Linguistsics anyone?

[Olowkow]

I did my graduate degree in theoretical linguistics using a theory called "transformational generative grammar" in the '70s when Chomsky, Ross, and Lakoff held sway. They were heady times for the study of syntax, phonology and semantics. Seeing the handwriting on the wall after completing my degree, the field was going nowhere, I abandoned this field for electronics. Seems like it has not gone very far in the past 35 years.

Now if no one here is interested in this, I understand perfectly and am happy to let this thread die of its own weight. It is an arcane discipline/science at best. The only data we use is our "linguistic intuitions" as native speakers.

To start things off, you'll need to know that the asterisk (*) is used by linguists to mean that a sentence is "ill formed". I remember a conference where someone stood up, could have been George Lakoff, and said, "Linguists of the world unite! You have nothing but your ass to risk!" You had to be there.

The "ill formed" sentence is a useful tool for proving linguistic hypotheses in syntax....actually, it's the only tool!

A typical question that would intrigue a linguist is, "what is wrong with sentence (1)?" The answer is more interesting than, "it ends with the word and". Similarly, why is sentence (2) well formed? What constraint in English is violated in (1), possibly a universal constraint in all languages, but not in (2)?

(1) *Who did Henry frighten the burglar and?
(2) Henry frightened the burglar and who?

Hint: (3) Henry frightened the burglar and his accomplice.

(Don't worry about "who" vs. "whom". Irrelevant.)

Check out Wiki to get up to speed on "generative or transformational grammar"

http://en.wikipedia.org/wiki/Transformational_grammar
http://en.wikipedia.org/wiki/Generative_grammar

 

[drzeus99]

Huh?

 

[Complexity]

I started studying Chomsky's transformational work in the seventies as well. Very cool.

I got into Chomsky's later work in government-binding theory in the last eighties. More cool.

One person in particular on this forum really dislikes Chomsky's linguistic work and has contempt for anyone who respects it. She isn't wrong about everything, but this is one of her most irrational bad points. Don't let her get to you.

 

[PixyMisa]

All the working linguists I've met (which, admittedly, comes to a grand total of three, only one of whom I've conversed with on multiple occasions) consider Chomsky's Universal Grammar to be precisely the reason why linguistics hasn't advanced since the 70's, and I agree with them. The basic problem is that it's unfalsifiable, modified ad-hoc to fit whatever observations come in.

I'm less familiar with Lakoff's history, though he currently seems to be peddling linguistic woo to anyone who'll listen, and I don't know Ross at all.

Real linguistics is fascinating, but Chomsky's later work is linguistic homeopathy.

 

[Abe_the_Man]

me no good at english? thats' unpossible!

 

[Complexity]

Real linguistics is fascinating, but Chomsky's later work is linguistic homeopathy.

Nonsense.

 

[Tokenconservative]

Nonsense.

Nevertheless, he proved himself in the discipline (one that has always fascinatied me, especially as it applies to advertising) in the 70s, then went on to make a fool of himself with his political ranting.

Tokie

 

[volatile]

All the working linguists I've met (which, admittedly, comes to a grand total of three, only one of whom I've conversed with on multiple occasions) consider Chomsky's Universal Grammar to be precisely the reason why linguistics hasn't advanced since the 70's, and I agree with them. The basic problem is that it's unfalsifiable, modified ad-hoc to fit whatever observations come in.

There's an article in New Scientist this week about Daniel Everett and his disagreements with Chomsky, particularly on the linguistics of the Pirahã in Brazil, but his theory is hardly rock-solid either. The alternatives to Chomskian thought are hardly overwhelming, or convincing.

Chomskian linguistics grew as a repsonse to behaviorism - Chomsky didn't pluck his ideas out of thin air, they're a refinement and improvement on previous theories of language, and they were based, at least in the beginning, on an empirical framework.

Also, it's not unfalsifiable, as Everett's work suggests that if his theories on the Piraha langauge are correct, then the Chomskian theory is falsified. From Wikipedia:

"The language is claimed to have no relative clauses or grammatical recursion, but this is not clear. Should the language truly feature a lack of recursion, then it would be a counterexample to the theory proposed by Chomsky, Hauser and Fitch (2002) that recursion is a crucial and uniquely human language property."

I'm less familiar with Lakoff's history, though he currently seems to be peddling linguistic woo to anyone who'll listen, and I don't know Ross at all.

Philosophy in the Flesh and Metaphors We Live By are hardly woo now, are they? Come on.

 

[Olowkow]

Thanks for all the interesting comments. I'm surprised at the response.

