2026-3

VSNiP

Problems of Contemporary Science and Practice. Vernadsky University

Вопросы современной науки и практики. Университет имени В.И. Вернадского

2541-853X 1990-9047

ФГБОУ ВО «Тамбовский государственный технический университет»

10.17277/voprosy.2026.03.pp.139-152

13382

Version of Record

Cross-Linguistic Interference in AI-Mediated Academic Writing Instruction

Межъязыковая интерференция в условиях ИИ-опосредованного обучения академическому письму

Gunina

N. A.

academic writing English academic writing large language models generative artificial intelligence AI-mediated learning cross-linguistic interference feedback learner autonomy.

академическое письмо англоязычное академическое письмо большие языковые модели генеративный искусственный интеллект ИИ-опосредованное обучение межъязыковая интерференция обратная связь языковая автономия.

2026

3 квартал 2026 года

3 3

2026

N. A. Gunina

The article examines the didactic potential and risks associated with using large language models (LLMs) in teaching academic writing in English. A key challenge in writing academic texts in a foreign language is cross-linguistic interference, which manifests at the lexico-semantic, grammatico-syntactic, discursive, and genre-rhetorical levels. Based on a theoretical analysis of domestic and international research, the concept of "cross-linguistic interference" is refined in the context of writing edited with the aid of large language models. The study demonstrates that LLMs can reduce surface-level linguistic errors, provide immediate feedback, and support the processes of editing and reflection. However, they may also preserve or mask source-language characteristics, reinforce formulaic academic writing, diminish learner autonomy, and create an illusion of overcoming linguistic challenges. The article substantiates the need to distinguish between formal linguistic correctness and the discursive adequacy of an academic text. A step-by-step procedure for using AI, comprising independent analysis of the source text, targeted prompting of the language model, verification of its recommendations, post-editing, and reflection, is proposed; it. It is concluded that the educational value of AI is determined not by the mere fact of its use, but by the organization of learning activities and the maintenance of learner responsibility for the content, structure, and linguistic presentation of the academic text.

Рассмотрены дидактические возможности и риски использования больших языковых моделей в обучении академическому письму на английском языке. Одной из проблем, которая возникает при написании научного текста на иностранном языке, является межъязыковая интерференция, проявляющаяся на лексико-семантическом, грамматико-синтаксическом, дискурсивном и жанрово-риторическом уровнях. На основе теоретического анализа отечественных и зарубежных исследований уточняется содержание понятия «межъязыковая интерференция» применительно к письму, редактируемому с помощью языковых моделей. Показано, что большие языковые модели способны снижать количество поверхностных языковых ошибок, обеспечивать оперативную обратную связь и поддерживать процессы редактирования и рефлексии. Вместе с тем они могут сохранять или маскировать особенности исходного языка, усиливать шаблонность академического текста, снижать самостоятельность обучающегося и создавать иллюзию преодоления языковых трудностей. Обоснована необходимость разграничения формальной языковой корректности и дискурсивной адекватности академического текста. Предложена последовательность работы с ИИ, включающая самостоятельный анализ исходного текста, направленный запрос к языковой модели, проверку ее рекомендаций, постредактирование и рефлексию. Сделан вывод о том, что образовательная ценность ИИ определяется не самим фактом его использования, а организацией учебной деятельности и сохранением за обучающимся ответственности за содержание, структуру и языковое оформление академического текста.

1. Lin L.H.F., Morrison B. Challenges in academic writing: Perspectives of Engineering faculty and L2 postgraduate research students, English for Specific Purposes, 2021, vol. 63, pp. 59-70. doi: 10.1016/j.esp.2021.03.004

2. Shpit Е.I. [Developing Professional Skills in Academic Writing with Postgraduate Students], Nepreryvnoye obrazovaniye: XXI vek [Lifelong Education:

The 21st Century], 2023, vol. 3, no. 43, doi: 10.15393/j5.art.2023.8650 (In Russ., abstract in Eng.)

3. Knyazeva О.V. [Problems of Teaching Academic English Genres to Technical University Students], Moskovskiy pedagogicheskiy zhurnal [Moscow Pedagogical Journal], 2025, no. 2, pp. 50-62. doi:10.18384/2949-4974-2025-2-50-62 (In Russ., abstract in Eng.)

4. Zavarykina L.V. [Teaching Various Academic Writing Genres in a Modern University], Nepreryvnoye obrazovaniye: XXI vek [Lifelong Learning:

The 21st Century], 2025, vol. 13, no. 2. doi: 10.15393/j5.art.2025.10404 (In Russ., abstract in Eng.)

5. Chen C., Gong Y. The role of AI-assisted learning in academic writing:

A mixed-methods study on Chinese as a second language students, Education Sciences, 2025, vol. 15, no. 2, art. 141. doi: 10.3390/educsci15020141

6. Ho N.N.Y., Rappa N.A., Wishart J. Using LLM chatbots for feedback in L2 English academic writing: a scoping review, Innovation in Language Learning and Teaching, 2026, vol. 1, no. 23. doi: 10.1080/17501229.2026.2698704

7. Cheng D., Li M., Lee T. Generative AI for second language writing feedback, assessment and learning, 2026. doi: 10.1017/9781009657334

8. Burhan K. Generative AI in second-language writing: A systematic review of empirical research, JRTI, 2025, vol. 10, no. 4, pp. 186-199. doi: 10.29210/30036974000

9. Ding L., Zou D., Kohnke L., Xie H. Harnessing Generative AI for L2 Written Corrective Feedback: A Systematic Review of Current Research and Future Direction. In: 2025 2nd International Conference on Artificial Intelligence and Teacher Education (ICAITE): Proceedings, 2025, pp. 43-48. doi: 10.1109/ICAITE68636.2025.11442331

