|
 |
Krashen & some failures for Massive Input | ||
Share with:
| ||
 Language Learning Forum : Learning Techniques, Methods & Strategies |  | |
| patrickwilken Senior Member Germany radiant-flux.net Joined 4536 days ago 1546 posts - 3200 votes  Studies: German | Message 1 of 67 26 October 2014 at 1:22pm | IP Logged | |||||||||||
|
What I did and didn't learn from Massive Input after 2.5 years
Since I am in the process of changing my learning style, I thought it might be of interest to share some observations of what I have learnt and not learnt from a process of using massive input to learn German. Caveat: The process of Massive Input was not pure. At the start of learning German two-and-a-half years ago I was about A0+ or A1-. I had lived on and off in Germany for a few years, but always worked in English speaking jobs and had made some attempts to learn German via language schools. I am also married to a native German speaker, but we spoke English for a long time (we lived on and off for years in the US and UK as well) and my wife's English is a strong C2. We did make attempts on and off to speak German but we'd invariably switch back to English as it was faster and more comfortable. When I started learning German two-and-a-half years ago I re-read and partly worked through an A1 and half an A2 course book, before switching purely to native materials. I also read through a short German grammar in the first couple of months to get a more full overview of how cases and tenses etc worked, For exactly one year I used Anki to drill words (L1>L2 and L2>L1). About a third of the cards were constructed from sentences I was trying to memorize. Over the last two-and-a-half years I have read almost 5 million words (nearly 20000 pages) of German, mostly intensively (i.e., ebooks with pop-up dictionary),and watched about 1000 hours of movies and TV shows intensively (i.e., without subtitles). I stopped using Anki after exactly one year, at which point I had read 2273 pages and watched 163 films: So about 90% of what I have read, and about 75% of what I listened to, has occurred since I stopped using SRS, and have been doing a purely input-based approach to learning. Over the last two months I have switched to talking almost solely in German to my wife (who only minimally corrects my speech). In the last week have started doing a daily writing exercise (the first time I have written in 2.5 years), and will probably hire a tutor fairly soon to work on my speech, so it's probably a good time to assess how far Massive Input has brought me. How well can I decode input? In general I have a good understanding of spoken German. For instance I saw the latest John Le Carre film, 'A most wanted man', staring the late great Philip Hoffman in the cinema yesterday and estimate I understood about 99% of speech. I am certainly in the 95% range for most movies/TV shows now. I have recently started to listen to Deutschland Funk (something like the PBS/BBC) that has very clear German, and there my understanding is less, though I can follow stories OK. This is probably because the vocabulary is a bit harder/different from what I am used to (and of course there are no pictures to help give context). I still find German undubbed films sometimes a bit different. I assume that's because of the use of more regional words, and sometimes some what poorer quality sound (this is particularly obvious in some TV shows). Reading is of course harder, and there are plenty of texts that I would still struggle with, but in general I can read lots of things well. I prefer reading extensively, as I feel I pick up the meanings of words much faster, but I read intensively too. At the moment I am reading Steven Baxter's 'The Time Ships' on the Kindle, for which I know at least 95%-98%. I also read articles intensively in the newspaper die Zeit on the weekend (roughly equivalent to the New York Times or the Guardian in English). For instance, this weekend I have read articles on the remarkable new CRISPR gene-slicing technology, a very photo-project in Israel, and the egg-freezing offer from women employees by Apple and Facebook. My reading of these articles is slower, and there are certainly gaps in my vocabulary, but I understand most of what I read. So I am very satisfied with my progress to date in terms of my understanding of German. I have the strong sense that the more I read and listen the easier these tasks will become, it could be quite a while before I understand as much as I like, but if just keep doing what I am doing I will eventually be able to read whatever I like and understand pretty much everything I read. How well can I make output? I can talk OK, I recently had an hour long conversation with a German language teacher I hadn't seen for about two years, and she thought my spoken German was good, and if pressed I think she would say a was B2/B2+ for spoken German (though see below for why this isn't really true). At this point I can talk about a wide range of topics (e.g., health risks of Ebola, the Green politics in Germany, childcare practices etc). I am not as fluent as I would like, and there are gaps in my vocabulary, but overall I can communicate about pretty