How a co-actor's task affects monitoring of own errors: evidence from a social event-related potential study

Efficient flexible behavior requires continuous monitoring of performance for possible deviations from the intended goal of an action. This also holds for joint action. When jointly performing a task, one needs to not only know the other's goals and intentions but also generate behavioral adjustments that are dependent on the other person's task. Previous studies have shown that in joint action people not only represent their own task but also the task of their co-actor. The current study investigated whether these so-called shared representations affect error monitoring as reflected in the response-locked error-related negativity (Ne/ERN) following own errors. Sixteen pairs of participants performed a social go/no-go task, while EEG and behavioral data were obtained. Responses were compatible or incompatible relative to the go/no-go action of the co-actor. Erroneous responses on no-go stimuli were examined. The results demonstrated increased Ne/ERN amplitudes and longer reaction times following errors on compatible compared to incompatible no-go stimuli. Thus, Ne/ERNs were larger after errors on trials that did not require a response from the co-actor either compared to errors on trials that did require a response from the co-actor. As the task of the other person is the only difference between these two types of errors, these findings show that people also represent their co-actor's task during error monitoring in joint action. An extension of existing models on performance monitoring in individual action is put forward to explain the current findings in joint action. Importantly, we propose that inclusion of a co-actor's task in performance monitoring may facilitate adaptive behavior in social interactions enabling fast anticipatory and corrective actions.