Ma(r)king concessions in English and German
What it is about
The paper is about choosing a concessive discourse marker (although, though, nevertheless, despite; German obwohl, trotzdem, zwar ... aber) when a generator has to signal a concession. The authors first separate three pragmatic uses from corpus concordances: convincing the hearer by conceding a counter-argument (Concede-I), denying a false implicature after the main point (Concede-II), and informing about a surprising correlation (Concede-III). They then give one knowledge pattern for all concessions: A implies C, B implies not-C, and the A-B link is the common ground. The main finding is that the pragmatic class fixes the order of the parts and the syntax, but not the marker; within each syntactic group the markers are largely interchangeable, differing in intensity, register and conditionality. They propose a generator in which an RST-like tree is linearised into sentence plans and then realised by systemic networks for English and German, with separate choice points for intensification, conditionality and formality.
Ideas
- I1. Encode the English concessive markers as a closed Rust enum with three classes (a conjunctive adjunct, a subordinator, a preposition), not as strings. Use the class to pick the order of the two clauses and the syntax, then check the marker choice with the rule that the register and conditionality features give.
- I2. A rule for the minor and main act: "on the one hand ... on the other hand" makes the last clause the main act. Implement this in
prag/src/roles.rsor the discourse module, and show which clause was chosen as main and why. - I3. A concession detector over English text: a cue set (although, though, even though, despite, nevertheless, however, still) with the three classes as labels. Measure the precision of each class on a hand-coded English test set, and use a negative control of non-concessive uses of though ("as though", "though" as a final adverb).
- I4. Denial-of-implicature check (Concede-II): a sentence with "That doesn't mean ..., though ..." should be read as a repair of an inference, which the frame reader can mark as a denial with the inferred clause.
For Mova
The chapters are 24. Coordination and subordination (subordinate although, despite as a preposition), 36. Discourse relations and coherence (contrast, concession, the cue-to-relation mapping) and 38. Dialogue, discourse markers and reported speech (nevertheless, though as discourse markers). The code is en/src/dict.rs (marker lexicon), prag/src/infer.rs (the failed-expectation inference) and the discourse relations in world. The paper is about generation, so the measures are for analysis: class precision and recall for the concession detector on English, and the share of cases where the order rule and the gold clause order agree.
Cite
@misc{grote1995ma,
title = {{Ma(r)king concessions in English and German}},
author = {Grote, Brigitte and Lenke, Nils and Stede, Manfred},
year = {1995},
eprint = {cmp-lg/9506001},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/cmp-lg/9506001}
}References
- J. Sinclair (ed.). Collins Cobuild English Language Dictionary. Collins, London, 1987. find
- DUDEN: Deutsches Universalworterbuch. Dudenverlag, Mannheim 1989. find
- M. Elhadad, K.R. McKeown. Generating Connectives. In: Proceedings of COLING-90 , Helsinki, 1990, pp 97-101. find
- R. Glas. Ein Textkorpus fur deutsche Gegenwartssprache. In: Linguistische Berichte 40, 1975, pp 63-66. find
- B. Grote. Generierung in der systemisch-funktionalen Grammatik: Die Behandlung von Prapositionen. Magisterarbeit, Universitat Trier, 1993. (To appear as Technical Report, FAW Ulm) find
- M.A.K. Halliday. Introduction to Functional Grammar. Edward Arnold, London, 1985. find
- M.A.K. Halliday, R. Hasan. Cohesion in English. Longman, Burnt Mill, 1976. find
- G. Helbig, J. Buscha. Deutsche Grammatik. Langenscheidt / Verlag Enzyklopadie, Berlin, 1991. find
- E. Hovy. Generating Natural Language under Pragmatic Constraints. Lawrence Erlbaum, Hillsdale/NJ, 1988. find
- R.T. Kasper. A exible interface for linking applications to PENMAN's sentence generator. In: Proceedings of the DARPA Workshop on Speech and Natural Language , 1989. find
- A. Knott, R. Dale. Using Linguistic Phenomena to Motivate a Set of Coherence Relations. In: Discourse Processes 18:1, 35-62, 1994. find
- E. Lang. Syntax und Semantik der Adversativkonnektive. In: B. Kunzmann-Muller, E. Lang (eds.): Studien zu Funktionswortern . Berlin, 1989. find
- N. Lenke, B. Grote, M. Stede. Concessions and how to make them. Paper presented at the Fourth International Colloquium on Cognitive Science, Donostia-San Sebastian, Spain, 1995. find
- W. Mann, S. Thompson. Rhetorical Structure Theory: A Theory of Text Organization. In: L. Polanyi (ed): The Structure of Discourse . Ablex, Norwood, 1987. find
- J. Martin. English Text { System and Structure. John Benjamins, Philadelphia/Amsterdam, 1972. find
- The Penman Documentation . Unpublished Documentation of the Penman Sentence Generation System. USC/ISI, Marina del Rey/CA, 1989. find
- R. Quirk, S. Greenbaum, G. Leech, J. Svartvik. A Grammar of Contemporary English. Longman, Burnt Mill, 1972. find
- Dietmar Rosner and Manfred Stede. Customizing RST for the automatic production of technical manuals. In R. Dale, E. Hovy, D. Rosner, and O. Stock, editors, Aspects of Automated Natural Language Generation. Springer, Berlin/Heidelberg, 1992. find
- D. Scott and C. de Souza. Getting the message across in RST-based text generation. In: R. Dale, C. Mellish, M. Zock (eds.): Current Research in Natural Language Generation. Academic Press, London, 1990. find
- M. Stede. A contrastive analysis of some contrastive discourse markers. In: J.J. Quantz, B. Schmitz (eds.): Ambiguity and Strategies of Disambiguation (Proceedings of a Workshop held at the Annual Conference of the Deutsche Gesellschaft fur Sprachwissenschaft). Technical Report KIT-120, FB Informatik, TU Berlin, 1994. find
This page in Markdown → https://md.mova.elamur.ai/free/en/docs/papers/grote-1995-ma-r-king-concessions-english/