Elamur Research⌕
☰
home / papers / 2026 / Semantic Change Characterization with LLMs using Rhetorics

Semantic Change Characterization with LLMs using Rhetorics

Summary

Most work on lexical semantic change only detects whether a word's meaning has shifted. This paper instead tackles characterizing the type of shift between two usages of a word. It follows Traugott's typology, grouped into three poles: dimension (broadening/narrowing), relation (metaphorization/metonymization) and orientation (amelioration/pejoration).

The authors prompt LLMs to reason with classic rhetorical devices rather than plain few-shot or Chain-of-Thought prompts. For example, a model tests whether two uses can be joined in a zeugma without an awkward pun, or uses antanagoge to rank which sense sounds more positive. To evaluate this, they built three new public sentence-pair datasets, one per pole, with a native English-speaking linguist. The dimension set is curated from WiC in DUREL format, and the relation set draws partly on a metaphor detection corpus.

On LLaMA-3-8b, LLaMA-3-70b and Phi-3, rhetorical prompting generally beats both baselines, and the models' reasoning is easy to inspect. The approach needs no task-specific fine-tuning and covers all change types. That makes it relevant for meaning-sensitive applications such as translation, chatbots and sentiment analysis.

Key points

  • Baselines are 3-shot prompting and CoT; the proposed method adds rhetorical reasoning (zeugma, antanagoge, figurative inference). The authors present this as the first quantitative test of such devices as semantic diagnostics.
  • Relation dataset accuracy (3-shot → rhetorical):
    • LLaMA-3-8b: .38 → .52
    • Phi-3: .49 → .56
    • LLaMA-3-70b: .52 → .59 (best overall)
  • Orientation is the hardest pole:
    • Phi-3 rises from .30 to .40.
    • LLaMA-3-70b goes from .36 to .39.
    • LLaMA-3-8b stays at .37.
    • Human agreement is also low (Fleiss' Kappa 0.3104, Krippendorff 0.3115).
  • The rhetorical strategy correlates least with the other prompting strategies. The authors read this as evidence that models' default intuitions about meaning change differ from established linguistic methods.
  • Reported limitations:
    • Models sometimes build zeugmas incorrectly or give reasoning that contradicts their answer.
    • Explanations may be memorized from training data.
    • The evaluation covers only English and only well-contextualized sentences.

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

Fields: Linguistics & languages

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.lchange-1.9

@inproceedings{sa2026semantic,
  title = {{Semantic Change Characterization with LLMs using Rhetorics}},
  author = {Sá, Jáder Martins Camboim de and Lee, Jooyoung and Silveira, Marcos Da and Pruski, Cedric},
  booktitle = {The Proceedings for the 6th International Workshop on Computational Approaches to Language Change (LChange'26)},
  year = {2026},
  doi = {10.18653/v1/2026.lchange-1.9},
  url = {https://doi.org/10.18653/v1/2026.lchange-1.9}
}

