It suggests, as does the material from other techniques, that his surface identification with powerful figures and groups is at least in part a means of maintaining his sense of mastery and of
allaying
his anxiety over bodily harm.
Adorno-T-Authoritarian-Personality-Harper-Bros-1950
4 70.
0 76.
4 Nb 10 16 10 0 5
72. 1 41)
0
~ ~
65 33
32
65
43. 9 71. 2
>z
"d
M
~
en 0 z
> 20) t"" ~ 15)21. 2% . . . ,
Na-NHL 2
1 9
2
6
20 )17. 4% ~ ~
62. 1 Nb37316 Na-NHL 4 1 4 3 3
L% 69. 7
71. 2
63. 6
H% 81. 2 78. 1 81. 2 76. 6 75. 0
Nb 6 10 4 3 4 Na-NHL 0 0 6 1 8
%A 75. 4 60. 8 76. 1 69. 2 73. 1
Nb 9 17 7 4 10 Na-NHL 4 1 10 4 11
78. 4
27)
15 )26. 3%
70. 9
47)
30 )23. 7%
31) . . , 11)23. 3% ! I:
-<
? 7. All groups combined Low quartile
High quartile
Total
L% 76. 7 63. 0 75. 8 68. 4 65. 8 63. 0 57. 0 85. 9 69. 1
Nb 9 19 8 3 14 7 0 1 61 6. 6% Na-NHL 8 8 22 20 12 11 13 1 95 10. 3%
ER 146 146 155 155 155 73 64 32 926 116
m; 84. 0 79. 9 71. 3 84. 4 86. 0 7i. 3 85. 9 72. 7 80. 7
Nb 15 33 17 9 11 7 2 5 99 10. 6% Na-NHL 4 4 27 9 15 3 4 2 68 7. 3%
ER 144 144 157 157 157 77 64 33 933 117
%A 80. 3 71. 4 73. 6 76. 5 76. 0 69. 3 71. 5 79. 2 74. 9
Nb 24 52 25 12 25 14 2 6 160 8. 6% Na-NHL 12 12 49 29 27 14 17 3 163 8. 8%
ER 290 290 312 312 312 150 J 2 8 65 1859 232
Score 1 2 3 4 5 6 7 8 Total Mean
aFor Groups 2, 3, and 4 the number of subjects varies somewhat from item to item because two questionnaire forms were used. The number indicated in the "N" column is the total N in each group or quartile. The N in parenth~ses refers to those subjects who received the complete set of projective questions. Thus, in Group 2 all 51 subjects received items 1 - 5 (Forms 45 and 40); only 10 Low and 14 High received all 8 questions (Form 45), so that the N's on items 6, 7, and 8 are 10 and 14 respectively. as indicated in parentheses in the "N" column. Similarly,
of the 14 Lows in Group 3, only 9 received the full battery, the remaining 5 receiving a form which did not in? clude items 1, 2, 7, and 8. For the high quartile in Group 3, there is an Nof 8 on items 1, 2, 7, and 8. In Group 4, the N on these same items is 13 for Lows and 11 for Highs.
Group 1 received a form which did not include item 8 (see text). Groups 5 and 6 received Form 40 which contained items 1-5 only.
Thus, while there were 312 subjects in all, the N per item varied. This is shown in the table under ~All groups combined," where the row "ER" gives the number of subjects answering each item.
bin the computation of all means, the component values are weighted by N.
Key to questions: 1 (Moods), 2 (Desires), 3 (Great People), 4 (Drive Nuts), 5 (Crimes), 6 (Embarrassing), 7 (Last Six Months), 8 (A we-inspiring).
? THE AUTHORITARIAN PERSONALITY
The average agreement varies among the six groups from 67? 9 per cent for the Psychiatric Clinic Men to S2. 6 per cent for the Employment Service Veteran Men. It appears that the scorers, try as they did to be unbiased, were systematically influenced by the fact that they knew the E-quartile standing of the Psychology Women and the Veteran Men. Thus, the %A for these two groups is about So per cent, whereas for the remaining four groups
(scored blindly) it averages slightly over 70 per cent. The relatively low %A for the Clinic Men is consistent with the low scoring reliability for this group (see Table I(XV)). The data for the Clinic Women and the Middle- Class Men and Women probably best represent what can be expected with the present Scoring Manual; the average agreement of about 72 per cent for these groups might have been slightly higher had all subjects received questions 6, 7, and S.
The over-all average agreement for the individual items varies from 69. 3 to So. 3 per cent. The poorest questions are 2 ("Desires"), 6 ("Embarrass- ing"), and 7 ("Last six months"). In all three of them the L% is particularly low (57-63 per cent), probably because of the brevity of the answers, as discussed above. Correction of this error should lead to considerably better results, particularly for Item 7? The agreements for the individual items re- veal again the great consistency of the highs. Thus, the L% varied within a range of 29 points (57-S6 per cent), whereas the H% covered a range of only 15 points (7I-S6 per cent). It is of some interest that the two most dif- ferentiating items, 1 ("Moods") and S ("Awe-inspiring"), deal with issues which, in their literal meanings, are completely removed from ideology about group interaction.
Table 2(XV) provides the empirical basis on which omissions (Nb) of Questions I, 2, 4, 7, and S were converted into H scores in the statistical treat- ment. On each of these questions approximately two-thirds or more of the total number of omissions were made by the high quartiles. Moreover, it is consistent with the differential trends found that the highs should omit, more often than the lows, questions dealing with inner life (moods, desires, drive nuts) and with emotionally intense experiences (last six months, awe-inspir- ing). Question 3 might also have been scored in this way since some 70 per cent of the omissions were by highs. However, omissions on this item were not scored H because they did not appear to fulfill the requirement of theo- retical consistency. Should the highs continue, in future groups, to make most of the omissions, Nb would have to be scored H and a theoretical rationale found. The scoring of Nb as H would, in the present groups, have raised the over-all agreement several points.
Having considered the degree to which groups are distinguished with respect to the H and L categories, we may consider briefly how the scoring. scheme applies to the individual. It is possible to give each subject a total score which is the sum of his individual item scores. This total may be called
? PROJECTIVE QUESTIONS
the L-H score, and is computed as follows: One point is given for each H score, zero points for each L score, and one-half point for each Neutral score (except for Nb on Items I, 2, 4, 7, and 8 where, as noted above, Nb is con- verted to H). Thus, the L-H scores may range, for eight questions, between o (all L scores) and 8 (all H scores), with a mid-point of 4. 0. A subject receiving 4 H scores, 3 L's and I Na has an L-H score of 4. 5, i. e. , just on the high side of center. The over-all H% of 80. 7 for the high quartiles can be converted to an average L-H score of 6. 5 for the eight items. Similarly, the
over-all L% of 69. I for the low quartiles becomes an L-H score of 2. 5. The computing of L-H scores for each subject provides a means of deter- mining the amount of overlap between the low and high quartiles. This has been done in the case of the Middle-Class Men and Women, who have a combined N of 135, and who received a battery of five Projective Questions. The L-H scores of the low quartile ranged between o. o and 4? 5, those of the high quartile between 1. 0 and 5. 0, the mean for the total group being about 2. 7 (slightly more H than L scores were assigned). Using 2. 7 as a dividing point, we may then say that all L-H scores of 2. 5 and below will be called "low," all scores of 3. 0 and above will be called "high. " On the basis of this criterion, 2 2 per cent of the low quartile members would be considered high in terms of Projective Question score, while I4 per cent of the high quartile members have a low L-H score. These are the exceptions. Or, to put it posi-
tively, 78 per cent of the anti-ethnocentric group, and 86 per cent of the ethnocentric group, would be correctly diagnosed on the basis of total Pro- jective Question score. ?
The above data do not, of course, include the middle scorers on E. How- ever, it is not unlikely that a longer Projective Question Test of perhaps fifteen or twenty items, applied to all subjects and not merely to the extreme quartiles, might yield correlations in the neighborhood of ? 7 between L-H score and E. The Projective Questions might then be used, like the F scale, as an indirect measure of ethnocentrism-a measure in which no reference to current social issues need be made. They also, like the F scale, permit one to go beyond the immediate limits of the scoring scheme and to make numerQus inferences regarding individual dynamics. That the variables in the two techniques are so consistent is an important argument in favor of the validity of each.
In summary, the results indicate that the Projective Question Test meets current standards of reliability, under the most demanding of scoring conditions. It has also been demonstrated that the categories denoted as "high" are in fact characteristic of the high scorers on the Ethnocentrism scale, the "low" categories characteristic of the low quartile on E. There are, however, many highs who get L scores on single items, and many lows who get scores of H, the average PQ-E agreement being in the neighborhood of 75 per cent. In terms of individual L-H scores, the ethnocentrists and anti-
? Anger. (Na)
THE AUTHORITARIAN PERSONALITY
ethnocentrists could, in the two groups considered, be diagnosed with ap- proximately 82 per cent accuracy.
