Linguists find proof of sweeping language pattern once deemed a 'hoax'

<a href="https:&#x2F;&#x2F;archive.md&#x2F;YYH8X" rel="nofollow">https:&#x2F;&#x2F;archive.md&#x2F;YYH8X</a>

&quot;for the frozen white stuff we lump under a single term.&quot;<p>From my perspective this is the hoax. I come from the alps and we have dozens of terms for snow. Only those people without snow might have one word, because they have no need to describe different versions of snow. I remember Sulz, Firn, Neu, Kunst, Matsch, Harsch, Papp, Pulver, ... (left 35 years ago).

The hoax is the interpretation that the language you speak has a significant impact on how you can think. This article seems to argue it&#x27;s the inverse relationship, which is not nearly as controversial.<p>English-speaking skiers have more words for &quot;snow&quot; than Inuktitut-speakers. It&#x27;s the culture that shapes the language, not the language that shapes the mind.

&quot;The researchers analyzed bilingual dictionaries between English and more than 600 languages, looking for what they call “lexical elaboration,” in which a language has many words related to a core concept. It’s the same phenomenon that fueled the Inuit debate. But this study brings a twist: rather than the number of words, it measured their proportion, the slice of dictionary real estate taken up by a concept.&quot;<p>This seems inadequate to make the kinds of claims the researchers are quoted as asserting in the article.

I happened to write my bachelors thesis on the effect of mother-tongues on cognitive processes in 2012 and found the literature very vague on this issue.<p>At the time the literature suggested that the cognitive processes are the same across populations of different mother-tongues but that language can influence the data those processes work upon, EG: exposed to the same events, what details get picked up, built into narratives and remembered.<p>I would move that language constitutes a very strong mnemonic anchor if nothing else.

So, the Innuit may not have 100 words for &quot;snow&quot; after all. But the Hacker&#x27;s Dictionary really does contain 216 synonyms for &quot;broken&quot;.<p>[*] <a href="https:&#x2F;&#x2F;hackersdictionary.com&#x2F;html&#x2F;index.html" rel="nofollow">https:&#x2F;&#x2F;hackersdictionary.com&#x2F;html&#x2F;index.html</a>

Hmm... Americans have <i>many</i> different words for &quot;penis.&quot;<p>I&#x27;ll just get my coat...

The British have umpteen ways to describe rain storms&#x2F;showers. Drizzle, deluge, mizzle, pouring, bucketing, lashing, and quite a few words for wind. Speaking of which the wind, which is the major feature of the weather in Arab countries... has given rise to umpteen different words for explaining what kind of wind they are experiencing, some of which don&#x27;t really have an English translation. The &quot;proof&quot; was always right there in front of them.

Flake, avalanche, snow, zastrugi, powder, firn, dump, pillow, iceberg, chop, snowball, flurry, yukimarimo, piste, ice, snirt, corn, blizzard, cornice, drift, freshie, smud, penitentes, frost, hardpack, slurry, berm, chowder, hoar, icicle, neve, slush, styrofoam, glacier, sleet, graupel, crust, crud, dendrites<p>I heard the Eskimos have over 50 words for a bad example<p>^ my favorite t-shirt.<p>So many of these studies also abuse compound words and misunderstand agglutination to produce their shocking counts.

The data exploration tool linked-to from the article:<p><a href="https:&#x2F;&#x2F;charleskemp.com&#x2F;code&#x2F;lexicalelaboration.html" rel="nofollow">https:&#x2F;&#x2F;charleskemp.com&#x2F;code&#x2F;lexicalelaboration.html</a>

in a polysynthetic language like inuit, &quot;words&quot; aren&#x27;t really a useful category to measure. i&#x27;d hedge my bets the amount of words for snow approaches infinity. do they perhaps mean &quot;roots&quot;?

One way to improve the rigor of this research would be to embed each word in each language in a common latent space then look for clusters.<p>But creating that latent space and the corresponding embedding algorithm is hard in the first place. Today’s embedding models could be terrible for the fringe languages this research is about, and we wouldn’t know because we don’t know how to evaluate overall semantic accuracy.<p>Am I off piste here?

Words are more than just symbols; they represent concepts and patterns we observe in the world, in our society, and inside ourselves.<p>Translation is only possible because we are all humans and have experienced broadly similar concepts, but there&#x27;s a limit to it, especially in social milieu and in how we conceptualize ourselves in society.<p>To truly understand another people and culture at a deep level, you need to learn their native tongue and their living environment -- This is what I&#x27;ve internalized as a long-time learner and teacher of languages.

This is somewhat similar to the language vector embedding, isn&#x27;t it?<p>And the article asks the reasonable question &quot;what is the difference between having a single word for a thing versus a commonly understood cluster of words?&quot;. It&#x27;s not a hard boundary.<p>Every translation loses a little bit of information but potentially brings in different connotations. The things that translators and localizers argue about endlessly: do we look for the words that most closely match the other words, or do we look for feeling and meaning that most closely matches the original intent?

This is an over-broad headline on a credulous article about a simplistic study of heavily biased data sources. It certainly isn’t “proof” or “sweeping”. Truly terrible article.

I can’t find the actual words. I’d like to see the four french synonyms for abandonment that they counted.

