TL;DR
This paper investigates the subtle timing differences in how nasal vowels are produced in Northern and Southern French, revealing that these variations can be explained by the precise coordination of articulatory gestures rather than fundamental differences in phonological representation. The findings suggest that Southern French nasal vowels are not inherently more complex but rather involve a specific temporal desynchronization of velum and oral gestures.
What Problem Does This Paper Address?
The production of nasal vowels in French has long been a topic of linguistic debate, particularly concerning Southern French varieties. While nasal vowels are thought to have evolved from sequences of an oral vowel followed by a nasal consonant (e.g., /VN/), their precise phonetic realization and underlying phonological representation in Southern French remain contentious. Some linguists propose that Southern French nasal vowels are realized as a three-part sequence: an oral vowel, a nasalized vowel, and a nasal "appendix" ([VN]), suggesting a more complex underlying structure (/VN/) compared to Northern French. This paper aims to clarify this debate by providing a detailed, comparative gestural analysis of nasal vowel production across different French regional varieties, specifically Northern French (Liège, Tournai) and Southern French (Marseille, Toulouse). The core problem is to determine whether the observed acoustic differences in Southern French nasal vowels reflect a more complex phonological representation or a nuanced temporal coordination of articulatory gestures.
Methods
The researchers employed an experimental phonology approach, using acoustic data to infer articulatory gestures. They recorded 48 speakers from four distinct regional varieties (Liège, Tournai, Marseille, Toulouse) reading a carefully constructed corpus of French bi-syllabic words. This corpus included minimal pairs designed to contrast oral and nasal vowels in various phonetic contexts (e.g., C₁V.C₂V vs. C₁.NV).
A key methodological innovation was the manual segmentation of acoustic signals using Praat to label phonetic segments and "pseudo-segments" corresponding to the hypothesized three parts of Southern French nasal vowels. This segmentation focused on identifying the onset of vowel nasalization and the presence of the nasal appendix.
The core analysis involved measuring the absolute and relative durations of these segments. Statistical comparisons, including MANOVAs and t-tests, were used to assess significant differences in these durations across regional varieties and between different phonetic contexts. These quantitative measures were then used to test two competing hypotheses about the phonological representation of Southern French nasal vowels.
Datasets
The study utilized a newly created corpus of French bi-syllabic (pseudo-)word triplets. This corpus comprised 12 triplets, each repeated three times by 48 speakers, resulting in a total of 5184 words. The design specifically aimed to create minimal pairs contrasting oral and nasal vowels (e.g., /ɛ~ẽ/, /a~ã/, /ɔ~ɔ/) and to control for homorganicity between nasal appendices and nasal consonants. The corpus was adapted to each regiolect to ensure consistent phonetic realization.
Key Insights and Ideas
- Gestural Timing is Key: The study strongly suggests that the perceived differences in Southern French nasal vowels are primarily due to the precise timing of articulatory gestures, particularly the velum opening gesture, rather than a more complex underlying phonological structure.
- Southern French Nasal Vowels are Not Necessarily "More Complex": The data challenges the hypothesis that Southern French nasal vowels have a fundamentally different, more complex phonological representation (/VN/) compared to Northern French (/Ṽ/). Instead, the observed acoustic features can be explained by a temporal desynchronization.
- Unified Account Possible: The findings support a unified account of nasal vowel production across these French varieties by focusing on the interplay between velum, glottal, and oral gestures.
- Velum Gesture Consistency: The duration of the velum opening gesture appears to be relatively consistent across Northern and Southern French, contradicting the idea that a longer velum gesture would be needed to produce a distinct nasal segment in Southern French.
- Nasal Appendix vs. Nasal Consonant: In Southern French, the "nasal appendix" observed in C₁V.C₂V items is significantly shorter than a plain nasal consonant in C₁V.NV items, further supporting the idea that it's not a fully realized nasal consonant.
- Shortened Following Consonant: The consonant following a nasal vowel (C₂) in Southern French is shorter than when it follows an oral vowel, indicating a compensatory shortening rather than the maintenance of a full /VN/ sequence.
Key Results
- Regional Variation is Significant: Statistical analyses confirmed significant variations in the durations of nasal vowel segments (vowel nasalization, nasalized portion, nasal appendix) across the four regional varieties studied.
- Velum Opening Gesture Duration: The duration of the velum opening gesture was not found to be significantly longer in Southern French compared to Northern French, whether measured in absolute terms or relative to the nasal vowel duration.
- Nasal Appendix Shorter than Plain Nasal: In Southern French, the nasal appendix in C₁V.C₂V items was significantly shorter than the plain nasal consonant in C₁V.NV items (e.g., Marseille: t(503)=-5.828, p<.0001; Toulouse: t(359)=-5.759, p<.0001).
- Following Consonant Shortened: The duration of the consonant C₂ was significantly shorter when following a nasal vowel in Southern French compared to when following an oral vowel (e.g., Marseille: t(503)=49.463, p<.0001; Toulouse: t(359)=48.356, p<.0001).
Limitations and Future Work
- Inferred Gestures: The study infers articulatory gestures from acoustic data, which is an indirect method. Direct articulatory measurements (e.g., EMA, ultrasound) could provide more definitive evidence.
- Limited Corpus: While extensive, the corpus focused on specific phonetic contexts (C₁V.C₂V, C₁.NV) and vowel qualities. Further research could explore a wider range of phonetic environments and vowel types.
- Future Work: The authors suggest that future research could explore the aerodynamic and articulatory correlates of these timing differences more directly. They also imply that this gestural approach could be extended to other phonological phenomena involving complex timing relationships.
Summary updated