An Analysis of Cohesion and Coherence in Analytical Exposition Texts Generated by AI Chatbot Auto-generative System

Abstract

The advancement of technology can either nurture the learning or bring setbacks to the education. One of the results of the technology advances is the development of Artificial Intelligence (AI) in which it is used world-widely to help learners answer their learning problem through an auto-generative system known as ChatGPT. This study was aimed at investigating the quality of cohesion and coherence in analytical exposition texts made by AI Chatbot Auto-generative System. Cohesion and coherence play a significant role in the clarity and the comprehensibility of a written text and this study evaluated the cohesive device and coherence of the texts it generates. As many as nine texts produced by AI Chatbot were employed in this qualitative case study. To determine the coherence and cohesion of the texts, the analysis of the texts employed systemic functional linguistics (SFL), with a particular emphasis on the schematic structure and linguistic features focused on elements that contribute to the cohesion of the texts, including Theme progression and cohesive devices. The findings of this study indicate that the AI Chatbot has the ability to produce analytical exposition texts through various themes in terms of cohesion and coherence. It is also found five cohesive devices that contribute to producing cohesive text such as reference, lexical, conjunction, substitution, and ellipsis. To create coherence especially at the sentence level, the texts that it produced also employ two types of themes namely topical and textual themes, and thematic development consistency that involve the constant and zigzag patterns. Based on the findings, it can be said that AI Chatbot such as ChatGPT can be used as the learning model in creating the written texts especially the analytical exposition texts for the EFL classroom context.