|
 |
Should a linguist be a polyglot? | ||
Share with:
| ||
 Language Learning Forum : Philological Room |  | |
 Iversen Super Polyglot Moderator Denmark berejst.dk Joined 6489 days ago 9078 posts - 16473 votes  Speaks: Danish*, French, English, German, Italian, Spanish, Portuguese, Dutch, Swedish, Esperanto, Romanian, Catalan Studies: Afrikaans, Greek, Norwegian, Russian, Serbian, Icelandic, Latin, Irish, Lowland Scots, Indonesian, Polish, Croatian Personal Language Map | Message 17 of 29 12 October 2012 at 10:59am | IP Logged | |||||
Much better. That Lynne Murphy found a quick and dirty way to make people who mistrust monolingual linguists shut up, but she overlooked that nobody would trust a doctor who only knew how to diagnose common cold. EMK's misgivings about linguists who write about languages they don't know well enough or pretend to teach language learning without having learnt any language since their early childhood are well founded. And only the perpetrators of those practices can't see they have a problem. In contrast the remarks about Chomsky (and the references to MIT in other messages) are equally well motivated, but there you are up against most of the American (and to some extent European) linguists - at least those who can be described as theoretical linguists because the simple fact is that all the complicated models of Chomsky and his followers have had next to no impact on the authors of materials for practical language learning. Not even the online translation systems could use all that theoretical navel-gazing to anything - for instance Google Translate is based solely on large-scale statistical analysis. I do think that its translations would gain from the use of some look-up lists (proper names, fixed expressions), a few rules of thumb (keep language names, check that negations don't disappear) and maybe even a set of simple one-layer transformation rules ... but nobody in that industry would dream of writing translation software that followed one of the avatars of the complete theory of transformational grammar. Inside the ivory towers of linguistics the spirit of Chomsky and his disciples may loom large, but the land in between is shared by irreverent square-eyed statisticians and a heterogenous bunch of traditional grammarians and dictionary builders. So from that point of view all the theoretical discussions are about as relevant as the scholastic discussions about the true nature of angels. It is fair enough to assume that there is an inherent language learning mechanism in mankind, but trying to pinpoint specific universal rules is doomed because every counterexample would lead to the death of that rule, and the absence of a counterexample could be explained away by descent of and contact between languages or by some rule of thumb concerning efficiency. If there really is a simple set of rules that no human language can disregard then why is is so difficult to find a simple list with X simple rules? Those rules I have seen proposed (one here, one there) in scientific articles are generally fairly contrived and look decidedly like something left over after all simple rules have been rejected due to counterexamples. In my opinion the search for linguistic universals is a dead end. It would be more logical just to describe everything there is and then leave aside the things you haven't found yet. Edited by Iversen on 12 October 2012 at 11:17am 9 persons have voted this message useful | ||||||
| Ari Heptaglot Senior Member Norway Joined 6368 days ago 2314 posts - 5695 votes Speaks: Swedish*, English, French, Spanish, Portuguese, Mandarin, Cantonese Studies: Czech, Latin, German | Message 18 of 29 12 October 2012 at 1:03pm | IP Logged | |||||
|
How about "Asking a linguist how many languages she speaks is like asking a mechanic how many racecar medals she has won"? Or maybe better "Asking a linguist how many languages he speaks is like asking an exercise physiologist how much he can bench press"? Or maybe "like asking an aerospace engineer how many flight hours she has"?
5 persons have voted this message useful | ||||||
| Serpent Octoglot Senior Member Russian Federation serpent-849.livejour Joined 6383 days ago 9753 posts - 15779 votes 4 sounds Speaks: Russian*, English, FinnishC1, Latin, German, Italian, Spanish, Portuguese Studies: Danish, Romanian, Polish, Belarusian, Ukrainian, Croatian, Slovenian, Catalan, Czech, Galician, Dutch, Swedish | Message 19 of 29 12 October 2012 at 3:34pm | IP Logged | |||||
|
Excellent comparisons, Ari!!!
