I Am Bold, and Do Not Care for Social Norms: The Role of Faux Pas Recognition in the Relation Between Psychopathy and Aggression

ABSTRACT Psychopathy is found to be associated with aggression. However, the role of “understanding the mental states of others” in this relation is not clearly understood yet. The current study aimed to increase our understanding of a specific ability of understanding the mental state of others, namely recognizing a faux pas, and understand the relation of this “social misstep” in self-reported aggressive behavior. The sample consisted of 105 individuals from the community (both males and females) who were tested with the Faux Pas test and several self-report questionnaires assessing triarchic psychopathic personality traits and aggression. We found a positive relationship between the triarchic constructs and self-reported proactive and reactive aggression. Additionally, although all psychopathy constructs were negatively related to self-reported empathy, only the boldness trait was negatively associated with faux pas recognition. No mediation effect of faux pas on the relation between psychopathy and aggression was found. Current findings underline the importance of differentiating different Theory of Mind and empathic abilities in relation to psychopathy and aggression, which is essential to develop effective interventions.


Introduction
Psychopathy is a complex constellation of personality traits including impulsive and antisocial behavior, callousness, and unemotionality (Cleckley, 1976;Hare, 1991).Research has found that psychopathy is associated with aggressive behavior, and proactive aggression particularly (e.g., Cima & Raine, 2009;Porter & Woodworth, 2018), and these associations are also found in the general population (e.g., Neumann & Hare, 2008;Van Dongen et al., 2017, 2022).That is, psychopathic personality traits are found to be dimensional, and to lie on a continuum, with psychopathic offenders on the extreme (Edens et al., 2006) Although different mechanisms underlying the relation between psychopathy and aggression have been previously studied, these mechanisms are not clearly understood yet.One such mechanism is "understanding the mental states of others," which relates to someone's "Theory of Mind" (or mentalizing), which entails representation of mental states (Bird & Viding, 2014).Since "understanding the mental states of others" (i.e.ToM) is a very complex process itself, the relation between psychopathy and the role of ToM in aggressive behavior is quite complex.The current study aimed to increase our understanding of a specific form of understanding a mental state of others, namely recognizing "faux pas," and understand the relation of this "social misstep" in self-reported aggressive behavior.

