" The hypoth- esis was that agreement with this statement, which expresses very firm belief in the
supernatural
and an attitude of submission toward it, would be asso- ciated with prejudice.
Adorno-T-Authoritarian-Personality-Harper-Bros-1950
62 5 2.
80 6
2. 23 0
2 3. 45 140 0 52 1 4. 70 42 0 63 3 3. 87 297
3. 32 1. 43 . . . , 3. 34 1. 48 II:
Totals:
Groups taking Form 60:
Univ. of Oregon Student Women Univ. of Oregon and Univ. of
California Student Women Univ. of Oregon and Univ. of
California Student Men Totals:
Groups taking Form 45:
Maritime School Men Psychiatric Clinic Men Psychiatric Clinic Women
Totals:
Groups taking Form 40:
Geo. washington Univ. Women Maritime School Men Middle-Class Women Middle-Class Men Working-Cla. ss Men
Totals: Over-all Totals:
. . . ,
Relation Between Subject? s and Mother's Religion MReligious S Religious Neither
Same Different s Not MNot Reli&ious Blank Over-all
N Mean N Mean N Mean N Mean N Mean N Mean N Mean S,D.
82 3. 50 41 38 3. 37 5
1. 3) 4. 00
2. 33
2. 85 3. 21
1. 67 2. 79
1 1. 20 1 1. 60 2 1. 05 4 1. 23
2 1. 17 1 2. 33
2. 75 4 1. 85
1 4. 50 10
0 6
4 1. 18 7
5 1. 84 23 3. 64
M
31 2. 74 21 170 3. 37 77
1. 37 > cj
II:
~
. . . . . . . . . ,
21 4. 22 16 28 3. 45 18
25 3. 15 22
1 2. 50 3 2. 78 7 2. 98
3. 43 54 3. 25
57 2. 95 158 3. 19
178 4. 36
50 3. 67
~
74
3. 57 56
129 4. 56 16 24 4. 16 14 40 4. 25 11
193 4. 44 41
85 4. 27 33 128 5. 10 15 84 4. 22 47 40 4. 49 13 33 4. 16 7 370 4. 56 115
807 4. 19 289
3. 72 3. 88 3. 30
1. 59 1. 63
1. 58 1. 76 1. 96 2. 08 1. 72 1. 90
z
3. 51 136
2. 76 21 3. 07 19 2. 03 60 3. 53
1. 40
5. 00 4. 40
2. 95 2. 70 3. 40
2. 92 2 5. 20 7 5. 17
1 3. 405
1 1. 20 11
1 1. 00 2
1 1. 00 2
6 2. 83 27 3. 53
3 3. 33 47
1. 46
1. 38 1. 29
1. 26
1. 32 >t:l
M
~
CJl 1. 60 0
3. 40 2. 57 3. 18
1. 36
71 3. 65 1. 60 t"'
299 4:07
132 4. 04 164 5. 08 154 3. 64
69 3. 89
59 3. 83 578 4. 19
1332 3. 82
. . . . . . . . . ,
. . . . . . > z
>
><
? Categories of Response
"High" categories 6 and 7 combined
"Low? categories
1, 2, 3, 4, and 5 combined
N Mean S. D. --
A-S score
65 180. 7 54. 5
58 115. 4 41. 8
ETHNOCENTRISM AND RELIGIOUS ATTITUDES 217
chology Class at the Unive~sity of California. 5 The responses of the subjects were categorized according to the following scheme:
I. Generally and without qualification against both religion and the Church. 2. "Not important," with no qualifications given.
3? Agnostic; emphasis on values, ethics, way of living fostered by religion and
the Church.
4? Emphasis on religion as a source of inner strength and satisfaction.
5? Acceptance of religion but rejection of the Church; emphasis on such con-
cepts as faith and God.
6. "Mildly important," with no qualifications given.
7? Religion and the Church both important; acceptance of prayer, church
attendance, religious rituals.
Mean A-S score for subjects giving each of these categories of response was calculated. Means for categories 6 and 7 were relatively high, means for all the other categories relatively low. These results were not, however, satis-
TABLE 5 (VI)
MEAN A-S SCORES OF GROUPS GIVING DIFFERENT CATEGORIES OF RESPONSE TO THE QUESTION: "HOW IMPORTANT ARE RELIGION AND THE CHURCH? "
factory from a statistical point of view, mainly because several of the cate- gories were represented by very few subjects. It seemed justifiable to com- bine categories 6 and 7, on the basis that both described a favorable attitude toward religion and the church while making no distinction between the two; similarly, all the remaining categories could be combined on the basis that they represented something other than uncritical acceptance of formal- ized religion. The means for these two higher-order categories appear in Table 5(VI). Subjects who believe that religion and the church are both important have a mean A-S score of r8o. 7, and the other subjects, those who are against religion and the church generally, those who accept religion but not the church, and those who emphasize the personal, ethical, and ra- tional aspects of religion, have a mean of I I 5+ The difference between these two means is 65. 3 and the C. R. is 7-3-
5 The over-all data for this group were presented in Chapter III.
? 218 THE AUTHORITARIAN PERSONALITY
These data may be taken to mean that for subjects high on A-S, religion is important primarily as a matter of convention. Subjects who are low on A-S may either be against religion or have a religion that is relatively per- sonal and internalized. The fact of acceptance or rejection of religion is not as important as the way in which it is accepted or rejected.
3. SCALE ITEMS
Three items pertaining to religious ideology were included in the scale which attempted to measure potentially fascistic trends in the personality. This scale will be described in the next chapter but it seems appropriate to consider here those items which touch upon the topic of religion. The first item was the following: "The modern church with its many rules and hypocrisies, does not appeal to the deeply religious person; it appeals mainly to the childish, the insecure, and the uncritical. " The hypothesis here was that disagreement with the item would indicate uncritical acceptance of the church and, hence, ethnocentrism, and that agreement with the item would indicate either an antireligious attitude or a genuinely religious but more intellectual point of view from which the church might be criti- cized-something which we should expect to go with low scores on the scales for measuring prejudice. The results were disappointing. The mean score on the item for the 295 subjects taking Form 78 of the questionnaire was 4. 67 and the D. P-As was - r. r 8. It appeared in retrospect that the state- ment was too complex and awkward, and hence, frequently misunderstood. More than this, the high mean score indicates that relatively few people were ready to criticize the church in this way; this, of course, would lower the chances of the item's being discriminating. Needless to say, the item was not retained in the questionnaire after its first trial.
