Page:Bergson - Matter and Memory (1911).djvu/197

 when the argument is inverted (although this ought to be legitimate on the hypothesis adopted), that is to say, when the intensity of the sensation is decreased instead of the intensity of pure memory being increased. For then, if the two states differ merely in degree, there should be a given moment at which the sensation changed into a memory. If the memory of an acute pain, for instance, is but a slight pain, inversely an intense pain which I feel will end, as it grows less, by being an acute pain remembered. Now the moment will come, undoubtedly, when it is impossible for me to say whether what I feel is a slight sensation which I experience or a slight sensation which I imagine (and this is natural, because the memory-image is already partly sensation); but never will this weak state appear to me to be the memory of a strong state. Memory, then, is something quite different.

But the illusion which consists in establishing only a difference of degree between memory and perception is more than a mere consequence of associationism, more than an accident in the history of philosophy. Its roots lie deep. It rests, in the last analysis, on a false idea of the nature and of the object of external perception. We are bent on regarding perception as only an instruction addressed to a pure spirit, as having a purely speculative interest. Then, as memory is itself essentially a knowledge of this kind, since its object is no longer present, we can only find between