As a part of the case study of an individual an analysis in terms of more unique
personality
vari- ables was made, the material here being considered in close conjunction with findings from the interview.
Adorno-T-Authoritarian-Personality-Harper-Bros-1950
conception is necessary to explain consistency of behavior in widely varying
situations, to explain the persistence of ideological trends in the face of contradicting facts and radically altered social conditions, to explain why people in the same sociological situation have different or even conflicting views on social issues, and why it is that people whose behavior has been changed through psychological manipulation lapse into their old ways as soon as the agencies of manipulation are removed.
The conception of personality structure is the best safeguard against the inclination to attribute persistent trends in the individual to something "innate" or "basic" or "racial" within him. The Nazi allegation that natural, biological traits decide the total being of a person would not have been such a successful political device had it not been possible to point to numerous instances of relative fixity in human behavior and to challenge those who thought to explain them on any basis other than a biological one. Without the conception of personality structure, writers whose approach rests upon the assumption of infinite human flexibility and responsiveness to the social situation of the moment have not helped matters by referring persistent
(
? INTRODUCTION
7
trends which they could not aP. prove to "confusion" or "psychosis" or evil under one name or another. There is, of course, some basis for describing as "pathological" patte~ of behavior which do not conform with the most common, and seemingly most lawful, responses to momentary stimuli. But this is to use the term pathological in the very narrow sense of deviation from the average found in a particular context and, what is worse, to suggest that everything in the personality structure is to be put under this heading. Actually, personality embraces variables which exist widely in the popula- tion and have lawful relations one to another. Personality patterns that have been dismissed as "pathological" because they were not in keeping with the most common manifest trends or the most dominant ideals within a society, have on closer investigation turned out to be but exaggerations of what was almost universal below the surface in that society. What is "pathological" today may with changing social conditions become the dominant trend of tomorrow.
It seems clear then that an adequate approach to the problems before us must take into account both fixity and flexibility; it must regard the two not as mutually exclusive categories but as the extremes of a single continuum along which human characteristics may be placed, and it must provide a basis for understanding the conditions which favor the one extreme or the other. Personality is a concept to account for relative permanence. But it may be emphasized again that personality is mainly a potential; it is a readi- ness for behavior rather than behavior itself; although it consists in disposi- tions to behave in certain ways, the behavior that actually occurs will always depend upon the objective situation. Where the concern is with antidemo- cratic trends, a delineation of the conditions for individual expression re- quires an understanding of the total organization of society.
It has been stated that the personality structure may be such as to render the individual susceptible to antidemocratic propaganda. It may now be asked what are the conditions under which such propaganda would increase in pitch and volume and come to dominate in press and radio to the exclusion of contrary ideological stimuli, so that what is now potential would become actively manifest. The answer must be sought not in any single personality nor in personality factors found in the mass of people, but in processes at work in society itself. It seems well understood today that whether or not antidemocratic propaganda is to become a dominant force in this country depends primarily upon the situation of the most powerful economic inter- ests, upon whether they, by conscious design or not, make use of this device for maintaining their dominant status. This is a matter about which the great majority of people would have little to say.
The present research, limited as it is to the hitherto largely neglected psychological aspects of fascism, does not concern itself with the production of propaganda. It focuses attention, rather, upon the consumer, the indi-
? 8 THE AUTHORITARIAN PERSONALITY
vidual for whom the propaganda is designed. In so doing it attempts to take into account not only the psychological structure of the individual but the total objective situation in which he lives. It makes the assumption that people in general tend to accept political and social programs which they believe will serve their economic interests. What these interests are depends in each case upon the individual's position in society as defined in economic and sociological terms. An important part of the present research, therefore, was the attempt to discover what patterns of socioeconomic factors are asso- ciated with receptivity, and with resistance, to antidemocratic propaganda.
At the same time, however, it was considered that economic motives in the individual may not have the dominant and crucial role that is often ascribed to them. If economic self-interest were the only determinant of opinion, we should expect people of the same socioeconomic status to have very similar opinions, and we should expect opinion to vary in a meaningful way from one socioeconomic grouping to another. Research has not given very sound support for these expectations. There is only the most general similarity of opinion among people of the same socioeconomic status, and the exceptions are glaring; while variations from one socioeconomic group to another are rarely simple or clear-cut. To explain why it is that people of the same socioeconomic status so frequently have different ideologies, while people of a different status often have very similar ideologies, we must take account of other than purely economic needs.
More than this, it is becoming increasingly plain that people very fre- quently do not behave in such a way as to further their material interests, even when it is clear to them what these interests are. The resistance of white-collar workers to organization is not due to a belief that the union will not help them economically; the tendency of the small businessman to side with big business in most economic and political matters cannot be due entirely to a belief that this is the way to guarantee his economic indepen- dence. In instances such as these the individual seems not only not to con- sider his material interests, but even to go against them. It is as if he were thinking in terms of a larger group identification, as if his point of view were determined more by his need to support this group and to suppress opposite ones than by rational consideration of his own interests. Indeed, it is with a sense of relief today that one is assured that a group conflict is merely a clash of economic interests-that each side is merely out to "do" the other- and not a struggle in which deep-lying emotional drives have been let loose. When it comes to the ways in which people appraise the social world, irra- tional trends stand out glaringly. One may conceive of a professional man who opposes the immigration of Jewish refugees on the ground that this will increase the competition with which he has to deal and so decrease his income. However undemocratic this may be, it is at least rational in a limited sense. But for this man to go on, as do most people who oppose Jews on
? INTRODUCTION
9
occupational grounds, and accept a wide variety of opinions, many of which are contradictory, about Jews in general, and to attribute various ills of the world to them, is plainly illogical. And it is just as illogical to praise all Jews in accordance with a "good" stereotype of them. Hostility against groups that is based upon real frustration, brought about by members of that group, undoubtedly exists, but such frustrating experiences can hardly ac<;ount for the fact that prejudice is apt to be generalized. Evidence from the present study confirms what has often been indicated: that a man who is hostile toward one minority group is very likely to be hostile against a wide variety of others. There is no conceivable rational basis for such generalization; and, what is more striking, prejudice against, or totally uncritical acceptance of,
a particular group often exists in the absence of any experience with mem- bers of that group. The objective situation of the individual seems an unlikely source of such irrationality; rather we should seek where psychology has already found the sources of dreams, fantasies, and misinterpretations of the world-that is, in the deep-lying needs of the personality.
Another aspect of the individual's situation which we should expect to affect his ideological receptivity is his membership in social groups-occu- pational, fraternal, religious, and the like. For historical and sociological reasons, such groups favor and promulgate, whether officially or unofficially, different patterns of ideas. There is reason to believe that individuals, out of their needs to conform and to belong and to believe and through such deviCes as imitation and conditioning, often take over more or less ready-made the opinions, attitudes, and values that are characteristic of the groups in which they have membership. To the extent that the ideas which prevail in such a group are implicitly or explicitly antidemocratic, the individual group mem-
ber might be expected to be receptive to propaganda having the same general direction. Accordingly, the present research investigates a variety of group memberships with a view to what general trends of thought-and how much variability-might be found in each.
It is recognized, however, that a correlation between group membership and ideology may be due to different kinds of determination in different individuals. In some cases it might be that the individual merely repeats opinions which are taken for granted in his social milieu and which he has no reason to question; in other cases it might be that the individual has chosen to join a particular group because it stood for ideals with which he was already in sympathy. In modem society, despite enormous communality in basic culture, it is rare for a person to be subjected to only one pattern of ideas, after he is old enough for ideas to mean something to him. Some selec- tion is usually made, according, it may be supposed, to the needs of his personality. Even when individuals are exposed during their formative years almost exclusively to a single, closely knit pattern of political, economic, social, and religious idei(s, it: is found that some confor! ll while others rebel,
? 10 THE AUTHORITARIAN PERSONALITY
and it seems proper to inquire whether personality factors do not make the difference. The soundest approach, it would seem, is to consider that in the determination of ideology, as in the determination of any behavior, there is a situational factor and a personality factor, and that a careful weighing of the role of each'will yield the most accurate prediction.
