Electrophysiological correlates of preparation and implementation for different types of task shifts.

The ability to flexibly shift between tasks is central to cognitive control, but whether the same brain mechanisms mediate shifting across different tasks is unknown. In this study, we investigated whether variations in stimulus-dimensions or response-mapping might influence task switching in terms of its preparatory processes, as reflected in cue-locked event-related potentials (ERPs), and its implementation processes, as reflected in stimulus-locked ERPs. Participants judged pairs of digits as same or different in one of two conditions. In one condition, the task-relevant stimulus-dimension was either repeated or switched across trials while the response-mapping rule was kept constant. In the other condition, the task-relevant stimulus-dimension was kept constant while the response-mapping rule was repeated or switched across trials. The length of the preparatory interval was manipulated. Data revealed switch-related preparatory ERP components (including N2 and a late slow positivity) that were associated with both types of task shifting and an N400-like negativity that distinguished between the two types. Several switch-related implementation ERP components associated with both types of task shifting were found at posterior sites. Distinct frontal modulations of the N1, P2, and N2 were found to associate with the implementation of the response-mapping shift, whereas a slow positivity was associated with the implementation of the stimulus-dimension shift. Therefore, this study demonstrates that there are shared and distinct processes across different types of task shifting. Finally, because the same transition-cue was used for different task shifts, the distinct processes cannot be explained simply by differences in cue processing.