Semester of Graduation

Spring 2024


Master of Arts (MA)



Document Type



Previous research on language acquisition has raised two competing hypotheses: Starting Small and Starting Big. Starting Small states that limitations (small working memory capacity, simple structure, short string length, etc.) are beneficial for language learning, which explains why children are able to acquire language more easily than adults. Starting Big proposes the opposite, that input of chunks of language rather than individual units creates a stronger and more fluid understanding of grammatical relations between words. The current study aimed to investigate these theories in the context of an artificial language paradigm using a forced-choice task, a production task, a working memory task, and a phonological memory task. The participants were trained in the artificial language in 1 of 5 possible segmentation conditions to manipulate input segmentation and across three types of grammar manipulations (subject-verb agreement, grammatical gender, and article contrast) to understand how the type of grammatical input may relate to these theories. Our results show that there may be benefits to both learning strategies. Starting Big, learning in chunks of language, was helpful with memorization, but could create frozen learning of some grammar structures that could not be generalized. However, Starting Small, breaking language into individual components, was the most beneficial for generalization because it creates a more fluid learning of grammar relations and prevents frozen forms. Overall, Starting Small may be the most helpful for the learning of natural language, which has implications for the way that children process language, and the way that second languages are taught to adults.



Committee Chair

McDonald, Janet

Available for download on Friday, April 04, 2025