Another item in the scale was the following: "Every person should have a deep faith in some supernatural force higher than himself to which he gives total allegiance and whose decisions he does not question.
" The hypoth- esis was that agreement with this statement, which expresses very firm belief in the supernatural and an attitude of submission toward it, would be asso- ciated with prejudice. The mean score for the 295 subjects taking Form 78 was 3? 97 and the D. P-As, 1. 26. When the item, slightly revised, was used in Form 6o of the questionnaire (N = 226) the mean score was 3. 6o and the D. P. E, r. 38. These D. P. 's are large enough to indicate that the relation of this item to ethnocentrism approaches statistical significance. The item was significantly correlated with the remainder of the scale in which it appeared, and this total scale was highly correlated with ethnocentrism.
The third item was as follows: "Sciences like chemistry, physics, and medicine have carried man very far but there are many important things that can never possibly be understood by the human mind. " It was considered that subjects in agreeing with this item conceived of a mysterious spiritual
? ETHNOCENTRISM AND RELIGIOUS A TTITUDES
realm of things with respect to which investigation was taboo and toward which their attitude was one of reverence. This outlook was expected to correlate with ethnocentrism. The mean in the case of Form 78 was 4? 35 and the D. P. As, ? 97? In the case of Form 6o, where the item appeared in a slightly revised version, the mean was 4. 98 and the D. P. E, 1. 32. Here, once again, is evidence of a relationship between a particular religious idea and ethnocentrism. This item, like the "supernatural force" item discussed above, was significantly correlated with the remainder of the scale for measuring implicit antidemocratic trends and it was employed throughout the course of the study.
These results suggest that had it been possible to express a variety of religious beliefs, ideas, and sentiments in the form of scale items, more im- pressive quantitative results bearing on the relations of religious ideology to ethnocentrism would have been obtained. This is a matter which might well be the topic of future research.
C. DISCUSSION
Belonging to or identifying oneself with a religious body in America today certainly does not mean that one thereby takes over the traditional Christian values of tolerance, brotherhood, and equality. On the contrary, it appears that these values are more firmly held by people who do not affiliate with any religious group. It may be that religious affiliation or church attendance is of little importance one way or the other in determining social attitudes, that the great majority of middle-class Americans identify themselves with some religious denomination as a matter of course, without thinking much about it. This would be in keeping with the facts that the mean scores and the variability for the large religious denominations are very similar to those found in our sample as a whole. It may be argued, however, that this conventional approach to religion expresses enough identification with the status quo, submission to external authority, and readiness to emphasize moralistically the differences between those who "belong" and those who do not, to differentiate, in terms of E score, members of the large denominations from the nonreligious and from the members of those minor groups which actually stand for trends of an opposite character. At the same time, mem- bers of the major denominations seem to differ widely among themselves with respect to trends of this kind, and where there are signs that the ac- ceptance of religion has been determined primarily by conventional or ex- ternal considerations, E score tends to go up. Thus it is that agreement between the parents in the matter of religious affiliation, a circumstance that might lessen the chances of an awakening on the part of the subject to the issues involved, and sameness of the subject's religion and that of the mother, something that might be indicative of submissiveness toward au-
? 220 THE AUTHORITARIAN PERSONALITY
thority, tend to be associated with ethnocentrism. But among the members of the major denominations there are many subjects whose religion would appear to be "genuine," in the sense that it was arrived at more or less independently of external pressure and takes the form of internalized values. These subjects, it seems, tend to score low, often very low, on ethnocentrism. Subjects with this same outlook probably predominate in the low-scoring Protestant denominations and often, ? no doubt, they profess to no religious affiliation at all.
It seems that we can approach an understanding of the relations between religion and ethnocentrism by paying attention to what the acceptance or the rejection of religion means to the individual. When the problem is ap- proached from this point of view the psychological factors which appear as most important are much the same as those which came to the fore in the preceding chapters: conformity, conventionalism, authoritarian submission, determination by external pressures, thinking in ingroup-outgroup terms, and the like vs. nonconformity, independence, internalization of values, and so forth. The fragmentary data on religious ideology afforded by the scale items lend themselves to the same mode of interpretation. An attitude of com- plete submissiveness toward "supernatural forces" and a readiness to accept the essential incomprehensibility of "many important things" strongly sug- gest the persistence in the individual of infantile attitudes toward the parents, that is to say, of authoritarian submission in a very pure form. Psychological variables of the kind discussed here are investigated directly in the next chapter.
D. SUMMARY AND CONCLUSION
Subjects who profess to some religious affiliation express more prejudice than those who do not; but mean A-SorE scores for all the large denomina- tions are close to the theoretical neutral point. The vast majority of our subjects do identify themselves with some religious group, and the variability with respect to ethnocentrism among these subjects is almost as great as it is in our sample as a whole. The factor of religious denomination does not prove to be very significant. Among the largest denominations no differ- ences of any significance appear; but Unitarians, who seem to be distin- guished by their liberalism, and a group of minor Protestant groups, in the case of which there might be some spirit of nonconformity or some lack of identification with the status quo, score lower than the others. Frequency of church attendance is also not particularly revealing; however, the finding that those who never attend obtain lower E scores than those who do attend is added evidence that people who reject organized religion are less prej- udiced than those who accept it.
When the religious affiliation of the subject is considered in relation to that
? ETHNOCENTRISM AND RELIGIOUS A TTITUDES
22I
of his parents, it appears that ethnocentrism tends to be more pronounced in subjects whose parents presented a unified religious front than in cases where the religious influence from the parents was inconsistent, partial, or nonexistent. Furthermore, there is an indication that agreement between the subject and his or her mother in the matter of religion tends to be asso- ciated with ethnocentrism, disagreement with its opposite. These results sug- gest that acceptance of religion mainly as an expression of submission to a clear pattern of parental authority is a condition favorable to ethnocentrism.
