Is there hypermnesia and reminiscence for information intentionally forgotten?

Memory performance was examined across consecutive tests in three directed-forgetting experiments. Following word-method or list-method cueing to forget, significant directed forgetting was observed for all tests: Free recall for remember cue words always exceeded free recall for forget cue words. Moreover, following either cueing method, similar magnitudes of hypermnesia (improved free recall across tests) and reminiscence (recovery of words across tests) were observed for both word types. Regardless of cueing method, after an initial free recall test, the level of recovery for both word types did not differ significantly. That is, directed forgetting was not observed for the reminiscence data. Taken together, the pattern of results suggest that cues to forget impair the encoding of information but, after an initial memory test, they do not interrupt the accessing of that information. These findings are consistent with the selective rehearsal account but not the retrieval inhibition account of directed forgetting.