Matt Vs Japan.... Has Anybody had this Experience?
I've had that experience with French. From no speaking to pretty decent speaking overnight.
Repetition of AUDIO material is key. If you hear something hundreds of times it's easy to reproduce as long as your mouth can do the actual mechanics of the sounds.
The audio aspect of this comment is the important part. We learn to speak through our ears.
I think Steve said this in one of his videos. Which is why I think he pushes for us to listen to the mini stories many many times.
Yep, it definitely works. Like i said as long as your mouth can make the actual mechanics of the phrases you can just say it without thinking.
I found my problem wasn't being able to say stuff, it was stringing together coherent thoughts in the language where i could just talk without having to think of what to say. In English i can just talk and talk without much thought but in French there's much more thinking going on, and it's not thinking in terms of translating the English into French or anything, it's more a delay where it doesn't roll off the tongue quite as quickly. It's hard to explain.
Agreed. IMO It's literally what "fluent" means. In English *and* Spanish, the words just flow out without thought. In French there is a delay for me which is not quite "fluid". In Russian I have to line up what I want to say and then slot in the words. Then I have to think about the grammar which is unclear. So essentially I guess it's sentence construction which is anything but fluid.
Yeah. 100%. I think the audio mini-stories are the most valuable part of lingQ at least in the early stages.
I only have a vague idea who Matt is and I think YMMV but Matt seems to clarify what he means starting around 26:30 in this video: https://www.youtube.com/watch?v=c21-nCVmieA&t=1655s
Thanks for finding this. I agree 100% with Matt's points here. There is a role for study and speaking practice in language learning, even if comprehensible input is the best place to spend vast majority of your time to acquire a language.
I'm pretty sure Matt had several years experience in speaking basic Japanese, before; he began using AJATT methodology to study the language. Which would explain his sudden jump from basic to fluent output. But that being said, I've definitely had the experience of studying through input for 1-2 years then opening my mouth and having natural sounding, fluid language come out. It isn't necessarily 100% correct all the time, but it is what I would describe as 'robust fluency'.
Isn’t this what krashen says? however I think Steve says you do need to speak. Personally I don’t want to practice speak so I’m just gonna do input. I’ll report back in a year or two what happens. Lol.
My method (for the first half to three quarters) is purely audio input. I find I can remember how words sound and I can say them more or less how I remember them sounding like.
I can pronounce more or less correctly and I can string words together in "sentences".
My spoken grammar sucks though. I can't reproduce grammar *at all*.
But according to the only native Russian I've spoken to live I am intelligible. (If somewhat childlike in my utterances).
I'm a fan of Matt's videos. I haven't seen them too recently so I don't recall when he first output for "real." But he took Japanese in high school, also did a study abroad program in Japan where he home stayed with a Japanese family and went to a Japanese high school. It was supposed to be for a year but he cut it short (I forget how short). He wasn't "fluent" then but I think he could talk just not in a way that made it enjoyable and he avoided doing it (but he was still in Japan so it's hard to imagine him not speaking any Japanese at all unless he locked himself in his room). He then went back to the states and continued inputting but not outputting... until finally he did output and it was like you said. So there was some output before he suddenly spoke Japanese but clearly there was a huge difference where he got so much better just inputting.
I had a similar experience with German I think. I got started and hooked on studying German when my dad paid for me to go study German in Berlin one summer when I was in college - I was living in Berlin for a few months. I actually ditched the course I signed up for a few weeks in and just bought books in German and tried to read them with a little paper Langenscheidt dictionary. I sort of tried to speak German when I went shopping, once in a while when I spoke to my landlord or randomly showed up at school. Most of my studying was on my own. At the end of the summer I visited some relatives near Frankfurt am Main (in Germany) and there I got to output for a week or so before coming back to the US. I was able to have a very simple conversation and this blew my mind. I continued studying (mostly just reading) German over the years after I went back to Boston. There was nobody to talk to and so it was weird to test myself on my speaking progress when I got the rare chance to actually speak. I basically leveled up between each conversation by reading/listening to German broadcasts on the Internet.