A quantitative approach to religious ideology was made by including in one form of the questionnaire an open-ended question concerning the im- portance, in the subject's mind, of religion and the church. When a cate- gorization of the answers to this question was made and mean A-S scores cal- culated, it turned out that the subjects who considered both religion and the church important were very considerably more anti-Semitic than were sub- jects who considered neither important or emphasized the ethical aspects of religion or differentiated between the church and "real" religion and, while rejecting the former, stressed the more personal and the more rational aspects of the latter.
Two scale-items pertaining to religious ideology appeared to be slightly correlated with prejudice. The more agreement with statements to the effect that people should have "complete faith in some supernatural force" and that "there are some things that can never be understood by the human mind," the higher did the A-S score tend to be.
In general, it appeared that gross, objective factors-denomination and frequency of church attendance-were less significant for prejudice than were certain psychological trends reflected in the way the subject accepted or rejected religion and in the content of his religious ideology. These trends -conventionalism, authoritarian submission, and so forth-were generally the same as those which came to the fore in preceding chapters, and we turn now to our attempt to investigate them directly.
? CHAPTER VII
THE MEASUREMENT OF IMPLICIT ANTIDEMOCRATIC TRENDS
R. Nevitt Sanford, T. W. Adorno, Else Frenkel-Brunswik, and Daniel f. Levinson
A. INTRODUCTION
At a certain stage of the study, after considerable work with the A-S and E scales had been done, there gradually evolved a plan for constructing a scale that would measure prejudice without appearing to have this aim and without mentioning the name of any minority group. It appeared that such an instrument, if it correlated highly enough with the A-S and E scales, might prove to be a very useful substitute for them. It might be used to survey opinion in groups where "racial questions" were too "ticklish" a matter to permit the introduction of an A-S or E scale, e. g. , a group which included many members of one or another ethnic minority. It might be used for measuring prejudice among minority group members themselves. Most important, by circumventing some of the defenses which people employ when asked to express themselves with respect to "race issues," it might provide a more valid measure of prejudice.
The PEC scale might have commended itself as an index of prejudice, but its correlations with the A-S and E scales did not approach being high enough. Moreover, the items of this scale were too explicitly ideological, that is, they might be too readily associated with prejudice in some logical or automatic way. What was needed was a collection of items each of which was correlated with A-S and E but which did not come from an area ordinarily covered in discussions of political, economic, and social matters. The natural place to turn was to the clinical material already collected, where, particularly in the subjects' discussions of such topics as the self, family, sex, interpersonal relations, moral and personal values; there had appeared numerous trends which, it appeared, might be connected with prejudice.
At this point the second-and major-purpose of the new scale began to 222
? MEASUREMENT OF ANTIDEMOCRA TIC TRENDS
2 23
take shape. Might not such a scale yield a valid estimate of antidemocratic tendencies at the personality level? It was clear, at the time the new scale was being planned, that anti-Semitism and ethnocentrism were not merely matters of surface opinion, but general tendencies with sources, in part at least, deep within the structure of the person. Would it not be possible to construct a scale that would approach more directly these deeper, often un- conscious forces? If so, and if the scale could be validated by means of later clinical studies, would we not have a better estimate of antidemocratic po- tential than could be obtained from the scales that were more openly ideo- logical? The prospect was intriguing. And experience with clinical tech- niques and with the other scales gave considerable promise of success. In attempting to account for the generality of A-S and of E, to explain what it was that made the diverse items of these scales go together, we had been led to the formulation of enduring psychological dispositions in the person -stereotypy, conventionalism, concern with power, and so forth. Study of the ideological discussions of individuals, e. g. , Mack and Larry, had had the same outcome: there appeared to be dispositions in each individual that were reflected in his discussion of each ideological area as well as in his dis- cussion of matters not ordinarily regarded as ideological. And when clinical- genetic material was examined, it appeared that these dispositions could fre- quently be referred to deep-lying personality needs. The task then was to formulate scale items which, though they were statements of opinions and attitudes and had the same form as those appearing in ordinary opinion- attitude questionnaires, would actually serve as "giveaways" of underlying antidemocratic trends in the personality. This would make it possible to carry over into group studies the insights and hypotheses derived from clinical investigation; it would test whether we could study on a mass scale features ordinarily regarded as individualistic and qualitative.
This second purpose-the quantification of antidemocratic trends at the level of personality-did not supersede the first, that of measuring anti- Semitism and ethnocentrism without mentioning minority groups or cur- rent politico-economic issues. Rather, it seemed that the two might be realized together. The notion was that A-S and E would correlate with the new scale because the A-S and E responses were strongly influenced by the underlying trends which the new scale sought to get at by a different approach. Indeed, if such a correlation could be obtained it could be taken as evidence that anti-Semitism and ethnocentrism were not isolated or specific or entirely superficial attitudes but expressions of persistent tendencies in the person. This would depend, however, upon how successful was the attempt to exclude from the new scale items which might have been so frequently or so automatically associated with anti-Semitism or ethnocentrism that they might be regarded as aspects of the same political "line. " In any case, how- ever, it seemed that the discovery of opinions and attitudes, in various areas
? 224 THE AUTHORITARIAN PERSONALITY
other than the usual politico-socioeconomic one, that were associated with anti-Semitism and ethnocentrism, would give a more comprehensive grasp of the prejudiced outlook on the world. The new instrument was termed the F scale, to signify its concern with implicit prefascist tendencies.
On theoretical grounds it was expected that the correlations ofF with A-S and E would not approach unity. It was hoped that the F scale would catch some of the antidemocratic potential that might not be expressed when sub- jects responded to items which dealt directly with hostility toward minority groups. True, the items of the present A-S and E scales were, for the most part, so formulated as to allow the subject to express prejudice while main- taining the feeling that he was being democratic. Yet it was recognized that a subject might score relatively low on A-S or E and still, in the interview, where a confidential relationship was established and the interviewer was very permissive, reveal that he was prejudiced. More than this, it had to be admitted that a subject might refuse altogether to express hostility against minority groups and yet reveal features, e. g. , a tendency to think of such groups in a stereotyped way or a tendency moralistically to reject social groups other than ethnic ones, which had to be taken as susceptibility to anti- democratic propaganda. If the F scale were to be regarded as a measure of antidemocratic potential-something which might or might not be ex- pressed in open hostility against outgroups-then it could not be perfectly correlated with A-S or E. Rather, the demand to be made of it was that it single out individuals who in intensive clinical study revealed themselves to be receptive to antidemocratic propaganda. Although it was not possible within the scope of the study to use the F scale alone as the basis for selecting interviewees, it was possible to relate F scale score to various other indices of antidemocratic personality trends as brought to light by other techniques. Such trends, it seemed, could exist in the absence of high A-S or E scores.
