(Non)culmination by abduction

Recent literature has aimed to explain (non-)culminating accomplishment inferences, which often involve the perfective aspect, but can also involve the imperfective. The goal of our paper is to explore how these inferences come about with the Hindi perfective and the Russian imperfective. We propose that abduction, that is, inference to the best explanation, is ideally suited for this task. We show how the occurrence of a (non-)culminated event is abduced in the relevant cases based on a semantic analysis which adopts the distinction between culminated and maximal events, as well as a set of non-defeasible rules encoding general mereological principles. We also show how our abductive framework can take into account facts about the conversation. This, among other things, allows us to make more nuanced predictions about what speakers will infer and when, thereby addressing possible worries of overgeneralization that an abductive framework inevitably faces. We end the paper with two outstanding issues warranting further research. First, we raise questions about the nature of (non-)culminating accomplishment inferences, which have previously been taken to be conversational implicatures. Second, we take some preliminary steps towards extending our analysis to defeasible causatives in Germanic and Romance languages.