Light From Psychology on Intercultural Relations

Muzafer Sherif

[Born in Odemis, Izmir, Turkey, Dr. Sherif did his undergraduate work there at International College. He holds the M.A. degree from both Istanbul University and Harvard, and the Ph.D. degree from Columbia University. He has been instructor in psychology at Gazi Teachers College, Ankara, then assistant Professor and finally full professor of psychology at Ankara University. In 1945, he returned to the United States on a fellowship from the State Department and went to Princeton University as a research fellow. In 1947 he was appointed a research fellow at Yale University. He is now professor of social psychology at the University of Oklahoma. He is the author of several books in English.]

I CONSIDER it a great honor to be here tonight and I feel quite at home on two grounds. First, I used to hang around here in my student days when I was working for my degree and worrying about my experiments. Then, I also feel at home because I am interested in the same problems that you are, both as a social psychologist and as a person who has seen human conflict and misery of all kinds in one of the very delicate spots of the world—my own country of Turkey. I cannot go into concrete details as I would like to tonight, but prefer to stay on the theoretical level. I hope the implications will be clear. In this discussion of the staggeringly intricate problems of intergroup relations, I propose to follow three lines. I will begin with a brief summary outline of what seems to me the only feasible approach in social psychology to intercultural problems. Second, I small consider the essential features that characterize the formation and functioning, of any group, for example an in group. My remarks are based on intensive studies of small groups, but my hunch is that they apply in

(111) all essentials to larger social units as well. Finally, I shall have something to say about intergroup tensions and conflicts.

I have organized my material under the first two heads, on the approach to intergroup relations and on the main features of in-groups, so as to lead directly to my third topic of tensions between groups. My findings have been chosen selectively, in other words, with considerations of intergroup relations in mind. Before going any further, perhaps there is need for some clarification of terms. In my discussion I shall simply use the word "group" instead of "cultural group," because every human social unit—whether relatively simple or complex, relatively slightly or highly developed—has a culture of its own. I take "culture" as referring to the accumulated products of interaction within the group, such as the status hierarchy, the social organization, the division of labor and work routines or techniques, the standards of living, beliefs in magic and myth and religion, the language and all art forms, standards of conduct, and any other social values or norms. This characterization is in line with the definition of culture given by Professor Melville Herskovits, in his recent book Man and His Works, as the man-made part of the environment.


When a person starts with methodological questions the implication is that his field is not yet a finished science. There are things to be ironed out. So I must begin with a few words about our approach, in social psychology, to the problem of relations within and between groups. The momentous events of human history in the last decade, and particularly the great changes brought about during the last war and as a consequence of it, have created a hitherto unprecedented state of disequilibrium and flux in the relations of social units and of nations. Modern means of transportation and communication, along with other developments derived from technological advance, have made the world too small for tic isolated existence and functioning of nations, small or large, no matter how distant from the trouble centers or how self-sufficient these groups might have previously been. In this general world setting, no human social unit can function as a closed

( 112) system today. No human grouping, no matter how weak or how powerful, has an entirely independent existence. The current state of world affairs is ever bringing all social units into closer and closer functional relation.

The rapidly increasing concern in many quarters (both academic and more practical) over the vital and frequently grim problems of intergroup relations is an inevitable product of this worldwide situation. A scholar has to keep an eye on international events if his social psychology is to have any realistic contact with the world of reality. Time and again, psychologists have displayed a tendency to be so absorbed in their own abstractions that they have lost sight of vitally important factors in real life. Consequently, their psychological formulations have not always been of appreciable use to men in economics, politics, administration, and other more concrete fields. The one healthy viewpoint that social psychologists can afford in this connection is to keep in mind, at every step, that the problem of intergroup relations is not solely psychological and that, hence, they cannot unaided reach generalizations or write prescriptions for resolving tensions. We have to keep a keen eye on concrete developments in the economic, political, military, and related fields. Because of the lack of such realistic perspective in human events, psychological accounts of the causes of group friction—of war and peace or of human cooperation and competition—have been nothing but one-sided abstractions. I feel that it is necessary to make a frank admission of this state of social psychology, to tell you at the outset that one should not expect too much as yet from us social psychologists. We cannot yet produce sufficient light for everyday practitioners to see by. We cannot deliver what we do not have. Social psychology has not yet established itself with the comprehensive and synthesizing perspective I have mentioned. It is still in the state of formation, erratic and inconsistent at many points. I am sure you are finding that out for yourselves as you read or listen to various social psychological presentations.

