On Motivation *)

Human motivation is, of course, immensely complex, and it is impossible to do justice to all its facets and layers. Simplified assumptions are necessary, particularly in macroscopic analyses of human behavior. In this vein, economics has been successful by assuming only one motive: greed. The postulation of the profit motive by early economic theorists was consistent with the hedonistic philosophy that represented the most reasoned standpoint of the emerging science of psychology in those days. The desire to maximize profit and cut losses was seen as a simplified version of the general principle that man acts to gain pleasure and to avoid pain.

Psychology has made great strides since the days when the economists first consulted it and obtained their postulate of the profit motive. When we approach it today for a similarly simplified but useful approximation to human motivation, we may be helped by guidelines from a much more sophisticated science. However, there are many competing emphases in psychology today which complicate our choice. There is the tradition from Pavlov with its formal, very elegant theories of conditioning and reinforcement; here, Homans (1961) has already convincingly shown the usefulness of Skinner's (1953) theory of learning for elementary sociological processes. There is the tradition from Gestalt psychology that emphasizes perceptions and their changes; here, Festinger's (1958) theory of cognitive dissonance has already brought light to several sociological topics. And there is the tradition from Freud which has proved its relevance for sociology, for example, in the work of Adorno, et al. (1950). Thus, when a sociologist approaches psychology for guidance about useful motivational assumptions, he is bewildered by a cafeteria counter of ideas, nearly all of which have been certified useful by colleagues in anthropology or sociology who have already tried them. Our choice must of necessity be arbitrary and can only be justified in terms of the theoretical advances it brings our own field of sociology. Mine is taken from the realm of ego-psychology that has emerged in the wake of psychoanalysis.1)

In ego-psychology we again and again find the idea that man acts to realize his self-image, and is particularly concerned with the preservation of his self-evaluation. This is apparent in the so-called "defense mechanisms" [A. Freud, 1942]. These may be viewed as typical ways in which our perceptions and talk change when our favorable opinion of ourselves is threatened. For example, sometimes a man does not admit to anyone, not even himself, that he hates his wife, but keeps insisting that everything is all right in his marriage; he has repressed his hostility. "Repression" is a drastic step; it implies a refusal to acknowledge what is apparent to intimate associates, in order to preserve a favorable self-evaluation. Other steps might lead to the same end. A worker who is doing a poor job may blame it on his poor equipment. Instead of saying, "I have done a poor job," he says "The tools were inadequate." Actions that are incompatible with his favorable self-evaluation tend to be described by the actor as acts of other agents: this is the well-known process of "projection." Its typical expression is that "they" are the objects of blame, not "me." Another example is the sexually weak man who never misses an opportunity to tell others of his sexual adventures. To protect the evaluation he enjoys, he pictures his actions as being the opposite of what they really are. Actions that are incompatible with his favorable self-evaluation thus tend to be described as opposite to, or different from, what they really are: this is the phenomenon of "inversion" (or reaction-formation). In clinical experience one can observe whole chains of defense mechanisms. In our culture, the homosexual, for example, tends to avoid recognition of his homosexuality in order to preserve his self-respect (repression). In addition, he may deny that the homosexual impulse comes from himself, saying by no means, "I am homosexual," but rather, "You are homosexual" (projection). Furthermore, he may convert his feeling of affection into its opposite, saying not, "I love him," but "I hate him" (inversion). Also, he may transfer the latter into "He hates me" (new projection) and thus develop paranoid ideas [Sears, 1951, Ch. 7].

The underlying motivation for all defense mechanisms may now be stated as a formal postulate:

This will be called the Identity Postulate: it indicates the more stable core – the attitudes toward oneself – of the bundle of activities we call a person, and it suggests that preservation of this core governs all other activities. We shall now explore some manifestations of this motive in society.

Our way of formulating the Identity Postulate conveys several simplifications and restrictions. The first of these is implied in the word "self-evaluation." For it is clear that in our self-image, aspects other than our self-evaluation may also have motivational significance. We act to preserve also our self-descriptions (the cognitions we have about ourselves) and our self-prescriptions (the exhortations we give ourselves, that is, the expression of our will). However, the preservation of self-evaluations seems by far the most potent. Thus, we make a deliberate simplification by saying that we act to maintain self-evaluations, rather than the entire self-image.

