The unidirectional prosaccade switch-cost: Correct and error antisaccades differentially influence the planning times for subsequent prosaccades

- Prosaccade latencies are longer when preceded by an antisaccade (i.e., switch-cost).
- We examined if correct vs. error antisaccades differently influence the switch-cost.
- Results showed that correct, but not error, antisaccades elicit a switch cost.
- Correct antisaccades inhibit the planning of stimulus-driven oculomotor networks.

Antisaccades produce longer reaction times (RT) than their prosaccade counterparts and this latency increase has been linked to an oculomotor 'pre-setting' that prevents the evocation of a stimulus-driven prosaccade. Moreover, a consequence of oculomotor pre-setting is a lengthening of the RTs associated with a subsequent prosaccade. The goal of the present study was to determine whether the constituent elements associated with planning a correct antisaccade (i.e., response suppression and vector inversion) imparts a residual delay that inhibits the planning of a subsequent prosaccade. To that end, participants alternated between pro- and antisaccades in a pseudo-randomized task-switching schedule (e.g., AABBAAB…) and responses were cued via a paradigm that was designed to evoke frequent error antisaccades (i.e., a saccade initially, and incorrectly, planned to the target stimulus). Results showed that RTs for correct antisaccades were longer than error antisaccades and that prosaccades preceded by the former, but not the latter, trial-type were associated with a reliable increase in RT (i.e., prosaccade switch-cost). In other words, error antisaccades were associated with a failure to withhold a stimulus-driven prosaccade and did not delay the planning of a subsequent prosaccade. Based on these findings we propose that the prosaccade switch-cost is not related to an explicit awareness of task goals; rather, our results are consistent with the assertion that a consequence of response suppression and vector inversion is a residual inhibition of stimulus-driven oculomotor planning networks.