Valuing Emotions

Marcia Cavell

[This paper was presented on June 11, 2004, at the Annual Meeting of the Rapaport-Klein Study Group]


"I fear that the animals regard man as a creature of their own kind which has in a highly dangerous fashion lost its healthy animal reason - as the mad animal, as the laughing animal, as the weeping animal, as the unhappy animal." (Nietzsche)

Until recently the emotions were consigned to a small backyard of the mental that was visited chiefly by philosophers. Aristotle named the features of the emotions, set the questions, that continue to guide inquiry: the relations between emotion, belief and desire, emotion and action; the place of bodily arousal and physiological expression; the connection of emotion with pain and pleasure. Aristotle was the first to ask: Can emotions be said to be rational or irrational? Are they passive or active, in or out of our control? Are they educable?

Then somewhere in the sixties the subject heated up. Philosophers again were in the vanguard, joined by psychologists, neuroscientists, and, of course, psychoanalysts, who, having long lamented their lack of a coherent theory of the emotions, suddenly acquired many allies in the task of constructing one. In philosophy the movement began with a challenge to the traditional distinction between reason and passion, attempting to rescue the emotions from the irrational and the a-rational by construing them as some combination of belief, desire, and, appraisal. But it now seems that while cognition often plays an essential, constitutive role in the emotions, it does not always; and that on the other hand, even construed generously, belief desire, and appraisal leave something out. We are still far from a general theory.

The aim of this essay is not to provide one. Indeed there may be insufficient similarities among the different things we call emotion for it to be a useful psychological concept. [Footnote 1]  (Think how murky the distinctions are between emotion, affect, and feeling.) I want rather, first, to set out what strike me as some salient features of the emotions that any satisfactory theory will have to accommodate; and second, to argue that emotions are a kind of felt perception of things in the  world and in ourselves that matter to us, a perception that like visual perception grasps reality, though it is also, like perception, open to error.

Emotion is what accounts for subjectivity, I will say: the way in which my experience is peculiarly mine, yours peculiarly yours. But what the emotion perceives - its intentional object - is not itself necessarily subjective: emotion is often a perception of something outside oneself which has value of some sort; and an emotion can be said to be more or less appropriate to the object or situation to which it is a response.

The ordinary, pre-theoretical view of the emotions

The emotions figure in our ordinary every day explanations of human experience and behavior. Stripped down, our explanations often go something like this: Maria is angry with you because you insulted her, and it is because she was angry that she slapped you. Hans is frightened because he thinks his father is going to punish him, and it is because he has this belief that he hides. John is blushing because he thinks he just made a faux pas. Sylvia  is ashamed because she is envious of her best friend's success.

As they stand these explanations are far too simple. For one thing, we do not always act on our emotions: we stay our hand, or the impulse to slap or to hide does not arise in the first place; which suggests, for another, that anger by itself does not explain the slapping or the hiding, but only as the emotion interacts with others of the agent's mental attitudes. Nevertheless, the ordinary view suggests a number of basic ideas about the emotions that I believe are right. Let me flesh them out.

1) The causal complexity of emotions. Maria's and Sylvia's beliefs are part of the causal story of their feeling as they do, and so are their desires: Maria's anger may be related to her urgent wish to be esteemed, her chronic fear of being insulted, a fear that may go deep into her character, setting a horizon against which she perceives, and mis-perceives, herself and others. - For the moment I am not distinguishing between emotions and feelings. I will come back to this later.

Emotions often breed other emotions. Sylvia's envy, together with beliefs about her friend, her own ideals for herself, her fantasies, cause yet other emotions to arise in her: shame, perhaps a fear of retaliation, worry about being herself in an enviable position, therefore open to the sorts of attacks on herself that, in her envious state, she would like to launch on others.

Our emotions are causally linked both to other  'inner' states like beliefs and desires, and also to the world around us: James has done something Maria perceives, perhaps correctly, as an insult. The frown on his father's face frightens Hans. You are leaving and I am sad. At least in human beings with relatively developed conceptual capacities, an emotion is part of a complex causal nexus which relates emotions to other emotions, to other mental states like beliefs and desires, to perceptions, and to the external world, the whole of which forms a disposition to act in certain ways in certain circumstances. The causal connections go every which way: from beliefs and desires to emotions, from emotions to emotions, from emotions to world, from world to emotion, from mind to body, and from body to mind.

Deeply enmeshed in other mental states and dispositions as we ourselves are enmeshed in the external world, our emotions express our vulnerability to the ever-changing reality around and within us. We are born to a world not of our making, and we become believing, desiring, impassioned creatures only through our relations with other creatures, relations over which we have little control.

2) The constitutive role of beliefs and desires in the emotions. Emotions differ from each other according to their degree of cognitive sophistication. Certainly such sophisticated emotions as pride, guilt, and remorse, are always formed partly from beliefs. Not only are beliefs and desires part of the causal story in these cases; they help identify the emotion as this particular emotion. If Toni is proud, there must be something she believes she has done well; if John is outraged, he must believe that someone has done something outrageous. We distinguish envy from jealousy, guilt from remorse, and remorse from a fear (merely) of being caught, in part through the role of belief in the emotional state. Freud acknowledges as much in writing: "When there is a mˇsalliance... between an affect and its ideational content... a layman will say that the affect is too great for the occasion - that it is exaggerated... On the contrary, the [analytic] physician says: 'No. The affect is justified'"

I take this to mean that in the absence of the relevant beliefs - in guilt, for example, the belief that you have done something wrong by your own lights - you cannot be feeling guilt. Envy is by definition a feeling we have towards someone who has something we value and that we believe we lack; pride is, by definition, a feeling we have when we believe we have done something to our credit. Where a cognitive component is present, the relevant beliefs and perceptions may be ghosts from the past, ghosts of which the agent himself is not aware. He doesn't understand why he feels guilty, or about what. The 'justification' is unconscious; but it is present.

