Towards a Universal Dependency Corpus for Old Saxon (Old Low German)

What it is about

Old Saxon sits in the West Germanic dialect continuum between Old English and Old High German. Its main text, the 9th-century Heliand, has been annotated three times, each time in a different tradition: Penn-style constituency (HeliPaD, 46,000 tokens, built on the Old English Penn scheme), tier-based morphosyntax (DDD), and partial phrase chunks (B4). The authors annotate a small UD gold set by hand, about 2,000 tokens from Heliand fit 1 and a Genesis fragment, and write a declarative Penn-to-UD converter. The converter has 12 head-finding rules (21 SPARQL updates), each logged with the rule that fired. The editions of the two corpora are tokenized differently, so the converted HeliPaD is aligned to the gold before scoring, and only 678 tokens can be compared. On those, UPOS is 96.9%, UAS 76.8% and LAS 58.6%. Most label errors are differences of convention rather than of analysis: nmod:gen against nmod, det:poss against det, and/or/but as cc or mark, and datives as iobj where UD v2 recommends obl for German-like languages. Fixing these would raise LAS to 68.4%.

Ideas

• I1. Head rules as a logged, ordered list. A Penn-to-UD converter for historical English trees (PPCEME-style brackets, for evaluation only if the licence allows) can be a short ordered rule list in Rust where each attachment records the rule that made it. That is the same why trace Mova keeps elsewhere, and it makes a converter error findable.
• I2. Before reading a low LAS as parser error, sort the label disagreements into convention classes. The paper's three classes (genitive subtype, possessive subtype, cc vs mark for coordinators) can become a tools/parsecmp report that counts "same arc, convention label" separately from real errors.
• I3. Decide lemma-based labels for coordinators once: and, or, but → cc, whatever the clause analysis. A small gate in tools/treeagree can enforce this on the school silver and on the Torah parses (and-chains are the backbone of biblical style).
• I4. Dative arguments: when the biblical register gives an archaic dative-like reading ("he gave him bread", "the ethical dative" in 8. Word order and phrase structure), follow UD v2 English practice (iobj for the bare second object, obl for PP datives) and test it with a count over the KJV parses.

For Mova

8. Word order and phrase structure (datives, and-chains) and 11-biblical-style.md; 24. Coordination and subordination for the cc/mark decision. Code: tools/treeagree (gates, I3), tools/parsecmp (convention-class report, I2), en/src/udreg.rs for allowed subtypes. Measure: LAS with and without the convention classes on EWT dev and on the Torah parses.

Cite

The paper itself, not this card. BibTeX from the library's record of the work. DOI 10.63317/592dfv44cgke

@inproceedings{chiarcos2026universal,
  title = {{Towards a Universal Dependency Corpus for Old Saxon (Old Low German)}},
  author = {Chiarcos, Christian and Siewert, Janine},
  booktitle = {Proceedings of the Ninth Workshop on Universal Dependencies (UDW 2026)},
  year = {2026},
  doi = {10.63317/592dfv44cgke},
  url = {https://doi.org/10.63317/592dfv44cgke}
}

