An appraisal of personality-oriented approaches to prejudice[1]

William R. Hood and Muzafer Sherif
Institute of Group Relations, University of Oklahoma

The objective of this paper is to give an evaluation of how the tremendous volume of work done in recent years along authoritarian personality lines fits into the larger scheme of social psychology. No attempt will be made to go into the details of technique or methodology used in these studies. Rather, attention will be concentrated on seeing how the findings relate to other data. By and large, this approach has aimed at understanding prejudice, which is one aspect of intergroup relations, in terms of the authoritarian personality, one segment of the area of personality psychology.

Let there be no question of the importance of personality factors in intergroup relations. Of course, such factors as frustration-tolerance, projectivity, intelligence, destructiveness, flexibility, and conventionalism are important ; but, even so, we must remember that these are not self-generating, that they must also be subjected to developmental analysis. The question of how these personality factors come into being is also a major problem, but here the question is exactly how these factors operate in intergroup affairs—in what functional relations.

Research along authoritarian personality lines has been an outgrowth, as have many other trends, of rapidly increasing concern over the vital and frequently grim problems of intergroup relations in the world today. Prejudice, ethnocentrism, and related topics are crucial aspects of these problems, and any light that can be shed on these complex areas is certainly desirable. We must, however, always keep in mind that the problems of intergroup relations are not solely psychological problems.

( 80) Social psychologists cannot alone reach generalizations and write prescriptions for resolving intergroup conflict and tension, and we should admit frankly that we have not even, as yet, established a comprehensive and integrated perspective for dealing with these problems.

However, a glance back over the attempts to deal with problems of group relations reveals systematic errors and inadequacies which we are now in a better position to avoid. Historically, many approaches and doctrines have been promulgated, consciously or unconsciously, in the interests of particular groups. Currently, we see academic alliances and theoretical preconceptions which are carried into research work resulting in one-sided approaches that deal only with certain aspects of the picture but which, nevertheless, claim to offer comprehensive treatments. Examples of these latter are the so-called "group-centered," "personality-centered," and "culture-centered" approaches. More specific biases such as the psychoanalytic and behavioristic orientations so prevalent today are seen not only in psychology but also in sociology, anthropology, and political science. When taken as comprehensive approaches to human relations, such one-sided emphases result in ignoring facts or twisting them into the narrow framework provided.

Factors coming from the individual—that is, personality factors and effects of frustration and other experiences—are of course important in understanding human relations—when they are taken as they fit in the interaction process. These factors will help illuminate the individual variations in exhibiting discrimination and prejudice within a given reference scale. However, in the tremendous volume of research done using the Bogardus Social Distance Scale [2] and its variations, we find some of the best-documented facts in social research to date. The social distance pattern in the United States, for example, is accepted by practically all groups and classes with only minor variations, and it persists over long periods, any noteworthy changes being clearly related to catastrophic world events. It is difficult to conceive of these consistencies as depending on widespread and long-lasting similarities in authoritarianism and frustration in the life histories of each generation in all the various groups.

The individual variations in degree, intensity, etc., within a given dimension may be determined, in part, by personality factors, but these alone cannot determine the existence or nonexistence of the scale. If the claims of the proponents of authoritarian personality went only so far, their findings could be more easily fitted into the over-all scheme of social psychology.


It is unfortunate that the carefully qualified statements of Adorno at certain points in the introductory and summary chapters of the book The Authoritarian Personality, [3] which is the chief proponent of this view under discussion, have so few counterparts throughout the rest of the volume. The over-all emphasis on personality factors of a particular type to the exclusion of other important factors is attested by such remarks as "...personality is essentially an organization of needs," which excludes cognitive factors, and "...both ideology and. . .group membership seem to express deeper trends in the individual," placing these factors in a secondary role.

But rather than quote at length in justification of this appraisal, let us examine some conclusions of authors of Studies in the Scope and Methods of "The Authoritarian Personality." Hyman and Sheatsley,[4] in their methodological critique, say that the point of view presented in The Authoritarian Personality tends to perpetuate the implication that the level of organization of attitudes is a kind of universal, an intrapsychic process which bears little relation to environmental conditions. They also speak of the authors instituting a devious search for some hidden determinant when, in a national sample, a particular attitude is almost universal, and of making the "personality theory" they offer identical with all of psychology, which, of course, it is not. For the same reasons Shils [5] complains that they assume political behavior is a function of deeper personality characteristics, social structure only playing the role of setting off the chain of personality-impelled actions.