A second restriction in our version of the Identity Postulate is conveyed by the word "maintain." Earlier, we thought that the proper word would be "enhance," so that we would have a complete parallel to economic theory: maximization of profit in economy should correspond to maximization of self-evaluation in sociology. Interestingly enough, this is contradicted by research: the normal state is one in which we maintain rather than enhance self-evaluations. Malewski (1962), in a study of girls who spent about four weeks in a summer camp, provides an illustration of this point. At the onset of his study, the girls were given the opportunity to rank their liking for a variety of gifts to be given to them later, in return for their cooperation. Girls with a positive self-image developed a liking for the gifts they had chosen when they actually received them, and tended to dislike the items they had initially rejected. But the girls with a negative self-image, which was confirmed in the course of the experiment, did the opposite. In the end it was more in keeping with their unfavorable conception of themselves to find those gifts more appealing that they initially had ranked as unattractive. 2) The problem of enhancing self-evaluation (achievement motivation) will be treated later as a special case of the motivation to maintain self-evaluation.

A third limitation of the Identity Postulate is found in the notion of "action repertoires." A person's action repertoire is defined as all actions of his of which he can have cognitions at one period of time. Our action repertoire normally grows with experience – this is one reason why older people usually are wiser than younger – but it may also be reduced over time by amnesia, senility, and the drying-up of imagination. One reason for our concern with the biographies of the persons we study derives from our need to know their action repertoires: we want to know their past encounters, what they have seen and done. Some actions within our repertoire may be imaginary, in addition to those in which we have actually engaged. Reading biographies and fiction, listening to tales and anecdotes, watching others in real life, on stage or on screen, all these experiences enhance our action repertoire. People with perception, intelligence, empathy, and imagination can take special advantage of such possibilities to broaden their action repertoire. The Identity Postulate does not assume that a person will do "anything" to maintain his self-evaluation; it merely suggests that he will do something within his action repertoire to achieve this end.

Our conception of the action repertoire gives a play to creativity that is not readily available in theories of learning or the sociological systems that have been built on them [Skinner, 1953; Homans, 1961]. Of course, when one is at a loss about what to do one may conceivably perform a random array of actions to see which one works or is rewarded. But normally we simply think about a situation, have daydreams or fantasies, doubts, and hesitations. This is basically a play with symbols (and the images they stand for), in which we try to ascertain how the various alternatives within our repertoire stack up within our imagination. We may let out trial balloons to outsiders, but most trials remain in our brain as our private, unspoken simulations. Thus within our action repertoire – that is, all actions we know or can imagine – we strike the ones most compatible with our self-image. They need not be repeats of past activities; they may be unique, creative, and original. 3)

The Identity Postulate suggests that an individual's actions emerge to the extent that they maintain his self-evaluation. However, one of Cooley's propositions holds that a person's self-evaluation tends to be a consequence of the evaluations of him made by others in his milieu (evaluative internalization). Hence, the Identity Postulate and Cooley's proposition render the deduction:

This deduction is important to sociology since it indicates that our associates can manipulate our motivation simply through their evaluations of us, that is, through their approval or disapproval. This Theorem of Social Motivation, as we shall call it, enters as the simplest motivational assumption in sociology. It says that among the rewards in social life are the favorable esteem and attitudes we might receive from others; public opinion of us is thus a key to our motivation.4) William James saw the motivational significance of evaluations by associates as early as 1890, when his famous dictum appeared: "A man's social me is the recognition which he gets from his mates.... Properly speaking, a man has as many social selves as there are individuals who recognize him and carry an image of him in their mind. To wound any one of these images is to wound him" (1890, Ch. 10, Sec. 1b).

Concern with favorable evaluations is a broad tent under which we shall soon find such things as preoccupation with approval, recognition, admiration, good will, esteem, love, rank, honor, as well as all the honorific garnishings that come with money, power, competence, holiness. It is abundantly evident in the hubris, the excessive pride that ancient Greek dramatists assumed to be the root of all human disaster. And it is equally evident among the forces that lift men to new heights of achievement in economy, polity, science, art, religion, and morals. In short, a desire for favorable evaluations goeth before a rise as well as a fall.

