BEHAVIORAL EVIDENCE FOR COLOR DISCRIMINATION IN CAT

Abstract

Three cats were trained to discriminate "red" from "blue," "red" from "green" and "red" from "yellow" using an operant conditioning procedure. The brightness of each spectral stimulus (wratten filters) was estimated with a Macbeth illuminometer and attenuated by the addition of neutral density filters to produce wide ranges of brightness (about 20 db) with maximal overlap. Six different brightnesses of each wavelength were used in each experimental session and these were arranged in an irregular order. Differential stimulus control over rate of responding was clearly established for each cat for each color discrimination problem. The rate of responding in the presence of the wavelength correlated with reinforcement (SD) was consistently higher than the rate during wave-lengths never correlated with reinforcement (S[DELTA]) in both training sessions and extinction tests. Response distributions obtained during extinction tests were independent of brightness (i.e. flat gradients were obtained) and the total response output was consistently higher to SD than to S [DELTA] across brightnesses. These behavioral data are consistent with the abundant electrophysiological evidence attesting to the spectral sensitivity of the peripheral and central receptors in cat.