Situational factors, chiefly economic condition and social group member- ships, have been studied intensively in recent researches on opinion and atti- tude, while the more inward, more individualistic factors have not received the attention they deserve. Beyond this, there is still another reason why the present study places particular emphasis upon the personality. Fascism, in order to be successful as a political movement, must have a mass basis. It must secure not only the frightened submission but the active cooperation of the great majority of the people. Since by its very nature it favors the few at the expense of the many, it cannot possibly demonstrate that it will so improve the situation of most people that their real interests will be served. It must therefore make its major appeal, not to rational self-interest, but to emotional needs-often to the most primitive and irrational wishes and fears. If it be argued that fascist propaganda fools people into believing that their lot will be improved, then the question arises: Why are they so easily fooled? Because, it may be supposed, of their personality structure; because of long-
. established patterns of hopes and a~pirations, fears and anxieties that dispose them to certain beliefs and make them resistant to others. The task of fascist propaganda, in other words, is rendered easier to the degree that antidemo- cratic potentials already exist in the great mass of people. It may be granted that in Germany economic conflicts and dislocations within the society were such that for this reason alone the triumph of fascism was sooner or later inevitable; but the Nazi leaders did not act as if they believed this to be so; instead they acted as if it were necessary at every moment to take into account the psychology of the people-to activate every ounce of their anti- democratic potential, to compromise with them, to stamp out the slightest spark of rebellion. It seems apparent that any attempt to appraise the chances of a fascist triumph in America must reckon with the potential existing in the character of the people. Here lies not only the susceptibility to antidemo- cratic propaganda but the most dependable sources of resistance to it.
The present writers believe that it is up to the people to decide whether or not this country goes fascist. It is assumed that knowledge of the nature and extent of antidemocratic potentials will indicate programs for demo- cratic action. These programs should not be limited to devices for manipu- lating people in such a way that they will behave more democratically, but they should be devoted to increasing the kind of self-awareness and self- determination that makes any kind of manipulation impossible. There is one explanation for the existence of an individual's ideology that has not so far been considered: that it is the view of the world which a reasonable man,
? INTRODUCTION
with some understanding of the role of such determinants as those discussed above, and with complete access to the necessary facts, will organize for himself. This conception, though it has been left to the last, is of crucial importance for a sound approach to ideology. Without it we should have to share the destructive view, which has gained some acceptance in the modern world, that since all ideologies, all philosophies, derive from non- rational sources there is no basis for saying that one has more merit than another.
But the rational system of an objective and thoughtful man is not a thing apart from personality. Such a system is still motivated. What is distinguish- ing in its sources is mainly the kind of personality organization from which it springs. It might be said that a mature personality (if we may for the moment use this term without defining it) will come closer to achieving a rational system of thought than will an immature one; but a personality is no less dynamic and no less organized for being mature, and the task of describing the structure of this personality is not different in kind from the task of describing any other personality. According to theory, the person- ality variables which have most to do with determining the objectivity and rationality of an ideology are those which belong to the ego, that part of the personality which appreciates reality, integrates the other parts, and operates with the most conscious awareness.
It is the ego that becomes aware of and takes responsibility for nonra- tional forces operating within the personality. This is the basis for our belief that the object of knowing what are the psychological determinants of ideology is that men can become more reasonable. It is not supposed, of course, that this will eliminate differences of opinion. The world is suffi- ciently complex and difficult to know, men have enough real interests that are in conflict with the real interests of other men, there are enough ego- accepted differences in personality to insure that arguments about politics, economics, and religion will never grow dull. Knowledge of the psycholog- ical determinants of ideology cannot tell us what is the truest ideology; it can only remove some of the barriers in the way of its pursuit.
B. METHODOLOGY
1. GENERAL CHARACTERISTICS OF THE METHOD
To attack the problems conceptualized above required methods for de- scribing and measuring ideological trends and methods for exposing person- ality, the contemporary situation, and the social background. A particular methodologicar challenge was imposed by the conception of levels in the person; this made it necessary to devise techniqu~s for surveying opinions, attitudes, and values that were on the surface, for revealing ideological
r -l
? ii. THE AUTHORITARiAN PERSONALITY
trends that were more or less inhibited and reached the surface only in indirect manifestations, and for bringing to light personality forces that lay in the subject's unconscious. And since the major concern was with patterns of dynamically related factors-something that requires study of the total individual-it seemed that the proper approach was through intensive clinical studies. The significance and practical importance of such studies could not be gauged, however, until there was knowledge of how far it was possible to generalize from them. Thus it was necessary to perform group studies as well as individual studies, and to find ways and means for integrating the two.
Individuals were studied by means of interviews and special clinical tech- niques for revealing underlying wishes, fears, and defenses; groups were studied by means of questionnaires. It was not expected that the clinical studies would be as complete or profound as some which have already been performed, primarily by psychoanalysts, nor that the questionnaires would be more accurate than any now employed by social psychologists. It was hoped, however-indeed it was necessary to our purpose-that the clinical material could be conceptualized in such a way as to permit its being quan- tified and carried over into group studies, and that the questionnaires could be brought to bear upon areas of response ordinarily left to clinical study. The attempt was made, in other words, to bring methods of traditional social psychology into the service of theories and concepts from the newer dy- namic theory of personality and in so doing to make "depth psychological" phenomena more amenable to mass-statistical treatment, and to make quan- titative surveys of attitudes and opinions more meaningful psychologically.
In the attempt to integrate clinical and group studies, the two were car- ried on in dose conjunction. When the individual was in the focus of atten- tion, the aim was to describe in detail his pattern of opinions, attitudes, and values and to understand the dynamic factors underlying it, and on this basis to design significant questions for use with groups of subjects. When the group was in the focus of attention, the aim was to discover what opinions, attitudes, and values commonly go together and what patterns of factors in the life histories and in the contemporary situations of the subjects were commonly associated with each ideological constellation; this afforded a basis on which to select individuals for more intensive study: commanding first attention were those who exemplified the common patterns and in whom it could be supposed that the correlated factors were dynamically related.
In order to study potentially antidemocratic individuals it was necessary first to identify them. Hence a start was made by constructing a question- naire and having it filled out anonymously by a large group of people. This questionnaire contained, in addition to numerous questions of fact about the subject's past and present life, a variety of antidemocratic statements with which the subjects were invited to agree or disagree. A number of individuals who showed the greatest amount of agreement with these state-
? INTRODUCTION
13
ments-and, by way of contrast, some who showed the most disagreement or, in some instances, were most neutral-were then studied by means of interviews and other clinical techniques. On the basis of these individual studies the questionnaire was revised, and the whole procedure repeated.
The interview was used in part as a check upon the validity of the ques- tionnaire, that is to say, it provided a basis for judging whether people who obtained the highest antidemocratic scores on the questionnaire were usually those who, in a confidential relationship with another person, expressed anti- democratic sentiments with the most intensity. What was more important, however, the clinical studies gave access to the deeper personality factors behind antidemocratic ideology and suggested the means for their investi- gation on a mass scale. With increasing knowledge of the underlying trends of which prejudice was an expression, there was increasing familiarity with various other signs or manifestations by which these trends could be recog- nized. The task then was to translate these manifestations into questionnaire items for use in the next group study. Progress lay in finding more and more reliable indications of the central personality forces and in showing with increasing clarity the relations of these forces to antidemocratic ideological expression.
2. THE TECHNIQUES
The questionnaires and clinical techniques employed in the study may be described briefly as follows:
a. THE QuEsTioNNAIRE METHOD. The questionnaires were always pre- sented in mimeographed form and filled out anonymously by subjects in groups. Each questionnaire included (I) factual questions, (2) opinion- attitude scales, and (3) "projective" (open answer) questions.
I. The factual questions had to do mainly with past and present group memberships: church preference and attendance, political party, vocation, income, and so on. It was assumed that the answers could be taken at their face value. In selecting the questions, we were guided at the start by hypoth- eses concerning the sociological correlates of ideology; as the study pro- gressed we depended more and more upon experience with interviewees.