3) The motivational character of the emotions. Emotions often motivate behavior - the slapping, the apology, the hiding - or are at least a part of the motivational story. (The root of 'emotion' is the Latin 'movere', to move.)  We should note that blushing with shame, perspiring with fear, unlike striking out and saying the hurtful word, are not motivated. That is, I do not perspire with fear in order to achieve something I want. Of course culture and custom play a complicated role here. In some societies the mourner  is taught not to cry, or on the contrary to shriek with pain; these cultural conditions can affect behavior. And the actor can imaginatively work herself into a state of mind in which blushing, perspiring, crying, 'come naturally'. Nevertheless, crude as it is, some distinction between emotional expressions that are not -volitional and those that are needs to be made. The emotions are mental states or dispositions - I'll pick up the distinction in a moment -  that may find overt expression in volitional doings, like the slapping and the hiding, but also in physiological responses over which we have no, or little, control.

But also an emotion may find no overt expression. We do not always blush when we are embarrassed; we can often hide our anger, or control it. So bodily arousal cannot, I believe, be made a necessary condition for emotion. Neither can overt behavior.

4) Emotions and rationality. Because beliefs and perceptions are partially constitutive of some emotions, those emotions are, like beliefs, subject to error, to education, and to rational appraisal.  Beliefs can be true or false, reasonable or unreasonable, wishful or realistic; they can be tested, appraised, and adjusted in the light of reason. And if beliefs can be tested and appraised, so can the emotions into which beliefs enter, even if the process of adjustment is typically much slower. In some cases we judge an emotion irrational, mis-guided, mis-placed, inappropriate, self-deluded; sometimes even 'genuine', or 'false'. Is Allen's pride, or guilt, or sorrow, genuine? And is this a question about Allen's belief that he is feeling proud or guilty, or about his pride or guilt itself? The line between what we believe about what we are feeling and what we are feeling seems to me not always clear. We might have reason to think that Allen was earlier not capable of feeling proud or guilty, but now is. Now these feelings are genuine, by which we might mean something like this: Allen can now discern those situations for which pride and guilt are appropriate, and register those discernments in his feelings.

5) Mis-identifying one's emotions. People  often deceive themselves about a feeling, or mis-identify it, or some combination of the two. Sylvia tells us, and apparently believes, that she is simply delighted about her friend's success; Maria assures us she's not angry about the insult. But other things they say or do belie their self-descriptions. Jane Austen's Emma is concerned about Harriet's interest in Mr. Knightly because, Emma thinks, Harriet will be hurt, since of course Mr. Knightly cannot be interested in Harriet. But the interest is reciprocated, Harriet assures Emma. Austen writes:

"Emma's eyes were instantly withdrawn. A few minutes were sufficient for making her acquaintance with her own heart. A mind like hers, once opening to suspicion, made rapid progress. She touched, she admitted, she acknowledged the whole truth. Why was it so much worse that Harriet should be in love with Mr. Knightly than with Frank Churchill? Why was the evil so dreadfully increased by Harriet's having some hope of a return? It darted through her with the speed of an arrow that Mr. Knightly must marry no one but herself" (p. 418).

6) Mental episodes and mental dispositions. [Footnote 2] We invoke emotional dispositions when we speak of someone as quick to anger or to take offense; to emotional episodes when we refer to moments of joy, disgust, disappointment.  Many mental states come in both varieties. My belief just now that a hummingbird is approaching the feeder is episodic; my long-standing, perhaps unconscious belief, that I will inevitably be disappointed, is a disposition.

Mental episodes are transient, part of what James calls the stream of consciousness. They occur at a particular time, and include things like Maria's anger over the insult, pleasure in listening to a Bach motet, passing thoughts of a friend's imminent departure. In contrast we might call mental dispositions those more-or-less persisting attitudes of the mind that underlie a mental state or succession of mental states. These dispositions have complex histories, as is briefly illustrated above with the emotions of Maria and Sylvia.

Following this distinction I would call unconscious fantasies dispositions that are triggered into play by emotions, in particular by anxiety. Fantasies are perhaps the most complex dispositions of all, a weave of emotion, belief, desire, generated from the deepest of one's fears and longings, defending against unconscious fears, condensing more of one's history, saying more about how one is disposed to see the world, than any other sort of disposition. Consider Maria's readiness to feel demeaned, and the many emotional memories and old beliefs that readiness itself contains. Perhaps she has good reasons, given what has happened to her, to feel that somehow she is deficient, defective, not worth attention. She may then respond to her feelings of humiliation with a defensive, unconscious fantasy of omnipotent superiority, which then breeds the fear of disappointment, a disposition to avoid trying to succeed at anything because the fear of failure, as measured against her fantasies of transcending success, is so great. Connected to the illusions of omnipotence may be an irrational fear of causing harm to others. The fantasy then may become part of a motivational structure which leads to behavior - not trying, holding herself back, sabotaging her efforts to succeed - that only confirms the feelings of insufficiency which generated the fantasy in the first place.

7) Emotions as the register of what we care about. We have emotions because we care about what happens to us, and care about things in the world as they relate to our vital interests. Caring is implicit in all the emotions: If we didn't care about what we do, about how people treat us, about who we are and about things and people in the world, we wouldn't feel anger, fear, anxiety, pride, or envy, or any of the other emotions. Nor would a care-less creature be rational in any sense we care about. Any account of  rationality must then provide a place for the emotions.

Mind, body, and the unconscious

The subject of the emotions is thorny in part because it forces us to the mind-body problem in a way that other mental states like belief do not. For one thing, as we have seen, the emotions are often, though not always, experienced viscerally, our connection with the external world vibrating in our bodies. Yet unlike itches, tingles, and nausea, emotions often, though not always,   represent the world as being a certain way, as ominous, exhilarating, disappointing, irritating, as a journey into an exciting future, or as desolate reminder of an irrevocably gone past. This is the sense in which emotions, unlike pains and pleasures, have intentionality: We are sad, or angry, or guilty, about something, even though we may not know what it is.

