In visual search tasks using pop-out stimuli, memory of the target features and positions in the past trials affects the present search performance. This is called “priming of pop-out” or “repetition effect in visual search” in which facilitation for the repeated and inhibition for the changed lasts over the following three to eleven trials. It has been suggested that the number of trials the effects persist reflects the capacities of a memory system mediating the effect (Yoshida, et al., 2000, Jpn. J. Psychon. Sci.: Maljkovic & Nakayama, 2000, Vis. Cogn.). To compare the capacities and the stored representations between visuospatial working memory and the memory system, we used a modified visuospatial n-back task (Carlson, et al., 1998, Cereb. Cortex) concurrently with the visual search task. Observers’ task was to quickly respond to the shape of an odd-colored target in a display, and to judge whether the instructed one characteristic (color or position) of the current target was the same as the target that had been presented n trials before. Observers successfully performed color n-back task or position n-back task with the visual search task at most to the 2 or 3 trials back. This number implies capacity of the visuospatial working memory to the target characteristic. During each types of the n-back task, the reaction time in the visual search task showed robust priming effects lasting three to twelve trials with respect to “both” target color and position, showing there were three to twelve memory traces to the past target events and the memory contained several target characteristics independently of the n-back task’s demand. So representations which one can consciously store and update in visuospatial working memory are limited to past a few targets and to the task-relevant aspect of them, even though reaction time indicates our visual system represents more objects and their multiple characteristics.

Supported by JSPS to TY