Abstract
One claim of the recent Boolean map theory of visual attention (Huang & Pashler, 2007) is that visual information that can be consciously accessed at one moment is limited to one feature, linked to possibly multiple locations (i.e., a map). We report new tests of the predictions. First, it should be efficient to detect a missing location, but not a missing feature, because multiple locations but not multiple features can be simultaneously represented (thus allowing a missing location, but not a missing color to be spotted). Second, the claim of “single feature access” should hold for a wide range of features including orientation, spatial frequency, motion direction, etc. Third, we should be unable not only to access multiple features in newly appeared objects but also to continuously monitor multiple features after they have been seen (i.e. instantly report a change on any of them), because the claim of “single feature access” implies only one single feature can be included in the content of visual access at one time, not one single feature can be added into the content of visual access at one time. Fourth, we should have greater difficulty judging the simultaneity of two features (appearing in objects) than judging the simultaneity of two locations (filled by objects), because the serial nature of feature access disrupts the direct verification of simultaneous presence. Fifth, it should take longer to retrieve objects of different colors from working memory than to retrieve objects of the same color from working memory, because retrieving objects of different colors has to be done sequentially whereas retrieving objects of the same color can be done simultaneously. These predictions were tested and generally confirmed. Their implications will be discussed.