Definiteness and implicatures in Tagalog∗

The use of an indefinite NP often implies that the NP’s descriptive content is instantiated by more than one individual. But do indefinites conventionally signal this notion of nonuniqueness or does nonuniqueness arise pragmatically, via reasoning about scalar alternatives? In this paper I argue that in Tagalog, indefinite bare NP patients are only able to implicate nonuniqueness if there is a grammatical alternative utterance employing a definite. In cases where the alternative utterance with a definite is blocked by the rules of Tagalog grammar (in particular word order configurations), the nonuniqueness implicature does not arise. I show that these results provide evidence for the pragmatic view of nonuniqueness. Building on this result, I discuss a model of implicature calculation which is sensitive to the syntactic well-formedness of pragmatic alternatives. In focus is the interpretation of bare NPs in Tagalog. Tagalog demonstrates a puzzling paradigm (noted by McFarland, 1978; Schachter and Otanes, 1982; Adams and Manaster-Ramer, 1988, amongst others). The definiteness of the patient NP appears to be sensitive to the order of constituents within its containing clause. In verb-initial sentences, genitive case-marked patients are indefinite (1a). In actor-initial sentences, such as the cleft structure in (1b), this interpretive constraint on the genitive patient NPs is lifted, and it becomes compatible with a definite interpretation.