The difference between percentages of women with mixed neurosis falling into the high and low halves of the E-scale distribution is statistically
significant
above the 2 per cent level of confidence (Table 4 (XXII)).
Adorno-T-Authoritarian-Personality-Harper-Bros-1950
0 14.
0
t'"'
'<:I
> (Jl
C'l . . . . (Jl
~
'0
0
. . . .
5. 3 > z
M
z . . . . ,
? THE AUTHORITARIAN PERSONALITY TABLE 4 (XXII)
PERCENTAGE OF THE UPPER AND OF THE LOWER HALVES OF THE E-SCALE DISTRIBUTION FALLING INTO VARIOUS PSYCHIATRIC CATEGffiiES
Psychoses:
Schizophrenia Manic depressive Other psychoses
Total psychoses
Psychoneuroses:
Psychoneurosis mixeda Reactive depression Anxiety statea Anxiety hysteria HYsteria Hypochondriasis Obsessive-compulsive Obsessive-compulsive
ruminative state
Total neuroses
Other disorders:
Psychopathic personality Alcoholism
Ulcers
Others
Total other disorders
N = 32
15. 7 12. 2 3. 1
31. 0
22. 0
- 18. 8 6. 3 9. 4
-
-
6. 3 62. 8
3. 1 2. 9 -
-- 2. 9 3. 1
5. 8 6. 2
Men WQmen
Low Half High Half Low Half High Half
N = 24
8. 3 4. 2
- 12. 5
29. 2
N = 24
12. 5 8. 3 -
20. 8
N = 34
14. 6 8. 9 -
23. 5
50. 0 2. 9 2. 9 2. 9 4. 2 4. 2 2. 9 - - 2. 9 5. 9
- 33. 3
12. 5 4. 2 20. 8
-
- 4. 2
- 66. 7
-
20. 9 33. 3
- 70. 4
8. 3 4. 2 - 4. 2 25. 0 12. 5 -
-
Rin only 2 cases are the differences between the high and low halves statistically significant. For the "psychoneurosis mixed type" there are significantly more low-scoring than high-scoring women (C. R. =2. 4; 2% level). There are significantly more high-scoring than low-scoring women with "anxiety state" (C. R. : 2. 1; 5% level),
total number of cases in each quartile but only on the number of psycho- neurotic cases in each quartile. Not all cases had such finer descriptions and many cases had more than one of these features. Therefore, the vertical columns in Table 5 (XXII) do not add up to wo. Table 6 (XXII) gives the same results as Table 5 (XXII), but for the upper and lower halves of theE distribution rather than for the four quartiles.
The relation between ethnocentrism and psychiatric diagnosis, as sum- marized in Tables 3-6 (XXII), may now be considered under two main headings: (1) ethnocentrism in relation to neurosis vs. psychosis, and (2) ethnocentrism in relation to specific diagnostic categories.
-
- 45. 9
'
? TABLE 5 (XXII)
PERCENI'AGE OF NEUROTIC PATIENTS IN EACH E-SCALE QUARTILE SHOWING VARIOUS NEUROTIC FEATURES
>(j r;f)
>-<
(")
:I: 0 t"" 0 C'l
cs
> t""
High N=14- t::
:I: 7. 1 M
7. 1 >t"" . . . ,
Depressive
Anxiety and phobias Obsessive-compulsive Neurasthenic
Hysterical conversion Hypochondriacal Psychopathic
Schizoid
Paranoid
Homosexual (or perversion)
:I:
>z
I:'
"d
~
14. 2 . . . , >t""
7. 1
Low N=8
50. 0 25. 0
Low Middle N=8
12. 5
High Middle N=7
28. 5 28. 5
High Low Low Middle N=4 N=14 N=10
50. 0 10. 0 50. 0 64. 3 30. 0 50. 0 10. 0
14. 2
7. 1 20. 0
~5. 0 10. 0 7. 1
50. 0 7. 1
WOMEN
High Middle
N=6
33. 3 50. 0
16. 6
33. 3 16. 6 16. 6
t""
25. 0 25. 0
>rj
> r;f)
0 r;f)
~
'0
0 ""'
12. 5
12. 5 25. 0
7. 1
7. 1
M z
? THE AUTHORIT ARIAN PERSONALITY TABLE 6 (XXII)
PERCENTAGE OF NEUROTIC PATIENTS IN THE UPPER AND LOWER HALVES OF THE E-SCALE DISTRIBUTION SHOWING VARIOUS NEUROTIC FEATURES
Men Women
Low Half High Half Low Half High Half
N = 16 Depressive 31. 3
N = 11
18. 1 36. 2 18. 1
9. 1
18. 2
N = 24
N = 20
Anxiety and phobias Obsessive-compulsive Hysterical conversion Hypochondriacal Psychopathic
12. 5
33. 3 20. 0 50. 0 20. 0 4. 2 5. 0 12. 5 5. 0 4. 2 5. 0 4. 2 5. 0 20. 0 5. 0
8. 4
4. 2 10. 0
NEUROSIS
6. 3 Schizoid 12. 5
Paranoid
Neurasthenic
Homosexual (or perversion)
1. ETHNOCENTRISM
AND PSYCHOSIS
6. 3 25. 0
IN REL A TION T O
In our total group, there was a preponderance of psychoneurotics over psychotics, the ratio being 62:24 per cent. Table 4 (XXII) shows that this ratio is somewhat greater in the low than in the high half of the E distribution. The trend appears even more markedly in Table 3 (XXII), especially in the figures for the women. There are practically no psychotics, but a relatively very large number of neurotics in the low quartile, with the proportion of psychotics increasing, that of neurotics decreasing, in the low middle and high middle quartiles. The largest number of psychotics is in the high middle quartile, but relatively few are in the high quartile. The same trends appeared in the male and female groups.
Several hypotheses can be offered to explain the drop in the proportion of psychotic subjects from the high middle to the high quartile. One hypothesis <, is that the drop is caused entirely by certain factors of sampling. It can be argued that the proportion of psychotics in the high quartile would actually be equal to or even higher than that in the high middle quartile were it not that many of the very high-scoring individuals were eliminated from the group through certain external circumstances. We know from experience at the Clinic and from work with other groups that high scorers are in general less cooperative, because they are relatively more suspicious and more afraid
of any infringement of their privacy. Thus, they often left unanswered parts
of the questionnaire dealing, for instance, with income or father's income and
? PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM 905
other "private" topics, even when their anonymity was assured. By these and similar means high scorers avoid dealing with topics which might remind them of problems and emotions they are trying to keep in repression or which might expose their weaknesses to others.
It is probable, therefore, that a great many extremely prejudiced people could never be induced to go to a psychiatric clinic for help. Perhaps this holds particularly for those ethnocentric individuals who are most disturbed, that is, for those with psychotic or near-psychotic disturbances. If it is true that there are relatively many potential high scorers among severely psy- chotic patients, the relatively low incidence of psychotics in the high quartile could be caused by the fact that the Clinic excludes violent psychotic cases and many of the cases with very poor prognosis.
Lastly, there were a few ward patients who were either too disturbed to fill out a questionnaire properly or who refused to do so. But there were not enough cases like this to explain entirely the decrease of psychotics from the high middle to the high quartile. Another line of speculation, to which we shall return later, is that prejudice in its extreme degree may be an expression of certain ego-defenses which the person has invested with a great deal of energy, because without them he would suffer a psychotic breakdown. Per- haps the high scorers are very similar to our "high middles" except for some- what greater ego-strength and better working defenses.
