Abstract
We observed that adaptation to a series of similar Glass patterns causes an unstructured test stimulus to appear to take on illusory structure locally perpendicular to that of the adaptor. We measured the magnitude of this aftereffect using a method analogous to that devised by Blake & Hiris (1993) to assess the strength of the motion aftereffect. After adaptation to a particular global form, subjects reported the perceived structure of Glass patterns varying in their coherence. A test pattern coherence of around 30% was required to null the aftereffect. Three lines of evidence indicate that this effect is mediated predominantly at the level of local oriented filtering of the image. 1. The objectively measured magnitude of the aftereffect appears essentially independent of the global structure of the adaptor, even though humans are reportedly more sensitive to concentric and radial Glass patterns than to translational ones. 2. The aftereffect transfers only partially between the two eyes, implicating monocular V1 neurons. 3. The aftereffect transfers almost completely across contrast polarities, a characteristic shared with the direct tilt aftereffect, a supposedly low-level phenomenon. However, the aftereffect of adaptation to Glass patterns does show weak position invariance such that small “phantom” aftereffects are observed in unadapted regions of the visual field. These phantom aftereffects indicate that adaptation is also occurring at a more global level of form analysis. Thus, it appears that the aftereffect of adaptation to Glass patterns is mediated at multiple levels of the cortical visual processing hierarchy. We argue that this effect is distinct from those reported by McKay (1957ab) and discuss its implications for our understanding of the coding of spatial image structure in the human visual system.
BlakeR.HirisE. (1993) Vision Res., 33, 1589–1592.
McKayD. M. (1957a) Nature 180, 849–851.
McKayD. M. (1957b) Nature 180, 1145–1146.