Prosodic effects on factive presupposition projection

Abstract

This paper investigates the interaction between prosodically mediated pragmatics and factive presupposition projection. In particular, it addresses a set of proposals, articulated most clearly in Abrusán (2011, 2016); Simons et al. (2010), and Simons et al. (2017), which argue that prosodically mediated focus, as a signal of the Question Under Discussion (Roberts, 1996, 2012), determines whether or not particular content becomes presupposed (Abrusán) or ends up projecting from the scope of entailment-targeting operators (Simons et al.). We present experimental results demonstrating that the predictions made by these proposals are too strong: although focus is shown to have an effect on factive presupposition projection, it does not completely eliminate the factive inference, as argued by these authors. Rather, we find that the main factor determining whether or not a factive inference projects is the identity of the predicate. We argue that this supports a view whereby factive presuppositions are lexically triggered, and may only be cancelled in a particular set of embedded contexts via local accommodation (Heim, 1983). However, focus may give rise to inferences via the QUD, to the effect that the factive inference is weakened (although not completely eliminated).