This paper examines the relationship between lexical frequency and phonological processes, focusing on rendaku in Japanese. Recently, the effect of lexical frequency on linguistic processes, either direct or indirect, has been confirmed in a growing body of studies. However, little attention has been paid to the potential effect of lexical frequency on rendaku. With this background, I examined the effect of lexical frequency on the applicability of rendaku, and developed an analogy-based model by incorporating lexical frequency. The results demonstrate (i) that lexical frequency affects the applicability of rendaku, (ii) less frequent compounds follow the existing patterns that the exemplar offers, and (iii) that rendaku is psychologically real; in other words, rendaku is productively applied to innovative forms, and such an application can be accounted for by the current model.
ASJC Scopus subject areas