LightofAbraxas LightofAbraxas

Recursion and the Universal Grammar

▼

Recursion and the Universal Grammar

from Elemental Forums

http://www.wired.com/wiredscience/2007/06/amazon_tribe_sh/

A group of people living in isolation from the rest of the world until very recently has been found not to have recursion in their language.

Who cares, right?

Well, actually, a lot of experts think that this could be a pretty important finding. I don't think that I'm exaggerating when I say that this puts modern linguistic theory on pretty shaky ground. This finding suggests that recursion emerges because it is a useful feature of communication rather than an evolved capability. This, together with a host of other reseach on language primitives in nonhuman animals (see: Vocalize to Localize and others), removes some of the apparent distance between the linguistic capabilities of humans and other primate (and other!) species.

On a more anthropological note, I personally feel like Noam Chomsky is a good example of all that can go wrong with science. He is undoubtedly intelligent, but very dogmatic, and one of the essential features of being a good scientist is being willing to re-evaluate pet theories when contradictory data emerges. Data that contradicts the hypothesis of a universal grammar isn't new. Herb Simon suggested as much 60 years ago in the Architecture of Complexity, and supporting data has been rolling in from the comparative cognition lit pretty steadily for the past 25 years or so.

Anyone else have any thoughts on this?

misc
111,807 views 29 replies +1 Loading…
Reply #26 from Elemental Forums Top

Ah, I think that I understand what you're saying now. I think that people are getting caught up in the phrase 'logical construct', since science doesn't (obviously) demonstrate logical relationships, only empirical ones. Let me know if this is accurate. 

Science 'works' means that science's methodology produces conclusions of a certain kind -- namely, conclusions that're consistent with each other and with other agreed upon norms like verifiability, simplicity, etc. In other words, science's methodology just does produce conclusions of a certain kind because producing conclusions of a certain kind just is a significant part of that methodology.

The lack of recursion means that a language is semantically finite as long as it is conceptually finite.
End of quote

Thanks for the explanation. I think I'm beginning to understand what's meant by recursion. I'm still struggling to wrap my head around what a non-recursive language would look like, though.

Does it mean that it wouldn't be able to contain concepts like 'large' in English, since 'large' is just another way of saying 'not-small'? How would this work with concepts like individual identity, since individual identity is exclusionary?

Reply #27 from Elemental Forums Top

Does it mean that it wouldn't be able to contain concepts like 'large' in English, since 'large' is just another way of saying 'not-small'? How would this work with concepts like individual identity, since individual identity is exclusionary?
End of quote

I'm still banging my aging brain on it too, but I'm pretty sure that it is about grammar, not vocabulary. In English, we can say a single sentence like "John's father's large head makes it hard for John to buy a hat for him." Without recursion, it would need to be more like "John has a father. His father has a large head. Buying a hat for his father is hard for John."

But I'm still mainly guessing because most of what I've skimmed re this thread is math stuff and the linguistics scraps have been very thin, mass-audience stuff.

p.s. Now I'm wondering if prepositional phrases count as recursion. I'm pretty sure that nested possessives are a recursive form.

The Political Machine 2024
Reply #28 from Elemental Forums Top

Hmm... I think you're right about prepositional phrases since they can be used to construct an infinite number of sentences so long as there's at least one noun and at least one preposition in the language...  And I think that if I'm reading the above correctly, saying that a language is nonrecursive is equivalent to saying that it includes a countable number of sentences.

 

Reply #29 from Elemental Forums Top

Quoting zigzag, reply 26

Ah, I think that I understand what you're saying now. I think that people are getting caught up in the phrase 'logical construct', since science doesn't (obviously) demonstrate logical relationships, only empirical ones. Let me know if this is accurate. 

Science 'works' means that science's methodology produces conclusions of a certain kind -- namely, conclusions that're consistent with each other and with other agreed upon norms like verifiability, simplicity, etc. In other words, science's methodology just does produce conclusions of a certain kind because producing conclusions of a certain kind because producing conclusions of a certain kind just is a significant part of that methodology.
End of zigzag's quote

 

Probably accurate.  The scientific method is a logical construct.  Science itself, may or may not demonstrate logical relationships, though in the context of the SM the relationships should still be logical.

 

Your second sentance is a bit obtuse though, at least I am having some difficulty parsing it.  I think I agree with it, but I'm not quite sure ;)

 

It's not about 'Science' working, it's about the Scientific Method working.  With that change I think I agree with you.  Of course the SM is definitional, you do things a certain way, and if you get a certain result you claim that the hypothesis held, if you get another result you discard (or refine more often than not) the hypothesis and test a revised one again.  You don't see me using the word 'conclusion' because many people will think it means the same thing as 'proof' or 'fact'.  Science works when you can collect facts (which are, of course, still based in some set of underlying assumptions) and apply those facts to one or several hypotheses.  The facts can never validate (meaning prove) the hypothesis, all they can really do is invalidate it, if you can find facts which contradict the hypothesis.  'Facts' can also be replaced by 'measurements' or 'observations'.

 

Again, this is an inductive process, but a process which has it's own consistant internal logic.  Most bad science happens when researchers violate that logic.  Intelligent Design is a 'theory' (though it's really not a scientific theory, but consider the word in the vernacular) which violates the SM because it is inherently untestable, and unfalsifiable.  To be sure, there are interesting 'theories' which fall outside the bounds of the SM, but they are not nominally considered 'science' either.  At least until technology or understanding has reached a point where they are able to be tested.