically, some feature of the object of which you are afraid. An object of an emotion, in this sense, could be a particular thing or person (that pudding, this man), an event or an action (the earthquake, your hitting me), or a state of affairs (my being in an aeroplane).

The nature of the object of the emotion is much more complex in what Hume called the indirect passions: emotions such as pride and humility (Treatise, bk. II, pt. i). Here, taking my pride in my house as an example, Hume talks of ‘object’ and ‘cause’, where the ‘object’ of the emotion is the self, and the ‘cause’ is my house: something which causes me to feel pleasure and which is related to me in some way. My approach may be contrary to Hume and to a certain philosophical tradition, but I prefer the more natural idea that the object of an emotion is that onto which one's thoughts and feelings are typically directed, and to which they typically return, so the object of my pride in this example is not just myself, nor just my house, but my‐house‐which‐belongs‐to‐me. This approach, as well as more naturally identifying the intentional object, also allows considerable variation in the focus of my thoughts and feelings when they are directed, proudly, onto their object: I can focus sometimes on the house and those of its features which explain my pride in it (the number of its storeys, say); sometimes on the relationship between me and the house (one of freehold ownership, say); sometimes on myself with the house and its features in the background, so to speak (like one of those Country Life pictures of the lord of the manor sitting proudly in his study with its grand coffered ceiling); and so on.

Moods, as contrasted with emotions, will be the topic of a later discussion in Chapter 6, but I should note here that what, in part, distinguishes emotions from moods is that emotions have more specific objects than moods. The distinction is thus a matter of degree. This avoids two unsatisfactory implications: that emotions necessarily have a specific object; and that moods necessarily are not directed towards an object and thus lack intentionality. Your fear on waking may have no very specific object—the dark, the shape of the curtains, (p.18) the strange noise which woke you—but it is still an emotional experience and not a mood. And the next morning, when your fear is gone but you remain anxious, it is natural to say you are anxious about everything this morning, or about nothing in particular, or that you are anxious about everything and nothing. Making the distinction in this way between emotion and mood allows, as should be the case, for two things: for a mood to develop into an emotion—anxiety into fear, for example—by focusing on a more specific object rather in the way that crystals can form around an object which is dropped into a liquid; and for an emotion to develop into a mood—fear into anxiety—rather in the way that smoke diffuses in the atmosphere, leaving just a haze through which all sorts of objects in the world are seen.

Intentionality gives rise to many philosophical problems, and it is not my wish to engage with these problems here: there is quite enough to be said without getting into those minefields. But I should just point out that, as is generally agreed, there are two features of intentional states of mind which apply as much to the emotions as they do to other sorts of intentional state. First, the object of an emotion has to be identified in a sufficiently fine‐grained way to capture why the person feels that emotion about that object: Oedipus might be delighted that he has married Jocasta, but would not be delighted that he has married his mother. Secondly, the object of an emotion need not exist: Jimmy might be afraid of the Abominable Snowman, when there is no such creature.

Many philosophers who discuss the intentionality of the emotions seek to capture the intentionality of the emotions in terms of beliefs, or beliefs and desires. (See, for example, Kenny 1963, Pitcher 1965, Alston 1967, Green 1972, Davidson 1976, Farrell 1980, Lyons 1993, and G. Taylor 1976 and 1985.) I think that this is a mistake. It runs the risk of leaving feelings out of emotional experience, for these beliefs and desires could be feelingless, by which I mean they could be characterized, perhaps impersonally, without any reference to what it is like from the point of view or perspective of the person experiencing the emotion—the point of view from which feelings are ineliminable. I agree that when I have an emotion, there will often be beliefs and desires which can be ascribed to me and which will play a role in making intelligible both my emotion and what I do out of that emotion. But the mistake is to think that these feelingless beliefs and desires, perhaps characterized impersonally, (p.19) exhaust the intentionality of emotional experience, and that they are therefore sufficient to make sense of emotion and action out of emotion. What I want to do is to emphasize an intentional element which is neither belief nor desire, and which is, in many respects, more fundamental to emotional experience than either of these. This sort of intentional element involves feelings which are directed towards objects in the world, typically towards the object of the emotion. To reflect the fact that this intentional element is both intentional and involves feelings, I will call it feeling towards. Feelings, as part of emotional experience, will be the topic of the next chapter, and I will there be comparing and contrasting feelings towards with bodily feelings, such as feeling your heart beating in your chest. But it will be helpful to the purpose of this chapter to make a few introductory remarks about feeling towards.

Feeling towards is thinking of with feeling, so that your emotional feelings are directed towards the object of your thought. So, for example, if I feel disgusted by the pudding, my feelings of disgust are directed towards some perceived or imagined property or feature of the pudding—its sliminess, perhaps—which I apprehend as disgusting. (As will emerge in due course, this is not to say that you cannot think of something as being, say, disgusting, yet not feel disgust.)

For each sort of emotion, there will be a broadly characteristic qualitative nature of these feelings. But individual emotional experiences can vary widely in how they feel. Any suggestion that our emotional feelings towards things can be understood as, or analysed into, simple terms (such as attraction or aversion, pleasure or distress, feeling comfortable or uncomfortable, positive or negative evaluation on some rating scale) should be strongly resisted. Surely we all know that emotional feelings are not that simple. If you ask me to say what the feelings are like when one is feeling disgusted or jealous or angry or in love, I refuse to answer: if you have experienced the emotion, then you know very well what these sorts of feeling can be like, and you do not need me to tell you; if you have not experienced the emotion and want to get at least some idea of what it feels like, then, as Harold Macmillan once said to a young politician, I suggest you read a good novel. It is, emphatically, not a requirement of my philosophical account that I should attempt such a thing.

Feeling towards, as it is thinking of with feeling, is a sort of thinking of. One can come to think of something as being a particular (p.20) way; certain features become salient. And thinking of is related to seeing an aspect, in that it is an intentional episode which can involve imagination, or perception, or some combination of imagination and perception, often mutually influential: one can come to see the Moon's surface as being face‐like, or to see the duck–rabbit picture as a picture of a duck; and one can think of one's neighbour as a witch and her two black cats as her familiars. (Cf. Wittgenstein 1958: II. xi, Budd 1989: ch. 4, Scruton 1974, and Wollheim 1980 and 1987.) The imagination which can be involved in these sorts of episodes of thought is not imagining that, as one might imagine that Beethoven wrote ten symphonies. The sort of imagining I have in mind is much closer to perception—it has a perceptual quality; but this is not, of course, to suggest that the imagining need be imagistic: if you imagine smelling newly baked bread or, like Yeats, hearing lake water lapping with low sounds by the shore, a visual image of the bread or the Lake of Innisfree need not feature in what you imagine (although, of course, it may well do). Nor, if the imagining is imagistic, do I want to suggest that what is perceived or imagined is an image; rather, as Sartre insists (1948), I am talking of the way the object of thought (the pudding, being in an aeroplane) appears to me in consciousness; imagining something in this way can be understood as visualizing something in your mind, or, as the saying goes, ‘in your mind's eye’.

Having thus introduced feeling towards, I now want to consider in some more detail the thoughts which can be involved in the emotions. In doing this I hope to achieve two things, one positive and one negative. First, I will try to show what is essentially right about the idea that the emotions are intentional, namely that an understanding of the thoughts involved can play a role in making sense of an emotional experience. Emotions are not brute feelings like toothache, which we cannot make sense of; all we can do is give toothaches a causal explanation. Secondly, and negatively, I will try to show up some of the difficulties faced by those accounts of the emotions which, so to speak, take the feelings out of thought by treating feelingless beliefs and desires as the only intentional episodes involved, thus over‐intellectualizing emotional experience.

The relation between an emotion and the beliefs involved is not a contingent one; that is to say, there are conceptual relations between an emotion and the beliefs which ground it. Anthony Kenny puts the point in a way which at first seems rather curious: ‘One cannot be afraid of just anything’ (1963: 192). It is clear, though, (p.21) what he means: ‘If a man says that he is afraid of winning £10,000 in the pools, we want to ask him more: does he believe that money corrupts, or does he expect to lose his friends, or to be annoyed by begging letters, or what? If we can elicit from him only descriptions of the good aspects of the situation, then we cannot understand why he reports his emotion as fear and not as hope. Again, if a man says that he feels remorse for the actions of someone quite unconnected with him, or is envious of his own vices, we are at a loss to understand him’ (1963: 191). However, as Kenny continues, ‘it is possible to be envious of one's own fruit trees; but only if one mistakenly believes that the land on which they stand is part of one's neighbour's property; just as it is possible to feel remorse for the failure of the crops in Vietnam if one believes that it was due to the inadequacy of one's own prayers. What is not possible is to envy something which one believes to belong to oneself, or to feel remorse for something in which one believes one has no part’ (1963: 193).

