Abstract
Traditional theories of attention typically posit that visual recognition requires focal attention. Accumulating findings in the last decade, however, suggest that real-world objects, relative to meaningless arbitrary items, can be detected and categorized even when presented very briefly at an unattended location. Most of the studies supporting this view of attention-free processing use dual-task paradigms, in which participants perform a highly demanding task at fixation, while simultaneously detecting a pre-specified object category (e.g., an animal or vehicle) at the periphery. A major limitation of this method is that the supposedly 'unattended' peripheral object is in fact a target stimulus, thus participants intentionally allocate attention to its location.
To overcome this limitation, we assessed object categorization under conditions in which an object is strictly irrelevant to task-requirements and to response-selection processes. Participants performed a forced-choice classification task of briefly presented pairs of objects (e.g., 'is there a nonsense object in the display, or not'). Within pairs of real-world objects, items either belonged to the same category (e.g., a tiger and a spider; a car and a boat) or to different categories (e.g., a dog and a jeep; a bus and a mouse). When participants performed the task for the two objects in a pair (i.e., both objects were attended), RT for same-category pairs was significantly shorter than for different-category pairs, indicating that object category was registered. When participants were cued to respond to only one of two objects in a pair, such that the other object served as an unattended distractor, no categorical effect was observed. Subsequent experiments revealed that unattended object distractors affect behaviour only if they are relevant to response-selection processes (i.e., compete with response to target). Our results suggest that when unattended objects are strictly task-irrelevant they are not automatically categorized, thus refuting previous claims of attention-free processing.
Meeting abstract presented at VSS 2012