Abstract
Existing analyses of NPI licensing in questions instantiate two differentapproaches. One approach holds that questions are NPI licensers in their own right(Kadmon & Landman 1990; Krifka 1995, 2003; van Rooy 2003); the other holdsthat, in virtue of their syntax, questions host silent expressions that do the licensingfor them, such as a silent version of exclusive only (Nicolae 2013, 2015) or negationnot (Guerzoni & Sharvit 2014). Based one a pattern of NPI licensing in alternativequestions, this paper presents a case for the former approach. Specifically, it offersan argument for the analysis developed in Krifka 1995, 2003 and van Rooy 2003,which centrally refers to questions’ information theoretic entropy (Shannon 1948).
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