There are various terms which have been used by philosophers to make this general point about how an emotion is related to the beliefs which are involved. Emotions have been said to have ‘formal objects’ or ‘proper objects'; they have been said to involve identificatory beliefs, which identify the emotion as being what it is, and explanatory beliefs, which give grounds for, or which make intelligible, the identificatory beliefs; and emotions have been said to involve beliefs about the object of the emotion as having determinable and determinate features, and these beliefs too make the emotion an intelligible one to experience in the circumstances (see, respectively, Kenny 1963, G. Taylor 1985, and Taylor 1976). In my view, not much hangs on what terminology is used; but from now on I will usually express the point in terms of beliefs about the object's determinable and determinate features. (A standard example of the contrast between determinable and determinate, is this: being coloured is a determinable property, and being red is a determinate property under the determinable of being coloured.) The idea can be illustrated with a simple example. If Peter is afraid of the bull, then there will be some feature of the bull which Peter thinks it has. If this feature is a determinable one (its being dangerous), then he will think it is dangerous in virtue of its having certain determinately dangerous features (having long horns which could harm him, perhaps), even if he is not able to say what they are. And if the feature which he thinks it has is a determinate one, then it ought to be possible for Peter to explain why this determinate feature falls under the determinable one. Thus, (p.22) according to this idea, if Peter did not believe that there was something potentially harmful about the bull's long horns, then it would be puzzling, to say the least, why he should feel fear of the bull in virtue of its having this feature.

In spite of these conceptual relations between Peter's fear of the bull and his beliefs about it, I suggest that it is possible for no relevant belief to be present when fear is experienced: I can (changing the example) be afraid of a snake which I do not believe to be danger‐ous in any way. Against this, Roger Scruton (1971) has argued that having a belief that an object is dangerous is a necessary condition of being afraid; this is an example of the over‐intellectualization of emotion which I want to resist, and I will be showing in the next chapter how it is even perfectly possible to be afraid of something and yet believe that it is not dangerous. Moreover, others, in particular Robert Gordon (1987), have argued, in respect of certain emotions which Gordon calls factive, that what is required is not just belief but true belief, if it is to be the case that the person does indeed have the emotion. This view is also mistaken (cf. Wollheim 1999: 103–10 for a thorough discussion and ultimate rejection of Gordon's view). Again, this is a sort of over‐intellectualization of the emotions. The idea is based in part on the idea that emotions must have propositions as their objects: if you are angry, you must be angry that p. And, according to this idea, if I assert that, for example, you are angry that James stole your shoelaces, this assertion presupposes the truth of the proposition that James stole your shoelaces. Perhaps it does. But once it is accepted, as it should be, that the object of an emotion need not be a proposition, and can be, for example, a person, then it becomes clear that I can perfectly coherently say that you are angry with James (adding, perhaps, that this is because you think he stole your shoelaces) whilst at the same time insisting that your emotion is entirely ungrounded.

Explaining, or making sense of, an emotion involves more than just establishing the beliefs which putatively ground it. Doing this is merely to make the emotion intelligible, as your envy of your own fruit trees can be made intelligible by your false belief that the fruit trees are on your neighbour's land.3 Intelligibility in this sense is a (p.23) very thin notion. For one thing, an emotion such as your envy of your own fruit trees may be intelligible but also irrational if the false belief which putatively grounds it is one which was not arrived at by a suitably rational process. For another thing, it is possible for an emotion to be intelligible but either inappropriate or disproportionate given the beliefs which putatively ground it. It is, for example, perfectly intelligible for me to be proud of my ancestors, including, as Hume puts it, their “riches and credit” (Treatise, 308), whereas it would be unintelligible if I were to say that I felt proud of the ancestors of someone totally unconnected to me, such as William Hague. But my pride can also be evaluated according to whether or not it is appropriate or proportionate. These days, it is, I think, inappropriate to be proud of your ancestral lineage however great their riches and credit—surely I ought to have left all that behind. Yet, in some other culture, even strong feelings about your ancestors would be entirely appropriate and proportionate. So the right thing to say so far as concerns an emotion's appropriateness and proportionateness is that they can be, to a large extent, culturally determined. To consider another example, we might agree that Harriet's feelings of guilt about forgetting to feed her hamster, thereby bringing about its death, are intelligible, appropriate, and involve beliefs which are rationally arrived at, but, given the importance and value of so many other things in her life, her feelings might have reached a level and duration where they have to be said to be disproportionate. These are distinct and complementary ways of making sense of emotions, and it is a mistake to try to cram them all into one single notion, such as rationality or intelligibility. And finally, there is the great importance of a person's mood and character in making sense of someone's emotional responses: for example, judgements about the intelligibility of a particular person's emotional response are made relative to that person's narrative, extending beyond this particular incident to include his other emotions, his moods, and his character. Your anger at your wife's harmless remark might only become intelligible in the light of your enduring jealousy of her relationship with another man, or in the light of the irritable mood you are in after such a terrible day at the office, or in the light of your general disposition to be irritable.

Emotions can involve certain wants or desires (what I will sometimes call emotional desires) as well as certain beliefs of the sort I have been discussing. I will call the view that the emotions involve (p.24) beliefs and desires the belief‐desire account; I prefer this term to the more frequently used cognitivism because this latter term is, I think, more accurately applied to the view that the emotions are judgements. (Cognitivism is a view held by Robert Solomon: he says ‘My shame is my judgement to the effect that I am responsible for an untoward situation or incident’ (1993: 187).) It is often through detailed consideration of the thoughts which can be involved in an emotion that we come to see what sort of emotion it is that we are experiencing; this is what is profoundly right about the view that the emotions are intentional. I will show what I mean by considering envy, an emotion which I dare say we have all felt at some time or another. (In this discussion, I owe much to Gabrielle Taylor's work, especially Taylor 1988). But first, a few words about the contrast between beliefs and desires.

Beliefs and desires are often contrasted in terms of their having opposing directions of fit. Mark Platts puts the idea like this: ‘Beliefs aim at the true, and their being true is their fitting the world; . . . beliefs should be changed to fit the world, not vice versa. Desires aim at realisation, and their realisation is the world fitting them; . . . the world, crudely, should be changed to fit with our desires, not vice versa’ (Platts 1979: 256–7; cf. Platts 1991 for further qualification). This way of contrasting beliefs and desires is, as Platts rightly insists, highly metaphorical, but it does bring out the role of beliefs and desires in explaining action, exemplified in means–end reasoning: the desire is aimed at some end, and the belief is about there being some means of achieving that end.

However, there are two cautionary points to be borne in mind when applying the idea of direction of fit to emotional desires. First, many of our emotional desires are intimately related to our agency: often an emotional desire is not just that the world be changed to fit our desire, but rather that it should be the agent—the person who has the emotion—who brings it about that the world is so changed. Thus, taking Aristotle's famous example of a desire for revenge which is involved in anger (Rhetoric 1378a30), you do not just desire that she suffer; rather, you desire that it be you who make her suffer. It is thus perfectly intelligible for you, an angry person, to want to bankrupt someone as a means of getting revenge, and then be disappointed to find that the person who is the object of your anger has already become a bankrupt before you were able to put your plans for revenge into motion. Maybe the point can be (p.25) accommodated within the direction‐of‐fit metaphor by saying that what you want is for the world to be changed so that the person has been bankrupted by you. But this accommodation still seems, to my ear, to omit the actual doing of the action from what is desired, even if it is, after all, logically equivalent to a formulation given simply in terms of states of affairs. Anyway, I will not argue the point any further.

