THE INTERACTION OF OBJECTIVE AND SUBJECTIVE ORGANIZATIONS IN A LOCALIZATION SEARCH TASK

We investigated how both objective and subjective organizations affect perceptual organization and how this perceptual organization, in turn , influences observers' performance in a localization search task. Two groups of observers viewing exactly the same stimuli (objective organ ization) performed in significantly different ways, depending on how t hey were induced to parse the display (subjective organization). In Ex periments 1 and 2, the observers were asked to describe the location o f a tilted target among a varying number of vertical or horizontal dis tracters. Subjective organization was induced by instructing observers to parse the display into either three horizontal regions (rows) or t hree vertical regions (columns). The position of the target was critic al: location performance, as assessed by reaction time and errors, was consistently impaired at the locations adjacent to the boundaries def ining the regions, producing what we refer to as the subjective bounda ry effect. Furthermore, the extent of this effect depended on whether the stimulus-driven and conceptually driven information concurred or c onflicted. This made location information more or less accessible. In Experiment 1, the strength of objective grouping was a function of the proximity of the items (near or far conditions) and their orientation in a 6 x 6 matrix. In Experiment 2, the strength of objective groupin g was a function of similarity of color (items were color coded by row s or by columns) and the orientation of the items in a 9 x 9 matrix. T he subjective boundary effect was more pronounced when the display pro moted grouping in the direction orthogonal to that of the task (e.g., when observers parsed by rows but vertical distractors were closer tog ether [Experiment 1] or color coded [Experiment 2] to induce global co lumns). In contrast, this effect decreased when the direction of both objective and subjective organizations was parallel (e.g., when observ ers parsed by rows and horizontal distracters were closer together [Ex periment 1] or were color coded [Experiment 2] to induce global rows). A localization search task proved to be an ideal forum in which objec tive and subjective organizations interacted. We discuss how these res ults indicated that observers' performance in a localization task was determined by the interaction of objective and subjective organization s, and that the resulting perceptual organization constrained coarse l ocation information.