Control over recollection varies with context-type: ERP evidence from the exclusion task.

The left-parietal ERP old/new effect—an index of recollection—is often larger for classes of item to-be-endorsed as old (targets) than to-be-rejected items (nontargets), and this has been interpreted as an index of selective retrieval. The question of interest here was whether selective retrieval would be more pronounced when targets are allocated according to distinct conceptual encoding tasks than when designated according to spatial location. Participants saw words on the left/right side of fixation and made a pleasantness or function judgment to each. Across test-blocks, target designation varied according to the kind of task judgment or the study location. Robust target old/new effects were observed for both classes of target designation but the nontarget amplitude was smaller when conceptual information was targeted. The current data indicate that the class of to-be-retrieved information determines the extent to which recollection can be controlled when all other factors are held constant.