Elamur Research⌕
☰
home / papers / 2020 / The Johns Hopkins University Bible Corpus: 1600+ Tongues for Typological Exploration

The Johns Hopkins University Bible Corpus: 1600+ Tongues for Typological Exploration

Summary

In NLP, "low-resource" often still means hundreds of thousands of parallel sentences, yet most of the world's languages have only a few thousand. The authors respond with the Johns Hopkins University Bible Corpus (JHUBC), a multi-parallel resource built on the Bible. Its book–chapter–verse numbering gives translations a shared alignment unit, and its text covers much core vocabulary. They merged earlier collections (Mayer and Cysouw, Asgari and Schütze, the CMU Wilderness Corpus) with new scraping. They then recovered verse boundaries, applied Unicode normalization, stripped notes and cross-references, tokenized the text and removed near-duplicate translations. The result is 4272 Bibles in 1611 languages from more than 90 families, far larger than any Bible corpus available before it. The share of languages per family closely tracks Ethnologue, which suggests the corpus is a reasonable typological sample. As a demonstration, the authors project pronoun distinctions such as clusivity through alignments onto languages that lack them. The resource supports typology, cross-lingual transfer and MT for languages with almost no other text.

Key points

  • JHUBC holds 4272 Bibles in 1611 languages, including 27 English translations. Deduplication cut 4998 collected Bibles down to this number.
  • Compared with the largest earlier Bible collection (1556 Bibles in 1169 languages, now unavailable), JHUBC has about 2.6× more translations and 38% more languages.
  • For 86% of the Bibles, the New Testament (roughly 8k verses) is the sole content. The rest also include the Old Testament (over 31k verses).
  • Family proportions match Ethnologue closely. Niger–Congo makes up 19.43% of JHUBC vs. 20.63% of Ethnologue, and Austronesian 17.19% vs. 16.82%.
  • The release format is plain text with one verse per line, blank lines for missing verses, and a master file mapping lines to verse IDs. Because of copyright, access is granted on request.

Interesting ideas

  • I1. Recovering IDs via longest common subsequence: When verse numbers are mixed into plain text alongside ordinary numbers, aligning the expected sequence 1…m with the extracted numbers via dynamic programming separates true markers from noise. The same trick could restore section or article numbering in scraped laws, manuals or OCR output.
  • I2. Cheap near-duplicate detection: The authors compare edit distance on a shared sample of aligned segments, ignoring non-letter characters, and drop pairs above a similarity threshold. This catches re-uploads that differ only in tokenization or punctuation, which is useful for any web-scraped multilingual collection.
  • I3. Alignment preserved by position: Missing verses are kept as blank lines, and a single global index maps line numbers to verse IDs. Every file therefore stays line-aligned without per-pair alignment files, a simple design for any multi-parallel dataset.
  • I4. Checking representativeness against a reference census: Comparing the corpus's distribution over language families with Ethnologue's gives a quick, quantitative check of sampling bias. An analogous check could compare a corpus against regional or genre distributions.
  • I5. Using many translations of the pivot language: Using several English versions of the same text, rather than one, lets projection methods take a consensus. This reduces noise from any single translation's quirks when transferring annotations such as pronoun features or parses.

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

Cited by

70 papers of our library cite it (of the 140673 whose bibliographies we read); 6 of them have a page here, 64 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work.

@inproceedings{mccarthy2020johns,
  title = {{The Johns Hopkins University Bible Corpus: 1600+ Tongues for Typological Exploration}},
  author = {McCarthy, Arya D. and Wicks, Rachel and Lewis, Dylan and Mueller, Aaron and Wu, Winston and Adams, Oliver and Nicolai, Garrett and Post, Matt and Yarowsky, David},
  booktitle = {Proceedings of the Twelfth Language Resources and Evaluation Conference},
  year = {2020},
  url = {https://aclanthology.org/2020.lrec-1.352/}
}