ToM, psychopathy, and aggression
Generally, the broader concept of empathy is regarded as the ability to understand and to be sensitive to the mental states, intentions, and feelings of other persons, with cognitive empathy (ToM) as taking the perspective of another person's mental states and/or feelings (e.g., Decety, 2010;Decety & Jackson, 2004).Affective empathy (mostly referred to as "empathy") on the other hand, refers to the ability to feel the same as the other person does (emotional contagion), and/or sympathize with the other person (emotional concern).
Importantly, according to the theoretical framework of Shamay-Tsoory et al. (2010), ToM is a construct that can also be divided into cognitive and affective aspects.Cognitive ToM resembles what is generally referred to as metalizing, while the affective part refers to the ability to infer on other's feelings and therefore relates to both affective and cognitive empathy.It is important to note that affective ToM differs from affective empathy, in that affective empathy also includes emotional contagion (feeling the same feeling as the other person does).And this distinction between cognitive and affective aspects of ToM might explain some inconsistencies that are found when examining the relation between psychopathy and ToM.
Patrick and colleagues (Nelson et al., 2016;Patrick et al., 2012Patrick et al., , 2013;;Yancey et al., 2016) have proposed the Triarchic Model of Psychopathy (TriPM; Patrick et al., 2009) for advancing our understanding of psychopathic personality.The triarchic model posits that psychopathy reflects three distinctive etiological phenotypic constructs, namely "boldness," "meanness," and "disinhibition."Individuals high on boldness report less fear, have the ability to remain calm under stressful circumstances, and have a high level of social influence and dominance (Patrick, 2010).Meanness is viewed as a more malignant phenotypic expression of fearlessness, combined with a lack of empathy (e.g., callousness), low social closeness, and cruel behavior toward humans and animals.The third construct, disinhibition, is defined as the common dispositional tendency underlying general externalizing conditions of differing types and entails weak inhibitory control and deficient affect regulation.
Although the inclusion of boldness as a trait of psychopathy is debated (Lilienfeld et al., 2012), some scholars postulate that this construct is an important trait related to the successful aspects of psychopathy and may be found in those persons high on the social ladder and in corporate institutions (Sörman et al., 2016).Additionally, Wall and colleagues (Wall et al., 2015) found that boldness has incremental validity in the prediction of psychopathy, especially the interpersonal traits within psychopathy (for example, manipulating others).In relation to recognizing others' mental states, one would speculate that boldness would be related to a better cognitive understanding of others mental states (i.e.mentalizing or Theory of Mind), but unrelated, or negatively related to affective aspects of empathy (emotional contagion or emotional distress).
Meanness on the other hand represents a constellation of traits that includes cruelty toward others and a lack of empathy.The concept of meanness in psychopathy is also very closely related to the inability to identify others' signals of distress, which is thought to be essential for experiencing empathic concern and modulating social behavior (Blair, 1995(Blair, , 2007(Blair, , 2015)).Consequently, a lack of empathy in meanness might therefore lead to deficient inhibition of morally unacceptable behaviors, such as (proactive) aggression (Blair, 1995).This is consistent with previous published literature, which show a relation between both cognitive and affective empathy, and aggressive behavior (e.g., Euler et al., 2017;Jolliffe & Farrington, 2004).Conversely, boldness is found not to be related to aggression (Patrick & Drislane, 2014;Van Dongen et al., 2017).
From the above description of boldness and meanness as defining constructs of psychopathy, one could speculate that those who are manipulative and social effective, would be good in "reading" other persons, while those who are violent seem to lack the ability to feel with other persons.This distinction is also reflected in the different facets that make up empathy and ToM, and therefore, it may be no surprise that previous research on the relationship between empathy, ToM, and psychopathy are inconsistent (Brook & Kosson, 2013;Lishner et al., 2012).Robinson and Rogers (2015) for example, found that psychopathic criminals had the capacity of showing cognitive empathy, but no affective empathy.Also, Sandoval and colleagues (Sandoval et al., 2000) found a negative relationship between self-reported affective empathy and psychopathy, but no relationship with cognitive empathy.However, there are also studies in which no relations or negative associations were found between both affective and cognitive empathy and psychopathy (Brook et al., 2013;Brook & Kosson, 2013;Domes et al., 2013).So, while most research found no lack of cognitive empathy in psychopathic individuals (R. Blair et al., 1996;Dolan & Fullam, 2004;Richell et al., 2003), while but Brook and Kosson (2013) did find a lack of ToM in psychopaths.Since this lack concerned negative emotions such as fear and sadness, this finding would now be interpreted as evidence for a diminished affective ToM in psychopathy.

Current study
The Faux Pas test (Stone et al., 1998) measures both cognitive and affective aspects of ToM in adults and has already been used to measure ToM deficits in Autism Spectrum Disorders (Spek et al., 2010), and also to investigate whether ToM influences one's ability to pretend to be mentally ill (Di Girolamo et al., 2021##) A "faux pas" is a slip or blunder in social rules or etiquette, manners, or conduct.To understand that something is a faux pas, we need to use cognitive ToM to understand that someone better did not had to say or do the thing that was said or done.To know the impact of the faux pas, one needs to first recognize the faux pas, and consequently understand how the other person is feeling because of the faux pas, thus involving affective ToM.To our knowledge, only Dolan and Fullam (2004) used the Faux Pas test to study ToM in psychopathy.They found that individuals with psychopathy had no difficulties in general ToM, but had difficulties understanding the emotional impact of a faux pas (i.e.affective ToM).
Dysfunctions in ToM in persons with psychopathic traits are thus subtle and may be interpreted in a way that is not done so in previous studies.Also, which traits of psychopathy are related to these aspects of ToM (compared to facets of affective empathy) and how these relate to aggressive behavior is not previously studied yet.Therefore, the aim of this current study was to investigate the role of ToM in the relation between the triarchic psychopathy constructs (boldness, meanness, or disinhibition) and proactive and reactive aggression A second aim was therefore, to examine whether differences in the recognition of faux pas (i.e.cognitive ToM) and the emotional impact of the faux pas (affective ToM), mediate the relation between psychopathic traits and aggression.Based on previous research (e.g., R. Blair, 1995;Dolan & Fullam, 2004; R. J. R. Blair, 2001;Porter & Woodworth, 2018), it is expected that psychopathic traits are positively related to proactive aggression and that understanding of faux pas (both cognitive and affective) mediates this relation.

