We have seen that this assumption is taken for granted in the account which is ordinarily given (or would be given if people were in the habit of putting their common sense assumptions into words) of how it is that facts come to be classified: facts are supposed to fall into classes because they share common qualities, that is because, in the changing fact directly known, the same qualities recur over and over again. There is no doubt that the fact with which we are directly acquainted can be classified, and it is equally undeniable that this fact is always changing, but if this change has the form of creative duration then its classification cannot be based upon the repetition of qualities at different “stages” in its course. It follows that either the fact with which we are directly acquainted does not change as a creative process, or else that we are quite wrong in assuming, as we ordinarily do, that we actually know qualities directly and that it is these qualities which form the basis of classification, and hence of all description and explanation. We have already seen that this assumption, though at first sight one naturally supposes it to be based on direct acquaintance, may really depend not on any fact directly known but on our preoccupation with explanation rather than with mere knowing. But if we never really are acquainted with qualities, if qualities are, as Bergson says, mere abstractions, how come we to be able to make these abstractions, and why do they apply to actual facts? If classification is not based on common qualities discovered by analysis and repeated over and over as actual facts directly known, on what is it based? We certainly can classify facts and these abstract common qualities, if abstractions they be, certainly correspond to something in the facts since they apply to them: what is the foundation in directly knowu fact which accounts for this correspondence between abstractions and facts if it is not qualities actually given as part of the facts? These questions are so very pertinent and at the same time so difficult to answer satisfactorily that one is tempted to throw over the view that the changing fact which we know directly forms a creative duration. This view is impossible to express without self-contradiction and it does not fit in with our accustomed habits of mind: nevertheless if we do not simply reject it at once as patently absurd but keep it in mind for a while and allow ourselves time to get used to it, it grows steadily more and more convincing: we become less and less able to evade these difficult questions by accepting the common sense account of what we know directly as consisting of a series of qualities which are repeated over and over, and more and more driven to regard it as a process in creative duration which does not admit of repetitions. There is no difficulty in seeing, the moment we pay attention, that what we know directly certainly does change all the time: but if we try to pin