Motivation and Intergroup Aggression: A Persistent Problem in Levels of Analysis

Muzafer  Sherif and Carolyn W. Sherif

The age-old problem of man's inhumanity to man has been analyzed by philosophers, commentators, moralists, and legal experts as well as social scientists and psychologists. Most accounts have recognized that intense emotions and motivational states are associated with aggressive deeds.

This discussion concerns the levels of analysis required for specifying the essential variables that generate aggressive behavior between members of different human groups. Many psychological accounts of aggression have failed to differentiate between interpersonal aggression, on the one hand, and intergroup aggression, on the other. Conversely, some writers who deal with aggression in intergroup and international affairs have contended that psychological variables are irrelevant to the analysis.

The present thesis is that motivations to aggression toward members of a different human group involve variables at a different level of analysis than those in strictly interpersonal aggression, as well as psychological variables. By recognizing the levels of analysis required, an integrated account of aggressive motives comes within our grasp. In an age when technological tools provide an aggressor with the power to destroy much of mankind and render the environment unsafe for those remaining, a proper grasp of these problems becomes essential.


In accounting for aggression or cooperation in human relations, traditional theorizing fell into two main camps, each with variations on its theme: the instinctivist camp and the environmentalist camp.

( 564)

The instinctivist camp explained the state of human relations in terms of primordial dispositions inherent in human nature. These dispositions, it was suggested, seek releasers that have no function other than to trigger an eruption. Thus, Hobbes depicted a human nature that is selfish, sneaky, and aggressive. Rousseau, disgusted by human relations in the hands of a degenerate and decaying aristocracy, depicted human nature as essentially innocent and good, but distorted by the shackles of the corrupt social order of his age. The environmentalist camp, on the other hand, put the praise or blame for aggression entirely on the cultural, social, political and economic conditions surrounding man, discounting the significance of his motivations and emotions.

If our aim is to understand human motives and ambitions as men transact in group units, we fail utterly if we limit the analysis entirely to environmental conditions. On the other hand, the instinctivist position amounts to leaving out the environmental conditions, hence gives a mutilated picture of the problem.

Until very recently, the most influential psychological theories tenaciously put the blame for destruction and murder committed by human groups on destructive and aggressive dispositions inherent in human nature. For example, one well-known list of innate dispositions was posited by William McDougall. McDougall included in his list the instincts of ascendance and submission, along with instincts of pugnacity and acquisition (McDougall, 1923).

The "instinctive" terms included in such lists varied according to the social philosophy and pessimistic or optimistic outlook of the author himself. The presence or absence of aggressive dispositions in such lists is one index of the conception of human relations held by particular authors. According to its social arrangements, each society has its "myth" about the Nature of Man, as cogently noted by Tolman. Being a peace-loving man in a war-torn world, Tolman included "loyalty to the group" and "sharing" in his own list of social drives (Tolman, 1942).

First, we shall examine the diverse means by which aggression has been sought as a fundamental part of man's nature, starting with the unabashed positing of a "death instinct." The pitfalls of limiting the analysis to psychological variables will be indicated. Finally, the levels of analysis will be specified that are required for adequate understanding of man's inhumanity to those he sets apart invidiously as beyond the pale of his moral standards.

"Death Instinct"

The instinctivist position explains the vindictiveness of group toward group on the basis of blind aggressive forces erupting from the depths of human nature. Perhaps its most influential proponent was Sigmund Freud,

( 565) the towering figure of the psychoanalytic movement. In his earlier work, Freud saw aggression as a response to the frustration of impulses that he then conceived as more basic. In his later writings, however, he posited an "innate, independent, instinctual disposition in man" towards aggression (Freud, 1930,^p. 102). Specifically, as problems of group relations became of greater concern to him, he developed the notion of two classes of instincts: "Eros or the sexual instincts" and a "death instinct, the task of which is to lead organic matter back into the inorganic state" (1927, p. 55).

For Freud, it was Eros that held human beings together in groups, but Eros was easily ravaged by the death instinct's aggressive impulses. Freud was very explicit on this point (1930, pp. 85-86): 

This aggressive cruelty usually lies in wait of some provocation, or else it steps into the service of some other purpose, the aim of which might as well have been achieved by milder measures. . . . It also manifests itself spontaneously and reveals men as savage beasts to whom the thought of sparing their own kind is alien. . . . The existence of this tendency for aggression . . . makes it necessary for culture to institute its high demands. Civilized society is perpetually menaced with disintegration through this primary hostility of men towards one another.

