A v firvn’xx vuu‘ ‘31 a. v .u’ .au .1 v "u 1 ivy-nu; ”3'8 ...,. 11"- .1... In]. ‘a rim. “HI-hi ;.1 1.1.4 flaws-13": '19:? . :. :§ $11 kg‘. .3 ""3? ’éww-Iiflqu ~ 32:31“ .:- 4 3. y 3 ,‘t .13. " on;- ‘2‘; :3 1 d.» ‘2 ‘ ‘fihil 1 . 1 .3 151-... .1. 3 . ' . “355x: ‘ . 1-.. 11 v ‘~ 2- 3.2: \11. 1. ’ 3 M” Nam s~-g:: {‘5‘ 3 A 15“.““I ‘4 1. “(3% 11‘- “'Ez‘afi‘k: .1€%’1~°1:;" :2: 112.5% ~22 -{{-~“3- " 53139:, 313.1. 131-1“? . vi r+ 1.:13‘ , $3111". 1:? nit-«4%?- #:3131332 "$4.3m: .1 3: 1w. ‘4 VA .1 ram-«'1:- ' I J- 33,331,343'321... i . %H1a“!%\1' .35- v 1‘}. F, < s' ‘1‘.“ ;-'\ K 1;“ w 31* a: W *1 ‘ , ti. . ",{vfl'}. (:3 r m- '. . 7 . .1.- .1.'1'1€.'.:'.i-' “.1 , A :1 MN! .':..r'...l_ .-.-"’,'r"5: ‘ a"? '57:" *1 ”1mm! 3;“..41m-1 "W 31:1} :1“ ”.134"? .323"..- an M "flit”. "3.03.”. . r. 0mm”. . ”1- v-rn-n 1.1., A. z £“!}1$"V"”:Yf" 1 rfvao" "-1- rem-w 7 ‘1 .g" ,3. -...-:'..;. 31.,..r..,2.n:§"“ ..~ 976369.17 IHIWUIHHIII”I”NHIHHHIHIIHIIIIUUIIHHIHIIHI 300774 3127 LIBRARY Michigan State University J p — This is to certify that the dissertation entitled Gender Differences in Narcissism;. Shame, Grandiosity, Object Relations and Measurement presented by Carol S . Schwartz has been accepted towards fulfillment of the requirements for Ph.D. degreein Clinical Psychology Major pmfg : Date 6/2;/f/ MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 PLACE IN RETURN BOX to remove thie checkout from your record. TO AVOID FINES mum on or More due due. 19-5517 ' . r.” I‘_ —+—7s h ’t MSU Is An Affirmative Action/Equal Opportunity institution GENDER DIFFERENCES IN NARCISSISM: SHAME, GRANDIOSITY, OBJECT RELATIONS AND MEASUREMENT BY Carol 8. Schwartz A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Psychology 1991 ABSTRACT GENDER DIFFERENCES IN NARCISSISM; SHAME, GRANDIOSITY, OBJECT RELATIONS AND MEASUREMENT BY Carol S. Schwartz Narcissism has been conceptualized as a character pathology in which two extreme states of awareness exist, one state characterized by inflated self esteem and the other by deflated self esteem. Several authors have suggested that across narcissistic individuals, a favored state may develop during childhood, such that female narcissists may be more likely to develop deflated self esteem as the dominant self perception, while male narcissists would tend to develop the inflated self perception position. These tendencies would then interact when male and female narcissistic characters form relationships with each other, such that the less conscious self perception would be projected into the partner. Six hypotheses were tested on 200 college students using 4 self report measures and three TAT cards, administered in a group setting. From among the subjects scoring high on narcissism, it was predicted that (1) females would have more conscious shame than males (2) females would have more unconscious grandiosity than males, while males would have more unconscious shame than females, (3) females would project unconscious grandiosity into males while males would project unconscious shame into females, (4) males and females would project images of females as psychological extensions of males, (5) males would score higher on the NPI, a leading instrument used to measure narcissism which favors the grandiose defense, and (6) males and females would show different response patterns on the OMNI, an instrument which measures the shame and grandiose defense separately. All hypotheses tested yielded nonsignificant findings. While females scored significantly higher than males on conscious shame, no significant interaction emerged between conscious shame and level of narcissism. The discrepancy between male and female scores on unconscious shame and grandiosity increased in the directions predicted as level of narcissism increased, but did not increase to the level of significance. Males and females scoring high on narcissism tended to project female TAT characters as extensions of male TAT characters, but this finding was just short of significance. The total male sample scored significantly higher on the NPI, but subjects identified as high in narcissism showed no significant sex difference in NPI score. No sex differences were found on the OMNI scale scores . To my children Michael and Aaron iv ACKNOWLEDGMENTS My thanks go to the people who actively shared my interest in these ideas; Chuck, Nick, Wes, Charles, Mary, Robert and Doug; to my Committee members who provided guidance, encouragement and support without exception, Bert, Ellen, Elaine and Lee; to those who helped me with the myriad of tasks, large and small, Sigmund, Mary, Chuck, Sharon, and Suzie, and to those who provided emotional, financial and practical assistance on a more personal level, my family and Chuck. Special thanks to Doug for opening his extensive library to me and for providing an introject for the essential humor required to sustain a task such as this one: to Sue for patiently teaching me how to analyze my data and for opening her home to me for three months so that I could use her computer for that task; and to my Chairperson, Bert, for being my friend and mentor, helping me to remember what I love most about this field, and reminding me to keep an eye toward the future. Most of all, my gratitude to Chuck, for listening, questioning, considering, and suggesting; for acting as computer and printer consultant, and for providing himself as an on-going idealizable and mirroring object. LIST OF TABLES . . . . . . . INTRODUCTION . . . . . . . Statement of Problem . Literature Review . . . Definition, History, Major Theories Phenomenological and Clinical Theory . Theories about two Types of Narcissism Sex Differences in Object Relations Borderline vs. Narcissistic Patterns Summary of Literature Review . . . Research and Instruments Current Research . METHOD . . . . . . . . . . . Research Design . . . . Research Instruments . Procedure . . . . . . . Data Analysis . . . . . RESULTS . . . . . . . . . . Summary of Main Results Specific Hypotheses . . DISCUSSION . . . . . . . . . vi viii 10 16 31 39 55 59 81 85 85 86 92 97 102 102 103 110 APPENDICES A. B. C. F. G. H. I. J. LIST OF Thematic Apperception Test Cards . . . . TAT Scoring Criteria for Shame and Grandiosity . . . . . . . . . . . . Criteria for Coding Extension . . . . . . Narcissistic Personality Inventory . . . Narcissistic Personality Disorder Scale . O'Brien Multiphasic Narcissism Inventory Internalized Shame Scale - Modified . . . Correlations . . . . . . . . . . . . . . Consent Form . . . . . . . . . . . . . . Debriefing Form . . . . . . . . . . . . . REFERENCES 0 O O O O O O O O O O O O O O 0 vii 119 123 126 128 132 134 136 137 138 139 141 Table Page 1. ANOVA-CS. Shame By Sex and Narcissism Level . . 103 2. ANOVA-UCS. Grandiosity By Sex and Narcissism mvel O O O O O O O O O O O O O O O O O O O 104 3. ANOVA-UCS. Shame By Sex and Narcissism Level . . 105 4. Female Subjects-Projection of Grandiosity into Male Characters . . . . . . . . . . . 106 5. Male Subjects-Projection of Shame Into Female Characters . . . . . . . . . . . . . . . . 106 6. All Subjects - Projection of Female Extension Characters . . . . . . . . . . . . . . . . 107 7. ANOVA-NPI Score by Sex and Narcissism Level . . 108 8. ANOVA-Sex Differences in OMNI-NA Factor by Narcissism Level . . . . . . . . . . . . . 109 9. ANOVA-Sex Differences in OMNI-NC Factor by Narcissism Level . . . . . . . . . . . . . 109 10. Mean UCS Grandiosity Scores by Sex and Narcissism Level . . . . . . . . . . . . . 114 11. Mean UCS Shame Scores by Sex and Narcissism Level 0 O O O O O O O O O I O O O O O O O I 115 viii IEIBQDHQIIQN STATEMENT OF PROBLEM Many clinicians have written that the narcissistic character structure is a system in which two contradictory states of awareness exist, one state characterized by inflated self esteem, and the other by deflated self esteem (Bach 1978; Kohut 1971, 1977, 1981; Kernberg, 1975 in Goldstein, 1985). Bach discusses the etiology and phenomenology of these two states. Kohut observes these same two distinct self states through the transference relationship in psychotherapy. Several authors have suggested that there are two Lypee of narcissistic pathology (Johnson 1987; Shulman 1986: 0'Leary 1986; Broucek 1982; Miller 1981, 1984, 1985, in O'Brien, 1988). These two types are differentiated on the basis of affect, defense and object relations. While the two types are described somewhat differently by the authors above, there is significant overlap between them. Two of the authors above suggest that the two types of narcissism are gender linked. (Johnson, 1987, 0'Leary, 1986). The two self states described by Bach and Kohut (above), again characterized by inflated and deflated self esteem, resemble 2 the two types of narcissism discussed by these authors. Another approach to the "two-state" "two-type" phenomenon is offered by Philipson (1985) and Reich (1953). These two authors write specifically about eex differegeee in object relations patterns of the narcissistic individual, and focus upon the uniqueness of the female narcissistic style. Philipson has written that the common definition of narcis- sism (DSM-III-R) describes a disorder that is primarily experienced by men, while women have "asymmetrical yet intersecting problems" as a consequence of individuation differences. Philipson argues that the female narcissistic style is not described in the literature on narcissism and is therefore not considered narcissism. Her description of the female narcissist resembles one of the two typee of narcissism discussed by the authors above, that type being attributed to the female more than the male narcissist by those authors. Still another perspective on different types of, or sex differences in narcissism is offered by Masterson (1981), Akhtar and Thompson (1982) and Haaken (1983). Like Philipson and Reich, they consider that individuation differences promote distinct symptom pictures for males and females. Unlike the authors above, however, they frame these differences by arguing that there is a tendency toward a higher incidence of narcissism in men, and a higher inci- dence of DQIQQIILDQ pathology in women. Their description 3 of borderline pathology bears some likeness to the female "type" of narcissism discussed by some of the authors above. There are other arguments concerning apparent differences or sex difference in narcissistic pathology (Lachmann, 1981). Few studies, however, have investigated these differences. A review of this literature indicates that much confusion exists. This confusion may lead to misdiagnosis and/or treatment error. In summary, four perspectives on a divergence in narcissistic pathology have been briefly outlined. These are: 1. There are two states of awareness which alternate within each narcissistic individual. One state is characterized by idflated self esteem, and the other by deflated self esteem. 2. There are two types of narcissism (a) which may be differentially represented by gender. These two types resemble the two etegee in #1 (above). 3. There are sex differences in object relations patterns of the narcissistic individual. (a) The female narcissistic style is fundamentally different from the male style, and is not included in the literature on narcissism. 4. There is more narcissistic pathology in men, and more borderline pathology in women. The current research is designed to explore 2(a) and 3 above. That is, (2a) are there two types of narcissism which are divisible by gender and (2) are there measurable gender differences in object relations patterns of the 4 narcissistic individual such as those described by Philipson and Reich. In order to explore these questions, it will be necessary to go beyond the ways in which narcissism has been conceptualized to the specifics of how it has been measured. We can then explore, in addition, what contribution the measuring instruments have made to the ideas discussed above. In the following pages the relevant literature will be reviewed concerning patterns of narcissistic defense and style which may be related to gender. Certain questions will then be posed in order to investigate the presence of gender differences in a sample identified as narcissistic. 5 LITERATURE REVIEW The following review of the literature on theories about differences or gender differences in narcissism will serve as background on the issues mentioned above. Before discussing different types of narcissism, however, some historical and contemporary theories of narcissism will be presented as an introduction. 1. DEFINITION, HISTORY, MAJOR THEORIES Narcissism, the term and the concept, has had many different definitions in the 90 years in which it has appeared in the literature in clinical psychology. Pulver (1986) writes that in early psychoanalytic literature the word was used in four different ways: (1) to describe a sexual perversion, (2) to define a developmental stage, (3) to identify types of object relationships and (4) to denote various aspects of the complex ego state of self esteem. The latter three usages remain in the contemporary literature. According to Morrison (1986) Havelock Ellis, in 1889, was the first to use the word Narcissism to describe a clinical disorder. He borrowed the term from the Greek myth of Narcissus. According to this myth, Narcissus was a creature who rejected the affection of others and was consequently, as punishment for his insensitivity, cursed to fall in love with his own elusive and unresponsive reflection. He became emotionally paralyzed by his 6 obsession with his reflection and ultimately died from this paralysis. Ellis found a similarity between the characteristics of Narcissus in the myth and certain of his clinical patients. In his words, narcissism was described as "that tendency which is sometimes found, more especially perhaps in women, for the sexual emotions to be absorbed, and often entirely lost, in self-admiration." Freud's first use of the term came in 1914 with the publication of "On Narcissism: An Introduction." In this paper, narcissism was defined in much the same way as Ellis had defined it. In Freud's words, narcissism was "the attitude of a person who treats his own body in the same way in which the body of a sexual object is ordinarily treated - who looks at it, that is to say, strokes it and fondles it till he obtains complete satisfaction through these activities." Freud's principle contributions in this paper included (1) defining narcissism and narcissistic object choice (the kinds of relationships a narcissistic patient would engage in), (2) defining primary and secondary narcissism (differentiating the natural self absorption of the infant from that of the adult), and (3) defining the creation of the ego ideal or the idealizing tendencies of the ego in the formation of narcissistic psychic structure. Freud was not satisfied with his early conceptualization of narcissism and said as much in letters to colleagues. He did not, however, as he had originally intended, return to 7 the topic of narcissism to further refine and develop his theory. The concept of narcissism was next taken up by the ego psychologists. Since that time the term has often been used either as a synonym for self esteem, or to refer to a tendency toward self absorption. The most esteemed contemporary theories, those of Kernberg and Kohut, describe narcissism as a deficit in the regulation of self esteem. It is still used today, however, to describe a personality type dominated by high self esteem (Harder, 1979; Raskin, 1981). Cooper, (1986) in contemplation of this inconsistency writes, It has become increasingly apparent that the term is overburdened with the baggage of its past that it has perhaps outlived its usefulness. The descriptive or explanatory . . ideas behind the term are not uniformly agreed upon, and often the word is used as if it explained a phenomenon. (p. 118) Despite this documented confusion, the term is more widely used today than ever before. In now appears with great frequency in the literature, both theoretical and empirical. e c' s Freud did not write specifically about the male narcissist. Rather, it is assumed that when he writes about the narcissist, he is talking about his male patients. In his discussion of female narcissism he writes specifically about how the female narcissist is different. Both Freud (1925) and Deutsch (1944, 1945) considered that male and female narcissism had different roots. They wrote that the 8 seeds of female narcissism grew out of the realization of, and disappointment over, not having a penis (Lachmann, 1982). This perceived anatomical injury was believed to have caused a narcissistic injury from which women never recovered. To compensate, extra attention was devoted to appearance of the face and body. According to Lachmann (1982), Deutsch was the first to formulate a comprehensive psychoanalytic statement of the role narcissism plays in female psychosexual development. She considered that passivity, masochism and narcissism were biologically given characteristics of femininity. The masochism was exacerbated by narcissistic rage, about "castration", which was turned inward. What Freud and Deutsch did not account for adequately was the importance of the early mother-child relationship, preoedipal factors, and the time prior to the girl's discovery of anatomical differences (Lachmann, 1982). According to Lachmann (1982) Jacobson, in 1964, contributed a fundamental change to theories about female narcissism. She considered that gender identity developed preoedipally in girls as a consequence of their identification with mother, and thus developed separately from their awareness of anatomical differences. She argued further that the issue of power, which is central in narcissistic pathology, is renegotiated during the oedipal phase but is largely determined by the father's attitude toward the little girl's emerging need for self assertion. 9 Thus, if the father is delighted by his daughter's emerging sense of power, the fact that she has no penis will be of little consequence to her. gentemeererx_lhe2rie§ Jacobson's theory foreshadowed the current, predominant theories of narcissism; those of Kernberg and Kohut. While Kernberg and Kohut disagree on many aspects of their respec- tive theories, they agree on many others. Both theories are gender neutral. Both consider, like Jacobson, that narcissistic personality disorder evolves as a consequence of inadequate parenting prior to the oedipal phase, and stress the object relationship between mother and child as the primary causative agent. Kernberg and Kohut also agree on many of the symptoms of the narcissistic individual. These symptoms include difficulties with adaptation to reality, impulse control, frustration tolerance, affective stability, reality testing, interpersonal relationships, defensive splitting, feelings of omnipotence, and primitive idealization. Points of disagreement include the role of aggression, the differential diagnosis between the borderline and the narcissist, the specific developmental precursors, and the treatment. Both of these theories are well represented in the original and revised edition of the Third Edition of the Diagnostic and Statistical Manual of the American Psychiatric Association (DSM-III) (APA, 1980), (DSM-III-R) (APA, 1986) (Goldstein, 1985). Kohut's theory 10 will be discussed in the following section, and will be used the most extensively in this study as the theoretical basis for defining narcissism. 2. PHENOMENOLOGICAL & CLINICAL THEORY: TWO INTERNAL STATES Bach (1976) describes the phenomenology of the narcissistic state as follows. The patient experiences two distinct states of awareness which are not integrated and so alternate in consciousness and cannot be experienced simultaneously. Bach considers that the narcissistic state of consciousness can be characterized most accurately on a phenomenological level as a defect in reflective self awareness. He describes two states of reflective self awareness: ebjegtixe and eebjeeEiye. According to Bach, the narcissistic individual suffers from a deficit in the ability to make a smooth transition between the two states of ebieeeige and egbieetiye self awareness. Rather, he gets stuck in one or the other, experiences each in an extreme way, and is comfortable in neither state. The interpersonal experience with the caretaker during infancy and childhood provides the stimulus for these extreme self experiences. Bach describes these experiences as follows: (A) The caretaker, usually the mother, characteristically imposes her own initiatives upon the child rather than mirroring the child's real self back to him. For example, the child may be absorbed in his play with a toy when the 11 mother enters the room abruptly and exclaims, "Look what a mess you have made!" While the child is engaging in a spontaneous process with the toy he might be described as "lost" in the experience. His experience of "I" is firmly located inside himself as he loses himself in the me-and-my- toy experience. When the mother abruptly disrupts him, demanding that he see himself from her perspective, the child experiences himself suddenly as something which is being looked at from the outside. The experience of "I" is then firmly located outside himself in the other. According to Bach, the continuous imposition of mother's perspective on the child, over the child's perspective, may promulgate a self experience of the ”I” as being lived by forces external to the self. The child repeatedly exposed to this experience loses the capacity to engage in spontaneous play. Rather he/she becomes uncomfortably preoccupied with how he/she appears to others, a state of consciousness Bach labels "objective self awareness." A parallel self experience develops, however, which is the promulgator of a state of extreme self absorption. (B) The mother who imposes her own initiatives upon her child is also one who is repeatedly unavailable for emotional refueling and may thus lead the child to feel he has no recourse other than himself. The child may as a consequence be prone to states of extreme self absorption or hypercathexis of the self. Within this state, there is a feeling of omnipotence which arises as a defense against the 12 awareness that the powerful mother is unavailable. Within the model of developmental psychoanalysis the self would be characterized as merged with the powerful mother as an unconscious defense mobilized to avoid the too—painful awareness of being a small, separate person without assistance in a very scary world. While in this state of consciousness, objective self awareness is suspended and Bach would consider the individual subjectively self aware. These two self states then, subjective and objective self awareness, alternate in consciousness. The narcis- sistic individual tends to feel either uncomfortably aware of how he/she appears to others and thus conscious of his/ her separateness and vulnerability, or lost within him/ herself, unaware of vulnerability, oblivious to reality, and grandiose in the sense of experiencing exaggerated powerfulness. Both states are experienced in an extreme way and transitions between them are not smooth or predictable. The individual, in fact, is scarcely aware of making these transitions or of having two distinct self perceptions that are quite contradictory. In one state self-consciousness or shame is conscious and grandiosity is disavowed. In the other state, self-absorption, insensitivity to others and thus grandiosity is conscious and shame is disavowed. Bach does not discuss the possibility of a sex difference in the extent to which one or the other state of awareness manifests itself. 