Publication Date

Spring 5-6-2026

School

College of Arts and Sciences

Major

TESL

Keywords

AI, writing center, education, linguistics, error correction, feedback

Disciplines

Bilingual, Multilingual, and Multicultural Education | Educational Technology | Higher Education

Abstract

This study examines how the large language model (LLM) ChatGPT can support teachers of Turkish learners of English as a Second Language (ESL) in addressing one of the most persistent cross-linguistic challenges in academic writing: the misuse of the definite article “the.” Because Turkish lacks a grammatical article system entirely, Turkish learners frequently omit or misapply the definite article in ways that reflect deeper structural differences between English and Turkish rather than simple errors. Writing samples were elicited from two proficiency levels of English in Turkish university English courses and analyzed through ChatGPT to determine whether AI could accurately diagnose the linguistic roots of article misuse and generate pedagogically useful information for teachers. Analysis of chat transcripts and AI-generated responses showed that ChatGPT performed somewhat reliably in identifying surface-level article errors and offered explanations that were largely accurate. However, it showed notable inconsistencies when distinguishing between article errors and article use in idiomatic phrases, and its feedback often lacked the metalinguistic depth needed for sustained use. These findings suggest that ChatGPT can serve as a useful supplementary tool for TESL (Teaching English as a Second Language) educators at the university, particularly for initial error identification and categorization, but will require critical scaffolding by TESL practitioners to compensate for gaps in precision and reliability.

Download

Share

COinS