On the Relevance of Social Psychology[1]

Muzafer Sherif[2]
Department of Sociology, Pennsylvania State University

MY title, "On the Relevance of Social Psychology," seems appropriate for an APA convention organized around the theme "Psychology and the Problems of Society." In choosing the topic, I took my signals from today's youth, who are raising serious questions and challenges about the relevance of things as they are. The salience of the signals was heightened because our experimental and field research on group formation and intergroup relations led us to concentrate on the rise of social movements.

The relationship between intergroup relations and social movements should be obvious. Intergroup relations involve either friendly encounters or hostile confrontations between identified members of two or more groups. The development of a social movement always implies, sooner or later, the confrontation of partisans committed to conflicting positions.

Typically, on one side of the confrontation, there are partisans of an established orthodoxy-stabilized even to the point of rigidity. Such partisans represent the classic problems of conformity to an established group: They are satisfied, even complacent, with the practices, human arrangements, values, or norms of the Establishment. These set the "natural" boundaries for what is to be expected, what is acceptable, respectable, prestigious, and even desirable. What is relevant to them is defined within such circumscribed boundaries.

On the other side, there are partisans to a new social movement, questioning and challenging the Establishment. The taproot of their partisanship flourishes in unrest, dissatisfaction, and frustration with things as they are. When we inquire how the unrest nourishes a social movement seeking change, the relationship between the rise of a new social movement and problems of norm formation, of group formation, and of intergroup relations becomes evident.

The term "relevance" is a nebulous one. Perhaps the only satisfactory definition of the word today lies in the function it serves. Of course, such a characterization is primitive, similar to the definition by usage acceptable for young children on a standard intelligence test, for example, defining the word "ball" as "it bounces." Raising the question of relevance reveals dissatisfaction, disaffection, and, if you will, alienation. Raising the question of relevance serves the function of questioning the prevailing state of things and signifies a groping toward change.


A social psychology that is relevant must do much more than conduct research on significant social problems after they have already become urgent business to administrators, policy makers, and a general public alarmed by them. Yet, until recent years, concentration on problems of social change trailed behind the study of compliance, conformity, interpersonal relations, and personal consistency in relatively stable social systems.

The neglect of persistent problems of social change is not peculiar to social psychology. The social sciences in general seem to suffer from similar neglect. Coleman (1969) made the point in a recent symposium devoted to the current state of social science in the following words: "The current neglect [of social change] leads one to suspect that the whole discipline . . . has evolved toward the study of social statics, and becomes impotent in the face of change . . . social change, social movements, conflict, and collective behavior . . . are the underdeveloped areas of social research. They are not only backward at present; they are not catching up [p. 112]."

When one looks at the state of things in social psychology, the question of relevance becomes compelling. On the whole, the thriving activity

(145) and ever increasing volume of research and publication present a picture of inconclusive and contradictory results. For illustration, I shall consider two active areas of research that are closely related to problems of change, namely, small group research and attitude change in response to communication.

In a systematic review of 1,279 research reports with analysis of 250 studies on small groups, McGrath and Altman (1966) reached discouraging conclusions. Observing a "staleness" in the literature, they concluded that "the problem seems to arise because research in the small group field is so segmented-in the form of idiosyncratic variables, tasks, and measures peculiar to the individual investigators-that no one has a common base from which to argue [p.80] ."

Searching for the cause of this jungle of activity shooting in all directions, McGrath and Altman discarded the charge of inadequate technical competence, for they found techniques "ingenious" and analytical tools often highly sophisticated. Instead, they concluded that the state of affairs reflected the value pattern of researchers, which they called "the entrepreneurial ethic." This ethic stresses "quantity at the expense of quality in research, rigor of method at the expense of creative-theoretical aspects of science, research funds at the expense of research ideas . . . . Together, these values and role changes issuing from the entrepreneurial ethic say to the researcher as creative artist, `Paint what sells. Paint for the highest bidder. Paint as fast and as furiously as you can. Believe in what you paint. Get a good sales force and an efficient home office behind you' [p. 92 ] ."

In research on attitude change in response to communication, the inconclusive and contradictory results have long been noted. Again, it would be misleading to conclude that this state of affairs is primarily a technical problem and not of substance. The contradictory results are not confined to comparisons between experimental and survey research, a contradiction discussed, for example, by Hovland (1959) in this series of APA addresses. They are not found only when comparing research in the laboratory with that in naturalistic settings. Within the laboratory itself, there is remarkably little convergence in findings on attitude change.

For example, in their recent critical analysis of research on attitude change, Kiesler, Collins, and Miller (1969) compiled a list of experimental studies on the extent of attitude charge in response to communication presenting positions discrepant in varying degrees from the initial attitudes of those exposed to communication. Some studies reported greater attitude change with greater discrepancy between the position advocated in (communication and the initial positions of recipients, but others reported greater attitude change with moderately discrepant communications and decreased change with larger discrepancies.

As we have noted elsewhere (C. Sherif, M. Sherif, & Nebergall, 1965), such contradictory findings on the same topic become less of a puzzle when the conditions in which each was obtained are brought into the picture. The apparently contradictory findings on attitude-communication discrepancy fall into a meaningful pattern when the social situation is specified. For example, the content of communication makes a difference, particularly with respect to how ego involving it is for the person; as does the structure of the communication; as does the communication situation, including the communicator; as do the clarity and number of response alternatives, etc.