The second cautionary point is concerned with whether we should always try to ‘change the world’ to fit our desires.4 The point can be illustrated by a passage from Proust's A la recherche du temps perdu. Albertine has left Marcel, the narrator, and he longs for her return, whether he ‘felt well and not too miserable’, or whether he ‘was in a sombre mood’, with ‘all [his] anger with her revived’. But he sees that having her back will not satisfy him. He says:

But the outcome of these two opposite moods was identical: it was essential that she should return as soon as possible. And yet, whatever joy I might feel at the moment of her return, I sensed that very soon the same difficulties would recur and that to seek happiness in a desire of the mind was as naive as to attempt to reach the horizon by walking straight ahead. The further the desire advances, the further does real possession recede. So that if happiness, or at least the absence of suffering, can be found, it is not the satisfaction, but the gradual reduction and the eventual extinction of desires that one should seek. One seeks to see the beloved object, but one ought to seek not to: forgetfulness alone brings about the ultimate extinction of desire. (v. 514)

Thus even if the notion of realization of a desire (in the sense of its coming true) is the same as the notion of satisfaction of that desire (which I doubt; see Wollheim 1999: 30–8), the satisfaction of a desire is not the same as the satisfaction of the person. This is obviously true of some cravings: I crave this ice‐cream, but eating it will not satisfy me. One might think that in Proust's example Marcel, realizing that satisfaction of his desire will not satisfy him, also desires not to desire Albertine's return, that he wants the first‐order desire (p.26) to be eliminated. But Proust is at pains to show that this is not so: one can have, as Marcel does, an emotional desire for something, and endorse that desire in second‐order thought, yet at the same time accept that satisfaction of this first‐order desire will not bring you satisfaction, and that the world should not—even cannot—be changed to fit that desire. He continues:

And I should have been so afraid of being robbed (had anyone been capable of so robbing me) of this need of her, this love for her, that I convinced myself that it was a precious necessity in my life. To be able to hear, without being charmed and pained by them, the names of the stations through which the train passed on its way to Touraine would have seemed to me a diminution of myself (for no other reason than that it would have proved that I was becoming indifferent to Albertine). It was right, I told myself, that by incessantly asking myself what she could be doing, thinking, wishing, at every moment, whether she intended, whether she was going to return, I should keep open the communicating door which love had opened up in me, and feel another person's life flooding through open sluices to fill the reservoir which must not again become stagnant. (v. 515)

Having registered these notes of caution about the notion of direction of fit as a way of contrasting beliefs and desires, I will now consider envy. What will emerge is that detailed consideration of the thoughts which can be involved in envy can help us to tell just what sort of envy someone is experiencing.

A's envy of B will include the thought that B has a feature X, and either the thought that B's X is, in some determinate respect, superior to that of A, or the thought that A does not have X: Alex envies Bill his new fast car, and he feels envy because he thinks that Bill's car is superior to his in virtue of its being faster. Such a thought seems to be common to what are often called non‐malicious and malicious envy, but the two sorts of envy differ in certain of the desires which are involved; thus thought with a belief‐like direction of fit is not sufficient to distinguish all types of envy. Only if the envy is malicious will Alex also desire that Bill in some way not have the car that Bill does have, perhaps by its being stolen, written off in a crash, or repossessed, and this will be desired regardless of the fact that satisfaction of the desire involves no non‐relative improvement in Alex's position. Furthermore, with some sorts of envy, your going up in the world is not sufficient to satisfy you—you want the other to go down. This was what Gore Vidal had in mind when he said that it is not enough to succeed—others must fail. Furthermore, it (p.27) may be the case that the desire involved in malicious envy is a desire that you be the agent of his failure: ‘If I can't make it to the top, at least I can make damn sure he won't either’.

Malicious envy can be thought of as the obverse of Schadenfreude: roughly, malicious envy involves resentment at the fortune of another, and Schadenfreude involves joy at the misfortune of another. Sometimes the frisson of Schadenfreude that we feel at another's downfall will reveal in us a previously existing malicious envy of which we were not aware, lurking behind the frequently expressed ‘I knew it couldn't last’ sort of response.

If Alex envies Bill his fast car, then the envy may be particular or general. In particular envy, Alex wants the very car which Bill has; whereas with general envy, Alex wants the same sort of car. Non‐malicious general envy can be quite harmless so far as the person envied is concerned, and can be an effective spur for the envious person. This sort of envy is close to role‐model emulation: you might envy your cricket coach's cover drive and try to get yours to be as good as his. Aristotle pithily says that ‘Emulation makes us take steps to secure the good things in question, envy [of a malicious sort] makes us take steps to stop our neighbour having them’ (Rhetoric 1388a35).

Analysis of an emotion along these lines, by subtly identifying the thoughts which are involved, opens up space to consider the ethical, and sometimes the political, dimensions of an emotion. For example, is envy necessary for a successful market economy? Is envy always a sin? It might be that envy is a perfectly intelligible human emotion which we cannot completely eliminate from our psychology, and it might also be that my envy is an ineliminable spur to my ambition and drive, which are central to my conception of myself, but it does not follow from either of these points that envy is an emotion which we should accept as entirely appropriate, or (as the above discussion of direction of fit has already suggested) that the desires involved in envy are desires whose satisfaction we should aim for, even if they are endorsed in second‐order thought. In Chapter 8, I return to some of these questions, as they bear on envy, and, especially, on jealousy.

Book II of Aristotle's Rhetoric is replete with definitions of emotions in terms of the beliefs and desires involved. It should be no surprise that his detailed discussion of the emotions is in a book of this name: the point, of course, being that in rhetorical discourse we often seek to appeal to the emotions in order to sway people's minds. (p.28) If, as a politician, I can get you to think that the industry for which you work is threatened by competition from abroad, and that this is unjust or unfair, then you will very likely feel fear and resentment; and if you do, you will consider your emotions to be justified by the thoughts which ground them, and to be appropriate and proportionate. Or if, as a defending barrister, I can get a juror to think that my client was wrongly treated by the police, then he is likely to feel anger at the injustice that was done; and again he will think his emotion to be justified, appropriate, and proportionate. Then, once these emotional responses have been engaged, your further thoughts, feelings, judgements, and actions will be influenced by your emotions in ways which the rhetorician hopes will serve his interests and purposes.

What Aristotle's discussion of the emotions in his Rhetoric brings out is the relationship between beliefs on the one hand, and desires on the other; or, more broadly, between recognition and response, where the notion of response includes not just motivating thoughts such as desires but affective elements of the emotion too. Even if, metaphorically, beliefs and desires have opposing directions of fit, and even if some relevant recognition on an occasion does not logically or conceptually imply the presence of any response, there is, in our emotional experiences, an intimate relationship between recognition and emotional response which needs to be understood. It is to this that I now turn.

Education of the Emotions and the Recognition–Response Tie

The essential idea is that our emotions can be educated: we can be taught to recognize, and to respond emotionally, as part of the same education. For example, we can be taught to recognize things as dangerous and to respond, appropriately and proportionately, with fear. From now on I will call the relationship between recognition and response the recognition–response tie. In order to consider further the nature of the relationship between recognition and response through the recognition–response tie, I must first examine in some detail what is involved in having a grasp of two distinct sorts of concept: concepts of features such as dangerousness—what I will (p.29) call concepts of emotion‐invoking determinable features—and concepts of emotions such as fear. But before turning to that, I should say something quite general about our conceptual capacities or abilities.

Like other abilities, such as the ability to ride a bicycle, our conceptual abilities can come in degrees: one can have the ability to ride a bicycle, but be unable to ride it up Snowdon, or to ride it with one hand on the handlebars, or to ride it when drunk; but it would still be true that you are able to ride a bicycle, as you can pretty well get around town on one. Similarly, you can have less than a complete grasp of a concept, yet you can quite reliably use the concept in everyday thought and talk (cf. Burge 1979 and 1986). It might not unreasonably be suggested that if an incomplete grasp of a concept is possible, then there must be something that it is to have a complete grasp or mastery of a concept. These expressions, however, would seem to me to be terms of art and open to significant interpretation; analogously, it is open to interpretation what it would be to have a complete grasp of bicycle‐riding: perhaps being able to ride one up Snowdon whilst drunk and with one hand on the handlebars. Nevertheless, I think it is definitely possible to say one thing, at least in respect of the sort of concepts I am concerned with here: the fact that you are able reliably to identify things of the relevant sort is not sufficient for having a complete grasp of the concept. Consider a colour concept: for example, what is involved in having the concept of red. If I were colour‐blind, but nevertheless could by various means reliably identify red things in the world (through prosthetic vision, or through having a constant companion who reliably pointed out to me all and only red things), then I would have a less than complete grasp of this concept. The reason in this case is that I would be lacking the ability to have a certain sort of visual experience in the presence of red things, and without that ability I cannot know what it is for things to look red. We might say that my use of the term ‘red’ is ‘piggybacking’ on the usual use of that term. Now let me apply this general point about concept possession to concepts of emotion‐invoking determinable features such as dangerousness, and to concepts of emotions such as fear, in order to show how we can have a sufficient grasp of these concepts to get the extension of the concept right, whilst this grasp is still less than complete. I will begin with a discussion of emotion‐invoking determinable features.

