Elamur Research⌕
☰
home / papers / 2025 / Generative AI and linguistic diversity in academic writing and publishing: Perspectives from World Englishes

Generative AI and linguistic diversity in academic writing and publishing: Perspectives from World Englishes

Paper · 2025 · Kingsley Ugwuanyi, Christian Mair, Sender Dovchin et al.

Summary

This article examines whether generative AI in academic writing and publishing (AWP) widens or narrows room for the many Englishes used in scholarship. It reports no experiment. Instead, a convenor (Ugwuanyi) put five guiding questions to five sociolinguists working in or near World Englishes: Mair, Erdocia, Dovchin, Kuteeva and doctoral researcher Airemionkhale. The questions covered GenAI's overall influence, bias toward dominant varieties, institutional duties, peer review and ethics. Their answers are printed unedited, and the convenor adds commentary on where they agree and differ. The shared conclusion is that LLMs trained mostly on standard, often American, English tend to produce a single uniform register. That pushes scholarly prose toward Anglophone conventions and risks treating other varieties as mistakes. The contributors also see openings: tools can demystify hidden genres, and writers can rework outputs to keep their own voice. They call for equity-minded journal and university policies, critical AI literacy, and co-design of tools with diverse communities. The piece matters because it links AI governance to the question of which Englishes count as legitimate.

Key points

  • The paper is a structured scholarly dialogue organised around 5 questions. It appeared in the Journal of English for Research Publication Purposes, 6(2), 265–289.
  • Contributors bring named concepts into the discussion:
    • Mair's "World System of Englishes"
    • Dovchin's "algorithmic colonialingualism," together with her view of GenAI as double-edged
    • Kuteeva & Andersson's (2024) argument that LLMs drive scholarship toward standardisation and convergence
  • Contributors report a generational gap in uptake. Older scholars use GenAI little or not at all, while junior scholars sometimes treat it as a co-researcher. According to a survey they cite, people over 55 used AI at only half the rate of those over 35.
  • Airemionkhale describes avoiding words such as "delve" and "underscore" because reviewers may read them as signs of AI authorship. This puts pressure on human writers to avoid machine-like style.
  • Mair names a "standardisation paradox": English becomes more homogeneous and more heterogeneous at once. The authors also acknowledge a limitation of their own panel: every contributor is based at a Global North institution.

Interesting ideas

  • I1. Departures from a dominant standard should be read as evidence of a legitimate variety, not as errors to fix, because "correcting" them quietly erases real usage.
  • I2. Mair limits machine translation to languages he can judge himself, using a tool's output only where a competent human can check it.
  • I3. Overused LLM words like "delve" are fingerprints of machine text, so heavy model output can shift the lexical profile of a whole corpus.
  • I4. The standardisation paradox shows that a pipeline can make text more uniform even while appearing to accommodate variety. Systems should therefore measure how much voice survives processing.
  • I5. LLM advice tends to encode Anglophone rhetorical conventions as if they were universal. Explicitly marking whose norms a resource reflects makes that bias visible.

For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

  • I1 (tree): The parser and UD conversion could tag nonstandard but systematic constructions as variety-specific instead of forcing them into the closest standard analysis, keeping the tree faithful to what the writer actually wrote.
  • I2 (learn): LLM teacher tiers could be used only on phenomena where witness parsers or negative controls can verify the labels, and left out of areas where no independent check exists.
  • I3 (learn): Silver data from LLM teachers should be checked for drift in word frequencies against human-written text, so these stylistic tics do not leak into lemma or frame statistics.
  • I4 (back): The text → graph → text compressor could be evaluated on how well it preserves dialectal and register markers in regenerated text, not only on whether the propositions survive.
  • I5 (frames): Speech-act and implicature rules could record which cultural or rhetorical convention they assume, so explanations of politeness or indirectness name their norm instead of presenting it as universal.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓Carnegie Mellon UniversityPittsburgh · United States539 works · 179,533 citations🎓Hamad bin Khalifa UniversityDoha · Qatar141 works · 14,175 citations🎓Dublin City UniversityDublin · Ireland70 works · 9,447 citations🎓Macquarie UniversitySydney · Australia42 works · 5,776 citations🎓Stockholm UniversityStockholm · Sweden21 works · 1,832 citations

Also on the byline: Curtin University.

