Personality Disorders: Theory, Research, and Treatment © 2011 American Psychological Association 2011, Vol. 2, No. 2, 113–127 1949-2715/11/$12.00 DOI: 10.1037/a0021870 Parallel Syndromes: Two Dimensions of Narcissism and the Facets of Psychopathic Personality in Criminally Involved Individuals

Michelle Schoenleber, Naomi Sadeh, and Edelyn Verona University of Illinois at Urbana–Champaign

Little research has examined different dimensions of narcissism that may parallel psychopathy facets in criminally involved individuals. In this study, we examined the pattern of relationships between grandiose and vulnerable narcissism, assessed using the Narcissistic Personality Inventory–16 and the Hypersensitive Narcissism Scale, respec- tively, and the four facets of psychopathy (interpersonal, affective, lifestyle, and antisocial) assessed via the Psychopathy Checklist: Screening Version. As predicted, grandiose and vulnerable narcissism showed differential relationships to psychopathy facets, with gran- diose narcissism relating positively to the interpersonal facet of psychopathy and vulnerable narcissism relating positively to the lifestyle facet of psychopathy. Paralleling existing psychopathy research, vulnerable narcissism showed stronger associations than grandiose narcissism to (a) other forms of psychopathology, including internalizing and substance use disorders, and (b) self- and other-directed aggression, measured with the Life History of Aggression and the Forms of Aggression Questionnaire. Grandiose narcissism was none- theless associated with social dysfunction marked by a manipulative and deceitful inter- personal style and unprovoked aggression. Potentially important implications for uncover- ing etiological pathways and developing treatment interventions for these disorders in externalizing adults are discussed.

Keywords: grandiose narcissism, vulnerable narcissism, psychopathy, forensic, aggression

In its original Greek form, narcissism re- (e.g., Kernberg, 1975). Thus, individuals scoring flected a perception and presentation of oneself high on grandiose narcissism are poised, self- as superior and worthy of admiration and spe- confident, and exhibitionistic (Wink, 1991). In cial treatment. Current conceptualizations of contrast, those scoring high on vulnerable narcis- and research regarding narcissism, however, sism can be exploitative and entitled, but the hall- have suggested that not all individuals with mark symptoms involve hypersensitivity to criti- narcissism solely display a grandiose sense of cism, fear of rejection, and a fragile self-esteem self (American Psychiatric Association [APA], (Akhtar & Thomson, 1982; Wink, 1991). 2000). Rather, two fairly distinct dimensions of Both narcissism dimensions are currently narcissism seem to exist: grandiose narcissism embedded in narcissistic personality disorder as and vulnerable narcissism (e.g., Wink, 1991). a single syndrome (APA, 2000). However, con- Grandiose narcissism refers to the openly arro- flating the grandiose and vulnerable dimensions gant and superior presentation, whereas vulner- is potentially problematic for clinicians because able narcissism is characterized by conflict be- these dimensions may have very distinct func- tween a grandiose and an inferior sense of self tional impairments that require different treat- ment interventions. The empirical evidence on the two dimensions of narcissism is scarce, This article was published Online First March 7, 2011. making it difficult to ascertain the clinical sig- Michelle Schoenleber, Naomi Sadeh, and Edelyn Verona, nificance of the grandiose and vulnerable man- Department of Psychology, University of Illinois at Urbana– Champaign. ifestations of the disorder. Moreover, the re- We thank Sheriff Walsh, jail staff, and state parole and search that has been conducted on narcissism is county and federal probation offices in Champaign County relatively limited in scope in that it has relied for their assistance. primarily on convenience samples without se- Correspondence concerning this article should be addressed to Edelyn Verona, Department of Psychology, University of vere functional impairments (e.g., undergradu- Illinois at Urbana–Champaign, 603 East Daniel Street, Cham- ate samples) and measures of narcissism that paign, IL 61820. E-mail: [email protected] mostly assess grandiose traits (e.g., the Narcis-

113 114 SCHOENLEBER, SADEH, AND VERONA sistic Personality Inventory; Bushman et al., suggesting that these disorders may share under- 2009; Pincus et al., 2009). Although these stud- lying vulnerabilities for personality pathology. For ies provide preliminary data, more research instance, psychopathy is a marked by a constella- aimed at delineating the nomological networks tion of interpersonal (e.g., grandiosity, deceitful- of the grandiose and vulnerable dimensions is ness, superficial charm), affective (e.g., shallow needed to begin to understand the clinical rele- affect, callousness, remorselessness), lifestyle vance of these constructs. (e.g., risk taking, irresponsibility, aggressiveness, This distinction between grandiose and vul- unrealistic goals), and antisocial (e.g., criminal nerable narcissism to some extent parallels re- versatility, repeated incarcerations) traits that, like search on primary and secondary variants of narcissism, increase risk for engagement in anti- psychopathy, which have a rich theoretical and social behavior (e.g., Bushman & Baumeister, empirical literature (e.g., Blackburn, 1975; 1998; Cleckley, 1976). Moreover, narcissism and Karpman, 1941; Lykken, 1995; Skeem, Johans- psychopathy converge at the trait level of person- son, Andershed, Kerr, & Louden, 2007). Pri- ality in that both disorders are marked by grandi- mary psychopathy is theorized to stem from a osity, manipulativeness, and callousness (APA, core set of manipulative interpersonal traits and 2000; Hare, 2003) and with the five-factor model deficits in empathy, whereas secondary psy- dimensions of low agreeableness and high extra- chopathy is theorized to develop out of an im- version (e.g., Lynam, 2001; Samuel & Widiger, pulsive–aggressive lifestyle and deficits in emo- 2008). Finally, as described earlier, research has tion regulation (e.g., Hare, 1991). Given the suggested that psychopathy is also multidimen- potential parallels between the narcissism di- sional (Cooke & Michie, 2001; Hare, 2003; mensions and psychopathy variants, applying Hill, Neumann, & Rogers, 2004), with the pri- an established psychopathy framework in the mary and secondary variants of psychopathy study and understanding of grandiose and vul- representing separable etiological pathways to nerable narcissism may be useful. Thus, we the disorder. Unlike with narcissism, however, sought to expand current conceptualizations of extensive research has been conducted to delin- grandiose and vulnerable narcissism by drawing eate the nomological networks associated with on the extensive research literature on psychopa- the primary and secondary dimensions of psy- thy to form and test hypotheses about the narcis- chopathy, making the psychopathy literature a sism dimensions as they manifest in criminally potentially informative framework for examin- involved individuals. Narcissism and psychopathy ing distinctions between grandiose and vulner- have been linked in the literature already, as evi- able narcissism. denced by research on the “dark triad” (i.e., nar- Empirical investigations of the external cor- cissism, psychopathy, and Machiavellianism; e.g., relates of the primary and secondary dimensions Paulhus & Williams, 2002). In addition, psychop- validate their theoretical distinctions. For in- athy was selected as a theoretical framework for stance, research has indicated that individuals examining narcissism in an externalizing sample with primary psychopathic traits exhibit lower because of symptom overlap among the disorders levels of anxiety, social withdrawal (Skeem et and their respective associations with heightened al., 2007), and emotional reactivity (Blair, risk for aggression and engagement in illegal be- 2005; Kiehl, 2006; Patrick, Bradley, & Lang, havior (Bushman & Baumeister, 1998; Hare, 1993) than those with secondary psychopathic 2003; Edens, Hart, Johnson, Johnson, & Oliver, traits. In contrast, secondary psychopathy is asso- 2000; Jones & Paulhus, 2010). Thus, we sought to ciated with emotional and behavioral dysregula- determine whether using psychopathy as a frame- tion, including increased risk for suicidality, alco- work for understanding narcissism dimensions, hol and substance use disorders, and impulsivity particularly among criminally involved individu- (Lorber, 2004; Patrick, Hicks, Krueger, & Lang, als, is fruitful. 2005; Sellbom & Verona, 2007; Smith & New- man, 1990; Vanman, Mejia, Dawson, Schell, & Using Psychopathy as a Theoretical Raine, 2003; Verona, Hicks, & Patrick, 2005; Framework Verona, Patrick, & Joiner, 2001). In terms of violence, both primary and secondary dimensions Narcissism and psychopathy in forensic sam- of psychopathy have been found to predict reac- ples show both conceptual and empirical overlap, tive violence and aggression (Hart, 1998; Walsh NARCISSISM AND PSYCHOPATHY 115

