How do speakers with and without aphasia use syntax and semantics across two discourse genres?

Abstract

ABSTRACTBackground: Discourse is an increasing focus of assessment in clinical and research settings because it reflects everyday communication. Everyday communication is likely to include a range of different discourse genres, e.g., describing a scene, or reflecting on life experiences. It is likely that speakers use verbs differently in these different discourse genres, but very little is known about this.Aims: To explore whether there were differences in how two groups of speakers (with and without a communication impairment) used verbs in two different discourse genres, in terms of syntax and semantics.Methods & Procedures: Data from people with aphasia (PWA) were taken from an earlier study, and neurologically healthy people were recruited for the current study. Participants produced discourses from two genres: a picture description (the Western Aphasia Battery “Picnic Scene”) and personal narrative (reflective responses to quality of life [QoL] questions). Discourses were analysed using measures of argument structure (mean predicate argument structure score), verb weight (% heavy verbs), and verb semantic category (% mental and relational verbs). Comparisons were made for each measure between genre and group using a series of two mixed two-way ANOVAs.Outcomes & Results: Data from 26 PWA and 27 neurologically healthy people were analysed. For predicate argument structure, there was a main effect of genre, significant interaction between group and genre, and main effect of group. For the semantic measures, there was a main effect of genre for % mental verbs but no effects or interactions for % heavy and % relational verbs. Post hoc correlations explored associations between the variables.Conclusions: Genre exerts no demonstrable impact on semantic weight, in either speaker group, but does exert influence on the semantic category of verbs used because, for both speaker groups, the picture description genre elicited a smaller percentage of mental verbs than the personal narratives produced in response to QoL questions. For PWA only, genre also exerted an influence on argument structure, with QoL narratives eliciting significantly less complex argument structures. This has implications for clinical assessment. Discourses of different genres should be sampled to fully assess a speaker’s syntactic and lexico-semantic skills; and the genre of discourse used for assessment and therapy materials should align with the client’s communication goals.