The present results can probably be improved in the future by modifica- tions of the Scoring Manual, by an increase in the number of items (which, other things being equal, will improve the reliability of the L-H score), and by instructions which lay greater stress on full answers (two or three sen- tences would do). Moreover, in its practical application the test may be scored in a clinically more meaningful way, once the scorers have demon- strated their competence. If all the items for a given individual were scored at once, many apparently ambiguous responses might be interpreted in the light of the total pattern. This would not only improve the quantitative scor- ing but would also stimulate further differentiation and elaboration within the broad framework now conceived. There is also considerable research which needs to be done regarding additional categories, relative frequencies of specific categories, sex differences, various patterns among lows and among highs, differences among various religious, political, and other groupings, and so on.
3. VALIDATION BY MEANS OF CASE STUDIES: MACK AND LARRY
Throughout the present research the attempt has been made to develop techniques that would yield statistically significant relationships among numerous variables and that would, as well, provide clinically meaningful material regarding the individual. Accordingly, following the presentation of statistical (group) results for each technique, the protocols of the high man, Mack, and the low man, Larry, have been considered in relation to the group data. To the extent that their results on a given technique are consistent with those for the total sample, and to the exent that interpreta- tions based on the technique are congruent with those derived from other techniques, further evidence of validity appears to be indicated.
The Projective Question responses of Mack and Larry are as follows (the scores being given at the end of each response):
MACK (High)
LARRY (Low)
1. Moods
Physical weakness, perhaps due to A lonesome mood, or a feeling that ill health over the last 4 years. (H) I am not progressing toward my goal, or a feeling that I have hurt
2 .
someone. (L)
Desires
When someone is persecuted un- justly, or to see a Negro service- man endure unjust discrimination and prejudice. (L)
? It depends a lot on definition, but if crazy is meant, such a thing as losing my wife and children would most closely approximate it. (H)
A person might become insane over too much interest in love, re- ligion, money, etc. Any obsession carried too far might drive a person insane. (L)
Murder, rape. (H)
Not included in this form.
5? Worst crimes
Murder, rape and a person that will incite hate toward another people, and groups that incite wars. (Nhl)
6. Embarrassing
7? Last six months
PROJECTIVE QUESTIONS
593
3? Great people
Lincoln, Lee, Gen. Geo. C. Mar- Lincoln, Willkie, Washington, Sta-
shall, Edison. (H)
lin, Chiang Kai Shek, Churchill, MacArthur, Eisenhower, and the great scientists of past and present. (H)
4? Drive nuts
Seeing all of the world possible, Trying to enjoy life as I have al- with a particular person to see it ways wanted; travel, meet im- with me. (H) portant people, have lots of friends,
go to a lot of parties. All this with- out harming anyone. (L predomi- nantly, despite some H trends. )
8. Awe-inspiring
The viewing of great natural won- ders of the world, new scientific achievements, meeting really great men and women and the sight of rare jewels and metals. (H)
T o see American people practice true democracy. Such incidents as a member of one race protects a member of another race. True comradeship between races. (L)
The scoring, done blindly and with complete agreement between two independent judges, gives Mack an L-H score of 6. 5 out of 7 (or an H% of 93), Larry an L-H score of 1. 5 (or an L% of 79). Mack's responses are rela- tively typical of those given by the high men; if he shows no bizarre features, neither does he show much individuality. Given an opportunity to be emo- tionally expressive (Items 7 and 8), he responds in a shallow, conventional- ized manner; his concern with "looking," with a minimum of differentiation or focal affect, appears to be based on a deep but inhibited curiosity for which he has few constructive expressions. His tendency to align himself with power and success is expressed in Item 3 and again in Item 8 ("meeting really great men and women"). Item I reveals his anxiety over bodily harm and his conception of ill health as a form of weakness (rather than, for example,
? 594
THE AUTHORIT ARIAN PERSONALITY
a barrier in the way of achievement). The equation of sickness with weak- ness is particularly interesting in the light of his mother's and his own weakness during childhood (see Chapters II and XX).
It suggests, as does the material from other techniques, that his surface identification with powerful figures and groups is at least in part a means of maintaining his sense of mastery and of allaying his anxiety over bodily harm. It is not clear from his Projective Question responses alone whether Mack's aggression is more a surface defense against the admission of passive dependency or, rather, a strong underlying need. That he has aggressive impulses which he cannot easily assimilate into the ego is suggested by his "Murder, rape" on Item 5 and "Anger" on 2. Moreover, that these impulses are partially directed against family (ingroup) members is suggested by his response to Item 4: his fear of "losing my wife and children" (particularly since he is not yet married) would seem to be based on unconscious hostility toward them-hostility which is projected onto the "threatening world. " (His fantasy, while he is still unmarried, of the death of wife and children may also be a projection of his own fear as a child that he might die when his mother did. ) However, the over-all impression given is that of a conventional, deferent, pseudo- independent fac,;ade, and that what lies beneath the surface is primarily anxiety
and dependency rather than active destructiveness.
Larry's responses, here as elsewhere, are less characteristically low than
Mack's are high. Like Mack, Larry is attracted to those who have power, but his conception of power is different from Mack's. Thus, Larry can admire foreign as well as American leaders (Item 5: in 1945 Chiang Kai Shek was still conceived as a democratic leader). And his relationship to power figures seems to be based more on the open expression of dependency and need for support than on defense against fear of his own weakness. Indeed, Larry is openly and intensely identified with the weak and the helpless, and he can therefore be opposed to social authority when it mistreats Negroes and others (Items 2, 5, and 8). He is also characteristically low in his intra- ception, intrapunitiveness, achievement values, conscious guilt, and the like.
Larry's "timid dependency" is clearly expressed in his response to Item 7: what he wants most is to be loved and protected-"all this without harming anyone. " The great fear of hurtng anyone, expressed also on Item 1, seems to imply deep-lying hostility which he must at all costs inhibit. Thus, his inhibition of aggression, combined with his tremendous love-seeking de- pendency, probably prevent Larry from fighting actively even for those things in which he believes most. He is disturbed by discrimination and he would be awe-inspired by the attaining of democratic group relations in America, but is not likely to be able militantly to oppose those who violate his basic values.
It cannot be said that the interpretations above were made in the absence of other clinical material on these two cases. However, the scoring was done
? PROJECTIVE QUESTIONS
595
in a controlled manner and many of the inferences follow fairly directly from the theory contained, in the Scoring Manual. The conclusions reached here . regarding Mack and Larry are in general agreement with those derived from other techniques, for example the F scale (Chapter VII) and the Thematic Apperception Test (Chapter XIV). It would seem, then, that the Projective Question technique may fruitfully be used not only for purposes of group research but also as an aid in the intensive study of the individual case.
E. CONCLUSIONS
It was apparent in the Scoring Manual that certain themes were repeated, with perhaps minor variations in form or content, in many of the Projective Question categories. There are, moreover, several instances in which two or more recurring' themes, taken together, permit inferences regarding deeper-lying trends and processes. The results and theoretical constructions derived from the application of this technique are similar to those obtained from the other techniques used in the present research. In some cases there is almost exact duplication of variables; in others there is a more complemen- tary or congruent relation, the variables from several techniques expressing diverse facets of a single, inclusive structure. Because the amount of duplica- tion is considerable, a very brief discussion of the theoretical implications of the present results will suffice. Differential trends for high and low scorers on the E scale seem to exist in the following areas.
1. General Ego Functioning. Highs and lows differ markedly in their man- ner of handling deep-level trends such as aggression, sex, dependency, anxiety, and the like. We are not yet in a position to say whether one group or the other shows a greater total amount of any given trend; what is clear is that both groups exhibit all of these trends to a significant degree. The primary difference seems to lie in the ego functioning, and particularly in the relation of the ego to the deeper levels of personality. In the lows, as other techniques have shown, the underlying trends are more ego-assimilated, in the highs more ego-alien.
The lows appear to differ from highs in at least the following respects. The relations between the various levels of personality are more fluid, the boundaries more permeable. The ego defenses of the lows are relatively more impulse-releasing: at best we find considerable sublimation, to perhaps a greater degree we find that impulses have been assimilated into the ego with- out being fully integrated-witness the recurring Projective Question cate- gory, "Conscious conflict and guilt. " In the highs, on the other hand-and the analysis of interviews led to the same conclusion-the ego defenses are characteristically more countercathectic; there is less sublimation and more use of defenses such as projection, denial, and reaction-formation, defenses which aid the individual in maintaining a moral fa<;:ade at the expense of self- expression and emotional release.