For another, the emotions run a gamut from primitive rat fear to the sophisticated complexity of Marcel's jealousy or Emma boredom. At both ends of the gamut there are problems. For most of the century after Darwin wrote The Expression of the Emotions in Man and Animals  (published in 1872), scientists ignored his work because they thought that in attributing emotions to animals Darwin had anthropomorphized them. His work on facial expression in men and animals has been taken up, notably by Paul Ekman, and developed into what is sometimes called the 'affect program theory'. The affect programs correspond roughly to the folk-psychological concepts of surprise, anger, fear, sadness, joy, and disgust. They are stereotyped responses involving different bodily systems. On this theory affects are sources of motivation that can be relatively independent from the agent's other mental states. Affect theory explains our sense that the emotions often overwhelm us, render us passive, irrupt inside of us. But it doesn't explain the features in our pre-theoretical view that we listed above. [Footnote 3]

Emotion in animals is a question I will only touch on here. My own line is that rats and a great many other animals surely experience fear. And just as surely, fear is not for them a mental state. We might even say that in their case fear is not an emotion, despite the fact that it shares elements of the emotion of fear as it occurs in us. Furthermore, because they lack concepts, animals can be afraid of a much more limited range of things than we can. Perhaps it is because our carings are more complex than those of other animals that they wonder at us as the laughing, weeping creatures who have lost our 'healthy animal reason'.

Every mind is a brain (though not merely), but not every brain is a mind. This is because the concept of mind is needed to capture the special complexity that enters into the description and explanation of behavior of creatures who, like us, can be self-reflective and have thoughts of a propositional character. The fact that beliefs enter into the 'higher' human emotions greatly expands the emotional repertoire available to us: whereas animals can have beliefs only about real, physical objects, the objects of human beliefs may be observed or unobservable; physical or mental; real or imagined: "If we had taken the left turn instead of the right we would be home by now"; "I am an evil person for wishing him harm"; "Oz must be beautiful"; "I am afraid of dying"; "Perhaps my success is responsible for your failure".

Furthermore, need, wish, and emotion often inspire in us flights of imagination, as they surely do not in other animals, that in turn engender other emotions. Any adequate theory of the emotions must allow for some continuity between human and other creatures, and also for some enormous gaps; it must also allow for the intricacies of the mind-body question.

Darwin emphasized the continuum linking us to other animals, addressing as he did so the behavioral manifestations of emotions. But as I noted earlier, we know that not all emotions have an outward show. Even if they did, describing it would tell us little about what an emotion is. This is the question addressed by the introspectionist psychologists, Wundt and Titchner. The study of an emotion, they held, should consist in a systematic phenomenological description of one's state of consciousness during the experience of an emotion. Darwin told us how an emotion looks, the introspectionists how it feels. But their story doesn't get at such basic conceptual questions as we have been discussing: What makes an emotion a mental state? Of what mental elements does it consist? What distinguishes the mental state of emotion from other mental states like belief and desire? (We often have beliefs and even desires in the absence of emotion.)

James argued against the introspectionists that they left out the body, and he was equally critical of the 'classical' views of Locke and Hume for their 'atomistic' view of mental life. Locke assumed there are primary states of mind which consist in simple sensations, themselves  of two kinds: sensations of external objects and sensations of 'internal' things like pleasure and pain. Sensations of sight, sound, smell, and so on, belong to the former; basic emotions to the latter. Locke and Hume as well thought that these individual units of thought recur in different combinations and configurations to make the emotions in all their great variety. [Footnote 4]

But our emotions and sensations are no more made up of atoms of sensation than rivers are made up of individual drops of water, James protested. Just as one cannot step into the same river twice, so a sensation is also an unrepeatable experience, part of a stream of consciousness. Or to change the metaphor, every mental event is part of an unduplicable life narrative.  "The trouble with the emotions in psychology," James writes, "is that they are regarded too much as absolutely individual things... But if we regard them as products of more general causes... the mere distinguishing and cataloguing become of subsidiary importance". Is it not, indeed, because of their embeddedness in such a life narrative, made possible by our cognitive capacities, that emotions are mental states?

James' own view of the emotions is perverse, though it has been and remains highly influential.  On the common view, as we saw, emotions have a causal role in our explanations of behavior: the causal direction goes from the mental perception of some fact to the mental state we call the emotion, then from the emotion to the bodily expression, either in the form of blushing, tears, increased heart rate, or voluntary action. James argues that this has things backwards. The causal direction is not from mind to body, but body to mind. Something causes us to blush, grow pale, tingle. Then if and when we become aware of this physiological state, we experience embarrassment, excitement, fear. 

James is not necessarily denying that beliefs and desires have a role to play in the emotions. Perhaps he would agree that in explaining Maria's anger our first step might be to say that she has seen or heard you do something she takes to be an insult. But since this might happen without her experiencing an emotion, our quarry has escaped our net. James argues: "Without the bodily states following on the perception, the latter would be purely cognitive in form, pale, colorless, destitute of emotional warmth" (p. 243). On James' view, there is no emotion without physiological arousal, and feeling just is the perception of the arousal.  It is fine to infer from someone's weeping that he is sad, not because sad thoughts are often expressed in weeping but because weeping often causes sadness.

James is right to insist that feelings are felt and are in that sense necessarily conscious. He is also right to call attention, under a different terminology, to the distinction between emotional states and emotional dispositions that I drew earlier. "An emotion," James writes, "is a tendency to feel, and an instinct is a tendency to act, characteristically, when in presence of a certain object in the environment" (p. 240). On James' account, instinct operates as follows: A creature becomes aware of an object, internal or external, absent or present, real or imaginary. Because of the creature's instinctual endowment, the awareness triggers an impulse, which spreads through the body, expressing itself successively in excitation, overt bodily expression, and finally, perhaps, voluntary muscular action. Thus feelings occur within a dispositional structure. As Wollheim puts it, "having called feeling emotion, James now calls this dispositional structure not emotion, but instinct".

