Defensive projection and paranoid delusions

Defensive projection and paranoid delusions

J. psychial. Res., Vol. 20, No. 3, pp. 161-173, Printed in Great Eritian. DEFENSIVE 1986 OOZ-3956/86 $3.00+ .OO Pergamon Journals Ltd. PROJECTION ...

1MB Sizes 0 Downloads 76 Views

J. psychial. Res., Vol. 20, No. 3, pp. 161-173, Printed in Great Eritian.

DEFENSIVE

1986

OOZ-3956/86 $3.00+ .OO Pergamon Journals Ltd.

PROJECTION

AND PARANOID

DELUSIONS

ALFRED B. HEILBRUN, JR, ROBYNE S. DILLER and VALERIE S. DODSON Psychology

Department,

Emory

University,

Atlanta,

GA 30322, U.S.A.

(Received 16 October 1984; revised 2 July 1985) Summary-The present study considered the implication of a new explanation of defensive projection for the thematic qualities of schizophrenic delusions. Projection was proposed to be a form of selfdeception practiced by people who make excessive use of social comparison in self-evaluation. Defensive projection occurs when an unacceptable quality is minimized by distorting this quality in other people used as standards of comparison. Reactive paranoid schizophrenics were expected to show projection as a distortion of social comparison and process paranoid schizophrenics were not. This was tested in terms of a number of thematic qualities that should appear in actual delusions of reactive paranoids but not in those of process paranoids (i.e. more specific social references, attribution of untrustworthiness by men, attribution of hostility by women, concern with female identity in women). The results confirmed these predictions and supported the social-comparison interpretation of projection as a defense in reactive schizophrenics that contributes to the development of delusions.

and GLICK (1984), in reviewing the century-old effort to establish useful subclassifications in schizophrenia, considered three different approaches that have figured prominently. Clinical-observational and statistical ordering of symptoms represent two of these, whereas subdividing schizophrenics on the basis of information relating to life history or course of illness represents the third. The reactive-process distinction, that separates schizophrenics in terms of the quality of their premorbid adjustment, has been one of the major nonsymptomatic classification systems. Reviews of the schizophrenia research literature in the 1960s and early 1970s (HIGGINS, 1969; NEALE and CROMWELL, 1970; PAYNE, 1973) suggested the promise of the process (poor premorbid adjustment) and reactive (good premorbid adjustment) classification in contributing toward an eventual understanding of schizophrenia. However, the process-reactive distinction has dwindled in importance and usage in more recent years. The clearest reason for this change is the tendency to preclude reactive schizophrenia from the diagnostic systems used by researchers and clinicians. For example, the disgnostic criteria for schizophrenia proposed for psychiatric research by FEIGHNER et al. (1972) stress the importance of the patient’s unmarried status, poor premorbid social history, and a family history of schizophrenia’. All of these tend to identify the process schizophrenic and exclude the previous category of reactive schizophrenia. Despite the diminished emphasis upon process and reactive status in schizophrenia research, we continue to believe that maintaining this distinction can promote the understanding of schizophrenia. The basis for this belief comes from a number of studies within our own program of research that have demonstrated differing patterns of cognitiion

ZIGLER

161

162

ALBERTB. HEILBRUN, JR et a/.

in process and reactive schizophrenics suffering from the same psychotic symptoms. The process and reactive patterns, observed in studies of delusions (HEILBRUN and HEILBRUN, 1977) and auditory hallucinations (HEILBRUN, 1980; HEILBRUN and BLUM, 1984; HEILBRUN et al., 1983; 1985), offered different explanations of cognitive vulnerability to these serious psychotic symptoms in the two types of schizophrenics. One of our studies in particular provided the impetus for the present investigation. HEILBRUN et al. (1985) considered the role of defensive projection in paranoid conditions observed in shizophrenics. They found that only reactive paranoid schizophrenics displayed evidence of projective defenses, whereas process paranoid schizophrenics and psychiatric controls did not. The process paranoid patients, on the other hand, responded in a singularly idiosyncratic fashion on another task relative to reactive paranoids and controls. These results suggest that the dynamics of delusional ideation may be different depending upon the premorbid status of the schizophrenic. Only the reactive schizophrenic would be expected to show the denial of threatening personal qualities and their attribution to others, believed to play a central role in the development of paranoid delusions in Freud’s original formulation (N~DERLAND, 1974) as well as in neopsychoanalytic explanations of paranoid problems (e.g. CAMERON 1963). The delusions of process schizophrenics could be explained in terms of their autistic ideation and peculiar views of events. There is considerable evidence that schizophrenics, broken down into process and reactive status, are characterized by different styles and quality of thought (DEWOLFE, 1974; ZIGLER and LEVINE, 1981). Reactives generally have been found capable of more articulated and organized thought, whereas process schizophrenics tend toward more diffuse, disorganized, and self-preoccupied (autistic) thinking. Traditional clinical descriptions of paranoid schizophrenia usually emphasize the reactive type of cognition, and, indeed, there are many concordance studies that have found paranoid symptoms more evident among reactive than process schizophrenics (GOLDSTEINet al., 1968; MCCREARY, 1974; NEALE et al., 1972; ZIGLER et al., 1976). However, there is contradictory concordance evidence that either suggests no difference in rates (ZIGLER and LEVINE, 1973) or just the opposite linkage between process status and paranoid symptoms (EISENTHAL et al., 1972; JOHANNSEN et al., 1972). The inconsistency in concordance studies might be explained in part by the different dynamics of delusions proposed for process and reactive schizophrenics. Both types of schizophrenics could show delusional symptoms (and be diagnosed as paranoid), but the delusions may differ in keeping with their distinct thought capabilities and styles. If so, the concordance figures between process or reactive schizophrenia and paranoid symptoms might be expected to vary with patient population differences and diagnostic emphasis. Besides the possibility that projection is instrumental in the genesis of only some delusions, the role of projections in paranoid disorders remains at issue because this defense is not well understood as a complex cognitive activity. If more were known about how this process of denial and attribution is accomplished, it might be easier to specify why projection should be related to the development of delusions and why this should be true for only some paranoids. The HEILBRUN et a/. (1985) study, in which reactive paranoids were found singularly committed to the use of projection, also proposed and tested an explanation of the mechanics of defensive projection. The projective defense was viewed as evolving from an overcommitment to social comparison as the basis for self-conceptual judgements.

