Error monitoring is related to processing internal affective states

Detecting behavioral errors is critical for optimizing performance. Here, we tested whether error monitoring is enhanced in emotional task contexts, and whether this enhancement depends on processing internal affective states. Event-related potentials were recorded in individuals with low and high levels of alexithymia—that is, individuals with difficulties identifying and describing their feelings. We administered a face word Stroop paradigm (Egner, Etkin, Gale, & Hirsch, 2008) in which the task was to classify emotional faces either with respect to their expression (happy or fearful; emotional task set) or with respect to their gender (female or male; neutral task set). The error-related negativity, a marker of rapid error monitoring, was enhanced in individuals with low alexithymia when they adopted the emotional task set. By contrast, individuals with high alexithymia did not show such an enhancement. Moreover, in the high-alexithymia group, the difference in the error-related negativities between the emotional and neutral task sets correlated negatively with difficulties identifying their own feelings, as measured by the Toronto Alexithymia Scale. These results show that error-monitoring activity is stronger in emotional task contexts and that this enhancement depends on processing internal affective states.