QUESTION PRODUCTION IN ITALIAN-SPEAKING PWAGR

Abstract

This paper presents the results of an elicitation task run with Italian-speaking people with agrammatic aphasia (PWAgr). Linguistic theories of aphasia have analysed the production of PWAgr as stemming either from a deterioration of grammatical knowledge or a limitation of extra-linguistic capacities which affect language. Among the former, the Tree Pruning Hypothesis (Friedmann and Grodzinsky, 1997) relates the height of the projections involved in the production of a structure with the chances of it being produced: the stronger the impairment in the speaker, the less likely they are to produce a structure involving higher nodes. Alternatively, syntactic knowledge may be preserved but its use is compromised by Working Memory (WM) limitations (e.g., Miyake et al., 1994; Jakubowicz, 2005). The two approaches make different predictions with respect to question production in Italian: according to the TPH, the production rates of yes/no-questions and who-questions should be comparable in moderate and mild aphasia. Why-questions involve a higher node and should thus not be available. Under a WM-limitation approach, the production rates of why-questions and yes/no-questions should be comparable and higher than those of who-questions. To determine which, if any, of the two approaches makes the correct predictions, three adult Italian speakers diagnosed with Broca’s aphasia and one with anomia performed an elicitation task targeting these three types of structures. Overall, yes/no-questions were the most frequent structures to be produced. The lack of who-questions, as opposed to the production of yes/no-questions, can be accounted for only if derivational complexity is taken into account, but the scarcity of why-questions suggests that the height of the nodes involved may play a role as well.