Understanding semantic and phonological processing deficits in adults with aphasia: effects of category and typicality

Abstract

Background: Semantic and phonological processing deficits are often present in aphasia. The degree of interdependence between the deficits has been widely studied with variable findings. Within semantic processing, category and typicality are proposed to influence accuracy and response time on semantic tasks in both healthy and aphasic subjects. Aims: This study examines the nature of semantic-phonological access in aphasia by comparing adults with aphasia to healthy control subjects. Three semantic tasks and three phonological tasks containing typical and atypical items of six semantic categories were used to assess the difference in category and typicality effects between persons with aphasia and healthy adults. Finally, we aim to identify demographic factors and formal language measures that correlate with semantic and phonological processing performance. Methods: Twenty patients with aphasia and ten neurologically healthy adults were administered six tasks: category superordinate, category coordinate, semantic feature verification, syllable judgment, rhyme judgment, and phoneme verification. Accuracy and reaction time data were collected and analyzed as three conditions: 1) phonological no name, 2) phonological name provided, and 3) semantic. Results: Patients with aphasia performed with significantly lower accuracy than controls, with greater between-group difference on phonological tasks than on semantic tasks. Patients were significantly slower than control on semantic and phonological no name conditions, but showed no difference on the name provided condition. Both patient and control groups showed category effect on semantic accuracy. The only category effect found on RT was controls on the phonological no name condition. Control showed an effect of typicality on the semantic condition for accuracy while patients showed it for RT. Correlations were found between language measures and education and task performance. Conclusions: Patients demonstrated greater phonological than semantic deficits. Both patient and control groups showed effect of category, but patients showed a reduced effect of typicality. Category and typicality effects are robust in semantic tasks, but not in either phonological task conditions, providing support for discrete serial processing models of lexical processing. Education level was found to be a predictor for semantic boundary knowledge, but not for phonological processing skills.