The idea of emotions as dispositions opens the door to Freud, who is often taken to have denied that emotions can be unconscious.  If, as he claimed, emotion is a discharge of instinctual energy, then, it is argued, no emotion could occur without the person's being to some extent aware of it. The passage that is often cited is the following from his 1915 Chapter on "The Unconscious":

"Strictly speaking... there are no unconscious affects as there are unconscious ideas. But there may very well be in the system Ucs. affective structures which like others become conscious. The whole difference arises from the fact that ideas are cathexes... whilst affects and emotions correspond to processes of discharge, the final manifestations of which are perceived as feelings."

Freud is clearly here denying that the states of mind to which the term 'unconscious emotion' applies are analogous to the states to which 'emotion' in its standard use applies; in this respect, unconscious emotions are unlike unconscious ideas. But Freud is not denying that  'unconscious emotion' is applicable to any state of mind.  In the text preceding the passage above, Freud writes:

"·in comparison with unconscious ideas there is the important difference that unconscious ideas continue to exist after repression as actual structures in the system Ucs., whereas all that corresponds in the system to unconscious affects is a potential beginning which is prevented from developing (p. 178)."

An unconscious emotion is a mental state that is potentially conscious, that can be inhibited from becoming conscious, that can interact with other mental states so as to qualify their meaning, and that can motivate both thought and behavior in the absence of consciousness.

On Freud's view the story of what any particular emotion is, along the path it travels from unconsciousness to full awareness, has many twists and turns. For one thing, an affect may have been in the first place pre-verbal and pre-symbolic, like the early traumatic anxiety provoked by a prolonged absence of the mother; yet it may haunt the mind, affecting later emotions, beliefs, perceptions, desires, motivations. Then, thoughts of all kinds, including those that enter into the emotions, are subject to distortion, condensation, and displacement, not necessarily because of repression, but because the mind has different ways of working. Finally, there are the tricks of memory, the way in which, for example, earlier events may be 'remembered' in the light of ones that occurred later (nachträglichkeit), confounding the events of one time, with all their emotional associations, with the events of another, and giving them a particular shading.

All these ideas of Freud's bring me to the recent neurobiological accounts of LeDoux.

A model for emotional irrationality and emotional learning

Work by LeDoux and others clarifies both the psychoanalytic view of the emotions and the common view sketched above. LeDoux's work is primarily about fear, but he believes that his account generalizes to all the emotions. His basic thesis is that the emotional system is in charge of appraising the meaning of a stimulus in terms of one's own welfare or the welfare of something or someone one cares about, and that much of this appraisal may go on beneath conscious awareness. That is, the organism can evaluate the emotional meaning of an event as 'good' or 'bad' before it knows just what the event is and what is bad about it. [Footnote 5] Most emotional processing takes place out of conscious awareness; feelings emerge only as a byproduct, "frills that have added icing to the emotional cake". Yet he also writes that "the capacity to have feelings is directly tied to the capacity to be consciously aware of one's self and the relation of oneself to the rest of the world" (p. 125), a remark which suggests a more substantial role for feelings in  human life than the pastry metaphor suggests.

On the account generally accepted now by neurobiologists, there are two neural pathways from stimulus, let's say a snake in the road, to the amygdala, the part of the brain that causes visceral and behavioral emotional responses. Often these two paths converge; but they need not. The 'low road' is more direct; and it is older in evolutionary terms in that it does not require cognition. The 'high road' goes from stimulus to amygdala via the cortex, the 'thinking' part of the brain. The high road is not required for some types of fear, or rage, or sadness; but it is required for such cognitively sophisticated emotions as guilt and pride. Emotional processing thus has an immediate and a mediated form: the former, 'Type I', though distinct from 'Type II', contributes to it. Type I emotions are immediate and a product of the evolutionary experience of the species; we share them with other animals. Type II are variable, specific to us, subject to learning, and to voluntary control.  Memory enters into both.

It seems that memory systems are of different kinds. One kind contains implicit, procedural rules that may have been learned prior to symbol formation. Other kinds of implicit memories involve particular ways in which memories are organized, for instance, by the kinds of associations that Freud called dreamwork. [Footnote 6]  Another system contains episodic, declarative memories, that typically presume a grasp of symbols and that are conscious. 

The relevance of these distinctions to the emotions is that one may have explicit, declarative memories of emotionally laden experiences  without feeling anything now. It is with implicit memory that feeling kicks in. You do not merely recall that you felt anxious when you found yourself alone in the dark street; you feel anxious. And if the memory is explicit, you also know what it was you feared then, and can appraise whether there is reason to be fearful now.

In the best of cases, implicit and explicit emotional memories meet in working memory where they can be refined and modified. But the gap between our advanced cognitive capacities and the older emotional and motivational systems, makes such retrospective shaping difficult. Furthermore, working memory can itself be disturbed by chronic or extreme stress; the emotional triggers or their implications may go unnoticed; and events that were merely part of the context of something dangerous in the past are not now discriminated from the danger itself, a discrimination which it is up to the more advanced cognitive capacities to make.

