Résumé
Aphasia is an acquired and specific language disorder, caused by a neurological lesion. It is rehabilitated by speech and language therapists, who use therapies based on neuroplasticity to lead their patients to recover their language as much as possible. Aphasia can concern speakers of all languages, including signed languages such as French Sign Language (LSF). To this day, no protocol specifically thought for aphasic LSF speakers exists. Existing therapies for aphasia in oral languages cannot be used directly with a signing patient, because of fondamental formal differences due to the use of an audio-phonatory channel in one case, and visuo-gestual in the other. However, studies led in the neuro-anatomy and aphasiology fields show parallels between cerebral organisation of oral and signed languages, that let us suppose that current therapies for oral aphasia might be adapted for sign aphasia rehabilitation. This master thesis presents an adaptation of the Phonological Components Analysis (PCA), which is a well known therapy for anomia, a systematic symptom in aphasia. We test this adaptation, in addition with executive exercises, with 3 LSF adult speakers, who became aphasic after strokes. We scale the evolution of their naming abilities and subjective feelings about their communication, first without training, then during an intensive training phase, and at last after a new training-less phase.Results indicate a training’s tendency to improve naming for treated items, generalising to untreated items in some cases, which seems to last in time. This improvement is more obvious in participants with a mild or moderate executive disorder, and seems to benefit less when the initial language labilities remain high (mild aphasia). These conclusions lay on the qualitative analysis of the participants scores and statements, and our methods don’t allow us to show statistically significant results. Researches deepening the topic could obtain more informative results regarding the protocol's efficiency.