Page:Bergson - Matter and Memory (1911).djvu/129

 consists in the common element seized and disengaged by the mind, there is a vague and in some sort objective resemblance, spread over the surface of the images themselves, which might act perhaps like a physical cause of reciprocal attraction. And should we ask how it is, then, that we often recognize an object without being able to identify it with a former image, refuge is sought in the convenient hypothesis of cerebral tracks which coincide with each other, of cerebral movements made easier by practice, or of perceptive cells communicating with cells where memories are stored. In truth, all such theories of recognition are bound to melt away, in the end, into physiological hypotheses of this kind. What they were aiming at, first, was to make all recognition issue from a bringing together of perception and memory; but experience stands over against them, testifying that in most cases recollection emerges only after the perception is recognized. So they are sooner or later forced to relegate to the brain, in the form of a combination between movements or of a connexion between cells, that which they had first declared to be an association of ideas; and to explain the