意味論研究会2015年5月22日

In this talk, I will present a portion of my dissertation “Interpreting questions under attitudes”, focusing on the chapter on exhaustivity of embedded questions.

Question-embedding predicates (e.g., “know” vs. “surprise”) vary in the strength of exhaustivity involved in the interpretation of their interrogative complements (Heim 1994; Beck and Rullmann 1999). This observation has led theories to posit flexibility in the semantics of question-embedding (Beck and Rullmann 1999; George 2011), but no theory has succeeded in predicting the strength of exhaustivity given the lexical semantics of the embedding predicates. This talk presents a semantics of question-embedding that achieves this prediction. I will argue that there is only one semantic derivation for embedded questions, i.e., a derivation involving “matrix exhaustification”. All readings fall out from this derivation once the lexical semantics of embedding predicates is taken into account. Specifically, matrix exhaustification derives the so-called “intermediate exhaustivity” (Spector 2006, Klinedinst & Rithschild 2011) only if the embedding predicate is upward monotonic. This reading can then be strengthened into the so-called “strong exhaustivity” by the mechanism of neg-raising.