Published in the Journal of Aesthetics and Art Criticism, vol. 13 (1954-55): 203-214. Transcribed into hypertext by Andrew Chrucky, Feb. 21, 2001.

    Emotions, as I shall maintain, are cognitions with a certain kind of psychical quality. Cognitions are a subclass of experiences. I shall therefore begin by classifying experiences in general and cognitions in particular. It is only after that has been done that one can profitably discuss emotions and sentiments.


    Experiences may be divided into those which do not and those which do have an 'epistemological object.' The former may be called Pure Feelings. The natural question to ask with regard to a feeling is: 'How are you feeling?' And the natural answer is to utter some adjective (or, more properly, adverb), such as 'Hot,' or 'Tired,' or 'Cross.' To feel tired is to be feeling in a certain way; it is not to be aware of a certain object, real or fictitious. On the other hand, there are many experiences about which it is natural to ask: 'What is the object of your experience?' or 'What is it about?' If a person says that he is seeing or hearing or thinking, it is natural to ask: 'What are you seeing?' or 'What are you hearing?' or 'What are you thinking about? And the answer that one expects is the utterance of some substantive or phrase equivalent to a substantive, e.g., 'A red flash,' 'A squeaky noise,' 'The square-root of minus 1.' I shall say that experiences of the latter kind 'have an epistemological object' or are 'epistemologically intentional.' All such experiences may be called Cognitions.

    It is important to notice that an experience may be epistemologically intentional, even if it be a delusive quasi-perception or a thought of something which does not and perhaps could not exist. A person who in a dream ostensibly sees a man pointing a revolver at him is having an epistemologically intentional experience, although there is no ontological object (i.e., no actual man, pointing an actual revolver at him at the time) corresponding to it. Similarly, a person who is thinking of a phoenix is having an epistemologically intentional experience. He is certainly thinking of something and he could describe what he is thinking of. If he were thinking of a dragon, instead of a phoenix, he would be thinking of something different and would give a different description. And that, in spite of the fact that there never have been, and perhaps could not be, in nature either phoenixes or dragons.

    So we begin by dividing experiences into Pure Feelings and Cognitions. The former are those which have only psychical qualities and do not have epistemological objects. The latter are those which have epistemological objects. Cognitions may have psychical qualities as well as epistemological objects; some of them certainly do, and perhaps all of them do. A pure feeling cannot significantly be described either as veridical or as delusive. These alternatives can be significantly predicated only of cognitions. A cognition is veridical if there is an ontological object answering to the description which the experient would naturally give of its epistemological object, i.e., the description which he would offer in answer to the question: 'What are you cognising?' It is totally delusive, if there is no ontological object answering even remotely to this. It is more or less delusive, if there is an ontological object which answers in certain respects more or lees closely to the description which the experient would naturally give of its epistemological object, but which fails in other respects to do so.


    For our purpose cognitions may be sub-divided into intuitive, perceptual, and conceptual. Intuitive cognition is direct prehension of particular existents. These always present themselves to the person who prehends them as having certain qualities, e.g., redness, squeakiness, etc., or as standing in certain mutual relations, e.g., spatial adjunction or separation, temporal overlapping or complete sequence, and so on. So far as we know, a human being is capable of rehending particulars of three and only three kinds, viz., sensibilia (i.e., color expanses, sounds, smells, etc.), his own mental images, and his own experiences.

    Perceptual cognition may be described as cognition of particular existents which seems prima facie to be purely intuitive, but which is found on careful consideration to be not wholly so. It always involves prehension of particulars, but it also involves non-inferential beliefs or quasi-beliefs, which are psychologically based on that prehension, but go beyond the information which it by itself supplies. The three most important kinds of perceptual cognition (when thus defined) are Sense-perception, Reminiscence, and Self-perception. The intuitive bases of these are respectively the sensing of sensibilia, the imaging of mental images, and reflexive acquaintance with one's own experiences. In each case the presence of intuitive cognition, and the absence of explicit inference or even of a noticeable process of associative transition, is likely to make it seem that the cognition is wholly intuitive.

