Elamur Research⌕
☰
home / papers / 2026 / GlobLingDiv: A global dataset linking linguistic diversity and digital support to reveal landscapes with under-resourced languages for NLP

GlobLingDiv: A global dataset linking linguistic diversity and digital support to reveal landscapes with under-resourced languages for NLP

Summary

Global linguistic diversity estimates that account for both how many languages a country has and how evenly speakers are spread across them have been scarce. Most entropy-based studies cover single regions or countries. The authors built GlobLingDiv, an openly released country-level database. It merges people-group data from the Joshua Project with Ethnologue speaker counts and aligns entries through ISO language and country codes. Where sources disagreed, the lower count was kept, and L1 and L2 speakers were not distinguished. For each country, the database reports language richness, Shannon entropy H, and exp(H), which is read as the "effective number of languages." It also includes a speaker-weighted score based on Ethnologue's Digital Language Support (DLS) scale. Ranking countries by exp(H) rather than raw counts reshapes the global picture. Across countries, higher diversity goes with weaker digital support. The resource helps NLP researchers locate under-resourced linguistic landscapes and supports quantitative work in the humanities and social sciences. The authors caution that country-level aggregation hides sub-national patterns, speaker counts are uncertain, and the correlation does not establish causation.

Key points

  • Coverage: The dataset holds 6,745 languages, about 7.6 billion speakers, 239 countries and 12,249 language–country pairs. It spans 214 language families and 153 isolates and covers roughly 95% of Ethnologue's richness, with extinct languages excluded. It is distributed as CSV tables on the TU Wien GitLab.
  • Richness vs. evenness: Papua New Guinea is the clear hotspot with exp(H) ≈ 96.65. Cameroon (289 languages, exp(H) ≈ 53.6) ranks above Vanuatu (115 languages, exp(H) ≈ 54.3) on richness, but Vanuatu comes out more diverse on exp(H). Switzerland likewise beats Germany on exp(H) despite having about half as many languages.
  • Digital support score: The five DLS categories (Still to Thriving, based on presence across 216 digital tools) were recoded as 1–5 and averaged per country, weighted by speaker numbers.
  • Score extremes: Australia, the US and the Netherlands reach 4.99. The Republic of the Congo scores 1.77 and Vanuatu 1.66, and much of Africa scores low.
  • Correlation: A Spearman correlation between exp(H) and log-transformed digital support is significant and negative (r = −.47, p < .001).

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 · History & humanities

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.latechclfl-1.8

@inproceedings{zeh2026globlingdiv,
  title = {{GlobLingDiv: A global dataset linking linguistic diversity and digital support to reveal landscapes with under-resourced languages for NLP}},
  author = {Zeh, Katharina and Essfors, Hannes and Benson, Juliane and Tüver, Lale and Baumann, Andreas and Fellner, Hannes A.},
  booktitle = {Proceedings of the 10th Joint SIGHUM Workshop on Computational Linguistics for Cultural Heritage, Social Sciences, Humanities and Literature 2026},
  year = {2026},
  doi = {10.18653/v1/2026.latechclfl-1.8},
  url = {https://doi.org/10.18653/v1/2026.latechclfl-1.8}
}

