Vision

Here's a nice little argument that strongly suggests that limitations and biases must exist. If you look inside an eye, you find that on the back surface there are a bunch of little photoreceptors, the rods and cones. Each photoreceptor produces a signal that tells how much light it's exposed to, and that's where we get all our visual information. As a result, our raw visual input is essentially an irregular array of pixels.

However, we don't see it that way. We don't have to think about comparing neighboring pixels and figuring out where the boundaries are, we just have immediate perception of boundaries and objects. Clearly, there is some intermediate processing going on before we as conscious observers receive our visual input. This intermediate processing is not within our conscious control, and is the source (or one source) of the limitations and biases I'm talking about.

At this point, it would be nice if I could point to a specific example and explain both the biological basis and the psychological impact. Unfortunately, although I actually went and did a small amount of reading on the subject, and found some suggestive tidbits about the biological basis of vision, I didn't find exactly what I was looking for. As a result, my argument here, as in many other places, is just some plausible ideas confirmed by introspection.

Babies, I think, would have been the perfect example. The psychological impact is clear: babies are cute, and I like to look at them, more so if they're related to me. It's also clear that paying attention to babies in this way is necessary for the survival of the species, since babies can't fend for themselves. It would be natural, then, for paying attention to babies to have a biological basis—there ought to be a set of neurons, part of the visual system rather than the conscious mind, that computes the cuteness of the raw image. As far as I know, nobody has identified such neurons, but they ought to exist.

By the way, these baby-identification systems of ours aren't infallible. Sometimes they misfire, and label as cute things that aren't actually babies … things that just have, say, big sad eyes, such as kittens and puppies.

Another good example is that the visual system is particularly sensitive to faces. Again, I don't have good low-level evidence for a biological basis, so I have to argue for the fact using plausibility. For one thing, Neal Stephenson mentions it in Snow Crash, so it must be true. For another, I remember I once read about a study that had been done with babies. The babies' eye motions in response to a variety of pictures had been recorded and analyzed, and it turned out that they paid more attention to faces than to other things. The idea was that the babies were young enough that the behavior had to be built in, not learned.

With those examples in mind, I think the general principle behind the limitations and biases will now make sense: the visual system is adapted so as to perceive best those things that are most relevant to us … or, I should say, those things that were most relevant to us an evolutionary time scale ago. I don't know what that time scale is, but apparently it's quite long. Here's a relevant quote from Constraints on Perfection.

Returning to the time-lag effect itself, since modern man has drastically changed the environment of many animals and plants over a time-scale that is negligible by ordinary evolutionary standards, we can expect to see anachronistic adaptations rather often.

Consequently, I suspect that the same general principle explains why nature scenes, although full of detail, are not distracting to us: we are adapted to perceive nature as a background.

As a final example, consider the idea that men's visual systems are designed to take special notice of women. It's common knowledge that boys, as they're growing up, suddenly come to notice girls; this sudden change, after years and years of exposure to the same external influences, implies a real internal change. Of course, it's also well known that hormones are involved, but that's not a complete explanation, any more than it would be to say that boys notice girls because of an increase in blood sugar level. So, my guess is that, among other things, the hormones are responsible for a physical change in the visual system, the creation or activation of a set of neurons that recognize women.

In this context, it's clear that the Islamic custom of concealing women in shapeless robes is not, as I once thought, a pointless relic, but rather a simple, practical application of the principle Out of Sight, Out of Mind. Is it a good idea? I'm not sure. The choice seems to be between having half the population constrained in choice of dress and having the other half always distracted by sex. Speaking in favor of the Western status quo, at least the distraction is pleasant.