Partial independence of brightness induction and brown inducton suggests a two-stage model for brightness induction
Author
Vincent, Joris
Metadata
Show full item recordAbstract
Bright contexts surrounding achromatic stimuli generally induce perceptual darkness in those targets. In yellowish targets, such contexts also induce a qualitative brownness. This raises the question whether brownness induction is mediated by the same mechanisms as darkness and brightness induction in achromatic targets - whether brown is just dark yellow. While darkness induction by simple, uniform bright surrounds is usually explained by contrast mechanisms such a lateral inhibition, several brightness/darkness induction effects cannot solely be explained by these mechanisms. White's (1979) illusion consists of a high-contrast square-wave grating in which one target replace part of the white bars and another part target replaces of the black bars. The target replacing the white bars shares a larger border with the flanking black bars, but is usually perceived as brighter than the other target - contrary to the predictions of contrast mechanisms. Instead, this induction has been explained separately by low-level (e.g., multiscale spatial filtering accounts), mid-level (e.g., based on T-junctions, or Gestalt grouping), and higher-level processing (e.g., scission of the scene into different depth planes). This complex illusion is well-suited for studying the association between brightness induction and brownness induction. Participants judged both the brightness and brownness induction in White's classic illusion, a checkerboard variant, and a radial variant. Across participants, both brightness and brown induction were in the direction of assimilation more often than in the direction of contrast, although the frequencies differed among the different stimuli. On the classic illusion, brownness induction and brightness induction were independent in their direction of effect. Achromatic brightness induction was almost exclusively observed in the assimilation direction, while brownness induction in either direction was observed across participants. This would suggest that brownness induction and brightness induction are mediated by independent mechanisms, producing uncorrelated effects. On the radial and checkerboard variants, participants showed congruent direction of brightness and brown induction more often than expected by chance. In contrast to the classic illusion, these results would suggest that brownness induction and brightness induction are mediated by shared mechanisms, producing the correlated effects. To resolve these conflicting results, a framework is suggested with at least two stages (lower- and higher-level) of processing underlying induction. A shared initial stage of visual processing is involved in both brightness and brownness induction, but the second stage consists of independent mechanisms for brightness and brownness. Different aspects of a stimulus display drive the stages differentially. The radial and checkerboard variants are suggested to drive the initial stage more strongly than the second stage, explaining the lack of independent brightness and brownness induction effects for these stimuli. The classic illusion is suggested to drive the second stage more strongly (than the radial and checkerboard variants), producing independent induction effects. Specifically, the second stage mechanisms would bias the brightness induction, but not the brownness induction, towards the assimilation direction. Several previous explanations of White’s illusion are discussed in relation to these two stages, providing a framework for integrating theories of brightness processing at different levels. The two-stage framework also generates several hypotheses about how well the brightness or brownness induction in a given stimulus predicts performance on another stimulus, or the other induction.
Collections
- Psychology [367]