Colour Perception 1978–1997
In the past twenty years, the spectral sensitivities of the three types of cone have been established with some certainty: direct measurements by microspectrophotometry and electrophysiology are in fair agreement with psychophysical estimates. Particularly significant was the publication of DNA sequences for the four opsins of the human eye, by Jeremy Nathans and colleagues in 1986. This work was soon to transform the understanding of retinitis pigmentosa and other retinal dystrophies, and it has given many insights into the evolution of colour vision; but, curiously, the explanations of dichromacy and anomalous trichromacy have not proved as straightforward as we all expected in 1986. What is clear, however, is that normal colour vision exhibits a genetic polymorphism: much of the intersubject variance in colour matches can be traced to differences in the amino-acid sequence of the opsins for the long-wave and middle-wave cone pigments. The last two decades have seen a major change in the status of opponent processes. In the 1970s it was still common for professors to tell undergraduates that the Young - Helmholtz theory of colour vision held at the receptor level and the Hering theory at the level of the retinal ganglion cells. It is now clear that the chromatically antagonistic processes revealed electrophysiologically and psychophysically in the early visual system do not correspond to the red - green and yellow - blue processes that Hering postulated on the basis of phenomenological observations. The existence of four unique hues is today one of the unexplained mysteries of colour science. In one salient respect, research in colour vision has been changed by instrumental advances. Computer-controlled monitors (though offering splendid pitfalls to the unwary) have allowed the study of spatially and temporally complex chromatic displays, notably in the field of colour constancy. Most recently there has been interest in the chromatic statistics of natural scenes: how well is the visual system matched to the statistics of the world and can it adapt to the gamut of chromaticities present in a given scene?