I take LeDoux's principal points to be as follows (some of them refer to ideas I spoke of earlier in introducing earlier 'the common view' of the emotions). 1) In both man and other animals, perception plays a principal role in the emotional system. (There is, I believe, a slight over-emphasis in Le Doux on the relations between emotion and external world. Feelings that arise from within, say hunger, or pain, can also give rise to fear and anxiety.) 2)These perceptions need not be conscious or verbal. 3) In the first instance the perceptions are of things in the external world, which are appraised in terms of their importance to the organism. 4) Emotions, therefore, vitally engage us with the external world, and cannot be construed as merely internal interruptions. 5) The appraisals may be automatic, unconscious, non-verbal, implicit, and non-declarative; but for creatures capable of propositional thought they can be at the same time declarative, cognitively laden, triggered by belief, and conscious. 6) Emotions have a history, the complexity of which will reflect the mental complexity of the organism.  7) Their biological function is to allow us to act quickly in the presence of danger. Thus they form part of the motivational system. 8) But because of the ways in which emotions have a history and are embedded in contexts of memory and association, they can be maladaptive, leading the organism to respond to something as a danger which in fact is not. 8) They are subject to learning, though to a limited extent.

LeDoux's account of the emotions harmonizes with Freud's: something may be going in us emotionally that we cannot feel;  we may have a feeling we cannot understand;  we may misconstrue an event as 'good' for us or as 'bad' not on the basis of a present appraisal of its value but because of its history and its associations. It is not only the external world that is appraised, but the internal world as well: one's own emotions, beliefs, and desire, can themselves seem dangerous. In either case, memory may say, 'Danger!', even though the danger in mind is over and past. There is a signal of anxiety, barely registered, and then the various defense mechanisms against registering the anxiety in full conscious awareness are put into play.

If, however, the anxiety can be put into words, then misplaced, destructive emotional triggers can often be identified and given an appropriate context. Reflection can go to work; and with luck the emotional triggers and the automatic defenses against them can be unlearned. But the situation must be right: the emotions must be aroused, that is, felt, not avoided through a nameless anxiety. They must be aroused in an environment safe enough so that they do not take over, and so that the agent can notice how he defends himself against them. This is just the environment that the analytic situation attempts to provide. A space is made in which the emotions can be both experienced in feeling, then clarified, and appraised. In this process of emotional growth, the past is acknowledged as past; old fears no longer prompt automatic, now inappropriate responses; one's emotions themselves are more finely discriminated; freed from the constraints of the past, one's emotional vocabulary expands.

Objectivity and subjectivity

James held that feeling is the perception of an inner, subjective, emotional state. Mark Solms takes a similar position:

"Only you can feel your emotions. This also applies to consciousness in general... but it applies to emotion in a special way. It is not only the perception of emotion that is subjective. What emotion perceives is subjective too. What you perceive when you feel an emotion is your own subjective experience of an event - not the event itself."

And Le Doux:

"My idea about the nature of conscious emotional experiences, emotional feelings, is incredibly simple. It is that a subjective emotional experience, like the feeling of being afraid, results when we become consciously aware that an emotion system of the brain, like the defense system, is active (1996, p. 268)."

Of course both the feeling and the emotion are subjective; but this doesn't mean that either emotion or feeling is a perception of an 'inner', subjective event. Let's say one has a  visual perception of the snake in the path, the angry expression on the father's face, the shaming parent. Emotion just is what the perception arouses, a perception that often ends in feeling.

Emotions often take propositions as their objects - 'I feel ashamed that I want to embarrass you', '·proud of your success.'  In such cases emotion has  intentionality. But emotions can also be directed immediately to persons and things: in loving you it may be you that I love you, and not some fact about you, even though I see you under certain descriptions that have to do with what I love about you: as witty, kind, forceful. Even when the emotion does have an intentional object, to construe the emotion simply as a propositional attitude would omit the subjective, first-person, narrative, phenomenal character of an emotion, how it feels to be disposed toward the world, or a part of it, in an angry or a sad or a jealous way. 'The glass that is half full' and 'the glass that is half empty' describe the same glass; but the two perceptions may carry a very different emotional quality. Someone who feels deprived, forlorn, emptied out, may see the glass as empty, that is, half-empty; another person who feels grateful for what he has is more inclined to see the glass as full, that is, half-full. The presence of intentionality distinguishes emotions from the kinds of feelings that are bodily pain; the peculiar nature of their intentionality, and the fact that sometimes it is absent altogether, distinguishes emotions from beliefs and desires (Goldie).

To get at the peculiar nature of the intentionality inherent in many emotions we might ask, as Wollheim does, what the respective roles in the mind of desire, belief, and emotion are. The role of belief, he answers, is to provide the creature with "a picture of the world it inhabits" (p. 13) (see Williams, 1973). Not just any picture, but one depicting the world more or less as it is. The role of desire is to provide the creature with objective ends, things at which to aim. Desire is not simply an internal drive, as Freud sometimes says, but a response to something outside that attracts us. Now, Wollheim continues, "the stage is set for the emotions. The role of emotion is to provide the creature... with an orientation, or an attitude to the world. If belief maps the world, and desire targets it, emotion tints or colors it: it enlivens it or darkens it, as the case may be" (p. 15).

Moran similarly claims that in comparison to belief  "an emotional attitude constitutes something closer to a total orientation of the self, the inhabiting of a particular perspective". An emotion may set the horizon, as it were, against which we see the world. More than beliefs and desires, emotions are what make my world peculiarly mine, and yours peculiarly yours.

I have now answered two of the questions posed above. The first was: What makes an emotion a mental state? I have answered that an emotion is a mental state whenever it has a cognitive component that itself counts as a mental state. Where people disagree about what counts as mental, they presumably will also disagree on what counts as an emotion.

The second question was: How does an emotion differ from other mental states like belief and desire? The suggestion is that emotions, unlike beliefs and desires, are total orientations toward the world. They have a narrative character that burrows deep into our individual histories and that disposes us both to perceive and to feel about the world in a certain way. incline. They are, I believe, the clue to what we mean by subjectivity.

Does the fact that emotions are peculiarly subjective suggest that they restrict what we see? Certainly sometimes. Anger, greed, jealousy, tend to narrow one's focus. By the end of the play, Macbeth's unbridled ambition and his ever-increasing guilt have drawn his interest away from the world entirely, imprisoning him in madness. Are emotions our impositions on reality? Sometimes, certainly. Once unleashed, Othello's jealousy finds confirmation everywhere.

