Abstract

The earliest thinking about lightness assumed that perceptual experience correlates with local stimulation. Wallach’s celebrated disk/annulus experiments showed that lightness depends on relative luminance, not absolute luminance, but his ratio theory recognized perceptual structure in only a minimal way. Work following the cognitive revolution showed that lightness depends on depth perception, the classification of illumination edges and reflectance edges, and the integration of multiple edge ratios to compute remote ratios. The role of perceptual structure has been featured in two kinds of theories – decomposition models and framework models. Decomposition models, invoking the logic of inverse optics, seek to decompose the retinal image into those separate factors, like reflectance and illumination, from which the image is initially synthesized. Framework models seek to organize the retinal image into frameworks, or perceptual groups, that represent regions of common illumination. This kind of grouping is different from, indeed orthogonal to, the more familiar kind of grouping for object segmentation. Framework theorists claim it can explain lightness illusions, as well as lightness constancy. Virtually all of these advances, made since Wallach’s paper, were anticipated by the gestalt theorists.