Only actions that are "visible," that is, those that can be accurately described by associates, can be evaluated. Thus, people tend to make visible their favorable attributes, and exaggerations enter social life. And, of course, to avoid unfavorable evaluations, men are apt to keep some actions or attributes hidden. In this way, a certain evasiveness seems to enter all social intercourse. However, it follows with equal logic that men concerned with maintaining a given level of evaluation will tend to reject anything but gentle flattery and to keep a supremely favorable evaluated action or attribute somewhat under the barrel. In this way, understatements become part of social intercourse. The fact that variations in visibility have motivational significance is one of the bases for the power of publicity. The manipulation of visibility of actions is a gentle yet effective device to change people's behavior without issuing new prescriptions or appearing "bossy" [cf. Blau, 1955, pp. 34-44].

The associates whose opinion of us we particularly value are called "significant others." To know anyone well we need to know who his or her significant others are.

Psychoanalysis has stressed the importance we attach to obtaining favorable evaluations from childhood associates, particularly from parents; some contemporary thinkers have stressed the point that certain persons strive primarily for favorable evaluations from peers, that is, contemporary associates resembling themselves in terms of the positions they occupy. In a well-known discussion of the American character, such persons are called "other-directed" [Riesman, 1953]. Another classification of the sources of significant evaluations contrasts associates in neighborly social relations with associates who are relative strangers. Those who seek favorable evaluations from the former have been termed "local" in their outlook; those concerned with the evaluations of the latter have been termed "cosmopolitan" in their outlook [Merton, 1957a, Ch. 10].

The two classifications overlap. Thus, we have four character types (see chart below) into which individuals can be placed on the basis of the source of the evaluation they would like to maintain.

Type I, the inner-directed local, is represented by the person who all his life arranges his actions so that his parents will approve of him. He strikes the observer as traditional. Type II, the other-directed local, is represented by the person who shifts his actions in order to be appreciated by whoever happens to be close to him. He strikes the observer as obliging. Type III, the inner-directed cosmopolitan, arranges his actions so as to obtain the approval of ideals he has learned about in the past from people he has not known intimately, for example, in the course of his higher education. He strikes the observer as principled. Type IV, the other-directed cosmopolitan, adopts a pattern of action that is approved by whoever is at the moment the spokesman in his organizations, or whoever dominates the mass media or marketplace. He strikes the observer as flexible. There is no basis, in my opinion, for the assumption that all of contemporary society is becoming other-directed. For one thing, the rapid growth of the medical, legal, teaching, and engineering professions in modern society represents, in all likelihood, an increase of people that can be called principled (Type III).

One should also be open to the possibility that a person's significant others are only anticipated, or even imaginary. It is not only in fairy tales that a girl strives today to please a Prince Charming who may appear one day. The striving businessman has already in his imagination anticipated his meeting with some great banker and adjusted his actions to win the latter's approval. The artist, misunderstood by his contemporaries, continues to create, upheld by the belief that he will please a future public that finally will appreciate him. The statesman becomes concerned over the judgment of future historians. There are, in other words, "anticipated others" whose evaluation of us may matter greatly.

Positions can be arranged in hierarchies from the highly evaluated ones to the lowly, and we use the word "rank" to indicate the evaluation of a position. In practice, ranks become convenient bundles of evaluations of their occupant. Instead of saying that a certain person in Cambridge, Massachusetts, is a gentleman of some learning with sound judgment of the work of others, that he speaks and writes with a sense for the essentials, and is well able to teach others the knowledge of his specialty, instead of saying all this, we simply say he is a "professor at Harvard." By thus mentioning his rank we include all these and other evaluations. Had we called him "graduate student" or "bookstore clerk at Harvard" we would have caught different bundles of evaluations. The fact that ranks are evaluations makes it possible to state a special case of our Theorem of Social Motivation:

This Theorem of Rank Motivation is basic to the control of motivation in hierarchical organization. A company of equals, such as a university faculty, can control the motivation of its members only through public opinion, that is, the exercise of approval and disapproval, the giving and withholding of esteem. A company with a variety of ranks has an additional hold on its members' motivation; it can play on the fear of demotion or fear of being bypassed at times of promotion. The more sustained and thorough control over motivation an organization needs, the more it has to resort to hierarchy of ranks. A very hierarchical structure like the army acquires such control over the motivation of its officers and soldiers that they can be sent to face their own death and destruction.