(p.30) There are two points to be made about the dangerous. The first is that the class of dangerous things has no unifying feature or features which can be captured in the language of the sciences. This is because dangerous things are picked out relative to our particular human, or, more locally, cultural interests: having a concept of the dangerous is essential to us if we are to find our way safely around the world. One need not be mystified by this sort of feature just because it is not one that can be unified by the sciences. Roger Scruton (1980) discusses an example which brings out this point. Being an ornamental marble was an important property for stone merchants searching for stones of the sort which would look good on rockeries. The class of ornamental marbles groups together a heterogeneous mixture of geological and chemical types: for example, it includes marble (a carbonate), onyx (an oxide), and porphyry (a silicate), whilst excluding limestone (which is chemically identical to marble). The property of being an ornamental marble is thus not one which will be unified by, nor will it even feature in, the geological or chemical sciences. But this fact alone does not make it suspect or in any sense second‐rate. Similarly, the concept of the dangerous groups together all sorts of things which have no features in common other than that of being dangerous: bulls with long horns, dogs with rabies, exposed electric fires, icy roads, strangers with sweets, certain ideas, Lord Byron.

This leads to the second point to be made about the dangerous: it is an evaluative property. By an evaluative property I mean a property whose recognition merits a certain sort of response (McDowell 1979 and 1985b). (So, on this notion of an evaluative property, the first point might apply to being red and being an ornamental marble, but still they are not evaluative properties.) The response which is merited by something which is dangerous is not just a recognition or judgement that it is dangerous (for, on this view, all properties would turn out to be evaluative), nor is it simply moving to get out of the way of what is dangerous; the response which is merited is one of fear, with all that this emotional experience involves, including thought, feeling, and action. The process of teaching a child how to identify things which are dangerous is typically one and the same process as teaching that child when fear is merited: pointing to the electric fire, we say, in a fearful and urgent tone of voice, ‘You shouldn't touch that, it's dangerous’, or ‘Don't go near that fire, it'll burn you’. Whilst recognition and response are distinct, and can come (p.31) apart—as I will show shortly—they are related because the emotional response will be of the sort which someone educated in this way ought to have in dangerous circumstances, and his emotional response will not be intelligible independent of his conception of the circumstances as dangerous. (This ‘ought’ is, I will later show, both normative and predictive.) Recognition and response will feature as part of the narrative structure of the person's emotional experience, and when he acts out of the emotion, they will serve to explain the action, so that it can be understood as intelligible, and appropriate and proportionate from the agent's perspective.5 Let us say, to use the relevant term of art and for the sake of argument, that someone so brought up has a complete grasp of both sorts of concept: of the emotion‐invoking determinable property—dangerousness; and of the emotion—fear. Now, there is more than one way in which someone can fall short of having a complete grasp of these concepts. Let me begin with the anthropologist.

It has been argued in a number of places, correctly in my view, that it is not possible to grasp an evaluative concept if you do not appreciate its evaluative point; given the absence of any unifying determinate features, you would not be able to go on to identify new situations as falling under the relevant concept without such an appreciation (cf. McDowell 1979 and 1987, Williams 1985, and Wiggins 1987: Essay III and 1996: 262–4). But it need not follow from this requirement that you yourself have to be a full participant in the evaluation: the recognition–response tie need not be part of your psychology. To see why, consider Allan Gibbard's made‐up story (1992) of the anthropologist who is visiting the Kumi tribe. He has the ability reliably to judge that situations are gopa, being such as to make appropriate a pleasurable feeling of glory (or some feeling like glory); Kumi tribesmen consider such an emotional response to be appropriate, for example, on returning from a successful headhunting expedition. And, let us say, the anthropologist also has the ability reliably to tell when tribesmen are experiencing the relevant emotion. He has these conceptual abilities without being able to respond with what the Kumi consider to be the appropriate feelings, as his emotions have not been educated according to the Kumi way. (p.32) But even though the recognition–response tie therefore need not be psychologically effective for him in order to get the extension of the concepts right, if he is to have at least this grasp of the relevant concepts (a less than complete grasp, but sufficient to get the extension right), the anthropologist nevertheless needs some appreciation of the significance or evaluative point for them of the properties picked out by the concept so that he can, at least to some extent, enter into the Kumi way of thinking of these matters; without that, the concept would, as Charles Taylor puts it, ‘get no grip whatever’ (1985: 53; cf. Williams 1986).

Something of a real‐life counterpart to Gibbard's story is to be found in the sensitive writing of the anthropologist Catherine Lutz about the people of the Ifaluk atoll in Micronesia: there, fago, which she roughly translates as compassion, love, or sadness, is an emotion which is appropriately felt when another person is gafago, which she translates as needy, or ‘lacking one or more of the qualities necessary for a good life’ (1995: 237). Echoing Taylor's point, she says that the relationship between neediness and nurturing feelings ‘is a deep and consistent theme running through Ifaluk culture’ (ibid.). If you are someone who has been brought up to have the right sort of fago feelings, then you ought to be able to spot when another person is gafago; and if you have been brought up to recognize when someone is gafago, then you ought to feel fago on those occasions. But the anthropologist might be able, without being educated in this way, to gain a grasp of both concepts in a way which is sufficient reliably to get their extensions right, and she can do this by coming to understand the cultural and ethical relevance of the recognition–response tie for the islanders.

Of course, what is true of the imaginary concept of gopa and the Ifaluk concepts of gafago and fago is equally true of our domestic concepts. In the light of what I have said so far, one might at this point be persuaded that we can grasp, albeit incompletely, an emotion‐inducing concept without ever having experienced the related emotion, but surely, one might go on to protest, one cannot know what someone is feeling if one has not oneself experienced that emotion: for example, how could I know what you are feeling in a situation which I recognize to be jealousy‐invoking if I have never myself felt jealous? This protest is justified in one sense of the phrase ‘know what someone is feeling’, but not in the other sense. Let me explain.

(p.33) For each sort of emotional experience there will be a paradigmatic narrative structure—paradigmatic recognitional thoughts, and paradigmatic responses involving motivational thoughts and feelings, as well as bodily changes, expressive activity, action, and so forth—and this will be epistemologically central to the concept of that emotion and to how we learn to apply it.6 On the strength of this learning, you can come reliably to judge that a person is experiencing a particular emotion, jealousy for example. To do this it is not necessary to have experienced jealousy yourself. This, then, is one sense in which we say that we ‘know what someone is feeling’: the sense in which what is required is that we be able correctly to answer the question ‘What is he feeling?’ (cf. Austin 1946: 96–7). You might thus perhaps answer ‘He is jealous’, whilst at the same time admitting that you yourself have never had jealous thoughts or feelings, and have no idea what it would feel like to be jealous. Similarly, as Christopher Peacocke points out, it is perfectly possible reliably to judge that people are seasick without ever having felt seasick or without knowing, or being able to imagine, what it would be like to feel seasick (see Peacocke 1985: 33–4). Of course, experiencing an emotion is much more complex in its intentionality than experiencing seasickness, but the same point applies. But there is a second sense in which we use the phrase ‘know what someone is feeling’ when we mean that we know what it is like to have that feeling, and in this sense I would not be speaking truly if I said ‘I know what you're feeling’ (adding, perhaps, ‘poor old chap’) if you were feeling violently seasick and I had always had the sea‐legs of a Captain Bligh, or if you were feeling terrible pangs of jealousy and I had never had a jealous thought or feeling in my life.