It is not my intent to dispose easily with inconclusive research findings, but to illustrate a problem at the core of the relevance issue in social psychology. This core problem concerns the definition of the stimulus in social psychology, that is, the definition of a social situation.


Any social situation is composed of sets of factors that far exceed the variables deliberately introduced into the situation or deliberately accounted for by a researcher. These sets of factors include the relationships among the persons present. They include the properties of the task, problem, or activity, for example, its familiarity, difficulty, structuredness, etc. They include the physical and social location of interaction, with all its potentially limiting and facilitating features, including social definitions of the use to which they are to be put. Furthermore, all of these sets of factors in a social situation are interrelated. The individual participant's appraisal of others, his judgment of the task or problem, and the way he sizes up the location and its facilities (including the persons in charge of them) significantly affect his behavior in the situation.

(146) Whether in the laboratory or in a natural setting, a research situation is also a social situation, contaminated by all these sets of factors, whether or not the researcher takes note of them. We have become increasingly aware of the unmistakable effects of such interdependent factors in the research situation, many of them not accounted for by the researcher. As a consequence, a research movement developed for more systematic study of what is known as the "social psychology of the psychological experiment [ Riecken, 1962) ." The movement is represented by the research of Orne (Orne, 1962; Orne & Evans, 1965; Orne & Scheibe, 1964), which demonstrated the "demand character" of a research situation, and the studies by Rosenthal (1966) on effects of the researcher's theoretical bias on the behavioral outcome in experiments.

Research in natural settings is not exempt. In the interview, the characteristics and behavior of the interviewer as well as the type of questions have long been known to affect both quantity and quality of data in survey research (Cantril, 1944) and in case study (Pearl, 1965). In tests of personality and attitudes, the individual's expectation that his person is being studied produces "socially desirable" responses (Edwards, 1957) . On tests of intelligence and information, the race and social class of tester and subject affect test scores (Deutsch, 1964; Pettigrew, 1964). In field observation, the consequence of introducing an observer is so striking, even when he is not identified as a researcher, that we referred to the phenomenon as "the observer being observed" (M.. Sherif & C. Sherif, 1964, 1965).

The researcher himself has the power to affect behavior in a research situation profoundly (cf. Milgram, 1965). If he is committed to a theoretical position or model, explicity or implicitly, his commitment is a strong determinant of what is included in the situation, what is measured, and what is made salient. This is to be expected from established facts on psychological selectivity. What falls readily within the bounds of the model is seized upon as relevant. In effect, the researcher stages his own scenario.

This state of affairs is not a new discovery. For example, it was evident in early attempts by different schools at analysis of psychological content, as noted in the following passage by Boring (1942):

Within the school, agreement is facilitated. Wundt's students confirmed the tridimensional theory of feelings; other: did not. Wurzburg never found images; Cornell did. 1:, feeling a sensation or not? Laboratory atmosphere largely determined what would be found in answer to that question, and laboratory atmosphere often extended from parent laboratory to its offspring . . . . There can be no doubt that within the Zeitgeist there are local Geister which determine which theory you shall apply to your experimental findings or even how you shall record your data [pp. 612-6131.


Probably this state of affairs will continue until at least the fundamentals are established as to what has to be relevant for all researchers in a problem area, setting bounds within which their research scenarios shall be staged. Such needed fundamentals cannot be laid simply on tree basis of hypothetical models or logical deductions from speculation.

The fundamentals of a field are its demonstrated achievements that provide paradigms for studying its persistent problems. This was the main theme in Kuhn's (1962) book The Structure of Scientific Revolutions, as stated in the following words:

Somehow, the practice of astronomy, physics, chemistry, or biology normally fails to evoke the controversies over fundamentals that today often seem endemic among, say, psychologists or sociologists. Attempting to discover the source of that difference led me to recognize the role in scientific research of what I have since called "paradigms." These I take to be universally recognized scientific achievements that for a time provide model problems and :solutions to a community of practitioners LP. x].

Conversely, Kuhn observed that "In the absence of a paradigm or some candidate for paradigm, all of the facts that could possibly pertain to the development of a given science are likely to seem equally relevant [p. 15 ] ."


The fragmented findings in many problem areas of social psychology are in large part a result of selective choice of variables considered to be equally relevant. Yet, the definition of what is relevant in a social situation is not solely a social-psychological problem. There are components in a social situation, even that considered strictly interpersonal such as interaction within a family or among friends, whose study requires more than the social psychologist's wares. Such components include institutionalized power differentials among partici-

(147) -pants, role expectations, culturally defined systems of exchange, technological facilities, value orientations, and normative regulation of the interaction process.

In other words, any social situation includes components that are subject matters for social scientists -sociologists, political scientists, anthropologists, etc. If we are to grasp the nature and the scope of factors in social situations, we have to borrow from these related social sciences.

I will not be surprised if this plea for interdisciplinary borrowing by social psychology sounds stale and incongruous, in view of recent reverses suffered by some institutionalized interdisciplinary programs. Some have epitomized their scornful view in the conclusion that "cross-disciplinary fertilization ends in cross-sterilization."

A number of studies can easily be chosen to illustrate the vantage point gained when the researcher considers the problems, viewpoints, and findings of neighboring disciplines. I shall confine myself to a single example chosen in part because it is a laboratory experiment, therefore a less likely candidate to benefit from interdisciplinary borrowing. The experiment was on the formation of social norms, using the autokinetic set-up (M. Sherif, 1936).

