Unitariness and Partial Identification in the Bella Coola Middle Voice

Abstract

The Bella Coola suffix -m has been analyzed in the literature as two or even three separate morphemes, based on the variable effects it has on the transitivity of its base. In this paper, I argue for -m as a single morpheme with a unified meaning-specifically, as a marker of a special case of Kemmer’s (1993) definition of the middle voice which I will refer to as the "non-unitariness" of event-participants. The cross-linguistically unusual "transitivizing" uses of -m are shown to fall out from the presence of a second, individuable entity in a semantically typical middle clause, independent of the application of the middle-marker itself.