A Classifier of Word-Level Variants in Witnesses of Biblical Hebrew Manuscripts

What it is about

Stemmatology reconstructs how manuscripts descend from one another from the variants between witnesses. The authors work on Ben Sira, with nine witnesses from the Dead Sea Scrolls and the Cairo Geniza, and annotate aligned word pairs as plus/minus, inversion, morphological, lexical or unclassifiable, with subcategories such as determination or preposition (4,046 words). Because the data is small, they add synthetic pairs generated by hand-written rules from randomised Qumran words. The strongest classifier is a Random Forest over the two words, POS tags, Levenshtein distance and rule-based features (F1 0.80); neural classifiers on DictaBERT did not win. Silver POS labels from DictaBERT-morph agree with the gold only 0.25 of the time, while number and gender agree 0.9 or more.

Ideas

• I1. The variant categories as a deterministic classifier over aligned Hebrew word pairs: the rule set plus a Levenshtein threshold gives a first pass, with its errors listed.
• I2. Treat automatic POS on ancient Hebrew from Modern Hebrew models as unreliable: check any feature that uses POS against gold labels first.
• I3. Generate minimal pairs by rule (a suffix pronoun added or removed, a preposition added, a number changed) as balanced test items for a morphology rule.

For Mova

I1 is a tool beside tools/ocrfix (witness comparison), if ever taken up. I2 is a caution for 13. Biblical Hebrew in Mova: the tools. I3 is a generator of test items for tools/hbo.

Cite

The paper itself, not this card. BibTeX from the library's record of the work. DOI 10.18653/v1/2025.findings-acl.1098

@inproceedings{nikolovastoupak2025classifier,
  title = {{A Classifier of Word-Level Variants in Witnesses of Biblical Hebrew Manuscripts}},
  author = {Nikolova-Stoupak, Iglika and Amblard, Maxime and Robert-Hayek, Sophie and Rey, Frédérique},
  booktitle = {Findings of the Association for Computational Linguistics: ACL 2025},
  year = {2025},
  doi = {10.18653/v1/2025.findings-acl.1098},
  url = {https://doi.org/10.18653/v1/2025.findings-acl.1098}
}

References

25 references, 5 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Akademie der Wissenschaften zu Göttingen. 2021. Qumran-digital: Ein komplettes philologisches qumran-lexikon zum hebräischen und aramäischen. Accessed: 18.10.2022. find
2. Pancratius C. Beentjes. 1997. Book of Ben Sira in Hebrew, volume 68 of Vetus Testamentum, Supplements. Brill. find
3. Zeev Ben-Hayyim. 1973. The Book of Ben Sira: Text, Concordance and an Analysis of the Vocabulary. Academy of the Hebrew Language and the Shrine of the Book. find
4. Frédérique Michèle Rey and Eric Reymond. 2024. A Critical Edition of the Hebrew Manuscripts of Ben Sira: With Translations and Philological Notes. Brill, Leiden. find
5. Bernard Cerquiglini. 1983. Eloge de la variante. Langages, (69):25–35. find
6. Philipp Roelli, editor. 2020. Handbook of Stemmatology. De Gruyter Reference. De Gruyter, Berlin/Boston. find
7. Avihay Chriqui and Inbal Yahav. 2021. Hebert and hebemo: a hebrew bert model and a tool for polarity analysis and emotion recognition. CoRR, abs/2102.01909. ↗ our card
8. William M. Schniedewind. 2013. A Social History of Hebrew: Its Origins Through the Rabbinic Period. find
9. Michael Collins, Jan Hajic, Lance Ramshaw, and Christoph Tillmann. 1999. A statistical parser for czech. In Proceedings of the 37th Annual Meeting of the Association for Computational Linguistics (ACL). find
10. Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. Bert: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers), pages 4171–4186. Association for Computational Linguistics. find
11. Eylon Gueta, Avi Shmidman, Shaltiel Shmidman, Cheyn Shmuel Shmidman, Joshua Guedalia, Moshe Koppel, Dan Bareket, Amit Seker, and Reut Tsarfaty. 2023. Large pre-trained models with extra-large vocabularies: A contrastive analysis of hebrew bert models and a new one to outperform them all. Preprint, arXiv:2211.15199. ↗ our card
12. Tuomas Heikkilä. 2023. Computer-Assisted Stemmatology. Routledge. find
13. Katherine L.(Ed.) Jansen. 1990. The new philology. Speculum, 65(1). find
14. Geoffrey Khan, Shmuel Bolozky, Steven E. Fassberg, Gary A. Rendsburg, Aaron D. Rubin, Ora R. Schwarzwald, and Tamar Zewi, editors. 2013. Encyclopedia of Hebrew Language and Linguistics, 1 edition, volume 1-4. Brill, Leiden. find
15. Matej Klemen, Luka Krsnik, and Marko Robnik-Šikonja. 2023. Enhancing deep neural networks with morphological information. Natural Language Engineering, 29(2):360–385. find
16. Yuval Marton, Nizar Habash, and Owen Rambow. 2010. Improving arabic dependency parsing with lexical and inflectional morphological features. In Proceedings of the NAACL HLT 2010 First Workshop on Statistical Parsing of Morphologically-Rich Languages, pages 13–21. find our card
17. Ray Pritz. 2016. Biblical hebrew and modern hebrew: How much do they understand? Journal of Biblical Text Research, 38:203–219. find
18. Miguel Pérez Fernández and John F. Elwolde. 1999. An Introductory Grammar of Rabbinic Hebrew. Brill. find
19. Avi Shmidman, Joshua Guedalia, Shaltiel Shmidman, Moshe Koppel, and Reut Tsarfaty. 2020. A novel challenge set for hebrew morphological disambiguation and diacritics restoration. In Findings of the Association for Computational Linguistics: EMNLP 2020, pages 3316–3326, Online. Association for Computational Linguistics. find our card
20. Shaltiel Shmidman, Avi Shmidman, and Moshe Koppel. 2023. Dictabert: A state-of-the-art bert suite for modern hebrew. Preprint, arXiv:2308.16687. ↗ our card
21. Eveline van Staalduine-Sulman. 2005. Vowels in the trees: The role of vocalisation in stemmatology. Aramaic Studies, 3:215–240. find
22. Angel Sáenz-Badillos. 1993. A History of the Hebrew Language. Cambridge University Press, Cambridge. find
23. Angela Taylor. 2019. A contrastive analysis between biblical and modern hebrew in the context of the book of ruth. Journal of Biblical Studies. find
24. Reut Tsarfaty and Khalil Sima’an. 2007. Three-dimensional parametrization for parsing morphologically rich languages. In Proceedings of the Tenth International Conference on Parsing Technologies, pages 156–167. find
25. École Biblique et Archéologique Française de Jérusalem, editor. 1955-1982. Discoveries in the Judean Desert. Clarendon Press, Oxford. find

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/findings-2025-stoupak-classifier-word-level-variants-witnesses/