Page:Bergson - Matter and Memory (1911).djvu/82

 interests us. But we have to take into account the fact that our body is not a mathematical point in space, that its virtual actions are complicated by and impregnated with real actions, or, in other words, that there is no perception without affection. Affection is, then, that part or aspect of the inside of our body which we mix with the image of external bodies; it is what we must first of all subtract from perception to get the image in its purity. But the psychologist who shuts his eyes to the difference of function and nature between perception and sensation,—the latter involving a real action, and the former a merely possible action,—can only find between them a difference of degree. Because sensation (on account of the confused effort which it involves) is only vaguely localized, he declares it unextended, and thence makes sensation in general the simple element from which we obtain by composition all external images. The truth is that affection is not the primary matter of which perception is made; it is rather the impurity with which perception is alloyed.

Here we grasp, at its origin, the error which leads the psychologist to consider sensation as unextended and perception as an aggregate of sensations. This error is reinforced, as we shall see, by illusions derived from a false conception of the rôle of space and of the nature of extensity. But it has also the support of misinterpreted facts, which we must now examine.

It appears, in the first place, as if the localiza-