![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Entry tags:
Languages (and sewing and elephants)
I read this New Yorker article about the Piraha tribe and language yesterday - long but fascinating.
I've never really bought the Chomsky universal grammar thing. More than anything else, it just seems to be unnecessarily complicated. Children are exposed to a lot of language from the moment they leave the womb; to me it would seem more surprising if they didn't pick it up so readily. (And, really, we're talking about 5-6 years to adult competence, which is quite a long time.)
There's also not been very much research done on the extent to which surrounding adults model and correct children's language; and although there has been some research on the sorts of error that children make (which Chomsky claims in support of his thesis, on the basis that certain sorts of grammatical error are very rare/never seen), again, given the extent to which they have models around them, I'm not sure that this means much either. If you never hear construction X, though you hear constructions Y, Z, and A through D, surely it would be more surprising if you experimented with construction X, rather than making mistakes in the correct usage of the other constructions? (The connectionist people are working on computer models of this, with some success.)
Anyway: the Piraha are I think interesting in the extent to which culture and language work together. One of the pro-Chomsky arguments is about the sorts of sign language which deaf children will self-develop; it would be interesting to know what sort of sign language deaf Piraha children would self-develop. My suspicion is that even without being able to hear, the cultural background would affect the grammar; and so Piraha sign language would not fit the Chomskyan universal grammar either.
In other news: I finished my summer dress, which has come out pretty well.
And yesterday I went along the south bank of the Serpentine to work, & encountered elephants! Anyone in the vicinity should go visit them.
I've never really bought the Chomsky universal grammar thing. More than anything else, it just seems to be unnecessarily complicated. Children are exposed to a lot of language from the moment they leave the womb; to me it would seem more surprising if they didn't pick it up so readily. (And, really, we're talking about 5-6 years to adult competence, which is quite a long time.)
There's also not been very much research done on the extent to which surrounding adults model and correct children's language; and although there has been some research on the sorts of error that children make (which Chomsky claims in support of his thesis, on the basis that certain sorts of grammatical error are very rare/never seen), again, given the extent to which they have models around them, I'm not sure that this means much either. If you never hear construction X, though you hear constructions Y, Z, and A through D, surely it would be more surprising if you experimented with construction X, rather than making mistakes in the correct usage of the other constructions? (The connectionist people are working on computer models of this, with some success.)
Anyway: the Piraha are I think interesting in the extent to which culture and language work together. One of the pro-Chomsky arguments is about the sorts of sign language which deaf children will self-develop; it would be interesting to know what sort of sign language deaf Piraha children would self-develop. My suspicion is that even without being able to hear, the cultural background would affect the grammar; and so Piraha sign language would not fit the Chomskyan universal grammar either.
In other news: I finished my summer dress, which has come out pretty well.
And yesterday I went along the south bank of the Serpentine to work, & encountered elephants! Anyone in the vicinity should go visit them.
no subject
Ooh, nice!
(And also wow for elephants!)
no subject
no subject
no subject
Lovely dress Juliet!
no subject
no subject
See http://nationalrail.co.uk/service_bulletins/currentAndFuture.html if you want to try and make more sense of it. You'll have to search for Cambridge to find the relevant sections.
no subject
no subject
It's a complete nuisance given there were a couple of things this month and next I quite fancied bringing my bike down for - they don't take bikes on rail replacement buses for planned engineering work.
no subject
It might be meaningful. It very much depends on whether you expect the distribution of errors to be related to the distribution of prior experiences. You're suggesting that the two would be similar (and moreover that errors are experiments rather than accidents), but it could be that errors are more "random" than that. But I think the only way you're going to come up with any idea of the plausible distribution of errors is through neural network modelling, as you mention.
no subject
no subject
http://www.sciencemag.org/cgi/content/full/317/5845/1674
(that link probably only accessible from a university network)
no subject
no subject
not commenting on the brain / language thing bcos I don't get it...