much any topic (albeit some worse than others, and not all fluently). However, my knowledge of German grammar is quite uneven, which is perhaps the most interesting result of this experiment. On the plus side I have picked up a lot of German grammar from simply reading/listening, for instance: How to form various tenses (simple perfect, past perfect, future, passive etc); whether to use haben or sein for a particular verb in the past perfect; which verbs take the dative; which verbs are irregular and how to conjugate them; a general sense of how a German sentence is put together, etc. I also know (roughly) 90% of the genders of words. In general I have a good intuitive understanding of how the language works, and so reading a grammar book is like reading about a well known place. However, where I fall down completely is declension. Even something so basic as "Der Mann hat einen Hund" (the man had a dog) sounds the same as "Der Mann hat ein Hund" to me. Why I have a completely failed to learn declensions? A simple answer might be simply idiosyncratic: Australians tend to 'swallow' the last syllable of words, so perhaps I am just unconsciously ignoring the declensions as I read? However, I think a better explanation may come from studies masking to study unconscious language processing. It's well known that if you briefly present a word/image followed by a random pattern, people will only report seeing the random pattern. If you subsequently measure how quickly someone responds to a related vs unrelated image/word you can get a sense of how far the masked word has been processed. These studies have shown that even the semantics of the word the word can be processed unconsciously and automatically - so if you show "dog" people respond faster to words like "cat", "walk", "leash", "pet" as compared to semantically unrelated "god", "highrise", "freeway" etc. What researchers have not been able to show (at least using masking) is the automatic semantic processing of word phrases. It appears that single words prime fine, but groups of words are treated as individual units in the brain, not as semantic wholes (this must happen at some point, obviously, but later than point probed by priming). So perhaps (esp. because I am coming at this as a native English speaker) I am just not sensitive to the way the various parts of a German sentence come together to create the correct declination, and no matter how much input my brain never seems to learn how to put the various parts of sentence correctly together (and importantly this is not necessary to extract out information from the sentence). So even for a simple sentence like: "Der Mann geht mit dem schönen Hund spazieren" (the man goes for a walk with the beautiful dog), you need to know that (1) the preposition 'mit' takes the dative, (2) 'Hund' is masculine singular, (3) that the adjective is proceeded by a definite rather than indefinite article, and that together these make "dem schönen" (rather than "der/die/das/den schöne/schöner/schönes/schön" Hund). Perhaps it also important to emphasis that I am not mentally translating back into English as I read. I just don't think my brain treats declinations as that important. In the above sentence, whatever the declination, I know that the man went for a walk with a beautiful dog. There is no "useful" information missing - though this is not necessarily the case for more complex constructions. And of course, perhaps I do have some vague sense of declensions. I can't say with certainty that if you held a gun to my head that I wouldn't guess above chance if presented with differently correctly/incorrectly declined phrases (I wouldn't say "Der Mann hat dem Hund" for instance - but more likely because I associate independently of case "der" not "dem" with Hund), but it seems obvious from my current progress that I would have to read many many more tens-of-thousands of pages before I could hope (if ever) to learn this really critical ubiquitous aspect of German grammar. Conclusions I certainly wouldn't say Massive Input has been a failure. I have learnt a lot of vocabulary and a lot of grammar, and it should be straightforward to learn declination via writing, and personally I prefer to do this later, after I have developed a reasonable vocabulary and better sense of grammar overall. However, it does prove in my mind that the strong form of Krashen's hypothesis, that input via reading alone is sufficient to train up the brain to make correct language output, is false. I would be really interested to hear what other people think might be a reasonable explanation for the gap in my grammar learning, and also whether they have experienced anything similar when learning a language predominantly via reading/listening. Edited by patrickwilken on 26 October 2014 at 5:31pm 12 persons have voted this message useful | ||||||||||||
 emk Diglot Moderator United States Joined 5535 days ago 2615 posts - 8806 votes Speaks: English*, FrenchB2 Studies: Spanish, Ancient Egyptian Personal Language Map | Message 2 of 67 26 October 2014 at 2:33pm | IP Logged | |||||||||||
|
Congratulations on using more German at home! Speaking at home made a huge difference for me very quickly.
