Abstract
The present study, which is conducted within the general framework of Generative Grammar (Chomsky, 1981; 1995 and thereafter), investigates the ability of three Greek-speaking agrammatic aphasics to handle (in both production and comprehension) a number of functional categories: Complementizer (C), Negation (Neg), Agreement (Agr), Tense (T), and Aspect (Asp). To test these categories, a series of structured tasks were administered (e.g. question elicitation task, constituent ordering task, truth-value judgement task, sentence completion task, sentence grammaticality judgement task, sentence completion task, sentence grammatically judgement task, sentence repetition task, etc). Apart from the afore-mentioned categories, the effect of some other variables on the agrammatic performance was also explored, such as the frequency of occurrence and the morphological regularity of the verb, as well as the structural complexity of the sentence. One of the principle goals of this study is to test ...
The present study, which is conducted within the general framework of Generative Grammar (Chomsky, 1981; 1995 and thereafter), investigates the ability of three Greek-speaking agrammatic aphasics to handle (in both production and comprehension) a number of functional categories: Complementizer (C), Negation (Neg), Agreement (Agr), Tense (T), and Aspect (Asp). To test these categories, a series of structured tasks were administered (e.g. question elicitation task, constituent ordering task, truth-value judgement task, sentence completion task, sentence grammaticality judgement task, sentence completion task, sentence grammatically judgement task, sentence repetition task, etc). Apart from the afore-mentioned categories, the effect of some other variables on the agrammatic performance was also explored, such as the frequency of occurrence and the morphological regularity of the verb, as well as the structural complexity of the sentence. One of the principle goals of this study is to test some accounts of agrammatism proposed so far, such as the Tree Pruning Hypothesis (TPH) (Friedmann & Grodzinsky, 1997) and the Interpretable Features’ Impairment Hypothesis (IFIH) (Nanousi, Masterson, Druks, & Atkinson, 2006; Varlokosta, Valeonti, Kakavoulia, Lazaridou, Economou, & Protopapas, 2006). According to TPH, agrammatism arises from deletion or ‘pruning’ of the syntactic tree at the T node. TPH predicts that everything above T is lost while material below is preserved. On the other hand, IFIH is not structural/hierarchical in nature, as it predicts that in agrammatism the most severe impairment will involve the functional categories with interpretable features (such as T and Asp) rather than the ones with uninterpretable features (such as Agr). The results of the present study are in line with IFIH, since T, Asp, Neg and C –which all bear interpretable features– were found to be severely impaired, while Agr was found relatively spared. In contrast, given that Asp is lower that T and Agr in the Greek syntactic hierarchy (Philippaki-Warburton, 1998), these results cannot be accountted for by TPH. Apart from reinforcing the descriptive adequacy of IFIH, this study aimed at developing its interpretative enrichment as well. More specifically, on the basis of the quantitative and qualitative aspects of the agrammatic patterns of our participants, it is argued that they have two deficits: a linguistic one, which is mainly syntactic in nature affecting functional categories across the board, and a “psychological/cognitive” one, namely a limitation of processing resources. Therefore, it is suggested that what renders the categories with interpretable features more vulnerable than those with uninterpretable features is the synergistic effect of three facts: i) agrammatic speakers (at least the ones participated in the present study) suffer from a reduction of processing resources, ii) they have a syntactic deficit affecting functional categories, and iii) the categories with interpretable features are more demanding in terms of processing resources, as they require integration of grammatical and conceptual knowledge. In contrast, implementation of only grammatical knowledge is sufficient for “uninterpretable” categories like Agr. Finally, similar but not identical patterns of performance were observed across different modalities (comprehension and production), reinforcing the weak (and not the strong) parallelism between them and, thus, providing empirical support to Grodzinsky’s claim (2000: 18) that there is anatomical proximity, but functional separation between production and comprehension mechanisms.
show more
