Autobiographical memory is considered the most durable memory component in the human memory system. This thesis aimed to examine the characteristics of autobiographical memory performance and its relationship with executive functioning in normal and pathological ageing. The particular focus was on two aspects of autobiographical memory - recall specificity and fluency - that have been little investigated in cognitive ageing research to date. The literature reviewed included the domains of cognitive and clinical psychology, neuropsychology, and gerontology. It is clear from the review of the literature that considerable uncertainty exists regarding the status of autobiographical memory in healthy ageing and dementia. Moreover, little is known about the involvement of executive processes in the retrieval of personal remote memories in old age. A series of cross-sectional quasi-experimental studies is reported. The autobiographical memory and executive functioning of 136 participants were assessed by utilising a variety of neuropsychological instruments, and a range of statistical techniques were used in the analysis of the interactions between these two cognitive systems. The findings are discussed in terms of theoretical and practical implications for cognitive, ageing, and clinical research. To the author's knowledge, no research to date has attempted to investigate the specificity of autobiographical recall in a sample of Alzheimer's patients - an aspect this study addressed, thus providing an original contribution to knowledge. In addition, it has been established that the autobiographical memory overgenerality identified in Alzheimer's Disease was expressed through an excess of categoric memories. These findings were followed by an attempt to establish whether autobiographical memory performance is influenced by age-associated changes in executive functions in healthy ageing. The thesis adopted as its main theoretical framework the executive decline hypothesis, the proposal that a subtle decline in executive functions could account for age-related decreases in memory (Troyer, Graves & CulIum, 1994; Crawford, Bryan, Luszcz, Obonsawin & Stewart, 2000). For the first time, the investigation of this hypothesis has been extended to autobiographical memory recall. The results indicated that the executive functions, when indexed by a composite score, failed to mediate ageassociated retrieval of personal memories in older adults. However, a specific executive function - verbal fluency - predicted a significant amount of age-related variance of autobiographical memory recall. These findings were interpreted in terms of the mediator-specificity requirements in mediational models. Further, the thesis examined the impact of verbal fluency strategic mechanisms, clustering and switching, on autobiographical memory retrieval. The results indicated that the strategic mechanisms employed in verbal fluency failed to predict autobiographical memory recall in older adults, as opposed to the original measure of total words generated across the verbal fluency tasks. This finding supported previous research questioning the sensitivity of the new verbal fluency scoring methodology (Troyer, Moscovitch & Winocur, 1997). An additional factor that has not been fully explored in previous research is the possibility that certain aspects of executive function could be important for different components of autobiographical memory. By analysing the existence of different subcomponents within the two cognitive systems and their association, this thesis provided additional evidence for the debate over the unity and diversity of executive functions. To summarise, it has been established that verbal fluency mediates age-related autobiographical memory recall in older adults, manifested through autobiographical fluency. This finding is explained in terms of the common source proposal, referring to the processes of strategic search underlying performance on both tasks. Further investigation is needed to clarify these issues and examine the age effects on intra-individual variability in executive functioning and autobiographical memory.

Il existe un lien bidirectionnel entre identité et mémoire, la mémoire autobiographique contenant toutes les informations qui permettent à un individu de se définir. Or, un point important de notre identité est de se définir en tant qu'homme ou femme. De plus, d'un point de vue cognitif, hommes et femmes présentent des différences de performances à un certain nombre de tâches dont des tâches de mémoire autobiographique. Dans cette thèse, nous avons souhaité investiguer à partir de méthodes comportementales et en neuroimagerie fonctionnelle les différences entre hommes et femmes dans la mémoire autobiographique épisodique et sémantique et ses liens avec l'identité, afin de déterminer si ces différences sont plutôt dues à des facteurs biologiques inhérents au sexe ou sociaux inhérents à l'identité de genre. Nos résultats convergent vers l'idée que l'identité de genre est plus déterminante que l'appartenance à un sexe dans la construction des souvenirs. / There is a bidirectional link between identity and memory because autobiographical memory contains all the information that allows an individual to define oneself. However an important matter about our identity is to be defined as a man or a woman. In addition, from a cognitive point of view, men and women show differencies in their performance in some tasks including autobiographical memory tasks. In this thesis, we wanted to investigate the differences between men and women in episodic and semantic autobiographical memory with behavioural and functional neuroimaging methods and their links with identity so as to determine if these differences are rather due to biological factors inherent to sex or to social factors inherent to gender identity. Our results converge on the idea that gender identity is more crucial than membership to one sex in the construction of memories.

