Regi, Ed, Frank: I have not had time to follow all aspects of this thread, but this discussion of perception caught my interest. I think that when you, Regi, say:
Unfortunately, the Objectivist description of “how it works,” is incorrect, and introduces an unnecessary problem. Just how does anything know what data belongs to what percepts and how it ought to be integrated before it knows what sensory data comes from or pertains to which entities? Evolution does not explain that because evolution would have to prepare the mystical faculty of integration with knowledge of every possible entity the organism was ever going to encounter.
you are turning an error of Peikoff’s into one of Rand’s. His mistake is obvious when, on page 53, he says “The result is your ability, when you look out, to see not merely a patch of brown, but a table.” I suspect that Peikoff momentarily lost his focus, and started speaking of perception as though it contained conceptual knowledge, when all he really needed to explain was how we perceive entities—not how we perceive entities of certain conceptual types. Unless he himself has misunderstood Rand.
When Ayn Rand said that sensations are integrated into perceptions, she meant the perception of entities as such: that is, physical wholes from which effects and actions proceed. Not tables, balls, people—these notions come later—but the subverbal awareness of “brown thing there,” “red thing I feel,” “big thing up there”—in a word, objects. Perception transforms the gestalts of sensation into things that can act and have significance for the higher life forms. As Rand said, it is largely automatic; and this is not surprising, because the fact that actions and effects are close in space and time to their source is metaphysical in nature and presumably would get wired in by evolution.
It is well to keep in mind that we are not only talking about the sense of sight but also those of touch, hearing, smell, and taste. And that the sensations being integrated, even if we assume that only one sense is operating, are extremely numerous and complex. Moreover, the information has to be highly dynamic: a paralyzed baby born with only one eye and whose crib was surrounded by objects that never moved or changed might be unable to learn to perceive the world in terms of entities. In other words, to focus on the sense of sight for a moment, in order to see an entity it is not enough for the eye to be hit with its colors and other data; the visual information must be gathered from different angles and distances and lighting conditions, the observer and the thing must change their locations with respect to each other—then only will there be enough data for the inherent logic that is wired in by evolution to scream “entity!”
Finally, I would like to observe that Rand did not specify how fast perceptual integration takes place. She merely was concerned to point out that perceptions were ultimately based upon the information provided by the senses, and upon nothing else. It is my belief, and I think it was hers, that there is a definite stage for humans during which the brain must process the raw sensory data into usable wholes—that is, entities as such, not as things with names such as table etc. I think she would have agreed that this stage might possibly be almost instantaneous in some animals, just as some mammals are able to stand up at birth. So that a kitten might attack its mother’s tail right away. The information it needs may be less for its sense of entity (which may be less sophisticated than man’s) than what man needs.
(Edited by Rodney Rawlings on 5/17, 1:16pm)