The Effects of Semantic Context on Word Retrieval Performance in Aphasia

Abstract

One method of studying normal word production is to manipulate the context in which production occurs. For example, in the semantic priming task, a related prime word presented prior to the production trial is generally found to facilitate naming. In contrast other tasks may elicit interference effects. In Humphreys et al.'s (1995) 'post-cue' procedure, where two pictures are displayed, and one is later cued to be named, naming latencies are actually slowed if the pictures are related. Since these tasks appear to tap into different aspects of the word production process, they may prove useful for studying aphasic word production. In this study, seven aphasics (three anomic, two Broca's one transcortical motor, one borderline conduction/Wernicke's), were compared with age-matched controls on two tasks, one designed to elicit semantic facilitation, and the other interference. In Experiment 1, participants completed a priming task in which they had to say aloud a written/spoken prime word, then name a picture. For the controls, naming latencies were significantly shorter following semantic primes. However, of the seven aphasics studied, only three showed significant effects: two demonstrated facilitation, and the other interference. In Experiment 2, the same participants were tested on a 'post-cue' type procedure, in which they saw two pictures and were then cued to produce both their names in a specified order. In contrast to the semantic priming study, no effects of semantic relatedness were found for controls. Of the seven aphasics, only one (the same individual who had displayed semantic interference in Experiment 1), was affected by the experimental manipulation, demonstrating an initial semantic facilitation effect which slowly transformed across sessions to semantic interference. We discuss implications of these results for models of word production in general, and theories of aphasic word production impairments in particular.