PROJECTION ANDPARANOID DELUSIONS

163

Social-comparison theory, proposed earlier by FESTINGER(1954), suggests the selfevaluations are often based upon comparisons between our own qualities and those of others about us. Defensive projection was hypothesized to be a distortion of the social-comparison process that allows the person to avoid the threat of recognizing an unacceptable personal quality. The shift of social-comparison standards toward the unacceptable quality by selective perception or outright misattribution would allow projectors to minimise unacceptable qualities in themselves or even avoid their recognition. Several implications of this proposal may be noted, using hostility as an example of an unacceptable quality. First, selective perception of hostility in or misattribution of hostility to people that serve as comparison standards precede denial and make it possible, a reversal of the order within the original Freudian version. In other words, denial is not assumed to be a discrete cognitive act, but the outcome of faulty observations regarding one’s social environment. Second, the social-comparison hypothesis no longer makes it necessary to argue whether the trait being projected is unconscious, an issue in earlier discussions of this particular defense (CHALUS,1976; HOLMES, 1968). The projector may be aware of his/her own hostility, but minimize it by escalating the perceived level of hostility in others. Finally, this hypothesis continues to place a premium upon the unconscious nature of defensive operations. The projector may be aware of the hostility, but not of the distortion introduced into the revised set of social-comparison standards. To be aware would compromise selfdeception and render the defense useless. Some evidence supporting the social-comparison explanation of defense projection has been reported. HEILBRUNet al. (1985) examined several ways in which normals could reach self-evaluative judgments, including social comparison, It was found that subjects who scored higher on a measure of defensive projection also made greater use of social comparison in reaching self-evaluative judgments. This was especially true when evaluating themselves on a negative trait. Another investigation (HEILBRUNand PEPE, 1985) of the role of awareness in defensive projection confirmed the importance of the unconscious nature of the defense as a requisite for effectiveness. Projection was associated with better stress management in normals only when the subjects were unaware of their use of this defense. Finally, HEILBRIJN and CASSIDY(1985) confirmed that defensive projection occurred in normal subjects who were less skilled in interpreting cues in their social environment. Poorer social cognition would facilitate the perceptual distortions required for revising standards of comparison. What, then, is the theoretical tie between the use of projection by reactive paranoid schizophrenics and the hypothesis that projection is a defensive strategy promoted by an overcommitment to social comparison in self-evaluation? There is considerable empirical evidence that reactive schizophrenics, along with paranoids, are excessively attentive to external social events, whereas attention in process and nonparanoid schizophrenics is less displayed outwardly and more oriented inwardly (DEWOLFE, 1974; SILVERMAN,1964; VENABLES1964). It would follow that reactive paranoid schizophrenics would be especially likely to use social comparison in self-evaluation and, by our analysis, to defend themselves by distortions within this process. Furthermore, these perceptual distortions by definition denigrate the projector’s social environment and leave him/her in a position where a new understanding of the relationship between the projector and that altered environment is

164

ALBERTB. HEILBRUN, JR et al.

necessary. This understanding could play a critical role in the formulation of a delusional system. The present study sought empirical support for this social comparison model of projection in the delusional behavior of reactive schizophrenics. A series of predictions regarding the quality and substance of delusional content were considered. It is hoped that supportive evidence would not only contribute to the understanding of how some paranoid delusions evolve but also lend credibility to our argument that the process-reactive distinction in schizophrenia has conceptual value. The first prediction regarding the thematic qualities of paranoid delusions follows directly from the social-comparison hypothesis that proposes projection to be a defensive revision of social standards that allows the person to deny or minimize an unacceptable personal quality. Given that these perceptual distortions become involved in delusions, the following prediction would be made:

If reactive paranoid schizophrenics, who have a history of concern with their social environment, use actual people as comparison standards in evaluating themselves, the distortion of these standards in defensive projection should implicate more specific people within delusional content. Process paranoids, having a history of restricted investment in their social environment and less inclination to judge themselves bv means of social comparison, should make more vague references to those people implicated in their delusions. Additional predictions are to be made, but these require consideration of a study reported by HEILBRUN (1982). This study involved the development of a psychometric scale measuring defensive projection using the trait terms on the self-descriptive Adjective Check List (GOUGH and HEILBRUN, 1980). The adjectives that differentiated the self-descriptions of normal subjects scoring high and low on a laboratory projection task were identified statistically and combined into male and female projection scales. The individuallyadministered laboratory task has received substantial research usage (HEILBRUN, 1972; 1977; 1978) as well as evidence of its construct validity. The latter would be represented by the isolation of projective defenses within a reactive paranoid schizophrenic group (HEILBRUN et al., 1985), the suggestion of effective stress management only in projectors who are unaware of this defense (HEILBRUN and PEPE, 1985), and the explanation for projective attribution by demonstrating deficiencies in social cognition among projectors (HEILBRUN and CASSIDY, 1985). The projection scales were offered as a feasible means of obtaining projection scores when only group-testing was possible. No consideration was given in the scale-development paper to the meanings of the traits chosen for self-description by projectors or nonprojectors. Whether the self-descriptions of males and females who use defensive projection, especially their patterns of denial, would suggest especially sensitive areas compensated by this defense remained unexplored. Table 1 presents the traits that distinguished normals scoring high on projection from those scoring low, separately by sex. The favorability of each trait, using values reported by W~LLUS and BEST (1977), is also included. Two observations regarding these differentiating traits appear warranted. First, the lists are almost totally different for the two sexes. Only three traits out of the 64 appear on

PROJECTIONSAND PARANOID DELUSIONS

TABLE

1. BEHAVIORAL TRAITSDISTINGUISHINGNORMAL MALESAND

165

FEMALESVARYING IN THEIR USE OF DEFENSIVE

PROJECTION Females

Males Trait

Favorability*

Projectors more:

greedy Projectors less: active adaptable charming cheerful clever deceitful determined dignified disorderly dissatisfied enthusiastic good-natured hasty inventive jolly kind loyal mannerly meek noisy obliging opportunistic optimistic pleasant pleasure-seeking praising preoccupied quick reasonable reflective responsible self-confident self-seeking selfish sexy shrewd sly sophisticated spendthrift trusting undependable

343

641 634 610 641 623 347 603 580 396 413 631 654 408 594 613 645 631 591 449 399 539 535 621 619 559 580 427 568 614 567 641 595 464 353 594 498 449 549 444 620 362

Trait

Favorability*

Projectors more: foresighted hard-headed silent timid worrying Projectors less: adaptable affectionate attractive clear-thinking feminine flirtatious hostile inventive mischievous obliging poised resourceful self-controlled sincere snobbish sociable tactful

*Favorability scores normed with a mean of 500 and standard favorable and below 500 are considered unfavorable.

deviation

598 412 464 444 393

634 611 620 636

530 437 356 594 464

539 594 611 607 630

352 599 606

of 100. Scores above 500 are considered

166

ALBERT B. HEILBRUN, JR et al.

both lists. Second, both lists show projectors to be self-critical. This shows up mainly in the failure to endorse positive traits (above 500) and to some degree in the greater endorsement of unfavorable traits (below 500). The probabthty that projectors would show this pattern of unfavorable endorsement relative to nonprojectors by chance is exceedingly small (male x2 (l)= 12.19, p < 0.001; female x2 (1)=27.45, p < 0.001). Current interest, given these observations, was whether there still remained pockets of defensive self-deception within the otherwise unfavorable self-descriptions of projectors. If so, these could suggest special areas of sensitivity that might systematically trigger projection in males and females. Such lucunae in projectors might then appear as genderrelated thematic qualities in reactive paranoid schizophrenics who use projection as a defense. Theoretically, if projection as a distortion of social-comparison standards has served its defensive function by creating these pockets of self-esteem in otherwise self-critical people, the defense would have achieved this end by the attribution of the opposite (unacceptable) quality to others. These unacceptable qualities, by this rationale, might then be expected to make their thematic appearance in the delusions of reactive paranoids who use projection. The three most unfavorable traits on the male list, all denied as personal characteristics by projectors, include deceitful (374), selfish (353), and undependable (362). This suggests the special unacceptability of untrustworthiness. In the case of females, hostile (356) and snobbish (352) were exceptionally unfavorable traits emphatically denied by otherwise nondefensive subjects. Being hostile may be viewed as especially unacceptable to the female projector. The second prediction, then, is that:

The thematic content of paranoid delusions in reactive schizophenrics will vary in emphasis by sex. Men will place more emphasis than women upon their inability to trust others, whereas women will emphasize the hostility of others more than men. These projected qualities would allow for denial of untrustworthiness in males and denial of hostility in females. Process paranoid schizophrenics, not given to the use of projection, should not show the same gender-based patterns. Reexamination of Table 1 offers encouragement to this speculative analysis of trust and hostility as holding special significance for male and female projectors, respectively. If male projectors have problems in trusting others, as would be expected if untrustworthiness were a projected attribute, the absence of both loyal and trusting from their self-descriptions could be explained. The deficit in social and affectional capabilities for female projectors (i.e. not affectionate, attractive, sociable, or tactful) would be consistent with viewing others as more hostile than they are. It is difficult to relate positively to a presumably hostile social environment. These implications are especially serious for the female. According to most contemporary sex-role theorists, expressiveness, the ability to foster and maintain warm and mutuallyrewarding social relationships, lies at the heart of her identity as a woman. (Note the failure to endorse feminine on Table 1 for the female projector.) If her own hostility represents a critical trigger for projection in a woman and attributed hostility interferes with her sense of social/affectional competence, she may encounter serious concerns about her female identity. The third set of predictions to be tested proposes: Female reactive paranoid schizophrenics will emphasize gender-defining material in their

PROJECTIONS AND PARANOID DELUSIONS

167

delusions and female process paranoids will not. Specifically, reactive paranoid females should (1) place greater importance upon sexual themes in which they initiate or are the object of sexual interest and (2) specify more objects in their delusions that clearly stamp them as women (e.g. female sex organs, apparel, jewelry, etc.). In summary, this study involved the analysis of delusional content provided by reactive and process paranoid schizophrenics. The purpose of the study was to test a series of predictions regarding reactive-process differences in delusional content that follow from: (1) The interpretation of defensive projection as a distorted social-comparison process and (2) the assumption that defensive projection is a dynamic in the development of reactive paranoid delusions but not in the forming of process paranoid delusions. METHOD

Subjects Delusional materials collected from 32 schizophrenics (15 females and 17 males) were analyzed in this study. All were volunteer patients at the Eastern State Hospital, Williamsburg, Virginia, and presented primary or secondary paranoid symptoms. Diagnoses were reached by hospital psychiatrists in keeping with DSM-II (AMERICAN PSYCHIATRIC ASSOCIATION, 1968) which was current at the time the materials were collected. The schizophrenics were separated into process and reactive types by using median scores on the Ullmann-Giovannoni self-report scale (ULLMANNand GIOVANNONI, 1964), a validated measure of process or reactive premorbid status (JOHNSON and RIES, 1967; MCCREARY, 1974; MEICHENBAUM, 1969). Within-sex cutting scores were required (between 11 and 12 for males and 14 and 15 for females) in order to create roughly equal numbers of process and reactive subjects for each sex. Comparison of the reactive and process groups (combined over sex) found reactives on the average to be somewhat younger (36.8 vs 41.8 yr), better educated (13.6 vs 11.2yr), and less chronic (4.6 vs 8.9 yr of hospitalization) than the process schizophrenics. The two groups were receiving almost identical daily dosages of phenothiazines (reactives = 321 mg, process = 319mg) at the time of their participation in the study.

Con tent analysis procedures Raw delusional materials. Delusional

materials were obtained from tape-recorded interviews conducted by a senior clinical psychologist.’ The goal of the unstructured interviews was to elicit delusional content with the means for doing so left to the judgment and skill of the interviewer. Verbatim transcripts were edited by the same senior psychologist so that only content he judged to be relevant to the delusions remained. These edited, verbatim typescripts were used for a prior study of delusional form and content (HEILBRUN and HEILBRUN, 1977), although no consideration was given in that study to the present content variables. Rating scales. The content of each patient’s delusion was analyzed according to four categories of information. First, ratings were obtained as to the type of delusion in terms of the traditional distinctions between persecution, grandiosity, control and jealousy. Fourpoint scales ranging from “definitely not implied” (score = 1) to “definitely implied” (score = 4) were used. The presence of more than one type within single delusions was the

168

ALBERT B. HEILBRUN, JR et al.