Sartre argues that an emotion is an active "transformation of the world" that we imaginatively perform in the face of an obstacle to desire:

"When the paths traced out become too difficult, or when we see no path, we can no longer live in so urgent and difficult a world. All the ways are barred. However, we must act. So we try to change the world. That is, to live as if the connection between things and their potentialities were not ruled by deterministic procedures,  but by magic."

Sartre compares fear with looking for a face concealed in a picture puzzle. Through a change in attention - turn of our eyes, the focusing on the upper left-hand corner -  we apprehend an object in a new way. Emotion, Sartre suggests, should be conceived along the same lines. It comes into play with frustration, transforming the meaning of the world so that it is more tractable to desire: I am unable to have the grapes I want, so I perceive them as sour, no longer worth having. Or I flee from a bear in order to escape, making the bear, as it were, disappear. Or, faced with personal ruin, instead of seeking ways to deal with it, I become sad: sadness "aims at eliminating the obligation to seek new ways" of approaching the world (p. 65). In general, Sartre claims, emotions are defenses against those aspects of the world that challenge our desire.

Sartre is surely right that emotions are aroused by one's appraisals of the world in terms of its values to oneself; right that emotions tend to transform the way the world looks and feels to us; right about the person who, repeating the past in a neurotic way, looks for confirming evidence that he is once again being neglected, or insulted. But emotions can disclose as well as foreclose reality. To return to the simple example of the snake in the road: If I am a creature to whom snakes are apt to be truly dangerous, my fear is appropriate, and it discloses an aspect of reality in its relation to my interests. If I do not feel afraid of the snake, something may be wrong with me. Or I may feel guilty because in fact I have harmed you; my feeling of guilt is a way of acknowledging, taking in, that harm. Or I am angry because I perceive, correctly, that you are trying to belittle me. Or I am sad because you are leaving.

As friends, moralists, therapists, we sometimes feel that another's emotions are too strong for the occasion, or misplaced. But equally often we feel that another lacks an emotion it would be appropriate for her to have. Emotion fails her perhaps because she does not see, or will not see, the harm she has done, or that has been done to her; or perhaps because something inhibits her response.

I am unhappy with the implication of the second suggestion, however, which is that one's emotions may be undeveloped though one's perceptions of the relevant occasion be perfectly clear. I want to say rather that if you don't feel fear, or guilt, or anger, where it's called for, you haven't really taken in the facts of the matter. Inhibiting the feeling is also inhibiting the perception, the knowing. Let's say you know perfectly well, in the ordinary sense of 'know', that someone you love is gone; but you feel little or nothing. Then you haven't really acknowledged the loss. Mourning is perhaps the paradigmatic process of taking in what you know.

My reader may object: But surely whether guilt or anger is called for is itself a subjective judgment; who's to say which feeling is appropriate to what situation, and when? This is, in effect, the large venerable question: Are value judgments 'objective' or 'subjective'. I will not say much about it here; but I want to suggest two positions from which it might be considered.

The first inquires into how we learn an emotional vocabulary in the first place. By an emotional vocabulary I mean both a semantics of the emotions, and also an emotional  repertoire. Take the first: We learn words for emotions only in our communications with other people, as we are responding to a shared, common world, out of interests that we also hold to some extent in common. If you prick yourself with a sharp pin and say 'Ouch', I know what you are talking about if I have also had run-ins with pins that are sharp.  We are able to acquire the words for pain, sorrow, shame, because they are expressed in outward behavior.

The word 'express' here is tricky: Are the emotions, beliefs, and desires that cause the grimace or a frown related only contingently to that outward expression? Or is there some more essential connection? Can we imagine, for example, a world in which the facial expression we now identify as fear expressed joy instead?

Wollheim writes:

"Let us note that there is a hidden complexity to this seemingly obvious issue. When we think of joy and smiling as expressively linked, the link has two aspects, neither of which has priority over the other: We link joy at once with the smile on the face, which we know from the outside, and with the activity of smiling, which we know from the inside. However these two aspects of the link - are themselves linked. For the bodily activity of smiling is partly identified as the process that issues in the facial configuration of the smile. Conversely, the facial configuration of the smile cannot be fully identified except as the product that issues from the bodily activity of smiling· This gives us the scale of the changes required if the prevailing expressive links are to reconfigure."

People can attempt to mask their fear, or misery, or anger, with a smile. But if the facial expression we now identify as fear were to be identified instead as joy, such massive changes in our perceptions of others and ourselves, including our proprioceptive feelings, would be required that we would simply have to be living in a different world. The link between the 'inner' and the 'outer' is nowhere clearer than in the emotions. It is such that we cannot think of the outward behavior and of the body as something that just happens to hook up in some ways with the emotions.

Let's turn from the naming of emotions to the matter of emotional repertoire, what the emotions are of which any one person is capable. We are tempted to think of emotions as belonging so entirely to our 'inner' lives that our communication with others plays no role in our acquiring them.  But as we learn to discriminate those characteristics of a situation that make it an occasion for feeling pride, or shame versus guilt, or envy versus jealousy, we also become capable of feeling pride, shame, guilt, envy, jealousy; as our discriminations of the situations in the world become finer and more various, so also may the emotions to which we are susceptible.  Conversations of all kinds - with friends, analysts, authors - may further emotional growth.

Aristotle held that character is acquired through habit, which is trained in the nursery. That is where we learn to curb our anger. It is also where envy and resentment, or on the other hand, kindness and empathy, are bred; where we learn to distinguish, if we do, between harm intentionally and accidentally done, between failing someone else's standards and failing our own, between a gift with a price tag and a gift of love, between a blow to our vanity and a mortal injury, between shame, embarrassment, and guilt.