There's an article in New Scientist this week about Daniel Everett and his disagreements with Chomsky, particularly on the linguistics of the Pirahã in Brazil, but his theory is hardly rock-solid either. The alternatives to Chomskian thought are hardly overwhelming, or convincing.

I read a long article about Everett's life with the Piraha. Not sure what to make of it, but he was certainly dedicated.

Here is one article, not the long one. Can't seem to find it.
http://www.crystalinks.com/piraha.html

The part about their not having a number system, and unable to learn to count is really fascinating. Also, he was a missionary, and when he tried to explain god, I guess they just thought he was nuts!

As for the lack of relative clauses, I'd have to know many more details, but he waited something like 20 years to publish. Some languages like Chinese just treat such embedded clauses as adjectives:

"The cat that ate the rat", becomes "the ate rat + [possessive particle] cat".

Chomskian linguistics grew as a repsonse to behaviorism - Chomsky didn't pluck his ideas out of thin air, they're a refinement and improvement on previous theories of language, and they were based, at least in the beginning, on an empirical framework.

A welcome improvement at the time too! Far from "homeopathy", but not the ultimate answer either. It's one of those things where you feel that it contains elements of what the correct theory would have to contain.

"The language is claimed to have no relative clauses or grammatical recursion, but this is not clear. Should the language truly feature a lack of recursion, then it would be a counterexample to the theory proposed by Chomsky, Hauser and Fitch (2002) that recursion is a crucial and uniquely human language property."

Yes, "recursion" is at the heart of generative syntax in the Chomskian view. I agree with those who fault him for his political views.

Haj Ross was one of his students at MIT. Bright guy. Did a lot with universals and constraints. Most notably, the "complex noun phrase constraint", which everyone now violates! Interesting. Lakoff has done a lot of work recently with his popularizing of certain claims. Kind of interesting. Not woo, just observations.

The example in the OP shows the constraint preventing the movement of a questioned element from a conjoined structure, by the way. I am told by my friends still in the field that the biggie these days is "minimalism". I don't get it, because I thought that was always the case. A sort of Occham's razor.

One thing for sure is that phonology theory definitely works, and has produced a lot of interesting work. Generative semantics never went far, unfortunately. There was a ton of interesting data in that field at the time.

Those who scoff at the word "grammar" should understand that it does not mean what you learned in high school. Instead, it is a set of "rules", better stated a set of principles, which describe how speakers actually behave linguistically (descriptive), not how they should behave (prescriptive). Theoretical linguists don't care if you say "ain't"! They just report it and analyze it!



My research claimed a "grammar" of bilingualism. That is, for bilingual members of a community of bilinguals who code switch, there is a rule governed behavior for such sentences. That is, code switching is not just random use of elements from both languages. There are "ill formed" and "well formed" utterances with regard to this as well.

 

[volatile]

I read a long article about Everett's life with the Piraha. Not sure what to make of it, but he was certainly dedicated.

Here is one article, not the long one. Can't seem to find it.
http://www.crystalinks.com/piraha.html

The part about their not having a number system, and unable to learn to count is really fascinating. Also, he was a missionary, and when he tried to explain god, I guess they just thought he was nuts!

And he became an atheist, divorced his wife and alienated his children as a result...!

As for being unable to count, Everett's thesis, as far as I can tell, is that the Piraha could count, but don't want to. Or something.

As for the lack of relative clauses, I'd have to know many more details, but he waited something like 20 years to publish. Some languages like Chinese just treat such embedded clauses as adjectives:

"The cat that ate the rat", becomes "the ate rat + [possessive particle] cat".

Well, exactly. I haven't read any of this in detail, and I'm sure Chomsky must have written his own rebuttal, but it strikes me that you'd have to make a really, really strong case that the Piraha have no grammatical equivalent of relative clauses to undermine Chomsky. Remember, Everett was not originally a linguist by training, but only became one because of his interest in documenting the Piraha.

Nevertheless, that he even has a case to make shows that the Chomskian hypothesis is at least falsifiable in principle.

A welcome improvement at the time too! Far from "homeopathy", but not the ultimate answer either. It's one of those things where you feel that it contains elements of what the correct theory would have to contain.

I agree. I think the answer's going to come out of neuroscience or neurolinguistics rather than pure linguistics, though.

Yes, "recursion" is at the heart of generative syntax in the Chomskian view. I agree with those who fault him for his political views.

I have some sympathy with Chomskerian politics, but his politics are utterly irrelevant to his linguistic theories.

Haj Ross was one of his students at MIT. Bright guy. Did a lot with universals and constraints. Most notably, the "complex noun phrase constraint", which everyone now violates! Interesting. Lakoff has done a lot of work recently with his popularizing of certain claims. Kind of interesting. Not woo, just observations.

