Presupposition Accommodation and Informativity Considerations with Aspectual Still

This paper deals with a newly observed phenomenon which lies at the interface of the semantics and pragmatics of aspectual still (as in John is still asleep), namely the fact that still is systematically infelicitous when the time described by the sentence (or more technically, the reference time) is not given a specific characterization by a temporal adverbial or the utterance context. The main claim I make is that this kind of infelicity results from the fact that in such constructions the use of still is uninformative. The analysis, which is couched within a DRT framework (Kamp 1981), also argues for two more general theoretical points. First, I show that the data cannot be accounted for by using current definitions of the presupposition triggered by still. I propose a modified definition of still which, following Ippolito’s 2007 analysis, uses one eventuality variable in both the assertion and presupposition of sentences with still, but, unlike this theory, does not require the denotation of this variable to be contextually salient. Second, as opposed to many current theories, I show that the data on still supports a view where tense in past tense sentences is not necessarily anaphoric or pronominal, but rather potentially existentially closed and novel. The novel data is described in section 1. Section 2 gives background about the anaphoric vs. novel characterization of tense and reference times and about the semantics of still. After examining some potential explanations of the data in section 3, and pointing out their shortcomings, I turn to my own proposal for the semantics of still in section 4, and show how it can account for the novel data. Section 5 examines the compatibility of the analysis with current claims on presuppositions and informativity. Section 6 concludes the paper and examines some remaining questions and directions for further research.