Pourquie, M.
Laboratoire Jacques Lordat - Octogone EA 4156 – Université Toulouse II-Le Mirail. France
We present and interpret some results obtained from a cross-linguistic case study of agrammatism in Basque and French. From such database, we assess the validity of current hypotheses developed in the context of aphasiology on the nature of agrammatism and verb processing in particular. Like Thompson (2003) in the “Argument Structure Complexity Hypothesis”, we document in agrammatic verb processing, the existence of an influence of argument structure complexity according to the number of arguments of the verb (Basque data are particularly clear on this point). However, we do not observe an increasing specific difficulty in lexical verb production but rather in morphosyntactic implementation only. Functional parts (or projections) of the verb appear specifically impaired and inflectional errors seem to increase according to the verb argument structure complexity.
What kind of underlying deficit might be put forward: a structural and/or a procedural one? A deficit in verb class access or a deficit in verb processing? In order to answer this question, we claim that whether we consider that words are inflected where they are stored or not is crucial to locating and determining the nature of the underlying problem. We treat this question from a neuropsycholinguistic (interdisciplinary) point of view, that is, by confronting studies at the crossroads of linguistics and psycholinguistics, within the context of a hopefully coherent model of language structure and processing. In summary, the structural and functional architecture of language is essential in order to determine the nature of language breakdowns; likewise, the study of aphasia and of agrammatism, in this particular case, is liable to provide “external evidence” for theoretical models of language.