Guidance by visual and verbal representations during visual search

Visual search is guided by target representations held in visual working memory (attentional templates). While several studies have explored how visual-perceptual representations of target defining features guide attention during visual search, much less is known about visual search guided by verbal representations. In this study, we used the contralateral delay activity (CDA) and the N2pc component of the event-related potential to measure template activation and target selection, respectively, during search guided by visual-perceptual versus verbal templates. Each trial started with the presentation of a cue display indicating one or two task-relevant colour(s) for the upcoming search. In different blocks, cues were either coloured squares (visual-perceptual cues) or the initial letters of the colour words (R for red; verbal cues). After a retention period of 1000ms, search displays with six differently coloured bars appeared and participants reported the orientation of the bar in (one of) the cued target colour(s). Target N2pcs were virtually identical in the visual-perceptual and verbal cueing tasks, both in one- and two-colour search. However, CDA components, measured during the retention period, were substantially increased in response to verbal as compared to visual-perceptual cues. These results suggests that the verbal cues were translated into visual-perceptual representations during the retention period (CDA), so that when the search display arrived, visual-perceptual colour representations were available to guide target selection equally efficiently in both search task (N2pc). We tested this hypothesis in a follow-up experiment in which we shortened the retention period to 300ms to reduce the time for the translation from verbal code to visual-perceptual representations. N2pc components were now substantially attenuated and delayed in the verbal versus visual-perceptual cueing task, demonstrating impaired target selection during visual search when verbal target representations cannot (sufficiently) be translated to a visual-perceptual representation before the arrival of the visual search display.