References

21 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Antonios Anastasopoulos, Christopher Cox, Hilaria Cruz, and Graham Neubig. 2020. Endangered languages meet modern nlp. In Proceedings of the 28th International Conference on Computational Linguistics: Tutorial Abstracts, pages 39–45, Online. International Committee on Computational Linguistics. find
  2. Gábor Bella, Paula Helm, Gertraud Koch, and Fausto Giunchiglia. 2023. Towards bridging the digital language divide. Preprint, arXiv:2307.13405. — [arXiv:2307.13405](https://arxiv.org/abs/2307.13405) find
  3. Juliane Benson, Katharina Zeh, Hannes Essfors, Hannes Fellner, Julia Neidhardt, and Andreas Baumann. 2025. Linguistic diversity and digitalization: An ambivalent relationship. In Digital Humanism, number 16319 in Lecture Notes in Computer Science, pages 358–365. Springer Nature. find
  4. Damian Blasi, Antonios Anastasopoulos, and Graham Neubig. 2022. Systematic Inequalities in Language Technology Performance across the World’s Languages. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 5486–5505, Dublin, Irelan… find our card
  5. Michaël Boissonneault, Adam Tallman, Volker Gast, and Simon J Greenhill. 2025. Projected speaker numbers and dormancy risks of canada’s indigenous languages. Royal Society Open Science, 12(2):241091. find
  6. Lindell Bromham, Russell Dinnage, Hedvig Skirgård, Andrew Ritchie, Marcel Cardillo, Felicity Meakins, Simon Greenhill, and Xia Hua. 2022. Global predictors of language endangerment and the future of linguistic diversity. Nature ecology & evolution, 6(2):163–173. find
  7. Alexandre Duchêne and Philippe N Humbert. 2018. Surveying languages: the art of governing speakers with numbers. International Journal of the Sociology of Language, 2018(252):1–20. find
  8. David M. Eberhard, Gary F. Simons, and Charles D. Fennig. 2022. Ethnologue: Languages of the World, 25 edition. Dallas: SIL International, Dallas, TX. find
  9. Federico Gaspari, Owen Gallagher, Georg Rehm, Maria Giagkou, Stelios Piperidis, Jane Dunne, and Andy Way. 2022. Introducing the digital language equality metric: Technological factors. In Proceedings of the Workshop Towards Digital Language Equality (TDLE 2022), co-located with LREC 2022, pages 1–12… find
  10. András Kornai. 2015. A new method of language vitality assessment. In Linguistic and Cultural Diversity in Cyberspace. Proceedings of the 3rd International Conference, pages 132–138, Moscow, Russia. Inter-regional Library Cooperation Centre. find
  11. Michael C Gavin, Carlos A Botero, Claire Bowern, Robert K Colwell, Michael Dunn, Robert R Dunn, Russell D Gray, Kathryn R Kirby, Joe McCarter, Adam Powell, and 1 others. 2013. Toward a mechanistic understanding of linguistic diversity. Bio-Science, 63(7):524–535. find
  12. Yoshiki Mikami and Shigeaki Kodama. 2012. Measuring linguistic diversity on the web. In Laurent Vannini and Hervé Le Crosnier, editors, Net.lang (version en anglais): Toward the Multilingual Cyberspace, pages 118–139. C&F Éditions. Chapter in edited volume on multilingual cyberspace. find
  13. Megan L. Gibson. 2015. A framework for measuring the presence of minority languages in cyberspace. In Linguistic and Cultural Diversity in Cyberspace. Proceedings of the 3rd International Conference, pages 61–70, Moscow, Russia. Interregional Library Cooperation Centre. find
  14. Claude E Shannon. 1948. A mathematical theory of communication. The Bell system technical journal, 27(3):379–423. find
  15. Gary F. Simons, Abbey L. Thomas, and Chad K. White. 2022. Assessing digital language support on a global scale. Preprint, arXiv:2209.13515. find
  16. François Grin and Guillaume Fürst. 2022. Measuring linguistic diversity: A multi-level metric. Social indicators research, 164(2):601–621. find
  17. Hanna Tuomisto. 2010. A diversity of beta diversities: Straightening up a concept gone awry. part 1. defining beta diversity as a function of alpha and gamma diversity. Ecography, 33(1):2–22. find
  18. Annika Grützner-Zahn and Georg Rehm. 2022. Introducing the digital language equality metric: Contextual factors. In Proceedings of the Workshop Towards Digital Language Equality (TDLE 2022), co-located with LREC 2022, pages 13–26, Marseille, France. find
  19. A Harald Hammarström, Robert Forkel, Martin Haspelmath, and Sebastian Bank. 2024. Glottolog 5.1. https://glottolog.org. Max Planck Institute for Evolutionary Anthropology, Leipzig. ↗
  20. This appendix summarizes the quantitative measures used to capture linguistic diversity and digital language support at the country level. find
  21. James Hutson, Pace Ellsworth, and Matt Ellsworth. 2024. Preserving linguistic diversity in the digital age: a scalable model for cultural heritage continuity. Journal of Contemporary Language Research, 3(1). find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/zeh-2026-globlingdiv-global-dataset-linking-linguistic/