Another hypothesis is that the decrease in the proportion of psychotics from the high middle to the high quartile is a true one, and that it can be ex- plained by certain features of the psychotic process itself which would tend to produce middle rather than extreme scores. In support of this hypothesis some observations on psychotic clinic patients should be mentioned here.
Several of the psychotic subjects were interviewed, and it was found that only one of them had even some slight knowledge of current events and of the social issues of the present. Even this one case, a professional person, was concerned mainly with abstract ideology and never talked in terms of po- litical reality. It seemed as if these psychotic subjects-all mild cases in their first psychotic episode-were emotionally too removed from social reality to pay much attention to it or to form any strong and consistent ideology about it. Apparently this did not produce enough inconsistency of response to have much lowering effect on scale reliabilities, but enough to produce various deviant patterns and unintense responses, resulting in "middle" scores.
A related hypothesis would be that the tendency toward "middle" E scores in our psychotic group was due in largest part to certain special, tempo- rary factors arising out of the circumstance that these patients were all in acute psychotic episodes-or had just recovered from one. Here, too, we have some supporting observations. One patient, for instance, responded only in terms of +1 and -1. When released from the hospital, greatly improved
? THE AUTHORITARIAN PERSONALITY
and free from delusions, the patient said in an interview that he thought his constricted responses had been due to his extreme lack of self-confidence at the time, which prevented him from expressing himself in a more definite fashion, and that he might respond differently now.
Another subject, who at the time suffered from paranoid delusions and who showed great hostility towards his hospital environment and resentment toward the test procedure, answered only in terms of +3 and -3 (further emphasized by exclamation marks, underlinings or negativisitic comments). Such a pattern of responses would also lead to a score closer to the mean than the subject's actual attitude would warrant.
Temporary characteristics of the illness itself may be important in some cases, but it is the author's impression that they are not likely to influence the E scores of most subjects to any considerable extent. The subjects who at the time the scale was administered were very much out of contact with reality usually refused to participate or produced records which could be immediately recognized as invalid. These either had many omissions or bizarre comments, or they showed that the person was not able to follow the directions properly.
Assuming that the trends obtained with the present small sample are valid, the data show a negative relationship between psychosis and strong opposition to prejudice, a positive relationship between psychosis and mod- erate prejudice, but a relatively low incidence of psychosis among the ex- tremely high scorers.
To explain this trend we favor the following hypothesis, which fits in with many of the findings discussed in previous chapters and which is sup- ported also by the clinical findings to be discussed later: Strong opposition to prejudice, as measured by the E scale, appears to be related to certain per- sonality structures wpich, under stress, are more likely to lead to psycho- neurotic than to psychotic disturbances. This hypothesis will be discussed in more detail later in the present chapter.
2. ETHNOCENTRISM IN REL A TION T O SPECIFIC DIAGNOSTIC CATEGORIES
Our sample does not contain enough c~ses, in each of the more frequent diagnostic categories, to draw very specific conclusions. On the whole, how; ever, it appears that ethnocentrism is not correlated very highly with any given psychiatric syndrome, at least as the latter was defined at Langley Porter Clinic. There were both high and low scorers among the schizo- phrenics, manic depressives, anxiety states, hysterias, obsessive-compulsives, and, of course, "mixed neuroses. " However, certain quantitative relationships between E and psychiatric diagnosis, as well as qualitative difference~ between high and low scorers 'Within the same diagnostic category, were found,
? PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM 907
The quantitative relations inay be considered first. (1) The low scorers, especially those in the low quartile, were concentrated mostly in the "mixed neurosis category," and distinguished most often by depression and conscious anxiety, and sometimes by neurasthenic features. This was particularly true of the women.
The difference between percentages of women with mixed neurosis falling into the high and low halves of the E-scale distribution is statistically significant above the 2 per cent level of confidence (Table 4 (XXII)). (2) There were many more high-scoring than low-scoring women classified as "anxiety state. " This difference is significant above the 5 per cent level of confidence (Table 4 (XXII)). The trend is less marked in men, where many low scorers were considered "anxiety states. " As will be seen below, important qualitative differences exist between the high- and low- scoring men with anxiety state. (3) Seven of our subjects were men with stomach ulcers, taken from a research project in psychosomatic medicine. Not one of these made a low score. One had a low middle score, but turned out to be strongly prejudiced against Negroes, although not in regard to other groups. Two were "high middles," and four fell into the high quartile. This is a very marked trend, though of course not conclusive because of the small number of cases. However, this result is interesting because the modern psychoanalytic theory concerning the dynamics of ulcer has much in com- mon with the dynamic formulations about the character structure of highly prejudiced men, as advanced in this book. This theory emphasizes underlying dependency which is held in repression by counteractive defenses, a mascu- line fa<;ade, much drive for activity, and so forth.
We may now consider the qualitative differences between high and low scorers in the same psychiatric category.
a. "MIXED NEUROSis. " This seemed to be the most frequent single diag- nosis of our low-quartile women. It also occurred in one-third of the "low middle" women. There were eleven low-scoring and only three high-scor- ing women with the diagnosis "mixed neurosis. " Among the low scorers, eight reported feelings of depression and inferiority, mood swings, crying; the rest complained of tiredness and/or dysmenorrhea and difficulties in social relationships. One case had other physical symptoms-probably on an hys- terical basis-and inhibitions in group situations. Of the three high-scoring women, none reported depressed feelings, two denied all psychological diffi- culties. One was a psychopath, who also complained of menstrual difficul- ties. She was brought in by her husband for drinking, spending money ex- cessively, and going out with men from bars. She did not feel any need of help. Two were psychosomatic cases: one an extremely tense young woman who had had a thyroidectomy and denied psychological problems; the other originally came in for a chronic (psychosomatic) skin rash, but soon ad- mitted sexual (marital) maladjustment of long standing.
When the cases of stomach ulcer are not included, there are fewer men
? THE AUTHORITARIAN PERSONALITY
with the diagnosis "mixed neurosis. " But the trend is similar to that in the women, with z5 per cent and 33 per cent in the two lower groups, only 15 per cent and 9 per cent in the two higher quartiles. There were three cases of "mixed neurosis" in the low quartile, one in the high. Two of the three low scorers were reported as suffering from neurotic depression. All three seemed to be soft, openly dependent characters whose difficulties had been precipitated by rejection by a love object. This was recognized by the sub- jects and brought out in the first interview. The one high scorer in this category was depressed but also showed obsessive-compulsive and anxiety symptoms. The anxiety centered on the idea that he might harm himself and his baby. It had appeared suddenly and left him subject to recurrent attacks.
b. ANXIETY STATE
Women: Of all single categories, this one contains the largest percentage
of high scorers (zS per cent)-no low scorers, with few cases in the middle quartiles. (C. R. between percentage falling in the high and in the low halves of the E distribution is significant at the 5 per cent level). Five high scorers were classed as "anxiety state," and two very similar ones were classed as "anxiety hysteria. " Five of these seven suffered from "spells" of tension, ir- ritability, or hyperventilation symptoms often including dizziness and faint- ing. There was characteristic hypochondriacal concern, fear of death, of heart attack, and so on. Two women were afraid they would harm their children during the spells; one actually had choked her children on such occasions. In the picture of the two cases which did not have "spells," the hypochondriacal preoccupations with physical symptoms stood out and were combined with some depressed affect, in one case with schizo-affective reaction. There were no low scorers in this category.
