Elamur Research⌕
☰
home / papers / 2022 / Pragmatic reasoning and semantic convention: A case study on gradable adjectives

Pragmatic reasoning and semantic convention: A case study on gradable adjectives

Paper · 2022 · Ming Xiang, Christopher Kennedy, Weijie Xu et al.

Summary

The paper asks how listeners handle unmodified gradable adjectives ("long", "empty", "bent"), whose cutoff point is set by context and often uncertain. It separates two questions: (S) under what conditions such a sentence counts as true, and (P) what it actually tells us about an object's degree. It compares answers from formal semantics, which posits threshold conventions (absolute adjectives tied to scale maxima or minima, relative ones context-driven), with answers from Bayesian RSA-style pragmatics, specifically the LG (Lassiter & Goodman) and QF (Qing & Franke) models. Three experiments collect human data: prior degree expectations (Exp. 1), truth-value judgments (Exp. 2) and inferred posterior degrees after hearing the adjective (Exp. 3). These cover absolute maximum, absolute minimum and relative adjectives. The Bayesian models capture what is communicated better than what is judged true. Their weakness on truth judgments comes mainly from misestimating where thresholds lie. The authors propose adding the semantic threshold conventions to the probabilistic models. This matters for any system that must decide whether a vague predicate holds.

Key points

  • The paper distinguishes truth conditions (S) from communicated information (P) for positive-form gradable adjectives.
  • It uses three adjective classes: absolute maximum (straight, empty, flat), absolute minimum (bent, open, striped) and relative (long, heavy, big).
  • Exp. 1 supplies priors, Exp. 2 truth-value judgments and Exp. 3 posterior degree judgments. LG and QF predictions are compared per adjective.
  • A pragmatic speaker model chooses between saying the adjective and staying silent. Grid search over λ (1–10) and cost (0–10) gave by-item R² values of 0.67 (2 parameters), 0.59 (3 parameters) and 0.68 (4 parameters). A finer grid left these results unchanged.
  • Data, code and materials are released on OSF (osf.io/nr6a4).

Interesting ideas

  • I1. Measure "is it true?" and "what does it tell me?" as separate targets, because a model can do well on one and fail the other.
  • I2. Give vague adjectives a conventional anchor where one exists: scale endpoints for absolute adjectives, and a context-relative threshold for relative ones.
  • I3. Model the speaker's alternative to an assertion as silence, which turns the listener's baseline into simply their prior over degrees.
  • I4. Derive the listener's degree belief by updating a prior over a comparison class, so the same word yields different expectations for different nouns (a long pole versus a long knife).
  • I5. Check that model fits survive a finer parameter grid as a robustness test.

For Mova

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

  • I1 (answer): Evaluate vague-predicate questions in two separate ways: yes/no truth judgments and estimates of an entity's degree.
  • I2 (world): Tag adjectives in the compiled measure/value knowledge as max-, min- or relative-threshold, so that states like "empty" or "bent" get explainable truth conditions.
  • I3 (frames): Compute implicatures of scalar adjectives in prag by comparing against a "said nothing" baseline drawn from world priors, keeping the reasoning traceable.
  • I4 (world): Store per-noun degree priors, such as typical lengths, and update an entity's state when a gradable adjective is asserted.
  • I5 (learn): Re-run threshold and parameter sweeps at finer resolution before accepting tuned settings, alongside the existing negative controls.

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

🎓University of California, IrvineIrvine · United States29 works · 36,049 citations🎓University of ChicagoChicago · United States28 works · 49,773 citations

📚 Shelves 6The field of science this work belongs to (OpenAlex fields).Computer Science · 9,086 worksArtificial Intelligence · 8,889 worksArts and Humanities · 9,296 worksLanguage and Linguistics · 9,206 worksPsychology · 798 worksExperimental and Cognitive Psychology · 463 works

🧭 Topics 1Research subjects and languages the work is about.Reasoning · 510 works

