Elamur Research⌕
☰
home / papers / 2026 / Meaning Annotation Experience. A Tribute to Petr Sgall

Meaning Annotation Experience. A Tribute to Petr Sgall

Summary

This paper, honouring the centenary of Petr Sgall, reports on an ongoing effort to annotate fine-grained spatial meanings in Czech. It also considers how vague natural-language meaning can be captured in a semantic scheme. The work is grounded in Functional Generative Description and the multi-layer Prague Dependency Treebank (PDT) framework. There, circumstantial modifiers already carry coarse functors on the tectogrammatical (meaning) layer, such as LOC ("where") or DIR3 ("where to"). The authors refine these functors with narrower subfunctors (e.g., inside, above, behind, near), assigned task by task according to form, such as a given preposition and case. Over 126,500 circumstants in Prague Dependency Treebank – Consolidated 2.0 received multiple independent annotations. Using confusion matrices of competing labels, the authors study where annotators diverge. They find that much of the disagreement stems from genuine vagueness rather than annotator error. Their proposed responses are:

  • weighted annotations,
  • intermediate labels,
  • fuzzy labels,
  • merging only categories whose boundaries are truly blurred.

These strategies aim to keep nuance in the planned 2028 corpus release.

Key points

  • Corpus: PDT-C 2.0, about 3 million tokens of genre-diverse Czech. Subfunctor annotation covers over 126,500 circumstants and is scheduled for a 2028 release.
  • Example task (LOC expressed by v+6 'in'): 14 subfunctors plus optional abstract/idiom labels. It covered 23,522 instances, each annotated at least three times, with up to two values per annotator. The most frequent label was inside, with 50,385 occurrences.
  • ingroup was paired with institute (2,702 times) more often than with itself (1,694 times). Because both also have many clear-cut cases, the authors propose an intermediate ingroup-institute label rather than merging them.
  • work, media and actinfo overlap heavily, with actinfo co-occurring more with work than with itself. The authors therefore suggest merging these three into a single category.
  • An opening example with English "on the tower" shows that a single prepositional phrase can denote quite different real-world locations. This illustrates the gap between linguistic meaning and extra-linguistic content.

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 10.63317/4v4gn44mvftb

@inproceedings{mikulova2026meaning,
  title = {{Meaning Annotation Experience. A Tribute to Petr Sgall}},
  author = {Mikulová, Marie and Štěpánek, Jan and Štěpánková, Barbora and Panevova, Jarmila and Hajicova, Eva},
  booktitle = {Proceedings of The Seventh International Workshop on Designing Meaning Representations (DMR 2026) @ LREC 2026},
  year = {2026},
  doi = {10.63317/4v4gn44mvftb},
  url = {https://doi.org/10.63317/4v4gn44mvftb}
}