Thus, the ultimate reason for the existence of human culture as well as the greatest threat to that culture were seen in man's innate destructiveness. Society, in Freud's eyes, was an inevitable enemy of the individual. The individual conscience was essentially "dread of society" (1922, p. 10).

For Freud, interaction in groups and collective encounters did not produce creative outcomes. On the contrary, they served only to release man's instinctive impulses. 

From our point of view we need not attribute so much importance to the appearance of new characteristics. For us it would be enough to say that in a group the individual is brought under conditions which allow him to throw off the repressions of his unconscious instincts. The apparently new characteristics which he then displays are in fact the manifestation of his unconscious, in which all that is evil in the human mind is contained as a pre-disposition. (1922, pp. 9-10.)

Freud's contention is, of course, in sharp opposition to results obtained in research on social interaction by social scientists during the last thirty years.

Doctrine and Fact on Aggressive Instincts

All of us, psychologists and laymen alike, seem irresistibly fascinated with musing over ourselves, over our desires, and over the attachments

( 566) and fantasies we had when we were children and adolescents with kaleidoscopic dreams. There is fascination in abandoning the usual restrictions and revealing ourselves freely to a trusted friend, to a beloved person, or in the permissive atmosphere of a confessional or therapeutic session.

The range of conscious awareness is very limited, as many studies have established. The happenings of our past are inseparable parts of our self-identity, whether they are horrid or pleasant, whether the source of our guilt, shame, inner conflict, or gratification. Beset with tedious preoccupations and role conflicts in a modern life that is casually, even contradictorily patterned, we develop a craving to unearth those parts of ourselves hidden by the limited awareness of the living present. This fascination may be partially responsible for the appeal of dramatic accounts of unbridled impulses lying beneath conscious awareness, such as presented in the aggression doctrine under consideration.

The doctrine of all-powerful aggressive impulses inhering in a "death instinct" had its critics even in the heyday of the psychoanalytic movement. both within the fold of the movement and outside it. Today, experimental. developmental and other research evidence on intergroup attitudes and deeds renders the explanation of aggression in human relations on the basis of a death instinct completely untenable.

In his comprehensive survey on aggression, Berkowitz (1962) summarized theories of aggression and critically evaluated them in the light of available research findings from several disciplines. He concluded that research evidence does not support Freud's conception of a "death instinct" whose energy must be released and whose inherent tendency is toward return to the "quiescence of inorganic matter." In fact he noted, very few psychoanalysts today accept the "death instinct." "Dramatic though it may be, the concept of an innate drive for destruction, as Freud posited it. is scientifically unwarranted. . . . There are a number of bases upon which the hypothesis can be attacked, some logical, others factual" (ibid., p. 8).

The Error of Attributing Man's Aggression to Biological Nature

Another lingering misconception is the postulation of a fighting impulse inherent in biological organisms. This misconception seems to arise from the undeniable fact that frequent fights occur among some species of animals. A fighting instinct has been postulated without detailed examination of the causes of these fights.

The entire concept of animal "instinct" is being subjected to searching examination today in the light of recent research (see Schneirla, 1964). Much animal behavior that was called "instinctive" because it occurred with relative invariance among members of a species turns out to involve a complex learning process by organisms developing in particular environ-

( 567) -ments and with species-specific capacities for perception, locomotion, etc.

With respect to a fighting instinct, it is important to note that the occurrence of fighting among animals is not without reference to environmental events. Indeed, Scott (1958) has observed in his survey of studies on aggression that the impetus for fighting is closely related to environmental conditions. Contrasting the sequence of events associated with the physiology of hunger, he concluded that there was "no physiological evidence of any spontaneous stimulation for fighting arising within the body" and that "the chain of causation in every case eventually traces back to the outside" (ibid., p. 62).

As Schneirla has demonstrated in a series of penetrating analyses, the criteria for drawing valid analogies between animal behavior and human behavior cannot rest solely on similarity of outcomes (e.g., injury to another), or complexity of performance (witness the army ant's maneuvers), or degree of adaptation to the environment. The criteria for proper analogies between behavior of different species must be found in the comparability of the processes underlying the behavior (Schneirla, 1946; 1951; 1953). According to the level of evolutionary development and capacities of the species, the variables that affect the animal's behavior differ. Thus, man's dominance-aggression patterns, while superficially comparable to the pecking order of fowls in some conventionalized situations, are by no means governed by the same variables, physiological or environmental (Schneirla, 1953).