We might then be tempted to say that it is necessary to know what it is like to feel an emotion if one is to have a complete grasp (p.34) of the concept of that emotion. (This temptation might be especially strong if one began, as I did, with an analogy with colour concepts.) But it is in this area that I think one must be particularly careful about the notion of having a ‘complete grasp’ of an emotion concept. Emotions can be experienced with greater or lesser intensity, and in this respect at least they are like feelings such as toothaches, headaches, and seasickness. My father claimed never to have experienced a headache, but most of us have had a headache and thus know what it is like to have one. But perhaps few of us have had a headache of such intensity that words cannot describe how awful it is. Most of us have had to go hungry for a day or two, but few of us, dear Reader, have felt the pangs of hunger of the sort felt by a person who is starving. Similarly, perhaps few of us have felt every emotion to an extreme: abject fear, utter despair, amour fou, gnawing envy, insane jealousy. I would not want to say of us that we do not have a complete grasp or full mastery of the concept of fear, despair, love, envy, or jealousy just because we have not felt them to this intensity. But at the same time, if, for example, the most intense despair Harriet had ever experienced was on the occasion of losing her pet hamster, and this feeling was over in a couple of days, she would surely not be speaking truly if she said to someone in utter despair at the death in a car‐crash of his whole family, ‘I know what you are feeling’, in the sense of knowing what it is like. Imaginative extrapolation from the one experience will not get Harriet there, any more than I can extrapolate from the experience of missing both lunch and supper to knowing what it is like to be starving. Concept possession, as I said at the beginning of this discussion, is an ability which comes in degrees, and I can now add that part of what it is to have a grasp of an emotion concept, namely knowing what it is like to experience the emotion, comes in degrees also.

So far, then, I have been discussing the notion of the recognition–response tie in relation to a typical sort of upbringing, and comparing it with how someone can gain a less than complete grasp of the relevant concepts without this typical sort of upbringing. There are other ways we can be brought up too. Someone can be badly brought up, as poacher rather than as gamekeeper so to speak: he can be brought up to recognize situations as emotion‐invoking but to respond in quite a different sort of way to how the rest of us are brought up—to respond with a quite different and inappropriate emotion perhaps. For example, one might be brought up to respond (p.35) to the dangerous with a thrill of excitement. Or one might not be brought up a poacher, but rather come to be a poacher in later life—for example, the torturer who learns to be truly sensitive to the suffering of his victim, and who is moved not to sympathy but to pleasure at inflicting such suffering. And someone can be brought up badly in another way: to respond in the same sort of way as the rest of us, that is with the same sort of emotion, but disproportionately. For example, a child who is brought up by very timorous parents might respond with fear to all sorts of thing which the rest of us consider to be not really dangerous at all: the occasional wasp or a horse in a nearby field at the picnic; the shooting gallery or the dodgems at the fair; the perfectly friendly dog in the park. This timorousness might well continue in later life, so that his responses, although certainly intelligible, are not proportionate. Adequately to explain or make sense of why he, a mature individual, is so afraid of this wasp at the picnic, it is not sufficient to point to the determinate features of the wasp (its having a painful sting), for the rest of us agree that it has these features, and that these features merit fear to some degree, but we do not consider it to merit this degree of fear. What we need to do to explain why he feels as he does is to look into the aetiology of his disposition to respond in this way—the narrative of his upbringing, and of his failure to shake off the timorousness which he picked up from his parents. Thus we cannot fully make sense of the emotional experience without taking into account the larger narrative of which it is a part, and, in particular, without taking account of his character trait.

The gamekeeper‐poacher and the timid picnicker are examples of how someone can, on an occasion, respond inappropriately or disproportionately as a result of having a certain sort of disposition: their recognition–response tie is not as it should be. This is a subject to which I will return in Chapter 6. But what I want to turn to now is how a person whose emotions have been educated as they should be could, on an occasion, have the relevant recognitional thought yet fail to have the related emotion. At this point, we must distinguish two elements of the response which have so far not been distinguished: feelings of fear, and motivating thoughts—typically desires. (One might broadly characterize these elements as the affective and the conative.) I shall discuss these two elements separately here, although I will return to the first in the next chapter.

(p.36) It is indeed possible to recognize something as, say, dangerous, and yet not feel fear. (Remember that I am considering here the person brought up normally, not the person whose use of the concept piggybacks on the normal use.) There might be one or more of a variety of possible explanations of why this happens on an occasion, and I consider some of them in Chapter 3. But there is a special, emotionally charged way of recognizing something as dangerous which does entail that I will feel fear of it, and this way is what I have called feeling towards: feeling fear towards something is thinking of it as dangerous in that special way which involves feeling fear. What I will argue in Chapter 3 is that the content of the recognition in feeling towards is different from the content of the recognition where no emotion is involved. There is thus, in respect of the feeling element of the emotional response, this very intimate connection between recognition and response, simply because feeling towards is thinking of with feeling, and differs in respect of content from merely thinking of without feeling.

The second element of the response is the motivation. Here it seems to me that the connection is weaker: I can recognize something as dangerous and not be motivated at all to act as I ought—in the way I was brought up to act in response to what is dangerous. And this could be so even if I had feelings of fear towards that thing. Again, there are a variety of possible explanations of why I might not be motivated to act as I ought. I might at the time be experiencing different emotions which motivate me otherwise, not just by outweighing my motivations out of fear (for this would not be a counter‐example), but by eliminating them entirely or by preventing them from ever arising. For example, a soldier in battle might be so committed to his squadron's aim of taking that machine‐gun nest that he feels no motivation appropriate to fear as he advances, even though he sees the dangers. One might deny the possibility of such an example, insisting that it must be explained either by his not seeing the situation as dangerous, or by his being motivated to some extent by fear. But there is no reason for this denial other than the prior prejudice that a certain failure to be motivated appropriately is impossible. Consider a different example: I might love someone very much and have the opportunity to see her in very congenial circumstances, and yet on this occasion have no motivation at all to be with her. Does this mean that I do not really love her, or that (as someone might say rather pompously) I do not know what love (p.37) is? Again, it is surely only prior prejudice which would lead one to say such things. To repeat, someone can be educated to recognize situations as emotion‐invoking, and to respond with appropriate motivations, and can still, on an occasion, fail to be so motivated. This failure does not show he is not having the emotion after all, or that he does not have a proper grasp of the concept. The dangerous is, indeed, what merits fear, but it does not follow from this conceptual truth that I will always respond fearfully to what I recognize as dangerous. The answer to the question of whether or not someone has a grasp of a concept is to be found through considering the history of his psychological development, and not every time a situation ‘presents itself’. An occasional failure to respond appropriately is consistent both with having the emotion and with having the concept of the emotion.7

I now turn to a more detailed consideration of the explanatory role of the intentional elements of an emotion. Even if mental states such as toothaches are intentional, in the sense of being directed towards an object, the intentional elements of an emotion play a much wider explanatory role in our psychology and our behaviour: they ought to be able to make intelligible the emotion of which they are a part, as we saw earlier with envy; and they ought to be able to explain action out of the emotion, showing why, from the agent's perspective, it was the thing to do. It is to these explanations that I turn, and to further dangers of belief‐desire accounts leading to over‐intellectualization of the emotions.

Explanation of an Emotion and of Action out of Emotion

Jane hits Jim. Why did she do this? Perhaps she did it out of anger, and she was angry because she thought that Jim had insulted her. I will explore an account which fleshes out these explanations in terms (p.38) of the thoughts involved, and then I will make a number of criticisms of it. Essentially I want to resist the over‐intellectualization of the emotions and action out of emotion which these sorts of explanation can suggest. The sort of account I have in mind is something like the one put forward by Donald Davidson in respect of what he calls ‘propositional emotions’ (1976). Whether it is, precisely, Davidson's account which I outline does not concern me; for it is, I believe, one which is commonly held, either explicitly or implicitly. Having criticized this sort of over‐intellectualizing account, I put forward an alternative—one which seeks to make an emotional response intelligible from the agent's perspective.

I will begin with an explanation of the emotional episode: Jane's becoming angry with Jim. What are the intentional elements involved? One suggestion is as follows: they are attitudes and beliefs, which are related syllogistically. There is thus a major or universal premise which is a dispositional attitude, and a minor or particular premise which is a belief; and these premises lead to a conclusion which is the motive or primary reason. So Jane might have the following thoughts:

  • (U) Jane has an unfavourable attitude towards people who make unjustified insults.

