Finiteness in Movima

This paper investigates to what extent finiteness is a useful category for characterizing predicates in Movima (isolate, Bolivia). In the traditional definition, finiteness enables a predicate to function as the main predicate of an independent declarative clause, and finite predicates are verbs containing temporal, aspectual, and/or modal information as well as person indexing. In Movima, however, main-clause predicates are not necessarily verbs, and they are not systematically marked for the above-mentioned categories. In contrast, complement and adverbial clauses, which involve nominalization, convey temporal and aspectual information and show more systematic person encoding; they thereby display more traditional finiteness criteria than main predicates. Furthermore, unmarked predicates can also occur in DPs, which then refer to participants in states and events. I argue that the difficulty to detect finiteness in main-clause predicates is due to a fundamental characteristic of Movima: unmarked predicates characterize entities, i.e. event/state participants rather than the states/events themselves, and this is incompatible with the (traditional) idea of finiteness, which is based on the state/event-denoting character of predicates. Moreover, the obligatory spatio-temporal localization of referents contributes to the “finite” character of pronominal predicates and of complement and adverbial clauses.