Any social psychologist who takes his work seriously should preface his communication to a Serious conference such as This with a frank description of tie state of social psychology Today. (:Iii and authoritarian generalizations in the name of science have not

( 113) advanced the welfare of human relations much, nor our scientific knowledge for that matter; therefore, it is only fair to you to ask that you take all my remarks in this discussion as tentative conclusions reached by one social psychologist. Some of the evidence on which I base my remarks can be found readily in the works I have appended to this paper.

Before getting to the more positive side of my discussion, I shall point out briefly what seem to me examples of influential, but one-sided approaches to the problems of human relations, especially human relations on an intergroup level. As the first illustration of such defective reasoning, I take theories which place great weight on alleged hereditary differences in psychological and biological endowment between human groupings. Here, of course, I am referring to racist theories which attempt to explain intergroup relations—the advantageous and dominating position of one group over another, the friction or understanding between two groups, or any other pattern of relationship—on the basis of congenitally given or innate psychological traits of the parties involved. Such theories are nothing but self-glorifying justifications for domination and exploitation in practice. A glance at representative works, such as Julian Huxley's We Europeans and Klineberg's Race Differences, is enough to convince one that the alleged superiority or inferiority of human groupings with respect to certain biological and psychological traits, or the very existence of in-born racial traits, is not grounded in solid scientific fact. Therefore, any social practices which are based on the remnants of the White Man's Burden attitude, or its more modern and subtle versions, are doomed to failure. As has been demonstrated time and again in the colonial world, in recent times, racist theory brings not only misery to the objects of its practice, but also disillusionment and frustration to its practitioners.

Now I shall call your attention to an approach which is more seriously considered, especially in academic quarters, but which nevertheless is equally one-sided. Here, I refer to theories of primary instincts. The social psychologies of William McDougall and Freud, which posit a number of basic instincts or inherent tendencies toward love, (lest: ruction, or aggression, are outstanding exponents of this line of thought. This approach, which may be

( 114) characterized as individualistic, finds in human instincts or alleged deep-seated drives the only effective explanation of harmony and friction in all human contacts, in intergroup as well as interpersonal relations. Surely no sane person can deny the importance for human relations of such basic human needs as hunger and thirst or sexual desire. Acceptance of these cravings as essential in the life of individuals is one thing; but to use them to explain everything in the experience and behavior of groups is quite another thing.

Certain well-established facts, verified time and again, force us to reject such instinctive drives as the sole explanation of human interaction. For example, it has been stressed for decades by sociologists and more recently by psychologists that the experience and behavior of individuals take on new qualities in group situations. In the monumental work of Murphy, Murphy and Newcomb, which is the best summary of social psychology up to 1937,

and in the recent volume on Readings in Social Psychology, which brings the subject down to 1947, one can readily find ample evidence of this fact. It is factually erroneous, therefore, to posit instinctive human drives and to examine them in isolation and out of social context. The individual interacts functionally and on a reciprocal basis with all other members of a group. His experience and behavior cannot be adequately understood without taking into full account the frame of reference of the social situation. In other words, human behavior is shaped jointly by two sets of factors that are both organically related to the social environment: those coming from within—motives, attitudes, and the like—and those received from contacts with the group. Both sets have to be taken into account in an attempt to explain any human reaction. This is especially true in the analysis of intergroup relations. The search for man's aboriginal motives as the only explanation for group friction simply leads up a blind alley.

The foregoing considerations were presented by way of defining my methodological approach to the main discussion. I shall try to give you in a nutshell the substance of what I was intending to establish.. Explanations of group or intergroup behavior on the basis of congenital traits or deep-seated instincts are alike one sided and doomed to failure. This is, however, not to deny the import-

(115) -tance of human needs in the analysis of interpersonal or intergroup relations. But motives can be more realistically appraised if they are studied in their actual functioning in a social context. All of our major drives come into play in a group setting of some kind or against a group background. In order to understand intergroup tension, therefore, we must start with the psychological analysis of social units.