? THE AUTHORITARIAN PERSONALITY
These formulations are supported by many of the Projective Question results. Compare, for example, the low category, "Conscious conflict and guilt," with the contrasting high category, "Rumbling from below" (Ques- tions I, "Moods," and 4, "Drive nuts"). The former category refers to im- pulses which, disturbing though they may be, the individual at least to some extent recognizes and tries to handle. If there is conflict, there is also an attempt to integrate; if there is much that remains unconscious, there is also a willingness to look within and an attempt to assimilate. For most highs, on the other hand, there are few focal conflicts but there is a deep sense of anxiety and distress. The conflict is covered over by a moral fac;ade or by symptomatic behavior; the disturbance is explained on the basis of a dis- tressing (overdemanding, boring) external situation or of poor bodily con- dition, and the conflict is never faced in psychological terms. The difficulty of the highs in assimilating many important needs is shown by several other categories. Thus, when asked about highly satisfying emotional experiences
("Last six months" and "Awe-inspiring") they show much less intensity and inner vitality than the lows, that is, much less ability to utilize their psychic energies for constructive and ego-satisfying purposes. When the highs refer more directly to needs such as dependency, sex, and aggression ("Desires" and "Worst crimes"), they are more crude, impersonal, primitive, object-less and ego-less.
2. Specific Properties of the Ego. These properties are, of course, inti- mately bound up with ego functioning as discussed above. As might be expected from their use of primarily countercathectic defenses, the highs have comparatively narrow, circumscribed egos. One manifestation of this narrowness is constriction of fantasy (particularly in Question 6, see also Chapter XIV); this is probably related to the highs' emphasis on "sticking to the facts," to their extraceptiveness, and to their rejection of "imagina- tion" and "emotion. "
The highs also diifer from the lows in their greater concreteness of think- ing and in their less differentiated emotional experience. Thus, the responses of the highs characteristically refer to specific behavioral acts or situations, while the lows tend, to a greater degree, to describe more subtle experiences and more abstract, generalized situations. On Question 3 ("Great people"), for example, the lows, more than the highs, give the qualities that they admire in a person, with or without specific examples. The same is true for Question 5 ("Worst crimes"). Evidence of the greater complexity and abstractness of the lows has been given in the interview chapters (Part II) and in Chapter VIII (showing negative correlations between ethnocentrism and intelli- gence). These results are consistent with those showing the highs to be more stereotyped in their ideology (Chapters III, IV, XVI, XVII), more rigidly concrete in their solutions of arithmetical and spatial probkms (Rokeach
(98)) 0
? PROJECTIVE QUESTIONS
597
The lows tend, in a frustrating situation, to blame themselves (intrapuni- tiveness), the highs to idealize themselves and to see the evil as existing in the external world (extrapunitiveness).
Finally, the greater intraception of the lows and the great anti-intracep- tion of the highs is apparent in their responses to the Projective Questions as well as in the material elicited by the interviews, the T. A. T. , and the F scale.
3? Achievement Values vs. Conventional Values. This distinction has been essential for the scoring of the Projective Questions. It is important not only because of the difference in the content of the values, but also because the values themselves express significant aspects of the personality dynamics of the two groups. It should be noted first that the two sets of values are not entirely mutually exclusive; few individuals will have only one set or the other. It is, rather, a matter of degree and of primacy. In most individuals one set of values is likely to be primary and most potent, the other to be of secondary importance. An individual who is struggling to decide between these two value orientations is, we believe, essentially struggling to decide, consciously or unconsciously, between conflicting needs and between con- flicting conceptions of himself as a total person.
Achievement values found predominantly in the lows place primary em- phasis on self-expression. Abstract and open-ended, they always leave room for further development and they can never be defined in terms of simple behavior formulae or rigid rules. Their main emphasis is on long-range goals, and the attainment of a given goal leads always to the formulation of new, higher goals expressing the same basic values. Examples of achievement values, taken from the Projective Question material as well as from the inter- views and other material, include the following: Value for scientific, intel- lectual, and aesthetic achievement, and for understanding for its own sake, regardless of immediate practical application. Creativity is valued above efficiency, constructiveness above practicality; productive living, even if it involves inner conflict, is preferred over good adjustment at the expense of
self-expression; richness and intensity of inner experience are valued more than "mere contentment. " In personal relationships, as other techniques have shown, there is concern with giving as well as taking, and with the exchange of love rather than the exchange of things. What is particularly important here is that recognition of one's own individuality is the basis for recognition of the individuality of everyone, and for the democratic concept of the dignity of man. 8 These values are expressed ideologically in terms of op- position to all social structures (milita1y, religious, educational, politico-
8 This point has also been made by Fromm (43). His distinction between "humanistic" and "authoritarian" ethics corresponds very closely to the present one between "achieve- ment" and "conventional" values, and is based on a similar attempt to distinguish two broad psychological approaches to man and society.
? THE AUTHORITARIAN PERSONALITY
economic) which are based on the principle of absolute authority, which value power more than love, which engage in group suppression and ex- ploitation, in short, which prevent man from developing his innate poten- tialities to a maximum degree. Once again we find anti-ethnocentrism as but one facet of a larger psychological framework.
What has been called "conventional values" might also have been called "conditional values," since their main function is to place limitations or con- ditions on the expression of needs rather than to stimulate need-experience. They might also have been called "authoritarian values," since they are based on the assumption of conformity to external authority rather than on inner moral responsibility. Whatever the name given, both of the above meanings, as well as others, belong to its definition. The prototypic examples of this value system are the Emily Post book of etiquette, the military "rules of behavior," and certain custom-ridden cultures, literate and nonliterate. 9 The main content of these values, at least for individuals with a strong middle- class identification, deals with conformity and loyalty to ingroup standards.
The difference between achievement values based on inner authority (in- ternalized conscience), and conventional values based on external authority (and thus replaceable when the authority changes), results also in a difference in reaction to value-violations. This is the difference between guilt and shame. It would be an exaggeration to state that the lows feel no shame, the highs no guilt. Nevertheless, there is considerable evidence from the Projective Questions, as well as from the ratings of interviews, that guilt is most charac- teristic of lows, shame of highs. The low categories for Questions 1, z, 4, and 6 are for the most part concerned with personal violations of achievement values, with practically no reference to "being caught" or to external author- ity. For the same items there are high categories referring to violations of conventional values most of which require, almost by definition, an external observing and punishing agent. Indeed, the explicit idea of inner conflict in any form is practically lacking from the high responses. Moreover, it is much more common for the lows to refer to their own personal violations of values, whereas the highs refer either to violations by others, or, more often, to
events which have, explicitly, neither moral nor motivational significance. 10 These results are in keeping with the findings reported in earlier chapters that the highs, particularly those who are more conventional and "middle- class," have a punitive but poorly internalized superego. The ego, submitting
9 See Kardiner (;g, 6o) and Benedict (15) for nonliterate societies, and Reich (g6); these are but a few of many examples.
10 W e should have to predict-and there is already considerable supporting evidence- that studies of specific cultures will show a connection between shame as a predominant emotion, threat of punishment or isolation as a means of discipline in raising children, em- phasis on ritual and custom, an authoritarian conception of God, relative lack of achieve- ment values, and ethnocentrism in group relations within the culture and with other cultures. The higher E scale means made by the more ritualized, fundamentalistic religions in our own culture (see Chapter VI) may be cited in support of this hypothesis.
? PROJECTIVE QUESTIONS
599
out of fear, must constantly forego conscious, constructive impulse gratifi- cation; instead, it finds morally acceptable ways of gaining indirect satisfac- tion (e. g. , aggression by means of ethnocentrism and moral indignation, dependency through submission to powerful authority), and it "cheats" the superego when fear of detection is minimized (e. g. , at conventions). Again we have a contradiction in levels: The highs, so moral on the surface, are essentially most concerned with underlying anxiety and with the gratification of impulses which, being ego-alien, have developed but little beyond their primitive, infantile form; whereas the lows, often so rebellious and so op- posed to traditional morality on the surface, have more fully internalized moral principles and in their emotional functioning are more troubled with moral conflict.
4? The Handling of Dependency as an Underlying Trend. General differ- ences in the ego functioning of lows. and highs have been discussed above. We may turn now to a particular disposition, namely dependency, and see what light the Projective Questions shed on its differential ego-assimilation in lows and highs. It should be noted that there are certain sex differences here, since the expression of dependency is culturally permitted, even valued, in women, whereas in men it is opposed and inhibited. For convenience the following discussion will focus on high versus low men, with the under- standing that for women some of these differences are somewhat reversed, while others hold equally well and still others are not found.
Dependency in lows is expressed mainly in the form of concern with love; many of their Projective Question responses deal with love-giving (nurtur- ance) and love-seeking (active, focal dependency). They seem highly con- cerned with emotional exchange in their personal relationships.