However, the distinction between potential and manifest should not be overdrawn. Given emotionally determined antidemocratic trends in the person, we should expect that in general they would be evoked by the A-S and E items, which were designed for just this purpose, as well as by the F scale and other indirect methods. The person who was high on F but not on A-S or E would be the exception, whose inhibitions upon the expression of prejudice against minorities would require special explanation.
B. CONSTRUCTION OF THE FASCISM (F) SCALE 1. THE UNDERLYING THEORY
The 38 items of the original F scale are shown in Table I (VII), num- bered in the order of their appearance on Form 78. If the reader considers that most of what has gone before in this volume was either known or
? MEASUREMENT OF ANTIDEMOCRA TIC TRENDS 225
thought about before construction of the F scale began, it will be apparent that in devising the scale we did not proceed in a strictly empirical fashion. We did not consider starting with hundreds of items chosen more or less at random and then seeing by trial and error which ones might be associated with A-S and E. For every item there was a hypothesis, sometimes several hypotheses, stating what might be the nature of its connection with prejudice.
The major source of these hypotheses was the research already performed in the present study. Available for the purpose was the following material: results, such as those given in preceding chapters, from the A-S, E, and PEC scales; numerous correlates of E derived from questionnaire studies, that is, from responses to factual and short essay questions pertaining to such topics as religion, war, ideal society, and so forth; early results from projective questions; finally, and by far the most important, material from the inter- views and the Thematic Apperception Tests. Another important source of items was research in fields allied to the present one in which the authors had previously had a part. Principal among these were several studies performed at the University of California on personality in relation to war morale and ideology (19, 20, 102, 107, 108, 109), and researches of the Institute of Social Research such as content analyses of speeches of anti-Semitic agi- tators and a study on anti-Semitic workers (2, 3, 56, 57, 57A, 57B). Finally, there was the general literature on anti-Semitism and fascism, embracing both empirical and theoretical studies.
It will have been recognized that the interpretation of the material of the present study was guided by a theoretical orientation that was present at the start. The same orientation played the most crucial role in the prepara- tion of the F scale. Once a hypothesis had been formulated concerning the way in which some deep-lying trend in the personality might express itself in some opinion or attitude that was dynamically, though not logically, re- lated to prejudice against outgroups, a preliminary sketch for an item was usually not far to seek: a phrase from the daily newspaper, an utterance by an interviewee, a fragment of ordinary conversation was usually ready at hand.
(As will be seen, however, the actual formulation of an item was a technical proceeding to which considerable care had to be devoted. )
As to what kinds of central personality trends we might expect to be the most significant, the major guide, as has been said, was the research which had gone before; they were the trends which, as hypothetical constructs, seemed best to explain the consistency of response on the foregoing scales, and which emerged from the analysis of clinical material as the likely sources of the coherence found in individual cases. Most of these trends have been mentioned before, usually when it was necessary to do so in order to give meaning to obtained results. For example, when it was discovered that the anti-Semitic individual objects to Jews on the ground that they violate con- ventional moral val~es, one interpretation was that this individual had a
? 226 THE AUTHORIT ARIAN PERSONALITY
TABLE x (VII) THE F SCALE: FoRM 78
z. Although many people may scoff, it may yet be shown that astrology can explain a lot of things.
3? America is getting so far from the true American way of life that force may be necessary to restore it.
6. It is only natural and right that women be restricted in certain ways in which men have more freedom.
9? Too many people today are living in an unnatural, soft way; we should return to the fundamentals, to a more red-blooded, active way of life.
xo. It is more than a remarkable coincidence that Japan had an earthquake on Pearl Harbor Day, December 7, 1944.
12. The modern church, with its many rules and hypocrisies, does not appeal to the deeply religious person; it appeals mainly to the childish, the insecure, and the uncritical.
14. After we finish off the Germans and Japs, we ought to concentrate on other enemies of the human race such as rats, snakes, and germs.
17. Familiarity breeds contempt.
19. One should avoid doing things in public which appear wrong to others, even
though one knows that these things are really all right.
20. One of the main values of progressive education is that it gives the child
great freedom in expressing those natural impulses and desires so often
frowned upon by conventional middle-class society.
23. He is, indeed, contemptible who does not feel an undying love, gratitude, and
respect for his parents.
24. Today everything is unstable; we should be prepared for a period of constant
change, conflict, and upheaval.
28. Novels or stories that tell about what people think and feel are more interest-
ing than those which contain mainly action, romance, and adventure.
30. Reports of atrocities in Europe have been greatly exaggerated for propa-
ganda purposes.
3I. Homosexuality is a particularly rotten form of delinquency and ought to be
severely punished.
32? It is essential for learning or effective work that our teachers or bosses outline
in detail what is to be done and exactly how to go about it.
35? There are some activities so flagrantly un-American that, when responsible officials won't take the proper steps, the wide-awake citizen should take the
law into his own hands.
38. There is too much emphasis in college on intellectual and theoretical topics,
not enough emphasis on practical matters and on the homely virtues of living. 39? Every person should have a deep faith in some supernatural force higher than himself to which he gives total allegiance and whose decisions he does
not question.
42. No matter how they act on the surface, men are interested in women for only
one reason.
43? Sciences like chemistry, physics, and medicine have carried men very far,
but there are many important things that can never possibly be understood
by the human mind.
46. The sexual orgies of the old Greeks and Romans are nursery school stuff
compared to some of the goings-on in this country today, even in circles where people might least expect it. ?
? 47?
50.
53?
55.
;6.
5s. 59? 6o.
No insult to our honor 'should ever go unpunished.
Obedience and respect for authority are the most important virtues children should learn.
There are some things too intimate or personal to talk about even with one's closest friends.
Although leisure is a fine thing, it is good hard work that makes life interest- ing and worthwhile.
After the war, we may expect a crime wave; the control of gangsters and ruffians will become a major social problem.
