KNOWLEDGE AS IDEALISATION. 387
if it were confined to a few special cases of eye and ear, and as if in these cases the sensations, as existences, were only double, or triple or quadruple. In fact, as existences, they are indefinitely multiple in every case. As I touch the table, how many distinct sensations do I have ? As the ray of light affects my retina, consider what a chaos of sensations is stimulated. I may remark incidentally that a large num- ber of the psychologists who have occupied themselves with the problem of space-perception do not seem to have realised the elements of the problem. They first talk as if the problem were: How to get space-relations out of sensations, as existences ? and secondly, as if the problem were : Given isolated sensations as equivalent to isolated points in space, to tell how these come to be connected with each other in complex space-forms ? But the problem in the first place is : How do we interpret sensations into spatial meanings ? and secondly : How do we interpret some sensations as isolated points and others as connected bodies ? We do not start with separate points which are to be combined through the medium of motion, or in any other way. The separate point is as much an inference, an interpretation of the sensation, as the connected line, surface, or solid. Our experience of one is built up along with that of the other. Sensations, as existences, in spatial perceptions as in all perceptions, are naught ; sensations, in their symbolic quality, as inference is put into them and they become meaning, are all. Our fundamental position is that sensation, as existence, and the process, as psychical occurrence, by which sensations are connected, never enter into knowledge. Knowledge is both the sensation and the process in their significant or sign-bearing quality. But what is the sense in calling the sign-borne content inferential, and in separating it from the sensational basis as immediate ? The general ground is the fact that the sensuous clustering is all that is present by way of im- mediate existence, and it is convenient to have a term to express that which is present by way of being signified or symbolised. The sensuous basis stands for, conveys to intelligence, the content of the experience, and the meaning is present only as thus represented. The sensations, as immediate existences, have no more meaning than letters of the alphabet or than vocal noises. The meaning is read into them or out of them, as one may prefer to state it. But more specifically, this element may be called mediate or inferential because it is present as the result of a process