Elamur Research⌕
☰
home / papers / 2015 / The Role of Expectedness in the Implicitation and Explicitation of Discourse Relations

The Role of Expectedness in the Implicitation and Explicitation of Discourse Relations

Written for the Research library (claude-opus-5-5) from the paper text, in our own words: the license "CC BY-NC-SA 3.0 (ACL Anthology, before 2016)" does not allow redistributing the text itself. Original — aclanthology.org/W15-2505.pdf

Summary

Human translators drop or add discourse connectives much more freely than statistical MT systems, and MT systems that omit connectives can distort meaning. This paper asks whether a relation's expectedness determines how readily it shifts between explicit and implicit marking in translation. Expectedness is grounded in Murray's continuity hypothesis and Sanders' causality-by-default hypothesis. The authors drew English–Dutch and English–German fragments from the Europarl Direct corpus, covering both translation directions. They focused on prototypical causal, negative and conditional connectives. Each connective was manually coded for how it was rendered in the other language: explicit, paraphrase, underspecified connective, syntax, or implicit. These five codes were then collapsed into implicitation/explicitation versus explicit-to-explicit and analysed with log-linear models. Causal relations, the expected type, were the ones most often added and most often removed. Negative relations came next, and conditional relations shifted least. This held regardless of language pair. The findings suggest that MT models should tell apart truly implicit renderings from alternative explicit encodings, and should model when a relation may safely stay unmarked.

Key points

  • Connectives studied: because/although/if (EN), omdat/hoewel/als (DU), weil/obwohl/wenn (GE). The sample had 250 instances per connective, giving 3000 ST–TT pairs in total.
  • Two annotators double-coded 6×50 fragments and reached κ = 0.84. One annotator then completed the rest.
  • Causal relations showed more implicit counterparts than negative ones (z = 6.21), and negative more than conditional (z = 4.72). Relation type did not interact with language-pair differences.
  • EN–GE showed more explicitations than implicitations, while EN–DU was roughly balanced.
  • Overall implicitation was 3.6%, rising to 11.2% if paraphrases are counted. This is far below the up-to-18% that Meyer and Webber (2013) report for human translators, which the authors attribute to their broad paraphrase category.

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

Topics: Discourse analysis

Cited by

8 papers of our library cite it (of the 140673 whose bibliographies we read); 2 of them have a page here, 6 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/w15-2505

@inproceedings{hoek2015role,
  title = {{The Role of Expectedness in the Implicitation and Explicitation of Discourse Relations}},
  author = {Hoek, Jet and Evers-Vermeul, Jacqueline and Sanders, Ted J.M.},
  booktitle = {Proceedings of the Second Workshop on Discourse in Machine Translation},
  year = {2015},
  doi = {10.18653/v1/w15-2505},
  url = {https://doi.org/10.18653/v1/w15-2505}
}

References

22 references, 4 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Fatemeh T. Asr and Vera Demberg. 2012. Implicitness of discourse relations. Proceedings of COLING. Mumbai, India. find our card
  2. Viktor Becher. 2011. When and why do translators add connectives? Target, 23(1):26– 47. find
  3. Bruno Cartoni, Sandrine Zufferey and Thomas Meyer. 2013. Using the Europarl corpus for linguistics research. Belgian Journal of Linguistics, 27:23–42. find
  4. Bruno Cartoni, Sandrine Zufferey, Thomas Meyer and Andrei Popescu-Belis. 2011. How comparable are parallel corpora? Measuring the distribution of general vocabulary and connectives. Proceedings of the 4 th Workshop on Building and Using Comparable Corpora, 78–86. Portland, Oregon. find
  5. Debopam Das and Maite Taboada. 2013. Explicit and implicit coherence relations: A corpus study. Proceedings of the 2013 Annual Conference of the Canadian Linguistic Association. Victoria, Canada. find
  6. Sandra Halverson. 1996. Norwegian-English translation and the role of certain connectives. Translation and Meaning, part 3: Proceedings of the Maastricht Session of the 2 nd International Maastricht-Lodź Duo Colloquium on ‘Translation and Meaning’, 128–139. Maastricht, the Netherlands. find
  7. Sandra Halverson. 2004. Connectives as a translation problem. In Harald Kittel, Armin Paul Frank, Norbert Greiner, Theo Hermans, Werner Koller, José Lambert and Fritz Paul. (Eds.), An International Encyclopedia of Translation Studies, 562–572. Berlin/New York: Walter de Gruyter. find
  8. Silvia Hansen-Schirra, Stella Neuman and Erich Steiner. 2007. Cohesive explicitness and explicitation in an English-German translation corpus. Languages in Contrast 7(2):241–265. find
  9. Jet Hoek and Sandrine Zufferey. 2015. Factors influencing the implicitation of discourse relations across languages. Proceedings 11 th Joint ACL-ISO Workshop on Interoperable Semantic Annotation (ISA-11), 39–45. London, United Kingdom. find our card
  10. Philipp Koehn. 2005. Europarl: A parallel corpus for statistical machine translation. Proceedings of the 10 th Machine Translation Summit, 79– 86. Phuket, Thailand. find
  11. Arnout W. Koornneef and Ted J. M. Sanders. 2013. Establishing coherence relations in discourse: The influence of implicit causality and connectives on pronoun resolution. Language and Cognitive Processes, 28(8):1169–1206. find
  12. Gina R. Kuperberg, Martin Paczynski and Tali Ditman. 2011. Establishing causal coherence across sentences: An ERP study. Journal of Cognitive Neuroscience, 23:1230–1246. find
  13. Junyi Jessy Li, Marine Carpuat and Ani Nenkova. 2014. Assessing the discourse factors that influence the quality of machine translation. Proceedings of the 52 nd Annual Meeting of the Association for Computational Linguistics, 283–288. Baltimore, MD, USA. find
  14. Willem M. Mak and Ted J. M. Sanders. 2013. The role of causality in discourse processing: Effects of expectation and coherence relations. Language and Discourse Processes, 28(9):1414–1437. find
  15. Thomas Meyer and Bonnie Webber. 2013. Implicitation of discourse connectives in (machine) translation. Proceedings of the 1 st DiscoMT Workshop at ACL 2013, 33–42. Sofia, Bulgaria. find our card
  16. John D. Murray. 1997. Connectives and narrative text: The role of continuity. Memory and Cognition, 25:227–236. find
  17. Rashmi Prasad, Nikhil Dinesh, Alan Lee, Eleni Miltsakaki, Livio Robaldo, Aravind Joshi and Bonnie Webber. 2008. The Penn Discourse Treebank 2.0. Proceedings of the 6 th International Conference on Language Resources and Evaluation (LREC 2008), 2961–2968. Marrakech, Morocco. find our card
  18. Ted J. M. Sanders. 2005. Coherence, causality and cognitive complexity in discourse. Proceedings/Actes SEM-05, First International Symposium on the Exploration and Modelling of Meaning, 105–114. find
  19. Ted J. M. Sanders and Leo G. M. Noordman. 2000. The role of coherence relations and their linguistic markers in text processing. Discourse Processes, 29:37–60. find
  20. David Steele and Lucia Specia. 2014. Divergences in the usage of discourse markers in English and Mandarin Chinese. Proceedings of the 17th International Conference on Text, Speech and Dialogue, 189–200. Brno, Czech Republic. find
  21. Yannick Versley. 2013. A graph-based approach for implicit discourse relations. Computational Linguistics in the Netherlands Journal, 3:148– 173. find
  22. Sandrine Zufferey and Bruno Cartoni. 2014. A multifactorial analysis of explicitation in translations. Target, 26:361–384. find

↑ back to top