On June 22, Anastasia Tsilia successfully and masterfully defended her dissertation entitled The Grammar of Defeated Expectations: the case of wh-exclamatives. Here is the abstract:
This dissertation investigates how the grammar expresses defeated expectations of the speaker or attitude holder through matrix and embedded wh-exclamatives, such as Wow, how smart Maria is! and George is surprised at how smart Maria is. It asks two central questions. First, why do wh-exclamatives morphosyntactically resemble wh-interrogatives cross-linguistically, while behaving differently from both interrogatives and declaratives? Second, what is the role of negation in wh-exclamatives when it appears redundant, not reserving the truth value of the sentence? I show that wh-exclamatives are not merely exclamatory uses of declaratives or interrogatives, and I disentangle the speech act of exclamation from the clause type of wh-exclamatives. I argue that the resemblance between wh-exclamatives and wh-interrogatives is not accidental: exclamatives are built from interrogatives plus surprise (or a predicate with an equivalent meaning). A wh-exclamative expresses the surprise of the speaker (for root exclamatives) or the attitude holder (for embedded exclamatives) at the answer to the question. I propose that the structure of root exclamatives is wow-type particle + Q, where wow a speaker-anchored surprise/amaze/can’t believe, and that of embedded exclamatives surprise-type predicate + Q. This accounts for their interrogative-like morphosyntax without reducing them to questions semantically: a question asks for an answer, whereas an exclamative expresses or describes surprise at an answer. The analysis has consequences for question embedding. I discuss an expanded typology of clause-embedding predicates, and I introduce a paradigm showing that declarative- and question-embedding surprise as well as matrix exclamatives all have different discourse contributions. I argue that surprise-type predicates need to have access to their prejacent set of propositions, and thus that answerhood operators should not be applied to the question complement before it composes with the predicate.
The second part of the dissertation turns to a cross-linguistically recurrent negation in wh-exclamatives, which I call Cover Negation (e.g., How often have I (not) watched this series!). Examining the phenomenon cross-linguistically, I show that this negation is not expletive and has certain cross-linguistically recurring properties: it resists stress, does not license negative polarity items, permits positive polarity items, often appears with a discourse-initial and or wh all-type material, and is blocked by relative-clause complementizers. I argue that Cover Negation contributes a coverage reading: the speaker is surprised that the actual wh-values instantiate sufficiently many contextually salient categories. Formally, it performs complementation over a contextual cover, a structured set of categories in the wh-domain. I discuss other corners of the grammar where such a wh-domain is attested, as well as open questions.
Congratulations, Anastasia!
