This study is a quantitative analysis of prosodic features of Māori from three groups of male speakers from different generations. It has been argued that under the influence of English the prosody of Māori has undergone drastic changes over the last century which in earlier studies have been studied impressionistically and also perceptually. In the current study we first determined the most frequent syllabic structures of words of Māori, extracted phrases that embed frequent words from the MAONZE corpus and then applied the quantitative Fujisaki model to the decomposition of F0 contours. This allows a comparison of the three sub-corpora on the global level, but also with respect to the F0 contours of individual target words. Our findings indicate a significant difference between young speakers on one side and present-day and historic elders on the other. Older speaker seem to form larger intonational units than the younger ones in terms of the duration of accent commands. Their F0 range is also significantly larger. In contrast, syllabic durations are quite similar. Altogether our results show, that intonational gestures are decoupled from lexical word stress and rather serve a segmenting purpose on the phrase level. This is supported by our analysis of individual words whose F0 contours are usually flat and only affected by phrase stress.
Cite as: Mixdorff, H., Watson, C., Keegan, P. (2018) Quantitative Analysis of Māori Prosody from Three Generations. Proc. Speech Prosody 2018, 631-635, doi: 10.21437/SpeechProsody.2018-128
@inproceedings{mixdorff18_speechprosody,
author={Hansjörg Mixdorff and Catherine Watson and Peter Keegan},
title={{Quantitative Analysis of Māori Prosody from Three Generations}},
year=2018,
booktitle={Proc. Speech Prosody 2018},
pages={631--635},
doi={10.21437/SpeechProsody.2018-128}
}