So I think you can make lots of progress by just reading or immersing without talking to anyone. Matt claims this prevented him from making bad habits. I think he's just smart and fixed his bad habits as he went. Seriously, he's way smarter than most people and you have to take that into consideration.
I find it hard to believe one could start to speak fluently in Japanese after 2 years of constant input (without practicing ouput). Matt's been at Japanese for over 10 years from what I've heard - getting fluent takes time.
You can check a short video of me speaking French after 95% input and 5% output in 23 months. I started speaking after 9 months of silence. In the infobox you'll find a translation into English.
Thanks for the video. I've understood your French video without any problem and you speak very well Italian too. Congrats and very inspiring.
How many hours per day did you average?
1,5 hours: 1 hour reading, 0,5 hour listening.
Obviously I can't confirm if you're reading off of cards but your pronunciation is pretty decent. I did exactly the same thing as you did with French for six months. I can understand spoken French at an intermediate level. Your French is better than mine.
The pure audio input method works.
An Okinawan friend of mine moved to Argentina with her family when she was young. She didn't speak a word for three years after she got there. Of course she spoke in Okinawan dialect with her family, but she didn't speak a word in Spanish with her neighbors. She watched dramas for three years and only started speaking when she was sure she could already speak. She said that if you start speaking right after you start studying, your Spanish will be weird, so it's better to start speaking when you can speak properly.
Matt talks about having exclusively Japanese friends in college and taking tour groups of Japanese tourists, not to mention his infamous six month period in Japan. So he really did not just start speaking with only input. He would have had hundreds of hours practicing with native speakers. I guess to Matt this output time was so small compared to thousands of hours of input, so he is talking relative times.
What is the actual story of his idol Khatzumoto?
At least you can learn old Norsk with heavy input ;)
(one of my all time favourite language learning scenes, even though it is not realistic... :)
Definitely, always loved that scene. :D
I reached fluency in Spanish by reading Harry Potter for 10 days: https://www.youtube.com/watch?v=TSpEU2f9tOw&ab_channel=DerPauleglot
All jokes aside though, I have no idea how people can claim to reach fluency without lots of speaking practice, getting corrected and, at least half-hearted, grammar study. I have more than 5,000 hours of Czech input under my belt but my speaking has taken quite a hit since corona started. This even happens to Expats who (virtually) stop using their native language. Well, and then there´s receptive bilingualism.
Don´t get me wrong, input improves your speaking, both fluency and accuracy, but I feel like you hit a ceiling at some point and then you need speaking, corrections and some grammar.
Long tie no see^^
I have been hiding from you. Had had enough of your antics. But I was wrong and now I am back for more!
The real issue is we have no clear definition of "fluency".
I find most people who are saying "you can't" are actually saying "you can't get to university level as in C1 or C2".
You can *definitely* get to "it flows out of my mouth, broken but intelligible" at the same time as "I can understand most youtubers speaking at normal speed" if the language is close to english in six months.
I didn't quite make that in six months for Russian but I'd say "it struggles to flow out of my mouth, I fumble for a word 1 in 10 and I can speak broken but intelligible" and understand a limited number of youtubers. I suspect I'll be at the same place with Russian as I am with French by the summertime.
But if it was "can you get to C1/C2 quickly" hell no.
Presumably Matt agrees that practicing speaking is needed since refold has an entire stage dedicated to output (stage 3) once your comprehension is good enough.
I don't know about this video but my current experience with German is saying that listening and speaking improve automatically just by increasing a lot of input. That's why I'm switching my learning experience using a lot more LingQ this year.
Even if I've learnt some German with bad mental focus, low energy and done not so well, I can understand basic conversations and I can formulate basic speaking sentences without having trained any of those. Which surprised me a lot.
So, I believe that doing a better job in increasing inputs in this "LingQ" way could give astonishing results. And reduced a lot the time needed in the target country. Increasing the initial inputs a lot before the output and then doing a short full immersion in the country could potentially have incredible results and being a lot cheaper.