rule rather than the exception. However, as commonly observed, the mean ratings revealed that persecution themes are most prevalent: (1) persecution = 3.73; (2) grandiosity = 2.19; (3) control = 2.16; (4) jealousy = 1.97. Having identified the type(s) of delusion within each protocol, judges were asked to identify the source or object implicated in each type of thema: Persecuted by whom or what? Grandiose compared to whom or what? Controlled by whom or what? Jealous of whom or what? They then rated the level of specificity for each source or object on the following scale: (1) person(s) specified and known to patient (e.g. wife); (2) person(s) specified but unknown to patient (e.g. the FBI); (3) person(s) not specified but presumable; (4) impersonal (no human source specified). A score of 4 was assigned to the most specific rating down to 1 for the impersonal rating. Specificity ratings were summed within each protocol when more that one type of delusion was represented. The second and third categories of information that were analyzed within the delusional content related to hostility from or distrust in the social environment. Two types of hostility ratings were obtained. Judges first considered whether the patient made reference to hostile action or intent toward him/her from the environment and then to what extent the hostile action or intent was a critical feature of the delusion? Each was rated on a four-point scale from “definitely not” = 1 to “definitely” = 4. Distrust of others also was rated along the same two scales of presence and importance. Judges were instructed to rate distrust as present only if explicitly mentioned and not to infer it from the patient’s attributions of hostility to others. This was done to make the rating of hostility and distrust as independent as possible. (The final hostility and distrust scores correlated only 0.16.) The final category of information analyzed was the relevance of the content to the patient’s identity as a woman. The involvement of sexuality in the delusions of reactive paranoid women was was tested by three ratings: (1) Whether the delusion made reference to the patient as the initiator of sexual behaviour, (2) delusional reference to the patient as the object of any type of sexual behaviour, and (3) how critical the sexual behaviour was to the delusion. All ratings were on the same four-point “definitely not” to “definitely” probability scales as described in the previous section. A second type of judgment relating to gender identity was obtained for female paranoids only. This involved the identification of all objects or events mentioned within the delusion which were indisputably female. Judges. One undergraduate senior and one graduate student in psychology served as judges. Each judge worked independently following instruction by the senior investigator on the meaning of each variable and the rating procedures for scoring them. These judges worked without knowledge of the hypotheses under test or the process-reactive status of the patients. All 32 protocols were rated by each judge. The product-moment correlations between their ratings will be reported in the Results section. Where necessary, rated variables were discarded from the study when interjudge reliability was too low. In all cases of disagreement, the finial score represented the average ratings of both judges.

PROJECTIONS AND PARANOID DELUSIONS

169

RESULTS

Data analysis The distributions of scores analyzed in this study were either demonstrated heterogeneous variances. Accordingly, nonparametric (SIEGEL, 1956) were employed in all comparisons.

highly Fisher

skewed or exact tests

Specificity of people in the delusions The first prediction proposed that reactive paranoid schizophrenics would involve more specific people in their delusions than process paranoid schizophrenics. The ratings of delusional type by the two judges correlated 0.72 @ < O.Ol), and their ratings for specificity of source or object showed a correlation of 0.60 @ < 0.01). The specificity scores for the entire sample were split at their median (7.50), and the proportions of reactive paranoids and process paranoids falling below or at/above this cutting point were determined. Eleven reactive paranoids had high specificity scores and 5 had low; in contrast, 5 process paranoids had high scores and 11 had low. The exact probability of finding proportional differences this extreme (or more extreme) given by the Fisher test was 0.039. Reactive paranoids were significantly more specific in identifying people within their delusions than process paranoids, as predicted.

Hostility and distrust themas The second set of predictions specified that women would place thematic emphasis upon hostility from the social environment, whereas men would emphasize social distrust in their delusions. Furthermore, it was predicted that these gender-related themes would be found only among reactive schizophrenics who are prone to use defensive projection and not among process schizophrenics. Interjudge agreement regarding the presence of hostility from others in the delusions (r= 0.77, p < 0.01) or the presence of distrust in others as a theme (r= 0.50, p < 0.01) was acceptable. The judges were unable to agree satisfactorily regarding how critical hostility (r = 0.29, ns) or distrust (r = 0.12, ns) was to the delusion, so these ratings were discarded from this analysis. A thematic pattern score was adopted to emphasize the gender relevance of the themes by subtracting the distrust score from the hostility score for each subject. Predicted emphasis upon hostility for women would result in plus pattern scores, and emphasis upon distrust for men would result in minus pattern scores. The rated hostility scores for the entire sample were transformed to a standard score distribution with a mean of 50 and a standard deviation of 10. The same procedure was followed with the distrust scores. These conversions were necessary to bring the two distributions into line, because the distrust thema scores were lower overall than the hostility scores. This was due most likely to the requirement that raters ignore attributions of hostility as evidence of untrustworthiness. The median hostility-minus-distrust pattern score was - 6.13. Seven of the eight reactive paranoid women exceeded this score, whereas 6 of the 8 reactive paranoid men fell at or below this figure. The Fisher exact probability of this proportional difference (or one more extreme) is 0.040. As predicted, comparison of the sexes found the reactive women more committed to hostility themes than to themes of distrust with the opposite pattern in reactive

170

ALBERT B. HEILBRUN, JR et al.

men. A parallel analysis of process paranoid schizophrenics revealed no difference between men and women. Fifty-six per cent of the process men had scores above the median ( - 5 .OO) compared with 43% of the women (p=O.501).

