Article révisé par les pairs
Résumé : Consistent evidence nowadays indicates that sleep protects declarative memory from lexical interference. However, little is known about its effect against emotional interference. In a within-subject counterbalanced design, participants learned a list of word pairs after a mood induction procedure (MIP), then slept or stayed awake during the post-learning night. After two recovery nights, half of the list was recalled after a similar mood induction than at the encoding session (no interference condition) and the other half after a different mood induction (interference condition). Amongst participants for whom the MIP was effective, an emotional interference effect appeared only in the sleep-deprived condition, with a lower recall of word pairs subjected to contextual interference than of the other pairs. These findings support the hypothesis of a decoupling between memories and their "affective blanket" during post-learning sleep, protecting recent memories against emotional contextual interference.