Abstract:

Sense of agency refers to the subjective experience of controlling one’s actions and, through them, external events, which is the core element that makes individuals responsible for their behavior to foster social cohesion. Recent studies have revealed that people potentially experience reduced sense of agency for harming others by claiming that they were obeying orders. However, little is known about the cognitive neural mechanism behind the reduced sense of agency when individuals are forced to inflict physical harm on others. This study adopted a temporal estimation task to investigate the internal mechanism of voluntary action on sense of agency and the moderating effect of social distance.

In Experiment 1, participants were asked to trade monetary gains for themselves off against painful electric stimuli experienced by friends or strangers, subjectively estimated the perceptual temporal interval between keypress actions (i.e., free, or coercive actions) and consequent neutral outcomes, and rated the feeling of control and responsibility (see Figure 1).

The results found that when obeying commands, compared to shocking strangers (427.11 ± 156.34), the estimated time interval for shocking friends (394.50 ± 139.89) was shorter, t (34) = 3.29, p = 0.002, Cohen’s d = 0.22, indicating a stronger implicit sense of agency. However, during free choice, the difference in estimated time intervals between shocking friends (413.48 ± 161.41) and shocking strangers (407.80 ± 155.74) was not significant, t (34) = 0.48, p = 0.638 (see Table 1).

In terms of the feeling of control, the results showed a significant main effect of voluntary action, F(1, 34) = 25.56, p < 0.001, η²_p = 0.43, with participants reporting lower ratings of feeling of control over electric shocks when obeying commands (4.01 ± 1.58) compared to when they had free choice (5.57 ± 1.22).

In terms of the feeling of responsibility, the results also showed a significant main effect of voluntary action, F(1, 34) = 9.81, p = 0.004, η²_p = 0.22, with participants reporting lower ratings of feeling of responsibility for the electric shocks when obeying commands (4.87 ± 1.30) compared to when they had free choice (5.53 ± 1.04).

Experiment 2 further investigated the cognitive neural mechanism of the interaction between voluntary action and social distance on sense of agency using event-related potential (ERP) technology and attempted to examine the robustness of explicit sense of agency.

Figure 2 illustrates the differences in the ERP components elicited by electric shock decisions across different experimental conditions, followed by a repeated measures analysis of variance (ANOVA). In the decision-induced N1 component (100~150 ms), the results showed that when obeying commands, shocking strangers (−0.61 ± 2.03 µV) elicited a more negative N1 amplitude compared to shocking friends (0.29 ± 2.65 µV), t (22) = 2.97, p = 0.007, Cohen’s d = 0.38. However, there was no significant difference between the two when participants had free choice, t (22) = 1.00, p = 0.326 (see Figure 3).

In the decision-induced N2 component (250~350 ms), the results indicated that the main effect of voluntary action was significant, (Frontal: t (22) = 2.81, p = 0.010, Cohen's d = 0.36; Frontal-Central: t (22) = 3.45, p = 0.002, Cohen's d = 0.50; Central: t (22) = 3.91, p = 0.001, Cohen's d = 0.67), suggesting that the N2 component was more negative in the free condition compared to the coercive condition (see Figure 3).

Regarding the decision-induced P3 component (350~450 ms), the results found that when obeying commands, shocking friends (5.12 ± 2.85 µV) elicited a larger P3 amplitude compared to shocking strangers (3.60 ± 3.81 µV), t (22) = 2.12, p = 0.045, Cohen's d = 0.45. However, there was no significant difference between shocking friends and shocking strangers when participants had free choice, t (22) = 1.23, p = 0.231 (see Figure 3).

In the tone-induced N1 component (70~200 ms), it was only found that when obeying commands, shocking friends elicited a more negative N1 amplitude compared to shocking strangers (Parietal: t (22) = 2.77, p = 0.011, Cohen's d = 0.50; Parietal-Occipital: t (22) = 2.67, p = 0.014, Cohen's d = 0.49) (see Figure 4).

Additionally, there was a significant correlation between the ratings of feeling of control and the decision-induced N2, r = −0.33, p = 0.025, whereas the implicit and explicit senses of agency were not correlated in both Experiment 1 and Experiment 2, ps > 0.05.

These results indicate that social distance moderates the implicit sense of agency in the face of serious unethical outcomes, while participants’ feeling of control and responsibility were not affected by social distance due to self−serving bias. When obeying orders, individuals consciously exhibit moral disengagement. The present study has demonstrated that implicit and explicit sense of agency have different processing mechanisms, which broadens the previous insight into understanding the sense of agency. We discuss the implications of utilizing free choice and closing the social distance with others as significant strategies for those experiencing the abnormal sense of agency.

Key words: sense of agency, voluntary action, social distance, time interval estimation, event-related potential (ERPs)