Production frequency and the verification of quantified statements

Abstract

In Experiment I subjects were required to generate nouns and adjectives which formed true completions of sentences of the form (Quantifier) S are. The quantifiers used were All, Many, Some, Few, and No. Three further experiments examined the effect of production frequency of the predicate on verification RT for quantified statements. True sentences with high-frequency category (Experiment II) or property (Experiments III and IV) predicates were verified more quickly than those with low-frequency predicates when the quantifier was All, Many, Some, or Few; this difference reversed for No-statements. False RT was fastest in all cases when the false sentence was semantically anomalous; but in Experiment IV, when degree of relatedness of subject and predicate words was varied within false but meaningful sentences, statements with high-related predicates were rejected more quickly. In general, the effect of semantic relatedness reversed for negative (Few and No) as opposed to positive quantifiers. An ordered attribute-search model was proposed to account for subjects' performance during both predicate production and sentence verification.