Anscombe on Memory

The person believes, that is assents to, the content of this experience. This assent does not, as in our previous argument, involve judgement of what the memory experience presents. For no one has the idea of 'the past' except in the first place from memory, and hence if memory is an experience the idea of the past must simply be in the experience: one cannot bring any primary judgement about the past to bear on the presentation. It is indeed difficult to see what the belief or assent consists in, as it cannot involve assessment of the experience. This was a problem which Hume felt strongly, and could not solve. It seems that assent can only be: allowing the memory experience to feel 'solid'. Plainly we have fallen into idealism.

(from Memory, Experience and Causation (Contemporary British Philosophy ed. H.D. Lewis publ.1974))

The empiricist orientation which seeks a mark of some kind to show that an experience is a memory experience throws its weight against the solid supposedly idealist view. A house has fallen on you and you have been buried for 3 days. I would say that such was a solid memory marks or no marks. You don't need to have the impossible re-experienceable original for that. Nevertheless there's a spectrum from the very definite to the imaginary re-constitution of the original. Both count as memory because the same power is at work. Woozley's point that even if the original experience is not around anymore memory is still involved with it seems to be non informitively obvious. It's a route on which one might meet Bergson on his way back.