Contact des langues et diffusion aréale des langues sinitiques dans la région de la Grande Baie – CLDA

The project’s initial work organization is structured around two principal work packages (WPs) and four core tasks, with a clearly defined timeline, division of responsibilities, and methodological framework.

WP1 (Data Collection) focuses on fieldwork for three under-documented Sinitic varieties – Tanka (Hong Kong), Southern Pinghua (Guangxi), and Cantonese (Hong Kong and Guangxi). Fieldwork is conducted in three phases using structured questionnaires, in-depth interviews, and high-quality audio recordings. Informants are selected according to strict sociolinguistic criteria (age, gender, locality, and language background) to ensure representativeness. The data collection encompasses phonology (core wordlists and extensive lexical items), grammar (syntactic sentences), and naturalistic speech (oral histories, conversations, and narratives), all transcribed in IPA with Chinese character glosses. In addition, substantial ethnographic documentation, including intangible heritage (Saltwater songs), photographs, videos, and unpublished manuscripts, is gathered to ground linguistic analysis in community practices.

WP2 (Data Analysis) addresses four research questions: contact between Cantonese and Tanka; contact between Cantonese and Pinghua; typological features of Sinitic languages in the Greater Bay Area (GBA); and the linguistic status of GBA languages within Mainland Southeast Asia. The analytical approach combines synchronic and diachronic perspectives, employing grammaticalization studies, comparisons with historical texts, and a language-gene coevolution framework for Pinghua.

Human and material resources are allocated as follows: two Principal Investigators (Dr. Chong Qi, CRLAO, France; Dr. Cong Wang, EdUHK, Hong Kong), 7 co-investigators, 1 postdoctoral fellow, 5 research assistants, 6 doctoral candidates, and 88 undergraduate/master’s student assistants. Equipment includes three computers and six sets of recording devices, supplemented by institutional support from both partner laboratories. Regular coordination is ensured through weekly online meetings (two per week, with additional sessions as needed), annual joint field missions, and active participation in conferences and lectures.

Project activities have included a kick-off meeting with 30 international scholars, a public lecture series (“Sinitic Languages in Southern China: Contact, Evolution, and Protection”) comprising eight talks across Hong Kong, Macau, Mainland China, and France, as well as two invited lecture series featuring 15 distinguished experts from multiple countries and disciplines (molecular anthropology, anthropology, history, and AI). We have also hosted five workshops on language contact and typology and, through the CLDA platform, organized two international emerging-scholar forums and six doctoral workshops to foster early-career researchers. The project will conclude with a final international conference in Paris, with 25 invited scholars.

Our research addresses four tasks. For Task 1, a systematic review and empirical study show that HK Tanka has converged with Cantonese under sociolinguistic pressure but retains its own innovations in phonology, maritime lexicon, and grammar; rhymes, tones, and SFPs differ significantly, though youth show growing overlap. Lexical convergence with Cantonese and divergence from Mandarin indicate contact not ancestry; hence Tanka is not a dialectal extension but genetically separate, with similarities from convergence (Dai & Wang, 2026; Wang, 2024; Wang et al., 2025).

For Task 2, three main results emerge: classifier systems of Southern Pinghua and Guangxi Cantonese have unusual features (generic CL-N, definite lone classifiers, etc.) absent outside Guangxi but present in Zhuang, confirming areal convergence via language shift, with dual-layer substratum and overlay (Qi & Peng, 2026a); disposal constructions in Baihe Pinghua show VCO/VOC from Middle Chinese but Zhuang contact led to convergence with VOC, non-systematic and without metatypy (Qi & Peng, 2026b); and Baihe Pinghua has a rare intransitive passive with marker zhe indicating external force, offering a new typological perspective (Peng et al., 2026). These findings confirm systematic contact effects and provide a foundation for integrating biological data.

For Task 3, analysis of 40 lexical items across 161 varieties plus phonological, grammatical, and toponymic data shows: Tanka-Cantonese correspondences due to contact (Wang, Dai & Ran, 2026); Pinghua bifurcation (Southern aligns with Cantonese, Northern with Xiang/Mandarin admixture) (Wang & Dai, 2026c); HK Tanka diverges phonologically but is mutually intelligible, with distinct SFPs yet youth convergence (Wang, 2025); toponymic 排 maps to Zhuang substrate (Wang & Dai, 2026b); and Baihe disposal variation is non-systematic (Qi & Peng, 2026a). Overall, both show Zhuang substrate and Cantonese adstrate, with areal diffusion shaping typology without overriding historical layers.

For Task 4, Tanka mediates between GBA and MSEA, with geographical dispersal from Fujian to Southeast Asia and bidirectional diffusion, as evidenced by cultural vocabulary such as Psidium guajava (Wang & Dai, 2026a; Wang, Dai & Peng, 2026; Wang, 2027).

