Keywords

Introduction

In most people’s eyes, narcissists are arrogant, selfish, exploitive, entitled, and aggressive. In a word, narcissism is “…‘bad’ and predicts other ‘bad’ things” (Campbell & Foster, 2007, p. 116; Lasch, 1979). Indeed, narcissism has been treated as a pathological disorder ever since its introduction into psychology (Ellis, 1898; Freud, 1914/1957; Kernberg, 1975; Kohut, 1977). Yet, decades of research on narcissism in normal populations has suggested that to some extent and in some aspects, narcissism could also be desirable and adaptive (Back, Schmukle, & Egloff, 2010; Sedikides, Rudich, Gregg, Kumashiro, & Rusbult, 2004; Watson & Biderman, 1993). For example, narcissists tend to be confident, assertive, extraverted, energetic, and happy (Watson & Biderman, 1993), and they are more likely to have high self-esteem and less likely to experience depression and anxiety (Sedikides et al., 2004). Conscious of both the pros and cons associated with narcissism, researchers recently have attempted to distinguish between adaptive and maladaptive narcissism and study them separately (Barry, Frick, Adler, & Grafeman, 2007; Hepper, Hart, & Sedikides, 2014; Hill & Yousey, 1998).

In this chapter, we elaborate on evidence that supports a distinction between adaptive and maladaptive narcissism and discusses the implications therein. In doing so, we focus on narcissism in the normal population (i.e., not the clinical disorder) and a specific variant known as “grandiose narcissism ” (in contrast to “vulnerable narcissism”). Grandiose narcissism is characterized by an inflated self-view, agentic orientation, selfishness, and a sense of specialness (Campbell & Foster, 2007).

Distinction Reflected in the Research Tradition

Exploration about narcissism has followed two traditions: clinically based and personality-based . While clinical psychologists have long treated narcissism as a pathological disorder that concerns clinical populations (Kernberg, 1975; Kohut, 1977; Pincus, Cain, & Wright, 2014), personality psychologists have largely considered it as a medley of adaptive and maladaptive components that are observed in normal populations (Raskin & Terry, 1988; Emmons, 1987). Almost from the nascence of the personality tradition, researchers have proposed two types of narcissism (Emmons, 1984; Watson & Biderman, 1993). One type is maladaptive, echoing the clinical tradition to some extent and encompassing defensiveness, aggressiveness, and egotism (Bushman & Baumeister, 1998). The other type is adaptive, reflecting the healthy components of narcissism and characterized by successful self-exhibition, acceptable self-aggrandizement, and high confidence (Kernberg, 1975; Watson & Biderman, 1993). Consistent with this proposal, decades of research has yielded a large body of evidence supporting a distinction between adaptive and maladaptive narcissism.

Distinction Reflected in the Measure of Narcissism

The Narcissistic Personality Inventory (NPI, Raskin & Hall, 1979) has served as the primary measure of grandiose narcissism. NPI scores , moreover, are often the basis of conceptualizations of grandiose narcissism (Cain, Pincus, & Ansell, 2008). The NPI was developed in conjunction with descriptions of Narcissistic Personality Disorder (NPD) in the Diagnostic and Statistical Manual of Mental Disorders (DSM-III ; American Psychiatric Association, 1980). The scale originally included 220 items, mostly tapping grandiose expressions of pathological narcissism, and eventually was refined and reduced to 40 items (Raskin & Hall, 1979; Raskin & Terry, 1988). Factor analyses have revealed diverse factor structures underlying the NPI, with factors of 2 (Power and Exhibitionism, Kubarych, Deary, & Austin, 2004; Leadership/Authority and Exhibitionism/Entitlement, Corry, Merritt, Mrug, & Pamp, 2008), 3 (Power, Exhibitionism, and Specialness, Kubarych et al., 2004; Leadership/Authority, Grandiose/Exhibitionism, and Entitlement/Exploitativeness, Ackerman et al., 2011), 4 (Exploitativeness/Entitlement, Leadership/Authority, Superiority/Arrogance, and Self-Absorption/Self-Admiration, Emmons, 1984, 1987), and 7 (Authority, Self-Sufficiency, Superiority, Exhibitionism, Exploitativeness, Vanity, and Entitlement, Raskin & Terry, 1988).

