The Linguistic Characteristics of ChatGPT-Assisted Academic Writing in Higher Education: A Lexico-Grammatical and Discourse-Analytical Perspective in EFL Contexts

Abstract

The accelerated integration of artificial intelligence-based tools into higher education has fundamentally reshaped academic writing practices, particularly within English as a Foreign Language (EFL) context.Among these technologies, ChatGPT has emerged as a widely adopted generative language model capable of supporting text production, revision, and organization at a level that closely resembles human academic writing.While existing research has largely emphasized pedagogical effectiveness, learner perceptions, and ethical concerns, comparatively limited attention has been devoted to the linguistic properties of ChatGPT-assisted academic texts.This conceptual and literature-based study examines the linguistic features of ChatGPT-assisted writing in higher education, with a specific focus on lexicogrammatical patterns, syntactic complexity, discourse organization, and stylistic tendencies observable in AI-supported academic texts.Drawing on recent empirical and theoretical studies in applied linguistics and educational technology, the article synthesizes current findings to explore how ChatGPT influences language production, coherence, stancetaking, and the expression of authorial voice in academic writing.The analysis suggests that ChatGPT-assisted texts tend to exhibit high levels of grammatical accuracy, cohesive discourse markers, and conventional academic structures, which may enhance clarity and formal appropriateness, especially for EFL learners.However, these advantages are accompanied by potential limitations, including stylistic homogeneity, reduced individuality, and a tendency toward formulaic expression.The study argues that ChatGPT does not replace human authorship but rather reshapes linguistic output through mediated interaction between human writers and AI systems.The article concludes by discussing the implications of these linguistic characteristics for academic writing instruction, assessment, and research, and calls for further corpus-based and discourse-analytical studies to better understand the long-term impact of AImediated writing on academic communication in higher education.

Ähnliche Arbeiten