Abstract
A salient feature of executive function is the ability to rapidly alternate between tasks to allow online reconfiguration of attentional and motor goals. Notably, task-switching efficiency is asymmetrically dependent on the demands of the trial preceding a switch. For example, a prosaccade preceded by an antisaccade (i.e., task-switch trial) results in longer reaction times (RT) than when a prosaccade is preceded by its same task counterpart (i.e., task-repeat trial). In contrast, RTs for antisaccade task-repeat and task-switch trials do not reliably differ. The unidirectional prosaccade switch-cost has been attributed to the inertial activation of an executive-mediated non-standard task-set. It is, however, unclear whether the task-set inertia derives from the antisaccade requirements of inhibiting a pre-potent response (i.e., response suppression) or the decoupling of stimulus-response spatial relations (i.e., vector inversion). To address this issue, we arranged stimulus-driven (SD: i.e., saccade at target onset) and minimally delayed (MD: i.e., saccade at target offset) prosaccades in an AABB paradigm. MD prosaccades were employed because they provide the same response suppression as antisaccades without the need for vector inversion. Results showed that RT for SD task-switch trials were longer – and more variable – than their task-repeat counterparts, whereas values for MD task-repeat and task-switch trials did not reliably differ. Moreover, SD task-repeat and task-switch movement times and amplitudes did not vary and thus demonstrates that the aforementioned switch-cost was unrelated to a speed-accuracy trade-off. Accordingly, our results suggest that response suppression engenders the persistent activation of a non-standard task-set that proactively – and selectively – delays the planning of a subsequent stimulus-driven prosaccade.