2 persons have voted this message useful | ||||||
| Medulin Tetraglot Senior Member Croatia Joined 4454 days ago 1199 posts - 2192 votes Speaks: Croatian*, English, Spanish, Portuguese Studies: Norwegian, Hindi, Nepali | Message 20 of 29 12 October 2012 at 3:38pm | IP Logged | |||||
|
Medical students learn all (major) diseases in medical school. They have all medical specialties as subjects.
But eventually (after graduating) they choose one specialty/field, and they specialize in it. Most residency training programs offer only one field: psychiatry, neurology, dermatology, radiology etc. There are some combined residency programs (for example, in the US: family medicine + psychiatry) but are relatively rare. So, if you want to be a psychiatrist, there are 6 years of medical school, where you learn all specialties. Plus internship. Plus 5 years of residency in psychiatry. 12 years of training to be a psychiatrist in the US and Europe. (Pre-med school not included) Most linguists didn't learn 20 major world languages of different families, nor they specialized in only one of them. Language professors would be much better, if they were trained like physicians, for a professor of Spanish it should look like this: 3 years of general linguistics and education theory/pedagogy 3 years of Romanic linguistics (when they would have to learn LATIN+ITALIAN+FRENCH+SPANISH+PORTUGUESE, all to the B2/C1 level) 3-5 years of specializing in Spanish language, and literature For linguists it should be like this: 2 years of general linguistics 4 years of special linguistics, they should learn: 1) one Romance language; 2) one Slavic language; 3) one Germanic language; 4) one more Indoeuropean language from another family; 5) Arabic or Hebrew (they can choose); 6) one Dravidian language; 5) one Bantu language; 6) one Polynesian or one Native American language (they can choose); 7) Ancient Greek or Ancient Latin (they can choose) ; 8) Chinese/Japanese/Korean/Thai/Vietnamese (they can choose) For 1) 2) 3) C1 level should be required for other languages B1 *** I'm amazed at Croatian linguists (armchair linguists?)...who write articles on VOSEO and USTEDEO in Central America, without even setting a foot there and interviewing/recording the local community (speaking to them in Spanish): http://www.findthatpdf.com/search-37630620-hPDF/download-doc uments-marko-kapovic-voseo-y-ustedeo.pdf.htm For a serious linguistic field work, you need to travel, not just stay in your armchair, and/or ask random interviews questions via Skype. When you ask many people in Central America and Chileans directly, -Do you use VOS?- All of them will say: NO WAY, WE DON'T. But secret recordings of colloquial speech tell us a much different story, Chileans do use VOS but they are too ashamed to acknowledge that when confronted directly: ¿Usa usted el pronombre VOS? Linguists with no field work are just using information from available literature and systematizing them. They are not serious scientists, but more like statisticians specialized in meta-analysis. Edited by Medulin on 12 October 2012 at 4:00pm 3 persons have voted this message useful | ||||||
| Serpent Octoglot Senior Member Russian Federation serpent-849.livejour Joined 6383 days ago 9753 posts - 15779 votes 4 sounds Speaks: Russian*, English, FinnishC1, Latin, German, Italian, Spanish, Portuguese Studies: Danish, Romanian, Polish, Belarusian, Ukrainian, Croatian, Slovenian, Catalan, Czech, Galician, Dutch, Swedish | Message 21 of 29 12 October 2012 at 4:23pm | IP Logged | |||||
|
Well, I have to admit that even I wouldn't want to be a linguist if it all was required. I also think two languages from the same group give you a better idea of how languages develop. Or more like one cousin of your native language and their common ancestor, unless you actually want to study Spanish, Latin AND a third Romance language.