People's preoccupation with rank is too well known to require elaboration. What needs elaboration is rather the diversions we may use against such preoccupations. We may meet our associates as superiors, equals, or inferiors in rank. The rewards in these social relations differ. The good superior receives "deference" and the good inferior receives "concern." Both deference and concern have a favorable evaluation as a common denominator. But true equals have "fun," the delight and joy of being together in play. Only equals meet in "play." When inferiors and superiors meet there is always "work" of one kind or another. Because fun is available only in equalitarian relations, however temporary they may be, it is an antidote to rank and to concern over favorable evaluations from others. From time to time most superiors give up the deference of their high position to enjoy the fun and joy of play with equals, and if left alone, inferiors are likely to turn work into play to share its joy. However, the fun of play is not lasting; it disappears as soon as someone fails to act as an equal or as soon as someone, through a tactless remark, makes the participants aware of their differences in rank, or in riches, power, education, or taste.

Economy, polity, intellect, religion, art, and morals are the institutional realms of society. Each one has an institutional value: prosperity, order, knowledge, sacredness, beauty, and virtue are the corresponding institutional values. Persons who control about the same degree of an institutional value belong to the same stratum. Each institutional realm has its typical mode of stratification. We say that command of prosperity defines riches, control of order defines power, command of knowledge defines competence; similarly, command of the sacred indicates holiness, command of beauty indicates taste, and command of virtue indicates rectitude. The Strata-Prestige Postulate now states:

Thus men of riches, power, competence, sacredness, taste, and rectitude are more likely to receive favorable evaluations than men who do not command these values. Perhaps it is even likely that women are more apt to fall in love with such men because of what they represent than with others.

A concerted effort to nullify the Strata-Prestige Postulate was made in Calvinist doctrine. Here one held that membership in the strata of the elect and the damned is not subject to human knowledge; it is invisible. Hence the elect cannot enjoy any special prestige among other human beings. The force of the process depicted in the Strata-Prestige Postulate could, however, not be contained. Weber (1901-02) suggests that the effect of Calvinist doctrine was simply that one began to use known positions in realms other than the religious as indicators of a person's position in the religious realm; for example, one's economic standing became a rough sign or index of one's religious standing.

Each institutional realm has potentially a typical "reward pattern" which provides for a regular relation between placement in a given stratum and the receipt of specified favorable evaluations known as honorific rewards. Changes in the degree of control a person or group has over institutional values are thus more or less automatically related to changes in specific honorific rewards. Such patterns are strategic objects of sociological study. 5) In the Western economy, the signs of success honored by the larger community consist of visible goods and services, number of residences (rated as to size and location), number of employees and labor-saving devices, annual charity and (in some countries) taxation contributions, and, in the case of firms, annual reports of earnings. Often the economic reward pattern also allows the successful individual to attach his name to his enterprise (e.g., the Ford Motor Company, the Krupp Works).6) In the contemporary polity the reward pattern centers on symbols of position (such as titles and uniforms), on constant publicity and evaluation by mass media, on approval from cheering masses, on ceremonial rites and decorations. Successful men may also have cities, roads, bridges, public buildings, acts of legislation, and the like named in their honor; they may have statues, portraits, and memorial plaques created to commemorate their deeds. An ultimate evaluation, here as in other fields, may be the judgment of future historians. In the realm of intellect, a firmly established pattern ties the name of the scientist to his published contributions to knowledge. Ideally, scientific articles and monographs get into print only if they contain new knowledge, and a scientist's own publications are often more dear to him than his worldly possessions. And any new scientific report is expected to recognize in text or footnotes the authors of the more relevant ideas entering as parts of a discovery, technique, or argument.

In religion various signs make visible how close a person is to the sacred (for example, special gifts of tongue or admissions to a graded series of holy rites), and call forth reverence from the community. The ultimate basis of the religious reward pattern is, of course, the evaluation of people by the divine assumed by the believers. In art the reward pattern at present is less clear cut. For a contemporary painter, for example, it would include the number of his private shows, the number and kind of reviews by critics, the rating of the galleries in which his paintings have been exhibited, the number and prominence of the collectors who have acquired his paintings, the number of his paintings that hang in museums. In the realm of morals, Western culture has not developed any elaborate reward pattern. The badges of rectitude are few; while virtue is, of course, appreciated, to do it visible honor is widely thought to cancel it. Only history will tell whether this lack in Western culture will prove significant or not.

The institutional reward patterns thus order and make visible the evaluations that come with membership in low and high strata. They show the way in which the Strata-Prestige Postulate manifests itself in practice. If we now combine the Strata-Prestige Postulate with our Theorem of Social Motivation, we obtain a Theorem of Strata Motivation:

Thus the dynamics of motivation become located not only in fear of bad public opinion or loss of rank, but in fear of downward social mobility in the stratification system of the larger society.