The problem of the norm-formation experiment did not arise from the laboratory or laboratory research. It was suggested directly by sociological accounts of norm formation. Specifically, Emile Durkheim contributed the guiding idea that new norms arise during interaction in fluid, uncertain, or out-of-the-ordinary conditions. This generalization was also supported by sociologists like Frederic Thrasher and Clifford Shaw, who reported that distinctive group codes emerge during interaction among youth in the unstable and fluid conditions of interstitial urban areas. This formulation of the problem defined the conditions needed to study norm formation. The only remaining difficulty was to devise an experimental set-up that embodied these conditions, after which the conduct of the experiment itself was a relatively easy matter.


To an important extent, the locus of change lies in the interaction of people with people. The greatest and most lasting changes in institutions, as well as in human relations and attitudes, are products of interaction in social movements.

A social movement does not appear full blown, but develops from widespread unrest in formative phases through a broad range of events over time and, cross-sectionally, in geographically scattered locations at a given time. From initial discontent, fluidity, and uncertainty, a social movement takes shape as groups form within its fold, groups that are born and inevitably shaped b), intergroup confrontations between the establishment of the time and partisans seeking change. Therefore, our understanding of social movements requires a basic grasp of the process of group formation and intergroup relations. We shall consider problems of attitude change within this context.


One reason why group research in social psychology is fragmented and shooting tin all directions with little coherence is that the essential properties defining the domain of groups are neglected.

We have learned from experimental studies of group formation (M. Sherif, 1966 ; M. Sherif & C. Sherif, 1953) and from research on groups in natural settings (M. Sherif & C. Sherif, 1964, 1965) that human groups have properties with significant impact in forming and changing attitudes. Whatever other properties a human group may have, invariably any group has (a) a structure or organization of roles and power differentials among members, and (b) a set of values or norms defining what is conformity and what is deviation, over and above what members may share with other groups in society.

Human groups tend toward exclusiveness. Whether directed toward socially desirable or socially objectionable activities, they tend to regard their give-and-take a private affair. For any group, there is a threshold at which the entrance of an outsider, including a researcher, is an intrusion. This fact raises a serious methodological dilemma.

When the researcher barges in with his paraphernalia to study a group, he has already changed significantly what he set out to study, even to the point of mutilating it. Therefore, if a researcher is interested in studying groups as they form and interact toward their objectives through their own designs, he has no choice but to study the processes without putting members on their tees to the fact that they are being studied. He can do so by using unobtrusive methods of observation and indirect measures for assessing attitudes, as we have been

(148) for over 20 years. In this regard, an ethical concern is raised.

If the study of group processes without member awareness is sponsored by any agency with power, directly or indirectly, over the group and its members, or by a researcher interested in serving such a client, it represents an unwarranted invasion of privacy-in short, it is spying. On the other hand, in basic research, the identity of particular members and the particular location is not an essential issue. Are the indirect or unobtrusive methods of research, which do not put the people studied on guard, to be used only in studying trivial topics and not in studying such vital problems as conformity-deviation, conflict-cooperation, or obedience-autonomy?

Let me close this issue by suggesting that the great concern over human dignity and privacy ran be served best if basic research discovers the exact conditions and processes underlying the manipulation of people toward deeds of blind obedience, unthinking conformity, hatred, and inhumanity to man. Denial of such scientific study amounts to reserving these problems as the exclusive domain of men of power, bent on using people for their own ends regardless of the cost in human dignity and even human life.

The essentials of group formation can be stated here only in bold strokes without shadings. Individuals with common motives or schemes tend to enter into recurrent interactions. In the course of their repeated give-and-take, the behavior of the various members in dealing with one another becomes increasingly predictable from one situation, one activity, and one time to the next, indicating the formation of directive attitudes toward what to expect of one another and toward relevant activities and objects of consequence to the group. Their "groupness" is proportional to the degree of stabilization in these respects.

A significant aspect of the roles emerging in the interpersonal relations of participants is the power dimension. The person's rank in the power structure, which necessarily becomes hierarchical, defines his status in the group, as measured by the relative effectiveness of the initiative he displays (a) in controlling interaction, activities, and decision making, and (b) in applying sanctions in case of nonparticipation and noncompliance.

As status-role differentiation proceeds, a set of norms arises and becomes stabilized. A social norm is defined as an evaluative scale designating an acceptable latitude and an objectionable latitude for behavior, activity, beliefs, or any other object of concern to the group. Especially in analyzing the rise of new groups, as in the context of a developing social movement, one finds that the process of norm formation is simultaneously the formation of directive attitudes by members, proportional to the importance of the group in the member's life.

The latitude of acceptable behavior for members is not the same in all matters. When a norm pertains to the most vital goals, loyalty to the group or its continued existence, the acceptable range is most restricted and sharply defined, particularly for members who achieve high status.. In such matters, compliance is expected as a matter of course; even slight deviations are reacted to promptly and strenuously. Therefore, in matters of vital importance, such as war and peace, a simple majority consensus is not tolerable for long, even when the majority has the power to enforce its views. Very likely, this fact of group dynamics is reflected in the requirement of a two-thirds majority vote by some deliberative bodies on major issues.

