Connectionist, UG, or both?

[Heath]

I can't seem to find any really balanced overviews of the current state of linguistics in relation to which is seen as the more accurate theory - connectionist or innatist.

Some authors seem to be 100% in support of a universal grammar (to the extent that they bend over backwards to explain the exceptions), and nearly all the other authors suggest that Chomskian approaches have fallen out of favour.

The odd author here and there brings up things like "Can they go hand in hand?" with a weighing up of why they can and can't, and occasionally highlight research in favour of each to show that no-one's really sure yet; but these are usually light asides.

Does anyone know if there is a clear sway to either approach (or a combination of the two), and if there are any good books sampling research supporting or challenging either of the two?




(nb. I'm 100% for connectionism without UG; but still need to learn more).

[fluffyhamster]

Perhaps try the Harley book mentioned here: http://forums.eslcafe.com/teacher/viewt ... 3057#23057

Also, see what you can glean from searches for 'connectionism' and 'connectionist' in say Richard Hudson's Language Networks: The New Word Grammar (previewable on Google Books), though Hudson is quite "anti-UG" (or rather, pro- exploring the alternatives).

I'm guessing that "pro-UG" yet still apparently open-minded, "synthesizing" linguists like Jackendoff, and maybe Pinker, would at least try to be somewhat balanced and reasonable in any discussion they might provide of connectionism, but obviously I'm not yet familiar enough with their writings to know quite what/how much they might have said about it.

[Heath]

Thanks Fluffy.

[fluffyhamster]

'UR wellcum mai Heefclif' texted FluffyKaff.

[J.M.A.]

I can't seem to find any really balanced overviews of the current state of linguistics in relation to which is seen as the more accurate theory - connectionist or innatist.

I think scientific consensus in the area of how humans evolved "language" (really various ancient systems) is at least moving away from the idea that there was a sudden cognitive revolution 50 000 years ago. Few are defending that 50 000 year claim anymore. We are moving towards a view of language as a symbolic system and relating discussions of its emergence and development to increasing capacities for social structures of greater complexity which developed in stages over long periods of time. Thus one could postulate Homo Erectus had some type of protolanguage, for example. However there are very good authors who you will find talking about UG. In any case the discussion at this level is decidedly interdisciplinary, so you will find experts weighing in from various fields, not just linguistics. As a result, sometimes linguistics gets treated rather simplistically by those outside the field, but on the other hand it can help to put linguistics in its place and keep things in perspective. You might enjoy this book:

http://ukcatalogue.oup.com/product/9780 ... estMatches
or this
http://ukcatalogue.oup.com/product/9780 ... estMatches

[J.M.A.]

I suppose I should add that in Applied Linguistics my general impression is that Emergentism or Connectionism holds more sway than Innatism. I think it's fair to say that Rod Ellis is fairly critical of Innatism and concludes that it is losing ground in AL in his revised Study of SLA, for example. I think you can see of few of the major AL researchers with new books with emergentist or connectionist slants. My impression is that this probably holds true for the field in general, though lots of folks interested in theory have read authors like Pinker and are influenced by their ideas.

[Heath]

Urgh, that Heathcliff thing. I haven't even read that book. But, anyway, it's just Heath. Apparently my folks used to love cowboys. I don't know why I, the first, was named after some unknown TV cowboy when my younger brother got Cody from Buffalo Bill Cody... but neither of us were particularly happy to relate our names to cowboys anyway.

So overly sentimental feminine novel character or cowboy. Just great.


Oh, and thanks for that summary J.M.A. I felt like the general field should be tending towards emergent/connectionist beliefs, but only because it seems much more intuitive for me, not because I'd come across anything clear research on the matter. So I did wonder.

[fluffyhamster]

Sorry for the overly-sentimental, feminine-novel character allusion, Now-not-so-dark-brooding-moor-walking-sheep-dipping-hero-like-one. It could be worse though - you could be called Heather, for instance. All of which reminds me of that time I met a new colleague (male) called Robin. 'Hi', I said, simply. 'Ah, I can tell you're alright, because you didn't cry "Where's Batman?"' replied the Boy Wonder, little realizing that I was then (due to him having implanted the notion!) everso slightly suppressing the need to then say exactly that - that, or 'Robin's a girl's name, and fighting talk, where I come from!'.