LEARNING LANGUAGE DISTINCTIONS THROUGH COMPETITION

Abstract

Successful use of language requires that one learns which linguistic formulations are conventional and which are not. In my dissertation, I investigate the role of competition as a key mechanism in learning these language distinctions. In chapter 2, I describe how I used an artificial language paradigm to find that competition leads to the suppression of generalization for exception items over the course of learning. In an ongoing fMRI experiment using the same paradigm, I aim to characterize the neural representation of generalizations and exceptions and how they may change over the course of learning. In chapter 3, I investigate the role of competition in learning unacceptability or learning what not to say in a new language for adult learners. I find that adult language learners have a specific difficulty learning unacceptability across different languages and populations. However, I also find that competition between conventional formulations and unconventional formulations can lead language learners to appropriately recognize unconventional formulations as unacceptable. In chapter 4, I further investigate how competition between items works in language learners, using a memory task. I find that compared to native speakers, they display less effect of competition between related memories. Together, these findings explain how competition is used to learn language distinctions at multiple linguistic levels, from words to sentences. I use both a neural method (fMRI) and multiple behavioral methods (mouse-tracking, acceptability judgment, memory recognition, self-paced reading, and memory retrieval) with various populations and languages (native English speakers, native Spanish speakers, nonnative English speakers of over 70 native languages, and Spanish learners in the classroom) to study the effect of competition on learning language distinctions.