Cognitive Control as Alternation of Activation and Suppression in the Stroop Task

Execution of tasks in which information rapidly changes its relevance employs both excitatory and inhibitory top-down control. This assertion is derived from empirical data and computational cognitive modeling of a screen-based Stroop task. Empirical data showing both within-trial (interference and facilitation) and between-trial (negative priming) effects are presented. A computational cognitive model developed within the ACT-R architecture is proposed. This model uses a spreading activation mechanism that activates information pertaining to the current trial and suppresses information pertaining to the preceding trial. A good fit of this model to human data is reported and the neuro-cognitive plausibility of top-down suppression is discussed.