2. Opinion-attitude scales were used from the start in order to obtain quan- titative estimates of certain surface ideological trends: anti-Semitism, ethno- centrism, politico-economic conservatism. Later, a scale was developed for the measurement of antidemocratic tendencies in the personality itself.
Each scale was a collection of statements, with each of which the subject was asked to express the degree of his agreement or disagreement. Each statement concerned some relatively specific opinion, attitude, or value, and the basis for grouping them within a particular scale was the conception that taken together they expressed a single general trend.
? 14 THE AUTHORITARIAN PERSONALITY
The general trends to which the scales pertained were conceived very broadly, as complex systems of thought about wide areas of social living. To define these trends empirically it was necessary to obtain responses to many specific issues-enough to "cover" the area mapped out conceptually-
and to show that each of them bore some relation to the whole.
This approach stands in contrast to the public opinion poll: whereas the poll is interested primarily in the distribution of opinion with respect to a particular issue, the present interest was to inquire, concerning a particular opinion, with what other opinions and attitudes it was related. The plan was to determine the existence of broad ideological trends, to develop instruments for their measurement, and then to inquire about their distribution within
larger populations.
The approach to an ideological area was to appraise its grosser features
first and its finer or more specific features later. The aim was to gain a view of the "over-all picture" into which smaller features might later be fitted, rather than to obtain highly precise measures of small details in the hope that these might eventually add up to something significant. Although this emphasis upon breadth and inclusiveness prevented the attainment of the highest degree of precision in measurement, it was nevertheless possible to develop each scale to a point where it met the currently accepted statistical standards.
Since each scale had to cover a broad area, without growing so long as to try the patience of the subjects, it was necessary to achieve a high degree of efficiency. The task was to formulate items which would cover as much as possible of the many-sided phenomenon in question. Since each of the trends to be measured was conceived as having numerous components or aspects, there could be no duplication of items; instead it was required that each item express a different feature-and where possible, several features- of the total system. The degree to which items within a scale will "hang together" statistically, and thus give evidence that a single, unified trait is being measured, depends primarily upon the surface similarity of the items- the degree to which they all say the same thing. The present items, obviously, could not be expected to cohere in this fashion; all that could be required statistically of them was that they correlate to a reasonable degree with the total scale. Conceivably, a single component of one of the present systems could be regarded as itself a relatively general trend, the precise measure- ment of which would require the use of numerous more specific items. As indicated above, however, such concern with highly specific, statistically
"pure" factors was put aside, in favor of an attempt to gain a dependable estimate of an over-all system, one which could then be related to other over-all systems in an approach to the totality of major trends within the individual.
One might inquire why, if we wish to knO\v the intensity of some ideolog-
? INTRODUCTION
ical pattern-such as anti-Semitism-within the individual, we do not ask him directly, after defining what we mean. The answer, in part, is that the phe- nomenon to be measured is so complex that a single response would not go very far toward revealing the important differences among individuals. Moreover, anti-Semitism, ethnocentrism, and politico-economic reactionism or radicalism are topics about which many people are not prepared to speak with complete frankness. Thus, even at this surface ideological level it was necessary to employ a certain amount of indirectness. Subjects were never told what was the particular concern of the questionnaire, but only that they were taking part in a "survey of opinions about various issues of the day. " To support this view of the proceedings, items belonging to a partic- ular scale were interspersed with items from other scales in the questionnaire. It was not possible, of course, to avoid statements prejudicial to minority groups, but care \Vas taken in each case to allow the subject "a way out," that is to say, to make it possible for him to agree with such a statement while maintaining the belief that he was not "prejudiced" or "undemocratic. "
Whereas the scales for measuring surface ideological trends conform, in general, with common practice in sociopsychological research, the scale for measuring potentially antidemocratic trends in the personality represents a new departure. The procedure was to bring together in a scale items which, by hypothesis and by clinical experience, could be regarded as "giveaways" of trends which lay relatively deep within the personality, and which con- stituted a disposition to express spontaneously (on a suitable occasion), or to be influenced by, fascist ideas.
The statements in this scale were not different in form from those which made up the surface ideology scales; they were direct expressions of opinion, of attitudes, or of value with respect to various areas of social living-but areas not usually touched upon in systematic presentations of a politico- socioeconomic point of view. Always interspersed with statements from other scales, they conveyed little or nothing to the subject as to the nature of the real question being pursued. They were, in the main,_ statements so designed as to serve as rationalizations for irrational tendencies. Two state- ments included in this scale were the following: (a) "Nowadays with so many different kinds of people moving around so much and mixing together so freely, one has to be especially careful to protect himself against infection and disease" and (b) "Homosexuality is an especially rotten form of delin- quency and ought to be severely punished. " That people who agree with one of these statements show a tendency to agree with the other, and that people who agree with these two statements tend to agree with open anti- democratic statements, e. g. , that members of some minority group are basic- ally inferior, is hardly to be explained on the basis of any obvious logical relation among the statements. It seems necessary, rather, to conceive of some underlying central trend which expresses itself in these different ways.
? THE AUTHORITARIAN PERSONALITY
Different people might, of course, give the same response to a statement such as the above for different reasons; since it was necessary to give the statements at least a veneer of rationality, it was natural to expect that the responses of some people would be determined almost entirely by the rational aspect rather than by some underlying emotional disposition. For this reason it was necessary to include a large number of scale items and to be guided by the general trend of response rather than by the response to a single statement; for a person to be considered potentially antidemocratic in his underlying dynamic structure, he had to agree with a majority of these scale items.
The development of the present scale proceeded in two ways: first, by finding or formulating items which, though they had no manifest connec- tion with open antidemocratic expressions, were nevertheless highly cor- related with them; and second, by demonstrating that these "indirect" items were actually expressions of antidemocratic potential within the personality as known from intensive clinical study.
3? Projective Questions, like most other projective techniques, present the subject with ambiguous and emotionally toned stimulus material. This ma- terial is designed to allow a maximum of variation in response from one subject to another and to provide channels through which relatively deep personality processes may be expressed. The questions are not ambiguous in their formal structure, but in the sense that the answers are at the level of emotional expression rather than at the level of fact and the subject is not aware of their implications. The responses always have to be interpreted, and their significance is known when their meaningful relations to other psychological facts about the subject have been demonstrated. One projec- tive question was, "What would you do if you had only six months to live, and could do anything you wanted? " An answer to this question was not regarded as a statement of what the subject would probably do in actuality, but rather an expression having to do with his values, conflicts, and the like. We asked ourselves if this expression was not in keeping with those elicited by other projective questions and by statements in the personality scale.
Numerous projective questions were tried in the early stages of the study, and from among them eight were selected for use with most of the larger groups of subjects: they were the questions which taken together gave the broadest view of the subject's personality trends and correlated most highly with surface ideological patterns.
b. CLINICAL TECHNIQUES. 1. The interview was divided roughly into an ideological section and a clinical-genetic section. In the first section the aim was to induce the subject to talk as spontaneously and as freely as possible about various broad ideological topics: politics, religion, minority groups,
? INTRODUCTION
income, and vocation. Whereas in the questionnaire the subject was limited to the topics there presented and could express himself only by means of the rating scheme offered, here it was important to know what topics he would bring up of his own accord and with what intensity of feeling he would spontaneously express himself. As indicated above, this material af- forded a means for insuring that the questionnaire, in its revised forms, more or less faithfully represented "what people were saying"-the topics that were on their minds and the forms of expression that came spontaneously to them-and provided a valid index of antidemocratic trends. The interview covered, of course, a much wider variety of topics, and permitted the ex- pression of more elaborated and differentiated opinions, attitudes, and values, than did the questionnaire. Whereas the attempt was made to distill from the interview material what seemed to be of the most general significance and to arrange it for inclusion in the questionnaire, there was material left over to be exploited by means of individual case studies, qualitative analyses, and crudely quantitative studies of the interview material by itself.
The clinical-genetic section of the interview sought to obtain, first, more factual material about the subject's contemporary situation and about his past than could be got from the questionnaire; second, the freest possible expressions of personal feelings, of beliefs, wishes, and fears concerning him- self and his situation and concerning such topics asparents, siblings, friends, and sexual relationships; and third, the subject's conceptions of his childhood environment and of his childhood self.
