Structural differences between common and rare words: Failure of equivalence assumptions for theories of word recognition

Abstract

Theoretical discussions of word frequency effects often assume words of different usage frequencies to be equivalent in other respects relevant to perception. Such assumptions were found to be unwarranted. In Study I, common as compared to rare words were found to be confusable with a greater number of other words by a substitution of a single letter. Moreover, the average usage frequency of such “neighbors” was higher for common words. In Study II, common and rare words were found to contain different distributions of phonemes and graphemes. In Study III, one-syllable words containing phonemes typical of common as compared to rare words were found to be more intelligible, ceteris paribus. The relation of these findings to theories of word-frequency effects is discussed.