We May See Color So We Can Understand Each Other

Jun 30, 2017

Humans and other primates see color thanks to three different kinds of cells in the retina.

By responding differently to short-, medium- and long-wavelength light, these cells provide the information the brain needs to figure out color in the environment.

This is how we do it. It's also how the birds and the bees do it.

But it turns out that our eyes do this imperfectly.

Whereas the cells in the eye of the birds and the bees sample light evenly across the spectrum, the receptors in our eyes are tuned less-regularly. In particular, long-wave cells (that is, cells that respond more readily to light from the long-wave, "red" end of the spectrum) are only a very narrow band apart from the medium-wave cells that are tuned to "green" light.

It seems, then, that evolution has given us the short end of the color-vision stick. But, thanks to work over the last decade or so, it is now possible to understand why this, actually, is not really true. Human (and primate) color vision may, in fact, be superbly well-adapted to its primary purpose.

The trick is getting clear about what that purpose is.

For many years — nearly a century, in fact — it was believed that color vision was for spotting ripe fruit. But there may be a better adaptive hypothesis: We see colors so that we can see each other.

Color is full of meaning to old-world primates such as ourselves. Changes in skin color reveal all manner of information about ourselves — are we happy or sad, sexually receptive, ovulating, angry?

And it turns out — as is argued in a paper published this month by an international team of researchers from Japan, the U.S., Canada, and the UK — that the narrow separation of bandwidth sensitivities of long- and medium-wave cones may be the best way for a trichromatic system, such as ours, to discriminate facial hues.

As Chihiro Hiramatsu of Kyushu University in Japan explains in an email, a narrow separation means a large overlap and that means, in turn, "that those cones respond to similar reflected spectra from objects, and could be able to compare subtle difference of the spectra at long wavelength."

What Hiramatsu and his colleagues did was present human perceivers with images of faces designed to mimic the effects of different kinds of color vision. For example, they compared our distinct form of trichromacy with one in which the wavelength sensitivities of the cones was more evenly tuned and also with one in which the distance between the sensitivities was even more narrow. They also looked at color systems in which there are only two effective kinds of receptors (dichromacy).

What they found is striking. We do it best. Our irregular distribution of sensitivities turns out to be just what we need to be sensitive to blushes and blanches.

This doesn't prove that the "social signaling" hypothesis is right, the authors cautiously acknowledge. Maybe trichomacy of our sort "became fixed early on in catarrhine evolution, due to fitness benefits from improved foraging ability, and that subsequently the red-green color channel became co-opted for socio-sexual signaling."

But these findings do demonstrate that our sort of trichromacy "gives individuals a clear benefit in exploiting the variation in skin coloration associated with important aspects of individual condition, such as intra-cycle variation in female fertility."

The authors continue:

"To our knowledge, this is the first empirical support that a necessary condition of social signal hypothesis is met, indicating that the relative spectral positioning of the M and L photoreceptors in catarrhine trichromatic visual system is well suited for detecting facial skin color variation in non-human primates."

So we may have developed color vision to see faces. But this raises another question: Why do faces fluctuate in color in the first place?

Presumably evolution wouldn't have favored creatures who visibly manifest, in the form of colorful change, how they feel, if there were not already others around to take in the spectacle? We blush for others. So you need color vision to have color. And you need color to have color vision?

This is a beautiful puzzle, one the theory of evolution is well able to explain.


Alva Noë is a philosopher at the University of California, Berkeley, where he writes and teaches about perception, consciousness and art. He is the author of several books, including his latest, Strange Tools: Art and Human Nature (Farrar, Straus and Giroux, 2015). You can keep up with more of what Alva is thinking on Facebook and on Twitter: @alvanoe

Copyright 2017 NPR. To see more, visit http://www.npr.org/.