Abstract
Aim: Prolonged exposure to an oriented line oppositely shifts the perceived orientation of a subsequently observed line, a phenomenon known as the tilt aftereffect (or TAE). The TAE transfers across luminance contrast polarity (Magnussen & Kurtenbach, 1979, Perception 8(5), 523-528), suggesting that either the polarities of local detectors are discarded before combining to encode the global orientation of the line, or that polarity is discarded after the global orientation of the line is encoded. Here we test between these two alternatives. Methods: The TAE was measured using line stimuli constructed from strings of ‘dark’ or ‘white’ elongated Gaussian micropatterns. Adaptors were either all white, all dark, one-alternating or two-alternating polarities. Test lines were either all white or one-alternating polarity. Results: (i) TAEs with all-white tests were small with one-alternating polarity adaptors, increased slightly for two-alternating polarity adaptors, and were largest with all-white or all-black adaptors. (ii) TAEs were also large when the test was one-alternating, irrespective of the adaptor type. Conclusion: The fact that alternating polarity adaptors produced only weak TAEs in all-white tests supports the idea that for 1st-order line orientation mechanisms polarity is discarded only after global line orientation is encoded. However the existence of strong TAEs for alternating polarity tests suggests that alternating polarities are combined by 2nd-order line orientation mechanisms. The asymmetry in the magnitude of TAE between alternating-polarity adaptors with white tests, and white adaptors with alternating-polarity tests, can be explained by an imbalance in the population of neurons sensitive to 1st- and 2nd-order stimuli, in that the 2nd-order stimuli are encoded by a subset of the mechanisms sensitive to 1st-order stimuli. The similar size TAE obtained with white adaptor/test and alternating polarity adaptor/test suggests that there are no neurons exclusively sensitive to 2nd-order stimuli.
Meeting abstract presented at VSS 2013