Page:Bergson - Matter and Memory (1911).djvu/322

 store up recollections or images. Thus, neither in perception, nor in memory, nor a fortiori in the higher attainments of mind, does the body contribute directly to representation. By developing this hypothesis under its manifold aspects and thus pushing dualism to an extreme, we appeared to divide body and soul by an impassable abyss. In truth, we were indicating the only possible means of bringing them together.

II. All the difficulties raised by this problem, either in ordinary dualism, or in materialism and idealism, come from considering, in the phenomena of perception and memory, the physical and the mental as duplicates the one of the other. Suppose I place myself at the materialist point of view of the epiphenomenal consciousness: I am quite unable to understand why certain cerebral phenomena are accompanied by consciousness, that is to say, of what use could be, or how could ever arise, the conscious repetition of the material universe I have begun by positing. Suppose I prefer idealism: I then allow myself only perceptions, and my body is one of them. But whereas observation shows me that the images I perceive are entirely changed by very slight alterations of the image I call my body (since I have only to shut my eyes and my visual universe disappears), science assures me that all phenomena must succeed and condition one another according to a determined order, in which