Rhetorical Move Analysis of Human vs AI-Generated Research Abstracts: A Corpus-Based Study
DOI:
https://doi.org/10.70670/sra.v4i2.2180Keywords:
Research Abstracts, Artificial Intelligence, Academic Writing, Genre Analysis, Rhetorical Move Structure, Hylands 5-Move Framework, AI, Corpus Linguistics, Academic DiscourseAbstract
This study looked at the differences between research abstracts written by humans and those created by intelligence. The goal was to see how well artificial intelligence can mimic the way humans write in settings. In today's world of publishing, abstracts are very important because they help decide what gets published. While tools like ChatGPT are good at writing text that sounds nice, they often do not understand the social context of what they are writing about. This means they can produce text that sounds too perfect and formulaic. To study this, a collection of research abstracts called the Human vs. AI Rhetorical Abstract Corpus was made. It had 80 abstracts, 40 written by humans and 40 created by intelligence using ChatGPT. The human abstracts came from known journals about applied linguistics. The artificial intelligence abstracts were created using the titles, keywords, and structure of the human abstracts. Each sentence in the abstracts was looked at. Categorized using a framework developed by Ken Hyland. This framework has five parts: introduction, purpose, method, product, and conclusion. The results showed that both human and artificial intelligence abstracts usually had a purpose and product section. However, there were differences in how often they had an introduction and a conclusion. The artificial intelligence model always included these sections but human writers did not always use them. This was because human writers wanted to make their abstracts shorter. These differences were statistically significant, which means they were not just random. The artificial intelligence abstracts also followed a strict pattern, while the human abstracts did not. This shows that artificial intelligence is not good at adapting to contexts and produces text that is too rigid. Journal editors can use these findings to tell if an abstract was written by an artificial intelligence. Future studies should look at abstracts from different fields to see if the same patterns hold true.
References
Anthony, L. (2024). AntConc (Version 4.2.4) [Computer Software]. Laurence Anthony's Lab.
Biber, D., & Egbert, J. (2024). Register, genre, and style: Multidimensional perspectives on text varieties (2nd ed.). Cambridge University Press.
Crossley, S. A., & McNamara, D. S. (2020). Moving beyond traditional linguistic indices: Assessing lexical diversity and text quality in L2 writing. Assessing Writing, 43, Article 100441. https://doi.org/10.1016/j.asw.2020.100441
Egbert, J., Larsson, T., & Biber, D. (2022). Doing corpus linguistics: Methodological considerations and practical applications. Routledge.
Flowerdew, J., & Mahlberg, M. (Eds.). (2021). Corpus linguistics and writing development: Perspectives from learner corpora. John Benjamins.
Hyland, K. (2000). Disciplinary discourses: Social interactions in academic writing. Longman.
Kyle, K., Crossley, S. A., & Jarvis, S. (2021). Assessing the validity of computational lexical diversity metrics in second language writing. Language Learning, 71(3), 677–715. https://doi.org/10.1111/lang.12444
Lund, B. D., & Agbese, M. (2023). Chatting about ChatGPT: How may AI affect academic writing? Library Hi Tech News, 40(3), 27–30. https://doi.org/10.1108/LHTN-01-2023-0009
McEnery, T., & Hardie, A. (2012). Corpus linguistics: Method, theory and practice. Cambridge University Press.
Melander, B., Swales, J. M., & Fredrickson, K. (1997). Journal abstracts from three academic fields. Text & Talk, 17(2), 251–274. https://doi.org/10.1515/text.1.1997.17.2.251
Nation, I. S. P. (2022). Learning vocabulary in another language (3rd ed.). Cambridge University Press.
Rudolph, J., Tan, S., & Tan, S. (2023). ChatGPT: Bullshit spewer or the end of traditional assessments in higher education? Journal of Applied Learning & Teaching, 6(1), 1–22. https://doi.org/10.37074/jalt.2023.6.1.9
Samraj, B. (2005). An exploration of a genre set: Research article abstracts and introductions. English for Specific Purposes, 24(2), 141–156. https://doi.org/10.1016/j.esp.2003.08.004
Susnjak, T. (2022). ChatGPT: The end of online exam integrity? arXiv preprint arXiv:2212.09292. https://doi.org/10.48550/arXiv.2212.09292
Swales, J. M. (1990). Genre analysis: English in academic and research settings. Cambridge University Press.
Swales, J. M. (2004). Research genres: Explorations and applications. Cambridge University Press.
