2024, Vol.25, No.

Modern English Education 2023 KCI Impact Factor : 1.55

pISSN : 1598-0782 / eISSN : 2586-6141
https://journal.kci.go.kr/meeso

Within the emergentist framework, which posits simple processors such as structurebased sentence-level processor and pragmatic processor at the heart of language processing, English reflexive pronoun interpretation by native Korean learners of L2 English is examined using the same test sentences in two different conditions: 1) within a single isolated sentence; and 2) within a larger story-based context. Results demonstrated that highly proficient learners tended to process English reflexives using sentence-based processor in the first condition as native speakers would, whereas the pragmatic processor seemed to interfere in the second condition leading to illicit interpretation. On the other hand, low proficiency learners did not show much difference in their English reflexive interpretations under the two conditions. Thus, contextual information influenced the way L2 English reflexive pronouns were interpreted and the degree of influence seemed to depend on L2 English proficiency levels. This observation on the development process of L2 English reflexives suggests that L2 learners ultimately settle on the sentence-level processor at the expense of a weakening pragmatic processor after starting out as competition between the two processors. Implications for classroom instruction involving English reflexive pronouns are also suggested.

This study explored the impact of ChatGPT on L2 academic writing by focusing on the use of conjunctive adjuncts (or “conjunctions,” e.g., and, but, since, however, as a result, on the other hand) among L1-Korean college students. A mixed-methods approach was adopted to compare a treatment group interacting with ChatGPT (n = 36) with a control group interacting with peers (n = 40) during instructional activities designed to teach these structures’ usage. Quantitative results indicated no significant difference in the frequency or types of conjunctions used between the two groups, suggesting that ChatGPT’s influence on conjunction usage patterns was minimal. However, a manual analysis revealed that the treatment group demonstrated more formal and structured uses of conjunctions, aligning with academic writing norms, while the control group more frequently deviated from these norms, by, for instance, starting sentences with coordinating conjunctions and using incorrect structures such as sentence fragments. While this study had a small sample size and a narrow focus on only one aspect of grammar, its results suggested a need to more comprehensively assess AI’s potential to complement traditional language learning methods.