The Semantics-Discourse-Pragmatics Interface: Theory and Applications
Understanding the meanings of utterances of sentences requires understanding three potentially distinct kinds of meaning: 1) semantic (or truth-conditional) meaning; 2) discourse moves/context updates (e.g. by asserting a sentence, you propose that the context be updated in a particular way); 3) pragmatic inferences (e.g. what can we conclude about the speaker’s communicative intent on the basis of their choosing to make the utterance they made).

Some have argued for conflating 1 and 2: various frameworks under the umbrella of dynamic semantics (Kamp 1981, Heim 1983, Veltman 1996, etc.) propose that we can understand the semantic content of sentences in terms of the way they update contexts. Others have argued for conflating 2 and 3: perhaps we can derive the effects of assertions entirely from pragmatic reasoning about a speaker’s decision to say a sentence with particular truth conditions (Lauer 2013). Others develop models in which the three are kept separate (Farkas & Bruce 2010).

This course begins by providing background on these theories: what are these three levels of meaning? How are they cashed out in the above frameworks? Why might we want to conflate them? Why might we want to keep them separate?

The course then works through three concrete applications to tricky phenomena currently being argued about in the literature on semantics and pragmatics, showing that taking these interfaces seriously can lend insight to problems that appear intractable if our attention is constrained to semantics proper. These applications are as follows:

I. Perspectival centering with matrix epistemic modals.
A sentence like “Paul might be at the party” is true iff Paul being at the party is compatible with what is known. But known by who? Intuitively, when I say this sentence, I’m saying that it’s compatible with what I know. But if you reply by saying, “No, you’re totally wrong,” you seem to be saying that it’s not compatible with what you know. Why does the perspective seem to flip from speaker to addressee? Is the source of this flexibility in the semantics of epistemic modals? Or does it come from what we’re doing when we assert and reply to epistemic claims?

Relevant literature:
Kratzer 1977: What ‘must’ and ‘can’ must and can meanYalcin 2007: Epistemic modalsStephenson 2007: Judge dependence, epistemic modals, and predicates of personal tastevon Fintel & Gillies 2011: `Might’ made rightMacFarlane 2014: Assessment sensitivityYanovich 2014: Standard contextualism strikes backMandelkern 2019: How to do things with modalsRudin 2019: Asserting epistemic modals

II. The meaning of English intonational tunes
A declarative sentence accompanied by a steep monotonic rise asks a biased question, rather than making an assertion: “You have a daughter?” Rising intonation doesn’t change the truth conditions of the sentence, but does change the way its utterance affects the context. How can we analyze this? As is the biased nature of the question encoded in the way the utterance affects the context, or does it come from pragmatic reasoning about the speaker’s choice to make this discourse move instead of some other one?

Relevant literature:
Bolinger 1989: Intonation and its usesGunlogson 2001: True to formNilsenova 2006: Rises and fallsMalamud & Stephenson 2015: Three ways to avoid commitmentsKrifka 2015: Bias in Commitment Space SemanticsFarkas & Roelofsen 2017: Division of labor in the interpretation of declaratives and interrogativesWestera 2018: Rising declaratives of the Quality-suspending kindJeong 2018: Intonation and sentence type conventionsRudin 2019: Intonational commitments

III. The semantics of imperatives
There is very little consensus about the most basic analysis of the semantics of imperatives: what kind of semantic object do they denote? Is their directive force a part of their semantics? What explains their illocutionary variability—i.e. the fact that some are commands, some are offers, some are wishes, and so on. How should the meaning of imperative utterances be parceled out into denotation, context update, and pragmatic inferences? How can we probe this question empirically?

Relevant literature:
Schmerling 1982: How imperatives are special, and how they aren’tPortner 2004: The semantics of imperatives within a theory of clause typesCondoravdi & Lauer 2012: ImperativesBarker 2012: Imperatives denote actionsKaufmann 2012: Interpreting imperativesCharlow 2014: Logic and semantics for imperativesStarr 2018: A preference semantics for imperativesRudin 2018: Rising imperatives