Selective effects of emotion on the phenomenal characteristics of autobiographical memories Alexandre Schaefer and Pierre Philippot University of Louvain, Belgium The present study investigates the emotional determinants of the phenomenal characteristics of auto- biographical memories. A total of 84 participants completed the Memory Characteristics Questionnaire MCQ, Johnson, Foley, Suengas, & Raye, 1988) after retrieving and orally describing a negative, a positive, and a neutral autobiographical memory. In addition, self-report and physiological measures of emotional state at retrieval were recorded. Results suggest that recall of perceptual, sensory, and semantic elements is better for emotional memories than for neutral ones. This difference is not significant for contextual and temporal aspects, suggesting that emotional memories are more vivid but no more specific than are neutral ones. In addition, positive memories yielded higher MCQ ratings than did negative memories for sensory, temporal, and contextual aspects. Finally, correlations suggest a positive relation between emotional state at retrieval and level of phenomenal detail of retrieved memories. Results are interpreted in terms of multilevel models of emotion and of Conway and Pleydell-Pearce's 2000) model. Inconsistent results have often fed the old debate over the effects of emotion on the quality of memory recall. Indeed, two apparently contra- dictory viewpoints can be found in the literature. On the one hand, several lines of research suggest that emotion increases the richness of subjective details of a memory. First, eyewitness studies have shown that people who attend to an emotional event report vivid and detailed memories for this event e.g., Christianson & Hubinette, 1993; Yuille & Cutshall, 1986; Yuille & Tollestrup, 1992). Second, flashbulb memory studies yielded similar results. The ``flashbulb effect'' denotes the fact that individuals keep a very detailed and vivid memory of the context in which an emotional public event e.g., the attack on the World Trade Center) is first learned e.g., Bohannon, 1988; Bohannon & Symons, 1992; Christianson, 1989; Conway et al., 1994; Curci, Luminet, Finkenauer, & Gisle, 2001; Finkenauer, Luminet, Gisle, El- Ahmadi, Van der Linden, & Philippot, 1998; Pil- lemer, 1984; Rubin & Kozin, 1984). Third, experiments using the ``remember-know'' R-K) paradigm found that emotional pictures and emotional words yielded more ``remember'' responses than did neutral ones, especially for negative pictures and words Dewhurst & Parry, 2000; Ochsner, 2000). In this paradigm, a ``remember'' response indicates a specific mem- ory of the event during which the item was learned, including thoughts, feelings, and sensory details. The ``know'' response is characterised by a feeling of having previously encountered the item without any specific remembrance of the learning event. On the other hand, other lines of research suggest that emotion has a deleterious effect on the retrieval of specific details of a memory. For instance, several studies have shown that people who witness highly negative events often have a MEMORY, 2005, 13 2), 148±160 # 2005 Psychology Press Ltd http://www.tandf.co.uk/journals/pp/09658211.html DOI:10.1080/09658210344000648 Correspondence should be sent to Alexandre Schaefer, Department of Psychology, Washington University, One Brookings Drive, Campus Box 1125, St. Louis, MO 63130, USA. Email: aschaefe@artsci.wustl.edu The writing of this article has been facilitated by grants from the ``Fonds National de la Recherche Scientifique de Belgique'' FNRS) 8.4512.98 and 8.4510.99. The authors wish to thank Regis Lepage and Sandrine Charlet for helping with the data collection.