Like all other descriptions and explanations, such an account must, of course, be expressed in terms of abstractions, and so is liable to be misunderstood unless the false implications of these abstractions are allowed for and discounted. According to Bergson the only actual reality is the changing fact itself, everything else is abstraction: this reality however is not confined to the fragment called “our present experience” which is in the full focus of consciousness and is all that we usually suppose ourselves to know directly; it includes besides everything that we are in a sense aware of but do not pay attention to, together with our whole past: for Bergson, in fact, reality coincides with the field of virtual knowledge, anything short of this whole field is an abstraction and so falsified. Even to say “we know this fact” is unsatisfactory as implying ourselves and the fact as distinct things united by an external relation of knowing: to say “the fact is different from the abstraction by which it is explained” similarly implies logically distinct terms in an external relation of difference, and so on. If Bergson is right in claiming that the actual fact is non-logical then obviously all attempts to describe it, since they must be expressed in terms of abstractions, will teem with false implications which must be discounted if the description is to convey the meaning intended. Bergson’s claim is that if we allow ourselves to attend to the changing fact with which we are actually acquainted we are driven to a theory of reality different from the theory of things and relations accepted by common sense. The two abstractions by means of which he attempts to express this new theory are matter and memory. In the actual fact Bergson would hold that both these notions are combined by synthesis in such a way as no longer to be distinct, or rather, for this implies that they started distinct and then became merged, it would perhaps be better to say that these two notions are abstractions from two tendencies which are present in the actual fact. In the actual fact they combine and, as it were, counteract one another and the result is something different from either taken alone, but when we abstract them we release them from each other’s modifying influence and the result is an exaggeration of one or other tendency which does not really represent anything which actually occurs but can be used, in combination with the contrary exaggeration, to explain the actual fact which may be described as being like what would result from a combination of these two abstractions. We will take matter first.