The interview was conducted in such a way that the material gained from it would permit inferences about the deeper layers of the subject's person- ality. The technique of the interview will be described in detail later. Suffice it to say here that it followed the general pattern of a psychiatric interview that is inspired by a dynamic theory of personality. The interviewer was aided by a comprehensive interview schedule which underwent several revisions during the course of the study, as experience taught what were the most significant underlying questions and what were the most efficient means for evoking material bearing upon them.
The interview material was used for estimation of certain common vari- ables lying within the theoretical framework of the study but not accessible to the other techniques. Interview material also provided the main basis for individual case studies, bearing upon the interrelationships among all the significant factors operating within the antidemocratic individual.
2. The Thematic Apperception Test is a well-known projective technique in which the subject is presented with a series of dramatic pictures and asked to tell a story about each of them. The material he produces can, when inter- preted, reveal a great deal about his underlying wishes, conflicts, and mech- anisms of defense. The technique was modified slightly to suit the present
? THE AUTHORIT ARIAN PERSO~ALITY
purposes. The material was analyzed quantitatively in terms of psychological variables which are found widely in the population and which 'vere readily brought into relation with other variables of the study.
As a part of the case study of an individual an analysis in terms of more unique personality vari- ables was made, the material here being considered in close conjunction with findings from the interview.
Though designed to approach different aspects of the person, the several techniques actually were closely related conceptually one to another. All of them permitted quantification and interpretation in terms of variables which fall within a unified theoretical system. Sometimes two techniques yielded measures of the same variables, and sometimes different techniques were focused upon different variables. In the former case the one technique gave some indication of the validity of the other; in the latter case the adequacy of a technique could be gauged by its ability to produce measures that were meaningfully related to all the others. Whereas a certain amount of repeti- tion was necessary to insure validation, the main aim was to fill out a broad framework and achieve a maximum of scope.
The theoretical approach required in each case either that a new technique be designed from the ground up or that an existing one be modified to suit the particular purpose. At the start, there was a theoretical conception of what was to be measured and certain sources-to be described later-which could be drawn upon in devising the original questionnaire form and the preliminary interview schedule. Each technique then evolved as the study progressed. Since each was designed specifically for this study, they could be changed at will as understanding increased, and since an important pur- pose of the study was the development and testing of effective instruments for diagnosing potential? fascism, there was no compulsion to repeat without modification a procedure just in order to accumulate comparable data. So closely interrelated were the techniques that what was learned from any one of them could be applied to the improvement of any other. Just as the clinical techniques provided a basis for enriching the several parts of the questionnaire, so did the accumulating quantitative results indicate what ought to be concentrated upon in the interview; and just as the analysis of scale data suggested the existence of underlying variables which might be approached by means of projective techniques, so did the responses on projective techniques suggest items for inclusion in the scales.
The evolution of techniques was expressed both in expansion and in con- traction. Expansion was exemplified in the attempt to bring more and more aspects of antidemocratic ideology into the developing picture and in the attempt to explore enough aspects of the potentially antidemocratic per- sonality so that there was some grasp of the totality. Contraction took place. continuously in the quantitative procedures as increasing theoretical clarity
? INTRODUCTION
permitted a boiling dowrt so that the same crucial relationships could be demonstrated with briefer techniques.
C. PROCEDURES IN THE COLLECTION OF DA T A
I. THE GROUPS STUDIED
a. THE BEGINNING WITH CoLLEGE STUDENTS. There were enough prac- tical reasons alone to determine that the present study, which at the begin- ning had limited resources and limited objectives, should start with college students as research subjects: they were available for the asking, whether singly or in groups, they would C<_)operate willingly, and they could be reached for retesting without much difficulty. At the same time, other con- siderations favored the use of college students in a study of ideology. In the first place, the intellectual and educational level is high enough so that there needed to be relatively little restriction with respect to the number and nature of issues that might be raised-a very important matter in a study that emphasized breadth and inclusiveness. One could be fairly certain that col- lege students had opinions about most of the various topics to be considered. In the second place, there could be relative certainty that all the subjects understood the terms of the questions in the same way and th~t the same responses had uniform significance. In the third place, however large a population one might be able to sample he would probably find that most
of his generalizations had in any case to be limited to various relatively homogeneous subclassifications of the total group studied; college students form one group that is relatively quite homogeneous with respect to factors that might be expected to influence ideology. And they represent an im- portant sector of the population, both through their family connections and through their prospective leadership in the community.
It is obvious, however, that a study which used only college students as subjects would be seriously limited in its general significance. Of what larger population could a group of students at a state university be regarded as an adequate sample? Would findings on this sample hold for all the stu- dents at this university? For college students generally? For young people of the middle class? It depends upon what kind of generalization is to be made. Generalizations about the distribution of particular opinions or about the average amount of agreement with this or that statement-the kind of in- formation sought in poll studies-could hardly go beyond the students at the university where the survey was made. Results from an Eastern uni- versity or from a privately endowed institution might be quite different. The present concern, however, was not so much with questions of dis- tribution as with questions of relationship. For example, there was less
? 20 THE AUTHORITARIAN PERSONALITY
interest in what per cent of the general population would agree that "labor unions have grown too powerful" and that "there are too many Jews in government agencies" than in whether or not there was a general relation- ship between these two opinions. For the study of how opinions, attitudes, and values are organized within the individual, college students had a great deal to offer, particularly in the early stages of the work where the emphasis was upon improving techniques and obtaining first approximations of gen- eral relationships. This work could proceed without hindrance so long as the factors to be studied were present, and varied sufficiently widely from one individual to another. In this regard, the limitations of the college sample were that the relatively high intellectual and educational level de- creased the number of extremely prejudiced individuals, and that some of the factors which were presumed to influence prejudice were rarely or never present.
These considerations made it necessary to study various other groups of subjects. As it turned out, the strength of the various ideological trends was found to vary widely from one group to another, while the relationships found in the college group were very similar to those found elsewhere.
b. THE GENERAL NoNcoLLEGE PoPULATION FROM WHICH OuR SuBJECTs WERE DRAWN. When it became possible through increased resources to expand the scope of the study, there began an attempt to obtain as subjects a wide variety of adult Americans. The aim was to examine people who pos- sessed in different degrees as many as possible of the sociological variables presumed to be relevant to the study-political, religious, occupational, in- come, and social group memberships. A list of all the groups (college and noncollege) from whom questionnaires were collected is given in Table I (I).
The group within which a subject was functioning at the time he filled out the questionnaire was, of course, not necessarily the most important or representative of the various groups to which he belonged. The questionnaire itself was relied upon to give information about the group memberships deemed most relevant to the study, and subjects could be categorized on this basis regardless of the group through which the questionnaires were collected.
The emphasis throughout was upon obtaining different kinds of subjects, enough to insure wide variability of opinion and attitude and adequate coverage of the factors supposed to influence ideology. The subjects are in no sense a random sample of the noncollege population nor, since there was no attempt to make a sociological analysis of the community in which they lived, can they be regarded as a representative sample. The progress of the study was not in the direction of broadening the basis for generalization about larger populations, but rather toward the more intensive investig-ation
? INTRODUCTION 21
TABLE I(l)
GRoUPS FRoM WHOM QuESTIONNAIRES WERE CoLLECTEDa
I. Form 78 (January to May, I945)
University of California Public Speaking Class Women . . . . . . . . . . . . I40 UniversityofCaliforniaPublicSpeakingClassMen . . . . . . . . . . . . . . . 52 University of California Extension Psychology Class (adult women). 40
? Professional Women (public school teachers, social workers, public healthnurses) (SanFrancisco area) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Jrotal 295
II. Form 6o (Summer, I945)
UniversityofOregonStudentWomen . . . . . . . . . . . . . . . . . . . . . . . . . . 47 University of Oregon and University of California Student Women. 54 University of Oregon and University of California Student Men . . . . 57 Oregon. Serv_ice Club Men (Kiwanis, Lions, Rotary Clubs) (Total
quesnonnarre) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Oregon Service Club Men (Form A only)" . . . . . . . .
situations, to explain the persistence of ideological trends in the face of contradicting facts and radically altered social conditions, to explain why people in the same sociological situation have different or even conflicting views on social issues, and why it is that people whose behavior has been changed through psychological manipulation lapse into their old ways as soon as the agencies of manipulation are removed.