Social units may be composed of many individuals, as is the case with ethnic groups or nations, or they may consist of a few members, as is true of cliques, gangs, fraternities and sororities, a neighborhood church, or closely knit clubs of various kinds. Social psychology, which deals with individuals in relation to social situations, is not yet equipped to handle effectively the immensely intricate large groups represented by whole peoples or nations. The same may be said with respect to the huge problems involved in the study of national character. However, even small groups show the essential structural properties of groups of any size. By "structural properties" I do not mean any vague abstraction but the relation among the group's members which is reflected in their behavior and affected by the functioning of the whole. Since small groups lend themselves much more readily to direct observation, and since they possess the essential features of any group, large or small, it will be rewarding at this early stage of developments in social psychology as a science to concentrate on the study of these smaller units. After establishing certain generalizations, we may test their implications for larger groups at a later stage.

It is, in my opinion, especially true that small in-groups formed under the stress of motives shared by the individuals concerned may serve as prototypes for in-groups of more complicated kinds—with respect to the organizational properties revealed in the hierarchy from the leader on down, to their cultural products, such as codes of behavior and the distribution of social roles among the members, and to their positive and negative relations with other groups of a similar nature. Cliques of adolescents which abound almost everywhere in societies in a state of transition, gangs of

( 116) youngsters such as are so prevalent in large cities, at least in America, and the newly organized clubs of socially climbing societies are all excellent examples of the small, spontaneously formed in-groups I have in mind. The particular activities which these groups serve—such as contacts between the sexes, social climbing, economic gain, or providing a sense of security—are immaterial for us in the present context. Our interest is, for the moment, confined to the excellent opportunity they afford for studying in-group formation and the relation between groups on a miniature scale. Another type of group which is especially good for study purposes is that of a set of people who have become isolated, or set apart, through some extraordinary event—for example, through shipwreck or confinement together as prisoners of war. Under such circumstances we find developing some sort of informal organization and code of conduct which are valid for the particular group during the period of its isolation. Standardized jokes and catchwords are among the group products that tend to arise very quickly in such cases.

With these considerations in mind, I have for the last ten years or so been observing group activities and collecting material bearing especially on the formation of cliques and gangs. As a result, I can suggest a list of minimum essential properties that characterize such groups and their interrelations. A brief glance at these essential features will help to clarify our conception as we proceed toward the most complicated of all problems, viz., intergroup relations. The four characteristics to follow are certainly typical of any small in-group: (1) shared motives of some kind; (2) differing effects of group interaction on individual behavior and experience; (3) a more or less definite structure with established status positions and social roles for all members; and (4) a set of standardized values, mores, or norms. I am using the word "norm" generically to denote any standard of conduct, social value, fad, or fashion accepted by the group. I shall have time only for a few words about each of these features.

First, as to motivational factors. In bringing and holding the group together, shared motives of some sort are basic. These may be rooted in biological need —such as to secure a livelihood, food, shelter, clothing, sexual satisfaction, and the like—or they may be

(117) socially derived. The group may give a sense of security to individuals floundering in unstable social relationships. It may center around gaining distinction or prestige, or around common interests, such as hobbies, social ideals, literature, and the like. Since socially derived motives involved in group formation vary from culture to culture and from class to class, their number and kind multiply with the development and differentiation of the larger social structure. Any one or any combination of human motives, whether of biological or of social origin, may function in bringing a group together and in maintaining its identity. And of course these motivational factors are weighty determiners of the direction a group's activities will take, of the nature of its structure, and of the values or norms which will result from interaction among its members.

However, no matter what the original motives were for bringing the group members together, once a group is formed, other motives enter the picture in an important way. For as the individual participates in group activities and achieves some sort of status, he develops a feeling of "belongingness" which is based chiefly on his functional part in the total structure. Thus each member's sense of security becomes closely tied to his feeling of belongingness and to his particular status and role. His very identity and conception of himself are defined in relation to the group. He strives to maintain and enhance his position by conformity to and active support of the group's norms. Even the satisfaction of biological needs, such as hunger and sex, comes to be regulated by these norms. The manner of achieving satisfaction, for example, is prescribed in terms of the individual's position in the group structure and the social obligations and expectations attached to it.

