Every man is mortal.
Socrates is a man.
Therefore, Socrates is mortal.
This syllogism involves the repetition of man, mortal, Socrates, and is. In order for this syllogism to be valid, we must be correct in identifying the first instance of each of these terms with the second instance, and this involves an act of memory. As Santayana explained:
Thus any survey which is analytic, so that it gives foothold for demonstration, or any definition following upon such analysis, presupposes the repetition of the same essences [i.e., concepts, ideas, etc.] in different contexts. This presupposition cannot be justified by the [datum] occupying the mind at any one time. No more can the assurance that a term remains the same in two successive instances and in two different contexts, nor that what is asserted by a predicate is asserted of the very subject which before had been intuited without that predicate. Explication is a process, a deduction is an event ; and although the force of logical analysis or synthesis does not depend on assuming that fact, but rather on ignoring it, this fact may be deduced from faith in the validity of demonstration, which would lapse if this fact were denied. The validity of demonstration is accordingly a matter of faith only, depending on the assumption of matters of fact incapable of demonstration. I must believe that I noted the terms of the argument separately and successively if I am to assert anything in identifying them or pronouncing them equivalent, or if the conclusion in which they appear now is to be relevant in any way to the premises in which they appeared originally. Thus any survey which is analytic, so that it gives foothold for demonstration, or any definition following upon such analysis, presupposes the repetition of the same essences in different contexts. [Scepticism and Animal Faith, 118]
Since Rand’s axioms cannot be identified without assuming memory, it would seem that memory would have to be included at the base of knowledge. So why didn’t Rand include it? Obviously, this is a major oversight on Rand’s part.
If Rand had not failed to include memory as part of her “base of knowledge,” how would she have gone about “validating” it? Most likely, she would have resorted to the same trick as with her other axioms: she would have contended that any attempt to deny memory must first assume it. However, not only is this line of reasoning open to the objections I raised to it in the previous post, it has about it a sophistical flavor which would be far more transparent were it not used to defend views which all sane people accept. People tend to accept any argument, no matter how suspect, when it is used to defend core beliefs. To argue that a principle is “self-evident” and “axiomatic” because it cannot be “denied” without first assuming it involves presuppositions that can easily be doubted and challenged. It hardly forms a criterion of “self-evidence,” and it is mere presumption to contend that it does. For if it did, why didn’t Rand resort to it more often? She could have made use of it, for instance, in “validating” concepts: for isn’t it true that concepts can only be denied by first making use of them? Yet Rand did not make use of this mode of argumentation in validating her concepts, instead opting for an argument that featured even greater opacity. Why was this? Perhaps because even she did not have complete trust in it.
The real error behind all this sterile rationalization upon which Rand attempted to build her metaphysics is the notion that matters of fact can be determined by logical, rhetorical, or moral constructions. Knowledge does not grow from such otiose web spinning. Knowledge arises to fill pressing animal needs, and grows and develops under the stress of the practical demands of everyday life and from scientific experimentation. Knowledge neither requires, nor can be justified, in logic or “self-evidence.” Foundationalism is as unnecessary as it is false and empty. Knowledge can only be justified (provisionally) either in daily practice or (better yet) through rigorous empirical (i.e., scientific) tests.