Deontological and altruistic guilt: evidence for distinct neurobiological substrates

Abstract

The feeling of guilt is a complex mental state underlying several human behaviors in both private and social life. From a psychological and evolutionary viewpoint, guilt is an emotional and cognitive function, characterized by prosocial sentiments, entailing specific moral believes, which can be predominantly driven by inner values (deontological guilt) or by more interpersonal situations (altruistic guilt). The aim of this study was to investigate whether there is a distinct neurobiological substrate for these two expressions of guilt in healthy individuals. We first run two behavioral studies, recruiting a sample of 72 healthy volunteers, to validate a set of stimuli selectively evoking deontological and altruistic guilt, or basic control emotions (i.e., anger and sadness). Similar stimuli were reproduced in a event-related functional magnetic resonance imaging (fMRI) paradigm, to investigate the neural correlates of the same emotions, in a new sample of 22 healthy volunteers. We show that guilty emotions, compared to anger and sadness, activate specific brain areas (i.e., cingulate gyrus and medial frontal cortex) and that different neuronal networks are involved in each specific kind of guilt, with the insula selectively responding to deontological guilt stimuli. This study provides evidence for the existence of distinct neural circuits involved in different guilty feelings. This complex emotion might account for normal individual attitudes and deviant social behaviors. Moreover, an abnormal processing of specific guilt feelings might account for some psychopathological manifestation, such as obsessive-compulsive disorder and depression.

Figure 1

fMRI paradigm. The timing of each event in each trial is illustrated schematically. Each trial included the presentation of an emotional or neutral face, followed by a contextual sentence. The content of the sentence leads to two types of guilt (deontological and altruistic) and two control conditions (anger and sadness). At the end of each trial, subjects were asked to indicate, in a forced yes/no choice, whether they experienced guilt. Two examples are illustrated showing a trial with emotional face plus sentence inducing deontological guilt (A), and a trial with emotional face plus sentence inducing altruistic guilt (B).

Figure 2

fMRI behavioral responses. Behavioral responses showing frequency of target‐guilt answers across all eight conditions. (A) Neutral (N) and emotional (E) faces plus sentence inducing deontological guilt and anger, respectively. (B) Neutral (N) and emotional (E) faces followed by statements inducing altruistic guilt and sadness.

Figure 3

Effect of emotional faces. Brain activation for the comparison of all trials containing an emotional face (columns filled in red) versus trials with neutral faces (columns filled in black). Significant activation was observed in right superior frontal gyrus (BA6), in the inferior frontal gyrus (BA45), extending to the middle frontal gyrus (BA9) and the right orbitofrontal cortex (BA 47, not visible in this projection), and the insula bilaterally. The plot shows the BOLD signal changes in the right superior frontal gyrus for all experimental conditions. The activation map is rendered at a threshold of P _corr. = 0.05, at the cluster level. Deont., deontological; Altr., altruistic.

Figure 4

Effects of deontological and altruistic guilty feelings. Average activity for deontological and altruistic guilt as compared with the control conditions (anger and sadness) revealed greater activation in the anterior cingulate extending to the medial frontal gyrus, plus a cluster in the posterior cingulate. The localization of these clusters is highlighted on the sagittal anatomical section (red coloured; display threshold: P _corr. = 0.05, at the cluster level). The signal plot shows the level of activation at the peak of the anterior cluster, with the critical four guilt conditions highlighted in red. When each type of guilt (deontological or altruistic) was compared separately to the corresponding basic emotion (anger or sadness respectively), two different patterns of brain activation were identified. (A) Deontological guilt showed greater activation in the anterior cingulate bilaterally (in yellow) (B), while altruistic guilt induced greater activation in the left medial frontal gyrus (in blue) (C). For these additional comparisons, the two activation maps are displayed at a voxel‐level threshold of P _unc. = 0.005. Deont., deontological; Altr., altruistic.

Figure 5

Selective activation of the insula for deontological guilt. Anatomical localisation and signal plot of the left insula that showed a selective effect for deontological guilt (columns filled in red), i.e. the interaction: (deontological guilt vs. anger) > (altruistic guilt vs. sadness). A similar pattern of activation was found in the right insula (also visible in the coronal section), where this effect only approached statistical significance (see main text). The activation map is rendered at a voxel‐level threshold of P _unc. = 0.002, to reveal also the activation of the right insula. Deont., deontological; Altr., altruistic; L, left; R, right.