--- title: "Nonliteral understanding of number words — Kao 2014" description: "Not read: this card is written from what is reliably known about the paper, and every idea is marked to check against the text. Cited by 24 papers here." authors: ["Justine T. Kao", "Jean Y. Wu", "Leon Bergen", "Noah D. Goodman"] year: 2014 venue: "PNAS 111(33):12002–12007" source: "https://doi.org/10.1073/pnas.1407479111" doi: "10.1073/pnas.1407479111" licence: "no open legal copy — no copy here" summary_licence: "Elamur Cosmic License" canonical: "https://research.elamur.ai/free/en/papers/kao-2014-nonliteral-understanding-number-words/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/kao-2014-nonliteral-understanding-number-words/) [home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2014](https://md.research.elamur.ai/free/en/papers/2014/) / Nonliteral understanding of number words # Nonliteral understanding of number words **Authors:** Justine T. Kao, Jean Y. Wu, Leon Bergen and [Noah D. Goodman](https://md.research.elamur.ai/free/en/authors/noah-d-goodman/) · **Year:** 2014 · **Venue:** PNAS 111(33), 12002–12007 **Link:** [doi.org/10.1073/pnas.1407479111](https://doi.org/10.1073/pnas.1407479111) · **Licence:** copyrighted (PNAS), no open legal copy fetched — card only, summary only · **Status:** not read [**189** citations](https://openalex.org/works?filter=cites:W2029138396) OpenAlex, 2026-10-14 · [**24** citations in our library](https://md.research.elamur.ai/free/en/papers/kao-2014-nonliteral-understanding-number-words/cited-by/) of the 140,884 papers whose bibliographies we read · [PNAS 111(33):12002–12007](https://openalex.org/S125754415) - [Publisher's page](https://doi.org/10.1073/pnas.1407479111) - [PDF](https://www.pnas.org/content/pnas/111/33/12002.full.pdf) - [DOI 10.1073/pnas.1407479111](https://doi.org/10.1073/pnas.1407479111) - [OpenAlex](https://openalex.org/W2029138396) **Authors (4)** 1. [Justine T. Kao](https://openalex.org/A5108531622) [Semantic Scholar](https://www.semanticscholar.org/author/1900374) · h 12 · 658 citations · 27 works 2. [Jean Y. Wu](https://openalex.org/A5047624579) [Semantic Scholar](https://www.semanticscholar.org/author/2110402830) · h 5 · 6,902 citations · 8 works 3. [Leon Bergen](https://openalex.org/A5078977334) [ORCID](https://orcid.org/0009-0001-1445-5247) · [Semantic Scholar](https://www.semanticscholar.org/author/2819425) · h 18 · 2,644 citations · 96 works 4. [Noah D. Goodman](https://openalex.org/A5001961716) [ORCID](https://orcid.org/0000-0002-9176-8802) · [Semantic Scholar](https://www.semanticscholar.org/author/144002017) · h 66 · 20,375 citations · 379 works Citations and author metrics: OpenAlex, 2026-10-14; authors with the same name are not always told apart. ## What it is about Not read: this card is written from what is reliably known about the paper, and every idea is marked to check against the text. The paper models hyperbole and the precision of number words with a Rational Speech Act (RSA) model. A speaker who says that an item cost some amount may be addressing the actual price, or how they feel about it (for example, that it was expensive), or both; the listener infers which question is under discussion together with the price and the affect. The model uses price priors for everyday items (kettles, watches, laptops) elicited from people, and a cost that makes round numbers cheaper to say than sharp ones. It predicts that an implausibly high price is read as hyperbole that conveys affect, and that round numbers get loose readings while sharp numbers get exact ones (the "pragmatic halo"). Human judgements in several experiments fit the model well. ## For Mova - I1. Hyperbole candidates from priors: when a stated quantity lies far outside the `onto` default range for the thing measured, `prag` marks a hyperbole reading that conveys an attitude, and the explanation names the prior that was exceeded. Negative control: plausible quantities must not be marked. (to check: how the paper sets the threshold, if any) - I2. Run the paper's setting in `prag/src/rsa.rs`, which already does exact RSA over small closed sets: worlds as (price, affect) pairs, a goal (QUD) variable, and a cost per utterance. Compare the L1 readings with the paper's qualitative findings. (to check: the paper's exact model equations and parameter values) - I3. Round versus sharp numbers in `math` and `world`: a round number in a narrative gets a tolerance interval, a sharp number an exact value, and word problems are checked for the case where this changes an answer. (to check) - I4. Ask the teacher, not the SLM, for small elicited priors on everyday quantities in tale texts, and store them as `onto` defaults with their source, so the hyperbole rule has data. (to check) The chapters are [languages/en/34-irony-and-figurative-language.md](https://mova.elamur.ai/free/en/docs/languages/en/34-irony-and-figurative-language/) and [languages/en/33-implicature-and-presupposition.md](https://mova.elamur.ai/free/en/docs/languages/en/33-implicature-and-presupposition/). The code is `prag/src/rsa.rs`, `prag/src/irony.rs`, `onto` defaults and `math`. The paper's experimental data, if released, is for tests only until its licence is checked. ## Institutions [🎓**Stanford University** Stanford · United States 248 works · 214,804 citations](https://md.research.elamur.ai/free/en/universities/stanford-university/) · [🎓**Massachusetts Institute of Technology** Cambridge · United States 125 works · 60,031 citations](https://md.research.elamur.ai/free/en/universities/massachusetts-institute-of-technology/) ## Cite Cite as Justine T. Kao, Jean Y. Wu, Leon Bergen and Noah D. Goodman. 2014. Nonliteral understanding of number words. PNAS 111(33):12002–12007. The paper itself, not this card. BibTeX from the library's record of the work. [DOI 10.1073/pnas.1407479111](https://doi.org/10.1073/pnas.1407479111) ```bibtex @inproceedings{kao2014nonliteral, title = {{Nonliteral understanding of number words}}, author = {Justine T. Kao, Jean Y. Wu, Leon Bergen and Noah D. Goodman}, booktitle = {PNAS 111(33):12002–12007}, year = {2014}, doi = {10.1073/pnas.1407479111}, url = {https://doi.org/10.1073/pnas.1407479111} } ``` - [← previous N-gram Counts and Language Models from the Common Crawl](https://md.research.elamur.ai/free/en/papers/buck-2014-n-gram-counts-language-models/) - [next → Recursive Deep Models for Discourse Parsing](https://md.research.elamur.ai/free/en/papers/li-2014-recursive-deep-models-discourse-parsing/) - [↑ back to papers of 2014](https://md.research.elamur.ai/free/en/papers/2014/)