What a man does is not so important so long as he does it well.
Human nature being what it is, there will always be war and conflict.
2. 23 0
2 3. 45 140 0 52 1 4. 70 42 0 63 3 3. 87 297
3. 32 1. 43 . . . , 3. 34 1. 48 II:
Totals:
Groups taking Form 60:
Univ. of Oregon Student Women Univ. of Oregon and Univ. of
California Student Women Univ. of Oregon and Univ. of
California Student Men Totals:
Groups taking Form 45:
Maritime School Men Psychiatric Clinic Men Psychiatric Clinic Women
Totals:
Groups taking Form 40:
Geo. washington Univ. Women Maritime School Men Middle-Class Women Middle-Class Men Working-Cla. ss Men
Totals: Over-all Totals:
. . . ,
Relation Between Subject? s and Mother's Religion MReligious S Religious Neither
Same Different s Not MNot Reli&ious Blank Over-all
N Mean N Mean N Mean N Mean N Mean N Mean N Mean S,D.
82 3. 50 41 38 3. 37 5
1. 3) 4. 00
2. 33
2. 85 3. 21
1. 67 2. 79
1 1. 20 1 1. 60 2 1. 05 4 1. 23
2 1. 17 1 2. 33
2. 75 4 1. 85
1 4. 50 10
0 6
4 1. 18 7
5 1. 84 23 3. 64
M
31 2. 74 21 170 3. 37 77
1. 37 > cj
II:
~
. . . . . . . . . ,
21 4. 22 16 28 3. 45 18
25 3. 15 22
1 2. 50 3 2. 78 7 2. 98
3. 43 54 3. 25
57 2. 95 158 3. 19
178 4. 36
50 3. 67
~
74
3. 57 56
129 4. 56 16 24 4. 16 14 40 4. 25 11
193 4. 44 41
85 4. 27 33 128 5. 10 15 84 4. 22 47 40 4. 49 13 33 4. 16 7 370 4. 56 115
807 4. 19 289
3. 72 3. 88 3. 30
1. 59 1. 63
1. 58 1. 76 1. 96 2. 08 1. 72 1. 90
z
3. 51 136
2. 76 21 3. 07 19 2. 03 60 3. 53
1. 40
5. 00 4. 40
2. 95 2. 70 3. 40
2. 92 2 5. 20 7 5. 17
1 3. 405
1 1. 20 11
1 1. 00 2
1 1. 00 2
6 2. 83 27 3. 53
3 3. 33 47
1. 46
1. 38 1. 29
1. 26
1. 32 >t:l
M
~
CJl 1. 60 0
3. 40 2. 57 3. 18
1. 36
71 3. 65 1. 60 t"'
299 4:07
132 4. 04 164 5. 08 154 3. 64
69 3. 89
59 3. 83 578 4. 19
1332 3. 82
. . . . . . . . . ,
. . . . . . > z
>
><
? Categories of Response
"High" categories 6 and 7 combined
"Low? categories
1, 2, 3, 4, and 5 combined
N Mean S. D. --
A-S score
65 180. 7 54. 5
58 115. 4 41. 8
ETHNOCENTRISM AND RELIGIOUS ATTITUDES 217
chology Class at the Unive~sity of California. 5 The responses of the subjects were categorized according to the following scheme:
I. Generally and without qualification against both religion and the Church. 2. "Not important," with no qualifications given.
3? Agnostic; emphasis on values, ethics, way of living fostered by religion and
the Church.
4? Emphasis on religion as a source of inner strength and satisfaction.
5? Acceptance of religion but rejection of the Church; emphasis on such con-
cepts as faith and God.
6. "Mildly important," with no qualifications given.
7? Religion and the Church both important; acceptance of prayer, church
attendance, religious rituals.
Mean A-S score for subjects giving each of these categories of response was calculated. Means for categories 6 and 7 were relatively high, means for all the other categories relatively low. These results were not, however, satis-
TABLE 5 (VI)
MEAN A-S SCORES OF GROUPS GIVING DIFFERENT CATEGORIES OF RESPONSE TO THE QUESTION: "HOW IMPORTANT ARE RELIGION AND THE CHURCH? "
factory from a statistical point of view, mainly because several of the cate- gories were represented by very few subjects. It seemed justifiable to com- bine categories 6 and 7, on the basis that both described a favorable attitude toward religion and the church while making no distinction between the two; similarly, all the remaining categories could be combined on the basis that they represented something other than uncritical acceptance of formal- ized religion. The means for these two higher-order categories appear in Table 5(VI). Subjects who believe that religion and the church are both important have a mean A-S score of r8o. 7, and the other subjects, those who are against religion and the church generally, those who accept religion but not the church, and those who emphasize the personal, ethical, and ra- tional aspects of religion, have a mean of I I 5+ The difference between these two means is 65. 3 and the C. R. is 7-3-
5 The over-all data for this group were presented in Chapter III.
? 218 THE AUTHORITARIAN PERSONALITY
These data may be taken to mean that for subjects high on A-S, religion is important primarily as a matter of convention. Subjects who are low on A-S may either be against religion or have a religion that is relatively per- sonal and internalized. The fact of acceptance or rejection of religion is not as important as the way in which it is accepted or rejected.
3. SCALE ITEMS
Three items pertaining to religious ideology were included in the scale which attempted to measure potentially fascistic trends in the personality. This scale will be described in the next chapter but it seems appropriate to consider here those items which touch upon the topic of religion. The first item was the following: "The modern church with its many rules and hypocrisies, does not appeal to the deeply religious person; it appeals mainly to the childish, the insecure, and the uncritical. " The hypothesis here was that disagreement with the item would indicate uncritical acceptance of the church and, hence, ethnocentrism, and that agreement with the item would indicate either an antireligious attitude or a genuinely religious but more intellectual point of view from which the church might be criti- cized-something which we should expect to go with low scores on the scales for measuring prejudice. The results were disappointing. The mean score on the item for the 295 subjects taking Form 78 of the questionnaire was 4. 67 and the D. P-As was - r. r 8. It appeared in retrospect that the state- ment was too complex and awkward, and hence, frequently misunderstood. More than this, the high mean score indicates that relatively few people were ready to criticize the church in this way; this, of course, would lower the chances of the item's being discriminating. Needless to say, the item was not retained in the questionnaire after its first trial.