This leads me to the second characteristic feature in my list, the fact that each individual's experience and behavior is affected, and in varying degrees and ways, by the entire interaction within the group. For not only motives but all psychological functions—perception, judgment, thinking, emotions, and the like-are modified in group situations. The differential effects of interaction are not merely something added or taken away, not a snipping off of veneer or the addition of color, in each member's general personality. Rather, they produce modifications in behavior that are

(118) different in kind from the individual's previous conduct. Of course, qualitative changes in behavior are not unique to group situations; they develop in any individual whenever important new factors enter into his experience. In group situations, the most salient new determinant is the developing social structure.

Here we have my third feature: the definite structure of each in-group. If interaction goes on long enough among human beings a marked system of relations gets established, usually in hierarchical order with the leader at the top and specific positions assigned to all individual members. This is the case, in fact, regardless of whether or not the group is formally organized and the members are fully aware of the structure's existence. The leadership position is attained, not by the mere possession of certain traits or capacities but by the individual's ability to meet the demands of the particular situation in which the group functions. Under certain circumstances, physical prowess or bravery may be the necessary requirement for leadership; under others it may be a question of intelligence or some special talent, of family background or of social prestige.

As the group organization crystallizes and the relative positions of status get established, power relations inevitably arise. Power is vested in the different members of the group in direct proportion to their position in the status hierarchy. Power thus ordinarily resides chiefly in the leader and those around him, both for the conduct of affairs within the group and for relations with other groups. However, the higher the position in the group structure, the greater is the demand on its possessor that he fulfill his obligations and live up to the expectations of the other members. Once he lets the others down, he is likely to fall in the hierarchy. This means that the leadership position is not a fixed, immutable throne. The leader, like everybody else, is subject to regulation by the group. He leads only so long as he stays within a reasonable range of group standards or norms. If he deviates from the social expectations of his fellows too much, he simply loses his weight in group affairs. Under such circumstances, leadership may change hands. I should think that he case of Ramjet McDonald in England, and the tragic fate of Marshal Pétain in France, were telling illustrations of this point in public affairs_ In view of these

(119) considerations, attempts to explain developments, achievements, exploits, catastrophic events, conflict or harmony among groups in terms of the personality of leaders alone have had very limited validity.

The formation of a group structure generates differential in-group and out-group attitudes. Along with a sense of belongingness, the usual group experience gives rise to a "we-feeling," that is, to a sense of solidarity and loyalty to the interests of the group in spite of various degrees of rivalry and friction among its members. If such competition is carried too far, the individuals concerned face various kinds of correctives, such as demotion, punishment, and ostracism. Along with the development of a "we" experience, there inevitably comes about a differentiation of the in-group from all out-groups. This "they" experience may or may not involve tension and hostility, depending chiefly on the nature of the interests of the groups. I shall return a little later to these in-group and out-group attitudes and their consequences.

Finally, my fourth characteristic feature of small groups is their tendency to generate codes and social values or norms. The individual's most important attitudes are formed by the norms of the group and his own status position. It is the functional relation of the group structure and its particular standards rather than any formal pronouncement of aims that is the crucial factor. (This circumstance may explain some existing discrepancies between what we practice and what we preach.) In groups that start without any such codes of behavior these come to the fore in the natural course of interaction. The norms may include the group's purposes, its values, its notions about its own worth and the worth of out-groups, a scale of social distance by which to measure itself and other groups, methods of enforcing conformity, catchwords, slogans, and similar ideas held in common. It is in the acceptance of such norms, and in the appraisal of his own status by them, that the individual experiences himself as part of the group. His conception of himself is drawn ill) in terms of the group's standards. Its sense of personal identity does not and cannot exist independently of the group setting. In short, he individual cannot be considered apart from, or in contrast to, the groups of which he is a part. So the old but still popular dichotomy of individual c versus group simply

( 120) evaporates into thin air. The individual gains his identity through joining groups and accepting their values and norms as his own.


