Abstract
Deploying selective processing affords us the ability to identify and localize objects in a complex scene while non-selective processes such as extraction of global image properties and summary statistics allow for access to scene categories and ensemble representations. Together they give us a rich perceptual experience. Evidence suggests that the two types of processing do not work on the same time scale, but we still do not know whether the two interact and how. We employed an attentional blink paradigm examining how the temporary gap in visual awareness is modulated when performing two different tasks during rapid visual serial perception (RSVP). The stimuli were real scenes with the selective task requiring identification and localization to a quadrant of an object in the scene and the non-selective task requiring the categorization of the scene itself. In different blocks, observers are asked to perform two selective or non-selective tasks or two different combinations of these two tasks spaced at four different lags (300-1200 milliseconds) in the RSVP. There is a deep attentional blink when both tasks are selective with no full recovery even at lag 4. In contrast an attenuated blink with recovery at lag 2 when the selective task is preceded by a non-selective task. The pattern is different when the second task requires non-selective processing, with no blink for categorizing two scenes in succession and a deep blink that persists even at lag 4 when the scene categorization is preceded by object localization in a scene. There is a similar pattern of results when we use a much simpler field of dots stimuli. The findings show that sequentially allocating attention in a different manner to perform tasks that do or do not require selective processing results in interactions that indicate that selective and non-selective processing is not performed in parallel.