Accounts of human aggressiveness must include the process, specific to the human species, by which man learns and responds to external stimulation in accordance with verbally formulated rules or generalizations. Because of this process, man's aggression to man does not vary systematically according to the same conditions as it does in lower species, such as food shortages and ecological arrangements. On the contrary, the most violent acts of human aggression seem to involve individuals and groups with greater food supplies, more space, and greater facilities for adaptation without aggression.

Error of Generalizing from Atypical Populations

Probably one reason that psychoanalysts postulated a self-generated instinct of aggression was that they based their theorizing on a select sample of individuals-disturbed, deranged, at odds with their fellow men. We borrow this inference from a leading researcher on human development, Robert R. Sears, who spent years studying children in their friendly and hostile interpersonal relations. Sears (1960, p. 96) drew attention to the pitfall of making wholesale inferences from data on a select and unrepresentative sample:

( 568)

Much of what we know-or at least hypothesize-about the moral properties of human behavior comes from the clinic. Only recently have more experimental and more replicable investigations begun to examine these matters. Quite naturally, such studies have begun with the clinically obtained hypotheses. And clearly, the clinic historically has drawn its main clientele from the ill, the disturbed, the badly socialized. These are the ones who resist temptation too little or too much, and who are in trouble one way or another with their feelings of guilt. A clinic population does not draw so heavily from those who are achieving their own ideals. Even among the ill, the therapeutic focus seems to fall more intensively on lapses from grace than on failures to achieve the ideal.

A select sample of clinical cases, who admittedly are at odds with their fellow men, is certainly an inadequate basis for generalizations concerning the nature of human motivation. There is a related habit, hard to overcome because it is so strongly ingrained. The tendency is to jump to conclusions and to evaluate human beings everywhere on the basis of the standards of value inculcated in us within the confines of human arrangements prevailing in our own culture.

Although modern anthropology has documented the wide variations in competition, submission, and aggression in different cultures, some psychological research on these topics proceeds uncritically from the assumption that the pattern prevailing in the researcher's society is typical. The use of any human culture as a prototype for generalizations about man's social motives is risky, as we shall see.

The Error in Seeking the Roots of Aggressiveness in Primitive Cultures

In theorizing about human nature, it has been almost an article of faith that the pristine traits of humanity would be revealed more clearly among peoples in more primitive cultures. Technically developed societies should be further removed from original human impulses, according to this reasoning. Therefore, if one would know whether human nature is innately aggressive, look to man in a less developed society.

Especially since the latter half of the 19th century, treatises marshaling evidence from various primitive cultures were written in efforts to prove the particular author's assumptions about the traits of original human nature. The selectivity of the cases included by various authors, in line with their particular assumptions, is a fascinating example of the intrusion of subjective factors in shaping discrimination and decision, which has also been found in a number of laboratory experiments.

Psychological theorizing on influences contributing to friendship or en-

( 569) -mity between human groups gains greater perspective if it starts from the vantage point provided by cultural anthropology as to the occurrence, frequency and scope of friendly or hostile intergroup events. Especially helpful for the student of intergroup relations is the orientation pursued for years by Leslie White and his associates. White and his associates do not present psychological analyses; on the contrary, they provide a notion of culture at its own level of organization. Their conception of culture is not reductionistic; i.e., culture is not reduced to unrelated, meaningless bits of behavior with no context. Culture is conceived as a meaningful structure. From a psychological viewpoint, relevant aspects of this pattern are parts of the stimulus world the individual confronts. His perceptions can be assessed more effectively within this cultural context.

A distinctive feature of White's orientation makes it particularly pertinent for assessing the human and physical arrangements contributing to positive or negative relations between human groupings. Instead of considering each cultural group as a more or less closed "culture pattern," each a law virtually unto itself, this orientation puts cultures in a time perspective. Lawful recurrences over time enable comparisons between cultures at different times, in terms of the actualities of prevailing conditions, techniques, modes of life, and premises regulating human arrangements (e.g., White, 1959).