13 Kohut (1971, 1977) describes the internal contradictory polar states (above) through their transference manifesta- tions in psychotherapy, and in fact considers that the development and recognition of both of these transferences is the only way to definitively diagnose a narcissistic character. The two transference manifestations which Kohut originally introduced were the Idealizing and the Mirroring transferences. He considered that both of these transferences were normal developmental phenomena which needed to be realized, understood by the caretaker and gratified in order for a child to satisfy his needs and thereby grow to realize his true potential. Each transference represents a state of consciousness, a way of feeling about the self, a way of feeling about the other, and specific affects and behaviors which are manifestations of these distinct states. Kohut considers that the achievement of realizing and integrating one's separateness and vulnerability requires two things: to have a powerful caretaker to look up to who functions as a protector and a role model, end to have the eel; mirrored in a positive, self-enhancing way. If the toddler feels he has a protector he can look up to, and that he himself is strong and good, he will be able to tolerate the growing awareness of his separateness and vulnerability. He will be able to slowly internalize the functions of the mother which help him to maintain his sense of strength. He 14 will be able to soothe himself when alone the way his mother does when they are together. If both or either of these maternal functions are inconsistently available to him, he may not be able to internalize his mother's strength and integrate it with his developing sense of self. He will, thus, be unable to tolerate his separateness and vulnera- bility and so will defend against it with symbiotic illusion and/or grandiosity. The need for an object (person) to idealize and an object to mirror (reflect) his strength will not disappear. In fact, according to Kohut, the narcis- sistic individual projects these qualities onto all persons, forming what he terms idealizing and mirroring transferences with the world in an effort to get his early developmental needs met. It is the task of the therapist to recognize these transferences in the therapeutic situation, recognize the underlying needs, understand and explain the patient's need to be mirrored and to have an object to idealize, and thereby enable the individual to grow beyond his early developmental fixation point. The phenomenological and behavioral state of the narcissistic individual while in these two transferential states resembles the statee of narcissism discussed above. While in the idealizing transference the individual is looking up to the other, feeling as though the other is all powerful, beautiful, strong, capable. The self, if sepa- rated from the idealized other, feels small, powerless, 15 weak: feels in fact the two-year-old self within, which never grew beyond these childhood perceptions. The individ- ual in this state is conscious of their shame. While in the mirroring transference, by contrast, the function of the other is to mirror the self, to make the self feel real, alive, important, powerful, etc. In this state, vulner- ability is less conscious if conscious at all. In fact, the feelings which are conscious while engaged in a mirror transference are the same feelings which are projected while in the idealizing transference. The narcissistic.individual carries within him a pre-ambivalent perception of self and other. Both self and other are experienced as one- dimensional. She/He does not perceive integrated, realistic images of others or self with both strengths and weaknesses. Strengths and weakness, or grandiosity and shame are separated internally. One is always conscious, the other always projected. In summary, Bach describes two nonintegrated states of awareness which are presented according to their etiology and phenomenology. Kohut describes two nonintegrated states of awareness which are observable through the transference in psychotherapy. Both of these clinical descriptions of internal states resemble the description of two clinical types (shame conscious and grandiose conscious) presented by Shulman, Broucek, O'Leary and Miller (below). 16 3. THEORIES ABOUT TWO DISTINCT TYPES OF NARCISSISM Johnson (1987) considers that there are two types of narcissistic pathology, differentially represented in males and females, and also argues for two states of awareness within the narcissistic individual: the "false, grandiose self" and the "symptomatic self.” The false self is built as a defense against the symptomatic self. The "real self" remains unconscious until the therapeutic process brings it into awareness. Johnson (1987) considers that men and women emerge from Mahler's (1975) Rapprochement phase (15-24 months) with dif- ferent character styles if trauma or frustration during that time period prevents need satisfaction. The different styles of defense which are adopted are attributable to sex differences in the interpersonal climate between mother and infant, and cultural patterns which reinforce these differences. A full explanation of Johnson's ideas follows. Mahler (1975) labels the period 15 to 24 months 'The Rapprochement with reality'. The task of this developmental period involves confrontation of one's separateness and vulnerability. In the developmental period just prior to the Rapprochement, (the Practicing period) the child is not aware of his separateness and vulnerability, but experiences himself as one with his caretaker and thus, just as powerful as he feels his caretaker to be. According to Mahler, the toddler at this stage feels as though the "world is his 17 oyster". Because this power is borrowed rather than real, the toddler is said to be "grandiose". The grandiosity and elation of the practicing period provides a natural defense against the realization during the rapprochement period of separateness and vulnerability. In other words, if the realization of separateness and vulnerability is too scary, the individual will defend against the fear by remaining stuck psychologically at the practicing phase, continuing to preserve the illusion that he is powerful and remaining unaware of his real separateness and vulnerability. As mentioned above, Johnson considers that two distinct character styles may evolve as a consequence of severe environmental frustration during the rapprochement period: the Narcissistis_gharaster and the Symbietis_§harasfer. Johnson states: "The Synnieeie_gne;ee§e;, who resembles most clearly the type of borderline individual described by Masterson (1976, 1981) suffers primarily from the failure to resolve the issue of separateness. Insufficiently individuated, the symbiotic character can feel or know herself (sic) only in immediate relation to another. Thus, she will tend to want to merge with or alternatively push away from a signi- ficant other in order to keep some continuing sense of her own existence, boundaries, or identity. Like the youngster she emulates, she will be characterized by attempts to coerce others to take care of her, respond to her, fight with her, and in all of these ways affirm her otherwise fragile existence. (p. 28) "The Nersissisfis.§herester derives primarily from a failure to accommodate around the issue of grandiosity and limitation. More often a male in this culture, he finds a bulwark for his uncertain sense of self more in the pursuit of his grandiose illusion than in the pur- suit of symbiotic illusion. Due to one kind of environmental frustration or another, the narcissist is arrested around the neutralization of grandiosity. 18 His grandiose view of self has been walled off from the necessary repeated exposure to limitation. While his grandiosity may be very apparent to others, the narcissist is often largely unaware of how truly infan- tilely grandiose he is. Still, any threat to his false self, which is essentially an expression of that grand- iosity, is experienced as a threat to his very existence." (p. 29) Johnson writes that there tends to be a "very profound" sex difference in the frequency of the "borderline" problem or symbiotic character in women and the narcissistic problem in men. These differences arise for two reasons. Because the young boy is both anatomically and chemically different from his mother he experiences a greater pull to separate and is thus more vulnerable to adapting a defensive style characterized by grandiosity. He has a need to feel invulnerable and powerful in order to separate and identify with his father and men in general. Further, the culture reinforces these adaptations by rewarding the male for his power, individuality, and rebelliousness. The girl, however, being the same sex as her caretaker, is more likely to be used by her caretaker as an object to gratify ongoing needs for a symbiotic-like attachment, and less likely to be reinforced for being powerful or rebellious. Johnson acknowledges that while narcissistic women and symbiotic men do exist, there is a profound cultural trend in the opposite direction. Johnson does not actually say there are two type: of narcissism, but rather that similar developmental failures are manifested by "the narcissistic character" in men and "the symbiotic character" in women. l9 O'Leary (1986) argues that there are two tynee of narcissism and that they are differentially represented by gender. He presents an argument for two distinct kinds of narcissism which are mediated by the extent to which grandiosity or shame is consciously experienced by the individual. In one type grandiosity is conscious and central, and shame is disavowed. The second type of individual shows a heightened sensitivity to shame, and grandiosity is not consciously present. O'Leary considers that males are over-represented in the first category and females cluster in the second. The grandiose type is further described as using projection as one of the major defenses, and often experiencing the therapist's empathic responses or interpretations as humiliating. O'Leary cautions that in working with this type of individual it is very important that the therapist continually communicate acceptance. The second type, in contrast, is able to use the therapist's observation that shame is present, acknowledge the shame, and respond to the therapist's attempts at empathic connection. O'Leary observes that females are most often the shame conscious individuals and draws on the literature about sex differences in relational capacity to explain this difference. According to this literature, relatedness is a more powerful force in the lives of women (Chodorow, 1978, Dinnerstein, 1976: Gilligan, 1982: Miller, 1976). Empirical 20 studies have shown women to associate achievement with violence while men associate affiliation with violence (Pollack and Gilligan, 1982). While this research has been challenged, it has also been defended and replicated (Benton, C.J., et al. (1983): Pollack, S., & Gilligan, C. (1983): Weiner, B., et al. (1983): Pollack, S., & Gilligan, C. (1983), Schwartz (1987)). 0'Leary concludes that because it has been argued and demonstrated that women value and pursue interpersonal connectedness and associate competition with violence, women are more sensitive to others and, therefore, others will be able to make women feel ashamed "because shame is an affect that is other connected" (p. 332). 0'Leary also draws on empirical work which demonstrates that women are more prone to experiencing shame (Gottschalk & Gleser, 1969 in 0'Leary). 0'Leary concludes that women will be less likely to reject an attempt for empathic connection from the therapist, or others, and less likely to defend with detachment. Likewise, because the culture encourages separateness and power and discourages dependency for the male, men will be more likely to defend with grandiosity and the disavowal of shame. Shulman (1986) suggests that there may be two distinct types of narcissistic patient, one who is described as suffering from developmental arrest (Kohut's model) and the other who is considered to be suffering from instinctual conflict (Kernberg's model). Unlike 0'Leary, he does not 21 consider either type to be gender specific. Shulman attempts to resolve the controversy between Kohut and Kernberg about the etiology and treatment of narcissism by suggesting that Kohut and Kernberg may have actually been writing about different populations. According to Shulman, the two types of narcissism can be differentiated from each other by (a) the patient's response to the analyst's empathic interventions, (b) the analyst's countertrans- ference in relation to being an "audience" to the narcissistic patient, and (c) the manner in which aggression is conveyed in the treatment situation. Shulman's Kohutian patient feels " . . . in perfect harmony, (and as if he and the therapist are) . . . perfectly tuned instruments" (p. 144), in response to empathic interventions. The therapist does net feel used, exploited and superfluous in the therapeutic relationship, and the patient's relationships with the therapist and others are nonaggressive. In contrast, the Kernbergian patient refuses empathic connection, protesting that any true understanding of him/her would expose his/her "badness, lack of concern for others . . . what I have been trying to cover up . . . I just wanted you to sit there, not bother me and not look too carefully" (p. 144). The therapist deee feel used, exploited, superfluous to the patient and impotent. The patient's aggression is extreme and a major focus of most relationships. 22 Shulman's two types of narcissist seem to differ importantly in their capacity for connecting with others. The Kohutian patient's relational capacity is much stronger. This author would suggest that this difference in capacity for connecting with others may be related to what both Johnson and 0'Leary have noticed and written about, and attributed to the female more than the male narcissist. Broucek (1982) makes an argument for two types of narcissism which includes the ideas of both 0'Leary and Shulman. Like 0'Leary, Broucek argues that the two types of narcissism are mediated by the extent to which grandiosity or shame is consciously experienced by the individual. Unlike 0'Leary, Broucek does not divide these two types of narcissism along a gender dimension. He makes no connection between relational capacity and shame proneness. Like Shulman, he considers that the shame conscious type resembles the individual which Kohut presents, while the grandiose type resembles the Kernbergian narcissist. Broucek calls the shame conscious type "disassociative" and the grandiose conscious type "unconflicted egotistical." Broucek, like Johnson, specifies that the narcissistic character is formed as a defensive adaptation to failures of maternal empathy during the rapprochement period. To elaborate on Broucek's view, he writes that at around 18-24 months the child develops the cognitive capacity for objective self awareness. At this point the 23 child is first able to take himself as an object - to see himself through the eyes of others and to experience his real separateness from the caretaker. An acute sense of smallness, weakness, dependency, etc. ensues if maternal functioning is not optimally soothing. In other words, if the mother does not convey to him a sense of his own strength, but rather is inconsistently responsive to his very urgent need to feel capable of this newly discovered separateness, he will not be ready emotionally for the cognitive realization of separateness. The shame affect which is elicited as a consequence of feeling one's separateness under these less than optimal circumstances may be defended against with grandiosity. Alternatively, shame may come to be experienced as the real self, and the defensive grandiosity, which Broucek considers to always be part of this experience, may be split off from awareness. Broucek thus considers that these two different solutions to the experience of acute shame become the basis for two types of narcissism which, again, he calls the "unconflicted egotistical type" (shame split off) and the "dissociative type" (grandiosity split off). Broucek states: The central sector of the personality can refuse recognition of either the grandiose self or the objectively derived 'actual' self (shame conscious) or alternately recognize both, but cannot recognize both simultaneously. When the central sector recognizes and embraces the grandiose self while disowning 'actual' self-—we see a type of narcissist who is unabashedly self-aggrandizing and is utterly shameless. This type, which I will tentatively designate the unconflicted 'egotistical' type, displays a seemingly total lack of tension between the grandiose self and the actual self. 24 The denial or disavowal of discrepancy between actual self and grandiose self is maintained by a selective inattention to all negatively toned critical reactions of others along with projection of already internalized negative self images. This type of narcissist has won a victory over shame, but at the price of impaired- interpersonal sensitivity and defective superego formation: he suffers from what Media referred to as 'that worst of all human diseases--the loss of shame.‘ When the central sector of the personality recognizes the actual self and disowns the grandiose self, the type of narcissistic disturbance resulting from the split might be best designated the dissociative type. In the dissociative type, the grandiose self is split off and when not behaviourally operative, is projected in the form of an idealized and omnipotent object. Broucek's two types of narcissist, again, are not different from O'Leary's. The only difference is that Broucek does not propose a relationship between gender and the particular style of narcissistic defense which is chosen. Alice Miller (1984) has proposed two dimensions of narcissistic personality disorder which she distinguishes from the narcissistic personality disorder described in DSM-III. Of all the theories presented above, Miller's is the only one to be empirically tested. The results of two studies testing her ideas (O'Brien, 1987, 1988) provide some evidence that these two dimensions exist, that they are discrete, and that they are measurable. Miller argues for a type of narcissistic personality which she calls the "narcissistically abused personality." This personality type is characterized as approval seeking, experiencing problems with belongingness and recognizing others' needs as being more important than one's own. While 25 this description, taken from O'Brien's work, is limited, it fits clearly into the shame conscious rather than the grandiose conscious category proposed by the authors above. The dimension she considers to reflect the DSM-III description of narcissism fits clearly into the proposed grandiose conscious category. O'Brien (1987, 1988) tested Miller's ideas with a scale he constructed, the O'Brien Multiphasic Narcissism Inventory (OMNI). He found that the "narcissistically abused" dimension (NAPD) appeared as a discrete, orthogonal factor in his factor analysis of scale scores. O'Brien initially tested a normal population, and later tested an outpatient sample of subjects diagnosed as Narcissistic Personality Disorder. Findings from these studies include: (1) congruence between the two populations on the discrete dimensions of narcissism revealed from a factor analysis of the scale, (2) a significantly higher mean score for the clinical sample; (3) a significant correlation between the dimension purported to measure narcissism as depicted in DSM-III (and DSM-III-R) and another widely used scale of narcissism based on DSM-III and DSM-III-R (NPI, Raskin, 1979): and (4) a slight negative correlation between the "narcissistically abused” dimension and this same scale (NPI, above). 26 These studies are of critical importance for several reasons. First, O'Brien's scale taps three dimensions of narcissism, finding that two dimensions do not correlate with a widely used Narcissistic Personality Disorder scale. Second, construct validation has been achieved using a clinic sample diagnosed Narcissistic Personality Disorder. Third, Miller's Narcissistically Abused Personality Disorder has characteristics which resemble the type of narcissism which 0'Leary has attributed to female more than male narcissistic personalities. (Her theory also resembles the theory of two authors described below.) O'Brien has not looked at sex differences in the endorsement of specific scale items of the OMNI. The OMNI will be discussed further in the following chapter on Research and Instruments. In summary, Johnson, 0'Leary, Shulman and Broucek all argue that there are two types of narcissism which are differentiated on the basis of conscious affect and style of defense. (See chart on page 29 for a summary of the various positions.) Johnson and 0'Leary divide two types along a gender dimension, arguing that gender differences in individuation and relational capacity account for these differences. 0'Leary and Broucek consider that their two types of narcissism are mediated by the extent to which grandiosity or shame is consciously present, 0'Leary con- cluding that women are more vulnerable to shame and thus to 27 the "shame conscious" pathology. Shulman and Broucek both consider that Kohut and Kernberg's different theoretical positions can be accounted for by the fact that they are actually looking at these two distinct types of individual, a grandiose disconnected and a more shame conscious, connected individual. Miller has proposed, and O'Brien has confirmed three orthogonal factors within a scale designed to measure narcissism. Miller's three dimensions are gender neutral, and the only ones to be tested empirically. 0'Leary, Broucek and Shulman's two types are quite similar. They are labeled Shame Conscious and Grandiose Conscious (0'Leary): Kohutian and Kernbergian (Shulman); and Dissociative and Unconflicted Egotistical (Broucek). Johnson's proposed two types are based more on differences in object relations (Symbiotic and Narcissistic). Two of Miller's three types (Narcissistically Abused and Narcissistic) resemble the shame and grandiose type discussed by 0'Leary and Broucek. Returning to our earlier focus, the question of interest for the current study is whether there ene gender differences in narcissism. Assuming that the two types of narcissist discussed above, by all five authors, can be divided according to gender, it might be proposed that the female narcissist is somehow more connected to others, although the specific nature of the connection ranges from "symbiotic like" to simply "empathically connected" and that 28 females experience more conscious shame while males experience more conscious grandiosity. The following section will outline theory which addresses how the unconscious affect (grandiosity for men and shame for women) is defended against. 1. Two Internal States RAE! Aflé£2fl§§§ TWO internal states 59!!! Transference Two internal states W W -self absorbed -uncomfortably pre- -unaware of others occupied with how -feeling of omni- he/she looks to potence as defense others against -opinions, needs and separation anxiety feelings of others' more important No speculation about favored states emphasis upon "alternating" states within the narcissistic individual No division according to gender HI . I E e e I: J' . -self feels important -others feel more powerful, strong important powerful strong No speculation about favored states emphasis upon "alternating" states within the narcissistic individual No division according to gender 2. Conscious Affect and Style of Defense lQEEQQ! Two types 9113131 Two types NezeieeieEic -failure to accommo- -failure to accommo- date around the date around the issue of ggendieeiey issue of sepezetion Differences due to gender differences in individuation and relational capacity £2§E_2££29_!