Abstract
Binocular vision requires integrating neural signals from both eyes. For over 40 years (Vidyasagar, Nature 1976; Blake et al., J. Exp Psychol. HPP 1981; Wolfe & Held, Vision Res. 1981), and therefore for much of Donald MacLeod’s career, a sharp distinction has been drawn between binocular responses that are ‘purely binocular’, meaning they require simultaneous corresponding excitation from both eyes (sometimes called binocular logical AND responses), and ‘simple binocular’ responses, meaning they can be driven by the left eye alone or right eye alone or both eyes (binocular logical inclusive OR). Evidence for ‘purely binocular’ (AND) mechanisms is a typical research aim because it implies a ‘central’ level of neural response requiring signals from both eyes.
Surprisingly, grouping of distinct objects in view is accounted for by the oft belittled simple binocular (OR) response. Binocular OR responses account for grouping together distinct objects – so that all objects appear the same color – when each object is presented in chromatic binocular rivalry. A binocular OR response, but not a binocular AND response, is revealed when the rivalrous chromaticities are swapped between the two eyes ~5–10 times a second with, importantly, the various objects in view swapped between their chromaticities at different temporal phases or unequal temporal frequencies. Grouping would follow from a binocular OR response because it is not selective for the eye of origin of the different chromaticities presented to each eye. Experiments verify the binocular OR predictions for grouping, including failure of grouping after adding objects without chromatic rivalry.