Working from the viewpoint of cultural evolution, Marshall D. Sahlins (1960) in a paper on the "Origin of Society" reached a conclusion (pp. 8182) about the search for the roots of aggression in primitive societies. It should serve as a corrective to statements about original human impulses formulated without regard to the prevailing mode of life and human arrangements:

Territorial relations among neighboring human hunting-and-gathering bands (a term used technically to refer to the cohesive local group) offer an instructive contrast. The band territory is never exclusive. . . . Warfare is limited among hunters and gatherers. Indeed, many are reported to find the idea of war incomprehensible. A massive military effort would be difficult to sustain for technical and logistic reasons. But war is even further inhibited by the spread of a social relation-kinship-which in primitive society is often a synonym for "peace." Thomas Hobbes' famous fantasy of a war of "all against all" in the natural state could not be further from the truth. War increases in intensity, bloodiness, duration and significance for social survival through the evolution of culture, reaching its culmination in modern civilization. Paradoxically the cruel belligerence that is popularly considered the epitome of human nature reaches its zenith in the human condition most removed from the pristine. By contrast, it has been remarked of the Bushmen that "it is not in their nature to fight."

( 570)

Human Motives and Their Social Context

Early in this century, theory in academic psychology was primarily intellectualistic. Reacting against this heritage, the emphasis shifted toward the psychodynamics of instinctual impulses conceived as the driving force of man's deeds, even though he is not aware of them. Reaction against the over-intellectualistic tradition was carried too far. In some quarters, emphasis on the vicissitudes of instincts is still very much alive. In others. motivational forces, even acquired motives, are taken as absolute determinants of behavior. An impressive series of assertions by well-known therapists and other psychodynamicists could be quoted to document the excessive emphasis on motives.

As Eysenck (1950, p. 64) stated:

In their excitement about the discovery of the powers of "emotion" over "intellect," many psychologists have gone to extremes, portraying the "man in the street" as the mere plaything of uncontrollable unconscious forces which cannot in any way be influenced by reason. Such a view is no less contrary to fact than the previous over-estimation of rationalistic influences; what is needed is a more realistic appraisal of the relative importance of these two factors in each individual case.

The psychodynamicists have tended to view the individual as a selfcontained powerhouse, generating his own impulses without regard to transactions with his cultural setting; to view the development of human conscience as merely the "dread of society"; to view society itself solely as the agent of oppression to the individual's impulses and passions. If we have pretense of a scientific approach, we must heed the indications of converging lines from psychiatry, studies of human development, animal studies, social psychology, and cultural anthropology. A balanced view must consider society as more than a system of suppressive prohibitions and conscience as more than "dread of society," but also composed of the values and imperatives of social groups of which the individual is an integral part (cf. Pear, 1950, p. 134).

Because of past emphases, a social scientist is still expected to be an apologist for the group or culture, and the psychiatrist and psychologist partial to psychodynamic explanations, ignoring group and social influences. Developments in recent theory and research are eliminating the one-sided emphasis in both social science and psychology. The adequate study of man and his relations requires taking into account his intimately felt motives, yearnings, aspirations, and cognitive processes in the context of his affiliations with other people and the sociocultural setting of which he is a

( 571) part (Sherif and Koslin, 1960; Sherif and Sherif, 1964). No man feels, yearns, thinks, hates or fights altogether in isolation.

For the individual, the social context constitutes a stimulus situation that must be included in analysis of his experience and behavior. It is also true that man can affect the social context itself by his actions. However, it is erroneous and misleading to conclude that events occurring in the social context are simply manifestations of psychological processes. On the contrary, a different level of analysis is required when we move from the individual as a unit of analysis to aspects of his social context: other persons, groups, institutions, social values, technology, and so forth. This does not mean that the two levels of analysis-psychological and sociocultural---are contradictory. Instead, each is necessary for a rounded account of individual behavior and sociocultural events.

What is uniquely individual and what is sociocultural must be integrated in the study of man's relation to man, including his aggressive confrontations. This view has gained substantial support in recent years, factually and theoretically (see Cloutier, 1965). A recent expression is found in a monograph on Psychiatric Aspects of the Prevention of Nuclear War prepared by the Group for the Advancement of Psychiatry (1964). After reviewing a considerable body of relevant material the authors recapitulate their conclusion (p. 118)

War is a social institution; it is not inevitably rooted in the nature of man. Although war has traditionally served as an outlet for many basic human psychological needs, both aggressive and socially cohesive ones, the increasing mechanization and automation of modern warfare has rendered it less and less relevant to these needs. There are other social institutions and other means of conducting conflict between groups of people, or between nations, that can serve these psychological needs more adaptively in our modern world.