§12 M2§£_2££2£_££2312 WM wane ' o s -defends w/projection -can acknowlg shame -interps experienced -connects empathi- as humiliating cally w/therapist Differences due to gender differences in individuation and relational capacity E2s§_efies_uele 5222.2ffen_zeuele Two types Two types EILLIB Two types 30 Unconflicted W -grandiosity cs. -shame conscious -Kernbergian type -Kohutian type -defends w/projection -defends w/proj. of split-off shame of split off grandiosity No speculation that differences are due individuation and relational capacity patterns No division according to gender -suffers from -suffers from instinctual conflict developmen'l arrest -refuses empathic -feels "in perfect connection w/therapist harmony w/therp." -aggressive and shaming -not shaming or of therapist aggressive w/therp. No speculation that differences are due to individuation differences No division according to gender N§I§i§81§§i§ N§221§§1§§i§§112 -grandiose -approval seeking -recognizes others' needs as more imp. -problems w/sense of belongingness No speculation that differences due to individuation differences No division according to gender *The left and right column in the chart above are arranged to emphasize the overlap between theories. Reading degn each column will provide a summary of each "type" of narcissism. The aggregate of symptoms will combine to form a theoretically consistent type. 31 4. SEX DIFFERENCES IN OBJECT RELATIONS PATTERNS OF THE NARCISSISTIC INDIVIDUAL As mentioned above, Johnson and 0'Leary both consider that there are two types of narcissism which are gender linked. These two types are differentiated on the basis of shame, grandiosity, defense and object relations. Several authors have written more specifically about the etiology and manifestation of sex differences in object relations patterns of the narcissistic individual. These authors include Illene Philipson (1985), and Annie Reich (1953). The ideas they present (below) resemble some of the ideas discussed above. Philipson (1985) argues that narcissism as a personality type and pathological disorder, describes a way of being that is primarily experienced by men. As evidence she points out that in three of the most prominent works on narcissism (Kernberg, 1975; Kohut, 1975, 1977), only 5 out of a total of 29 clinical presentations depict women. This is especially surprising in light of the common assertion within the feminist as well as the academic literature that women have higher rates of mental illness than men (Gove, 1980). While the evidence for this is mixed within a broader definition of mental illness, there is no evidence that men have higher rates of mental illness than women. Philipson summarizes the danger of assuming gender neutrality in the description of narcissism: 32 . . . all too often . . . supposed gender neutrality thinly masks a developmental theory that is truly only descriptive of boys. When this happens a number of outcomes can occur: "Homology theories," in which male development is considered the standard, are devised in order to include women (the female Oedipus complex is the best example of this): women can be seen both theoretically and clinically as "deviant cases, " wherein their needs, problems, and behaviors are judged against a theory that is implicitly premised on male experience: or women's particular developmental story can be rendered invisible, and their psychological conflicts and actions can be viewed through a theoretical prism that fundamentally misperceives their meaning. It is this latter outcome that seems to best describe the prevailing psychoanalytic theory of narcissistic development. (P. 219) Philipson's argument for a feminine brand of narcissism is based, as is that of 0'Leary and Johnson, on gender differences in the mother-son, mother-daughter relationship. Drawing on Kohut's work, Philipson reminds the reader that the seeds of narcissistic pathology are contained within an inconsistently responsive maternal attitude, or as Kohut labels it "faulty maternal empathy." Again, faulty empathy means the inability to respond to the child according to his or her unique needs as he/she is feeling them. This inability is in part due to projection of the characteristics of others onto the child and a response pattern which follows from the mother's unique relationship to that other person or persons. Examples might be the projection of characteristics of husband or father onto a male child and the projection of self onto a female child. What is crucial to this argument, as originally presented by Chodorow, is that male children will be perceived as 33 "others” by a female parent, and female children will perceived, at least in part, as a mother's extension of herself. Chodorow is talking about a general response pattern of female parents which may manifest itself to a greater or lesser extent, depending upon the degree of social isolation, the frustration level, and the character structure of the individual parent. According to Chodorow, even healthy mothers have different perceptions of and relational patterns with female and male children. The female will have a longer separation process than the male. Her sense of self develops in the context of a close relationship and, therefore, women have a greater tendency to define themselves, or find their identity, in relation to others. When maternal empathy is unpredictable, the child will be unable to internalize the mother's soothing and ”smiling" (p. 216) functions, and will, therefore, have difficulty establishing a sense of self worth and feelings of self esteem. The child, and ultimately the adult, will then only be able to feel valued when it is esteemed by and through external sources. Philipson argues that the styles of extracting external valuation will differ for males and females because of the differences in the mother-child relationship. She considers that a girl who has experienced faulty maternal empathy can gain self esteem by ee;1ng_ee_en exteneien of her mother. Philipson states: "She does not have to feel esteemed for herself, but as part of one who is 34 perceived as omnipotent" (p. 223). Philipson then argues that it becomes a common pattern of defense for narcissistic females to attach themselves to someone they perceive as omnipotent, and derive their own self valuation through their identification with this person. Philipson concludes: " . . . it seems fair to assert that for a female "narcissistic personality," male love partners can become part of the woman's self and replace the mother whose legacy to her daughter consists of ill-defined ego boundaries and insufficient autonomy . . . For male narcissists, however, female love partners tend not to become parts of the self: in fact, females are more likely to be used to admire and esteem men's grandiosity, while male partners can be constitutive of women's grandiosity. Women esteem men, and men are the vehicles through which they frequently attempt to find their self esteem." (p. 224) Philipson continues: . . . it is possible to claim that while both males and females may experience mothers' emotional inconsistency or faulty empathy, their characteristic ways of reacting to this are different. In most cases what psychoanalysts and social theorists describe and define as narcissism refers to the male's narcissistic defense. Certainly the low self esteem, the deficient psychic structure, and the deeply unconscious hunger for love that is at the root of the narcissistic dilemma is shared by both women and men. But because women are reared primarily by individuals of the same gender, they develop ego boundaries that allow or compel them to feel esteemed and loved through identification or "fusion" with others, prompting them to deal with this psychological impoverishment by acting as an extension of another. Because men erect rigid ego boundaries in order to establish their gender identities in opposition to the women who rear them, their manner of managing such impoverishment is different. Drawing on Philipson's argument then, it might be argued that female narcissists will "identify," "fuse," or "act as an extension" of another person in order to gain 35 self esteem and defend against conscious awareness of the enene that is characteristic of the narcissistic individual. Male narcissists will also have to draw their self esteem from external sources, but will do so in a different way. Philipson's characterization of the female narcissist as one who (1) does not feel valuable - or feels enene, and (2) who defends against shame by acting as an engeneien of the idealized other, is reminiscent of one type of, or one state within, narcissism, as mentioned by the authors above. She is reminiscent of Johnson's eynnieeie_ene;eeee;, who can e- .7 14. ,- s- . f run-2*.t- -_.tI-1 . a . -_ of 0'Leary's female who WW and is W. of Broucek's W time who W. of Shulman's thutian type who W Moist and WM £h§1§21§§, of Miller's NageissiSEieally Abused Tyne who WWW. of Bach's described state of anee§ively Self Awezeneee where ene s a ' s o o e s' o ' nt, and of Kohut's description of the 1deelizing_1reneferenee where the WW Annie Reich (1953) also writes specifically about the relationship patterns of female narcissists. Her arguments are very similar to those of Philipson. In both of the two relationship patterns common to female narcissists which she 36 discusses, the female places herself in what Reich terms a "subservient” position vis a vis the male love object. The first type of relationship Reich discusses is characterized by "primitive idealization" of the male love object by the female. The woman feels she is unable to live without the man, unable to acknowledge any of his less than perfect characteristics, and willing to make any and all sacrifices to keep him. (In other words, she projects a grandiose or idealized image on to the man which is a carryover from an earlier developmental period in which Mother was experienced as this perfect, powerful, beautiful other.) Reich considers that the function of the male partner in this relationship dyad is to serve as the embodiment of the female's split off grandiose self. Her unity with the real object provides an inner sense of wholeness. The male partner is experienced as providing for the female on the outside what she cannot provide for herself independently. It is the striving for self cohesion that drives the attachment and fuels the intensity of the connection. The second type of relationship dyad described is also characterized by a tendency toward overvaluation of objects. In contrast to the first type described above, however, these relationships are short lived. Male partners are idealized for a short time, until some flaw is discovered. They are then quickly devalued, only to be replaced by 37 another partner with whom the female narcissist can then repeat the cycle. During the period of idealization, Reich describes the quality of the attachment as follows: These women "fall in love" with men whom they "deify" and without whom they consider life unbearable. They take over the man's personality, interests and values completely: it is as if they had no judgment of their own, no ego of their own. Reich draws on Deutsch's description of the ”as if" personality as an example of this type of individual. Deutsch describes these persons as incapable of loving anybody, and as able to relate to external objects only via a primitive form of identification. Reich considers that in both of these relationship dyads what the woman is looking for is her own power - her own grandiosity. The male lover is dropped if he fails to bring the wished for power that association with him promised. A better object is then sought. Here is it,important to note once again the considerable overlap between theories. In particular, the issue of the defensive aspects of object relations patterns are highlighted by Philipson, Johnson and Reich. Specifically, they write that female narcissists defend against realizing their separateness and vulnerability by "acting as an extension of", "fusing," "identifying,” (Philipson), "symbiosing" (Johnson) or relating through a form of "primitive idealization" (Reich). Female narcis- sists experience their grandiosity vicariously through men. 38 Male narcissists defend with grandiosity and project their shame into the women they are close to. Summer! In sections 2, 3 and 4 above, nine authors have been discussed. Five argue for two types of narcissism which are based on differences in shame and grandiosity, two for two contradictory states within the narcissistic individual, and two for sex differences in object relations patterns of the narcissistic individual. Of these nine authors, four state specifically that there are sex differences in conscious affect, defensive style and object relations patterns. Taken together, these four suggest that (1) females are more shame conscious, (2) females are more likely to project their grandiosity into a consequently idealized other: (3) females are more likely to be connected to others in an intense way, i.e., "fused," "symbiosed," "identifying," "acting as an extension." Males, by contrast, are (1) more likely to be grandiose, (2) more likely to project their split-off shame into others: and (3) less likely to be empathically connected to others. The three authors who argue for two types of narcissism without speculating that a gender difference exists (Broucek, Shulman and Miller), nonetheless describe types of narcissism which are consistent with the gender-linked types of narcissism proposed by others. In addition, the two 39 states of consciousness which alternate within the individual, described by Bach and Kohut, are also consistent with the two gender-linked types. Again, this supports the notion that there may be a gender-based, favored state. This review, therefore, supports the conclusion that many clinicians have noticed discrete types of narcissism, and that an exploration for a gender link is warranted. Further, it supports the likelihood that crucial aspects of narcissistic experience (female experience that is) may have been either overlooked, as Philipson suggests, or misdiagnosed. The issue of misdiagnosis, an already significant problem, is presented below. 5. BORDERLINE VS. NARCISSISTIC PATTERNS Are these "fusing," "shame conscious," "identifying," "symbiosing," "primitive idealizing" women borderline - and these "grandiose," "shame projecting," "non-connecting" men nezeieeietie? This is the conclusion that Masterson, Haaken and Johnson suggest. To consider this possibility, it is important to review how these diagnostic terms are used. Toward this end, Masterson and Kohut's conceptualizations of borderline pathology are summarized below. Masterson considers that the borderline individual experiences an abandonment depression whenever he or she takes any step toward autonomous functioning. The etiology of this dynamic is that the mother, e la and 40 eeneieeenELy, abandoned the child emotionally whenever the child behaved autonomously. At the same time, the child was rewarded with love whenever he or she clung to Mom. This pattern leads to a style wherein the borderline individual becomes fearful and anxious whenever he/she acts autono- mously because of abandonment fear, and a characteristic form of clinging behavior which is an unconscious attempt to gain approval. Masterson differentiates between the border— line and the narcissist by arguing that the borderline can- not separate enough to love and work without experiencing an abandonment depression. The narcissist, by contrast, is not aware of his/her real separateness, because he/she has not advanced beyond the Practicing period (Mahler) wherein the child and the caretaker were perceived as one person. Thus, believing themselves to be powerful as they run on this "borrowed" power, the narcissist can function quite well. The puzzle for Masterson is that the borderline appears to be more advanced developmentally in that he/she is aware of his/her separateness and vulnerability, whereas the narcis- sist is not. Yet, at the same time that the narcissist is more developmentally advanced, he/she cannot function as well. Masterson considers that women are more likely to be borderline, and men more likely to be narcissistic. Johnson likens the female "symbiotic character" to Masterson's borderline personality because the "symbiotic 41 character . . . has not resolved the issue of separateness" and ". . . can feel or know herself only in relation to another ...” (p. 28). As I will argue below, I consider that there are many problems with this comparison. These problems include no distinction between "using others as self objects" (i.e., persons used as containers for split-off aspects of the self) (Kohut, 1971, 1977, 1981), and being "symbiotic" with others: in other words, no attempt to differentiate rela- tional capacity across types of pathology in this argument. Johnson concludes that because women use their relationships with men to complete themselves, they are borderline. But even a cursory reading of psychoanalytic theory of narcissism reveals that the narcissist is by definition dependent upon others to complete the self. As detailed above, Kohut, Bach and Broucek all consider that the narcissist feels either shame or grandiosity, but cannot feel both at the same time, and thus is not integrated internally. They consider that self-objects (persons) are sought to 'project' the unconscious part of the self into. When joined with the self-object, then, the narcissist feels whole. These self-object connections are sought and main- tained with life and death fervor because the loss of the self-object causes 'fragmentation of the self," i.e., feels like a complete and devastating loss of the self (Kohut, 1971, 1977, 1981). 42 Kohut's conceptualization of the etiology and manifestation of borderline pathology is quite different from Masterson's. Kohut considers that the origin of the borderline problem is a maternal attitude characterized by a ”flat response," to the child and a leek of empathy. The consequence for the child and later adult is a "basic hollow core" which is surrounded by defenses. For Kohut the borderline's adaptation to treatment is to "borrow the whole of the analysts personality organization in order to survive." Masterson and Kohut clearly disagree on the pattern of the maternal response, and the 'self' which emerges as a consequence. For Masterson, the self is thwarted from individuating. For Kohut there is no self. For Masterson, the treatment of choice is confrontation of the constella- ‘ tion of affects that emerge in response to efforts to individuate, with the goal that the patient be able to recognize and conquer these affects rather than being controlled by them. For Kohut, there can be no interpretive treatment. Because Kohut's borderline has no self, the therapist's role is simply to support. To remove the defenses would be to expose a hollow core, and perhaps elicit psychosis. Once these theories are examined at close range several issues emerge. (1) The "borderline personality" is by no means a unitary construct. Any reference to it, therefore, 43 should be qualified according to specific theorist. (2) Even within the inconsistencies, the borderline individual's symptoms are generally considered more eegege than the narcissists. The borderline, which all too often may mean the fenele, is generally considered to be more impaired than the narcissist (male). The political implications of this will not be discussed in the current research except insofar as to mention here that the potential effects of such a generally-held misconception are more harmful to the female than to the male. Returning once again to the literature, Haaken (1983) states directly that problems in the early female individ- uation process are more likely to produce borderline personality disorders, and disturbances in the early male individuation process are more likely to produce narcissistic personality disorders. To explain sex differences in individuation, Haaken cites Lachman (1982) who cites Barglow and Schaefer's (1976) argument that the innate aggressiveness of males promotes earlier individuation, and perhaps premature individuation, by adding physical frustration to the mother-son dyad. She cites Lewis (1976) who argues that while some innate sex differences may contribute to differences in the individua- tion process, it is social and cultural forces which are most crucial in shaping individuation and relational capacity differences. Most of her arguments, however, are 44 based on Chodorow's theories of differential gender socialization, summarized above, including the preoedipal and oedipal periods as well as the long period of adulthood during which the culture may reinforce existing personality patterns. She concludes that in talking about sex differences in separation pathology, several possibilities arise as to their etiology. She argues that it may be that similar forms of maternal care have different effects on the two sexes, that maternal ambivalence may be expressed differently with daughters v. sons, and/or that later socialization experiences impact on early experience to further either cohesion or fragmentation of the self. In reading her arguments it seems reasonable that all three may be true. It does not follow to this author, however, that women should be given a different, more pejorative diagnosis, or that the differences in adaptation to failures of maternal empathy during the same develop- mental period should result in differences in the sevegity of the ensuing pathology. Several issues arise when arguments for more severe pathology within women are examined at close range. Haaken and Johnson's argument that women are more prone to develop- ing borderline pathology is made on the basis that women may have weaker ego boundaries as a consequence of the prolonged "symbiosis" or closeness with their caretaker during child- hood. Because women form their identity in the context of a 45 close, same-sex relationship with their caretaker, it is argued that they are considered to be vulnerable to "knowing themselves only in relation to others." There are several things wrong with this argument. First, there is some evidence, though from only one study, that closeness to others is correlated with high individuation. Specifically, in one study using a sample of normal subjects, women were both more relationally oriented and more highly individuated than men (Schwartz, 1987). This finding is in the opposite direction of what would be predicted if prolonged closeness actually encouraged identity diffusion. This empirical work supports arguments posed by Jean Baker Miller and others that closeness to others enhances ones sense of self and facilitates individuation. To this author there is a critical and little-cited difference which seems pivotal in explaining the specific style of relatedness which might be attributed to the female narcissist. Specifically, it seems that because the female child is in a prolonged "symbiosis" with a EQIQifiélELiQ parent that she learns to W by acting as an engeneien to the significant other. It is when one combines theory of differential gender socialization with theory of individuation pathology that one can postulate a unique form of individuation pathology as a consequence of prolonged attachment to a nereissiseie 46 parent. Thus, it is not prolonged attachment itself which is responsible for a unique pattern of narcissistic develop- ment. Still, this is an argument for a different style of "connecting" for the female narcissist. This does not make the female "borderline.” The second problem with the argument that women may be more prone to borderline pathology is that relational styles which are, at close range, quite distinct are compared and equated at times in the literature. For instance, can we determine if there is a difference between borderline pathology, as defined by Johnson (above), and relating to others as "self objects," as discussed by Kohut? The object relations of the narcissistic individual are characterized by an inability to relate to others as whole persons with separate feelings, motives and needs. Instead, the narcissistic person "uses” others to satisfy needs of the self which cannot be satisfied independently. Thus, the female narcissist may use a male love object as a self- object to satisfy her needs for power. Through her rela- tionship with him, she is enabled to feel powerful. Stated another way, her split-off grandiosity is projected onto him. She can then only feel whole when attached to him. The male narcissist, conversely, in Kohutian terms, is more likely to use the female love object as a "self object" to reflect his power back to him, and to contain his split-off feelings of inadequacy. According to Johnson or Masterson, 47 the female dynamic may be conceptualized as the woman being ”symbiotically” attached to the male, and thus being ”borderline", whereas the male narcissist is simply using the female to contain split-off aspects of the self. In fact, it seems quite likely that the two styles may leek different, but may represent similar levels of developmental arrest. The distinction between borderline and narcissistic object relations patterns is not easy to make. It does not appear, however, that these relational patterns can be differentiated on the basis of using others to complete the self. It does not seem valid to consider that a female who uses others to complete her sense of self is borderline, whereas the male who does the same is narcissistic. Further, from the preceding review of theory on proposed sex differences in object relations patterns of the narcissistic individual, it appears likely that a number of distinct styles of relating to objects may be present. While Reich's proposed style of projecting split-off grandiosity into the male is consistent with the dominant theories of narcis- sistic object relations, it is less clear how Johnson's proposal that a woman "can feel or know herself only in relation to others" fits into this same theory. Philipson's argument that the woman extracts external valuation by acting as an extension of a significant other eeeeuse she was valued by her narcissistic mother only when she complied 48 with the mother's demand that she serve as an extension, makes the most sense within the framework of theory on narcissism. Reich's other suggestion that narcissistic women relate through "primitive identification" is another dynamic which does not fit theory pertaining to object relations patterns of the narcissistic character. From my reading of Kohut, I would speculate that he would call this a borderline dynamic, revealing the absence of any true self. The problem of course is that there are no clear operational definitions or consistent usages of many of these terms or constructs. One cannot clearly distinguish between "knowing oneself only in relation to others" and feeling whole only when united with the self object containing split off fragments of the self. To begin to distinguish these patterns seems the most appropriate and meaningful task for the future. To actually label them borderline or narcissistic seems premature in light of the considerable confusion around the usage of the terms. Further, as mentioned above, it is important to recognize the political implications of such labeling, i.e., that it leads to women being diagnosed in a manner which suggests that they are more impaired. It may also lead to different treatment decisions including resignation to supportive therapy only, if one follows Kohut's recommen- dations. This is a significant problem. 