Toward a clarification of the concept of narcissism

Narcissism was initially conceptualized as a unitary construct referring to self-love, inflated self-images, self-serving bias, and demanding display of entitlement (Campbell et al., 2006; Kernberg, 1967; Twenge & Campbell, 2009). In the 1990s, however, this conceptualization of narcissism was broadened into two subdimensions, namely grandiose (or overt) and vulnerable (or covert; Wink, 1991). Although these two forms of narcissism share some common characteristics such as grandiose fantasies, a sense of entitlement, and the exploitation of others (Pincus et al., 2009), they can manifest in very different ways. Individuals with high narcissistic grandiosity are characterized by arrogance, grandiosity, egoism, and a lack of empathy. When threatened, they employ maladaptive strategies, such as the devaluation of others to keep their self-esteem high (Pincus et al., 2014). In contrast, individuals with high narcissistic vulnerability are more likely to be shy, embarrassed, and anxious; their self-esteem is fragile and accompanied by anger and irritability when threatened (Dickinson & Pincus, 2003; Pincus et al., 2009). In addition to grandiose and vulnerable dimensions, Ettema and Zondag (2002) suggested a possible third dimension of narcissism called isolation. This dimension relates to the separation between self and others and the feeling of not being known and understood by them. Individuals who feel isolated believe they are being criticized by people who have little respect for who they really are and are disappointed that others do not realize this. This isolation dimension usually concurs with vulnerable narcissism (Ettema & Zondag, 2002). Like vulnerable narcissism, isolation narcissism is considered dysfunctional (Ettema & Zondag, 2002), while grandiose narcissism is seen as a healthier subtype of narcissism compared to these two (Sedikides et al., 2004). Although many scholars have indeed considered the multifaceted nature of narcissism in their research (Dickinson & Pincus, 2003; Miller & Campbell, 2008; Wink, 1991), most of the studies conducted to date on the role of narcissism in the quality of interpersonal relationships have treated narcissism as a unitary construct (Barelds et al., 2017; Chin et al., 2017). Therefore, research is needed on the associations of different forms of narcissism with relational capacity. In addition, it remains unclear what mechanisms may underlie this association. In the present study, we attempt to fill these gaps in the literature.

Narcissistic traits and relational capacity

There is evidence that grandiose, vulnerable and isolated narcissism affect the individual’s relational capacity and the ability to relate to others. In the present study, relational capacity refers to the capacity to sincerely care for and connect with others, feel a sense of belongingness, and communicate personal experiences in (long-term) relationships (albeit not necessarily exclusive; Verheul et al., 2008). Specifically, higher levels of grandiose narcissism were linked to less avoidance within interpersonal relationships (Campbell et al., 2006; Dickinson & Pincus, 2003; Rohmann et al., 2012), while higher levels of vulnerable narcissism were linked to more social withdrawal, feelings of insecurity and anxiety about possible rejection (Dickinson & Pincus, 2003; Rohmann et al., 2012; Smolewska & Dion, 2005). Grandiose narcissists have also been found to score higher on social-cognitive skills, such as perspective-taking, emotional intelligence, social reasoning, and empathy (Vonk et al., 2013). It is believed that they use these skills to manipulate others to reinforce their endless identity needs for superiority (Jonason & Webster, 2012). In addition, individuals high in narcissistic grandiosity are more likely to report low interpersonal distress and secure attachment styles associated with positive self-representation. When interacting with others, they tend to be dominant and assertive and generally see themselves positively, although others prefer to describe their actions negatively. Therefore, grandiose narcissists are considered to have difficulty understanding the impact of their behaviors on others (Dickinson & Pincus, 2003; Miller et al., 2007), and any conflict in the environment is attributed externally and not seen as a consequence of their behavior and unrealistic expectations.