Despite the complexity and inconclusiveness of the factors underlying the NPI, researchers have observed that the NPI includes both healthy and unhealthy factors (e.g., Emmons, 1984; Raskin & Terry, 1988). This distinction is most evident in the seven-factor model: authority and self-sufficiency are healthy and associated with such desirable traits as self-confidence and assertiveness , whereas entitlement, exploitativeness, and exhibitionism are unhealthy and associated with poor psychological well-being and social adjustment (Raskin & Terry, 1988; Watson & Biderman, 1993). Using these factors, researchers have developed two NPI subscales that gauge adaptive and maladaptive narcissism separately (Barry, Frick, & Killian, 2003). The two subscales have exhibited acceptable reliability (i.e., internal consistency) and validity (i.e., predictive validity and construct validity) (Barry et al., 2007; Cai, Shi, Fang, & Luo, 2015; Hepper et al., 2014). Most evidence we review in the sections below employs this measurement scheme.

Distinction Reflected in Personality Nomologic Networks

Research on overall grandiose narcissism has established that it is positively correlated with extraversion, openness, and conscientiousness but negatively with neuroticism and agreeableness (for a review, Miller & Maples, 2011), with the magnitude of the correlations varying from small for conscientiousness (0.08) to moderate for extraversion (0.39). Research based on factors of grandiose narcissism has consistently demonstrated that all factors, whether they be healthy or unhealthy, are positively correlated with extraversion and negatively with agreeableness. The healthy and unhealthy factors differ, however, in their relationship with neuroticism, while relatively healthy factors, such as leadership and authority , are negatively associated with neuroticism; relatively unhealthy factors such as entitlement and exploitativeness are positively associated with it (Ackerman et al., 2011; Brown, Budzek, & Tamborski, 2009; Corry et al., 2008; Hill & Roberts, 2012). Thus, although healthy and unhealthy factors share some similarities in terms of a nomologic foundation for personality, they differ in their associations with neuroticism .

Distinction Reflected in Associations with Intrapersonal Adaptions

Research has revealed that healthy and unhealthy components of grandiose narcissism manifest distinct associations with intrapersonal adaptions. Individuals with higher scores for exploitativeness or entitlement are more likely to be self-conscious (Watson & Biderman, 1993), to report larger actual-ideal discrepancies (Rhodewalt & Morf, 1995), and to possess lower self-esteem (Brown et al., 2009). Higher levels of exploitativeness or entitlement have been linked to increased mood variability and emotional intensity (Emmons, 1987), greater neuroticism (Emmons, 1984), and higher scores on the Narcissistic Personality Disorder scale (Emmons, 1987; Watson, Grisham, Trotter, & Biderman, 1984). In contrast, individuals who score higher on the Leadership/Authority dimension report a higher level of self-awareness (Watson & Biderman, 1993) and self-esteem (Brown et al., 2009; Emmons, 1984; Rhodewalt & Morf, 1995; Watson & Biderman, 1993; Watson, Little, Sawrie, & Biderman, 1992) and a lower level of neuroticism (Emmons, 1984; Rhodewalt & Morf, 1995) and actual-ideal self-discrepancy (Emmons, 1984; Raskin & Terry, 1988; Rhodewalt & Morf, 1995). Furthermore, Leadership/Authority is negatively associated with indices of poor psychological well-being, such as anxiety, social anxiety, depression, and personal distress (Emmons, 1984; Watson & Biderman, 1993; Watson & Morris, 1991). Taken together, adaptive and maladaptive components of narcissism are associated with intrapersonal adaptions in opposite directions: while the former is beneficial, the latter is detrimental.

Distinction Reflected in Associations with Interpersonal Adaptions

Grandiose narcissism can be toxic in interpersonal situations. Not all components of grandiose narcissism, however, are problematic. Two lines of evidence are available so far. The first line of evidence involves the relationship between narcissism and aggression. It is well-known that people with high grandiose narcissism are often high in aggression. When confronted with failure, social rejection, or any other source of threat to the ego, they often respond in aggressive ways (Baumeister, Bushman, & Campbell, 2000; Morf & Rhodewalt, 2001). For instance, they may denigrate evaluators, punish competitors, and even act antagonistically toward innocent others (Bushman & Baumeister, 1998; Horton & Sedikides, 2009; Martinez, Zeichner, Reidy, & Miller, 2008; Twenge & Campbell, 2003). Exploration of the relationship between aggressiveness and specific components of narcissism, however, have shown that aggressiveness is mainly associated with unhealthy components such as entitlement and exploitativeness rather than the relatively healthy self-sufficiency and superiority components (Moeller, Crocker, & Bushaman, 2009; Reidy, Zeichner, Foster, & Martinez, 2008; Washburn, McMahon, King, Reinecke, & Silver, 2004; but see Blinkhorn, Lyons, & Almond, 2015).

