Deep in the Brazilian Amazon, fewer than 800 people speak Pirahã (pronounced roughly pee-da-HAN). It is a language isolate with no proven relatives, and its reported grammatical properties are debated in modern linguistics.
The fieldworker who triggered the storm is Daniel Everett, a missionary-turned-linguist who lived with the Pirahã for years. According to his reports, the language is missing several things that 20th-century linguists had argued were universal in human language:
- No exact number words. The community uses only hói ("a small amount") and hoí ("a larger amount") — a contrast of quantity, not of count. In experiments, Pirahã speakers can lay out a matching row of objects one-for-one, up to ten, almost without error, but they fail as soon as the set has to be carried across space or time: hidden behind a screen, turned at right angles, or dropped one by one into a can.
- No basic colour vocabulary. Colours are described compositionally — "like blood", "unripe", "dirty" — rather than with fixed labels like "red" or "green".
- No grammatical recursion. Sentences cannot be embedded inside other sentences. This directly challenges Noam Chomsky's claim that recursion is the defining feature that separates human language from animal communication. If Everett is right, Pirahã breaks the theory.
- No creation myths, no fiction, no stories about ancestors no living person knew. A strong cultural norm — Everett calls it the "immediacy of experience" principle — restricts speech to what the speaker or a living interlocutor has personally witnessed.
Stranger still, Pirahã is one of a small handful of languages that can be transmitted across five parallel channels: it can be spoken, whistled, hummed, sung, or shouted. Because Pirahã has only three vowels and eight consonants, its prosody (pitch and rhythm) carries an unusually heavy informational load — enough that mothers can hum entire sentences to babies without articulating a single consonant.
"The Pirahã people communicate almost as much by singing, whistling, and humming as they do using consonants and vowels." — Daniel Everett, Current Anthropology 46(4), 2005, p. 622
Everett's claims are hotly contested. Later researchers, including Andrew Nevins, David Pesetsky and Cilene Rodrigues, have argued that Pirahã does have recursion — just rarely used — and that the number facts can be explained without exotic cognitive consequences. Everett has fired back; the debate is still open.
The cultural norm restricting speech to directly experienced events, and communication by whistling, humming, singing, and shouting, are less contested but still studied. Whatever the outcome of the recursion debate, Pirahã has kept an older question in view: how do language, culture, and thought shape one another?