The present study investigated the extent to which psychopathic traits would influence one’s tendency to empathize with people in a wide variety of different social situations. To this end, on each trial participants were presented with a description of people in various scenarios and given the free choice to either: a) empathize for the target’s benefit, b) empathize for their own benefit, or c) merely observe. Two a priori hypotheses were formed about the relationship between empathic choices and psychopathic traits: 1) level of psychopathic traits would be unrelated to total empathic choices, but 2) level of psychopathic traits would influence the type of motivation that underlay the choice to empathize when it was chosen. Specifically, we hypothesized that PPI-R-40 scores would correlate positively with the frequency that empathy was chosen for non-virtuous motives, and (potentially) negatively with the frequency that empathy was chosen for virtuous motives.
Hypothesis one was supported. Overall, participants chose to empathize on 74% of all trials, and this was largely stable across the spectrum of PPI-R-40 scores. Thus, psychopathic traits were largely unrelated to the overall frequency of choices to empathize. This counters a considerable literature that has reported negative relationships between psychopathic features and levels of empathy [e.g. 26, 27, 28, but see 29]. However, it should be noted that most of this existing work has evaluated empathic abilities, not propensities [see 25]; and further, that the work that has evaluated propensities has rarely done so within a free-choice environment. Our study thus stands somewhat separate from most existing work, and in so doing suggests that – regardless of their actual ability to do so – those with heightened psychopathic traits may choose to empathize just as often as anyone else (see also [30] regarding psychopathy and empathy not being mutually exclusive and [31] regarding labelling high psychopathy / high empathy as Dark Empaths).
Our second hypothesis was not entirely supported, however. While some hypothesized effects were seen at the PPI-R-40 subscale level (e.g. Machiavellian egocentricity correlated positively with non-virtuous motives; Social influence correlated negatively with virtuous motives), the hypothesized relationship with total PPI-R-40 scores was not found. Thus, all participants, regardless of total psychopathic traits chose to empathize for virtuous and non-virtuous reasons with approximately equal frequency. In some respects, this too supports the notion that those with heightened psychopathic traits don’t differ in their predispositions towards empathy. However, the non-relationship between psychopathic traits and frequency of non-virtuous empathy did run somewhat counter to participants’ own self-reported motivations to empathize on the Motivation to Empathize scale (within which those with heightened psychopathic traits did show expected higher levels of non-virtuous motivations). Thus, there appears to be some discrepancy between participants’ self-reported motivations to empathize and their actual choices made on the Empathic Choice Task. Such self-reported/behavioural discrepancies are not unusual – indeed, it has long been recognized that self-reported intentions do not necessarily predict real-world behaviour [e.g. 32], in part because social desirability and impression management can lead people to self-report what they believe they should do, rather than what they would actually do [20; see also 21, 22, 23 regarding self-report of empathy specifically]. However, it is interesting that this discrepancy between MtE and ECT scores increased as a function of psychopathic traits. One possibility is that the self-reports from individuals with heightened psychopathic traits are less trustworthy (due either to intended deception or lack of insight). Alternately, it could be that individuals with psychopathic traits were more likely to try to manage their responses on the ECT in line with social desirability demands. While this may seem counter to the concept of the psychopathic individual as a violator of social norms, recent research suggests that psychopathic individuals may be particularly likely to employ impression management strategies to try to manipulate the perceptions of others (but only in public; in private, the opposite appears true [33] (see also [34] which has reported that the self-concepts of individuals with dark triad traits tend to be more state- than trait-dependent).