    Under the head of conceptual cognition I include all those processes which operate with general ideas or abstract concepts. By means of it an individual can think of things and persons and events and situations which he is not prehending and is not perceiving or remembering. He does this by thinking of a certain combination of characteristics, which together constitute a description of a certain possible thing or person or event or situation. He then thinks of the object as 'a so-and-so' or as 'the so-and-so' which answers to this description. We can of course imagine or suppose that there is something answering to a certain description without actually believing that there is. We can do so when we positively know that there is not. This happens, e.g., in either composing or understanding an admittedly fictitious narrative. A great deal of cognition, which seems prima facie to be purely perceptual, turns out on closer inspection to be partly conceptual. It seems likely that all one's cognitions of other persons' minds and of their experiences consists of conceptual cognition based on one's perception of their bodies, their gestures, their speech, and so on.


    We are now in a position to consider the nature and the subdivisions of Emotion. Suppose that a person were to say: 'I am having an emotion.' Then there are two questions which it would be sensible to ask: — (1) 'What kind of emotion?,' and (2) 'Towards what object?.' The answer that we should expect to the first question would be: 'One of hatred,"One of fear,' and so on. The answer we should expect to the second would be: 'Towards Smith,' 'Towards a ghost,' and so on.

    Every emotion is an epistemologically objective or intentional experience, i.e., it is always a cognition, either veridical or wholly or partly delusive. But every emotion is something more than a mere cognition. An emotion is a cognition which has one or more of the specific forms of a certain generic kind of psychical quality which we will call emotional tone. To be fearing a snake, e.g., is to be cognising something — correctly or incorrectly — as a snake, and for that cognition to be toned with fearfulness. In general, to be fearing X is to be cognising X fearingly; to be admiring X is to be cognising X admiringly; and so on.

    An emotion, then, as I have defined it, always has an epistemological object. But, corresponding to the various kinds of emotion, there are certain experiences called emotional moods. E. g., the mood which corresponds to the emotion of anger is crossness. One may feel cross without being angry with anyone or anything, and one may feel alarmed without being frightened at anyone or anything. I think that an emotional mood is either a pure feeling or else an emotionally toned cognition with an extremely vague indeterminate object. It might, e.g., be one's cognition of things in general or of one's present total environment. The connection between an emotional mood and the corresponding emotion is this. The pure feeling or the extremely vague cognition, which is the emotional mood, has the same kind of emotional tone as the determinate cognition which is the emotion.


    Since all emotions are cognitions, we shall expect to find a division among them corresponding to the division of cognitions into intuitive, perceptual, and conceptual.

I do not think that most purely intuitive cognitions have any marked emotional tone. But, then, purely intuitive cognition are very rare in grown persons. Intuitive cognitions occur mainly as constituents of perceptual or conceptual cognitions. Perhaps the primitive fear, which all babies are said to exhibit on hearing any loud sudden noise, such as a clap of thunder, would be an example of an emotion which is purely intuitive on the cognitive side.

    Perceptions, on the other hand, are often strongly toned with emotional qualities. One may, e.g., perceive with fear an object which one takes to be a snake; and so on.

    Almost any emotional quality which can qualify a perception can also qualify a conceptual cognition. Thus a human being can fear things or persons or events which he is not perceiving or remembering but is only expecting or believing to exist or feigning to exist. A result is that the emotions which we share with animals are felt by us towards a much wider range of objects.

    There are some kinds of emotion which, from the nature of their objects, can be felt only by a being who is capable of conceptual cognition. Hope and anxiety e.g., can be felt only by a being who can conceive alternative possible future states of affairs and anticipate them with various degrees of conviction. Religious awe can be felt only by a being who can think of the description of a supernatural power, either personal or impersonal, and can believe that there is an object answering to that description. And so on.


    I shall now consider a number of distinctions which it is important to recognize and define in discussing emotion. They are the following:

  1. Motived and unmotived emotions;
  2. Misplaced emotions;
  3. Appropriate and inappropriate emotions;
  4. First-hand and second-hand emotions; and
  5. Pure and mixed emotions.

I will take these in turn.

(1) Motived and Unmotives Emotions

    One may feel an emotion towards an object without consciously distinguishing any attributes of it with regard to which one could say: 'I feel this emotion towards that object in respect of those attributes of it.' You may, e.g., just feel an emotion of dislike in presence of a person, without being able to mention an attribute in respect of which you dislike him. But very often one can mention certain attributes, which one believes rightly or wrongly to be present in the object, and which one believes rightly or wrongly to be calling forth the emotion which one feels towards it. You may be able to say, e.g., 'I dislike so-and-so for his ugly voice and bad manners.' To dislike a person in respect of certain qualities, which one takes him to possess, is a more complex experience than just to dislike him for no assignable reason. Presumably all the emotions of animals are of the latter kind, whilst many human emotions are certainly of the former.