49 My final argument against speculating that women belong in the borderline camp while men belong in the narcissistic camp is drawn from another author's observation that the usage of these diagnostic terms, in the absence of a unitary conceptual framework, contributes to confusion and is, in some ways, utterly meaningless. Meissner (1976) comes to the following conclusion in his review of Kernberg and Kohut: Kernberg writes as though the diagnostic spectrum stretching between the psychoses and the neuroses were filled by nothing but forms of borderline personality. At the same time, Kohut writes as though the same diagnostic vacuum were filled by nothing but forms of narcissistic pathology. . . . The upshot of the ambiguities in these approaches is that there is considerable diagnostic confusion and difficulty in making appropriate discriminations between these various forms of psychopathology (p. 405). W To conclude the review of the literature on theory, and to provide a clinical example of the confusion mentioned above, Lachmann's case presentation detailed below provides an example of female narcissistic behavior which could be interpreted by the authors presented above as reflective of many different psychological processes. Lachmann's interpretation of the behavior, however, provides an interesting and logical explanation which does not hinge on sex differences in severity of individuation pathology. It is based on a female narcissist's use of a male love partner to further an idealized self perception. 50 To begin with, Lachmann (1982) cautions the reader that among the various studies of female psychological develop- ment, the role of narcissism has been subject to the most frequent and extensive revisions. He cites Freud and Deutsch's anatomical damage arguments to provide examples of the radical theoretical revisions which have taken place. Lachmann cites Stolorow's (1982) functional definition of narcissism as gender neutral: "Mental activity is narcissistic to the degree that its function is to maintain the structural cohesiveness, temporal stability, and positive affective coloring of the self representation" (1982, p. 48). Using this definition, Lachmann argues that there can be no sex differences in narcissistic mental activity per se. He argues, quite logically, however, that there can be sex differences in the content and elaboration of the self- representation and in defensive or compensatory styles. He states: "Any mental act has, among its multiple functions a narcissistic one--the maintenance or restoration of vulnerable aspects of the self-representation. This self-structure is neither masculine nor feminine, though we would postulate a core gender identity to which masculine qualities or feminine qualities accrue as its contents. Indeed, examples of the narcissistic function can be found in those activities whose specific purpose is to shore up a precarious sense of gender identity." Lachmann challenges Deutsch's argument that women are by nature, and as a consequence of their discovery that they have no penis, masochistic. He argues that what may look masochistic may actually function to further eeneeien of the self rather than fragmentation. 51 Lachmann cites the case history of a 30-year-old woman, Judith, who attempts to gratify her unmet dependency needs by caring for men. The defect in Judith's self structure was most evident in her inability to be alone and in her excessive submissiveness toward men. When Judith did not have an ongoing relationship with a man, she would become depressed and experience quasi-hallucinatory experiences in which the corners of her bedroom would become elongated. This was interpreted as a distortion derived from a childhood fantasy of living in an orphanage where her parents had threatened to place her as punishment for bad behavior. In her relationships with men Judith took a self- sacrificial stance, nurtured them as she wished she would have been nurtured, and obtained vicarious gratification through identification with the cared for man. According to Lachmann, this behavior would have been interpreted as masochistic by Deutsch, and as part of Judith's nature as a woman. Lachmann argues that Judith's behavior was a consequence of inadequate structuralization of the self. Through her masochistic submissiveness she maintained and consolidated a self-representation which was crucially different from the representation of her mother (who was not nurturing). "Self esteem was thus vulnerably established but dependent upon and maintained by the utilization of needy men whose presence reassured her that she was "satisfying" and not "needy," self sacrificing and not "vain," and nurturing and not dependent (p. 52)." 52 Once again, the potential for different interpretations of this dynamic is probable. Lachmann's patient's behavior might be interpreted as demonstrating the use of the eynnieELe defense which Johnson talks about, which Haaken and Masterson might call a bezdexline personality dynamic, which Philipson might call attempting to get external valuation by acting as an ex;eneien_of the male partner, and what Kohut might call, in part, a developmentally normal need to establish an ideelieing_ezenefiezenee. According to Lachmann, she was attempting to further cohesion of the self by using a relationship with a man to project her dependency needs, and to reflect back to her an idealized image of herself as satisfying and nurturing. Lachmann argues that narcissistic activity cannot be masochistic, or used to harm the self. By definition, narcissistic activity is engineered to serve an integrating, bolstering function. Judith‘s self sacrificial stance toward her partner served to bolster an idealized image of herself. So, while her behavior may appear on the outside to be a manifestation of shame, Lachmann would consider it a manifestation of grandiosity (the ideal self). This argument has profound implications when considered alongside both the clinical observations discussed above (women more often carry the persona of shame), and with theory concerning the acculturation of femininity (what does the culture define as perfect femininity?) Putting these 53 two together leads to the question, if being strong in a female sense has come to mean supporting others at the expense of the self, does performing that perfectly mean complete self denial? Karen Horney's (1950) theory of neurosis and the "Search for Glory" provides an interesting perspective on this question, while not addressing it directly. While she does not discuss female and male grandiose styles as such, she talks about different styles of relating to others which emerge from a "false self" position, and which are motivated by the need to create a "perfect" "glorious" or ”idealized" self image. Two of these styles are reminiscent of the male and female types of narcissism discussed above. While one style emulates a shame persona and the other a power persona, both are presented as emerging from the grandiose position by Horney. According to Horney, inadequate reflection and support of the self during childhood leads to feelings of not belonging, or being inferior, and of what she calls basic anxiety. As more and more energy becomes funneled into defending against this anxiety, the natural unfolding of a healthy self is arrested. As energy is diverted for the creation and maintenance of these defenses, alienation from the real self progresses until the subjective feeling of 'self' becomes conceptual and disconnected rather than spontaneous and congruent. Maintaining the conceptual false 54 self comes to require a compulsive, rigid style of relating to others. Feelings of exclusion are transferred into feelings of elevation and superiority over others. The ultimate personal goal is transformed into what Horney calls "The Search for Glory" or the creation and maintenance of a conceptually perfect self, rather than the continual growth and unfolding of the real self. Horney identifies three basic styles of relating to others out of this rigid and conceptual self: movement toward, away from and against others. She describes movement toward others as developing to an extreme of clinging, and against to an extreme of purposely hurting others. The individual who moves toward others is described in the extreme as ”compliant, self effacing, suffering, and devoted", while the individual who moves against is portrayed as "exploitative, ambitious, expansive and domineering." Both styles contain grandiosity. Perfection and glory can be achieved through compliance, suffering and devotion as easily as through ambition and domination. The devoted individual may liken him/herself to a Saint, while the ambitious one may imagine him/herself a conquering hero. Again, combining Lachmann's conceptualization of Judith as one who glorifies her image by assuming the position of selflessness, and Horney's portrait of the Search for Glory through devotion, an image emerges of a culturally supported feminine style which seems to contain both a shame persona 55 (self effacing, suffering, devoted) and grandiosity ("Only a Saint could suffer as I do.") simultaneously. 6. SUMMARY OF LITERATURE REVIEW ON DIFFERENCES IN NARCISSISM In this section I have reviewed five perspectives on the argument that there are two types of narcissistic pathology (Johnson, 1987: 0'Leary, 1986: Broucek, 1982: Shulman, 1986: Miller, 1981, 1984, 1985, in O'Brien, 1988). Of these five, two authors suggest that these differences are gender linked (Johnson, 0'Leary). I have also reviewed two major theoretical positions on the etiology, phenomenol- ogy and transference manifestation of the narcissistic character structure (Bach, 1976: Kohut, 1971, 1977). The two types of narcissism which some consider to be gender linked were compared to the two states of consciousness ex- perienced by the narcissistic individual (See chart, p. 29). It was suggested that there may be a sex difference in the extent to which a given individual experiences one or the other of these two contradictory states of awareness. It was also suggested that there may be considerable overlap between the ideas presented by the seven authors discussed. Second, several ideas were presented concerning gender differences in object relations patterns of the narcissist (Philipson, 1985: Reich, 1953: Lachmann, 1982). It was suggested that (1) female narcissists may seek external validation by acting as an QEEEESIQD of an idealized other; (2) female narcissists 2r2iect_their_§plit:off_srandio§itx 56 onto their nele partners (and thus may be more likely to seek external validation from them), while male narcissists WW into their female partners (Philipson): (3) narcissistic women form ezin1e12e ideneifieeeiene with idealized others because they lack a real self (Reich): (4) female narcissists, like male narcissists, use others to further cohesion of their self structure. The specific form of defense and compensation follows from gender role determined aspects of their identity (Lachmann). Lastly, the argument that females are more likely to manifest borderline disorders, while men are more likely to manifest narcissistic disorders was reviewed (Johnson, 1987: Masterson, 1976; Haaken, 1983). These arguments were based on proposed differences in individuation which are said to leave women with "weaker ego boundaries" as a consequence of prolonged symbiosis with the caretaker. These arguments were refuted on several grounds including (1) prolonged "closeness" to the caretaker has not been shown to result in "weaker ego boundaries," (2) female children of narcissistic mothers may develop a unique narcissistic style of relating to idealized others as an "extension" of themselves, but this does not make them "borderline," (3) the language used to describe borderline relational patterns is vague, and not clearly different than language used to describe narcissistic relational patterns, (4) the definition of 57 bendenline and nezeieeieELe disorders is inconsistent within the literature and (5) the borderline diagnosis generally carries with it (a) inconsistent, but often negative conclusions about prognosis, and (b) a rather pejorative connotation. From the review of the literature on theory several issues arose which will be the subject of the current investigation. These are: 1. Do female narcissists feel more conscious shame, while male narcissists feel more conscious grandiosity? 2. Do female narcissists have more unconscious grandiosity, while male narcissists have more unconscious shame? 3. Is there a distinct quality of object relations functioning in males v. females who are diagnosed narcissistic personality disorder? Specifically, do females project grandiosity into male partners while males project shame into female partners? Do female narcissists feel themselves to be "extensions" of their male partners? 4. How is narcissism diagnosed? If there is indeed a female narcissistic style which has not been written about, is there a gender bias in the instruments currently used to measure narcissism? These, again, are the questions which were explored in the current study. Before looking at specific hypotheses, 58 however, it is important to review current methods used in the measurement of narcissism. This will be done in order to explore the interaction between inconsistencies in definition, inconsistencies in measurement, and speculations about sex differences in prevalence of, as well as distinct patterns within, the disorder. 59 RESEARCH AND INSTRUMENTS Because the purpose of the current study was to explore sex differences in narcissism, it was first necessary to identify a narcissistic population and then to test this population for differences. Toward this end, current methods or scales used to identify narcissistic individuals were reviewed. The review of these scales and the research in which they were used reveals several things. First, the scales are very different from each other. Thus, the differences in the definitions of narcissism, mentioned above, are represented in the scales. Second, the individual scale biases may be seen as contributing to the perception that male narcissists outnumber female narcis- sists. The most widely used scale seems to measure mostly the grandiose defense. I have reviewed most of the scales used to measure narcissism below. Some will be used in the current study and others are included to document the history of empirical work on narcissism. Also documented is the confusion which continues to surround the use of the term. Over the last 10 years significant attention has been focused on narcissism as a clinical phenomenon. Around the time of the inclusion of the Narcissistic Personality Disorder in the third edition of the Diagnostic and Statistical Manual of the American Psychiatric Association (APA, 1980), several attempts were made to establish an 60 instrument to measure narcissism. To date, these instru- ments include, but are not limited to, the Narcissistic Personality Inventory (NPI) (Raskin, 1979): a scale for use with the Thematic Apperception Test (TAT), (NP) (Shulman and McCarthy, 1986): the Millon Multiaxial Clinical Inventory (Millon, 1980): a scoring criterion for use with the Rorschach (Exner, 1969: Harder, 1977: Urist, 1977): an MMPI Narcissistic Personality Disorder Scale (Ashby, Lee and Duke, 1979, reported by Emmons, 1987) and the O'Brien Multiphasic Narcissism Inventory (OMNI), (O'Brien, 1987). The NPI has been used the most extensively for research purposes. Narcissistic Personality Inventory - s o ' n Because the NPI has been widely used, I will review the empirical work which has been carried out using the NPI below. Before doing so, however, is important to look carefully at the construction of the instrument and the clinical symptoms upon which it is based. The NPI was developed in direct response to inclusion in the DSM-III of the Narcissistic Personality Disorder, and so was constructed to reflect narcissism as depicted therein. Therefore, much of the empirical work that has been done on narcissism defines narcissism in accordance with the DSM-III criteria. While Goldstein (1985) considers these criteria to be a fair representation of both Kernberg 61 and Kohut's work, Akhtar and Thompson (1982) have written that the DSM-III criteria leave out the central feature of the condition, namely: "the co-existence of mutually contra- dictory stances seen in almost all areas of functioning." (Akhtar and Thompson, p. 17). By "mutually contradictory stances" Akhtar and Thompson are referring to the two diametrically opposed self states: one characterized as grandiose and the other as shame ridden. This author would point out further that, in addition to leaving out the feature of mutually contradictory stances, what is mostly emphasized in these criteria are the symptoms of the grandiose defense. The DSM-III criteria are printed below in their entirety. The following traits must be characteristic of the individual's current and long term functioning: A. Grandiose sense of self-importance or uniqueness, e.g., exaggeration of achievements and talents, focus on the special nature of one's problem. B. Preoccupation with fantasies of unlimited success, power, brilliance, beauty, or ideal love. C. Exhibitionism: The person requires constant attention and admiration. D. Cool indifference or marked feelings of rage, inferiority... W. or emptiness in response to criticism, indifference of others, or defeat. In addition: At least two of the following are characteristic of disturbances in interpersonal relationships: 62 (1) entitlement: expectation of special favors without assuming reciprocal responsibilities: e.g., surprise and anger that people will not do what is wanted: (2) interpersonal exploitativeness: taking advantage of others to indulge own desires or for self-aggrandizement: disregard for the personal integrity and rights of others: (3) relationships that characteristically alternate between the extremes of e_e;ideelineeien and devaluation: (4) lack of empathy: inability to recognize how others feel: e.g., unable to appreciate the distress of someone who is seriously ill." Only three words (underlined) included in these criteria are representative of the underlying feelings of inferiority. It is this author's observation that none of these words or the state of consciousness they represent, are discussed in depth in work done with the NPI. This represents an important limitation of the scale which will be discussed further below. The scale itself has two versions, a 40-item and a 54- item format. The 40-item version was developed after factor analytic studies determined which of the items in the original 54-item scale accounted for most of the variance within each of four factors. The individual items consist of two dichotomous statements. The subject must endorse one. Each item consists of a narcissistic and a non- narcissistic statement. For example: 1. v at a e fo ' uenc'n e le. B. I am not good at influencing people. 63 The scale score is the total number of narcissistic statements endorsed. Emmons' (1984) factor analysis of the NPI revealed four factors: Exploitativeness/Entitlement, Leadership/ Authority, Superiority/Arrogance, and Self-absorption/Self- admiration. This factor analysis was later replicated (Emmons, 1987). The NPI scale has been used on college students and a clinic sample. Significantly different means were obtained between the two samples. Raskin considers that his studies with college students are measuring Narcissistic Personality Iliil rather than 215929211 Prifitera and Ryan (1984) found that the NPI distinguished between narcissistic and non- narcissistic psychiatric patients in a clinic sample. 3W Two studies have tested and supported the construct validity of the NPI (Prifitera, 1984: Emmons, 1987). Eight- week alternate form reliability of the scale has also been demonstrated (Raskin and Hall, 1981). A number of studies have shown a significant relationship between NPI scores and other constructs (Raskin, 1980; Emmons, 1981: Raskin & Hall, 1981: Raskin, 1981, Emmons, 1984: Watson, 1984). One study investigated the relationship between NPI scores and basic dimensions of personality (Emmons, 1984). Sex differences in narcissism have been demonstrated in one study using the NPI (Emmons, 1984:). All of these studies are presented in detail below. 