Dependency plays a much different role in the personality of high men. It remains for the most part an ego-alien trend which can seldom be expressed directly because it violates the image (ego ideal) of the normal, masculine man: rugged, practical, realistic, earthbound, independent, "normally sexual and acquisitive, ready to take an active part in the bitter competition de- manded by human nature, and eager to rise to the top of the ladder of suc- cess. " It is apparent that not all men who have this self-ideal are high with respect to E or personality. Nevertheless, the present results indicate that most high men have this ideal, and that most men with this ideal are high.
Part of the high man's defense against ego-alien passive dependency-it is not the love-seeking dependency of the lows-is the rigid value for work, and anxiety over dependency is expressed in the idea that overwork would ? drive one nuts (Projective Questions 1 and 4).
If dependency promotes a concern with love in the lows, it promotes a concern with power in the highs. One of the more direct forms in which high dependency is expressed is submission to power figures. Whereas the ego-assimilated dependency of the lows is expressed in their value for equali-
? 6oo THE AUTHORITARIAN PERSONALITY
tarian relationships and social structures, the ego-alien dependency of the highs leads to the acceptance of absolute authority and to a value for authori- tarian forms of social interaction. As shown in Projective Question 3 ("Great people"), what the highs admire most in others is power, strength, authority, rugged masculinity. While the aggressive-assertive needs of authoritarian individuals are the most conspicuous ones, the dependent-submissive needs are equally if not more important. In the Projective Questions, particularly 7 ("Last six months") and 8 ("Awe-inspiring"), it was the highs' deference to authority which was expressed most strongly.
The ego-alien passive dependency of the highs is expressed in another form in Projective Questions I ("Moods") and 4 ("Drive nuts"). In Question I, the most unpleasant moods are those involving a feeling of helplessness and dependency in the face of a threatening or barren environment. The dependency comes to the surface but meets an unwelcome reception in the weak, superego-ridden ego: there are no persons toward whom it can be satisfyingly expressed; there is no differentiated affect to make it an enrich- ing experience; there is only a vague anxiety which is actually based not on the external situation but on a deep inner conflict between the superego and the upsurging primitive dependency. It is as if to express the impulse would be to lose one's masculinity or even to undergo bodily harm. Similar conflicts are expressed in Question 4? Again we find the inability to look inward and the vague rumblings of ego-alien trends, including body anxiety and aggres-
sion as well as dependency.
5? The Handling of Other Trends. There is evidence that aggression, sex,
curiosity, homosexuality, and other trends are handled by highs and lows in ways similar to those discussed above. One additional point should be made regarding aggression. It appears not to have undergone much real socializa- tion in highs but has, rather, remained relatively crude, destructive, punitive, unsoftened by ego-assimilation. If the term "hostility" be used for the ag- gression found in lows, then the term "destructiveness" seems most appropri- ate for what exists in highs, especially as revealed in Questions 2 ("Desires"), 5 ("Crimes"), and 8 ("Awe-inspiring").
The results and interpretations discussed above must not be applied in a stereotyped way. It should be understood that, while most highs show most of the high variables, and similarly for the lows, there are numerous excep- tions and numerous variations on the central theme. The personality proto- type above was that of the pseudodemocratic high, and other patterns, such as the "fascist leader" or the "psychopath" will be different in many respects. It is believed, nevertheless, that the Projective Question technique has yielded results comparable, and congruent, with those of other techniques, and that it may profitably be used for the study of other personality structures.
? PART IV
INTRODUCTORY REMARKS
The present volume has so far offered findings from our research ranging from surface ideology to largely unconscious psychological traits of our subjects. The direction of research and the order of presentation were sug- gested by the nature of the ideological data themselves; they could not be derived solely from external factors, such as economic status, group mem- bership, or religion; but rather. the evidence pointed unmistakably to the role played by motivational forces in the personality. However, the study did not move mechanically from the ideological to the psychological; rather, we were constantly aware of the structural unity of the two. It thus seems permissible that we reverse the procedure now and ask: what is the mean- ing of the subjects' overt opinions and attitudes in the areas covered by the A-S, E, and PEC scales, when they are considered in the light of our psycho- logical findings, particularly those deriving from the F-scale and the clinical sections of the interviews? By answering this question we may come closer to an integration of the various aspects of a study which is centered in the problem of the relationship between ideology and personality.
As was natural, the material for this task was mainly taken from the non- clinical parts of the interviews. Not only did these data promise to yield additional evidence bearing on the major issues discussed thus far, but the wealth of detailed and elaborate statements which our subjects had formu- lated spontaneously and in their own way, offered numerous psychological leads. There is good reason to believe that the non-clinical sections of the interviews constitute through their inherent structure a link between ideology and personality. However, attention was not limited to this inter- relationship; at the same time an attempt was made to obtain a more colorful picture of the various ideologies themselves than was possible as long as we limited ourselves to the standard questionnaires.
Since the data from the questionnaire and from the Thematic Apper- ception Test and the clinical parts of the interviews had been subjected to thorough statistical treatment, quantification of the present material, though desirable, did not seem necessary. The aim, rather, was to develop for the problem areas under consideration, a phenomenology based on theoretical formulations and illustrated by quotations from the interviews. This pro- cedure, it was hoped, would yield not only more information about the specific structure of the ideologies and the manner in which personality is expressed in them but also a further differentiation of the guiding theoretical concepts themselves.
The advantages of this supplementary procedure are several. It permits 603
? THE AUTHORITARIAN PERSONALITY
us to exploit the richness and concreteness of "live" interviews to a degree otherwise hardly attainable. What is lost for want of strict discipline in in- terpretation may be gained by flexibility and closeness to the phenomena. Rare or even unique statements may be elucidated by the discussion. Such statements, often of an extreme nature, may throw considerable light on potentialities which lie within supposedly "normal" areas, just as illness helps us to understand health. At the same time, attention to the consistency of the interpretation of these statements with the over-all picture provides a safeguard against arbitrariness.
A subjective or what might be called speculative element has a place in this method, just as it does in psychoanalysis, from which many of our categories have been drawn. If, in places, the analysis seems to jump to con- clusions, the interpretations should be regarded as hypotheses for further research, and the continuous interaction of the various methods of the study should be recalled: some of the measured variables discussed in earlier chapters
were based on speculations put forward in this part.
In view of the discussions in Chapters III and IV it was not deemed nec-
essary to differentiate between A-S and E in the treatment of the interview material. While the generally close correlation of anti-Semitism and ethno- centrism could be taken for granted on the basis of previous results, more specific accounts of the nature of their interrelation, as well as of certain deviations, were incorporated into the first chapter of the present part
(Chapter XVI).
The chapter which discusses various syndromes found in high and low
scorers (XIX) is also included in this part. Although from a strictly logical point of view it may not belong here, it seemed nevertheless appro- priate to include it, since it is based almost entirely on interview material and focused on the interconnection between ideology and personality. The syndromes evolved in this chapter should be followed up by quantitative investigation. 1
1 We have not deemed it necessary to establish cross references between interviewees' statements presented here-under interview numbers-and those given in Part II under code numbers (see Chapter X, p. 342). Therefore, some quotations may appear here which have already been given there, in a different connection. However, as twelve of the San Quentin inmates are dealt with as a special group in a later chapter (XXI), a Key linking the interview numbers used here with the fictitious names assigned to them there has been inserted on the bottom of Table r (XXI).
? CHAPTER XVI
PREJUDICE IN THE INTERVIEW MATERIAL
T. W. Adorno
A. INTRODUCTION
Our study grew out of specific investigations into anti-Semitism. As our work advanced, however, the emphasis gradually shifted. We came to re- gard it as our main task not to analyze anti-Semitism or any other anti- minority prejudice as a sociopsychological phenomenon per se, but rather to examine the relation of antiminority prejudice to broader ideological and characterological patterns. Thus anti-Semitism gradually all but dis- appeared as a topic of our questionnaire and in our interview schedule it was only one among many topics which had to be covered.
Another investigation, carried through parallel to our research and partly by the same staff members of the Institute of Social Research, i. e. , the study on anti-Semitism within labor (57B), concentrated on the question of anti- Semitism, but at the same time was concerned with sociopsychological issues akin to those presented in the present volume. While the bulk of the material to be discussed in this chapter is taken from the section on prejudice of the Berkeley interviews, an attempt was made to utilize, at least in a supplemen- tary form, some of the ideas of the Labor Study as hypotheses for further investigation. This was done as a part of the work carried out in Los Angeles. In collaboration with J. F. Brown and F. Pollock we drew up an additional section of the interview schedule devoted to specific questions about Jews. These questions were derived for the most part from the material gathered through the "screened interviews" of the Labor Study. The aim of this new section of the interview schedule was to see if it was possible to estab- lish certain differential patterns within the general structure of prejudice.