Men: The numerical trend was less clear here, with a slightly greater pro- portion of cases in the low and low middle quartiles. There were four low scorers, three high scorers. One of the low scorers suffered from hyper- ventilation symptoms and fainting spells in certain situations of friction with
a brutally aggressive father. The other three were similar to the low-quartile cases classed as "mixed neurosis," except for clinically more marked anxiety, , with signs of much self-dissatisfaction, depression, social and sexual malad- justment, work disturbances, and some schizoid withdrawal.
The three high scorers showed physical anxiety symptoms with little conscious content. In one case this was coupled with much hypochon- driacal concern, and in another, with some paranoid trends involving anxiety dreams and fear of attack by a certain person. Two of the three attributed the onset of symptoms to accidents.
c. ScHIZOPHRENIA 1 Women: There were high- and low-scoring schizophrenics. None of the J high-quartile women fell into this group. The "high middles" that were j
? PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM 909
classed as "schizophrenia" did not fall into any of the schizophrenic "types" (e. g. , hebephrenia, paranoia). They showed paranoid, catatonic, obsessive- compulsive, and other features. The difference between the high and low scorers seems to be similar for men and for women. The high scorers appear to be very infantile, constricted, narrow personalities, often classed as "schizophrenia, simple type. " Among the "high middle" women, several sudden post-partum psychoses of withdrawal in previously schizoid or com- pulsive+ schizoid personalities, were found. The low-scoring schizophrenics were more of the hypersensitive, introspective sort, with relatively much interest in their own and others' psychological lives, and with relatively much insight into their own illness.
With regard to paranoia, the following observations have been made. Our group did not include any cases diagnosed as "true paranoia," but it included several schizophrenics (and others) with paranoid ideas. Among these were high, low, and middle scorers. However, the paranoid symptoms of the low scorers appeared to be qualitatively different from those of the high scorers, in that the low scorers more often combined ideas of being persecuted with severe inferiority feelings-"others are threatening, rejecting, or ridiculing me because of my symptoms, because I am inferior. " The "devil" is not only threatening from outside, but is largely perceived as inside the person. The high scorers, on the other hand, tended towards more highly projective types of fantasies, sometimes accompanied by bragging, self-aggrandisement, and self-righteousness. Consciously, at least, the "devil" or evil forces were seen as only outside.
d. OBSESSIVE-COMPULSIVE NEuRosis. T h e r e w e r e o n l y I m a n a n d 4
women in this group. The man fell into the high-middle quartile on E. Of ' the 4 women, 2 were low, I was high middle, and I was high. One of the two obsessive-compulsive cases appearing in the low group was just on the borderline between the low and low-middle group and had an F score slightly above the mean. The other case, an all-round low scorer, showed no typical obsessive-compulsive pattern, but had a phobic tendency and much conscious anxiety and feeling of inadequacy. The 2 high-scoring women (I high middle, I high) were both classed as "obsessive-compulsive ruminative state" because of particularly rigid preoccupations, and constant ruminative thinking of schizoid quality. Extensive data are available only on the high-scoring case, a fifteen-year-old girl with preoccupations of a sexual character. She was worried because of fantasies about intercourse and pregnancy. During her stay at the hospital most of her conscious anxiety and ruminative thinking were centered about her physiological functions, particularly constipation,
and imagined somatic changes (enlargement of abdomen).
From the above description it can be seen that a number of psychological trends differentiate the ethnocentric from the non-ethnocentric patients,
? 910 THE AUTHORITARIAN PERSONALITY
regardless of formal classification. These and other trends will be studied more directly, and interpreted in relation to our general theory, below.
E. ETHNOCENTRISM IN RELATION TO THE MINNESOTA MULTIPHASIC PERSONALITY INVENTORY
The question of possible relationships between ethnocentrism and psychi- atric diagnoses was approached in a preliminary way also by means of a test procedure known as the Minnesota Multiphasic Personality Inventory (from now on referred to as MMPI). This is an improved inventory of the Bern- renter type which has been validated against psychiatrists' diagnoses of care- fully studied cases. It contains several scales, each made up of items which differentiate statistically between patients showing a given clinical syndrome -such as conversion hysteria, paranoid conditions, or schizophrenia-and the general population. It was thought that the test scores might provide more valid and more reliable criteria than the diagnoses that had been made of our subjects by many different physicians with varying orientations, training, and experience.
Test results were available for 34 men and 48 women, that is, for 68 per cent of our total group. Because of the fact that men and women had to be treated separately, the number of subjects is far too small to lead to conclusive results.
Comparison of average scores on the various MMPI scales for the four E quartiles, and preliminary inspection of individual and group test profiles, failed to show large or consistent relationships between E and psychiatric syndromes as measured by this inventory. The results show a few trends suggesting that further research along the same lines might be well worth while.
The following are the names of the scales and brief descriptions of the principal psychological or psychiatric dimensions they are supposed to measure. The descriptions are condensations of those given in the far more detailed test manual. For a description of the test and its interpretations see Hathaway and McKinley (so).
r. Hypochondriasis-Scales I (HCh) and II (Hs). Both scales purport to measure the degree of abnormal concern about bodily functions; many of the symptoms mentioned are vague or belong among the list of common physical expressions of anxiety. Scale I is more valid be- cause it is less highly correlated with Sc and also contains an age correction.
z. Hysteria (Hy). A preliminary scale, measuring the degree of similarity between the subject (S) and patients who have developed conversion- type hysteria symptoms.
3? Depression (D). "Measures depth of clinically recognized symptom
? PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM
9I I
complex, depression. "? "A high score indicates poor morale (of the emotional type) with a feeling of uselessness and inability to assume the normal degree of optimism regarding the future. "
4? Hypomania (Ma). "Measures the personality factor characteristic of persons with marked overproductivity of thought and action. "
5? Psychasthenia (Pt). "Measures the similarity of subject to psychiatric patients who are troubled by phobias or compulsive behavior. " Mild degrees of this tendency may "be manifested merely in a mild de- pression, excessive worry, lack of confidence or inability to concen- trate. "
6. Paranoia (Pa). The preliminary scale, differentiating normals from a group of clinic patients characterized by suspiciousness, oversensitiv- ity and delusions of persecution with or without expansive egotism. Their diagnoses were usually paranoia, paranoid state or paranoid schizophrenia.
7? Schizophrenia (Sc ). Preliminary scale measuring similarity of subject's responses to those of patients who are characterized by bizarre and unusual thoughts or behavior.
8. Psychopathic Deviate-Scale I (Pd). Measures "absence of deep emo- tional response, inability to profit from experience and disregard of social mores. " (Revised) Scale II (Pd,) contains in addition a rather large group of items expressing a feeling of estrangement from the self and others, and is more highly correlated with Sc than is Scale I (Pd).
9? Mfr? Measures masculinity or femininity of interest pattern.
The scales are arranged in such a way that the means are so with Standard Deviations of 10. Deviations from soin the direction of o are usually disre- garded. Scores around 70 (i. e. , 2 S. D. above the mean), are usually consid- ered of borderline significance, scores above So as high. Elevations to 6o can be regarded as clinically significant when occurring in individual profiles in which most scores are close to so (or below).
