Abstract

The problem of forgotten evidence consists of a pair of scenarios originally proposed by Alvin Goldman. In the “forgotten good evidence” and “forgotten bad evidence” scenarios, subjects hold the same memory belief while irreversibly forgetting its original, though different, pieces of evidence. The two scenarios pose a series of challenges to current time slice (CTS) theories, which posit that memory beliefs are justified solely by contemporaneous states. Goldman’s two scenarios pose an apparent dilemma to CTS theories given a naïve picture of how a memory belief is successfully retained while its evidence is irreversibly forgotten. In my view, however, CTS theories may find a solution to the apparent problem by carefully examining the conditions under which a memory belief is successfully retained while its evidence is completely forgotten. Namely, the two scenarios overlook an important difference between forgetting good evidence and forgetting bad evidence.