References

35 references, 9 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Agić, Ž. and Vulić, I. (2019). JW300: A wide-coverage parallel corpus for low-resource languages. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 3204–3210, Florence, Italy, July. Association for Computational Linguistics. find our card
  2. Agić, Ž., Hovy, D., and Søgaard, A. (2015). If all you have is a bit of the Bible: Learning POS taggers for truly low-resource languages. In Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Proces… find
  3. Asgari, E. and Schütze, H. (2017). Past, present, future: A computational investigation of the typology of tense in 1000 languages. In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 113–124, Copenhagen, Denmark, September. Association for Computational… find
  4. Black, A. W. (2019). CMU wilderness multilingual speech dataset. In ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 5971–5975, May. find
  5. Chang, P.-C., Galley, M., and Manning, C. D. (2008). Optimizing Chinese word segmentation for machine translation performance. In Proceedings of the Third Workshop on Statistical Machine Translation, pages 224–232, Columbus, Ohio, June. Association for Computational Linguistics. find our card
  6. Christodouloupoulos, C. and Steedman, M. (2015). A massively parallel corpus: the Bible in 100 languages. Language resources and evaluation, 49(2):375–395. find our card
  7. Duh, K. (2018). The multitarget TED Talks task. http://www.cs.jhu.edu/˜kevinduh/a/multitarget-tedtalks/. ↗
  8. Eberhard, D. M., Simons, G. F., and Fennig, C. D. (2019). Ethnologue: Languages of the World. Twenty-second edition. Dallas, Texas: SIL International. find
  9. Francis, W. N. and Kucera, H. (1964). Brown corpus. Department of Linguistics, Brown University, Providence, Rhode Island, 1. find
  10. Graddol, D. (1997). The future of English?: A guide to forecasting the popularity of the English language in the 21st century. British Council. find
  11. Guzmán, F., Chen, P.-J., Ott, M., Pino, J., Lample, G., Koehn, P., Chaudhary, V., and Ranzato, M. (2019). The FLORES evaluation datasets for low-resource machine translation: Nepali–English and Sinhala–English. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processin… find our card
  12. Kanungo, T., Resnik, P., Mao, S., Kim, D.-W., and Zheng, Q. (2005). The Bible and multilingual optical character recognition. Commun. ACM, 48(6):124–130, June. find
  13. Kirov, C., Cotterell, R., Sylak-Glassman, J., Walther, G., Vylomova, E., Xia, P., Faruqui, M., Mielke, S. J., McCarthy, A., Kübler, S., Yarowsky, D., Eisner, J., and Hulden, M. (2018). UniMorph 2.0: Universal morphology. In Proceedings of the 11th Language Resources and Evaluation Conference, Miyaz… find
  14. Koehn, P. (2005). Europarl: A parallel corpus for statistical machine translation. In MT summit, volume 5, pages 79–86. find
  15. Levy, O., Søgaard, A., and Goldberg, Y. (2017). A strong baseline for learning cross-lingual word embeddings from sentence alignments. In Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers, pages 765–774, Valencia, Spain… find
  16. Littell, P., Mortensen, D. R., Lin, K., Kairis, K., Turner, C., and Levin, L. (2017). Uriel and lang2vec: Representing languages as typological, geographical, and phylogenetic vectors. In Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Vol… find our card
  17. Mayer, T. and Cysouw, M. (2014). Creating a massively parallel Bible corpus. In Proceedings of the Ninth International Conference on Language Resources and Evaluation (LREC’14), pages 3158–3163, Reykjavik, Iceland, May. European Language Resources Association (ELRA). find our card
  18. McCarthy, A. D., Li, X., Gu, J., and Dong, N. (2019). Improved Variational Neural Machine Translation by Promoting Mutual Information. arXiv e-prints, page arXiv:1909.09237v1, Sep. — [arXiv:1909.09237](https://arxiv.org/abs/1909.09237) find
