I am currently writing on the following skeptical paradox:

(1) A belief is (epistemically) reasonable only if one has good, presently-available evidence for it.
(2) We have no good, presently-available for most of the things we recollect.
(3) Therefore, most of our recollections are not (epistemically) reasonable.

I call this the paradox of absent evidence.

The point of epistemic reasonability is that it is aimed at true beliefs and not, say, at beliefs that are pleasant, practical, prudent, et cetera. It might be prudent to believe a proposition for which one has no evidence but it would not, at least if (1) is true, be epistemically rational to believe it.

It seems that (1) and (2) are true and that (3) must be true if (1) and (2) are. So where is the mistake? Or should we just accept skepticism about memory?