Role of Antonymy Relations in Semantic Judgments

Decision times for sentences or word pairs involving direct (e.g., boy-girl) or indirect (e.g., boy-sister) antonyms were measured in a sentence verification task and a lexical decision task. In Experiment 1 false sentences involving the direct antonyms were disconfirmed faster than sentences involving the indirect antonyms, even though the former were rated as more closely related in meaning. In Experiment 2 a smaller advantage for the direct antonyms was found in a lexical decision task, although "word" decisions were made more quickly for both types of antonym pairs than for pairs of unrelated words. Experiment 2 also investigated the degree of semantic facilitation obtained for category-instance word pairs. No significant latency differences were found for lexical decisions involving word pairs consisting of a category name (e.g., bird) and either a high frequency (e.g., robin) or low frequency (e.g., chicken) instance. The locus of the positive effect of semantic relatedness on falsification time for sentences involving antonyms is discussed in light of these results, as is the locus of the "typicality effect" obtained in previous semantic memory studies.