Gender-identity

concerns in female delusions

The third set of predictions concerned the emphasis upon gender identity of women in the delusions of female reactive schizophrenics. One prediction was that reactive females would place greater importance upon sexual involvement in their delusions than process females because of concern about their identity as women. This gender-identity effect was not expected for male schizophrenics. Three aspects of sexuality, as represented in delusional content, were considered in the sexual score. Ratings of whether the patient introduced sex into her/his delusion where the patient was the initiator or the object were combined with ratings of how critical the sexual themas were to the overall delusion. Interjudge agreement over the sexuality ratings was satisfactory (r= 0.65, p < 0.01). It was predicted that reactive paranoid women would be singularly committed to sexuality themes with neither the process paranoid women nor either male paranoid group placing similar emphasis upon sex in their delusions. This was tested by comparing the ratio of reactive women having sexuality scores above the sample median of 3.5 and at/below this figure (7: 1) to the ratio for the remainder of the sample (9: 15). The exact probability (0.018) for this disproportionality confirmed the predicted emphasis upon sex in the delusions of female reactive paranoid schizophrenics only. The final prediction in this study proposed that reactive women would specify more distinctly female items or activities in their delusions than process women because of genderidentity concerns. Examples of such items or activities that were found were tampons, a dress, jewelry, purse, vagina, uterus and menstruation. Interjudge agreement regarding the presence (score = 1) or absence (score = 0) of such items or activities was 0.87 @ < 0.01). The ratio of reactive paranoid women specifying at least one distinctly female item or activity in their delusion to the number not doing so (5:3) was compared with the same ratio for process paranoid women (1:6). The difference approached significance @ < 0.078) and was in line with prediction; reactive schizophrenic women placed greater emphasis upon gender-defining content. This analysis tends to underestimate the differential inclusion of gender-defining terms in the expressed delusions of these two types of patients, since reactive women always made multiple references to such terms (i.e. 2-4 instances) if reference was made at all. If the ratios involved the number of female paranoids who made multiple references to gender-defining terms in their delusions, the proportions of reactives who did and did not (5:3). compared with process (0:7), would achieve stastistical significance (p=O.O19). DISCUSSION The present study provided results that supports the major assumptions under test. The first of these assumptions represented a new interpretation of defensive projection and its role in paranoid delusions. Projection was proposed to be a cognitive process by which the person perceptually revises the characteristics of other people that serve as comparison

PROJECTIONSAND PARANOID DELUSIONS

171

standards for self-evaluation. Perceptual revision occurs by selective perception or misattribution of the negative trait in question and always involves degrading the quality of the social environment. Social comparison, under this circumstance, allows projectors to minimize or deny the unacceptable trait in themselves. The faulty attribution may require projectors to reconsider their relationships with the social environment in order to accommodate their misperceptions. If, for example, hostility is the distorted attribute, the projector must come to terms with the hostility of those people that serve as standards of social comparison. These explanations by their nature must be unrealistic, and in the case of more disturbed people (e.g. schizophrenics) may initiate or feed into delusional systems. Since only those who are sensitive to their social environment would be expected to emphasize social comparison in self-evaluation, this explanation should be applicable to the development of paranoid delusions in reactive (good premorbid) schizophrenia but not in process (poor premorbid) schizophrenia. The results were consistent with the hypothesis that projection is a distortion of the socialcomparison process. People implicated in the delusions of reactive paranoid schizophrenics were more clearly specified than was true for their process counterparts. Social comparison should be a more preferred basis for self-evaluation in reactives and would be expected to lend itself to clearer specification, because actual people serving as standards are likely to become implicated in the delusion. Analysis of self-descriptions provided by normals tending to use projection as a defense suggested further differences in delusional content given the functional role of projection in reactive delusions but not process delusions. Men, especially concerned about being considered untrustworthy, should project untrustworthiness. Women in an effort to deny hostility should project hostility. Examination of the thematic content of delusions revealed these gender-linked projected qualities in the delusions of reactive paranoids; reactive males placed more emphasis upon their inability to trust others in their delusions and reactive females emphasized hostility toward them from others. These thematic patterns were not found in male and female process paranoids. The self-descriptions of projectors also led to the prediction that reactive paranoid women would be concerned about their basic female identity. It was expected that delusional content would affirm this concern by themas emphasizing their sexuality and by reference to items and activities that are specific to women, Reactive paranoid females, compared with the remaining patients, placed more emphasis upon sexuality and more frequently made multiple references to female-defining terms than process paranoid females. A question that remains, given the thema results, is why the sense of not being worthy of trust in men and the sense of being hostile and lacking in female identity in women should be vulnerable self-conceptual areas in projectors and specificaly defended by projection? Normal projectors were generally self-critical and able to tolerate the admission of many other unfavourable qualities. One possibility rests on the logical assumption that people who are sensitive enough to their social environment to invoke social comparison in order to evaluate their own traits are also likely to depend heavily upon the reactions of others to them for self-definition. These self-perceived qualities, being trustable and being womanly, may be especially vulnerable, because they are confounded with one’s ability to establish and maintain effective and congenial social relationships. You must relate to