One thing I don't know about is the writing experience. For now I find it very difficult without training. But I'm going to change a lot the way I've tackled this language in this 2022 and LingQ is going to be the center of it.
My 2c is that writing is the hardest skill of them all for me. I can speak very good Spanish (as in it feels like English to me). I can write but it is seriously broken. That said I did it mostly via audio and TV. So maybe if I had spent a bunch of time reading it might be different.
I don't know how much he exaggerates this, at least a little bit, though I think it's possible to some extent.
But I always concern about what a style of fluency comes with particular input. I really appreciate the ability to express myself in thr easy-going way in Russian and I want the same in English.
I'm afraid that literature, books, articles can build up such a type of fluency, that I would need to overcome with the additional effort in the end, that's why I'm currently trying at least to shift my attention onto the conversational type of input. After all, this is exatcly how we acquire our native language.
In other words, I don't want to start a sentence with "Once upon a time" instead of just "In past" or somethin like that. Sure you'll understand me but I don't want you to get the impression, that I'm your granny telling you the story of Goldilocks.
For example, in Russian there is a HUGE difference between how even our modern literature is written and how people really speak to each other. The 80% of complex grammar structures will confuse people in the speaking form. The same with too academical words. You can even отхватить по роже from the working-class if you seem too smart :D
You are not kidding. As you said at some point, some of the famous youtubers like Kate Clapp don't speak proper Russian. That said, I can kind of understand her now. I know way more about miniature handbags and makeup than I ever wanted to LOL.
>I've kind of done the same thing (although not for even close to the number of hours Matt did) and I'm trying to figure out if output (if not fluent, in some form) will eventually come naturally with time without much deliberate practice, which I know would be frustrating/hard work to me.
Yes, output will come with enough input. Some practice will help, but listening is a hugely important foundation.
Using myself as an example: my Cantonese speaking has made great progress, in spite of almost never speaking it. But my listenng is still not great. I'm at the stage where I can understand nearly all daily conversations, but if I watch a youtube video about history I get a bit lost. I have zero doubt that when my listening gets as good as my Mandarin listening my output will be better. A bit of practice helps, but I honestly believe that if you spend the 1000 hours necessary to get to "fluent" listening then speaking will come easily with like 20 hours practice.
If you spent years doing input, a 3-6 month transition to conversational fluency would be very possible, and would seem "sudden" in hindsight. I think he is saying he had a foundation based on input that made that transition relatively quick.
As a counterpoint to Matt's experience, I read zero Manga, had never watched anime and didn't use an input-"only" focused method when I learned Japanese. I learned all 4 modes (reading writing listening speaking) of the language in parallel through a combination of university classes, both in the US and Japan, an internship in Japan and taking my first job there after graduating. I was conversationally fluent in 2 years, could use workplace Japanese proficiently after 3 years (including my internship) and started my first job as an engineer communicating with my co-workers and with local engineers on a day to day basis in Japanese.
That whole process took 4 calendar years, but I wasn't putting in nearly the hours Matt was. My major was engineering, not Japanese, and I had a full life outside of Japanese as well.
That said, I do think that speaking in our daily university conversation classes with teachers who would correct our language, including pronunciation, was an advantage. I do take Matt's point that speaking in another environment without that corrective feedback before you have adequate input could lead to developing some bad habits that could be hard to correct later.
Also, I don't have time to even to do the approach I did with Japanese (much less what Matt is recommending). So, I am trying a comprehensible input approach with Spanish, only for about 30-60 minutes per day. I do hope at some point I'll look back in hindsight to think the transition to fluency was sudden, because at the moment it's pretty rough lol.
I'm fairly sure he said he found he could speak fluently (not perfectly) "right out of the gate," or words to that effect. I'm not sure he'd would've said that had it been 3-6 months of practice, even it that would seem to him like a short space of time. It was definitely implied, IMO, that it was almost instant language coming out of him.
That's interesting. I think neither matt or Krashen (or Steve?) believe that being corrected is of any use whatsoever. Though I've heard many testimonies like yours of people saying it worked well for them. I suspect the 'input only' crew would put that down to the input you received, and that you've mistakenly perceived that it was the corrections that got your speaking to a good level.
