|
 |
Revolutionary approach to learning langua | ||
Share with:
| ||
 Language Learning Forum : General discussion |  | |
| Cainntear Pentaglot Senior Member Scotland linguafrankly.blogsp Joined 6014 days ago 4399 posts - 7687 votes  Speaks: Lowland Scots, English*, French, Spanish, Scottish Gaelic Studies: Catalan, Italian, German, Irish, Welsh | Message 113 of 129 14 February 2009 at 12:49pm | IP Logged | |||||
|
Reineke,
There's arguably a big difference between an infant learning by input only (I'm sure Boydell will have been exposed to language from the very beginning, like the rest of us) and an adult learning by input only. The Boydell case is only peripherally relevant to Krashen, as the core assumption to be proven is the idea that adults can learn like children, which is better exemplified by feral children. As far as I'm aware, there is no record of a wolf-boy, wild-child or any other feral gaining L1-equivalent fluency in any language when discovered after the onset of puberty. In fact, the quality of language produced by a feral child discovered so late is, I have been led to believe, worse than that of many successful adult foreign language learners. This suggests very strongly that adult language learning is fundamentally different from infant learning. That is not to say that learning through immersive input is not possible, but just that there is very little on which to base the argument that it is optimal, or even "natural". 1 person has voted this message useful | ||||||
| slucido Bilingual Diglot Senior Member Spain https://goo.gl/126Yv Joined 6678 days ago 1296 posts - 1781 votes 4 sounds Speaks: Spanish*, Catalan* Studies: English | Message 114 of 129 14 February 2009 at 1:42pm | IP Logged | |||||
|
As I said before, I don't believe that Krashen tale.
By the way, how can we assess what are people doing in his mind? Two people can receive the same input, listening or reading, without any output, but we don't know if one of them practice a lot inside his mind. Sure the results will be different. Maybe one of them practices sentences in his head. Maybe he experiments with words, mentally reviewing, trying new combinations, comparing, imagining different situations, talking to himself in silence and never ever writing or talking aloud. Apparently this person is an example of "input hypothesis" but he (or she) "practices a lot" (output) from the very beginning, but only using his mind. I think this can be a very big "confusion factor" in any scientific research about this subject. Edited by slucido on 14 February 2009 at 1:43pm 1 person has voted this message useful | ||||||
| reineke Senior Member United States https://learnalangua Joined 6450 days ago 851 posts - 1008 votes Studies: German | Message 115 of 129 14 February 2009 at 3:05pm | IP Logged | |||||
Boydell's case is very relevant because the core idea is that humans CAN learn from comprehensible input only and be capable of "refined" native output. Feral children have been deprived of all normal human interaction, including comprehensible input. They have serious problems "acquiring" a language AND/or learning it through all other means. They have problems socializing. Such abuse often results in severe mental retardation and is therefore problematic as a proof for language acquisition theories. A feral child does not compare well with an adult language learner. Most feral children are not happy little tarzan boys but poor creatures that have been chained to a wall or locked up for years, severely malnourished, abused and deprived of light. Krashen on comprehensible output: http://www.sdkrashen.com/articles/comprehensible_output/inde x.html Edited by reineke on 14 February 2009 at 3:11pm 1 person has voted this message useful | ||||||
| Minder Newbie United States Joined 5765 days ago 11 posts - 11 votes Speaks: English* Studies: Spanish, Japanese | Message 116 of 129 14 February 2009 at 10:35pm | IP Logged | |||||
|
I just read this thread and finally decided to stop lurking and start posting. I think that appropriate language learning follows in the order of listening -> reading -> writing/speaking. This is why. The idea of reinforcement states that anything you do is reinforced. This means that if you begin reading before you know the way a language flows and sounds you are reinforcing poor pronunciation. If you attempt to write or speak your own thoughts before you know how a native would say it you are not expressing yourself in the manner a native would. For this reason I think that listening to the language initially is very important.