References

49 references, 11 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Eneko Agirre and Philip Edmonds. 2007. Word sense disambiguation: Algorithms and applications, volume 33. Springer Science & Business Media. find
  2. Keith Allan. 2013. The Oxford handbook of the history of linguistics. OUP Oxford. find
  3. Stefano Baccianella, Andrea Esuli, and Fabrizio Sebastiani. 2010. SentiWordNet 3.0: An enhanced lexical resource for sentiment analysis and opinion mining. In Proceedings of the Seventh International Conference on Language Resources and Evaluation (LREC’10), Valletta, Malta. European Language Resour… find our card
  4. Lukas Berglund, Meg Tong, Max Kaufmann, Mikita Balesni, Asa Cooper Stickland, Tomasz Korbak, and Owain Evans. 2023. The reversal curse: Llms trained on "a is b" fail to learn "b is a". ArXiv, abs/2309.12288. find
  5. Andreas Blank et al. 2003. Polysemy in the lexicon and in discourse. Trends in linguistics studies and monographs, 142:267–296. find
  6. Vladimir V. Bochkarev, Stanislav V. Khristoforov, Anna V. Shevlyakova, and Valery D. Solovyev. 2022. find
  7. Neural network algorithm for detection of new word meanings denoting named entities. IEEE Access, 10:68499–68512. 3130–3148, Toronto, Canada. Association for Computational Linguistics. find
  8. Jader Camboim de Sá, Dimitra Anastasiou, Marcos Da Silveira, and Cédric Pruski. 2024. Socio-cultural adapted chatbots: Harnessing knowledge graphs and large language models for enhanced context awarenes. In Proceedings of the 1st Worskhop on Towards Ethical and Inclusive Conversational AI: Language… find
  9. Simon Hengchen, Nina Tahmasebi, Dominik Schlechtweg, and Haim Dubossarsky. 2021. Challenges for computational lexical semantic change. ArXiv, abs/2101.07668. — [arXiv:2101.07668](https://arxiv.org/abs/2101.07668) find
  10. Jie Huang and Kevin Chen-Chuan Chang. 2023. Towards reasoning in large language models: A survey. In Findings of the Association for Computational Linguistics: ACL 2023, pages 1049–1065, Toronto, Canada. Association for Computational Linguistics. find our card
  11. Pride Kavumba, Ana Brassard, Benjamin Heinzerling, and Kentaro Inui. 2023. Prompting for explanations improves adversarial NLI. is this true? Yes it is true because it weakens superficial cues. In Findings of the Association for Computational Linguistics: EACL 2023, pages 2165–2180, Dubrovnik, Croat… find
  12. Jader Martins Camboim de Sá, Marcos Da Silveira, and Cédric Pruski. 2024. Survey in characterization of semantic change. ArXiv. find
  13. Erik Cambria, Rui Mao, Melvin Chen, Zhaoxia Wang, Seng-Beng Ho, and San Murugesan. 2023. Seven pillars for the future of artificial intelligence. IEEE Intelligent Systems, 38:62–69. find
  14. Kate Kearns. 2006. The Handbook of English Linguistics. Wiley Online Library. find
  15. Lyle Campbell. 2013. Historical linguistics. Edinburgh University Press. find
  16. Pierluigi Cassotti, Stefano De Pascale, and Nina Tahmasebi. 2024. Using synchronic definitions and semantic relations to classify semantic change types. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 4539–4553, Bangkok, Thail… find
  17. Guizhen Chen, Liying Cheng, Anh Tuan Luu, and Lidong Bing. 2023. Exploring the potential of large language models in computational argumentation. ArXiv, abs/2311.09022. find
  18. Minjin Choi, Sunkyung Lee, Eunseong Choi, Heesoo Park, Junhyuk Lee, Dongwon Lee, and Jongwuk Lee. 2021. MelBERT: Metaphor detection via contextualized late interaction using metaphorical identification theories. In Proceedings of the 2021 Conference of the North American Chapter of the Association f… find
  19. Ishita Dasgupta, Andrew Kyle Lampinen, Stephanie C. Y. Chan, Antonia Creswell, Dharshan Kumaran, James L. McClelland, and Felix Hill. 2022. Language models show human-like content effects on reasoning. ArXiv, abs/2207.07051. find
  20. Lauren Fonteyn and Enrique Manjavacas. 2021. Adjusting scope: A computational approach to case-driven research on semantic change. In Workshop on Computational Humanities Research. find
  21. Mario Giulianelli, Iris Luden, Raquel Fernandez, and Andrey Kutuzov. 2023. Interpretable word sense representations via definition generation: The case of semantic change analysis. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pag… find our card
  22. George Lakoff and Mark Johnson. 2008. Metaphors we live by. University of Chicago press. find
  23. Emmy Liu, Chenxuan Cui, Kenneth Zheng, and Graham Neubig. 2022. Testing the ability of language models to interpret figurative language. In Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 4437–4452,… find
  24. Chenyang Lyu, Yongxin Zhou, and Tianbo Ji. 2022. MLLabs-LIG at TempoWiC 2022: A generative approach for examining temporal meaning shift. In Proceedings of the First Workshop on Ever Evolving NLP (EvoNLP), pages 1–6, Abu Dhabi, United Arab Emirates (Hybrid). Association for Computational Linguistics… find
  25. Rowan Hall Maudslay and Simone Teufel. 2022. Metaphorical polysemy detection: Conventional metaphor meets word sense disambiguation. In Proceedings of the 29th International Conference on Computational Linguistics, pages 65–77, Gyeongju, Republic of Korea. International Committee on Computational Li… find
