effects of verb-argument cues on verb production in persons with aphasia using a verb-final language

Introduction Verbs are regarded as one of the critical linguistic constituents, given that verbs carry both syntactic and lexical-semantic features in a sentence (Druks, 2002). There have been attempts to determine the nature of verb deficits in persons with aphasia (PWA). Thompson and colleagues proposed the argument structure complexity hypothesis, and they found that PWA showed greater difficulties in verb production as the number of arguments increased (Kim & Thompson, 2000; Thompson, 2003; Thompson et al., 1997). However, it has not been examined whether providing cues on verb arguments facilitate verb production abilities in PWA. The current study investigated the effects of verb-argument cues on verb production in PWA using a verb-final language. Korean is a verb-final language, in which verbs are placed following the verb arguments. Therefore, employing a verb-final language is a good way to test the hypothesis whether the verb-argument cues help PWA to retrieve the verb in a sentence, given that the verb arguments can be provided prior to the verb. It is hypothesized that PWA would benefit from cues especially for verbs with more arguments. Methods Participants: Eleven PWA (mean age=50, SD=13) and 11 normal elderly individuals (NEI) (mean age=61, SD=1) participated in the study. They were all Korean-monolingual speakers. The NEI group showed normal range of performance on the Korean Mini-Mental State Examination (K-MMSE) (Kang, Na, & Hahn, 1997). All individuals with aphasia suffered a single, left hemisphere stroke. The diagnosis of aphasia was based on administration of the Korean version of Western Aphasia Battery (Kim, & Na, 2001). The verb confrontation naming task: It consisted of three verb types with 1-place, 2-place, and 3-place arguments. The task contained a total of 24 stimuli with 8 pictures for each type. The verb completion task: The structure of the task was consistent with the naming task. In the completion task, examiners verbally presented the verb arguments. Participants were required to complete the sentence with the verb. Results A three-way ANOVA (group x argument structure x task-type) was performed. A main effect for task-type was significant (F(1, 20=27.138, p<.001) with better performance on the verb-completion than the naming task. PWA significantly performed worse than NEI (F(1, 20)=29.131, p<.001). The three-way interaction was significant (F(3, 60)=5.036, p<.005). As a post-hoc analysis, two separate two-way ANOVAs were performed for each group. PWA showed a significant two-way interaction (F(2, 20)=4.702, p<.05), indicating that they benefited from argument cues especially in verbs with 2- and 3-place arguments than 1-place verbs. However, these effects were not significant in NEI. Discussion PWA showed greater difficulties in producing verbs with more argument structures compared to the normal group. In contrast, PWA presented significantly increased performance on the verb-completion task as the number of verb arguments increased since more cues were provided for 2- or 3-place verbs than 1-place verbs. The current results suggested that the features of more semantic and syntactic units to be activated may induce greater difficulties in retrieving verbs with more arguments, and PWA benefited from the verb-argument cues in verb production.