Some remarks on the dialects of north Kerintji: a link with the Mon-Khmer languages
E.O. Van Reijn’s 1974 article in the Journal of the Malaysian Branch of the Royal Asiatic Society presents a comparative phonological analysis of North Kerintji dialects, arguing that their characteristic sound-shifts are not independent developments but reflect a Mon-Khmer substratum in the highlands of West Sumatra. Drawing on field data from fourteen Kerintji dialect localities and extensive cross-referencing with Mon-Khmer, Munda, and other Austronesian languages, Van Reijn demonstrates that the parallelisms are too systematic to be attributed to chance.
Summary
The article systematically documents six categories of sound-shifts shared between North Kerintji dialects and the Mon-Khmer family. The most striking is the reduction of originally bisyllabic words to monosyllabic forms through the loss of the first-syllable vowel, producing initial consonant clusters in which the first consonant reduces to a glottal stop that merges with the following consonant to form a preglottalized implosive stop. This development is paralleled in Cham, Bahnar, Thai, and the Besisi dialects of the Malay Peninsula. Additional shifts include the regular change of final -s to -h (often with diphthongization of the preceding vowel), the loss of final -l, and a series of characteristic vowel diphthongizations (u > au, i > aː or ɛː, i > uː or ɔː, a > aː, a > au, a > ea or ɛa).
A second major section addresses the development of final unvoiced stops into preglottalized nasals, which may then give rise to new unvoiced stops or reduce to a bare glottal stop. Van Reijn traces these chains through multiple Kerintji dialects and compares them with parallel developments in Semang, Sakai, Bahnar, and Stieng forms. The article also distinguishes between “pausal” and “non-pausal” forms in Kerintji, noting that the non-pausal forms often preserve vowels and consonants lost in the pausal variants.
The concluding section extends the argument to intervocalic consonant clusters, showing that the frequent appearance of intervocalic m, n, p, t, k in Achinese, Cham, Redjang, and Batak dialects (where Malay has mb, nd, mp, nt, ŋk) is best explained as the differentiation of an original implosive voiced stop rather than as assimilation of a homorganic nasal. Van Reijn invokes Haudricourt’s work on preglottalized consonants in Indochina and Pinnow’s historical phonology of Kharia to support this reconstruction, arguing that the “m with an implosive attack” described by Jaspan for Redjang is the same implosive voiced stop documented by Schmidt for Mon.
Key Findings
- Six distinct sound-shift categories are documented across fourteen Kerintji dialect localities (Semurup, Kemantan, Sungai Tutung, Koto Payang, Koto Lanang, Rawang, Koto Baru Rawang, Dusun Baru, Sungai Penuh, Kumun, Tanjung Pauh, Pulo Tengah, Tanjung Tanah, Hiang), each with parallel developments in Mon-Khmer languages (pp. 130–135).
- The bisyllabic-to-monosyllabic reduction produces preglottalized implosive stops, a development shared with Cham, Bahnar, Thai, and Besisi of Sepang; for example, Kerintji (TT) ‘rau^N corresponds to Malay harum “sweet-smelling” (p. 130).
- The shift -s > -h with diphthongization is regular across Kerintji, Kelantan, Cham, Achinese, and Minangkabau, as in Kerintji (KL) alauh = Malay halus “soft, delicate” (p. 131).
- Final unvoiced stops develop into preglottalized nasals through a chain: e.g., Malay bukit “hill” yields Kerintji (R.s.) bukuʔŋ < buku1ʔŋ < buku1ʔn < bukIʔn < bukIt, demonstrating a full phonological chain (p. 135).
- The intervocalic m/mb variation in Achinese, Cham, Redjang, and Batak dialects is reconstructed as the outcome of an original implosive voiced stop *b, not as nasal assimilation, supported by Pinnow’s data on Kharia and Haudricourt’s analysis of Indochinese preglottalized consonants (pp. 136–137).
- Pausal and non-pausal forms in Kerintji are distinguished throughout; non-pausal forms often preserve final-syllable vowels and consonants lost in pausal forms, a phenomenon first noted by Voorhoeve for Redjang (p. 130, n.1).
Conclusion
Van Reijn’s definitive takeaway is that the systematic parallelism between the sound-shifts of certain Sumatra languages and the Mon-Khmer family (including the dialects of the aboriginal tribes of the Malay Peninsula) “cannot be due to mere chance, but points to the existence of a Mon-Khmer substratum in the highlands of West Sumatra” (p. 136). While acknowledging that much further research is needed, he argues the hypothesis is already useful for elucidating specific dialectal peculiarities, particularly the status of intervocalic consonant clusters in the region.
Context
- The primary data consist of Van Reijn’s own field recordings and transcriptions from fourteen North Kerintji dialect localities, supplemented by published comparative wordlists (Skeat & Blagden’s Comparative Wordlist), Jaspan’s Redjang Ka-Ga-Nga-Texts (1964), and Pinnow’s Versuch einer historischen Lautlehre der Kharia-Sprache (1959).
- The article contributes to the long-standing debate on the linguistic prehistory of the Malay Peninsula and Sumatra by providing phonological evidence for a Mon-Khmer substratum, building on earlier work by Kuiper (1948) and Haudricourt (1950) while extending the argument specifically to the Kerintji dialect continuum.