The second position calls into question the distinction itself between the valuing, emotional, thinking, feeling subject, and an objective world that is value-neutral. It is a distinction that many philosophers in this century have attempted, successfully I think, to undermine (I have particularly in mind Heidegger and Davidson.) But though Freud was not aware of it, the psychoanalytic picture of the relationship between mind and world also points the way to new ideas of subject and object, subjectivity and objectivity.

Look for a moment at subjectivity as Descartes conceives it. The Cartesian I is already an accomplished thinker when he begins his epistemological quest for certainty. But he finds that nothing can be known for certain: everything he thinks he knows he finds he can doubt. The one indubitably existing thing, he discovers,  is his I, not his self as an object in the world but as a mere subject of thought, the rememberer remembering, the doubter having doubts. This I  is a thing that has thoughts, that can question, form images, have desires, doubt the existence of everything beyond himself. So in discovering his I Descartes also discovers that he is radically cut off from the objective world that he longs to know: he can know it neither through acquaintance, nor through propositional knowledge about it. He can posit the existence of an objective world, but it is forever beyond his grasp.

We think of the Cartesian I as a point of radical subjectivity; yet ironically, as a Cartesian subject I can have no knowledge of my inner world, since that requires being able to distinguish inner from outer; and this is just the distinction I am incapable of making. If I'm lucky, thoughts in my head correspond to things in the external world, but only God can know when this correspondence holds.

Contrast this Cartesian story with Freud's. Descartes assumed, as have most philosophers after him, that we need know nothing about the history of the I in order to know what the I is. The I is somehow beyond space and time. Freud's story, on the other hand, is a developmental story: the I that can ask, What can I know?, has a history which is an essential part of what the I as questioner is now. I become an I through an ever more complex process of identifying with the valuing, loving, and loved object, at first barely registered as an object. This I begins by taking in the object, the real object; that is, fantasies of incorporation, if such there are at a primitive developmental stage, are spun around an object - say the breast -  that truly exists outside the infant's self. Psychologically speaking, there must be for the infant a proto-I, and equally a proto-object, for neither could arise out of a unity lacking all differentiation.

This proto- I becomes ever more aware that its objects are separate from itself. And as the separation-individuation process unfolds, the object is increasingly loved not only for what the I can get from it, but for its own sake, because the object is lovable. I take this to be the developmental importance of gratitude: it feelingly acknowledges certain fundamental facts about one's existence in the world, among them that one is incomplete and dependent on the object, without a fall into the self-hatred of envy; and it feelingly acknowledges the goodness of the given and the giver.

The Freudian I requires pleasure and pain, love and loss, for its development. It is formed through such emotions as anxiety, anger, sadness, hope, fear, envy, and gratitude. Without them the I could not distinguish 'inner' from 'outer', fantasy from reality. The earliest proto- I begins when the infant discovers that what he wants most of all, the breast, sometimes eludes him, and that he cannot answer his needs all by himself. The I is then constructed out of increasingly complex emotional processes involving a dialectic between real bodily self and real objective world.

The concept of objectivity is essential to the Freudian picture. As knowers we can be more and less objective; we become more objective as we become more mindful of the role of our fears and wishes in our perceptions; more empathic towards others; less resistant to taking in painful facts; and less insistent on having the world be just as we want it to be. The distinctions between more and less objective can only be made from within the world of possible experiences, and these are always value laden.

So let's ask again: Is it the act itself that is just, or heroic, or stingy, or rude? Or are all such judgments really about me? Am I moved by a Bach chorale because it is moving? Or is this merely my personal response? If you do not find a Charlie Chaplin movie funny, is it because laughter is 'just a matter of taste'? Or might you be missing something in the movie?

Instead of talking about where values are,  in here or out there - an unintelligible question, to my ear -  we might instead talk about how we support evaluative judgments. Sometimes and in some cases they can be argued along rational lines, and people can be brought to value things that they did not before. Insofar as a statement attributing a quality to a thing can be rationally defended, it is, in an important sense of the word, objective. Values, including emotional values, are more public than our skepticism about the objectivity of value suggests, a conclusion that was implied in my earlier remarks about how one learns both to name and to have emotions.

We might think of the emotions as a kind of sixth sense, allowing us to perceive aspects of the world that are perceptible in no other way. When a situation amuses me, or makes me sad, or indignant, I may be responding to qualities of it that cannot be articulated in words and that you may not be picking up if it leaves you cold.  If the emotions do constitute such a sense, then an impoverished emotional life and an impoverished perception of reality go hand in hand.

A consensus has been growing in psychoanalytic theory that places emotion (rather than 'drive') at the heart of a theory of motivation (on this point, see Westen, 1997). This emerging theory accommodates the fact that the child does not come equipped with a fixed repertoire of motives; it helps us understand both the flexibility and the fixity of human motivation; and, most important, it sees the individual as embedded in the world, becoming what she is only through interactions with other persons and things beyond her own skin.

In the name of reason we have been at war with the emotions for years. Passions 'blind', 'oppress' wrench us from our true selves ('I am beside  myself  with anger - grief - worry'). Hume said that reason is and must always be the slave of the passions, suggesting a friendlier view of the emotions but one that remains adversarial. In defense of 'reason' we suppress and repress, join monasteries, disown and deny, pray, repent and punish, ignore.

Without question the emotions frequently distort and mislead perception. And though they arise from the self and were surely designed by evolution to serve the self's interests, they are frequently destructive of both personal happiness and interpersonal solidarity. But as Elster eloquently remarks: "Emotions matter because if we didn't have them, nothing else would matter" (p. 403). Turn it around: emotion is how mattering shows up; the emotions help us know what matters to us, and sometimes why. They call into question a picture of reason itself, reason as cold, unmotivated, at a distance from the world, detached from the body; of objectivity as a stance bereft of feeling. The emotions are that aspect of our mental life which gives us the best understanding of subjectivity, understood as what it is that makes one's experience uniquely hers, and often so difficult to capture in words.


