Empathy is a complex construct that determines our abilities for social interaction. It includes an affective component (affective empathy) that refers to the capacity of sharing emotions and respond to the emotions of others and a cognitive component (cognitive empathy) that partly overlaps with the concept of the theory of mind (ToM), namely, the individual ability to understand what other people think, and impute desires, intentions, and beliefs to oneself and others (Decety and Jackson, 2004; Decety and Moriguchi, 2007; Young et al., 2007; Blair, 2008; De Waal, 2008). ToM is an essential parameter in normal social interactions such as cooperating with colleagues and family members, thinking about others in their absence and anticipating their emotional reactions. As such, ToM is not only a cognitive construct but also involves an affective dimension based on the empathetic appreciation of the listener's emotional state (Shamay-Tsoory et al., 2009). Not surprisingly, ToM is severely affected in a variety of psychiatric disorders including autism, schizophrenia, bipolar disorder, and attention-deficit/hyperactivity disorder (Corcoran et al., 1995; Kerr et al., 2003; Abikoff et al., 2004; Blair, 2005; Bora and Pantelis, 2016; Maoz et al., 2019). Recent insights proposed that in neurotypical individuals, there are two ToM systems: one controlled that acts when we deliberately considers other's thoughts and emotions and the other implicit that involves the automatic analysis of their viewpoints even when such analysis is irrelevant for task processing (Onishi and Baillargeon, 2005; Surian et al., 2007; Kovacs et al., 2010; Schneider et al., 2012). Implicit ToM is thought to be developed early during development and remain stable over life span (Schneider et al., 2012). Although there is a wide agreement that adult humans are able to engage in unconscious analyses of others' mental states (Schneider et al., 2017), several studies led to conflicting data regarding the reliability of the implicit ToM using non-verbal measures such as violation of expectation paradigms and interactive and anticipatory looking tasks (for review, see Kulke and Hinrichs, 2021).

Automatic perspective taking (APT) seems to be the most reliable facet of implicit ToM. Using the dot-perspective-taking task developed by Samson et al. (2010, Experiment 1), previous studies demonstrated that people automatically take into account others' perspective even when it prevents them from achieving their own goals (Qureshi and Monk, 2018; Qureshi et al., 2020). Although there is an ongoing theoretical debate of whether automatic interference effects in the dot-perspective task are the product of domain-specific perspective-taking processes or of domain-general submentalizing processes (for review, see Cole and Millett, 2019 and Westra et al., 2021), recent lines of evidence suggest that this human ability may be of key importance for the ascription of mental states and social interactions (Furlanetto et al., 2016; Gao et al., 2019). In particular, Drayton et al. (2018) reported that psychopaths are able to represent others' perspective in goal-conducive tasks but show a striking ability of ignoring it in non-goal-relevant situations. They postulated that their lack of the ability to automatically represent the belief states of others when it does not serve their own ends may be at the origin of their maladaptive social behavior. However, this study included only offenders from a high-security hospital without control groups.

Automatic perspective taking is a complex phenomenon that may be impacted by attention and global intelligence as well as levels of social cognition (Schneider et al., 2012; Burnside et al., 2017; Pineda-Alhucema et al., 2018). To date, there is no study addressing the cognitive and emotional determinants of APT in healthy adults. Our hypothesis is that APT is independent of explicit ToM as well as affective empathy performance but may be affected by impulsivity and attention as well as low levels of global intelligence. To address this issue, we examined the performance in the Samson et al. APT task in 91 healthy adults who underwent a detailed neuropsychological testing including assessment of their general intelligence, attention and impulsivity, alexithymia, and measures of affective empathy and explicit ToM.

The descriptive data of the present sample are summarized in Table 1. Among the independent variables included in the regression models, only the CPT-II parameters were significantly associated with the mean reaction times after Benjamini–Hochberg correction for multiple comparisons.

CPT-II HRT and hit RT SE (two measures of inattentiveness and impulsivity) were positively associated with the mean reaction time in the Self condition [regression coefficient 5.58 (95% CI: 1.52, 9.63) and 5.33 (95% CI: 1.78, 8.87), respectively]. The mean reaction time in the Other condition was positively related to CPT-II HRT [regression coefficient 4.71 (95% CI: 1.40, 8.02)] (Table 2).

The WAIS global score was negatively associated with mean reaction time in the Self condition [regression coefficient for the Self condition: −2.48 (95% CI: −4.40, −0.56)]. In the Other condition, this association was also statistically significant [regression coefficient for Other condition −2.37 (95% CI: −4.15, −0.59)]. Similar data were obtained for some of the WAIS subscales. However, these associations did not persist after correction for multiple comparisons (Table 2).

The increased number of commissions as well as increased detectability and HRT-SE in CPT-II were all related to increased number of errors in Self condition (regression coefficients from 0.03 to 0.04). CPT-II HRT-SE was also related to worse performances in Other condition (regression coefficient 0.02) (Table 3).

The GeSoCS score was negatively associated with the number of errors in Self condition [regression coefficient: −0.07 (95% CI: −0.12, −0.03)] (Table 3).

