CoChat

Languages Support Efficient Communication about the Environment: Words for Snow Revisited

Open access

The DOI matches its Crossref record; 47 of the 48 references it cites were found in a registry or an index.

Abstract

The claim that Eskimo languages have words for different types of snow is well-known among the public, but has been greatly exaggerated through popularization and is therefore viewed with skepticism by many scholars of language. Despite the prominence of this claim, to our knowledge the line of reasoning behind it has not been tested broadly across languages. Here, we note that this reasoning is a special case of the more general view that language is shaped by the need for efficient communication, and we empirically test a variant of it against multiple sources of data, including library reference works, Twitter, and large digital collections of linguistic and meteorological data. Consistent with the hypothesis of efficient communication, we find that languages that use the same linguistic form for snow and ice tend to be spoken in warmer climates, and that this association appears to be mediated by lower communicative need to talk about snow and ice. Our results confirm that variation in semantic categories across languages may be traceable in part to local communicative needs. They suggest moreover that despite its awkward history, the topic of "words for snow" may play a useful role as an accessible instance of the principle that language supports efficient communication.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1371/journal.pone.0151138

Cites

47 of 47 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig 1 Results of cross-language surveys. Top panels: Library survey. (a) Locations associated with 13 languages that use the same term for ice and snow, and (b) 37 languages that use different terms. (c) Temperatures associated with the locations shown in (a) and (b). Bottom panels: IDS+ data. (d) Locations associated with 21 languages that use the same term for ice and snow, and (e) 145 languages that use different terms. (f) Temperatures associated with the locations shown in (d) and (e). Map data for this and all subsequent figures are from naturalearthdata.com.
Fig 1 Results of cross-language surveys. Top panels: Library survey. (a) Locations associated with 13 languages that use the same term for ice and snow, and (b) 37 languages that use different terms. (c) Temperatures associated with the locations shown in (a) and (b). Bottom panels: IDS+ data. (d) Locations associated with 21 languages that use the same term for ice and snow, and (e) 145 languages that use different terms. (f) Temperatures associated with the locations shown in (d) and (e). Map data for this and all subsequent figures are from naturalearthdata.com.CC BY · PubMed Central
Fig 2 Results of an exhaustive analysis of meaning pairs in the IDS+ dataset. For each pair of meanings, we conducted a mixed effects logistic regression across languages, with mean temperature as the independent variable, and use of a single linguistic form for the two meanings as the dependent variable. For each pair, we compared models with and without temperature as a fixed effect, and recorded the p value associated with this comparison. Two histograms are displayed back-to-back: the gray histogram counts pairs for which the use of a single form is non-positively (zero or negative coefficient) associated with temperature, and the black histogram counts pairs for which the use of a single form is positively (positive coefficient) associated with temperature. The x-axis for each histogram shows minus natural log of the resulting p value, such that higher values correspond to greater significance; this quantity increases to the right for the positive histogram, and to the left for the non-positive histogram. The y-axis shows the number of IDS+ pairs for which that -ln(p) value was obtained.
Fig 2 Results of an exhaustive analysis of meaning pairs in the IDS+ dataset. For each pair of meanings, we conducted a mixed effects logistic regression across languages, with mean temperature as the independent variable, and use of a single linguistic form for the two meanings as the dependent variable. For each pair, we compared models with and without temperature as a fixed effect, and recorded the p value associated with this comparison. Two histograms are displayed back-to-back: the gray histogram counts pairs for which the use of a single form is non-positively (zero or negative coefficient) associated with temperature, and the black histogram counts pairs for which the use of a single form is positively (positive coefficient) associated with temperature. The x-axis for each histogram shows minus natural log of the resulting p value, such that higher values correspond to greater significance; this quantity increases to the right for the positive histogram, and to the left for the non-positive histogram. The y-axis shows the number of IDS+ pairs for which that -ln(p) value was obtained.CC BY · PubMed Central
Fig 3 Results of cross-language Twitter analysis. (a) Locations associated with the 18 languages represented in the Twitter dataset we consider. (b) The natural log probability of mention of ice or snow in a given language as a function of the mean temperature where that language is spoken. Mixed effects logistic regression revealed that temperature is negatively associated with probability of mention of ice or snow, as a fixed effect (β = −0.29, χ2(1) = 21696, p < 10−15), when including random intercepts for each language family.
Fig 3 Results of cross-language Twitter analysis. (a) Locations associated with the 18 languages represented in the Twitter dataset we consider. (b) The natural log probability of mention of ice or snow in a given language as a function of the mean temperature where that language is spoken. Mixed effects logistic regression revealed that temperature is negatively associated with probability of mention of ice or snow, as a fixed effect (β = −0.29, χ2(1) = 21696, p < 10−15), when including random intercepts for each language family.CC BY · PubMed Central
Fig 4 Results of US Twitter analysis. (a) Locations associated with the 25 urban areas represented in the Twitter analysis of American English. (b) Natural log probability of mention of ice or snow in a given urban area as a function of the mean temperature in that area. Logistic regression revealed that temperature is negatively associated with probability of mention of ice or snow (β = −0.060, χ2(1) = 322, p < 10−15).
Fig 4 Results of US Twitter analysis. (a) Locations associated with the 25 urban areas represented in the Twitter analysis of American English. (b) Natural log probability of mention of ice or snow in a given urban area as a function of the mean temperature in that area. Logistic regression revealed that temperature is negatively associated with probability of mention of ice or snow (β = −0.060, χ2(1) = 322, p < 10−15).CC BY · PubMed Central
Fig 5 Analysis of missing forms in the IDS+ data. Locations associated with (a) 20 languages that are missing forms for ice or snow or both and (b) 140 languages that include forms for both meanings. (c) Temperatures associated with the locations shown in (a) and (b).
Fig 5 Analysis of missing forms in the IDS+ data. Locations associated with (a) 20 languages that are missing forms for ice or snow or both and (b) 140 languages that include forms for both meanings. (c) Temperatures associated with the locations shown in (a) and (b).CC BY · PubMed Central
Fig 6 Regression coefficients and -ln(p) values for the exhaustive analysis of meaning pairs in the IDS+ dataset. Labels are included for all non-dominated pairs that are associated with a positive regression coefficient. Two of these pairs include “male,” which is a meaning associated with animals rather than people.
Fig 6 Regression coefficients and -ln(p) values for the exhaustive analysis of meaning pairs in the IDS+ dataset. Labels are included for all non-dominated pairs that are associated with a positive regression coefficient. Two of these pairs include “male,” which is a meaning associated with animals rather than people.CC BY · PubMed Central
Fig 7 Monte Carlo results. The first column shows that the first (lowest) temperature quartile of a lumper sample is almost always higher than the first quartile of a splitter sample. The second column shows that the differences between third quartiles of the samples are less skewed. The third column confirms that differences between the first quartiles are greater than differences between the third quartiles. Results for the library and IDS+ datasets are shown in the first and second rows respectively. The x axis in each plot shows differences in degrees Celsius, and the y axis shows number of samples.
Fig 7 Monte Carlo results. The first column shows that the first (lowest) temperature quartile of a lumper sample is almost always higher than the first quartile of a splitter sample. The second column shows that the differences between third quartiles of the samples are less skewed. The third column confirms that differences between the first quartiles are greater than differences between the third quartiles. Results for the library and IDS+ datasets are shown in the first and second rows respectively. The x axis in each plot shows differences in degrees Celsius, and the y axis shows number of samples.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 47 references

