Basic gestalt laws cannot explain uncrowding

Visual crowding is the inability to perceive elements within clutter. Traditional crowding models, such as pooling, predict that performance deteriorates when flankers are added. However, this prediction has been disproved. For example, performance was found to deteriorate when a Vernier was surrounded by a single square but also to improve when more squares were added. This phenomenon is termed “uncrowding.” Previous studies showed that it is not the number of flankers that matters for uncrowding but the configuration. To understand how a configuration leads to crowding or uncrowding, we presented a Vernier surrounded by a square in the center of the screen. To that we added squares and stars that constructed different configurations according to the Gestalt laws of symmetry, closure, and good continuation. We did not find any evidence that the Gestalt laws we tested play an important role in crowding. To test for low-level factors, we also used a pixel-wise clustering method (k-means algorithm). However, we could not find evidence for the involvement of low-level factors either. We conclude that neither Gestalt laws nor basic processing can explain crowding and uncrowding. Likely, more complex aspects of display matter.