Researcher ORCID Identifier

0000-0002-1056-4403

Graduation Year

2022

Document Type

Campus Only Senior Thesis

Degree Name

Bachelor of Arts

Department

Linguistics

Reader 1

Robin Melnick

Reader 2

Michael Diercks

Reader 3

Galia Bar Sever

Terms of Use & License Information

Terms of Use for work posted in Scholarship@Claremont.

Rights Information

© 2022 Emily M Clarke

Abstract

The structure and nature of Bananagrams provides an interesting and complicated illustration of linguistic cognitive processes. It’s speed forces unconscious and instinctive word choices, a combination of lexical access and orthographic processing. While the crossword product and dynamic gameplay make Bananagrams a challenging scenario to study, some of the interesting tasks presented by word recognition and selection within a large set of unordered letters while under time pressure can be investigated on a smaller scale. This project attempts to do so and by analyzing the words created and the order in which they are, given a large unordered set of letters, investigates the question: what are the interaction effects between semantic relation, word frequency, and lexical similarity in a lexical item production task? To investigate this question, I designed a novel lexical access task, and semantic, frequency, and lexical similarity scoring mechanisms. 51 Claremont College students were asked to play a word game three times. The word game involved making words as they saw them as quickly as possible out of a set of 30 letter tiles. Participant data was scored using semantic word embeddings, COCA frequency data, and Levenshtein distance. The data was then compared to the scores of 100,000 randomly generated word sets. Using this scoring mechanism, I found a strong faciliatory effect of both frequency and lexical similarity but a slight inhibitory effect of semantic similarity.

Download

This thesis is restricted to the Claremont Colleges current faculty, students, and staff.

Share

COinS