elwoodblues

That's certainly the way it looks in my philosophy of language class. We're looking at his main thesis, and it seems essentially irrefutable. Which is a damned serious allegation.

His basic idea is this. There is a universal grammar that we are all born with. It's an actual structure _somewhere_ in the brain. He then goes on to admit, quite up front, that these structures aren't detectable by modern science. But he asserts they're there, and at some undetermined point in the future his claims will be supported by evidence.

So, he's making a specific claim. "There are structures in the brain that determine basic grammar." In order to refute it, you need to confirm the general negative claim, "There is NO structure in the brain that determines basic grammar." This would be extremely difficult if neurology were sufficiently advanced. As it is, it's basically impossible.

So there is no way to disprove his theory.

That doesn't make it a good theory. In fact, it makes it very, very bad. In any hard science, this would be a fatal flaw in any theory. Is linguistics theory really that different?

Have any linguists raised this point? What're Chomsky's answers to this?

Feather

No, in order to refute "it," "it" would have to actually be something more than just an opinion. There is nothing to refute until he produces evidence for it.

elwoodblues

I simplified and cut out quite a bit. There is a fair amount of convincing experimental evidence for something going on. Look at competence/production distinction for this. However, there is no direct evidence for structures in the brain. Chomsky basically asserts it, on the implied ground that all other possible mechanisms had been eliminated.

Don't take my statement as a comprehensive overview of Chomsky's position by any means. It's a quick and dirty run down. However, he states quite unashamedly that there is no direct evidence for his postulated physical structures of universal grammar.

beausoleil

Actually, I find his arguments quite convincing - humans are born with a tendency to adopt certain grammatical structures and not others. All those examples of the sorts of mistakes people don't make\are certainly thought provoking.

I agree there is no explanation of exactly how the physical structures of the brain achieve this selectivity, but the selectivity itself is falsifiable, and that is what the universal grammar is based on, as I understand it.

You could turn the argument round...

Human grammar is a subset of possible grammars
All human cognitive features map to brain structures
We can't find the brain structure corresponding to human grammar.

Maybe it is the second part of the argument rather than the first which is unfalsifiable. Failure to find a brain structure is not seen as a problem - 'we just didn't find it yet'

echidna

Well I suppose one would need to analyse the common grammatical elements between the world’s languages & the countless syntaxes in existence. But to propose that there is a single grammar we are born with seems unlikely if I understand correctly. Logic would seem to dictate the existence of concepts such as noun, verb, adjective, adverb, while syntax itself varies impossibly all over the world. There seems no common word order. I really don’t understand what he means by a single grammar.

Although at the same time, I think it can be debated (I acknowledge with great difficulty), that communication promotes some syntaxes over others in the evolutionary sense. But there seems little need for anything innate.

Possibly one could begin to analyse MRI’s & brain damage cases for what effects these might have on language.

Certainly there appear to be language centres in the brain, and brain damage in specific areas can affect language perception and performance.

This site : <a href="http://www.med.harvard.edu/publications/On_The_Brain/Volume4/Number4/F95Lang.html," target="_blank">http://www.med.harvard.edu/publications/On_The_Brain/Volume4/Number4/F95Lang.html,</a> lists the region around the Sylvian Fissure as the brain’s language centre, and while not dissected into a grammar zone, may be the closest to Chomsky’s assertion.

Now further, many very basic human sign languages used for non-verbal intellectual disabilities strongly resemble the sign language successfully taught to Coco and her disciples. So do gorilla brains also contain this “structure” ? It would seem so.

seanie

I've never read any Chomsky on linguistics but I've read 'The Language Instinct' by Steven Pinker.

I couldn't really do justice in trying to describe his arguments but he's firmly of the opinion of an innate grammar. That can be expressed in different ways hence different actual grammar systems but there is a fundamental core that binds them together.

He gives two 'empirical' examples to support the assertion that grammar is innate.

The first is indentured slaves in the Phillipines at the turn of the century (I think?). There was an influx of workers from a vartiety of backgrounds. They developed a pidgin language. This was fairly crude and grammmatically limited.

However the next generation spoke a much richer, fully grammatical creole. The children who had only ever been exposed to the pidgin had re-ordered it in a more grammatical form. There was a spontaneous transfomation into a fully fledged language.

The second example was Nicaragua in the late 70's early 80's. When the Sandanistas took over there were no facilities for the deaf. They decided to open a deaf school. They attempted to teach a standard Nicaraguan Sign Language to the adults.

Similar to the first generation in the Phillipines the adults struggled and effectively had a pidgin sign language. Useful but grammatically limited. The children, even though predominantly exposed to this poorly grammatical language, developed their own, much richer, fully fledged sign language. I think Nicaragua ended up having two distinct though related sign languages. The one they originally attempted to teach and the one the kids developed themselves.

The argument was there was an innate grammar skill in children whereby, when exposed to a comparatively crude pidgin, they instinctively transfromed it into a fully grammatical language.

However this ability wore off in adolesence.

I don't know enough to properly judge but it was a fascinating book.

Godless Dave

I majored in linguistics, and my professor's take was basically "Chomsky's a brilliant man, but don't believe everything he says." He tends to take some ideas to an extreme and is fond of making wild assertions. Kind of like his political ideas. To be fair, the best way to get attention in the social sciences is to make wild counterintuitive assertions. Yet another reason I didn't go to grad school for linguistics.

MortalWombat

How can one distinquish between "unversal grammar" as an actual brain construct versus being a consequence of humans living in a world in which there are objects (nouns), actions (verbs), and the recognition that objects and actions can be categorized (adjectives and adverbs)? My guess would be that there are a limited number of ways in which languages could be constructed given the above, which may artificially give the impression of a "universal grammar."

Godless Dave

I think the concept of a "universal grammar" is a helpful way of looking at human language ability, but I see it more as an analogy than an accurate description of what's going on in the brain. My opinion is that the nature of human cognition is what gives shape to grammar, and that, along with Mortal Wombat's point, is why we see the similarities between unrelated languages.

Even "incorrect" grammar has its own rules. Black English/aka Ebonics has consistent grammatical rules that people learn just from hearing it and interacting with other speakers. Kids learning English figure out rules on their own and then have to learn the exceptions - they'll say "mans" as plural for "man" because they've figured out that plurals are made by adding an "s" or "z" sound. This is "wrong" but it follows the rules they've learned so far - before their parents realize they're teaching them any rules.

doubtingt

The inability to directly observe physical structures of universal grammer is in no way a requirement to test his theory.
Many, many theoretical entities in science cannot be directly observed. All that matters is that we can make predictions from these assumptions about events that we CAN observe. This happens all the time in every branch of science.

As you point out, there is already empirical evidence that is predicted by and can be explained by this theory. While definitive answers may still be some time away, the current state of the evidence certainly seems to favor some version of a language instinct over the
"blank slate" hypothesis.