218 PSYCHOLOGY. wliat the Societies for ' Psychical Research ' are collect- ing and trying to interpret in the most reasonable way. If the dream were the only one of the kind the subject ever had in his life, if the context of the death in the dream differed in many particulars from the real death's context, and if the dream led to no action about the death, unques- tionably we should all call it a strange coincidence, and naught besides. But if the death in the dream had a long- context, agreeing point for point with every feature that attended the real death ; if the subject were constantly having such dreams, all equally perfect, and if on awaking he had a habit of acting immediately as if they were true and so getting ' the start ' of his more tardily informed neighbors, — we should probably all have to admit that he had some mysterious kind of clairvoyant power, that his dreams in an inscrutable way knew just those realities which they figured, and that the word * coincidence ' failed to touch the root of the matter. And whatever doubts any one preserved would completely vanish if it should appear that from the midst of his dream he had the power of inter- fering with the course of the reality, and making the events in it turn this way or that, according as he dreamed they should. Then at least it would be certain that he and the psychologist were dealing with the same. It is by such tests as these that we are convinced that the waking minds of our fellows and our own minds know the same external world. The psychologist's attitude towards cognition will be so important in the sequel that we must not leave it until it is made perfectly clear. It is a thoroughgoing dualism. It supposes two elements, mind knowing and thing known, and treats them as irreducible. Neither gets out of itself or into the other, neither in any way is the other, neither makes the other. They just stand face to face in a common world, and one simply knows, or is known unto, its counter- part. This singular relation is not to be expressed in any lower terms, or translated into any more intelligible name. Some sort of signal must be given by the thing to the mind's brain, or the knowing will not occur— we find as a matter
