SOME BEMARKS ON MEMORY AND INFERENCE. 161 with itself. He admits the extreme fallibility of memory in detail, but contends that at least it cannot be wrong in its assertion of my past existence. But how far, and in what sense, when bared of or transformed in detail, my past existence remains mine, is a matter not discussed, nor, apart from this, is there any evidence produced for the truth of the contention. If wherever else a witness can be tested he is shown to be fallible, you can hardly assume him to be infallible in or beyond a certain point, simply be- cause in or beyond that point you have in fact always found him to be right. And with regard to memory of my past existence the case stands as follows. All the memories that we can examine belong to minds which have had some previous existence, and it is very probable that memory can exist only as the result of some foregoing psychical develop- ment, however short. And, if this is so, then memory will be for this extraneous reason, and will be so far, infallible. It will be infallible, we may say, accidentally and in fact, but not in principle. Its evidence will depend on and be restricted to that which is otherwise known. And such an infallibility is, I presume, for Prof. Ladd's purpose useless. And even so much as this can, perhaps, not be demonstrated. For that memory should supervene suddenly at a certain point of physiological development in such a way that its report of a past psychical self would be wholly mistaken, seems not clearly and in principle to be impossible. If so, even the limited infallibility of memory seems not proved, but in any case, even if proved, I have shown its dependent nature. 1 From this obscure and unsafe position Prof. Ladd passes to a second, which, itself untenable, seems not even consistent with the first. All reasoning, he argues, goes from premises to a conclusion, and our knowledge of the conclusion depends upon our memory of the premises. Hence, if that is fallible, every possible act of reasoning is discredited. Far then from being able to show that memory is fallible, we have even to assume the opposite if we intend to have any conclusion whatever. And with this we have a sure and certain remedy, Prof. Ladd argues, against the disease of scepticism. But the ground of the argument seems to me incorrect, and the 1 If a man mistakenly remembers events ten years before he was born, is it satisfactory to add : There you see at once that his memory is really infallible, for he had, as a fact, some actual past (as you saw) before he made that mistake about his past '? And even this amount of de facto infsdlibility rests on the assumption I have noticed in the text. It is therefore so far precarious, as well as in any case derivative. 11