The study of behavior in groups in relation to the properties of groups will proceed more advantageously if we utilize established principles of behavior, when available, instead of improvising new concepts and models at the drop of a hat. My example concerns leader-follower relations.

In the formation of a group, the top and bottom positions in the status hierarchy are the fist to stabilize (M. Sherif & C. Sherif, 1964). The top status is the leadership position. It follows that the leader becomes a leader relative to other members, the activities engaged in, and the emerging norms, and not so much through personal characteristics or styles that he may display apart from them.

Similarly, when an observer assesses the differentiated positions of any group, it is the top and bottom positions that first become salient. This salience of the top and bottom positions in the group reveals the operation of a general principle of judgment, namely, the salience of end points or end anchors in a series of graded stimuli, which is indicated by lower variability or error in judgment (Volkmann, 1951).


Figure 1 gives the average variability in judgments of member status by several observers, each rating a different natural group. The base line shows the average rank received by a member during the entire study period. The curves indicate the locations of mean variability for successive blocks of observation periods (10 periods of observation for each block). Note that early in the study (Time a) all judgments are more variable than later in the study period (Time n). End anchoring is revealed in each curve by the lower mean variability in the observers' judgments at the highest and lowest status positions.

As in judgment of neutral stimuli, the observers' confidence in their judgments of member status is greater at the highest and lowest statuses than for intermediate positions. As Figure 2 shows, observers' ratings of their confidence increase over time (Time a to Time n), but are consistently highest at the extreme status positions. These findings were confirmed this summer in ratings by three observers, making intensive observations of 38 groups as they formed during a summer camp. In this research, the rating dimension was the relative standing of the groups themselves, rather than the status of the members within the groups.

Additional research indicates that end anchoring operates also in members' judgments of each others' performance (Koslin, Haarlow, Karlins, & Pargament, 1968) and in their sociometric choices according to the effective initiative displayed by fellow members (M. Sherif & C. Sherif, 1969).

The implications are twofold: First, an important theoretical link is established between judgment of ordered, neutral stimuli and of social structure. Second, interaction over time invariably becomes differentiated in terms of relative power, even in informal groups in which members disclaim any status differentials and even when researchers fail to study the process.


Ingroup formation does not typically occur in isolation. The process is speeded in the context of intergroup relations. Inevitably, the formation of a group in a contest of other groups implies comparison among the groups by the members. In our research on group formation conducted in summer 1969, such comparisons were assessed indirectly through asking members of seven groups to estimate how long it would take each group to complete a novel activity. Each person first made the estimates individually, then reached a consensus with other members of his own group on a group estimate for each group. Figure 3 gives the means of individual estimates and of group estimates for the time expected for each group's performance by members of the other groups. The groups are arranged along the base line on the basis of the observer's ratings of the groups' relative prestige and solidarity after observing them daily for the three previous weeks. The group estimates (solid circles) reveal definite upper, middle, and lower status levels, in line with the observer's ratings, even though neither bore much resemblance to subsequent performance in the event. Similarly,

(150) the individual estimates (open circle) reflect comparative evaluation of the groups, anchored firmly at the upper and lower levels.

Not all group comparison, are invidious, yet most social-psychological theories deal exclusively with negative prejudices and stereotypes, accompanied by doctrinal ideologies of superiority that justify for members their hatred and outright aggression toward the despised group. As a result, many of them have focused on factors that are neither necessary nor sufficient for invidious comparison to occur. For example, a glance at history suggests that groups practicing democratic and egalitarian relations within their own fold are no more immune to wars and racial suppression than those with authoritarian patterns (AY. Sherif, 1966). Similarly, in examining evidence about the personalities involved in historic conflicts, one finds nearly a complete range of individual patterns, as the historian Brinton (1965) observed.

Our experiments on group formation and intergroup relations in naturalistic settings were specifically designed to preclude explanation of the results on the basis of undue frustrations or neurotic tendencies of individuals, of the authoritarian nature of relations within groups, or of immutable and innate instinct (M. Sherif, Harvey, White, Hood, & Sherif, 1961; M. Sherif & C. Sherif, 1953; At. Sherif, White, & Harvey, 1955; cf. M. Sherif & C. Sherif, 1969, chap. 11). Experimental units were composed of normal, well-adjusted, and stable individuals, matched as nearly as possible for the range of individual differences. Groups were formed by creating conditions that promoted cooperation and grassroots participation within the fold. Both the formation of intergroup hostility and its change were studied.

In brief, the guiding theory was that, once in-groups formed, the sufficient condition for conflict, mutually negative attitudes, and stereotyped image is prolonged interaction between groups toward goals that only one group can attain and only then at the expense of the other. The experiments confirmed the prediction, revealing through a combination of unobtrusive research methods the rise of hostile attitudes, as well as aggressive acts of vandalism, violent confrontation, preparations for deterrence of possible future aggression, and self-glorification justifying the deeds within their respective folds.

The dynamics of intergroup confrontations are relevant to events during the rise of social movements defying the status quo. In the experiments, as intergroup conflict grew, its impact on the respective ingroups was undeniable. Ingroup solidarity increased, as members closed ranks. Leader-follower relations that had stabilized before the conflict changed to elevate those most skilled in conflict and to demote the weak apt heart. Preparation for future conflicts preoccupied members almost to the exclusion of erstwhile peaceful pursuits.