  19. McCarthy, A. D., Kirov, C., Grella, M., Nidhi, A., Xia, P., Gorman, K., Vylomova, E., Mielke, S. J., Nicolai, G., Silfverberg, M., Arkhangelskij, T., Krizhanovsky, N., Krizhanovsky, A., Kylachko, E., Sorokin, A., Mansfield, J., Ernstreits, V., Pinter, Y., Jacobs, C., Cotterell, R., Hulden, M., and Y… find our card
  20. Mueller, A., Nicolai, G., McCarthy, A. D., Lewis, D., Wu, W., and Yarowsky, D. (2020). An analysis of massively multilingual neural machine translation for low-resource languages. In Proceedings of the Twelfth International Conference on Language Resources and Evaluation (LREC 2020), Marseilles, Fra… find
  21. Nicolai, G. and Yarowsky, D. (2019). Learning morphosyntactic analyzers from the Bible via iterative annotation projection across 26 languages. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pages 1765–1774, Florence, Italy, July. Association for Computat… find
  22. Nicolai, G., Lewis, D., McCarthy, A. D., Mueller, A., Wu, W., and Yarowsky, D. (2020). Fine-grained morphosyntactic analysis and generation tools for more than one thousand languages. In Proceedings of the Twelfth International Conference on Language Resources and Evaluation (LREC 2020), Marseilles,… find
  23. Qi, Y., Sachan, D., Felix, M., Padmanabhan, S., and Neubig, G. (2018). When and why are pre-trained word embeddings useful for neural machine translation? In Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies… find our card
  24. Resnik, P., Olsen, M. B., and Diab, M. (1999). The Bible as a parallel corpus: Annotating the ‘book of 2000 tongues’. Computers and the Humanities, 33(1):129–153, Apr. find
  25. Schlichtkrull, M. and Søgaard, A. (2017). Cross-lingual dependency parsing with late decoding for truly low-resource languages. In Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics: Volume 1, Long Papers, pages 220–229, Valencia, Spain, April… find
  26. Summers, D. and Gadsby, A. (1995). Longman dictionary of contemporary English. Longman Harlow. find
  27. Sylak-Glassman, J., Kirov, C., Yarowsky, D., and Que, R. (2015). A language-independent feature schema for inflectional morphology. In Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Vo… find
  28. Tiedemann, J. (2012). Parallel data, tools and interfaces in OPUS. In Proceedings of the Eighth International Conference on Language Resources and Evaluation (LREC-2012), pages 2214–2218, Istanbul, Turkey, May. European Language Resources Association (ELRA). find
  29. Ure, J. (1971). Lexical density and register differentiation. Applications of linguistics, pages 443–452. find
  30. Wagner, R. A. and Fischer, M. J. (1974). The string-tostring correction problem. J. ACM, 21(1):168–173, January. find
  31. Wu, W. and Yarowsky, D. (2018). A comparative study of extremely low-resource transliteration of the world’s languages. In Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018), Miyazaki, Japan, May. European Language Resources Association (ELRA). find
  32. Wu, W., Vyas, N., and Yarowsky, D. (2018). Creating a translation matrix of the Bible’s names across 591 languages. In Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018), Miyazaki, Japan, May. European Language Resources Association (ELRA). find our card
  33. Wu, W., Nicolai, G., and Yarowsky, D. (2020). Multilingual dictionary based construction of core vocabulary. In Proceedings of the Twelfth International Conference on Language Resources and Evaluation (LREC 2020), Marseilles, France, May. European Language Resources Association (ELRA). find
  34. Xia, P. and Yarowsky, D. (2017). Deriving consensus for multi-parallel corpora: an English Bible study. In Proceedings of the Eighth International Joint Conference on Natural Language Processing (Volume 2: Short Papers), pages 448–453, Taipei, Taiwan, November. Asian Federation of Natural Language P… find
  35. Yarowsky, D., Ngai, G., and Wicentowski, R. (2001). Inducing multilingual text analysis tools via robust projection across aligned corpora. In Proceedings of the First International Conference on Human Language Technology Research. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/mccarthy-2020-johns-hopkins-university-bible-corpus/