Participants
The data has been obtained from a sample of the general population (N = 105; 52.4% male).The mean age was 28.3 years (SD = 11.7;range 16-66 years).The test subjects were collected by means of convenience sampling.The subjects did not receive compensation for participating in the investigation and signed a written informed consent before they started the study.This study was conducted in accordance with the ethical standards of the Declaration of Helsinki and to the guidelines of the ethics committee of the Department of Psychology, Education and Child Studies of the Erasmus University Rotterdam.

Materials psychopathy traits
The Dutch version of the Triarchic Psychopathy Measure (TriPM, Patrick, 2010;Van Dongen et al., 2017) is used to measure psychopathic traits.The TriPM consists of 58 items that subjects rate on a 4-point scale to indicate the extent to which claims can be used to describe themselves (where 1 = false, 4 = true).The items measure three subscales of psychopathy: boldness, meanness, and disinhibition.There is evidence of good psychometric properties of this questionnaire (Van Dongen et al., 20042017), see also Table 1 for an overview of reliabilities (i.e.Cronbach's α) for the current sample.

Empathy traits
The Dutch version of the Interpersonal Reactivity Index (IRI; De Corte et al., 2007;Davis, 1980) was used to measure trait empathy.The questionnaire consists of 28 items that subjects on a 5-point scale to indicate to what extent the statements apply to themselves (from 0 = do not describe me, 5 = describes me very well).The items measure four subscales: perspective taking (PT), fantasy (FS), empathic concern (EC), and personal distress (PD).De Corte et al. (2007) found that the IRI is a reliable and valid measure for the assessment of empathy.

Reactive and proactive aggression
The Dutch Reactive and Proactive Aggression Questionnaire (RPQ; Cima et al., 2013;Raine et al., 2006) was used to measure reactive and proactive aggression.The questionnaire consists of 23 items and the subjects had to indicate for each question how often it occurs (0 = never, 1 = sometimes or often = 2).Previous research found good validity and internal consistency for this measure (Cima et al., 2013), and confirmatory factor analysis also has shown significant evidence a two-factor solution (reactive and proactive) of the tool.

Theory of mind
The Dutch version of the Faux Pas Test (Spek & van Berckelaer-Onnes, 2010;Stone et al., 1998) is a test to assess ToM in adults.It consists of nine stories of which five contain a faux pas.(see description below).Participants have to answer six questions following the different stories, of which five questions assess the recognition of the faux pas (i.e.cognitive ToM), and one assesses the impact of the faux pas (i.e.affective ToM).The original version of the test is found to have good interrater reliability (Gregory et al., 2002).In our current sample, Cohen's κ was run (N = 17) to determine if there was agreement between the two raters (i.e.interrater reliability) when scoring Faux Pas total scores and the score on item 6.We found here was fair agreement between the two raters on the total scores, κ = .346,p < .001,and substantial agreement on scores for item 6, κ = .667,p < .001.Internal consistency of the Faux Pas test is typically not calculated because of the low number of items and type of task (i.e. a behavioral task).