Your experiences with pure input also match mine: It's possible to acquire a pretty decent understanding while glossing over a lot of grammatical details. I find that two things help a lot with that: 1. Training myself to notice some of the little details I was ignoring. With just a little bit of study, these details will start "popping out" of the input. 2. Once my mental model of French comprehension started to fully incorporate those little details, my own errors started sounding very jarring. I have several guesses about the grammar problems: 1. It takes kids a long time to internalize grammar, too. 2. Adult brains may be desensitized to finer details of phonetics and grammar. In my experience, this can largely be overcome by listening more carefully and noticing what I'm actually hearing. 3. Output doesn't arise as naturally from input as Krashen thinks. Rather, the speaker compares their own output to their "model" of the language, and adjusts their output to match. I agree with Krashen that vast quantities of comprehensible input are essential. But I personally know people who provide very tidy counter-examples to a "pure input" hypothesis. For example, one of my friends heard a language from birth until age 20, and she can watch native police shows comfortably, but she can't actually speak—and she wishes she could. My guess is that she could learn to speak very quickly, because she would basically just need to say lots of stuff, and notice whether or not it sounded "weird" to her ears. 5 persons have voted this message useful | ||||||||||||
| Serpent Octoglot Senior Member Russian Federation serpent-849.livejour Joined 6600 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 3 of 67 26 October 2014 at 2:41pm | IP Logged | |||||||||||
|
This stuff doesn't get any easier if you learn it with grammar books as a beginner/low intermediate learner. At least it wasn't for me. It only started to click after input, despite my familiarity with cases from my native Russian and later also Finnish and Latin. (yes, if it helps, I really found the Finnish 15 cases much easier!)
I also tend to think that the Super Challenge etc is an overkill, frankly speaking. As in, I strongly suspect that you could get 50% or 75% less input and lose only a very small fraction of your current progress. I'm not criticizing your strategy - being in L2 country and learning only one language, this certainly makes sense. I just think you shouldn't feel discouraged if you see someone who gets less input and yet has better language skills. This is not necessarily due to doing more grammar etc - you're simply far closer to the point of saturation than most learners are. In Spanish and Italian (where my learning has been more input-based) I use SRS and especially cloze deletion to deal with this sort of thing. You could also try making your own exercises (shorter term cloze deletion without or prior to SRS) and playing with the sentences. As for Krashen, heritage learners etc, I tend to think the key issue is getting enough input, both listening and reading. Many heritage learners don't read. Speaking of that, maybe varying your reading strategy can help? Popup dictionaries are great for the vocabulary, but extensive reading is better for absorbing the grammar, as your brain can't focus on everything at the same time. Edited by Serpent on 26 October 2014 at 2:48pm 4 persons have voted this message useful | ||||||||||||
| patrickwilken Senior Member Germany radiant-flux.net Joined 4536 days ago 1546 posts - 3200 votes Studies: German | Message 4 of 67 26 October 2014 at 2:44pm | IP Logged | |||||||||||
I guess this will become obvious over the next few months as I start writing more. I certainly hope the errors I make start sounding obvious! :)
I don't buy time as a big factor here. I know kids to a lot of input, but I have made almost ZERO progress after seeing somewhere between a half-a-million to a million declinations. Something else is going on here to stop me learning this. Input is simply not enough.
