Helping or punishing strangers: neural correlates of altruistic decisions as third-party and of its relation to empathic concern

Abstract

Social norms are a cornerstone of human society. When social norms are violated (e.g., fairness) people can either help the victim or punish the violator in order to restore justice. Recent research has shown that empathic concern influences this decision to help or punish. Using functional magnetic resonance imaging (fMRI) we investigated the neural underpinnings of third-party help and punishment and the involvement of empathic concern. Participants saw a person violating a social norm, i.e., proposing unfair offers in a dictator game, at the expense of another person. The participants could then decide to either punish the violator or help the victim. Our results revealed that both third-party helping as well as third-party punishing activated the bilateral striatum, a region strongly related with reward processing, indicating that both altruistic decisions share a common neuronal basis. In addition, also different networks were involved in the two processes compared with control conditions; bilateral striatum and the right lateral prefrontal cortex (lPFC) during helping and bilateral striatum as well as left lPFC and ventral medial prefrontal cortex (vmPFC) during punishment. Further we found that individual differences in empathic concern influenced whether people prefer to help or to punish. People with high empathic concern helped more frequently, were faster in their decision and showed higher activation in frontoparietal regions during helping compared with punishing. Our findings provide insights into the neuronal basis of human altruistic behavior and social norm enforcement mechanism.

Keywords: empathic concern; reward; striatum; third-party help; third-party punishment.

Figure 1

Illustration of trial procedures (A) in the decision condition and (B) in the control condition. ISI, inter-stimulus interval; ITI, inter-trial interval.

Figure 2

Correlation (A) between IRI_EC scores (X-axis) and average help/punish rate (Y-axis) and (B) between IRI_EC scores (X-axis) and the difference in RT between help and punish (Y-axis). IRI_EC, empathic concern subscale of interpersonal reactivity index scale; RT, response time; ^*p < 0.05.

Figure 3

Separate and conjunction mapping of regions involved in third-party help and punishment (A) and timecourse of percent signal change in the local peak voxel of left striatum (B) and right striatum (C) in four conditions (i.e., help, help_control, punish, punish_control). Error bars: SEM.

Figure 4

Correlation between the contrast of help vs. punish and IRI_EC scores (A) and plots of the positive correlation between IRI_EC scores and contrast values in local peak voxel of left lPFC (B) and that of left IPL/AG (C). IRI_EC, empathic concern subscale of interpersonal reactivity index scale; lPFC, lateral prefrontal cortex; IPL, inferior parietal lobule; AG, angular gyrus.

Figure 5

Regions showing increased functional connectivity with bilateral striatum during third-party help decisions (compared with control conditions; (A) and plots of parameter estimates of PPI in the local peak voxel of right lPFC with left (B)/right (C) striatum in four conditions (i.e., help, help_control, punish, punish_control). Abbreviations: PPI, psycho-physiological interaction; lPFC, lateral prefrontal cortex; Stri, striatum; Error bars: SEM.

Figure 6

Regions showing increased functional connectivity with bilateral striatum during third-party punishment decisions (compared with control conditions; (A) and plots of parameter estimates of PPI in local peak voxel of left lPFC with left (B)/right (C) striatum and that of vmPFC with right striatum (D) in four conditions (i.e., help, help_control, punish, punish_control). PPI, psycho-physiological interaction; lPFC, lateral prefrontal cortex; vmPFC, ventral medial prefrontal cortex; Stri, striatum; Error bars: SEM.