Abstract
If a vertical line is surrounded by a tilted frame, it is typically perceived as being tilted in the opposite direction. This rod-and-frame illusion is thought to be driven by two distinct mechanisms. Large frames cause a distortion of the egocentric reference frame, with perceived vertical biased in the direction of the frame's tilt (i.e., a visuovestibular effect). Small frames are thought to drive the illusion through local contrast effects within early visual processing. Wenderoth and Beh (1977) found that the visuovestibular effect could be induced by a stimulus consisting of only two lines, indicating that an intact frame was not necessary to achieve the illusion. Furthermore, Li and Matin (2005) demonstrated that the Gestalt of an intact frame provided no additional impact to the illusion, as the visuovestibuar effect of an intact frame was less than the sum of its parts. It is unclear whether the same is true for the local contrast effects caused by small frames. Participants performed a perceptual task in which they reported the orientation of a target line (12' in length) presented in the context of either an intact frame (32' on a side, tilted ± 15°) or partial frame (that is, flankers consisting of either the top/bottom of the frame in collinear locations with respect to the target line, or the left/right sides in lateral locations). Significant contrast effects occurred for all stimulus conditions, with the top/bottom flankers causing an effect substantially larger than that of the left/right flankers. Indeed, the effect of the top/bottom flankers even surpassed that of the intact frame, indicating that the overall effect of the frame was a weighted average of the two flanker conditions. These findings suggest an underadditivity of the cellular mechanisms responsible for the contextual effects of lateral and collinear flankers.
Meeting abstract presented at VSS 2016