Conversational Implicature in Chatbot Interactions: A Relevance Theory Study among University Students in Lahore

Abstract

As AI chatbots become embedded in everyday academic and social life, university students routinely draw implied meaning from machine-generated responses despite chatbots' lack of genuine communicative intention. This study examines how university students in Lahore derive conversational implicature from interactions with ChatGPT and Google Gemini, guided by Relevance Theory (Sperber & Wilson, 1995; Wilson & Sperber, 2004), which holds that hearers seek interpretations yielding adequate contextual effect for least processing effort. Data were collected through elicited chatbot transcripts, five prompts per chatbot per participant, and semi-structured interviews, and analyzed using a relevance-theoretic coding framework distinguishing strong implicature, weak implicature, and contextual assumption. The population comprised approximately 1,200 students from the University of the Punjab, the University of Management and Technology, and the National University of Modern Languages, Lahore, from which a stratified purposive sample of 21 students, 7 per institution respectively, was drawn. Findings indicate that strong implicature predominated in academic prompts, arising largely from evaluative silence, while weak implicature and contextual assumption were more prevalent in social and ambiguous-request exchanges. The study offers a relevance-theoretic classification of student-derived implicature types and contributes new empirical insight to pragmatics, human-computer interaction, and Pakistani higher education discourse studies.