A new view on NAmE high negated questions: Experimental evidence and analysis

Abstract

The consensus on English High Negated Questions (HNQs) is that they necessarily convey the speaker’s bias for the prejacent. I argue that it is the speaker’s bias for the addressee’s belief in the prejacent that is necessarily conveyed by HNQs. I support this claim with experimental results, which show that HNQs are not deemed felicitous without the speaker’s bias for the addressee’s belief in the prejacent even when there is speaker bias for the prejacent, and that they are felicitous when there is speaker bias for the addressee’s belief in the prejacent but no speaker bias for the prejacent. I propose a theoretical analysis which combines independently motivated elements to predict this phenomenon and shed further light on two issues that are unresolved in the literature.