Vulnerable narcissists maintain their self-esteem by avoiding criticism and judgment from others, which leads to more insecure behaviors (Vonk et al., 2013). Research has shown that they are less able to understand the feelings and thoughts of others (Baron-Cohen et al., 2001; Vonk et al., 2013). Thus, their heightened sensitivity to the judgment of others may stem partly from a tendency to misinterpret others’ feelings and thoughts, which can lead to severe self-focus and an inability to respond appropriately (Aradhye & Vonk, 2014). Interpersonal coldness and social avoidance of vulnerable narcissists mainly result from their difficulty in managing their vulnerability in interpersonal relationships. To ensure this, they are more likely to withdraw socially, either in an avoidant or a cold, distanced way (Cooper, 1981, 1998). Similarly, isolated narcissists are more likely to separate themselves from others because of their excessive need for self-preoccupation (Ettema & Zondag, 2002). In addition, they distance themselves from others because they believe they cannot recognize and understand their worth (Ettema & Zondag, 2002). Furthermore, compared to their grandiose counterparts, vulnerable narcissists reported greater interpersonal distress and were more likely to experience conflict around their entitled expectations (Dickinson & Pincus, 2003). This can result in a fast response to frustration and a risk of violence. Indeed, narcissistic vulnerability (but not grandiosity) was found to be an important driver for aggressive and or criminal behavior, fueled by suspiciousness, dejection and angry rumination (Krizan & Johar, 2015; Pincus et al., 2009). Although previous studies that treated narcissism as a one-dimensional construct generally reported that narcissistic individuals are overrepresented among criminals (Hepper et al., 2014; Larson et al., 2015), a recent study found that violent offenders were characterized by higher levels of vulnerable narcissism, while higher levels of grandiose narcissism were more characteristic of the community controls (Bogaerts et al., 2021). The isolation dimension has been rarely investigated, but due to its similarity to the vulnerable dimension, it has been suggested that isolated narcissists may display more maladaptive behaviors in various areas of life, including interpersonal functioning compared to their grandiose counterparts (Ettema & Zondag, 2002).

Despite a scarcity of studies in this area, the findings have shown that grandiose narcissists have a better capacity to interact with others than vulnerable and isolated narcissists. Further research is needed to support these findings. In addition, more insight into the mechanisms underlying these associations is needed to better understand why these three forms of narcissism may be differently associated with relational capacity. Therefore, in this study, we propose two factors that may explain the associations of three narcissistic types (i.e., grandiose, vulnerable, and isolated) with the ability to interact with others.

Identity integration and social concordance as possible explanatory mechanisms

The first factor is identity integration which represents a coherence of identity and the capacity to see oneself and one’s own life as stable, integrated, and purposeful (Verheul et al., 2008). A recent study showed that grandiose narcissism was positively associated with identity integration, while vulnerable narcissism was negatively associated (Bogaerts et al., 2021). It may be that individuals who score high on narcissistic grandiosity have a more integrated identity, as they are more likely to regulate their self-esteem through overt self-enhancement and devaluation of others. Hence, their self-image is less dependent on the perceptions of others (Malkin, 2016). In contrast, the sense of identity of individuals high on narcissistic vulnerability largely depends on the opinions of others, making them hypersensitive to negative feedback from others (Zeigler-Hill et al., 2008). With regard to isolated narcissism, there is no empirical evidence on how it relates to identity integration. However, a stronger sense of coherence has often been associated with lower levels of social isolation (Chu et al., 2016; Limarutti et al., 2021). Thus, because of their tendency to distance themselves from others, isolated narcissists might be expected to have a less integrated identity. Furthermore, integrated identity plays an important role in interpersonal relationships; it is one of the precursors of authentic intimacy with another person. Only individuals with a firm and coherent sense of identity can participate in close, warm, communicative, and committed interactions (Beyers & Seiffge-Krenke, 2010; Bogaerts et al., 2004; Keybollahi et al., 2012; Ragelienė, 2016). Someone with high self-differentiation is better able to form and maintain interpersonal relationships (Rassart et al., 2012) and is less controlling (Adams et al., 1984). Besides, individuals with higher levels of identity integration perceive social interactions as safer and have better quality of peer relationships, more satisfying romantic relationships, and fewer feelings of loneliness (Bogaerts et al., 2006; Doumen et al., 2012). Conversely, individuals with lower levels of identity integration have a chaotic sense of self, unrealistic self-esteem, and unstable perceptions of self and others. As a result, they are less able to integrate positive and negative aspects of images of self and others, which can lead to difficulties in intimate relationships (Di Pierro et al., 2018). It has recently been shown that a disintegrated identity can also lead to violent behavior, but only via self-control. That is, individuals with low levels of identity integration may have a diminished capacity for self-control, making them more prone to criminal behavior (Bogaerts et al., 2021). Taken together, it could be that grandiose narcissists have a superior capacity to relate to others because of their more integrated identity compared to vulnerable and isolated narcissists. To our knowledge, no previous studies to date have investigated whether identity integration can underly the associations of grandiose, vulnerable and isolated narcissism, with relational capacity.

