This brief post by Daniel Kernell, author of Colours and Colour Vision: An Introductory Survey, at the Cambridge University Press blog Fifteen Eighty Four gives a nice introduction to color vision. As Kernell himself is “color blind” he has a special relationship to his subject. He explains, in so far as this is possible, the differences in color perception between a trichromat and a dichromat. (Link via Publishing Cambridge.)

The mechanics of color vision are pretty well understood: rods and cones, and then neurons sorting it all out in the brain. The hard problem is working out how the neurons arrange for us to “see” the things in our brain. The old idea of perception as some kind of tiny homunculus sitting inside our heads watching a sort of unfolding movie, has obviously got to be nonsense. But coming up with a reasonable alternative is tough, leading philosophers of mind into equally crazy convolutions.

Cutting the Gordian knot, Kevin O’Regan* theorizes that color vision, all vision, results from our interacting with the scene being viewed. We feel the scene’s reality by sampling it: we scrutinize this bit and then that bit. O’Regan compares the process to the way in which your hand recognizes a penknife while it remains invisible in your pocket. It’s a bit like the kid’s game of touch bag where you get to identify objects by feeling them inside a bag. By feeling various bits of an object you are able to reconstruct the idea of the whole. But if the object is just laid on your hand you are unable to tell what it is. Similarly, when we see, we don’t have any sort of photographic representation inside our brain. In fact, the overall clarity of photos misrepresents what we actually see: our eyes are in fact relatively course-grained for most of their coverage. Our belief that what we see is really what we see as a sharp image results not from our having a clear image of the whole, but from our confidence that by moving the focus of  our eyes we can establish the details of any part of the scene. In other words by knowing we can sharpen up any bit of the scene whenever we want, we make the assumption (and adjustment) that every bit of the view is crystal clear, even though brief introspection of the visual evidence of what you are looking at will easily demonstrate that what we are seeing is actually a little central area of clarity surrounded by a large extent of vagueness. O’Regan’s theory takes the task of perception out of the brain and turns it into an interaction between the brain and the object being viewed.

O’Regan’s sensorimotor approach seems to me to get past many of the barriers which have prevented us from coming up with a good theory of vision/perception/consciouness.

* J. Kevin O’Regan: Why Red Doesn’t Sound Like a Bell: Understanding the Feel of Consciousness, 2011