Another item in the scale was the following: "Every person should have a deep faith in some supernatural force higher than himself to which he gives total allegiance and whose decisions he does not question.
" The hypoth- esis was that agreement with this statement, which expresses very firm belief in the supernatural and an attitude of submission toward it, would be asso- ciated with prejudice. The mean score for the 295 subjects taking Form 78 was 3? 97 and the D. P-As, 1. 26. When the item, slightly revised, was used in Form 6o of the questionnaire (N = 226) the mean score was 3. 6o and the D. P. E, r. 38. These D. P. 's are large enough to indicate that the relation of this item to ethnocentrism approaches statistical significance. The item was significantly correlated with the remainder of the scale in which it appeared, and this total scale was highly correlated with ethnocentrism.
The third item was as follows: "Sciences like chemistry, physics, and medicine have carried man very far but there are many important things that can never possibly be understood by the human mind. " It was considered that subjects in agreeing with this item conceived of a mysterious spiritual
? ETHNOCENTRISM AND RELIGIOUS A TTITUDES
realm of things with respect to which investigation was taboo and toward which their attitude was one of reverence. This outlook was expected to correlate with ethnocentrism. The mean in the case of Form 78 was 4? 35 and the D. P. As, ? 97? In the case of Form 6o, where the item appeared in a slightly revised version, the mean was 4. 98 and the D. P. E, 1. 32. Here, once again, is evidence of a relationship between a particular religious idea and ethnocentrism. This item, like the "supernatural force" item discussed above, was significantly correlated with the remainder of the scale for measuring implicit antidemocratic trends and it was employed throughout the course of the study.
These results suggest that had it been possible to express a variety of religious beliefs, ideas, and sentiments in the form of scale items, more im- pressive quantitative results bearing on the relations of religious ideology to ethnocentrism would have been obtained. This is a matter which might well be the topic of future research.
C. DISCUSSION
Belonging to or identifying oneself with a religious body in America today certainly does not mean that one thereby takes over the traditional Christian values of tolerance, brotherhood, and equality. On the contrary, it appears that these values are more firmly held by people who do not affiliate with any religious group. It may be that religious affiliation or church attendance is of little importance one way or the other in determining social attitudes, that the great majority of middle-class Americans identify themselves with some religious denomination as a matter of course, without thinking much about it. This would be in keeping with the facts that the mean scores and the variability for the large religious denominations are very similar to those found in our sample as a whole. It may be argued, however, that this conventional approach to religion expresses enough identification with the status quo, submission to external authority, and readiness to emphasize moralistically the differences between those who "belong" and those who do not, to differentiate, in terms of E score, members of the large denominations from the nonreligious and from the members of those minor groups which actually stand for trends of an opposite character. At the same time, mem- bers of the major denominations seem to differ widely among themselves with respect to trends of this kind, and where there are signs that the ac- ceptance of religion has been determined primarily by conventional or ex- ternal considerations, E score tends to go up. Thus it is that agreement between the parents in the matter of religious affiliation, a circumstance that might lessen the chances of an awakening on the part of the subject to the issues involved, and sameness of the subject's religion and that of the mother, something that might be indicative of submissiveness toward au-
? 220 THE AUTHORITARIAN PERSONALITY
thority, tend to be associated with ethnocentrism. But among the members of the major denominations there are many subjects whose religion would appear to be "genuine," in the sense that it was arrived at more or less independently of external pressure and takes the form of internalized values. These subjects, it seems, tend to score low, often very low, on ethnocentrism. Subjects with this same outlook probably predominate in the low-scoring Protestant denominations and often, ? no doubt, they profess to no religious affiliation at all.
It seems that we can approach an understanding of the relations between religion and ethnocentrism by paying attention to what the acceptance or the rejection of religion means to the individual. When the problem is ap- proached from this point of view the psychological factors which appear as most important are much the same as those which came to the fore in the preceding chapters: conformity, conventionalism, authoritarian submission, determination by external pressures, thinking in ingroup-outgroup terms, and the like vs. nonconformity, independence, internalization of values, and so forth. The fragmentary data on religious ideology afforded by the scale items lend themselves to the same mode of interpretation. An attitude of com- plete submissiveness toward "supernatural forces" and a readiness to accept the essential incomprehensibility of "many important things" strongly sug- gest the persistence in the individual of infantile attitudes toward the parents, that is to say, of authoritarian submission in a very pure form. Psychological variables of the kind discussed here are investigated directly in the next chapter.
D. SUMMARY AND CONCLUSION
Subjects who profess to some religious affiliation express more prejudice than those who do not; but mean A-SorE scores for all the large denomina- tions are close to the theoretical neutral point. The vast majority of our subjects do identify themselves with some religious group, and the variability with respect to ethnocentrism among these subjects is almost as great as it is in our sample as a whole. The factor of religious denomination does not prove to be very significant. Among the largest denominations no differ- ences of any significance appear; but Unitarians, who seem to be distin- guished by their liberalism, and a group of minor Protestant groups, in the case of which there might be some spirit of nonconformity or some lack of identification with the status quo, score lower than the others. Frequency of church attendance is also not particularly revealing; however, the finding that those who never attend obtain lower E scores than those who do attend is added evidence that people who reject organized religion are less prej- udiced than those who accept it.
When the religious affiliation of the subject is considered in relation to that
? ETHNOCENTRISM AND RELIGIOUS A TTITUDES
22I
of his parents, it appears that ethnocentrism tends to be more pronounced in subjects whose parents presented a unified religious front than in cases where the religious influence from the parents was inconsistent, partial, or nonexistent. Furthermore, there is an indication that agreement between the subject and his or her mother in the matter of religion tends to be asso- ciated with ethnocentrism, disagreement with its opposite. These results sug- gest that acceptance of religion mainly as an expression of submission to a clear pattern of parental authority is a condition favorable to ethnocentrism.
A quantitative approach to religious ideology was made by including in one form of the questionnaire an open-ended question concerning the im- portance, in the subject's mind, of religion and the church. When a cate- gorization of the answers to this question was made and mean A-S scores cal- culated, it turned out that the subjects who considered both religion and the church important were very considerably more anti-Semitic than were sub- jects who considered neither important or emphasized the ethical aspects of religion or differentiated between the church and "real" religion and, while rejecting the former, stressed the more personal and the more rational aspects of the latter.
