Yet another challenge to Chomsky on language

Learning a language is not easy, as I can personally testify. So the ease with which children all over the world learn the language of their environment, despite the fact that the languages are very different, has always been a source of great curiosity for researchers. Linguist Noam Chomsky’s theory revolutionized the field. To the extent that I understand it, he argued that there was something known as a Universal Grammar, common but deep rules that all languages shared, and that our brains were hardwired and coded to be receptive to them. As children heard adults around them speaking, various switches were thrown in the brain that converted these general rules into the specific ones for that particular language, enabling children to quickly pick up the syntactic structures of that language.

Over time, the question of what constituted those basic rules of grammar has shifted but the basic idea of some kind of universal grammar has remained. But there have been periodic attempts to upend the Chomsky model as the dominant paradigm and this article is one such attempt, though the author Vyvyan Evans, while claiming to target Chomsky, seems to be aiming his fire more at Stephen Pinker’s interpretation and popularization of Chomsky’s ideas. Another article critiques Evans’s arguments.

While I find the subject fascinating, I simply do not have the expertise to evaluate the competing theories or to judge whether Chomsky’s or any of the alternative view is correct but I pass along the articles because they discusses the issues involved fairly clearly.


  1. says

    With current technology, it is impossible to examine how human brains process language. Thus all we have is speculation and non proven theories. Universal Grammar is a well known theory, but nowhere near universally accepted. This theory might be correct. Or it might be false. We have no way to find out for sure. Personally, I am skeptical of Universal Grammar. And the Poverty of the stimulus argument, which is often used in support of Universal Grammar, seems implausible for me personally.

    In order to accept and perceive as plausible some theory that claims to explain something, I want evidence. For example, personally, I am willing to believe that the evolution theory is correct, because the evidence supporting it is overwhelming. For Universal Grammar the existing evidence is flimsy, hence I remain skeptical.

  2. Venkataraman Amarnath says

    My niece’s second daughter started talking around 16 months, nothing unusual. While all around her -- parents, big sister, grand parents talked Tamil, she developed her own language. I saw her for about two weeks when she was 25 months old. It was amazing the flurry of sounds she made, but we could understand not a single word. I thought she was not just babbling by observing her facial expression and body language consistent with whatever she was trying to convey. She cried a lot and I believe she was exasperated that nobody understood her language. She would have made a good subject for linguistic studies, but with proper local treatment and extra care she learnt proper Tamil and English.

  3. says

    It’s a very silly article.

    Lake correctly notes that children who never experience language, don’t learn language, unlike spiders who spin webs without needing to see it. He also notes that signed languages exist -- seeming to think he disagrees with Chomsky on both these points.

    Lake then denies the existence of linguistic universals, before naming universals like verbs, and changes of word order. He’s right that some languages lack adjectives, missing the point that the function of describing a noun is fulfilled in other ways.

    Lake says “There should be spurts in grammatical complexity every time a new rule kicks in” but in the next paragraph “children appear to pick up their grammar in quite a piecemeal way”, as though the two were in irreconcilable contradiction.

    Finally, he says neuroscience has disproven the notion of a single module in the brain which learns language -- as though anyone had ever proposed the “Language Aquisition Device” as a literal chunk of tissue.

  4. cafebabe says

    I have always had a problem with public discussions of Chomsky’s work. How many articles have we seen over the years by right-wing dumb-asses who want to attack Chomsky’s activism and can’t resist adding something along the lines of “and anyway he’s not such a great thinker because all his theories have been disproved”. This from people who wouldn’t know a phrase structured grammar if you strangled them with it.
    In my second career, as a computer scientist, I lived and breathed the insights that his work provided for the study of formal languages but I am happy that for natural languages we may very well find out that, as the T-shirt says, I think you’ll find that it’s more complicated than that.
    In the meanwhile, I find the explanatory power of Chomsky’s language model compelling, and I am not ready to replace it until I see an alternative with greater explanatory power.

  5. mailliw says

    @4 cafebabe

    I learned about tranformational generative grammar as part of my degree back in the 1970s. The idea of universal grammar struck me as very persuasive.

    I think we still have a lot to learn about how the brain and language actually works, but I agree that the problem with the critics is that they don’t present an alternative coherent hypothesis to Chomsky.

    Natural languages are rich and diverse and full of potential ambiguities, formal languages are tight and compact and have means of detecting and eliminating contradictions. There are simpilarities between natural and formal languages, but there are also huge differences.

Leave a Reply

Your email address will not be published. Required fields are marked *