64 Seale_§orrelations NPI/MCMI Prifitera (1984) found that the total NPI score correlated significantly with the Millon Clinical Multiaxial Inventory (MCMI). Emmons (1987) found that only one factor of the NPI (Exploitativeness/Entitlement) correlated significantly with the MCMI. NPI/NPD Watson (1984) found no relationship between the NPI and the Narcissistic Personality Disorder subscale of the MMPI (NPD) (Soloman, 1982). Emmons (1987) reinvestigated the relationship between the NPD and the NPI, after factor analyzing the NPI, and found a significant positive correlation between the NPD and one factor (above) of the NPI, again, Exploitativeness/ Entitlement. This one factor, then, is the one most representative of pathological narcissism as measured by the NPD and, according to one study, the MCMI. NPI/SELFIBM scans Emmons found a significant positive correlation between the total score on the NPI and the Selfism Scale (Phares and Erskine, 1984). NPI/arse Emmons' most interesting finding in the study above was a correlation between the Exploitativeness/ Entitlement factor of the NPI and a subscale of the Self-Focus Sentence 65 Completion Test (SFSC) (Exner, 1973) called Negative Self Focus. This is the only study which investigates construct validity using a projective test. This finding is interest- ing because it juxtaposes a conscious, behavioral manifesta- tion of narcissism (exploitativeness/entitlement) with the underlying, unconscious self perception (negative self focus). In other words, both the vulnerability and the defense against it are revealed at the same time by using an objective and a projective test simultaneously. - ' w' 0 V Narcissism, as measured by the NPI, has been found to correlate significantly with egeeeiziex (Raskin, 1980) as measured by the Barron Symbolic Equivalents Test: with eenee;1en_eee§ing (Emmons, 1981) as measured by the Sensation Seeking Scale: with the gzexegeneien and Regenegieien subscales of the Eysenck Personality Questionnaire (Raskin and Hall, 1981). NPI scores have also been found to correlate with the use e1 fins; negsen ginge- le;_n£enenne (Raskin, 1981, reported by Emmons, 1987); and with peer ratings of narcissism (Emmons, 1984). A negeeiye correlation was found between the NPI and three empathy scales (Watson, et. al., 1984): and with intrinsic neligious xelnee as measured by the Allport and Ross Religious Orientation measure (Watson, et al, 1984). Emmons (1984) correlated NPI scores with basic dimensions of personality as measured by four prominent 66 personality inventories. He found that NPI scores correlated positively with W -°' ~ '9 _; "Se—=12: sT ,-'_'_" ° 2° —;3°,° -_ 1712?; A- 395.1!h121fi12nign: and negatively with abasement, deference and social anxiety. A look at the words underlined above (creativity, sensation seeking, extroversion, psychoticism, dominance, independence, self-esteem, self-monitoring, exploitativeness and exhibitionism) reveals that narcissism, as measured by the NPI, has so far been correlated with variables which are most representative of the grandiose self state, or the grandiose pole of the “mutually contradictory stances" found in the narcissistic disorder. The scale is negatively correlated with the shame/inadequacy pole of the disorder. Emmons comments that these results suggest that narcissism, as measured by the NPI, seems to reflect many traits which seem to be adaptive. Here again, he is referring to narcissism as a trait as opposed to a personality disorder. He further suggests that there may be a curvilinear relationship between narcissism and health. Adaptiveness aside, it should be emphasized that narcissism as defined by Kohut and Kernberg is a pathological as well as a painful state. Further, when narcissism is defined as the adaptation of a false self to cope with the perceived demands of reality while leaving the real self covered over (Kohut, 1971, 1977, 1981, Johnson, 67 1987, Miller, in Morrison, ed., 1986) it is clear that there could be no curvilinear relationship between this state of consciousness and health. The confusion over whether or not some narcissism is healthy is furthered by the fact that studies using the NPI, as mentioned above, have not attended to the vulnerability, weakness and fear of the narcissistic personality. This again is probably because the DSM-III and DSM-III-R criteria do not emphasize the vulnerability of the narcissistic personality. H21 3 E E'EE I n . i Sex differences in narcissism, as measured by the NPI, were found by Watson (1984). Watson (1984) explored Akhtar and Thomson's (1982) hypothesis that males are likely to be more narcissistic than females. Using the NPI, Watson did find significantly higher NPI scores for males in a college student sample. _ S _ 't . _ , . s' , A look at the DSM-III and DSM-III-R criteria for narcissism at close range reveals the missing pieces in this description of the narcissistic state. As mentioned above, Goldstein (1975) has compared and contrasted the theories of Kohut and Kernberg and how the two are represented in DSM- III. He concludes that DSM-III is an accurate representa- tion of narcissism as described by both Kohut and Kernberg. A careful review of Goldstein's argument, however, indicates that while it is true that there is nothing in DSM-III that 68 is not mentioned by Kohut and Xernberg, it is also true that both Kohut and Kernberg emphasize symptomatology which is not in DSMIII. These criteria are: Kohut: Chief complaints which are vague and ill defined -1n_the_§eml_spher_ei perverse fantasies or lack of sexual interest. -In_the_sp_oial_sphene_z. work inhibitions. inability to form and maintain significant relationships, delinquent activities -In_th.e_man1fest_pers.onal1m lack of humor . . . lack of sense of proportion. . . pathological lying -In_1hep_exohpsomatie_§pherez hypochondriacal preoccupations Kernberg:-A search for gratifications confirming one's grandiosity -No enjoyment in life except through praise from others plus one's own grandiose fantasies -Idealization of those who give narcissistic supplies, and devaluation of those who do not (DSMIII just says relationship alternating between the extremes of idealization and devaluation - does not say what about the objects elicits these differential responses.) -Aternation of feelings of inferiority and insecurity with those of grandiosity and omnipotence. All but one of the criteria above focus on the underlying vulnerability of the narcissist: the feelings of inferiority, lack of enjoyment in life, work and social inhibitions, etc. The argument that the DSM-III and DSM-IIIR, and thus the NPI, are incomplete measures of narcissism is supported by recent empirical work by O'Brien (1987, 1988). As mentioned above, a factor analysis of O'Brien's scale (OMNI) revealed three factors which emerged congruently across 69 normal and clinical populations. Only one of these factors correlated significantly with the NPI. One of them correlated negeniyely with the NPI. Summary In summary, the NPI, which is based on DSM-III criteria for narcissistic personality disorder, has been factor analyzed into four factors. The scale as a whole has been found to correlate significantly only with the Selfism Scale. There is some evidence that it also correlates significantly with the MCMI. Only one factor, Exploitativeness/Entitlement, correlates significantly with the MCMI (Emmons, 1984), the NPD and the negative self focus scale of the SFCT. The NPI scale as a whole also correlates with personality traits including creativity, sensation seeking, extroversion, psychoticism, dominance, independence, self esteem, self monitoring, and exploitativeness. It correlates negatively with three empathy scales, religious values, abasement, deference and social anxiety. Sex differences in narcissism were found in one study using the NPI. In this study, males scored higher than females on the NPI in a college student sample. It is concluded that only one factor of the NPI measures pathological narcissism. Further, the NPI does not tap the underlying shame pole of the narcissistic personality disorder. Rather, it appears to measure the 7O grandiose defense against that shame. If, in fact, females are more prone to experiencing the shame pole of the narcis- sistic disorder, as 0'Leary has suggested, they will not be as likely to test positive on this instrument, and therefore will not be considered narcissistic. Projective Tests Rorschach Exner (1969) found that reflection responses, pair type responses, C and CF responses on the Rorschach were related to narcissism (diagnosed by clinicians and measured separately but concurrently by a sentence completion test devised by Watson (1965)). This finding validated both the impulsivity and the self absorption of the narcissist which is discussed in the literature (Goldstein, 1985). - e ' s In 1973 Exner constructed the Self Focus Sentence Completion Test (SFCT) from the best items on the sentence completion test devised by Watson. This instrument is used specifically to measure self absorption and will be presented in detail. The SFCT is scored on four mutually exclusive dimen- sions which are: Self focus, External focus, cher (neither self nor external focus), Ambivalence (both self and other focus), and two qualifying dimensions: ginl-negative self focus, §(e1;external focus -affective. Exner provided normative data for the SFCT, using 2,592 nonpsychiatric 71 subjects and 273 psychiatric patients. He found significant differences between these two groups in the d%, or the difference between the number of Self Focus scores and External Focus scores. Exner calls the ratio between these two scores the "Egocentric balance". He cites theory including that of Freud, Jung, Erikson, Piaget, Rogers and Maslow which maintains that self and other focus should be balanced in the healthy individual. Exner found that a difference (d%) score of 3 or less was found in 75% of the normative sample, whereas a score of 4 or more was found in 75% of the psychiatric sample. Exner also found that in three independent studies of schizophrenics, acting out adolescents and psychosomatics, that patients who improved significantly with treatment showed a corresponding decrease in their d% when tested with the SFCT both before and after treatment. This scale offers some important characteristics which make it a valuable tool. While the SFCT was not specific- ally constructed to measure narcissism, the parent scale from which it was derived was found to correlate signifi- cantly with one factor (Exploitativeness/Entitlement) on the NPI and with a high number of reflection responses on the Rorschach. Further, the narcissist is consistently referred to in the literature as self focused. As such, we could easily predict that the narcissist would show a d% larger than 4. While this scale offers no obvious features which 72 would enable the differential identification of the two types of narcissism which have been separated according to gender (above), it does offer the possibility of identifying a self focused population from which further testing could be done. The concept of d% eliminates the need to identify the narcissistic personality by either the grandiose or the shame conscious defense when seeking to simply identify a narcissistic population. Both defenses are likely to manifest themselves similarly in terms of self focus. Unfortunately, Exner found that psychiatric groups diagnosed as schizophrenic, adolescent behavior problem, and psycho- path also showed a d% larger than 4. These findings should not necessarily rule out the use of this scale to identify a narcissistic population in a college student sample. This is argued on the grounds that schizophrenics and psychopaths are not likely to be found in a college student sample and a diagnosis of adolescent behavior problem is not inconsistent with a diagnosis of Narcissistic Personality Disorder. It does, however, indicate that the population identified would not be a narrow one and so would introduce a probable source of bias if used as an instrument for identifying a narcis- sistic population. fielder Harder (1979) developed a scale to measure the Ambitious-Narcissistic Character style with three projective tests: the Early Memories Test, Thematic Apperception Test 73 and the Rorschach. The Ambitious-Narcissistic style, originally described by Reich (1933) is redefined by Harder as follows: "The ambitious narcissistic style in the adult corresponds to strong and pleasureful elements of self- confidence, ambition, mastery, strength, prowess, intrusiveness, bodily exhibitionism, and urethral erotic fantasy content" (p. 25). This definition is included here for two reasons. First, it illustrates the confusion over whether narcissism is a desirable constellation of adaptative personality traits, or whether it represents a defense against underlying leek of self esteem. This confusion, mentioned above in reference to DSM-III criteria, occurs elsewhere in the literature and thus makes it necessary to define specifically what is meant by the word narcissism in each and every study. Second, it illustrates a common misperception about narcissism. While Kohut does make reference to "healthy narcissism", by which he means healthy self development, this study provides another example of research which does not address the basic polarity of the narcissistic character dleezden, but instead focuses only on the grandiose defense. Thus, the research has become misleading, or at least, confusing. Again, if it can be shown that the grandiose defense is more typical for males, research which focuses primarily on the grandiose defense may eliminate females. 74 £111.55 Urist (1977) developed a scale for use with the Rorschach to assess object relations functioning. The focus of the scale was measurement of a particular style of relatedness which Urist terms ”mutuality of autonomy." According to Urist, mutuality of autonomy is a type of relatedness in which the individual has attained the "capacity to attribute to others an autonomous, inherent identity and to cathect them in their own right." This is in marked contrast to using others for the purpose of filling one's own needs as the narcissist does. Urist's scales focus on the developmental progression toward separation—individuation, with particular emphasis given to the issue of the autonomy of others vis-a-vis the self and the self vis-a-vis others. Particular scale points refer to levels of the capacity to experience self and others as mutually autonomous within relationships. Urist was able to demonstrate the validity of this scale as well as the enduring consistency of the individual's conception of human relationships across several types of ratings. While this method of measurement assesses the unconscious determinants of object relations and so renders a more complete picture of the subject's inner world, it's major limitation is the difficulty of Rorschach administration, scoring and interpretation of a large number of subjects for research purposes. 75 H22 Shulman and McCarthy (1986) developed a projective measure of narcissism which uses responses to TAT cards and early memories. This scale, called the N-P, has four parts. These parts include stories written to each of two TAT cards (cards 1 and 13MF), and written descriptions of the subject's earliest memory and most striking childhood memory. Each of the four protocols is scored according to six criteria taken from the DSM-III section on Narcissistic Personality Disorder. The four protocols are scored 0, 1 or 2 depending on how descriptive the protocol content is of the six DSM-III criteria. These criteria include: 1. Grandiosity or fantasies of ideal love, perfect beauty, or unlimited or unrealistic success. 2. Idealization and/or devaluation of people. 3. Entitlement or interpersonal exploitativeness. 4. Lack of empathy. 5. Oversensitivity to criticism, that is, rage or coolness in response to others' criticism or indifference. 6. Need for attention and/or admiration. Interjudge agreement on scoring has been established at the .05 level, and clinical validity has been indicated by an 85% classification agreement between N-P and interview ratings, by a clinical psychologist, on narcissism. Shulman and Ferguson (1988) found the following in their efforts to establish construct validity of the N-P. The NPI and one of its factors (Self Absorption/Self Admiration) has demonstrated statistically significant associations with the N-P and an interview assessment of 76 narcissism. A second factor of the NPI, Exploitativeness/ Entitlement, which has been identified as the one scale factor representing pathological narcissism, did not attain statistical significance, but did show an association with the N-P and interview rating. Shulman and Ferguson (1988) consider that the N-P is a better measure of pathological narcissism than the NPI. While the N-P appears to be measuring something which is somewhat distinct from the NPI, it is still based on the DSM-III criteria which emphasizes the grandiose defense over the shame ridden self state. As such the N-P is considered by this author to be an instru- ment which might also have a gender bias. Other Self Report Measures The MMPI Narcissistic Personality Disorder Subscale is the one self-report scale which uses Kernberg's criteria, end has been matched with measurements of the low self esteem pole of the narcissistic disorder for construct validation. Ashby, Lee and Duke, who constructed the scale, drew on the following elements of Kernberg's criteria: "chronic uncertainty and dissatisfaction about oneself, grandiose fantasies existing side-by-side with feelings of inferiority (1975, p. 264)." Solomon (1982) correlated the MMPI NPD scale with the Tennessee Self-concept Scale (TSS), involvement in a satisfying love relationship, and frequency of nightmares. Solomon found that high scores on the NPD correlated negatively with high scores on the T88, with involvement in a satisfying love relationship, and 77 correlated positively with high frequency of nightmares. In other words, those who were high on narcissism according to the NPD, were 39; high on self esteem, were nee involved in a satisfying love relationship, and gene subject to frequent nightmares. The significant findings were all within the group which scored high on narcissism and so were more at the narcissistic personality dleexden end of the continuum, as opposed to those who merely showed narcissistic traits. Solomon's subjects, however, were drawn from a college sample. This study is important for the purposes of the current research for two reasons. First, it demonstrates that a group of highly narcissistic persons can be found within a normal population of college students. Second, it measures both polarities of the narcissistic disorder. Q'E I H 1!. l . H . . I ! The O'Brien Multiphasic Narcissism Inventory (OMNI) (O'Brien, 1987) was developed to measure Alice Miller's hypothesized dimensions of narcissism. Miller, according to O'Brien, argues for two dimensions of narcissism which can be distinguished from the Narcissistic Personality Dimension outlined in DSM-III and DSM-III-R. These additional dimensions include (1) the Narcissistically Abused Dimension (NAD) and (2) the Poisonous Pedagogy Dimension (PPD). The Narcissistically Abused Dimension is, according to O'Brien: ...marked by tendencies toward looking for others' approval for self validation, experiencing problems with belongingness, and recognition of others' needs as being of greater importance than one's own. (p. 500) 78 The Poisonous Pedagogy Dimension is described as follows: ...reflects an unconscious need to control others, by assaulting and potentially blocking the normal narcis- sistic development of the other person. Specifically, Miller argued that in child-rearing, poisonous pedagogy takes the form of rigid disciplinary manipulations purportedly carried our for the child's "own good" (Miller, 1984). (p. 500) The third dimension, as mentioned, is considered to be like that described in DSM-III. O'Brien's (1987) factor analysis of the OMNI revealed three orthogonal factors, as above (NAD, PPD, NPD). Later, these same three factors emerged when the scale was tested on a clinic population (O'Brien, 1988). Validity studies with the OMNI show important findings. The Narcissistic Personality Dimension is the enly factor which correlates significantly with the NPI. This factor also correlates positively and significantly with the Eysenck Personality Inventory Extroversion subscale. In contrast, the Narcissistically Abused Dimension showed an lnyexee correlation with the Extroversion subscale. In addition, the NAD correlated -.02 with the NPI. Importantly, all three factors correlated significantly with clinician diagnosis of Narcissistic Personality Disorder when validity studies were carried out using a clinic population (O'Brien, 1988). O'Brien did not look at sex differences in response pattern in any of the four studies in which the OMNI has been used. 79 The O'Brien scale has some important features which make it useful for the current study. First, it has been shown to be a valid measure of narcissistic personality disorder, but clearly measures mere than the NPI. Second, the dimension which does correlate with the NPI is clearly noted to represent narcissism as depicted in DSM-III, and is clearly descriptive of the grandiose defense. Third, the dimension which has an inverse correlation with both extroversion and the NPI scale, appears to address the self state of narcissism which 0'Leary has argued is more common in female narcissists, and which specifically addresses the perception that "others' needs ...are...of greater importance than one's own." This description clearly resembles the position that Philipson considers female narcissists to occupy, specifically as a consequence of being in a prolonged, close relationship with a narcissistic parent. As such, the OMNI scale is considered by this author to be well suited for use in the exploration of the above hypothesized gender patterns in narcissism. Summary of Research and Scales In this section I have reviewed several scales which have been developed to measure Narcissistic Personality Disorder. While the NPI is the most widely used instrument to date, it has been found to correlate with constructs and personality traits which represent a range of functioning from healthy self esteem to the grandiose defense. Only one 80 scale factor correlates consistently with measures of pathological narcissism. It does not correlate with personality traits which suggest vulnerability and shame. Because both grandiosity and vulnerability are assumed to co-exist and alternate in consciousness in the narcissistic character, it was suggested that the NPI is an incomplete measure of narcissism. Critically, it was noted that if women are more prone to experiencing the shame pole of the narcissistic disorder, they will not test positive on this instrument. One study of sex differences in prevalence of narcissism using this instrument did, in fact, find a sig- nificant difference between male and female scores, with males scoring higher. Other scales reviewed include the Rorschach, SFCT, N-P, NPD and the OMNI. While the Rorschach tests an interesting aspect of object relations functioning, it is not considered practical for use with a large number of subjects. The SFCT is a good measure of self absorption and could be used to identify a self focused population, but is not narrow enough in focus for use in research which is designed to explore sex differences in a narrow diagnostic category. The N-P, like the NPI, is based on the DSM-III and DSM-IIIR criteria for narcissism, and as such, is considered by this author to be biased toward the grandiose defense, and thus, perhaps, male narcissism. The NPD represents both poles (grandiosity and shame) of the Narcissistic Personality Disorder and 81 correlates with three independent measures of low self esteem. The OMNI reveals three orthogonal factors within the scale, two of which do not correlate with the NPI. One factor has characteristics of the shame pole of narcissism, which 0'Leary and Broucek have suggested may be particularly common in women. This scale also correlates well with the independently given clinical diagnosis of Narcissistic Personality Disorder, its critical importance lies in the fact that it is measuring an aspect of narcissism which is not represented in the NPI, nor in the DSM-III. It was concluded from this review that the best instruments to use in the current research to identify a narcissistic population which is neither narrowly focused on grandiosity, or widely focused on self absorption, are the Narcissistic Personality Disorder Scale of the MMPI (NPD) and the O'Brien Multiphasic Narcissistic Inventory (OMNI). CURRENT RESEARCH For the purposes of this study it is assumed that for subjects diagnosed narcissistic personality disorder, women and men share certain common features. Using Kohut's theoretical basis, it is assumed that both male and female narcissists use others to further cohesion of their self structure. In Kohut's terminology, others are used as "self objects" rather than being perceived as whole objects in their own right. Further, it is assumed that both men and women experience the mutually contradictory stances which 82 Akhtar and Thompson consider to be the central feature of the condition. These two dynamics then interact. The mutually contradictory stances are experienced in alterna- tion. The stance which is split off from consciousness is projected into the self object. When joined with the self object, a feeling of wholeness ensues. The mutually contra- dictory stances are characterized as grandiose and shame ridden. Thus, when feeling grandiose, the self object will contain split-off feelings of shame and be devalued. When feeling shame ridden the self object will be idealized as it contains the split-off feelings of grandiosity. So long as self and object perceptions are split in this way, the real self which might be realized in part by integrating these two self perceptions, remains submerged. The narcissist then does not live out of his real self. Rather, he/she lives out of a false self which is constructed to maximize external validation. Stolorow's functional definition, again, captures the rudiments of gender neutral narcissistic mental activity: Mental activity is narcissistic to the degree that its function is to maintain the structural cohesiveness, temporal stability and positive affective coloring of the self representation. This definition would include the use of self objects to complete the self, the experience of and defense against mutually contradictory stances, and the need for external validation. All of these are considered to apply equally to male and female narcissists. What may differ for males and 83 females, according to the foregoing literature review, is the specific use of the self object, a preference for one pole of the contradictory stances, and the means of achieving external validation. From the preceding review of the literature it is suggested that (1) female narcissists are more likely to be conscious of their shame and not conscious of their grandiosity, while male narcissists are more likely to be conscious of their grandiosity and not conscious of their shame, (2) female narcissists are more likely to use the male self-object to