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Arts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 worksLiterature and Literary Theory · 519 worksHistory and Philosophy of Science · 53 works

🧭 Topics 1Research subjects and languages the work is about.English · 2,263 works

Cite

Cite as Kingsley Ugwuanyi, Christian Mair, Sender Dovchin, Iker Erdocia, Maria Kuteeva and Esther Airemionkhale. 2025. Generative AI and linguistic diversity in academic writing and publishing: Perspectives from World Englishes. arXiv:2607.28505 · Generative AI and linguistic diversity in academic writing and publishing: Perspectives from World Englishes. Journal of English for Research Publication Purposes, 2025, 6(2), 265-289.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.1075/jerpp.00035.ugw · arXiv 2607.28505

@misc{ugwuanyi2025generative,
  title = {{Generative AI and linguistic diversity in academic writing and publishing: Perspectives from World Englishes}},
  author = {Kingsley Ugwuanyi, Christian Mair, Sender Dovchin, Iker Erdocia, Maria Kuteeva and Esther Airemionkhale},
  year = {2025},
  doi = {10.1075/jerpp.00035.ugw},
  eprint = {2607.28505},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.1075/jerpp.00035.ugw}
}

References

40 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Agarwal, D., Naaman, M., & Vashistha, A. (2025). AI suggestions homogenize writing toward Western styles and diminish cultural nuances. In Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems (pp. 1-21). find
  2. Bender, E.M., Gebru, T., McMillan-Major, A. & Shmitchell, S. (2021). On the dangers of stochastic parrots: can language models be too big? In Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency (FAccT ‘21). Association for Computing Machinery, NY, USA, 610–623. doi:10.1145/3442188.3445922 ↗
  3. Bisht, G. (2023). Contradictions of decolonial literacy in an ESL writing context. In C. Chinn, F. Kaplan, & A. Wise (Eds.), Proceedings of the 15th International Conference of the Learning Sciences (ICLS 2023) (pp. 784–791). International Society of the Learning Sciences. https://repository.isls.org/bitstream/1/10329/1/ICLS2023_784-791.pdf ↗
  4. Blommaert, J. (2010). The sociolinguistics of globalization. Cambridge University Press. ↗
  5. Botes, E., Dewaele, J. M., Colling, J., & Teuber, Z. (2025). Initial indications of generative AI writing in linguistics research publications (preprint). PsyArXiv. https://doi.org/10.31234/osf.io/4yvbp_v1 ↗
  6. Canagarajah, S. (2024). Decolonizing academic writing pedagogies for multilingual students. TESOL Quarterly, 58(1), 280-306. ↗
  7. ChatGPT. (2025, May 6). Please help me rewrite this in Nigerian English [Large language model]. OpenAI. Personal communication. find
  8. Crawford, K. (2021). The atlas of AI: Power, politics, and the planetary costs of artificial intelligence. Yale University Press. find
  9. Dai, D. W., & Hua, Z. (2025). When AI meets intercultural communication: new frontiers, new agendas. Applied Linguistics Review, 16(2), 747–751. ↗
  10. de Fine Licht, K. (2025). Rethinking the ethics of GenAI in higher education: A critique of moral arguments and policy implications. Journal of Applied Philosophy. Early view: https://doi.org/10.1111/japp.70026 ↗
  11. Dovchin, S. (2024). Afterword: Artificial intelligence in Applied Linguistics: A double-edged sword. Australian Review of Applied Linguistics, 47(3), 410-417. find
  12. Erdocia, I., Migge, B., & Schneider, B. (2025). Language is not a data set—Why overcoming ideologies of dataism is more important than ever in the age of AI. Journal of Sociolinguistics, 28(5): 20-25. ↗
  13. European Commission (2019). High-level expert group on artificial intelligence. Ethics Guidelines for Trustworthy AI. European Commission. find
  14. Faisal, F., & Anastasopoulos, A. (2024). An efficient approach for studying cross-lingual transfer in multilingual language models. arXiv preprint arXiv:2403.20088. ↗
  15. Fleisig, E., Smith, G., Bossi, M., Rustagi, I., Yin, X., & Klein, D. (2024). Linguistic Bias in ChatGPT: Language models reinforce dialect discrimination. arXiv preprint arXiv:2406.08818 ↗
  16. Fricker, M. (2007). Epistemic injustice: Power and the ethics of knowing. Oxford University Press. ↗
  17. Graham, P. [@paulg]. (2024, April 7). Someone sent me a cold email proposing a novel project. Then I noticed it used the word “delve.” [Tweet]. X. https://x.com/paulg/status/1777030573220933716 ↗