  • (P) Jane believes that Jim insulted her without justification.

  • (C) Jane has an unfavourable attitude towards Jim in so far as he has insulted her without justification.

And Jane's unfavourable attitude towards Jim (C) is her anger with Jim.

But (C) alone is not sufficient to explain her action, although (C) will explain her having a further pro‐attitude or desire:

  • (D) Jane desires to get her own back on Jim.

Jane has this desire because getting one's own back is the sort of thing which angry people want to do. This desire (D) then combines with a belief, and specifically with a means–end belief:
  • (B) Jane believes that the best means of getting her own back on Jim is to hit the person whom she believes to be Jim.

And (D) and (B) then serve to explain the action:
  • (A) Jane hits Jim.

(p.39) So, there are two distinct explanations here, the first of Jane's anger at Jim (C), and the second of Jane's action, her hitting Jim (A). Thus, reading backwards as it were, we can say that Jane hits Jim because she wanted to get her own back, and because she believed that this was the best means of doing so, given the circumstances. Note that this may sound a trivial sort of explanation, and in a sense it is, but it would not sound trivial if Jane hit Herbert, thinking that Herbert was Jim; thus (B) would be at fault. And the reason—the primary reason—why Jane wanted to get her own back on Jim is that she was angry with him, and she was angry with him because she believed that she had been insulted by him. Again, this may sound trivial, but it would not sound trivial if Jane's belief, (P), were false and she had really been insulted by Herbert.

Perhaps the first thing that might strike one about this sort of account is that the explanations proffered are perfectly consistent with Jane experiencing no emotion at all. And if this is so, surely something has gone badly—and fundamentally—wrong. I think a large part of the difficulty here can be traced to a particular philosoph‐ical conception of an action and thus of what is to be explained. Consider two particular actions of a hitting of B by A: in the first action, A is angry with B; and in the second action, A is not angry with B—imagine here the Mafia chief who might have an unfavourable attitude towards people who do not treat him with due respect, and who might seek his revenge coldly, without experiencing any anger at all. According to this conception, both actions can be fully explained by reference to the sorts of belief and desire which I have been discussing. So what could be provided, on this account, to distinguish the angry action from the action not done out of anger? Well, the explanation might mention anger, perhaps: so we might say that (C) above is anger, even though it is perfectly possible to deduce (C) from (U) and (P), neither of which makes any mention at all of anger. And then, if there are further features of the angry action which have to be explained—such as, say, a certain sort of trembling as the blow is struck—then these can be separated off as perturbations of rational activity and causally explained by reference perhaps to the physiology of emotion.

This sort of conception seems to me to be quite alien to the nature of emotion and of emotional experience. There is an important and fundamental contrast between an action done out of an emotion and an action which merely ‘goes through the motions’. It is not true (p.40) that precisely the same sort of action can be done, sometimes with emotion and sometimes without, where the emotion is taken as comprising feelingless belief and desire, plus something which is not directed towards the object of the emotion—a psychological add‐on (a visceral feeling perhaps), or a purely physiological add‐on (a visceral change perhaps), or both.8 The person who thinks that additions like these are sufficient to distinguish emotional thought and action from unemotional thought and action is the add‐on theorist. Against this view stands the phenomenology. Consider doing these things unemotionally: striking a blow; making love; seeking safety. Now consider, and contrast, acting when you act out of emotion: angrily striking the blow; making love passionately; fearfully running away. The phenomenology of such actions—what it is like for the agent—is fundamentally different in character. And an action done with feeling can be distinct in its phenomenology not just for the agent, but also for others involved directly or peripherally in the action; one just has to think what it is like to be made love to with feeling for this to be obvious: it is not like being made love to without feeling, plus feeling. Acting out of emotion is not acting without emotion (explained by feelingless beliefs and desires) plus some added‐on ingredient or ingredients. Rather, when an action is done out of an emotion, the whole action, and the whole experience of the action, is fundamentally different.

This, then, is my diagnosis of the attraction of those add‐on theories which appeal to feelingless beliefs and desires as the only explanatory intentional states which are involved: action out of emotion is not fundamentally different from other sorts of action; so it can and must be explained, or made intelligible, just like other sorts of action, by appealing to feelingless beliefs and desires; then, once these explanatory beliefs and desires are in place, postulation of any further intentional state—one essentially involving feeling—in order to explain the action comes to seem otiose. So, on this diagnosis, the intentionality of feeling involved in emotional thought comes to be forced out of the picture by the apparent explanatory sufficiency of feelingless belief and desire to explain emotional action. This will be the subject of a more extended discussion in the next chapter, where the notion of feeling towards will help to explain what is so (p.41) fundamentally different about action out of emotion and action not out of emotion but performed for some other reason. Once this notion has been accepted, it will become clear that the presence of feelingless belief (or belief and desire) is not, after all, sufficient—or even necessary—for an emotional experience: feelings, and in particular feelings towards the object of the emotion, are central to emotional experience, and to exclude these feelings from an explanation of emotion or of action out of emotion is to over‐intellectualize them.

This distinction, between action out of emotion and action not out of emotion, is from the personal point of view: there is a difference in what it is like to act out of emotion and not out of emotion, but in both cases there is something that it is like. But there is a further distinction to be made, which is between action as such (whether out of emotion or not), and movements of bits of the body, characterized impersonally. If one were, first, to conflate action out of emotion with action not out of emotion (involving, perhaps, a gross description of what is done), and then, further, to conflate action, as such, with mere bodily movements, the idea that an impersonal characterization can be provided of action out of emotion might seem more tempting. This, in turn, could lead to a further diagnosis of the appeal of add‐on theories. They are, I think, particularly appealing to a philosophy of mind which divides the mental, and the causes and effects of the mental, into two contrasting parts: that part which can be characterized impersonally, postulating inputs, states of the organism, and outputs in broadly information‐processing terms; and that part which can only be characterized personally, in terms of what it is like. In an important paper, Naomi Eilan (1998) has questioned this way of dividing things, and has suggested that the impersonal characterization, sometimes described as the ‘easy’ problem (Chalmers 1996), cannot be completed without reference to what it is like, to ‘phenomenal consciousness’. If I am right about these things: that intentionality, including feeling towards, is entirely and fundamentally personal; that emotion both is directed towards an object and involves feelings; and that emotional thought and action do not comprise states of the organism and bodily movements, impersonally characterized, plus something else (say, in this case, phenomenal consciousness); then this would lend weight to her point.

A second and related way in which these explanations can over‐intellectualize the emotions is to be found in the nature of the beliefs and desires which are deployed in the means–end explanation (p.42) of Jane's action out of anger: they are (again) consistent with Jane not being angry, and with her hitting of Jim not being a hitting out of anger (cf. the discussion in Stocker 1981 and in Smith 1998). (There is a further issue in this area to do with actions which are expressions of emotion, such as kicking a chair in anger; expression of emotion will be the subject of Chapter 5.) The point can best be explained by an example concerning an action out of love. Your wife has just come home after a terrible journey, and sinks exhausted into a chair. You get up and make her a cup of tea, and this is an action you do out of love for her. It is therefore, as I have argued above, a fundamentally different sort of action from one not done out of love, and thus calls for a distinct sort of explanation. But it does not follow from this requirement that the two explanations cannot also contain many thoughts in common—ones which explain why you made the cup of tea—set out as means–end belief‐desire explanations. There could be a desire to cheer her up and a belief that making her a cup of tea is the best way of doing this; and there could be a desire to make her a cup of tea, and a belief that the best way of doing this is to do what you are doing (put the kettle on, and so forth). And there may well be more explanatory thoughts in common: in both cases, perhaps, you saw that she was exhausted and you remembered the sort of tea she likes. (Of course if you did not love her you would, perhaps, be less likely to have had these thoughts, but obviously you could have had them if you did not love her.) But there is one explanation that only has a place where the action is done out of love, and another explanation that only has a place where it is not the case that the action was done out of love. The first sort of explanation might be given where someone (a child, a philosopher) rather tiresomely keeps on and on asking you why you did what you did, and you have given all the reasons you can think of: ‘She looked exhausted, I felt awful for her, and I thought a cup of tea would cheer her up’ and so forth. Then in despair you shrug your shoulders and say ‘Because I love her’. This is not to give a further reason, for there is no further reason; it is rather to put in context all the reasons that you have already given—all the episodes of thought and feeling which are involved—by placing them in the narrative as part of the love you have for her. The complex web of thoughts and feelings is thus summarized, or concertinaed so to speak, into a single explanatory phrase: ‘Because I love her’, and your having these thoughts and feelings are made primitively (p.43) intelligible by reference to your emotion and to the dispositions which the emotion involves. An emotional thought or feeling is primitively intelligible if it cannot be better explained by anything else other than the emotion of which it is a part; thus, Jane's desire to get her own back on Jim (D) is primitively intelligible—it is just the sort of desire which angry people paradigmatically have. The second sort of explanation, one which only has a place where the action is not done out of love, is that you did it in order to do a loving act, because this is the sort of thing which loving people do, or that you did it for the sake of love. Of course, it is consistent with loving a person that you do something for the sake of your love, but it is not consistent with an action out of love that it be done for the sake of love. If something is done for the sake of love, the explanatory thoughts and feelings are not made primitively intelligible by reference to the emotion: further explanation is necessary. Where the answer is that you made her a cup of tea for the sake of your love, there is always room to ask for a further explanation which is ‘outwith’ your love, to use the Scots phrase: perhaps you did it for the sake of your love in order to reassure her that you love her; and perhaps you wanted to reassure her because you thought she was not sure that you loved her; and so on. But where the reason is ‘Because I love her’, there is no need for further explanation ‘outwith’ your love. Acting out of emotion is not the same as, and is not reducible to, acting for the sake of an emotion.