A Second Look at the Springs of Aggressive Acts

In the light of accumulating research, psychologists and social scientists have felt the necessity in recent years of examining the springs of aggressive deeds more carefully. After World War II, there was a crop of post-mortem verdicts in professional journals (including the Journal of Abnormal and Social Psychology published by the American Psychological Association) analyzing the horrors of the recent war in such terms as "collective guilt," "early childhood frustrations" of Axis leaders, and "unleashing of hidden aggressive tendencies" accumulated during the critical period preceding the war. Shortly thereafter, psychologists, social scientists and psychiatrists turned to more serious study of the causes of aggressive tendencies and

( 572) -deeds. Necessarily, such study required considerations outside of the provincialism of their particular specialty. They gave more receptive consideration to evidence from various disciplines.

They found, for example, studies of the American soldier conducted during the war (Stouffer et al., 1949). Included were answers to questions about reasons for fighting. The soldiers were the men actually doing the fighting, in situations where the expression of aggressive feelings toward the enemy was socially permissible. Yet the majority of the soldiers said they fought to "get the job done," or because they did not want to let their outfits down. Only 2 percent said they fought out of anger, revenge, or "fighting spirit." Another 3 percent gave replies that might be interpreted as aggressive, such as "making a better world," "crushing the aggressor," "belief in what I am fighting for" (p. 109). Certainly aggressive and vindictive feelings were not the most salient reported by these American fighting men.

Such reports led to a closer look at the locus of aggressive and destructive emotions, since quite clearly they were not universal. Pear (1950) commented on this point, asking whether wars were the outcome of aggressive impulses of the general populace or whether such impulses had to be fanned in the name of things sacred and just to a people. He reached the following conclusion about a theory of war based on individual impulses toward aggression:

It fails to distinguish between the aggressiveness of the warmakers, which can be very real indeed (though frequently personal greed, still socially disapproved if found out, masquerades as socially approved aggressiveness) and the attitudes of the general population, many of whom may not know of the impending war, of the combatant, the semi-combatant soldiers, and of the victims. In a war involving more than half the population of the world, a vast number of people who had nothing to do with declaring war suffered passively. Often aggressiveness had to be stirred up and intensified even in the fighters (we have recently read about the experimental army "hate school" abolished as a result of psychiatrists' reports) [;] in the uniformed sections many people of both sexes lived an unaggressive life and yet helped to win the war[;] "backroom boys" and scientists are unlikely to have done their best thinking if viscerally stirred: "beating the enemy" cannot have been a constant day-and-night goal giving incentive to all non-combatants, as the excellent book War Factory, among others. showed.

At about the same time, another eminent psychologist known for his continued interest in "personal factors" wrote on the "role of expectancy" of war in bringing about conflict. Gordon Allport of Harvard University concluded (in Cantril, 1950, p. 43):

( 573)

The people of the world-the common people themselves-never make war. They are led into war, they fight wars, and they suffer the consequences; but they do not actually make war. Hence when we say that "wars begin in the minds of men" we can mean only that under certain circumstances leaders can provoke and organize the people of a nation to fight. Left alone people themselves could not make war.

Foundations and organizations have arranged several joint committees of experts to pool their findings on the causes of aggressive passions and deeds. For one of these, the sessions and comments by psychologists, psychiatrists, and social scientists of various scientific positions and ideologies were published for UNESCO with Hadley Cantril as editor (1950). (The participants were Gordon W. Allport, Gilberto Freyre, Georges Gurvitch, Max Horkheimer, Arne Naess, John Richman, Harry Stack Sullivan, and Alexander Szalai. )

Despite their differences on several points, all participants agreed on several fundamental issues. Their common statement included agreement on the following conclusion: "To the best of our knowledge, there is no evidence to indicate that wars are necessary and inevitable consequences of `human nature' as such" (p. 17).

Neither the psychodynamicist nor the newspaper commentator who concludes that wars are inevitable because of "human nature" has provided adequate evidence for his theory. Their arguments begin with the observation that men make wars, then state that men make wars because they are aggressive by nature. The proof of men's aggressive nature lies in the fact that they make wars. Nothing has been added to the first observation but words to confuse the unwary.