References

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

1. Þórunn Arnardóttir, Hinrik Hafsteinsson, Einar Freyr Sigurðsson, Kristín Bjarnadóttir, Anton Karl Ingason, Hildur Jónsdóttir, and Steinþór Steingrímsson. 2020. A Universal Dependencies conversion pipeline for a Penn-format constituency treebank. In Proceedings of the Fourth Workshop on Universal Dependencies (UDW-2020), pages 16–25, Barcelona, Spain (online). find
2. Otto Behaghel and Burkhard Taeger. 1984. Heliand und Genesis, 9 edition. Max Niemeyer Verlag, Tübingen. find
3. Cathy Bow, Baden Hughes, and Steven Bird. 2003. Towards a general model of interlinear text. In Proceedings of E-MELD Workshop 2003: Digitizing and annotating texts and field recordings, pages 11–13, East Lansing, Michigan. find
4. Christian Chiarcos and Christian Fäth. 2017. CoNLL-RDF: Linked corpora done in an NLP-friendly way. In International Conference on Language, Data and Knowledge, pages 74–88. Springer. find
5. Christian Chiarcos and Christian Fäth. 2019. Graph-based annotation engineering: towards a gold corpus for Role and Reference Grammar. In 2nd Conference on Language, Data and Knowledge (LDK 2019). Schloss Dagstuhl-Leibniz-Zentrum fuer Informatik. find
6. Christian Chiarcos, Christian Fäth, and Maxim Ionov. 2022. Querying a dozen corpora and a thousand years with Fintan. In Proceedings of the Thirteenth Language Resources and Evaluation Conference (LREC-2022), pages 4011–4021, Marseille, France. find
7. Christian Chiarcos and Luis Glaser. 2020. A Tree Extension for CoNLL-RDF. In Proceedings of the 12th Language Resources and Evaluation Conference (LREC-2020), pages 7161–7169, Marseille, France (online). find
8. Christian Chiarcos, Maxim Ionov, Luis Glaser, and Christian Fäth. 2021. An Ontology for CoNLL-RDF: Formal Data Structures for TSV Formats in Language Technology. In 3rd Conference on Language, Data and Knowledge (LDK 2021). Schloss Dagstuhl-Leibniz-Zentrum für Informatik. find
9. Christian Chiarcos and Niko Schenk. 2018. The ACoLi CoNLL libraries: Beyond tab-separated values. In In Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018), Miyazaki, Japan. find
10. Christian Chiarcos and Janine Siewert. 2026. Consolidating syntactically annotated corpora with LLOD technology. An experiment in the Old Saxon Heliand. In Tenth Workshop on Linked Data in Linguistics (LDL-2026), Palma de Mallorca, Spain. Co-located with LREC 2026. find
11. Stefanie Dipper, Karin Donhauser, Thomas Klein, Sonja Linde, Stefan Müller, and Klaus-Peter Wegera. 2013. HiTS: Ein Tagset für historische Sprachstufen des Deutschen. Journal for Language Technology and Computational Linguistics, 28(1):85–137. find
12. Stefanie Dipper, Cora Haiber, Anna Maria Schröter, Alexandra Wiemann, and Maike Brinkschulte. 2024. Universal Dependencies: Extensions for modern and historical German. In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), pages 17101–17111. find
13. Christian Fäth, Christian Chiarcos, Björn Ebbrecht, and Maxim Ionov. 2020. Fintan - Flexible, integrated transformation and annotation engineering. In Proceedings of the 12th Language Resources and Evaluation Conference (LREC-2020), pages 7212–7221, Marseille, France (online). find
14. Cora Haiber. 2024. A Crosslingual Approach to Dependency Parsing for Middle High German. In Proceedings of the 20th Conference on Natural Language Processing (KONVENS 2024), pages 23–31, Vienna, Austria. find
15. Caitlin Light. 2012. The information structure of subject extraposition in Early New High German. University of Pennsylvania Working Papers in Linguistics, 18(1):20. find
16. Sonja Linde. 2009. Aspects of word order and information structure in Old Saxon. In Roland Hinterhölzl and Svetlana Petrova, editors, Information structure and language change: New approaches to word order variation in Germanic, pages 367–389. Walter de Gruyter. find
17. Sonja Linde and Roland Mittmann. 2013. Old German Reference Corpus: Digitizing the knowledge of the 19th century. In Paul Bennett, Martin Durrell, Silke Scheible, and Richard J Whitt, editors, New Methods in Historical Corpora, pages 235–246. Narr Francke Attempto Verlag, Tübingen. find
18. Rosemarie Lühr. 2025. Reflexivität im Altsächsischen. In Norbert Kössinger, editor, Altsächsisch: Beiträge zur altniederdeutschen Sprache, Literatur und Kultur. Walter de Gruyter. find
19. Svetlana Petrova and Michael Solf. 2009. On the methods of information-structural analysis in historical texts: A case study on Old High German. In Roland Hinterhölzl and Svetlana Petrova, editors, Information structure and language change: New approaches to word order variation in Germanic, pages 121–203. Walter de Gruyter. find
20. Svetlana Petrova, Michael Solf, Julia Ritz, Christian Chiarcos, and Amir Zeldes. 2009. Building and using a richly annotated interlinear diachronic corpus: The case of Old High German Tatian. In Traitement Automatique des Langues, Volume 50, Numéro 2: Langues anciennes [Ancient Languages], pages 47–71. find
21. Susan Pintzuk and Ann Taylor. 1997. Annotating the Helsinki Corpus: The Brooklyn-Geneva-Amsterdam-Helsinki Parsed Corpus of Old English and the Penn-Helsinki Parsed Corpus of Middle English. In Tracing the Trail of Time: Proceedings from the Second Diachronic Corpora Workshop, New College, University of Toronto, Toronto, May 1995, pages 91 – 104. Brill, Leiden, Niederlande. find
22. Eiríkur Rögnvaldsson, Anton Karl Ingason, and Einar Freyr Sigurðsson. 2011. Coping with Variation in the Icelandic Parsed Historical Corpus (IcePaHC). Language Variation Infrastructure, 3:97–112. find our card
23. Alessio Salomoni. 2021. A UD Literary Treebank for German. Ph.D. thesis, Università degli studi di Bergamo, Italy. find
24. Beatrice Santorini. 1993. The rate of phrase structure change in the history of Yiddish. Language Variation & Change, 5(3). find
25. Christopher D. Sapp, Elliott Evans, Rex Sprouse, and Daniel Dakota. 2024. Introducing a Parsed Corpus of Historical High German. In Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), pages 9224–9233, Torino, Italia. ELRA and ICCL. find
26. Silke Scheible, Richard J Whitt, Martin Durrell, and Paul Bennett. 2011. A gold standard corpus of Early Modern German. In Proceedings of the 5th Linguistic Annotation Workshop (LAW V), held in conjunction with ACL-HLT 2011, pages 124–128, Portland, Oregon, USA. find our card
27. Mariana Scott. 1966. The Heliand: Translated from the Old Saxon. University of North Carolina Press, Chapel Hill. find
28. Edward Henry Sehrt. 1925. Vollständiges Wörterbuch zum Heliand und zur altsächsischen Genesis. Vandenhoeck & Ruprecht. find
29. Eduard Sievers. 1878. Heliand. Buchhandlung des Waisenhauses, Halle. find
30. Petrova Svetlana, Donhauser Karin, and Linde Sonja. 2009. Heliand (Version 1.0). Humboldt-Universität zu Berlin. find
31. Ann Taylor, Mitchell Marcus, and Beatrice Santorini. 2003. The Penn Treebank: An overview. In Anne Abeillé, editor, Treebanks: Building and using parsed corpora, pages 5–22. Kluwer Academic Publishers, Dordrecht/Boston/London. find
32. Robert D Van Valin. 2014. Role and reference grammar. In The Routledge Handbook of Syntax, pages 579–603. Routledge. find
33. George Walkden. 2016. The HeliPaD: a parsed corpus of Old Saxon. International Journal of Corpus Linguistics, 21(4):559–571. find
34. Lars Erik Zeige, Gohar Schnelle, Martin Klotz, Karin Donhauser, Jost Gippert, and Rosemarie Lühr. 2021. Deutsch Diachron Digital - Referenzkorpus Altdeutsch (Version 1.2). Humboldt-Universität zu Berlin. find

This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/chiarcos-2026-universal-dependency-corpus-old-saxon/