Yakkha complex predicates and the grammar/lexicon distinction

Schackow, Diana
Journal Title
Journal ISSN
Volume Title
This presentation deals with the problem of the grammar/lexicon distinction, illustrated by example of the complex predicates (CPs) of Yakkha, an endangered Tibeto-Burman language of Nepal (currently documented as part of the author's PhD thesis). CPs have tended to receive marginal attention in grammaticography, as they occupy an intermediate position between grammar and lexicon (Schultze-Berndt 2006). The presentation will give an overview of the highly productive and multifunctional system of complex predication in Yakkha, arguing that a grammar which neglected these CPs would miss a vital component of the 'genius' of this language. A CP in Yakkha consists of at least two verbs: a lexical verb (V1) and one or more of thirty ‘function verbs’ (V2), responsible for the fine-tuning of a predicate, regarding the argument structure, temporal structure, spatial orientation or modality of an event. This description, however, simplifies the heterogenous internal structure of the CPs, as grammaticalizations and lexicalizations have equally contributed to their formation. A particular V2 may have a productive grammatical function, and also participate in non-compositional, idiomatic verb- verb combinations. For instance, the verb pipma ‘give’ is a benefactive marker and it marks events by which some participant is affected involuntarily. This semantic component is also found in lexicalizations such as yoŋdipma ‘be frightened’ (‘shake-give’). The verb tsama ‘eat’ has developed into a reflexive marker, but it also occurs in dozens of lexicalized predicates, often with typical middle semantics (actions involving the body, social interactions), e.g. intsama ‘play’ (‘revolve-eat’) or lemtsama ‘cheat’ (‘flatter-eat’). Many V2 occur in intermediate stages between lexicalization and grammaticalization. These examples serve to illustrate that due to the heterogenous structure of the CPs and the various ways in which V1 and V2 may interact, it is not a trivial question where to draw the line between grammar and lexicon. The more general lesson for language documentation that we can draw from this particular (but surely not unique) case is that pre-established concepts (like 'grammar' and 'lexicon'), as useful as they are, may have to be re-thought and further developed in order to provide a just description of a particular language.
Access Rights
Email libraryada-l@lists.hawaii.edu if you need this content in ADA-compliant format.