Author ORCID Identifier

https://orcid.org/0000-0002-4805-7984

Date of Award

8-8-2024

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Applied Linguistics and English as a Second Language

First Advisor

YouJin Kim

Second Advisor

Eric Friginal

Third Advisor

Ute Römer-Barron

Fourth Advisor

Marije Michel

Abstract

Despite recent first language (L1) human-computer interaction alignment research (e.g., Branigan et al., 2010), there remains a paucity of second language (L2) research on alignment during learner-computer interaction. Therefore, it is unclear whether linguistic alignment is operative between L2 learners and Artificial Intelligence (AI) chatbots, and further whether alignment promotes L2 development. In addition, in previous alignment research, primes (model examples) and recast primes have been identified as sources of alignment, with primes predominantly having been examined. L2 researchers have claimed that learners need both positive and negative evidence for successful L2 acquisition (Gass, 1997). Accordingly, receiving only correct forms might not always facilitate L2 development; however, receiving recast primes, which involve both positive and negative evidence, might promote L2 learning more effectively (McDonough & Mackey, 2008). The goal of this dissertation is three-fold: (1) to compare the effects of two sources of alignment on the occurrence of syntactic and pragmatics alignment during L2 learner-chatbot interactions and further on L2 English grammar and pragmatics development; (2) to examine the influence of learner-related factors on linguistic alignment and alignment-driven learning; (3) to explore how L2 learners perceive chatbot-based interaction for their language learning. 84 Korean English as a foreign language (EFL) college students were assigned to one of three conditions: prime, recast prime, and control. Participants performed three chatbot-based tasks in their condition and took a pretest, two posttests, a C-test, language aptitude tests, and chatbot perception surveys over four online sessions. The results indicated that linguistic alignment is present during L2 learner-chatbot interaction to varying degrees, depending on the target feature, and that alignment in general promoted L2 grammar and pragmatics development. For the alignment sources, recast primes were found to be more beneficial in promoting alignment and alignment-driven learning. Furthermore, the findings suggested that language aptitude, prior knowledge, L2 proficiency, and the amount of target feature production mediated the role of alignment in L2 grammar and pragmatics development with differing degrees of influence. Overall, learners had positive perceptions of chatbots; however, they were split when it comes to their preference for interaction partners between human interlocutors and chatbots. Implications for the role of linguistic alignment during chatbot-based interaction for L2 development are discussed.

DOI

https://doi.org/10.57709/37129935

File Upload Confirmation

1

Available for download on Saturday, May 30, 2026

Share

COinS