The CLDA project has laid a strong foundation for our future research agenda. Over the past three years, we have compiled a rich, systematically structured linguistic dataset spanning a wide range of Southeast Chinese varieties, with particular emphasis on the under-documented Tanka and Pinghua dialects. The resulting corpus, which covers phonological, lexical, and syntactic information, has become one of the most comprehensive empirical resources available for the study of coastal and insular Sinitic languages.

Building on this achievement, we plan to submit a new joint application to the ANR-RGC this year. Our new proposal will pursue three interconnected research directions designed to (i) refine historical classification, (ii) test areal-typological hypotheses for Southeast China with the Greater Bay Area (GBA) at the center, and (iii) explain how maritime mobility may have driven diffusion and contact-induced change.

1) Reconstructing relationships among coastal varieties Our preliminary findings indicate that the genetic and areal relationships among Tanka, Danzhouhua, Southern Pinghua, Weitouhua, and the Yue (Cantonese) dialect group are more complex than traditional classifications suggest. Rather than behaving as peripheral varieties, these systems appear to reflect layered histories of both shared innovations and shared retentions that cut across established subgroup boundaries. In the next phase, we will conduct a fine-grained comparative analysis using both classical historical-comparative methods and computational phylogenetics. We will aim to clarify the position of each variety within the broader Sinitic family while identifying the relative contributions of:

substrate effects (local non-Mandarin or non-Han linguistic influences),

superstrate influences (later/overlying standard or dominant varieties),

chronological stratification of sound changes.

The goal is to propose a more nuanced model of dialect divergence and convergence in the coastal zone.

2) Testing an areal typology for Southeast China (with the GBA as core). While the first strand clarifies historical relationships, this second strand moves toward a holistic areal-typological framework. The GBA is not only an economic hub but also a long-standing contact zone where multiple Sinitic and non-Sinitic languages have coexisted, competed, and mutually transformed one another over centuries. By integrating our new CLDA data with existing descriptions of Min, Hakka, She, and Zhuang varieties, we will investigate whether the GBA and adjacent regions constitute a distinct linguistic area (Sprachbund). Specifically, we will test whether certain typological patterns—such as tone-sandhi profiles, classifier systems, or serial-verb constructions—are better explained by shared areal pressure and contact dynamics than by inheritance alone. This line of inquiry addresses a broader theoretical question: how language contact and areal pressure shape structural convergence.

La région de la Grande Baie (GBA) en Chine, qui abrite quelque 80 millions de locuteurs de plusieurs langues sinitiques, telles que le cantonais, le hakka, le min, le pinghua, le mandarin du sud-ouest, le tankas, le mai, le danzhou et les langues minoritaires zhuang, dong , hlai et hmong-mien, ainsi que diverses autres langues, est l'une des régions de Chine les plus riches en langues. Parmi ces langues sinitiques, le cantonais a longtemps servi de langue locale principale dont la population native reste la plus importante dans la GBA. Il existe néanmoins des contacts de longue date entre le cantonais et d'autres langues sinitiques. Afin de clarifier la relation entre le dialecte cantonais et les autres langues sinitiques et d’observer les caractéristiques typologiques des langues dans la GBA, ce projet vise à examiner le contact des langues et l’évolution des langues en danger à travers une étude typologique à grande échelle sur des langues dans la GBA tels le cantonais, le tankas et le pinghua du Sud.
Ce projet adopter une méthodologie de recherche mixte, incluant une vaste étude sur le terrain, une approche d’analyse à la fois diachronique et synchronique, ainsi que la théorie de la coévolution linguistique et génétique. Trois aspects interdépendants ont été considérés lors de la planification du projet. Le premier consiste à mener trois phases de terrain sur les langues sinitiques, dont le cantonais et le dialecte tankas à Hong Kong ainsi que le cantonais et le pinghua du Sud dans le Guangxi. Le deuxième aspect consiste à explorer davantage la relation entre les two langues sinitiques susmentionnées et le cantonais du point de vue du contact des langues, afin d'examiner les caractéristiques typologiques des langues dans la GBA. Enfin, le troisième aspect consiste à étudier les langues sinitiques de la GBA dans le contexte de l'Asie du Sud-Est continentale.
Ce projet vise à explorer des données linguistiques de première main sur les tankas et le pinghua du Sud de la GBA, qui sont encore des langues peu connues de la communauté scientifique, et à nous permettre de comprendre les mécanismes de contact et de changement de langues ainsi que le problème de la divergence et de la convergence des langues dans cette région particulière de l’Asie du Sud-Est. Ce projet mettra également en lumière le rôle clé du cantonais dans l’évolution des langues de la GBA et permettra à approfondir la recherche sur les langues sinitiques. Enfin, dans ce projet, l’étude sur les origines des langues et les gènes liés aux langues de la GBA contribuera de manière significative à la recherche sur l'évolution ethnique en fournissant aux communautés scientifiques de riches données sur le tankas et le pinghua du Sud.