Why Everyone Should Study Linguistics

[fluffyhamster]

I am familiar with the arguments for innateness (as are probably most regulars here). One thing that always gets me right away is that 'provided there is a minimum'. Anyway, the empiricist counterarguments are in my view just as if not more compelling (see most notably Sampson), especially when the focus shifts from generalities to specifics: nativists are often plain wrong when they proceed to theorize (as they must surely feel compelled to, if they are to remain in the business of linguistics as they practise it) about what language should be like on the basis of innateness (versus how it actually is).

It must seem unfair that empiricists have an easier time of it (especially of late), but then, they have chosen the simpler theory to explain what should be the same set of facts.

[revel]

Hey all.

Oh boy, I think I'm going to get myself into some hot water here, as I really don't know who all these experts are besides the "Ch" man, and have only browsed his ideas and that several years ago.

But Stephen took the words right out of my fingertips when he pointed out that drilling in L 2 acquisition is important, and he even won more of my support by mentioning piano playing.

I can not imagine any pianist who has not spent hours drilling scales. The physical and psychological exercise involved in such work must certainly contribute to the final interpretation of a musical piece on the keyboard as a form of communication. I can not imagine any L2 learner who has not spent times learning basic schemes and structures through repetition of such, to later be able to take advantage of them in the substitution necessary to communicate what they need and want despite the obviously simple nature of such exercise. Just the physical work of wrapping one's mouth around strings of sounds is beneficial and can only contribute to later interpretation of the language.

I can't go on much more about this because I am out of my water here, been so long since I've read a linguistic text, not very useful to me right now. But I had to add my $.02 as I find that I hardly ever agree with Stephen, or at least with his way of writing (and I suppose the feeling is somewhat mutual). Having learned to speak Spanish fluently thanks to personal drilling and learning of basic patterns, I know that what he says is right, at least in my case. May not work for everyone, but we are not talking about Taliban children memorizing the Koran here, we are talking about turning musical scales into interpretative sonnets.

peace,
revel.

[woodcutter]

I like repeating too. What I'm trying to say is that you don't need the innate theory to basically ignore how kids learn. If I had run around repeating at the top of my voice, asking people to read me simple texts, asking people what was inside their bags all day long for 5 years then I would probably be excellent at Korean. However, I have to use English, and that activity would land me in the loony bin.

It is impossible to learn as children do - time constraints, social constraints, L1 interference, lack of friendly cooperation from others.

[fluffyhamster]

Cor, I've mentioned Sampson several times now on Dave's (on these threads that touch on Chomsky and UG), and although I still think that his books (especially The 'Language Instinct' Debate) are well-written and well-argued enough to be worth a read, I really do wish he could restrain himself sometimes with what he posts online:
http://www.grsampson.net/AMiu.html

(Time now for the subthread title: Why Would Anyone Read This, or Twenty Minutes of Your Life That You'll Never Get Back)

I now image Sampson as some aspirational yob on horseback galloping off after foxes, innatists and immigrants (he seems to think that mentioning dubious racial research would be a good way to open up a debate on immigration into the UK: http://www.grsampson.net/CDissident.html ).

What he needs to do is calm down and stick to the linguistics, or make his case as well if not better online as he does in his books (which will probably remain an impossible task, unless he somehow makes them available for free); as it is, I must admit that he comes across a bit of a crank in his online missives at least (that is, his linguistics starts to seem not so much empirically as racially motivated - or is it just Sapir-Whorf in so many words?).

So I guess that where I want to disagree with Chomskyan linguistics, I will in future seek to find other authors to quote than Sampson (e.g. Pullum) to illustrate the same point or points (although few cover quite so many or in as sustained or sometimes even succinct a fashion as Sampson does).

Anyway, would you say that linguistics is actually not the best arena in which to settle ("settle"? Argue? Establish "revolutionary" but potentially unscientific research paradigms in etc) the nature versus nuture debate?

[woodcutter]

I didn't read the article yet, but does Sampson make the points I have made here in the book, and say that the special situation young children have is a superb learning environment basically impossible to replicate (whether or not you have "motherese" exactly")?

