Abstract
Many common tasks require us to individuate in parallel more than one object in a complex scene. Although the mechanisms underlying our abilities to count the number of objects, remember the visual properties of objects and to make saccadic eye movements towards objects have been studied separately, each of these tasks require selection of individual objects. To investigate the links between the mechanisms underlying these three abilities, we measured the capacity in number of items for counting, visual working memory and trans-saccadic perception, as well as the interference between tasks. The capacity of trans-saccadic perception—as measured by the transfer of adaptation aftereffects across gaze shifts—was around four items. The addition of a VWM or counting task that involved additional stimuli, however, reduced trans-saccadic capacity to only one item. However, a task requiring estimating the numerosity of a large number of dots (16 to 64) did not interfere with remapping four items. Likewise, maintaining a memory set of two or four items impaired the ability to quickly count the number of items in a brief display. Overall, the pattern of results across the four experiments was not compatible with the predictions of either “slot” (fixed number of objects) or “resource” (finite task resources across tasks) models. Instead, our results suggest that our abilities to count and remember small groups of stimuli are grounded in a sensorimotor “saliency map” of the scene which is also used for integrating information across eye movements.
Ministero dell'Istruzione, dell'Universitá e della Ricerca.