& Kosson, 2008). However, individuals with nalizing disorders (64.2%, substance use disor- primary psychopathic traits are more likely to der overall; 30.6%, alcohol dependence; 25%, engage in unprovoked, instrumental forms of ag- antisocial personality disorder; Gunderson, gression than those with secondary psychopathic Ronningstam, & Smith, 1991; Stinson et al., traits (Skeem, Poythress, Edens, Lilienfeld, & 2008; Widiger & Corbitt, 1993). However, Cale, 2003; Williamson, Hare, & Wong, 1987; given that narcissistic personality disorder com- Woodworth & Porter, 2002). In sum, the extant bines grandiose and vulnerable features, which literature has indicated that primary psychopathy aspects of narcissism account for comorbidity is related to emotional stability, lack of anxiety, with internalizing and externalizing disorders is and unprovoked–instrumental aggression, unclear. A few studies have examined narcis- whereas secondary psychopathic traits are associ- sism dimensionally, and at least one study ated with psychopathological dysfunction marked found a positive relationship between depres- by emotional and behavioral dysregulation. sion and vulnerable narcissistic traits in clinical samples (r ϭ .57; Watson, Sawrie, Greene, & Narcissism–Psychopathy Parallels Arredondo, 2002). Moreover, a small literature Preliminary research on the association of has suggested that vulnerable, but not grandi- narcissism with psychopathy has produced ose, narcissism shows a positive relationship mixed results. Although several studies have with shame (e.g., Cheek & Hendin, 1996; indicated that narcissism is associated with pri- Gramzow & Tangney, 1992; Hibbard, 1992), mary psychopathy (e.g., Benning, Patrick, Blo- which is ubiquitous in internalizing psychopa- nigen, Hicks, & Iacono, 2005; Cale & Lilien- thology (e.g., Tangney & Dearing, 2002). These feld, 2002; Hart & Hare, 1989; McHoskey, findings suggest that, as with secondary psy- Worzal, & Szyarto, 1998; Rutherford, Alter- chopathy, vulnerable narcissism, but not gran- man, Cacciola, & McKay, 1996), the associa- diose narcissism, represents risk for internaliz- tion of narcissism with secondary psychopathy ing psychopathology. is less consistent. For instance, although some Research in undergraduates has shown that studies have found comparable relationships be- narcissistic traits may also increase risk for ag- tween narcissism and primary and secondary gression (e.g., Jones & Paulhus, 2010; Locke, psychopathy (e.g., Rutherford et al., 1996), sev- 2009; Paulhus, Robins, Trzesniewski, & Tracy, eral studies have reported relatively weaker 2004), and theory has suggested that aggressive (Benning et al., 2005, Study 1; McHoskey et al., tendencies influence the degree to which narcis- 1998; Miller, Dir, Gentile, Wilson, Pryor, & sism is pathological (Kernberg & Caligor, Campbell, 2010) or nonexistent relationships 2005). For instance, Bushman, Baumeister, and between narcissism and secondary psychopathy colleagues (e.g., Bushman et al., 2009; Bushman (Benning et al., 2005, Study 2; Hart & Hare, & Baumeister, 1998) have provided evidence that 1989). The mixed findings across studies may the grandiose features of narcissism can promote be because the grandiose and vulnerable dimen- aggression, especially when an individual’s self- sions of narcissism were not differentiated. In- esteem is threatened. The empirical evidence on deed, with a few exceptions (e.g., Miller et al., the relationship between vulnerable narcissism 2010), existing research has focused on grandi- ose narcissism without giving sufficient atten- and aggression is very limited, but preliminary tion to vulnerable narcissism. data from the Pathological Narcissism Inventory Another way to examine parallels between have indicated that vulnerable narcissism is asso- psychopathy and narcissism dimensions is to ciated with homicidal ideation and self-directed investigate whether grandiose and vulnerable violence, including suicide attempts (Pincus et narcissism show similar relationships to exter- al., 2009). More systematic research is needed nal correlates as primary and secondary psy- to determine whether vulnerable narcissism, chopathy, respectively. Epidemiological re- like secondary psychopathy, relates to reactive search has indicated that narcissistic personality forms of aggression and self-harm, whereas disorder evidences high comorbidity with inter- grandiose narcissism, like primary psychopa- nalizing disorders (20.6%, major depressive thy, is associated with mostly unprovoked acts disorder; 54.7%, anxiety disorders) and exter- of violence. 116 SCHOENLEBER, SADEH, AND VERONA