Two scale-items pertaining to religious ideology appeared to be slightly correlated with prejudice. The more agreement with statements to the effect that people should have "complete faith in some supernatural force" and that "there are some things that can never be understood by the human mind," the higher did the A-S score tend to be.
In general, it appeared that gross, objective factors-denomination and frequency of church attendance-were less significant for prejudice than were certain psychological trends reflected in the way the subject accepted or rejected religion and in the content of his religious ideology. These trends -conventionalism, authoritarian submission, and so forth-were generally the same as those which came to the fore in preceding chapters, and we turn now to our attempt to investigate them directly.
? CHAPTER VII
THE MEASUREMENT OF IMPLICIT ANTIDEMOCRATIC TRENDS
R. Nevitt Sanford, T. W. Adorno, Else Frenkel-Brunswik, and Daniel f. Levinson
A. INTRODUCTION
At a certain stage of the study, after considerable work with the A-S and E scales had been done, there gradually evolved a plan for constructing a scale that would measure prejudice without appearing to have this aim and without mentioning the name of any minority group. It appeared that such an instrument, if it correlated highly enough with the A-S and E scales, might prove to be a very useful substitute for them. It might be used to survey opinion in groups where "racial questions" were too "ticklish" a matter to permit the introduction of an A-S or E scale, e. g. , a group which included many members of one or another ethnic minority. It might be used for measuring prejudice among minority group members themselves. Most important, by circumventing some of the defenses which people employ when asked to express themselves with respect to "race issues," it might provide a more valid measure of prejudice.
The PEC scale might have commended itself as an index of prejudice, but its correlations with the A-S and E scales did not approach being high enough. Moreover, the items of this scale were too explicitly ideological, that is, they might be too readily associated with prejudice in some logical or automatic way. What was needed was a collection of items each of which was correlated with A-S and E but which did not come from an area ordinarily covered in discussions of political, economic, and social matters. The natural place to turn was to the clinical material already collected, where, particularly in the subjects' discussions of such topics as the self, family, sex, interpersonal relations, moral and personal values; there had appeared numerous trends which, it appeared, might be connected with prejudice.
At this point the second-and major-purpose of the new scale began to 222
? MEASUREMENT OF ANTIDEMOCRA TIC TRENDS
2 23
take shape. Might not such a scale yield a valid estimate of antidemocratic tendencies at the personality level? It was clear, at the time the new scale was being planned, that anti-Semitism and ethnocentrism were not merely matters of surface opinion, but general tendencies with sources, in part at least, deep within the structure of the person. Would it not be possible to construct a scale that would approach more directly these deeper, often un- conscious forces? If so, and if the scale could be validated by means of later clinical studies, would we not have a better estimate of antidemocratic po- tential than could be obtained from the scales that were more openly ideo- logical? The prospect was intriguing. And experience with clinical tech- niques and with the other scales gave considerable promise of success. In attempting to account for the generality of A-S and of E, to explain what it was that made the diverse items of these scales go together, we had been led to the formulation of enduring psychological dispositions in the person -stereotypy, conventionalism, concern with power, and so forth. Study of the ideological discussions of individuals, e. g. , Mack and Larry, had had the same outcome: there appeared to be dispositions in each individual that were reflected in his discussion of each ideological area as well as in his dis- cussion of matters not ordinarily regarded as ideological. And when clinical- genetic material was examined, it appeared that these dispositions could fre- quently be referred to deep-lying personality needs. The task then was to formulate scale items which, though they were statements of opinions and attitudes and had the same form as those appearing in ordinary opinion- attitude questionnaires, would actually serve as "giveaways" of underlying antidemocratic trends in the personality. This would make it possible to carry over into group studies the insights and hypotheses derived from clinical investigation; it would test whether we could study on a mass scale features ordinarily regarded as individualistic and qualitative.
This second purpose-the quantification of antidemocratic trends at the level of personality-did not supersede the first, that of measuring anti- Semitism and ethnocentrism without mentioning minority groups or cur- rent politico-economic issues. Rather, it seemed that the two might be realized together. The notion was that A-S and E would correlate with the new scale because the A-S and E responses were strongly influenced by the underlying trends which the new scale sought to get at by a different approach. Indeed, if such a correlation could be obtained it could be taken as evidence that anti-Semitism and ethnocentrism were not isolated or specific or entirely superficial attitudes but expressions of persistent tendencies in the person. This would depend, however, upon how successful was the attempt to exclude from the new scale items which might have been so frequently or so automatically associated with anti-Semitism or ethnocentrism that they might be regarded as aspects of the same political "line. " In any case, how- ever, it seemed that the discovery of opinions and attitudes, in various areas
? 224 THE AUTHORITARIAN PERSONALITY
other than the usual politico-socioeconomic one, that were associated with anti-Semitism and ethnocentrism, would give a more comprehensive grasp of the prejudiced outlook on the world. The new instrument was termed the F scale, to signify its concern with implicit prefascist tendencies.
On theoretical grounds it was expected that the correlations ofF with A-S and E would not approach unity. It was hoped that the F scale would catch some of the antidemocratic potential that might not be expressed when sub- jects responded to items which dealt directly with hostility toward minority groups. True, the items of the present A-S and E scales were, for the most part, so formulated as to allow the subject to express prejudice while main- taining the feeling that he was being democratic. Yet it was recognized that a subject might score relatively low on A-S or E and still, in the interview, where a confidential relationship was established and the interviewer was very permissive, reveal that he was prejudiced. More than this, it had to be admitted that a subject might refuse altogether to express hostility against minority groups and yet reveal features, e. g. , a tendency to think of such groups in a stereotyped way or a tendency moralistically to reject social groups other than ethnic ones, which had to be taken as susceptibility to anti- democratic propaganda. If the F scale were to be regarded as a measure of antidemocratic potential-something which might or might not be ex- pressed in open hostility against outgroups-then it could not be perfectly correlated with A-S or E. Rather, the demand to be made of it was that it single out individuals who in intensive clinical study revealed themselves to be receptive to antidemocratic propaganda. Although it was not possible within the scope of the study to use the F scale alone as the basis for selecting interviewees, it was possible to relate F scale score to various other indices of antidemocratic personality trends as brought to light by other techniques. Such trends, it seemed, could exist in the absence of high A-S or E scores.
