La mémoire de source constitue l’ensemble des processus cognitifs impliqués dans la reconnaissance de l’origine de l’information et participe de ce fait au fonctionnement adapté dans la vie quotidienne. Cet article a pour objectif de faire une revue de la littérature des modèles théoriques et neuro-anatomiques proposés pour expliquer la mémoire de source, des différentes méthodes d’évaluation utilisées, des facteurs (démographiques, cognitifs) qui influencent ce phénomène mnésique, ainsi que ses déficits dans différentes pathologies psychiatriques et en particulier la schizophrénie. L’étude de la mémoire de source s’est montrée particulièrement pertinente pour l’étude de la schizophrénie en proposant des modèles explicatifs explorant les symptômes positifs, en particulier les hallucinations auditives. Toutefois, le déficit en mémoire de source n’est pas spécifique de la schizophrénie puisque des déficits ont été mis en évidence, d’une part, au cours d’autres pathologies psychiatriques et, d’autre part, en population générale en corrélation avec des traits schizotypiques. La mémoire de source est un concept intéressant pour les recherches portant sur l’identification des substrats neurocognitifs des symptômes positifs. Toutefois, des améliorations méthodologiques, notamment une homogénéisation des paradigmes utilisés et une meilleure sélection des populations étudiées, sont nécessaires.

Source monitoring refers to the ability to remember the origin of information. Three source monitoring processes can be distinguished: external source monitoring, internal or self-monitoring and reality monitoring (i.e. discrimination between internal and external sources of information). Source monitoring decisions are based on memory characteristics recorded such as perceptions, contextual information or emotional reactions and heuristic or more controlled judgement processes.

Several studies suggested that specific structures in the prefrontal and the mediotemporal lobes are the main areas implicated in source monitoring.

A typical source monitoring paradigm includes an items generation stage and a second stage of recognition of items (old versus new) and identification of their sources: external (usually the examiner) or internal (the subject). Several indices can be calculated based on the raw data such as the number of false alarms, attribution biases or discrimination indexes. To date, there is no standardized source monitoring task and differences in the type of items used (words, pictures), in the cognitive or emotional effort involved or in the delay between the two test stages, contribute to the heterogeneity of results.

Factors such as age (either very young or very old) and emotions influence source monitoring performances. Influence of gender was not properly explored, whereas the role of IQ and selective attention is still debated.

Source monitoring deficits are observed mainly in disorders affecting frontotemporal areas, such as frontal trauma, Alzheimer’s disease or frontotemporal dementia.

Source monitoring errors (e.g. external misattribution of self-generated information) are observed in schizophrenia and seem to correlate with positive symptomatology, in particular auditory hallucinations, thought intrusion and alien control symptoms. These results are of particular interest in clinical research because source monitoring is one of the rare cognitive tests showing a correlation with the positive dimension. Source monitoring deficits have been proposed as a potential explanation for the positive symptoms and some, but not all studies lent support to this hypothesis. Heterogeneity of studied samples, in particular different criteria to define hallucinating subjects (e.g. currently versus anytime during their lives), could explain the discordant results.

Source monitoring impairments were observed in pharmacological models of psychosis, in first degree relatives of schizophrenic patients, and also in the general population associated with schizotypal dimensions. These results support a relationship between source monitoring deficits and some of the symptomatic dimensions of the schizophrenic spectrum but still await replication.

Some studies found source monitoring deficits in other psychiatric conditions such as mania or obsessive–compulsive disorder. Thus, those studies suggest that source monitoring deficits may be not specific to schizophrenia.

Source monitoring competencies are critical for good (i.e. adapted) everyday functioning. Source monitoring deficits have been suggested as a potential explanation for some (or all) positive psychotic symptoms. However, to date, methodological inconsistencies (especially with regard to test design and choice of subjects’ samples) have precluded firm, definite conclusions.