Seeing the maps was interesting. Pretty sure there are like 2 dozen words for weed…

Article title:<p>&gt; Linguists Find Proof of Sweeping Language Pattern Once Deemed a ‘Hoax’<p>Abstract from the cited paper [0]:<p>&gt; our work suggests that large-scale computational approaches to the topic can produce non-obvious and well-grounded insights about language and culture.<p>I think I&#x27;ll continue to be sceptical of the Sapir–Whorf hypothesis.<p>[0] <a href="https:&#x2F;&#x2F;osf.io&#x2F;preprints&#x2F;psyarxiv&#x2F;qmgn8_v2" rel="nofollow">https:&#x2F;&#x2F;osf.io&#x2F;preprints&#x2F;psyarxiv&#x2F;qmgn8_v2</a>

There are about a dozen types of snow. It’s quite reasonable for people who care about the difference to be able to describe them in language. Anyone who has shoveled snow can tell you there’s a difference between a cold light snow and a heavy wet snow. Anyone who has walked on snow crust can recall the feeling.<p>Ask anyone who skis what his favorite type of snow is. His least favorite: Champaign powder, fat wet flakes, cold fluff, icy crust, I could probably talk for an hour about the different types of snow and the conditions that lead to them. Some types of snow lead to avalanche conditions. Some are dangerous to drive in. Some are a dream to ski, some make you turn around and go home.<p>Maybe we don’t have singular words for it, but we certainly can describe the differences in language. It would be insane to think otherwise.

So someone named Geoff Pullum called this a hoax. Now that claim may be wrong. Did the journalist find some explanation of why Pullum said that? I&#x27;m curious.

Skiers also have many words for snow: powder, slush, corn, corduroy, crust, ice (not ice), blue ice (actually ice), windpack, popcorn (unrelated to &quot;corn&quot;), and of course the California favorite: cement.

I love how languages reveal cultural and societal features. A popular example is that in English the word &quot;free&quot; means both &quot;gratis&quot; and &quot;libre&quot;, which, in my opinion, are very distinct meanings in today&#x27;s world. I like to imagine the kind of worldviews that generated a given language feature.<p>I&#x27;ll share some other revealing or at least interesting examples I liked; I&#x27;ll paste below some cherry-picked excerpts from a conversation I had with an LLM:<p><pre><code> Japanese: Honne vs. Tatemae Words: 本音 (honne, true feelings) vs. 建前 (tatemae, public facade) Cultural significance: Japanese society values harmony and social cohesion. The existence of specific terms for “what you really think” vs. “what you say to maintain face” reflects the high cultural importance of context-sensitive communication and emotional restraint. Korean: Nuanced honorifics System: Verbal endings, titles, and pronouns change based on age, status, and relationship Cultural significance: The extreme granularity of politeness levels in Korean reflects a hierarchical, Confucian-influenced society where social status, age, and respect are central to daily interactions. Russian: Degrees of truth and lies Words: ложь (lozh, a lie), неправда (nepravda, untruth), and правда (pravda, truth) Cultural significance: Russian distinguishes between lies and non-truths—which can imply omission, alternative interpretations, or state-controlled narratives. The prominence of pravda (also the name of a Soviet newspaper) shows how central truth and its manipulation are in Russian cultural-political life. Spanish: Ser vs. Estar (to be) Words: Ser (essential being) vs. Estar (temporary state) Cultural significance: The fact that Spanish makes a grammatical distinction between inherent traits (ser feliz – being a happy person) and current states (estar feliz – feeling happy now) may reflect a worldview that embraces fluidity in personal and social identity. Danish: Hygge Word: Hygge — cozy, intimate, contented atmosphere Cultural significance: This untranslatable term reflects a cultural emphasis on modest comfort, emotional safety, and communal well-being—especially during long, cold winters. It&#x27;s not just a word but a cultural ideal. Finnish’s lack of a future tense Finnish uses present-tense forms to talk about future events, relying on context or adverbs instead of a separate future-tense verb. Some linguists argue this encourages a more present-focused worldview, though opinions vary. Tsimané (Amazonian): No Fixed Future vs. Past Distinction Grammar: Many Amazonian languages (like Tsimané) have clear past vs. “non-past” rather than past vs. future. Cultural significance: Reflects a worldview where the future is not an ontological category—reinforcing an orientation toward present action and community relationships rather than distant plans. French savoir vs. connaître French distinguishes “knowing how” (savoir) from “knowing someone or being familiar with something” (connaître). English’s single “know” hides this nuance, whereas French speakers constantly signal whether they’re referring to factual&#x2F;learned knowledge or personal acquaintance&#x2F;experience. Georgian: Evidentiality Markers Grammar: Verb prefixes or particles indicate how the speaker knows what they’re saying (e.g., witnessed vs. heard vs. inferred). Cultural significance: The need to signal source of knowledge underlines a communal emphasis on accuracy, trustworthiness, and relational nuance. Quechua: Three-way Evidentiality Markers: Distinguish whether information is firsthand (-mi), hearsay (-si), or inferred (-chá). Cultural significance: Highlights a worldview where knowing how one learned something is as important as the information itself—rooted in oral tradition and communal storytelling.</code></pre>