A second factor that may explain why grandiose narcissists have better relational capacities than vulnerable and isolated narcissists is social concordance. Social concordance is defined as the ability to control aggressive impulses toward others, deal with frustration, cooperate well with others, and value and honor others (Verheul et al., 2008). We argue that grandiose narcissists will display higher levels of social concordance compared to vulnerable and isolated narcissists. There is ample evidence that individuals high in grandiose narcissism are more able to control and regulate their emotions (e.g., Casale et al., 2019; Czarna et al., 2021) and show less distrust and hostility towards others (e.g., Czarna et al., 2021; Krizan & Johar, 2015) compared to vulnerable narcissists. It is also believed that vulnerable narcissists are less cooperative due to their hostility and poor anger management, while isolated narcissists are self-preoccupied and do not care much about others (Ettema & Zondag, 2002). One study found that individuals with high narcissistic vulnerability were less likely to cooperate in the prisoner’s dilemma game — used to investigate competitive versus cooperative behavior — and were less likely to initiate or respond to collaboration as the game progressed (Malesza, 2020). They also care less about supporting others, while expecting others to support them (Miller et al., 2011; Pincus et al., 2009). Unlike vulnerable narcissists, grandiose narcissists were more likely to make a good first impression and maintain collaborative relationships (Casale et al., 2019; Malesza, 2020). However, their likeability and willingness to cooperate decreased as the game proceeded. It is plausible that individuals with high grandiose narcissism use cooperation as a strategy to boost their self-esteem, at least at the beginning of the game and until the game moves in the desired direction (Malesza, 2020).

Furthermore, lower levels of social concordance can also impact the quality of interpersonal relationships and lead to poor relations with others. In other words, the inability to control emotions and anger contributes to interpersonal problems and social maladjustment (e.g., Brackett et al. 2006; Caspi et al., 1987). Emotional stability can instead help people avoid fruitless conflicts or mitigate the negative impact of personal antagonisms (Lopes et al., 2011). Likewise, the willingness to cooperate supports the functioning of a wide variety of relationships, including couples (Murray & Holmes, 2009), groups (Fehr et al., 2002), and societies in general (Nowak, 2006). Finally, treating others with dignity and respect presages the ability to give and receive love and build relationships based on mutual caring and trust. Overall, higher levels of social concordance may account for a better capacity to relate with others in individuals high in narcissistic grandiosity. Note that grandiose narcissists rely heavily on others to bolster and maintain their self-esteem. As a result, they engage in behaviors that provide immediate gratification to their desires for social status, positive affect, and ego involvement in achievement domains. However, their actions usually have long-term negative consequences (Jones & Paulhus, 2011; Malesza & Ostaszewski, 2016). Compared to grandiose narcissists, vulnerable narcissists may have less relational capacities due to their lack of cooperation and a tendency to exhibit hostile and aggressive behavior towards others. Similarly, isolated narcissists may have less relational capacities due to their preoccupation with themselves and disinterest in reaching others. Nevertheless, the mediating effect of social concordance in the associations of grandiose, vulnerable, and isolated narcissism with relational capacity has never been tested before.

