False but Slow: Evaluating Statements with Non-referring Definites

Abstract

One central debate in the analysis of definite descriptions concerns the truth-value of sentences where there is no entity that meets the description in the definite. Classical Russellian accounts predict them to be plain false, whereas presuppositional accounts predict them to be infelicitous. Recent discussions have homed in on the factors that affect actual judgment behaviour in relation to the underlying status posited by different accounts. This article presents experimental evidence for a presuppositional view based on response times for judging statements with non-referring definites to be ‘false’, which were longer relative to control statements where existence was asserted. I discuss the theoretical implications of these results, as well as of other findings from the literature, arguing that they support a presuppositional view of definites that sees the existence presupposition as conventionally encoded. The article also makes a methodological contribution, as systematic evidence on speakers' judgments in these cases turns out to be hard to come by. Finally, the results inform the more general issue of the online processes involved in the interpretation of presupposed, as opposed to asserted, content.