Abstract
Previous studies suggest that 1) pre-activating a visual representation of an item in visual working memory (VWM) prioritizes access to visual awareness for this item and that 2) VWM can contain representations of bound items instead of separate features. It is currently unclear whether VWM affects access to visual awareness at a feature level, a bound conjunction level, or both. To investigate this question, we conducted a series of experiments in which we combined a delayed match to sample task with a breaking Continuous Flash Suppression (b-CFS) task. In each trial, subjects memorized an object consisting of a disk with two halves with different colors for the later recall test and, between them, had to detect the location of a target presented under suppression. We varied the congruence in colors between the memory representation and to-be-detected target. Our results show that memory congruent objects (consisting of a conjunction of features) break CFS faster than memory incongruent objects. Interestingly, we also observed this congruence effect when we presented the memorized object in a horizontally-mirrored configuration of colors. There was no difference between breakthrough RTs when only a single feature in the target was congruent with the memory representations. Our results suggest that VWM prioritizes bound objects for access to visual awareness. We are currently investigating the reason for the similarity of RTs between the memory congruent conjunction and mirrored memory congruent conjunction conditions in additional experiments.