Individuals with sufficient degrees of maladjustment to seek psychiatric help usually score high (2 S. D. above the average) on more than one of these scales. Recent clinical experience with the inventory seems to indicate that profiles or patterns of scores have more diagnostic significance than the single scores taken by themselves. In general, cases falling into the psychoneurotic group have their maximal scores on the HCh, Hs, Hy, D, and Pt scales (with secondary elevations on any of these), whereas psychotics on the whole have profiles with peaks on D, Sc, Ma, Pa, Pd, and Pdr.
Slight borderline elevations on the "psychotic" scales occur frequently in a great variety of conditions without clinical evidence of psychotic manifes- tations. At present, their significance is not quite clear. Harris and Christian- sen (48), in a study on the effect of short psychotherapy, have found that patients showing elevations on the psychotic scales and Pd, but without
? THE AUTHORITARIAN PERSONALITY
clinical evidence of psychotic tendencies, responded less well to psycho- therapy than others who did not have such scores.
Means for each of the MMPI scales were computed for each E quartile, with men and women being treated separately. These means are shown in Tables 7 and S(XXII). The number of cases in each of these subgroups was so small that no measure of variability was computed. However, profiles for individual cases were drawn for high and low quartile. They showed great variability with regard to magnitude of score as well as to type of profile. This means (I) that differences between means would have to be very large to be statistically significant, and (2) that even significant differ- ences between means for single scales would be hard to interpret, if one takes the view that only profiles, and not single scale values, have much psy- chological meaning. (a) On the whole, the low scorers made somewhat less abnormal scores. (b) This was especially true for the men on the scales Hypochondriasis I, II and Depression, and for men and women on Psychas- thenia, Paranoia, and Schizophrenia. The low-scoring women were some- what higher on Hysteria; the low-scoring men on Femininity of Interests. Because of the nature of the differences mentioned under (b), it was thought necessary to determine whether some of these trends were caused by pres- ence of psychotic cases in the group, especially since there were somewhat more psychotics among the high half (especially "high middles"). When the means for nonpsychotic subjects were computed separately, the average pat- terns and differences remained much the same. Because the number of cases was again reduced by this procedure, . and also because the patterns for high and high-middle scorers and those of low and low-middle scorers were similar in most respects, the figures for the two low quartiles and the two high quartiles, respectively, were combined (Figure I (XXII)).
The average profiles for the high scorers-especially for the men-re- semble most closely the "severe neurotic" pattern described by Harris and Christiansen in their study of the effects of brief psychotherapy; whereas the means for the low scorers resemble a more mildly neurotic pattern. The "severe neurosis" pattern, in which HCh and/or Hs, Hy, and D stand out as a pattern, with definite secondary elevation in Pd and with Pa, Sc approach- ing significance, but below the means for the first four scales, was found to be correlated with relatively poor prognosis for brief psychotherapy (of the sort administered at Langley Porter Clinic).
One difference between our high's average pattern and Harris and Christian- sen's poor prognosis pattern lies in the prominence in their group of second- ary elevated scores on Pd I, II. In our group there is little difference on these scales between our high- and low-scoring women and, for the men, the high scorers exceed the low scorers only on Pd II.
In the Harris and Christiansen study a question was raised concerning the
? WOMEN
N=34
"'
FIGURE 1 (XXII)
AVERAGE MMPI PROFILE FOR NON- PSYCHOTIC PSYCHIATRIC PATIENTS FALLING INTO EACH HALF OF THE E- SCALE DISTRIBUTION 1'0
~
5
8
C"l
~
1:"'
~ ~
ti
:I: z>
"
80
60
:;50 . . .
0
(. )
(/) 80
60
50
MEN
N=26
/
H1ghs (N=16)
gor-------------------------------------------------------~
'---Lows (N=16)
40. _------~----~--_. _ _ _ _. _ _ _~----~--------~------~--~
HCh Hs Hy D Ma Pt Pa Sc Pd Scales
~ ~. . . .
> 1:"'
? Quartile
Low(N=3) Lowmiddle (N=7) Highmiddle(N=7) High (N=9)
Half
Lowhalf(N=10) Highhalf(N=16)
HCh
Hs HY
Pa Sc Pd Pdr Mfr
58 67 59 63 72 566062 61 66 697563 70 63 616760 66 59
576261 61 68 657061 68 61
Mean (HCh, Hs,
HY, D)
63 64 72 74
63 73
Mean (Ma, Total Pa, Sc) Mean
63 64 57 61 68 70 61 67
59 62 64 68
TABLE 7 (XXII)
MEAN SCORES ON THE SEVERAL SCALES OF THE MMPI FOR SUBJECTS FALLING INTO EACH QUARTILE AND INTO EACH HALF OF THE E-SCALE DISTRIBUTION
Nonpsychotic Male Patients
Scales
D Ma Pt
54
56 60 66 73 55 60 59 71 69 90 61 76 68 74 71 84 54 73
55 62 66 71 58 63 64 73 70 86 57 74
66 65
67 65 70
? Quartile
Low(N=10) Lowmiddle(N=6) High middle (N =7) High(N=11)
Half
LOWhalf(N=16) Highhalf(N=18)
Pdr Mfr
Mean (HCh, Hs, Hy, D)
64 68 62
68
65 66
Mean (Ma, Total Pa, Sc) Mean
56 59 59 63 63 61 65 65
57 61 64 64
TABLE 8 (XXII)
MEAN SCORES ON THE SEVERAL SCALES OF THE MMPI FOR SUBJECTS FALLING INTO EACH QUARTILE AND INTO EACH HALF OF THE E-SCALE DISTRIBUTION
Nonpsychotic Female Patients
Scales
HCh Hs Hy D Ma Pt Pa Sc Pd
? 53 59 72 71 52 58 59 58 60
64 67 74 66 58 60 58 62 70
54 55 67 71 62 65 62 65 61
59 70 69 75 56 68 67 71 63 67 54
57 62 73 69 54 59 58 60 64 67 46 57 64 68 73 58 67 65 69 62 65 52
65 47 71 46 62 49
? THE AUTHORITARIAN PERSONALITY
psychological meaning of elevations on such scales as Ma, Pa, and Sc in the absence of clinically discernible psychotic trends. Therefore the items on each scale were (arbitrarily) grouped into subscales, each of which was de- signed to measure some common general attitude. None of the subscales for the so-called neurotic scales (HCh, Hs, D, Hy) differentiated between the poor and good prognostic groups. The subscales that did differentiate came from Pd and Ma, Sc, Pa. Their content revealed a common "feeling of being victimized," a "tendency towards perceiving one's problems as imposed from outside and resulting in a feeling of lack of control of senses and motorium. "
The difference between the neurotic and psychotic scales-apart from items referring to specific symptoms (e. g. , delusions)-seems to lie in the relationship of the ego to the world, and to the body.
Perhaps the finding that our high scorers are somewhat higher on the psychotic scales may be interpreted in the same way. It would certainly fit in with trends described earlier in this book, e. g. , the tendency to externalize and project unacceptable impulses, ideas, and affects. However, in order to test such an interpretation, an analysis of our data in terms of the Barris- Christiansen subscales would have to be made. This was not thought worth while mainly because of the small number of cases in our sample. Further study along these lines should prove rewarding.
One very unexpected result was that the high-scoring men obtained ex- tremely high scores on the depression scale, whereas clinically the low scorers, and especially the low-scoring women, showed the greatest tendency towards neurotic depressive symptoms. On the MMPI, the low-scoring women did not make particularly high D scores. (See also the discussion of the clinical material in Section G of the present chapter. )
These apparently contradictory results suggest that the D scale does not measure the same psychological tendencies as were observed clinically in our low scorers, who characteristically suffer from subjective depressions, feel- ings of inferiority and failure.
t'"'
'<:I
> (Jl
C'l . . . . (Jl
~
'0
0
. . . .