The list of questions follows. Not all of these questions were put to every subject, nor was the exact wording of the questions always the same, but most of the ground marked off by the questions was covered in each case.
72. 1 41)
0
~ ~
65 33
32
65
43. 9 71. 2
>z
"d
M
~
en 0 z
> 20) t"" ~ 15)21. 2% . . . ,
Na-NHL 2
1 9
2
6
20 )17. 4% ~ ~
62. 1 Nb37316 Na-NHL 4 1 4 3 3
L% 69. 7
71. 2
63. 6
H% 81. 2 78. 1 81. 2 76. 6 75. 0
Nb 6 10 4 3 4 Na-NHL 0 0 6 1 8
%A 75. 4 60. 8 76. 1 69. 2 73. 1
Nb 9 17 7 4 10 Na-NHL 4 1 10 4 11
78. 4
27)
15 )26. 3%
70. 9
47)
30 )23. 7%
31) . . , 11)23. 3% ! I:
-<
? 7. All groups combined Low quartile
High quartile
Total
L% 76. 7 63. 0 75. 8 68. 4 65. 8 63. 0 57. 0 85. 9 69. 1
Nb 9 19 8 3 14 7 0 1 61 6. 6% Na-NHL 8 8 22 20 12 11 13 1 95 10. 3%
ER 146 146 155 155 155 73 64 32 926 116
m; 84. 0 79. 9 71. 3 84. 4 86. 0 7i. 3 85. 9 72. 7 80. 7
Nb 15 33 17 9 11 7 2 5 99 10. 6% Na-NHL 4 4 27 9 15 3 4 2 68 7. 3%
ER 144 144 157 157 157 77 64 33 933 117
%A 80. 3 71. 4 73. 6 76. 5 76. 0 69. 3 71. 5 79. 2 74. 9
Nb 24 52 25 12 25 14 2 6 160 8. 6% Na-NHL 12 12 49 29 27 14 17 3 163 8. 8%
ER 290 290 312 312 312 150 J 2 8 65 1859 232
Score 1 2 3 4 5 6 7 8 Total Mean
aFor Groups 2, 3, and 4 the number of subjects varies somewhat from item to item because two questionnaire forms were used. The number indicated in the "N" column is the total N in each group or quartile. The N in parenth~ses refers to those subjects who received the complete set of projective questions. Thus, in Group 2 all 51 subjects received items 1 - 5 (Forms 45 and 40); only 10 Low and 14 High received all 8 questions (Form 45), so that the N's on items 6, 7, and 8 are 10 and 14 respectively. as indicated in parentheses in the "N" column. Similarly,
of the 14 Lows in Group 3, only 9 received the full battery, the remaining 5 receiving a form which did not in? clude items 1, 2, 7, and 8. For the high quartile in Group 3, there is an Nof 8 on items 1, 2, 7, and 8. In Group 4, the N on these same items is 13 for Lows and 11 for Highs.
Group 1 received a form which did not include item 8 (see text). Groups 5 and 6 received Form 40 which contained items 1-5 only.
Thus, while there were 312 subjects in all, the N per item varied. This is shown in the table under ~All groups combined," where the row "ER" gives the number of subjects answering each item.
bin the computation of all means, the component values are weighted by N.
Key to questions: 1 (Moods), 2 (Desires), 3 (Great People), 4 (Drive Nuts), 5 (Crimes), 6 (Embarrassing), 7 (Last Six Months), 8 (A we-inspiring).
? THE AUTHORITARIAN PERSONALITY
The average agreement varies among the six groups from 67? 9 per cent for the Psychiatric Clinic Men to S2. 6 per cent for the Employment Service Veteran Men. It appears that the scorers, try as they did to be unbiased, were systematically influenced by the fact that they knew the E-quartile standing of the Psychology Women and the Veteran Men. Thus, the %A for these two groups is about So per cent, whereas for the remaining four groups
(scored blindly) it averages slightly over 70 per cent. The relatively low %A for the Clinic Men is consistent with the low scoring reliability for this group (see Table I(XV)). The data for the Clinic Women and the Middle- Class Men and Women probably best represent what can be expected with the present Scoring Manual; the average agreement of about 72 per cent for these groups might have been slightly higher had all subjects received questions 6, 7, and S.
The over-all average agreement for the individual items varies from 69. 3 to So. 3 per cent. The poorest questions are 2 ("Desires"), 6 ("Embarrass- ing"), and 7 ("Last six months"). In all three of them the L% is particularly low (57-63 per cent), probably because of the brevity of the answers, as discussed above. Correction of this error should lead to considerably better results, particularly for Item 7? The agreements for the individual items re- veal again the great consistency of the highs. Thus, the L% varied within a range of 29 points (57-S6 per cent), whereas the H% covered a range of only 15 points (7I-S6 per cent). It is of some interest that the two most dif- ferentiating items, 1 ("Moods") and S ("Awe-inspiring"), deal with issues which, in their literal meanings, are completely removed from ideology about group interaction.
Table 2(XV) provides the empirical basis on which omissions (Nb) of Questions I, 2, 4, 7, and S were converted into H scores in the statistical treat- ment. On each of these questions approximately two-thirds or more of the total number of omissions were made by the high quartiles. Moreover, it is consistent with the differential trends found that the highs should omit, more often than the lows, questions dealing with inner life (moods, desires, drive nuts) and with emotionally intense experiences (last six months, awe-inspir- ing). Question 3 might also have been scored in this way since some 70 per cent of the omissions were by highs. However, omissions on this item were not scored H because they did not appear to fulfill the requirement of theo- retical consistency. Should the highs continue, in future groups, to make most of the omissions, Nb would have to be scored H and a theoretical rationale found. The scoring of Nb as H would, in the present groups, have raised the over-all agreement several points.
Having considered the degree to which groups are distinguished with respect to the H and L categories, we may consider briefly how the scoring. scheme applies to the individual. It is possible to give each subject a total score which is the sum of his individual item scores. This total may be called
? PROJECTIVE QUESTIONS
the L-H score, and is computed as follows: One point is given for each H score, zero points for each L score, and one-half point for each Neutral score (except for Nb on Items I, 2, 4, 7, and 8 where, as noted above, Nb is con- verted to H). Thus, the L-H scores may range, for eight questions, between o (all L scores) and 8 (all H scores), with a mid-point of 4. 0. A subject receiving 4 H scores, 3 L's and I Na has an L-H score of 4. 5, i. e. , just on the high side of center. The over-all H% of 80. 7 for the high quartiles can be converted to an average L-H score of 6. 5 for the eight items. Similarly, the
over-all L% of 69. I for the low quartiles becomes an L-H score of 2. 5. The computing of L-H scores for each subject provides a means of deter- mining the amount of overlap between the low and high quartiles. This has been done in the case of the Middle-Class Men and Women, who have a combined N of 135, and who received a battery of five Projective Questions. The L-H scores of the low quartile ranged between o. o and 4? 5, those of the high quartile between 1. 0 and 5. 0, the mean for the total group being about 2. 7 (slightly more H than L scores were assigned). Using 2. 7 as a dividing point, we may then say that all L-H scores of 2. 5 and below will be called "low," all scores of 3. 0 and above will be called "high. " On the basis of this criterion, 2 2 per cent of the low quartile members would be considered high in terms of Projective Question score, while I4 per cent of the high quartile members have a low L-H score. These are the exceptions. Or, to put it posi-
tively, 78 per cent of the anti-ethnocentric group, and 86 per cent of the ethnocentric group, would be correctly diagnosed on the basis of total Pro- jective Question score. ?
The above data do not, of course, include the middle scorers on E. How- ever, it is not unlikely that a longer Projective Question Test of perhaps fifteen or twenty items, applied to all subjects and not merely to the extreme quartiles, might yield correlations in the neighborhood of ? 7 between L-H score and E. The Projective Questions might then be used, like the F scale, as an indirect measure of ethnocentrism-a measure in which no reference to current social issues need be made. They also, like the F scale, permit one to go beyond the immediate limits of the scoring scheme and to make numerQus inferences regarding individual dynamics. That the variables in the two techniques are so consistent is an important argument in favor of the validity of each.
In summary, the results indicate that the Projective Question Test meets current standards of reliability, under the most demanding of scoring conditions. It has also been demonstrated that the categories denoted as "high" are in fact characteristic of the high scorers on the Ethnocentrism scale, the "low" categories characteristic of the low quartile on E. There are, however, many highs who get L scores on single items, and many lows who get scores of H, the average PQ-E agreement being in the neighborhood of 75 per cent. In terms of individual L-H scores, the ethnocentrists and anti-
? Anger. (Na)
THE AUTHORITARIAN PERSONALITY
ethnocentrists could, in the two groups considered, be diagnosed with ap- proximately 82 per cent accuracy.
