Levels of Brightness Perception

Most vision scientists are comfortable with the idea that perception is a multi-level process. For example we take it for granted that the two best understood properties of human colour vision, trichromacy and colour-opponency, are underpinned by physiological mechanisms operating at different stages in the visual pathway. The observation that any colour can be matched by a suitable mixture of three primaries-the definition of trichromacy-is understood to be a consequence of having three cones rather than two or four. On the other hand, our inability to perceive reddish greens or bluish yellows, one of the hallmarks of colour-opponency, is understood to be a consequence of the particular way the three cones are combined post-receptorally. In other words trichromacy and colour-opponency have independently measurable behavioural consequences that reflect their different physiological origins. In this chapter I will argue that a multi-level approach is also the right approach for brightness and lightness perception. No single process mechanism can, in my view, account for the many fascinating brightness/lightness phenomena that presently fill the pages of journals and textbooks alike. This admittedly unglam-orous viewpoint is not, as one might expect, shared by all. Notably, Gilchrist and Economou (this volume) argue that all brightness/lightness phenomena can be explained within a single theoretical framework. Their approach, inspired by Gestalt psychology, rejects the very idea of " levels " in perception. Their viewpoint has come to the fore at the same time as a renaissance of interest in contextual effects on surface colour appearance (e.g., see the recent special editions of Perception, 1997, Vol. 26, Nos. 4 and 7). Today's emphasis is on configurational relationships, and these are believed to be the major, if not the sole determinant of the perceived pattern of brightness/lightness variations in the image. I will argue that such contextual effects are best considered within a multi-level framework that includes both low-level contrast and mid-level configura-tional mechanisms. While the famous Gestalt maxim " the whole is greater than the sum of the parts " undoubtedly holds for brightness/lightness perception, it will 20 Frederick A. A. Kingdom be argued that the parts, when considered as levels or stages, nevertheless do exist and can be behaviourally identified. It should be emphasised that his chapter is not a review of the rapidly expanding literature on lightness and brightness perception. Nor does it present a fully-fledged model of lightness/brightness perception. It is essentially a polemic, and only …