Present Study We expected grandiose narcissism to be associ- ated with fewer psychopathology symptoms, par- Given the similarities between the narcissism ticularly internalizing, and higher unprovoked and psychopathy constructs and the lack of sys- aggression, whereas we expected vulnerable nar- tematic research on the narcissism dimensions, cissism to be associated with heightened emo- we were interested in further delineating the tional problems, comorbid psychopathology, and relationships between grandiose and vulnerable reactive aggression. narcissism and the facets of psychopathy. In our effort to examine the parallels between narcis- sism and psychopathy, we investigated not only Method whether the facets of psychopathy were differ- Participants and Procedures entially associated with dimensions of narcis- sism but also whether the latter would show Participants consisted of 343 men (67.3%) and relationships to external criteria (e.g., psycho- women (32.7%) ranging in age from 18 to 61 logical disorders, forms of aggression) that (M ϭ 30.23, SD ϭ 9.0). To enhance recruitment would be expected if grandiose and vulnerable of externalizing individuals, we recruited adults narcissism paralleled primary and secondary (a) incarcerated in county jails (n ϭ 140; 40.8%), psychopathy, respectively. (b) in substance use treatment (n ϭ 27; 7.9%), and Research on narcissism to date has relied (c) on parole or probation or with a history of legal almost exclusively on a diagnosis of narcissism, conviction (n ϭ 176; 51.3%). Despite the different which confounds grandiose and vulnerable nar- recruitment sites, the three subsamples evidenced cissism, or on the Narcissistic Personality In- similar demographic characteristics and similar ventory (NPI; Raskin & Hall, 1979, 1981), histories of violence and legal involvement.1 which measures only grandiose aspects of nar- About half of the participants were African Amer- cissism. We sought to remedy this limitation by ican (n ϭ 186, 54.2%), followed by Caucasian assessing both grandiose and vulnerable narcis- (n ϭ 131, 38.2%), mixed ethnicity/other (n ϭ 17, sism using two separate measures designed spe- 5.0%), and Hispanic (n ϭ 9, 2.6%). The sample cifically to assess these dimensions—the 16- was diverse in terms of education level, with ap- item version of the NPI (NPI–16; Ames, Rose, proximately one third (35.6%) not completing & Anderson, 2006) and the Hypersensitive Nar- high school (n ϭ 122), 20.7% earning a high cissism Scale (HSNS; Hendin & Cheek, 1997), school diploma (n ϭ 71), 33.8% completing some respectively. Another contribution of this study college (n ϭ 116), 5.5% holding a bachelor’s involves examining relationships between nar- degree (n ϭ 19), 4.1% currently in college (n ϭ cissism and psychopathy in a forensic sample 14), and .3% obtaining a graduate-level degree recruited through legal and treatment agencies (n ϭ 1). Individuals with a lifetime diagnosis of a in the community, instead of relying on a con- psychotic (non–substance induced), bipolar, or venience sample as has been typical in research pervasive developmental disorder were ineligible on dimensions of narcissism. More important, a criminally involved sample should evidence greater variability in the symptoms and behav- 1 Because regression analyses indicated that recruitment iors of interest (narcissism, psychopathy, ag- sample (county jail, probation or parole, substance use center) did not interact with the PCL: SV facets to predict gression) than would a typical undergraduate grandiose (|␤s| Ͻ .12, ps Ͼ .50) or vulnerable (|␤s| Ͻ .24, sample. ps Ͼ .13) narcissism, all analyses were collapsed across We expected that the two dimensions of nar- recruitment samples. We found no group differences on the cissism would be differentially associated with the main study variables between those recruited through the facets of psychopathy, with grandiose narcissism substance use treatment center and those recruited while on parole or probation. However, individuals recruited while in showing relationships to the facets of primary jail had mean levels on grandiose narcissism, as well as on psychopathy (interpersonal and affective) and vul- the affective and antisocial facets of psychopathy, that were nerable narcissism showing relationships to the significantly higher than those of individuals recruited while facets of secondary psychopathy (lifestyle and an- on parole or probation or through substance use treatment centers (ts ϭ 2.1–3.5, all ps Ͻ .05). Jail inmates were also tisocial). We further hypothesized that the two lower on vulnerable narcissism and the lifestyle facet of dimensions of narcissism would parallel research psychopathy relative to the substance use sample (ts ϭ 2.3 on psychopathy in terms of external correlates. and 2.3, respectively, ps Ͻ .05). NARCISSISM AND PSYCHOPATHY 117 to participate because the acute effects of these The number of narcissistic responses endorsed disorders can artificially inflate scores on mea- was summed to create a NPI–16 total score. We sures of psychopathy (e.g., antisocial behavior obtained a Cronbach’s alpha of .72 for the total during mania). Those who participated but were score in the present sample, which is compara- later found to have one of these exclusionary ble to that of the validation studies (␣s ϭ .65– criteria were omitted from analysis (fewer than .78; Ames et al., 2006). 3% of participants). In addition, due to missing The HSNS (Hendin & Cheek, 1997) consists data on some of our variables of interest, the of 10 items that were specifically selected to number of participants available for the different assess vulnerable or covert forms of narcissism. analyses varied from N ϭ 320 to 342. We used the HSNS in this study because items Participants were administered question- on this scale related more to hypersensitivity to naires to measure narcissistic personality fea- rejection (“My feelings are easily hurt by ridi- tures, a semistructured diagnostic interview to cule or the slighting remarks of others”) than to assess for Axis I disorders, and a separate life self-centeredness or grandiosity. In fact, the history interview to assess psychopathy, antiso- HSNS is expected to be generally uncorrelated cial personality disorder, conduct disorder, and with measures of grandiose narcissism, and this lifetime history of aggression and violence. The lack of correlation has been found in previous interviews and self-report questionnaires were work (e.g., r ϭ .02, Hendin & Cheek, 1997). administered during one session that lasted ap- For the HSNS, participants were asked to indi- proximately 2–4 hr. Participants provided in- cate how often they generally experience these formed consent, and the study was approved by feelings on a scale ranging from 1 (very unchar- the university’s Institutional Review Board be- acteristic or very strongly disagree)to5(very fore data collection. characteristic or very strongly agree). Re- sponses were summed to create a HSNS total Independent Measures of the Narcissism score (␣ϭ.75). Dimensions

The NPI–16 (Ames, et al., 2006) was created Psychopathy Facets from the longer 40-item NPI (Raskin & Terry, The Psychopathy Checklist: Screening Ver- 1988) to index the grandiose narcissism do- sion (PCL: SV; Hart, Cox, & Hare, 1995) is a mains of self-ascribed authority, superiority, en- measure designed to index psychopathic traits titlement, and self-absorption (Ames et al., in samples without access to institutional files, 2006). Ames et al. (2006) found that the NPI–16 such as prison records. Although we conducted had good face, predictive, and discriminant va- a review of public criminal records, use of the lidity and is thus a good alternative to the NPI PCL:SV (rather than the full Psychopathy for use in a long assessment protocol such as the Checklist—Revised) was necessary because the one conducted for this study. As would be ex- information gathered from these sources was pected, the NPI shows positive relationships to limited. Information about psychopathic traits measures of narcissistic personality, such as the was collected from a semistructured interview scales from the Schedule for Nonadaptive and and used to rate participants on 12 psychopathic Adaptive Personality (r ϭ .52), the Minnesota traits. Criminal history information provided by Multiphasic Inventory–2 (r ϭ .62), the Millon the participant was verified from the public Clinical Multiaxial Inventory–III (r ϭ .58), and criminal records review (online county, state, the Personality Diagnostic Questionnaire–4 (r ϭ and federal criminal records), which were un- .29 [Samuel & Widiger, 2008] and rs ϭ .20 and 2 available for 11.4% of the sample. Each trait .43 [Miller & Campbell, 2008]). For each item on the NPI, participants were asked to read two statements and choose the one that best de- 2 We examined whether the absence of collateral in- scribed them. One choice represents a narcissis- formation in the form of public records affected the tic response (“I know that I am good, because relationship of the PCL:SV facets and narcissism dimen- sions. Regression analyses indicated that access to public everybody keeps telling me so”) and the other a record information did not interact with the PCL:SV non-narcissistic response (“When people com- facets to predict grandiose (|␤s| Ͻ .25, ps Ͼ .32) or pliment me, I sometimes get embarrassed”). vulnerable (|␤s| Ͻ .22, ps Ͼ .25) narcissism. 118 SCHOENLEBER, SADEH, AND VERONA on the PCL:SV was rated on a 3-point scale psychological disorders of major depressive ranging from 0 (not at all characteristic)to2 disorder (MDD), generalized anxiety disorder (extremely characteristic) by trained doctoral (GAD), lifetime alcohol dependence (AD), and students or doctoral-level raters. We calculated lifetime drug dependence (DD) as described in a total PCL:SV psychopathy score by summing the Diagnostic and Statistical Manual of Mental across ratings for these 12 traits. Because re- Disorders (4th ed., text revision, or DSM–IV– search (e.g., Hill et al., 2004) has indicated that TR; APA, 2000). Interviews were conducted by the PCL:SV has as many as four facets, with the trained doctoral students or doctoral-level rat- interpersonal and affective facets representing ers. Interrater reliability was available for 26% aspects of primary psychopathy and the lifestyle of interviews, which resulted in the following and antisocial facets representing aspects of sec- intraclass correlations for symptom counts: ondary psychopathy, we calculated separate MDD ϭ .94, GAD ϭ .91, AD ϭ .98, and DD ϭ scores for the interpersonal, affective, lifestyle, .98. For the analyses in this study, we used and antisocial facets. The interpersonal facet symptom counts for these Axis I disorders and score included ratings for superficial charm, AAB and CD (described earlier) to create two grandiosity, and deceitfulness/conning. The af- composite psychopathology scores—an inter- fective facet included lack of remorse, lack of nalizing symptom composite and an externaliz- empathy/shallow affect, and failure to assume ing symptom composite. The internalizing responsibility for antisocial behavior. The life- symptom composite consisted of symptom style facet included impulsivity/need for stimu- counts for (a) current MDD, ranging from 0 lation, irresponsibility, and lack of reasonable to 9, and (b) current GAD, ranging from 0 to 3. goals. Finally, the antisocial facet included lack of The externalizing symptom composite consists behavioral control and antisocial behaviors during of symptom counts for AD, DD, and AAB, all adolescence and adulthood. Interrater reliability ranging from 0 to 7, as well as past CD, ranging was available for 20% of interviews, which re- from 0 to 15. To ensure that all of these disor- sulted in an intraclass correlation of .98 for PCL: ders received equal weight in the calculation of SV total score and of .95, .94, .85, and .97 for the their respective composite scores, we trans- interpersonal, affective, lifestyle, and antisocial formed the symptom counts to z scores before facets, respectively. summing. Antisocial personality disorder (APD) and Of the participants in this sample for whom conduct disorder (CD) were also assessed using we had complete data on these measures (N ϭ interview data, and threshold symptom counts 321), 24 (7.5%) met criteria for current MDD, (i.e., the number of symptoms endorsed at and 22 (6.9%) met criteria for current GAD. As threshold levels) of adult antisocial behavior would be expected from a sample of individuals (AAB) and CD were coded separately. Interra- with a criminal history, seeking substance use ter reliability was available for 20% of inter- treatment, or both, a substantial number of par- views. Intraclass correlations in the sample for ticipants met criteria for substance use disor- AAB and CD symptom counts (our primary ders. In this sample, 141 (43.9%) met criteria measures) were .93 and .96, respectively. APD for lifetime AD, and 192 (59.8%) met criteria and CD were reliably diagnosed, as evidenced for lifetime DD. in the diagnostic concordance rates of 96% for APD and 98% for CD across raters. As would Other- and Self-Directed Aggression be expected in a largely forensic sample, of the 335 participants with complete data on these Forms of Aggression Questionnaire (FOA; measures, 174 individuals (53.4%) met criteria Verona, Sadeh, Case, Reed, & Bhattachar- for past CD and 164 (51.3%) met criteria for jee, 2008). The FOA consists of 40 items that APD. measure engagement in various forms of reactive aggression, including physical aggression (“I start Symptoms of Psychopathology fights”; ␣ϭ.92), verbal aggression (“I curse them out”; ␣ϭ.89), relational aggression (“I ruin their We used the Structured Clinical Interview for friendships with other people”; ␣ϭ.85), passive– DSM–IV–TR Axis I Disorders (First, Spitzer, rational aggression (“I take my time doing Gibbon, & Williams, 2002) to assess for the things they want me to do, just to show them”; NARCISSISM AND PSYCHOPATHY 119