The parallel between the Theorem of Rank Motivation and the Theorem of Strata Motivation is obvious. The former indicates how the imposition of a hierarchy of ranks on an organization or market controls the motivation of the participants; the latter shows how the stratification in the society as a whole gives structure to the motivation of the population.

The fact that hierarchy and stratification are linked to motivation has led some sociologists to argue that hierarchy and stratification are inevitable: "The main functional necessity explaining the universal presence of stratification is precisely the requirement faced by any society of placing and motivating individuals in the social structure" [Davis and Moore, 1945, p. 243]. But this is only a partial truth, although admittedly important. As we have already seen, there are forms of social motivation based solely on public opinion of the performers which are unrelated to hierarchy and stratification [cf. Tumin, 1953].

More important, hierarchy and stratification can acquire full motivational force only in organizations in which individuals can advance freely and in societies in which mobility is the rule. If ranks are fixed in advance and strata are inherited from parents and organized as castes, they lose their motivational dynamics; they cannot be gained or lost by any action in any action repertoire. Ascription thus tends to nullify the Theorems of Rank Motivation and Strata Motivation. For this reason, where high degrees of ascription prevail, social structures cannot arouse much motivation. A business firm in which nepotism prevails is not full of highly motivated executives. A caste society is stamped by a spirit of fatalism. When ascription loses its grip, new motivational raw material can be tapped. The armies of the French Revolution dropped their caste system so that others besides noblemen could become officers, and thus tapped new motivational resources needed for victory. The American South is finally giving its Negroes a chance to break the caste barrier and is removing the old fatalistic attitudes of the Negroes as it desegregates. Organizations and societies on the move thus unleash their motivational resources by removing ascription.

The opposite process – increasing ascription – dries up motivation. The Reformation stressed that man could do very little to achieve his own salvation. The system of confession, special masses, sacraments, tithes, letters of indulgence, and pilgrimages that the Catholic Church had developed for the laity as steps on the ladder to heaven was destroyed wherever the Reformation was victorious. Salvation was instead seen as a gift of God which man might receive but which he could not claim by any religious rites or good deeds. This element of ascription in religious stratification greatly undercut the potency of the religious reward system. The Protestant clergy made several efforts to introduce some elements of achievement into the system. For example, during the period of orthodoxy it was particularly strongly stressed that only those who held the right beliefs would receive the gift of grace. This made the achievement of knowledge of the doctrine and the defense of the "true" doctrine against "false" ones highly rewarded pursuits.

The Lutheran doctrine emphasized the belief that one could by one's own efforts, if not gain grace, at least fall from grace. The fear of downward mobility – the core of the Strata-Motivation Theorem – thus could to some extent still be operative within the religious realm under Lutheranism. In the orthodox Calvinist doctrine of predestination, however, complete ascription prevails; the selection of the elect and the damned is already made and will not change. The Theorem of Strata Motivation is here entirely out of commission. Religious rewards in this setting thus become less effective. The stage is set for any form of secularization because more competitive rewards are available in other realms of society. This is another development of Weber's previously cited study of the relation between the Protestant ethic and the spirit of capitalism.

Outside the realm of ascription, the Theorems of Social, Rank, and Strata Motivation play on a man's fear of loss in standing. The risk of such losses may be small or great, but some losses seem inevitable; this is the factual basis of our statement that it is only human to fail, and of the pragmatic view of life that takes not an absolute but a statistical view of success and failure. It follows plainly from our assumptions that men want to insure themselves against such losses. The basic principle of insurance is to spread the risks, to avoid putting all eggs into one basket. This applies also to evaluations: the greater the number of a person's actions that are known about and evaluated by others, the smaller the risk of his being unfavorably evaluated for any single or occasional failure or slip.

This Principle of Insurance makes it possible for us to identify some situations in which the nature of the social relations affects the maintenance of evaluations. The "familiar" relation, by definition, is one in which our associates know a large share of our actions and thus are able to evaluate them. Another way of stating the Principle is to say that evaluations fluctuate less in social relations marked by high familiarity. Combining the Principle with the Theorem of Social Motivation we obtain:

We might call this the Theorem of Affiliation since it tells of a need for joining into friendly relations.