5. 3 > z
M
z . . . . ,
? THE AUTHORITARIAN PERSONALITY TABLE 4 (XXII)
PERCENTAGE OF THE UPPER AND OF THE LOWER HALVES OF THE E-SCALE DISTRIBUTION FALLING INTO VARIOUS PSYCHIATRIC CATEGffiiES
Psychoses:
Schizophrenia Manic depressive Other psychoses
Total psychoses
Psychoneuroses:
Psychoneurosis mixeda Reactive depression Anxiety statea Anxiety hysteria HYsteria Hypochondriasis Obsessive-compulsive Obsessive-compulsive
ruminative state
Total neuroses
Other disorders:
Psychopathic personality Alcoholism
Ulcers
Others
Total other disorders
N = 32
15. 7 12. 2 3. 1
31. 0
22. 0
- 18. 8 6. 3 9. 4
-
-
6. 3 62. 8
3. 1 2. 9 -
-- 2. 9 3. 1
5. 8 6. 2
Men WQmen
Low Half High Half Low Half High Half
N = 24
8. 3 4. 2
- 12. 5
29. 2
N = 24
12. 5 8. 3 -
20. 8
N = 34
14. 6 8. 9 -
23. 5
50. 0 2. 9 2. 9 2. 9 4. 2 4. 2 2. 9 - - 2. 9 5. 9
- 33. 3
12. 5 4. 2 20. 8
-
- 4. 2
- 66. 7
-
20. 9 33. 3
- 70. 4
8. 3 4. 2 - 4. 2 25. 0 12. 5 -
-
Rin only 2 cases are the differences between the high and low halves statistically significant. For the "psychoneurosis mixed type" there are significantly more low-scoring than high-scoring women (C. R. =2. 4; 2% level). There are significantly more high-scoring than low-scoring women with "anxiety state" (C. R. : 2. 1; 5% level),
total number of cases in each quartile but only on the number of psycho- neurotic cases in each quartile. Not all cases had such finer descriptions and many cases had more than one of these features. Therefore, the vertical columns in Table 5 (XXII) do not add up to wo. Table 6 (XXII) gives the same results as Table 5 (XXII), but for the upper and lower halves of theE distribution rather than for the four quartiles.
The relation between ethnocentrism and psychiatric diagnosis, as sum- marized in Tables 3-6 (XXII), may now be considered under two main headings: (1) ethnocentrism in relation to neurosis vs. psychosis, and (2) ethnocentrism in relation to specific diagnostic categories.
-
- 45. 9
'
? TABLE 5 (XXII)
PERCENI'AGE OF NEUROTIC PATIENTS IN EACH E-SCALE QUARTILE SHOWING VARIOUS NEUROTIC FEATURES
>(j r;f)
>-<
(")
:I: 0 t"" 0 C'l
cs
> t""
High N=14- t::
:I: 7. 1 M
7. 1 >t"" . . . ,
Depressive
Anxiety and phobias Obsessive-compulsive Neurasthenic
Hysterical conversion Hypochondriacal Psychopathic
Schizoid
Paranoid
Homosexual (or perversion)
:I:
>z
I:'
"d
~
14. 2 . . . , >t""
7. 1
Low N=8
50. 0 25. 0
Low Middle N=8
12. 5
High Middle N=7
28. 5 28. 5
High Low Low Middle N=4 N=14 N=10
50. 0 10. 0 50. 0 64. 3 30. 0 50. 0 10. 0
14. 2
7. 1 20. 0
~5. 0 10. 0 7. 1
50. 0 7. 1
WOMEN
High Middle
N=6
33. 3 50. 0
16. 6
33. 3 16. 6 16. 6
t""
25. 0 25. 0
>rj
> r;f)
0 r;f)
~
'0
0 ""'
12. 5
12. 5 25. 0
7. 1
7. 1
M z
? THE AUTHORIT ARIAN PERSONALITY TABLE 6 (XXII)
PERCENTAGE OF NEUROTIC PATIENTS IN THE UPPER AND LOWER HALVES OF THE E-SCALE DISTRIBUTION SHOWING VARIOUS NEUROTIC FEATURES
Men Women
Low Half High Half Low Half High Half
N = 16 Depressive 31. 3
N = 11
18. 1 36. 2 18. 1
9. 1
18. 2
N = 24
N = 20
Anxiety and phobias Obsessive-compulsive Hysterical conversion Hypochondriacal Psychopathic
12. 5
33. 3 20. 0 50. 0 20. 0 4. 2 5. 0 12. 5 5. 0 4. 2 5. 0 4. 2 5. 0 20. 0 5. 0
8. 4
4. 2 10. 0
NEUROSIS
6. 3 Schizoid 12. 5
Paranoid
Neurasthenic
Homosexual (or perversion)
1. ETHNOCENTRISM
AND PSYCHOSIS
6. 3 25. 0
IN REL A TION T O
In our total group, there was a preponderance of psychoneurotics over psychotics, the ratio being 62:24 per cent. Table 4 (XXII) shows that this ratio is somewhat greater in the low than in the high half of the E distribution. The trend appears even more markedly in Table 3 (XXII), especially in the figures for the women. There are practically no psychotics, but a relatively very large number of neurotics in the low quartile, with the proportion of psychotics increasing, that of neurotics decreasing, in the low middle and high middle quartiles. The largest number of psychotics is in the high middle quartile, but relatively few are in the high quartile. The same trends appeared in the male and female groups.
Several hypotheses can be offered to explain the drop in the proportion of psychotic subjects from the high middle to the high quartile. One hypothesis <, is that the drop is caused entirely by certain factors of sampling. It can be argued that the proportion of psychotics in the high quartile would actually be equal to or even higher than that in the high middle quartile were it not that many of the very high-scoring individuals were eliminated from the group through certain external circumstances. We know from experience at the Clinic and from work with other groups that high scorers are in general less cooperative, because they are relatively more suspicious and more afraid
of any infringement of their privacy. Thus, they often left unanswered parts
of the questionnaire dealing, for instance, with income or father's income and
? PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM 905
other "private" topics, even when their anonymity was assured. By these and similar means high scorers avoid dealing with topics which might remind them of problems and emotions they are trying to keep in repression or which might expose their weaknesses to others.
It is probable, therefore, that a great many extremely prejudiced people could never be induced to go to a psychiatric clinic for help. Perhaps this holds particularly for those ethnocentric individuals who are most disturbed, that is, for those with psychotic or near-psychotic disturbances. If it is true that there are relatively many potential high scorers among severely psy- chotic patients, the relatively low incidence of psychotics in the high quartile could be caused by the fact that the Clinic excludes violent psychotic cases and many of the cases with very poor prognosis.
Lastly, there were a few ward patients who were either too disturbed to fill out a questionnaire properly or who refused to do so. But there were not enough cases like this to explain entirely the decrease of psychotics from the high middle to the high quartile. Another line of speculation, to which we shall return later, is that prejudice in its extreme degree may be an expression of certain ego-defenses which the person has invested with a great deal of energy, because without them he would suffer a psychotic breakdown. Per- haps the high scorers are very similar to our "high middles" except for some- what greater ego-strength and better working defenses.