    I will now try to analyse these notions rather more fully. Suppose that a person feels an emotion E towards an object O, and that this appears to him to be evoked by his knowledge or belief that O has a certain attribute P. Then I shall say that this emotion is ostensibly motivated, and I shall describe P as the ostensible motivating attribute. Next suppose that a person's emotion E towards O really is evoked by his knowledge or belief that O has a certain attribute P. Then I shall say that this emotion is actually motived, and I shall describe P as the actual motivating attribute. Suppose, next, that a person's emotion E towards O does not appear to him to be evoked by any knowledge or belief that he has about the attributes of O. . Then I shall say that this emotion is ostensibly unmotivated. Suppose, lastly, that the emotion E really is not evoked by any knowledge or beliefs which the person who feels it towards O has about the attributes of O. Then I shall say that it is actually unmotived.

    We must now notice the following possibilities of mistake: (1) An ostensibly motived emotion may be really unmotived. I may think, e.g., that my emotion of dislike for Smith is motived by my knowledge that he is an atheist. But really it may be caused, not by this or by any other knowledge or beliefs that I have about his attributes, but by some peculiarity in his voice or appearance, which I have never explicitly noticed, but which arouses my dislike through some unpleasant association which it has for me.

    (2) An ostensibly motived emotion may be actually motived, but the actual motivating attribute may differ from the ostensible one. I may think, e.g., that my dislike of Smith is evoked by my knowledge that he is an atheist. But really may be evoked, not by this, but by my belief that he is a communist or by my knowledge that he is a successful rival in business.

    (3) Even if an ostensibly motived emotion is actually motived and if the ostensible motivating attribute is the same as the actual one, it may be that the object does not really possess that attribute. I may think, e.g., that my dislike Smith is evoked by my belief that he is a communist, and it may really be evoked by that belief. But the belief may be false; Smith may really be a conservative.

    (4) An ostensibly unmotived emotion may be actually motived. This can happen in two ways. (i) I may have a number of conscious beliefs and bits of knowledge about Smith's attributes, and I may think that none of them evokes my dislike of him. But I may be mistaken. It may be that one or other of them does evoke it. (ii) Even if I am correct in thinking that none of them do so, it may be that I have certain unconscious beliefs or bits of knowledge about Smith's attributes, i.e., some which exist only in a dispositional form or which for some reason I fail to notice. And it may be that one or other of these is what evokes my dislike of Smith.

    An emotion which starts by being actually unmotived will very often generate beliefs about the attributes of its object. It may thus become an ostensibly motived emotion. We shall begin to believe that the object has the kind of attributes which generally evoke that kind of emotion; and then we may begin to believe that what evokes the emotion is our knowledge that the object has these attributes. At length our belief that it has these attributes may become at least a part-cause maintaining and perhaps heightening the emotion which we feel towards it. At that stage the emotion has become, not only ostensibly motived, but to some extent actually motived. The following would be an example of this process. One may start with an unmotived emotion of love towards a person. This may in part be evoked by some obscure and quite unrecognised bodily or mental quality in him. One will then be very liable to believe that he is unusually beautiful or brave or clever. One may then come to think that one loves him because of one's awareness of these qualities in him. And eventually one's love for him may in fact be maintained partly or wholly by these beliefs about his qualities. I take it that this is at any rate part of what is meant by the word 'rationalisation.'

    Beliefs generated in this way are often false. But they are also quite often true. One may begin with an unmotived repulsion for a person. This may generate the belief that he is dishonest, and we may often find in the end that he really is a crook. On the other hand, an emotional mood, such as crossness, may be due to purely internal causes, such as a disordered liver. Once started it is very liable to crystallise into the corresponding emotion, viz., in this case anger, towards the first suitable object which happens to be available. And then it is liable to generate quite false beliefs about that object. Jealousy is the stock example of an emotion which is specially liable to generate false beliefs about its objects and thus to provide itself with motives.

    It seems to me that, when a belief about an object is generated by an emotion felt towards that object, one has generally a suspicion at the back of one's mind that it will not bear critical inspection. We tend to refuse to inspect such beliefs critically ourselves, and to feel resentment if other persons attempt to do so. In fact, beliefs about objects which are generated by emotions towards those objects are generally themselves emotionally-toned beliefs.

    Emotions which are conceptual on the cognitive side, i.e., emotionally-toned beliefs, expectations, imaginations, etc., are, I think, generally motived. If one thinks of an object which one is not perceiving or remembering and perhaps could not perceive or remember, one must do so by thinking of it as the owner of such and such qualities or as a term standing in such and such relations. If one's cognition of such an object is emotionally toned, the emotion will generally be felt in respect of some of these qualities and relations. Cf., e.g., our emotions towards Charles I with those of a person like Cromwell or Strafford, who had actually met him. We can cognise Charles I only conceptually, viz., by thinking of him as a person who had such and such qualities, stood in such and such relationships, and did and suffered such and such things. If we feel emotions towards him, they must be motived by our beliefs about his qualities and relations. But some of the emotions which Cromwell, who had met Charles I and talked with him, felt towards him, might have been evoked by certain peculiarities in his personal appearance or his voice or manner, which Cromwell had never explicitly noted. So some at least of Cromwell's emotions towards Charles I might have been unmotived, even if they were all ostensibly motived; whilst all our emotions towards Charles I are both ostensibly and actually motived.

(2) Misplaced Emotions

    An emotion may be said to be 'misplaced' if either

  1. it is felt towards an object which is believed to exist but does not really do so, or
  2. it is felt towards an object which really does exist in respect of attributes which do not really belong to it.

In the first case it may be said to be totally misplaced, in the second partially misplaced.

    Let us first consider emotions which are perceptual on the cognitive side. A perception, or at any rate a quasi-perception, may be completely hallucinatory as, e.g., a dream. In a dream one may have an hallucinatory ostensible perception of a man chasing one with a revolver, and this may be strongly toned with fear. Such an emotion is totally misplaced.

    Again, a perception may not be hallucinatory but it may be largely delusive. There may be a certain physical object corresponding to one's perception, but one may be misperceiving it to a considerable degree. One may, e.g., perceive a certain physical object, which is in fact a tree of curious shape in twilight. One may misperceive it as a man lying in wait. The perception will then be toned with fear, but the fear will be misplaced. If one had perceived the object correctly as a tree, one would not have perceived it with fear. Of course there are real qualities in the tree and its surroundings which cause one to mistake it for a man lying in wait. These are certain shapes and spatial relations, certain arrangement of light and shade, and so on. These real attributes give rise to the false belief that it has certain other attributes, which it does not in fact have. And it is the false belief that it has these latter attributes which is the immediate cause of one's perceiving it with fear.

    Let us next consider emotions which are conceptual. It is evident that these may be completely misplaced. For there may be nothing answering to the description of an object which one believes to exist, and yet the belief may have strong emotional tone. Completely hallucinatory perceptions are very rare in sane waking healthy persons. But beliefs in the existence of objects which do not in fact exist are, and have always been, quite common among sane waking men. Indeed a large part of the life of humanity has been occupied in feeling strong emotions towards beings who never existed, e.g., the gods Jupiter or Moloch or Huitzilopochtli (to go no further); or towards beings who do exist, e.g., Hitler or Stalin, in respect of attributes which they do not possess. We must notice that all emotions which are felt towards other persons in respect of their supposed mental or moral qualities must be in part conceptual on the cognitive side. For one cannot literally perceive another person's mind or his experiences or his dispositions or his motives. One can only conceive them, and we are very liable to be mistaken in our beliefs about them, and thus to have misplaced emotions.

(3) Appropriate and Inappropriate Emotions

    As we have seen, there are two aspects to every emotion. In its cognitive aspect, it is directed towards a certain object, real or imaginary, which is cognised, correctly or incorrectly, as having certain qualities and standing in certain relationships. In its affective aspect, it has an emotional quality of a certain kind and of a certain degree of intensity. Now some kinds of emotional quality are fitting and others are unfitting to a given kind of epistemological object. It is appropriate to cognise what one takes to be a threatening object with some degree of fear. It is inappropriate to cognise what one takes to be a fellow man in undeserved pain or distress with satisfaction or with amusement. Then, again, an emotion which is fitting in kind to its epistemological object, may be unfitting in degree i.e., inordinate. A degree of fear which would be appropriate to what one took to be a mad bull would be inappropriate to what one took to be an angry cow.

    It should be noticed that an emotion which is misplaced may be appropriate to its object, as that object is misperceived or misconceived. If a short-sighted person takes what is in fact a harmless but excited cow for a mad bull, it is appropriate for him to cognise it with a high degree of fear. Conversely, an emotion, which is veridical on the cognitive side, may be unfitting in kind or inordinate in degree. A woman who panics in presence of what she correctly takes to be a mouse illustrates this fact.

    This notion of a certain fittingness or unfittingness, in kind or in degree, between emotional tone and epistemological object, is plainly of the utmost importance to ethics and to aesthetics. I think that it still awaits an adequate analysis.

(4) First-hand and Second-hand Emotion

    This is an important distinction which arises in connexion with emotions which are conceptual on the cognitive side. Let us take as an example the emotion of religious awe towards God. This would be a first-hand emotion, if and only if the person who felt it was really thinking at the time of the qualities and relations which constitute a description of God, e.g., a being of infinite wisdom and power, who has created and governs and maintains everything, and if he were really believing at the time that there is something answering to that description.

    But most concepts which have been fairly often used have had names attached to them, and it is possible to use and to react to these names consistently and correctly without thinking of the characteristics which they connote. Now in many cases a certain name has become associated through early training with a certain kind of emotional mood. If one now hears or sees or uses that name, the associated emotional mood tends to be excited. One will then tend to think that one is feeling a certain emotion towards a certain object in respect of certain of its attributes, when really one is not thinking of the object or of its attributes at all. This is what I call 'second-hand' emotion.

    Many words and symbols, particularly those associated with religion, morality, and politics, are almost devoid of cognitive meaning for most people at most times. But they have become extremely powerful stimulants of second-hand emotions. It is obvious that a great deal of the emotions which we feel are second-hand, and there is always a likelihood of emotions, which were first-hand, becoming second-hand. A typical example is the sorrow felt by a bereaved person. It begins by being first-hand, and in the course of nature it tends to fade away after a while. But often the bereaved person cannot face this fact, and so pumps up a second-hand emotion to replace the vanished first-hand one.

    It is important to remember, however, that nearly all second-hand emotion depends on the existence of a corresponding first-hand emotion in someone at some time in the past. If no one had ever believed in God with a first-hand emotion of awe, it is unlikely that anyone would now have a second-hand emotion of awe called up by the word 'God.' But the first-hand ancestor of a second-hand emotion may be a very long way back in the past history of an individual or of a race.

(5) Pure and Mixed Emotions

    I think we may fairly assume that there is a certain fairly small number of primacy species of emotional tone, just as there is a limited number of primary colors, and that the vast majority of human beings are born with dispositions corresponding to each of them. Let us call these 'primary emotional dispositions.' I should suppose that the emotional tones of fear and of anger, e.g. are certainly primary, and that the corresponding emotional dispositions are innate in the individual and common to the race. Probably some innate emotional dispositions do not come into action until certain stages of development, e.g., puberty, have been reached.

    Now these primary emotional dispositions are either very specialised or very generalised in respect of the stimuli which originally excite them. The disposition to feel fear, e.g., seems to be excited at first only by sudden loud noises and by the experience of falling. So the original stimulus is here very specialised. The disposition to feel anger, on the other hand, is aroused from the first by the thwarting of any impulse. So here the original stimulus is highly generalised. In course of experience these primary emotional dispositions become generalised or specialised, as the case may be. We acquire, e.g., the disposition to fear snakes, to fear policemen, and to fear ghosts, in addition to fearing sudden noises and falls. Conversely, one acquires the disposition to feel angry at injustice done to others, beside feeling angry at being thwarted oneself.

    I do not think that a given kind of emotional tone remains completely unaltered in quality as the objects of the emotion become extended and more subtle. No doubt there is a qualitative likeness, e.g., between fearing a sudden noise, fearing an interview with one's headmaster, and fearing God. They all resemble each other in a specific way, in which, e.g., the experiences of fearing a sudden noise and being angry at a sudden blow do not resemble each other. But there is a difference in the emotional qualities of these various experiences of fear. This might be compared to differences of shade between various instances of the same color, e.g., scarlet, rose-colored, pink, etc. I think, then, that we must say that the various primary kinds of emotional tone become differentiated in shade as the cognitions which they qualify become more complex and more abstract.

    Suppose now that one perceives or thinks of an object which has several characteristics. In respect of one of them it may excite one emotional disposition, e.g., that of fear, and in respect of another of them it may excite another emotional disposition, e.g., that of anger. One's perception or thought of the object will then be toned with an emotional tone which is a blend of the fear-quality and the anger-quality.

    I think that the best way to conceive of blended emotions is by analogy with blended colors, such as purple or orange. Any shade of purple resembles pure blue to some degree and pure red to some degree, and there is a continuous series of possible shades of purple, stretching from pure blue at one end to pure red at the other. A sensation of purple is produced when the same part of the retina is stimulated at the same time by a stimulus which would produce a sensation of pure red, if it acted alone, and a stimulus which would produce a sensation of pure blue if it acted alone. In the same way there are many different shades of blended emotional tone, stretching, e.g., from pure fear without anger to pure anger without fear. The particular shade of blended emotion which is felt on any particular occasion will presumably depend on the relative degree of excitement of the various emotional dispositions, e.g., the fear-disposition and the anger disposition.

    The following remarks are worth making about blending.

  1. It may be that certain primary emotional dispositions, e.g., those of anger and of fear, are directly linked from the first. Others become linked only indirectly in the course of experience.
  2. Probably a grown person hardly ever has an experience with a pure primary emotional tone. The notions of the pure primary emotions, like the notions of the pure primary colors, are ideal limits.
  3. Whilst some of the primary emotional qualities blend readily with each other, as do the colors red and blue or red and yellow, it may be that others will not blend. The latter would have to each other the kind of opposition which there is between complementary colors, such as red and green or blue and yellow.

    Lastly, it is worth while to notice that there are certain emotional adjectives, such as 'sad' and 'cheerful,' which apply to a total phase of experience as, a whole rather than to any part of it. We might call such qualities emotional 'pattern qualities.' They depend on the qualities and relations of the constituent parts of the whole, e.g., on the emotional tones of the various experiences included in a total phase of experience. But they are not reducible to these. Very often superficial introspection will catch the emotional pattern-quality of the phase as a whole, and will fail to reveal the emotional qualities of the constituent experiences. One may notice that one feels sad or elated without knowing why. More elaborate introspection will reveal the emotional qualities of the constituent experiences, but it may lose sight of the emotional pattern-quality of the total phase as a whole.


    Suppose that a certain object has been repeatedly perceived or thought of by a person. Suppose that it is complex in its nature and structure, and that this person has perceived it or thought of it in many different contexts on various occasions. These various cognitions of the object will have produced a highly complex trace, i.e., a very complex dispositional idea of the object. Suppose that this trace has become associated with the traces of certain names, phrases, or symbols, which have often been heard or seen or uttered in intimate connexion with perceiving or thinking of this object. Lastly, suppose that, on many occasions when this object has been perceived or thought of, strong emotions have been felt towards it by this person. When he perceived or thought of it in certain situations, or when he specially attended to certain aspects of it, his cognition of it had the emotional tone X. When he perceived or thought of it in certain other situations, or when he specially attended to certain other aspects of it, his cognitions of it had the emotional tone Y. And so on. The result is that the dispositions corresponding to the emotions X, Y. etc., will have become associated with his dispositional idea of this object. Henceforth anything that excites the dispositional idea of the object, e.g., perceiving of it, thinking of it, or perceiving or thinking of any word or phrase or symbol connected with it, will tend to excite all these emotional dispositions. We sum all this up by saying that this person has 'formed a sentiment' about this object.

    When a sentiment is aroused the emotional tone of the experience will be some shade of a blended tone. The particular shade will vary according to the past conditions under which the sentiment was formed and the present circumstances which are exciting it. It is of course possible that some of the associated emotional dispositions are such that the corresponding emotional qualities are like complementary colors and will not blend. It may be, e.g., that fear and contempt will not blend; and yet a certain object may have come to arouse both of them In that case the person who cognises that object may have the two kinds emotional tone rapidly alternating with each other. Or he may distinguish certain features in the object, and at the same time feel pure fear of it, in respect of some of them, and pure contempt for it in respect of others.

    Sometimes a sentiment gets concentrated on one particular symbol for the object instead of on the object itself. Or it may become concentrated on one particular part of the object instead of on the object as a whole. We then say that that symbol or that part has become a fetish. Fetishism is a fairly common aberration of sexual sentiment.

    Presumably there are no innate sentiments. But there are certain sentiments which practically every human being will inevitably acquire quite early in life. One is a sentiment about himself and his own powers, defects, achievements, and failures. Another is a sentiment about his parents or parent-substitutes, such as nurses, and about the members of his household in general. Another is a sentiment about the social groups, other than his household, of which he is a member. Everyone is a self; everyone had parents, and started life as a helpless infant kept alive and trained by them or by substitutes for them; and nearly everyone grows up as a member of several social groups. It is therefore inevitable that reflexive, filial, family, and other social sentiments should be formed fairly early in practically everyone.

    Certain reflexive emotions, such as remorse, self-approval, etc., are obviously very important to ethics. It is worth while to notice that we have emotions and sentiments, which are not only reflexive, but are about our own emotions and sentiments.. These may be called 'second-order reflexive emotions.' A person may, e.g., be ashamed of being afraid, or afraid of being ashamed, or afraid of being afraid, or ashamed of being ashamed. Again a person may feel angry with himself in respect of his sentiment of love for a person whom he knows to be worthless and unfaithful to him. This is just one more instance of the extreme complexity of human life and experience as compared with anything that occurs or could occur in animals. I take leave to doubt, for this reason among others, whether even an exhaustive study of the emotions of rats in mazes furnishes a very adequate or a very secure foundation for conclusions about the emotions and sentiments even of the quite ordinary human beings who pursue that study. To introspect carefully, to note sympathetically the talk and the behavior of one's fellow-men in their intercourse with each other and with oneself; to read autobiographies and the novels of great novelists; and to study and to watch performances of the plays of great playwrights; these are the only effective ways of learning about emotion and sentiment in their specifically human forms.

    In conclusion there are two points worth noting about the names which are used in ordinary life for various emotions and sentiments:

    (i) We have an enormous number of such names, e.g., 'envy, 'jealousy,' 'contempt,' 'pity,' 'awe,' etc. But we must not rashly assume that there are different kinds of emotional quality, either pure or blended, corresponding to each of these. Two emotions which have the same quality, or two sentiments which when aroused give rise to two such emotions, may have different names because they have different kinds of object. 'Envy,' e.g., is the name of a certain kind of emotion called forth by witnessing another person getting what one wants for oneself. 'Jealousy' seems to be the name of an emotion of much the same quality, where what one wants for oneself and what the other person gets is the affection of some third person. I do not say that there is no shade of difference in the emotional quality in the two cases; but the different names are certainly given in respect of the different kinds of object, and not in respect of the difference, if any, in the shade of emotional quality.

    (ii) Because a certain sentiment is distinguished from others by a certain name, e.g., 'love,' we must not rashly assume that the blended emotion connected with it contains any emotional constituent that is peculiar to it. It is certain, e.g., that the blended emotion which one feels when one is in love with a person and when that sentiment is aroused, has several factors which occur in other blended emotions connected with different sentiments. It is quite possible that there may be no single factor in this blended emotion which does not also occur in some other blended emotion. It may be that what distinguishes this emotional quality from all others is some pattem-quality due to the particular proportion in which emotional factors, each of which occurs elsewhere, are here blended.

    Even when the blended emotion characteristic of a certain sentiment does have a peculiar emotional constituent, it may be that this by itself is somewhat trivial. Suppose, e.g., that lustful emotion is a peculiar constituent of the blended emotion connected with the sentiment of sexual love for a person. Suppose that every other constituent of this blended emotion can occur as a constituent of some other blended emotion. It might still be the case that these other emotional features, though less characteristic of erotic emotion when taken severally, were collectively as essential as the constituent of lustful emotion. Mere lustful emotions, if it should occur unblended with these other constituents, might not suffice to constitute that peculiar emotion which is felt when one is in love with a person and when that sentiment is aroused.

    In such cases as these we are rather liable to give the same name

  1. to the blended emotional quality characteristic of a certain sentiment, and
  2. to any emotional quality which is held (rightly or wrongly) to be a peculiar constitutent in that blend.

The name 'erotic emotion' might be given, e.g., either

  1. to the blended emotion which is felt by a person towards another whom he is in love with, or
  2. to the purely lustful emotional quality which is, perhaps, the only constituent peculiar to that blended quality. If this happens and we fail to notice it, we are certain to be landed sooner or later in tiresome controversies which are really purely about words.