Tuesday, July 03, 2007

Chomsky Falsified?

I wanted to link to it earlier but I have finished reading it only now. This is an article in The New Yorker (caution: it is very long) about an American anthropologist who claims that the existence of a tribal language in the Amazon disproves the central tenets of Chomskyan linguistics - recursive enumerability and the existence of universal grammar and linguistic universals. The langauge in question, Piraha, lacks words for such universal abstractions like numbers or colours and their whole worldview is completely empiricist. They seem to be incapable of think of something removed from their immediate experience. In mathematical terms -- they are incapable of recursive thought. This appears to give more weight to the pre-Chomsky Sapir-Whorf hypothesis which says that language and grammar are determined by culture and are not immutable or apriori, like an organ in the brain.

I have a very shallow and elementary understanding of these concepts (even though, rather shamefully, I had to study some of these in college) but to me it just sounds like a case of one of those intermediary forms in the Darwinian evolution. A recursive language can't just appear out of nowhere. There should be some link between the animal communication and the language of Shakespeare. That's why much of the subject of linguistics, specially when it is used in the subject of philosophy of mind and human nature, feels unintuitive. There is too much mathematics and too little biology. Anyway, this is really fascinating subject. Worth exploring further.

Also see a slideshow in The New Yorker and an interview of Everett and a few other links on The Edge.


orenda said...

Hmm...it seems to me that they do, despite the article's claims, have a term which is used outside of "the present" which the author says this tribe lives in. It would seem -- at least to me -- that the type of ostensive description involved in the act of saying "this looks like blood" one moment and following it with "this looks like red berries" the next (to describe the same object) has one very important thing that hasn't changed: the act of pointing and the notion of saying "this." And this (as Hegel might say) might well be the most universal term of all, encompassing in its uses all thises.

I also think Sellars had something along these lines with his "Myth of Our Rylean Ancestors":

"Imagine a stage in pre-history in which humans are limited to what I shall call a Rylean language, a language of which the fundamental descriptive vocabulary speaks of public properties of public objects located in Space and enduring through Time."


I bring these two points up -- the existence of a universal term in the tribe's vocabulary -- and Sellars's essay, because I don't believe, as the author of the article holds, that the linguistic differences here are so earthshattering. Clearly, the language has both universals and intentionality, which makes it (at most) a second cousin of any language of ours, if not directly related in a next-door sort of way. I'm not sure that the claim that the tribe's language embodies a "pure immediacy of experience" means quite what they hope it to mean. They are talking about thoughts, as Sellars says, in the same way that we do.

Anonymous said...

Have look at
http://ling.auf.net/lingBuzz/000411 for simple contradict of facts!

Alok said...

Orenda: That's interesting what you say about "this". He is saying that it is just an assertion about what they see and what they lack is the ability to embed and generate sentences through recursion... It is reflected in the absence of creation myths or their inability to understand or create fiction.

When I asked Everett if the Pirahã could say, in their language, “I saw the dog that was down by the river get bitten by a snake,” he said, “No. They would have to say, ‘I saw the dog. The dog was at the beach. A snake bit the dog.’ ” Everett explained that because the Pirahã accept as real only that which they observe, their speech consists only of direct assertions (“The dog was at the beach”), and he maintains that embedded clauses (“that was down by the river”) are not assertions but supporting, quantifying, or qualifying information—in other words, abstractions.

That Sellars article looks interesting. My understanding of these things alas is still pretty basic. But hopefully soon...

anonymous: thanks the pdf article looks fantastic. read only about 15-20 pages so far, already a comprehensive rebuttal of most of the claims.

tom said...

The most interesting part to me is the idea that words determine thought. I first encountered this in Lev Vygotsky's "Thought and Language". It seems likely that as newborns, we "think" about the world very differently, and as we learn "how to think", our perceptions of the world are progressively whittled down to the useful and the necessary - we learn how to communicate, how to survive. In the end, words become dominant in our minds and filter our experiences to a great extent. We use "art" to try and express our impressions without words, but words get in the way even there.

Also, it seems that any normal person, raised in any culture from birth, learns its language. Chomsky's idea that attributing anything to "natural selection" is the same as saying nothing, also seems to apply to his own universal principle. A lot of academia is mere territorial tussling.

Richard said...

I haven't read the New Yorker article but most of the press reports I read seemed to make the case rather poorly (too many unhelpful attempts to bring Sapir into it). In response to Orenda's point, I think it might be best to look at one of Everett's own articles:


The key passage is this one:

"One answer that's been given when I claim that Piraha lacks recursion, is that recursion is a tool that's made available by the brain, but it doesn't have to be used. But then that's very difficult to reconcile with the idea that it's an essential property of human language—if it doesn't have to appear in a given language then, in principle, it doesn't have to appear in any language. If it doesn't have to appear in one part of a language, it doesn't have to appear in any part of a language... If you go back to the Pirahã language, and you look at the stories that they tell, you do find recursion. You find that ideas are built inside of other ideas, and one part of the story is subordinate to another part of the story. That's not part of the grammar per se, that's part of the way that they tell their stories.

So the evidence is still being collected, the claims that I have made about Pirahã lacking recursion and the fact that Piraha is an evidence that there probably isn't a need for universal grammar. Contrary to Chomsky's proposal that universal grammar is the best way to think about where language comes from, another possibility is just that humans have different brains that are different globally from those of other species, that they have a greater general intelligence that can be exploited for all sorts of purposes in human thinking and human problem-solving... The ongoing investigation of these claims and alternatives to universal grammar, an architectonic effect of culture on grammar as whole, and the implications of this for the way that we've thought about language for the last 50 years are serious. If I am correct then the research so ably summarized in Steve Pinker's book The Language Instinct might not be the best way to think about things."

I think that's a rather more nuanced claim and one that probably needs to be taken quite seriously.