Semantics and Pragmatics of Evidentials in Turkish

Date of Completion

January 2011


Language, Linguistics|Middle Eastern Studies




This dissertation investigates the semantic and pragmatic characterization of the evidential morphology in Turkish. That evidentials are parasitic to tense/aspect morphology is well noted (Izvroski 1997, Matthewson et al. 2007 a.o.), and the traditional literature describes the morphemes that encode tense and aspect in Turkish, namely [-ml;] and [-DI] as also employing evidentiality (Lewis 1967, Underhill 1976 a.o.) The main hypothesis pursued in this dissertation is that evidential morphology is semantically decomposed as tense or aspect and epistemicity in Turkish. The evidential morphology differs in its compatibility with specific temporal adverbs, and there are differences between the assertability of each subtype of evidentials. Both of these facts argue for the existence of semantically two distinct indirect evidential forms in Turkish language. A comparison of English Present Perfect Aspect and the indirect evidentiality shows that Turkish is part of the well noted Present Perfect Puzzle. This would be unexpected if Turkish did not possess semantically distinct two indirect evidential morphemes. I claim that the English-like Present Perfect Aspect meaning is contingent on the availability of inferential evidence in Turkish. Cross-linguistically, the level of meaning evidential forms are interpreted in varies. I argue in this dissertation that evidentials are presuppositional operators in Turkish language. To that end, in addition to the regular truth conditionality tests, two arguments are used; (i) a comparison of the Free Choice Any in English and the herhangi bir in Turkish shows that the type of evidentiality involved in a statement is determinant in the pragmatics of this item, (ii) evidentials can be embedded, and this is unexpected if evidentials were operative at the speech act level. The pragmatic distribution of each evidential form suggests that evidentials convey information about speaker's commitment levels to the truth of a proposition, and each evidential subtype exhibits different pragmatic properties. ^ One other aspect of evidential morphology discussed in this dissertation is the semantic contribution of reduplication of the indirect evidential morphology in Turkish. Reduplication facilitates a new layer of meaning, which I claim is regulated by an abstract REDUP morpheme. ^