The conception of personality structure is the best safeguard against the inclination to attribute persistent trends in the individual to something "innate" or "basic" or "racial" within him. The Nazi allegation that natural, biological traits decide the total being of a person would not have been such a successful political device had it not been possible to point to numerous instances of relative fixity in human behavior and to challenge those who thought to explain them on any basis other than a biological one. Without the conception of personality structure, writers whose approach rests upon the assumption of infinite human flexibility and responsiveness to the social situation of the moment have not helped matters by referring persistent
(
? INTRODUCTION
7
trends which they could not aP. prove to "confusion" or "psychosis" or evil under one name or another. There is, of course, some basis for describing as "pathological" patte~ of behavior which do not conform with the most common, and seemingly most lawful, responses to momentary stimuli. But this is to use the term pathological in the very narrow sense of deviation from the average found in a particular context and, what is worse, to suggest that everything in the personality structure is to be put under this heading. Actually, personality embraces variables which exist widely in the popula- tion and have lawful relations one to another. Personality patterns that have been dismissed as "pathological" because they were not in keeping with the most common manifest trends or the most dominant ideals within a society, have on closer investigation turned out to be but exaggerations of what was almost universal below the surface in that society. What is "pathological" today may with changing social conditions become the dominant trend of tomorrow.
It seems clear then that an adequate approach to the problems before us must take into account both fixity and flexibility; it must regard the two not as mutually exclusive categories but as the extremes of a single continuum along which human characteristics may be placed, and it must provide a basis for understanding the conditions which favor the one extreme or the other. Personality is a concept to account for relative permanence. But it may be emphasized again that personality is mainly a potential; it is a readi- ness for behavior rather than behavior itself; although it consists in disposi- tions to behave in certain ways, the behavior that actually occurs will always depend upon the objective situation. Where the concern is with antidemo- cratic trends, a delineation of the conditions for individual expression re- quires an understanding of the total organization of society.
It has been stated that the personality structure may be such as to render the individual susceptible to antidemocratic propaganda. It may now be asked what are the conditions under which such propaganda would increase in pitch and volume and come to dominate in press and radio to the exclusion of contrary ideological stimuli, so that what is now potential would become actively manifest. The answer must be sought not in any single personality nor in personality factors found in the mass of people, but in processes at work in society itself. It seems well understood today that whether or not antidemocratic propaganda is to become a dominant force in this country depends primarily upon the situation of the most powerful economic inter- ests, upon whether they, by conscious design or not, make use of this device for maintaining their dominant status. This is a matter about which the great majority of people would have little to say.
The present research, limited as it is to the hitherto largely neglected psychological aspects of fascism, does not concern itself with the production of propaganda. It focuses attention, rather, upon the consumer, the indi-
? 8 THE AUTHORITARIAN PERSONALITY
vidual for whom the propaganda is designed. In so doing it attempts to take into account not only the psychological structure of the individual but the total objective situation in which he lives. It makes the assumption that people in general tend to accept political and social programs which they believe will serve their economic interests. What these interests are depends in each case upon the individual's position in society as defined in economic and sociological terms. An important part of the present research, therefore, was the attempt to discover what patterns of socioeconomic factors are asso- ciated with receptivity, and with resistance, to antidemocratic propaganda.
At the same time, however, it was considered that economic motives in the individual may not have the dominant and crucial role that is often ascribed to them. If economic self-interest were the only determinant of opinion, we should expect people of the same socioeconomic status to have very similar opinions, and we should expect opinion to vary in a meaningful way from one socioeconomic grouping to another. Research has not given very sound support for these expectations. There is only the most general similarity of opinion among people of the same socioeconomic status, and the exceptions are glaring; while variations from one socioeconomic group to another are rarely simple or clear-cut. To explain why it is that people of the same socioeconomic status so frequently have different ideologies, while people of a different status often have very similar ideologies, we must take account of other than purely economic needs.
More than this, it is becoming increasingly plain that people very fre- quently do not behave in such a way as to further their material interests, even when it is clear to them what these interests are. The resistance of white-collar workers to organization is not due to a belief that the union will not help them economically; the tendency of the small businessman to side with big business in most economic and political matters cannot be due entirely to a belief that this is the way to guarantee his economic indepen- dence. In instances such as these the individual seems not only not to con- sider his material interests, but even to go against them. It is as if he were thinking in terms of a larger group identification, as if his point of view were determined more by his need to support this group and to suppress opposite ones than by rational consideration of his own interests. Indeed, it is with a sense of relief today that one is assured that a group conflict is merely a clash of economic interests-that each side is merely out to "do" the other- and not a struggle in which deep-lying emotional drives have been let loose. When it comes to the ways in which people appraise the social world, irra- tional trends stand out glaringly. One may conceive of a professional man who opposes the immigration of Jewish refugees on the ground that this will increase the competition with which he has to deal and so decrease his income. However undemocratic this may be, it is at least rational in a limited sense. But for this man to go on, as do most people who oppose Jews on
? INTRODUCTION
9
occupational grounds, and accept a wide variety of opinions, many of which are contradictory, about Jews in general, and to attribute various ills of the world to them, is plainly illogical. And it is just as illogical to praise all Jews in accordance with a "good" stereotype of them. Hostility against groups that is based upon real frustration, brought about by members of that group, undoubtedly exists, but such frustrating experiences can hardly ac<;ount for the fact that prejudice is apt to be generalized. Evidence from the present study confirms what has often been indicated: that a man who is hostile toward one minority group is very likely to be hostile against a wide variety of others. There is no conceivable rational basis for such generalization; and, what is more striking, prejudice against, or totally uncritical acceptance of,
a particular group often exists in the absence of any experience with mem- bers of that group. The objective situation of the individual seems an unlikely source of such irrationality; rather we should seek where psychology has already found the sources of dreams, fantasies, and misinterpretations of the world-that is, in the deep-lying needs of the personality.
Another aspect of the individual's situation which we should expect to affect his ideological receptivity is his membership in social groups-occu- pational, fraternal, religious, and the like. For historical and sociological reasons, such groups favor and promulgate, whether officially or unofficially, different patterns of ideas. There is reason to believe that individuals, out of their needs to conform and to belong and to believe and through such deviCes as imitation and conditioning, often take over more or less ready-made the opinions, attitudes, and values that are characteristic of the groups in which they have membership. To the extent that the ideas which prevail in such a group are implicitly or explicitly antidemocratic, the individual group mem-
ber might be expected to be receptive to propaganda having the same general direction. Accordingly, the present research investigates a variety of group memberships with a view to what general trends of thought-and how much variability-might be found in each.
It is recognized, however, that a correlation between group membership and ideology may be due to different kinds of determination in different individuals. In some cases it might be that the individual merely repeats opinions which are taken for granted in his social milieu and which he has no reason to question; in other cases it might be that the individual has chosen to join a particular group because it stood for ideals with which he was already in sympathy. In modem society, despite enormous communality in basic culture, it is rare for a person to be subjected to only one pattern of ideas, after he is old enough for ideas to mean something to him. Some selec- tion is usually made, according, it may be supposed, to the needs of his personality. Even when individuals are exposed during their formative years almost exclusively to a single, closely knit pattern of political, economic, social, and religious idei(s, it: is found that some confor! ll while others rebel,
? 10 THE AUTHORITARIAN PERSONALITY
and it seems proper to inquire whether personality factors do not make the difference. The soundest approach, it would seem, is to consider that in the determination of ideology, as in the determination of any behavior, there is a situational factor and a personality factor, and that a careful weighing of the role of each'will yield the most accurate prediction.
