Investigation of the acoustic effects of orthographic consistency on Japanese homophones

Abstract

Research has claimed that the phonetic realization of word duration in homophone pairs, such as time and thyme, systematically differs based on their lexical characteristics, including part of speech, frequency, and orthographic consistency. This research has shown that higher orthographic consistency is associated with shorter homophone duration in English (Gahl, 2008). The present study builds on this line of research in two ways. First, we extend it to Japanese, examining whether orthographic consistency effects on homophone duration generalize beyond alphabetic languages (e.g., English). Unlike alphabetic languages, Japanese employs logographic (or morphographic) orthography, in which logographic characters do not reliably correspond to any sub-lexical phonological unit (Wydell, 1998). Second, we investigate two directions of orthographic consistency: spelling-to-sound (feedforward) and sound-to-spelling (feedback) (Ziegler et al., 1997). While previous research has demonstrated effects of sound-to-spelling consistency on word production (e.g., Ziegler et al., 2003), studies on acoustic duration have focused exclusively on spelling-to-sound consistency (Brewer, 2008; Muschalik and Kunter, 2024), overlooking possible effects of sound-to-spelling consistency on word duration.