(University of Helsinki)
Syllable Structures North to South in the Altaic World
68th Annual Meeting of the PIAC, Bangkok 2026
The prototypical (Ural-)Altaic non-derived word root for non-deictic functions is a simple bisyllabic sequence of consonants (C) and vowels (V) with no initial cluster, no final consonant, and with a maximum of two successive consonants (CC) in intervocalic position: (C)V(C)CV. This is still synchronically the dominant type in Mongolic and Tungusic (as well as in conservative Uralic languages like Finnish). However, in many modern languages this structure has been simplified by deleting the final vowel, with monosyllabic sequences of the type (C)V(C)C as the result. This has happened only relatively recently in several Mongolic and Tungusic (as well as in many Uralic) languages, and it is also observed in early Turkic at the level of the pre-protolanguage (as noted already by Ramstedt 1948: 99, but not properly recognized by some later scholars).
However, to the south of the Altaic sphere, a different syllable structure prevails, based on a monosyllabic sequence in which the initial consonant (C) can be preceded by a pre-initial (H) and followed by a medial (M), while the final consonant (F) can be followed by a post-final (S), with the maximal six-segment sequence ((H)C)(M)V((F(S) as the result. This is the originally dominant syllable type in languages of the Tibetan (Bodic) sphere (as in Classical Tibetan). At a deeper chronological level these monosyllables may have developed from sesquisyllabic structures of the type HVC(M)V(F)S. This was probably also the original syllable type of Chinese (and probably of other languages with a Sinitic typology in SE Asia), though later developments have introduced tonal distinctions at the expense of the complexity of segmental sequences and distinctions.
The two syllable types have met at the intersection of northern and southern Central Asia, that is, Mongolia and Tibet. There are indications that the Mongolic languages spoken in the historical context of the Amdo Province of Tibet (modern Gansu and Qinghai Provinces of PR China), also known as the Shirongolic languages, have in the course of the last few centuries undergone a change in their phonotactic orientation and started to introduce initial consonant clusters of the Tibetan type, containing an initial and a preinitial, of which the preinitial can have been subsequently lost. This has yielded unexpected correspondences, such as Central Mongolic (Khalkha) shud [ɕut] vs. Shirongolic (Bonan) dung [tuŋ] < rtung [ʂtuŋ] < Proto-Mongolic *sidü/n ’tooth’. However, these developments have affected the different Shirongolic languages at different time levels, in different ways, and even in different lexemes (as discussed by Nugteren 2011: 91–92 and passim).
The present paper will attempt to analyse the reorientation of syllable structure in the Shirongolic languages in some more detail and to place this phenomenon in an areal and chronological context.
Nugteren, Hans (2011). Mongolic Phonology and the Qinghai-Gansu Languages. LOT 289. Utrecht: Netherlands Graduate School of Linguistics.
Ramstedt, G. J. (1948). Über Stämme und Endungen in den altaischen Sprachen. Journal de la Société Finno-Ougrienne 55 (2): 98–105. Helsinki 1951.