Cite

Cite as Ming Xiang, Christopher Kennedy, Weijie Xu and Timothy Leffel. 2022. Pragmatic reasoning and semantic convention: A case study on gradable adjectives. Semantics and Pragmatics.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.3765/sp.15.9

@article{xiang2022pragmatic,
  title = {{Pragmatic reasoning and semantic convention: A case study on gradable adjectives}},
  author = {Ming Xiang, Christopher Kennedy, Weijie Xu and Timothy Leffel},
  journal = {Semantics and Pragmatics},
  year = {2022},
  doi = {10.3765/sp.15.9},
  url = {https://doi.org/10.3765/sp.15.9}
}

References

38 references, 3 of them with a card of ours. ↗ leads to the work.

  1. Bumford, Dylan & Jessica Rett. 2021. Rationalizing evaluativity. Sinn und Bedeutung (SuB) 25. 187–204. https://doi.org/10.18148/sub/2021.v25i0.931. ↗
  2. Bürkner, Paul-Christian. 2017. brms: An R package for Bayesian multilevel models using Stan. Journal of Statistical Software, Articles 80(1). 1–28. https://doi.org/10.18637/jss.v080.i01. ↗
  3. Burnett, Heather. 2016. Gradability in natural language: Logical and grammatical foundations (Oxford Studies in Semantics and Pragmatics). Oxford: Oxford University Press. https://doi.org/10.1093/acprof:oso/9780 198724797.001.0001. ↗
  4. Chen, Sherry Yong & Bob van Tiel. 2021. Every ambiguity isn’t syntactic in nature: Testing the rational speech act model of scope ambiguity. Proceedings of the Society for Computation in Linguistics (SCiL) 4(1). 254–263. https://doi.org/10.7275/h3rp-m711. ↗
  5. Chierchia, Gennaro, Danny Fox & Benjamin Spector. 2012. Scalar implicature as a grammatical phenomenon. In Claudia Maienborn, Klaus von Heusinger & Paul Portner (eds.), Semantics: An international handbook of natural language meaning, vol. 3. Berlin: de Gruyter. https://doi.org/1 0.1515/9783110253382.2297. ↗
  6. Doran, Ryan, Gregory Ward, Meredith Larson, Yaron McNabb & Rachel E. Baker. 2012. A novel experimental paradigm for distinguishing between what is said and what is implicated. Language 88(1). 124–154. https://do i.org/10.1353/lan.2012.0008. ↗
  7. Ferreira, Victor S. 2019. A mechanistic framework for explaining audience design in language production. Annual Review of Psychology 70. 29–51. https://doi.org/10.1146/annurev-psych-122216-011653. ↗
  8. Foppolo, Francesca & Francesca Panzeri. 2011. When straight means relatively straight and big means absolutely big. Paper presented at the 31st Incontro di Grammatica Generativa, Rome, Italy, February 24-26. https://doi.org/10281/24104. ↗
  9. Fox, Danny. 2007. Free choice and the theory of scalar implicatures. In Uli Sauerland & Penka Stateva (eds.), Presupposition and implicature in compositional semantics, 71–120. Hampshire: Palgrave Macmillan. https://doi.org/10.1057/9780230210752_4. ↗
  10. Frank, Michael C. & Noah D. Goodman. 2012. Predicting pragmatic reasoning in language games. Science 336(6084). 998–998. https://doi.org/10.1126/science.1218633. Pragmatic reasoning and semantic convention Franke, Michael, Fabian Dablander, Anthea Schöller, Erin Bennett, Judith Degen, Michael Henry Tessler, Justine T. Kao & Noah D. Goodman. 2016. What does the crowd believe? A hierarchical approach to estimating subjective beliefs from empirical data. In CogSci 2016: The annual meeting of the Cognitive Science Society, vol. 38, 2669–2674. https://cognitivescience society.org/wp-content/uploads/2019/03/cogsci2016_proceedings.pdf. ↗