Situational factors, chiefly economic condition and social group member- ships, have been studied intensively in recent researches on opinion and atti- tude, while the more inward, more individualistic factors have not received the attention they deserve. Beyond this, there is still another reason why the present study places particular emphasis upon the personality. Fascism, in order to be successful as a political movement, must have a mass basis. It must secure not only the frightened submission but the active cooperation of the great majority of the people. Since by its very nature it favors the few at the expense of the many, it cannot possibly demonstrate that it will so improve the situation of most people that their real interests will be served. It must therefore make its major appeal, not to rational self-interest, but to emotional needs-often to the most primitive and irrational wishes and fears. If it be argued that fascist propaganda fools people into believing that their lot will be improved, then the question arises: Why are they so easily fooled? Because, it may be supposed, of their personality structure; because of long-
. established patterns of hopes and a~pirations, fears and anxieties that dispose them to certain beliefs and make them resistant to others. The task of fascist propaganda, in other words, is rendered easier to the degree that antidemo- cratic potentials already exist in the great mass of people. It may be granted that in Germany economic conflicts and dislocations within the society were such that for this reason alone the triumph of fascism was sooner or later inevitable; but the Nazi leaders did not act as if they believed this to be so; instead they acted as if it were necessary at every moment to take into account the psychology of the people-to activate every ounce of their anti- democratic potential, to compromise with them, to stamp out the slightest spark of rebellion. It seems apparent that any attempt to appraise the chances of a fascist triumph in America must reckon with the potential existing in the character of the people. Here lies not only the susceptibility to antidemo- cratic propaganda but the most dependable sources of resistance to it.
The present writers believe that it is up to the people to decide whether or not this country goes fascist. It is assumed that knowledge of the nature and extent of antidemocratic potentials will indicate programs for demo- cratic action. These programs should not be limited to devices for manipu- lating people in such a way that they will behave more democratically, but they should be devoted to increasing the kind of self-awareness and self- determination that makes any kind of manipulation impossible. There is one explanation for the existence of an individual's ideology that has not so far been considered: that it is the view of the world which a reasonable man,
? INTRODUCTION
with some understanding of the role of such determinants as those discussed above, and with complete access to the necessary facts, will organize for himself. This conception, though it has been left to the last, is of crucial importance for a sound approach to ideology. Without it we should have to share the destructive view, which has gained some acceptance in the modern world, that since all ideologies, all philosophies, derive from non- rational sources there is no basis for saying that one has more merit than another.
But the rational system of an objective and thoughtful man is not a thing apart from personality. Such a system is still motivated. What is distinguish- ing in its sources is mainly the kind of personality organization from which it springs. It might be said that a mature personality (if we may for the moment use this term without defining it) will come closer to achieving a rational system of thought than will an immature one; but a personality is no less dynamic and no less organized for being mature, and the task of describing the structure of this personality is not different in kind from the task of describing any other personality. According to theory, the person- ality variables which have most to do with determining the objectivity and rationality of an ideology are those which belong to the ego, that part of the personality which appreciates reality, integrates the other parts, and operates with the most conscious awareness.
It is the ego that becomes aware of and takes responsibility for nonra- tional forces operating within the personality. This is the basis for our belief that the object of knowing what are the psychological determinants of ideology is that men can become more reasonable. It is not supposed, of course, that this will eliminate differences of opinion. The world is suffi- ciently complex and difficult to know, men have enough real interests that are in conflict with the real interests of other men, there are enough ego- accepted differences in personality to insure that arguments about politics, economics, and religion will never grow dull. Knowledge of the psycholog- ical determinants of ideology cannot tell us what is the truest ideology; it can only remove some of the barriers in the way of its pursuit.
B. METHODOLOGY
1. GENERAL CHARACTERISTICS OF THE METHOD
To attack the problems conceptualized above required methods for de- scribing and measuring ideological trends and methods for exposing person- ality, the contemporary situation, and the social background. A particular methodologicar challenge was imposed by the conception of levels in the person; this made it necessary to devise techniqu~s for surveying opinions, attitudes, and values that were on the surface, for revealing ideological
r -l
? ii. THE AUTHORITARiAN PERSONALITY
trends that were more or less inhibited and reached the surface only in indirect manifestations, and for bringing to light personality forces that lay in the subject's unconscious. And since the major concern was with patterns of dynamically related factors-something that requires study of the total individual-it seemed that the proper approach was through intensive clinical studies. The significance and practical importance of such studies could not be gauged, however, until there was knowledge of how far it was possible to generalize from them. Thus it was necessary to perform group studies as well as individual studies, and to find ways and means for integrating the two.
Individuals were studied by means of interviews and special clinical tech- niques for revealing underlying wishes, fears, and defenses; groups were studied by means of questionnaires. It was not expected that the clinical studies would be as complete or profound as some which have already been performed, primarily by psychoanalysts, nor that the questionnaires would be more accurate than any now employed by social psychologists. It was hoped, however-indeed it was necessary to our purpose-that the clinical material could be conceptualized in such a way as to permit its being quan- tified and carried over into group studies, and that the questionnaires could be brought to bear upon areas of response ordinarily left to clinical study. The attempt was made, in other words, to bring methods of traditional social psychology into the service of theories and concepts from the newer dy- namic theory of personality and in so doing to make "depth psychological" phenomena more amenable to mass-statistical treatment, and to make quan- titative surveys of attitudes and opinions more meaningful psychologically.
In the attempt to integrate clinical and group studies, the two were car- ried on in dose conjunction. When the individual was in the focus of atten- tion, the aim was to describe in detail his pattern of opinions, attitudes, and values and to understand the dynamic factors underlying it, and on this basis to design significant questions for use with groups of subjects. When the group was in the focus of attention, the aim was to discover what opinions, attitudes, and values commonly go together and what patterns of factors in the life histories and in the contemporary situations of the subjects were commonly associated with each ideological constellation; this afforded a basis on which to select individuals for more intensive study: commanding first attention were those who exemplified the common patterns and in whom it could be supposed that the correlated factors were dynamically related.
In order to study potentially antidemocratic individuals it was necessary first to identify them. Hence a start was made by constructing a question- naire and having it filled out anonymously by a large group of people. This questionnaire contained, in addition to numerous questions of fact about the subject's past and present life, a variety of antidemocratic statements with which the subjects were invited to agree or disagree. A number of individuals who showed the greatest amount of agreement with these state-
? INTRODUCTION
13
ments-and, by way of contrast, some who showed the most disagreement or, in some instances, were most neutral-were then studied by means of interviews and other clinical techniques. On the basis of these individual studies the questionnaire was revised, and the whole procedure repeated.
The interview was used in part as a check upon the validity of the ques- tionnaire, that is to say, it provided a basis for judging whether people who obtained the highest antidemocratic scores on the questionnaire were usually those who, in a confidential relationship with another person, expressed anti- democratic sentiments with the most intensity. What was more important, however, the clinical studies gave access to the deeper personality factors behind antidemocratic ideology and suggested the means for their investi- gation on a mass scale. With increasing knowledge of the underlying trends of which prejudice was an expression, there was increasing familiarity with various other signs or manifestations by which these trends could be recog- nized. The task then was to translate these manifestations into questionnaire items for use in the next group study. Progress lay in finding more and more reliable indications of the central personality forces and in showing with increasing clarity the relations of these forces to antidemocratic ideological expression.
2. THE TECHNIQUES
The questionnaires and clinical techniques employed in the study may be described briefly as follows:
a. THE QuEsTioNNAIRE METHOD. The questionnaires were always pre- sented in mimeographed form and filled out anonymously by subjects in groups. Each questionnaire included (I) factual questions, (2) opinion- attitude scales, and (3) "projective" (open answer) questions.
I. The factual questions had to do mainly with past and present group memberships: church preference and attendance, political party, vocation, income, and so on. It was assumed that the answers could be taken at their face value. In selecting the questions, we were guided at the start by hypoth- eses concerning the sociological correlates of ideology; as the study pro- gressed we depended more and more upon experience with interviewees.
2. Opinion-attitude scales were used from the start in order to obtain quan- titative estimates of certain surface ideological trends: anti-Semitism, ethno- centrism, politico-economic conservatism. Later, a scale was developed for the measurement of antidemocratic tendencies in the personality itself.
