Elamur Research⌕
☰
home / papers / 2026 / MultiCoS: A Multilingual Dataset of Connective Semantics with Context–Sentence Compatibility

MultiCoS: A Multilingual Dataset of Connective Semantics with Context–Sentence Compatibility

Summary

Cross-linguistic work on clausal coordinators such as and and or relies on typological claims, for example that no language has a dedicated exclusive "or" or a NAND connective. Existing resources rarely provide the evidence needed to check such claims. Connective lexicons list items and their senses but give no felicity data. Grammar databases like WALS, Grambank and Terraling record mostly morphosyntactic features per language. MultiCoS addresses this gap. It annotates connectives from 24 typologically varied languages for:

  • Boolean meaning (AND, OR, NOR, NAND, XOR)
  • contrast marking
  • polar vs. alternative question readings
  • free-choice effects

Every annotation is backed by native-speaker judgments of whether a coordinated sentence fits a given context, including contexts where it does not fit. The data were gathered with one consultant per language, using translation-in-context tasks and compatibility questionnaires. The main contribution is the resource itself. Because each label is tied to explicit evidence, others can verify claims about connective senses, test proposed universals, and look for counterexamples. The authors also suggest the data could serve as a probe of how NLU models judge felicity, though they present this use as exploratory only.

Key points

  • Covers 24 languages from many families, including Basque, Mapudungun, Kîîtharaka, Igbo, Telugu and Tagalog, all collected with a uniform elicitation protocol.
  • The authors present it as the first multilingual connective resource whose labels rest on context–sentence compatibility judgments, including infelicitous cases.
  • Elicitation ran online over video calls in two questionnaire sessions per language, with one paid consultant each (9 GBP/hour).
  • The annotation targets distinctions such as inclusive vs. exclusive disjunction and AltQ vs. PolQ readings (as with Basque ala vs. edo).
  • The data are released on GitHub (EdinburghMeaningSciences/multicos). The NLU felicity benchmarking is only preliminary, without comparisons across models or prompts.

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 10.63317/28ccty7yu9hn

@inproceedings{mucha2026multicos,
  title = {{MultiCoS: A Multilingual Dataset of Connective Semantics with Context–Sentence Compatibility}},
  author = {Mucha, Anne and Qing, Ciyang and Uegaki, Wataru},
  booktitle = {Proceedings of the Fifteenth Language Resources and Evaluation Conference},
  year = {2026},
  doi = {10.63317/28ccty7yu9hn},
  url = {https://doi.org/10.63317/28ccty7yu9hn}
}

