Page:Mind (New Series) Volume 8.djvu/173

SOME BEMABKS ON MEMORY AND INFERENCE. 159 A memory, we have seen, is a state of mind which differs from a mere imagination of the past, and in passing from one to the other we are aware that we take a new attitude. But how in the end can we tell that in memory our attitude ia justified, and that our remembrance really is any better than mere fancy ? So far, indeed, as we can apply inference and can rationally construct the past order, we seem to stand on safe ground. But when we are left at last with an idea of the past which shows no visible inconsistency, but about which we are able to find no further evidence, what test can we apply ? The answer must be that we do not possess any valid criterion. There are marks which give us a certain degree of probability, and there are characters which more or less strongly impel us to take the idea as real, but there is in the end no criterion which is not fallible. I will briefly mention the characters which usually distinguish what we call a memory from a mere imagination. The interest of the subject is in the main confined to psychology, we should find some difficulties there into which I shall not enter, and the order of my statement does not pretend to be systematic. We may place first the characters of clearness and strength, and in the next place fulness of detail, a detail which is not visibly rational. Next may come the sense of familiarity, and after that fixity of connexion, and I will then go on to add a few remarks, (i.) I will not venture to ask here what clearness and strength are to mean, but, whatever they mean, a mere imagination may have as much or more of them than a memory, and so much as this seems plain, (ii.) The same may be said with regard to mere fulness of detail, for a simple imagination may be very full in comparison with a memory. The character of the detail is, however, a sign to be noticed. If the particulars are many and yet appear as an accidental conjunction, not depending upon any general idea but all seemingly irrelevant, that, so far as it goes, is a mark of genuine memory. But this mark of irrational detail is, however, no test, (iii.) The sense of familiarity is again deceptive. Its nature has been much discussed, 1 but I think we may represent it as follows. There is in memory an absence of strangeness. The detail comes without shock to a mind which does not expect it and yet is already adjusted to receive it. And this adjustment points to an associative disposition set up by past experience, but it x The word " assimilation " tends to introduce us here, in the pages of Wundt and others, into a world of what I will venture to call the merest mythology.