On the syntax and semantics of mirativity: Evidence from Spanish and Albanian.

摘要

In this dissertation, I examine mirative constructions in Spanish and Albanian, in which past tense morphology is used to convey speaker's surprise and does not seem to contribute its usual temporal meaning to the asserted proposition. I put forward an analysis that makes the following claims.;First, mirative sentences are assertions that include a modal component. This modal component brings up the speaker's beliefs in a way that entails the opposite of what the assertion expresses. Thus, a clash is generated between the speaker's beliefs and the assertion, and this triggers a sense of surprise.;Second, the past tense morphology is analyzed as being a real past tense, following recent proposals for counterfactual conditionals. In the case of miratives, the past tense keeps its normal semantics, but is interpreted in the CP domain as the time argument of the modal base, rather than in TP. The beliefs that are contrasted with the assertion are therefore past beliefs up to the discovery time (which usually coincides with the speech time), in which the actual state of affairs is encountered by the speaker.;Third, a syntactic Agree relationship is established between the interpretable past tense feature in C and the uninterpretable past tense feature in T. This Agree relationship must meet locality requirements, such that there must not be intervenors between C and T. I claim that the participle movement observed in Albanian miratives happens in order to overcome intervention, so that Agree can take place.;The dissertation also accounts for the role of aspect in deriving differences in the nature of the surprise, at least in Spanish, and for the aspectual requirement that miratives seem to have cross linguistically. Finally, I show that miratives cannot be analyzed in the same way as apparently similar constructions, in terms of usage, such as exclamations and exclamatives. On the contrary, the analysis presented in this dissertation calls for a closer relationship between miratives and counterfactual conditionals, which I also explore.