Procedure
After participates agreed to participate in the study, they gave written informed consent.After filling out the different questionnaires, participants were asked to listen to the different stories that the research assistant read to them.Participants had the possibility to read the stories on paper and could reread the stories when they wanted to.Subjects were instructed that they would get nine short stories and had to answer six questions following each of the story.Five of the nine stories contained a faux pas (stories 1, 2, 3 and 8 did not contain a faux pas).
An example of a story with a faux pas was: Kim's cousin, Stefan, would come to visit and Kim had for him made an apple pie.After dinner, she said, "I made a cake especially for you.Its in the kitchen."Hmmmm," Stefan said."It smells lovely!I like cake a lot, except apple pie of course." An example of a story without faux pas was: Ellen was waiting at the bus stop.The bus was late and she was waiting for a long time.She is 65 and was tired of standing so long.When the bus finally came it was full and all seats were taken.She saw a neighbor, Paul, standing in the aisle of the bus."Hi Ellen."He said."Did you have to wait long?" "About twenty minutes," she replied.A young man stood up and said: "Ma'am, would you like to have my seat?" Then different questions were asked: "Has anyone said anything inappropriate or something he or she had better not say?."The following questions were asked when the subject correctly detected a faux pas (the names were adapted to the story): (1) Who said anything inappropriate or something he or she had better not say?(2) Why did that person it better not say?
(3) Why do you think that person has said it?(4) A question that is focused on the understanding that the person who has said is not aware that it was awkward, for example: When he smelled the pie, he knew it was an apple pie?(5) How do you think Kim felt?Also two control questions were asked to see whether the participant had understand the story: (1) What cake had Kim baked in the story?(2) How Stefan and Kim know each other?Questions 2 and 4 were only asked when the subject recognized the faux pas correctly (i.e."Yes" on Question 1).For every correct answer, the participants got one point.For correctly reject a faux pas, the participants got two points.The total score was calculated by adding all points by adding together.There is also calculated a sub-score of question 6, this question measures affective theory of mind.

Statistical analyses
Statistical analyses were performed using SPSS Statistics version 23.First, because of the skewed distribution of the scores of the Faux Pas Test, and proactive aggression, relations between the different constructs (psychopathic traits, empathy, and faux pas) was examined using Spearman's rho correlation analyses.Then, using the PROCESS custom dialog box (Hayes, 2012), we performed mediation analyses on standardized scores to examine the relation between psychopathic traits and aggression and whether understanding of faux pas recognition mediate this relation.
For all analyses, α was set to .05.Although, when testing hypotheses, adjusting the alpha level when doing multiple comparisons is required (to control for inflation of the results), such adjustments are not necessary when analyses are exploratory in nature (as with the correlation analyses) (see Bender & Lange, 2001).

Results
Descriptive statistics of all the scales and subscales are presented in Table 1.

Correlational analyses
Correlation analyses were used to explore the relation between psychopathic traits, faux pas (ToM), trait empathy, and aggression (see Table 2).Results showed that total scores on the Faux Pas Test (faux pas recognition) were related only to the personal distress facet of the IRI.Scores on question 6 of the Faux Pas Test (impact of faux pas; affective ToM), did not correlate with trait empathy scores.
With respect to the relation between psychopathic traits and faux pas recognition, it was found that, only the boldness construct of the TriPM was negatively related to the total scores on the Faux Pas.This means that persons scoring higher on boldness, seem to be less likely to recognize a faux pas.Meanness and disinhibition were not related to scores on Faux Pas.Psychopathy construct scores were not correlated to question 6 of the Faux Pas Test, meaning psychopathic traits were not related to affective ToM.
When triarchic constructs were correlated with the trait empathy facets as measured with the IRI, it was found that boldness was only negatively related to personal distress (regarded as a facet of affective empathy), while meanness was associated with all the four facets of empathy, and disinhibition only with perspective taking (regarded as a cognitive aspect of empathy; ToM).
When psychopathic traits were related to self-reported aggression, it was found that both meanness and disinhibition correlated positively with reactive, and proactive aggression, while boldness was only positively related to proactive aggression.

