Eduard,
A lot of these criticisms I have no problem with. My problem with them is that, as stated, they frequently presuppose things about Chomsky's thought that aren't so.
1. The deductive/inductive issue deals with the definition of a generative grammar. A generative grammer defines the possible sentences of a language in the same way that a formula like n=2m-1 defines the set of odd numbers. A generative grammar is not a process or a machine that grinds out sentences. It's merely a kind of formal definition.
2. The choice of syntax as the starting point is basically methodological. Sentences are observable formal objects and that makes them suitable for formal definition, a quality that semantic representations don't have.
3. UG is a hypothesis. Hypotheses are by definition unproven. That's not a problem, merely a fact. I happen to believe that when the hypothesis is formulated in a sufficiently explicit way it will be falsified, like any scientific hypothesis worthy of the name. In fact, my statement is a tautology. That doesn't mean that the UG hypothesis can't generate a great deal of interesting research.
4. Basically the same comment. The IH attempts to explain the speed and ease with which first language acquisition takes place. It's an interesting attempt. I just don't happen to think it's necessary.
5. I think modularity is an artifact of the model, not a fact of language or cognition. However, it grows out of language research, not brain research, which is a very different area whose findings don't, in the present state of science, falsify a linguistic hypothesis.
6. Chomsky does not hold that meaning is irrelevant, only that a grammar has to start with structure, not semantics. However, generative grammarians have since the mid 60s had a very strong interest in semantics and have done interesting work in the area.
7. Linguistic dualism has been around for a long time. It's one of Charles Hockett's design features, and it's hard to find purer American Structuralist than Hockett. Deep and Surface Structure are hypotheses that have long since been rejected in generative grammar, in fact almost forty years ago. They've been replaced by much better constrained and articulated hypotheses.
8. Again, we're dealing with hypotheses, which can only be falsified. But once again, the notion transformation gradually faded from generative grammar precisely because it was too powerful and too difficult to constrain.
9. Why should a theory of language lend itself well to language teaching? This is science vs. engineering. Engineering is about making things work in the very messy real world. Science is about describing and explaining nature. The connections between the two tend to be indirect, and their practitioners have very different mindsets, interests, and ways of working.
I disagree with Chomsky on a lot, but one of the things that commonly makes discussions of his work and its impact difficult is the frequency and extravagance with which his work gets misunderstood and misrepresented. He and the research agenda he initiated have made fundamental contributions to linguistics and to modern thought, but we have to deal with his thought, not popular accounts of it.
Herb
-----Original Message-----
From: Assembly for the Teaching of English Grammar on behalf of Eduard C. Hanganu
Sent: Mon 9/4/2006 7:15 AM
To: [log in to unmask]
Subject: Re: On innate knowledge of language
Herb:
I recognize my biases and limitations. Due to the incomplete nature
of my knowledge all messages I write will present a distorted, and
biased perspective on issues. But I don't think that Simpson iss a
major theoretician of the cognitive grammar model. I mentioned him
simply because he offeres a pertinent critique of Chomsky's T-G and
MP main claims. Van Valin and Williams also point to the same issues
in Chomksy's perspective on language. Some of the criticism points
which I consider quite important are:
1. His deductive,and not inductive, approach to language.
2. The syntactocentric perspective, which limits the study of
language to the study of grammar.
3. The unproven idea of a UG ( which is different from the idea that
languages have many similar characteristics.
4. The unproven Innateness Hypothesis.
5. The idea that language processing in the brain is modular ( in
contradiction to the latest brain research).
6. The notion that meaning is largely irrelevant in language, and
that the goal of language research is to discover the structures that
underlie language.
7. The unproven notion of a dual language layer (deep/surface
structure).
8.The unproven, "natural," transformations and movement rules.
9. The extremely theoretical and abstract nature of his theory of
language which makes T-G and MP impractical for language education.
*****
There might be some other things which could be said about Chomsky's
language theory, but the above suffice to show that there is very
little practical, that is, instructional value, in his perspective.
Eduard
On Sun, 3 Sep 2006, Herbert F.W. Stahlke wrote...
>Eduard,
>
>You very slightly misrepresent the status of arguments on the
innateness =
>hypothesis. While I agree with Sampson that the Innateness
Hypothesis =
>is seriously underdetermined by the data, Sampson's alternative is
also =
>too vaguely stated to be called a hypothesis. While I tend to opt
for =
>the cognitivist position, that's a matter of opinion and
temperament, =
>not of fact. We don't as yet know enough about either language or =
>cognition to claim that a certain relationship exists between them.
=
>Such claims may be interestingly argued but cannot be anywhere near =
>conclusive. Sampson doesn't falsify (and you didn't claim this) =
>Pinker's position. Rather he shows that we can't yet support one =
>position or the other empirically. In such a situation I incline =
>towards the less specific. There is a strong philosophical
tradition =
>beyond inclining towards the more specific too, especially since the
=
>stronger position is more explicit and therefore easier to falsify.
=
>While I have an opinion, I am agnostic as to the outcome of the
debate. =
>I suspect both sides are wrong.
