Cette recherche s’intéresse aux mécanismes d’activation impliqués dans le traitement des informations émotionnelles, en référence aux perspectives épisodique et fonctionnelle de la mémoire. Deux épreuves d’amorçage de répétition ont été utilisées impliquant la présentation d’expressions faciales émotionnelles (EFE). Dans ces tâches d’amorçage, la manipulation de facteurs tels que le type de lien ou le décours temporel entre l’amorce et la cible (SOA) a permis de sonder les processus impliqués dans l’activation des informations mnésiques. Les résultats montrent un effet d’amorçage lors du traitement de la cible pour les SOA courts (100ms), quelle que soit l’émotion présentée (tristesse ou joie). En revanche, pour les SOA plus longs (500ms et 1000ms), l’effet d’amorçage n’est présent que pour les essais où l’amorce présentait une expression faciale de tristesse. Ces résultats sont discutés en référence aux traitements automatiques des informations émotionnelles.

Short-term priming paradigms are usually used to investigate the nature and organization of information in memory. The processes implemented in the activation of information in memory can be explained by manipulating factors, such as the link or the delay between prime and target. Models of embodied cognition (Wilson, 2002) and grounded cognition (Barsalou, 2008) integrated perceptual processes in the emergence of concepts and in memory processes. The purpose of the present study is to investigate on the relationship between perceptual and memory mechanisms in a context of memory activation for specific stimuli such as emotional facial expressions (EFE). We conducted two repetition priming tasks manipulation, using EFE in congruent versus non congruent relations. The first task was designed as a go–no go task using joy and sadness in congruent versus non congruent relation. In a counterbalance procedure, participants were requested for part of the experiment, to respond as fast as possible to targets representing joyful expressions and for the other part, to targets representing sad expressions. Surprisingly, results revealed priming effect only for sad prime but not for joyful prime. For the second experiment, we used different delays of SOA between the target and the prime presentation. The manipulation of SOA is closely linked to the hypothesis of automaticity of cognitive processes. More precisely the priming effect is posited to be triggered by short, ephemeral processes, because this effect is observed for short SOAs (less than 300ms), and it tends to disappear when increasing the SOA. Based on the automaticity hypothesis we employed an evaluative decision task by manipulating delay of presentation of primes, short delay (100ms) and longer delay (500ms). Participants were required to make decisions on the nature of expressions. Results of the second experiment, revealed priming effect for a short delay between prime and target (SOA 100ms). However, for the longer delay (SOA 500ms) the priming effect was obtained only when primes represented sad facial expressions. For joyful facial expressions, the inverse pattern of results was observed. Results are discussed in terms of inhibition of specific joyful emotions known to cause particular behavior and in line with the literature on the time course of automatic processing for emotional information.