Christie, in the same volume,[6] concludes that the general argument in The Authoritarian Personality emphasized purely personality determinants and discounted contemporary social influences. He says, further, that the discussion regarding the relationship between group membership and individual ideology in that volume strongly emphasized the lack of importance of group membership as a factor. Each statement by these critics is extensively documented from the original work, and generally the one-sided emphasis on personality factors can be documented at almost any length ; but this should be sufficient to show that no straw man or dead horse is involved.


Views which explain prejudice primarily through personality factors permit the individual to make of the world what he will through wish fulfillment, fantasy, and autism. While the deviate individual may succeed in ignoring reality to some extent, structuring and restructuring of perceptions by the great bulk of people are somewhat held in check by actual conditions such as, for example, group sanctions, which set certain compelling limits. The point is that intergroup relations of consequence today are not primarily matters of deviate behavior. And at any rate, as we shall see, we cannot extrapolate from the properties of individuals to the characteristics of group situations.

Still the question remains: Exactly how do personality factors operate —in what functional relations? Arguments on this point are now going on in all quarters with culture-, group-, and personality-oriented workers busily building cases for their preferred explanations. As an example which is pertinent in this context, Hyman and Sheatsley say of The Authoritarian Personality, "...the mistakes and limitations—no one of them perhaps crucial—unfortunately operate in favor of the authors' assumptions, and cumulatively they build up a confirmation of the theory which, upon examination, proves to be spurious.[7]

Findings such as those in The Authoritarian Personality and other studies based on psychoanalytic preconceptions point up the dangers of coming to the area of intergroup relations with fixed notions about the relative importance of various factors in any over-all way. Frustration, for example, is an important item in intergroup relations, and we must seriously consider the views of "displacement," "scapegoat," and "frustration-aggression" theories; but surely not as sovereign principles that can explain the whole of intergroup relations.

In Psychological Factors of Peace and War, [8] Eysenck and Himmelweit, clinical psychologists at Maudsley Hospital, London, have surveyed the literature in clinical and social areas dealing with theories which would make factors coming from the individual the sole or major determinant of aggressive attitudes in human relations. They point to the inadequacies of such a one-sided approach, favoring instead a functional approach to the situation as a whole.

Probably, all other theories advanced with a sovereign principle suffer from similar undue emphases on their particular brand of influence, whatever that may be. The scientist, unfortunately, is not a different breed from the rest of humanity. He also operates on a system of tacit

( 83) assumptions, concepts, generalizations, in terms of which he views the social world. Thus, it would seem that there is no satisfactory solution in arguments or disjointed evidence. If we are to achieve a vantage point which will give us valid criteria to apply, it must be through consideration of basic principles on which there is general agreement.

A body of knowledge relating to such basic principles has been accumulating for a number of years. There are facts in the psychology of perception which are so well established that they are no longer controversial topics. Most of these findings are in the line of development which attempts to tie up attitudes with some major findings in perception—especially perceptual selectivity as influenced by motivational factors. A tremendous amount of experimental evidence indicates that perception, taken as a prototype of experience, is jointly determined by the totality of functionally related external factors (objects, persons, events) and internal factors (motives, attitudes, complexes) operative at a given time. However, the factors which enter into such a frame of reference or system of relations—by which an individual item of experience or behavior can be understood—cannot be taken in a mechanical or additive way. Findings along the lines of "membership character" make it crystal clear that the parts of a structure have properties different from those they would have in other structures and different from their properties when studied in isolation.

If simple discrete weights presented to the individual in close temporal sequence, or simple forms, are judged or perceived in a relational way, it is reasonable to raise the question of how much more crucial it becomes to study the individual in the framework of interaction. We cannot legitimately extrapolate the effects of motivational urges of the individual to group situations as if the group situation were a void. Likewise, the compelling material conditions (technology, socioeconomic forces) influence human relations as affected by the existing organizational structure and by the system of beliefs or norms. These points are further elaborated in Sherif and Sherif, Groups in Harmony and Tension.[9]

Most of the theories proposed to explain intergroup relations, of which we must remember prejudice is only one aspect, deal with factors that do enter into determining the character of relationships between groups, and therefore between individual members. These factors enter jointly in various degrees in the shaping of the final product, each with a relative weight in a specific situation, but these relative weights are

( 84) not fixed quantities. The various items have their place in functional wholes which must receive first consideration if the parts are to be fitted into the larger picture in a meaningful way.

