Implicit cueing of cognitive control states

According to classic models, control can be recruited or relaxed on demand (Braver, 2012; Norman & Shallice, 1986). In recent years, this assumption has been tested using congruency cues, namely by explicitly informing participants about congruency of the upcoming trial (e.g., Bugg & Smallwood, 2016; Wühr & Kunde, 2008). Overall, these studies suggest that while participants can exploit congruency cues to relax control (i.e., a cueing benefit emerges for congruent trials), it is much harder to recruit control on demand (i.e., cueing benefit mostly do not emerge for incongruent trials). However, the inconsistent benefits of congruency cueing may stem from the fact that such procedures rely exclusively on explicit, instruction-based information to induce control adjustments. Support for this idea comes from paradigms in which control adjustments emerge implicitly, as a function of prior exposure to congruent and incongruent trials, even in the absence of awareness (Braem et al., 2019). These findings suggest that control recruitment and relaxation may depend more strongly on implicitly learned regularities than on explicit expectations. The present study, therefore, tests this possibility by examining whether control states can be prepared implicitly through learning a regular sequence of congruency, without providing any explicit information about upcoming conflict. In order to reconcile these findings, the present experiment tests whether the explicit nature of congruency cues is the limiting factor for control recruitment in conflict tasks.