Present study

The present study sought to bridge these gaps in the literature by investigating whether identity integration and social concordance can mediate the associations of grandiose, vulnerable, and isolated narcissism with relational capacity. Because our sample includes individuals convicted of criminal behavior as well as individuals from the community and is considerably heterogeneous in terms of age, we included age and belonging to a criminal or community group as controlling variables in the proposed mediation model. In addition, the data further suggest that relational capacity increases with age and therefore it is important to consider age when researching this variable (e.g., Kotiuga et al., 2022). Several hypotheses have been formulated on the basis of the aforementioned literature. First, we hypothesized that grandiose narcissism would have a direct positive effect on relational capacity (e.g., Campbell et al., 2006; Dickinson & Pincus, 2003; Vonk et al., 2013; path c1), while vulnerable (e.g., Dickinson & Pincus, 2003; path c2) and isolated narcissism (e.g., Ettema & Zondag, 2002; path c3) would have a direct negative effect on relational capacity. In line with the recent study (Bogaerts et al., 2021) we further hypothesized that grandiose narcissism would have a direct positive effect (path a11), while vulnerable narcissism would have a direct negative effect (path a21) on identity integration. There is no empirical evidence on the effect of isolated narcissism on identity integration, but given that isolated narcissists tend to distance themselves and that lower levels of social isolation were found to be associated with a stronger sense of coherence (Chu et al., 2016; Limarutti et al., 2021), we expected that isolated narcissism like vulnerable narcissism, would have a direct negative effect on identity integration (path a31). Similarly, we hypothesized that grandiose narcissism (e.g., Casale et al., 2019; Czarna et al., 2021; path a12) would have a direct positive effect, whereas vulnerable (e.g., Krizan & Johar, 2015; path a 22) and isolated narcissism (Ettema & Zondag, 2002; path a32) would have a direct negative effect on social concordance. Consistent with previous evidence, we expected that both identity integration (e.g., Beyers & Seiffge-Krenke, 2010; Keybollahi et al., 2012; path b1) and social concordance (Fehr et al., 2002; Murray & Holmes, 2009; path b2) would have a direct positive effect on relational capacity. Finally, based on the discussed literature, we argue that grandiose narcissists would have a better capacity to relate to others, probably due to their better sense of self and lower hostility and better cooperative skills. Therefore, we hypothesized that identity integration and social concordance would be significant mediators in the positive association between grandiose narcissism and relational capacity, and the negative association of vulnerable and isolated narcissism with relational capacity. See Fig. 1, for a graphical representation of the model.

Fig. 1
figure 1

Mediating Roles of Identity Integration and Social Concordance in the Associations Between Narcissism Traits and Relational Capacity

Full size image

Shedding light on the mechanisms underlying the associations of these three narcissistic types with relational capacity could benefit the treatment of narcissistic individuals and potentially make them less harmful to others in social and intimate relationships.

Methods

Participants and procedure

This study is part of a larger research project investigating personality and relationships among participants from both forensic psychiatric institutions and the community. The sample comprised 222 male participants. Participants ranged in age from 20 to 60, with a mean age of 37.71 years (SD = 13.25). Most of the participants self-reported a Dutch nationality (67.6%, n = 150), living alone (28.2%, n = 62), and having an income from paid employment (65.4%, n = 138). The most common finished level of education was intermediate vocational education/MBO (31.1%, n = 69), followed by higher professional education/HBO (28.4%, n = 63), higher general secondary education/HVO (9%, n = 20), and secondary education/VWO (9%, n = 20). Of this sample, 157 (70.7%) were participants from the community and 65 (29.3%) were outpatients in treatment at four Dutch outpatient forensic centers. Forensic patients were attending mandatory treatment which was imposed by the judge due to a committed offense. The index offenses included physical aggression (45.3%, n = 29); domestic violence (31.3%, n = 20); verbal aggression (20.3%, n = 13); and other offenses (3.1%, n = 10). More details on sample characteristics can be found in Table A1 in the appendix.

