Abstract
Crowding is the breakdown in object recognition that occurs in cluttered visual environments and the fundamental limit on peripheral vision, affecting identification within many visual modalities and across large spatial regions. Though frequently characterized as a disruptive process through which object representations are suppressed or lost altogether, we demonstrate that crowding systematically changes the appearance of objects. In particular, target patches of visual noise that are surrounded ("crowded") by oriented Gabor flankers become perceptually oriented, matching the flankers. This was established with a change-detection paradigm: under crowded conditions, target changes from noise to Gabor went unnoticed when the Gabor orientation matched the flankers (and the illusory target percept), despite being easily detected when they differed. Rotation of the flankers (leaving target noise unaltered) also induced illusory target rotations. Blank targets led to similar results, demonstrating that crowding can induce apparent structure where none exists. Finally, adaptation to these stimuli induced a tilt aftereffect at the target location, consistent with signals from the flankers "spreading" across space. These results confirm predictions from change-based models of crowding, such as averaging, and establish crowding as a regularization process that simplifies the peripheral field by promoting consistent appearance among adjacent objects.