Each scale was a collection of statements, with each of which the subject was asked to express the degree of his agreement or disagreement. Each statement concerned some relatively specific opinion, attitude, or value, and the basis for grouping them within a particular scale was the conception that taken together they expressed a single general trend.
? 14 THE AUTHORITARIAN PERSONALITY
The general trends to which the scales pertained were conceived very broadly, as complex systems of thought about wide areas of social living. To define these trends empirically it was necessary to obtain responses to many specific issues-enough to "cover" the area mapped out conceptually-
and to show that each of them bore some relation to the whole.
This approach stands in contrast to the public opinion poll: whereas the poll is interested primarily in the distribution of opinion with respect to a particular issue, the present interest was to inquire, concerning a particular opinion, with what other opinions and attitudes it was related. The plan was to determine the existence of broad ideological trends, to develop instruments for their measurement, and then to inquire about their distribution within
larger populations.
The approach to an ideological area was to appraise its grosser features
first and its finer or more specific features later. The aim was to gain a view of the "over-all picture" into which smaller features might later be fitted, rather than to obtain highly precise measures of small details in the hope that these might eventually add up to something significant. Although this emphasis upon breadth and inclusiveness prevented the attainment of the highest degree of precision in measurement, it was nevertheless possible to develop each scale to a point where it met the currently accepted statistical standards.
Since each scale had to cover a broad area, without growing so long as to try the patience of the subjects, it was necessary to achieve a high degree of efficiency. The task was to formulate items which would cover as much as possible of the many-sided phenomenon in question. Since each of the trends to be measured was conceived as having numerous components or aspects, there could be no duplication of items; instead it was required that each item express a different feature-and where possible, several features- of the total system. The degree to which items within a scale will "hang together" statistically, and thus give evidence that a single, unified trait is being measured, depends primarily upon the surface similarity of the items- the degree to which they all say the same thing. The present items, obviously, could not be expected to cohere in this fashion; all that could be required statistically of them was that they correlate to a reasonable degree with the total scale. Conceivably, a single component of one of the present systems could be regarded as itself a relatively general trend, the precise measure- ment of which would require the use of numerous more specific items. As indicated above, however, such concern with highly specific, statistically
"pure" factors was put aside, in favor of an attempt to gain a dependable estimate of an over-all system, one which could then be related to other over-all systems in an approach to the totality of major trends within the individual.
One might inquire why, if we wish to knO\v the intensity of some ideolog-
? INTRODUCTION
ical pattern-such as anti-Semitism-within the individual, we do not ask him directly, after defining what we mean. The answer, in part, is that the phe- nomenon to be measured is so complex that a single response would not go very far toward revealing the important differences among individuals. Moreover, anti-Semitism, ethnocentrism, and politico-economic reactionism or radicalism are topics about which many people are not prepared to speak with complete frankness. Thus, even at this surface ideological level it was necessary to employ a certain amount of indirectness. Subjects were never told what was the particular concern of the questionnaire, but only that they were taking part in a "survey of opinions about various issues of the day. " To support this view of the proceedings, items belonging to a partic- ular scale were interspersed with items from other scales in the questionnaire. It was not possible, of course, to avoid statements prejudicial to minority groups, but care \Vas taken in each case to allow the subject "a way out," that is to say, to make it possible for him to agree with such a statement while maintaining the belief that he was not "prejudiced" or "undemocratic. "
Whereas the scales for measuring surface ideological trends conform, in general, with common practice in sociopsychological research, the scale for measuring potentially antidemocratic trends in the personality represents a new departure. The procedure was to bring together in a scale items which, by hypothesis and by clinical experience, could be regarded as "giveaways" of trends which lay relatively deep within the personality, and which con- stituted a disposition to express spontaneously (on a suitable occasion), or to be influenced by, fascist ideas.
The statements in this scale were not different in form from those which made up the surface ideology scales; they were direct expressions of opinion, of attitudes, or of value with respect to various areas of social living-but areas not usually touched upon in systematic presentations of a politico- socioeconomic point of view. Always interspersed with statements from other scales, they conveyed little or nothing to the subject as to the nature of the real question being pursued. They were, in the main,_ statements so designed as to serve as rationalizations for irrational tendencies. Two state- ments included in this scale were the following: (a) "Nowadays with so many different kinds of people moving around so much and mixing together so freely, one has to be especially careful to protect himself against infection and disease" and (b) "Homosexuality is an especially rotten form of delin- quency and ought to be severely punished. " That people who agree with one of these statements show a tendency to agree with the other, and that people who agree with these two statements tend to agree with open anti- democratic statements, e. g. , that members of some minority group are basic- ally inferior, is hardly to be explained on the basis of any obvious logical relation among the statements. It seems necessary, rather, to conceive of some underlying central trend which expresses itself in these different ways.
? THE AUTHORITARIAN PERSONALITY
Different people might, of course, give the same response to a statement such as the above for different reasons; since it was necessary to give the statements at least a veneer of rationality, it was natural to expect that the responses of some people would be determined almost entirely by the rational aspect rather than by some underlying emotional disposition. For this reason it was necessary to include a large number of scale items and to be guided by the general trend of response rather than by the response to a single statement; for a person to be considered potentially antidemocratic in his underlying dynamic structure, he had to agree with a majority of these scale items.
The development of the present scale proceeded in two ways: first, by finding or formulating items which, though they had no manifest connec- tion with open antidemocratic expressions, were nevertheless highly cor- related with them; and second, by demonstrating that these "indirect" items were actually expressions of antidemocratic potential within the personality as known from intensive clinical study.
3? Projective Questions, like most other projective techniques, present the subject with ambiguous and emotionally toned stimulus material. This ma- terial is designed to allow a maximum of variation in response from one subject to another and to provide channels through which relatively deep personality processes may be expressed. The questions are not ambiguous in their formal structure, but in the sense that the answers are at the level of emotional expression rather than at the level of fact and the subject is not aware of their implications. The responses always have to be interpreted, and their significance is known when their meaningful relations to other psychological facts about the subject have been demonstrated. One projec- tive question was, "What would you do if you had only six months to live, and could do anything you wanted? " An answer to this question was not regarded as a statement of what the subject would probably do in actuality, but rather an expression having to do with his values, conflicts, and the like. We asked ourselves if this expression was not in keeping with those elicited by other projective questions and by statements in the personality scale.
Numerous projective questions were tried in the early stages of the study, and from among them eight were selected for use with most of the larger groups of subjects: they were the questions which taken together gave the broadest view of the subject's personality trends and correlated most highly with surface ideological patterns.
b. CLINICAL TECHNIQUES. 1. The interview was divided roughly into an ideological section and a clinical-genetic section. In the first section the aim was to induce the subject to talk as spontaneously and as freely as possible about various broad ideological topics: politics, religion, minority groups,
? INTRODUCTION
income, and vocation. Whereas in the questionnaire the subject was limited to the topics there presented and could express himself only by means of the rating scheme offered, here it was important to know what topics he would bring up of his own accord and with what intensity of feeling he would spontaneously express himself. As indicated above, this material af- forded a means for insuring that the questionnaire, in its revised forms, more or less faithfully represented "what people were saying"-the topics that were on their minds and the forms of expression that came spontaneously to them-and provided a valid index of antidemocratic trends. The interview covered, of course, a much wider variety of topics, and permitted the ex- pression of more elaborated and differentiated opinions, attitudes, and values, than did the questionnaire. Whereas the attempt was made to distill from the interview material what seemed to be of the most general significance and to arrange it for inclusion in the questionnaire, there was material left over to be exploited by means of individual case studies, qualitative analyses, and crudely quantitative studies of the interview material by itself.
The clinical-genetic section of the interview sought to obtain, first, more factual material about the subject's contemporary situation and about his past than could be got from the questionnaire; second, the freest possible expressions of personal feelings, of beliefs, wishes, and fears concerning him- self and his situation and concerning such topics asparents, siblings, friends, and sexual relationships; and third, the subject's conceptions of his childhood environment and of his childhood self.
