التعليمية
Volume 15, Numéro 3, Pages 1034-1048
2025-12-26

Students’ Use Of Compliments In Response To Chatgpt Output: A Case Study

Authors : Khelloufi Abdennour . Hammoudi Abdelhak .

Abstract

The incorporation of Large Language Models (LLMs) such as ChatGPT into our daily interactions has opened up new frontiers for research within the field of pragmatics. There has been some recent research exploring the pragmatic awareness of AI. Still, the production of compliment speech acts directed towards these non-human interlocutors remains a significant gap in the literature. As a means of addressing this gap, this study examines how Algerian EFL university students respond to ChatGPT output through the speech act of complimenting. This study adopted a mixed-method approach to collect data from 31 fifth-year university students at the École Normale Supérieure of Setif, Algeria, by using an online questionnaire and a Discourse Completion Task (DCT). A systematic analysis of the questionnaire data reveals the frequency, contexts, and motivations for complimenting ChatGPT, while a systematic analysis of the DCT data is undertaken using Yuan’s (2002) framework of compliment strategies. The major findings show that compliments are frequent and are motivated primarily by performance and accuracy of the AI's output rather than socio-relational objectives. In addition, the analysis of the compliment strategies found a significant preference for compliments to be expressed through direct, explicit, and formulaic expressions. It appears that a functional and formal adaptation of the compliment speech act in human-AI communication occurs, in which the absence of interpersonal face-work results in a simplification of pragmatic strategies. Therefore, this study contributes to the fields of pragmatics and human-AI interaction by providing an empirical analysis of a core complimenting behavior in a new communicative context.

Keywords

Pragmatics ; Compliments ; Speech Act Theory ; Human-AI Interaction ; ChatGPT