Functional categories and syntactic operations in (Ab)normal language acquisition.

This article argues in favor of the hypothesis that computational complexity determines order of acquisition of functional categories by normal children and patterns of impairment vs. relative preservation of these categories in children with Specific Language Impairment. Complexity is defined in terms of the properties of functional categories of the language the (ab)normal child is acquiring. The results of a study on present/past tense production and comprehension, carried out with children with and without Specific Language Impairment confirm the predictions of the computational complexity hypothesis. It is shown that it is easier to compute kernel functional categories than those that introduce further semantic modifications in the sentence.

Processing of number and gender inflections by French-speaking aphasics.

This study investigates French-speaking aphasics' sensitivity to gender (of human nouns) and number marking in a sentential context. Using a forced picture choice task, we tested sentences in which grammatical marking surfaced either on a function word or on a content word, within or outside a noun phrase (NP) whose gender or number was required to be identified. Ten fluent and 10 nonfluent aphasics together with 20 adults without neurological history were tested. Results showed that neither group of aphasics presented an across-the-board deficit. Nonfluent patients were strongly impaired when marking surfaced on a content word outside NP (verbs and adjectives), but their sensitivity to grammatical marking was relatively well preserved in the within NP condition (determiners and nouns) and for function words in the outside NP condition (the copula). Fluent patients showed a specific impairment in dealing with semantic gender (as opposed to number information); and their difficulty was exacerbated when the information conveyed by the suffix of a content word must be integrated into an higher order semantic representation, as in the outside NP condition (adjectives). These results are consistent with the view that for both nonfluent and fluent aphasics, the functional locus of their impairment lies on a reduction in the computational resources available to the language processor, which is more severe in the former than in the latter group.