However, the distinction between potential and manifest should not be overdrawn. Given emotionally determined antidemocratic trends in the person, we should expect that in general they would be evoked by the A-S and E items, which were designed for just this purpose, as well as by the F scale and other indirect methods. The person who was high on F but not on A-S or E would be the exception, whose inhibitions upon the expression of prejudice against minorities would require special explanation.
B. CONSTRUCTION OF THE FASCISM (F) SCALE 1. THE UNDERLYING THEORY
The 38 items of the original F scale are shown in Table I (VII), num- bered in the order of their appearance on Form 78. If the reader considers that most of what has gone before in this volume was either known or
? MEASUREMENT OF ANTIDEMOCRA TIC TRENDS 225
thought about before construction of the F scale began, it will be apparent that in devising the scale we did not proceed in a strictly empirical fashion. We did not consider starting with hundreds of items chosen more or less at random and then seeing by trial and error which ones might be associated with A-S and E. For every item there was a hypothesis, sometimes several hypotheses, stating what might be the nature of its connection with prejudice.
The major source of these hypotheses was the research already performed in the present study. Available for the purpose was the following material: results, such as those given in preceding chapters, from the A-S, E, and PEC scales; numerous correlates of E derived from questionnaire studies, that is, from responses to factual and short essay questions pertaining to such topics as religion, war, ideal society, and so forth; early results from projective questions; finally, and by far the most important, material from the inter- views and the Thematic Apperception Tests. Another important source of items was research in fields allied to the present one in which the authors had previously had a part. Principal among these were several studies performed at the University of California on personality in relation to war morale and ideology (19, 20, 102, 107, 108, 109), and researches of the Institute of Social Research such as content analyses of speeches of anti-Semitic agi- tators and a study on anti-Semitic workers (2, 3, 56, 57, 57A, 57B). Finally, there was the general literature on anti-Semitism and fascism, embracing both empirical and theoretical studies.
It will have been recognized that the interpretation of the material of the present study was guided by a theoretical orientation that was present at the start. The same orientation played the most crucial role in the prepara- tion of the F scale. Once a hypothesis had been formulated concerning the way in which some deep-lying trend in the personality might express itself in some opinion or attitude that was dynamically, though not logically, re- lated to prejudice against outgroups, a preliminary sketch for an item was usually not far to seek: a phrase from the daily newspaper, an utterance by an interviewee, a fragment of ordinary conversation was usually ready at hand.
(As will be seen, however, the actual formulation of an item was a technical proceeding to which considerable care had to be devoted. )
As to what kinds of central personality trends we might expect to be the most significant, the major guide, as has been said, was the research which had gone before; they were the trends which, as hypothetical constructs, seemed best to explain the consistency of response on the foregoing scales, and which emerged from the analysis of clinical material as the likely sources of the coherence found in individual cases. Most of these trends have been mentioned before, usually when it was necessary to do so in order to give meaning to obtained results. For example, when it was discovered that the anti-Semitic individual objects to Jews on the ground that they violate con- ventional moral val~es, one interpretation was that this individual had a
? 226 THE AUTHORIT ARIAN PERSONALITY
TABLE x (VII) THE F SCALE: FoRM 78
z. Although many people may scoff, it may yet be shown that astrology can explain a lot of things.
3? America is getting so far from the true American way of life that force may be necessary to restore it.
6. It is only natural and right that women be restricted in certain ways in which men have more freedom.
9? Too many people today are living in an unnatural, soft way; we should return to the fundamentals, to a more red-blooded, active way of life.
xo. It is more than a remarkable coincidence that Japan had an earthquake on Pearl Harbor Day, December 7, 1944.
12. The modern church, with its many rules and hypocrisies, does not appeal to the deeply religious person; it appeals mainly to the childish, the insecure, and the uncritical.
14. After we finish off the Germans and Japs, we ought to concentrate on other enemies of the human race such as rats, snakes, and germs.
17. Familiarity breeds contempt.
19. One should avoid doing things in public which appear wrong to others, even
though one knows that these things are really all right.
20. One of the main values of progressive education is that it gives the child
great freedom in expressing those natural impulses and desires so often
frowned upon by conventional middle-class society.
23. He is, indeed, contemptible who does not feel an undying love, gratitude, and
respect for his parents.
24. Today everything is unstable; we should be prepared for a period of constant
change, conflict, and upheaval.
28. Novels or stories that tell about what people think and feel are more interest-
ing than those which contain mainly action, romance, and adventure.
30. Reports of atrocities in Europe have been greatly exaggerated for propa-
ganda purposes.
3I. Homosexuality is a particularly rotten form of delinquency and ought to be
severely punished.
32? It is essential for learning or effective work that our teachers or bosses outline
in detail what is to be done and exactly how to go about it.
35? There are some activities so flagrantly un-American that, when responsible officials won't take the proper steps, the wide-awake citizen should take the
law into his own hands.
38. There is too much emphasis in college on intellectual and theoretical topics,
not enough emphasis on practical matters and on the homely virtues of living. 39? Every person should have a deep faith in some supernatural force higher than himself to which he gives total allegiance and whose decisions he does
not question.
42. No matter how they act on the surface, men are interested in women for only
one reason.
43? Sciences like chemistry, physics, and medicine have carried men very far,
but there are many important things that can never possibly be understood
by the human mind.
46. The sexual orgies of the old Greeks and Romans are nursery school stuff
compared to some of the goings-on in this country today, even in circles where people might least expect it. ?
? 47?
50.
53?
55.
;6.
5s. 59? 6o.
No insult to our honor 'should ever go unpunished.
Obedience and respect for authority are the most important virtues children should learn.
There are some things too intimate or personal to talk about even with one's closest friends.
Although leisure is a fine thing, it is good hard work that makes life interest- ing and worthwhile.
After the war, we may expect a crime wave; the control of gangsters and ruffians will become a major social problem.
What a man does is not so important so long as he does it well.
Human nature being what it is, there will always be war and conflict.
