Core Arguments in Universal Dependencies
Paper · 2017 · Daniel Zeman
Summary
The paper examines how, in Universal Dependencies, to distinguish core arguments of a verb from oblique ones when a language has both case morphology and prepositions. The UD guidelines merely list possible criteria and give no clear definition. The English rule "there is a preposition, hence it is not core" does not work for such languages. The author takes the methodology of Andrews (2007) as a basis. First, prototypical transitive verbs are identified, i.e., those that have an agent and a patient. Then one examines how the language encodes their arguments and extends this encoding to the remaining verbs. This methodology is applied to German, Czech, and Russian. Statistics from 26 UD treebanks show that annotators draw the boundary between core and oblique in different ways. As a result, annotation across corpora is inconsistent. The author proposes refining the guidelines: defining the properties of core arguments separately for each language and not abandoning the argument–adjunct distinction entirely. For Ukrainian, which also has a rich case system, this bears directly on the annotation of objects in oblique cases.
Key points
- The methodology rests on the functions S, A, and P from Andrews (2007). Arguments with these functions become
nsubjorobjin UD, while the remaining nominal phrases receiveobl. - German helfen and Czech pomoci govern the dative case. Nevertheless, their prepositionless arguments are recognized as core. Czech verbs can have a patient in the instrumental (hýbat) as well as in the genitive (dotknout se).
- Tables for UD Czech and two versions of Russian SynTagRus (1.3 and 2.0) show that the corpora annotate identical case forms very differently. For example, in SynTagRus 2.0 the prepositionless dative is almost always (about 95%) labeled
obl. - Genitive subjects in Slavic languages arise because of numerals, not because of the verb. The Russian dative in constructions such as «мне холодно» fails all three subject tests: case, agreement, and subject inheritance in an infinitival construction.
- The author proposes introducing the relation
obl:argto preserve the argument–adjunct distinction where it can be established.
Institutions
Fields, topics and concepts
📚 Shelves 6The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 worksPsychology · 1,552 worksExperimental and Cognitive Psychology · 870 works
🧭 Topics 1Research subjects and languages the work is about.Universal Dependencies · 184 works
Cite
Cite as Daniel Zeman. 2017. Core Arguments in Universal Dependencies. Proceedings of the Fourth International Conference on Dependency Linguistics (Depling 2017).
@inproceedings{zeman2017core,
title = {{Core Arguments in Universal Dependencies}},
author = {Daniel Zeman},
booktitle = {Proceedings of the Fourth International Conference on Dependency Linguistics (Depling 2017)},
year = {2017},
url = {https://aclanthology.org/W17-6532/}
}References
- Avery D. Andrews. 2007. The major functions of the noun phrase. In Timothy Shopen, editor, Language Typology and Syntactic Description. Volume I: Clause Structure. Second edition, pages 132–223. Cambridge University Press, Cambridge, UK. find
- Robert M. W. Dixon. 2012. Basic Linguistic Theory. Oxford University Press, Oxford, UK. find
- Matthew S. Dryer. 2007. Clause types. In Timothy Shopen, editor, Language Typology and Syntactic Description. Volume I: Clause Structure, pages 224– 275. Cambridge University Press, Cambridge, UK. find
- William A. Foley. 2007. A typology of information packaging in the clause. In Timothy Shopen, editor, Language Typology and Syntactic Description. Volume I: Clause Structure, pages 362–446. Cambridge University Press, Cambridge, UK. find
- Jan Hajič, Jarmila Panevová, Eva Hajičová, Petr Sgall, Petr Pajas, Jan Štěpánek, Jiří Havelka, Marie Mikulová, Zdeněk Žabokrtský, and Magda Ševčíková-Razímová. 2006. Prague Dependency Treebank 2.0. CD-ROM, Linguistic Data Consortium, LDC Catalog No.: LDC2006T01, Philadelphia. find
- Bohuslav Havránek and Alois Jedlička. 1966. Stručná mluvnice česká. Fortuna, Praha, Czechia. find
- Gerhard Helbig and Joachim Buscha. 1998. Deutsche Grammatik. Ein Handbuch für den Ausländerunter-richt (18. Auflage). Langenscheidt, Leipzig, Germany. find
- Petr Karlík, Marek Nekula, Jana Pleskalová, et al. 2016. Nový encyklopedický slovník češtiny. Nakla-datelství Lidové noviny, Praha, Czechia. find
- (6) (1) nsubj (7) (8) case obj:nacc Drak pohnul skálou Nom Ins Dragon moved rock (9) nsubj obl expl:pv det Drak se vrací každý rok Nom Acc Acc Dragon returns every year obj nsubj iobj Král dal Jiřímu zlato Nom Dat Acc King gave George gold (11) nsubj:pass obl:agent aux:pass iobj Drak byl zabit Jiřím… find
Updated
This page in Markdown → https://md.research.elamur.ai/free/en/papers/zeman-2017-core-arguments-universal-dependencies/