Combined processing of what and where information within the visuospatial scratchpad

Abstract

Considerable evidence has accumulated to suggest that within visuospatial working memory (WM) there are two functionally independent subsystems for the object-based and the spatial dimension of information. Less is known, however, how these subsystems cooperate in conditions with combined maintenance requirements. This question was addressed in the present study. In Experiment 1, subjects were presented with two one-dimensional (object-based or spatial) WM tasks and a two-dimensional task in which both dimensions were to be remembered. Subjects had to maintain polygon shapes, or their location, or both their shapes and locations. The effect of two-dimensional storage across the two subsystems was asymmetric: Whereas performance for spatial information declined, performance for object-based information was quite unaffected. In Experiment 2 the amount of information to be remembered from the same dimension was varied, again in a spatial and in an object-based WM task. The results suggest that whereas recognition performance for one single item was comparable across both WM subsystems, the shapes of the polygons may represent more complex stimuli than their locations and thus require more dimension-specific storage capacity. The dimension-specific costs on spatial WM in two-dimensional storage conditions may be explained by a competition for general resources resulting from a strategic focus on the more complex object dimension.