Eskimos have a hundred words for snow
The old claim turned out to be wrong and was rejected by evidence.
TaughtEskimos have dozens, or hundreds, of words for snow, proving that language shapes how you perceive the world.
NowThe claim collapsed under examination. Inuit and Yupik languages are polysynthetic, meaning they build long words from roots and suffixes, so counting "words" is meaningless: any language can do the same by that method. The root inventory for snow is comparable to English's.
What actually happened
The chain of custody is the fun part. Franz Boas mentioned four snow-related roots in 1911. Benjamin Lee Whorf inflated it to seven in 1940 and implied more. From there it grew by pure repetition, hitting nine, then fifty, then a hundred, then four hundred, with each retelling raising the bid because nobody was checking.
Geoffrey Pullum's 1991 essay "The Great Eskimo Vocabulary Hoax" traced the whole lineage and is why this is dated to 1991. His central point is structural: in a polysynthetic language a single word can encode what English needs a phrase for, so "snow that fell yesterday and is now crusted" is one word. English has snow, sleet, slush, powder, blizzard, flurry, hail, frost, drift, and every skier knows a dozen more. The comparison was never like for like.
The honest nuance: Igor Krupnik and others have argued that some Arctic dialects do have unusually rich, precise snow and ice terminology, which makes sense for people whose safety depends on distinguishing ice conditions. That is a modest, plausible claim about specialized vocabulary, the same way sailors have many words for ropes. It is not the sweeping Whorfian claim that was taught, and strong linguistic determinism has not held up either.
Newsletter
Get new weekly revisions
Sources
Who was taught this
Still standard through 2005, so anyone who finished school between 1950 and 1991 learned the earlier version.