EFFECTS OF CONTEXT-DEPENDENT AND CONTEXT-INDEPENDENT ASSOCIATIVE STRENGTH BETWEEN PRIME AND TARGET WORDS ON THE PROCESSING OF LEXICAL AMBIGUITY

The time course of lexical ambiguity resolution was investigated using the phonological priming paradigm between polysemous phonogramic Kana words (primes) and compound ideogramic Kanji words (targets) in Japanese. Subjects made the lexical decision under rapid serial visual presentation of sentence contexts, primes, and targets. In Experiment 1, the context-dependent associative strength of target words and prime-target stimulus onset asynchronies (SOAs) were manipulated. At 300-ms SOA, results of reaction time showed facilitation effect for both contextually appropriate and inappropriate target words. At 500-ms SOA, while the facilitation of reaction time held for contextually appropriate targets, this effect was not obtained for contextually inappropriate targets. In Experiment 2, the context-independent associative strength was manipulated in addition to the two variables of Experiment 1. Results of reaction time replicated the multiple access at short SOA and the context-dependent selective access at long SOA. The results of error rates indicated a greater context-independent facilitation effect at 300-ms SOA than at 500-ms SOA. These findings provide evidence in support of the transition from multiple access, which is influenced by context-independent associative strength, to context-dependent access in the process of resolving lexical ambiguity.