References

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

  1. Simon Dobnik, Robin Cooper, Adam Ek, Bill Noble, Staffan Larsson, Nikolai Ilinykh, Vladislav Maraev, and Vidya Somashekarappa. 2022. In Search of Meaning and Its Representations for Computational Linguistics. In Proceedings of the 2022 CLASP Conference on (Dis)embodiment, pages 30–44, Gothenburg, Sw… find
  2. Kira Droganova and Daniel Zeman. 2019. Towards Deep Universal Dependencies. In Proceedings of the Fifth International Conference on Dependency Linguistics, pages 144–152, Paris, France. Association for Computational Linguistics. find
  3. Luke Gessler, Austin Blodgett, Joseph C. Ledford, and Nathan Schneider. 2022. Xposition: An Online Multilingual Database of Adpositional Semantics. In Proceedings of the Thirteenth Language Resources and Evaluation Conference, pages 1824–1830, Marseille, France. European Language Resources Associati… find
  4. Jan Hajič, Eduard Bejček, Alevtina Bémová, Eva Buráňová, Eva Fučíková, Eva Hajičová, Jiří Havelka, Jaroslava Hlaváčová, Petr Homola, Pavel Ircing, Jiří Kárník, Václava Kettnerová, Natalia Klyueva, Veronika Kolářová, Lucie Kučová, Markéta Lopatková, David Mareček, Marie Mikulová, Jiří Mír… find
  5. Eva Hajičová, Marie Mikulová, Barbora Štěpánková, and Jiří Mírovský. 2022. Advantages of a Complex Multilayer Annotation Scheme: The Case of the Prague Dependency Treebank. In Proceedings of the 16th Linguistic Annotation Workshop, pages 70–78, Marseille, France. ELRA. find
  6. Eva Hajičová and Petr Sgall. 1980. Linguistic Meaning and Knowledge Representation in Automatic Understanding of Natural Language. In COLING 1980: The 8th International Conference on Computational Linguistics, pages 67–75. find
  7. Martin Haspelmath. 2010. Comparative Concepts and Descriptive Categories in Cross-Linguistic Studies. Language, 86(3):663–687. find
  8. Kenneth Lai and Shira Wein, editors. 2025. Proceedings of the 6th Workshop on Designing Meaning Representations. Association for Computational Lingustics, Prague, Czechia. find
  9. Bolei Ma, Yuting Li, Wei Zhou, Ziwei Gong, Yang Janet Liu, Katja Jasinskaja, Annemarie Friedrich, Julia Hirschberg, Frauke Kreuter, and Barbara Plank. 2025. Pragmatics in the Era of Large Language Models: A Survey on Datasets, Evaluation, Opportunities and Challenges. In Proceedings of the 63rd Annu… find
  10. Marie Mikulová. 2024. Fine-grained Classification of Circumstantial Meanings within the Prague Dependency Treebank Annotation Scheme. In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation, pages 7314–7323, Torino, Italia. ELRA and I… find
  11. Marie Mikulová and Eduard Bejček. 2018. ForFun 1.0: Prague database of forms and functions – an invaluable resource for linguistic research. In Proceedings of the 11th International Conference on Language Resources and Evaluation, Miyazaki, Japan. ELRA. find
  12. Marie Mikulová, Jiří Mírovský, Milan Straka, Pavlína Synková, Barbora Štěpánková, Jan Štěpánek, and Jan Hajič. 2026. Prague Dependency Treebank - Consolidated 2.0: Enriching a Complex Annotation Scheme. In Proceedings of the 15th Language Resources and Evaluation Conference, Palma de Mallorca, S… find
  13. Marie Mikulová, Jan Štěpánek, and Jan Hajič. 2025a. Label Bias in Symbolic Representation of Meaning. In Proceedings of the 19th Linguistic Annotation Workshop, pages 142–159, Vienna, Austria. ACL. find
  14. Marie Mikulová, Barbora Štěpánková, and Jan Štěpánek. 2025b. From Form to Meaning: The Case of Particles within the Prague Dependency Treebank Annotation Scheme. In Proceedings of the 31st International Conference on Computational Linguistics, pages 2163–2175, Abu Dhabi, UAE. Association for Compu… find
  15. Jarmila Panevová. 1980. Formy a funkce ve stavbě české věty. Academia, Prague, Czechia. find
  16. Steven T. Piantadosi, Harry Tily, and Edward Gibson. 2012. The Communicative Function of Ambiguity in Language. Cognition, 122(3):280–291. find
  17. Barbara Plank. 2022. The “Problem” of Human Label Variation: On Ground Truth in Data, Modeling and Evaluation. In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 10671–10682, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics. find our card
  18. Valentina Pyatkin, Daniel Fried, Elias Stengel-Eskin, Alisa Liu, and Sandro Pezzelle, editors. 2024. Proceedings of the 3rd Workshop on Understanding Implicit and Underspecified Language. ACL, Malta. find
  19. Michael Roth and Dominik Schlechtweg, editors. 2025. Proceedings of Context and Meaning: Navigating Disagreements in NLP Annotation. International Committee on Computational Linguistics, Abu Dhabi, UAE. find
  20. Zacchary Sadeddine, Juri Opitz, and Fabian Suchanek. 2024. A Survey of Meaning Representations – From Theory to Practical Utility. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 2877–2892, Mexic… find
  21. Petr Sgall. 1965. Generation, Production, and Translation. In COLING 1965. find
  22. Petr Sgall. 1995. From Meaning via Reference to Content. In Karlovy Vary studies in reference and meaning, pages 172–183. Filosofia Publications, Prague, Czech Republic. find
  23. Petr Sgall. 2002. Freedom of Language: Its Nature, Its Sources, and Its Consequences. In Prague Linguistic Circle Papers: Travaux du cercle linguistique de Prague nouvelle série. Volume 4, pages 309–329. John Benjamins Publishing Company. find
  24. Petr Sgall. 2006. Valence jako jádro jazykového systému. Slovo a slovesnost, 67(3):163–178. find
  25. Petr Sgall, Eva Hajičová, and Jarmila Panevová. 1986. The Meaning of the Sentence and Its Semantic and Pragmatic Aspects. Academia/Reidel, Prague/Dordrecht. find
  26. Alexandra N Uma, Tommaso Fornaciari, Dirk Hovy, Silviu Paun, Barbara Plank, and Massimo Poesio. 2021. Learning from disagreement: A survey. Journal of Artificial Intelligence Research, 72:1385–1470. find
  27. Jens EL Van Gysel, Meagan Vigus, Jayeol Chun, Kenneth Lai, Sarah Moeller, Jiarui Yao, Tim O’Gorman, Andrew Cowell, William Croft, Chu-Ren Huang, Jan Hajič, James H. Martin, Stephan Oepen, Martha Palmer, James Pustejovsky, Rosa Vallejos, and Nianwen Xue. 2021. Designing a Uniform Meaning Representat… find
  28. Amir Zeldes, Tatsuya Aoyama, Yang Janet Liu, Siyao Peng, Debopam Das, and Luke Gessler. 2025. eRST: A Signaled Graph Theory of Discourse Relations and Organization. Computational Linguistics, 51(1):23–72. *Figure 3: Confusion matrix for the spatial meaning “where” (functor LOC) expressed by v +6 ‘in… find our card

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/mikulova-2026-meaning-annotation-experience-tribute-petr/