Competition and inhibition in lexical retrieval: Are common mechanisms used in language and memory tasks?

Abstract

The following series of experiments examined whether common mechanisms are involved in word retrieval within language and memory domains. Four patients with short-term memory (STM) deficits were examined; however, two of the patients showed a consistent impairment in inhibiting irrelevant verbal information as well. To the extent that repeatedly retrieving verbal items from the same category would require the capacity to suppress competing items to select the target, we investigated whether patients with a verbal inhibitory deficit, in addition to a reduced STM capacity, would be impaired in retrieving items in a semantic context relative to STM patients who do not display a similar verbal inhibition deficit and normal control subjects. Experiments 1- 4 consisted of language tasks which required the repeated naming or matching of items in a semantic or unrelated context. The findings revealed that verbal inhibition patients showed the greatest degree of difficulty during picture naming relative to the matching tasks in a semantic context, suggesting that they are susceptible to interference from semantic competitors to a greater extent at a lexical level. Experiments 5 - 7 consisted of recall and recognition memory tasks with items in a semantically related or unrelated context. Experiments 5 and 6 showed that while STM patients and controls displayed a similar degree of interference for items in a semantic context, STM patients can recall and recognize items near or within the range of controls when demands on STM capacity are minimal during encoding. However, Experiment 7 showed that recall can decline for patients with STM deficits when items are processed more rapidly during encoding. The results from Experiments 5 - 7 suggest that interference among items from the same category can occur in memory tasks (at a conceptual level), but verbal inhibition patients are not affected to a greater degree than control subjects. The overall findings are interpreted within the framework of spreading activation models, and provide implications for potential differences in competition and selection demands at lexical and conceptual levels of representation.