␣ϭ.85), and aggression directed toward prop- rated by a diagnostic interviewer on a scale erty (“I start a fire that causes damage”; ␣ϭ ranging from 0 to 5 according to how frequently .83). Participants rated each item on a 5-point the individual had engaged in that type of be- scale ranging from almost never to always to havior since the age of 13 (ranging from never convey how frequently they act aggressively to too many times to count). Participants were when they are upset or angry. Participants in asked to report the number of times they (a) this study endorsed range of severity on the attacked another person without provocation FOA subscales; for example, scores on the (e.g., “jumping” someone); (b) injured them- Physical Aggression subscale ranged from 6 selves on purpose without the intention of kill- to 38 (out of a possible 40). Research has indi- ing themselves (e.g., burning, cutting); and (c) cated that the FOA shows convergent validity attempted to kill themselves (whether or not it with other measure of aggression, including the occurred during a depressive episode). Many of Aggression Questionnaire (Buss & Warren, the participants in the present sample reported 2000; r ϭ .65 with FOA total score), and taps engaging in these behaviors frequently; for ex- into personality traits related to low agreeable- ample, 49.6% reported engaging in unprovoked ness, low conscientiousness, and high neuroti- assaults, 18.3% reported self-harming, cism (Verona et al., 2008). Thus, this measure and 21.5% reported attempting suicide. assesses mostly hostile or reactive aggression that is manifested in various forms (e.g., phys- Results ical, verbal, relational). Life History of Aggression (LHA; Coc- We examined the relationships among the caro, Berman, & Kavoussi, 1997). The dimensions of narcissism and the four facets of LHA was developed as an interview-based mea- psychopathy to begin to assess convergence sure of trait aggressive behavior, with research among these constructs. Zero-order correla- showing that it predicts overt aggression and tions, and descriptive statistics for these vari- personality disorders characterized by suicidal- ables, can be found in Table 1. Grandiose and ity and violence (Coccaro et al., 1997). In this vulnerable narcissism were not significantly as- study, the LHA was used to assess unprovoked sociated with each other (r ϭ .08, ns), consis- (nonreactive) violence against others, self- tent with previous research using these two harm, and suicide attempts, constructs not ex- measures (Hendin & Cheek, 1997) and suggest- plicitly assessed with the preceding measures of ing they tap into independent dimensions of psychopathology or aggression. Items were narcissism. These dimensions also showed differ-

Table 1 Descriptive Statistics and Correlations Between Narcissism Dimensions and Psychopathy Facets Narcissism PCL:SV Facets Variables Grandiose Vulnerable Interpersonal Affective Lifestyle Antisocial Total Narcissism ءء19. ء16. 06. ءء14. ءءGrandiose — .08 .21 ء14. 09. ءءVulnerable — Ϫ.04 .10 .27 PCL: SV facets ءء71. ءء33. ءء30. ءءInterpersonal — .49 ءء81. ءء56. ءءAffective — .40 ءء68. ءءLifestyle — .42 ءءAntisocial — .79 Total — M 4.5 27.1 2.2 2.6 3.5 3.5 11.8 SD 3.1 7.4 1.7 1.5 1.5 1.8 4.8 Minimum 0 10 0 0 0 0 1 Maximum 14 50 6 6 6 6 22 Note. N ϭ 340. PCL:SV ϭ Psychopathy Checklist: Screening Version. .p Ͻ .01 ءء .p Ͻ .05 ء 120 SCHOENLEBER, SADEH, AND VERONA ential relationships to psychopathic personality with the parole or probation sample not differing facets. The interpersonal, affective, and antisocial from the other two samples. facets of psychopathy were all associated with As for psychopathy effects, analyses revealed grandiose narcissism at the zero-order level that the interpersonal facet was primarily re- (rs ϭ .21, .14, and .16, respectively, all ps Ͻ sponsible for the relationship between psychop- .01), whereas the lifestyle facet was solely as- athy and grandiose narcissism (␤ϭ.24, p Ͻ sociated with vulnerable narcissism (r ϭ .27, .001). Therefore, grandiose narcissism was, as p Ͻ .01). expected, associated with psychopathic traits of charm, grandiosity, and duplicity. In contrast, Psychopathy Facets Contribute to elevations in vulnerable narcissism were posi- Narcissism Dimensions tively associated with the lifestyle facet (␤ϭ .24, p Ͻ .001), indicating that vulnerable nar- To determine whether the four psychopathy cissism is characterized primarily by impulsiv- facets would uniquely contribute to grandiose ity, irresponsibility, and unrealistic goals. More or vulnerable narcissism after taking demo- important, we found a similar pattern of results graphic variables into account, we ran separate across the three recruitment samples, although hierarchical multiple regressions for each of the the results for the sample recruited through the narcissism dimensions. We entered age, gender, substance use treatment center were all nonsig- and recruitment sample in the first step and the nificant because of the small sample size (N ϭ four PCL:SV facets in the second step. Results 27). In particular, grandiose narcissism showed are displayed in Table 2. Age was a significant a positive relationship to the interpersonal facet predictor of grandiose narcissism, with younger in all three samples (␤s ϭ .23–.25). Overall, participants scoring higher on the NPI–16. For vulnerable narcissism tended to be positively vulnerable narcissism, both gender and recruit- associated with the lifestyle facet (␤s ϭ .34 and ment sample were significant predictors. Women .35 for those on parole or probation and those in evidenced greater levels of vulnerable narcissism, substance use treatment, respectively), although and individuals recruited through the substance this association was somewhat weaker and non- use treatment center evidenced greater vulnerable significant for participants recruited while in jail narcissism than did those recruited while in jail, (␤ϭ.12).