I don't want to sound contrary, but I am not sure this has much to do with me having an adult brain, though it might have something to do with me having an English-L1 adult brain that's learning German, which doesn't understand/care about declinations.
I am sure output and error monitoring is helpful, but I have learnt a ton of grammar only with input, so this hypothesis much be false in some absolute sense. None of these explanations (not enough time; stupid adult brain; need for output) explains why I can conjugate lots and lots of irregular verbs correctly, or know which verbs take the dative case, or know which verbs take 'sein' and which 'haben' in the past perfect, etc. I have learnt a lot of grammar through input WITHOUT doing output. It's just that there is a very particular (and unfortunately important) part of German grammar that has not been learnt via input. And this is not through lack of trying! If you consider there is probably a declination in written German every 5-10 words, I have probably seen somewhere between 0.5-1 million declinations so far. Something every particular is going on to stop me learning this. My two big theories/guesses at the moment are: (1) declinations don't effect the overall meaning enough for me to pay attention to them; (2) declinations are constrained by too many different factors in a sentence for my brain to learn them (at least without prior exposure to the grammar rules). Number 1 seems to be false as I am apparently learning other aspects of grammar that add little to the understanding of meaning (e.g., gender and also things like the conjugations of irregular verbs). So my best guess is that declinations (and anything else where you have to take different parts of the sentence simultaneously into account) are simply too complex for an input-only approach to work. At which point I agree that some aspects of grammar can only be learnt by hitting the grammar books, doing exercises/cloze-deletions etc, and learning by monitoring errors in output. Edited by patrickwilken on 26 October 2014 at 3:20pm 1 person has voted this message useful | ||||||||||||
| patrickwilken Senior Member Germany radiant-flux.net Joined 4536 days ago 1546 posts - 3200 votes Studies: German | Message 5 of 67 26 October 2014 at 2:58pm | IP Logged | |||||||||||
I think it's just perhaps easier later when you have more vocabulary and other parts of the grammar are understood.
I guess it's a case of swings-and-roundabouts (hope that idiom makes sense). I don't think my progress is that so or that MI has a been a waste of time. I have learnt a lot of grammar and vocabulary. My aim is to get to C1+ as soon as possible, and I am not convinced I could get there any faster if I focussed more on grammar first; perhaps yes/ perhaps no. I certainly my understanding of complex texts has not been affected by my lack of understanding of declinations.
Cloze deletion is a good tip, but I am in the unusual situation of having a live-in tutor who can correct written text for me. So I hope I'll get the same out of just writing lots of essays.
Extensive reading is a good tip. It's a trade off in learning more words (where extensive reading is better) or more grammar (where intensive works better). I have started switching between the two now, and also am finding that the writing is making me much more conscious of the grammar as I read. Thanks Serpent and EMK for the helpful responses! Edited by patrickwilken on 26 October 2014 at 2:59pm 1 person has voted this message useful | ||||||||||||
| daegga Tetraglot Senior Member Austria lang-8.com/553301 Joined 4524 days ago 1076 posts - 1792 votes Speaks: German*, EnglishC2, Swedish, Norwegian Studies: Danish, French, Finnish, Icelandic | Message 6 of 67 26 October 2014 at 3:13pm | IP Logged | |||||||||||
This is one of the few occasions where extensive listening can be detrimental. As long as one doesn't speak in a very formal/stilted way, <ein> is pronounced [ain] and <einen> is pronounced [ain:], so just with a slightly longer n. You might not hear the difference, there might even be none for a particular speaker. So you hear evidence that is contrary to what you inferred from reading texts. If you don't have a clear set of rules in your mind, this might be confusing for your subconscious presentation of language knowledge. Edited by daegga on 26 October 2014 at 3:18pm 9 persons have voted this message useful | ||||||||||||
| Cavesa Triglot Senior Member Czech Republic Joined 5012 days ago 3277 posts - 6779 votes Speaks: Czech*, FrenchC2, EnglishC1 Studies: Spanish, German, Italian | Message 7 of 67 26 October 2014 at 3:22pm | IP Logged | |||||||||||
|
I strongly believe the most efficient way is to combine lots of input WITH explicit grammar study. I disagree