A second line of evidence has examined the relationship between grandiose narcissism and empathy . Overall, research has shown that high grandiose narcissism is associated with low empathy (Fan et al., 2011; Watson et al., 1984). For specific components of narcissism, however, research shows that lack of empathy is more likely to be associated with unhealthy components rather than the healthy ones. An early study examined the relationship between empathy and the various factors underlying the NPI (Watson & Morris, 1991). Results showed that exploitativeness/entitlement was negatively associated with empathic concern and perspective taking but others factors were not. Later, a study examined adaptive and maladaptive narcissism among adolescents directly and found that maladaptive narcissism was related to a constellation of callous-unemotional traits (e.g., failure to show empathy, constricted display of emotion), whereas adaptive narcissism was not (Barry et al., 2003). More recently, a series of studies examined how adaptive and maladaptive narcissism were differentially associated with state empathy (Hepper et al., 2014). Results showed that when exposed to a target person’s distress, individuals high in maladaptive narcissism (as opposed to those high in adaptive narcissism) displayed low momentary empathy as indicated by both self-reports (Study 1) and autonomic arousal (Study 3). Taken together, it is maladaptive narcissism rather than adaptive narcissism that is associated with interpersonal problems.

Distinction Reflected in Associations with Problem Behaviors

Two kinds of problem behaviors have been shown to be differentially associated with adaptive and maladaptive narcissism. One has to do with impulsive buying. Grandiose narcissism has been linked to problematic consumption behaviors (Rose, 2007). One of our recent studies, however, showed that it is maladaptive narcissism rather than adaptive narcissism that predicts a tendency of impulsive buying (Cai et al., 2015). In this research, we first demonstrated with an internet sample that impulsive buying is positively associated with maladaptive narcissism but not with adaptive narcissism (Study 1). We then replicated this finding with a twin sample and further showed that the association between maladaptive narcissism and impulsive buying had a genetic foundation (Study 2).

Another involves conduct in children and adolescence (Barry et al., 2003; Washburn et al., 2004). A longitudinal study investigated a group of children and young adolescents over a 4-year period (Barry et al., 2007). Results showed that while maladaptive narcissism predicted delinquency and police contact at all follow-ups, adaptive narcissism exhibited no significant correlation with delinquency and a significantly negative one with police contact. In summary, it is maladaptive narcissism but not adaptive narcissism that predicts problem behaviors.

Distinction Reflected in Developmental Trajectories

Only a few studies have examined the development of narcissism. A longitudinal study showed that in general, observer-rated narcissism increased from ages 14 to 18, followed by a slight but nonsignificant decline from ages 18 to 23 (Carlson & Gjerde, 2009); multiple cross-sectional studies have shown that narcissism is negatively correlated with age in adulthood (Foster, Campbell, & Twenge, 2003; Roberts, Edmonds, & Grijalva, 2010). These studies suggest that narcissism increases during adolescence but tends to decline during adulthood. Interestingly, Foster et al. (2003) also demonstrated that age-related decreases tend to be larger for the maladaptive components (i.e., exhibitionism, exploitativeness, and entitlement) than for the adaptive components (i.e., self-sufficiency and authority). A recent large cross-sectional study has investigated more than 20,000 people in China (Cai, Kwan, & Sedikides, 2012). This study, again, replicated the age-related downward trend for overall grandiose narcissism. It demonstrated, moreover, differential trajectories for adaptive and maladaptive narcissism: while adaptive narcissism remained stable across a life-span, maladaptive narcissism exhibited a decreasing tendency.Footnote 1 Together, these findings indicate that adaptive and maladaptive aspects of narcissism follow different developmental trajectories.

Distinction Reflected in Genetic and Environmental Bases

Two previous studies have examined grandiose narcissism from the perspective of behavioral genetics. Overall, substantial genetic influences on grandiose narcissism have been found in both Asian and Western samples (e.g., Luo, Cai, Sedikides, & Song, 2014; Vernon, Villani, Vickers, & Harris, 2008). Furthermore, non-shared environments (i.e., environments not shared by twin siblings, like life events), but not shared environments (i.e., environments shared by twin siblings, like living conditions), exhibited a pronounced influence on narcissism. Two recent twin studies shed light on how these effects might vary with whether grandiose narcissism is adaptive or maladaptive. One examined the etiology of grandiosity and entitlement, which are reflective of adaptive and maladaptive narcissism, respectively (Luo, Cai, & Song, 2014). These results showed that the genetic and environmental effects on grandiosity and entitlement were largely different: less than 10% of genetic and environmental effect were accounted for by common genetic and environmental factors. The other twin study examined adaptive and maladaptive narcissism directly. Results revealed that both aspects were heritable, with more than half of their variation accounted for by unique environments (Cai et al., 2015); more importantly, the majority of the genes (54%) and environments (85%) underlying adaptive and maladaptive narcissism were different.Footnote 2 These two studies provide both direct and indirect evidence for the distinct genetic and environmental foundations of adaptive and maladaptive narcissism.