FWIW, I'm unsure either way. I certainly don't think corrections would hurt, but I wonder if using that 'conversation practice' time would be better spent doing more input, at least until you're at that stage where you're ready to output. It seems to me like the idea of a long period of all input is to avoid forcing output, which will inevitabIy be littered with errors, even though some of those errors will be corrected in the moment.
I hope you find success with Spanish!
Thanks! I'm definitely on team input!
I have also heard this said about corrections. Most of the "input crew" will allow for the effectiveness of vocab study and even (gasp) grammar study in some cases. Steve says he doesn't do SRS or vocab lists, but what is LingQ if not a way to integrate vocab study with input to make it more comprehensible? Bill Van Patten says grammar instruction is effective "insofar as it makes input comprehensible". I think this is a key difference. These things are valuable only if they are timely and relevant to your use of the language.
I found a presentation that Van Patten gave (here, in Spanish around 23 minutes or so: https://www.youtube.com/watch?v=HgOABLFUJuI&t=3658s) where he talks about the type of correction I'm talking about. He says that language production has nothing to do with language acquisition, with one exception, which is when the communication is part of a meaningful interaction. He gives this example:
Native: ¿Donde viven tus padres?
Student: Mis padres, uh, es, uh - vive en Sacramento.
Native: ¿Viven en Sacramento?
They define this as input, but this is the type of correction I'm talking about. The student incorrectly uses the verb (vive instead of viven) and received immediate feedback because the Native restated his statement with an embedded correction in the form of a question. If the learner notices this in interactions like this one, according to Van Patten, he gets a clue through "input appropriate for his level from another person." I've found this valuable inside and outside the classroom.
My classroom experience at university was like this every day. You could say it was comprehensible input from day one, or what Krashen calls i+1 input from day 1. We also spoke from Day 1 in this way.
Also: one other thing.
It's possible to "cheat" not "studying" grammar but still learning about it *directly* through input.
Some of the youtubers that I find easiest to understand are in fact Russian language teachers who are doing youtube podcasts of wait-for-it.... grammar lessons.
"about russian in russian" is essentially this as is "maria petrova". Basically they just kind of use simpler words to talk about the grammar. Maria Petrova is great because she usually tells a story with flash cards first. Irina (lastname?) from about russian... is also good.
They are both teaching old school... writing stuff down on whiteboards and talking about grammar.
In Russian for example "cases" are supposed to be super-difficult. In fact I think conceptually if I understood both Maria Petrova and Irina correctly, they aren't that hard to understand. Just wierd and IMO crazy. Thanks Russians!
That said, since we don't have them in English, they are super hard to remember. When I'm trying to speak, my brain is overwhelmed by the processing power required to recall and just output the words in a string so cases go out the window.
I suspect Mandarin is going to be easier because my understanding is it has very limited grammar compared to Russian.
Right. There is a dichotomy between "fluent" but "not perfect". There are tons of very functional (and fluent) immigrants here in Canada whose first language is not English. But many of them make consistent wierd (to me) grammatical errors. I am in the same boat. My Spanish is fluent but not 100% correct. I am self taught so my grammar is likely wierd. In Russian I'm not fluent (although I can produce kind of) but my grammar is totally and completely borked.
In French it's similar although I'd say I can speak "fluently" in French it is no way perfect. Probably most of my grammar is essentially subconscious Spanish I have twisted into French.
So yeah to me one of the largest issues is "what do we mean when we say 'fluent'?"
As a disclaimer to this, I do not watch Matt vs. Japan. I have probably seen some videos, but that type of content generally has not been very interesting to me. I am familiar with Refold.la, but haven't read through it all.
I do not think Matt is exaggerating at all. I think in fact the opposite, that people see some of the numbers being advocated and assume it is exaggerating or overstating how much input we really need. IIRC Matt advocates for 5-6 hours of input daily for 1,5 years before starting output. That is 2.500 - 3.000 hours! If I spent 2.500 doing something I would hope I was reasonably good at it. Saying one fluently speaks a language after 100 hours of speaking practice when that is less than 4% of their overall time is a very different claim than they spent 100 hours of speaking practice and nothing else.
