Page:Mind (New Series) Volume 8.djvu/172

 158 F. H. BBADLEY: In the proper sense of inference then memory involves an inference but itself really is not one. 1 If, however, the term were used in a looser way, the answer might be different, and the whole sequence might perhaps be called an inference. It would be here as in a case which involves observation. I may see a man and recognise him as a certain person by a genuine inference, and I then may perceive him to act in a certain manner. I may, on this, attribute the perceived act to the inferred person, and this whole process might be termed an inference. And in the same way memory also might be called an inference, for the reason beside that it does not involve perception. I do not think, however, that we need here consider this looser use. Nor will I stop here to discuss a possible attempt to confound inference with memory on the ground that all inference in the end is irrational habit. For the secondary distinction between inference and memory proper would still remain, even if both were in the end mere results of memory in the sense of habit. I could not in this paper attempt to deal with such a fundamental question, 2 and must pass on to another branch of our enquiry. 1 The above and what follows may, I hope, justify the doctrine I have stated elsewhere, that memory in its essence involves an inference and so is inferential. I have never said or meant that memory consists in mere inference, and that you could make the goodness of the inference a test of memory. The question as to how memory, involving an inference, differs from inference proper, was not discussed or raised by me at all. The statement in my Logic, p. 75, as to the want of a point of identity in mere imagination, is certainly, as it stands, obscure and perhaps mis- leading. Whether my mind was clear when I wrote it I cannot now tell. What I should have said is that wherever we take ourselves merely to imagine, there not only is no intrinsical necessity attaching the result to the starting-place, but we also recognise that the identity of the subject is lost and that there is a breach in continuity. In memory, on the other hand, though the result is not taken as the necessary ideal development of the subject itself, yet we ignore the doubt as to a solution of con- tinuity. We connect the end of the process with and attribute it to the beginning, because the process comes to us from one end to the other without an apparent break or loss of the subject and without the suggestion of an alien intrusion, or again of a sufficient competing alternative. In imagination the connexion between subject and predi- cate is that of casual occupancy, but in memory we have possession which to such an extent is de facto that the question of title is not raised, or, if raised, it is assumed to be somehow satisfactorily settled. With regard to the distinction between inference and mere imagination that is given correctly in my Logic, p. 410. 2 A sceptical objection of this kind, if based on a psychological ground, seems (Appearance, p. 137) inconsistent with itself. The proper way to urge the objection is to compare the actual inferences which we must use with that ideal of inference which alone we can take as satisfactory.