The Discovery of Spectral Opponency in Visual Systems and its Impact on Understanding the Neurobiology of Color Vision

Abstract

The two principal theories of color vision that emerged in the nineteenth century offered alternative ideas about the nature of the biological mechanisms that underlie the percepts of color. One, the Young-Helmholtz theory, proposed that the visual system contained three component mechanisms whose individual activations were linked to the perception of three principal hues; the other, the Hering theory, assumed there were three underlying mechanisms, each comprising a linked opponency that supported contrasting and mutually exclusive color percepts. These competing conceptions remained effectively untested until the middle of the twentieth century when single-unit electrophysiology emerged as a tool allowing a direct examination of links between spectral stimulation of the eye and responses of individual cells in visual systems. This approach revealed that the visual systems of animals known to have color vision contain cells that respond in a spectrally-opponent manner, firing to some wavelengths of stimulation and inhibiting to others. The discovery of spectral opponency, and the research it stimulated, changed irrevocably our understanding of the biology of color vision.

Keywords:

• color vision
• color theory
• single-unit electrophysiology
• spectral opponency
• eye
• lateral geniculate nucleus
• Hering
• Young-Helmholtz

Notes

¹ Starting in January, 1960, the author was a student in the laboratory of R. L. De Valois. At least from that time forward, it had become abundantly clear that all of the LGN cells previously identified as having narrow-band spectral responsivity were actually of the spectrally-opponent variety.

² To be fair to these researchers, at the time of the discovery of spectrally-opponent neurons in the primate visual system, the spectral absorption properties of the primate cone photopigments still had not been accurately determined; indeed, there were continuing arguments as to whether there were three types of cones (e.g., Willmer, 1961), and it was still some years before the cone photoreceptors would be routinely identified as S, M, and L. It is also noteworthy that the tendency to use human hue designations to label biological structures or processes misleadingly has extended well beyond the naming of spectrally-opponent neurons. For example, the three types of cones of the trichromatic retina often have been identified as “blue, green, and red,” a practice still followed in some circles, despite the fact that it has long been appreciated that the locations of the peak sensitivity of these cones do not correspond to the spectral locations associated with these percepts and that, in any case, each class of photoreceptor responds univariantly and thus each must be “color blind.” And there are even more unforgiveable cases of this kind of label misattribution; for example, the genes that specify the opsin proteins that comprise the three classes of cone photopigments have at times also been labeled using human hue names (e.g., “red genes”).