I have left out a big one here too - the constant mental repetition and use of simple elements in target language that comes naturally to children

[fluffyhamster]

Well, Sampson doesn't quite say that 'the special situation young children have is a superb learning environment basically impossible to replicate', but he does devote the largest portion of TLID's second chapter ('The Original Arguments for a Language Instinct' i.e. those forwarded mainly by Chomsky from the start through to 1980, certainly prior to the arrival of Pinker and the rest of the "New Wave") to debunking the strand of the nativist argument that he (Sampson) dubs '(Response to the argument from) poverty of data' (i.e. the poverty of the stimulus).

It appears that e.g. Hornstein and Lightfoot make selective use of Newport & the Gleitman's findings regarding motherese, whilst Virginia Valian views ellipsis/informality in motherese as constituting ungrammaticality (and nativists then often veer towards arguing that lack of irrefutably "negative" i.e. non-occuring evidence is the issue when real data disproves their theories, but if this were so, scientific theorizing and discovery of any kind would be impossible in a world that provides only positive, occuring evidence e.g. we do not see heavy inert bodies float upwards).

Chomsky, as is well known, seems completely uninterested in establishing facts as opposed to consulting then quoting from his intuition, but has certainly been proved wrong with regard to 'S-before-MV questions' (see links below) - the data available to the child is richer than Chomsky supposes; then, to actually quote rather than paraphrase Sampson, 'But the detailed facts about the child's data scarcely matter, since the argument is in any case logically untenable. The obvious reply to anyone who asserts that a language has certain properties, no evidence of which is present in the data available to an individual learning the language, is "How do you know?"...The fact that Chomsky and his English-speaking readers all agree that questions formed according to hypothesis 1 [operate on the first finite verb, versus 2, operate on the finite verb of the main clause] are ungrammatical...proves that evidence refuting that hypothesis is encountered in the life of any English speaker. Clearly, this point is independent of the deficiencies which Chomsky or other linguists believe characterize the language learner's data. Oddly, this self-defeating aspect of the argument from the poverty of the data has not been discussed by Chomsky, so far as I have read.' (I've quoted Sampson on question structure there at some length because it 'is not just an example of the alleged poverty of linguistic data. It is overwhelmingly the example that nativists use to make their general idea more concrete and plausible'. (Readers wondering how many features of language supposedly become known to "learners" independently of any evidence might be surprised that, contrary to e.g. Neil Smith's claims of there being 'innumerable such examples' in generative textbooks, there are in fact just four doing the rounds - see the Pullum & Scholz paper below for further details - with the question structure example the ace of the pack)).

Sampson goes on to detail Stich & Hornstein's reasoning defending against this objection - to start with, they claim that adult linguists can introspect and elicit judgements beyond the data in ways that children can't - but he points out its circularity, for speakers' beliefs about their language surely ultimately derive from their experience of real examples.

And all this presupposes that a linear (process) versus constituent (completed product to be dissected and analyzed after the fact) view of especially speech is to be frowned at (see pp 20-21 of David Brazil's A Grammar of Speech, available on Google Book Search.

Actually, I've just seen that a fair bit of Sampson's TLID is also available! Specifically, most of pages 43-49).

Edit: There's also an abridged version of Sampson's TLID available on his website: http://www.grsampson.net/ATin.html

Anyway, some relevant links:
http://forums.eslcafe.com/teacher/viewt ... 2541#32541
http://lingphil080.dlp.mit.edu/~norvin/ ... Scholz.pdf
http://www.dcs.kcl.ac.uk/staff/lappin/s ... otes2.html
(second and third links above are just some of those produced by a Google search for 'Pullum Scholz'; second is quoted by Sampson)

And whilst I'm putting the boot in, how about these:
http://itre.cis.upenn.edu/~myl/language ... 00025.html
http://itre.cis.upenn.edu/~myl/language ... 00485.html

I'm pretty sure that Fernando Pereira (mentioned in both the LL posts above) also found his way into John Coleman's Introducing Speech and Language Processing (CUP 2005).

[woodcutter]

Mmmm, he does come across a bit cranky. Why post that first article? Why write it? You can easily point out the logical flaws in what people say and do about religion, culture or race, but it usually isn't wise, and probably won't help anyone.