The third and final way in which the explanations I have outlined tend to over‐intellectualize the emotions and action out of emotion is well brought out by Gabrielle Taylor (1985), and by an example she gives. She asks what part the universal pro‐attitude (U) plays in making sense of Jane's emotion. This is not the same question as whether a suitable universal is available. For what is at issue is whether this unfavourable attitude towards other people who make unjustified insults is a prior dispositional commitment of Jane's, or whether it is merely something which Jane is subsequently committed to, as a matter of consistency, in virtue of having this emotional response. To draw out this point, Taylor's central example is of humiliation, drawn from a passage near the end of James Joyce's short story ‘The Dead’. It is a marvellous example, and as I cannot hope to improve on it, I trust she will forgive me for using it here. Taylor sets the scene like this: ‘The protagonist, Gabriel, has spent the evening at a party given by his aunts, where his function was to (p.44) see that all went smoothly, and to make the after‐dinner speech. He is now alone with his wife Gretta and has just been listening to her reminiscences. She has been telling him of a boy, a boy employed in the gas‐works who years ago was in love with her and who died, she thinks, for her sake. Gabriel tries to stop her telling the tale by making ironic comments, but she does not even notice that this is what he is doing’ (1985: 8–9). Taylor then goes on to quote Joyce:

Gabriel felt humiliated by the failure of his irony and by the evocation of this figure from the dead, a boy in the gasworks. While he had been full of memories of their secret life together, full of tenderness and joy and desire, she had been comparing him in her mind with another. A shameful consciousness of his own person assailed him. He saw himself as a ludicrous figure, acting as a penny‐boy for his aunts, a nervous, well‐meaning sentimentalist, orating to vulgarians and idealizing on his own clownish lusts, the pitiable fatuous fellow he had caught a glimpse of in the mirror. (ibid. 9)

What is the universal to which Gabriel might be committed as a matter of consistency in virtue of his feeling of humiliation? Might it be the thought that anyone should feel humiliated if he is full of memories of a secret life together while he is being compared in her mind with another (ibid. 10)? Surely Gabriel might think that it is only people like him who would have such feelings. But perhaps it is only the presence of his other thoughts and feelings that very evening that explain why he now feels humiliated. So perhaps we need something like ‘for someone who already thinks of himself as inadequate in other respects, the present situation is mortifying’ (ibid. 11). But, as Taylor insists, this universal does not explain his feeling humiliated; this would be to get the order of explanation in reverse. The point is that even if Gabriel were committed to some such universal, and one can presumably be found on the grounds of the requirement for consistency, the universal would play no role in explaining his emotion which would not already be filled by his particular thoughts and feelings which arose that very evening. Rather than over‐intellectualizing the reasoning into a syllogistic form as does Davidson, with the universal as major premise, it would be better, as Taylor says, to look more widely to all sorts of earlier events and other thoughts and feelings which Gabriel has had in the past, and to his character more generally, in order to make sense of why he feels as he does on this occasion. As Taylor puts it, ‘the constraints (p.45) on what is explanatory are not provided by what must be the case in all rationality. The appeal is no longer to the wholly rational being; it is to the admittedly far less neat and precise notion of what it would be human and natural for a person to feel under certain circumstances, given that person's relevant other beliefs and attitudes’ (ibid. 14). My notion of narrative structure captures the same essential point. Gabriel's emotional experience cannot be adequately made sense of in abstraction from the narrative in which it is embedded, and, in the light of that narrative, including his other emotions, moods, and character, we can come to see his emotion as intelligible in the circumstances; and, to the extent that we might consider Gabriel's feelings to be inappropriate or disproportionate, we can also come to see why they might be so.

But I would go further than Taylor in her resistance to the over‐intellectualization of the emotions. Taylor wants, quite rightly, to resist the model when it relates the attitudes and beliefs which explain the emotion syllogistically; rather, there may be a complex, dynamic structure of episodes of thought which serve to make sense of the emotion. However, she still rather suggests a picture where the beliefs which are involved have to be firmly in place as part of the agent's psychology prior to, or at the same time as, the emotion if they are to be explanatory in the required way. (I infer this to be her view because her reason for rejecting the postulation of a universal as explanatory is that it need not be in place in this way.) As Taylor puts it in one example, ‘I am afraid of the snake because its bite is poisonous and poison would harm me’ (ibid. 2).

I believe that many of our emotional experiences are not like this: there seems to me to be too much talk of belief, and not enough talk of feeling, perception, and imagination. The point is not easy to express, but what often happens, I think, is that we first have an emotional response towards an object, a feeling which is often quite primitive in ways which I will be trying to elucidate in later chapters. Then, in self‐interpretation, when we become reflectively aware of this feeling towards the object of the emotion (as we reflective beings are sometimes able to do), we also normally seek to make it intelligible by looking for the identificatory and explanatory beliefs of the sort which Taylor discusses. What really comes first is the emotional response itself—the feeling of fear towards the snake—and not the thought that its bite is poisonous and the thought that poison would harm me.

(p.46) Perhaps there is a more primitive story to be told about many of our emotional responses, particularly those which, intuitively, we have in common with many other animals. Let us return for a moment to the recognition–response tie in respect of fear, and consider how an episode of fear might occur in the psychology of a zebra. Intuitively, one wants to say that there is some sort of primitive fearful recognition by the zebra of, say, the lion, and a response simply in terms of visceral reactions and an impulse towards flight. And perhaps it is just this sort of recognition and response which can be operative in we humans when we respond ‘primitively’.

Consider these two stories, both involving fear. In the first, you hear the news that the international pharmaceutical company for which you work is to merge with another of roughly equal size, and that there will be significant redundancies in the research and development department where you work. You have not been there for long, and so the redundancy cost of getting rid of you will be relatively low. Furthermore, you are aware that you have not had any notable research successes recently. As a result of these and other thoughts and deliberations, you become afraid that you will be made redundant, and you start to look for another job because you want to avoid being without work. This is, note, genuine fear that you now feel (as is only too clear when you wake at four o'clock in the morning, bathed in sweat), and it is fear which explains your action.

In the second story, you are wandering across the road with your thoughts—in that excellent phrase—‘miles away’, and you suddenly see a bus coming towards you. In fear, you throw yourself out of the way of the bus.

I think that beliefs and desires can be appealed to in order to explain both of these emotional responses. But there does seem to be some sort of significant difference in the role of beliefs and desires in explaining the two responses. Of course one difference is that, in the first story, you are very likely to be aware of your beliefs and desires at the time of the emotional experience, and perhaps aware of them as featuring in your practical reasoning, whereas in the second story this is less likely. But there is more to the difference than that: in the second story, there seems to be good reason to say that the beliefs and desires which make sense of the emotional response do not also causally explain the emotional response: as one might say, they come too late for that.