>
>Herb
>
>
>-----Original Message-----
>From: Assembly for the Teaching of English Grammar on behalf of
Eduard =
>C. Hanganu
>Sent: Sun 9/3/2006 4:29 PM
>To: [log in to unmask]
>Subject: Re: On innate knowledge of language
>=20
>Robert:
>
>I do not want to assume that you are not familiar with what
implies=20
>providing "bibliographical information" in support of a
hypothesis.=20
>You are not providing the information requested. Instead, you are=20
>making reference to Pinker's "The Language Instinct." But, as Herb=20
>has corroborated, Sampson has already provided evidence that
Pinker's=20
>case is too weak to be considered. You also mention an article
which=20
>you have not read. Are we moving into anecdotal? My request stands:
>Please provide BIBLIOGRAPHICAL INFORMATION, for the Innateness=20
>Hypothesis,as Chomsky and Pinker state it, that is, research
evidence=20
>that language is innate, and not, as cognitive linguistics
affirms,=20
>part of the human cognitive process.
>
>
>Eduard=20
>
>
>
>On Sun, 3 Sep 2006, Robert Yates wrote...
>
>>Eduard has an interesting challenge.
>>
>>>>> [log in to unmask] 09/03/06 7:46 AM >>>
>>
>>Please, be so kind and provide the bibliographical information
which=20
>>includes research that shows evidence that children "know=20
>>(unconsciously) what a noun [ or other part of speech] is." I=20
>haven't=20
>>found yet such evidence in all the language literature I have read.
=20
>>
>>***
>>If children did not know what nouns are unconsciously we might
expect
>>all kinds of "errors" around nouns. For example, we might have the
>>articles in very strange position, we might have the plural "s"=20
>attached
>>to words that can't be pluralized, we might expect comparative and
>>superlative morphology attached to nouns, we might expect
scrambled=20
>word
>>order in apparently noun phrases.
>>
>>I don't know of ANY research that shows children's confusion with
>>respect to nouns or any category. Perhaps, Eduard could share us
the
>>evidence that kids don't know what nouns are.
>>
>>Pinker, in The Language Instinct, notes that there is no child
data=20
>with
>>the following kinds of errors for yes-no questions. (See the
Chapter
>>Baby Born Talking, p. 276 in my edition for this discussion)
>>
>>He is smiling -- Does he be smiling?
>>She could go. Does she could go?
>>
>>If you teach ESL, you have heard such examples in the questions
of=20
>ESL
>>students. Why is it kids learning English understand how "do"
works=20
>for
>>questions and adult L2 learners can have very different principles?
=20
>If
>>language principles are not innate, we should expect some kids to=20
>have
>>"wild" grammars with respect to this property of the English=20
>auxiliary
>>system.
>>
>>Of course, there is PUBLISHED evidence that meets Eduard's=20
>challenge.=20
>>One example is summarized in Pinker (Chapter 5, pages 129 +). (I
have
>>not read the actual paper). It is work by Peter Gordon with
compound
>>nouns. Notice the following property with compound nouns. In the
>>compound, irregular plurals are possible; regular plurals aren't.
>>
>>1a) purple people eater
>> b) purple baby eater
>> c) *purple babies eater
>>
>>2 a) cookie monster
>> b) *cookies monster (What kind of monster would only eat ONE
>>cookie?)
>>
>>3) a) rat catcher
>> b) *rats catcher
>>
>>Actually, if I had a lot of rats in my house (in other words, it
was
>>rat-infested, but not *rats-infested) I would want all of the rats
>>caught, not just one.
>>
>>Gordon tested this contraint on compound structures on three and
five
>>year old kids with questions like the following:
>>
>>Experimenter: Here is a monster who eats mud. What do you call=20
>him? =20
>>Kid: A mud-eater.
>>
>>Experimenter: Here is a monster who eats mice. What do you call
him?
>>Kid: A mice-eater.
>>
>>And, the crucial question is the following:
>>Experimenter: Here is a monster who eats rats. What do you call him?
>>
>>According to Pinker, Gordon found that his 3 and 5 year old kids all
>>responded: A rat-eater.
>>
>>Think about the kind of knowledge a kid needs to have to
recognize=20
>that
>>even though irregular plurals can be used in such compounds but=20
>regular
>>plurals can't. And, remember the immediate INPUT.
>>
>>What do you call a monster that eats RATS? The input in this=20
>question
>>would favor *"rats-eater."
>>
>>I have no idea what the story is if kids don't know what a noun
is=20
>and
>>the different properties of IRREGULAR and REGULAR nouns.
>>
>>Perhaps, Eduard will let us know.=20
>>
>>Bob Yates, Central Missouri State University
>>
>>To join or leave this LISTSERV list, please visit the list's web=20
>interface at:
>> http://listserv.muohio.edu/archives/ateg.html
>>and select "Join or leave the list"
>>
>>Visit ATEG's web site at http://ateg.org/
>
>To join or leave this LISTSERV list, please visit the list's web =
>interface at:
> http://listserv.muohio.edu/archives/ateg.html
>and select "Join or leave the list"
>
>Visit ATEG's web site at http://ateg.org/
>
>To join or leave this LISTSERV list, please visit the list's web
interface at:
> http://listserv.muohio.edu/archives/ateg.html
>and select "Join or leave the list"
>
>Visit ATEG's web site at http://ateg.org/
To join or leave this LISTSERV list, please visit the list's web interface at:
http://listserv.muohio.edu/archives/ateg.html
and select "Join or leave the list"
Visit ATEG's web site at http://ateg.org/
To join or leave this LISTSERV list, please visit the list's web interface at:
http://listserv.muohio.edu/archives/ateg.html
and select "Join or leave the list"
Visit ATEG's web site at http://ateg.org/
|
