Structural Priming in Agrammatic Aphasia

Healthy speakers show structural priming, i.e., reuse of sentence structures previously produced. Agrammatic speakers also show this pattern, particularly for structures that are difficult for them. Research has found that healthy speakers show priming for sentence structures as well as thematic roles. Further, priming effects in healthy speakers appear to be maintained for up to ten intervening sentences between prime and target trials. Two different mechanisms of priming have been proposed. In the activation account, priming results from activation of syntactic, but not thematic, structures from previous production, hence transient priming effects. In the learning account, however, priming arises from learning of the interactive relationship between syntactic and thematic information, which eventually results in changes in the language production system, hence relatively long-lasting priming effects. However, it is not clear what mechanism of priming is used in agrammatic speakers.

Experiment 1 investigated the unit (syntactic/thematic) of structural priming using a modified priming paradigm (after Bock, 1986). Syntactic priming was examined by comparing production of dative sentences (either double object (DO) or prepositional object (PO)) following DO and PO prime sentences. To examine thematic role priming, production of dative sentences following provide-with primes were compared to dative primes. Thirteen healthy controls and 13 agrammatic participants repeated/read prime sentences and created target sentences using written words provided. Results indicated that both groups showed syntactic as well as thematic role (independent of structural dissimilarity and animacy cues) priming. Experiment 2 investigated the duration of priming by comparing production of dative sentences following DO and PO primes, with two (lag 2) or four (lag 4) intervening sentences between prime and target trials. Results showed that both groups exhibited priming effects at both lags, with the magnitude of priming not significantly different between the two lags. Importantly, in both Experiments 1 and 2, the magnitude of priming effects did not significantly differ between the two groups, and greater language impairments in agrammatic speakers resulted in greater priming effects. These findings suggest that priming results from implicit learning of abstract linguistic representations, and therefore, it would be beneficial to incorporate structural priming into aphasia treatment.