  18. Grammarly. (2025, April 9). Decoding AI language: Common words and phrases in AI-generated content. Grammarly. https://www.grammarly.com/blog/ai/common-ai-words/ ↗
  19. Hofmann, V., Kalluri, P. R., Jurafsky, D., & King, S. (2024). AI generates covertly racist decisions about people based on their dialect. Nature, 633(8028), 147-154. ↗
  20. Jones, R. (2025). Culture machines. Applied Linguistics Review, 16(2), 753–762 find
  21. Khalifa, M., & Albadawy, M. (2024). Using artificial intelligence in academic writing and research: An essential productivity tool. Computer Methods and Programs in Biomedicine Update, 5, 1-11. ↗
  22. Kuteeva, M. & Andersson, M. (2024). Diversity and standards in writing for publication in the age of AI – between a rock and a hard place. Applied Linguistics, 45(3), 561–567. ↗
  23. Mair, C. (2013). The World System of Englishes: Accounting for the transnational importance of mobile and mediated vernaculars. English World-Wide, 34(3), 253-278. find
  24. McKinley, J., & Rose, H. (2018). Conceptualizations of language errors, standards, norms and nativeness in English for research publication purposes: An analysis of journal submission guidelines. Journal of Second Language Writing, 42, 1-11. ↗
  25. Milmo, D. (2025, June 14). Workers in UK need to embrace AI or risk being left behind, minister says. The Guardian. https://www.theguardian.com/technology/2025/jun/14/workers-in-uk-need-to-embrace-ai-or-risk-being-left-behind-minister-says ↗
  26. Moorhouse, B. L., Consoli, S., & Curle, S. M. (2025). Generative AI and the future of writing for publication: Insights from applied linguistics journal editors. Applied Linguistics Review, 16(6), 2721–2747 ↗
  27. Nguyen, J. [@JeremyNguyenPhD]. (2024, March 30). Are medical studies being written with ChatGPT? Well, we all know ChatGPT overuses the word “delve”… [Tweet]. X. https://x.com/JeremyNguyenPhD/status/1774021645709295840 ↗
  28. Ofosu-Asare, Y. (2025). Cognitive imperialism in artificial intelligence: Counteracting bias with indigenous epistemologies. AI & Society, 40, 3045–3061. ↗
  29. Payne, A. L., T. Austin, & A. M. Clemons (2024). Beyond the front yard: The dehumanizing message of accent-altering technology. Applied Linguistics, 45(3), 553-560. ↗
  30. Pérez-Milans, M., Kaufhold, K., & Soler, J. (2025). English and academic publishing: Capitalist endeavours, colonial entanglements, and knowledge production. In J. Soler & K. Kaufhold (Eds.), Language and the knowledge economy: Multilingual scholarly publishing in Europe (pp. 195–209). Routledge. find
  31. Phillipson, R. (1992). Linguistic imperialism. Oxford University Press. find
  32. Piller, I. (2022). What exactly does an editor do? Multilingua, 41(6), 629-637. ↗
  33. Pinhanez, C. S., & Wornyo, E. (2025). Ethical Co-Development of AI Applications with Indigenous Communities. In Proceedings of the Extended Abstracts of the CHI Conference on Human Factors in Computing Systems (pp. 1-4). https://dl.acm.org/doi/abs/10.1145/3706599.3706649 ↗
  34. Qijia, Z., & Shikui, Z. (2025). Exploring the impact of AI on research practices: A postdigital ethnographic study of doctoral students’ engagement with ChatGPT. Studies in Continuing Education, 1–17. Early View: https://doi.org/10.1080/0158037X.2025.2535565 ↗
  35. Saulière, J. (2014). Corporate language: The blind spot of language policy? Reflections on France's Loi Toubon. Current Issues in Language Planning, 15(2), 220-235. find
  36. Soler, J., Erdocia, I., & Savski, K. (2023). (Im)possible change: Criticality and constraints in the infrastructures of the academic knowledge economy. Language, Culture and Society, 5(2), 167-181. find
  37. Tafazoli, D. (2024). Exploring the potential of generative AI in democratizing English language education. Computers and Education: Artificial Intelligence, 7, 100275. https://doi.org/10.1016/j.caeai.2024.100275 ↗
  38. Tupas, R. (Ed.). (2015). Unequal Englishes: The politics of Englishes today. Springer. find
  39. Weaver, K. D. (2024). The Artificial Intelligence Disclosure (AID) framework: An introduction. arXiv preprint arXiv:2408.01904 ↗
  40. Weber, M. (1978). Economy and society: An outline of interpretive sociology. University of California Press. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/ugwuanyi-2026-generative-ai-linguistic-diversity-academic/