[From Rick Marken (2005.11.12.1830)]
Fred Nickols (2005.11.12.1228 EST)--
Rick Marken (2005.11.12.0820)--
So I would say that the task we are (or should) all be faced with is
challenging our _beliefs_ using perceptual evidence. Perception should
be the ultimate arbiter of what we believe because it is the only thing
of which we can be certain: I perceive therefore I believe
Hmm. What about optical illusions? Why should I be so certain of my
Excellent question. I happen to have a book sitting next to me called "Visual Illusions" by Al Seckel (a gift from my father in law, who knows about my interest in perception) and it has all the cool illusions with which I am familiar and then some. I happen to be looking at a picture of the familiar Ames room where two people standing in opposite corners of the room appear to be of drastically different size. There is no question that this is what I perceive. The question is about what I believe to be the explanation for this perception. In particular, am I really seeing a perfectly proportioned little tiny person and another perfectly proportioned great big person. The reason the illusion is compelling is because we assume that it's _not_ a tiny and a large person in the room. We _believe_ that the people are of approximately the same size. And, in fact, our belief is correct. When you see the Ames room in person you can see when two people of approximately equal size get into the room and then see that one looks tiny and the other looks large when they are in the corners of the room.
This is why visual illusions are so cool. There is no question about what we perceive. We call them illusions because we _know_ that what we perceive is inconsistent with what we believe to be true based on other (also perceptual) grounds.
Our initial beliefs about everything are, of course, based on perceptual evidence. So, for example, there are illusions, like Mach bands, where to see things that aren't really there. In the case of Mach bands you see dark bands at the inflection points in a luminance gradient. The initial belief is that what you see is what is there: the bands correspond to something "out there". And there is no reason to suspect that this belief is wrong except that this was shown in a book on illusions where they tell you that this is not what's really out there. If you want to test your belief that the bands are really there you would have to do experiments to test it. And the experiments would have to be done properly.
For example, I know that when Ernst Mach first presented a demonstration of his bands (dark bands appear where there is no corresponding band of low luminance) the initial belief of many observers was that the bands were actually there in the luminance pattern being viewed. These people tested this by taking a photograph of the luminance pattern (created by a spinning disk) that resulted in the bands. The bands showed up in the photograph as well so the conclusion was that the bands were really there. But this was a poor test because the luminance pattern that produced the bands (using a spinning disk) was preserved in the luminance pattern in the photograph. Eventually, photometer measurements were used to prove that the perceived bands were not present in the luminance pattern. This test resulted in a change is beliefs but not in the perception. The perception of Mach bands remains even when other perceptual evidence (photometer readings) rejects the belief that these bands correspond to bands of luminance.
My experience is that perceptual evidence (as in the case of the photometer and Mach bands) can lead to a revision of belief but not of perception.
Richard S. Marken
Home 310 474-0313
Cell 310 729-1400