contain their split off grandiosity, while male narcissists are more likely to use the female self-object to contain their split off shame, and (3) female narcissists extract external validation by acting as an extension of their male partner. It is also suggested that the NPI may identify significantly more male narcissists than female narcissists, and that the OMNI-NP factor may identify more male narcissists, while the OMNI-NA factor may identify more female narcissists. These are the issues which will be addressed in the current research. The specific hypotheses are: H1: Male and female subjects identified as narcissistic will differ on a measures of conscious shame, with female subjects being more shame conscious than male subjects. H2: Male and female subjects identified as narcissistic will differ on a measures of unconscious shame and grandiosity, with males showing more unconscious shame and females showing more unconscious grandiosity. H3: H4: H5: H6: 84 Male and female subjects identified as narcissistic will show different patterns of defensive projection, with female subjects projecting unconscious grandiosity on to the male, and males projecting unconscious shame on to the female. Male and female subjects identified as narcissistic will project images of female characters as extensions of male partners. Male and female subjects identified as narcissistic will score significantly different on the Narcissistic Personality Inventory, with male subjects scoring higher than female subjects. Male and female subjects identified as narcissistic will score significantly different on the O'Brien Multiphasic Narcissism Inventory, with female subjects scoring higher on the Narcissistically Abused factor, and male subjects scoring higher on the Narcissistic Personality factor. MEIHQD RESEARCH DESIGN The purpose of the current research was twofold. First, it was designed to measure sex differences in shame and grandiosity, and the content of defenses and compensations in a sample identified as narcissistic. Second, it was to explore possible gender bias in an instrument which is currently widely used to measure narcissism. O'Leary's hypothesis that female narcissists experience more conscious shame was explored. Broucek's hypothesis that female narcissists experience more unconscious grandiosity and male narcissists experience more unconscious shame was also explored. Reich's and Philipson's argument that narcissistic women project their unconscious gran- diosity into male partners, while narcissistic men project their unconscious shame into female partners was also ex- plored. Phillipson's suggestion that narcissistic women find their source of external validation in acting as an extension of a significant other was also investigated. Lastly, the NPI and the OMNI scales were administered to explore sex differences in response patterns. To test the hypotheses 67 male and 134 female under- graduate students enrolled in an introductory psychology class were recruited to complete four self report measures and write responses to three Thematic Apperception Test 85 86 (TAT) cards, two of which depicted men and women in rela- tion. The OMNI scale was used to define a narcissistic population. The TAT cards were used to assess unconscious affect and object relations patterns. The ISS was used to assess conscious shame. The NPI and the OMNI were adminis- tered in order to examine sex differences in response patterns. The self report measures were computer scored. The TAT stories were rated by two independent raters for the affec- tive, defensive and compensatory themes mentioned above. RESEARCH INSTRUMENTS Thematic Apperception Test The Thematic Apperception Test (TAT), developed by Morgan and Murray (1935), is a projective test which requires the subject to tell a story in response to an image or picture. The cards are selected according to their portrayal of themes which are of interest to the clinician or investigator. Subject stories created in response to the specific cards reveal the subjects' inner world, including thoughts, feelings, fantasies, wishes and conscious and unconscious motives (Karon, 1981). Although the TAT reveals both conscious and unconscious thoughts and feelings, it was used in this study for the specific purpose of exploring the unconscious content, as well as to explore defensive patterns. The TAT was included specifically to explore the unconscious content which a projective instrument reveals in 87 order to contrast that material with the conscious content revealed in the self report measures. Four cards from the TAT (Cards 4, 10, 86F, and 17BM) were used to elicit stories from subjects (see Appendix A). Cards 4 and 10 were selected because they depict scenes of men and women in relation and are commonly used to elicit thoughts and feelings concerning the subject's experience with relating to the opposite sex (Karon, 1981). These cards were used to assess the projective defenses detailed in H3. Cards 8GF and 17BM were selected because they depict a woman (8GP) and a man (188M) alone, and were used to elicit thoughts and feelings which subjects high on narcis- sism have when they are alone. No specific hypotheses were formulated regarding sex differences in the thematic content of these cards. Female subjects wrote stories to cards 4, 8GF and 10. Males subjects wrote stories to cards 4, 17BM and 10. As mentioned above, the tendency for female narcis- sists to project grandiosity into the male character and for male narcissists to project shame and/or inadequacy into the female characters were explored. The tendency for narcis- sistic subjects to portray females as extensions of their male partners was also explored. Specific scoring criteria and instructions are included in Appendices B and C. H 's . !° E II! I ! The Narcissistic Personality Inventory (Raskin, 1979) was selected for use in this study because the current 88 review of the literature led to the possibility that there may be a gender bias in this instrument. Specifically, because this scale emphasizes the grandiose defense against underlying feelings of inadequacy, it is hypothesized that male narcissists may score significantly higher than female narcissists. This scale was administered to the entire population and then examined for differences in response patterns of subjects identified as high, medium and low on narcissism. The NPI was developed in direct response to inclusion in the DSM-III (American Psychiatric Association, 1980) of the diagnostic category Narcissistic Personality Disorder. Two studies have tested and supported the construct validity of the scale (Prifitera, 1984: Emmons, 1987). Eight-week alternate form reliability of the scale has also been demonstrated (Raskin and Hall, 1981). The scale has two versions, a 40-item and a 54-item format. The 40-item version was developed after factor analytic studies determined which of the items in the original 54-item scale accounted for most of the variance within each of four factors (Emmons, 1984). The 40-item version was used in the current study (see Appendix D). The individual items consist of two dichotomous statements. The subject must endorse one. Each item consists of a narcissistic and a non-narcissistic statement. The scale score is the total number of narcissistic statements endorsed. Prifitera and 89 Ryan (1984) found a mean score of 25.4 in subjects identi- fied as high in narcissism, and a mean score of 14.5 in subjects identified as low in narcissism. 's 's 'c s n 't D' o d S The MMPI Narcissistic Personality Disorder scale (Ashby, Lee and Duke, 1979) was selected for use in the current study because it is the one self report measure which has been matched with measurements of the low self esteem pole of the narcissistic disorder for construct validation (Soloman, 1982). This scale, however, was dropped from the study for several reasons. First, although the scale did correlate at a statistically significant level with the OMNI (the other scale used to measure narcissism), l; = .24, e <.001) NPD scale reliability was low (.42). Also, using the means reported from the clinic population, 80% of the current, normal sample was identified as clinically narcissistic. In addition to this being an improbable finding for a normal population, this finding in inconsistent with the findings from the OMNI scale, which identified only a small percentage of the sample as clinically narcissistic. For all of the above reasons, it was decided that the NPD scale should be dropped from the study, and the OMNI scale retained as the only instrument used to define the narcissistic population. 90 Wu The O'Brien Multiphasic Narcissism Inventory (OMNI) (O'Brien, 1987) was selected for use in the current study because a factor analysis of the scale (O'Brien: 1987, 1988) revealed three orthogonal factors: one of which shows a significant positive correlation with the NPI, and one of which shows a negative correlation with the NPI and appears to represent a dimension or type of narcissism which is theoretically consistent with a type of narcissism which some have considered to be more common in women (Johnson, 1987: 0'Leary, 1986: Philipson, 1985: Reich, 1953). OMNI scale scores were also examined for sex differences in response patterns. Differential endorsement of factors 1 (NPD) and 3 (NAD) according to gender was hypothesized. Three studies have been done which provide evidence for the construct validity of the scale. Factor 1, the Narcissistic Personality Dimension, was found to correlate significantly with the NPI and with the extroversion scale of the Eysenck Personality Inventory. Factor 3, the Narcissistically Abused Dimension, was found to have an inverse relationship to both the NPI and the extroversion scale of the Eysenck Personality Inventory. The scale as a whole was found to correlate significantly with independent diagnoses of Narcissistic Personality Disorder by clinicians in four separate out-patient clinics in New York City. Thus, it was concluded that the OMNI measures three separ- ate, orthogonal dimensions of Narcissism (see Appendix F). 91 Because only a small number of the total sample could be identified as clinically narcissistic according to the criteria identified by O'Brien (1986) using the OMNI scale, the total sample distribution was divided into thirds, with the upper third being defined as "high" in narcissism, the middle third being defined as ”moderate" in narcissism, and the lower third being defined as "low" in narcissism. WW1: A modified version of the Internalized Shame Scale (ISS) (Cook, 1986) was selected because it is the one shame scale which emphasizes the inner affective experience of the respondent rather than emphasizing situational cues and thereby measuring affects other than shame (Chang, 1988). The scale modification is based on Novak's (1986) factor analysis which revealed that factors other than shame were included in Cook's original scale. Based on Novak's find- ings, 11 items were selected from the original version of the scale (See Appendix G). Validity information on the entire scale is provided by Cook (1988). Validity infor- mation on this shortened version of the scale is provided by Chang (1988). The scale is in self—report format. Items are endorsed on a 5-point likert scale. The total scale score is the sum of individual item scores. 92 PROCEDURE Data Collection Subjects were recruited from the Michigan State Univer- sity Subject Pool. All participants were enrolled in an introductory psychology course and received course credit for their participation. Subjects were told that they were participating in a psychology research experiment entitled "Creative Story Telling." All data were collected in a group setting. One female experimenter was present at each session. Subjects were seated in a classroom and given a packet containing all four self report measures, 4 lined 8 1/2 x 11, blank sheets of paper on which to record responses to TAT cards, two pencils, one pen, a consent form (See Appendix I) and 1 answer sheet. Subjects were instructed to read and sign the consent form. Subjects were then asked to record their age, sex and marital status on the top of the packet of blank sheets of paper. All subjects were then instructed to make sure their packets were complete. Instructions for participation were then given as follows: This is a story of creative story telling which I think you will find enjoyable. Before we get started with the stories, there are four questionnaires which I would like you to fill out. Please complete each questionnaire on the computer sheet which you will find in your packet. Each questionnaire is numbered to match the computer answer sheet numbers, so that all questionnaire answers may be recorded on a single answer sheet. Please use only the pencils provided. 93 The first questionnaire is called the NPI. Please read the instructions and complete this questionnaire now. You will have 10 minutes. I will tell you when you have three minutes left and when your time is up. Please remember to answer the questions on the computer sheet - not on the questionnaire. Are there any ques- tions? (wait 7 minutes, announce time, wait 3 minutes and stop.) The second questionnaire is called the NPD. Please read the instructions and complete this questionnaire now. You will have 10 minutes. I will tell you when you have 3 minutes left and when your time is up. Use the computer sheet to record your answers beginning at item #41. Are there any questions? (Wait 7 minutes, announce time, wait 3 minutes and stop.) The third questionnaire is called the OMNI. Please read the instructions and complete this questionnaire now, beginning with item #60. You will have 10 minutes. I will tell you when you have 3 minutes left and when your time is up. Use the computer sheet to record your answers. (Wait 7 minutes, announce time, wait 3 minutes and stop.) The fourth questionnaire is called the 188. Please read the instructions and complete this questionnaire now, beginning with item #101. You will have 5 minutes. I will tell you when you have 1 minute left and when your time is up., Record your answers on the computer sheet. Are there any questions? (Wait 4 minutes, announce time, wait 1 minute and stop.) Now we are ready for the last part of the research. Please remove the pages with pictures on them from your packet. You should have three pictures. I am going to ask that you concentrate on one picture at a time. Do not look ahead at the other pictures until you are told to do so. Please look at the first picture. Your task is to make up as dramatic a story as possible about this picture and write that story on the first blank page included in your packet. The other two blank pages are for the next two stories you will write. Please make sure that these pages remain stapled together and that you clearly mark at the beginning of each story, "Story #1," "Story #2,” etc., so I will know which stories go to which pictures. Please use the pen included in your packet. Do not try to erase anything. If you make a mistake, just draw a line through the error and continue on. 94 When writing your story, please remember the following directions. Make up as dramatic a story as you can. Have fun with the story. Please include in your story what has led up to the event shown in the picture, describe what is happening at the moment, what the characters are feeling and thinking: and then give the outcome. Write your thoughts as they come to mind. You will have 15 minutes to record each story. You will be asked to write a total of three stories, one for each picture. In a moment, I will ask you to begin writing your first story. I will let you know when you have only 5 minutes left, and when to stop. Are there any questions? Please begin. (wait 10 minutes, announce time, wait 5 minutes and stop.) Please go on to the next picture. Again, make up as dramatic a story as you can. Have fun with it. Please include in your story what has led up to the event shown in the picture, describe what is happening at the moment, what the characters are feeling and thinking: and then give the outcome. Write your thoughts as they come to mind. You will have 15 minutes. I will tell you when you have 5 minutes left and when to stop. Please begin. Repeat for final story. At the end of the testing session students were debriefed about the purpose of the study and given some information about theory and research upon which the study is based. Subjects were also given a written debriefing sheet (see Appendix J) and informed that results would be made available to them upon completion of the study. Training of Coders One upper-level graduate student in clinical psychology, and one BA level psychology major were used as coders for this study. Both coders were blind as to the hypotheses and all identifying characteristics of the subjects. 95 The coder for shame and grandiosity on the TAT stories was trained at the BA level in psychology. He and the investigator jointly coded four sets of 20 stories in order to establish reliability. Reliability percentages established between the coder and the investigator during these four trials, in order of their occurrence, were 79%, 87%, 95%, and 90%. During the process of establishing reliability, decision rules were added to the original scoring criteria to reflect issues which arose in the process of coding actual data. The original scoring criteria, together with decision rules developed while working on reliability, are included in Appendix B. Once percent reliability had been established at an adequate level, the coder was given 100 stories to code over a one-month period. The remaining 300 stories were distributed 100 at a time, each to be completed over a one- month period. After each 100 stories was completed, the coder gave the investigator a list of stories which he found difficult to score. These stories were then coded by the investigator and the coder jointly, with a meeting following to compare coding and discuss differences, and arrive at a consensus decision. In this way the coder was in continual contact with the investigator, thus reducing the possibility of drift. 96 E ! . g 3' Extension coding was done by an upper level graduate student in clinical psychology. She and the investigator jointly coded five sets of 10 stories to establish reliability. Reliability percentages established between the coder and the investigator during these five trials, in order of their occurrence, were 80%, 90%, 80%, 60% and 80%. During the process of establishing reliability, decision rules were added to the original scoring criteria to reflect issues which arose in the process of coding actual data. The original scoring criteria, together with decision rules developed while working on reliability, are included in Appendix C. Once an acceptable reliability percentage had been established, the coder was given 70 stories to code over a two-week period. At the end of this period the coder and the investigator met to discuss protocols which were difficult to score. Because there were a large number of protocols which were difficult to score with the existing coding system, the coding system was further revised after this meeting. A second round of reliability checks was then carried out to ensure the reliability of the revised coding system. A total of three reliability checks was done. For each one, the coder and the investigator each coded 20 protocols. Reliability percentages on these three trials were 70%, 60% and 80%. 97 The coder then completed 35 stories each week, checking at the end of each week with the investigator to discuss protocols which were difficult to score. These protocols were then discussed by the coder and the investigator until a consensus decision could be made. In this way the coder and investigator maintained continual contact thereby reducing the possibility of drift. DATA ANALYSIS The raw data consist of subject responses to four self- report scales, and three TAT card stories. The self report measures are the Narcissistic Personality Inventory (NPI) (40 items), the O'Brien Multiphasic Narcissism Inventory (41 items), the Internalized Shame Scale (11 items) and the Narcissistic Personality Disorder Scale (19 items). The TAT cards were cards 4 and 10 for all subjects, and card 8GF for female subjects and 17BM for male subjects. All responses to self report questionnaires were recorded directly onto single computerized grid sheets which were then fed directly into the computer scanner. This process eliminated the necessity for transfer of information from questionnaires to computer forms, and thus reduced the probability of random error in this part of the process. Subject responses to TAT cards 4 and 10 were coded as detailed above. All subjects received scores for grandiosity, shame and extension. 98 Shame and grandiosity were scored from cards 4 and 10. Each subject received four scores: one for each character on each of two cards. The scores were tallied separately for female and male characters, and a total score including both characters was recorded for each subject. Extension was scored from card 4 only, as card 10 depicts intimacy and has been shown to elicit responses which suggest "merged" object relations functioning (Schwartz, 1987: Berry, 1985). In this study, it was thought that it would be difficult to separate "merged" object relations functioning from "extension" object relations functioning. Each subject received two scores: one for each character on card 4. The scores were recorded separately for female and male characters, and a total score including both characters was recorded for each subject. Each subject had a total of 120 scores, including the questionnaire items and the TAT story codes. In order to analyze the data, reliability analysis was first carried out for all four self-report scales. Using Cronbach's Alpha, reliability coefficients for each of the scales were as follows: Narcissistic Personality Inventory (Alpha = .78), Narcissistic Personality Disorder Scale (Alpha = .42), O'Brien Multiphasic Narcissism Inventory (Alpha = .67), Internalized Shame Scale (Alpha = .91). T-Tests were then computed to investigate sex differences in the whole population on all variables included in the research design. 99 The correlations between all four self-report scales (above), as well as three factors on the OMNI scale (Narcissistically Abused (NA), Narcissistic (NC), and Poisonous Pedagogy (PP)) were computed (See Appendix H). A narcissistic population was identified by dividing the total distribution of scores on the OMNI scale into three parts. The top 1/3 was labeled 'high', the middle 1/3, 'medium' and the bottom 1/3 'low'. Cell frequencies were adjusted to make cells in corresponding rows and columns proportional. A two-way Analysis of Variance was used to investigate the relationship between a number of dependent variables and two independent variables. The dependent variables were conscious shame (ISS score), unconscious shame (shame scored on the TAT), unconscious grandiosity (grandiosity scored on the TAT), grandiosity attributed to the male TAT character, grandiosity attributed to the female TAT character, and the three factors on the OMNI scale (Narcissistically Abused, Narcissistic Personality, Poisonous Pedagogy). To explore sex differences in several continuous variables at all levels of narcissism, Pearson correlations between these variables and narcissism were computed separately for male and female subjects. The two correlations were then examined to determine if they were significantly different. Chi-Square analysis was used to investigate the relationship between level of narcissism, sex and several 100 dichotomous variables. These dichotomous variables included the number of female TAT characters who felt shame, the number of male TAT characters who felt shame, the total number of TAT extension characters, female TAT extension characters and male TAT extension characters. There were no subjects who had both male and female extension characters, or more than one male or female extension character, so that Chi-Square was appropriate. Chi-Square analyses were done for male and female subjects separately, and male and female subjects combined. T-Tests were used to look at sex differences in means on all variables by level of narcissism. In other words, the ANOVA was broken down into three individual components, in order to examine the individual means for non-significant trends. Sample Characteristics Subjects for this research were recruited from the Michigan State University Subject Pool. The subjects ranged in age from 17 to 39, with a mean age of 20. The number of subjects in each age category were as follows: Number of Subjects Age 4 17 50 18 51 19 36 20 27 21 20 22 5 23 1 24,27,28,31,39 2 38 101 167 of the subjects were female and 67 of the subjects were male. Four of the subjects were married, one was divorced and 195 were single. RESULTS SUMMARY OF MAIN RESULTS Statistical analysis of the data indicates that (1) female subjects report significantly more conscious shame than male subjects, but there is not a significant interaction between level of narcissism and sex differences in conscious shame. (2) Male and female subjects identified as narcissistic do not differ significantly on measures of unconscious (TAT) shame and grandiosity, (3) Male and female subjects identified as narcissistic did not show different patterns of defensive projection. Female narcissistic subjects did not project more grandiose male characters, and male narcissistic subjects did not project more shameful female characters. (4) Both male and female subjects identified as narcissistic tend to project more extension characters than non-narcissistic subjects, and more female than male extension characters. These findings, also, however, did not reach statistical significance. (5) Male subjects scored significantly higher on the NPI. There was not a significant interaction, however, between level of narcissism and sex differences in NPI score. (6) Male and female subjects identified as narcissistic did not score significantly different on the OMNI NA and/or NC factors. 