In the first chapter of Matter and Memory, Bergson sets out to revise both the realist and idealist notions of how perception occurs, replacing them with a third account he believes easier to reconcile with the dictates of common sense and empirical observation. Both of the earlier doctrines fail, Bergson argues, because they equate perception with a kind of knowledge (28). In Bergson’s theory, the brain does not give birth to representations of any sort. Its faculties, rather, are always directed toward action, differing from the reflex functions of the spinal chord only by degree (23). Objects are perceived with regard to their utility, their relation to the body in the form of motor possibility. Not even distant or irrelevant objects are relegated to the realm of indifferent knowledge, since “even when the stimulation received is not at once prolonged into movement, it appears merely to await its occasion” (29).
The body, then, exists among an aggregate of images as a “zone of indeterminacy” that defines itself in relation to the nascent possibilities of movement implicit in all that it perceives. This perception, however, is not located in some corresponding “cerebral deposits” that mechanically translate stimulation from afferent nerves—surprisingly, Bergson asserts that perception takes place in the objects themselves. Here, he claims, it is possible to find resolution to the old debate between realism and idealism: perception does not represent the material world, but merely removes from it all that does not contribute to possible action. In a way, Bergson offers an inversion of the idealist view—the material world is not constituted entirely by our perception—our perception instead is ultimately constituted by the material world. As it passes from real to “virtual”, nothing can be added to the image, only taken away (37).
Although the brain cannot add to perceptions, it can augment them with memories, themselves nothing more than stored motor impulses awaiting an instance of relevancy to attach themselves to the present. No perception, in fact, can be said to be instantaneous (69). All perception involves some degree of duration, and therefore some amount of memory. The more memories that exist to assert a possible valency with perceived objects, the larger the zone of indeterminacy grows, and the more consciousness can resemble true volition, as opposed to a series of unmediated or mechanical reflexes.
By relating his theory to human agency (an idea he will develop in later chapters) Bergson seems preempt a fairly obvious objection to his theory of perception—that it is too deterministic, not responsive enough to problems of subjective states and irrational responses. In a less generous assessment, one could see his merely replacing one kind of mediation (the brain “representing” images rather than communicating or delaying their affect) with another (perceptions assaulted by the entirety of our memories searching for relevancy). The first model seems no less plausible than the first, especially when we consider the amount of memories that could never be instantiated in action, or the amount of perceptions to which we can assign no possibility of encounter or motor influence (how would one “act”, for example, on the light from a star, or the images produced by an electron microscope?)