Abstract
Purpose: Object recognition in visual search can be conceptualized as a 3-step process. First, a candidate object must be selected by attention. Second, attributes of the object must be bound into a recognizable item. Third, the bound item must be linked to a representation in memory. But what happens when attention is redeployed to a new object? Our prior work suggests that links to memory are broken when attention is redeployed allowing only one object to be recognized at a time (JEPHPP,26,2,693). What about binding? Does binding persist after attention is redeployed? Methods: We used a Repeated Search paradigm. Subjects viewed an unchanging display and were repeatedly queried about its contents. In this variant, Ss viewed 2, 4, or 6 realistic objects (e.g., hat, TV, clock etc, — Height range 2.5–5 deg; width range 3.5–6 deg) arranged on an invisible circle of 15 deg radius. On 50% of trials, one object became “unbound” meaning that its features were scrambled for the duration of the trial. Ss responded to the presence or absence of a scrambled object on each trial. We compared Repeated Search, where the same objects remained visible for an entire block of trials, to Unrepeated where new objects were presented on each trial. We hypothesized that, if Ss could see multiple bound objects simultaneously, then scrambling an object would immediately attract attention in the Repeated condition. Results: The scrambling of an item did not immediately attract attention. Search slopes were equally inefficient in both the Repeated and Unrepeated conditions, consistent with a serial examination of items. These results were replicated with various stimuli. Conclusion: Our results showed no indication of persistent binding. It appears that when attention is deployed away from an object the visual representation of the object returns to its preattentive state.