Conclusions, Implications, and Future Directions

To date, studies on grandiose narcissism have focused primarily on overall narcissism and relied on the NPI for analysis. In this chapter, we illustrated evidence indicating that grandiose narcissism actually includes two distinct components: adaptive and maladaptive narcissism. Based on the seven-factor model of the NPI (Raskin & Terry, 1988), research has established that while exhibitionism, entitlement, and exploitativeness are maladaptive, authority and self-sufficiency are adaptive (with superiority and vanity being neither adaptive nor maladaptive). Adaptive and maladaptive narcissism differ from each other in terms of how they correlate with other personality traits, inter- and intrapersonal adaptions, problem behaviors, developmental trajectories, and genetic and environmental foundations. People with high maladaptive narcissism are more likely to score higher in neuroticism, actual-ideal discrepancies, depression, anxiety, aggression, impulsive buying, and delinquency but lower in empathy and self-esteem. In contrast, people with high adaptive narcissism are more likely to manifest the opposite tendencies for almost every one of these traits and proclivities. Moreover, maladaptive narcissism declines with age, whereas adaptive narcissism does not. Particularly notable, adaptive and maladaptive narcissism differ substantially in their genetic and environmental bases. These findings provide convergent and consistent evidence for the distinctiveness between adaptive and maladaptive narcissism as well as the double-edged sword nature of grandiose narcissism.

Distinguishing between adaptive and maladaptive narcissism may help us better understand the complexity of grandiose narcissism as well as other relevant findings. First, we may gain a more nuanced understanding about grandiose narcissism and grandiose narcissists. People with extremely high grandiose narcissism must be high in both adaptive and maladaptive facets, that is, attractive but toxic; people with moderate narcissism may be high in either facet or moderate in both facets, that is, proud but not too annoying, annoying but not too proud, or somewhat proud and somewhat annoying; a low narcissist should be low in both facets, thus behaving in a modest and agreeable manner. These possibilities suggest that narcissists with similar scores on the NPI still may be quite different from each other. Second, we may have a better understanding about the mixed nature of the NPI and further its ambiguous correlations with many other constructs. Although the NPI’s latent factor structure is still inconclusive, as the chief measure of narcissism (although see Chap. 12 by Foster et al., this volume, for a review of additional measures of grandiose narcissism), two functionally distinct components emerge: adaptive and maladaptive. These two components may have correlations with other variables differing in magnitude (e.g., correlations with impulsive buying, Cai et al., 2015) or direction (e.g., correlations with neuroticism Ackerman et al., 2011; Corry et al., 2008). As a result, correlations based on the total score of the NPI are possibly confounded and may be misleading at times. These possibilities suggest that we should be cautious whenever we use the total score of the NPI as an index of grandiose narcissism and examining its relationship with other variables.

Evidence for the distinction between adaptive and maladaptive narcissism also suggests future directions for both empirical research and intervention practice. First, most studies on narcissism so far treat it as a singular construct. Given the distinct nature of adaptive and maladaptive narcissism, studies distinguishing between them are needed, particularly in cases where different results are expected to exist. For instance, research has suggested several contrasting self-regulation strategies employed by narcissists, including inter- versus intrapersonal processes (Morf & Rhodewalt, 2001), admiration versus rivalry approaches (Back et al., 2013), and primitive versus mature strategies (Roche, Pincus, Lukowitsky, Menard, & Conroy, 2013). Future study may examine how adaptive and maladaptive narcissism are differentially associated with these self-regulation strategies. Second, since current conceptualizations and operationalizations of adaptive and maladaptive narcissism are based almost exclusively on reformulations of the NPI, future studies should develop purpose-built measures of these two forms of narcissism. Third, since the dark side of narcissism mainly involves its maladaptive component, future intervention practices should treat adaptive and maladaptive narcissism independently and focus on how to curtail the maladaptive aspect while perhaps leaving the adaptive aspect intact (e.g., Hepper et al., 2014).