No CrossRef data available.
Published online by Cambridge University Press: 28 April 2020
Miri, an Austronesian language spoken in northern Sarawak, Malaysia, has two sets of vowel changes that are conditioned by voiced obstruents. In the first set, a last-syllable low vowel is fronted and raised to [e], or less commonly [i], if a voiced obstruent appears earlier in the word, while a penultimate low vowel immediately following the trigger is skipped. In the second, a high vowel in the final syllable undergoes breaking (diphthongisation) or lowering, depending upon specific conditions, unless there is a voiced obstruent anywhere earlier in the word. For both triggers and suppressors, this effect is cancelled by an intervening blocking consonant, which includes any nasal or voiceless obstruent except glottal stop. The challenge is to understand why voiced obstruents have this double function, acting as a trigger with low vowels and a suppressor with high vowels, given the lack of an a priori transparent relationship between low vowel fronting and high vowel breaking/lowering.
Thanks to two anonymous referees, the editors of Phonology, Thomas Kettig, and especially Erik Thomas for useful feedback that led to improvements in an earlier version of this paper. The reader should note that IPA symbols are used throughout, rather than those that are commonly used in the practical orthographies for languages of island Southeast Asia. Most importantly, this affects the palatals, where in the practical orthographies j is a voiced palatal affricate, c a voiceless palatal affricate, ñ a palatal nasal and y a palatal glide. Proto-Austronesian *R and its continuation in many later proto-languages was probably an alveolar trill (distinct from *r, which likely was a flap).