Jumping ahead for a moment, let me add that once such new roles, norms, and activities stabilized within the conflicting groups, they almost had a life of their own, extending well beyond the experimental period in which actual conflict prevailed. Once recognized as performing a service useful to the group in conflict, members who thus benefited attempted to continue the same function, even when it was neither needed nor desired by the rank and file. Fortunately, there were no opportunities in our experiments, as there are in actual life, for these eager warriors to establish a self-perpetuating bureaucracy with control over the group resources.

The stabilization of hostile group images and negative attitudes by members became a heavy hand shaping each new interaction episode with the outgroup. Encounters between groups invariably turned into a vicious circle of mutual recrimination, placing blame squarely on the adversary.

(151) Contiguity of the groups in separate but equal conditions for extrinsic reinforcement had no effect in reducing hostility, but served as an opportunity for renewed aggression. Renewed aggression did not produce catharsis, but intensified the aggression. Communication from neutral outsiders conveying accurate information about the other group was scorned as inaccurate and misleading. The attempt to negotiate differences was spurned by the adversaries and repudiated by fellow members.

In short, the piecemeal measures so widely studied in research on attitude change were simply irrelevant in the dynamics of conflict between groups whose members' directive attitudes toward the other group were firmly anchored within their own groups. Nevertheless, attitudes and behavior were changed experimentally.

In line with the theory of conflict, the conditions necessary for change proved to be a series of superordinate goals, that is, a succession of goals whose attainment was urgently and intensely desired by partisans on both sides of the conflict, but that could be attained only by their cooperation in joint endeavors as equals.

I should emphasize here that superordinate goals are not gimmicks to be added to the kit of the social technician for partisans on one side or the other to pull out as an antidote to conflict when the situation gets hot. Nor are they merely any goals that the groups have in common.

Superordinate goals are high in the scheme of concerns of both groups, even to the extent that they are more compelling than their provincial concerns within each group. Thus, superordinate goals establish conditions for genuine participation and cooperation across group lines that are conducive to fruitful negotiations, exchange of persons, unbiased reception of communication, and attitude change on both sides, not on one side alone.


In grappling with problems of attitudes and their change nearly 25 years ago, our point of departure was established findings in the experimental psychology of judgment. Since then, in collaboration with colleagues, a research program was conducted on social judgment and attitudes, with particular emphasis on personal involvement in the values of one's reference groups (C. Sherif & M. Sherif, 1967; C. Sherif, M. Sherif, & Nebergall, 1965; M. Sherif & Hovland, 1961; M. Sherif & C. Sherif, 1969).

The social judgment-involvement formulation led to a redefinition of the attitude concept, to several research procedures to assess attitudes and attitude change, and to definite predictions about attitude change. Attitude was defined as the individual's set of categories for evaluating a domain of social objects, persons, groups, ideas, etc., that he establishes as he learns about that domain and that relate him to subsets within the domain with varying degrees of positive and negative affect.

The structure of a person's attitude was partitioned into three ranges or latitudes, defined as follows:

    Latitude of acceptance: positions acceptable to him.

Latitude of rejection: positions objectionable to him.

Latitude of noncommitment: positions that he does not choose to evaluate one way or another when not forced to do so.

These latitudes provided a key to the discrepancy problem, that is, the discrepancy between the person's position on a topic and the position advocated in a communication. The general problem is illustrated in Figure 4. As, the diagram shows, Persons A, B, and C find the same position most acceptable (triangle), but their latitudes of acceptance for other positions differ. Obviously,

(152) the position of the communication, indicated by the arrow, is not equally discrepant for the three persons. The discrepancy of a communication is represented more adequately relative to the limit of the person's latitude of acceptance.

Further research revealed significant relationships between the relative sizes of the latitudes of acceptance, rejection, and noncommitment and the degree of the person's ego involvement in his attitude. The social judgment-involvement formulation unequivocally proposes the following operational definition for measurement of the degree of ego involvement: The greater the size of the latitude of rejection and the smaller the size of the latitude of noncommitment (approaching zero), the greater is the degree of the person's ego involvement in an issue. Correspondingly, the greater is his resistance to changing his attitude.

Research recently completed by Sarup (1969 ) using the Own Categories Method illustrates the relationships. The task was to categorize 50 statements in terms of how favorable or unfavorable the statements were to the respondent's reference group. Subjects could use any number of categories they chose and distribute the 50 statements in any way appropriate for this task. Criterion samples known by observation to be either highly active or inactive in their reference groups were compared. On the average, the more highly involved and active sample used only 3 categories in sorting the 50 statements, while the inactive and less ego-involved sample used 4.6 categories (p < .001) . Figure 5 compares the two samples according to the mean number of statements.., out of 50, placed in the latitudes of acceptance, rejection, and noncommitment. Confirming earlier findings, the figure shows that the latitude of rejection is greater and the latitude of noncommitment is smaller when ego involvement is high. Sarup's further research indicates that the Own Categories Method provides a reliable measure of the degree of involvement that has the advantage of being obtained indirectly or unobtrusively.

The cumulative findings on latitudes of acceptance, rejection, and noncommitment reveal what political scientists and practical politicians have long known, namely, that people are frequently much clearer about what they are against than what they are for. They suggest a model for predicting the extent of attitude change toward or away from a communication, and for predicting no change.