Among the variables studied, only the CPT-II parameters have a significant impact on egocentric and altercentric interference. Egocentric interference for the number of errors was negatively related to CPT-II HRT-SE [regression coefficient: −0.03 (95% CI: 0.05, 0.01), adjusted p = 0.0034]. Altercentric interference for the number of errors was negatively related to CPT-II Omissions [regression coefficient: −0.04 (95% CI: 0.05, 0.02), adjusted p = 0.0002], Commissions [regression coefficient: −0.04 (95% CI: 0.07, 0.01), adjusted p = 0.0086], and HRT-SE [regression coefficient: −0.04 (95% CI: 0.07, 0.02), p = 0.0005; Table 4]. No significant association was found between CPT-II variables and both egocentric and altercentric interference for the mean reaction times.

Our data show that CPT-II parameters of attention and impulsivity have a direct and strong impact on APT performances assessed with the dot-perspective task in healthy adults. Attention and impulsivity levels are associated with both mean reaction times and mean number of errors in the current experimental paradigm. Although intuitively evident, the impact of attention and impulsivity is quite different in Self compared to Other conditions. When participants focus on their own perspective, their mean reaction time increases as a function of the two measures of inattentiveness and impulsivity, namely, the hit RT, a measure of response time, and HRT-SE, a measure of the consistency of response time in CPT-II. In contrast, when participants report the avatar's view, their mean reaction time decreases in cases with higher number of omissions (that assess the level of inattentiveness). In other words, people with decreased levels of attention and higher impulsivity may need more time to define their own perspective (both in consistent and inconsistent trials) but tend to precipitate their response when they should take the perspective of others. Most importantly, CPT-II parameters assessing inattentiveness and impulsivity are associated with increased number of errors both in Self and Other conditions. Of importance, these parameters were also negatively related to egocentric and altercentric interference at the level of mean number of errors but not mean reaction time. Gardner et al. (2018) reported that attentional orienting contributes decisively to performance in the dot-perspective task (Gardner et al., 2018). More recently, O'Grady et al. (2020) postulated that APT is a rapid, unconscious, and involuntary phenomenon, referred to as spontaneous but not automatic, that needs preserved attention. The present findings parallel these observations, suggesting that people with lower attentional resources display worse performances in terms of number of errors, are less able to manage their reaction time both in Self and Other conditions, and are less responsive to both egocentric and altercentric interferences. A very recent study by Qureshi et al. (2020), using the same experimental paradigm, demonstrated that conflict indices (consistent-inconsistent perspective) are related to stop-signal inhibitory control task (but not go–no-go or shape-matching tasks) and indicate that this selection may occur very late in cognitive processing. In contrast, altercentric interference was not related to any of the executive tasks in this study. Taking together, these data indicate that attentional resources rather than inhibitory control abilities may impact on self–other interference when performing a visual perspective task. Pointing to the importance of adjusting for global empathy levels, the mean number of errors in our cohort was negatively associated with the GeSoCS score, a composite measure of explicit ToM and affective empathy. However, this global empathy score was not related to mean reaction times and did not impact on egocentric and altercentric interference. Neither global intelligence nor alexithymia have an effect on dot-perspective task performances. This observation agrees with several reports showing that dot-perspective task performances may be preserved in psychiatric pathologies affecting intelligence but also emotional regulation such as autism spectrum conditions and alcoholism (Pearson et al., 2013; Cox et al., 2016, 2018; Sijtsma et al., 2020).

To our knowledge, this is the first study that analyzes the cognitive and emotional determinants of APT performance in community-dwelling young adults. Our results indicate that APT mostly depends on attentional resources and impulsivity even in healthy adults. It is likely that patients with specific alterations of attention and impulsive control, such as patients with attention deficit and hyperactivity disorder (ADHD) and borderline personality, may fail to activate this human ability. The strengths of this report include its use of a large sample, independent variables assessing cognitive processes such as attention and fluid intelligence, general levels of empathy, but also emotional regulation (alexithymia, impulsivity) and stringent control for multiple comparisons across the regression models. Some limitations should also be noted. All of our cases were socially integrated young individuals without history of criminal convictions and substance abuse. The careful exclusion of neurological and psychiatric disorders as well as regular use of psychotropics and CPT-II performances within the normal range limits the generalizability of our observations. Moreover, no direct assessment of inhibitory control and cognitive flexibility was performed as part of the neuropsychological analysis. These parameters are known to impact performances on visual perspective taking (Qureshi and Monk, 2018; Qureshi et al., 2020; Li et al., 2021). Future studies in mixed series including patients with ADHD and borderline personality, investigation of social parameters (job loss, separation), and careful analysis of executive functions are needed to obtain a better insight into the social relevance of APT concept.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Commission cantonale d'éthique de la recherche (CCER). The patients/participants provided their written informed consent to participate in this study.

PG, AP, and M-LM: conceived the study. M-LM: conceived the visual perspective-taking task. CR and M-LM: recruitment. M-LM, CR, and PG: neuropsychology supervising. M-LM and FH: data preparation. FH, PG, and M-LM: analyzed the data. PG, AP, FH, and M-LM: manuscript writing. All authors contributed to the article and approved the submitted version.

This research was funded by the Swiss National Foundation (Grant number 320030_189247).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.