A preservation/generation distinction about memory

Abstract A number of authors, including ourselves, have defended the view (which we call the “generationism about memory”) that memory is a generative, rather than preservative, source of epistemic justification. This paper clarifies the very distinction that the whole debate rests on; i.e. the distinction between preservative sources and generative sources of epistemic justification. Our aim is to present a “substantial” or “demanding” definition of preservative/generative distinction such that the candidate cases that are only superficially generative (including many of the cases that have been presented by other researchers in defence of generationism) do not count as counterexamples to preservationism. After some methodological remarks (Section 2), we propose a preservative/generative distinction, called “PM/GM”, as a first approximation (Section 3). Then, we argue the PM/GM distinction has to be revised in several ways in order to capture generationism in a substantial sense (Sections 4 and 5). Finally, we compare our revised distinction, “PM*/GM*”, with other distinctions in the literature (Section 6).