Syntactic abilities in Malay adult speakers with aphasia: a study on passive sentences and argument structures

Abstract

ABSTRACT Background Within the last four decades, individuals with aphasia have been observed to face difficulties in understanding complex sentences despite having good understanding of single words. The difficulties observed have been proposed to follow a pattern predicted by the theory of grammar, making the deficit a special case of an underspecified language competence. Aims The purpose of this study is to examine sentence types – reversible actives and passives and argument complexity among Malay speakers with aphasia, looking at sources of difficulties with non-canonical sentences in a language with free word order. Methods & Procedures A group of five non-brain-damaged (NBD) adults and five adults with aphasia matched with age, education, and language were recruited. Subjects were tested in their sentence comprehension abilities via two picture matching tasks: a reversibility active/passive and an argument complexity task based on the number of arguments. Outcomes & Results Generally, the control adults had better sentence comprehension than the group of adults with aphasia, with significant differences in the mean scores of both active and passive reversible sentences. The analysis of the argument complexity task showed that the comprehension level in adults with aphasia was preserved although lower compared to their normal counterparts. Finally, an error analysis of the responses showed that individuals with aphasia had different sources of difficulties: with thematic roles in passive sentences based on syntactic similarities and with active sentences based on semantic similarities. Conclusions Individuals with aphasia, speakers of a free word order language demonstrated lower understanding of both reversible active and passive sentences with no apparent signs of canonicity patterns. A more detailed error analysis disentangled the source of difficulties with reversible errors for passives based on grammatical order and semantic-based errors for agent-driven sentences based on semantic preferences.