Cited by

See all 137

Metadata

Authors
Regier, T., Carstensen, A., & Kemp, C.
Affiliations
  • University of California, Berkeley (US)
  • Carnegie Mellon University (US)
Corresponding author
  • Regier, Terry
  • Carstensen, Alexandra
  • Kemp, Charles
Venue
PLOS ONE
Published
2016-04-13
Volume
11
Issue
4
Pages
e0151138
ISSN
1932-6203 (ISSN-L)
OpenAlex
W2339609277
Cited by
136
References
48
Citations by year
2026: 17 · 2025: 6 · 2024: 8 · 2023: 10 · 2022: 6 · 2021: 25 · 2020: 11 · 2019: 32 · 2018: 13 · 2017: 4 · 2016: 3
FWCI
24.67
Type
journal article
Language
en
Open access
Locations
  • PLoS ONE (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • KiltHub Repository (submittedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • eScholarship (California Digital Library) (submittedVersion)
  • Figshare (submittedVersion, cc-by)
  • Minerva Access (University of Melbourne) (submittedVersion, cc-by)
  • KiltHub Repository (cc-by)
Funders
  • National Science Foundation (1106400, 1041707)
  • University of California Berkeley (DGE-1106400)
Topics
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Categorization, perception, and language
  • Social Sciences › Cultural Studies › Language and cultural evolution
  • Social Sciences › Linguistics and Language › Language and Culture
Keywords
semantic categories, language and environment, communication efficiency, communication needs, Eskimo languages, language change
MeSH terms
Inuit, Female, Humans, Male, Semantics, Snow

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Languages Support Efficient Communication about the Environment: Words for Snow Revisited. https://app.cochat.ai/papers/10.1371/journal.pone.0151138
The paper
Regier, T., Carstensen, A., & Kemp, C. (2016). Languages Support Efficient Communication about the Environment: Words for Snow Revisited. PLOS ONE. https://doi.org/10.1371/journal.pone.0151138

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.