Descartes famously asked just how much one should trust one’s own senses. Is it not possible that an extremely powerful and malicious demon could trick you into thinking you see things that you do not? (Even more subtly: could a demon trick you into thinking you think things that you do not?)
As it so happens, demons are not necessary, because we are deceived by our senses all the time. Daniel Schacter argued in the Seven Sins of Memory that human memory should not be understood as a function that is meant to keep a perfect record of what has happened, but as a tool that is meant to be used to promote survival. As such, the “seven sins of memory” (common memory “faults” like forgetting things) are often byproducts of useful features or are even useful features in and of themselves. For instance, having a memory junked full of useless details is not efficient; it is better to remember only that which is important. It cuts down on time speant searching for an important memory. Anyone who has dealt with a file cabinet full of documents nobody will ever need should recognize this insight. (Full disclosure: Dr. Schacter is a professor in the Harvard Psychology Department, where I am a graduate student.)
The same argument applies to the senses. The sense did not evolve in order to give us perfectly accurate representations of the world. They evolved in order to help us cope with the world. Often, the two goals are one and the same, but they are not always. Visual illusions are examples where our eyes do not give us faithful representations of the world, usually because of tricks our brains employ to make our vision more useful, not less.
Other aspects of our senses may not be inaccurate per se, but they are not the only solutions to the problems. The typical human eye, for instance, has three different types of cones, which optimally respond to three different wavelenghts of light: 565 nm, 535, nm and 440 nm, which our brain perceives as red, green and blue. It is important to note that “red,” “green” and “blue” are perceptions that exist in our minds. As I understand it, our cones could preferentially respond to 566 nm, 536 nm, and 441 nm without our perception of red, green and blue being altered. We could only have two kinds of cones (as red-green colorblind people do), and thus have a more impoverished perception of color. We could have 4 types of cones, as some people do, and have a richer perception of color.
The point being made here is that our brains do not passively view the world, and our eyes do not just take photographs. What we “see” is a representation of the world that is faithful in some respects and unfaithful in others. Additionally, it is constantly touched-up. When a fashion photographer takes a photograph, she not only manipulates the lighting, the clothing, the angle, etc., all before the image reaches the camera, but after the picture is taken, the photograph is altered. This alteration — removing of blemishes, fading this, enhancing that — is not done randomly; it is done to improve the usefulness of the image — usefulness in terms of selling product or magazines or whatever. The eye and brain are similar.
The same argument The New York Times just ran a fantastic piece touching on just this issue. It is perfectly possible for something to be square in your field of view, but for you to completely fail to see it. I have this happen with my keys all the time. This is at least partly due to the fact that you don’t pay equal attention to everything in your line of sight. Your attention focuses in a particular place (or possibly places — this is actually a very complicated area of research). Magicians use this fact, manipulating your attention so that you “see” what they want you to see. Thus, magicians and perception researchers often have much to say to each other. I had heard of symposiums like the one described in the Times article before. Without spoiling any of the great stories in it, I recommend you read the article before it gets archived and starts costing money.
We recently completed an experiment probing the allocation of attention. This was one of my first Web-based experiments. The results are in, and as soon as I have a chance, I’ll share them here on this site. These will be the first results from our Web-based experiments we’ll be sharing on line.