The present results can probably be improved in the future by modifica- tions of the Scoring Manual, by an increase in the number of items (which, other things being equal, will improve the reliability of the L-H score), and by instructions which lay greater stress on full answers (two or three sen- tences would do). Moreover, in its practical application the test may be scored in a clinically more meaningful way, once the scorers have demon- strated their competence. If all the items for a given individual were scored at once, many apparently ambiguous responses might be interpreted in the light of the total pattern. This would not only improve the quantitative scor- ing but would also stimulate further differentiation and elaboration within the broad framework now conceived. There is also considerable research which needs to be done regarding additional categories, relative frequencies of specific categories, sex differences, various patterns among lows and among highs, differences among various religious, political, and other groupings, and so on.
3. VALIDATION BY MEANS OF CASE STUDIES: MACK AND LARRY
Throughout the present research the attempt has been made to develop techniques that would yield statistically significant relationships among numerous variables and that would, as well, provide clinically meaningful material regarding the individual. Accordingly, following the presentation of statistical (group) results for each technique, the protocols of the high man, Mack, and the low man, Larry, have been considered in relation to the group data. To the extent that their results on a given technique are consistent with those for the total sample, and to the exent that interpreta- tions based on the technique are congruent with those derived from other techniques, further evidence of validity appears to be indicated.
The Projective Question responses of Mack and Larry are as follows (the scores being given at the end of each response):
MACK (High)
LARRY (Low)
1. Moods
Physical weakness, perhaps due to A lonesome mood, or a feeling that ill health over the last 4 years. (H) I am not progressing toward my goal, or a feeling that I have hurt
2 .
someone. (L)
Desires
When someone is persecuted un- justly, or to see a Negro service- man endure unjust discrimination and prejudice. (L)
? It depends a lot on definition, but if crazy is meant, such a thing as losing my wife and children would most closely approximate it. (H)
A person might become insane over too much interest in love, re- ligion, money, etc. Any obsession carried too far might drive a person insane. (L)
Murder, rape. (H)
Not included in this form.
5? Worst crimes
Murder, rape and a person that will incite hate toward another people, and groups that incite wars. (Nhl)
6. Embarrassing
7? Last six months
PROJECTIVE QUESTIONS
593
3? Great people
Lincoln, Lee, Gen. Geo. C. Mar- Lincoln, Willkie, Washington, Sta-
shall, Edison. (H)
lin, Chiang Kai Shek, Churchill, MacArthur, Eisenhower, and the great scientists of past and present. (H)
4? Drive nuts
Seeing all of the world possible, Trying to enjoy life as I have al- with a particular person to see it ways wanted; travel, meet im- with me. (H) portant people, have lots of friends,
go to a lot of parties. All this with- out harming anyone. (L predomi- nantly, despite some H trends. )
8. Awe-inspiring
The viewing of great natural won- ders of the world, new scientific achievements, meeting really great men and women and the sight of rare jewels and metals. (H)
T o see American people practice true democracy. Such incidents as a member of one race protects a member of another race. True comradeship between races. (L)
The scoring, done blindly and with complete agreement between two independent judges, gives Mack an L-H score of 6. 5 out of 7 (or an H% of 93), Larry an L-H score of 1. 5 (or an L% of 79). Mack's responses are rela- tively typical of those given by the high men; if he shows no bizarre features, neither does he show much individuality. Given an opportunity to be emo- tionally expressive (Items 7 and 8), he responds in a shallow, conventional- ized manner; his concern with "looking," with a minimum of differentiation or focal affect, appears to be based on a deep but inhibited curiosity for which he has few constructive expressions. His tendency to align himself with power and success is expressed in Item 3 and again in Item 8 ("meeting really great men and women"). Item I reveals his anxiety over bodily harm and his conception of ill health as a form of weakness (rather than, for example,
? 594
THE AUTHORIT ARIAN PERSONALITY
a barrier in the way of achievement). The equation of sickness with weak- ness is particularly interesting in the light of his mother's and his own weakness during childhood (see Chapters II and XX).
It suggests, as does the material from other techniques, that his surface identification with powerful figures and groups is at least in part a means of maintaining his sense of mastery and of allaying his anxiety over bodily harm. It is not clear from his Projective Question responses alone whether Mack's aggression is more a surface defense against the admission of passive dependency or, rather, a strong underlying need. That he has aggressive impulses which he cannot easily assimilate into the ego is suggested by his "Murder, rape" on Item 5 and "Anger" on 2. Moreover, that these impulses are partially directed against family (ingroup) members is suggested by his response to Item 4: his fear of "losing my wife and children" (particularly since he is not yet married) would seem to be based on unconscious hostility toward them-hostility which is projected onto the "threatening world. " (His fantasy, while he is still unmarried, of the death of wife and children may also be a projection of his own fear as a child that he might die when his mother did. ) However, the over-all impression given is that of a conventional, deferent, pseudo- independent fac,;ade, and that what lies beneath the surface is primarily anxiety
and dependency rather than active destructiveness.
Larry's responses, here as elsewhere, are less characteristically low than
Mack's are high. Like Mack, Larry is attracted to those who have power, but his conception of power is different from Mack's. Thus, Larry can admire foreign as well as American leaders (Item 5: in 1945 Chiang Kai Shek was still conceived as a democratic leader). And his relationship to power figures seems to be based more on the open expression of dependency and need for support than on defense against fear of his own weakness. Indeed, Larry is openly and intensely identified with the weak and the helpless, and he can therefore be opposed to social authority when it mistreats Negroes and others (Items 2, 5, and 8). He is also characteristically low in his intra- ception, intrapunitiveness, achievement values, conscious guilt, and the like.
Larry's "timid dependency" is clearly expressed in his response to Item 7: what he wants most is to be loved and protected-"all this without harming anyone. " The great fear of hurtng anyone, expressed also on Item 1, seems to imply deep-lying hostility which he must at all costs inhibit. Thus, his inhibition of aggression, combined with his tremendous love-seeking de- pendency, probably prevent Larry from fighting actively even for those things in which he believes most. He is disturbed by discrimination and he would be awe-inspired by the attaining of democratic group relations in America, but is not likely to be able militantly to oppose those who violate his basic values.
It cannot be said that the interpretations above were made in the absence of other clinical material on these two cases. However, the scoring was done
? PROJECTIVE QUESTIONS
595
in a controlled manner and many of the inferences follow fairly directly from the theory contained, in the Scoring Manual. The conclusions reached here . regarding Mack and Larry are in general agreement with those derived from other techniques, for example the F scale (Chapter VII) and the Thematic Apperception Test (Chapter XIV). It would seem, then, that the Projective Question technique may fruitfully be used not only for purposes of group research but also as an aid in the intensive study of the individual case.
E. CONCLUSIONS
It was apparent in the Scoring Manual that certain themes were repeated, with perhaps minor variations in form or content, in many of the Projective Question categories. There are, moreover, several instances in which two or more recurring' themes, taken together, permit inferences regarding deeper-lying trends and processes. The results and theoretical constructions derived from the application of this technique are similar to those obtained from the other techniques used in the present research. In some cases there is almost exact duplication of variables; in others there is a more complemen- tary or congruent relation, the variables from several techniques expressing diverse facets of a single, inclusive structure. Because the amount of duplica- tion is considerable, a very brief discussion of the theoretical implications of the present results will suffice. Differential trends for high and low scorers on the E scale seem to exist in the following areas.
1. General Ego Functioning. Highs and lows differ markedly in their man- ner of handling deep-level trends such as aggression, sex, dependency, anxiety, and the like. We are not yet in a position to say whether one group or the other shows a greater total amount of any given trend; what is clear is that both groups exhibit all of these trends to a significant degree. The primary difference seems to lie in the ego functioning, and particularly in the relation of the ego to the deeper levels of personality. In the lows, as other techniques have shown, the underlying trends are more ego-assimilated, in the highs more ego-alien.
The lows appear to differ from highs in at least the following respects. The relations between the various levels of personality are more fluid, the boundaries more permeable. The ego defenses of the lows are relatively more impulse-releasing: at best we find considerable sublimation, to perhaps a greater degree we find that impulses have been assimilated into the ego with- out being fully integrated-witness the recurring Projective Question cate- gory, "Conscious conflict and guilt. " In the highs, on the other hand-and the analysis of interviews led to the same conclusion-the ego defenses are characteristically more countercathectic; there is less sublimation and more use of defenses such as projection, denial, and reaction-formation, defenses which aid the individual in maintaining a moral fa<;:ade at the expense of self- expression and emotional release.
