Page:Bergson - Matter and Memory (1911).djvu/74

 Hence it is inferred that all sensation is naturally and necessarily unextended, so that extensity is superimposed upon sensation, and the process of perception consists in an exteriorization of internal states. The psychologist starts, in fact, from his body, and, as the impressions received at the periphery of this body seem to him sufficient for the reconstitution of the entire material universe, to his body he at first reduces the universe. But this first position is not tenable; his body has not, and cannot have, any more or any less reality than all other bodies. So he must go farther, follow to the end the consequences of his principle, and, after having narrowed the universe to the surface of the living body, contract this body itself into a centre which he will end by supposing unextended. Then, from this centre will start unextended sensations, which will swell, so to speak, will grow into extensity, and will end by giving extension first to his body, and afterwards to all other material objects. But this strange supposition would be impossible if there were not, in point of fact, between images and ideas, the former extended and the latter unextended, a series of intermediate states, more or less vaguely localized, which are the affective states. Our understanding, yielding to its customary illusion, poses the dilemma, that a thing either is or is not extended; and as the