172

ALBERTB. HEILBRUN. JR et al.

other people in order for them to display trust in you or for them to react to you as a woman. Defensive perceptual distortions might be required in order to deal with concerns about trust and gender identity when the person is both sensitive to social appraisal and socially withdrawn. In this circumstance, more constructive efforts for men to prove themselves worthy of trust by others or for women to confirm their female identity are made difficult by their social withdrawal. This possibility requires research documentation, although it is not entirely speculative at this time. A rather powerful relationship was found for both male and female clients of a college mental health agency between their scores on the projection scales used in the present study and the MMPI Social Introversion scale (HEILBRUN, 1982). Those who used these defenses more extensively also displayed signs of social introversion. The second and more distant assumption underlying the present study is that the processreactive distinction remains a useful way of classifying schizophrenics as far as understanding symptom development. The present results and those reported by HEILBRUN and HEILBRUN (1977) reveal distinct differences in the form and substance of delusions expressed by process and reactive paranoid schizophrenics. Similarly, other studies (HEILBRUN, 1980; HEILBRUN et al., 1983; HEILBRUN and BLUM, 1984) have identified both common and distinct cognitive deficits in process and reactive schizophrenia demonstrating auditory hallucinations. Both series of studies imply that different explanations of symptom vulnerability in schizophrenia are required for process and reactive schizophrenics. Furthermore, process and reactive vulnerability to schizophrenic symptoms seem to have their origins in differential and extreme styles of attention deployment, as hypothesized earlier (HEILBRUN, 1973). Methodological weaknesses in the present study require comment. Limited sample size; use of different cutting scores to define process/reactive status in male and female patients so as to avoid even further problems with subgroup numbers; dependence upon DSM-II for the diagnosis of schizophrenia, the current standard when the delusional materials were collected; and only moderate interjudge agreement when these materials were analyzed are some of the flaws. Accordingly, the results should be treated conservatively until they are replicated. On the other hand, the fact that predicted relationships were disclosed despite these limitations increases our confidence that such a replication would be successful. Acknowledgement-The Psychology, College

authors would like to express their appreciation to Glenn Shean, of William and Mary, for making the basic patient data available.

Department

of

REFERENCES AMERICAN PSYCHIATRICASSOCIATION:Committee on Nomenclature and Statistics (1968) Diagnostic and Sfatistical Manual of Mental Disorders (DSM-II). APA, Washington DC. CAMERON, N. (1963) Personality Development and Psychopathology: A D.vnamic Approach. Houghton Mifflin, Boston. CHALUS, G. (1976) Relationship between paranoid tendencies and projective behaviour. Psycho/. Rep. 38, 1175-1181. DEWOLFE, A. S. (1974) Are there two kinds of thinking in process and reactive schizophrenics? J. abnorm. Psychol. 83, 285-290. EISENTHAL, S., HARFORD, T. and SOLOMON, L. (1972) Premorbid adjustment, paranoid-nonparanoid status, and chronicity in paranoid schizophrenic patients. 1. nerv. menf. Dis. 155, 227-231, FEIGHNER, J. P., ROBINS, E., GUZE, S. B., WOODRUITF, R. A., WINOKUR, G. and MUNOZ, R. (1972) Diagnostic criteria for use in psychiatric research. Archs gen. Psychiat. 26, 57-63. FESTINGER, L. (1954) A theory of social comparison processes. Hum. Relat. 7, 117-140. GOLDSTEIN, M., HELD, J. and CROMWELL, R. (1968) Premorbid adjustment and paranoid-nonparanoid status in schizophrenia. Psycho/. Bull. 70, 382-386.

PROJECTIONS AND

PARANOID DELUSIONS

173

COUGH,H. G. and HEILBRUN, A. B. (1980) The Adjective Check List Manual. Consulting

Psychologists Press, Palo Alto, CA. HEILBR~, A. B. (1972) Defensive projection in late adolescents: Implications for a developmental model of paranoid behavior. Devel. Psychol. 7, 880-891. HEILBRUN, A. B. (1973) Aversive Maternal Control: A Theory of Schizophrenic Development. John Wiley, New York. HEILBRUN, A. B. (1977) The influence of defensive styles upon the predictive validity of the Thematic Apperception Test. J. pers. Assess. 41, 486-491. HEILBRUN, A. B. (1978) Projective and repressive styles of processing aversive information. J. consult. c/in. Psycho/. 46, 156-164. HEILBRUN, A. B. (1980) Impaired recognition of self-expressed thought in patients with auditory hallucinations.