? THE AUTHORITARIAN PERSONALITY
These formulations are supported by many of the Projective Question results. Compare, for example, the low category, "Conscious conflict and guilt," with the contrasting high category, "Rumbling from below" (Ques- tions I, "Moods," and 4, "Drive nuts"). The former category refers to im- pulses which, disturbing though they may be, the individual at least to some extent recognizes and tries to handle. If there is conflict, there is also an attempt to integrate; if there is much that remains unconscious, there is also a willingness to look within and an attempt to assimilate. For most highs, on the other hand, there are few focal conflicts but there is a deep sense of anxiety and distress. The conflict is covered over by a moral fac;ade or by symptomatic behavior; the disturbance is explained on the basis of a dis- tressing (overdemanding, boring) external situation or of poor bodily con- dition, and the conflict is never faced in psychological terms. The difficulty of the highs in assimilating many important needs is shown by several other categories. Thus, when asked about highly satisfying emotional experiences
("Last six months" and "Awe-inspiring") they show much less intensity and inner vitality than the lows, that is, much less ability to utilize their psychic energies for constructive and ego-satisfying purposes. When the highs refer more directly to needs such as dependency, sex, and aggression ("Desires" and "Worst crimes"), they are more crude, impersonal, primitive, object-less and ego-less.
2. Specific Properties of the Ego. These properties are, of course, inti- mately bound up with ego functioning as discussed above. As might be expected from their use of primarily countercathectic defenses, the highs have comparatively narrow, circumscribed egos. One manifestation of this narrowness is constriction of fantasy (particularly in Question 6, see also Chapter XIV); this is probably related to the highs' emphasis on "sticking to the facts," to their extraceptiveness, and to their rejection of "imagina- tion" and "emotion. "
The highs also diifer from the lows in their greater concreteness of think- ing and in their less differentiated emotional experience. Thus, the responses of the highs characteristically refer to specific behavioral acts or situations, while the lows tend, to a greater degree, to describe more subtle experiences and more abstract, generalized situations. On Question 3 ("Great people"), for example, the lows, more than the highs, give the qualities that they admire in a person, with or without specific examples. The same is true for Question 5 ("Worst crimes"). Evidence of the greater complexity and abstractness of the lows has been given in the interview chapters (Part II) and in Chapter VIII (showing negative correlations between ethnocentrism and intelli- gence). These results are consistent with those showing the highs to be more stereotyped in their ideology (Chapters III, IV, XVI, XVII), more rigidly concrete in their solutions of arithmetical and spatial probkms (Rokeach
(98)) 0
? PROJECTIVE QUESTIONS
597
The lows tend, in a frustrating situation, to blame themselves (intrapuni- tiveness), the highs to idealize themselves and to see the evil as existing in the external world (extrapunitiveness).
Finally, the greater intraception of the lows and the great anti-intracep- tion of the highs is apparent in their responses to the Projective Questions as well as in the material elicited by the interviews, the T. A. T. , and the F scale.
3? Achievement Values vs. Conventional Values. This distinction has been essential for the scoring of the Projective Questions. It is important not only because of the difference in the content of the values, but also because the values themselves express significant aspects of the personality dynamics of the two groups. It should be noted first that the two sets of values are not entirely mutually exclusive; few individuals will have only one set or the other. It is, rather, a matter of degree and of primacy. In most individuals one set of values is likely to be primary and most potent, the other to be of secondary importance. An individual who is struggling to decide between these two value orientations is, we believe, essentially struggling to decide, consciously or unconsciously, between conflicting needs and between con- flicting conceptions of himself as a total person.
Achievement values found predominantly in the lows place primary em- phasis on self-expression. Abstract and open-ended, they always leave room for further development and they can never be defined in terms of simple behavior formulae or rigid rules. Their main emphasis is on long-range goals, and the attainment of a given goal leads always to the formulation of new, higher goals expressing the same basic values. Examples of achievement values, taken from the Projective Question material as well as from the inter- views and other material, include the following: Value for scientific, intel- lectual, and aesthetic achievement, and for understanding for its own sake, regardless of immediate practical application. Creativity is valued above efficiency, constructiveness above practicality; productive living, even if it involves inner conflict, is preferred over good adjustment at the expense of
self-expression; richness and intensity of inner experience are valued more than "mere contentment. " In personal relationships, as other techniques have shown, there is concern with giving as well as taking, and with the exchange of love rather than the exchange of things. What is particularly important here is that recognition of one's own individuality is the basis for recognition of the individuality of everyone, and for the democratic concept of the dignity of man. 8 These values are expressed ideologically in terms of op- position to all social structures (milita1y, religious, educational, politico-
8 This point has also been made by Fromm (43). His distinction between "humanistic" and "authoritarian" ethics corresponds very closely to the present one between "achieve- ment" and "conventional" values, and is based on a similar attempt to distinguish two broad psychological approaches to man and society.
? THE AUTHORITARIAN PERSONALITY
economic) which are based on the principle of absolute authority, which value power more than love, which engage in group suppression and ex- ploitation, in short, which prevent man from developing his innate poten- tialities to a maximum degree. Once again we find anti-ethnocentrism as but one facet of a larger psychological framework.
What has been called "conventional values" might also have been called "conditional values," since their main function is to place limitations or con- ditions on the expression of needs rather than to stimulate need-experience. They might also have been called "authoritarian values," since they are based on the assumption of conformity to external authority rather than on inner moral responsibility. Whatever the name given, both of the above meanings, as well as others, belong to its definition. The prototypic examples of this value system are the Emily Post book of etiquette, the military "rules of behavior," and certain custom-ridden cultures, literate and nonliterate. 9 The main content of these values, at least for individuals with a strong middle- class identification, deals with conformity and loyalty to ingroup standards.
The difference between achievement values based on inner authority (in- ternalized conscience), and conventional values based on external authority (and thus replaceable when the authority changes), results also in a difference in reaction to value-violations. This is the difference between guilt and shame. It would be an exaggeration to state that the lows feel no shame, the highs no guilt. Nevertheless, there is considerable evidence from the Projective Questions, as well as from the ratings of interviews, that guilt is most charac- teristic of lows, shame of highs. The low categories for Questions 1, z, 4, and 6 are for the most part concerned with personal violations of achievement values, with practically no reference to "being caught" or to external author- ity. For the same items there are high categories referring to violations of conventional values most of which require, almost by definition, an external observing and punishing agent. Indeed, the explicit idea of inner conflict in any form is practically lacking from the high responses. Moreover, it is much more common for the lows to refer to their own personal violations of values, whereas the highs refer either to violations by others, or, more often, to
events which have, explicitly, neither moral nor motivational significance. 10 These results are in keeping with the findings reported in earlier chapters that the highs, particularly those who are more conventional and "middle- class," have a punitive but poorly internalized superego. The ego, submitting
9 See Kardiner (;g, 6o) and Benedict (15) for nonliterate societies, and Reich (g6); these are but a few of many examples.
10 W e should have to predict-and there is already considerable supporting evidence- that studies of specific cultures will show a connection between shame as a predominant emotion, threat of punishment or isolation as a means of discipline in raising children, em- phasis on ritual and custom, an authoritarian conception of God, relative lack of achieve- ment values, and ethnocentrism in group relations within the culture and with other cultures. The higher E scale means made by the more ritualized, fundamentalistic religions in our own culture (see Chapter VI) may be cited in support of this hypothesis.
? PROJECTIVE QUESTIONS
599
out of fear, must constantly forego conscious, constructive impulse gratifi- cation; instead, it finds morally acceptable ways of gaining indirect satisfac- tion (e. g. , aggression by means of ethnocentrism and moral indignation, dependency through submission to powerful authority), and it "cheats" the superego when fear of detection is minimized (e. g. , at conventions). Again we have a contradiction in levels: The highs, so moral on the surface, are essentially most concerned with underlying anxiety and with the gratification of impulses which, being ego-alien, have developed but little beyond their primitive, infantile form; whereas the lows, often so rebellious and so op- posed to traditional morality on the surface, have more fully internalized moral principles and in their emotional functioning are more troubled with moral conflict.
4? The Handling of Dependency as an Underlying Trend. General differ- ences in the ego functioning of lows. and highs have been discussed above. We may turn now to a particular disposition, namely dependency, and see what light the Projective Questions shed on its differential ego-assimilation in lows and highs. It should be noted that there are certain sex differences here, since the expression of dependency is culturally permitted, even valued, in women, whereas in men it is opposed and inhibited. For convenience the following discussion will focus on high versus low men, with the under- standing that for women some of these differences are somewhat reversed, while others hold equally well and still others are not found.
Dependency in lows is expressed mainly in the form of concern with love; many of their Projective Question responses deal with love-giving (nurtur- ance) and love-seeking (active, focal dependency). They seem highly con- cerned with emotional exchange in their personal relationships.