Now we can proceed to a brief consideration of intergroup relations against this background of methodological remarks and of the essential characteristics of small groups. But we must not be too hasty in extending generalizations about in-groups to apply to relations between groups. Most psychological studies to date have dealt with relations within the group and have paid relatively little attention to the intergroup level. Many questions of relations simply cannot be answered unless in-groups are studied in actual contact with other in-groups.

For example, in a well-known study of the effects of different sorts of adult leadership on child interaction, Lewin and his students sought to compare groups led by democratic and autocratic techniques. While this study did reveal certain differential effects of such leadership, its implications are limited by the fact that functional relations between the groups were not brought into the picture. It is of comparatively little value to scientists in this field, or to practitioners either, to compare democratic and autocratic group structures as such if the special character of those structures has no influence either on behavior toward out-groups, or on individual behavior outside the group. As a matter of fact, during Lewin's experiment, a clash of interests occurred between two sets of boys. But even the members of the democratic group did not stop to think of a democratic way of settling the affair; this was something to be practiced within the in-group; the fight that developed was hardly a democratic method of solution. Indeed, attitudes that hold good within a group need not apply to other groups at all. For instance, an aristocratic group may profess and even practice democratic procedures within the bounds of its exclusive set, while at the same time behaving in a supercilious manner toward outside groups.

Groups do not function as closed systems. The external world and other. groups constantly impinge on them. When individuals come to a new group situation or atmosphere, they come with

( 121) certain attitudes formed earlier in other group situations. In the Lewin study just mentioned, one boy actually preferred the autocratic group to the democratic group. This boy was the son of an army officer and had been brought up to admire authoritarian methods. The personal commitment of this boy to a military family background was more decisive, in his development, than the pressures and demands of the experimental democratic situation.

The factors that really count in shaping our attitudes and regulating our behavior—at times against great odds—are those which are organically related to our deepest commitments. If our identifications are unfortunately with autocratic institutions, symbols, and idols, our attitudes and behavior will be regulated accordingly. The available reports on the records of German and Japanese soldiers in action seem to indicate that these men participated rather actively and eagerly in the service of their leaders' dreams of world conquest; otherwise, they would not have been able to fight the way they did. Even current newspaper reports from Germany and Japan are somewhat disillusioning in their revelation of persisting strong identification with blind and self-glorifying nationalism and vested interests. As Lapierre and Farnsworth point out, in their Social Psychology which was published just a few months ago, Lewin's experiments were carried out with American boys who were already steeped in the American democratic process. The results might have been different, they insist, if they had been carried out with German and Japanese boys.

These points lead me to conclude, first, that it is erroneous to draw conclusions about intergroup relations on the basis of results obtained solely from a study of in-group relations. And second, I believe it is misleading to make generalizations on the basis of a single group situation if the individual members have connections with other groups as well. Especially in the highly differentiated societies of the West, individuals are brought up in many different groups which may have contradictory or conflicting values. As they grow older, Westerners usually participate in still other groups. Of course, these various memberships are not insulated from each other. Indeed, it Is in my opinion these contradictory group memberships that are largely responsible for the rather difficult plight and confusion of many human beings today. It therefore becomes

(121) imperative to consider the consequence of in-groups in connection with other groups. This sort of approach will give us a more promising lead for handling intergroup problems, such as tension and group prejudice.

As has already been emphasized, one of the products of group formation is a delineation of "we" and "they," the "we" to include the members of the in-group. The "we" thus delineated comes to embody a whole set of qualities and values to be upheld, defended, and cherished dearly. Offenses from without or deviations from within are promptly reacted to with appropriate corrective, defensive or, at times, offensive measures. To all individuals and groups comprising the "they" are attributed a set of values or traits. These attributed traits may be favorable, unfavorable, or both, depending upon the nature of the relations between the groups in question. If the interests, directions, or goals of the intergroup contacts in question are harmonious, the "they" group is pictured in a positive or favorable light. But, if the activities and functional views of the two interacting groups clash, then the characteristics attributed to "them" are negative and derogatory. If one group takes the position that another group is in its way for any reason, interferes with its goals or interests, or should be working in its interests, all sorts of stereotypes are developed to justify this position. All doctrines of race superiority are deliberate or unconscious justifications of such a point of view.

