Elamur Research⌕
☰
home / papers / 2004 / Animacy Encoding in English: Why and How

Animacy Encoding in English: Why and How

Paper · 2004 · Annie Zaenen, Jean Carletta, Gregory Garretson et al.

Summary

The paper argues that animacy, meaning whether a referent is human, another living or agent-like being, or inanimate, matters for practical NLP and especially for generation and translation. In English, animacy does not separate grammatical from ungrammatical sentences, but it softly steers construction choices: Saxon versus of-genitive, double-object versus prepositional dative, and active versus passive.

The authors describe two medium-scale annotation efforts that used nearly the same scheme:

  • Boston: a team labelled 20,000 possessive NPs from the Brown Corpus. An automatic first pass relied on word lists and morphology. Pairs of coders then made consensus decisions with a decision tree.
  • Stanford-Edinburgh paraphrase project: this team annotated Switchboard dialogue and measured agreement between independent coders.

Overall agreement is high, but mostly because HUMAN and NONCONC dominate the data and are easy to tell apart. The intermediate classes (ORG, PLACE, TIME, CONCRETE) were coded inconsistently, and the rare ones were too sparse to judge. The resources therefore give reliable evidence for a binary human-versus-nonconcrete contrast. They also make it possible to study how animacy interacts with syntax and information status.

Key points

  • Scheme: the top-level split is three-way: human, other animate, inanimate.
    • Other animates are subdivided into organizations, animals, intelligent machines and vehicles.
    • Inanimates are subdivided into concrete, non-concrete, place and time.
  • Boston effort: 20,000 possessive NPs from the Brown Corpus. Automated coding came first, followed by manual consensus coding. This design allows reliability to be checked only between the automatic coder and the humans as a group. The data is released as the Boston University Noun Phrase Corpus.
  • Stanford-Edinburgh effort: the paraphrase corpus is built on Switchboard, which already carries syntactic markup. The LINK project adds information status to the same data. The paraphrase corpus is offered to Switchboard subscribers.
  • Sources of disagreement:
    • unclear HUMAN/ORG boundaries;
    • metonymic names such as IBM or Russia;
    • vague spoken "they";
    • unclear TIME cases such as "next time";
    • attention slips, possibly made worse by tools that auto-advance between markables.
  • Rare classes: ANIMAL, MAC and VEH were so rare that no significant reliability results could be obtained for them.

Interesting ideas

  • I1. Bootstrap annotation with an automatic word-list and morphology pass, then have humans correct it, and measure agreement between the machine and the human group.
  • I2. A high kappa can be misleading when two easy, dominant classes swamp the rest, so per-category confusion matrices should be examined.
  • I3. Keep animacy (an inherent property of the entity), definiteness (its status at a point in the discourse) and person (its relation to the speech participants) as three separate scales, even though phenomena often depend on several at once.
  • I4. When realizing a predicate, prefer the more animate argument for prominent slots such as direct object, but weigh this against person, recency and expression length.
  • I5. Animacy belongs to the referent, not the word: names of organizations or places, and vague pronouns like "they", are systematically ambiguous between groups of people, institutions and locations.

For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

  • I1 (learn): silver animacy or entity-type labels could be seeded cheaply this way, with each witness tagger's agreement against a small human-checked sample reported.
  • I2 (learn): evaluation of fine-grained labels such as semantic roles or entity types should report per-class agreement, so that weak middle categories are not hidden.
  • I3 (world): entities in the text world could store these as independent attributes, letting later stages combine them explainably instead of collapsing them into one salience score.
  • I4 (back): generation from the meaning graph could rank argument realizations with explicit, inspectable weights for animacy and these competing factors.
  • I5 (frames): coreference and frame filling could keep such mentions as underspecified ORG/HUMAN/PLACE and resolve them from context, rather than fixing a type from the lexicon.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓Stanford UniversityStanford · United States434 works · 314,419 citations🎓University of EdinburghEdinburgh · United Kingdom359 works · 77,352 citations🔬Palo Alto Research CenterPalo Alto · United States14 works · 8,123 citations🎓Boston UniversityBoston · United States26 works · 4,022 citations

 Fields, topics and concepts

📚 Shelves 4The 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 works

🧭 Topics 1Research subjects and languages the work is about.English · 2,263 works

Cite

Cite as Annie Zaenen, Jean Carletta, Gregory Garretson, Joan Bresnan, Andrew Koontz-Garboden, Tatiana Nikitina, M. Catherine O'Connor and Tom Wasow. 2004. Animacy Encoding in English: Why and How. Proceedings of the Workshop on Discourse Annotation.

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

@inproceedings{zaenen2004animacy,
  title = {{Animacy Encoding in English: Why and How}},
  author = {Annie Zaenen, Jean Carletta, Gregory Garretson, Joan Bresnan, Andrew Koontz-Garboden, Tatiana Nikitina, M. Catherine O'Connor and Tom Wasow},
  booktitle = {Proceedings of the Workshop on Discourse Annotation},
  year = {2004},
  doi = {10.3115/1608938.1608954},
  url = {https://doi.org/10.3115/1608938.1608954}
}

