Page:Mind (New Series) Volume 8.djvu/165

 SOME REMABKS ON MEMORY AND INFERENCE. 151 separated from and is ideally connected with my present, and this ideal separation and connexion is and must be performed always in the way which I have described. 1 In short, memory as immediate is to my mind a sheer miracle, and I cannot accept a miracle even where I am assured that it is due merely to the brain. The past, we have so far seen, is perceived by means of serial association, and, before I proceed, it is necessary to warn the reader here against a dangerous misconception. We have in the series a-b-c the association of b with a and of c with 6 ; but we have not merely these separate associa- tions, and, if we had no more than this, we should have no series at all. For every series which we know is known by us as one, and, if it had no real unity, the appearance of its oneness would be inexplicable. But this unity involves, so far as I can see, and consists in an ideal identity of character. There is some one content that is present through and is de- veloped by the series, and is qualified by and itself essentially qualifies this series. But, if so, the members of the series will be joined not merely by association with one another, for each one must be associated also with one and the same quality. There will hence in fact be no merely successive association any more than there is any merely successive perception. The division of association into that which is simultaneous and that which is merely successive is in principle vicious, and any enquiry based on it is foredoomed to failure. The succession should be represented not as Xa-6 X but rather as /. And so we perceive how the whole a b series may thus be thought of as one, and how the idea of the whole is united with and so may reproduce any of the members, singly or at irregular intervals, and again in either direction. For beside the mere association of member with member we have as its complement in every series the con- 1 If we wish to avoid mistake here, we must beware of confusion. We must distinguish the exciting cause of a reproduction from the ground of a memory. The ground of a particular memory is that which places it in connexion with a certain member of my past series. But it may be partially excited by that which cannot complete and so date it. A scent may, for instance, remind me of a certain flower, which then by associa- tion calls up its adjuncts involving a dated event in my life. The dating associations here are not those which excite, and the latter may be very fruil and slight indeed. The reproduced when excited then dates itself by association with what is constructed from my present. If on the contrary I go backwards or forwards retracing my life, the exciting cause of a memory and its ground may be the same.