  11. Goodman, Noah & Michael Frank. 2016. Pragmatic language interpretation as probabilistic inference. Trends in Cognitive Science 20(11). 818–829. https://doi.org/10.1016/j.tics.2016.08.005. ↗
  12. Grice, H. Paul. 1975. Logic and conversation. In Peter Cole & Jerry L. Morgan (eds.), Speech acts (Syntax and Semantics 3), 41–58. Leiden: Brill. https://d oi.org/10.1163/9789004368811_003. ↗
  13. Jasbi, Masoud, Brandon Waldon & Judith Degen. 2019. Linking hypothesis and number of response options modulate inferred scalar implicature rate. Frontiers in Psychology 10. https://doi.org/10.3389/fpsyg.2019.001 89. ↗
  14. Kao, Justine T., Jean Y. Wu, Leon Bergen & Noah D Goodman. 2014. Nonliteral understanding of number words. Proceedings of the National Academy of Sciences (PNAS) 111(33). 12002–12007. https://doi.org/10.1073/pnas.1407 479111. ↗
  15. Kennedy, Christopher. 1999. Projecting the adjective: The syntax and semantics of gradability and comparison. New York: Routledge. https://doi.org/10.4324/9780203055458. ↗
  16. Kennedy, Christopher. 2007. Vagueness and grammar: The semantics of relative and absolute gradable adjectives. Linguistics and Philosophy 30(1). 1–45. https://doi.org/10.1007/s10988-006-9008-0. ↗
  17. Kennedy, Christopher & Louise McNally. 2005. Scale structure, degree modification, and the semantics of gradable predicates. Language 81(2). 345– 381. https://doi.org/10.1353/lan.2005.0071. ↗
  18. Klein, Ewan. 1991. Comparatives. In Arnim von Stechow & Dieter Wunderlich (eds.), Semantik: Ein internationales Handuch der zeitgenössischen Forschung, chap. 32, 673–691. Berlin: de Gruyter. https://doi.org/10 . 1 515/9783110126969.8.673. ↗
  19. Lassiter, Daniel & Noah D. Goodman. 2013. Context, scale structure, and statistics in the interpretation of positive-form adjectives. Semantics and Linguistic Theory (SALT) 23. 587–610. https://doi.org/10.3765/salt.v23i0 .2658. ↗
  20. Xiang, Kennedy, Xu, Leffel Lassiter, Daniel & Noah D. Goodman. 2017. Adjectival vagueness in a Bayesian model of interpretation. Synthese 194(10). 3801–3836. https://doi.org/10 .1007/s11229-015-0786-1. ↗
  21. Lewis, Karen S. 2020a. Metasemantics without semantic intentions. Inquiry 65(8). 1–29. https://doi.org/10.1080/0020174X.2020.1847184. ↗
  22. Lewis, Karen S. 2020b. The speaker authority problem for context-sensitivity (Or: you can’t always mean what you want). Erkenntnis 85(6). 1527–1555. https://doi.org/10.1007/s10670-018-0089-2. ↗
  23. Noveck, Ira A. 2001. When children are more logical than adults: Experimental investigations of scalar implicature. Cognition 78(2). 165–188. https://do i.org/10.1016/S0010-0277(00)00114-1. ↗
  24. Papafragou, Anna & Julien Musolino. 2003. Scalar implicatures: Experiments at the semantics–pragmatics interface. Cognition 86(3). 253–282. https : //doi.org/10.1016/S0010-0277(02)00179-8. ↗
  25. Pickering, Martin J. & Simon Garrod. 2013. An integrated theory of language production and comprehension. Behavioral and Brain Sciences 36(4). 329– 347. https://doi.org/10.1017/S0140525X12001495. ↗
  26. Pinkal, Manfred. 1995. Logic and lexicon: The semantics of the indefinite. Dordrecht: Kluwer. https://doi.org/10.1007/978-94-015-8445-6. ↗
  27. Qing, Ciyang. 2020. Semantic underspecification and its contextual resolution in the domain of degrees. Stanford, CA: Stanford University dissertation. https://searchworks.stanford.edu/view/13751296. ↗
