Organisation mentale des connaissances conceptuelles relatives aux verbes d'action : études de cas uniques

Study one We report the case of LC, a brain-damaged patient presenting with a disproportionate production impairment for verbs in comparison with nouns. We showed that LC’s verb disproportionate impairment in naming tasks was neither due to the lower imageability of verbs, nor to their greater number of lexical competitors in comparison to nouns. Furthermore, we showed that, contrary to the semantic account put forward by Bird, Howard and Franklin (2000) to explain any grammatical category-specific deficits, LC’s verb disproportionate production deficit could not be attributed to a semantic origin. We report the case study of a patient JB with the frontal variant of frontotemporal dementia (fv-FTD), who was disproportionately impaired in naming and comprehending verbs in comparison with nouns. We examined to what extent the patient’s verb disproportionate deficit was dependent on the type of stimuli used to assess verb processing, that is, static depictions of actions, videotaped actions, or verbal stimuli. We found that the verb disproportionate deficit JB presented when her naming or comprehension was assessed from static depictions of actions (i.e., photographs) disappeared when naming or comprehension was assessed from videotaped actions or verbal stimuli. These results indicated that JB did not present disproportionate difficulties with verb processing per se (i.e., with retrieving the lexical and semantic features of verbs). Instead, the seemingly disproportionate verb deficit found in JB – and possibly also in other previously reported patients with executive resource limitation – was likely due to the picture stimuli used to probe verb versus noun naming and comprehension not being equal in executive resource demands. The finding of this study thus underscores the need of considering carefully the specific effects of task and type of stimuli in the patients’ performance with action pictures before making theoretical claims about the noun versus verb or object versus action lexical and semantic representation in the brain. Study 3 We report the results of a single-case study carried out with a brain damaged patient, GC, whose conceptual knowledge of living things (animals and plant) was significantly more impaired than his knowledge of both artifacts and actions, which was similarly impaired. This pattern provides evidence in support of the view that the “objects” vs. “actions” conceptual distinction is not relevant for theories of the neural and functional organization of conceptual knowledge. We then examined whether this pattern of conceptual impairment could be accounted for, first, by the “sensory/functional” and, second, the “manipulability” account for category specific conceptual impairments advocated within the Feature Based Organization theory, by assessing, first, the patient’s knowledge of sensory compared to functional and motor features and, second, his knowledge of non-manipulable compared to manipulable items. The findings showed that the patient’s disproportionate impairment for living things compared to both artifacts and actions was not associated with a disproportionate impairment of sensory compared to functional or motor knowledge or with a relative sparing of manipulable compared to non manipulable items. We propose that the concepts of artifacts and of actions are represented and processed by a domain-specific conceptual system that represents knowledge required for planning our own actions and understanding others’. Study 4 Research on the structure of conceptual knowledge has been essentially devoted to the study of concrete objects. Comparatively, little attention has been given to the empirical investigation of action conceptual structure. The present work addresses the question of the internal structure of action concepts through the case study of GC, a patient presenting with a semantic deficit. The aim of the study was to determine whether the various features of action concepts may be differentially affected by a semantic damage. This issue was examined by analyzing GC’s drawings of actions from name and from photograph. A detailed analysis of the patient’s drawings, as compared to controls’ drawings, showed that GC was able to retrieve only part of the semantic features of action concepts, that is, the features relating to the structure of the event such as the number and type of participants, and the body part involved in the achievement of the action. By contrast, GC could not retrieve the characteristics of the object or the instrument involved in the action, had difficulties in accessing the features reflecting body part configurations and the specific manner in which actions are achieved, and lastly, had difficulties in understanding the result of the action. We proposed that the patient’s performance might be accounted for by a greater vulnerability of action specific features in comparison to general ones. This hypothesis was partly supported by GC’s results in two additional tasks. These findings suggest that the deterioration of action conceptual features following brain damage may parallel the specific versus general breakdown previously shown for object concepts in neuropsychological literatureLe présent travail de recherche a tenté d’apporter des éléments de réponse à deux questions principales : (1) le système de représentation et de traitement des connaissances sur les actions est-il séparé sur les plans fonctionnel et neural de ceux des autres domaines de connaissances et, le cas échéant, quel(s) serai(en)t le(s) principe(s)/le(s) mécanisme(s) sous-jacent(s) à cette organisation ? (2) quelle pourrait être la structure des représentations conceptuelles des actions, c’est-à-dire les propriétés constitutives des concepts d’actions ? Npis avons répondu à ces questions en étudiant quatre cas de patients cérébrolésés (cf. résumé en anglais)(PSY 3) -- UCL, 201