Chomsky's Linguistic Theory

[nosos]

I think Wittgenstein's private language argument is more concerned with the nature of the perceived world, what it can't be like.

I mean, he's not really saying that one cannot talk to oneself (if one has an abstract grammatical system of signifiers etc). The implication is more that the world really is shared.

Precisely - the issue is not whether or not one has an 'inner voice', the issue is whether that 'inner voice' is tied into a web of interlocution. When you're 'speaking' in your 'inner voice' are you speaking (such as would be understood by another were you to verbalise it)? The alternative is to suggest that we 'think' in a 'private language'.

 

[Knotted]

A child born in a lightless cave will certainly have eyes, but the visual areas of the brain will not develop normally. (They will develop to some extent because the fetal retina generates patterns which play a role in the early wiring-up of the visual cortex, among other reasons.) We know this through experiments in which animals are reared in selectively impoverished visual environments.

Universal Grammar is the theory of the initial state of the language faculty. That state consists of a number of parameters, whose combinations of possible settings determine the range of human languages. Acquiring the syntax of a language requires that these parameters are set into a particular pattern through experience of that language. This is an active process on the part of the child.

The child seeks patterns in the language heard around him/her and then tries out rules to account for them, that is, the child experiments with different parameter settings. There is some evidence to support this. Young children can and do produce utterances which don't yet follow the grammar of the adult language, but which are consistent with UG.

Let's consider the position of the child surrounded by adult pidgin speakers and advance a hypothesis of our own. Pidgin languages have highly impoverished grammar. A single isolated child among adult pidgin speakers will therefore not have a good model onto which its own language faculty can map UG parameters. But a community of children among pidgin speakers will hear each others' hypothesis-testing utterances. Because these are more likely to be consistent with UG than the pidgin tongue, the children can work out their own more grammatically-rich language among themselves.

This all sounds reasonable, but I can't reconcile the last paragraph with an innate UG.

I think we need to be careful when we say that a grammar is impoverished. This does not mean that the syntax is impoverished. In fact it seems likely that this means the opposite. There are two types of trivial Turing Machine, one that does not recognise anything and one that recognises everything. So both the suppremely rich syntax (everything is allowed) and the suppremely poor syntax (nothing is allowed) are grammatical suppremely poor (there are no rules).

Does this matter? Well if the child is gaining positive examples of grammatical utterances then according the theory, this is all that is needed for the innate grammar to kick in. There is no need for negative examples - ie. examples of how grammar cannot work as this is already 'known' by the child. There is no need for the child to model anything here - this is already innate. Nor will there be any correction when the child produces the 'wrong' grammar because there is no established grammar. The child will therefore start speaking consistently with respect to the dictates of universal grammar regardless of whether anybody else is.

It really does require just one brain. There is one universal grammar machine each. It is not shared amongst a group using some sort of Vulcan mind meld.

 

[Knotted]

I think Wittgenstein's private language argument is more concerned with the nature of the perceived world, what it can't be like.

I mean, he's not really saying that one cannot talk to oneself (if one has an abstract grammatical system of signifiers etc). The implication is more that the world really is shared.

The argument is notoriosly difficult to interpret - nevermind assess.

What would be allowed though is a trivial, meaningless private language consisting purely of tautologies. If the private language is not meaningful (in Wittgenstein's sense) but somehow regulates thoughts which is necessary because of some deep cognitive function, then this is consistent with the PLA (I think).

 

[Knotted]

Precisely - the issue is not whether or not one has an 'inner voice', the issue is whether that 'inner voice' is tied into a web of interlocution. When you're 'speaking' in your 'inner voice' are you speaking (such as would be understood by another were you to verbalise it)? The alternative is to suggest that we 'think' in a 'private language'.

I agree with that. There are interpretations that insist that language must be social as opposed to social in principle. I don't think that it is clear that this is incorrect from Wittgenstein, rather that the argument would not work if it were the former.

 

[nosos]

I don't see how to draw a distinction between social and social in principle once you've rejected Cartesian subjectivity (as Wittgeinstein does) with an "inner" world of representations able to be principle make social.

 

[Knotted]

I don't see how to draw a distinction between social and social in principle once you've rejected Cartesian subjectivity (as Wittgeinstein does) with an "inner" world of representations able to be principle make social.

If you are Robinson Crusoe, with nobody to talk to, you could still talk as if there was someone there. This would be meaningful. Its only if you talk in such a way that it is impossible for anybody to understand you that is ruled out.

A may be going astray here, but that's what I understood you to mean.