  28. Qing, Ciyang & Michael Franke. 2014. Gradable adjectives, vagueness, and optimal language use: A speaker-oriented model. Semantics and Linguistic Theory (SALT) 24. 23–41. https://doi.org/10.3765/salt.v24i0.2412. ↗
  29. Ronai, Eszter & Ming Xiang. 2020. Pragmatic inferences are QUD-sensitive: An experimental study. Journal of Linguistics 57(4). 841–870. https://doi.org/10.1017/S0022226720000389. ↗
  30. Rotstein, Carmen & Yoad Winter. 2004. Total adjectives vs. partial adjectives: Scale structure and higher-order modifiers. Natural Language Semantics 12(3). 259–288. https://doi.org/10.1023/B:NALS.0000034517.56898.9a. ↗
  31. Schöller, Anthea & Michael Franke. 2017. Semantic values as latent parameters: Testing a fixed threshold hypothesis for cardinal readings of few & many. Linguistics Vanguard 3(1). https://doi.org/10.1515/lingvan-2016-0 072. ↗
  32. Solt, Stephanie. 2009. Notes on the comparison class. In Rick Nouwen, Robert Rooij, Uli Sauerland & Hans-Christian Schmitz (eds.), Vagueness in communication: International Workshop, VIC 2009, held as part of ESSLLI 2009, Pragmatic reasoning and semantic convention Bordeaux, France, July 20-24, 2009. Revised selected papers, 189–206. Heidelberg: Springer. https://doi.org/10.1007/978-3-642-18446-8_11. ↗
  33. Solt, Stephanie. 2012. Comparison to arbitrary standards. Sinn und Bedeutung (SuB) 16(2). 557–570. https://ojs.ub.uni-konstanz.de/sub/index.php/sub/article/view/447. ↗
  34. Syrett, Kristen. 2007. Learning about the structure of scales: Adverbial modification and the acquisition of the semantics of gradable adjectives. Evanston, IL: Northwestern University dissertation. https://doi.org/10.21985/N2CB1R. ↗
  35. Tessler, Michael Henry & Noah D. Goodman. 2019. The language of generalization. Psychological Review 126(3). 395. https://doi.org/10.1037/rev00 00142. van Tiel, Bob, Michael Franke & Uli Sauerland. 2021. Probabilistic pragmatics explains gradience and focality in natural language quantification. Proceedings of the National Academy of Sciences (PNAS) 118(9). https://doi.o rg/10.1073/pnas.2005453118. van Tiel, Bob, Emiel van Miltenburg, Natalia Zevakhina & Bart Geurts. 2016. Scalar diversity. Journal of Semantics 33(1). 137–175. https://doi.org/10.1 093/jos/ffu017. ↗
  36. Toledo, Assaf & Galit Weidman Sassoon. 2011. Absolute vs. relative adjectives: Variance within vs. between individuals. Semantics and Linguistic Theory (SALT) 21. 135–154. https://doi.org/10.3765/salt.v21i0.2587. ↗
  37. Unger, Peter. 1975. Ignorance. Oxford, UK: Clarendon Press. https://doi.org/10.1093/0198244177.001.0001. ↗
  38. Waldon, Brandon & Judith Degen. 2020. Modeling behavior in truth value judgment task experiments. Proceedings of the Society for Computation in Linguistics (SCiL) 3. 238–247. https://aclanthology.org/2020.scil-1.29. Xiang, Kennedy, Xu, Leffel Ming Xiang Department of Linguistics University of Chicago 1115 E. 58th Street Chicago, IL 60637 mxiang@uchicago.edu Christopher Kennedy Department of Linguistics University of Chicago 1115 E. 58th Street Chicago, IL 60637 ck@uchicago.edu Weijie Xu Department of Language Science University of California, Irvine 3151 Social Science Plaza A Irvine, CA 92697-5100 weijie.xu@uci.edu Timothy Leffel Seattle, WA tjleffel@gmail.com ↗

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/xiang-2022-pragmatic-reasoning-semantic-convention-case/