Damasio A.R. (1994). Descartes' Error: Emotion, Reason, and the Human Brain. New York: Avon Books.

de Sousa R. (1987). The Rationality of Emotions. Cambridge, MA: MIT Press.

Deigh J. (2001). "Emotions: The Legacy of James and Freud." The International Journal of Psychoanalysis, 82 (6): 1247-1256.

Ekman P. (1998). Introduction. The Expression of the Emotions in Man and Animals. Oxford: Oxford University Press.

Ekman P. & Davidson R. (1994). The Nature of Emotions, Fundamental Questions. New York: Oxford University Press.

Elster J. (1999). Alchemies of the Mind. Cambridge: Cambridge University Press.

Fonagy P., Gergely G., Jurist E.L. & Target M. (2002). Affect Regulation, Mentalization, and the Development of the Self. New York: Other Press.

Freud S. (1896). The Aetiology of Hysteria. S.E., 3.

Freud S. (1900). The Interpretation  of Dreams. S.E., 5.

Freud S. (1905). Three Essays on the Theory of Sexuality. New York: Basic Books.

Freud S. (1909). Notes on a Case of Obsessional Neurosis. S.E., 10.

Freud S. (1915). The Unconscious. S.E., 14.

Freud S. (1926). Inhibitions, Symptoms and Anxiety S.E., 20.

Goldie P. (2000). The Emotions, A Philosophical Exploration. Oxford: Clarendon Press.

Gordon R.M. (1987). The Structure of Emotions. New York: Cambridge University Press.

Griffiths P.E. (1997). What Emotions Really Are, The Problem of Psychological Categories. Chicago: University of Chicago Press.

Jacobson J. (1994). "Signal Affects and our Psychoanalytic Confusion of Tongues." Journal of the American Psychoanalytic Association, 42 (1): 15-42.

James W. (1884). "What Is an Emotion?" Mind, 9: 188-205.

James W. (1961). Psychology, The Briefer Course. New York: Harper & Row.

LeDoux J., Ed. (1998). The Emotional Brain. New York: Touchstone.

Moran R. (1988). "Making Up Your Mind: Self-Interpretation and Self-Constitution." Ratio (New Series): 135-151.

Nussbaum M. (1986). The Fragility of Goodness. Cambridge: Cambridge University Press.

Nussbaum M. (1994). The Therapy of Desire. Princeton: Princeton University Press.

Sartre J.-P. (1948). The Emotions, Outline of a Theory. New York: Philosophical Library.

Solms M. & Turnbull O. (2002). The Brain and the Inner World, An Introduction to the Neuroscience of Subjective Experience. New York: Other Press.

Stern D. (1977). The First Relationship: Infant and Mother. Cambridge: Harvard University Press.

Werner H. (1948). Comparative Psychology of Mental Development. New York: International Universities Press.

Westen D. (1997). Towards a Clinically and Empirically Sound Theory of Motivation. The International Journal of Psychoanalysis, 78 (3): 521-549.

Williams B. (1973). Deciding to Believe. Problems of the Self. New York and Cambridge: Cambridge University Press.

Wollheim R. (1999). On the Emotions. New Haven: Yale University Press.


1 This is Griffiths' (1997, #658) argument. He suggests that the concept of emotion may  resemble the concept of sublunary objects in Aristotle's day. There are such objects, but they have nothing especially in common to distinguish them from other arbitrary collections. On Griffiths' view the category of emotion subsumes three different kinds of psychological state: socially sustained pretenses, 'irruptive motivational complexes' that require higher cognition, and affect program responses like fear and anger. Some of these emotion categories, like anger, themselves reflect these fractures. By socially sustained pretenses Griffiths means a kind of "disclaimed action", like "falling in love," which is the adoption of a social role that permits the performance of certain behaviors (p. 245). This categorization omits the sorts of emotional dispositions on which I will focus, and underplays, I believe, the role of reason in our emotional life, or as I would rather put it, the extent to which an understanding of emotion must affect our idea of reason itself.

2 There are many ways of drawing this familiar philosophical distinction. Mine is indebted to Richard Wollheim.

3 See Paul Griffiths for a discussion of the strengths and weaknesses of affect program theory {Griffiths, 1997 #658}.

4 Hume distinguished between sensations, or impressions, and ideas, which he held to differ from sensations only in vivacity.

5 Damasio speaks of emotion as a combination of a mental evaluative process and a dispositional response to that process. According to his 'somatic marker' hypothesis, emotions are an essential part of practical reason, giving the creature a sense of how the world relates to its goals and projects. Something more than a merely intellectual rehearsal of possible outcomes enters our deliberations about what to do, and this something more is feelings. Somatic markers are feelings generated by what Damasio calls secondary emotions, the ones that learning has connected, via the high road LeDoux describes, to certain situations. Somatic markers assist the reasoning process by highlighting some options as dangerous, others as favorable (Damasio 1994, p. 174). If in the past an alternative similar to the one we are now considering had been connected with something 'bad', then, however fleetingly, we may experience an unpleasant gut feeling, which functions like a signal of anxiety, quickly alerting us to danger. This is the wisdom in the advice to trust our feelings. But feelings can also be untrustworthy in all the ways suggested by the vicissitudes of emotional memory.

6 Werner's idea of a syncretic mode of thought is similar: "If several mental functions or phenomena, which would appear as distinct from each other in a mature state of consciousness, are merged without differentiation into one phenomenon, we may speak of a syncretic function or a syncretic phenomenon" (Werner H., Comparative Psychology of Mental Development. New York: International Universities Press, 1948).

Marcia Cavell, Ph.D.
1570 Olympus Avenue
Berkeley, CA 94708
E-Mail <>

Web Editor: Paolo Migone. For suggestions or corrections, please e-mail to: