Instinct: A Study in Social Psychology
Chapter 19: The Emotions and Sentiments
Luther Lee Bernard
Emotions may be classified in a great many ways, according the viewpoint or interest of the one who makes the classifications. One of the most common classifications is based upon the feeling tone content of the emotion. From such a viewpoint emotions are either pleasant or unpleasant. This sort of a basis for the classification of the emotions is especially likely to be used when there is confusion in the mind of the classifier as to the distinction between feeling and emotion.
Another criterion of classification may be called the functional. This classification is based on the adjustment situation in which the emotion is supposed to serve. This sort of classification is illustrated by the one quoted from Warren below under the caption of "Human Emotions." A somewhat similar basis of classification separates emotions generally into the aggressive or assertive and the regressive, of which the defensive is a large subdivision, although the defensive emotions are sometimes also assertive and aggressive. Ordinarily, but not always, the aggressive emotions are pleasant and the regressive unpleasant. Thus fear is usually unpleasant, but when one is escaping from the object of which he is afraid the emotional complex may contain a large element of pleasant feeling. Also, anger may be either pleasant or unpleasant, according as one recognizes an advantage or a disadvantage with regard to one's relation to the object of anger. It is pleasant to defeat one's enemies, but it is unpleasant to be defeated by them. Another classification of the same general type may be made on the basis of emotions connected
(474) with approach toward an object and emotions connected with retirement from an object. This, however, is not always a usable distinction, because some emotions, like anger, may be associated with both types of action.
One may also speak of higher and lower emotions. Generally, the higher emotions are those in which the sensory content is organized into percepts or abstracted into concepts and the feeling content arises from the supplementation or interference of cortical or organized sensory processes. The lower emotions are constituted of lower organic and relatively unorganized sensory elements and of feeling proceeding from visceral, reproduction or other lower functional processes in correlation or interference. In some degree at least the higher emotions may be said to be more intellectualized and the lower ones to be concerned with the valuation of those adjustments connected with survival and perpetuation of the species. Such a classification does not necessarily imply a moral valuation, although in general, perhaps, we may say that the higher emotions are morally or socially more acceptable and aesthetically most approved. But doubtless there are important exceptions to this generalization. It is, of course, not impossible to classify emotions on a social or moral basis, but in doing so we pass from the neurological and physiological basis or account of emotions to a sociological consideration of them. The same objection which we earlier opposed to such a basis of classification of instincts holds also in large measure with reference to the emotions. They are not to be defined descriptively from the standpoint of their social value, but rather from their neurological and physiological correlation. We may, however, classify and evaluate them sociologically as a preliminary to selecting and utilizing them by means of our knowledge of them neurologically and organically, that is, for purposes of social control.
One of the most common classifications of emotions is with
( 475) reference to the kind of overt activity expression which accompanies them. Thus we speak of anger, fear, hate, the sex emotions, hunger, filial, family, retributive, altruistic, competitive, etc., emotions. Some of these overlap and, furthermore, from the standpoint of definition, they are open to the same criticism as the classifications on the basis of function and moral or social values. In fact there is much overlapping between this classification and that on the basis of function. Also, the emotions, classified on this basis, are not in themselves separate and distinct. They overlap and are easily transformed into one another. For example, it is not always possible to distinguish between anger and hate or even between anger and fear, or in some cases between love and anger or hate. Fear becomes anger when the subject of the emotion develops a recognition of capacity to overcome the object instead of recognizing himself as being in danger of being overcome by it. Love may be transformed into anger by the recognition that the object of the affection has betrayed the subject, and this emotion of anger may, by extending the recognition content, be deepened into hate.
Thus it may be said that the recognition content of the emotion largely, if not completely, dominates the identity of the emotion. Feeling dominates the definition of the emotion in so far as that definition is dependent upon the existence of a pleasant or unpleasant feeling tone. But the most numerous characterizing elements of emotion are those which we call sensory. This is partly because they are localized, while feeling is not localizable. This fact enables us to define the emotion in some degree according to its sensory reference, as sex emotions, love emotions, the emotion of anger and disgust, etc., according as the sensory concomitants are referable to primary and secondary physiological sex centers, the heart and breast, the flushing of the face, hot flashes through the
(476) body, tingling of the spine, twitching of the lips, etc., or to visceral disturbances which, in extreme cases, may eventuate in vomiting. In much the same way we speak of intellectual emotions when we recognize the sensory reference as being of a highly organized character which takes the form of the intensification or distortion of perceptions and concepts. This sensory content, which plays such a large part in the make-up of emotions, is the recognitive element in emotion, for not only does it enable us to localize the sensory and organic disturbance, but it also enables us to discriminate between emotions on the basis of the kinds of sensory elements involved. Thus the visceral and other more primitive organic sensations and sensory complexes, such as the kinaesthetic and muscular sensations and strains, generally characterize what we ordinarily call the lower emotions. At the other extreme we have the more definite and quantitative recognitive sensory elements, arising as visual and auditory percepts and their complexes, in the highest or most intellectualized emotions. The middle ground of peripheral sensations, especially pain, pressure and heat and cold perceptions, distribute toward either emotional extreme according to the circumstances in point.
The higher recognitive elements, such as percepts and concepts, may change the whole attitude of the organism toward the environmental pressures. Not only may the permutations of recognition constitute the difference between two closely related emotions, like anger and hate, or sex love and parental love, but they may even change the feeling tone and thus produce a major metamorphosis of emotional processes and attitudes. In the former case parental love and sexual love apparently differ only as we recognize the object of affection a5 offspring of as sexual mate. In the one case we have ä tender protective attitude with possibly certain localized sensory disturbances in the breast or other parts of the body which have frequently come in contact with the object of the
( 477) emotion. In the other case we have an attitude of tenderness and domination, possibly accompanied by certain sensory disturbances as before, but also in the sexual regions of the body, because of their organic connection with this kind of love. In both cases, especially if these types of love have been intellectualized and sublimated as is customary among the more civilized peoples, there are numerous perceptual and conceptual recognitive elements in consciousness which relate to memories of previous experiences, measurements of values in the relationship, potentialities of further developments or extensions of the relationship, and all sorts of things which one may "think" concerning child or lover. These cannot be detached from the emotional situation except by a process of abstraction. On the other hand, the recognition of opposition or unfaithfulness on the part of the object of the emotion will change the feeling content, with the result that the feeling content which is pleasant in the two cases mentioned above will become unpleasant. The other recognitive elements may remain much the same on the perceptual side, although some of the conceptual elements, such as valuation of the relationship and the measure of potentialities of development may change. When the opposition is recognized as coming from without and as being directed against both the subject and the object of the emotion, affection will usually remain and both pleasantness and unpleasantness will exist in the same emotion. Merely the character of the parental or sexual love emotion changes in such a case, for there can be as many varieties of these emotions as there are varieties of recognition of situation and relationships where they are developed. But if the opposition or unfaithfulness is recognized as coming from the object, affection may he changed wholly or in part into hate or aversion or disgust, while the major part of the higher recognitive elements remain much the same. The lower recognitive elements, the relatively unorganized sensory
(478) elements, are modified to correspond to the type of emotional expression; in fact they do much to constitute its identity, as do the higher elements.
Where an emotion of love is transformed into one of hate or disgust by a change in the higher recognitive elements, the previous emotion may be retained in part. It is not an uncommon experience to love and hate the same person and even to feel disgust with a loved person. But if the recognition elements which produce the new emotional attitude grow in volume and definiteness the transformation of the emotions becomes more and more complete. Most people have had the experience of having one emotional attitude transformed gradually into another, as the result of a progressive recognition of new elements in a situation. Sometimes the reverse process occurs. The process of exercising an emotion, especially when it is employed to excess, sets up sensory processes — usually of the lower unorganized sort, but possibly also of the higher intellectual type— which characterize and initiate or generate a new, often an antagonistic, emotion. Thus the excessive devotion of the lover or parent sometimes fosters an attitude of disrespect in the child or lover. Turning the other cheek sometimes generates excessive anger instead of gentleness and humility. Sex satiety often produces disgust, or even occasionally hate or disloyalty, in the subjects themselves. Many examples of this sort might be cited; but the prevailing relationship is of the kind previously stated, where the recognitive element initiates the emotion rather than proceeds from it.
So close is this relation between recognition content and emotion that we might even question the existence of any fixed forms of emotion whatever. The fact seems to be that emotions are constantly being transformed — into each other, primarily as their recognition content changes. This hypothesis is further substantiated by the fact that the overt or bodily expressions of emotions are not easily distinguished
( 479) on the basis of the emotions of which they are supposed to be characteristic. The expressions ordinarily attributed to one emotion merge into those attributed to another, just as the emotions themselves are transformed one into the other. The fiction of definite and distinct emotions arises from the logical device of classification or description. Seldom in life do we find definite divisions between phenomena, but we attribute clear-cut distinctions to them for the sake of logical manipulation in our thinking. The separateness of emotions is therefore a logical fiction, not a fact.
Finally, we may consider the classification into primary and secondary or derivative emotions. The writers who use this classification usually have in mind a distinction on the basis of heredity or the degree of primitiveness of the emotion. McDougall means by a primary emotion one which is the inherited affective correlate of the supposedly unchangeable central segment of the instinct mechanism. This assumption of an inherited conscious correlate we have already shown to be an illusion of interpretation. Those writers who employ the terms to cover the more primitive emotions may appear to have more of a justification for their usage. Those emotions which are characterized by strong visceral and other organic sensory disturbances and which arise in connection with reproductive, nutritive or other vegetative functioning of the organism, may properly be termed primary in the sense that they are probably the oldest emotions in the race. Also they are most likely to remain most nearly unchanged in their form of expression. That is, they are most stable in the physical or overt expression which characterizes them, in the sensory content and in the higher recognitive elements which go along with them. Even the feeling tone which accompanies them is likely to be less variable than in the case of the higher and more intellectualized emotions.
These more primitive emotions may be modified through the process of sublimation into derivative emotions. Thus, some writers have supposed that filial and parental love were sublimations of sex love. This, however, is not necessarily correct. Sublimation into derivative emotions does not ordinarily take place across recognition content, but through an extension of recognition content in the direction of intellectualization and definition. Thus the sublimation of the sex emotions would be into the derivative forms which we find in literature and art and which deal with sex in its intellectualized and technic expression. This process of sublimation of a crude primitive emotion is a process, not only of intellectualization, but also of Tsthetic refinement and development. The same tendencies toward intellectualization and refinement may be illustrated from other emotions also. Crude patriotism grows into a love for and loyalty to the higher aims of all humanity. Primitive devotion to offspring may be intellectualized and spiritualized into interest in and devotion to the welfare of all children, of all classes and races, everywhere. Some social thinkers fear the ultimate result of a progressive sublimation of emotions which turns the primitive or primary types into derivative forms and content. They anticipate that such a process of sublimation may eventuate in a deterioration of the native vigor and moral stamina of the race. But there is little justification for this fear, if the process of intellectualization and sublimation is guided by well-established values, that is, if it is carried out under the dominance of social science instead of at the behest of a hedonistic aesthetics or a subjectivistic ethics.
The more sublimated and intellectualized an emotion becomes, the inure secondary its character, the less easily it can be defined in terms of its lower organic and poorly organized sensory content; also the more variable it becomes. Thus the secondary emotions are not so readily and obviously distin-
(481) -guished from each other by their visible signs as are the primary or more primitive emotions. Their definition must be made in terms of their higher recognition content, and this is comparatively difficult for those not skilled in the process. Also, those who live nearer what we sometimes call an animal plane of existence do not experience the higher or more sublimated and intellectualized emotions sufficiently frequently to enable them to recognize or classify them within themselves with any great facility. Thus the higher emotions of social loyalty, devotion to social welfare and progress, or the less social emotions of aesthetic appreciation of classical music, "good" literature and art, are familiar affective content only to the "élite." The person of poor breeding does not recognize his poor taste. The so-called moral imbecile is not aware of the world of social obligation which lies beyond his horizon of thought and feeling. The typical ward politician has not dreamed of the wider obligations of man to man. The typical subjective religionist, who bends all his energies to saving his own or another's "soul," so frequently neglects alike the lesson of the good samaritan and his obligation to good citizenship of a higher order, and he is unable to understand why his brother with a quickened social conscience regards him with pity or contempt.
In general, we may say that the emotion has as much inner unity and constancy of form— is as definable as the activity or activities to which it corresponds or in connection with which it arises and from which it secures its organization, have unity and constancy of form. Thus the primitive or primary emotions are attached to fairly definite activity processes and are usually fairly easily recognized by their visible as well as subjective signs. But the more sublimated and derivative emotions have a vast number of forms which are constantly changing like the higher recognition and sensory and feeling contents which constitute them. McDougall assigns primary
( 482) emotions to his seven primary "instincts," but he is doubtful about the emotional correlates of his secondary "instincts," and he finds no definite emotional correlates whatever for the so-called general innate or instinctive tendencies, such as play, imitation, suggestion and sympathy. This is what we should expect in the light of our analysis. While his primary "instincts" are only habit complexes, they possess sufficient unity and constancy of function to embody a fair degree of unity and constancy of recognition content, especially of the higher orders. Hence we may recognize his corresponding emotions as possessing a certain unity of consciousness and of definition. But this unity is general. Within these emotional complexes, which we recognize as possessing a certain degree of inner similarity and unity, there is a wide range of variation. For example, anger varies widely according to its recognition content; and the same may be said of the other primary emotions of McDougall. They are not, as his discussion might lead one to suppose, identical under all conditions of manifestation. Nor are all of them truly primary or primitive emotions in the sense in which we have defined that term here. Some of them are sublimated and intellectualized emotions. Nevertheless, they are highly standardized, possessing recognitive unity and therefore capable of definite classificatory categories, because they correspond to well-recognized habit or value complexes imbedded in our social traditions and valuations, and they themselves constitute well-recognized value complexes.
McDougall was unable to find definable emotional complexes to correspond to his secondary "instincts" and general innate tendencies, because they lacked unity and constancy of content. They are not even habit complexes. In so far as they possess social-value unity and are recognized as having such, they have a diffused emotional value content to correspond, but this varies and transforms itself so rapidly that it
( 483) is difficult to place the analytical finger upon it and name it. The secondary "instincts" have such a decidedly variable value and activity content, but no one would deny that the quality of gregariousness, for example, has a diffused and variable emotional correlate, but its abstractness and vagueness make it difficult as yet to localize and name it. The almost total lack of unity, from an activity standpoint, of the so-called general innate tendencies precludes any unity of emotional content which would render it possible to assign names to their general emotional expression. However, the concrete activity functions which correspond to the various forms of activity content in imitation, suggestion, etc., have emotional correlates, often of a primary quality; for such activity contents are as frequently definite and primary in character.
The conception of emotion here set forth does not provide for its definite separation as a concept from sentiment, as that term is used by McDougall. In fact there is no distinction, except possibly one of degree of organization and of recognition of relationship between object and subject. A sentiment is an emotion in which definite feeling tones are organized about a more or less definite recognitive element or process with a value reference to the subject. In the sentiment the object is analyzed and more or less clearly defined. But the attitude toward the object of attention is not merely one of analysis and definition, such as might exist in a dominantly or purely intellectual attitude. The object is also definitely a source of pleasantness and unpleasantness— consequently the attitude is an affective one. In the lower forms of sentiment the object need not be recognized as a source of pleasantness or of unpleasantness, although it must be such a source. The affective or sentimental attitude can exist toward the object without an attempt to measure its hedonic effect upon the subject.
(484) But the subject must recognize his own attitude toward the object, at least more or less clearly. Even the very young child recognizes its sentiment or love for its parents; and it is possible that animals which can develop affective attitudes recognize some degree of affection for the objects of these attitudes, although doubtless not in reflective or analytical terms. However, the higher sentiments probably always carry with them a recognition, not only of affection for the object, but also an intellectual appreciation of the effect of the object upon the subject. The effect need not necessarily be stated in hedonic terms. The object may simply be called good or beautiful or fine or noble, or by some other term of appreciation or value, carrying more or less of the implication of the satisfaction to be expected from the object by the subject.
The lower emotions, the more primitive ones, may involve little or no sentiment, because they do not recognize or define the objective cause and therefore the subject does not develop a definite affective attitude towards it, much less an appreciation of its hedonic consequences for the subject. But sentimental attachments to objects which produce pleasure may arise even in connection with the lower emotions, in addition to the unorganized non-recognitive sensory content, when the evaluative element is clear and well defined. Thus even people very low in the cultural scale may have strong sentimental attachments for objects of sex love or for sources of food and drink. Sometimes these sentiments, largely because of the lack of more sublimated competing sentiments or emotions, are stronger in the ignorant and uncultured than they are in the highly civilized and cultivated. A sentiment is an emotion with a definite recognitive element acting together with a strong feeling tone as value consciousness, and the more highly developed the recognitive or intellectual element the more sublimated is the emotion, the finer the sentiment. The higher and finer sentiments are those which are most social-
( 485) -ized or defined on the attitudinal side. In them the attention is centered, not so much upon the self as a recipient of crude feeling satisfactions or distress as upon the wider social significance and value of the object, or at least upon its more refined and sublimated intellectual and aesthetic significance and value. Thus we generally regard the lower sex sentiment, lust, and the lower social sentiment, hate, with disapproval. With these we contrast favorably the sublimated sex devotion of a Dante for Beatrice or of a Petrarch for Laura. Similarly we contrast admiration with hero worship or an ethical patriotism, which is not afraid to acknowledge national wrong for the sake of righting it, with a jingoistic patriotism and narrow nationalism, to the disadvantage of the latter in each case. The dogmatist's devotion to his religion which causes him to burn or place upon the rack those who disagree is a low and unmoral sentiment as compared with the Christ-like or Buddha-like devotion which forgives and serves. Knowledge or science, that is, a broad and penetrating analysis of the object and its functional relationships within its environment, makes for a refinement of sentiment. The truly cultivated person is catholic and tolerant and his sentiments are refined, even if not always so vigorous.
McDougall properly speaks of the rôle of the self-regarding sentiment in the development of character. The self-regarding sentiment is indispensable to character formation. Character growth is a process of integrating the self-consciousness as a means to building up an internal control over conduct through the personality. The personality which possesses a strong self-consciousness, with an affective attitude toward the self— the essence of the self-regarding sentiment— is a strong character. Whether it is also a good character depends on whether the self-regarding sentiment is merely hedonistic
( 486) and self-seeking or is socialized. A good character is one which not only has a strongly organized self-consciousness but also has a strongly developed social consciousness, with a well balanced and coördinated affective attitude toward both the self and society. In other words, a social-regarding sentiment is quite as important in the development of a good and strong character as is the self-regarding sentiment. As Professor Cooley has pointed out, self and social consciousness properly grow up together and neither can be complete without the other. In the same way neither the self-regarding nor the social-regarding sentiment can reach the highest and finest fruition without the other. However, it is possible to develop either of these sentiments which are so important for character development in a lopsided and incomplete manner, just as we may develop either self or social consciousness out of proportion as regards the other. The old individualistic ethics was largely metaphysical and barren because it failed to see the importance of either the social consciousness or the social-regarding sentiment for a functional character development. And one may suspect McDougall of retaining in some degree this bias in his discussion of the growth of the self-consciousness and the advance to the higher plane of social conduct. He certainly lacks the appreciation of social values and pressures in the character-building processes which is characteristic of Cooley's writings. The more intellectualized and analytical these two character-forming sentiments become, provided they also retain their affective element, and value reference, the more serviceable they may be as control processes.
All of McDougall's primary emotions are sentiments, except possibly the crudest and most primitive forms of fear,
( 487) anger and wonder. In the lower forms of these emotions there may be absent all definite recognition and value content. One may be simply afraid without clearly recognizing any object of fear, or angry without being definitely conscious of being angry at anything or any person in particular. Possibly one may have a vague attitude of wonderment without being able to assemble one's consciousness about what causes the wonder; there may be simply a feeling or sense of being oppressed with more than one can understand, or the attitude may be one of seeking to understand with a high degree of nervousness at the difficulty of defining the stimuli. But ordinarily the subject recognizes the object of fear or anger or surprise and develops an affective attitude toward it. In the higher forms of emotion involved in righteous anger, or in safeguarding the public welfare and in the pleasant pursuit of scientific knowledge of puzzling problems, we have highly sublimated and intellectualized or socialized sentiments.
The other primary emotions of McDougall are perhaps always sentiments or types of sentiments. Thus negative and positive self-feeling obviously are affective interpretations of the self, with more or less reference to environing conditions. The emotions of disgust and tenderness represent general types of sentiments corresponding respectively to the unpleasantness and pleasantness affective attitudes which may be attached to any analyzed and defined object which the environment brings into consciousness. Consequently, disgust and tenderness are not so much emotions or sentiments as they are general emotional or sentimental attitudes towards objects. We are disgusted with objects which are unpleasant to us, and in extreme cases they produce the lower and unsublimated emotional expression of nausea and vomiting. In the cases where our attitudes are more intellectualized and sublimated we merely disapprove. Similarly we may have a tender attitude toward any object which affords pleasure to
( 488) us, and this attitude may vary from blind adoration, such as is illustrated in Shakespeare's story of Romeo and Juliet, to one of approval on intellectual or broad objectively stated aesthetic principles. Disgust and tenderness are not emotions and sentiments, but attitudes basic to as many specific emotions and sentiments as the permutations of the recognitive and feeling elements in consciousness can organize into distinct categories subject to our disapproval or approval. They are types of emotions and sentiments and within these two types we can range almost all specific emotional attitudes. Thus under disgust we may place the vast range of fear sentiments, each defined in terms of the object of the fear. Here also belong the hates and aversions, but only in part the sorrows, and disappointments, for there is an element of tenderness in these also. Under the tenderness type are to be found the loves, defined and distinguished according to the objects recognized, sympathies, joys, good wills, and the like. The objects toward which one may be tender are infinite in number, ranging from human beings, through material objects, to principles. The only constant or stable element in tenderness is the feeling tone, and even it may vary in intensity, if not in quality or kind. And the same is true of disgust. The differentiation of these attitudes into concrete or specific emotions and sentiments is accomplished through the differentiation of the recognitive element which functions as a complement of environmental pressures rather than of inheritance or instinct. The basic feeling tones remain constant in each type of cases; but the degree or intensity of the feeling tone varies according to the organization of the specific sentiment involved.
The secondary emotions which are listed by McDougall, and which he does not regard as sentiments, would, according
( 489) to the analysis here presented, be classed as sentiments as well as emotions. Of admiration he says, "This is certainly a true emotion, and it is as certainly not primary. It is distinctly a complex affective state and implies a considerable degree of mental development." These are just the conditions— an affective attitude towards a defined object with a recognized value significance for the subject— which render the emotion a sentiment. According to McDougall, admiration is a complex or derivative emotion made from the fusion of the two primary emotions, "wonder and negative self-feeling or the emotion of submission." It is possible to regard some emotions and sentiments as being compounded out of other emotions and sentiments, and we might therefore describe admiration as a sentiment composed of the sentiments of wonder and negative self-feeling, or of the emotion of wonder (when of too low an order to constitute a true sentiment) and the attitude of submission and of those other recognitive elements which definitely describe or locate the particular kind or form of admiration which is under consideration. There are, of course, as many varieties of admiration as there are of recognition and wonder and submission (if we accept as truly descriptive the composite or atomic theory of the sentiments and emotions) connected with it.
But one should be very careful about accepting the atomic theory of the sentiments and the emotions. In his discussion McDougall uses this basis of description altogether, treating each complex emotion as a compound of unitary and indivisible primary emotions and each sentiment as a complex of one or more emotions or sentiments organized around the idea of some object. This is too mechanical and artificial. Each emotion and sentiment is an organic whole; possessing its own individuality or complex of traits, according as it varies in character. That is, the composite units in an emotion or
( 490) sentiment are not emotions and sentiments which may be added or subtracted mathematically, like so many bricks in a wall, but they are feeling tones and sensory (including recognitive) elements, and, because the feeling elements vary in quality and intensity and the sensory content varies infinitely in variety of original units  and of their recognitive permutations, there can be as many forms and shades of emotion and sentiment as there can be variations and combinations of these basic elements— and this is infinite. Thus, each emotion and sentiment, and each variety of these, is built up from the basis of feeling and sensory (including recognitive) content into an individual organization, not from other supposedly unitary primary emotional units or bases. That is to say, the dynamic elements in emotions and sentiments are feelings and recognitive elements. Because of this fact we have a much greater variety of emotions and sentiments than would be possible if so-called primary emotions and sentiments were basic building material. The variety and shading is infinite.
Yet, in spite of this great variety and fluidity of emotions and sentiments, we have certain more or less standard or recognized forms of emotions and sentiments. Thus, besides the so-called primary emotions, which we have already analyzed, McDougall lists the following complex emotions.: 
(491) The first nine of these emotions, he says, do not necessarily involve sentiments. The remaining twelve do imply the existence of sentiments. According to our analysis, we would classify all of them as sentiments as well as emotions. McDougall regards all of these emotions as instinctive or as compounds of instinctive processes. The present analysis regards all of them as affective attitudinal complexes built up primarily under the domination of the environment, which works through the recognitive elements, as explained in earlier chapters. In this respect they are very similar to his supposed instincts.
Warren has a much lengthier list of emotions, which he breaks up into several subdivisions, as follows: [11
|1. Expressive (Nutritive)||2. Reproductive|
|+ Joy (Enthusiasm)
- Grief (Despair)
|3. Defensive||4. Aggressive|
- Timidity (Embarrassment)
|Flight and Hiding
|- Anger (Passion)
|5. Social||6. With Temporal Projection Retrospective Reference|
|Emotion||Instinct||- Regret (Remorse)|
+ Satisfaction (Elation)
|Social Emotion||Instinct||Prospective reference|
In this list he has indicated the feeling quality attached to each emotion, by means of plus and minus signs placed before the emotions, and also the psychic or activity organization to which it corresponds. Even this list by Warren does not exhaust the category of emotion, but it approaches more nearly toward a complete enumeration of the well-recognized or standardized organizations of emotion and sentiment. It also gives us some suggestion as to the proper basis for the recognition of separate emotions or emotion complexes.
These so-called emotions are really emotion complexes which are distinguishable from each other on the basis of the types of activity, attitudes and consciousness (recognitive-sensory and feeling processes) with which we find them associated. It is the activity, attitudinal and consciousness complex or organization which determines the perception of the organic character of the emotion complex and also our naming of it. And the organic character of the activity, attitudinal and consciousness complexes, that is, their functional and putative unity of organization, depends in part upon the unity of the physical organism which exercises the complex more or less as a unit. Thus, for example, the definition or characterization of lust, flight, shame or anger is dependent in part upon the functioning of certain parts of the body or of the whole body in certain ways. This is likely to be the case where primitive emotions
(493) are concerned, and not so likely where derivative and sublimated emotions are being considered. In such cases, and perhaps to some degree in all classes of emotions, the definition or identity of the emotion complex or aggregate is determined in terms of or with reference to the value of the accompanying activity, attitudinal and consciousness correlates for the individuals and society or group concerned. Here, as in the case of the complexes miscalled instincts, the recognition of value plays a large, often a dominant, rôle in characterization or definition. Because of such definition and organization into complexes on the basis of structural or activity and valuational correlations we are able to group emotions under such meaning categories as enables us to handle them as adjustment processes instead of allowing them to appear to us as anarchic manifestations. We have even gone further in the process of synthesis than the facts warrant, and often we do not recognize that a single term does not represent an emotional unit, but a vast complex of fluid emotional forms having a certain unity or similarity of manifestation or of value. Such an illusion of unity in emotion undoubtedly has its functional value, but it may also have its disadvantages for purposes of psychological analysis.
McDougall classifies the sentiments according to two general categories— according to their emotional constitution and according to the nature of their objects. Of the former method he says, "In dealing with the emotions, we named and classed them according to their nature as states of affective consciousness and as tendencies to action; and we may attempt to name and classify the sentiments also according to the nature of the emotional dispositions that enter into the composition of each one. But since, as we have seen, the same emotional dispositions may enter into very different sentiments, we can carry the naming and classification of them but
( 494) a little way on this principle, and we have accordingly buy very general names for the sentiments. We have the name; love, liking, affection, attachment, denoting those sentiment that draw one towards their objects, generally in virtue of the tender emotion with its protective impulse which is their principal constituent; and we have the names hate, dislike, and aversion, for those that lead us to shrink from their objects, those whose attitude or tendency is one of aversion, owing to the fear or disgust that is the dominant element in their composition. The two names love and hate, and the weaker but otherwise synonymous terms liking and dislike, affection and aversion, are very general; each stands for a large class of sentiments of varied, though similar, composition; the character common to the one class being the fundamental tendency to seek the object and to find pleasure in its presence, while that of the other class is the tendency to avoid the object and to be pained by its presence . . . . A third principal variety of sentiment . . . is primarily the self-regarding sentiment, and is, perhaps, best called respect. Respect differs from love in that, while tender emotion occupies the principal place in love, it is lacking, or occupies an altogether subordinate position, in the sentiment of respect. The principal constituents of respect are the dispositions of positive and negative self-feeling; and respect is clearly marked off from love by the fact that shame is one of its strongest emotions." 
In classifying the sentiments according to the nature of their objects, McDougall says, "They then fall into three main classes, the concrete particular, the concrete general, and the abstract sentiments— e. g., the sentiment of love for a child, of love for children in general, of love for justice or virtue.
( 495) Their development in the individual follows this order, the concrete particular sentiments being, of course, the earliest and most easily acquired. The number of sentiments a man may acquire, reckoned according to the number of objects in which they are centered, may, of course, be very large; but almost every man has a small number of sentiments— perhaps one only— that greatly surpass all the rest in strength and as regards the proportion of his conduct that springs from them . . . . When any one of the emotions is strongly or repeatedly excited by a particular object, there is formed the rudiment of a sentiment." One of the strongest and most complex sentiments in most people, McDougall says, is parental love. Affection for an equal and active sympathy are also very common. But the self-regarding sentiment is the most important of all in building up character: "There is only one sentiment which by becoming the master-sentiment can generate strong character in the fullest sense, and that is the self-regarding sentiment." McDougall devotes three of his most important chapters, those concerned with the constructive social direction of the individual, to the discussion of the growth of this sentiment. A criticism and evaluation of his treatment in this connection has already been offered in this chapter.
Warren's definition of sentiment corresponds pretty closely to our own: "A sentiment is a mental state whose leading components are feelings and imagery. It is due to the combination of systemic with ideational impulses . . . . Sentiments may be aroused by any sensory or ideational impulse." He classifies sentiments "according to the type of primary expe-
(496) -rience which arouses them." This method of classification is more closely related to
McDougall's second method than to his first, but it corresponds definitely to neither. Apparently
he does not accept the atomic view of McDougall which makes the sentiments a mere
mechanical combination of the emotions. He is thoroughly functional and builds up each
sentiment from the basic elements of experience in much the same way as has been insisted upon
in this chapter. He does not attempt a detailed list of the sentiments but classifies them generally
according to their supposed origins in experience as follows: 
While I regard Warren's definition of sentiment as on the whole a good one, an examination of his discussion of his classification headings or types  convinces me that he has in mind perception and general attitudinal content rather than true sentiments. He mentions only such concrete examples as certainty, doubt, a sense or sentiment of opposition, of being thwarted, of force or power in the environment, of the inevitable, of the sublime, or of right, of wrong and of the good or bad. These are recognitive terms primarily, although they may have affective aspects or attitudes attached to them. But before they can be true sentiments they must carry some recognition of the value of their objects for the subject.
An attitude is a more general term than the others which we have used. It is based on both acquired and instinctive proc-
( 497) -esses. It is sometimes affective in character; sometimes predominantly intellectual. At other times it is merely physical or organic, with little or no mental content. "An attitude is due to a repetition of neural processes of one and the same type . . . . The relation between the stimulus and attitude factors is illustrated by comparing the experiences of watching a bonfire and a destructive conflagration. The sensory stimuli may be substantially alike in the two cases, but our response is quite different. We assume entirely different emotional and motor attitudes."  Warren is here speaking of conscious attitudes rather than of merely physical or unconscious attitudes, and of course the conscious ones are the most significant from the standpoint of emotion and sentiment. Clearly it is the recognitive element, the recognition of significance, which here primarily distinguishes one attitude from another when the stimuli are alike. The recognition of significance is dependent upon past experience or memory.
An attitude is a condition of preparedness of the organism which in the last analysis reduces to preparedness for action. Action, however, is suspended, otherwise the process would be action instead of attitude. Attitudes cease in so far as they go into action. Since attitudes are fundamentally suspended action, with a recognitive element in the higher forms, which evaluates the relationship between the subject and object in either affective or intellectual terms, we can, in these higher forms, describe or classify attitudes "according to the type of mental state which characterizes them. There are attitudes corresponding to each type of experience." 
Warren outlines the general classes of attitudes, with their corresponding and characterizing bases, as follows: 
Social and Moral (Conscience)
" Proclivities "
Of Rational Action
In addition to this method of classification, others are possible also. For example, attitudes might be classified as intellectual and emotional or affective, as antagonistic and coöperative or sympathetic, or as good and bad. In this last case, however, the classification would be made, not in terms of the characteristics or structural and mental preparedness of the individual possessing the attitude, but in terms of the value of the attitude for society or any individual concerned. In such a case many attitudes possessing very different structural and mental components might be either good or bad, and vice versa. Such a classification is sociological rather than psychological. Other classifications, in terms of the posture of the individual concerned and in terms of the organs or technique employed, might be made.
In this discussion we are particularly interested in those conscious attitudes which have emotional correlates. One of the completest attempts at such a classification is presented by Warren under the general title of Human Dispositions, as follows: 
|1. Expressive||2. Reproductive|
|3. Defensive||4. Aggressive|
|5. Social||6. Instinctive and Sentimental|
Belief and disbelief
Practically all conscious attitudes will possess, not only recognitive elements, but also affective correlates. Only the abstract scientist or philosopher can detach himself from a feeling of consequences or affective values which relate either directly or indirectly to himself or to those dear to him, and not even he can achieve such detachment all the time. For the great mass of mankind self-consciousness dominates social consciousness in a recognitive situation and emotion arises in connection with the perception of self in the situation. Even when these perceptions are sublimated into social consciousness, or when we are observing external relationships of cause
(500) and effect objectively and in a detached manner, feeling may enter in according to the laws of feeling correlation, and sublimated emotions appear in consciousness. But where attitudes are concerned the self or self-consciousness is necessarily prominent and emotions attach strongly to this recognition of the self in relation to other objects. Also, since an attitude is fundamentally a suspended activity process, or a condition of preparedness for action, the absence of emotion of some sort is practically inconceivable in connection with a conscious attitude.
So many classifications of emotions and sentiments have been presented in this chapter primarily for the purpose of illustrating the lack of uniformity of treatment, on the one hand, and the lack of agreement as to the classification to which different processes belong, on the other hand. Thus, for example, McDougall lists affection as a sentiment, while Warren classifies it as an emotion. Warren means by sentiment something quite different from what McDougall understands by the term and other writers similarly. The cause of this confusion and lack of agreement is, apparently, the fact that psychology was developed as a form of logic before it was made the science of behavior. As a consequence, in its anxiety to understand human conduct and human attitudes, it refined logical distinctions on an a priori basis to a very great degree, without having the technique for testing out these logical distinctions and a priori classifications by reference to the actual physiological and neurological facts. Now that the science of human behavior has arrived, with its examination of neurological and physiological bases of action and mental behavior, the logical dress which it has inherited from the old metaphysical psychology does riot fit very well. Not only is it ill-fitting, but it is also excessive in amount.
In the logical and metaphysical stage of the development of any discipline, lack of accuracy and functional adaptation
(501) is in a manner compensated for by excessive refinement of descriptive classification. The scholastic attitude develops within it and the lack of functional understanding is to a degree hidden by the wordiness and refinement of logic. One of the first results of the invasion of a discipline or control situation by science, with the consequent displacement of metaphysics, is simplification. As a consequence understanding improves. This is what is happening in psychology, and soon a similar movement must take place in sociology. Much of the overlapping and confusion in terminology which we have found and exhibited in this chapter on emotion will disappear when we accept a neurological basis of discussion and reclassify our phenomena accordingly. Much can now be done to bring understanding out of confusion if we keep constantly and clearly in mind the distinction between instinct and habit and thus eliminate the many superstitions about instinct which we inherited from metaphysics. Also, if we remember that attitude is a general term for suspended activity, and that in its conscious aspects it must be either intellectual or emotional, another chief source of confusion will be avoided, especially the one which regards attitude as a subdivision of emotion or rational thought. Also we must keep in mind the fact that emotions are composites or complexes of feeling and sensory or cognition elements, and not independent or inherited mental elements. Also, that feelings and sensory processes are not logical concepts but neural correlates, and that out of the latter develop the vast complexes of intellectual content. And, finally, it is necessary to recognize that sentiments are but more highly developed and organized emotions, in which the object of the emotion is recognized as having a negative or positive feeling or welfare value for the subject. On the basis of these neuro-psychic facts we may rebuild our classifications without overlapping and remove the confusion which we have inherited from the metaphysics of
(502) logic which has so long served in the place of a behavioristic psychology.
Desire is a term which occurs so frequently in the literature— especially in the literature of sociology— and the mental attitude of desire is supposed to play so large a rôle in the mediation of adjustments, that it requires some definition in behavioristic and neurological terms. Desire is an attitude with a large emotional content or correlate. It is the recognition of an object, which is not possessed, as desirable, that is, as an object whose possession or imminence would give pleasure or satisfaction. In the same way aversion is the recognition that the possession of or close functional contact with an object would cause unpleasantness to the subject. Desire and aversion are conscious attitudes. The object desired may be a commodity or an activity experience, knowledge, or anything which can be identified with the subject.
The consciousness of desire, on its recognitive side, arises from the readiness of the organism for the adjustment which has not been made. This unsatisfied readiness of the organism produces only a poorly defined emotional disturbance, such as restlessness, nervousness, irritability, vague unhappiness, etc., if the organism has not before had the experience of satisfaction. Such often is the case where sex tendencies are repressed at puberty without giving them some direct or indirect and sublimated avenue of expression. But where there has been previous experience of satisfaction, the readiness of the organism for the satisfaction is recognized in terms of a consciousness of the process of satisfaction, which may be pictured in elaborate detail, if the inhihition of satisfaction continues sufficiently long. This is the recognition content of desire. Neurologically the picturing of the process of satisfaction is practically identical with the inner or neural aspect of the completed act, if the overt part of the action could also
( 503) be established. That is, the organism, in attempting satisfaction, initiates the first part of the act neurally, or in consciousness (it is all the more conscious because the neural process is not completed by overt response), and this consciousness without the completion, which is inhibited, is called desire. Not infrequently, where a successful completion of the overt phase of the act cannot be secured, the inner neural or conscious part is repeated over and over again and it becomes an end in itself. This is a pathological condition which sometimes reaches the magnitude of a major psychosis or insanity, as in dementia praecox, just as an abnormal substitution of anticipation consciousness in aversion may also result in such insanities as the phobias or possibly even paranoia.
Neurologically, under normal conditions of expression, desire is not so much the cause of an adjustment recognized as "desirable" as it is the conscious correlate of the attempt of the organism to achieve the "desired" adaptation by initiating the inner or neural stages of the adjustment act or series of acts. Desire has both its instinctive and acquired neural correlates. To speak of instinctive desire is misleading, for desire never appears, as will readily be seen from the previous discussion, if the instinctive mechanism is really successful in making an adjustment between subject and object. It is only when the instinctive readiness of the organism, or the inherited action pattern, fails to function successfully that desire arises. Desire indicates interruption, not fulfillment. Hence, desire correlates more particularly with acquired activity; but it may have an instinctive basis. That is, the desire may be in part the conscious correlate of the instinctive neural mechanism attempting to go into overt action. But there is always more than just this neural basis of instinctive origin There is also the consciousness of effort, of things to be overcome, of a technique of overcoming, which has as its neural correlate the modifications of the instinctive process by outside interference.
( 504) That is to say, most of the conscious content of the desire recognition has a modified or acquired neural basis. And this is also the explanation of the fact that in desire there is an undercurrent of unpleasantness along with the dominant feeling tone of pleasantness. There is interruption by modification, a secondary interruption, but the modification of the neural processes is a successful one— at least in the imagination.
Sentiment is somewhat under a shadow, according to the dominant categories of good taste, at least that exaggeration of sentiment which we call sentimentality. And yet sentiment of a sincere and properly chastened sort— sublimated and intellectualized— is one of the most valuable indicators of normality of adjustment or of attitude. It is the highest possible form of emotion, unless we account as emotion the almost intellectual appraisal of the beautiful and good or of the harmful and ugly, in which the subject practically detaches himself affectively from the object.
Too frequently we speak of the higher or more sublimated emotions and sentiments as if they always had a pleasant feeling tone. This is a mistake. Some of the strongest sentiments and emotions are unpleasant, even when of a high order. Such, for example, were the emotions of the ancient Hebrew prophets who agonized over the superficiality and perverted ideals of their people, or of Jesus on the Mount of Olives, and such might be the attitude of a modern moral leader who had some interest greater than filling successfully a well-paid position. The true patriot, in contrast with the professional optimist and the politician, will find much to suffer as well as to enjoy from his devotion to the highest ideals for his country and his civilization. The cynic is often not a bad man, but a man too good for his age, at least for the masses of mankind to understand and to profit by.
The conclusions reached in this and the preceding chapter concerning the nature and function of emotion may be stated briefly as follows: (r) Emotion is a composite concept and not an original and independent mental element. It is a varying synthesis of feeling tones and sensory elements of all grades of development from the lower visceral and other organic and peripheral pre-recognitive sensory elements and complexes to the cortically assembled recognitive processes built out of sensory elements as standardized percepts and concepts, under the general direction of environmental pressures.(a) At one extreme emotion is resolved into feeling and simple sensory elements and complexes. At this extreme we have the primitive or so-called lower animal emotions. At the other extreme the higher recognitive elements and the feelings arising from their operation may dominate, especially in the so-called higher or human emotions. With the superior development of the recognitive elements the emotional evaluation and adjustment merges into rational adjustment.
(3) Emotions are not the fixed or inherited conscious correlates of the instincts, as McDougall has supposed. They are, in part, conscious correlates of the modified processes which arise from interruption of previous neural sets and complexes. They also arise from the random sensory disturbances caused by overflow of energy and impulse resulting from interruption or inhibition of dominant neural and activity processes, more especially when this interruption arises from the outside. Thus emotion is seen to belong rather to the field of acquired than of inherited neuro-psychic organization and technique.
(4) The emotions are as variable and possess as much acquired content as the habit and value complexes which have so often been miscalled instincts. However, they are not identical with these acquired complexes, nor are they strictly their conscious affective correlates. The correlation of the emotions extends beyond the dominant activity or thought adjustment
( 506) processes into the field of random activity or sensorimotor discharge. Emotions, like all other forms of consciousness, indicate inhibition or delay or modification of action instead of being the correlates of uninterrupted activity, which has a minimum of conscious correlation.
(5) The classification of the emotions is relative to the type of content, and there may be an almost infinite variety of terms, especially in the higher or human sublimated types of emotions. We cannot classify emotions effectively on the basis of inherited or acquired elements, but best on the basis of the types of adjustment being mediated as indicated by the content. Thus the emotion varies primarily according to the changes in the recognitive element, especially of the higher or perceptual and conceptual kinds. The relatively fixed elements in emotion are the lower and relatively undifferentiated and non-recognitive elements and the feeling elements correlated with the exercise of these.
(6) If we are to speak of primary and secondary or derivative emotions, the distinction should be made on the basis of the degree to which the relatively primitive and fixed elements become sublimated and intellectualized through the development and differentiation of the recognitive elements. While the elements which enter into the more primitive and primary emotions have more of an instinctive basis than have those of the derivative emotions, there is no emotion, however primitive or primary, which in its organized expression is purely instinctive. The so-called primary emotions of the classifications are, like the corresponding so-called instincts, complexes, and are consequently derivative in character, that is, they are organized under the control of environmental pressures.
(7) Sentiments are merely emotions in which the recognitive element is more highly developed, especially with reference to the value of the object for the subject. Sentiments therefore generally fall within the range of the higher or more sublimated
( 507) and intellectualized emotions, and the quality of the sentiment is a fair indication of the quality of the adjustment being made or of the attitudes toward objects, or situations.
(8) Attitude is the most general term indicating suspended action or the preparedness of the organism for action; it is not a sub-category of emotion or of rational adjustment, as is sometimes claimed. It includes such terms as instinct, habit, emotion, sentiment and desire.
(9) Desire is the conscious correlate of the neural activation which occurs when an organism is ready for action but is prevented from achieving the "desired" adjustment by some factor in the situation. The correlation is not alone with the previously prepared neural activity set, but also with the neural extensions innervated by the interruption or inhibition. Hence desire includes the recognition of both the adjustment desired and of the obstacles to be overcome, as well as the feeling correlations arising from anticipated success or failure. Aversion is a corresponding conscious correlate of an activity recognized as undesirable. Desire is not so much a cause as an indication of the value of a suspended response.
(10) The discussion of emotion in its various aspects, as well as of other attitudes, has been confused by the abundance of terminology which does not fit the facts. This confusion of terminology has arisen from the attempt to analyze the conduct of the "mind" on a logical basis before behavior had been studied from a neurological and physiological standpoint. A proper study of behavior and classification of terms descriptive of that behavior must start from a neuro-psychic basis and disregard the old logical metaphysical classifications where they are in conflict. The result of a, behavioristic study, as is always the case where science is substituted for metaphysics, will be a great simplification of terminology and facilitation of effective thinking.
(11) Emotion is not the cause of activity, as is so frequently
(508) assumed, but one method of evaluating the suspended or inhibited activity for the organism or society. The most primitive form of evaluation is feeling. Feeling proving inadequate as a method of mediating the more complex adjustments of a highly differentiated organism to a rapidly changing environment, emotion appears as a union of feeling and sensory elements. The sensory aspect of emotion constantly develops in complexity and effectiveness, because of the multitude of combinations of technic processes of measurement with objective reference which it can develop in the form of perceptions. In its conceptual form the recognitive element becomes dominant in valuing and mediating adjustments and these are spoken of as rational. The evolution of evaluative and adjustment technique appears to be away from instinct and feeling towards rational or conceptual methods.