Dependency plays a much different role in the personality of high men. It remains for the most part an ego-alien trend which can seldom be expressed directly because it violates the image (ego ideal) of the normal, masculine man: rugged, practical, realistic, earthbound, independent, "normally sexual and acquisitive, ready to take an active part in the bitter competition de- manded by human nature, and eager to rise to the top of the ladder of suc- cess. " It is apparent that not all men who have this self-ideal are high with respect to E or personality. Nevertheless, the present results indicate that most high men have this ideal, and that most men with this ideal are high.
Part of the high man's defense against ego-alien passive dependency-it is not the love-seeking dependency of the lows-is the rigid value for work, and anxiety over dependency is expressed in the idea that overwork would ? drive one nuts (Projective Questions 1 and 4).
If dependency promotes a concern with love in the lows, it promotes a concern with power in the highs. One of the more direct forms in which high dependency is expressed is submission to power figures. Whereas the ego-assimilated dependency of the lows is expressed in their value for equali-
? 6oo THE AUTHORITARIAN PERSONALITY
tarian relationships and social structures, the ego-alien dependency of the highs leads to the acceptance of absolute authority and to a value for authori- tarian forms of social interaction. As shown in Projective Question 3 ("Great people"), what the highs admire most in others is power, strength, authority, rugged masculinity. While the aggressive-assertive needs of authoritarian individuals are the most conspicuous ones, the dependent-submissive needs are equally if not more important. In the Projective Questions, particularly 7 ("Last six months") and 8 ("Awe-inspiring"), it was the highs' deference to authority which was expressed most strongly.
The ego-alien passive dependency of the highs is expressed in another form in Projective Questions I ("Moods") and 4 ("Drive nuts"). In Question I, the most unpleasant moods are those involving a feeling of helplessness and dependency in the face of a threatening or barren environment. The dependency comes to the surface but meets an unwelcome reception in the weak, superego-ridden ego: there are no persons toward whom it can be satisfyingly expressed; there is no differentiated affect to make it an enrich- ing experience; there is only a vague anxiety which is actually based not on the external situation but on a deep inner conflict between the superego and the upsurging primitive dependency. It is as if to express the impulse would be to lose one's masculinity or even to undergo bodily harm. Similar conflicts are expressed in Question 4? Again we find the inability to look inward and the vague rumblings of ego-alien trends, including body anxiety and aggres-
sion as well as dependency.
5? The Handling of Other Trends. There is evidence that aggression, sex,
curiosity, homosexuality, and other trends are handled by highs and lows in ways similar to those discussed above. One additional point should be made regarding aggression. It appears not to have undergone much real socializa- tion in highs but has, rather, remained relatively crude, destructive, punitive, unsoftened by ego-assimilation. If the term "hostility" be used for the ag- gression found in lows, then the term "destructiveness" seems most appropri- ate for what exists in highs, especially as revealed in Questions 2 ("Desires"), 5 ("Crimes"), and 8 ("Awe-inspiring").
The results and interpretations discussed above must not be applied in a stereotyped way. It should be understood that, while most highs show most of the high variables, and similarly for the lows, there are numerous excep- tions and numerous variations on the central theme. The personality proto- type above was that of the pseudodemocratic high, and other patterns, such as the "fascist leader" or the "psychopath" will be different in many respects. It is believed, nevertheless, that the Projective Question technique has yielded results comparable, and congruent, with those of other techniques, and that it may profitably be used for the study of other personality structures.
? PART IV
INTRODUCTORY REMARKS
The present volume has so far offered findings from our research ranging from surface ideology to largely unconscious psychological traits of our subjects. The direction of research and the order of presentation were sug- gested by the nature of the ideological data themselves; they could not be derived solely from external factors, such as economic status, group mem- bership, or religion; but rather. the evidence pointed unmistakably to the role played by motivational forces in the personality. However, the study did not move mechanically from the ideological to the psychological; rather, we were constantly aware of the structural unity of the two. It thus seems permissible that we reverse the procedure now and ask: what is the mean- ing of the subjects' overt opinions and attitudes in the areas covered by the A-S, E, and PEC scales, when they are considered in the light of our psycho- logical findings, particularly those deriving from the F-scale and the clinical sections of the interviews? By answering this question we may come closer to an integration of the various aspects of a study which is centered in the problem of the relationship between ideology and personality.
As was natural, the material for this task was mainly taken from the non- clinical parts of the interviews. Not only did these data promise to yield additional evidence bearing on the major issues discussed thus far, but the wealth of detailed and elaborate statements which our subjects had formu- lated spontaneously and in their own way, offered numerous psychological leads. There is good reason to believe that the non-clinical sections of the interviews constitute through their inherent structure a link between ideology and personality. However, attention was not limited to this inter- relationship; at the same time an attempt was made to obtain a more colorful picture of the various ideologies themselves than was possible as long as we limited ourselves to the standard questionnaires.
Since the data from the questionnaire and from the Thematic Apper- ception Test and the clinical parts of the interviews had been subjected to thorough statistical treatment, quantification of the present material, though desirable, did not seem necessary. The aim, rather, was to develop for the problem areas under consideration, a phenomenology based on theoretical formulations and illustrated by quotations from the interviews. This pro- cedure, it was hoped, would yield not only more information about the specific structure of the ideologies and the manner in which personality is expressed in them but also a further differentiation of the guiding theoretical concepts themselves.
The advantages of this supplementary procedure are several. It permits 603
? THE AUTHORITARIAN PERSONALITY
us to exploit the richness and concreteness of "live" interviews to a degree otherwise hardly attainable. What is lost for want of strict discipline in in- terpretation may be gained by flexibility and closeness to the phenomena. Rare or even unique statements may be elucidated by the discussion. Such statements, often of an extreme nature, may throw considerable light on potentialities which lie within supposedly "normal" areas, just as illness helps us to understand health. At the same time, attention to the consistency of the interpretation of these statements with the over-all picture provides a safeguard against arbitrariness.
A subjective or what might be called speculative element has a place in this method, just as it does in psychoanalysis, from which many of our categories have been drawn. If, in places, the analysis seems to jump to con- clusions, the interpretations should be regarded as hypotheses for further research, and the continuous interaction of the various methods of the study should be recalled: some of the measured variables discussed in earlier chapters
were based on speculations put forward in this part.
In view of the discussions in Chapters III and IV it was not deemed nec-
essary to differentiate between A-S and E in the treatment of the interview material. While the generally close correlation of anti-Semitism and ethno- centrism could be taken for granted on the basis of previous results, more specific accounts of the nature of their interrelation, as well as of certain deviations, were incorporated into the first chapter of the present part
(Chapter XVI).
The chapter which discusses various syndromes found in high and low
scorers (XIX) is also included in this part. Although from a strictly logical point of view it may not belong here, it seemed nevertheless appro- priate to include it, since it is based almost entirely on interview material and focused on the interconnection between ideology and personality. The syndromes evolved in this chapter should be followed up by quantitative investigation. 1
1 We have not deemed it necessary to establish cross references between interviewees' statements presented here-under interview numbers-and those given in Part II under code numbers (see Chapter X, p. 342). Therefore, some quotations may appear here which have already been given there, in a different connection. However, as twelve of the San Quentin inmates are dealt with as a special group in a later chapter (XXI), a Key linking the interview numbers used here with the fictitious names assigned to them there has been inserted on the bottom of Table r (XXI).
? CHAPTER XVI
PREJUDICE IN THE INTERVIEW MATERIAL
T. W. Adorno
A. INTRODUCTION
Our study grew out of specific investigations into anti-Semitism. As our work advanced, however, the emphasis gradually shifted. We came to re- gard it as our main task not to analyze anti-Semitism or any other anti- minority prejudice as a sociopsychological phenomenon per se, but rather to examine the relation of antiminority prejudice to broader ideological and characterological patterns. Thus anti-Semitism gradually all but dis- appeared as a topic of our questionnaire and in our interview schedule it was only one among many topics which had to be covered.
Another investigation, carried through parallel to our research and partly by the same staff members of the Institute of Social Research, i. e. , the study on anti-Semitism within labor (57B), concentrated on the question of anti- Semitism, but at the same time was concerned with sociopsychological issues akin to those presented in the present volume. While the bulk of the material to be discussed in this chapter is taken from the section on prejudice of the Berkeley interviews, an attempt was made to utilize, at least in a supplemen- tary form, some of the ideas of the Labor Study as hypotheses for further investigation. This was done as a part of the work carried out in Los Angeles. In collaboration with J. F. Brown and F. Pollock we drew up an additional section of the interview schedule devoted to specific questions about Jews. These questions were derived for the most part from the material gathered through the "screened interviews" of the Labor Study. The aim of this new section of the interview schedule was to see if it was possible to estab- lish certain differential patterns within the general structure of prejudice.
The list of questions follows. Not all of these questions were put to every subject, nor was the exact wording of the questions always the same, but most of the ground marked off by the questions was covered in each case.