Psychopathology and Aggression Correlates Table 2 of Grandiose and Vulnerable Narcissism Summary of Hierarchical Multiple Regression Analyses With Psychopathy Facets Predicting Correlations between internalizing and exter- Dimensions of Narcissism nalizing symptom composites and the two dimen- Grandiose Vulnerable sions of narcissism are presented in Table 3. Also narcissism narcissism in Table 3 are the correlations between several forms of aggression, as measured by the LHA, R2 or R2 or Variable ß ⌬R2 ß ⌬R2 the FOA, and the two narcissism dimensions. We determined whether the correlations for ءء ءء Step 1 .06 .05 grandiose narcissism were significantly differ- Ϫ.06 ءءAge Ϫ.22 -ent from those for vulnerable narcissism, fol ءءGender .10 Ϫ.17 -lowing Steiger (1980). Because of multiple test ءءRecruitment sample .06 Ϫ.14 ing and because these analyses were exploratory ءء09. ءءStep 2 .06 Ϫ.11 in nature, we applied a Bonferroni correction to ءءInterpersonal facet .24 Affective facet Ϫ.01 .10 the alpha level for the correlations depicted in ءء Lifestyle facet .01 .24 Table 3. We used p values of .002 and .0003 Antisocial facet .03 .05 (i.e., .05/32 and .01/32, respectively) for the Note. N ϭ 340. Gender was coded as follows: female ϭ 0, zero-order correlations and .003 and .0006 (i.e., male ϭ 1. Sampling group was coded as follows: recruited .05/16 and.01/16, respectively) for the compar- from local jails ϭ 1, recruited while on parole or probation or with history of incarceration ϭ 0, recruited through isons of these correlations. substance use treatment ϭϪ1. As shown in Table 3, only vulnerable narcis- -p Ͻ .001. sism was associated with the internalizing com ءء .p Ͻ .05 ء NARCISSISM AND PSYCHOPATHY 121

Table 3 Psychopathology, Suicidality, and Violence Correlates of Grandiose and Vulnerable Narcissism Grandiose Vulnerable External Criterion Variables narcissism narcissism t (difference) ءء5.9 ءءInternalizing symptoms Ϫ.03 .40 ءء4.8 ءءMajor depressive disorder .01 .34 ءء5.2 ءءGeneralized anxiety disorder Ϫ.07 .32 1.6 ءء24. ءExternalizing symptoms .12 ء3.2 ءءAlcohol dependence Ϫ.00 .24 ءء3.3 ءءIllicit drug dependence Ϫ.02 .23 2.4 02. ءءConduct disorder .20 27. ءء16. ءAdult antisocial behavior .14 Life history of aggression 2.0 ءءAssaults on self .04 .19 ء3.2 ءءSuicide attempts Ϫ.04 .20 2.4 03. ءءUnprovoked assaults on others .19 Forms of Aggression Questionnaire ءء3.5 ءءVerbal aggression .10 .36 ءء3.9 ءءRelational aggression .10 .38 ءء3.3 ءء37. ءPassive–rational aggression .13 1.5 ءء23. ءءAggression against property .15 40. ءء21. ءءPhysical aggression .25 Note. Ns ϭ 320–342. p Ͻ .0003 for zero-order correlations and .0006 for ءء .p Ͻ .002 for zero-order correlations and .003 for differences in t ء differences in t. posite and MDD and GAD symptom counts, vulnerable narcissism was associated with acts with all of these correlations being significantly of self-harm according to the LHA, in the form different from those seen for grandiose narcis- of both nonsuicidal self-injury and suicide at- sism. Vulnerable narcissism was also associated tempts. Moreover, the correlation between sui- with the externalizing composite, specifically cide attempts and vulnerable narcissism was with lifetime alcohol and drug dependence. significantly greater than that for grandiose nar- More important, the correlations between vul- cissism. As for other-directed aggression, vul- nerable narcissism and drug and alcohol use nerable narcissism was associated with all symptom counts were significantly greater than forms of reactive aggression on the FOA, and those for grandiose narcissism. Grandiose nar- its associations with verbal, relational, and pas- cissism was instead associated with past CD, sive–rational aggression were significantly and this correlation did not differ between the greater than those seen for grandiose narcis- two narcissism dimensions. Both grandiose and sism. In contrast, elevations in grandiose nar- vulnerable narcissism were also significantly cissism were associated with more frequent un- related to AAB symptom counts. In summation, provoked assaults against other people on the vulnerable narcissism was more strongly asso- LHA and physical aggression (against people and ciated with other psychopathology, both inter- property) on the FOA.3 More important, even nalizing and substance use disorders, although when psychopathy facets are taken into account, grandiose narcissism was associated with child- all but one of the grandiose and vulnerable nar- hood conduct problems and AAB. Moreover, cissism relations to unprovoked attacks and reac- the associations between vulnerable narcissism and internalizing and substance use disorders tended to be significantly greater than those 3 We also considered whether age may be affecting the between grandiose narcissism and internalizing correlations between the narcissism dimensions and the and substance use disorders. aggression variables by separating our sample into two groups, individuals age 30 and younger (N ϭ 197) and those Finally, we examined the dimensions of nar- older than age 30 (N ϭ 146). Following Steiger (1980), a cissism in relation to both self- and other- comparison of the narcissism–aggression correlations for directed aggression. As depicted in Table 3, younger and older adults revealed no significant differences. 122 SCHOENLEBER, SADEH, AND VERONA tive aggression, respectively, remain. Only the both are characterized by superficial charm, very modest relationship between grandiose nar- haughty arrogance, and interpersonal duplicity, cissism and passive–rational aggression is nonsig- whereas vulnerable narcissism and secondary nificant after taking psychopathy into account. psychopathy share traits related to impulsivity Thus, vulnerable narcissism shows a stronger re- and irresponsibility in criminally involved lationship with self-harm and aggressive traits, adults. Although modest in size, these congru- particularly those that are reactive in nature, ous associations between the narcissism dimen- even when psychopathy facets are accounted sions and psychopathy facets suggest that the for. In contrast, although grandiose narcissism larger research literature on psychopathic per- showed some association with reactive physical sonality disorder can serve to inform the dis- aggression, it was not associated with self- tinctions between the grandiose and vulnerable directed aggression and was instead associated narcissism dimensions in externalizing adults. with unprovoked assaults on others—a relation- In particular, our results suggest that both ship not seen for vulnerable narcissism. narcissism and psychopathy share vulnerabili- ties that can be conceptualized in terms of broader theoretical constructs in the personality Discussion literature, including a recent theoretical formu- This study is one of the first to examine the lation of antisocial/psychopathic tendencies by relations between dimensions of narcissism and Patrick, Fowles and Krueger (2009). In their facets of psychopathy in a forensic sample, in model, three phenotypic components associated which these disorders are frequently diagnosed with antisocial/psychopathic tendencies are ad- and cause impairments in functioning, espe- vanced, including disinhibition (problematic cially in regard to illegal behavior. Conse- impulse control), meanness (callousness, cold- quently, the findings contribute to the literature heartedness, antagonism), and boldness (social by providing novel information regarding the dominance and emotional resiliency). In this extent to which narcissism and psychopathy regard, grandiose narcissism appears to relate to converge in a high-risk sample, research that the construct of boldness because we found that has implications for understanding the etiology grandiose narcissism was associated with social of and designing treatments for these disorders. dominance (e.g., charm and duplicity) and Our research supports studies that have found fewer problems in psychological functioning an association between grandiose narcissism (i.e., internalizing symptoms, alcohol and drug and primary psychopathy (Benning et al., 2005; problems, and self-directed violence) compared McHoskey et al., 1998) and advances knowl- with vulnerable narcissism, which parallels re- edge of vulnerable narcissism by informing its search on the differential correlates of the psy- parallels with secondary psychopathy, a rela- chopathy facets (Patrick et al., 2005; Verona et tively understudied area of research. al., 2001, 2005). By contrast, vulnerable narcis- sism seems to be marked by high levels of disinhibition. Like secondary psychopathy (Hill Implications for Personality Disorder et al., 2004; Skeem & Mulvey, 2001), vulnera- Research ble narcissism showed robust associations with outcomes generally associated with impulsivity, As predicted, the two forms of narcissism including reactive aggression and substance showed differential relationships to the psy- use. Both narcissism dimensions contain charac- chopathy facets, with grandiose narcissism re- teristics of meanness (e.g., callousness, coldheart- lating positively to the interpersonal facet and edness, antagonism), with grandiose narcissism vulnerable narcissism relating positively to the demonstrating more proactive (and perhaps more lifestyle facet. The divergent findings for gran- callous) forms and vulnerable narcissism showing diose and vulnerable narcissism underscore the more reactive forms of this trait. This type of need to disentangle the heterogeneous nature of conceptualization would be another step toward narcissistic personality disorder as manifested dimensional or spectrum models of personality to in externalizing individuals. Specifically, our characterize the overlap across the personality dis- results suggest that grandiose narcissism and order categories currently in the DSM–IV–TR. primary psychopathy converge to the extent that These results provide preliminary evidence that NARCISSISM AND PSYCHOPATHY 123 narcissism and psychopathy share vulnerabilities (Dolan & Park, 2002; Morgan & Lilienfeld, potentially consistent with dimensional models of 2000). Secondary psychopathy is theorized to personality. be conceptually similar to the externalizing spectrum of psychopathology (Krueger et al., Parsing the Etiology of Narcissism 2002; Patrick et al., 2005) and to share etiolog- via Psychopathy ical vulnerabilities with disorders on this spec- trum, including APD and substance use. Vul- Extensive work has been done on the etiology nerable narcissism may possibly be a related of psychopathy, with research suggesting dis- indicator of this latent externalizing dimension tinct etiological pathways for the primary and (e.g., Miller et al., 2010). A putative endophe- secondary manifestations of the disorder. For notype of the externalizing spectrum is reduced instance, primary psychopathy traits are theo- P300 amplitude (Iacono, Malone, & McGue, rized to result from deficits in the emotional 2003; Patrick et al., 2001), an event-related po- circuitry of the brain postulated to modulate the tential index of working memory functioning experience of fear (e.g., amygdala, paralimbic that is thought to have neural generators in the system; Birbaumer et al., 2005; Blair, 2005; lateral prefrontal cortex (Nieuwenhuis, Aston- Kiehl, 2006), and the allocation of attentional Jones, & Cohen, 2005) and anterior cingulate resources to motivationally salient stimuli to the cortex (Dien, Spencer, & Donchin, 2003). One neglect of contextual stimuli (Newman, 1998; potential avenue for future research is to explore Sadeh & Verona, 2008). To the extent that the the extent to which individuals high on vulner- construct of grandiose narcissism measures able narcissism evidence reduced P300 ampli- similar pathology to that of primary psychopa- tude, which would shed light on the etiological thy in criminally involved individuals, the same overlap between vulnerable narcissism and ex- or similar etiological mechanisms may be at ternalizing disorders. play for both diagnoses but result in different manifestations depending on other traits and Strengths, Limitations, and Clinical developmental pathways. Future research can Implications address this question by further examining whether the grandiose dimension of narcissism This study has several strengths, including is associated with comparable behavioral and the investigation of narcissistic and psycho- psychophysiological abnormalities as primary pathic personality traits in a large sample of psychopathy, such as deficient aversive condi- externalizing adults in whom extremes on these tioning (Flor, Birbaumer, Hermann, Ziegler, & personality features are prevalent and have se- Patrick, 2002), anticipation of threat (Hare, rious consequences for both the individual and 1965), attentional processing (Hiatt, Schmitt, & society. Moreover, we were able to examine Newman, 2004; Sadeh & Verona, 2008), and several external criteria in relation to the lesser startle reactivity to threatening cues (Levenston, studied narcissism dimensions, which provides Patrick, Bradley, & Lang, 2000; Patrick et al., insight into the nomological network associated 1993). Although preliminary research is being with the constructs of grandiose and vulnerable conducted with undergraduate samples on the dimensions. psychophysiological correlates of grandiose As with any investigation, however, this narcissism versus antisocial traits (Kelsey, Orn- study also has limitations. First, our sample duff, McCann, & Reiff, 2001; Kelsey, Ornduff, included more men than women, which may Reiff, & Arthur, 2002; Sylvers, Brubaker, Al- have overrepresented the manifestation of nar- den, Brennan, & Lilienfeld, 2008), additional cissism and psychopathy in men. Although the research is needed to fully explicate these rela- number of men in this sample is disproportion- tionships in ecologically valid samples. ate to the general population, our recruitment In contrast to the primary variant, research procedures oversampled women relative to their has indicated that secondary psychopathic traits representation in the legal system (Federal Bu- are associated with deficits in executive func- reau of Investigation, 2008). Second, the use of tioning and cognitive control (Raine et al., self-report data for our measure of the narcis- 1998; Sellbom & Verona, 2007), potentially sism dimensions is a limitation (e.g., Oltmanns governed by the dorsolateral prefrontal cortex & Strauss, 1998; Oltmanns & Turkheimer, 124 SCHOENLEBER, SADEH, AND VERONA