The scope of this proposition has been known for some time [cf. Rundblad, 1951]. The decline in ascription that was implied in the urbanization and industrialization of Western Europe brought about an upsurge in friendly voluntary association in the nineteenth century. Similarly, contemporary so-called underdeveloped countries seem to emerge with the same rich flora of associations when their structures change toward that of a modern society. And long before, de Tocqueville, observing the settlers from aristocratic (that is, ascribed) Europe in achievement-stressing America was surprised by the abundance of voluntary associations they created in their new land. Teenagers also illustrate the process: from the secure ascription of their parental homes where nothing they do could deprive them of their positions as sons or daughters and tenants, they branch out into an adult world in which schools fail them if they do not meet standards, landladies evict them if they do not pay rent, and employers fire them if they do not perform. Their abundance of cliques and friendships meets their thus generated need for familiar relations until they find this support in families of their own.

In all these instances, we might assume that the new associations help to maintain a certain inner security among the participants in the face of threats to their new status and the fluctuations of evaluations they receive in their new world. However, the Theorem of Affiliation merely states that certain aggregates of persons will have a motive to form or enter into some risk-spreading social relations. Whether or not they will actually do so is a different issue, one that depends on the opportunities open to them by virtue of their positions in the social structure. If they are already involved in some social relation characterized by familiarity, they may revitalize and intensify their participation in it. If they have access to already existing groups that meet this criterion, they may join them. If they do not know any such groups but interact with others who have the same motive, such a group may be formed.

Psychologists, of course, have reason to deal with many more aspects of the need to affiliate and with its individual variations. Our argument merely suggests in what social structures such needs will be strong – and this is usually all we need to know as sociologists. It is noteworthy that a broad proposition about affiliation for this purpose can be derived from the same motivational postulate as the rather different propositions of rank and strata motivation.

All scales of evaluation have a range. This is the difference between getting an A or an F in a college course, the gap between the very rich and the very poor, the distinction in military life between the private and the general. Through our encounters we become accustomed to a rather limited range: most students never meet one whose typical grade is F; most persons know no one who has an income of 10 million dollars a year, nor anyone with an income of 100 dollars a year; most people in the Army do not know generals, nor privates facing dishonorable discharge. Our encounters usually provide us with a limited, not a full range.

When a person acts to maintain the evaluation he receives in his encounters, this means, to begin with, that he maintains it within the limited range to which he is accustomed. To be thrown outside this accustomed range through a sudden loss of social station and resources, a sudden catapulting promotion, or stroke of financial luck, makes a man lose all his bearings. As a plant or an animal accustomed to a temperate zone runs the risk of perishing when suddenly transplanted to a tropic or Arctic climate, so a person also risks destruction when suddenly thrown below or above his accustomed range of evaluation. The territory outside this accustomed range was defined by Durkheim (1897) as anomie. People living in anomie do not respond to the Theorem of Social Motivation. Anomie thus results in the withdrawal of motivation on all levels that have concerned us. It reduces man to an animal unable to respond to any social reward. Anomie thus is the most dangerous poison to any human society.

It is the sudden change outside the accustomed range that brings about the anomic state. If the shift upward or downward is slow, there is time to acquire new anchorage points and units of evaluation, and thus extend one's scale to realms in which one previously did not know any of the bearings. Even those who rather suddenly find themselves in an anomic range but manage to survive the first confusion eventually build scales that fit their new circumstances. It is instructive to contemplate that a skid row, something that seems so anomic to the outsider (including most sociologists who have written about it), upon close analysis manifests its own distinct scale of evaluation and a hierarchy of status based upon it.

An evaluation is a complex communication that cannot be understood unless we know three separate components:

1. A unit. This may be a grade in school, items of visible consumer goods, a rank in an organization, et cetera.
2. An anchorage point. This may be the average grade in school, the average standard of living, the typical rank for one's age group, et cetera.
3. An evaluative score. Here we become specific, and speak of, for example, a B-plus student, a boy with generous allowance and car of his own, the president-elect of an organization.

The score is, of course, the evaluation proper. But it is clearly dependent upon the other two. Units may change. Better tests allow for more discriminating grades than A, B, C, D, and F and we may assign numerical grades ranging, say, from 50 to 100. Car models and other status symbols may change to make it easier (or more difficult) to see what is plain and what is fancy. Personal distinctions and ranks may multiply. Anchorage points may also change. A new admission policy may bring an influx of very bright students so that the average grade is pushed upward. The standard of living may rise so that every student has a car. The number of executives may rise. All this illustrates that the evaluation is a function of the size of units in use and the location of the anchorage point employed. Only as long as all the encounters a person has employ the same units and the same anchorage point, will the scale of evaluation remain stable.