Figure 6 is a diagram of predictions for three illustrative cases when other conditions are held constant. The ordinate represents extent of at-

(153) -titude change, with the horizontal line indicating no change, the curve above the line indicating change toward communication, and that below the line indicating change away from communication—a boomerang or contrast effect.

The base line represents discrepancies between the position advocated in communication and the position most acceptable to the person. The size of the discrepancy increases along the base line. Below it, three cases are diagrammed with different latitudes. Case A has a broad latitude of rejection and small noncommitment. The extent and direction of predicted changes at different discrepancies for Case A appear as the lowest curve on the figure. The most change over wider discrepancies is predicted for Case C, whose latitudes of acceptance, noncommitment, and rejection are about equal in size, while Case B is intermediate. In each case, however, as the communication falls within the latitude of rejection, the curve inflects toward less change, reaching the no change point, and finally changing away from the communication (contrast effect).

This diagram has some bearing on the communication process and its effects upon a rising social movement; a problem to which I now turn.


Social movements serve as the clearest indicators of social problems, and as the best indicators of the simmering directions of social change, if not always as the vehicle of change itself.

Every social movement, small or large, engages in collective behavior, such as demonstrations, protest meetings, boycotts, and even riots. Three points have to be made about collective behavior.

1. The collective behavior aspect of a social movement may be heroic as well as beastly. The authors partisan to social movements made salient the heroic; the reactionary authors, the beastly.

2. The frequent label for collective flare-ups has been in terms of sickness or pathology. The labels are misleading. Once a medical diagnosis is made in medical terms on social problems of such enormity, it erroneously appears that the solution is to call upon social "doctors" for a remedy with a relatively cheap price tag (say $5-10 billion).

3. Flare-ups of collective behavior (violent or nonviolent) are inseparable parts of a social movement, but only episodes and not everything in the multifaceted pattern.

What follows is a highly condensed account of a formulation on the rise of social movements developed over the last five years (M. Sherif & C. Sherif, 1969, chaps. 23-24).

There is an overlap between the rise of a social movement and the formation of small groups, especially in the initial stage. The point of overlap is the human tendency, demonstrated time and again, to move toward a new social formation when prevailing conditions prove to be exasperating, and change for the better is not in sight. But in social movements, we have to cope with large numbers of people, the geographic spread of events, and the characteristics of the status quo or the Establishment that the movement will confront sooner or later. The large number of individuals spread over wide stretches of geography raises problems of coordination.

A social movement makes sense only when considered as a developing formation over time, with ups and downs, through the phases off factionalism and angry protestation to the phase- of convergence on a more inclusive leadership and affirmative platform. In studying a social movement, or collective behavior episodes within it, it becomes necessary, therefore, to specify the point in time and formative stage in its natural history. The timing of the phases in the developing pattern is not a matter of a year or two or the, interval between elections.

Needless to say, what I have called phases in a social movement are not mutually exclusive in time. They are retarded or accelerated by surrounding circumstances, such as prosperity, peace, and solidarity within the system, on the one hand, or by war, depression, and within-system divisions, on the other. In fact, they are inevitably affected by the entire scheme of relations within the country and by its place in the world scene.

Motivational Base

A social movement always rises from a motivational base that signifies social problems. The phrase motivational base is used in a generic sense. It may consist of material destitution, such as hunger and miserable living conditions. It may consist of desperation caused by exploitation, racism, and colonialism. The motivational base may arise from the divorce between what is preached "day in and day out" and what is pursued

(154) relentlessly in fact. It may reflect the experience of relative deprivation, that is, a gap in the opportunities and privileges between groups that persists in spite of improvements in the plight of the underprivileged. For example, according to the United States Bureau of the Census (1967), the income gap between whites and nonwhites remained almost constant from 1947 to 1965. Human judgment is ever a comparison-making process. In this case, the comparison group is the dominant group and, therefore, is crucial in determining whether the plight of one's own group (reference group) is at the level it should be.

The motivational base, which is fed by social problems, is a necessary condition for the stirrings toward a social movement, but not a sufficient one. The necessary and sufficient are still separated by the difficult human problems of converging on a bill of gripes, converging on more inclusive leadership, and converging on an affirmative platform of changes along with strategy to chart the guidelines to action.

For the individual, reckoning with the problems of the motivational base by joining a social movement implies change. But tearing one's self away from the moorings of things as they are, even when things are miserable, is a painful process. Shedding the ties and prestige symbols ingrained during a whole lifetime by a large number of people is not a feat to be achieved overnight. The motivational strain, sooner or later, brings about alienation, and with it repudiation of things as they are, or at least of some prevailing institutions and practices. Such alienation temporarily, at least, leaves the individual in a state of normlessness, which is a state of tension not to be endured forever.

When a person is caught in a state of ambiguity, fluidity, and normlessness, the psychological tendency is toward some kind of stabilization, as numerous studies have shown. The end result is not doomed to be chaos. On the contrary, alienation and normlessness are conducive to the search for new alternatives. These may include switching to a new reference group, joining an ongoing movement, or throwing one's lot in with thousands of others toward initiating a new social movement.

Initial Phase

Even in the early years of a movement, there have to be heroes to be embraced, acceptable ideological formulas and slogans. In spite of this, the most salient characteristic of a new movement in its initial phase is that it is more clear about what it is against, what it rejects, than what it is for. For example, the student demonstrations in Paris in the spring of 1968 were most pronounced in what they were against, rather than in converging on an ideological platform. Probably, at this early phase, it is a mistake to identify such negativism with particular doctrines, such as anarchism, even though a number of anarchists may be involved. A great variety of social movements show this tendency to crystallize first on what they are against, while still divided into factions over specific changes to be instituted and strategies to be adopted.

