June Choe will be defending their dissertation proposal on Monday, May 13 at 11:00am. The defense will take place in person in the Linguistics department library, and over Zoom. All are welcome to attend!

The proposal document can be found here, and the abstract is included below.


Title: Acquiring words beyond the basic level

Supervisor: Anna Papafragou

Proposal committee: Charles Yang (chair), Kathryn Schuler, Florian Schwarz

Time: May 13 (Monday), 11:00am-12:30pm

Place: Linguistics department library, 3401C Walnut St, Suite 300 and Zoom.

Early word learning is characterized by mapping labels for objects to the so-called “basic” level of description, such as dog. This basic-level bias presents a challenge for the eventual acquisition of words that differ in semantic specificity, such as the narrower subordinate-level (e.g., dalmatian) and the broader superordinate-level (e.g., animal) meanings. Previous literature has typically treated the acquisition of non-basic-level nouns in separate and asymmetric ways. Subordinate nouns are often characterized in terms of the perceptual distinctiveness of their referents and thus their meanings are thought to be acquired via the convergence of physical cues from the world. By contrast, superordinate nouns are thought to be initially difficult due to the unavailability of the corresponding concepts but trivially acquired via induction upon further conceptual maturity in later childhood. Of course, such views have not gone unchallenged – a parallel line of work has reported strong contributions from the language input in overcoming the basic-level bias in acquisition, raising questions about the scope of the perceptual and conceptual challenges to learning words beyond the basic level.

Building on this linguistic perspective, in this dissertation we develop a pragmatically-motivated framework that unifies the seemingly disparate challenges to the acquisition of subordinate and superordinate nouns. I propose that hierarchical noun meanings form a scale on the basis of their pragmatic informativeness; this position predicts that the acquisition of these words is guided by the accessibility and relevance of specific linguistic-semantic alternatives that constrain the generalization of word meaning. In a series of studies, we test the consequences of this pragmatic position in the acquisition and interpretation of subordinate and superordinate nouns. Beginning with subordinates, in Study 1 we show that semantic alternatives modulate subordinate-level conjectures in adults, and in Study 2 we plan to test whether children can generate similar inferences in word learning, on the basis of contextual factors such as speaker knowledge. Turning to superordinates, in Study 3 we show that the distribution of superordinate nouns in the input both reflects, and is discoverable from, semantic-pragmatic properties of these words. Then, in Study 4 we plan to investigate this mapping problem as experienced in children, probing and leveraging the circumstances under which children expect the superordinate-level to be talked about. In conclusion, we show that the acquisition of words at different levels of specificity is in large part characterized by the task of reasoning about a speaker’s intent in their choice of a word among other alternatives. This dissertation furthers our understanding of the acquisition of words at and beyond the basic-level, and the development of adult-like pragmatic competence that governs the unspoken rules of their usage.