Now, our Theorem of Social Motivation suggests that we want a stable evaluation. We can spell out the prerequisite for this in a Theorem of the Preservation of the Reward System:

People accustomed to receiving high evaluations, for example, persons who occupy high ranks or strata, are particularly likely to defend vigorously their scales of ranking and the existing methods of scoring people's standing as high or low. This gives a special flavor to their outlook; they engage in "status politics" [Hofstadter, 1955]

Persons sharing a common style of life centered on their particular reward pattern form a vested interest or "status group." Their concern with the protection of their reward patterns generates much of the dynamics in a society. Max Weber focused a good deal of his attention on this problem. His historical writings contain analyses of patrimonial vassals, Junkers, officers, civil servants, priesthoods, and several other vested interests. One conclusion is that much of human history is a contest between various vested interest groups.

The fact that some men are more striving, competitive, and ambitious than others is a problem for the psychologist; the sociologist needs to know only when one society, community, or institutional realm becomes more concerned with achievement than another. We need to know the conditions when the motivation to maintain a favorable evaluation is transformed into the motivation to enhance evaluations for persons with all varieties of traits and childhood experiences.

Generally speaking, achievement motivation occurs when people fail to preserve a stable reward scale and have to keep up with a changing reward system. Such instances are easy to illustrate. For example, changes in the units and anchorage points of scales of evaluation are inherent in the age-grading that prevails in all human societies. The baby – at least, a physically fine baby – can count on much love or appreciation without regard to how well he performs. But as he grows older, more effort and ingenuity are needed to maintain the evaluation so freely accorded during childhood; with age, the anchorage point moves upward until old age again relaxes it. Societies in which prosperity, knowledge, and other institutional values are expanding are also ones in which the anchorage points for the corresponding reward patterns are pushed upward. In such societies one has "to keep up with the Joneses" to maintain evaluation. Here, as in age-grading, the process is beyond the individual's control.

The crucial element in such instances is our encounter with persons who use different units and anchorage points in making their evaluations of us. We may write the Theorem of Achievement Motivation in this way:

Thus we see the motivation to achieve as one in which we maintain our relative evaluation. Anchorage points and units change, so we must, like Alice in Wonderland, run faster and faster to stay in the same place. Thus we have derived the motivation to achieve from the motivation to maintain.

A possible cue for the terms of exchange is given by Homans' (1961) so-called Justice Proposition, which says that anger occurs in a man when his rewards are less than proportional to his investments. The theoretical importance of the Justice Proposition is not that it predicts when people become angry – although this is admittedly useful to know – but that it provides one point of equivalence between different scales of evaluation. This point of equivalence is the minimum reward for a given investment needed to avoid anger. By knowing this critical amount of reward we may begin to compare evaluative scores from different encounters. However, the Justice Proposition is rather metaphorical, particularly since "investment" has no clear meaning in sociology. Moreover, to convert scores from one evaluative scale to another we need a minimum of two points of equivalence, and the Justice Proposition provides only one. Let us therefore extend and generalize it to serve our purpose.

We define a person's "commitment" to a set of actions as the extent to which his self-image is dependent on his engaging in these actions. We know that a man is highly committed to the writing of poetry if he feels less than himself in periods when he is unable to form his verses. If his failure to write does not affect the way he feels about himself, then his commitment is low. Commitment thus implies that a person has "invested his ego" in some activities, that they are relevant and important to him.

The relation between commitments and rewards may be reviewed in four combinations:

Homans' Justice Proposition says that the third type represents angry men. Homans mentions in passing that what we have denoted as the second type contributions.

Anger and guilt are large words. The third type may not really be angry, merely irritated at his bad fortune. The second type may not really be guilty, merely elated over his luck. But it seems clear that we might expect an emotive reaction when commitments and rewards no longer are commensurate. The quality of the emotions involved will vary in the second and third type, but a psychogalvanometer would give a higher reading in the second and third types than in the first and fourth, where the reading would be more neutral.7) Thus we have two emotive points and we have extended the Justice Proposition to read:

Granted a given commitment, the two emotive reactions thus obtained define a kind of freezing point and boiling point for any scale of evaluation. Evaluations falling within the range between these points are thought of as "just"; evaluations outside this range are "unjust."