2009), and future research can include struc- facets of psychopathy from normal personality tured diagnostic assessments of narcissism traits: A step toward community epidemiological symptoms. Our use of clinician-rated psychop- investigations. Assessment, 12, 3–18. athy and psychopathology assessments, how- Birbaumer, N., Veit, R., Lotze, M., Erb, M., Her- ever, helped limit the extent to which common mann, C., Grodd, W., & Flor, H. (2005). Deficient fear conditioning in psychopathy: A functional method variance contributed to the associations magnetic resonance imaging study. Archives of between the narcissism dimensions and psy- General Psychiatry, 62, 799–805. chopathy facets. Blackburn, R. (1975). An empirical classification of Despite these limitations, this study’s results psychopathic personality. British Journal of Psy- provide important information about the extent chiatry, 127, 456–460. to which narcissism and psychopathy converge Blair, R. J. R. (2005). Applying a cognitive neuro- and are potentially mutually informative in ex- science perspective to the disorder of psychopathy. ternalizing samples. Our findings also have im- Development and Psychopathology, 17, 865–891. plications for clinicians treating individuals Bushman, B. J., & Baumeister, R. F. (1998). Threat- with narcissistic traits. If, as the findings sug- ened egotism, narcissism, self-esteem, and direct gest, grandiose and vulnerable narcissism and and displaced aggression: Does self-love or self- hate lead to violence? Journal of Personality and primary and secondary psychopathy manifest Social Psychology, 75, 219–229. similarly among externalizing individuals, cli- Bushman, B. J., Baumeister, R. F., Thomaes, S., Ryu, nicians attempting to treat grandiose narcissistic E., Begeer, S., & West, S. G. (2009). Looking traits may want to emphasize building empathy again, and harder, for a link between low self- for others and decreasing interpersonal exploi- esteem and aggression. Journal of Personality, 77, tation. In contrast, those treating vulnerable nar- 427–446. cissism may want to focus on controlling emo- Buss, A. H., & Warren, W. L. (2000). Aggression tional outbursts and impulsive behavior, as well questionnaire. Los Angeles, CA: Western Psycho- as the comorbid symptoms of depression that logical Services. are often observed in those with vulnerable nar- Cale, E. M., & Lilienfeld, S. O. (2002). Histrionic personality disorder and antisocial personality dis- cissism. A divergence in coping strategies used order. Journal of Personality Disorders, 16, 52– by vulnerable and grandiose narcissists could 72. explain the differential risk for internalizing Cheek, J. M., & Hendin, H. M. (1996, August). psychopathology, and this information on cop- Shyness, shame, and narcissism. In D. Paulhus ing strategies can be incorporated in treatment (Chair), The social cognition of shyness and social interventions that decrease emotional distress in anxiety. Symposium conducted at the meeting of the former and increase emotional engagement the American Psychological Association, Toronto, in the latter. Thus, this study advances the lit- Ontario, Canada. erature by illuminating commonalities in the Cleckley, H. (1976). The mask of sanity (5th ed.). St. manifestation and potential treatment of person- Louis, MO: Mosby. ality pathology in externalizing adults and sug- Coccaro, E. F., Berman, M. E., & Kavoussi, R. J. (1997). Assessment of life-history of aggression: gests that shared vulnerabilities may underlie Development and psychometric characteristics. narcissism and psychopathy in criminally in- Psychiatry Research, 73, 147–157. volved individuals. Cooke, D. J., & Michie, C. (2001). Refining the construct of psychopathy: Towards a hierarchical References model. Psychological Assessment, 13, 171–188. Dien, J., Spencer, K. M., & Donchin, E. (2003). Akhtar, S., & Thomson, J. A. (1982). Overview: Localization of the event-related potential novelty Narcissistic personality disorder. American Jour- response as defined by principal components anal- nal of Psychiatry, 139, 12–20. ysis. Cognitive Brain Research, 17, 637–650. American Psychiatric Association. (2000). Diagnos- Dolan, M., & Park, I. (2002). The neuropsychology tic and statistical manual of mental disorders (4th of antisocial personality disorder. Psychological ed., text rev.). Washington, DC: Author. Medicine, 32, 417–427. Ames, D. R., Rose, P., & Anderson, C. P. (2006). The Edens, J. F., Hart, S. D., Johnson, D. W., Johnson, NPI-16 as a short measure of narcissism. Journal J. K., & Olver, M. E. (2000). Use of the Person- of Research in Personality, 40, 440–450. ality Assessment Inventory to assess psychopathy Benning, S. D., Patrick, C. J., Blonigen, D. M., in offender populations. Psychological Assess- Hicks, B. M., & Iacono, W. G. (2005). Estimating ment, 12, 132–139. NARCISSISM AND PSYCHOPATHY 125