  26. George A. Miller. 1995. Wordnet: A lexical database for english. Commun. ACM, 38:39–41. find our card
  27. Arindam Mitra, Luciano Del Corro, Shweti Mahajan, Andres Codas, Clarisse Simoes, Sahaj Agrawal, Xuxi Chen, Anastasia Razdaibiedina, Erik Jones, Kriti Aggarwal, Hamid Palangi, Guoqing Zheng, Corby Rosset, Hamed Khanpour, and Ahmed Awadallah. 2023. Orca 2: Teaching small language models how to reason.… find our card
  28. Roberto Navigli. 2009. Word sense disambiguation: A survey. ACM Comput. Surv., 41:10:1–10:69. find
  29. Human and computational measurement of semantic proximity, sense clusters and semantic change. In Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations, pages 137–149, St. Julians, Malta. Association for Computational Lingu… find
  30. Francesco Periti, Pierluigi Cassotti, Haim Dubossarsky, and Nina Tahmasebi. 2024a. Analyzing semantic change through lexical replacements. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 4495–4510, Bangkok, Thailand. Associati… find
  31. Francesco Periti, Haim Dubossarsky, and Nina Tahmasebi. 2024b. (chat)GPT v BERT dawn of justice for semantic change detection. In Findings of the Association for Computational Linguistics: EACL 2024, pages 420–436, St. Julian’s, Malta. Association for Computational Linguistics. find
  32. Fabio Petroni, Tim Rocktäschel, Sebastian Riedel, Patrick Lewis, Anton Bakhtin, Yuxiang Wu, and Alexander Miller. 2019. Language models as knowledge bases? In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural… find our card
  33. Mohammad Taher Pilehvar and Jose Camacho-Collados. 2019. WiC: the word-in-context dataset for evaluating context-sensitive meaning representations. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volum… find our card
  34. Steven Pinker. 2003. The language instinct: How the mind creates language. Penguin UK. find
  35. Lidia Pivovarova and Andrey Kutuzov. 2021. Rushifteval: a shared task on semantic shift detection for russian. ArXiv. find
  36. Alessandro Raganato, Jose Camacho-Collados, and Roberto Navigli. 2017. Word sense disambiguation: A unified evaluation framework and empirical comparison. In Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers, pages 99–1… find our card
  37. Dominik Schlechtweg, Barbara McGillivray, Simon Hengchen, Haim Dubossarsky, and Nina Tahmasebi. 2020. SemEval-2020 task 1: Unsupervised lexical semantic change detection. In Proceedings of the Fourteenth Workshop on Semantic Evaluation, pages 1–23, Barcelona (online). International Committee for Com… find our card
  38. Dominik Schlechtweg, Shafqat Mumtaz Virk, Pauline Sander, Emma Sköldberg, Lukas Theuer Linke, Tuo Zhang, Nina Tahmasebi, Jonas Kuhn, and Sabine Schulte Im Walde. 2024. The DURel annotation tool: Gerard Steen, Lynne Cameron, Alan Cienki, Peter Crisp, Alice Deignan, Raymond W. Gibbs, Joe Grady, Zoltán… find
  39. James W. A. Strachan, Dalila Albergo, Giulia Borghini, Oriana Pansardi, Eugenio Scaliti, Saurabh Gupta, Krati Saxena, Alessandro Rufo, Stefano Panzeri, Guido Manzi, Michael S A Graziano, and Cristina Becchio. 2024. Testing theory of mind in large language models and humans. Nature human behaviour. find
  40. Nina Tahmasebi, Lars Borin, and Adam Jatowt. 2018. Survey of computational approaches to lexical semantic change. arXiv: Computation and Language. find
  41. Xuri Tang, Weiguang Qu, and Xiaohe Chen. 2013. Semantic change computation: A successive approach. World Wide Web, 19:375–415. find
  42. Elizabeth Closs Traugott. 2017. Semantic change. find
  43. Ruiyu Wang and Matthew Choi. 2023. Large language models on lexical semantic change detection: An evaluation. ArXiv, abs/2312.06002. — [arXiv:2312.06002](https://arxiv.org/abs/2312.06002) find
  44. Taylor W. Webb, Keith J. Holyoak, and Hongjing Lu. 2022. Emergent analogical reasoning in large language models. Nature Human Behaviour, 7:1526 – 1541. find
  45. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Fei Xia, Ed Chi, Quoc V Le, Denny Zhou, et al. 2022. Chain-of-thought prompting elicits reasoning in large language models. Advances in neural information processing systems, 35:24824–24837. find our card
  46. Janyce Wiebe and Rada Mihalcea. 2006. Word sense and subjectivity. In Proceedings of the 21st International Conference on Computational Linguistics and 44th Annual Meeting of the Association for Computational Linguistics, pages 1065–1072, Sydney, Australia. Association for Computational Linguistics. find
  47. Xi Ye and Greg Durrett. 2022. The unreliability of explanations in few-shot prompting for textual reasoning. In Neural Information Processing Systems. find our card
  48. Frank D. Zamora-Reina, Felipe Bravo-Marquez, and Dominik Schlechtweg. 2022. Lscdiscovery: A shared task on semantic change discovery and detection in spanish. ArXiv, abs/2205.06691. find
  49. Xulang Zhang, Rui Mao, Kai He, and Erik Cambria. 2023. Neuro-symbolic sentiment analysis with dynamic word sense disambiguation. In Conference on Empirical Methods in Natural Language Processing. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/sa-2026-semantic-change-characterization-llms-rhetorics/