PROMPT ENGINEERING AND STYLISTIC CONTROL IN LARGE LANGUAGE MODELS: A QUANTITATIVE ANALYSIS OF FORMALITY, EMOTIONAL TONE, AND PERSUASION

Authors

  • Hossain Md. Sayem Master's Student, Faculty of Humanities, National Research University Higher School of Economics (HSE University), Nizhny Novgorod, Russia
  • Arif Md. Ismail Hossain Graduate Student, Faculty of Economics, National Research University Higher School of Economics (HSE University), Nizhny Novgorod, Russia

DOI:

https://doi.org/10.46827/ejals.v9i2.746

Keywords:

large language models, prompt engineering, stylistics, computational linguistics, DeepSeek

Abstract

This study investigates the extent to which stylistic features of Large Language Model (LLM) output can be controlled through prompt engineering. Drawing on established frameworks in stylistics, we examine three core stylistic dimensions formality, emotional tone, and persuasive style across six controlled prompt conditions using the DeepSeek LLM. Through a mixed-methods approach combining qualitative textual analysis with quantitative measurement of six linguistic metrics (average sentence length, technical term density, emotion marker frequency, modal verb usage, contraction frequency, and a composite formality index), we demonstrate that systematic variations in prompt structure produce measurable and linguistically interpretable stylistic shifts in generated text. Our findings indicate that explicit stylistic instructions yield stronger effects than implicit contextual cues, and that different stylistic dimensions respond differently to prompt manipulations. Specifically, formal prompts produced texts with 137% longer sentences and 185% more technical terms than informal prompts; emotional prompts generated up to 7.19 emotion markers per 100 words compared to zero in neutral conditions; and directive persuasive prompts exhibited 445% more modal verbs than formal neutral texts. These results provide empirical support for applying classical stylistic theory to AI-generated text and offer practical guidance for linguistically informed prompt engineering.

Author Biographies

  • Hossain Md. Sayem, Master's Student, Faculty of Humanities, National Research University Higher School of Economics (HSE University), Nizhny Novgorod, Russia

    Hossain Md. Sayem is a Master's student in the Faculty of Humanities at the National Research University Higher School of Economics (HSE University), Nizhny Novgorod, Russia. His field of study is Fundamental and Applied Linguistics. His research interests include computational stylistics, prompt engineering, Large Language Models, and the linguistic analysis of AI-generated text. He holds a degree in Linguistics and is currently investigating the intersection of natural language processing and classical linguistic theory.

  • Arif Md. Ismail Hossain, Graduate Student, Faculty of Economics, National Research University Higher School of Economics (HSE University), Nizhny Novgorod, Russia

    Arif Md. Ismail Hossain is a graduate student at the Faculty of Economics, National Research University Higher School of Economics (HSE University), Nizhny Novgorod, Russia. His research interests include monetary economics, financial market development, emerging market macroeconomics and AI. He holds a BBA in Finance from Premier University, Chittagong, Bangladesh.

References

Adamopoulou E, Moussiades L, 2020. Chatbots: History, technology, and applications. Machine Learning with Applications 2: 100006. https://doi.org/10.1016/j.mlwa.2020.100006

Alboqami H, 2022. Factors affecting consumers adoption of AI-based chatbots: The role of anthropomorphism. In: Pehcevski J (Ed.), Chatbots and Text Generation, Arcler Press, pp. 29-56. https://doi.org/10.4236/ajibm.2023.134014

Biber D, 1988. Variation across Speech and Writing. Cambridge University Press, Cambridge, UK. Retrieved from https://doi.org/10.1017/CBO9780511621024

Ciesla R, 2024. The Book of Chatbots: From ELIZA to ChatGPT. Springer Nature. Retrieved from https://doi.org/10.1007/978-3-031-51004-5

Halliday MAK, 1985. An Introduction to Functional Grammar. Edward Arnold, London, UK. Retrieved from https://books.google.ro/books/about/An_Introduction_to_Functional_Grammar.html?id=JM3KAgAAQBAJ&redir_esc=y

Kolmogorova AV, Margolina AV, 2024. Written vs generated text: "Naturalness" as a textual and psycholinguistic category. Research Result: Theoretical and Applied Linguistics 10(2): 71-99. https://doi.org/10.18413/2313-8912-2024-10-2-0-4

Leech G, Short M, 2007. Style in Fiction: A Linguistic Introduction to English Fictional Prose, 2nd edn. Pearson, Harlow, UK. Retrieved from https://books.google.ro/books/about/Style_in_Fiction.html?id=mScMAjO9qjkC&redir_esc=y

O'Sullivan J, 2025. Stylometric comparisons of human versus AI-generated creative writing. Humanities and Social Sciences Communications 12: 1708. https://doi.org/10.1057/s41599-025-05986-3

Swales JM, 1990. Genre Analysis: English in Academic and Research Settings. Cambridge University Press, Cambridge, UK. Retrieved from https://books.google.ro/books/about/Genre_Analysis.html?id=shX_EV1r3-0C&redir_esc=y

Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez AN, Kaiser L, Polosukhin I, 2017. Attention is all you need. Advances in Neural Information Processing Systems 30: 5998-6008. https://doi.org/10.48550/arXiv.1706.03762

Zhao WX, Zhou K, Li J, Tang T, Wang X, Hou Y, Min Y, Zhang B, Zhang J, Dong Z, Du Y, Yang C, Chen Y, Chen Z, Jiang J, Ren R, Li Y, Tang X, Liu Z, Liu P, Nie JY, Wen JR, 2023. A survey of large language models. arXiv. Retrieved from https://arxiv.org/abs/2303.18223

Downloads

Published

2026-09-25

Issue

Vol. 9 No. 2 (2026)

Section

Articles

License

Authors who publish with this journal agree to the following terms:

Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.

Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.

Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).