Mediation analysis
Because exploration analyses showed only a relation between the boldness construct, faux pas recognition, and proactive aggression, we included these variables in the mediation analysis.Analysis showed that the model for predicting proactive aggression explained 12.3% of variation in proactive aggression scores and was significant, F(1, 102) = 8.12, MSE = .94,p =.005.In the model, boldness had a significant direct effect on proactive aggression scores (b = .37,SE = .10,95% CIs = .17,.56).Additionally, boldness was significantly related to faux pas recognition (b = −.28,SE = .10,95% CIs = −.47,−.08), however, faux pas recognition was not significantly related to proactive aggression scores (95% CIs = −.10,.28).This resulted in a non-significant indirect effect of boldness on proactive aggression (95% CIs = −.12,.05),meaning that faux pas recognition did not mediate the relation between boldness and proactive aggression scores.

Discussion
The aim of this study was to examine the role of ToM (and facets of empathy) in the relationship between the different triarchic psychopathic trait constructs and (proactive) aggression in a non-clinical population.Second, we had the aim to examine whether faux pas recognition would be a mediator in this relationship.Based on previous research, it was expected that psychopathic traits would have a positive relationship with proactive aggression and that a lack of affective ToM could explain part of this relationship.The first hypothesis was confirmed; the results showed that the different constructs of the TriPM were positively associated with proactive aggression, and also with reactive aggression.This result is in line with previous research (e.g., Bo et al., 2013;Cornell et al., 1996;Hecht et al., 2016).Because this relationship has been demonstrated in a sample of healthy community members (of which a large portion were students), this supports the assumption that psychopathy can be seen as a dimensional personality construct (Patrick et al., 2009).
Additionally, boldness, but not meanness and disinhibition, was negatively related to the total score of the faux pas test, reflecting mostly cognitive ToM.None of the triarchic psychopathy constructs were related to question 6 of the test (i.e.affective ToM).Since previous studies found a relation between psychopathy and affective empathy in particular, these findings are somewhat surprising.However, results did show that faux pas recognition was associated with personal distress, a facet of affective empathy, and not with cognitive aspects of empathy (i.e. with perspective taking, which is generally regarded as a measure of ToM).This would suggest that the Faux Pas Test assesses forms of affective empathy, and not so much (affective) ToM.Moreover, the finding that boldness, but not meanness is found to be related to faux pas scores was also not expected.
The hypothesis that e faux pas recognition would mediate the relation between boldness and proactive aggression was not conformed.No mediation effect of faux pas recognition was found.Because we found faux pas recognition to be related to affective aspects of empathy (and not ToM), this finding is in contrast with wat would be expected and also with previous results of Lozier et al. (2014), andTaubner et al. (2013), in which empathy did mediate the relation between psychopathic traits and proactive aggression.However, if we stick to what the Faux Pas test is supposed to measure (i.e.ToM), then findings are less surprising.Previous studies have found that cognitive ToM is not altered in individuals with psychopathic characteristics (e.g., R. Blair et al., 1996;Dolan & Fullam, 2004;Richell et al., 2003).Yet, affective ToM was affected in the study by Dolan and Fullam (2004).It is even more surprising that we found a negative association between boldness and faux pas, thus ToM, instead of better mentalizing abilities in persons scoring higher on boldness, because it was assumed that bold people would be better in mentalizing.But again, in our study faux pas is found to be related to affective aspects of empathy, and not with ToM.Also, the sub-score of the faux pas that would indicate affective ToM was not related to any self-reported empathy scores.These findings question the validity of the faux pas test in assessing mentalizing abilities.
When looking into the content of faux pas, one recognizes that it not only requires the person to take another person's perspective, it also requires that one does not care for the violation of social norms (i.e. a faux pas).This may be an explanation for the fact that in our study, persons scoring higher on boldness show less faux pas recognition, which may be consistent with the idea that bold persons in fact do not care about violating social norms.However, one would also expect this in persons scoring higher on meanness and disinhibition.
