'Anti-ergative' marking in Tibeto-Burman

Abstract

Using arguments based on the data on verb agreement (pronominalization) in Tibeto-Burman (TB), LaPolla 1989 (see also LaPolla 1992) claims that Proto-Tibeto-Burman (PTB) should be reconstructed as a language with no inflectional morphology. In that paper it is argued that the PTB system of grammatical relations' was close to the typical 'role-dominated' (Van Valin & Foley 1980) Burmese-Ylpho system (epitomized by Lahu—see Matisoff 1973); that is. a system where there is no definable 'subject' or 'direct object', and where semantic and pragmatic principles govern the organization of discourse, not syntactic functions. In this paper we look at the nature of 'objects' in TB languages, and find additional support for this role-dominated view of PTB grammatical relations. From a survey of ninety-five reliable grammars or descriptions of languages in the TB family, I have found 11 languages with no nominal object marking, 20 languages with nominal morphology consistently marking the patient as object, regardless of clause type, and 64 languages where the patient in monotranstive clauses Is often or always marked with the same postposition as the goal or beneficiary (dative) in ditransitive clauses. This type of marking is discussed in Dryer 1986 as Primary Object marking. I argue that this type of marking in the TB languages reflects the semantically based nature of grammatical relations in PTB.