102 103 SPECIFIC HYPOTHESES HI: Male and female subjects identified as narcissistic will differ on a measure of conscious shame, with females showing significantly more conscious shame than males. This prediction was not supported by the data. While it was found that there was a population-wide significant sex difference in conscious shame, with female subjects scoring higher than male subjects, analysis of variance results showed no significant interaction between the three levels of narcissism identified and conscious shame score (See Table 1). Table 1 ANOVA-CS. Shame By Sex and Narcissism Level Source of Variation F Significance of F Main Effects Level of Narcissism 18.981 .000 Sex 5.366 .022 Two-way Interactions Level by Sex 45.613 .350 32: Male and female subjects identified as narcissistic will differ on measures of unconscious shame and grandiosity, with female subjects showing more unconscious grandiosity, and male subjects showing more unconscious shame. This prediction was not supported by the data. Analysis of variance results for both predictions above showed a nonsignificant interaction for level of narcissism and unconscious affect. 104 s ' G ' 3' Table 2 shows the ANOVA results for level of unconscious (TAT) grandiosity by level of narcissism. As the table indicates, there is not a significant sex difference in unconscious grandiosity, and no significant interaction between sex differences in unconscious grandiosity and level of narcissism. Table 2 ANOVA-UCS. Grandiosity By Sex and Narcissism Level Source of Variation F Significance of F Main Effects Level of Narcissism .453 .64 Sex .352 .55 2-way Interactions Level by Sex 1.371 .26 s ' us A ham Table 3 shows the ANOVA results for level of unconscious shame by level of narcissism. As the table indicates, there is no significant sex difference in unconscious shame and no significant interaction between sex differences in unconscious shame and level of narcissism. 105 Table 3 ANOVA-UCS. Shame By Sex and Narcissism Level Source of Variation F Significance of F Main Effects Level of Narcissism .45 .63 Sex .35 .55 Two-way Interactions Level by Sex .22 .26 H3: Male and female subjects identified as narcissistic will show different patterns of defensive projection, with female subjects projecting unconscious grandiosity into males and male subjects projecting unconscious shame into females. This prediction was not supported by the data. Chi-Square analysis was used to investigate the relationship between level of narcissism, sex, and shame and grandiosity projected into opposite-sex TAT characters. Chi-Square analysis was done for male and female subjects separately on the shame and grandiosity projections. There were no significant findings on these variables. Tables 4 and 5 show the results of the Chi-Square analyses. 106 Table 4 Female Subjects - Proj. of Grandiosity into Male Characters Low Medium High Narcissism Narcissism Narcissism Male Character Grandiosity 6 5 2 No Grand. 35 36 39 Table 5 Male Subjects - Projection of Shame into Female Characters Low Medium High Narcissism Narcissism Narcissism Female Character Shame 2 1 1 No Shame 18 19 19 H4: Male and female subjects identified as narcissistic will project images of females as extensions of males. A strong trend in support of this prediction was revealed by the data, although again, findings did not reach statistical significance. Chi-Square analysis was used to investigate the relationship between level of narcissism, projection of 'extension' characters, and sex of characters depicted as 'extensions.' Table 6 shows the Chi Square results for projection of female extension characters by male and female subjects combined. The number of female extensions characters mm Kin—L.“ m 107 projected at the "high" level of narcissism is more than twice the number projected at the "low" level (n <.06). Table 6 All Subjects — Projection of Female Extension Characters Low High Narcissism Narcissism Number of Female Extension Characters 5 12* Number of Female Non- extension Characters 56 49 *n <.06 Table 6 (above) shows the Chi Square results for the comparison of the "high" and "low" groups. As noted above, these results are nearly significant, and thus show a strong trend in support of the hypothesis that narcissistic subjects project images of females as extensions of males. HS: Male and female subjects identified as narcissistic will score significantly different on the Narcissistic Personality Inventory, with male subjects scoring higher than female subjects. This prediction was not supported by the data. A significant sex difference was found population wide in the NPI score, with male subjects scoring higher than female subjects. Analysis of Variance showed no significant inter- action, however, between the three levels of narcissism identified and sex differences in the NPI score (see Table 7). 108 Table 7 ANOVA-NPI Score by Sex and Narcissism Level Source of Variation F Significance of F Main Effects Level of Narcissism 6.24 .002 Sex 3.89 .050 Two-way Interaction Level by Sex .57 .565 HG: Male and female subjects identified as narcissistic will score significantly different on the O'Brien Multiphasic Narcissism Inventory, with female subjects scoring higher on the Narcissistically Abused factor and male subjects scoring higher on the Narcissistic Personality factor. These predictions were not supported by the data. Analysis of variance was used to investigate sex differences in the two OMNI factors by level of narcissism. No signifi- cant sex differences were found on either factor, and no significant interaction was found between sex differences in the two factors and level of narcissism. Results of the analyses are shown in Tables 8 and 9 below. 109 Table 8 ANOVA-Sex Differences in OMNI-NA Factor Narcissism Level Source of Variation F Significance of F Main Effects Level of Narcissism 81.45 .000 Sex .83 .362 Two-way Interactions Level by Sex .32 .725 Table 9 ANOVA-Sex Differences in OMNI-NC Factor by Narcissism Level Source of Variation F Significance of F Main Effects Level of Narcissism 90.18 .000 Sex .01 .920 Two—way Interactions Level by Sex .70 .498 Additional Findings In addition to the specific hypotheses discussed above, one other finding adds measurable impact to the argument detailed above that the NPI scale is an incomplete measure of narcissism. Using the Pearson Product Moment Correlation, the Narcissistic Personality Inventory was found to correlate negatively with the Internalized Shame Scale at a significant level (; = -.289, n <.000). The exploration of themes from cards 8GF and 17BM revealed no findings for sex differences in content. DI§QH§§IQE Although none of the findings in this research reached statistical significance, an analysis of trends in the data reveal that there are some limited findings in the directions predicted. These findings will be detailed below. Overall, the data show that there were no sex significant sex differences in narcissism in a population of subjects scoring high on a clinically validated narcissism scale. Because these subjects were not considered clinically narcissistic, some of the trends revealed may be found to extend to a clinical population if tested. In any discussion of research based on a "category" of illness, it is important to remember that categories are attempts to reduce extremely complex phenomena to a manageable size. That complexity and nuance will be lost is assumed. In line with this, readers should be reminded that "narcissism" is a 20th century construct, which is only defined and used in this country. It is defined as a constellation of symptoms which are purported to reflect a particular underlying character structure. Within the literature, as noted extensively earlier in this work, it is defined inconsistently. Outside the literature, in common parlance, it is used pejoratively to describe someone who is self-absorbed and unempathic. With this as a backdrop, it is more than obvious that the construct has limitations, and as such can be misused. Importantly, the most relevant 110 111 misuse for the purposes of clinical practice is that it may serve to further misunderstanding of our clients who present for help with mystifying and painful life experiences. Given that the use of diagnostic categories is a first approach to understanding these life experiences, the practice of documenting experiences within this 'system of categories' must be carefully scrutinized. Toward that end, the theoretical and research literature on narcissism has been reviewed. Findings from this literature include the discovery by clinicians that what has been called narcissism is far more differentiated and complex than a one-word category can address. Further, in a culture where men and women have been treated differently, hold different role expectations and possess different strengths and weaknesses, it is only logical to suppose that different symptom pictures may emerge in reaction to similar failures of maternal/paternal empathy. To ignore that these differences may exist is to run the risk of distorting what is presented clinically, and thereby further the "distortion" process for the narcissist, who is already presenting as a stranger to his/her own needs and feelings. This research has been an attempt to further differentiate a very complex phenomenon, to attempt to demonstrate empirically what some clinicians have observed in their clinical practice, and to reduce distortion in the clinical and research literature of the female experience. 112 To review briefly, it has been argued that there are two types of narcissism. It has been suggested that the failure of integration of shame and grandiosity, which is the fundamental symptom of a narcissistic character, is the basis for these two types. Again, it is suggested that the narcissistic character feels shame or grandiosity, in alternating states, cannot integrate these discrepant states, and does not feel anything in between. This then becomes the foundation for the two types, one type who favors a shame state and the other who favors a grandiose state. The shame state has been associated with the need to cling to others, to act as an extension of others, and to hold the grandiosity at an unconscious level. The grandiose state has been associated with using others as an extension of self, and with holding shame at an unconscious level. This research has been focused not on the demonstration of there being a shame-based and a grandiose-based type of narcissism per se, but on the extent to which a shame-based and a grandiose-based type may be gender linked. There is some suggestion in the data that certain traits which manifested to a limited extent in the current sample, may manifest at a statistically significant level in a clinical sample. These include the tendency to project females as extensions of males, and the tendency for females to hold grandiosity and males to hold shame at an unconscious level. Both of these variables, as shown below, were seen to increase in terms of a sex discrepancy as the 113 level of narcissism increased. The conscious shame variable and the NPI variable, however, while approaching significant sex discrepancies, held those discrepancies almost constant when moving from middle to high levels of narcissism. Thus, while conscious shame may be more common to the average female than the average male, that difference does not reach statistical significance in a narcissistic sample because the male narcissist has so much shame relative to his non- narcissistic cohort. The same analogy holds for the sex discrepancy in the NPI variable. While males tend to score higher on the NPI when looking at the whole sample, narcis- sistic females will have enough of the narcissistic traits contained in the NPI scale that a significant sex difference may not emerge in a clinically narcissistic sample. It was not demonstrated in this study that the female narcissist projects her grandiosity into her male partner, although a very strong tendency was demonstrated for both males and females to project an image of a female who is an extension of the male. These findings are somewhat discrepant. Reich's clinical finding that the narcissistic female acts as an extension of her male partner whether or not she idealizes him provides one possible explanation for this finding. As discussed above, Reich considers that the narcissistic female may be on a continual quest for the perfect male, and may be disillusioned over and again by what she finds. Thus she has an even chance at any one time of being a disillusioned extension preparing to move on. 114 There is also a subtle distinction between projecting one's grandiosity onto a blank screen vs. the unconscious search for a truly grandiose person in one's environment. It is the latter which is assumed to underlie the less than coin- cidental pairing of opposites. Perhaps a research project which explored actual pairing behavior among shame-based and grandiose-based narcissists in the real world would yield a finding more in line with the sex differences hypothesized above. Trends The data on unconscious shame and grandiosity reveal a nonsignificant trend in the directions predicted, which may extend to a significant finding in a clinically narcissistic sample. To explore this hypothesis in greater depth, T- tests were used to compare male and female subjects on the unconscious shame and grandiosity variables at each of the three levels of narcissism. Table 10 shows the mean scores for men and women on unconscious grandiosity at each level of narcissism. Table 10 Mean UCS Grandiosity Scores by Sex and Narcissism Level Low Medium High Narcissism Narcissism Narcissism Males .80 .75 .50 Females .54 .78 .80 115 As table 10 shows, male and female subjects reverse scores when moving from low to high narcissism. Female subjects have more and more unconscious grandiosity as they become more narcissistic. Male subjects have less and less unconscious grandiosity as they become more narcissistic. Because the differences increase as the level of narcissism increases, a clinically narcissistic sample may show a sex difference in the direction predicted. Table 11 shows the results for the same analysis done with the unconscious shame variable. The mean scores for male and female subjects by level of narcissism are shown in the table. Table 11 Mean UCS Shame Scores by Sex and Narcissism Level Low Medium High Narcissism Narcissism Narcissism Males .15 .15 .25 Females .21 .14 .07 Again, note that male and female subjects reverse positions when moving from low to high narcissism. Male subjects have more and more unconscious shame as their level of narcissism increases, while female subjects' shame decreases. It is interesting to note that while female subjects increase in conscious (ISS) shame as the level of narcissism increases, as shown in H1 (above), these same subjects decrease in unconscious (TAT) shame. This finding lends support to the 116 effectiveness of the self report and projective measure as methods for eliciting conscious and unconscious material. RESEARCH IMPLICATIONS The Narcissistic Personality Inventory is the most widely cited instrument for measuring narcissism in the literature. Results from this study include the finding that the NPI is negatively correlated at a significant level with the Internalized Shame Scale. Thus, the higher a subject scores on the NPI, the lower they score on shame. Yet, according to the most widely cited theory on narcissism (Kohut, Kernberg), the mutually contradictory states of shame and grandiosity are considered to be the central feature of the disorder. These results support the argument made above that the NPI is an incomplete measure of narcissism. In addition, results from this study include the finding that there is a population wide significant sex difference in rate of response on the NPI. Men score significantly higher than women. While the finding for sex differences among the subjects high in narcissism was nonsignificant, a-trend in this general direction was found. These findings support the argument that the instrument is gender biased. In contrast to the NPI, the OMNI scale was found to be correlated at a significant level with both shame (ISS) (; = .39, e <.000) and grandiosity (NPI) (I = .24, n <.001). 117 Further, there was no sex difference in rate of response either for the population as a whole, or for the narcissis- tic population alone. This finding suggests that the OMNI scale is a more complete measure of the disorder, does not have a gender bias, and is thus a more meaningful research instrument. Directions for Further Research It is suggested that a larger number of subjects, a clinically narcissistic sample or more sensitive instruments might enhance the findings contained in this research, or yield findings which are statistically significant. In terms of instrument sensitivity, one explanation for the lack of findings in this study is the possibility that the self report measure, administered individually, is not a sufficient stimulus to elicit the narcissistic personality dynamics outlined in the literature review. Perhaps the defensive splitting, projection of disowned parts of the self and treatment of female partners as "extensions" are phenomena which are most present during actual interpersonal events. If so, the hypothesized sex differences would be most effectively tested with observational studies. Another suggestion for future research would be to divide subjects by gender personality type rather than sex. Gender-role inventories which measure the extent to which a given subject has feminine or masculine characteristics, might more effectively define a sample which has personality 118 patterns associated with the hypothesized sex differences in narcissism, than the simple definition of masculinity and femininity according to physical characteristics. APPENDIX A TAT CARDS 119 120 121 APPENDIX B TAT SCORING CRITERIA FOR SHAME AND GRANDIOSITY APPENDIX B TAT SCORING CRITERIA The four TAT protocols will be scored on three separate dimensions including: (1) presence of the projection of shame or grandiosity in protocols written in response to the two relationship stimuli, and (2) presence of characters portrayed as "extensions" of other characters in protocols written in response to only one of the relationship stimuli. (3) No specific scoring system will be used with the solitude cards. Instead, these protocols will be examined for the presence of any consistent themes. Scoring for Projection of Shame and Grandiosity These coding instructions pertain to cards 4 and 10. Keep each subject's TAT cards stapled together. Then for each subject code all protocols for shame. For each protocol, code the male and the female TAT character separately. After coding each protocol for presence of shame, begin again and code all protocols for grandiosity. Use the following guidelines to score for shame and grandiosity. fiflAME Score shame if the character feels as though he/she: is not good enough is somehow left out is looked down upon by other people doesn't measure up scolds self or puts self down sees self as small and insignificant wonders to self "how could anyone really love me or care about me?" is defective as a person, as if something is 123 124 basically wrong with him/her is intensely inadequate or full of self doubt sees self as striving for perfection only to continually fall short in comparison to others, is just not as important QBAEDIQ§IIX Score grandiosity if the protocol reveals: grandiosity or fantasies of ideal love, perfect beauty, or unlimited or unrealistic success devaluation of people entitlement or interpersonal exploitativeness lack of empathy oversensitivity to criticism, that is, rage or coolness in response to others' criticism or indifference need for attention and/or admiration Decision Rules for Coding Shame and Grandiosity 1. Be sure to base decisions only on what is written in criteria. 2. When in doubt, score for the healthiest capacity the subject or character shows. 3. If there is a third person in the story (someone not pictured in the TAT card), who would be scored either grandiose or shameful), note the sex of the character and enter a "1" to indicate a positive score under "Other." 4. Look at the process. Scoring for shame and grandiosity is, essentially, scoring for deficits in the regulation of self esteem. Where self esteem is poor, there is a tendency to react to setbacks by feeling either extremely bad about oneself (shame), or to defend against one's shame by feeling shame's polar opposite, grandiosity. Healthy regulation of self esteem entails the ability to bear setbacks (the break up of a relationship, death of a loved one, loss of status, i.e., loss of a job, money, home, etc.) without concluding that either (1) the self is to blame (shame) or (2) the self is enelEled to something better, and so feels outraged (grandiosity). In the TAT protocols look for a sense that the author (through the characters) is able to see recovery from life events as a process, where there may be initial pain, followed by a gradual 'building up of the self' or 'coming to terms' with the loss. The ability to conceptualize process resolution is a positive indicator that the 'self' is perceived, in some sense, as in transition - still being 125 built or improved. The self is not static, i.e., either inadequate or perfect. 5. In general, when a character chooses suicide or homicide as a way to resolve feelings, this would be scored 'grandiose,' but not 'shame' (unless the character states that his suicide is the result of feeling shame about the self or self hatred). Score grandiose on the basis that the character's oversensitivity to setbacks is what prompts the action. 6. The same is generally true for characters who choose to end a relationship in response to conflict, with no attempt to work on the feelings first, either alone or with the other person. Look at the context first, but use this action as a 'red flag' for grandiosity. 7. Remember that a character showing need would not necessarily be considered grandiose. Everyone has needs - it is only considered grandiose if the need is expressed inappropriately or, in other words, if a need appears extreme and/or is expressed in an extreme way. Likewise, a character who is, literally, not included in an interaction depicted in the story is not necessarily someone who feels left out all the time. A character who would be scored for "shame" on this criteria would be someone who feels left out and it is clearly stated in the story, or obvious that it is a feeling the character carries with him. APPENDIX C CRITERIA FOR CODING EXTENSION APPENDIX C CRITERIA FOR CODING EXTENSION Read through the story first to get a sense of it. Then consider, for each story, if the female character is depicted as an exeenelen of the male character, and then if the male character is depicted as an ekeenelen of the female character. Use the following criteria to determine if one character is an extension of the other. A character is considered an extension of another character if they derive their self esteem by: a) being part of an esteemed other and so, -over identifying with their love partner -fusing -living through the other -overinvesting -idealizing the other -being a good self-object (doing whatever the other wants to the detriment of the self) Below I will list some things we are nee measuring. I will also list some decision rules made in the process of scoring. 