Intonation adapts to lexical tone: the case of Kammu

In this paper, we investigate how lexical tones interact with intonation, using data from the Austroasiatic language Kammu, one of few languages with two dialects whose only major phonological difference is the presence or absence of lexical tones. Northern (and Western) Kammu have developed tones in connection with the merger of voiceless and voiced initial consonants, while the non-tonal Eastern dialect kept the segmental opposition with no tones. We found the following prosodic hierarchy: (1) lexical tones, (2) phrase-final boundary tone, (3) focus marking. The results strongly suggest that the intonational systems of the two Kammu dialects are basically identical, and that the main differences between the dialects are adaptations of intonation patterns to the lexical tones when the identities of the tones are jeopardized.