Mere juxtaposition of contributions from areas with different biases will not give the solutions to intergroup relations problems. The factors must be brought together in a functional way, not in a mechanical way. The results of an approach centered strictly in the individual, no matter how dynamic and functional internally, become merely mechanical juxtaposition when an attempt is made to relate them to contributions from other areas, and the value of the dynamic emphasis is lost.

While this overview is not new,[10] it has been especially on the rise since the middle thirties and is now beginning to find wider acceptance in the social sciences. Just one of the most recent examples is the Simpson and Yinger book.[11]

Even so, many of the views which accept multiple determinants only "make a place for" them. It cannot be too strongly emphasized that it is not sufficient simply to grant the fact that there are many factors coming from within the individual and from his surroundings that enter into determining the shape of intergroup relations. We have to go a step further and take these various factors in their functional relationships. And also we must keep in mind that the dominant factors or major anchorages will not be the same for all cases of intergroup relations or for all individuals involved in the same interaction. Statements made in advance of the facts, as to which will be the dominant factors under all circumstances, do a disservice to the development of a naturalistic social science. A view which emphasizes cultural, or group, or economic factors to the exclusion of individual differences can no more be condoned than its obverse.

In conclusion, we may summarize the high points of our discussion in a few sentences. On the whole, vital topics of man's relation to man and to groups have been formulated in terms of theories which may be labeled as primarily "group-centered" or "personality-centered." "Group-centered" approaches have stressed factors of group membership, socioeconomic factors, and cultural factors such as the existence of the prevailing values or norms. By and large, in this line of approach, personality factors have tended to be neglected.


On the other hand, "personality-centered" approaches have advanced theories which tried to explain the whole scheme of human relations primarily in terms of personality factors stemming from the particular life history of single individuals. Such theories have failed to give due recognition to the place of group membership, prevailing social distance scales, the nature of intergroup relations, and other sociocultural determinants.

Our assignment for this symposium has been a consideration of one current "personality-centered" approach to the problem of prejudice. It is pointed out in this particular approach, as in other psychoanalytically oriented approaches, that individuals, even in the same group, vary in the intensity and in the particular manners in which they reveal their prejudices. This is one thing, but trying to base the whole explanation on individual variations amounts to ignoring the limiting factors in the rise of the institution of prejudice.

The social distance scale of any given group is the product of positive or negative relations between groups. In determining the social distance scale of a given group, a whole host of socioeconomic and cultural factors come into the picture as they pertain to the relations of the group in question. The end result of relations between groups is standardized in terms of the social distance scales of the respective groups.

These scales define the limits within which individual variations will take place. An adequate personality theory will, at best, point up factors which contribute to the determination of these individual variations falling within the particular reference scale. But these factors do not determine the existence or nonexistence of the scale itself.

In actual research practice, this implies the order of attack which must be followed. The nature and extent of the prevailing scales must first be ascertained if the particular variations exhibited by given individuals are to be meaningful. Only by considering factors coming both from the sociocultural setting and from the individual in a functionally interrelated way, and by noting in every case the relative weights of factors coming from both sides, can we hope to attain an adequate functional account of intergroup phenomena such as prejudice.


  1. The main points of this paper were delivered at a symposium entitled "The Authoritarian Personality" at the annual meeting of the Southwestern Psychological Association, Oklahoma City, Oklahoma, December 17, 1954.
  2. E. S. Bogardus, "Measuring Social Distance," Journal of Applied Sociology, 9:299-308, 1925.
  3. T. W. Adorno, et al., The Authoritarian Personality (New York: Harper & Brothers, 1950).
  4. H. H. Hyman and P. B. Sheatsley, "The Authoritarian Personality—A Methodological Critique," in Studies in the Scope and Method of "The Authoritarian Personality" (Glencoe, Ill.: Free Press, 1954), edited by R. Christie and M. Jahoda.
  5. E. A. Shils, "Authoritarianism: 'Right' and 'Left,"' op. cit.
  6. Christie, "Authoritarianism Re-examined," op. cit.
  7. Hyman and Sheatsley, op. cit., p. 122 (emphasis in original).
  8. T. H. Pear, ed., Psychological Factors of Peace and War (New York: Philosophical Library, 1950), Chaps. 3 and 8.
  9. M. and C. W. Sherif, Groups in Harmony and Tension (New York: Harper & Brothers, 1953), Chaps. 6 and 8.
  10. M. Sherif, The Psychology of Social Norms (New York: Harper & Brothers, 1936).
  11. G. E. Simpson and J. M. Yinger, Racial and Cultural Minorities, An Analysis of Prejudice and Discrimination (New York: Harper & Brothers, 1953).

Valid HTML 4.01 Strict Valid CSS2