The roles of pattern orthogonality and color contrast in the generation of pattern-contingent color aftereffects.

Abstract

A previous report (Emerson, Humphrey, & Dodwell, 1985) presented evidence that several pattern-contingent color aftereffects could be established simultaneously in individual observers, using as induction patterns the three basic orthogonal pattern pairs of Hoffman’s (1966) Lie transformation group theory of neuropsychology. Contingent aftereffects (CAEs) generated with these patterns show little interference between pattern pairs. The present experiments explored the induction and test conditions that are sufficient and]or necessary for generating single and multiple CAEs. The data indicated that, for CAE generation to occur, it is necessary that two orthogonal inducing patterns not be viewed in light of the same color. It is sufficient to induce with two distinct patterns in different colors, but more CAEs are reported if these two patterns are orthogonal than if they are nonorthogonal. The scheduling of the inducing conditions is relatively unimportant. In testing, the probability of obtaining a CAE report is greatly increased by presenting the subject with a test pattern designed to elicit two CAEs of different hues; orthogonality of test pattern quadrants has little influence. The results support the view that pattern “channels” bounded by orthogonal pairmates exist in the human visual system and that operations in one channel are relatively independent of operations in another.