Since in-groups are often related to many other groups, a set of attributes and characteristics is standardized for each of them on a scale ranging from positive to negative. An important feature of this scale is the social distance at which various groups are placed from the "we" group, that is, the degree of intimacy or closeness allowed with a particular group in terms of the functional relation between them. Such a scale of social distances exists in most countries of the world today. Since there are many studies of social distance or prejudice in the United States the main features of such scales can be summarized from the American findings.

It has been well established that the notion of social distance is not limit lip primarily on the basis of factual experience or knowledge, but is handed down by old members to children or new members of the group, chiefly in the form of short-cut dic-

( 123) -tums. Once the scale is learned and accepted as a basis for defining the individual's position in relation to others, increased contact with members of groups lower in the scale, information about them, or educational attempts to alter the prejudice., simply have no power to affect it. In America, at least, the scale of social distance or prejudice is nationwide for most people in the general run of life, as can be seen from the findings of the social distance studies made by Professor Bogardus. He got essentially the same results in 1926 and again twenty years later in 1946, despite the war, the Four Freedoms, and everything else.

The scale cuts across class boundaries, social demarcations, occupational groupings, and regional lines. In the United States, it can be said that a very similar scale of prejudice exists among rich and poor, in the North and in the South. As the studies of Murphy and Likert and of Horowitz revealed, only those individuals and groups who have dissented in some degree with the larger American setting are free of this prejudice scale. Of course, this does not mean that all individuals or groups in America are equally responsible for the maintenance and perpetuation of prejudice. The problem involves not merely psychological factors, but also economic factors and power relationships which are beyond the scope of this paper.

Since attitudes of social distance are formed as the individual accepts the norms of his group, it is entirely reasonable to suppose that such attitudes can be altered only by the acceptance of membership and a functioning role in a new group. And this is exactly the case, as several studies have shown. The problem then becomes one of creating group situations which will carry over into daily life. For one can hardly expect that attitudes so long and so firmly established can be erased by temporary exposure to new situations. It is this problem that practitioners must face. As social psychologists, we can only say that the attitudes and loyalties generated in the new group situation must carry an emotional and motivational weight: greater than those carried over from the old, more persistent groups. Only by creating new loyalties with strong emotional appeal in the day-to-day activities of the individual, as well as in the group situation itself, can the old attitudes be wiped away and replaced.

( 124)


I should like to conclude with a brief advance report on an experiment still in progress.[1] It may help to clarify the points I have been trying to make. To the best of my knowledge, this experiment represents the first attempt to study both in-group and intergroup relations in one design. It is being carried out under my direction and with the active and generous backing of Professor Carl Hovland, chairman of the psychology department at Yale. In choosing our subjects we tried to get as homogeneous a group as possible. We selected twenty-four boys who were not already friends with one another but who came from the same sort of socio-economic background. They were all Protestants, in fact with few exceptions they were all Episcopalians. They were eleven and twelve years of age, and normal both in personality adjustment and in intelligence. They could not have had any prejudice against one another on socio-economic, racial, or religious grounds. Next we agreed upon a physically isolated camp site in northern Connecticut and took the boys with us. They had been tested in the psychology department at Yale ahead of time, so we had a fairly adequate account of their backgrounds, intelligence, and personalities to begin with. The experiment at camp was carried out in three stages. I shall briefly describe each and give you a rough picture of the general results.

Our first stage was that of spontaneous grouping. Since we had kept all social factors constant we wanted to see what sort of friendships, likes, and dislikes would come to the fore naturally. So we didn't interfere at this point. We put all the boys together in one bunkhouse, contrary to accepted practice at camps, and avoided any scheme of classification for their sleeping, eating, hikes, or for their headquarters on the playground. During the three days of this stage, the youngsters could choose their own beds, their own seats at table, and their own playmates on teams. Then we administered popularity ratings and made sociograms

( 125) of the groupings thus formed according to spontaneous preference and interest.