Bashing Chomsky is harmless and very popular fun though, (even if he is the world's most prominent Jewish thinker). "Poverty of the stimulus" arguments are beyond the reach of most people, so I don't understand why people don't more often explore the kind of avenues I have mentioned in this thread, rather than blithely accepting the magic abilities of kids. Hasn't anyone else ever taught them? They are bleedin' rubbish. They spend so long learning "Red" and "Dog" that they develop a decent accent along the way though, avoiding quickly running complex stuff through a dodgy phonemic system like a generally better learner will.

[fluffyhamster]

I posted the link to the article to show that even though I appear to criticize Chomsky all the time and never Samspon, I am not a completely uncritical reader or supporter when it comes to anything written by the latter author. (Wow I like stating the obvious sometimes). Or perhaps you were asking Sampson. (Probably!).

Cranky, not necessarily a crank. Sounds better eh, a bit more respectful (certainly, less disrespectful), for those who've read what seems to be his more considered, less personal stuff.

Hmm, Chomsky-bashing. It's not so much that I find it great fun in itself, it's just that I really don't think that TG > UG has ever held out the prospect of being much fun (or even that productive a use of time) for most "users".

I think that young children take on board a lot more than we think they are (it's just that they lack the social awareness to explicitly agree with or otherwise comment upon every last word that we say to them, beyond the functional implications/"meaning" of the form - certainly, it takes a while for any them them to transform into possible grammar mavens).

[woodcutter]

I think they take very little on board. I teach them shape names one week, and by the next week it is as if it never happened.

Get a group of good committed linguists ready to crazily throw off all inhibition, put them in a kindergarten environment with some helpful mummy figures 24 hours a day for a couple of years, and they will kick kiddie butt, and get pretty fluent in a target language. Now that would be an experiment.

[fluffyhamster]

Sorry woody, I'm not sure if you are talking about your experiences of teaching non-native kids in a classroom, or about your own children (I'm assuming the former, but 'teach' is a bit ambiguous - it could mean 'told' or 'mentioned', or you could even be supplementing your own kids schooling by sitting them down informally at home, you slavedriver). As for me, I was talking about native immersion at the end of my last post, but I'm always up for a discussion of non-native "immersion", as the link below shows (although that can get a bit too country-specific, and might be best reserved for the International forums - again, follow the link).
http://forums.eslcafe.com/job/viewtopic.php?t=53345&

[woodcutter]

It's all the same to me really. Young kids are not especially good language learners, brainwise. Put in any equal situation with most others and they will do worse. However, due to their personalities, social situation, time available etc their situation is usually much, much better. The reason that immersion language at a young age is worthwhile is that kids can learn language, and they can't do much else, so why not learn it, painfully slowly, and get a decent accent.

The trouble is you can't make the situation equal. That's why Sampson and a few of us, perhaps, should get into a romper suit for a couple of years and show the kiddies just how fast pops can get the foreign language down given their special situation, and shut Pinker etc up.

[fluffyhamster]

Sure, if adults put their mind to it, they can learn additional languages to a very high level and thereby dent aspects of the nativist case. But the nativists would still counter how amazing it is that children with no apparent mind unfailingly learn to a much higher level (whilst as you say overlooking the priveleged, supportive and rich environment in which children can "dedicate" themselves to learning their mother tongue(s), all the while getting plenty of things wrong in that gradual process).

[woodcutter]

Look round the next adult class you teach. Think what is going on in those heads. Ponder the great firewall of native language and self-defeating attitude that you face. No wonder children do much better. It is only remarkable if you don't think hard about it, so the nativists ought to do so.

The only aspect not "dented" in the nativist case by good adult learners is accent. And why is it that our foreign language accents are poor? Cos our brains is dull? Or because we speak good English, and our accents is Englishy?

Apparently elderly deaf learners have an "accent" in sign language, but I'd like to know what kind, because for the rest of us the type we have will be quite clear. It's a shame that all the nativists best evidence lies in places that are not easy to examine.

[Stephen Jones]

The only aspect not "dented" in the nativist case by good adult learners is accent

How does a successful adult learner, such as Conrad for example, make the least dent in the nativist case? Your argument lacks all logic.

[fluffyhamster]

...with the question structure example the ace of the pack.

Maybe I should've said 'joker' rather than 'ace'.