Another hypothesis is that the decrease in the proportion of psychotics from the high middle to the high quartile is a true one, and that it can be ex- plained by certain features of the psychotic process itself which would tend to produce middle rather than extreme scores. In support of this hypothesis some observations on psychotic clinic patients should be mentioned here.
Several of the psychotic subjects were interviewed, and it was found that only one of them had even some slight knowledge of current events and of the social issues of the present. Even this one case, a professional person, was concerned mainly with abstract ideology and never talked in terms of po- litical reality. It seemed as if these psychotic subjects-all mild cases in their first psychotic episode-were emotionally too removed from social reality to pay much attention to it or to form any strong and consistent ideology about it. Apparently this did not produce enough inconsistency of response to have much lowering effect on scale reliabilities, but enough to produce various deviant patterns and unintense responses, resulting in "middle" scores.
A related hypothesis would be that the tendency toward "middle" E scores in our psychotic group was due in largest part to certain special, tempo- rary factors arising out of the circumstance that these patients were all in acute psychotic episodes-or had just recovered from one. Here, too, we have some supporting observations. One patient, for instance, responded only in terms of +1 and -1. When released from the hospital, greatly improved
? THE AUTHORITARIAN PERSONALITY
and free from delusions, the patient said in an interview that he thought his constricted responses had been due to his extreme lack of self-confidence at the time, which prevented him from expressing himself in a more definite fashion, and that he might respond differently now.
Another subject, who at the time suffered from paranoid delusions and who showed great hostility towards his hospital environment and resentment toward the test procedure, answered only in terms of +3 and -3 (further emphasized by exclamation marks, underlinings or negativisitic comments). Such a pattern of responses would also lead to a score closer to the mean than the subject's actual attitude would warrant.
Temporary characteristics of the illness itself may be important in some cases, but it is the author's impression that they are not likely to influence the E scores of most subjects to any considerable extent. The subjects who at the time the scale was administered were very much out of contact with reality usually refused to participate or produced records which could be immediately recognized as invalid. These either had many omissions or bizarre comments, or they showed that the person was not able to follow the directions properly.
Assuming that the trends obtained with the present small sample are valid, the data show a negative relationship between psychosis and strong opposition to prejudice, a positive relationship between psychosis and mod- erate prejudice, but a relatively low incidence of psychosis among the ex- tremely high scorers.
To explain this trend we favor the following hypothesis, which fits in with many of the findings discussed in previous chapters and which is sup- ported also by the clinical findings to be discussed later: Strong opposition to prejudice, as measured by the E scale, appears to be related to certain per- sonality structures wpich, under stress, are more likely to lead to psycho- neurotic than to psychotic disturbances. This hypothesis will be discussed in more detail later in the present chapter.
2. ETHNOCENTRISM IN REL A TION T O SPECIFIC DIAGNOSTIC CATEGORIES
Our sample does not contain enough c~ses, in each of the more frequent diagnostic categories, to draw very specific conclusions. On the whole, how; ever, it appears that ethnocentrism is not correlated very highly with any given psychiatric syndrome, at least as the latter was defined at Langley Porter Clinic. There were both high and low scorers among the schizo- phrenics, manic depressives, anxiety states, hysterias, obsessive-compulsives, and, of course, "mixed neuroses. " However, certain quantitative relationships between E and psychiatric diagnosis, as well as qualitative difference~ between high and low scorers 'Within the same diagnostic category, were found,
? PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM 907
The quantitative relations inay be considered first. (1) The low scorers, especially those in the low quartile, were concentrated mostly in the "mixed neurosis category," and distinguished most often by depression and conscious anxiety, and sometimes by neurasthenic features. This was particularly true of the women.
The difference between percentages of women with mixed neurosis falling into the high and low halves of the E-scale distribution is statistically significant above the 2 per cent level of confidence (Table 4 (XXII)). (2) There were many more high-scoring than low-scoring women classified as "anxiety state. " This difference is significant above the 5 per cent level of confidence (Table 4 (XXII)). The trend is less marked in men, where many low scorers were considered "anxiety states. " As will be seen below, important qualitative differences exist between the high- and low- scoring men with anxiety state. (3) Seven of our subjects were men with stomach ulcers, taken from a research project in psychosomatic medicine. Not one of these made a low score. One had a low middle score, but turned out to be strongly prejudiced against Negroes, although not in regard to other groups. Two were "high middles," and four fell into the high quartile. This is a very marked trend, though of course not conclusive because of the small number of cases. However, this result is interesting because the modern psychoanalytic theory concerning the dynamics of ulcer has much in com- mon with the dynamic formulations about the character structure of highly prejudiced men, as advanced in this book. This theory emphasizes underlying dependency which is held in repression by counteractive defenses, a mascu- line fa<;ade, much drive for activity, and so forth.
We may now consider the qualitative differences between high and low scorers in the same psychiatric category.
a. "MIXED NEUROSis. " This seemed to be the most frequent single diag- nosis of our low-quartile women. It also occurred in one-third of the "low middle" women. There were eleven low-scoring and only three high-scor- ing women with the diagnosis "mixed neurosis. " Among the low scorers, eight reported feelings of depression and inferiority, mood swings, crying; the rest complained of tiredness and/or dysmenorrhea and difficulties in social relationships. One case had other physical symptoms-probably on an hys- terical basis-and inhibitions in group situations. Of the three high-scoring women, none reported depressed feelings, two denied all psychological diffi- culties. One was a psychopath, who also complained of menstrual difficul- ties. She was brought in by her husband for drinking, spending money ex- cessively, and going out with men from bars. She did not feel any need of help. Two were psychosomatic cases: one an extremely tense young woman who had had a thyroidectomy and denied psychological problems; the other originally came in for a chronic (psychosomatic) skin rash, but soon ad- mitted sexual (marital) maladjustment of long standing.
When the cases of stomach ulcer are not included, there are fewer men
? THE AUTHORITARIAN PERSONALITY
with the diagnosis "mixed neurosis. " But the trend is similar to that in the women, with z5 per cent and 33 per cent in the two lower groups, only 15 per cent and 9 per cent in the two higher quartiles. There were three cases of "mixed neurosis" in the low quartile, one in the high. Two of the three low scorers were reported as suffering from neurotic depression. All three seemed to be soft, openly dependent characters whose difficulties had been precipitated by rejection by a love object. This was recognized by the sub- jects and brought out in the first interview. The one high scorer in this category was depressed but also showed obsessive-compulsive and anxiety symptoms. The anxiety centered on the idea that he might harm himself and his baby. It had appeared suddenly and left him subject to recurrent attacks.
b. ANXIETY STATE
Women: Of all single categories, this one contains the largest percentage
of high scorers (zS per cent)-no low scorers, with few cases in the middle quartiles. (C. R. between percentage falling in the high and in the low halves of the E distribution is significant at the 5 per cent level). Five high scorers were classed as "anxiety state," and two very similar ones were classed as "anxiety hysteria. " Five of these seven suffered from "spells" of tension, ir- ritability, or hyperventilation symptoms often including dizziness and faint- ing. There was characteristic hypochondriacal concern, fear of death, of heart attack, and so on. Two women were afraid they would harm their children during the spells; one actually had choked her children on such occasions. In the picture of the two cases which did not have "spells," the hypochondriacal preoccupations with physical symptoms stood out and were combined with some depressed affect, in one case with schizo-affective reaction. There were no low scorers in this category.