Frustration-Aggression in Its Social Context

As mentioned earlier, Freud's position before he posited the "death instinct" as the source of hostile and destructive dispositions was that aggression is the result of frustration of basic drives, especially in early childhood. This earlier formulation by Freud found its most systematic expression in a highly influential book published over a quarter of a century ago (Dollard et al., 1939). It advanced the theory that all aggressive behavior (excluding only instrumental acts) was the outcome of frustration, and that frustration invariably led to aggressive tendencies.

In the face of criticism from various quarters, the formulation was subsequently qualified to indicate that every frustration does not necessarily lead to aggression. In the words of one of the principal proponents of the formulation, frustration "produces instigators to a number of different types of responses, one of which is an instigation to some form of aggression" (Miller, 1941, p. 338).


Of course, frustration suffered when activities directed toward attainment of goal objects are blocked is a state of arousal and tension with some consequences. But, in a systematic examination of the possible consequences, the following questions have to be faced:

1. (a) Is every act of aggression the outcome of frustration?  (b) Is aggression the invariable response to frustration?

2. Within what sets of circumstances and within what framework of interpersonal and group ties is frustration conducive or not conducive to aggressive deeds?

In relation to the first question, Berkowitz' critical survey of research (1962) led him to conclude unequivocally that "there are some aggressive acts. . . . that are not necessarily instigated by frustrations" (p. 30). Among such acts, he includes a number with important consequences, such as wholesale killing and destruction initiated as policies during wartime. He also cites the interesting research by Bandura and his co-workers demonstrating that children can acquire hostile modes of behavior merely by observing the aggressive actions by an adult, even when the adult is "nurturant" to the children.

J. P. Scott (1958) cites evidence from experiments on animal behavior in his own and other laboratories that there are important causes of aggressive behavior other than frustration. "For example, we have seen that the best way to train a mouse to be highly aggressive is not to frustrate him but to give him success in fighting" (p. 33). "Frustration leads to aggression only in a situation where the individual has a habit of being aggressive" (p. 35).

The further question of whether frustration invariably leads to aggressive behavior in animals is answered by Scott in these words: "In short, while frustration is highly likely to produce aggression, the result may also be other kinds of behavior" (p. 34).

Himmelweit (1950) reached a similar conclusion in her survey of studies of human frustration. She listed the following responses to frustration observed in different studies of human subjects: aggression, regression (lowering the level of performance), evading the situation by leaving it or daydreaming, apathy and resignation (especially for prolonged frustration), repression or "forgetting." Frequently several of these responses were observed in the same frustrating situation.

When applied to problems of hostility and prejudice between human groups the frustration-aggression hypothesis is an inadequate and even misleading explanation of aggressive behavior and of reaction to frustration. It is misleading because it directs investigation exclusively toward an aroused psychological state and a class of actions (aggressive). Any

( 575) adequate hypothesis about human behavior must include specification of the environmental circumstances in which a psychological state occurs and in which a class of actions occurs. As applied to frustration and aggression in group relations, this means that an adequate hypothesis must include statements about the culture and organizational context of both tension arousal and behavior. The second question raised above is pertinent here.

If the frustrating experience is strictly an individual affair within the context of day-to-day interpersonal relations, how can it become the basis for prejudice, injustice, and aggressive actions that follow existing dominance-subordination arrangements in society? Why is there not more variation in the targets of hostility? It is necessary to ask whether a frustration is individually experienced or is seen by other members of a group as a common frustration. If the frustration is seen as shared, does it have any bearing on the direction of the greatest prejudice and the most hostile actions toward other groups? According to an adapted version of the frustration-aggression hypothesis, the most vehement aggressive impulses should be generated among those who share the greatest frustrations.

Is the direction of prejudice and aggressive actions primarily from the most downtrodden and deprived groups toward dominant groups who maintain their mighty position, or are these not frequently directed from the mighty and powerful toward the underprivileged groups, who one may assume are more frustrated? Such questions direct serious consideration to the organizational or group context of frustration and of aggressive actions. Answers to such questions are necessary before we can conclude that frustration suffered by single individuals without becoming a common concern have a major role in intergroup aggression.

It should be clear that human societies vary a great deal in whom they select as proper targets of hatred and aggressive acts. These social variations have profound significance for the role of frustration in generating aggressive deeds. The individual who develops in a society where discrimination is directed against a particular group and who is not obviously a member of that group learns to conceive of himself as belonging to those groups that consider themselves to be superior to the discriminated group. If members of the discriminated group challenge his superior position, he may feel very frustrated indeed, but frustrated as a member of the dominant and superior kind. Thus, the social context is essential for analyzing both his experience of frustration and its consequences for his behavior.

