Tibeto-Burman "split cognates"

Abstract

The characteristic "split cognates" of Austro-Tal (ATLC:152) are the product of contrasting modes of canonical reduction [CR]: Miao-Yao (CRR) vs Kadai (CRL). Japanese regularly displays CRR: me 'eye', Old Japanese me < •mai (Benedict 1990). The monosyllabic roots of Sino-Tibetan hardly lend themselves to "splitting" of this kind, however, even though the actual forms tend to be disyllabic through prefixation.I This follows from the fact that the prefixes involved are weakly stressed, typically with schwa vocalism, hence can scarcely be expected to furnish a basis for a CRR "split" product of MY/Japanese style. It would appear that only a deviant ST language would behave in so unseemly a fashion and indeed this has proved to be the case.