Page:Bergson - Matter and Memory (1911).djvu/230

 them: this similarity of reaction following actions superficially different is the germ which the human consciousness developes into general ideas.

Consider, indeed, the purpose and function of our nervous system as far as we can infer them from its structure. We see a great variety of mechanisms of perception, all bound, through the intermediary of the centres, to the same motor apparatus. Sensation is unstable; it can take the most varied shades; the motor mechanism, on the contrary, once set going, will invariably work in the same way. We may then suppose perceptions as different as possible in their superficial details: if only they are continued by the same motor reactions, if the organism can extract from them the same useful effects, if they impress upon the body the same attitude, something common will issue from them, and the general idea will have been felt and passively experienced, before being represented.—Here then we escape at last from the circle in which we at first appeared to be confined. In order to generalize, we said, we have to abstract similarity, but in order to disengage similarity usefully we must already know how to generalize. There really is no circle, because the similarity, from which the mind starts when it first begins the work of abstraction, is not the similarity at which the mind arrives when it consciously generalizes. That from which it starts is a similarity felt and lived; or, if you prefer