 

[nosos]

Its only if you talk in such a way that it is impossible for anybody to understand you that is ruled out.

Sorry could you expand on this? I think we've crossed wires

 

[Knotted]

Sorry could you expand on this? I think we've crossed wires

From Russell's Philosophy of Logical Atomism lectures:

In a logically perfect language, there will be one word and no more for every simple object, and everything that is not simple will be expressed by a combination of words, by a combination derived, of course, from the words for the simple things that enter in, one word for each simple component. A language of that sort will be completely analytic, and will show at a glance the logical structure of the facts asserted or denied. … A logically perfect language, if it could be constructed, would not only be intolerably prolix, but, as regards its vocabulary, would be very largely private to one speaker. That is to say, all the names that it would use would be private to that speaker and could not enter into the language of another speaker.

Its reasonable to suggest that this is the view that Wittgenstein is polemicing about.

We could discuss it more if you want, but I would rather start a new thread.

 

[nosos]

I'm still slightly confused - no worries was just a small point.

 

[dash_two]

I think we need to be careful when we say that a grammar is impoverished. This does not mean that the syntax is impoverished. In fact it seems likely that this means the opposite.

I have been careless in my use of the words 'grammar' and 'syntax'. Strictly speaking, a grammar is the framework of a language, comprised of morphology (concerned with the internal structure of words) and syntax (concerned with the order of words in an utterance). Don't see why an impoverished grammar should imply an enriched syntax.

I have never heard a pidgin tongue in real life, but according to what I've read pidgins have little or no syntax or morphology. Because of this, they're only really effective in referring to the here-and-now or to already-familiar situations. (Newspaper headlines sometimes approach pidgin, the Edinburgh Evening News once came up with a good example: DRUG GANG WAR FEAR, though this still sort-of follows English syntax.)

Re. Turing machines, I don't really remember how they're supposed to work, so will need to pass them by!

It really does require just one brain. There is one universal grammar machine each. It is not shared amongst a group using some sort of Vulcan mind meld.

The positive examples have to come from somewhere though. iirc Chomsky's theories do not claim that positive examples are irrelevant to language acquisition, just that they are insufficient.

It is intriguing though to learn of a possible group size effect - some kind of equilibrium selection on the final language state is perhaps going on among the children. Normally of course adult speakers would lay that equilibrium on for them.

 

[Knotted]

I have been careless in my use of the words 'grammar' and 'syntax'. Strictly speaking, a grammar is the framework of a language, comprised of morphology (concerned with the internal structure of words) and syntax (concerned with the order of words in an utterance). Don't see why an impoverished grammar should imply an enriched syntax.

I have never heard a pidgin tongue in real life, but according to what I've read pidgins have little or no syntax or morphology. Because of this, they're only really effective in referring to the here-and-now or to already-familiar situations. (Newspaper headlines sometimes approach pidgin, the Edinburgh Evening News once came up with a good example: DRUG GANG WAR FEAR, though this still sort-of follows English syntax.)

OK I think I'm coming round to your view. The pidgin language may not have enough positive examples of grammar in order for the next generation to creolerise (if there's such a word).

[This seems at least against the spirit of Chomsky's theory. You don't need positive examples of every instance of grammar in order for the child to learn it. The child's innate abilities will generate the grammar.]

However, the way I've stated that, the question of the number of speakers is irrelevant. Simple pidgin should produce simple pidgin in the next generation. Sufficiently developed pidgin should produce creole in the next generation. There should be a critical level of sophistication. But I would suggest that the only way that the population size can tip this balance is by creating grammatical inconsistencies which are then ironed out. That is by negative examples, which contradicts the poverty of stimulus argument.

Having said that, I don't think that is clear that population size is the crucial factor in the case of the Bedouin signers. It seems to be just a reasonable suggestion.

 

[Knotted]

I have been careless in my use of the words 'grammar' and 'syntax'. Strictly speaking, a grammar is the framework of a language, comprised of morphology (concerned with the internal structure of words) and syntax (concerned with the order of words in an utterance). Don't see why an impoverished grammar should imply an enriched syntax.

I'm confusing issues here. I shouldn't have talked about grammar and syntax. Just the complexity of grammatical rules and the richness of the resulting language.

Simple rules might allow greater flexibility and thus a greater variety of examples of correct grammar. There might be more positive examples of grammar in a simple language.

I'm less sure now of what to conclude from that though.

 

[wild_philosophe]

Chomsky's linguistic theory is largely discredited now. I'm not sure even Chomsky believes it any more. Certainly, there is no evidence that we are hard-wired with a universal grammar.

i don't think that's the case.