Deficiencies in the Cambridge Grammar of the English Language

[Tdol]

Sorry- I didn't quote the post.

[rewboss]

Originally Posted by alienvoord

I mean the unconscious knowledge that we use to construct utterances. Chomsky believes that most of this knowledge is innate, but I don't think you have to believe that in order to subscribe to this idea. Whether the knowledge is innate or acquired really doesn't matter. Either way, it consists of rules that we follow unconsciously. We do not normally violate these rules - we put the verbs and nouns in a certain order, we don't put adverbs between certain words, etc.

Actually, I think Chomsky's argument is considerably weakened if these rules are not innate. If they are not innate they must come from somewhere -- they don't just pop into existence.

We don't put adverbs between certain words because those who taught us language (whether through formal instruction or natural methods) don't put adverbs between certain words. We do, however, sometimes violate these rules; the evidence is in the way all languages change and evolve.

A native speaker's understanding of grammar is unconscious, yes, but it has still been learned. If you drive a car, especially if you have been doing so for a long time, you do a lot of it instinctively, and you don't usually have to think very hard; however, you still had to learn to drive. (It's not a perfect analogy, of course, since when we learn to drive we do so by formal instruction first and then our skills are fine-tuned by experience, but it still illustrates the point.)

These syntactic rules are certainly not universal. In modern German declarative sentences, for example, the inflected verb is usually the second item in a main clause (and if the first item is not the subject, the subject follows the inflected verb), and one of the last items in a subordinate clause. A German doesn't have to think about these rules at all (except when the sentence becomes very long and constituent parts of a verb phrase end up a great distance from one another), but the difficulty most non-native speakers have with such structures indicates that this "innate" understanding of the underlying grammar is peculiar to German speakers. However, we know on other grounds that your biological heritage has no effect whatsoever on your ability to learn a specific language.

What you're basically saying is that we humans have the ability to formulate highly complex and subtle rules even though they are unable to explain them, and without being conscious of the process. But that doesn't mean that these rules do not have to be learnt.

This is where we get carried away by terminology if we are not careful. The moment we hear the word "unconscious", we understand "instinctive", and when we hear "instinctive", we understand "innate" and therefore "doesn't have to be learned".

And if all languages do have superficial similarities -- all languages having verbs, for example -- that doesn't mean much. Studying sufferers of various types of aphasia, for example, we know that different parts of the brain are responsible for certain parts of language, but that is about as far as it goes. In fact, different languages have very different ways of expressing the same thing, and different ways of using verbs. In European languages we inflect our verbs to indicate person, time, tense and so on; non-European languages may use completely different mechanisms, such as particles or honorifics. The only real similarity is that we all have words for describing actions, but how else are you going to develop a language? Besides, if our languages, as seems likely, are all related to one another, it's hardly a surprise that they have superficial similarities.

[alienvoord]

Originally Posted by rewboss

Actually, I think Chomsky's argument is considerably weakened if these rules are not innate. If they are not innate they must come from somewhere -- they don't just pop into existence.

OK, but what are we talking about now? Are you objecting to transformational grammar because you believe that these rules are not innate?

A native speaker's understanding of grammar is unconscious, yes, but it has still been learned. If you drive a car, especially if you have been doing so for a long time, you do a lot of it instinctively, and you don't usually have to think very hard; however, you still had to learn to drive. (It's not a perfect analogy, of course, since when we learn to drive we do so by formal instruction first and then our skills are fine-tuned by experience, but it still illustrates the point.)

I think a better analogy is learning to walk.

What you're basically saying is that we humans have the ability to formulate highly complex and subtle rules even though they are unable to explain them, and without being conscious of the process. But that doesn't mean that these rules do not have to be learnt.

This is where we get carried away by terminology if we are not careful. The moment we hear the word "unconscious", we understand "instinctive", and when we hear "instinctive", we understand "innate" and therefore "doesn't have to be learned".

Maybe that's what you think when you hear "unconscious," but I didn't say any of that. You haven't convinced me that they are or are not innate, or that it matters at the level of talking about case and coordinated pronouns. In the same way, perhaps, that biologists are not concerned with the origins of life when talking about evolution.

And if all languages do have superficial similarities -- all languages having verbs, for example -- that doesn't mean much.

I think you've got it backwards. One of the points of transformational grammar is that languages are superficially different, and are the same at a deeper level.

Your example about German word order has been used to demonstrate points of transformational grammar.

The fact is that we construe "who do you want to fight?" and "who do you wanna fight?" as having different meanings, without being able to articulate why this is so, and without being explicitly taught how to tell the difference. We have to account for that, and for many other pieces of data like it. Transformational grammar accounts for it.