All participants were introduced to the purpose of the study and gave written informed consent to take a voluntary part. They did not receive any payment or other benefits for their participation and were informed that participation may be discontinued at any time without penalty or loss. Participants from forensic centers were also informed that their decision to participate in the study would not have any influence on their treatment status, and that information gathered for this research would not be shared with their therapists. Forensic patients completed questionnaires during a treatment session. Furthermore, master-level students in clinical forensic psychology recruited the participants from the community by a means of a snowball sampling technique. Snowball sampling is a non-probability sampling technique where currently enrolled research participants assist researchers in identifying other potential participants. Inclusion criteria included being at least 18 years old and having sufficient knowledge of the Dutch language. Community participants were matched with the participants convicted of criminal behavior on two characteristics, namely age and level of education. Participants with a university degree were excluded from the community sample because this category was not present in the sample of criminals. To ensure that community participants did not belong to the criminal group, they were asked if they had ever been convicted of a criminal offense. In addition, community participants who had been in treatment with a psychologist, psychotherapist, or another care provider in the past three years were excluded. None of the community respondents reported a history of criminal behavior or a history of counseling. All participants were instructed to return the completed questionnaires in a sealed envelope. The data were anonymized and could not be traced to an individual participant. All procedures involving human participants were performed according to the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The Scientific Research Committee of FPC Kijvelanden and the Ethical Review Board of Tilburg University approved the study.

Measures

Grandiose, vulnerable, and isolated narcissism

The Dutch Narcissism Scale ([Nederlandse Narcisme Schaal]; NNS; Ettema & Zondag, 2002) was used to measure grandiose, vulnerable, and isolated narcissism. The 35-item NNS is a Dutch adaptation of the Narcissistic Personality Inventory (Raskin & Hall, 1979, 1981) and the Hypersensitive Narcissism scale (Hendin & Cheek, 1997). It is a self-report questionnaire measuring three different types of narcissism including vulnerable narcissism (11 items; e.g., “Small remarks of others can sometimes easily hurt my feelings”), grandiose narcissism (12 items; “Sometimes I feel like I got lucky with who I am anyway”) and isolation (12 items; “I often have the feeling that there is a shield separating me from others”; Ettema & Zondag, 2002). All items are rated on a seven-point Likert scale ranging from 1 = that is certainly not the case to 7 = that is certainly the case. The total scores were computed for each narcissism subscale by summing all the items that belonged to the corresponding subscale. Higher scores indicate greater levels of narcissism. The validity of the Dutch NNS was demonstrated by its relations with age, self-esteem, burnout, and empathy (Ettema & Zondag, 2002), the meaning of life (Zondag, 2005), satisfaction with life and depression (Nauta & Derckx, 2007), and boredom (Zondag, 2007). In previous research, internal consistency (Cronbach’s α) of grandiose narcissism ranged from 0.71 to 0.77, and of vulnerable narcissism from 0.77 to 0.87. Cronbach’s alpha in the present study was 0.71 for grandiose narcissism, 0.81 for vulnerable narcissism and 0.78 for isolated narcissism.

Identity integration, social Concordance, and relational capacity

The Severity Indices of Personality Problems – Short Form (SIPP-SF; Verheul et al., 2008) was used to measure identity integration, social concordance, and relational capacity. The SIPP-SF is a 60-item self-report questionnaire that measures five domains of maladaptive personality functioning, namely: self-control (12 items), identity integration (12 items), relational capacities (12 items), responsibility (12 items), and social concordance (12 items). All items are rated on a four-point Likert scale ranging from 1 = fully disagree to 4 = fully agree, with higher scores indicating greater impairments. The self-control and responsibility domains of the SIPP-SF were not included in this study. The identity integration domain (e.g., “I strongly believe that I am just as worthy as other people”) captures the capacity to see oneself and one’s own life as stable, integrated, and purposive. The social concordance domain (e.g., “I can work with people on a joint project in spite of personal differences”) captures the capacity to value someone’s needs and identity, resist aggressive impulses towards others, and work together with others. The relational capacity domain (e.g., “I have been able to form lasting friendships”) captures the capacity to genuinely care about others and to feel loved and recognized by others, as well as the ability to share mutual experiences with others often but not necessarily in the context of a long-term, intimate relationship. Total scores were calculated for each of these three domains by summing all items that belonged to that domain. Higher scores indicated higher levels of identity integration, social concordance or relational capacity. The SIPP-SF has demonstrated good psychometric properties in both Dutch and English samples (Arnevik et al., 2009; Rossi et al., 2017). In the current sample, all three domains have shown very good internal consistency with Cronbach’s alpha of 0.92 (identity integration), 0.85 (social concordance) and 0.87 (relational capacity).