References

31 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Moshe E Bar-Lev and Roni Katzir. 2025. Communicative stability and the typology of logical operators. Linguistic Inquiry, 56(3):415–437. find
  2. Margit Bowler. 2014. Conjunction and disjunction in a language without ‘and’. In Proceedings of Semantics and Linguistic Theory (SALT) 24, pages 137–155. find
  3. Enrico Flor, Nina Haslinger, Hilda Koopman, Eva Rosina, Magdalena Roszkowski, and Viola Schmitt. 2017. Cross-linguistic evidence for a non-distributive lexical meaning of conjunction. In Proceedings of the Amsterdam Colloquium, pages 255–264. find
  4. Max Glockner, Vered Shwartz, and Yoav Goldberg. 2018. Breaking NLI systems with sentences that require simple lexical inferences. In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers), pages 650–655, Melbourne, Australia. Association for… find our card
  5. Martin Haspelmath. 2007. Coordination. In Timothy Shopen, editor, Language Typology and Syntactic Description, 2nd ed., vol. 2: Complex constructions, pages 1–51. Cambridge University Press, Cambridge. find
  6. Laurence Robert Horn. 1972. On the semantic properties of logical operators in English. University of California, Los Angeles. find
  7. Hans Kamp. 1973. Free choice permission. Proceedings of the Aristotelian Society, 74:57–74. find
  8. Roni Katzir and Raj Singh. 2013. Constraints on the lexicalization of logical operators. Linguistics and Philosophy, 36(1):1–29. find
  9. Lisa Matthewson. 2004. On the methodology of semantic fieldwork. International journal of American linguistics, 70(4):369–415. find
  10. Lisa Matthewson. 2011. Methods in cross-linguistic semantics. In Claudia Maienborn, Klaus von Heusinger, and Paul Portner, editors, Semantics: An international handbook of natural language meaning, pages 268–284. de Gruyter, Berlin. find
  11. Caterina Mauri. 2008. The irreality of alternatives: Towards a typology of disjunction. Studies in Language, 32(1):22–55. find
  12. Moreno Mitrovic and Uli Sauerland. 2014. Decomposing coordination. In Proceedings of NELS, volume 44, pages 39–52. find
  13. Anne Mucha, Wataru Uegaki, Mary Amaechi, and Tim Jantarungsee. 2025. Disjunction with additives. In Proceedings of Sinn und Bedeutung 29, pages 1071–1087. find
  14. Aakanksha Naik, Abhilasha Ravichander, Norman Sadeh, Carolyn Rose, and Graham Neubig. 2018. Stress test evaluation for natural language inference. In Proceedings of the 27th International Conference on Computational Linguistics (COLING 2018), pages 2340–2353, Santa Fe, New Mexico, USA. Association f… find
  15. Swarnadeep Saha, Yixin Nie, and Mohit Bansal. 2020. ConjNLI: Natural language inference over conjunctive sentences. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 8240– 8254, Online. Association for Computational Linguistics. find
  16. Mario Saltarelli. 1988. Basque. Croom Helm, London. find
  17. Mark Sammons. 2015. Recognizing textual entailment. The Handbook of Contemporary Semantic Theory, pages 523–557. find
  18. Benjamin Spector. 2014. Global positive polarity items and obligatory exhaustivity. Semantics and Pragmatics, 7:1–61. find
  19. Wataru Uegaki. 2023. The informativeness/complexity trade-off in the domain of boolean connectives. Linguistic Inquiry, 55(1):174–196. find
  20. Alex Warstadt, Amanpreet Singh, and Samuel R. Bowman. 2019. Neural network acceptability judgments. Transactions of the Association for Computational Linguistics, 7:625–641. find
  21. Yoad Winter. 2002. Flexibility principles in Boolean semantics: The interpretation of coordination, plurality, and scope in natural language. MIT press. find
  22. Thomas Ede Zimmermann. 2000. Free choice disjunction and epistemic possibility. Natural Language Semantics, 8(4):255–290. 11. find
  23. Language Resource References Diptesh Das, Thomas Scheffler, Peter Bourgonje, and Manfred Stede. 2018. Constructing a lexicon of English discourse connectives (DiMLex-Eng). In Proceedings of the SIGDIAL 2018 Conference, pages 23–32. find our card
  24. Martin Haspelmath, Matthew S. Dryer, David Gil, and Bernard Comrie, editors. 2005. The World Atlas of Language Structures. Oxford University Press, Oxford. find
  25. Hilda Koopman and Cristina Guardiano. 2022. Managing data in TerraLing, a large-scale cross-linguistic database of morphological, syntactic, and semantic patterns. In The Open Handbook of Linguistic Management, pages 617–630. MIT Press. find
  26. Guillaume Marchal, Rudi Scholman, and Volker Demberg. 2021. Semi-automatic discourse annotation in a low-resource language: Developing a connective lexicon for Nigerian Pidgin. In Proceedings of the 2nd Workshop on Computational Approaches to Discourse, pages 84–94. find
  27. Ponrawee Prasertsom, Apiwat Jaroonpol, and Attapol T. Rutherford. 2024. The Thai Discourse Treebank: Annotating and modeling discourse relations in Thai. Transactions of the Association for Computational Linguistics, 12:1–25. find
  28. Hedvig Skirgård, Hannah J. Haynie, Damián E. Blasi, Annemarie Verkerk, Sam Passmore, Felix Tiedemann, Christoph Rzymski, Yulia Egorova, Tobias Hofer, Gerhard Jäger, Robert Forkel, Johanna Nichols, Michael Dunn, and Russell D. Gray. 2023. Grambank: A cross-linguistic database of morphosyntactic varia… find
  29. Manfred Stede, Thomas Scheffler, and Ana Mendes. 2019. Connective-Lex: A web-based multilingual lexical resource for connectives. Discours, 27:1–35. find
  30. Shun Wan, Peter Bourgonje, Heng Xiao, Chih-Wen Ho, and Manfred Stede. 2023. Chinese-DiMLex: A lexicon of Chinese discourse connectives. In Proceedings of the CODI Workshop at ACL 2023, pages 110–122. find
  31. Derya Zeyrek and Kübra Başıbüyük. 2019. TCL: A lexicon of Turkish discourse connectives. In Proceedings of the DMR Workshop at ACL 2019, pages 41–50. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/mucha-2026-multicos-multilingual-dataset-connective-semantics/