seanie

The examples given in Pinker's book were meant to make a particular point. The unusual circumstances gave an opportunity to evaluate whether or not grammar is innate to some extent.

In both cases it's claimed that the children didn't learn a grammatical language. Because in both cases they weren't exposed to a grammatical language.

They were actually exposed to a poorly grammatical pidgin language that they instinctively transformed. They had an ingrained ability to re-form and re-order the language they heard, correcting inconsistencies and inventing new tools, to produce a fully fledged grammatical language.

I think the argument is that the brain 'added-value'. I reordered and recategorised the information it was recieving in the pidgin to produce a richer creole.

As to the original question I don't see the problem in being unable to identify the structures in the brain responsible. The task is to evaluate whether there is in fact an innate grammar ability. If that is indeed the case the default assumption is that that ability resides in the brain somewhere.

That's far more likely than it residing in the colon for example.

Vorkosigan

Elwood, part of Chomsky's case rests on certain philosophical issues. For example, children appear to figure out grammar rules on very little information. They know that language comes in words, although if you listen to the syllables, there is little difference in time delay between syllables and between words, in spoken speech. Heck, they know language has grammar and they experiment to figure what the rules are. There's no way you could figure that out, it has to be innate. They produce language that is more sophisticated than that which they have been taught. Pinker's The Language Instinct is really the best introduction, because Pinker puts everything firmly into an evolutionary perspective, whereas Chomsky apparently thinks that evolution could not have produced these features.

Chomsky's language problem is actually a version of the Problem of Induction, the answer to which is that it is built in. Evidence from many different fields indicates that much is built-in, and triggered and enriched by exposure to environmental stimuli. Have you read much in Evolutionary Psychology?

Caveat: See The Symbolic Species for a reply to Pinker and Chomsky. Neural networks apparently can perform this way without a high degree of innateness. I do not agree with the author's conclusions, but the arguments are there.

Vorkosigan

ohwilleke

The certainly evidence for the notion that there is an innate mental ability to have some grammatical organization of language, but is a long shot from the notion that any particular grammatical structure is preferred.

For example, people who have no training in sign language do have a tendency to organize ad hoc sign languages in a subject-object-verb word order (me-there-go). But, there is nothing unnatural about the English language subject-verb-object word order if you've been raised in it (I want to go there). English speakers don't have a natural tendency to rearrange their words, and neither do speakers from languages with different word orders.

Similarly, I don't think that one can say that inflected, or non-inflected forms are natural, or that particular components of sentences need to agree. Some languages do fine without definite and indefinite articles (the, a, an), others use them. Some languages, like French, give nouns gender routinely, others do not.

Some languages, German, for instance, love compound words. Others languages like to break words up into little pieces.

Some languages have polite forms (Japanese, French, for instance), others don't.

Some languages have strong tonality (e.g. Mandarin Chinese), others use tonality only for limited purposes (e.g. the intonations of questions v. statements in English, "You are coming for dinner." v. "You are coming for dinner?").

Almost by definition, language consists of the used of words in an organized way to communicate, but to call that a universal grammar is to overreach.

The indication from real life experience is that grammar is mutable early in life, and hard to change later in life. A ten year old raised in China who learns to speak English will get word order and other English grammar "right", while a thrity year old will probably make mistakes all his life in English grammar no matter how hard he tries.

There is also indications from real life that there is no such thing as a tabla rosa. Children cannot grow into adults without carrying from parents who provide them care, nuture and language models. Without that contact, language does not develop to an advanced level at all.

I'm not sure that the spontaneous reorganization of language by a community experiencing the fusion of two languages says much for the universality of the particular result they reach. One community in a French-English merger, for instance, might end up with French rules for verbs and English rules for nouns, while another might do the opposite. Certainly, there is a need to have some organized way of communicating for clarity, but no need for it to be a particular way.

I have a hard time thinking of any rule that is really universal across all languages. And, to the extent that you find some common threads, basic as they may be, I'm not sure what a big deal that is. You can say, for instance, that all languages are composed of words. But, even there, you have to stretch the definition of word to apply it to sign language and idiom.

Usually people who assume that something is innate or has some de ex machina cause, simply haven't looked hard enough to discover the real mechanism. For example, the convergence of Nicarguan sign language or Philipino Pidgen, may flow as much from social structures (social leaders being emulated, for example), as from any spontaneous tendency working in the same way in multiple people as is implied from an innate universal grammar theory.

[ October 14, 2002: Message edited by: ohwilleke ]</p>

ksagnostic

Exactly.

ksagnostic

One thing that is important to remember when following this topic is that there are two statements which tend to be taken as synonymous in this discussion, but actually are not.

