Is consuming material that's beyond your "comprehensible input" level ever BETTER than understanding what you consume?

[Le Baron]

The thing is, you can reductio ad absurdum the whole thing, because doing a bunch of grammar exercises beforehand makes pretty much every text more comprehensible.

I wouldn't argue with that. Indeed there's not quite enough talk about what actually happens before anyone is even able to get near any texts/books at all. Every language I've started learning has never been just a matter of 'acquisition', because there is some sort an 'orientation' period (shorter or longer) where you just don't know what is going on, up from down or anything. This requires learning and ordering to some point before you could even have any ability to read anything. Arguably it can be more naturalistic or less so.

This is all well and good, but throughout the thread, the word level seems to be continuing the usage of the term as used be Khayyam in the thread title: "comprehensible input" level -- text at a level where it contains some new language to be acquired and not so much that you get stuck in a dictionary or grammar book. One of the concepts being explored in parts of this thread is what that level actually is.

Yes, a fair objection. These days I care so much less about ordering these sorts of things. 'Level' is a difficult continuum to measure out and 'your level' means to me simply if you can read it or not with some margin for incomprehensibility. What that margin will be seems to differ from person to person according to their tolerance. The high-tolerance learners might want to ask themselves: 'Am I looking up many words per page?' And thereby admitting the text is perhaps too complicated right now. The low-tolerance (perfectionist) learners conceding that looking up a few words every couple of pages isn't 'failure'.

'Levels' as created for class systems are also somewhat arbitrary and only made for logistics reasons. A good teacher can gauge a student's capacities, but we don't always have this luxury.

[emk]

Yeah but you’d still need some kind of learning resources to bring you up to the level where you’d be dining on Molière, Hugo, and Flaubert. Unless our hypothetical learner really just sits down at A0 and brute forces French literature classics with a dictionary and a grammar cheat sheet. Now that would be a terrifying level of commitment if someone were to pull that off.

Champollion did it with even less than that, after all! I'm in awe of the people who decipher dead languages in unfamilar scripts.

I've been under the impression that grad students in smaller dead languages sometimes don't get much more than a grammar, a glossary, and some parallel text. For example, I think most people who have learned Old Norse historically have "chained" off of a living related language or two. Languages like Latin and Greek and Middle Egyptian get high-quality courses. I get the impression that smaller dead languages are sometimes left to academic polyglots, who are expected to be able to learn from more limited resources.

[Cainntear]

It's not idiotic, it's statistics, which are themselves fuzzy.

It's idiotic to say that you always need to know 98% af the words in a text to feel comfortable reading it - even if there is a survey that says that some random group of readers needed to know 98% of the words (or word families) in some sample text to feel comfortable.

Or rather, the idiocy is in misinterpreting the evidence. Paul Nation was reacting to the claim that had been floating around associated with comprehensible input that you only needed to understand 80% of the language for it to qualify as CI. (I don't believe this was anything Krashen said, but just a guideline that emerged from teachers with no real research behind it.)

He was countering the 80% claim with statistics, demonstrating that the number of words you need to understand is extremely high. Many people unfortunately took his point wrong, and the 98% was interpreted as a proposal for improvement of CI practices rather than the disproof of the CI idea it was intended as.
Nation was only talking about vocabulary specifically, and that was because it was the easiest thing to measure objectively -- I'm pretty sure he wasn't intending to imply that grammar was irrelevant, but (and I can remember whether this was implied or explicitly stated) actually added to the difficulty of the text; the 98% vocabulary thing already relied on knowing the full grammar, so less grammar knowledge means you need even more vocab knowledge.

Krashen's CI seems like a bit of a false claim.

And it all goes back to the idea that I've said many times before: successful learners in every class do more than they're asked to do -- people who do well in target-language only environments are often actively working out rules consciously, and unavoidably doing silent translations in their heads. These are all things that are explicitly talked down in CI ideology -- "don't translate, even in your head" -- which means the poorer learners are left behind because they're not being taught to do what the strong learners do.

I think the one thing that Krashen doesn't get is that any success of his ideas doesn't prove the acquisition-learning hypothesis, but instead says that a lot of courses just teach the grammar either too quickly, too much at a time, or both. I don't really see much difference conceptually between CI and Assimil -- Assimil structures the grammar introductions around texts, and therefore paces itself better than TY-style stuff, where you're expected to learn all person conjugations in a particular tense at once, rather than one at a time. I mean... Assimil's still pretty dense stuff, but it's pretty decent for certain language pairs.

