Good Reason

It's okay to be wrong. It's not okay to stay wrong.

Talk the Talk: Universal Grammar

Next week’s Talk the Talk topic comes to us from the pages of New Scientist.

Many linguists are interested in the similarities between languages. Noam Chomsky once claimed that if a Martian visited Earth and looked at all the human languages, they’d be impressed not by the diversity, but by how similar all human languages are. (Falsify that claim.)

Linguists in the Chomskyan mold have postulated the existence of a Universal Grammar — a set of structural principles that undergird human language. It’s an appealing idea — not least because it could explain how children learn language so quickly, from nada to full sentences in about two or three years. Why so fast? The UG is already in there at birth, and kids will pick up the individual quirks of their native language as they go.

The New Scientist article (PDF) highlights the work of linguists who take a different view. For example, Chomsky felt that recursion was one of the fundamental properties of human language. You can repeat elements of English syntax in certain ways: “My mother’s doctor’s boyfriend’s cat.” No non-human animal communication system has this, and every human language has it.

Except Pirahã. Dan Everett, who’s worked among these Amazonian people for years, says there’s no recursion in Pirahã. You can’t say “My brother’s house”. You have to say “I have a brother. My brother has a house.” And so it goes; the more languages we know about, the more we find that violate these seemingly inviolable constraints.

Is the theory of Universal Grammar falling apart? If language isn’t innate in our human brains, then how do we do it? On the next Talk the Talk.


  1. yes of course. clearly universal grammar is totally refuted if one researcher in one place finds one potential exception.

  2. I've read this a few times, and I can't tell if you're serious. For one thing, you're right. Well, if we're talking about the very strong view.

    On the other hand, I'd like more than one counter-example from one researcher. One counter-example may not be enough to overturn a theory, but it may be enough to force the strong view to retreat.

    It seems to me that UG is looking less and less relevant in the face of these very unusual languages.

Comments are closed.

© 2018 Good Reason

Theme by Anders NorenUp ↑