1) Language is a uniquely human ability, and human beings have an innate ability to learn language.

This statement I tend to agree with. However, such a statement does not explain why human beings have this ability. The explanation of the language innatists, such as Chomsky, Pinker, Bellugi, etc., is a specific explanation among several of how humans acquire language.

2) The ability to acquire language is itself uniquely hard wired into human beings, and we humans have a unique "deep structure" template in our central system (aka, a "universal grammar", "language acquisition device", "language module", or "language instinct") that allows most of us to almost effortlessly learn the grammar of the language we grow up hearing, or as may be the case of people who are deaf, seeing.

I was unimpressed with this claim in undergraduate and graduate school, and have become steadily more unimpressed with it as time has gone on. For my part, I don't see any evidence that a universal grammar or "language instinct" is a necessary explanation for human language abilities.

I see no reason to think that the ability of human beings to learn to communicate with their fellows by acquiring the local language requires anything other than the generic ability of an intelligent creature to acquire his species' communication system. Humans do have some unique features, including a unique ability to rapidly produce a complex sound stream and an extremely long developmental period. I think those two features alone, along with the more general ability of intelligent social mammals to attend to and learn their species' communication signals, much moreso than a hypothetical innate language acquisition device, are probably sufficient to explain human language acquisition.

[ October 14, 2002: Message edited by: ksagnostic ]

[ October 15, 2002: Message edited by: ksagnostic ]</p>

Vorkosigan

But KS, it is well known that language exposure must take place during a specific period of life, or else it never takes. How does that square with your analysis?

ksagnostic

Because here is what I said:

"Humans do have some unique features, including a unique ability to rapidly produce a complex sound stream and an extremely long developmental period."

I think that the critical period of development is typical among mammals. It is interesting to note that Kanzi the bonobo learned to comprehend English, as well as to use lexigrams, as an infant during lessons that were obstensibly aimed at his adoptive mother, Matata. Matata failed to learn either. The difference is that we humans, largely because of our three times larger brain, have a critical period that lasts about five to six years (with further brain maturation going on for another fifteen years), instead of two years for other great apes (with further maturation going on for another five or six years). That extended critical period, plus a supralaryngeal speech tract, I believe, is sufficient to account for the development of language. A deep structure is superfluous.

The structures of the brain that have been linked to language have now been discovered in great apes. Presumed differences in proportion in the brain between humans and other great apes (such as relatively larger frontal lobes) have faded away under further analysis. What is left is a simply larger version of the same brain with a correspondingly longer developmental period, and a speech tract.

[ October 15, 2002: Message edited by: ksagnostic ]

[ October 15, 2002: Message edited by: ksagnostic ]

[ October 15, 2002: Message edited by: ksagnostic ]

[ October 15, 2002: Message edited by: ksagnostic ]

[ October 15, 2002: Message edited by: ksagnostic ]

[ October 17, 2002: Message edited by: ksagnostic ]</p>

seanie

Originally posted by ohwilleke:
The certainly evidence for the notion that there is an innate mental ability to have some grammatical organization of language, but is a long shot from the notion that any particular grammatical structure is preferred.

The long shot isn't what's being claimed.

I think Chomsky may have actually noticed that languages come in a wide variety. That's not really news. The argument is that whatever way a particular language is organised there's an innate mental ability that tracks/seeks out that organisation. A specific hardwired grammar ability. That's what's meant to be universal. Not whether or not doors are female.

There is also indications from real life that there is no such thing as a tabla rosa. Children cannot grow into adults without carrying from parents who provide them care, nuture and language models. Without that contact, language does not develop to an advanced level at all.

Well that kinda touches on the controversy. The debate really is about to what extent children are a 'tabula rasa'.

The opposite to the Chomskyan view would be that language is aquired as part of a general learning ability. That straightforward logic dictates the structure of a language and that children learn the particular rules of whatever linguistic culture they're in.

However that's the whole point of the examples Pinker puts forward. He's claiming they show that children aren't a blank slate that just learn what's around them. There's more to them than that.

In both cases the children grew up using a fully grammatical language.

However, it's claimed, that they didn't learn this fully grammatical language because they weren't exposed to a fully grammatical language.

The argument is that if they were blank slates that learnt from their parents, then they would've grown up speaking the poorly grammatical pidgin. But they didn't. They 'instinctively' turned it into a richly grammatical creole.

They achieved this, subconciously, because they had an innate language generating/analysing system that corrected, added to and transformed the language they were hearing into something richer.

Now I'm way out of my depth here. Grammar's not my strong point. I can't even remember where to put apostrophes.

But give the Pinker book a read. It's fascinating.