The Department of Linguistics is pleased to present Donka Farkas speaking on Canonical and non-canonical speech acts.

Zoom Information will be emailed on Thursday, November 19, 2020.

Abstract
The general issue addressed in this talk is how best to characterize canonical and non-canonical speech acts. The framework I will use is rooted in Farkas and Bruce (2010) and Farkas and Roelofsen (2017). The speech acts I concentrate on are assertions and questions.

The first part of the talk focuses on canonical assertions and questions. Pretheoretically, canonical, or typical, assertions are informing speech acts whereby a knowledgeable speaker informs her addressee of the truth of the proposition she expresses. Canonical, or typical, questions request information, i.e., an ignorant speaker requests her addressee to resolve the issue she raises. The question addressed in this part of the talk is why should canonical assertions and questions have these properties? I will attempt to answer it by showing that these properties follow from a context structure view based on Farkas and Bruce (2010) and the basic conventional discourse effects (CDE) declaratives and interrogatives are assigned in Farkas and Roelofsen (2017). CDE are defined as functions from the denotation of sentences and input context structures to output context structures. These functions affect the discourse commitments of the speaker, the conversational table and the future states of the conversation the move projects.

The second part of the talk considers ways in which assertions and questions can be non-canonical, i.e., ways in which declaratives and interrogatives can be used in contexts that override the canonical default assumptions discussed in the first part. It will be argued that such non-default cases can be either unmarked (as in the case of `quiz’ questions in English) or marked for various types of particular deviations from the canonical case, such as markers of bias in questions, which signal a departure from the speaker neutrality assumption in questions, or markers of non-categorical commitment, which signal departures from speaker knowledgeability in assertions. It will be argued that a promising way of treating
some of the linguistic means used in non-canonical speech acts is to treat them as force modifiers, i.e. as contributing special CDE, treated formally as functions from contexts C to contexts C’, where C is the result of applying the basic CDE of the sentence to its input context (see Faller, 2002; Murray, 2010).

For questions please email Maria Zimmer.