Federal Bureau of Investigation. (2008). Crime in the Iacono, W., Malone, S., & McGue, M. (2003). Sub- United States, 2007. Washington, DC: U.S. De- stance use disorders, externalizing psychopathol- partment of Justice. ogy and P300 event-related potential amplitude. First, M. B., Spitzer, R. L., Gibbon, M., & Williams, International Journal of Psychophysiology, 48, J. B. W. (2002). Structured Clinical Interview for 147–178. DSM–IV–TR Axis I Disorders (SCID-I). New Jones, D. N., & Paulhus, D. L. (2010). Different York: Biometrics Research, New York State Psy- provocations trigger aggression in narcissists and chiatric Institute. psychopaths. Social Psychological and Personal- Flor, H., Birbaumer, N., Hermann, C., Ziegler, S., & ity Science, 1, 12–18. Patrick, C. J. (2002). Aversive Pavlovian condi- Karpman, B. (1941). On the need of separating psy- tioning in psychopaths: Peripheral and central cor- chopathy into two distinct clinical types: The relates. Psychophysiology, 39, 505–518. symptomatic and the idiopathic. Journal of Crim- Gramzow, R., & Tangney, J. P. (1992). Proneness to inal Psychopathology, 3, 112–137. shame and the narcissistic personality. Personality Kelsey, R. M., Ornduff, S. R., McCann, C. M., & and Social Psychology Bulletin, 18, 369–376. Reiff, S. (2001). Psychophysiological characteris- Gunderson, J. G., Ronningstam, E., & Smith, L. E. tics of narcissism during active and passive coping. (1991). Narcissistic personality disorder: A review Psychophysiology, 38, 292–303. of data on DSM–III–R descriptions. Journal of Kelsey, R. M., Ornduff, S. R., Reiff, S., & Arthur, Personality Disorders, 5, 167–177. C. M. (2002). Psychophysiological correlates of Hare, R. D. (1965). Temporal gradient of fear arousal narcissistic traits in women during active coping. in psychopaths. Journal of Abnormal Psychol- Psychophysiology, 39, 322–332. ogy, 70, 442–445. Kernberg, O. F. (1975). Borderline conditions and Hare, R. D. (1991). Manual for the Hare Psychopa- pathological narcissism. New York, NY: Jason thy Checklist–Revised. Toronto, Ontario, Canada: Aronson. Multi-Health Systems. Kernberg, O. F., & Caligor, E. (2005). A psychoan- alytic theory of personality disorders. In M. F. Hare, R. D. (2003). Psychopathy Checklist–Revised Lenzenweger & J. F. Clarkin (Eds.), Major theo- (2nd ed.). Toronto, Ontario, Canada: Multi-Health ries of personality disorder (2nd ed.; pp. 14–156). Systems. New York, NY: Guilford Press. Hart, S. D. (1998). The role of psychopathy in as- Kiehl, K. A. (2006). A cognitive neuroscience per- sessing risk for violence: Conceptual and method- spective on psychopathy: Evidence for paralimbic ological issues. Legal and Criminological Psy- system dysfunction. Psychiatry Research, 142, chology, 3, 121–137. 107–128. Hart, S. D., Cox, D. N., & Hare, R. D. (1996). Krueger, R. F., Hicks, B. M., Patrick, C. J., Carlson, Manual for the Screening Version of Psychopathy S. R., Iacono, W. G., & McGue, M. (2002). Etio- Checklist Revised (PCL:SV). Toronto: Multi- logic connections among substance dependence, Health Systems. antisocial behavior, and personality: Modeling the Hart, S. D., & Hare, R. D. (1989). Discriminant externalizing spectrum. Journal of Abnormal Psy- validity of the Psychopathy Checklist in a forensic chology, 111, 411–424. psychiatric population. Psychological Assessment: Levenston, G. K., Patrick, C. J., Bradley, M. M., & A Journal of Consulting and Clinical Psychol- Lang, P. J. (2000). The psychopath as observer: ogy, 1, 211–218. Emotion and attention in picture processing. Jour- Hendin, H. M., & Cheek, J. M. (1997). Assessing nal of Abnormal Psychology, 109, 373–385. hypersensitive narcissism: A reexamination of Locke, K. D. (2009). Aggression, narcissism, self- Murray’s Narcissism Scale. Journal of Research in esteem, and the attribution of desirable and human- Personality, 31, 588–599. izing traits to self versus others. Journal of Re- Hiatt, K. D., Schmitt, W. A., & Newman, J. P. search in Personality, 43, 99–102. (2004). Stroop tasks reveal abnormal selective at- Lorber, M. F. (2004). Psychophysiology of aggres- tention in psychopathic offenders. Neuropsychol- sion, psychopathy, and conduct problems: A meta- ogy, 18, 50–59. analysis. Psychological Bulletin, 130, 531–552. Hibbard, S. (1992). Narcissism, shame, masochism, Lykken, D. T. (1995). The antisocial personalities. and object relations: An exploratory correlational Hillsdale, NJ: Erlbaum. study. Psychoanalytic Psychotherapy, 9, 489–508. Lynam, D. R. (2001). Psychopathy from the perspec- Hill, C. D., Neumann, C. S., & Rogers, R. (2004). tive of the five-factor model of personality. In P. T. Confirmatory factor analysis of the Psychopathy Costa & T. A. Widiger (Eds.), Personality disor- Checklist: Screening Version in offenders with ders and the five-factor model of personality (2nd Axis I disorders. Psychological Assessment, 16, ed., pp. 325–348). Washington, DC: American 90–95. Psychological Association. 126 SCHOENLEBER, SADEH, AND VERONA