For the second stage we divided the boys into two groups of twelve matched as far as possible in ability, personality ratings, and the evidence of the sociograms. For example, if we found four boys naturally hanging together, tine put two of them in each group to see if they would get absorbed into the new social unit. Then we introduced activities that required cooperation and interdependence from each member: if one boy failed to cooperate the whole activity was likely to fail. The result was the development of an in-group feeling in each set, the emergence of recognized leaders, and the formation of a status hierarchy—especially at the top levels. The strength of this in-group feeling was different in the two sets for reasons that I will not go into now. At the end of this stage another popularity test was given. The resulting sociograms or friendship preferences were predominantly in favor of in-group members and showed decided shifts away from the clusters originally developed during the first stage.

The third stage was that of intergroup relations—and it was the main part of the experiment. The two groups of twelve were brought into functional relation through highly competitive team games and through situations in which one group was mildly frustrating to the other. We did this with a view to studying. step by step, the mechanics of group prejudice. We can draw certain conclusions from the data so far available. First, it appears that when two groups compete and frustrate one another, the in-group feeling—the sense of solidarity and loyalty—is consolidated and even intensified. In-group cooperation is thus enhanced; there develops a greater tendency to depend on each other, to work for common goals, and to meet situations on an increasingly democratic basis. At the same time, there was a corresponding increase in hostility toward the out-group. This illustrates a point I was making earlier: democracy and cooperation within a group need not mean democracy and cooperation with outsiders.

I cannot help mentioning one fascinating instance of the sort of prejudice that developed among our boys. One youngster was running to get some water to drink when one of the counselors—not the counselor for the boy's own group but for the other one—

(126) advised him not to since he was very tired and perspiring. But one of the boys from the counselor's side called out: "He is not from our group; don't advise him—let anything happen to him!" This story illustrates the danger of jumping to conclusions about attitudes on the basis simply of observation of in-group activities.

Annotated by Muzafer Sherif

H. Cantril. The Psychology of Social Movements (New York: John Wiley, 1941).

Second part describes the rise of such social movements as the "kingdom of Father Divine," the Oxford movement, and Nazism. Emphasizes the emergence of such movements from chaos, insecurity, and deprivation, the false haven they temporarily offer, and their collapse in disillusionment on contact with the world of reality.

Kurt Lewin. Resolving Social Conflicts (New York: Harper & Brothers, 1498).

Chapter 9 describes the security derived from stable group memberships and the corresponding insecurity that accompanies the instability of group ties. Chapter 4 states that attitudes in intergroup relations cannot he changed by piecemeal methods, but only by a structural change in group identification.

Gardner Murphy. Personality, a Biosocial Approach (New York: Harper & Brothers, 1947).

Chapter 31 contains one of the best accounts of the differences in human behavior due to individual and to social factors. Chapters 32-37 provide an excellent synthesis of modern knowledge about the relation between individuals and groups.

T. M. Newcomb. Personality and Social Change (New York: Dryden Press, 1943).

Detailed account of a highly successful study on attitude change with emphasis on the role played by identification with a new group, as well as on the role of personal factors.

Muzafer Sherif. An Outline of Social Psychology (New York: Harper & Brothers, 1948).

Chapter 5 summarizes the main features of newly arising in-groups. Chapter 6 stresses the necessity of going beyond the properties of in-groups if study of intergroup relations is to be realistic. Chapter 14 reviews research findings on social distance and draws certain socio-psychological conclusions about the mechanism of prejudice.

Muzafer Sherif and H. Cantril. The Psychology of Ego-involvements, Social Attitudes and Identifications (New York: John Wiley, 1916).

Chapter 10 summarizes research on small in-groups with special emphasis on the influence on individual behavior of group memberships of various kinds.

W. H. Whyte. Street Corner Society (Chicago: University of Chicago Press, 1943).

This sociological field study of spontaneously formed in-groups is cull of implications and explicit directions for the concrete understanding of intergroup and intragroup relations.


A fuller account of this experiment (financed by a grant to Yale University from the American Jewish Committee) is given in the author’s chapter entitled “An Experimental Study of Intergroup Relations” in Social Psychology at the Crossroads, to be published early in 1951 by Harper & Brothers.

Valid HTML 4.01 Strict Valid CSS2