If you want to object to transformational grammar, fine, but be aware that you're talking about several decades of mainstream linguistic theory. That doesn't mean that it's right, but it means that it has a lot of prononents and a lot of explanatory power. If you have an alternative theory, your theory has to account for everything that transformatlional grammar accounts for.

[BobK]

Originally Posted by alienvoord

...
If you want to object to transformational grammar, fine, but be aware that you're talking about several decades of mainstream linguistic theory. That doesn't mean that it's right, but it means that it has a lot of prononents and a lot of explanatory power. If you have an alternative theory, your theory has to account for everything that transformatlional grammar accounts for.

This sort of argument makes me feel uncomfortable.

On the one hand, I agree with it; people can't just go around saying 'That's rubbish', dismissing something with a sound academic pedigree. Besides, when I had an academic lifestyle, and was up-to-date with the latest linguistic theories, I found a lot of the arguments of transformational-generative grammar persuasive - even, in some respects, unarguable.

On the other hand, I resent being bored into submission by people who are simply adept at producing arguments and citations (this isn't a personal jibe at anyone - it's just my reason for staying out of these arguments as far as I can [inviting though they sometimes are ]).

b

[alienvoord]

Originally Posted by BobK

On the other hand, I resent being bored into submission by people who are simply adept at producing arguments and citations (this isn't a personal jibe at anyone - it's just my reason for staying out of these arguments as far as I can [inviting though they sometimes are ]).

It's not my intention to bore anyone into submission! What I meant was: rewboss should read some of the literature (if they haven't already) and know what they're getting into if they want to argue about transformational grammar. I'm not sure that I can say anything more about it without getting out my old books and copying passages, but rewboss doesn't need me to do that.

[BobK]

No, I honestly didn't mean you AV. It's just that when I was a student I had more time for academic arguments (I even read Aspects... - that's how conscientious I was )

b

[alienvoord]

Originally Posted by BobK

No, I honestly didn't mean you AV. It's just that when I was a student I had more time for academic arguments (I even read Aspects... - that's how conscientious I was )

b

For me, the further I get away from my university years, the more interested I get this stuff.

I do think that an alternative theory needs to account for everything that the mainstream theory accounts for, though. I'm not sure if you were referring to that specifically or not. I think that applies in any scientific field.

[MikeNewYork]

Originally Posted by BobK

No, I honestly didn't mean you AV. It's just that when I was a student I had more time for academic arguments (I even read Aspects... - that's how conscientious I was )

b

I think Chomsky was very clever. I'm not so sure that his theories were clever, but he cleverly proposed things that can't really be disproven. Of course, they can't be proved either. One of his initial premises (one which I've never accepted) is that brain of children is not developed enough to produce unique language. First of all, I think that we know so little about how the brain works, that this premise is absurd on its face. There seem to be some cracks developing in some of Chomsky's foundation premises. Some say they are for the "birds".

ScienceDaily: The Birds And The B's: Challenging Chomsky, Starlings Learn 'Human-only' Syntax Patterns

[alienvoord]

Originally Posted by MikeNewYork

I think Chomsky was very clever. I'm not so sure that his theories were clever, but he cleverly proposed things that can't really be disproven. Of course, they can't be proved either. One of his initial premises (one which I've never accepted) is that brain of children is not developed enough to produce unique language. First of all, I think that we know so little about how the brain works, that this premise is absurd on its face. There seem to be some cracks developing in some of Chomsky's foundation premises. Some say they are for the "birds".

ScienceDaily: The Birds And The B's: Challenging Chomsky, Starlings Learn 'Human-only' Syntax Patterns

I have some problems with Chomsky's theory too; my problem is that linguistic competence cannot be directly observed. On the other hand, transformational grammar has a lot of explanatory power. I'm not familiar with any other linguistic theory as powerful.

I think Gentner reached the wrong conclusion with his birds. The people and Language Log think Gentner's claim that starlings know recursive embedding is too strong: Liberman tries to show that the same experiment, when performed on humans, will fail. More info here and here.

The birds became habituated to a pattern like AAAABBBB. I haven't read the paper but I think this means they assumed that AAABBB had the structure
[A[A[AB]B]B]
But there's nothing to suggest that the starlings weren't simply recognizing the numbers of As and Bs and responding when the numbers were the same.

I don't think this is a challenge to Chomsky at all. Something that might be a challenge is the claim that Pirahã has no recursive embedding.

[Tdol]

There have been so many claims about animals and language over the years, but they generally seem to fizzle out.