Fine-grained control of voice-onset time production, lexical usage-frequency, and phrasal context.

Abstract

Voice‐onset time (VOT) in English maps onto the phonological categories voiced or voiceless, but fine‐grained control of VOT production may vary as a function of additional linguistic features (e.g., lexical frequency, phrasal context, indexical features). This study investigated the patterns of VOT production in words that varied (a) by lexical usage‐frequency (high versus low) and (b) in selected linguistic contexts (isolation, carrier phrase, unfamiliar phrase, familiar phrase). Results show that talkers produced longer VOT for voiceless stops in low‐frequency words in unfamiliar and familiar phrases but not in isolation or carrier sentences. Neither lexical frequency nor phrasal context seemed to induce changes in VOT in voiced stops. Thus, it appears that talkers make use of fine temporal distinctions within the voiceless category but not the voiced. This suggests a complex asymmetry in how continuously varying VOT maps onto more discretely varying phonological categories.