Ce travail de recherche examine la représentation de l'expérience des violences extrêmes dans l'espace fictionnel contemporain de l'Afrique subsaharienne francophone. Les nombreuses fictions en prose produites dans le sillage des conflits armés des années 90 et du génocide au Rwanda soulèvent des interrogations liées à la représentation de la douleur, de la cruauté et de la mort ainsi qu’à l'éthique de l'art. Comment le texte littéraire met-il en récit les événements traumatiques ? Comment l'écrivain pense et problématise-t-il des crises extrêmes relevant de l'histoire immédiate ? Selon quelles modalités littéraires sont-elles constituées en un objet de connaissance et de sensibilisation ? Et quelles sont les stratégies langagières et esthétiques privilégiées pour transmettre la mémoire des atrocités dans une visée de témoignage ou de réflexion critique ? Cette thèse explore la mise en écriture de ces drames collectifs qui inaugurent, sur les plans historique et socioculturel, une nouvelle ère de violence en Afrique subsaharienne. Dans ce contexte, nous nous intéressons surtout aux ouvrages qui – en raison des choix formels et stylistiques singuliers adoptés par leurs auteurs – se caractérisent par une radicalisation du discours et des scénographies particulièrement violentes. Cette étude qui, au final, interroge les pouvoirs et les possibles limites de l'art dans la représentation de faits de violence extrême analyse un vaste corpus de romans et de nouvelles parus entre 1998 et 2010 et propose une approche pluridisciplinaire, puisant, à côté des théories littéraires et esthétiques, dans l'histoire, la sociologie, l'anthropologie et la psychiatrie. / This research project examines the representation of the experience of extreme violence in the contemporary fictional space of Sub-Saharan Francophone Africa. The numerous works of prose fiction written in the wake of the armed conflicts of the 1990s and the Rwandan genocide raise questions related to the representation of pain, cruelty and death as well as to the ethics of art. How do literary texts put into narrative traumatic events? How do writers think and problematize extreme crises of immediate history? By the means of what literary modalities are these crises constituted into an object of knowledge and awareness? And what esthetic and language strategies have been privileged to convey the memory of the atrocities in order toprovide testimony or aim at critical reflection? This thesis explores the writing of the collective tragedies that, from a historical and socio-cultural perspective, mark the start of a new period of violence in Sub-Saharan Africa. In this context, we are focusing predominantly on texts that are characterized – through the distinctive choices of form and style operated by the authors – by a radicalization of discourse and particularly violent plots and esthetics. This research which interrogates the powers and the possible limits of art in the representation of facts of extreme violence analyses an extensive corpus of novels and short stories published between 1998 and 2010 and suggests a multidisciplinary approach which, next to literary and esthetic theories, draws on history, sociology, anthropology and psychiatry.

This research was executed as three separate studies. Study 1 focused on the perception of pain and the semantic aspects of pain. Study 2 focused on the measurement of acute pain and mood states. Study 3 focused on the psychological treatment of cancer pain. In Study 1 a group of 66 Setswana-speaking adults were required to describe what they saw, what happened, and what would be the outcome with respect to three visually presented pain scenes using The Pain Apperception Test (PAT) A qualitative analysis of their responses shows that pain is experienced as an all encompassing experience affecting all aspects of their lives, such as the physical, emotional, social, and economic. This was found to be true, irrespective of gender and age with the exception of economic issues. A qualitative analysis of their responses to the Pain Eliciting Incidents Questionnaire (PEIQ) reveals that the Setswana pain descriptors are classifiable according to the three dimensions of pain namely, the sensory-discriminative, affective-motivational, and cognitive-evaluative. Sludy 2 applied the Profile of Mood States (POMS) preoperatively to a group of 58 female laparotomy (gynaecological) patients. These patients were also tested post-operatively with the Visual Analogue Scale (VAS) and the Wisconsin Brief Pain Questionnaire (WBPQ) as pain measures. The pain measures were taken at no medication and at the peak of medication. Factor analysis could not confirm the validity of the six POMS scales. These scales also did not show correlations with post-operative pain. Correlations between the pain measures showed acceptable reliability and validity of the VAS and the WBPQ. In Study 3 three groups of 15 cancer patients each, suffering from chronic pain, were treated over a period of two weeks with either cognitive behavioural therapy plus medication, reassurance therapy plus medication, or medication only. Comparison of before and after treatment pain measures showed that both cognitive behavioural therapy and reassurance therapy had a beneficial effect. Follow-up results three months later showed that the beneficial effect of reassurance therapy did not persist. Patients treated with cognitive behavioral therapy still showed the beneficial effects thereof. / Psychology / D. Litt. et Phil. (Psychology)

This research was executed as three separate studies. Study 1 focused on the perception of pain and the semantic aspects of pain. Study 2 focused on the measurement of acute pain and mood states. Study 3 focused on the psychological treatment of cancer pain. In Study 1 a group of 66 Setswana-speaking adults were required to describe what they saw, what happened, and what would be the outcome with respect to three visually presented pain scenes using The Pain Apperception Test (PAT) A qualitative analysis of their responses shows that pain is experienced as an all encompassing experience affecting all aspects of their lives, such as the physical, emotional, social, and economic. This was found to be true, irrespective of gender and age with the exception of economic issues. A qualitative analysis of their responses to the Pain Eliciting Incidents Questionnaire (PEIQ) reveals that the Setswana pain descriptors are classifiable according to the three dimensions of pain namely, the sensory-discriminative, affective-motivational, and cognitive-evaluative. Sludy 2 applied the Profile of Mood States (POMS) preoperatively to a group of 58 female laparotomy (gynaecological) patients. These patients were also tested post-operatively with the Visual Analogue Scale (VAS) and the Wisconsin Brief Pain Questionnaire (WBPQ) as pain measures. The pain measures were taken at no medication and at the peak of medication. Factor analysis could not confirm the validity of the six POMS scales. These scales also did not show correlations with post-operative pain. Correlations between the pain measures showed acceptable reliability and validity of the VAS and the WBPQ. In Study 3 three groups of 15 cancer patients each, suffering from chronic pain, were treated over a period of two weeks with either cognitive behavioural therapy plus medication, reassurance therapy plus medication, or medication only. Comparison of before and after treatment pain measures showed that both cognitive behavioural therapy and reassurance therapy had a beneficial effect. Follow-up results three months later showed that the beneficial effect of reassurance therapy did not persist. Patients treated with cognitive behavioral therapy still showed the beneficial effects thereof. / Psychology / D. Litt. et Phil. (Psychology)