I just wish the way languages are taught to linguists was different from teaching translators/interpreters/language teachers. Most German teachers I've had spoke only basic English and never tried to draw any parallels with it, especially in terms of vocabulary. The only times they mention English are when they're annoyed that students try to use English structures in German - which is largely THEIR fault because they don't bother to emphasize the differences. For me nowadays, linguistics (especially applied linguistics) are (is?) simply a pleasant way to get a good and respected degree and find out a few things I'd not know otherwise. It's a way to indulge my passion for several languages without making any serious "official" commitments. Of course my knowledge (even limited) of various languages makes me a better linguist. But I'd not be where I am if I had to study them all formally and do all sorts of projects and whatnot. I'd lose some or all motivation to learn things on my own. 3 persons have voted this message useful | ||||||
 emk Diglot Moderator United States Joined 5318 days ago 2615 posts - 8806 votes Speaks: English*, FrenchB2 Studies: Spanish, Ancient Egyptian Personal Language Map | Message 22 of 29 12 October 2012 at 4:46pm | IP Logged | |||||
There's actually a fairly heated debate between Chomsky, who essentially founded modern theoretical linguistics, and Peter Norvig, who works as the Director of Research at Google. Norvig has a summary of the debate on his website which I recommend highly (emphasis added):
Norvig came by his faith in statistics the hard way. Once upon a time, he loved formal, non-statistical models just as much as Chomsky, but that faith was broken on the hard rocks of engineering reality. This evolution in Norvig's thought can be seen in his two textbooks: Paradigms of AI Programming: Case Studies in Common Lisp, 1992. This is an introductory textbook covering some techniques used in the ill-fated 1980s AI boom. There's lots of logical inference, rule-based systems and all the other old-school AI paraphernalia. The best rebuttal of this approach is perhaps this essay by Clary Shirky. Norvig's first textbook is fun, but the only truly enduring chapter is the one on writing LISP compilers. Artificial Intelligence: A Modern Approach. The second edition of this book, published in 2002, is relentlessly probabilistic and statistical. It captures a trend that started in 1987 with the first workable system for face recongition and which became widely known with the first halfway-decent spam filter. Now, as Iversen points out, Google Translate currently uses a ridiculously naive statistical model of language. But as Novig admits in the quote above, everybody knows that this model eventually needs to incorporate tree structure and other higher-level issues. But so far, nobody's been able to achieve any practical gains by doing things correctly: The naive statistical techniques work very well, and the path forward is still unclear. Oh, and full disclosure: I'm on Norvig's side of this debate, for terribly prosaic reasons. I've worked for AI and robotics companies, and I've witnessed terribly smart people fail over and over again with 1980s-era AI techniques based on formal models. But even a relative dolt like me can get surprisingly good results with half-baked statistical techniques. 5 persons have voted this message useful | ||||||
| mahasiswa Pentaglot Groupie Canada Joined 4218 days ago 91 posts - 142 votes Speaks: English*, French, Spanish, German, Malay Studies: Arabic (Egyptian), Persian, Russian, Turkish, Mandarin, Hindi | Message 23 of 29 12 October 2012 at 6:32pm | IP Logged | |||||
I was translating a récit during the summer written in français français, and I accord to your siding, probabilistic translation is totally helpful. I was not aware of this particular debate of Chomsky's, but it's not surprising. My mind was made up before learning about it via your post though. In translating the short story, the few dubious portions of the original text I would run the original words and its synonyms and the target translation word and its synonyms through Google's Ngram viewer to try to align the frequencies of the original word or phrase and the one I had in mind. I admit that the final translation can't rely on this method 100% but it can be helpful in suggesting subtle differences. I am Canadian, so my French is very different from literary French from France. An excerpt from our correspondence may provide a good example of what we're talking about here:
1 person has voted this message useful | ||||||
| Iversen Super Polyglot Moderator Denmark berejst.dk Joined 6489 days ago 9078 posts - 16473 votes Speaks: Danish*, French, English, German, Italian, Spanish, Portuguese, Dutch, Swedish, Esperanto, Romanian, Catalan Studies: Afrikaans, Greek, Norwegian, Russian, Serbian, Icelandic, Latin, Irish, Lowland Scots, Indonesian, Polish, Croatian Personal Language Map | Message 24 of 29 15 October 2012 at 4:05pm | IP Logged | |||||
|
Language is a chaotic system where patterns emerge.
3 persons have voted this message useful |
|  | 
Printable version
|
| Copyright 2024 FX Micheloud - All rights reserved No part of this website may be copied by any means without my written authorization. |
