Page:Bergson - Matter and Memory (1911).djvu/104

 of memories. But if it could be positively established that the cerebral process answers only to a very small part of memory, that it is rather the effect than the cause, that matter is here as elsewhere the vehicle of an action and not the substratum of a knowledge, then the thesis which we are maintaining would be demonstrated by the very example which is commonly supposed to be most unfavourable to it, and the necessity might arise of erecting spirit into an independent reality. In this way also, perhaps, some light would be thrown on the nature of what is called spirit, and on the possibility of the interaction of spirit and matter. For a demonstration of this kind could not be purely negative. Having shown what memory is not, we should have to try to discover what it is. Having attributed to the body the sole function of preparing actions, we are bound to enquire why memory appears to be one with this body, how bodily lesions influence it, and in what sense it may be said to mould itself upon the state of the brain matter. It is, moreover, impossible that this enquiry should fail to give us some information as to the psychological mechanism of memory, and the various mental operations connected therewith. And, inversely, if the problems of pure psychology seem to acquire some light from our hypothesis, this hypothesis itself will thereby gain in certainty and weight.

But we must present this same idea in yet a