Dissertations, Theses, and Capstone Projects

Date of Degree

2-2025

Document Type

Thesis

Degree Name

M.A.

Program

Linguistics

Advisor

Kyle Gorman

Keywords

Language Acquisition, Syntactic Acquisition, Principles and Parameters, Computational Linguistics, Generative Grammar, Psycholinguistics

Abstract

This paper investigates the effects of sententially irrelevant syntactic information on language acquisition models and introduces a new variant of the variational learner (VL) called the early decision variational learner (EDVL). EDVL accelerates convergence by locking in parameter values once they reach a threshold, in contrast to VL’s continued adjustments until all parameter values reach threshold. Both learners are grounded in the generative approach to language acquisition. We examine both learners' responses when they only update their parameter values when the input contain the relevant information to determin the parameter value such as prepositions, wh-words, and auxiliary verbs. Chapter 1 provides some theoretical background on language acquisition and parameter setting algorithms, while Chapter 2 details the VL and EDVL models and results. Chapter 3 explores the impact of sentential irrelevance on learning, followed by two case studies, and Chapter 4 concludes the study with implications for language acquisition studies.

Share

COinS