McHoskey, J. W., Worzel, W., & Szyarto, C. (1998). Initial construction and validation of the Patholog- Machiavellianism and psychopathy. Journal of ical Narcissism Inventory. Psychological Assess- Personality and Social Psychology, 74, 192–210. ment, 21, 365–379. Miller, J. D., & Campbell, W. K. (2008). Comparing Raine, A., Meloy, R., Bihrle, S., Stoddard, J., LaC- clinical and social-personality conceptualizations asse, L., & Buchsbaum, M. (1998). Reduced pre- of narcissism. Journal of Personality, 76, 449– frontal and increased subcortical brain functioning 476. assessed using positron emission tomography in Miller, J. D., Dir, A., Gentile, B., Wilson, L., Pryor, predatory and affective murders. Behavioral Sci- L. R., & Campbell, W. K. (2010). Searching for a ences and Law, 16, 319–332. vulnerable dark triad: Comparing factor 2 psy- Raskin, R., & Hall, C. S. (1979). A narcissistic per- chopathy, vulnerable narcissism, and borderline sonality inventory. Psychological Reports, 45, personality disorder. Journal of Personality, 78, 590. 1529–1564. Raskin, R., & Hall, C. S. (1981). The Narcissistic Morgan, A. B., & Lilienfeld, O. (2000). A meta- Personality Inventory: Alternative form reliability analytic review of the relation between antisocial and further evidence of construct validity. Journal behavior and neuropsychological measures of ex- of Personality Assessment, 45, 159–162. ecutive function. Clinical Psychology Review, 20, 113–156. Raskin, R., & Terry, H. (1988). A principal- Newman, J. P. (1998). Psychopathic behavior: An components analysis of the Narcissistic Personal- information processing perspective. In D. J. ity Inventory and further evidence of its construct Cooke, A. E. Forth, & R. D. Hare (Eds.), Psychop- validity. Journal of Personality and Social Psy- athy: Theory, research and implications for soci- chology, 54, 890–902. ety (pp. 81–104). Dordrecht, the Netherlands: Klu- Rutherford, M. J., Alterman, A. I., Cacciola, J. S., & wer Academic. McKay, J. R. (1996). Validity of the Psychopathy Nieuwenhuis, S., Aston-Jones, G., & Cohen, J. D. Checklist–Revised in male methadone patients. (2005). Decision making, the P3, and the locus Drug and Alcohol Dependence, 44, 143–149. coeruleus-norepinephrine system. Psychological Sadeh, N., & Verona, E. (2008). Psychopathy traits Bulletin, 131, 510–532. and selective impairments in attentional function- Oltmanns, T. F., & Strauss, M. E. (1998). Peer as- ing. Neuropsychology, 22, 669–680. sessment of personality traits and pathology in Samuel, D. B., & Widiger, T. A. (2008). Conver- female college students. Assessment, 5, 53–65. gence of narcissism measures from the perspective Oltmanns, T. F., & Turkheimer, E. (2009). Person of general personality functioning. Assessment, 15, perception and personality pathology. Current Di- 364–374. rections in Psychological Science, 18, 32–36. Sellbom, M., & Verona, E. (2007). Neuropsycholog- Patrick, C. J., Bradley, M. M., & Lang, P. J. (1993). ical correlations of psychopathic traits in a non- Emotion in the criminal psychopath: Startle reflex incarcerated sample. Journal of Research in Per- modulation. Journal of Abnormal Psychology, sonality, 41, 276–294. 102, 82–92. Skeem, J. L., Johansson, P., Andershed, H., Kerr, M., Patrick, C. J., Fowles, D. C., & Krueger, R. F. (2009). & Louden, J. E. (2007). Two subtypes of psycho- Triarchic conceptualization of psychopathy: De- pathic violent offenders that parallel primary and velopmental origins of disinhibition, boldness, and secondary variants. Journal of Abnormal Psychol- meanness. Development and Psychopathology, 21, ogy, 116, 395–409. 913–938. Skeem, J. L., & Mulvey, E. P. (2001). Psychopathy Patrick, C. J., Hicks, B. M., Krueger, R. F., & Lang, and community violence among civil psychiatric A. R. (2005). Relations between psychopathy fac- ets and externalizing in a criminal offender sample. patients: Results from the MacArthur Violence Journal of Personality Disorders, 19, 339–356. Risk Assessment Study. Journal of Clinical and Patrick, C. J., Iacono, W. G., Bernat, E., Malone, Consulting Psychology, 69, 358–374. S. M., Krueger, R. F., & McGue, M. (2001). Brain Skeem, J. L., Poythress, N., Edens, J. F., Lilienfeld, potential correlates of the latent externalizing di- S. O., & Cale, E. M. (2003). Psychopathic person- mension of psychopathology. Psychophysiol- ality or personalities? Exploring potential variants ogy, 38, S75. of psychopathy and their implications for risk as- Paulhus, D. L., & Williams, K. M. (2002). The dark sessment. Aggression and Violent Behavior, 8, triad of personality: Narcissism, Machiavellian- 513–546. ism, and psychopathy. Journal of Research in Per- Smith, S. S., & Newman, J. P. (1990). Alcohol and sonality, 36, 556–563. drug abuse-dependence disorders in psychopathic Pincus, A. L., Ansell, E. B., Pimentel, C. A., Cain, and nonpsychopathic criminal offenders. Journal N. M., Wright, A. G. C., & Levy, K. N. (2009). of Abnormal Psychology, 99, 430–439. NARCISSISM AND PSYCHOPATHY 127

Steiger, J. H. (1980). Tests for comparing elements of Verona, E., Patrick, C. J., & Joiner, T. E. (2001). a correlation matrix. Psychological Bulletin, 87, Psychopathy, antisocial personality, and suicide 245–251. risk. Journal of Abnormal Psychology, 110, 462– Stinson, F. S., Dawson, D. A., Goldstein, R. B., 470. Chou, S. P., Huang, B., Smith, S. M., . . . Grant, Verona, E., Sadeh, N., Case, S. M., Reed, A., & B. F. (2008). Prevalence, correlates, disability, and Bhattacharjee, A. (2008). Self-reported use of dif- comorbidity of DSM–IV narcissistic personality ferent forms of aggression in late adolescence and disorder: Results from the Wave 2 National Epi- emerging adulthood. Assessment, 15, 493–510. demiologic Survey on Alcohol and Related Con- Walsh, Z., & Kosson, D. S. (2008). Psychopathy and ditions. Journal of Clinical Psychiatry, 69, 1033– violence: The importance of factor-level interac- 1045. tions. Psychological Assessment, 20, 114–120. Sylvers, P., Brubaker, N., Alden, S. A., Brennan, Watson, P. J., Sawrie, S. M., Greene, R. L., & Arre- P. A., & Lilienfeld, S. O. (2008). Differential en- dondo, R. (2002). Narcissism and depression: dophenotypic markers of narcissistic and antisocial MMPI-2 evidence for the continuum hypothesis in personality features: A psychophysiological inves- clinical samples. Journal of Personality Assess- tigation. Journal of Research in Personality, 42, ment, 79, 85–109. 1260–1270. Tangney, J. P., & Dearing, R. L. (2002). Shame and Widiger, T. A., & Corbitt, E. M. (1993). Antisocial guilt. New York, NY: Guilford Press. personality disorder: Proposal for DSM–IV. Jour- Vanman, E. J., Mejia, V. Y., Dawson, M. E., Schell, nal of Personality Disorders, 7, 65–77. A. M., & Raine, A. (2003). Modification of the Williamson, S., Hare, R. D., & Wong, S. (1987). startle reflex in a community sample: Do one or Violence: Criminal psychopaths and their victims. two dimensions of psychopathy underlie emotional Canadian Journal of Behavioral Science, 19, 454– processing? Personality and Individual Differ- 462. ences, 35, 2007–2021. Wink, P. (1991). Two faces of narcissism. Journal of Verona, E., Hicks, B. M., & Patrick, C. J. (2005). Personality and Social Psychology, 61, 590–597. Psychopathy and suicidality in female offenders: Woodworth, M., & Porter, S. (2002). In cold blood: Mediating influences of personality and abuse. Characteristics of criminal homicides as a function Journal of Consulting and Clinical Psychol- of psychopathy. Journal of Abnormal Psychology, ogy, 73, 1065–1073. 111, 436–445.