Why do we perceive colour? (i.e., distinct values?)

Question

This is a question I often asked myself, but never really found a satisfactory answer to. It may be that I always went about it from a wrong starting point because the question is induced by physics. But I think an evolutionary perspective may be the most satisfactory one to explain this.

Humans and other animals had much to gain from differentiating not only dark and light but also colors to identify edible things, or threats etc. For this purpose we developed multiple types of cone cells. But what reason is there to not perceive light as we perceive sound, in a spectrum? We could easily distinguish said things by just knowing how high or low the light frequency is (or for that matter, what the spectral 'fingerprint' of specific objects is). Why differentiate between an - albeit big - array of colors?

Or, to go one step further, why did we need an extra type of cone in the middle, the yellow-green wavelength? Is there an evolutionary reason? I'd settle for using the three more or less distinct cone sensitivities as explanation that the brain just has enough information to arbitrarily differentiate these into very discrete cognitive signals.

But evolutionarily, there has to be a purpose, so that one of the questions is the right one. The colors must bring something to the table. Or not?

Thanks in advance. I am also grateful for anyone nitpicking my arguments, because it always helps.

Answer 1

"But what reason is there to not perceive light as we perceive sound, in a spectrum?"

We do perceive light in a spectrum. Not just in a spectrum, but in a combination of spectral components (because we rarely in nature observe true single-wavelength light). If you are thinking about names of colors, those are just words we made up to describe colors to other people, and the range and spacing of those colors differs across language groups. We do the same for sound in the context of music: B-flat versus an E, for example.

Or, to go one step further, why did we need an extra type of cone in the middle, the yellow-green wavelength?

Colors are perceived in vision by the ratio of activity of different cones, and intensity and spectrum can be confused. If you have no "green" cones, there are big stretches of wavelengths that only activate one of the other two cones. In that situation, there is no way to discriminate between intensity and color.

If, on the other hand, you are wondering about transduction, it's because sound and light have different physical properties. The cochlea effectively "maps" sounds onto a physical arrangements of hair cells. You could do something similar with a prism, but you would lose spatial acuity (effectively, you would need a whole spectrum analyzer for the receptive field of each photoreceptor: your retina would have to be absolutely massive). Instead, color vision works by differential chemical sensitivity to different light wavelengths.

Your question doesn't seem to be much about evolution, besides the title, so I left off that part.