I think that the initial listening doesn't have to be comprehensible. All you have to do is focus on the flow of the sentences as a whole and the different phonemes present in each word. I do agree with the article that this is a valuable step that at the least familiarizes you with the way that the language sounds. I think that following this step listening + reading (simultaneously) is important. This continues to make sure that the sounds you are internalizing are correct. Now it is important that these sentences are comprehensible FOR FASTEST ACQUISITION. By comprehensible I mean that with whatever materials you have at hand, whether those be context, dictionary, pictures, etc. you can "understand" what is trying to be expressed. If the material is not understandable even with context and dictionary support then the material is too difficult. I think the argument is not that incomprehensible input provides no benefits, but that it provides veerrryyyy slow progress compared to comprehensible input. i+2 is not only 2 times as difficult as i+1. The difficulty progresses in a faster than linear fashion, meaning that pushing right next to your current level will yield the fastest and easiest results. When you read on your own you are subvocalizing and ingraining sentence patterns into your "brain language circuitry" or whatever. This means that yes in a fashion you are outputting! The difference from traditional output, however, is that there is no creating on your part (which could easily lead to incorrect ouput which would be reinforced). Subvocalizing or even reading it out loud, allows you to practice your ability to mechanically produce sounds and sentences while maintaining perfect grammar and native like style. I entirely agree with the fact that you must practice speaking to be able to speak, but I do think that individual creation of unique sentences should wait so that aspects of your L1 do not creep into and influence how you articulate concepts in your L2. Anyway, thats my personal opinion. I think that incomprehensible input can prime you for future learning and I think that there are benefits to it, but they are not immediately noticeable and pale in comparison to MORE COMPREHENSIBLE input. We don't need to make a distinction between incomprehensible and comprehensible, rather, we need to make the distinction between more comprehensible and less comprehensible. I think so long as they are both beyond your current limit the more comprehensible the material you have the more you will learn from it. 1 person has voted this message useful | ||||||
| Cainntear Pentaglot Senior Member Scotland linguafrankly.blogsp Joined 6014 days ago 4399 posts - 7687 votes Speaks: Lowland Scots, English*, French, Spanish, Scottish Gaelic Studies: Catalan, Italian, German, Irish, Welsh | Message 117 of 129 15 February 2009 at 5:07am | IP Logged | |||||
The problem is that this flies in the face of current academic thinking. Sulzberger's experiment, as summarised earlier by Raчraч Ŋuɲa and described clearly in the radio interview in his link, does nothing to disprove current thinking -- in fact it can be used to support the prevailing theory: that the adult learner cannot hear the meaningful sounds of a foreign language that don't exist in his native language. There is nothing in the experimental design that suggests any way of changing that. Meanwhile, neuroscience is opening up a whole new area for research: mirror neuron theory. Basically, this says that when we see or hear something, we experience it as though we were doing it ourselves. When applied to language, this would suggest that when we hear a phoneme, we experience it as though we did it ourselves -- but if we cannot produce the phoneme ourselves, what framework have we got against which to "mirror" the input? Let's consider the guitar example, as mentioned by a couple of posters earlier. Most people are capable of humming the main riff from Smoke on the Water or whistling the bass line from Stand By Me, but put a guitar or bass in their hands and they haven't got a clue -- we don't have the neural model of how to articulate the sounds on a guitar. Even the beginner to intermediate guitarist has difficulty doing this. He has to have it written out or he works it out note by note. But the expert musician can listen once to a riff like that, pick up his instrument and just play, because when he hears the tune, he immediately relates it to the method of articulation -- his mirror neurons mean he understands the tune as a guitar tune. So what about the beginner and the non-player? Well, they fall back on a method of articulation they're familiar with: humming, whistling, or la-la-la. One of the reasons for a musician reaching a plateau may well be that this model of articulation is so well refined that there is no impetus for him to develop the new model. He then never succeeds in learning to play by ear. What does this mean for the language learner? Well, if you start listening to French and you relate it to English, it's possible you'll never hear the distinction between AN, EN and ON, or relating Hindi to English you'll never hear the distinction between t, th, d, dh, T, TH, D and DH. So while many people claim that "listen first" is the way to avoid pronunciation errors, I feel that conscious early focus on (approximately) correct articulation is not only the best way to develop a good accent, but also the best way to improve comprehension.