Men: The numerical trend was less clear here, with a slightly greater pro- portion of cases in the low and low middle quartiles. There were four low scorers, three high scorers. One of the low scorers suffered from hyper- ventilation symptoms and fainting spells in certain situations of friction with
a brutally aggressive father. The other three were similar to the low-quartile cases classed as "mixed neurosis," except for clinically more marked anxiety, , with signs of much self-dissatisfaction, depression, social and sexual malad- justment, work disturbances, and some schizoid withdrawal.
The three high scorers showed physical anxiety symptoms with little conscious content. In one case this was coupled with much hypochon- driacal concern, and in another, with some paranoid trends involving anxiety dreams and fear of attack by a certain person. Two of the three attributed the onset of symptoms to accidents.
c. ScHIZOPHRENIA 1 Women: There were high- and low-scoring schizophrenics. None of the J high-quartile women fell into this group. The "high middles" that were j
? PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM 909
classed as "schizophrenia" did not fall into any of the schizophrenic "types" (e. g. , hebephrenia, paranoia). They showed paranoid, catatonic, obsessive- compulsive, and other features. The difference between the high and low scorers seems to be similar for men and for women. The high scorers appear to be very infantile, constricted, narrow personalities, often classed as "schizophrenia, simple type. " Among the "high middle" women, several sudden post-partum psychoses of withdrawal in previously schizoid or com- pulsive+ schizoid personalities, were found. The low-scoring schizophrenics were more of the hypersensitive, introspective sort, with relatively much interest in their own and others' psychological lives, and with relatively much insight into their own illness.
With regard to paranoia, the following observations have been made. Our group did not include any cases diagnosed as "true paranoia," but it included several schizophrenics (and others) with paranoid ideas. Among these were high, low, and middle scorers. However, the paranoid symptoms of the low scorers appeared to be qualitatively different from those of the high scorers, in that the low scorers more often combined ideas of being persecuted with severe inferiority feelings-"others are threatening, rejecting, or ridiculing me because of my symptoms, because I am inferior. " The "devil" is not only threatening from outside, but is largely perceived as inside the person. The high scorers, on the other hand, tended towards more highly projective types of fantasies, sometimes accompanied by bragging, self-aggrandisement, and self-righteousness. Consciously, at least, the "devil" or evil forces were seen as only outside.
d. OBSESSIVE-COMPULSIVE NEuRosis. T h e r e w e r e o n l y I m a n a n d 4
women in this group. The man fell into the high-middle quartile on E. Of ' the 4 women, 2 were low, I was high middle, and I was high. One of the two obsessive-compulsive cases appearing in the low group was just on the borderline between the low and low-middle group and had an F score slightly above the mean. The other case, an all-round low scorer, showed no typical obsessive-compulsive pattern, but had a phobic tendency and much conscious anxiety and feeling of inadequacy. The 2 high-scoring women (I high middle, I high) were both classed as "obsessive-compulsive ruminative state" because of particularly rigid preoccupations, and constant ruminative thinking of schizoid quality. Extensive data are available only on the high-scoring case, a fifteen-year-old girl with preoccupations of a sexual character. She was worried because of fantasies about intercourse and pregnancy. During her stay at the hospital most of her conscious anxiety and ruminative thinking were centered about her physiological functions, particularly constipation,
and imagined somatic changes (enlargement of abdomen).
From the above description it can be seen that a number of psychological trends differentiate the ethnocentric from the non-ethnocentric patients,
? 910 THE AUTHORITARIAN PERSONALITY
regardless of formal classification. These and other trends will be studied more directly, and interpreted in relation to our general theory, below.
E. ETHNOCENTRISM IN RELATION TO THE MINNESOTA MULTIPHASIC PERSONALITY INVENTORY
The question of possible relationships between ethnocentrism and psychi- atric diagnoses was approached in a preliminary way also by means of a test procedure known as the Minnesota Multiphasic Personality Inventory (from now on referred to as MMPI). This is an improved inventory of the Bern- renter type which has been validated against psychiatrists' diagnoses of care- fully studied cases. It contains several scales, each made up of items which differentiate statistically between patients showing a given clinical syndrome -such as conversion hysteria, paranoid conditions, or schizophrenia-and the general population. It was thought that the test scores might provide more valid and more reliable criteria than the diagnoses that had been made of our subjects by many different physicians with varying orientations, training, and experience.
Test results were available for 34 men and 48 women, that is, for 68 per cent of our total group. Because of the fact that men and women had to be treated separately, the number of subjects is far too small to lead to conclusive results.
Comparison of average scores on the various MMPI scales for the four E quartiles, and preliminary inspection of individual and group test profiles, failed to show large or consistent relationships between E and psychiatric syndromes as measured by this inventory. The results show a few trends suggesting that further research along the same lines might be well worth while.
The following are the names of the scales and brief descriptions of the principal psychological or psychiatric dimensions they are supposed to measure. The descriptions are condensations of those given in the far more detailed test manual. For a description of the test and its interpretations see Hathaway and McKinley (so).
r. Hypochondriasis-Scales I (HCh) and II (Hs). Both scales purport to measure the degree of abnormal concern about bodily functions; many of the symptoms mentioned are vague or belong among the list of common physical expressions of anxiety. Scale I is more valid be- cause it is less highly correlated with Sc and also contains an age correction.
z. Hysteria (Hy). A preliminary scale, measuring the degree of similarity between the subject (S) and patients who have developed conversion- type hysteria symptoms.
3? Depression (D). "Measures depth of clinically recognized symptom
? PSYCHOLOGICAL ILL HEALTH AND POTENTIAL FASCISM
9I I
complex, depression. "? "A high score indicates poor morale (of the emotional type) with a feeling of uselessness and inability to assume the normal degree of optimism regarding the future. "
4? Hypomania (Ma). "Measures the personality factor characteristic of persons with marked overproductivity of thought and action. "
5? Psychasthenia (Pt). "Measures the similarity of subject to psychiatric patients who are troubled by phobias or compulsive behavior. " Mild degrees of this tendency may "be manifested merely in a mild de- pression, excessive worry, lack of confidence or inability to concen- trate. "
6. Paranoia (Pa). The preliminary scale, differentiating normals from a group of clinic patients characterized by suspiciousness, oversensitiv- ity and delusions of persecution with or without expansive egotism. Their diagnoses were usually paranoia, paranoid state or paranoid schizophrenia.
7? Schizophrenia (Sc ). Preliminary scale measuring similarity of subject's responses to those of patients who are characterized by bizarre and unusual thoughts or behavior.
8. Psychopathic Deviate-Scale I (Pd). Measures "absence of deep emo- tional response, inability to profit from experience and disregard of social mores. " (Revised) Scale II (Pd,) contains in addition a rather large group of items expressing a feeling of estrangement from the self and others, and is more highly correlated with Sc than is Scale I (Pd).
9? Mfr? Measures masculinity or femininity of interest pattern.
The scales are arranged in such a way that the means are so with Standard Deviations of 10. Deviations from soin the direction of o are usually disre- garded. Scores around 70 (i. e. , 2 S. D. above the mean), are usually consid- ered of borderline significance, scores above So as high. Elevations to 6o can be regarded as clinically significant when occurring in individual profiles in which most scores are close to so (or below).
Individuals with sufficient degrees of maladjustment to seek psychiatric help usually score high (2 S. D. above the average) on more than one of these scales. Recent clinical experience with the inventory seems to indicate that profiles or patterns of scores have more diagnostic significance than the single scores taken by themselves. In general, cases falling into the psychoneurotic group have their maximal scores on the HCh, Hs, Hy, D, and Pt scales (with secondary elevations on any of these), whereas psychotics on the whole have profiles with peaks on D, Sc, Ma, Pa, Pd, and Pdr.