If the participants with psychopathic traits were particularly aware of the social/contextual demands of the study, then we may anticipate that differences in their empathic choices would reveal themselves in a context-dependent manner. To this end, exploratory analyses revealed that the empathic choices of individuals with heightened psychopathic traits were particularly sensitive to the power dynamic in the depicted scenarios (see also [35]). Specifically, participants with heightened psychopathic traits more frequently chose to empathize for non-virtuous reasons when the scenario placed them in a position of power over the target, yet more frequently chose to empathize for virtuous reasons when the scenario placed the target in a position of power over them. This suggests that when the participant with heightened psychopathic traits felt they were in a position of power they were more likely to act selfishly, for their own gain; in contrast, when they felt they were in a less powerful position they were more likely to act graciously, for the benefit of the other. This pattern was particularly so for individuals higher in the Machiavellian Egocentricity subscale of the PPI-R-40. It has long been recognized that a strategic sensitivity to the intricacies of state-dependent power dynamics can be useful – for instance, for building up one’s reputation or for gaining status and resources [36]. For instance, individuals with lower power may be more motivated to gain the goodwill of those in higher power, while individuals with higher power may be more motivated to gain information and exert influence [37]. We may then speculate that the tendency for those with heightened psychopathic traits to empathize virtuously when targets held power over them could represent an attempt to gain power (see [38]) or a profitable connection [39] by behaving cooperative or conciliatory. In contrast, their tendency to empathize non-virtuously when they held power over the targets could represent an attempt to exploit or dominate [see 40, 41, 42]. Of potential import, both of these cases may be viewed as examples of the individual with heightened psychopathic traits employing empathy in a contextually specific way, as a function of their own unique (but not illogical or irrational) situational goals and motivations [see 43]
This characterization of the psychopath as a logical and rational being has a consistent presence in both academic and popular writings. And yet this characterization has rarely infused itself into underlying etiological theories of the disorder. Rather, most theories continue to exist as deficit-based theories, wherein the psychopathic individual is deemed victim to cognitive or emotional dysfunction that prevents them from making wise, or socially appropriate, decisions [e.g. 4, 5, 6]. Certainly, the present findings would require replication; however, they also contribute to a growing body of work that has been increasingly questioning these deficit-based theories [e.g. 12, 18, 28, 45]. Instead, they support conceptualization of the psychopath as a rational agent whose emotional tendencies may be different, but not necessary dysfunctional (see [46, 47, 48]).
The study is not without its limitations, of course. First, while the sample size (N = 158) was largely appropriate to test hypothesized effects it may have been somewhat underpowered for identifying some of the interactive effects that were investigated. For instance, the power-related effects – which essentially represent a 3-way interaction – only reached trend status. We anticipate that a larger sample will have a greater opportunity to reveal the full nature of this effect, and welcome future work oriented towards this goal.
Second, recruitment relied exclusively on undergraduate psychology students from a post-secondary institution in Canada, and measurement of psychopathic traits was handled via a well-validated self-report instrument. A rich literature supports the validity of these recruitment and assessment procedures; nonetheless, the extent to which the present results will generalize to general populations or to forensic populations with PCL-R-assessed psychopathy, will require additional testing.
Third, the ECT is a relatively new measure of behavioural empathy that has yet to receive comprehensive validity/reliability testing. That said, we found considerable overlap between participants’ responses on the ECT and the MtE, which has previously received independent validation [16]. We welcome future work that aims to further assess the response validity of the ECT (our lab is currently invested in several related projects), however, this ECT/MtE relationship does provide valuable preliminary evidence for the validity of the task. Finally, while the ECT may be viewed as a significant improvement on self-reported assessments of empathy, by assessing decisions to at, the task itself still technically relied on self-report. That is, participants still self-reported how they would act in each presented scenario. Thus, ECT metrics may remain somewhat influenced by participants’ willingness and/or ability to accurately report how they would act in those scenarios.
Future studies could examine whether these results would be found among other populations, especially clinical psychopathy populations. Studies could further attempt to elucidate what contextual factors might promote engaging in virtuous or non-virtuous empathy among psychopathic populations. Would individuals with high psychopathy show a distinctive increase in non-virtuous empathy once specific goals are induced (such as those that align with values of power, excitement, or promotion)? Would a larger inclusion of others in their ‘in-group’ be associated with higher virtuous empathy (motivations or choices)? Would measuring participants’ perceptions of long-term benefit/gain when choosing to empathize reveal some choices denoted as ‘virtuous’ as being, instead, underlyingly non-virtuous (and would such underhandedly non-virtuous choices correlate with psychopathy)? Additionally, further research could seek to control for the influence or measure the contribution of factors such as emotional intelligence, social-desirability, and depth of engagement and mentalization in the social scenarios for individuals with psychopathy. Finally, as empathy is a mental behaviour (rather than observable), future studies could aim to further clarify the relationship between virtuous/non-virtuous empathic choices and pro-social/anti-social behaviour (as well as the impact of psychopathy and motivations).