Argument structure processing in aphasia: A cross-language and cross-population behavioral study.

• ¹ Université de Montréal / Basque Center on Cognition, Brain and Language, Faculté de médecine, Ecole d'Orthophonie et d'Audiologie, Canada

Verb processing is problematic in many aphasia subtypes, and verb argument structure effects have been observed on both verb lexical retrieval and morphosyntactic processing in different languages. For instance, there is evidence in the aphasia literature that aphasic speakers find verbs with increasing numbers of arguments more difficult to retrieve, and perform poorly compared to controls (Cruice, Pritchard and Dipper, 2014). Recent neuroimaging studies propose that the posterior perisilvian network is responsible for the processing and integration of arguments selected by the verb, while the left inferior frontal gyrus underlies the generation of their syntactic form (Thompson, Bonakpadour & Fix, 2010). In the context of aphasiology, this hypothesis would then predict that Broca’s, but not Wernicke’s aphasic individuals, have spared access to verb complex lexical entries while their syntactic structure implementation will be more difficult. Since verb argument structure is differentially implemented across languages (word-order, case morphology, verb agreement…), it is crucial to develop cross-language and cross-population studies of verb processing if we want to define the nature of "Verb Argument Structure Complexity" in aphasia. A psycholinguistic test was developed to tease apart lexical and inflectional verb processing in both production and comprehension in three languages (Basque, French and Spanish). It includes, for each language, 30 lexical verbs (10 intransitive, 10 transitive and 10 ditransitive), of which 15 (5 intransitive, 5 transitive and 5 ditransitive) have to be encoded and decoded in sentences with different agreement conditions (singular/plural subject; singular/plural direct object; singular/plural indirect object). The same list of verbs was used in 1. action naming, 2. sentence production and 3. sentence comprehension with the aim to determine whether argument structure complexity, defined as the number of arguments associated to a given verb, increases difficulties in lexical access or in morphosyntactic computation. The tool itself will be presented as well as preliminary data collected in French from a case study of agrammatism, and control adults (N=10). I will argue that cross-language behavioral studies are useful to complement neuroimaging data for investigating the nature of argument structure complexity in agrammatic aphasia and other aphasia types.