10. Ding L., Zou D., Kohnke L. Enhancing L2 writing with AI: Effects of ChatGPT feedback and the role of student engagement, Artificial Intelligence in Language Education, 2026, vol. 2, art. 103356. doi: 10.29140/aile.2026.103356

11. Ding L., Zou D., Kohnke L. ChatGPT as an automated writing evaluation tool: How students perceive it and how it affects their writing, Education and Information Technologies, 2025, vol. 30, pp. 26777-26799. doi: 10.1007/s10639-025-13775-3

12. Ren B., Dong Y., Zhang, B. From errors to excellence: AI-driven scaffolding for advancing L2 writing skills, Cogent Education, 2025, vol. 12, no. 1. doi: 10.1080/2331186X.2025.2588512

13. Cheng A., Calhoun A., Reedy G. Artificial intelligence-assisted academic writing: Recommendations for ethical use, Advances in Simulation, 2025, vol. 10,

art. 22. doi: 10.1186/s41077-025-00350-6

14. Wang C., Xu W., Tan X. Advanced multi-lingual writers’ self-directed use of generative AI in academic writing: Rethinking writing, authorship, and learning, Applied Linguistics, 2026, vol. 47, no. 2, pp. 388-405. doi: 10.1093/applin/amaf057

15. Cole R., Maher L., Rice R. Understanding and avoiding hallucinated references: A critical literacy assignment for students, 2026, available at: https://wacclearinghouse.org/repository/collections/continuing-experiments/august-2025/

ai-literacy/understanding-avoiding-hallucinated-references/ (accessed 15 September 2026).

1. Lin, L. H. F. Challenges in Academic Writing: Perspectives of Engineering Faculty and L2 Postgraduate Research Students / L. H. F. Lin, B. Morrison // English for Specific Purposes. – 2021. – Vol. 63. – P. 59–70. doi: 10.1016/j.esp.2021.03.004

2. Шпит, Е. И. Развитие профессиональных умений в академическом письме у аспирантов / Е. И. Шпит // Непрерывное образование: XXI век. – 2023. – Т. 3, № 43. doi: 10.15393/j5.art.2023.8650

3. Князева, О. В. Проблемы обучения студентов технического вуза жанрам академического английского языка / О. В Князева // Московский педагогический журнал. – 2025. – № 2. – С. 50–62. doi: 10.18384/2949-4974-2025-2-62

4. Заварыкина, Л. В. Обучение различным жанрам академического письма в современном университете / Л. В. Заварыкина // Непрерывное образование: XXI век. – 2025. – Т. 13, № 2. doi: 10.15393/j5.art.2025.10404

5. Chen, C. The role of AI-assisted learning in academic writing: A mixed-methods study on Chinese as a second language students / C. Chen, Y. Gong // Education Sciences. – 2025. – Vol. 15, No. 2. – Art. 141. doi: 10.3390/educsci15020141

6. Ho, N. N. Y. Using LLM chatbots for feedback in L2 English academic writing: a scoping review / N. N. Y. Ho, N. A. Rappa, J. Wishart // Innovation in Language Learning and Teaching. – 2026. – Vol. 1, No. 23. doi: 10.1080/17501229.2026.2698704

7. Cheng, D. Generative AI for second language writing feedback, assessment and learning / D. Cheng, M. Li, T. Lee. – 2026. doi: 10.1017/9781009657334

8. Burhan, K. Generative AI in second-language writing: A systematic review of empirical research / K. Burhan // JRTI. – 2025. – Vol.10, No. 4. – P. 186–199. doi: 10.29210/30036974000

9. Harnessing Generative AI for L2 Written Corrective Feedback: A Systematic Review of Current Research and Future Direction / L. Ding, D. Zou, L. Kohnke, H. Xie // In 2025 2nd International Conference on Artificial Intelligence and Teacher Education (ICAITE): Proceedings. – 2025. – P. 43–48. doi: 10.1109/ICAITE68636.2025.11442331

10. Ding, L. Enhancing L2 writing with AI: Effects of ChatGPT feedback and the role of student engagement / L. Ding, D. Zou, L. Kohnke // Artificial Intelligence in Language Education. – 2026. – Vol. 2. – Art. 103356. doi: 10.29140/aile.2026.103356

11. Ding, L. ChatGPT as an automated writing evaluation tool: How students perceive it and how it affects their writing / L. Ding, D. Zou, L. Kohnke // Education and Information Technologies. – 2025. – Vol. 30. – P. 26777–26799. doi: 10.1007/s10639-025-13775-3

12. Ren, B., Dong, Y., & Zhang, B. (2025). From errors to excellence: AI-driven scaffolding for advancing L2 writing skills // Cogent Education. – 2025. – Vol. 12, No. 1. doi: 10.1080/2331186X.2025.2588512

13. Cheng, A. Artificial Intelligence-Assisted Academic Writing: Recommendations for Ethical Use / A. Cheng, A. Calhoun, G. Reedy // Advances in Simulation. – 2025. – Vol. 10. – Art. 22. doi: 10.1186/s41077-025-00350-6

14. Wang, C. Advanced multi-lingual writers’ self-directed use of generative AI in academic writing: Rethinking writing, authorship, and learning / C. Wang, W. Xu, X. Tan // Applied Linguistics. – 2026. – Vol. 47, No. 2. – P. 388–405. doi: 10.1093/applin/amaf057

15. Cole, R. Understanding and avoiding hallucinated references: A critical literacy assignment for students / R. Cole, L. Maher, R. Rice. – 2026. – URL : https://wacclearinghouse.org/repository/collections/continuing-experiments/august-2025/ai-literacy/understanding-avoiding-hallucinated-references/ (дата обращения: 15.09.2026)