Well, I'll say this. Have you ever finished someone else's Yes, I'm sure you have. How can you do that if you aren't constantly creating language as you receive it? So it can be theorised that when we receive input, we understand it by recreating the process by which it was initially composed, and saying "what would I mean if I said that?" The problem with i+1 is that it includes things that you grammatically cannot say, so there is no answer to the question "what would I mean if I said that?" because you wouldn't say it. (Note, I have no problem with i+1 when we're talking about vocabulary items -- we are all well used to guessing and inferring the meaning of new words like "blog", "podcast" etc or new word bundles like "log in" and "connect to... network" -- my problem is with the suggestion that we can pick up the fundamental underlying grammar this way.) So yeah, I think all language comes first and foremost from production, and that the slow initial stages of a child's first language development are because there is no way of guiding production before the development of native language. However, once the native language is developed, we can "cue" development by properly designed syllabuses (but sadly most current syllabuses are very poorly designed IMO). Edited by Cainntear on 15 February 2009 at 5:08am 1 person has voted this message useful | ||||||
| Cainntear Pentaglot Senior Member Scotland linguafrankly.blogsp Joined 6014 days ago 4399 posts - 7687 votes Speaks: Lowland Scots, English*, French, Spanish, Scottish Gaelic Studies: Catalan, Italian, German, Irish, Welsh | Message 118 of 129 15 February 2009 at 7:05am | IP Logged | |||||
|
Someone mentioned an interesting case in the thread How fast can one learn a new language?: David Tammet, a mathematical and linguistic savant from England.
When a documentary was made about Tammet (entitled The Boy with the Incredible Brain), they focused mostly on his mathematical ability. As it was described in the program, Tammet does maths without conscious recourse to rules -- in cognitive terms he does it procedurally, not declaritively. That is to say that he doesn't "work it out" through a sequence of carefully ordered and consciously controlled steps, but his brain carries out the process almost in a "black box" style, with the individual steps hidden from scrutiny. So when he is multiplying, say 132 by 465, he doesn't consciously break it up into 100*465 + 30*465 + 2*465, but he just... well... does it. Now, at the end of the program, they sent him to Iceland where he had a week to study before going on a live talk show. They didn't give a great deal of detail on how he studied, but we do know he had a personal tutor, was in the country, and was reading from some a variety of materials, including some pretty old-school stuff. If his mathematical and linguistic skills have a common mechanism, as I suspect they would, then he must surely be internalising languages procedurally, even if he is starting from a declarative model. This individual case stands as further evidence of the incorrectness about the hard distinction Krashen draws between "learning" and "acquisition". Interestingly enough, once I started getting good at language, I started getting better with numbers. I became semi-aware during my study of Spanish and Gaelic that I was over-relying on my declaritive/memorised model of language and I was suppressing my procedural/internalised model. The right answer would run out of my procedurally memory and I would hold it back while I worked through it declaritively, and I wouldn't say anything until I knew why it was right. Once I became aware I was doing this, I managed to snap myself out of it, and started just trusting that the first answer I came up with was probably right. That was when I started speaking fluently. I was doing some work with numbers in the office (checking off serial numbers, checking stock numbers, prices etc) and I experienced the same feeling. I was trying to memorise the serial numbers and then actively recall them, but the back of my brain just recognised them and internalised them and when I trusted myself, I could just type them in without consciously knowing the number. The full answer to the sums would be in one part of my head while my declarative system was still at the first digit. So Krashen's stuff about overmonitoring is really correct, but as with many people, but he takes it too far when he suggests that the monitor is entirely separate from acquisition (which I call true learning but he confusingly calls acquisition). Now, when I talked about speaking fluently above, I think I need to clarify something. There is a hard distinction drawn by many teachers and researchers between "fluency" and "accuracy". They suggest that you can have a high "fluency" and a low "accuracy" -- I dispute this, and suggest that unbroken expression cannot usefully be called fluency. I suggest that very low accuracy indicates that the speaker may well be trying to employ the declarative model but recognising that this is too slow, avoids processing anything. His utterance is then a "brain dump" of the starting phase of the process, or one step in. IE, a list of words and the prototypes of the rules that they should be processed with. This would explain common errors like unconjugated infinitives. I would argue that the concept of fluency is only useful if it refers to the state of speaking from a procedural model. This doesn't preclude low accuracy, certainly, but does not support the extreme low levels of accuracy accepted by some teachers today. It also makes it impossible to accurately talk about fluency in the early stages of learning as we cannot detect a difference between fluent and non-fluent speech until we get to a stage where the difference in processing times between the declaritive and procedural models are large enough to provide an unambiguous indication of which model is being employed. Unless we use a Michel Thomas-like method, where we build the model rapidly and leave the content of the vocabulary until later. 1 person has voted this message useful | ||||||
| Cainntear Pentaglot Senior Member Scotland linguafrankly.blogsp Joined 6014 days ago 4399 posts - 7687 votes Speaks: Lowland Scots, English*, French, Spanish, Scottish Gaelic Studies: Catalan, Italian, German, Irish, Welsh | Message 120 of 129 15 February 2009 at 9:12am | IP Logged | |||||
How so? 1 person has 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. |