Slight borderline elevations on the "psychotic" scales occur frequently in a great variety of conditions without clinical evidence of psychotic manifes- tations. At present, their significance is not quite clear. Harris and Christian- sen (48), in a study on the effect of short psychotherapy, have found that patients showing elevations on the psychotic scales and Pd, but without
? THE AUTHORITARIAN PERSONALITY
clinical evidence of psychotic tendencies, responded less well to psycho- therapy than others who did not have such scores.
Means for each of the MMPI scales were computed for each E quartile, with men and women being treated separately. These means are shown in Tables 7 and S(XXII). The number of cases in each of these subgroups was so small that no measure of variability was computed. However, profiles for individual cases were drawn for high and low quartile. They showed great variability with regard to magnitude of score as well as to type of profile. This means (I) that differences between means would have to be very large to be statistically significant, and (2) that even significant differ- ences between means for single scales would be hard to interpret, if one takes the view that only profiles, and not single scale values, have much psy- chological meaning. (a) On the whole, the low scorers made somewhat less abnormal scores. (b) This was especially true for the men on the scales Hypochondriasis I, II and Depression, and for men and women on Psychas- thenia, Paranoia, and Schizophrenia. The low-scoring women were some- what higher on Hysteria; the low-scoring men on Femininity of Interests. Because of the nature of the differences mentioned under (b), it was thought necessary to determine whether some of these trends were caused by pres- ence of psychotic cases in the group, especially since there were somewhat more psychotics among the high half (especially "high middles"). When the means for nonpsychotic subjects were computed separately, the average pat- terns and differences remained much the same. Because the number of cases was again reduced by this procedure, . and also because the patterns for high and high-middle scorers and those of low and low-middle scorers were similar in most respects, the figures for the two low quartiles and the two high quartiles, respectively, were combined (Figure I (XXII)).
The average profiles for the high scorers-especially for the men-re- semble most closely the "severe neurotic" pattern described by Harris and Christiansen in their study of the effects of brief psychotherapy; whereas the means for the low scorers resemble a more mildly neurotic pattern. The "severe neurosis" pattern, in which HCh and/or Hs, Hy, and D stand out as a pattern, with definite secondary elevation in Pd and with Pa, Sc approach- ing significance, but below the means for the first four scales, was found to be correlated with relatively poor prognosis for brief psychotherapy (of the sort administered at Langley Porter Clinic).
One difference between our high's average pattern and Harris and Christian- sen's poor prognosis pattern lies in the prominence in their group of second- ary elevated scores on Pd I, II. In our group there is little difference on these scales between our high- and low-scoring women and, for the men, the high scorers exceed the low scorers only on Pd II.
In the Harris and Christiansen study a question was raised concerning the
? WOMEN
N=34
"'
FIGURE 1 (XXII)
AVERAGE MMPI PROFILE FOR NON- PSYCHOTIC PSYCHIATRIC PATIENTS FALLING INTO EACH HALF OF THE E- SCALE DISTRIBUTION 1'0
~
5
8
C"l
~
1:"'
~ ~
ti
:I: z>
"
80
60
:;50 . . .
0
(. )
(/) 80
60
50
MEN
N=26
/
H1ghs (N=16)
gor-------------------------------------------------------~
'---Lows (N=16)
40. _------~----~--_. _ _ _ _. _ _ _~----~--------~------~--~
HCh Hs Hy D Ma Pt Pa Sc Pd Scales
~ ~. . . .
> 1:"'
? Quartile
Low(N=3) Lowmiddle (N=7) Highmiddle(N=7) High (N=9)
Half
Lowhalf(N=10) Highhalf(N=16)
HCh
Hs HY
Pa Sc Pd Pdr Mfr
58 67 59 63 72 566062 61 66 697563 70 63 616760 66 59
576261 61 68 657061 68 61
Mean (HCh, Hs,
HY, D)
63 64 72 74
63 73
Mean (Ma, Total Pa, Sc) Mean
63 64 57 61 68 70 61 67
59 62 64 68
TABLE 7 (XXII)
MEAN SCORES ON THE SEVERAL SCALES OF THE MMPI FOR SUBJECTS FALLING INTO EACH QUARTILE AND INTO EACH HALF OF THE E-SCALE DISTRIBUTION
Nonpsychotic Male Patients
Scales
D Ma Pt
54
56 60 66 73 55 60 59 71 69 90 61 76 68 74 71 84 54 73
55 62 66 71 58 63 64 73 70 86 57 74
66 65
67 65 70
? Quartile
Low(N=10) Lowmiddle(N=6) High middle (N =7) High(N=11)
Half
LOWhalf(N=16) Highhalf(N=18)
Pdr Mfr
Mean (HCh, Hs, Hy, D)
64 68 62
68
65 66
Mean (Ma, Total Pa, Sc) Mean
56 59 59 63 63 61 65 65
57 61 64 64
TABLE 8 (XXII)
MEAN SCORES ON THE SEVERAL SCALES OF THE MMPI FOR SUBJECTS FALLING INTO EACH QUARTILE AND INTO EACH HALF OF THE E-SCALE DISTRIBUTION
Nonpsychotic Female Patients
Scales
HCh Hs Hy D Ma Pt Pa Sc Pd
? 53 59 72 71 52 58 59 58 60
64 67 74 66 58 60 58 62 70
54 55 67 71 62 65 62 65 61
59 70 69 75 56 68 67 71 63 67 54
57 62 73 69 54 59 58 60 64 67 46 57 64 68 73 58 67 65 69 62 65 52
65 47 71 46 62 49
? THE AUTHORITARIAN PERSONALITY
psychological meaning of elevations on such scales as Ma, Pa, and Sc in the absence of clinically discernible psychotic trends. Therefore the items on each scale were (arbitrarily) grouped into subscales, each of which was de- signed to measure some common general attitude. None of the subscales for the so-called neurotic scales (HCh, Hs, D, Hy) differentiated between the poor and good prognostic groups. The subscales that did differentiate came from Pd and Ma, Sc, Pa. Their content revealed a common "feeling of being victimized," a "tendency towards perceiving one's problems as imposed from outside and resulting in a feeling of lack of control of senses and motorium. "
The difference between the neurotic and psychotic scales-apart from items referring to specific symptoms (e. g. , delusions)-seems to lie in the relationship of the ego to the world, and to the body.
Perhaps the finding that our high scorers are somewhat higher on the psychotic scales may be interpreted in the same way. It would certainly fit in with trends described earlier in this book, e. g. , the tendency to externalize and project unacceptable impulses, ideas, and affects. However, in order to test such an interpretation, an analysis of our data in terms of the Barris- Christiansen subscales would have to be made. This was not thought worth while mainly because of the small number of cases in our sample. Further study along these lines should prove rewarding.
One very unexpected result was that the high-scoring men obtained ex- tremely high scores on the depression scale, whereas clinically the low scorers, and especially the low-scoring women, showed the greatest tendency towards neurotic depressive symptoms. On the MMPI, the low-scoring women did not make particularly high D scores. (See also the discussion of the clinical material in Section G of the present chapter. )
These apparently contradictory results suggest that the D scale does not measure the same psychological tendencies as were observed clinically in our low scorers, who characteristically suffer from subjective depressions, feel- ings of inferiority and failure.
