November 13, 2012: Hedde Zeijlstra (University of Amsterdam) — LUCL
We are very happy to announce that Hedde Zeijlstra (University of Amsterdam) will be giving a LUSH talk entitled Universal NPIs and PPIs.
Time: 11:00 – 13:00
Location: LEIDEN, Lipsius, room 002
In the domain of existential quantifiers over individuals, both Negative Polarity Items (NPIs), such as English anybody, and Positive Polarity Items (PPIs), such as someone, are attested. However, no universal quantifier over individuals seems to behave like an NPI or PPI.
In the domain of deontic modals, things are however, different: here some universals can be PPIs (e.g. English must) or NPIs (e.g. English need), but no existential deontic NPIs or PPIs have been attested.
In this talk, I argue that one of the well-known explanations for NPI-hood, namely Chierchia’s (2006, 2011) account in terms of domain widening and obligatory exhaustification) actually predicts the existence of universal PPIs. I demonstrate that those deontic modals that are PPIs are indeed PPIs for the same reason that English any-terms are NPIs.
This result gives rise to two questions. First why does it look like there are no quantifiers over individuals that exhibit PPI-hood? And, second, how should the NPI-/PPI-hood of those NPIs/PPS be explained that do not follow from Chierchia’s account.
In the remainder of this talk, I first discuss that, despite appearances, universal PPIs that quantify over individuals do exist, but that the morphosyntactic differences between such PPIs and the modal PPIs make that the former but not the latter display the typical PPI-like behaviour that is standardly taken as a diagnostic.
Second, I show that existential PPIs and universal NPIs must exhibit NPI/PPI-like properties for different reason than their mirror images (existential NPIs, universal PPIs) and that this makes testable predictions with respect to the strength of their NPI-/PPI-hood.