the SC is an ovekill because my language learning has jumped several levels when I started devouring heaps and heaps of native input and I am not the only one, even though the exact amount of input needed may significantly vary between individuals. It helps a lot, I get a feel for what is correct and what is not, similarily to the one in native language as it is the exactly same mechanism (that's why people who read a lot rarely struggle with grammar and vocabulary in their native language while most people unable to write a simple email withou mistake are almost always those who consider books to be a waste of time). What I think may be the base of your trouble is a disbalance similar to one I have noticed in my Spanish studies. I believe it is extremelly useful to use a grammar book and look at the logic, rules and applications of the things you encounter in the native input every day. These two processes are awesomely enforcing each other. You study a grammar chapter and remember having seen or heard the point before. You see the application of the rule in a book or hear it in a movie and the things get connected even if you don't bother remembering "oh, the actor just used subjunctive". And most importantly, both sources get combined and harvested when you speak or write. That is why my active skills in Spanish are far behind the passive ones. Sure, I can get a point across most times but with many mistakes, but my undestanding is in most cases almost perfect. I clearly see the difference in using grammar I have both studied and soaked through immersion and in using the grammar I have approached only fromone of these directions. I think most adult learners need to study the grammar. There are lot of differences. Some people need small chunks while others prefer the bigger picture first. People prefer various types of exercise. Some read grammar books cover to cover while others follow an order given by a course or another source. But the common point is that we are shown the logic and create a basic structure we cover in many layers of details and exemples as the months and years go. From what I've seen and heard, people who do not build this basic structure somehow usually end up with all those exemples piled up on ground, in a chaotic heap helpful enough for passive understanding but difficult for their brain to navigate through to pick the thing needed at the moment. 10 persons have voted this message useful | ||||||||||||
| emk Diglot Moderator United States Joined 5535 days ago 2615 posts - 8806 votes Speaks: English*, FrenchB2 Studies: Spanish, Ancient Egyptian Personal Language Map | Message 8 of 67 26 October 2014 at 3:28pm | IP Logged | |||||||||||
Keep in mind that French children don't process gender the same way adults do until they reach 5 or 6 years of age. That's between 15 million and 78 million words of input, going by the figures from another paper I can't find right now. The Super Challenge was closer to 2.5 million words. Kids get a ridiculous amount of input before anybody expects them to speak really well. I think a lot of fairly competent adult learners (the ones without years of immersion) are basically reasonably articulate 3-year-olds with bigger vocabularies, and more conscious ability to clean up their grammar. Articulate 3-year-olds can say a lot of grammatically-acceptable stuff, but then they'll turn around make huge errors 5 seconds later.
Well, in the case of French, one of the biggest long-term challenges for adult English speakers seems to be mastering gender. There are people who've been living in heavy immersion for 40 years who still make plenty of gender mistakes. This cannot possibly be a problem of input after all that time. I hypothesize that this arises from three things: 1. Lots of studies show that adults generally have a much harder time with phonetics than infants. 2. A lot of French gender information is expressed by phonetic distinctions that English speakers can't really hear very well: "on" versus "onne", for example. 3. Adults may have some subtle grammatical weaknesses, too—not necessarily overwhelming ones, but enough that we need to work a bit harder. I speak French considerably better than my kids do right now, but despite my best efforts, they actually sound more native than I do. I certainly don't mean any of this to be discouraging. I've had very good results simply by training myself to pay attention to distinctions that I didn't pick up automatically. My brain might be inclined to ignore the difference between "-on" and "-onne" at the end of French words, but nothing stops me from focusing on it for a while and seeing results. Edited by emk on 26 October 2014 at 3:29pm 7 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. |
