Abstract
The question of what makes a visual object is both intriguing and elusive. Using the contralateral delay activity (CDA), an ERP component sensitive to the number of objects represented in visual working-memory (VWM), we tested the effects of Gestalt grouping cues on the binding of features from different dimensions. Participants performed a change-detection paradigm, in which "objecthood" was manipulated using shared location and common-fate grouping cues across 4 experiments. In Experiment 1, stationary colors and orientations were presented in 3 conditions: 2 items in 2 separate locations, 4 items in 4 separate locations, and 4 items in 2 separate locations (creating 2 color-orientation conjunctions grouped by a shared location cue). The results suggested that sharing a location did not lead to an integration of the color and orientation to one object in VWM as indicated by the CDA. In Experiment 2, the colors and orientations moved for 1000 ms, either together or independently, and then remained stationary for 100 ms. There were 4 conditions: 2 separate items, 4 separate items, 2 "common-fate" items (i.e. 4 items creating 2 color-orientation conjunctions that moved together), and 4 separate items meeting to create 2 color-orientation conjunctions. Interestingly, even a common-fate grouping cue did not result in the binding of color and orientation. To further investigate the failure in binding features from different dimensions, we replicated these two experiments, using familiar shapes instead of orientations. In Experiment 3, colors and shapes were not integrated in VWM when they shared the same location. However, in Experiment 4, colors and shapes moving together were represented as integrated objects in VWM. Our results suggest that a potent Gestalt grouping cue of common-fate does not produce reflexive binding. Instead, binding also depends on the type of objects and presumably their familiarity.
Meeting abstract presented at VSS 2014