Chapter 4. Exploring the Potential of ChatGPT for Enhancing Pragmatic Instructions: Generating and Analyzing Conversation Closings
Yukie Saito
Abstract
Since the emergence of Generative AI, such as ChatGPT, educators have engaged in a global discussion about its advantages and disadvantages for educational purposes. Despite concerns about its use, English educators are exploring its potential integration. One application is creating pragmatically appropriate dialogues, such as conversation closings. Previous analyses highlighted shortcomings in ELT textbooks regarding conversation closings. Further study of the American TV show Friends and the Michigan Corpus of Academic Spoken English (MICASE) revealed more diverse and complex strategies not well represented in textbooks, though it is challenging for teachers to obtain this data. Closing a conversation appropriately, depending on the context, is as crucial as initiating it. English language learners often struggle with this aspect. In conversations, closings are often ritualized and more complex, with different steps than merely one realization of adjacency pairs. Major steps in closing a conversation are pre-closings, shutdowns, and terminal exchanges. To find better examples of conversation closings than those in textbooks, I generated conversations using ChatGPT. These were analyzed for pre-closings, shutdowns, and terminal exchanges, and compared to results from textbooks, Friends, and the MICASE. The analysis revealed that while the conversation closings created by ChatGPT include pre-closings, shutdowns, and terminal exchanges, the use of pre-closings and the variety of shutdowns used was limited compared to Friends and MICASE. Thus, it may be premature to rely on ChatGPT for generating conversation closings for pragmatic instruction, and English educators may still need to seek other resources, such as dramas and corpus data like MICASE. However, there is potential that with different prompts and contextual information about participants, ChatGPT could generate more diverse conversation closings. With the rapid advancement of Generative AI technology, including that of ChatGPT, English educators are encouraged to stay informed about its development to explore its potential for creating pragmatically appropriate conversations, including conversation closings.
About the Contributor
Yukie Saito obtained an M.A. in TESOL from Teachers College, Columbia University, an M.Sc. in SmartEdTech from Université Côte d’Azur, and a Ph.D. in Education from Temple University. She is a professor at the Faculty of Global Informatics, Chuo University. She coordinates an English course for first-year students and study abroad programs, and is in charge of a seminar focusing on EdTech and its application to English education. She has been interested in pragmatics, English teachers’ cognition and classroom practice, CEFR, and EdTech. Her main research interest is the application of Virtual Reality (VR) in English education.
Citation
Saito, Y. (2026). Exploring the potential of ChatGPT for enhancing pragmatic instructions: Generating and analyzing conversation closings. In R. Dykes, O. Edwards, D. Bollen, & T. S. W. Lin (Eds.), Artificial intelligence in Japan’s language learning classrooms (pp. 85–112). Candlin & Mynard. https://doi.org/10.47908/45/4
Since the emergence of Generative AI, such as ChatGPT, educators have engaged in a global discussion about its advantages and disadvantages for educational purposes. Despite concerns about its use, English educators are exploring its potential integration. One application is creating pragmatically appropriate dialogues, such as conversation closings. Previous analyses highlighted shortcomings in ELT textbooks regarding conversation closings. Further study of the American TV show Friends and the Michigan Corpus of Academic Spoken English (MICASE) revealed more diverse and complex strategies not well represented in textbooks, though it is challenging for teachers to obtain this data. Closing a conversation appropriately, depending on the context, is as crucial as initiating it. English language learners often struggle with this aspect. In conversations, closings are often ritualized and more complex, with different steps than merely one realization of adjacency pairs. Major steps in closing a conversation are pre-closings, shutdowns, and terminal exchanges. To find better examples of conversation closings than those in textbooks, I generated conversations using ChatGPT. These were analyzed for pre-closings, shutdowns, and terminal exchanges, and compared to results from textbooks, Friends, and the MICASE. The analysis revealed that while the conversation closings created by ChatGPT include pre-closings, shutdowns, and terminal exchanges, the use of pre-closings and the variety of shutdowns used was limited compared to Friends and MICASE. Thus, it may be premature to rely on ChatGPT for generating conversation closings for pragmatic instruction, and English educators may still need to seek other resources, such as dramas and corpus data like MICASE. However, there is potential that with different prompts and contextual information about participants, ChatGPT could generate more diverse conversation closings. With the rapid advancement of Generative AI technology, including that of ChatGPT, English educators are encouraged to stay informed about its development to explore its potential for creating pragmatically appropriate conversations, including conversation closings.
About the Contributor
Yukie Saito obtained an M.A. in TESOL from Teachers College, Columbia University, an M.Sc. in SmartEdTech from Université Côte d’Azur, and a Ph.D. in Education from Temple University. She is a professor at the Faculty of Global Informatics, Chuo University. She coordinates an English course for first-year students and study abroad programs, and is in charge of a seminar focusing on EdTech and its application to English education. She has been interested in pragmatics, English teachers’ cognition and classroom practice, CEFR, and EdTech. Her main research interest is the application of Virtual Reality (VR) in English education.
Citation
Saito, Y. (2026). Exploring the potential of ChatGPT for enhancing pragmatic instructions: Generating and analyzing conversation closings. In R. Dykes, O. Edwards, D. Bollen, & T. S. W. Lin (Eds.), Artificial intelligence in Japan’s language learning classrooms (pp. 85–112). Candlin & Mynard. https://doi.org/10.47908/45/4
Information About the Book
|
Title: Artificial Intelligence in Japan’s Language Learning Classroom
Editors: Robert Dykes, Oliver Edwards, Dave Bollen, and Tina Shu-wen Lin Publication date: June 2026 Read more... |