Comprehension and grammaticality judgement in Persian-speaking agrammatics.

This thesis examined the performance patterns of two Persian-speaking agrammatic patients and ten normal, Persian-speaking controls in two experiments, one syntactic and the other morphological. The goal of the Syntactic Experiment was to assess the patients I sensitivity to various simple and complicated sentence structures, as well as to agreement and word order violations in both Syntactic Comprehension and Syntactic Grammaticality Judgement Tasks. The Morphological Experiment focussed on the agrammatics' ability to process a certain set of morphological items, including clitic pronouns and null subject pronoun across three sub-experiments: a Morphological Comprehension Task, a Morphological Grammaticality Judgement Task and a Morphological Cloze Test. The design of the present study allowed us to investigate the extent to which aspects of morphology and syntax are differentially disrupted, and to see if agrammatics exhibit a predicted patterns of syntactic and morphological breakdown across language modalities. Moreover, it enabled us to examine the effects of language typology on the manifestation of agrammatism, as well as to obtain evidence that could distinguish between the adequacy of the two most divergent accounts of agrammatism, namely, representational and processing accounts. The combined results of the Syntactic and Morphological Experiments Provide strong evidence that our agrammatic subjects, despite presenting with impoverished output, retained a remarkably well-preserved tacit syntactic and morphological knowledge of Persian grammar. The patients' performance was inferior to that of the normal subjects. However, their high level of comprehension and grammaticality judgement performance and their remarkable sensitivity to grammatical violations indicate that they were able to recognize a wide variety of syntactic and morphological cues and to construct phrasal constituents which observe appropriate grammatical constraints in the various experimental tasks. The findings of the present study suggest that none of the representational accounts of agrammatism provides an adequate characterization of the performance patterns of our two agrammatic subjects. Instead, it is proposed that their agrammatic deficit may be attributed to a reduction or disruption in the efficiency of the language processing mechanism. We suggest that both complexity of the linguistic materials and slowness of the language processing mechanism have converged to diminish or reduce the linguistic performance of the two agrammatic patients