Sophia Malamud

An old joke begins, “When a diplomat says yes, he means maybe; when a diplomat says maybe, he means no.” The distinction between the direct meaning, and the ultimate import of an utterance points to the complex interaction between grammar, culture, and rationality. The present research investigates this interaction, and helps elucidate the way that an ultimate import of an utterance arises from the grammatically encoded meanings, the speakers’ understanding of their cultural context, and their assumptions about rational progress of conversation.

The goal of a semantic analysis is to determine units of meaning (meanings of words and parts of words), and to formulate ways in which these meanings grammatically combine to form the meanings of more complex expressions. While semantics is often defined as the study of conditions under which utterances are true, recent research has strived to apply its tools and insights to the study of expressions that cannot be said to be true or false, such as imperative sentences (e.g., Go!).

When the meaning of an expression varies with its extra-linguistic or discourse context, the scholar of meaning must determine whether this variation is due to underspecification or ambiguity in the meaning of the sentence per se (as in the case of “I”, “yesterday”, and “her” in I saw her yesterday) or whether some part of the information conveyed is due to inference in pragmatics. The goals of pragmatics are, in one definition, “the study of linguistic acts and the contexts in which they are performed” (Stalnaker). Pragmatics includes computations that derive the ultimate force of an utterance (sometimes termed speaker-meaning) based on its compositional semantics, cultural assumptions, and general rationality-based reasoning. Recent scholarship has strived to adopt the tools of decision and game theory to make such modelling of contextual inferences more precise. At the interface between semantics and pragmatics are phenomena where the ultimate force of an utterance arises through interaction of pragmatic inference and semantic composition.

I propose to extend and combine these lines of research at the semantics-pragmatics interface in a study of the meaning of four utterance modifiers: Mandarin particle ba, English reverse-polarity tags (John is here, isn’t he?) and rising intonation (John is here?), and English and Russian please. In this talk, I focus on Mandarin ba. I build on my prior work (Ettinger and Malamud 2013, Malamud and Stephenson 2014, inter alia) to offer a model of conversation where a unified semantics of clause types (cf. Starr 2010) constrains the range of interpretations for an utterance. A pragmatic conversational scoreboard tracks speakers’ public commitments to propositions, issues, and actions/preferences (cf. Farkas & Roelofsen forthcoming). These commitments constitute a target for collaborative updates; conversational moves may fall short of this target. Moves that fall short place their at-issue content on different parts of the Table, depending on the degree of authority the speaker and the hearer exercise over this content. Mandarin ba signals a lower degree of speaker authority, and potentially higher degree of hearer authority than would otherwise be expected for assertions and requests; pragmatic reasoning then derives the various effects of the utterance modifier.