Statistical analysis

All data were analyzed using IBM SPSS Statistics (version 25) and free software environment R (R Core Team, 2021). First, descriptive statistics were computed to summarize sample characteristics. Subsequently, the point-biserial correlation was used to measure the relationship between the continuous and a binary variable (i.e., criminal behavior), while the Pearson correlation was used to measure the relationship between pairs of continuous variables. Criminal behavior was used as the grouping variable where 0 = community participants and 1 = sample of criminals (i.e., a conviction of one or more violent crimes). Finally, we used the lavaan package in R (Rosseel, 2012) to test whether identity integration and social concordance mediate the association of grandiose, vulnerable and isolated narcissism with relational capacity, controlled for age and belonging to a criminal or community group through path analysis. Path analysis can be seen as a special case of structural equation modeling and involves only observable variables (Lei & Wu, 2007). We applied the maximum likelihood method to assess parameter estimates. In addition, the fit of the model was evaluated with the comparative fit index (CFI; values ≥ 0.90) and standardized root-mean-square residual (SRMR; <0.08; Hu and Bentler, 1999). Before the analyses, it was checked whether the data are normally distributed. Data are normally distributed if absolute values of skewness and kurtosis are not larger than 2 (Field, 2009).

Results

Descriptive statistics, including skewness and kurtosis, are presented in Table 1. All variables were normally distributed. Table 2 depicts the point-biserial and Pearson correlations. Relational capacity positively correlated with grandiose narcissism, identity integration and social concordance, respectively, and negatively correlated with vulnerable narcissism, isolated narcissism and criminal behavior. In addition, grandiose narcissism had a positive correlation with identity integration and social concordance, while vulnerable and isolated narcissism had negative correlations with these variables.

Table 1 Questionnaire Characteristics (n = 222)
Full size table
Table 2 Correlations Among the Study Variables
Full size table

The results of the mediation analysis are summarized in Table 3. The model fitted the data well (CFI = 0.94 and SRMR = 0.05). Considering direct effects, isolated narcissism (b = − 0.24, p < .001), and criminal behavior (b = − 2.30, p = .01) were negatively associated with relational capacity, while identity integration (b = 0.38, p < .001), and social concordance (b = 0.016, p = .01) were positively associated with relational capacity. Grandiose narcissism, vulnerable narcissism and age did not have a significant association with relational capacity. Furthermore, grandiose narcissism was positively (b = 0.32, p < .001), whereas vulnerable (b = − 0.10, p = .03) and isolated narcissism (b = − 0.37, p < .001) were negatively associated with identity integration. Finally, grandiose narcissism had a positive association (b = 0.09, p = .03), while isolated narcissism (b=- 0.33, p < .001) had a negative association with social concordance. Vulnerable narcissism was not significantly associated with social concordance.

Table 3 Summary of Direct and Indirect Effects
Full size table

When it comes to indirect effects, grandiose narcissism was significantly and positively associated with relational capacity through identity integration (b = 0.12, p < .001), while social concordance was not a significant mediator in this association. This means that a higher level of grandiose narcissism leads to a more coherent identity which in turn enhances the capacity to relate with others. Likewise, identity integration (b = − 0.03, p = .03) was a significant mediator in the negative association between vulnerable narcissism and relational capacity, whereas social concordance did not have a significant mediating role in this association. This indicates that individuals high in vulnerable narcissism have a less coherent sense of self which further negatively influences their relations with others. In contrast, both identity integration (b = − 0.014, p < .001) and social concordance (b = − 0.05, p = .01) significantly mediated the negative association between isolated narcissism and relational capacity. This finding signifies that a higher level of isolated narcissism leads to a less coherent identity and lower social concordance which consequently results in lower relational capacity.