1. We are not looking at relationships in which characters have poor boundaries. This is too broad a category. While it is true that characters who act as "extensions" of other characters have poor boundaries, it is also true that many characters with poor boundaries do not fit the criteria for being extensions of others. 2. We are not looking at dysfunctional relationships in general. Again, this is too broad a category. We are selecting a very specific type of dysfunctional relationship. WHAT WE ARE LOOKING FOR: Below are some decision rules based on a preliminary sample of stories. A character who is an extension of another character may feel or act in some or all of the following ways: 1. An extension character makes it their business to figure out what the other character wants or needs from them and then makes sure they do their best to do it et tnei; own eknenee or te ene detgimene of tnei; egn well being. Here the character is deriving their self esteem by being a good self object - as opposed to by taking care of themselves. 126 127 2. An extension character is not likely to be described as willful or argumentative - or as having a strong opinion or definite point of view that would counter the wish of the love partner. 3. An extension character will feel empty and desperate if the relationship fails. They may kill themselves or "never by the same" at the extreme. Less extreme would be the inability to resolve the trauma - i.e., there being no mention in the resolution that they ever got over it and went on with their lives. 4. These characters are not likely to have rich, individual thoughts and ideas when in relation to others. 5. Extension characters are likely to take on the values of those they are in a relationship with. 6. In some of the stories previously considered, it has been stated that a character "understood" some atrocity committed by the other character (with no process leading to the "understanding") and/or "agreed with everything said" or "would do anything to make it ok," etc. While these stories may have been coded "0" (non-extension) when other story elements were taken into consideration, these lines by themselves indicate strongly the "extension" self, and would come under the category of "being a good self-object," i.e., doing whatever the other wanted done. It might also come under the category of ”idealization." 7. If a character says "don't go - I will be empty /lost/ nothing /dead /etc. without you,” code extension. If the character says "don't go! I want to be with you," this is more likely to be coded "0," the difference being that the second character is making a choice out of preference rather than desperation or to avoid fragmentation of the self. 8. If a character leaves a bad relationship, this is an indication that that character is nee an extension. Since they had the resources to initiate termination of the relationship, they should not be considered an extension of it. 9. Score the healthiest capacity. See if extension-like behavior is resolved. If it is, do not score 'extension.' APPENDIX D NARCISSISTIC PERSONALITY INVENTORY (NPI) ... mflka-a'h-v 's‘ _ APPENDIX D NARCISSISTIC PERSONALITY INVENTORY INSTRUCTIONS: The NPI consists of a number of pairs of statements with which you may or may not identify. Consider this example: A. "I like having authority over people", verses B. "I don't mind following orders". Which of these two statements is closer to your own feelings about yourself? If you identify more with "liking to have authority over other people", then with "not minding following orders", then you would choose option "A". You may identify with both "A" and ”B". In this case you should choose the statement which seems closer to your personal feelings about yourself. Or, if you do not identify with either statement, select the one which is least objectionable or remote. In other words, read each pair of statements and then choose the one that is closer to your own feelings. Indicate your answer by writing the letter ("A" or "8") in the space provided to the right of each item. Please do not skip any items. 1. A. I have a natural talent for influencing people. 1. B. I am not good at influencing people. 2. A. Modesty doesn't become me. 2. B. I am essentially a modest person. 3. A. I would do almost anything on a dare. 3. B. I tend to be a fairly cautious person. 4. A. When people compliment me I sometimes get embarrassed. 4.___ B. I know that I am good because every- body keeps telling me so. 5. A. The thought of ruling the world frightens the hell out of me. 5.___ B. If I ruled the world it would be a much better place. 6. A. I can usually talk my way out of anything. 6. B. I try to accept the consequences of my behavior. 7. A. I prefer to blend in with the crowd. 7. B. I like to be the center of attention. 8. A. I will be a success. 8. B. I am not too concerned about success. 9. A. I am no better or no worse than most people. 9. B. I think I am a special person. 128 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. A. B. 129 am not sure if I would make a good leader. see myself as a good leader. am assertive. wish I were more assertive. like having authority over other people. don't mind following orders. find it easy to manipulate people. don't like it when I find myself manipulating people. HH HH HH HH I insist upon getting the respect that is due to me. I usually get the respect that I deserve. I don't particularly like to show off my body. I like to display my body. I can read people like a book. People are sometimes hard to understand. If I feel competent I am willing to take responsibility for making decisions. I like to take responsibility for making decisions. I just want to be reasonably happy. I want to amount to something in the eyes of the world. My body is nothing special. I like to look at my body. I try not to be a show off. I am apt to show off if I get the chance. I always know what I am doing. Sometimes I'm not sure what I am doing. I sometimes depend on people to ge things done. I rarely depend on anyone else to get things done. Sometimes I tell good stories. Everybody likes to hear my stories. I expect a great deal from other people. I like to do things for other people. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. A. B. A. B. A. B. A. B. A. B. A. B. A. B. B. 130 I will never be satisfied until I get all that I deserve. I take my satisfactions as they come. Compliments embarrass me. I like to be complimented. I have a strong will to power. Power for its own sake doesn't interest me. I don't very much care about new fads and fashions. I like to start new fads and fashions. I like to look at myself in the mirror. I am not particularly interested in looking at myself in the mirror. I really like to be the center of attention. It makes me uncomfortable to be the center of attention. I can live my life in any way I want to. People can't always live their lives in terms of what they want. Being an authority doesn't mean that much to me. People always seem to recognize my authority. I would prefer to be a leader. It makes little difference to me whether I am a leader or not. I am going to be a great person. I hope I am going to be successful. People sometimes believe what I tell them. I can make anybody believe anything I want them to. I am a born leader. Leadership is a quality that takes a long time to develop. I wish someone would someday write my biography. I don't like people to pry into my life for any reason. I get upset when people don't notice how I look when I go out in public. I don't mind blending into the crowd when I go out in public. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. A. B. 40. A. B. 131 I am more capable than other people. There is a lot that I can learn from other people. I am much like everybody else. I am an extraordinary person. 39. 40. APPENDIX E NARCISSISTIC PERSONALITY DISORDER SCALE APPENDIX E NARCISSISTIC PERSONALITY DISORDER SCALE Please answer the following questions by marking '1' if you consider the statement to be true or mostly true about yourself, and by marking '2' if you consider the statement to be false or mostly false about yourself. There are no right or wrong answers. You may consider that the statement is both true end false, or that the statement is nelene; true nor false. In these cases, you should still choose either true or false depending on which one is eleee; to your personal feelings about yourself. Please continue with the next number on your answer sheet which should be #41. If the next number on your answer sheet is nee #41, please raise your hand and request assistance from the experimenter. 41. I enjoy detective or mystery stories. 42. My sex life is satisfactory. 43. A minister can cure disease by praying and putting his hand on your head. 44. I used to like drop-the-handkerchief. 45. I believe that my home life is as pleasant as that of most people I know. 46. I certainly feel useless at times. 47. It makes me impatient to have people ask my advice or otherwise interrupt me when I am working on something important. 48. There is something wrong with my mind. 49. I seldom or never have dizzy spells. 50. I believe there is a Devil and a Hell in afterlife. 51. I can easily make other people afraid of me, and sometimes do for the fun of it. 52. Life is a strain for me much of the time. 53. My sex life is satisfactory. 54. Once in a while I think of things too bad to talk about. 132 55. 56. 57. 58. 59. 133 I cannot keep my mind on one thing. I have certainly had more than my share of things to worry about. I often feel as if things were not real. I have felt embarrassed over the type of work that one or more members of my family have done. I worry quite a bit over possible misfortunes. APPENDIX F O'BRIEN MULTIPHASIC NARCISSISM INVENTORY APPENDIX F O'BRIEN MULTIPHASIC NARCISSISM INVENTORY Use your answer sheet to answer the following questions with a X§§ or £91 There is no "right" answer. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. Would you rather try to please others than to have your own way? Would you rather give a gift than receive one? Do you find it easy to relax in a group? Do you tend to feel like a martyr? Do you tend to see people as being either great or terrible? Do you usually find it hard to settle down? Do you tend to get angered by others? Do you have a tendency to over-react? Are you jealous of good-looking people? Do you tend to be secretive about your personal life? Do you pay a lot of attention to the financial matters of others? Do you think that movie stars have better lives than you do? Do you try to avoid dramatizing your feelings? Does your life deserve special recognition? Will your experiences greatly guide others? When confused, do you think of your mother's wishes to help you resolve your conflict? Do you appreciate people to march to the beat of a different drummer? Do you try to avoid rejection at all costs? 134 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 135 Do you have fantasies about being violent without knowing why? Do you tend to feel humiliated when criticized? Do you know how to solve other people's problems? Would your secretive acts horrify your friends? Do people love you for the way you improve their lives? Do you find it easier to empathize with your own misfortunes than with those of others? Do your views of people change back and forth easily? Do you think that sexual intercourse is clean? Do you wonder why people aren't more appreciative of your goodness? Do you avoid telling people "what its all about"? Are you a perfectionist Is seduction the best part of your sex life? Do you find that going through life is like walking on a tightrope? Do you find yourself fantasizing about your greatness? Do you have problems that nobody seems to understand? Are you clever enough to fool people? Do you worry a lot about your health? Do you expect people who love you to spend money to show it? Is it important for you to know how other people spend their time? Do all your friends come from the same mold? Are you especially sensitive to success and failure? If you're tough on others, is it "for their own good'? Do you crave attention from others? APPENDIX G INTERNALIZED SHAME SCALE (ISS) APPENDIX G INTERNALIZED SHAME SCALE (ISS) MODIFIED VERSION Below is a list of statements describing feelings or experiences that your may have from time to time or that are familiar to you because you have had these feelings and experiences for a long time. Most of these statements describe feelings and experiences that are generally painful or negative in some way. Some people will seldom or never have had many of these feelings. Everyone has had some of these feelings at some time, but if you find that these statements describe the way you feel a good deal of the time, it can be painful just reading them. Try to be as honest as you can in responding. Read each statement carefully and blacken in on your answer sheet numbers 1-5 according to the following scale. The range 1-5 indicates the frequency with which you find yourself feeling or experiencing what is described in the statement. Use the scale below. DO NOT OMIT ANY ITEM., l-Never 2-Seldom 3-Sometimes 4-Frequently 5-Almost Always 101. I feel like I am never quite good enough. 102. I feel somehow left out. 103. I think that people look down on me. 104. Compared to other people I feel like I somehow never measure up. 105. I scold myself and put myself down. 106. I see myself as being very small and insignificant. 107. I say to myself, "how could anyone really love me or care about me?" 108. I feel defective as a person, as if something is basically wrong with me. 109. I feel intensely inadequate and full of self doubt. 110. I see myself striving for perfection only to continually fall short. 111. When I compare myself to others I am just not as important. 136 APPENDIX H CORRELATIONS APPENDIX H CORRELATIONS Table 1 NPI Scale/OMNI Scale and All OMNI Factors OMNI Scale NC Factor NA Factor PP Factor Total Only Only Only NPI Scale .2418 .1666 -.0761 .4044 Total P=.0001 P=.018 P=.283 P=000 Table 2 NPI Scale/NPD and Shame Scale NPD ISS (MMPI NPD Scale) (Shame Scale) NPI Scale .0207 -.2897 Total P=.770 P=.000 Table 3 OMNI Scale/Shame Scale and NPD Scale NPD ISS (MMPI NPD Scale) (Shame Scale) OMNI Scale .4990 .3907 Total P=.000 P=.000 137 APPENDIX I CONSENT FORM APPENDIX I CONSENT FORM Thank you for your interest in participating in this research project. This study is focused on an investigation of the experiences which accompany relationships with others. If you choose to participate, I will be inviting you to use your imagination in the creation of stories. I will also ask you to complete some questionnaires. About two hours of your time will be requested. If you would like to take part, your signature is required on this form to indicate that you have, or are now being, informed of the following: 1) 1Your participation in the study has been explained to your satisfaction and is understood. 2) You freely consent to participate. 3) You understand that you may discontinue the experiment at any time without recrimination. 4) All information which you supply is both confidential and anonymous. 5) There are judged to be no risks of any kind associated with participation in this study. 138 APPENDIX J DEBRIEFING FORM APPENDIX J DEBRIEFING FORM Thank you for participating in my research study. I hope that you find your experience to be both interesting and enjoyable. My study is designed to explore the emotional experiences of men and women in their relationships with each other when certain kinds of bonding experiences have occurred earlier in life. Specifically, some believe that women's closer relationships with their mothers early in life, when coupled with certain kinds of bonding experiences, may predispose them toward an impaired ability to sustain their sense of self when involved in an intimate relationship. These same women may carry a feeling of shame as a consequence of this and other aspects of their charac- ter. Men, conversely, when experiencing similar bonding patterns early in life, are thought to experience a false sense of power and prestige, rather than shame, and to use others to admire their accomplishments. Your participation in this study required that you fill out four questionnaires and write one story to each of three pictures, two showing a man and woman together, and one showing either a man or a woman alone. Your answers to the questionnaires and the stories you wrote will tell me something about your early bonding experiences and how they impact on your relationships with others today. By looking 139 140 carefully at your answers and your stories, I hope to determine if there are consistent differences between men and women whose early experiences may have predisposed them to certain thoughts and emotions which accompany being together with each other and being alone. If you are interested in finding out the results of this study, please contact me at the address below. I will take your name and address and mail a brief summary of the results to you when I complete the study. The study will not be completed for approximately one year. If you are interested in reading further on this subject, the books listed below provide an interesting introduction to the issues covered in the research. Thank you for your time and interest. Carol Schwartz 534 Sycamore Lane East Lansing, MI 48823 Chodorow. N- (1978) W W Berkeley: University of California Press. Philipson. I- (1985). W Psychology of Women Quarterly, (, 213-228. LI§I_QE_BEEEBEN§E§ Akhtar, S. & Thomson, A. (1982). Overview: narcissistic personality disorder. Amcricanalournal_of_£sychiatry. 112.(l). 12-20- American Psychiatric Association. (1980). ’ ' ' (3rd ed.). Washington,DC: APA. American Psychiatric Association. (1986). Qiegnee;ig_end zeyleedl. Washington, DC: APA. Ashby, H. U., Lee, R. R., & Duke, E. H. (1979). A narcissistic personality disorder MMPI scale. Paper presented at the meeting of the American Psychological Association, New York, September 3,1979. Benton, C. M., Hernandez, A. C. R., Schmidt ,A., Schmitz, M. D., Stone, A. J., & Weiner, B. (1983). Is hostility linked with affiliation among males and with achievement among females?: A critique of Pollack and Gilligan. denznel of_Rcrsona11ty_anc_§oc1al_£sycholocy 15(5). 1167- -1171. Bach. 8.8- (1985). Narc1ssist1c_scaccs_aad_thc_thcraocccic DIQEE§§1 New York: Jason Aronson, Inc. Broucek, F. J. (1982). Shame and its relationship to early narcissistic developments. ln;ezne§lenel_den:nel_e1 Psychoanalysis as. 369- -378. Chodorow, N. (1978). u ' e ' : Psychoanalys1s_aac_chc_sociolccy_of_ccnccr- Berkeley: University of California Press. Cooper, A. M. (1986). Narcissism. In Esseneiel 2a pens en Neneleelenl Andrew P. Morrison, M. D., Editor, New York University Press, pp. 348- 372. Emmons, R. A. (1987). Narcissism: theory and measurement. Iouraal_of_2crsona11ty_anc_§ocial_2sycholocy 52(1). 11-17. Emmons, R. A. (1984). Factor analysis and construct validity of the narcissistic personality inventory. Journa1_of Personality_Asscssmcnt 11(3). 291- ~300- Exner, J.E., Jr. (1969). Rorschach response as an index of narcissism. Iocrnal_of_2crsonality_1sscssmcat. 11. 324-330. 141 142 Exner, J. E., Jr. (1973). The self focus sentence completion: Astudy of egocentricity. den:nel_ef Personality_Asscssman 11. 437- 454- Freud. 8- (1914)- 9n_Harcissism1__An_Iatroccction1_ Standard Edition 14: 73-102, 1957. Goldstein, W.N. (1985). DSM III and the narcissistic personality- Amcrican_Iocraa1_of_£sychochcrapy. 22(1). 4-16 0 Gove, W.R. (1980). Mental illness and psychiatric treatment among women- 2sychclocy_of_Eomcn_ouartcrly. 5(3), 345-361. Haaken, J. (1983). Sex differences in narcissistic disorders. Thc_Amcr1can_Icarnal_of_£sychoanalysis. ge(4), 315-324. Harder, D. W. (1979). The assessment of ambitious- narcissistic character style with three projective tests: The early memories, TAT and rorschach. Jenznal cf_£crsonality_1sscssman 11(1). 23- 32- Horney. K- (1950). ncarosis_ana_haman_crocth1__1hc_stracclc tosarc_sclf:rcalication1_New York: Norton. Johnson, M. (1980). Mental illness and psychiatric treatment among women- 2sycholocy_of_Eomcn_Quartcrly. 1(3), 363-371. Johnson. S-M- (1987). haman1z1nc_chc_narciss1scic_§tylc. New York: W.W. Norton & Company. Karon, B.P. (1981). The thematic apperception tests (TAT). In A.I. Rabin (Ed.), Asseeenent wltn egejeetlve geennlgne_. New York: Springer Publishing Co., Inc., pp. 85-120. Kohut. H. (1971). Thc_analys1s_of_chc_sclfo New York: International Universities Press. Kohut, H. (1977). es 0 ' t e . New York: International Universities Press. Kohut, H. (1981). w o a 's u ? New York: International Universities Press. Lachmann, F.M. (1985). Narcissism and female gender identity: A reformulation. The Eeyenoenelyele Review, 69(1), 43- 61. 143 Lewis, H.B. (1979). Gender identity: Primary narcissism or primary process- Ba11cc1n_cf_thc_hcnninccr_c11n1c. 51(2), 145-160. Mahler, M.S., Pine, R. & Bergman, A. (1975). The ' New York: Basic Books. Masterson. J- (1981)- 2sychotheraox_of_thc_hordcrlinc edglgl New York: Bruner/Mazel. Masterson. J. (1981). Thc_narcissistic_and_hordcrlinc dleerdexee New York: Bruner/Mazel. Meissner, W.W. (1986). Narcissistic and borderline conditions: a differential diagnosis. In EeeenElel Andrew P. Morrison, M.D., Editor, New York University Press, pp. 348-372. Miller. 3.8. (1984). Thc_ac1cloomant_of_comcnls_scasc_of eelf. Work in progress, No. 12, Wellesley College, Wellesley, MA. O'Brien, M. L. (1987). Examining the dimensionality of pathological narcissism: factor analysis and construct validity of the O'Brien multiphasic narcissism inventorY- Psychologica1_8cports1 cl. 499- -510. O'Brien, M.L. (1988). Further evidence of the validity of the O'Brien multiphasic narcissism inventory. Psychological_8cports1 92. 879-882. 0'Leary, J. & Wright, F. (1986). Shame and gender issues in pathological narcissism. Esychoaaalytic_£sycholocy. 1(4), 327-339. Phillipson, I. (1985). Gender and narcissism. RSYQDQIQQY WI 2: 213-228 0 Pollack, S., & Gilligan, C. (1982L Images of violence in thematic .apperception test stories. den;nel_efi s , _;(1), 159-167. Pollack, S., & Gilligan, C. (1983). Differing about differences: The incidence and interpretation of violent fantasies in women and men. 2crsoaa11cy_aac_soc1al_£sycholocy. 15.(5). 1172-1175- Pollack, S. & Gilligan, C. (1983L Killing the messenger. Iournal_of_Ecrsoaal1tYLand_§oc1al_£sYcholocY. 45(5). 1176- 1178. 144 Prifitera, A. 8 Ryan, J. J. (1984L Validity of the narcissistic personality inventory (NPI) in a psychiatric sample. Ioarna1_of_c11n1ca1_2sycho1ocy. 19(1), 140- -142. Pulver, S. (1986). Narcissism: The Term and the Concept. In Esscnt1al.2apcrs_on_uarc1ss1sm1 Andrew P- Morrison. M.D., Editor, New York University Press, pp. 91-111. Raskin, R. & Hall, C.S. (1981). The narcissistic personality inventory: Alternate form reliability and further evidence of construct validity. lengnel_ef £crsonality_bsscssmcnt. 15(2). 159-162- Raskin, R. & Hall, C.S. (1979). A narcissistic personality inventory- Psychological_hcports. 15. 590. Reich, A. (1953). Narcissistic object choice in women. . 1. Schwartz, C.S. (1987). d' e 'v' e ' . Unpublished masters thesis, Michigan State University, East Lansing. Shulman, D.J. (1986). Narcissism in two forms: Implica- tions for the practicing psychoanalyst. Eeyeneenelyeie E§Y§h212§¥. 1(2). 133-147- Shulman, D.G. and Ferguson, G.R. (1988). Two methods of assessing narcissism: comparison of the narcissism- projective (N-P) and the narcissistic personality inventory (NPI) WW. 55(6), 857-865. Soloman, R. S. (1982). Validity of the MMPI narcissistic personality disorder soale- Psychological_3coorcs. 59, 463-466. Stoller, R. (1976). Primary femininity. den:nel_ef_§he Fonalc_£sycholocy. 24. 59-78. Watson, P.J., Grisham, S., Trotter, M.V. & Biderman, M.D. (1984). Narcissism and empathy: Validity evidence for the narcissistic personality inventory. den;nel_ef Personality_Asscssman 48(3). 301-305- Urist, J. (1977L The rorschach test and the assessment of object relationS- Ioaraa1_of_£crsonalicy_Asscssmcnc. 41, 3-9. MICHIGAN STATE UNIV. LIBRARIES ’lHIWlIWHIHIINWIHIHW\ll'NIIWIHIHHI 31293007743127