La mémoire autobiographique est vue comme un ensemble d'informations et de souvenirs personnels permettant de construire un sentiment d'identité. Elle se développe progressivement et apparaît très sensible aux pathologies neurodéveloppementales. La mémoire autobiographique est intimement liée au Soi lui permettant d'encoder et de récupérer toutes ses représentations et expériences. Ainsi, le Soi se constitue d'aspects explicites comprenant la mémoire autobiographique mais également d'aspects plus implicites relatifs aux expériences corporelles du sujet. L'atteinte des aspects implicites et explicites du Soi pourrait rendre compte de certains symptômes psychotiques et des difficultés d'adaptation des patients atteints de schizophrénie. La schizophrénie est une pathologie neurodéveloppementale qui débute à la fin de l'adolescence mais qui pourrait puiser son émergence dans des stades bien plus précoces. Le premier épisode psychotique qui signe l'entrée dans la phase active de la maladie est généralement précédé par une phase « prodromique » où des symptômes cliniques sont présents à un niveau infraliminaire du seuil de psychose. Ces sujets sont qualifiés de sujets à ultra haut risque de psychose (UHR). Les troubles du Soi sont bien documentés dans la schizophrénie, néanmoins très peu de données sont disponibles concernant les sujets UHR. Le but de cette thèse est multiple : (i) mesurer l'impact des anomalies neurodéveloppementales sur la mémoire autobiographique, (ii) objectiver des déficits de la mémoire autobiographique dès la phase prodromique, (iii) évaluer l'ensemble des composantes du Soi afin d'investiguer leurs interactions et l'impact des anomalies développementales sur celles-ci. Nous avons ainsi effectué trois études. Notre première étude a investigué le lien entre le poids des anomalies neurodéveloppementales et la mémoire autobiographique en comparant deux pathologies neurodéveloppementales, une à début tardif : la schizophrénie, et l'autre à début précoce : les troubles du spectre autistique. Nous avons pu mettre en évidence un pattern de performances similaires entre les deux populations bien que les mécanismes responsables des troubles en mémoire autobiographique apparaissent distincts. Dans notre deuxième étude, nous avons comparé les performances autobiographiques de patients atteints de schizophrénie par rapport à celles de sujets UHR. Nos résultats révèlent un déficit de la mémoire autobiographique aussi sévère chez les sujets UHR que chez les patients atteints de schizophrénie mettant ainsi en évidence une atteinte de cette fonction en amont du premier épisode psychotique. Dans la lignée de ces résultats, nous avons conduit une troisième étude. Le but étant de situer la mémoire autobiographique dans un contexte plus large, celui du Soi, tout en intégrant une composante développementale. Nous avons élaboré et proposé une batterie d'investigation examinant différents aspects du Soi implicites et explicites, combiné à l'évaluation d'anomalies du neurodéveloppement. Celle-ci a été administrée chez des sujets UHR en comparaison à des patients atteints de schizophrénie. Au final, nos résultats révèlent un impact de la charge neurodéveloppementale sur les différents aspects du Soi, la pertinence d'investiguer au sein d'un même protocole ces différents aspects et la présence d'anomalies du Soi déjà présents chez les sujets UHR, constituant potentiellement des marqueurs prédicteurs de transition psychotique et permettant d'améliorer la détection précoce de ces sujets et leur prise en charge. / Autobiographical memory is delineated as a set of personal information and experiences to build a sense of identity. It develops gradually and appears very sensitive to neurodevelopmental disorders. Autobiographical memory is intimately linked to the self, enabling it to encode and retrieve all its representations and experiences. Thus, the self is constituted of explicit aspects including autobiographical memory but also by more implicit aspects relating to the subject's body. Implicit and explicit self-aspects alterations may account for certain psychotic symptoms and adaptation difficulties in patients with schizophrenia. Schizophrenia is a neurodevelopmental disorder that begins at the end of adolescence but which could emerge in much earlier stages. The first psychotic episode that signs the beginning of the active phase of schizophrenia is usually preceded by a "prodromal phase" during which clinical signs are present at a sub-threshold level. Individuals experiencing these signs are considering as Ultra High Risk of psychosis (UHR). Self-disorders are well documented in schizophrenia, however very little is known regarding UHR subjects. The aim of this thesis is multiple: (i) to measure the impact of neurodevelopmental anomalies on autobiographical memory, (ii) to objectify autobiographical memory deficits in the prodromal phase, (iii) to evaluate all the self-components in order to investigate their interactions and the impact of developmental anomalies. We have conducted three studies. Our first study investigated the relationship between neurodevelopmental anomalies and autobiographical memory by comparing two neurodevelopmental disorders, one with late onset: schizophrenia and the other with early onset: autism spectrum disorders. Results revealed a pattern of similar performances between the two populations although the mechanisms responsible for autobiographical memory impairment do not appear the same. In our second study, we compared the autobiographical performances of patients with schizophrenia compared to those of UHR subjects. Our results highlighted a deficit of autobiographical memory as severe in UHR subjects as in patients with schizophrenia, thus revealing an impairment of this function upstream of the first psychotic episode. In line with these results, we conducted a third study. The aim was to situate the autobiographical memory in a wider context, the multi-componential Self, while integrating a developmental component. We developed and proposed a battery investigating different self-components, combined with the assessment of neurodevelopmental anomalies. This battery was administered in UHR subjects compared to patients with schizophrenia. Finally, our results reveal an impact of neurodevelopmental abnormalities on the different self-aspects, the relevance of investigating these different self-aspects within the same protocol and the presence of self-abnormalities already present in the UHR subjects, constituting potentially predictive marker of psychotic transition and improving the early detection of these subjects and the development of healthcare and reinsertion programs.