Learning what to (not) take for granted
Location: TSH 203
Guest Speaker: DR. ATHULYA ARAVIND
Bio: Dr. Athulya Aravind is a post-doctoral fellow at the Harvard Lab for Developmental Studies. The primary focus of her research is first language acquisition. In particular, she looks at their children’s developing understanding of what structures are possible in their language, how those structures are interpreted, and how they may be used in conversation. Starting in Fall 2019, she will be an Assistant Professor at MIT’s Department of Linguistics.
Abstract: The overall conveyed meaning of an utterance is a conglomerate of inferences, distinguishable both in how they arise (their semantics) and how they can be appropriately used in conversation (their pragmatics). In this talk, I focus on the distinction between asserted content, the main new information conveyed by an utterance, and presuppositions, background information that is taken for granted. For example, the utterance “I ate an apple, too” conveys, as part of its asserted meaning, that the speaker ate an apple, and presupposes that something else has been eaten. These meaning components are not explicitly labeled as such in ordinary conversation. How does a child learning language learn to distinguish between them and identify how they are factored into the overall interpretation?
In a series of behavioral experiments, I use the conversational principles governing assertion and presupposition to probe children’s ability to distinguish between these two layers of meanings (e.g. Stalnaker 1974, Grice 1975). Cooperative speakers adhere to different rules when asserting vs. presupposing something. It is uncooperative to assert something that your listener already knows. On the flip side, it is uncooperative to presuppose something your listener doesn’t already know. I show that 4-to-6-year-olds have adult-like biases about the knowledge state of the listener depending on the presupposed and asserted content of a speaker’s statement.