References

25 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Aissen, Judith, 2002. Differential object marking: Iconicity vs. economy. NLLT, 21 435-483. ↗
  2. Anschutz, A. 1997. "How to Choose a Possessive Noun Phrase Construction in Four Easy Steps." Studies in Language 21, 1, 1-35. ↗
  3. Bock, J. K., Loebell, H. and Morey, R. (1992) From conceptual roles to structural relations: Bridging the syntactic cleft. Psychological Review 99: 150–171. ↗
  4. Bossong, Georg. 1985. D i f f e r e n t i e l l e Objektmarkierung in den Neuiranischen Sprachen,Gunter Narr Verlag, Tübingen. find
  5. Bossong, Georg. 1991. 'Differential Object Marking in Romance and Beyond', in D. Wanner and D.Kibbee (eds.), New Analyses in Romance Linguistics: Selected Papers from theXVIII Linguistic Symposium on Romance Languages, John Benjamins, Amsterdam, pp. 143-170. find
  6. Bresnan, Joan, Dingare, Shipra, and Manning, Christopher D. 2001. Soft Constraints Mirror Hard Constraints: Voice and Person in English and Lummi. Proceedings of the LFG '01 Conference. CSLI Online. 20 pages. find
  7. Bresnan, Joan, Carletta, Jean, Crouch, Richard, Nissim, Malvina, Steedman, Mark, Wasow, Tom and Zaenen, Annie. 2002. Paraphrase analysis for improved generation, LINK project, HCRC Edinburgh-CLSI Stanford. find
  8. Carletta, Jean. 1996. Assessing agreement on classification task: the kappa statstic. Computational linguistics, 22 (2): 249-254 find
  9. Carletta, Jean, Shipra Dingare, Malvina Nissim, and Tatiana Nikitina. 2004. Using the NITE XML Toolkit on the Switchboard Corpus to study syntactic choice: a case study. In Proceedings of the 4th International Conference on Language Resources and Evaluation (LREC2004), Lisbon, May 2004. find
  10. Comrie, Bernard, 1989, Language Universals and Linguistic Typology, The University of Chicago Press Cueni, Anna, Joan Bresnan, Tatiana Nikitina and Harald Baayen. 2004. Predicting the Dative Alteration, Stanford University, ms. in preparation. find
  11. Dahl, Östen and Fraurud, Kari. 1996 Animacy in Grammar and Discourse. In Thorstein Fretheim & Jeanette K. Gundel (eds.), Reference and Referent Accessibility, 47-64. Amsterdam/Philadelphia: John Benjamins. ↗
  12. Garretson, Gregory, O'Connor, M. Catherine; Skarabela, Barbora; & Hogan, Marjorie. March 2004. Coding practices used in the project Optimal Typology of Determiner Phrases. http://npcorpus.bu.edu/documentation/index.ht ml Garretson, Gregory, and O’Connor, M. Catherine. 2004. A combined automatic-and-manual method for studying discourse features in corpora. Paper to be presented at the Fifth North American Symposium on Corpus Linguistics, May 21-23, Montclair State University, NJ. ↗
  13. Godfrey, J. Holliman, E. & McDaniel J. 1992 SWITCHBOARD: Telephone speech corpus for research and development. Proceedings of ICASSP-92, 517-520. find
  14. Leech, G., B. Francis, et al. (1994). The Use of Computer Corpora in the Textual Demonstrability of Gradience in Linguistic Categories. Continuity In Linguistic Semantics. C. Fuchs and B. Victorri. Amsterdam, John Benjamins Publishing Company: 57-76. ↗
  15. McDonald, J.L., Bock, K. and M. Kelly. 1993. Word and world order: semantic, phonological, and metrical determinants of serial position. Cognitive Psychology 25: 188-230. ↗
  16. Nissum, Malvina, Shipra Dingare, Jean Carletta and Mark Steedman. 2004. An annotation scheme for information status in dialogue. Submitted to LREC 2004
  17. O’Connor, M. Catherine. 2000. Optimality typology of the DP: Markedness within the nominal. NSP-grant. BCS-0080377. find
  18. O’Connor, M. Catherine., Anttila, Arto, Fong, Vivienne and Maling, Joan (2004). Differential possessor expression in English: Re-evaluating animacy and topicality effects. Paper presented at the Annual Meeting of the Linguistic Society of America, January 9-11, Boston, MA. find
  19. Rosenbach, Anette (2002) Genitive Variation in English. Conceptual Factors in Synchronic and Diachronic Studies. Berlin/New York: Walter de Gruyter.) ↗
  20. Rosenbach, Anette. 2003. Aspects of iconicity and economy in the choice between the s-genitive and the of-genitive in English. In: Günter Rohdenburg & Britta Mondorf (eds.). Determinants of Grammatical Variation in English Berlin/New York: de Gruyter. ↗
  21. Rosenbach, A., J. Aissen, and J. Bresnan. 2002. Pilot study of the influence of animacy on subject choice in a reading task. (Heinrich Heine University, UCSC, and Stanford) find
  22. Silverstein, Michael. 1976. Hierarchy of features and ergativity. In Richard Dixon, editor, Grammatical Categories in Australian Languages. Australian Institute of Aboriginal Studies. ↗
  23. Stefanowitsch, Anatol. 2000. Constructional semantics as a limit to grammatical alternation: The two genitives of English. in CLEAR (Cognitive Linguistics: Explorations, Applications, Research), 3. Trommer, Jochen. s.d. Direction Marking and Case in Menominee, http://www.ling.uni-osnabrueck.de/trommer/nim.pdf ↗
  24. Woolford, Ellen. 1999. Animacy hierarchy effects on object agreement, in Paul Kotey ed. New Dimensions in African Linguistics and Languages (Trends in African Linguistics 3), 203-216. find
  25. Yamamoto, Mutsumi, 1999, Animacy and Reference: a cognitive approach to corpus linguistics , John Benjamins. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/zaenen-2004-animacy-encoding-english-why-how/