Determining inhibition: Individual differences in the 'lexicon context' trade-off during lexical ambiguity resolution in working memory.

The current experiment assessed the relation between inhibitory mechanisms underlying language processing in working memory (WM) and the influence of a postulated internal cognitive trade-off between lexicon and context factors driving inhibition in sentence processing. Participants with high and low WM span (Daneman & Carpenter, 1980) were presented with sentences that began with an ambiguous noun (being both homographic and homophonic). A cue towards the relevant meaning of the ambiguous word was presented three words later by another noun. Immediately after the cue, a verb was presented that finally disambiguated the ambiguous noun. In half of the cases, the final disambiguation was congruent with the disambiguation cue. In the other half, subjects had to realize that it was not the cued but the other meaning, which turned out to be relevant in the actual sentence. This experimental paradigm was identical to the one used in Gunter et al. (2003). In contrast to the original experiment, the filler items in the present study were unambiguous. Both groups showed a smaller N400 component for the cue related to the dominant as compared to the subordinate meaning, suggesting that the dominant meaning was more active in WM than the subordinate one. In the original experiment, only the high-span subjects manifested this effect. Thus, a change in the global experimental setting changed the use of WM-related inhibitory processes. Data also suggest that, as a default strategy, the lexicon-context trade-off of high-span subjects relies more on lexicon information than does that of low-span subjects.