Whenever there is actual fact there is memory, and memory creates duration which excludes repetition. Perceived familiarity depends upon memory but memory, according to Bergson, does not work by preserving a series of repetitions for future reference. If we say that memory connects “the past” with “the present” we must add that it destroys their logical distinctness. But of course this is putting it very badly: there is really no “logical distinctness” in the actual fact for memory to “destroy”: our language suggests that first there was matter, forming a logical series of distinct qualities recurring over and over, and then memory occurred and telescoped the series, squeezing “earlier” and “later” moments into one another to make a creative duration. Such a view is suggested by our strong bias towards regarding abstractions as having independent existence apart from the real fact from which they have been abstracted: if we can overcome this bias the description will do well enough. According to Bergson, as we have just seen, every actual fact must contain some memory otherwise it would not be a fact but simply matter, since it is an act of memory that turns matter into perceived fact. Our ordinary more or less familiar facts, however, contain much more than this bare minimum. The facts of everyday life are perceived as familiar and classified from a vast number of points of view. When you look at a cherry you recognise its colour, shape, etc., you know it is edible, what it would taste like, whether it is ripe, and much more besides, all at a glance. All this knowledge depends on memory, memory gives meaning to what we might call bare sensation (which is the same thing as Bergson’s present matter) as opposed to the full familiar fact actually experienced. Now the meaning is ordinarily contained in the actual fact along with the bare sensation not as a multiplicity of memories distinct from the bare sensation, but, as we put it, at a glance. This peculiar flavour of a familiar fact can be analysed out as consisting of memories of this or that past experience, if we choose to treat it in that way, just as a fact can be analysed into qualities. According to Bergson this analysis of the meaning of a familiar fact into memories would have the same drawbacks as the analysis of a present fact into qualities: it would leave out much of the meaning and distort the rest. Bergson holds that wherever there is duration the past must be preserved since it is just the preservation of the past, the creation of fact by a synthesis of what, out of synthesis, would be past and present, which constitutes duration. The essential point about mental life is just the performing of this act of synthesis which makes duration: wherever there is mental life there is duration and so wherever there is mental life the past is preserved. “Above everything,” Bergson says, “consciousness signifies memory. At this