hal-05506291 The paths to rhotacization in Tangut, with new evidence from Qiangic

Since Nishida (1964), the term RHOTACIZATION refers to a feature characterizing a set of syllables in Tangut phonology, specifically those in the second and (most of the) third minor cycle. Beaudouin (2023a) synthesizes previous research, describing this set as the outcome of a merger involving three historical "r" sounds: preinitial *r-, medial *-r-, and coda *-r. To date, only the development of *-r into rhotacization has supporting evidence based on an allophonic realization /Vr/ [V˞], as observed in Geshiza Horpa (Honkasalo 2019). However, no similar evolutionary pathway has been established for the other two historical "r" sounds. Additionally, Tibetan and Chinese transcriptional data suggests that some preinitial "r" sounds might indeed have been (at least slightly) realized, challenging the unity of this rhotacized class. This paper addresses these two challenges in a tripartite way. First, it explores the phonological status of the second and third minor cycles through a thorough analysis of their internal distribution, and of their transcription in Tibetan and Northwest Chinese. Second, it recalls the three documented origins of these syllables by summing up comparative evidence from previous studies. Third, it draws from allophonic processes documented in Qiangic languages to investigate how these distinct segments could merge into one autochthonous category. A pattern of monosyllabicization appears, shedding light on some overlooked phonotactic and distributional constraints.