Word-initial consonant-vowel coordination in a lexical pitch-accent language.

Malin Svensson Lundmark,Susanne Schötz,Johan Frid,Gilbert Ambrazaitis

doi:10.1515/phon-2021-2014

Malin Svensson Lundmark, Susanne Schötz + Show 2 more

Open Access

https://doi.org/10.1515/phon-2021-2014

Copy DOI

Journal: Phonetica	Publication Date: Nov 17, 2021
Citations: 3	License type: CC BY 4.0

Affiliation: Lund University, Linnaeus University

Abstract

Previous research has acknowledged the effect of prosody on inter-gestural coordination, but specifically the effect of tones is still understudied. This paper has a two-fold purpose. First, it aims to explore effects of the Swedish word accents on word-initial consonant-vowel (CV) coarticulation. Second, it aims to revisit the existing evidence for tonal integration. Based on Articulatory Phonology, it has been suggested that tones- in tone languages- are integrated in the gestural organization of a syllable-initial CV sequence in the same manner as would an additional consonant (CCV), indicated by a time lag between the gestural onsets of the C and the V gesture (CV onset time lag). However, we argue that the existing evidence is inconclusive, because previous cross-linguistic research has used small-scale data sets (one to seven speakers), and we still lack a well-grounded consensus on how gestural onsets are to be measured. This study uses Electromagnetic Articulography (EMA) to investigate word-initial CV coordination in a lexical pitch-accent language (Swedish) with a binary tonal word accent distinction: a tonal fall and a tonal rise, respectively. A selection of 13 spatial, temporal or coordinative measures of bilabial and tongue body data from 19 speakers, and acoustic fo data, were examined to study the CV sequence /ma/. Mixed effects regression models revealed a longer tongue body movement in the rising tone context and small but significant differences in tongue body height, in the closing and the opening of the lips, as well as in the CV onset time lag between the two tonal contexts. We argue that these effects are biomechanical in nature, due to the physiological connections between the tongue, the jaw, and the larynx. In addition, our results suggest either synchronized CV onsets or a CV onset time lag (as in tone languages), depending on the timing landmarks used. In order to evaluate such results as evidence for or against the integration of tone in CV coarticulation, we argue that future research needs to compare data from a variety of languages using a considerable number of speakers. The present study provides new reference values for such comparisons.

Highlights

This study investigates how the articulation of the word-initial consonant–vowel (CV) sequence /ma/ is affected by the tonal context, using Electromagnetic Articulography (EMA)
In order to evaluate such results as evidence for or against the integration of tone in CV coarticulation, we argue that future research needs to compare data from a variety of languages using a considerable number of speakers
The falling and the rising tone of the Swedish word accents are executed mostly during the vowel segment, there is a significant difference between the fo environments of the word accents already when the tongue body is lowering, that is during the span that includes the majority of our articulatory measurements

Summary

Introduction

This study investigates how the articulation of the word-initial consonant–vowel (CV) sequence /ma/ is affected by the tonal context, using Electromagnetic Articulography (EMA). It has been shown that CV sequences exhibit a greater degree of coarticulation, and more phonetic change, than VC sequences (e.g., Fant 1969; Löfqvist and Gracco 1999; Macneilage and Declerk 1969; Recasens 2002; Stevens and Blumstein 1981), and articulatory coordination varies depending on the types of consonants and vowels involved (e.g., Fowler and Saltzman 1993; Kawahara et al 2014; Löfqvist and Gracco 1999; Mooshammer et al 2006; Recasens 2002) Effects of prosody, both concerning its grouping and its prominence functions, have been studied within this line of research (e.g., Byrd et al 2005; Cho 2002; de Jong et al 1993; Lindblom et al 2007; Mooshammer et al 2013). Recent contributions to the development of this field include work on lexical tones in tone languages, suggesting effects of tone on lip and tongue body coordination (Gao 2008; Hu 2016; Karlin and Tilsen 2015; Zhang et al 2019), as well as differences in tongue body height between the tones (Erickson et al 2004; Hoole and Hu 2004; Shaw et al 2016)

Objectives

Methods

Results

Discussion

Conclusion