4.3: There’s More to Seeing than Meets the Eye

When we talk about the input to the visual system, we naturally think about the eyes. Light passes in through the lens of each eye and is projected onto rods and cones at the back of the retina. But our perception is much different and much richer than this input. The lens reverses the image from left to right and turns it upside down. The image on the back of the retina is not homogeneous, but punctate (because the rods and the cones are separate, discrete units). Furthermore, the eye is constantly moving around, and we blink frequently.

In short, the image on the back of our retina is two-dimensional, upside down, punctate, and it jumps all over the place. But we see a threedimensional world, right side up, full of familiar objects that don’t constantly leap about.

A related point can be made about perceptual constancies, such as size, shape and distance constancies. Hold your hands up in front of your face, with the left hand close to your eyes and your right hand as far away as possible. Do your hands look the same size? Walk over to the wastebasket, then back away while keeping it in view. Does it seem to become smaller as you back away from it? Circle around and view it from different perspectives. Does it seem to change shape as you move? The fact that the wastebasket seems the same size as you back away from it (so that the smaller retinal image of it is smaller) is known as size constancy. The fact that it doesn’t seem to change shape as you circle around it (so that the retinal images of it do change shape) is known as shape constancy.

Provided you don’t get too far away, the size and shape of the wastebasket seem to remain constant; your perception of it is the same. Yet its image on your retina becomes smaller as you back away from it, and the image changes shape as you change your orientation with respect to the wastebasket. Here the input is different, but the appearance of the object remains the same. So, something is going on that involves more than just the images on your retinas. There is debate over how to explain size constancy, but there is some evidence that in part it is caused by the perceptual system’s enrichment of the sensory input.

Information Processing

In the cases considered thus far, the perceptual system seems to enrich the visual input. It does not just passively register it, it does something to it. It is often useful to think about such processes in terms of information processing. Suppose that you program your computer to accept certain inputs (say a series of number between 0 and 100), and then have it manipulate or process them so that it generates a certain output (say the average values of the numbers in the input).

Many cognitive processes also involve information processing. In the case of visual perception, the input consists of the stimulation of rods and cones at the back of the retina, the relative orientation of our eyes to one another, and perhaps various factors involving the orientation of our bodies. And the output is the perceptual experience we have when we see something.

In the cases considered so far, the perceptual system enriches the visual input in ways that do not involve our beliefs or desires. This is called bottom-up processing. The idea here is that our nervous system records sensory stimulations and passes the information on up to the brain. But we see with our minds as much as with our eyes; we also engage in top-down processing that involves something very much like inference. So, before turning to it, it will be useful to recall a few facts about inference.