[galaxyrocker]

I've been under the impression that grad students in smaller dead languages sometimes don't get much more than a grammar, a glossary, and some parallel text. For example, I think most people who have learned Old Norse historically have "chained" off of a living related language or two. Languages like Latin and Greek and Middle Egyptian get high-quality courses. I get the impression that smaller dead languages are sometimes left to academic polyglots, who are expected to be able to learn from more limited resources.

Going a bit off-topic here, but that's certainly the way it seems to me. There's a few textbooks for Old Irish, but a lot of it is simply, "Here's grammar, here's vocab, translate this stuff to and from OI". Middle Welsh is slightly better, with two modern-style textbooks released in the past few years. Old French is mostly lots of reading with glosses, same with Old Occitan from what I could find. Indeed, most languages hosted on UT Austin's website are like that, even in the physical book forms. It's really not for the faint of heart, and takes some linguistic knowledge.

And lord knows how things like Sumerian and Akkadian are, though I do know there's some more modern books for it (and Saksrit). Let alone, maybe, Old Tamil or any number of Prakits (outside Pali which has a few decent resources). No clue what Classical Chinese is like either, at least outside the English sphere.

[Severine]

In the other thread, I argued that there's a continuum from "inspirational teacher" to "official movement" to "bureaucratically-mandated practice." By the time teachers 1, 2 and 3 are having that discussion, you've already fallen pretty far down the continuum. Here are some much better questions the teachers could be arguing about:

1. "Should we have a library of level-appropriate books and media about varied and interesting topics?"
2. "Should we strongly encourage students to read more books and watch more TV series?"
3. "Johnny has read lots of books and watched lots of TV series, and he does well on comprehension tests. But his speaking skills are far below average. Should we try to get Johnny more conversation practice? Should we make him write more? Or should we give him more books and TV series?"

These are concrete questions about how to teach students, not arguments over the difference between "acquisition" and "learning." I believe that one of Krashen's biggest contributions was arguing that (1) and (2) were good policy. I also believe that one of Krashen's biggest failures was encouraging people in the educational system to downplay question (3), or even to give Johnny bad advice.

I've worked one-on-one with several Johhny-type EFL learners who had been so convinced that input and immersion were the only way to learn that they treated their failure to achieve conversational fluency by those methods alone as evidence of their stupidity.

Teachers should be open to a variety of different methods to support and encourage students' learning, and open to adapting their approaches according to individual needs. Dogma is the enemy of flexible experimentation.

[Khayyam]

Based on my experience with German, and now with Persian, I'm of the opinion that there is likely never a time when consuming material that's beyond my "comprehensible input" level (like letting a radio show play when I can only understand 10% of the words) is a smarter approach than focusing on a much smaller bit of material and really learning it.

Maybe I've missed something here having only read your OP, but since you never mention translation it sounds like you are relying on repetition only to learn, and this is extremely ineffective ime. If I re-read material many times, after two times, with few exceptions, my improvement stops.

On the other hand, if I read it using a reading tool, even just quickly noting L1 definitions when I don't know a word/phrase, I get a lot out of it. I can then read it without the tool and understand it significantly better. Huge improvement. The same holds for any reading method where translation is allowed. Stopping and looking up each word in a paper dictionary and writing crib notes, memorizing wordlists, creating anki decks, all such methods really increase my comprehension. Repetition without translation does very little in comparison.

I'll take it a step further and say I get more out of reading a passage twice, where I only know 10% of the vocabulary, using a reading tool than I get out of reading a passage several times, where I know 90% of the vocabulary, using no translation.

Oh, no--I don't just use repetition. I mark all the words in a section that I don't know or am uncertain about, I number all the marked words, and then I type up and print out a list of translations of all those words (along with often bizarre remarks about them to jog my memory) that I will then freely refer to as I repeatedly read and listen to the section. Of course I wouldn't expect just reading and listening to the section 10 times in a row, without looking up the words, to make the words' meanings any clearer! But repeatedly reading and listening to a section while referring to one of these lists sure does the job.

Whether I also refer to a translation of the text in a language I know well...it depends. When I was learning German, I almost never did. It's similar enough to English that if you know the meanings of the individual words, you can usually make out the meaning of a sentence. With Persian, OTOH, the sentence structure is so different than what I'm used to, and there are so many idioms, that it just makes sense to freely refer to an English translation along with my list of words.

As I continue to repeat a section, I mix things up every which way--I read it in English while listening in Persian, I listen to it with total concentration with my eyes closed, I listen to it while walking or doing chores, I read it in Persian without listening, I listen to it while only referring to my list of translations--every way I can think of. Soon enough, I simply know it.

So no, I don't merely repeat sections--I'm rather more resourceful and proactive than that.

[german2k01]

Based on my experience of learning German I have come up with own language acquisition theory; people who do not post on language forums they learn the language at a lightening speed.