Discussion

Despite the well-established multidimensional nature of narcissism, the association between narcissism as a multidimensional construct and the capacity to relate with others has never been tested before. In this study, we investigated how three different types of narcissism, namely grandiose, vulnerable, and isolated, are associated with relational capacity in 222 males from the community and forensic population. In addition, it was also investigated whether identity integration and social concordance may underly these associations. The proposed mediation model was tested in R and the analysis was adjusted for age and belonging to a criminal or community group. The results revealed that of the three narcissistic types, only isolated narcissism had a direct effect on relational capacity, while criminal behavior was the only significant covariate. In addition, both identity integration and social concordance had a positive direct effect on relational capacity. Furthermore, grandiose (+) and vulnerable narcissism (-) were significantly associated with relational capacity only through identity integration, while social concordance was not a significant mediator in this association. In contrast, both identity integration and social concordance significantly mediated the negative association between isolated narcissism and relational capacity.

Consistent with our hypothesis, individuals with higher levels of isolated narcissism showed a poorer capacity to relate with others. Isolated narcissists seem to be less capable of really caring about other people and sharing mutual experiences with them. This can be explained by their excessive narcissistic need for self-preoccupation and beliefs that others cannot recognize and understand their worth (Ettema & Zondag, 2002). However, we found no significant direct associations between grandiose and vulnerable narcissism with relational capacity. Grandiose narcissism did have a significant positive correlation with relational capacity, while vulnerable narcissism had a negative correlation. Previous research has shown that narcissistic grandiosity is associated with less avoidance within interpersonal relationships (Campbell et al., 2006; Dickinson & Pincus, 2003; Rohmann et al., 2012) and also with higher levels of perspective-taking, emotional intelligence, social reasoning, and empathy (Vonk et al., 2013); all these factors are assumed to positively influence social interactions. In contrast, narcissistic vulnerability is associated with greater social withdrawal, feelings of insecurity, and concern about possible rejection (Dickinson & Pincus, 2003; Rohmann et al., 2012; Smolewska & Dion, 2005), that is, the factors that may harm relations with others. In this study, we cannot make any statements about the direction of these significant associations of the two types of narcissism with a relational capacity based on correlations. Based on our empirical findings, our results provide some empirical support for the evidence that individuals with higher levels of grandiose narcissism are better able to get along with others in comparison to individuals with higher levels of vulnerable narcissism. The absence of significant direct effects of the two types of narcissism on relational capacity may be because we included identity integration and social concordance as mediators in the model. Given the significant bivariate correlations between the two different forms of narcissism and mediators, the lack of significant direct effects could therefore be explained by the common variance shared by these variables.

Indeed, in line with our hypotheses, identity integration explained the positive association between grandiose narcissism and relational capacity, as well as the negative association of vulnerable and isolated narcissism with relational capacity. This finding suggests that individuals high on grandiose narcissism have a more coherent sense of self and thus a better capacity to relate with others. This is consistent with previous findings showing that individuals with higher levels of grandiose narcissism have a better-integrated identity (Bogaerts et al., 2021). It also supports the idea that identity integration can be seen as a precursor to authentic intimacy with another person and that only individuals with high self-differentiation are able to form warm and close interpersonal relationships (Beyers & Seiffge-Krenke, 2010; Keybollahi et al., 2012; Rassart et al., 2012). In contrast, our results indicate that individuals high in narcissistic or isolated vulnerability have a less coherent identity, which may ultimately reduce their capacity to relate with others. The previous study also found that vulnerable narcissists have a chaotic sense of self and unstable perceptions of self and others (Bogaerts et al., 2021). Lower levels of identity integration of individuals with high narcissistic vulnerability were attributed to their enhanced sensitivity to negative feedback from others (Rogoza et al., 2018; Zeigler-Hill et al., 2008). Individuals with a disintegrated identity thus have more difficulty integrating positive and negative representations of self and others, which makes them less able to interact with others (Di Pierro et al., 2018) and makes them more prone to criminal behavior along with isolated narcissists (Bogaerts et al., 2021). Similarly, isolated narcissists may have a less integrated identity due to their increased need for social isolation (Chu et al., 2016; Limarutti et al., 2021). In brief, in addition to supporting existing findings, our study provides direct evidence that identity integration may explain why vulnerable and isolated narcissists have a lower capacity to relate to others than grandiose narcissists.