In the initial years, the vehement expression of a bill of gripes serves as the rallying point for unity and solidarity among participants in a new social movement. Thus, it would be misleading to view the frequent but sporadic flare-ups of collective behavior during this early phase apart from their particular place in the pattern of the movement. No matter how emotional, and even irrational, such collective actions may seem when evaluated as isolated incidents by outsiders, their logic and function become understandable when viewed as part and parcel of a developing social movement. The vandalistic destruction of a valuable cargo of tea in Boston Harbor is an example.

Convergence on Affirmative Platform

In time, the pattern of a developing social movement takes shape through the convergence of the motivational base and an ideological base, namely, the crystallization of both a bill of gripes and a platform of advocated changes and action strategies, formulated by a coordinated leadership that arches over various factions. This convergence requires intellectual leadership to translate the ideological premises into shortcut formulations and slogans that are relevant to ongoing actualities. It implies further differentiation of the roles and statuses within the movement, in short, organizational solidification even in the face of factional disputes that almost invariably develop around platform, strategy, and action.

During the convergence phase, the pattern of a social movement consists of much more than collective action and protest. Much of the activity involves planning, office work, communication, recruitment, and training, during which the char-

155) -acter of collective behavior becomes more deliberately planned, coordinated, and executed.


The path of a social movement is not like an open expressway strewn with rose petals. The moment it voices its protests and rejections in some form of collective action, it is confronted with counterreactions from the Establishment, with coercive and even repressive measures. Countermovements arise within and outside of the Establishment, both to justify and to execute repressive measures. This has been the predicament of almost every social movement of any scope, not excluding the Black Power movement and student unrest today.

The use of repressive force by the Establishment or by countermovements quickly foments the dynamics of intergroup conflict. In this case, the weight of power lies with the Establishment, with the resulting sense of victimization for the movement. On both sides, the latitudes of acceptance shrink further, and latitudes of rejection expand. The chance that communication across partisan lines will have any effect, other than negative, approaches zero. Within the movement itself, the margin for tolerating deviation shrinks, proportionally to the risks that membership has come to entail.

Part of the challenge that social movements present to social psychology is the opportunity to study groups in formation, interpartisan encounters, attitude formation, and change within the context of such events. In the process, some of our cherished notions are likely to be shaken. Take, for example, the prevailing conclusion that little attitude change is found in field studies, while laboratory studies yield the greatest change.

A recent report to the President by several Republican congressmen after a tour of university campuses includes an incident during the People's Park controversy at the University of California (Evans & Novak, 1969). While militant students demonstrated, two of the congressmen talked with moderate students nearby. At that moment, a National Guard helicopter started spraying tear gas over the campus. According to the congressmen, the effect on the nonmilitant students was instantaneous-"they were radicalized that moment." It is not merely a matter of whether research is in the laboratory or in the field that determines whether or not there will be attitude change.

The congressmen's report just cited illustrates another common misconception, namely, that the number of militant members at a particular time is an adequate indicator of a movement's viability and possible impact upon prevailing attitudes. The report noted that, although militant students constitute a relatively small minority, there is a large proportion of moderates who are similarly alienated by the Establishment's rigidity and similarly concerned about human values, and who take radical positions on specific issues, especially when repressive force is used (Sunday Boston Globe, 1969).

The significance of a movement as a force for change does not depend merely on the number of its committed partisans at a particular time, but on how widespread the social problems are that provide the motivational base. As the political scientist Deutsch (1964) observed, an assessment of the independence movement in the American colonies in terms of its active supporters before the Revolution would have been extremely misleading; they constituted no more than a third of the colonial population. What counted was the potential adherents of the movement-those affected by the motivational base, but not yet active. Here the independence movement outstripped those loyal to the King by recruiting eight times the number of adherents.

Pulling together the points made about the rise of a social movement, let me summarize in six steps that also serve to characterize any social movement, including countermovements:

1. A social movement is a formative pattern of attempts toward change that develops in phases over time.

2. It is initiated through interaction among people prompted by a motivational base that is fed by persisting social problems.

3. It is carried out by those directly affected and by others who throw their lot with them.

4. It develops through declaration of gripes and the formulation and proclamation of platform or ideology, which imply organization.

5. It develops for the purpose of bringing about evolutionary or revolutionary changes, or of suppressing changes (countermovements ).

6. Its efforts toward change are effected by means of appeals to the public, slogans, symbolisms, agita-

(156) -tion, episodes of collective action, and encounters with the opposition (strikes, rallies, resistance, boycotts, demonstrations, riots, insurrection, etc.).

A relevant social psychology should be concerned with the study of social movements produced by social problems, for it is these movements that are groping toward the shape of the future.


BORING, E. G. Sensation and perception in the history of experimental psychology. New York: Appleton-Century-Crofts, 1942.

BRINTON, C. The anatomy of revolution. (Rev. ed.) New York: Random House, Vintage Books, 1965.

CANTRIL, H. Gauging public opinion. Princeton, N. J.: Princeton University Press, 1944.