I have been a huge fan of Dr.Brown's listening approach(his methodology allows me to develop good pronunciation) and also a huge fan of Dr Krashen's narrow reading. Narrow reading allows me to acquire repetitive words naturally. I pick up good parts from every method, for example, I even pick up a good part of L-R system. if I really love a book and really want to cement it deeply on a subconscious level, I choose this method.

I have even attended language classes in Germany. As long as you are not dissecting the language and not doing exercises like math problems; you are probably following the right approach.

Language schools and publishers are a million dollar industry. Again you can not blame them I mean for example people who study languages as a profession they have to bring food as well on the table so dissecting the language is the best approach for them since they can evaluate the performance of each student through tests.

Language learning takes time, with time your subconscious mind will fill in the missing knowledge, the thing is. it takes time. 2% methodology 98% showing up every day.

No magical pills, unfortunately.

[leosmith]

repeatedly reading and listening to a section while referring to one of these lists sure does the job.

That makes a lot more sense. Personally, reading a passage more than twice in a row doesn't add any noticeable benefit. However, if I do something like read it once, harvest/memorize all the unknown vocab, read it again, wait a week or more, and repeat the exercise, there is improvement. I also think activities like reviewing all the grammar, listening and repeating, memorizing the whole passage, etc. would add benefit, but not as much as starting a new passage and only reading it twice. Plus those super-intense activities tend to drive me nuts. It's a matter of personal preference, but I wonder if you've tried limiting yourself to two reads?

[Khayyam]

repeatedly reading and listening to a section while referring to one of these lists sure does the job.

That makes a lot more sense. Personally, reading a passage more than twice in a row doesn't add any noticeable benefit. However, if I do something like read it once, harvest/memorize all the unknown vocab, read it again, wait a week or more, and repeat the exercise, there is improvement. I also think activities like reviewing all the grammar, listening and repeating, memorizing the whole passage, etc. would add benefit, but not as much as starting a new passage and only reading it twice. Plus those super-intense activities tend to drive me nuts. It's a matter of personal preference, but I wonder if you've tried limiting yourself to two reads?

It depends on how advanced I am in the language, and how difficult I find the material. If there are 50+ unknown words on a page, as was the case when I jumped into Genesis 1 as a Persian novice (because I'm an idiot), then I absolutely do derive benefit from repeating my reading/listening dozens of times. I was eventually able to walk around outside while listening to the narration and follow it almost perfectly (which is one way I measure success), despite the fact that my previous Persian experience consisted of nothing more than learning the alphabet and reading-listening to a Hans Christian Andersen story. We can debate whether it's wise to aim so high and grind so hard (the reasonable, not-prideful part of me kinda wishes I'd started with Assimil), but it can bring home the bacon.

If I'd read that twice and then let it percolate for a week before coming back to it, I'll bet I would have been back to square one, or close. I can see how that might work for easier material, though.

I'm sure there'll come a point where I limit myself to two reads, and then just one, but I prefer to let that happen naturally as my vocab grows from all the grinding. There's no set number of times I repeat a section; it's a matter of how many times I need to repeat it to achieve my goal of effortless comprehension of the audio while walking around outside.

As for being driven nuts by super-intense study--I do feel the same way sometimes, but I also feel that my receptive German would be far inferior to what it is if I had taken a more relaxed approach. I'm dying to bring my Persian to the same level.

[leosmith]

If I'd read that twice and then let it percolate for a week before coming back to it, I'll bet I would have been back to square one, or close. I can see how that might work for easier material, though.

I actually meant not coming back to it at all - that burden still being quite heavy for my preferences. I did something similar to what you're doing in 2005 with Japanese. I had just finished Heisig, which means I could recognize almost all kanji (Chinese characters), and I wanted to start reading so that they would stick in my mind. I found a (paper) book of short stories which I felt would cover most characters. The book was exactly what I wanted - normal Japanese on one page, and the same text but including furigana (phonetic script) above the kanji on the next. The third and sometimes fourth pages contained grammar and vocabulary explanations. Even though I had a base in spoken Japanese, and Japanese kana (phonetic scripts), this exercise was brutal. But I continued, and several months later completed the book. Then I re-read it several times, and thought I should be able to pretty much read anything Japanese. Wow, was I wrong. The next thing I read in Japanese was almost as hard as the First time I tried to read that first book.

Today, accomplishing this online is much easier, but I still wouldn't do it again. I've found reading things once or twice with the reading tool, then moving on, to be a much easier path, regardless of the difficulty level. I think since the total amount of material consumed is higher, that trumps the fact that not every little nuance is memorized. And I can control the intensity of it by putting more or less stuff in Anki.