Partial voicing neutralization in unstressed stop-nasal sequences

Abstract

We present the results of a production experiment that explores the realization of voicing in English stop-nasal sequences, a phonological environment with a low functional load. The results imply a separation between primary and secondary voicing cues. Primary cues – aspiration and vocal fold vibration during stop closure – robustly distinguish underlyingly voiced stops from and underlyingly voiceless ones. Meanwhile, secondary cues – vowel duration and stop closure duration – are limited in their use by phonological position or absent entirely. A principal component analysis of the data indicates that all speakers occasionally produce tokens that are ambiguous in voicing cues.