The interview was conducted in such a way that the material gained from it would permit inferences about the deeper layers of the subject's person- ality. The technique of the interview will be described in detail later. Suffice it to say here that it followed the general pattern of a psychiatric interview that is inspired by a dynamic theory of personality. The interviewer was aided by a comprehensive interview schedule which underwent several revisions during the course of the study, as experience taught what were the most significant underlying questions and what were the most efficient means for evoking material bearing upon them.
The interview material was used for estimation of certain common vari- ables lying within the theoretical framework of the study but not accessible to the other techniques. Interview material also provided the main basis for individual case studies, bearing upon the interrelationships among all the significant factors operating within the antidemocratic individual.
2. The Thematic Apperception Test is a well-known projective technique in which the subject is presented with a series of dramatic pictures and asked to tell a story about each of them. The material he produces can, when inter- preted, reveal a great deal about his underlying wishes, conflicts, and mech- anisms of defense. The technique was modified slightly to suit the present
? THE AUTHORIT ARIAN PERSO~ALITY
purposes. The material was analyzed quantitatively in terms of psychological variables which are found widely in the population and which 'vere readily brought into relation with other variables of the study.
As a part of the case study of an individual an analysis in terms of more unique personality vari- ables was made, the material here being considered in close conjunction with findings from the interview.
Though designed to approach different aspects of the person, the several techniques actually were closely related conceptually one to another. All of them permitted quantification and interpretation in terms of variables which fall within a unified theoretical system. Sometimes two techniques yielded measures of the same variables, and sometimes different techniques were focused upon different variables. In the former case the one technique gave some indication of the validity of the other; in the latter case the adequacy of a technique could be gauged by its ability to produce measures that were meaningfully related to all the others. Whereas a certain amount of repeti- tion was necessary to insure validation, the main aim was to fill out a broad framework and achieve a maximum of scope.
The theoretical approach required in each case either that a new technique be designed from the ground up or that an existing one be modified to suit the particular purpose. At the start, there was a theoretical conception of what was to be measured and certain sources-to be described later-which could be drawn upon in devising the original questionnaire form and the preliminary interview schedule. Each technique then evolved as the study progressed. Since each was designed specifically for this study, they could be changed at will as understanding increased, and since an important pur- pose of the study was the development and testing of effective instruments for diagnosing potential? fascism, there was no compulsion to repeat without modification a procedure just in order to accumulate comparable data. So closely interrelated were the techniques that what was learned from any one of them could be applied to the improvement of any other. Just as the clinical techniques provided a basis for enriching the several parts of the questionnaire, so did the accumulating quantitative results indicate what ought to be concentrated upon in the interview; and just as the analysis of scale data suggested the existence of underlying variables which might be approached by means of projective techniques, so did the responses on projective techniques suggest items for inclusion in the scales.
The evolution of techniques was expressed both in expansion and in con- traction. Expansion was exemplified in the attempt to bring more and more aspects of antidemocratic ideology into the developing picture and in the attempt to explore enough aspects of the potentially antidemocratic per- sonality so that there was some grasp of the totality. Contraction took place. continuously in the quantitative procedures as increasing theoretical clarity
? INTRODUCTION
permitted a boiling dowrt so that the same crucial relationships could be demonstrated with briefer techniques.
C. PROCEDURES IN THE COLLECTION OF DA T A
I. THE GROUPS STUDIED
a. THE BEGINNING WITH CoLLEGE STUDENTS. There were enough prac- tical reasons alone to determine that the present study, which at the begin- ning had limited resources and limited objectives, should start with college students as research subjects: they were available for the asking, whether singly or in groups, they would C<_)operate willingly, and they could be reached for retesting without much difficulty. At the same time, other con- siderations favored the use of college students in a study of ideology. In the first place, the intellectual and educational level is high enough so that there needed to be relatively little restriction with respect to the number and nature of issues that might be raised-a very important matter in a study that emphasized breadth and inclusiveness. One could be fairly certain that col- lege students had opinions about most of the various topics to be considered. In the second place, there could be relative certainty that all the subjects understood the terms of the questions in the same way and th~t the same responses had uniform significance. In the third place, however large a population one might be able to sample he would probably find that most
of his generalizations had in any case to be limited to various relatively homogeneous subclassifications of the total group studied; college students form one group that is relatively quite homogeneous with respect to factors that might be expected to influence ideology. And they represent an im- portant sector of the population, both through their family connections and through their prospective leadership in the community.
It is obvious, however, that a study which used only college students as subjects would be seriously limited in its general significance. Of what larger population could a group of students at a state university be regarded as an adequate sample? Would findings on this sample hold for all the stu- dents at this university? For college students generally? For young people of the middle class? It depends upon what kind of generalization is to be made. Generalizations about the distribution of particular opinions or about the average amount of agreement with this or that statement-the kind of in- formation sought in poll studies-could hardly go beyond the students at the university where the survey was made. Results from an Eastern uni- versity or from a privately endowed institution might be quite different. The present concern, however, was not so much with questions of dis- tribution as with questions of relationship. For example, there was less
? 20 THE AUTHORITARIAN PERSONALITY
interest in what per cent of the general population would agree that "labor unions have grown too powerful" and that "there are too many Jews in government agencies" than in whether or not there was a general relation- ship between these two opinions. For the study of how opinions, attitudes, and values are organized within the individual, college students had a great deal to offer, particularly in the early stages of the work where the emphasis was upon improving techniques and obtaining first approximations of gen- eral relationships. This work could proceed without hindrance so long as the factors to be studied were present, and varied sufficiently widely from one individual to another. In this regard, the limitations of the college sample were that the relatively high intellectual and educational level de- creased the number of extremely prejudiced individuals, and that some of the factors which were presumed to influence prejudice were rarely or never present.
These considerations made it necessary to study various other groups of subjects. As it turned out, the strength of the various ideological trends was found to vary widely from one group to another, while the relationships found in the college group were very similar to those found elsewhere.
b. THE GENERAL NoNcoLLEGE PoPULATION FROM WHICH OuR SuBJECTs WERE DRAWN. When it became possible through increased resources to expand the scope of the study, there began an attempt to obtain as subjects a wide variety of adult Americans. The aim was to examine people who pos- sessed in different degrees as many as possible of the sociological variables presumed to be relevant to the study-political, religious, occupational, in- come, and social group memberships. A list of all the groups (college and noncollege) from whom questionnaires were collected is given in Table I (I).
The group within which a subject was functioning at the time he filled out the questionnaire was, of course, not necessarily the most important or representative of the various groups to which he belonged. The questionnaire itself was relied upon to give information about the group memberships deemed most relevant to the study, and subjects could be categorized on this basis regardless of the group through which the questionnaires were collected.
The emphasis throughout was upon obtaining different kinds of subjects, enough to insure wide variability of opinion and attitude and adequate coverage of the factors supposed to influence ideology. The subjects are in no sense a random sample of the noncollege population nor, since there was no attempt to make a sociological analysis of the community in which they lived, can they be regarded as a representative sample. The progress of the study was not in the direction of broadening the basis for generalization about larger populations, but rather toward the more intensive investig-ation
? INTRODUCTION 21
TABLE I(l)
GRoUPS FRoM WHOM QuESTIONNAIRES WERE CoLLECTEDa
I. Form 78 (January to May, I945)
University of California Public Speaking Class Women . . . . . . . . . . . . I40 UniversityofCaliforniaPublicSpeakingClassMen . . . . . . . . . . . . . . . 52 University of California Extension Psychology Class (adult women). 40
? Professional Women (public school teachers, social workers, public healthnurses) (SanFrancisco area) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Jrotal 295
II. Form 6o (Summer, I945)
UniversityofOregonStudentWomen . . . . . . . . . . . . . . . . . . . . . . . . . . 47 University of Oregon and University of California Student Women. 54 University of Oregon and University of California Student Men . . . . 57 Oregon. Serv_ice Club Men (Kiwanis, Lions, Rotary Clubs) (Total
quesnonnarre) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Oregon Service Club Men (Form A only)" . . . . . . . .