(p.47) This is not as revisionary as it might sound. The idea is not that there is no causal psychological explanation of an action out of fear such as jumping away from the oncoming bus, for surely it was, after all, something that you recognized about the bus which causally explains your feelings about it and why you did what you did. It is, rather, just the idea that dubbing the psychological episodes involved as beliefs gives them all too intellectual a flavour in explaining such an action: for example, a person's beliefs ought to meet certain rationality constraints, such as being consistent and coherent. It is as though we almost have to post‐rationalize the bus story by ascribing the right beliefs, in order to throw light onto the emotional response in the context in which it is set. Post‐rationalization here does not imply ‘making up’ the beliefs we are to ascribe; ascription of the beliefs which Taylor talks of can be correct or incorrect, appropriate or inappropriate. But correct and appropriate ascription of such beliefs need not imply, I think, that these beliefs need play a causal role in explaining the emotional response.

The recognition–response tie which I have introduced is meant to be sufficiently protean that it can accommodate, at the one extreme, psychological episodes as complex and sophisticated as those involved in the fear of being made redundant, and, at the other extreme, psychological episodes as primitive and animal‐like as those involved in fear of the oncoming bus. This, I agree, leaves me with the tasks of saying just how some of our more primitive emotional responses are ‘animal‐like’, and of saying what the nature of these primitive responses is. I will be trying, rather inadequately, to fulfil these tasks in the next two chapters.


There are, then, a number of ways in which explanations of the emotions and of action out of the emotions in terms of beliefs and desires can over‐intellectualize the emotions. But, nevertheless, there is much that is profoundly right about accounts of emotion which place emphasis on their intentionality and on the sorts of commonsense psychological explanations of emotion and action out of emotion which I have been discussing—just so long as those explanations do not over‐intellectualize them. I will end this chapter by summarizing what is right about them.

(p.48) First, when we have an emotion, we are engaged with the world, grasping what is going on in the world, and responding accordingly. Psychological states like headaches or toothaches may be intentional, being directed towards a condition of one's body, but the thoughts and feelings in an emotional experience are (at least typically) directed towards objects in the world, beyond the bounds of our bodies.

Secondly, an emotional experience, and action out of the emotion, can be made intelligible by reference to the thoughts which are involved in it. An emotion can be groundless if it is based on false perceptions or beliefs, and it can be irrational if it is based on irrationally formed beliefs. And emotion, and action out of emotion, can be assessed for appropriateness and proportionality.

Thirdly, the emotions can be educated; in bringing up a child we use the child's capability for emotional experience, and our own emotional responses, to educate him or her to recognize certain things as meriting a certain sort of emotional response. And, through this process of education, the child's responses can come to be both appropriate and proportionate.

Fourthly, the emotions partially shape and determine what we value. As Hume says (Enquiries: 209): ‘If I have no vanity, I take no delight in praise: if I be void of ambition, power gives me no enjoyment: if I be not angry, the punishment of an adversary is totally indifferent to me. In all these cases there is a passion which points immediately to the object, and constitutes it our good or happiness’. It is thus wrong to say, as does, for example, Paul Griffiths, that the emotions are conceived of as ‘relatively unintegrated, irruptive sources of motivation’ (Griffiths 1997: 243), and that ‘The vernacular concept of emotion is an attempt to mark out a category of psychological states which produce behaviour not integrated into long‐term, planned action’ (ibid.). On the contrary, the emotion of being proud of my son for his academic achievement, or the emotion of loving my wife, hardly needs to be irruptive or to produce behaviour unintegrated into long‐term, planned action. There is a contrast in commonsense psychology which we would do well to bear in mind here: the contrast between being emotional and having an emotion (see Parrott 1995). Griffiths' remarks are, if anything, appropriate to being emotional.

Fifthly, and following on from this last point, our emotional responses can reveal to us what we value, and what we value might (p.49) not be epistemically accessible to us if we did not have such responses (cf. Stocker 1996). Recall how a feeling of Schadenfreude might reveal to you that you had been envious of the other's success. Or I, an Englishman, might think that I have contempt for everything English, and think that I have no pride in being English; yet when abroad I suddenly feel a surge of pride when someone says what a wonderful place England must be to live in. Or—a common enough experience, I think—I might be critical of my parents or siblings to others who are not part of the family; but if you make the mistake of joining in the criticisms, you will find that I turn angrily and vehemently against you, and start fiercely to defend my relations, for after all I really do love them, as my angry response reveals to me and to you.

Sixthly therefore, and finally, we should respect our emotional responses, and listen to what they have to say to us and about us. But they are not the final arbiter: our emotional responses should be held up for examination and reflection. Of course, this cannot be done from an emotionless, purely rational perspective, for there is no such standpoint, but it should be done in the light of reason and of our other emotional responses to the other things we value. And if this examination and reflection shows us that our emotional responses are not appropriate, then the emotion should cease. That it may not is important, and I will come to why and how it may not in the next chapter.


(1) Phobias, like claustrophobia and other phobic fears, are also states—dispositional states—although they are neither character traits nor emotions. A phobic fear, such as a phobia of dogs, is a disposition to respond and act in certain ways on sight of a dog. So when we say ‘He's afraid of dogs’, it is the disposition to be afraid of dogs which we are talking about.

(2) Throughout this book I will try to make clear the sense in which I use emotion terms like ‘jealous’. I will sometimes use the more convenient phrase ‘So‐and‐so's emotion’ rather than ‘So‐and‐so's emotional episode’ where it is clear from the context that it is the emotional episode which I have in mind. Wollheim (1999) also distinguishes emotion and emotional episode, but he insists that emotions are dispositions (which possess ‘psychological reality’). I agree that emotions involve dispositions, but his account, I think, leaves insufficient place for the narrative of past emotional episodes as being part of a continuing emotion.

(3) Where there is no relevant belief, as in my fear of the snake which I do not believe to be dangerous, there will still be some thoughts and feelings about it which will make the emotion intelligible. And, so far as fear of snakes is concerned (see Wilson 1984 and 1998), we do not have to look very far to guess what these might be.

(4) Wollheim (1999: 45–51) offers a superb critique of the idea of direction of fit. He says that, because the truth‐relation and the satisfaction‐relation are both symmetrical (if X satisfies Y, then Y is satisfied by X), the idea of direction of fit has no place. So the idea must be prescriptive of what we should try to do: it gives us, as he puts it, ‘an ethic of belief and desire’ (ibid. 49). But, leaving aside any difficulties with belief, this ethic ‘is surely far too indulgent towards desire’ (ibid. 50). What follows in the text illustrates this point, I think.

(5) Of course there are other possibilities: someone's response might be inappropriate or disproportionate from his own perspective; or the response might be appropriate and proportionate from his perspective, yet inappropriate or disproportionate in the opinion of others. These are possibilities I will come to later.

(6) It is thus not a contingent matter that when we are taught the meaning of an emotion term, we are naturally taught by reference to what is paradigmatic; what is non‐paradigmatic of one emotion will often be paradigmatic of, and epistemologically central to, some other emotion. But what is paradigmatic of an emotion is not constitutive of what that sort of emotion is; it does not ‘define the very character of our emotions’, as Ronald de Sousa claims (1987: 183). A narrative for a particular emotional experience can include what is non‐paradigmatic just as much as what is paradigmatic: one can laugh out of grief or kill a person out of love for her; analogously (the example often used against the paradigm or prototype view of concepts), penguins are just as much birds as robins or sparrows, even if they are not epistemologically central or paradigmatic of birds (cf. Rey 1990 and Gardner 1992; see also Foot 1983 and Lyons 1993).

(7) This discussion is closely related to issues concerning internalism in ethics. See Williams (1995: 569) and Blackburn (1998: ch. 3) for the view, parallel to the one I am putting forward, that there is an internal, but not strict, relation in ethical thought between recognition and motivational response; as Blackburn nicely puts it (1998: 65): ‘externalists can have individual cases, but internalism wins the war’. However, there is a significant difference between the two issues: so far as emotions are concerned the relevant response is not just motivation but also feeling.

(8) It might be said that the visceral feeling is intentional, as it is directed towards the viscera. This may well be true (I think it is), but the point remains that this sort of feeling is not directed towards the object of the emotion.