rowanexer wrote:So what gives? I like comprehensible input but I can't understand how they neglect other language skills, or why they think that speaking can be improved by listening. It's bringing back bad memories of the AJATT cult. Any thoughts?
Basically, it's all just another attempt to flesh out out Krashen's "Comprehensible Input" hypotheses despite decades or counter-evidence. The internet loves pseudoscience that's easy to fool yourself you understand and doesn't want to go any deeper.
I want to make it very clear that I am not against listening to things that are appropriate to your level; I'm only against "comprehensible input" as a reified concept the way Krashen pushes it.
The fact of the matter is... well you've already hit the nail on the head with one word: cult. People invest themselves so heavily in it that they refuse to critically examine evidence, and when anyone attempts to discuss inconvenient evidence, they're decried as a heretic who has been brainwashed by the "language teaching industry", which is presented as being a big conspiracy that's only interested in making money.
Le Baron wrote:I don't really see much use in listening like that at the start. My own view tallies with that 'Word Brain' fellow Sebastian Kamps. That especially at the beginning you should read what you're listening to (dialogue transcripts ideally). At least for a few rounds before you listen to it unassisted. The idea that you just listen and an unknown language, whose word separations you can't instinctively divine, unravels itself before you seems to me just voodoo rubbish. Best to make the link between the sounds and the written language from the start.
I agree with this... as might be expected from a sporadic Assimil user...!
But I would say that I'm firmly in the camp that says learning to produce is quite simply the best way to get grammar right (as might be expected from a Michel Thomas fan) and that I've had far more success with the Assimil style where the new language is similar to a language I already know (for example, having good French, passable Italian and excellent Spanish under my belt,I found Assimil Catalan excellent, and Assimil Corsican pretty good, but I struggled with Assimil Basque to the point that I never got even halfway through it).
The big thing with learning new languages is the need to pick up the sound system, and I don't think that can be reliably done without speaking. I suspect that people who do well with Assimil-like methods (heavy listening content supported by reading) are going to be unconsciously engaging in subvocalisation of what they're reading. That means they're
anticipating what they're going to hear, and the act of listening proves or disproves the hypothesis they've formed about pronunciation. But it still leaves a gap as to how learning to pronounce should get started. Assimil falls into the familiar trap of having a single chapter on pronunciation. Reading that just overloads your brain, so why doesn't Assimil follow its own logic and only tell students what they need at the time they need it?
Le Baron wrote:I know about the fossilised mistakes thing and whilst it's sometimes a bother to undo some things, I think it's very exaggerated. It's really not what happens. Does anyone learn any new skill and then make the mistakes they made at the start even though they kept on learning and practising? No.
I think it's important to recognise that this depends on the nature of the mistakes, but I do think the word "fossilised" misleads something chronic.
I knew a professional French tutor who could pronounce English phonemes perfectly, but sounded totally French when speaking casually.
'E said ziss sort of teeng habitually. We could have totally logical conversations about it, because he identified the source of the problem partially correctly (he blamed the non-native teacher, but accepted that he couldn't rule out that he wouldn't have had the same problem with a native teacher based on his classmates' errors) and he knew the way he could solve it. He argued that breaking the merger of two distinct phonemes into a single one would work, but would take over a year as it involved relearning every single word in the TH-vs-Z and the TH-vs-T group and just had no obvious reward because he was very high level and everyone understood him anyway. Given how many words he knew, it was no small feat, so I absolutely couldn't argue with that. He basically just reinforced my view that it's really important to make sure it's learned correctly from the start....
I've never met a cabinet maker who said: I made my first cabinet 6 weeks into my apprenticeship and now 20 years on all my cabinets look like those of a 6-week apprentice. It's just nonsense. I made loads of speaking errors in both German and Dutch, but they've disappeared over time. Quite quickly in fact.
The thing is, you probably were aware when your brain was letting two different phonemes "fall together" (eg "u" and "ü"), and you probably took action against it; and even if that action wasn't successful in that you might have pronounced them the same, because you were aware of it you filed it.
The problem with a "mistakes aren't important" mindset is that it encourages students to stop paying attention to their errors, so the phonemes fall into one box, which means the words all fall into one box. E.g. classmates on the OU French course who would not be able to tell the difference between E, É and È, and would ask which of the three to write, or would ask the pronounciation of words written with them, even though they were allegedly at the B1 or B2 level by that point.
Jethro Tull's frontman and flautist played for over 25 years with self-taught wrong fingering, then rapidly relearned with correct fingering in the 90s. Nothing is too ingrained to relearn. These are second languages not first languages.
Yes, and that's why I think the "fossilisation" talk is potentially dangerous, but changing the fingering is really more analogous to learning correct pronunciation when you already make the correct distinction between phonemes. There's a one-to-one mapping of "wrong fingering" to "right fingering", but if the phoneme map is wrong when you learn words, you're actually ending up having to split one phoneme into two, which means splitting all the words learned using the false phoneme into the correct box. Not easy, and as I said, I completely understood my French pal not putting the effort in and couldn't argue against it.
I still never see any convincing reasons for all the precise hour numbers these 'don't speak' gurus throw out.
Me neither. And I've seen a convincing reason for speaking early: if you don't test your hypothesis, you won't change your hypothesis. Therefore if you learn without speaking, you might well learn your false assumptions. Despite agreeing with you that there's no such thing as fossilisation, I think that learning false assumptions is just creating a heck of lot of work for yourself to unlearn the errors later.
So don't let the errors form.
Even six months might be shaky, but a year in if a person hasn't said a dicky bird they'll have a lot of catching up to do once they open their mouth, because the assumption that the vocal apparatus required to produce new sounds simply falls into place on the strength of listening is imaginary.
Seconded. I learned how to pronounce retroflex consonants be a regimen of physical stretch exercises in my tongue on the walk to work in the morning and home in the evening. Before I did it, I genuinely couldn't tell a retroflex T and a dental T apart. If nothing else, I learned to pronounce my colleagues names right if they were Indians in the Bangalore office. I couldn't have done that by listening.
[quoet]I assume their theory is that after this massive input something somewhere in your brain has primed you for making the
right sound rather than the
wrong sound, but there's no explanation of this. All this is just stretching the process out longer than it needs to be.[/quote]
Agree wholeheartedly.