The effects of verb retrieval therapy for people with non-fluent aphasia: evidence from assessment tasks and conversation.

Despite often impressive improvements on linguistic assessments, there is a lack of evidence of significant generalisation from impairment-focused aphasia therapy to everyday communication. The aim of the current study was to investigate the effect of a verb retrieval therapy across a range of levels of language production. Nine participants with chronic non-fluent stroke aphasia were recruited into this case series. Baseline assessment included naming a range of verbs (i.e., action verbs, semantically light verbs and personally relevant verbs) and sentence production. Multiple samples of conversation were collected from each participant and his/her partner. Consecutively failed verbs were divided across treatment and control sets; these sets were matched for salient psycholinguistic variables such as frequency, imageability and argument structure. A multi-component verb retrieval therapy was delivered, consisting of semantic feature analysis, gesture production and phonemic cueing. Following therapy, participants demonstrated significant and sustained gains in naming treated verbs; more modest effects were seen in untreated verbs. Mixed patterns of generalisation were evident in assessment of sentence production. In conversation, while group analysis suggested a lack of change, individual analyses indicated increased verb retrieval for three participants and qualitative changes related to the syntactic contexts of verbs retrieved.