INVESTIGATING THE NEXT-TEXT SPLIT IN SINGAPORE ENGLISH

Abstract

Linguists have observed that the DRESS lexical class, as identified by Wells (1982), is not being realized with a consistent vowel in Singapore English (SgE); specifically, some DRESS words appear to be raised towards the FACE vowel. However, little is known about the distribution and the possible reasons for the phonological variation in the SgE DRESS vowel. The present study investigates the phonological and social conditioning of variation in this feature, termed the NEXT-TEXT split by Deterding (2007). Through an online survey and a production study, this paper studies the distribution of raising in the DRESS class, the effects of age and education on DRESS raising, the effect of the SgE TRAP-DRESS merger on DRESS raising and finally whether the raised DRESS vowel is being merged with FACE. The findings show that the distribution of the DRESS raising is lexically selective and partially phonologically conditioned. Age and education level were found to have a significant effect on DRESS class raising. While younger speakers are found to raise DRESS words significantly more frequently, older speakers use a significantly higher vowel in raised DRESS words. More highly educated SgE speakers are found to raise DRESS words to a lesser degree than less-educated speakers. A speaker’s degree of TRAP-DRESS merger correlates positively with their rate of DRESS raising, but not with degree of raising, in terms of height or frontness of the vowel. Finally, statistical analysis reveals that the raised DRESS vowel is not merged with the FACE vowel, but rather is distinct in height, frontness, and duration. Overall, this study has found that the NEXT-TEXT split is expanding among young SgE speakers, thus widening the distance between SgE and the phonology of British English. This phenomenon is consistent with previous claims that SgE is in a phase of endonormative stabilisation (Schneider 2007), in which speakers are increasingly adopting local norms.