COLEMAN, J. S. Methods of sociology. In R. Bierstedt (Ed.), Design for sociology: Scope, objectives and methods. Philadelphia: American Academy of Political and Social Science, 1969.

DEUTSCH, K. W. External involvement in internal war. In H. Eckstein (Ed.),Internal war: Problems and approaches. (Monograph 9, 86-114) New York: Free Press, 1964.

DEUTSCH, M. Guidelines for testing minority group children. Journal of Social Issues, 1964, 20(Whole No. 2).

EDWARDS, A. L. The social desirability variable in personality assessment and research. New York: Holt, Rinehart & Winston, 1957.

EVANS, R., & NOVAE, R. Campus revolt hearings changing minds in House.Pittsburgh Post-Gazette, May 31, 1969.

HOVLAND, C. I. Reconciling conflicting results derived from experimental and survey studies of attitude change. American Psychologist, 1959, 14, 8-17.

KIESLER, C. A., COLLINS, B. E., & MILLER, N. Attitude change. New York: Wiley, 1969.

KOSLIN, B. L., HAARLOW, R. N., KARLINS, M., & PARGAMENT, R. Predicting group status from members' cognitions. Sociometry, 1968, 31, 64-75.

KUHN, T. S. The structure of scientific revolutions. Chicago: University of Chicago Press, 1962.

MCGRATH, J. E., & ALTMAN, I. Small group research: A synthesis and critique of the field. New York: Holt, Rinehart & Winston, 1966.

MILGRAM, S. Some conditions of obedience and disobedience to authority. In I. D. Steiner & M. Fishbein (Eds.), Current studies in social psychology.New York: Holt, Rinehart & Winston, 1965.

ORNE, M. T. On the social psychology of the psychological experiment with particular reference to demand characteristics and implications. American Psychologist, 1962, 17, 776-783.

ORNE, M. T., & EVANS, F. J. Social control in the psychological experiment: Antisocial behavior and hypnosis. Journal of Personality and Social Psychology, 1965, 1, 189-200.

ORNE, M. T., & SCHEIBE, K. E. The contribution of nondeprivation factors in the production of sensory deprivation effects: The psychology of the "panic button." Journal of Abnormal and Social Psychology, 1964, 68, 3-12.

PEARL, A. Youth in lower class settings. In M. 'Sherif & C. W. Sherif (Eds.), Problems of youth: Transition to adulthood in a changing world. Chicago: Aldine, 1965.

PETTIGREW, T. F. A profile of the Negro American. Princeton, N. J.: Van Nostrand, 1964.

RIECKEN, H. W. A program for research on experiments in social psychology. In N. F. Washburne (Ed.), Decisions, values and groups. New York: Pergamon Press, 1962.

ROSENTHAL, R. Experimenter effects in behavioral research. New York: Appleton-Century-Crofts, 1966.

SARUP, G. Reference groups: Some determinants and consequences. Unpublished doctoral dissertation., Pennsylvania State University, 1969.

SHERIF, C. W., & SHERIF, M. Attitude, ego-involvement and change. New York: 1967. (Chap. 7.)

SHERIF, C. W., SHERIF, M., & NEBERGALL, R. Attitude and attitude change.Philadelphia: Saunders, 1965.

SHERIF, M. The psychology of social norms. New York: Harper, 1936.

SHERIF, M. In common predicament: Social psychology of intergroup conflict and cooperation. Boston: Houghton Mifflin, 1966.

SHERIF, M., HARVEY, O. J., WHITE, B. J., HOOD, W. R., & SHERIF, C. W. Intergroup conflict and cooperation: The Robbers Cave experiment. Norman: Institute of Group Relations, University of Oklahoma, 1961.

SHERIF, M., & HOVLAND, C. 1. Social judgment: Assimilation and contrast effects in communication and attitude change. New Haven: Yale University Press, 1961.

SHERIF, M., & SHERIF, C. W. Groups in harmony and tension. New York: Harper, 1953.

SHERIF, M., & SHERIF, C. W. Reference groups: Exploration into the conformity and deviation of adolescents. New York: Harper & Row, 1964.

SHERIF, M., & SHERIF, C. W. (Eds.) Problems of youth: Transition to adulthood in a changing world. Chicago: Aldine, 1965.

SHERIF, M., & SHERIF, C. W. Social psychology. New York: Harper & Row, 1969.

SHERIF, M., WHITE, B. J., & HARVEY, O. J. Status in experimentally produced groups. American Journal of Sociology, 1955, 60, 370-379.

Sunday Boston Globe. Student majority agree with goals, if not methods of the campus radicals. July 13, 1969.

UNITED STATES BUREAU OF THE CENSUS. Current population reports. December 28, 1967, 4, No. 53.

VOLKMANN, J. Scales of judgment and their implications for social psychology. In J. H. Rohrer & M. Sherif (Eds.), Social psychology at the crossroads. New York: Harper & Row, 1951.


  1. Requests for reprints should be sent to Muzafer Sherif, Department of Sociology, Pennsylvania State University, University Park, Pennsylvania 16802.
  2. The author wishes to thank his wife, Carolyn Sherif, for substantial help in writing this paper. Unpublished research reported here was supported by a grant from the National Science Foundation. D. Schiermeister, T. MacIntyre, and M. Scott participated in the collection of data on individual and group judgments of performance (see Figure 3).




Valid HTML 4.01 Strict Valid CSS2