The evolution of vision

Recently, Scientific American came out with an article on the history of the evolution of the eye.  Intelligent Design proponents have used that as the basis for renewing old creationist arguments against the idea that the eye evolved.  See here and here.

The Scientific American article is about the evolution of the physical eye, rather than about the evolution of the visual system.  The ID response, too, is about the physical eye.  But it is the usual argument about irreducible complexity.  According to that argument, a change in the physical eye would have to have a corresponding change in the entire visual processing system, before there could be any benefit.  And this is part of why ID proponents and creationists see it as irreducibly complex.

The Scientific American article begins:

The human eye is an exquisitely complicated organ. It acts like a camera to collect and focus light and convert it into an electrical signal that the brain translates into images. But instead of photographic film, it has a highly specialized retina that detects light and processes the signals using dozens of different kinds of neurons.

That seems to be assuming Marr’s theory of vision.  And, indeed, it is a little difficult to see how vision could evolve based on Marr’s theory.  However, I think it likely that J.J. Gibson’s ecological theory of visual perception better describes how vision works.  And a rather simpler evolutionary path should suffice if vision is as described by Gibson.

To illustrate this, consider how the bar code scanner works at your local grocery store.  It typically uses a laser to emit light, and a single photo cell to detect the reflected light.  This suggests that it is should be possible to “see” something as complex as a bar code, with only a single retinal receptor cell.  The eye itself would have to move back and forth, so that light from multiple regions of the environment can get to that retinal cell during part of the eye’s motion.  And, indeed, the eye does move back and forth in saccades.  Adding some sort of focusing ability to the eye allows an increased precision of the resulting perceptual ability to discriminate between different parts of the environment.  And adding additional retinal receptors and visual neurons allows more of the visual work to be done in parallel, thus speeding up the rate at which information about the environment can be extracted.  These changes in focusing ability, and in the number of retinal receptors, are independent of one another, so that they do not need to all evolve at once.

In my view of vision, there is no need for an internal image to ever be formed.  Rather, the visual scanning is used to effectively partition the environment into smaller regions, as suggested in a recent post.  Our understanding and visual perception of our environment then derives from that hierarchical partitioning, rather than from a computational analysis of an internal image.