This project aims at identifying both the different strata of Altaic influence and the contact-induced changes in the hybrid dialects of the Gansu-Qinghai area, as well as questioning their classification among the North-Western branch. It involves highly endangered dialects that will probably disappear in the very near future, and will contribute to increase the visibility and awareness of regional languages and their communities of speakers. Traditionally, the North-Western Chinese dialects branch regroups all Mandarin varieties spoken from Xinjiang to Shanxi provinces (Yuan Jiahua 1960); further divisions were made in the late 1980s between the Lanying dialects, the Central Plains dialects, and later, the Jin languages, whose status is still disputed among linguists (Wurm, Li, Baumann and Lee 1987). The first two sub-branches geographically cover several dialects situated in the ‘Hexi Corridor’ (spread between Gansu and Qinghai provinces) that have often been qualified as “creoles”, “mixed languages” or “Chinese hybrid dialects” (Dwyer 1995, Wurm 1995, Slater 2003, Janhunen 2004, Bell 2020, Lefort forthcoming) due to several non-Sinitic traits that have been identified in these languages. Those varieties include the Xining, Wutun, Zhoutun, Gangou, Linxia, Dongxiang Chinese and Tangwang dialects, spoken by both Hui (Chinese Muslims) and Han communities.
During the early stages of the project, the relevance of Sino-Arabic writing systems to the research questions was identified as a significant new development. While the original plan was to rely primarily on historical sources from the Jin, Yuan, and Ming periods (notably the three editions of the Laoqida), the scope was subsequently expanded to incorporate additional materials. These included texts written in the Xiaojing script—an Arabic-Persian based system historically used by Chinese Muslim communities to transcribe Chinese dialects, particularly Linxia and Tangwang. Furthermore, earlier sources were integrated, such as studies of 16th-century Persian translations and linguistic materials linked to mosque-based teaching practices (jingtangyu), with records extending to the early 20th century.
Another strand of research focused on documenting Arabic and Persian lexical borrowings in both the Mongolic (Dongxiang) and Sinitic (Linxia and Tangwang) languages of the region. This documentation enabled a more precise assessment of the mechanisms and pathways of linguistic influence. The findings revealed that the non-Sinitic features found in the hybrid Chinese dialects of Southern Gansu were already present in earlier forms, suggesting reanalysis as the principal process of borrowing and the earliest stage of influence. Moreover, non-Altaic languages—particularly Persian—were found to have played a far more significant role in shaping the linguistic features of the region than previously acknowledged. The majority of Arabic-derived vocabulary in both Mongolic and Chinese hybrid varieties appears to have been transmitted via Persian and Turkic languages, indicating a shared borrowing pathway and common patterns of language contact.
The Hexi Corridor is also home to different ethnic minorities who speak various Mongolic, Turkic and Tibetan languages that have influenced one another to a greater or lesser extent, resulting in the mutual transfer of certain linguistic traits that are completely absent from the language families they belong to. On that account, some researchers consider having identified a linguistic area (Dwyer 1995, Janhunen 2004, Slater 2003, etc.), whereas others believe this region regroups several different isoglosses (Nugteren 2011). All these Chinese hybrid dialects share a number of linguistic features thought to have been induced through contacts with Mongolic (and possibly Turkic) languages spoken by neighbouring groups. Historical and linguistic elements suggest that those Mongolic languages may have emerged between the late 13th and early 14th century, at the time of the Mongolian invasions of Central Asia. Chinese speakers have been exposed to those languages ever since, hence the formation of the above-mentioned hybrid dialects.
Chinese scholars tend to argue for a one-to-one language influence (Luo 2004, Lei 2017, Yang & Zhao 2021, Min 2018, Zhou 2023, etc.), but do not put them into a broader perspective. Indeed, various other Chinese languages—those of the Shanxi and Shaanxi regions, for instance—also display some of these traits though they were not in contact with Mongolic languages in recent times. These traits are, most probably, the results of contacts with Altaic languages that occurred around the 10th–12th centuries, at the time of the Liao empire rule in Northern China, whose leaders were speakers of Khitan (Janhunen 2005, Yang 2015), or later, during the Mongolian rule of the Yuan dynasty. These dialects include those spoken in the Shanxi and Shaanxi regions, along the Yellow River, which belong to the Central Plains and Jin branches. Accordingly, some of the traits found in the Chinese hybrid dialects of the Gansu-Qinghai area might be evidence of a deeper stratum of influence—and perhaps of a common origin with other dialects—rather than changes induced by secondary contacts.
Collectively, these results advance the project’s objective by providing a more nuanced understanding of the multilingual interactions that have shaped the linguistic landscape of the Hexi Corridor. They also highlight the broader socio-historical role of Persian and Islamic cultural transmission in the development of regional dialects, thereby integrating relevant perspectives from the social sciences and humanities into the analysis.
