The status of automatized knowledge in his mind is experienced by man as if it had a direct, effortless, self-evident quality (and certainty) of perceptual awareness. But it is conceptual knowledge --- and its validity depends on the precision of his concepts, which require as strict a precision of meaning (i.e., as strict a knowledge of what specific referents they subsume) as the definitions of mathematical terms. (It is obvious what disasters will follow if one automatizes errors, contradictions and undefined approximations.)
This paragraph provides the very core of what is wrong with Rand's epistemology; and it will require a number of posts to identify and elucidate all the various fallacious branches that sprout from this one trunk. Right now I wish to concentrate on the final sentence, the one embalmed in parenthesis: It is obvious what disasters will follow if one automatizes errors, contradictions and undefined approximations. Is it really obvious? Rand's assertion assumes the following: (1) concepts can be erroneous; (2) concepts can be contradictory; and (3) improper concept formation leads to concepts that are ill defined and merely approximate. I will contend that these three presumptions are wrong.