J. abnorm. Psycho/. 89, 728-736. HEILBRUN, A. B. (1982) Psychological scaling of defensive cognitive styles on the ACL. J. pers. Assess. 46, 495-505. HEILBRUN, A. B. and BLUM, N. A. (1984) Cognitive vulnerability to auditory hallucination: impaired perception of meaning. Br. J. Psychiat. 148, 508-512. HEILBRUN, A. B., BLUM, N. A. and GOLDREYER, N. (1985) Defensive projection: An investigation of its roles in paranoid conditions. J. nerv. ment. Dis. 173, 17-25. HEILBRUN, A. B., BLUM, N. A. and HAAS, M. (1983) Cognitive vulnerability to auditory hallucination: Preferred imagery mode and spatial location of sounds. Br. J. Psychiat 143, 294-299. HEILBRUN, A. B. and CASSIDY, J. (1985) Toward an explanation of defensive projection in normals: The role of special cognition. J. sot. c/in. Psycho/. 3, 190-200. HE~LBRUN,A. B. and HEILBRUN, K. S. (1977) Content analysis and delusions in reactive and process schizophrenics.

J. abnorm. Psycho/. 86, 597-608. HEILBRUN, A. B. and PEPE, V. (1985) Awareness

of cognitive defenses and stress management.

Br. J. med. Psycho/.

58, 9-17. HIGGINS, J. (1969) Process-reactive schizophrenia: recent developments. J. nerv. ment. Dis. 149, 450-572. HOLMES, D. S. (1968) Dimensions of projection. Psycho/. BUN. 69, 248-268. JOHANNSEN, W., FRIEDMAN, S., LEITSCHUH, T. and AMMoNs, H. (1963) A study of certain schizophrenic dimensions and their relationship to double alternative learning. J. consult. Psychol. 27, 375-382. JOHNSON, M. H. and RIES, H. A. (1967) Validational study of the self-report scale for process-reactive schizophrenics. J. consult. Psychol. 31, 321-322. MCCREARY, C. P. (1974) Comparisons of measures of social competency in schizophrenics and the relation of social competency to socio-economic factors. J. abnorm. Psycho/. 83, 124-129. MEICHENBAUM, D. H. (1969) Validational study of martial status and the self-report scale for process-reactive schizophrenia. J. consult. c/in. Psycho/. 33, 351-356. NEALE, J. M. and CROMWELL, R. L. (1970) Attention and schizophrenia. ln Progress in Experimental Personality Research (Edited by MAHER, B.A.) pp. 37-66. Academic Press, New York. NEALE, J. M., KOPFSTEIN, J. H. and LEVINE, A. J. (1982) Premorbid adjustment and paranoid status in schizophrenia: Varying assessment techniques and influence of chronicity. Proceedings of the 80th Annual Convention of the American Psychiatric Association, Vol. 7, pp. 321-322. APA, Washington DC, NIEDERLAND, W. G. (1974) The Schreber Case: Psychoanalytic Profile of a Paranoid Personality. Quadrangle, New York. PAYNE, R. W. (1973) Cognitive abnormalities. In Handbook of AbnormalPsychology (Edited by EYSENCK, H. J.) pp. 420-483. Pitman, London. SIEGEL, S. (1956) Nonparametric Statistics for the Behavioral Sciences. McGraw-Hill, New York. SILVERMAN, J. (1964) Scanning-control mechanism and “cognitive filtering” in paranoid and nonparanoid schizophrenia. J. consult. Psycho/. 28, 385-393. ULLMANN, L. P. and GIOVANNONI, J. M. (1964) The development of a self-report measure of the process-reactive continuum. J. nerv. ment. Dis. 138, 38-42. VENABLES, P. H. (1964) Input dysfunction in schizophrenia. In Progress in Experimental Persona/ity Research, (Edited by MAHER, B.A.) pp. 1-47 Academic Press, New York. WILLIAMS, J. E. and BEST, D. L. (1977) Sex stereotypes and trait favorability on the Adjective Check List. Educ. psycho/. Measur. 37, 101-110. ZIGLER, E. and GLICK, M. (1984) Paranoid schizophrenia: an unorthodox view. Am. J. Orfhopsychiat. 54,43-70. ZIGLER, E. and LEVINE, J. (1973) Premorbid adjustment and paranoid-nonphranoid status in schizophrenia: a further investigation. J. abnorm. Psycho/. 82, 189-199. ZIGLER, E. and LEVINE, J. (1981) Premorbid competence in schizophrenia: What is being measured? J. consult.

clin. Psycho/. 49, 96-105. ZIGLER, E., LEVINE, J. and ZIGLER, B. (1976) The relation between premorbid nonparanoid status in schizophrenia: A methodological and theoretical critique.

competence

and paranoid-

Psycho/. BUN. 83, 303-313.