Back to list

Detail of contribution

Auteur: Lavi WOLF

Epistemic ambiguity across languages

Abstract/Résumé: Epistemic modals (henceforth EM) are viewed in the classical Kratzerian tradition as truth conditional quantifiers over possible worlds, thus part of the propositional content of utterances (Kratzer, 1981; Lewis, 1986). Recently, however, there has been a revival of a claim made by Lyons (1977) with regards to two types of EM, subjective and objective. While the latter behaves as a truth-conditional quantifier should, the former appears to have only wide scope readings (1.a), resist embeddings (1.b) and stay out of assents/dissents from previously uttered material (1.c): (1) a. #Every student may have left but not every one of them has. may >> every (von Fintel & Iatridou, 2003). b. #If Max might be lonely, his wife will be worried (Papafragou, 2006). c. John: This professor must be smart. Marsha: I don’t believe it (=I don’t believe that this professor is smart) (Papafragou, 2006). The subjective-objective distinction gets further substantiation from cross-linguistic data. It appears that while English favors the subjective interpretation of EM (Lyons, 1977), Dutch favors the objective (Huitink, 2008). This observation has specific linguistic implications, for example the possibility of narrow scope readings (2.a) and felicitous embeddings (2.b): (2) a. Iedere student kan vertrokken zijn, maar niet iedere student is vertrokken. Every student may left be but not every student be left. Every student may have left but not every one of them has. every >> may. b. Als Max wel eens eenzaam kan zijn, zal zijn vrouw zich wel zorgen maken. 'If Max may be lonely, his wife will be worried' (Huitink, 2008). Interestingly, Chinese has a particle which marks the scope of a given QP. This particle DOU interacts with epistemic modals as well, and does so in a manner that nicely fits the evidence from English and Dutch. When the modal has wide scope it is interpreted subjectively and when it has narrow scope, objectively (cf. Lin, 2012). In order to provide a uniform theory that accounts for both types of EM, I propose that EM can be asserted either expressively (thus being non truth-conditional speech act modifiers) or descriptively (thus being truth-conditional quantifiers over possible worlds). This proposal is influenced by, and modifies, Lyons' (1977) original distinction, and is further motivated in Wolf (2012). The formal account distinguishes between the propositional content of a speech act and the degree of strength by which it is performed (cf. Vanderveken, 1990-1). The cognitive importance of such a distinction is to differentiate between claims and evidence. When a speaker performs a descriptive EM assertion she makes a claim about possibilities, while expressive EM assertions mark the speaker as a source of evidence for a claim about realities.