The present study also provides evidence that not only identity integration, but also social concordance may be an explanatory mechanism in the negative association between isolated narcissism and relational capacity. Individuals high in narcissistic isolation are less able to appreciate the needs of others, resist aggressive impulses towards others, and cooperate with others. This can lead to interpersonal problems and social maladjustment (e.g., Brackett et al. 2006; Caspi et al., 1987). Contrary to our expectations, this study did not support the mediating effect of social concordance on the association between grandiose and vulnerable narcissism and relational capacity. It could be that social concordance was only a significant mediator in the association between isolated narcissism and relational capacity because isolated narcissists hold considerably more negative views of others than the other two types of narcissism. Specifically, grandiose narcissists only see others as a means to bolster their self-image (Pincus et al., 2014), while vulnerable narcissists tend to perceive others mainly as a threat to their fragile ego (Dickinson & Pincus, 2003; Pincus et al., 2009). In contrast, isolated narcissists have beliefs that others are overly critical of them, that they do not value their worth, and that they cannot be fully understood and valued by others (Ettema & Zondag, 2002). All these extremely negative views of other people can cause isolated narcissists to constantly withdraw from others and have more difficulty controlling their negative impulses toward others, making it more difficult for them to relate to others.

Furthermore, our findings also fit with previous studies showing that grandiose narcissism is a better-adapted narcissistic subtype than vulnerable or isolated narcissism (e.g., Ettema & Zondag, 2002; Kaufman et al., 2020; Ng et al., 2014; Rose, 2002). For example, previous research has linked grandiose narcissism to better life satisfaction, lower perceived stress (Ng et al., 2014), a tendency toward assertiveness, persistence, and achievement (Rose, 2002), and non-criminality (Bogaerts et al., 2021). Our research also showed that, as opposed to vulnerable or isolated narcissism, grandiose narcissism is associated with better identity integration, higher levels of social concordance, and a greater capacity to relate with others. Although these grandiose narcissists may seem more satisfied with their lives and more capable of interacting with others, their “good” intentions are usually manipulative and connected to their endless identity needs for superiority (Jonason & Webster, 2012. Hence, their actions usually have negative consequences for other people in the long run (Jones & Paulhus, 2011; Malesza & Ostaszewski, 2016).

Limitations and future direction

This study is not without limitations. We only included males and therefore the results cannot be generalized to females. In addition, we used the snowball sampling method to recruit participants from the community. Thus, the representativeness of the subsample of the controls cannot be guaranteed. Another limitation concerns the validity of our findings relying solely on self-reported data. It has been suggested that self-reported data can be influenced by external biases such as social desirability. Offenders are particularly thought to employ socially desirable responding styles (Mills & Kroner, 2006) and for this reason, we cannot rule out the possibility that reported results are partly influenced by intentional deception in our sample. Finally, when adjusting the analysis for the age and belonging to the criminal or community group, due to the small sample size, we considered only the adjustments related to the outcome variable, although these covariates could affect the other variables in the model as well. For example, it has been shown that narcissism is overrepresented in individuals convicted of criminal offenses as well as that it tends to decline with age (Foster et al., 2003). Future studies should therefore attempt to include a more representative and larger sample of both genders and intend to measure the constructs of interests more objectively, via direct observations as well as to control all variables in the model for the potential influence of age and criminal behavior.

Nevertheless, the findings of the present study could be of practical importance for enhancing the relational competence of grandiose, vulnerable, and isolated narcissists. Grandiose and vulnerable narcissism were not directly, but only indirectly related to relational capacity via identity integration. Thus, the intervention aimed at improving the relational capacity of individuals high on narcissistic grandiosity or vulnerability should focus on developing a more coherent identity. Similarly, for isolated narcissists, a successful relational intervention should focus on strengthening both identity integration and social concordance as they were significant in explaining mechanisms of diminished capacity to relate with others for this particular type of narcissist. Thus, by enhancing these two, it is highly likely that relational capacity would be improved as well.

To conclude, the present study adds to the understanding of how three different types of narcissism are associated with relational capacity as well as to what extent identity integration and social concordance could explain these associations. Consistent with the literature, grandiose narcissism can be viewed as a healthier type of narcissism compared to the other two types. In addition, the findings of this study may hold important clinical implications. However, before translating the findings of this study into practice, replication studies are necessary to confirm our results.