Working memory modulates attentional control as revealed by sequential effects

Sequential effect refers to the phenomenon that interference in conflict paradigms is smaller after incongruent trials than after congruent trials in tasks for which the stimuli contain irrelevant information. Currently, there are two major groups of interpretations of the sequential effects. One group suggests that enhanced attentional control after perceiving conflict in the preceding trial causes the modulations on attentional processing in the current trial and thus leads to sequential effects, a view also called control adjustment (Botvinick et al., 1999). A second group suggests that some non-attentional process contributes to the sequential effects (Hommel et al., 2004). Another issue is whether sequential effects can occur at the non-response level, as current studies have mainly focused on the response-level conflict. A third issue is related to the view of control adjustment: Does the enhanced attentional control after conflict work by enhancing the excitatory processing of the task-relevant information, or by modulating the inhibitory processing of the task-irrelevant information? Three experiments were designed to address these issues. Experiments 1 and 2 found that sequential effects can occur at both the non-response and response levels, but only when the conflict in the preceding trial and the current one are the same type. Experiments 2 and 3 found that interference changed not only with the preceding conflict but also with matching between the two consecutive trials. These effects were not observed in neutral trials. Further, variations on behavioral interference with preceding conflict and with matching were associated with parallel changes on N300 and SP, the two event-related potential components that have been suggested to be involved in conflict processing (West, 2003). Together, these findings suggest that sequential effects can occur at multiple levels and may be caused by control adjustment rather than non-attentional processes. Further, the control adjustment can be elicited by conflict from the perceptual stimuli and from the contents of working memory, with both modulations occurring via working memory, thus indicating the influence of working memory on attentional control. Finally, such modulation may occur by modulating the inhibitory processing to distractors rather than the excitatory processing to targets