A possible explanation for the discrepancy between the current findings and those from Dolan and Fullam (2004) may be that they examined a prison population with persons who met diagnoses of antisocial personality and psychopathy, while the current study comprised of a non-clinical sample.Dolan and Fullam (2004) concluded that the lack of affective ToM was very subtle in the persons with psychopathy.Therefore, a relation between meanness and affective ToM in our sample may not have been detected.That is, although the faux pas task is developed to measure subtle mentalizing difficulties (unlike other ToM tasks), the faux pas task may not have been sensitive enough to detect difficulties in affective ToM in this sample.But again, the faux pas scores were only related to trait empathy scores on the personal distress subscale of the IRI, not with other scales.
The current results can also be interpreted as in line with a general assumption about persons with psychopathic traits.That is, it may be that individuals with psychopathic traits know how another person might feel (affective ToM), but they are less likely to be negatively aroused by another person's feelings (personal distress or emotional contagion).This is what may have happened in the current study.Persons may not have recognize the faux pas, when they did not feel that the character in the story would be upset by the violation of the social norm (i.e. they said that they did not understand what the problem would be).This is supported by the fact that the faux pas scores were positively related to personal distress and boldness was found to be negatively related to personal distress.These findings may support the Violence Inhibition Model (VIM; R. Blair, 1995), and the Somatic Marker Hypothesis (SMH; Damasio, 1994).These models suggest that psychopathy stems from an inability to connect emotions and physical sensations of behavior with the consequences of behavior.Consequently, individuals with psychopathic traits are not affectively impacted by emotions of another person (e.g., victim), and aggressive behavior is not inhibited in these persons.Importantly, however, results of the current study showed no relation between faux pas and personal distress on the one hand and proactive aggression, on the other hand.Therefore, the current results must be interpreted with caution.
The results from the current study add to the literature and theory on psychopathy because the specific relationship between boldness and ToM has not been explored before.Previous studies have found that boldness has a unique contribution in differentiating psychopathy from antisocial personality (Wall et al., 2015).Also, Esteller and colleagues (Esteller et al., 2016) found that boldness, but not meanness and disinhibition, was related to a reduced startle reflex potentiation (a physiological response to fear).Since there has been little attention to the relationship between boldness and ToM (or empathy in general), it is important to further elucidate this relation in future research.
Although the current research has some notable strengths (e.g., it included a ToM task instead of only self-reports), some limitations must be taken into account.The first limitation is that participants are included using convenience sampling, and therefore reflects mainly higher educated individuals.Therefore, although, our sample scored comparable with other community samples (e.g., Van Dongen et al., 2017), the current findings limit generalizability to other populations, such as incarcerated individuals.Another limitation is the fact that we used a correlational design and self-report measures.This limited the interpretation of causal relations in the mediation analysis.And although the use of self-reports is sometimes debated, recent studies showed that self-report measures correlate in a substantial way with interviewbased measures of psychopathy such as the Psychopathy Checklist-Revised (Edens et al., 2008), and are found to be negatively rather than positively related to social desirability (Verschuere et al., 2014).
Despite the above limitations, this study has important implications.It shows that also in the general population, a relation can be found between particular psychopathic traits and different types of aggression, and that boldness personality traits are associated with a diminished ability to detect a faux pas, which might be explained by alterations in personal distress.
Ultimately, current findings underline the importance of distinguishing different ToM and empathic abilities in relation to psychopathy and aggression, which is essential to develop effective interventions.

Conclusions
In conclusion, the current study adds to what we know about the relation between psychopathic traits and mentalizing abilities, thereby contributing to our knowledge of the process of understanding mental states in others and its relationship with proactive aggressive behavior.Also, it adds to a growing body of data pointing to the fact that persons with psychopathic traits do know what is right, but simply do not care.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Table 1 .
Psychometric properties for all measures.

Table 2 .
Pearson correlations for study variables.