Cognitive control over unconscious cognition: flexibility and generalizability of task set influences on subsequent masked semantic priming.

Recent studies using an induction task paradigm indicated that unconscious automatic processes underlying masked semantic priming are susceptible to cognitive control influences. In this paradigm, participants first perform different induction tasks (semantic decision vs. perceptual decision), which serve to activate a corresponding task set. Thereafter, the masked prime and the target for the lexical decision task is presented. Previously, perceptual and semantic induction tasks were presented in separate blocks, and the response to the induction task was given immediately after the inducing stimulus. The present study, therefore, tested two possible boundary conditions, flexibility of cognitive control and completeness of task set execution, for the emergence of task set effects on masked semantic priming. In the first experiment, perceptual and semantic induction tasks were presented in a randomized fashion, to assess whether task set influences on masked semantic priming can occur on a trial-by-trial basis. The other two experiments tested whether task set effects on masked priming survive, when the response to the induction task is delayed. The present study yielded the same pattern of results irrespective of the variations in the induction task paradigm: When the masked prime was shortly presented after the induction task, masked semantic priming was larger subsequent to the semantic than subsequent to the perceptual induction task. The present study shows that task sets can configure unconscious processing streams rapidly on a trial-by-trial basis and demonstrates the generalizability of cognitive control effects on masked semantic priming across variations of the induction task paradigm.