Abstract
Crowding (the deleterious effect of flanks on target identification) is ubiquitous in spatial vision. The purpose of the present experiment was to test the hypothesis that crowding is simply contrast masking by remote flanks. To test this hypothesis we measured and compared crowding in a direction-identification experiment with contrast masking in a detection experiment. In the crowding experiment we measured contrast thresholds for identifying the direction of E-like and C-like patterns (i.e., up vs. down, or left vs. right), using a 2-alternative method of constant stimuli. The patterns were comprised of circular Gabor patches, and the surrounding flanks also consisted of Gabor patches. In the contrast masking experiment we measured contrast thresholds for detecting a single Gabor patch in the presence of surrounding flanks consisting of Gabor patches (similar to Polat & Sagi, 1993). In both experiments, we varied the distance of the flanks and their carrier orientation. In both experiments, we found that in the fovea, high contrast flanks resulted in threshold elevation when the flank-to-target distance was less than about four times the Gabor patch standard deviation. In the fovea, in both experiments threshold elevation was strong when the target and flanks had similar orientations, and was weak or absent when the target and flank orientations were orthogonal. Over a wide range of stimulus conditions we found that foveal threshold elevation in the crowding experiment was quantitatively similar to the threshold elevation in the contrast masking experiment, as predicted by a test-pedestal model. We conclude that foveal “crowding” can be explained by contrast masking. In contrast, in peripheral vision, crowding is much stronger and extends over a much larger spatial extent than predicted by contrast masking.