Perhaps the inadequacies of the frustration-aggression hypothesis for predicting violence within an organizational context are most striking when it is applied to lynchings or murders of dominated groups. Such deeds have been attributed to frustrations or deprivations suffered by the lynchers, who frequently, though not always, have included "poor whites."

There may be a grain of truth that personal frustrations have something

( 576) to do with the degree and extent of brutality by individual lynchers. But the presence or absence of lynching as an institution, to be engaged in without serious fear of punishment, is not adequately explained in these terms. When this brutal institution is outside the bounds of the organizational and normative orientation of the group, when it is forbidden and punishable, the mere thought of such deeds in concert with one's fellows is personally appalling, no matter what the individual's frustrations may be. The point was made well in a comparison reported by Klineberg (1950, p. 198): 

White Brazilians are, on the whole, much more economically frustrated than white Americans. The economic standards of the former are definitely much lower, and relatively many more of them live near or at a bare subsistence level. There are fluctuations in economic conditions in Brazil, just as in the United States. However, there are no lynchings of Brazilian Negroes. . . . This fact makes it clearly inadequate to explain aggression against Negro or, in more general terms, hostility against other groups (which may take the form of war in extreme cases) entirely in terms of the aggressive impulses developed within the individual as a result of his frustrations.

It would be difficult to contend that there are no white Americans frustrated for reasons of poverty or lack of status. But despite this, it may be safe to predict that lynching and publicly condoned murder will be driven out-of-bounds in the United States, including the South. If so, it will be because of a changing context for intergroup behavior. This changing context will reflect the determined efforts of many Americans toward equal civil rights for all citizens.

Relationships within and between human groups, which form the context for frustration and associated aggression toward others because of their group membership, set limits for the degree and targets of aggression and chart the direction of what is desirable, or even ideal, in intergroup action. Frustration and aggression are undeniable human phenomena. Unless they are assessed within the context of group relationships, researchers can scarcely hope to have more than fragments of data, useful for valid predictions only in the restricted interpersonal relationships of the laboratory.


The individual is the unit of psychological analysis, but any psychological analysis necessarily includes variables from the stimulational background and the immediate stimulus context of the individual. Attempts to deal with

(577) human aggression exclusively through motivational concepts amount to ignoring the stimulational background and context.

When aggression is directed toward members of other groups, the relevant stimulational background and context are organizational and sociocultural. As the student of vision specifies the properties of visual stimuli in non psychological units, so too must the student of human aggression attempt to specify social stimuli at a different level of analysis. The relevant units of analysis are not merely other individuals, but groups, social and economic organizations, institutions, cultures and societies. The properties of these units, of which the individual is a part, are as crucial to analysis of his behavior as any emotion or motive he may experience. Necessarily, then, assessment of the etiology of man's inhumanity to men in other groups requires juncture of the psychological and sociocultural levels of analysis.

This perspective permits specification of the conditions in which psychological factors, such as frustration, are and are not conducive to aggression toward another group or its members. In studies guided by this perspective, we have found that the sufficient condition for the rise of aggressive tendencies toward another group is conflict between the goals of the groups in question (Sherif and Sherif, 1953; 1969). To the extent that the group constitutes a stable pattern of interpersonal ties with binding standards of conduct for the members, the individual member moves to further group goals, and experiences pleasant emotions at its successes and personal frustration at its reverses. When, over time, the actions of another group appear to threaten a cherished goal of his group, the individual joins with enthusiasm into the formulation and execution of aggressive plans.

The research subjects in these experiments were normal, well-adjusted, and healthy in their usual environments, not warped or perverted. The crucial importance of the properties of the groups and their relationships was demonstrated by changing the conditions of their interaction. At a time when intergroup antagonism was so intense that each group refrained from contact with the other, they faced a series of experimentally introduced goals that had high appeal value for both groups, but could be attained only if both pooled their resources and efforts (superordinate goals). Over time, the erstwhile aggressors began to cooperate, then to change their attitudes toward one another. Friendship was extended across group lines (Sherif, 1966).

The experiments demonstrate the necessity of including variables at the sociological level of analysis (in this case, the properties of groups and intergroup relations) in dealing with events at the psychological level (in this case, the aggressive or friendly attitudes and behavior of the indi-

( 578) -vidual). The events that ensued cannot be accounted for by eliminating their social context and its change. Nor is the picture complete without considering the significant consequences of individual experience and action, for these were the medium by which the social context was established and then transformed from intense conflict to harmonious interplay between groups.