Yes - I agree. Metaphors We Live By, which I admit I've only skim read, seems to support this view, that Lakoff is describing linguistic usage rather than trying to develop some kind of metaphysical woo linguistics. Philosophy in the Flesh would seem to support the observation that he's fundamentally interested in a materialistic understanding of language (and everything else).

My research claimed a "grammar" of bilingualism. That is, for bilingual members of a community of bilinguals who code switch, there is a rule governed behavior for such sentences. That is, code switching is not just random use of elements from both languages. There are "ill formed" and "well formed" utterances with regard to this as well.

That's relatively uncontentious, I'm sure.

 

[Olowkow]

LOL!!!

That's relatively uncontentious, I'm sure.

It sure wasn't in the '70's! We "believers" took a lot of heat. It is now common knowledge I am told.

 

[Olowkow]

All the working linguists I've met (which, admittedly, comes to a grand total of three, only one of whom I've conversed with on multiple occasions) consider Chomsky's Universal Grammar to be precisely the reason why linguistics hasn't advanced since the 70's, and I agree with them. The basic problem is that it's unfalsifiable, modified ad-hoc to fit whatever observations come in.

I'm less familiar with Lakoff's history, though he currently seems to be peddling linguistic woo to anyone who'll listen, and I don't know Ross at all.

Real linguistics is fascinating, but Chomsky's later work is linguistic homeopathy.

One has to remember that in all scientific pursuits, you will have those whose only claim to fame is to attempt to "kill the current dragon". Yes, there are always the "believers" and "skeptics", as in any field. I think it is a stretch, however, to say that "Chomsky's ideas were the reason why linguistics hasn't advanced". I was there, and I could have come up with a better idea, but I didn't...not for lack of trying, nor did any one else. It was quite compelling at the time. Syntax is a very difficult problem to crack. Actually much more difficult than you might at "first blush" imagine.

Lakoff's metaphor stuff is interesting to me. It is an extension of what he always has been doing. Super intelligent guy, and lots of fun in person.

No, not homeopathy. LOL. Just wondering what "real linguistics" is in your opinion. To me, the phonology aspect is the most convincing...just amazing how the mind deals with phonemes and sounds. Just bizarre.

 

[volatile]

LOL!!!



It sure wasn't in the '70's! We "believers" took a lot of heat. It is now common knowledge I am told.

Well, it's what I was taught at Undergrad level a decade ago, so I guess it is!

 

[Complexity]

Chomsky has done some remarkable thinking and research on linguistics over many decades. I first encountered his work on phonology and speech production around 1971, when I was in high school. I studied his work in some linguistics classes in the mid-seventies at the University of Illinois at Urbana-Champaign, especially his work on transformational generative grammars. I was truly impressed, both by the work and by the way of thinking about language and the brain that had led to it.

I learned a great deal about model development, hypothesis generation and testing, and the scientific method from my study of linguistics.

When I was studying at Northwestern for my doctorate in the late eighties, I took some more linguistics course in which some of Chomsky's later work was central. Central to this later work of Chomsky's is his government-binding theory. He has striven to understand the nature of human language, what grammars have in common and how they differ, how we learn the languages we are exposed to as children, and how language provides a window on the brain.

I've heard criticism on this forum that his work is considered passe, that it isn't useful for natural language processing, and so forth.

Chomsky brought about at least two revolutions in linguistics. I'm not concerned with what is regarded as fashionable by some in linguistics, and his work was not developed for the implementation of natural language processing systems. Chomsky has sought to understand language and the mind on language. His work is seminal and unprecedented.

 

[Dancing David]

Nevertheless, he proved himself in the discipline (one that has always fascinatied me, especially as it applies to advertising) in the 70s, then went on to make a fool of himself with his political ranting.

Tokie

Does that mean you actualy listen to what Chomsky has to say? Or do you just disagree with him.

 

[Olowkow]

I studied his work in some linguistics classes in the mid-seventies at the University of Illinois at Urbana-Champaign, especially his work on transformational generative grammars.
I learned a great deal about model development, hypothesis generation and testing, and the scientific method from my study of linguistics.

Amen to that. I myself recall a "seminal" moment when I said to myself, "Oh, now I get it!". It took me a while to understand what the problem was.

[Chomsky] has striven to understand the nature of human language, what grammars have in common and how they differ, how we learn the languages we are exposed to as children, and how language provides a window on the brain.

I was involved for several years in a research project investigating the written language of deaf children. As frustrating as the tools were to use, transformational grammar did provide some insights.

...his work was not developed for the implementation of natural language processing systems.

True, but many tried to force this issue early on.