Final devoicing in Singapore English

Abstract

This study investigates the voicing contrast in word-final obstruents in Singapore English. Previous auditory accounts have suggested that word-final obstruents undergo complete devoicing in Singapore English and become identical to their voiceless counterparts, including neutralization of differences in adjacent vowel duration (among others, Bao 1998, Wee 2008). On the basis of acoustic data from the NIESCEA corpus (Low 2015), we examine the extent to which voicing-related differences are neutralized when reading wordlists, sentences and passages as well as in unscripted conversations. Results suggest that neutralization is phonetically incomplete in Singapore English, with underlying voicing being recoverable from such parameters as closure voicing ratio and durations of consonants and adjacent vowels. These findings are in line with the results for other devoicing languages, such as German and Russian, that also show incomplete neutralization of the voicing contrast in final obstruents (e.g., Roettger et al. 2014, Kharlamov 2014).