The traditional dichotomy between psychological and sociological levels of analysis dies hard. As long as its remnants influence theory and research, accounts of man's social behavior will be filled with contradictions and apparent paradox. These can be resolved by recognition that both levels of analysis are necessary and must be integrated in a comprehensive conception.


Berkowitz, L. 1962. Aggression: A social psychological analysis. New York: McGraw-Hill.

Cantril H., ed. 1950. Tensions that cause wars. Urbana: Univ. Illinois Press.

Cloutier, F. 1965. International tensions and mental Health. In M. Schwebel, ed., Behavioral science and human survival. Palo Alto, Calif.: Science and Behavior Books, Inc. Pp. 96-102.

Dollard, J., N. Miller, L. W. Doob, O. H. Mowrcr, R. R. Sears. 1939. Frustration and aggression. New Haven: Yale Univ. Press.

Eysenck, H. J. 1950. War and aggressiveness: a survey of social attitude studies. In T. H. Pear, ed., Psychological factors of peace and war. New York: Philosophical Library.

Freud, S. 1922. Group psychology and the analysis of the ego. London: Hogarth.

Freud, S. 1927. The ego and the id. London: Hogarth.

Freud, S. 1930. Civilization and its discontents. London: Hogarth.

Group for the Advancement of Psychiatry. 1964. Psychiatric aspects of the prevention of nuclear war. New York.

Himmelweit, Hilda. 1950. Frustration and aggression, a review of recent experimental work. In T. H. Pear, ed., Psychological factors of peace and war. New York: Philosophical Library.

Klineberg, O. 1950. Tensions affecting international understanding. New York: Social Sciences Research Council, Bull. 62.

McDougall, W. 1923. Outline of psychology. New York: Scribner.

Miller, N. 1941. The frustration-aggression hypothesis. Psychol. Rev. 48: 337-342.

Pear, T. H. 1950. Peace, war and culture patterns. In T. H. Pear, ed., Psychological factors of peace and war. New York: Philosophical Library.

( 579)

Sahlins, M. D. 1960. Origin of Society. Sci. Am. 203 (Sept. 1960): 76-87.

Schneirla, T. C. 1946. Problems in the biopsychology of social organization. J. Ahnorm. Soc. Psychol. 41: 385-402.

Schneirla, T. C. 1951. The "levels" concept in the study of social organization. In J. H. Rohrer and M. Sherif, eds., Social psychology at the crossroads. New York: Harper.

Schneirla, T. C. 1953. The concept of levels in the study of social phenomena. In M. Sherif and C. W. Sherif, Groups in harmony and tension. New York: Harper. (1966 edition, Octagon Books, New York.)

Schneirla, T. C. 1964. Instinctive behavior, maturation-experience or development. In N. R. F. Maier and T. C. Schneirla, Principles of animal psychology. New York: Dover.

Scott, J. P. 1958. Aggression. Chicago: Univ. Chicago Press.

Sears, R. R. 1960. The growth of conscience. In I. Iscoe and H. Stevenson, eds., Personality development in children. Austin: Univ. of Texas Press.

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

Sherif, M., and B. Koslin, 1960. The "institutional" vs. "behavioral" controversy in social science with special reference to political science. In M. Sherif, Social interactions: Process and products. Chicago: Aldine, 1967. Chapter 4.

Sherif, M., and C. W. Sherif. 1953. Groups in harmony and tension. New York: Harper. (1966 edition, Octagon Books, New York.)

Sherif, M., and C. Sherif. 1964. Reference groups: Exploration into conformity and deviation of adolescents. New York: Harper and Row.

Sherif, M., and C. W. Sherif. 1969. Social psychology. 3rd ed. New York: Harper and Row. Chapters 11-12.

Stouffer, S. A., A. A. Lumsdaine, M. H. Lumsdaine, R. M. Williams, Jr., M. B. Smith, I. L. Janis, S. A. Star, L. S., Cottrell. 1949. Studies in social psychology in World War II, vol. II. The American soldier. Princeton N. J.: Princeton Univ. Press.

Tolman, E. C., 1942. Drives toward war. New York: Appleton-Century.

White, L. A. 1959. The evolution of culture. New York: McGraw-Hill.



Valid HTML 4.01 Strict Valid CSS2