I would read "holy, s**t! I can speak Japanese." as an self affirmation more than anything else. After the early stages of language learning the improvement is so slow and unnoticeable that it is hard to recognize how much progress we are really making. Getting a job speaking Japanese to people from Japan is an objective marker that one might speak Japanese! We are all subject to cognitive biases, and this type of objectivity can do a lot of work to cut through those biases.
Right. By my reckoning about 500 hours of listening was enough to get French up to a decent level where I can understand almost all youtubers and speak coherently.
500 hours of listening is *not* enough for Russian but I can understand *some* youtubers and speak broken.
Do you really need 2000 hours to get to understanding combined with conversational fluency if all you are focusing on is listening and speaking? I don't believe so. But we'll see.
I have been following his methodology religiously. Exactly around the 18-month mark in a conversation with an old German lady, my subconscious mind came up with a few sentences fluently out of nowhere.
They just popped out of my mouth like water flows. I had no flipping idea about grammar tenses but the lady noticed my choice of word. In short, when I was standing up and going for my home she said I spoke well in German. The quality of the sentence impressed her maybe but I would not consider myself a true fluent at that point.
Around 2 0 month mark I got a job , I spoke a few sentences out of nowhere at work and my German bosses noticed them. Again the same issue they were grammatically correct sentences. There was no conscious effort involved in formulating those sentences. They just flew. Again they were kind of individual sentences. I will not still consider myself a true fluent speaker. But the quality of sentences I can not judge by myself but German bosses noticed it. I am feeling like I am getting better at speaking but I am not right there yet.
The funny thing about this is that I did not touch a single grammar book in German. Only input heavy. Matt started outputting at a 2-year mark according to him it was not that perfect initially and took another year of input before he felt comfortable speaking Japanese fluently.
My German boss told me a few days ago whatever I am doing to learn German I am on the right track.
OK, that's interesting. I'd like to see how you're doing in another 20 months. I'm personally starting to believe that one possibly needs 3-4 years of constant input. I can see the method working the best of all the methods out there, but only for those who have the luxury to put their native language aside for long parts of their day during those 3-4 years. At least for things to start to become automatic. It seems like all input for an hour/day, spending the other 15 hours/day in your native language, would never really get you very far in terms of automaticity. I might be wrong.
For very high fluency my guess is you're right on the nail.
That said, for acceptable fluency, you need less than half of that. A six month period of brutal slogging, 3-4 hours a day, then another six months of just an hour a day.
Pretty cool. That's motivating. I find words pop up in my head in Russian but not complete sentences. I'm only at 7 months though. Love to hear what you'll be like by the summer time (2 years?).
This March it will be 2 years and by March 2023 it will be 3 years in total and I think I will try to maintain the same intensity by then to assess how his method actually turns out.
His method is technically a full-time job and that's how a kid learns his native language just through sheer language exposure. As per his given timeline 3 years for fluency and 5 years for perfection but yeah after a three year I will keep learning the language on a daily basis but I am not sure if I can maintain the same intensity due to other factors like getting a professional job lets see how it goes
I think though, that like Matt, you're shooting for a very very high degree of functionality in the language because you need it for day to day life as well as work.
So yeah, there's no surprise that it's a full time job to get to that kind of level.
That said, after having already done a pretty functional level of Spanish in the past, I can say that the intensity definitely slows down after a while. I think the highest intensity for me was the first year to two years. By the time I hit the five year mark Spanish was like English for me. I could use it to do essentially everything.
instant output seems unlikely to me... but the ratio is likely completely different. If he invested 2k, 3k hours into japanese input, then 100 or 200 hours of practicing output might seems like nothing to him. I can believe that 100 hours of output would be sufficient.
Matt was also in Japan in high school for a year I believe. If he was able to build up up basic conversational abilities back then, then I do think the rest should come a lot quicker at more advanced levels.
I think he was also into AJATT and things like that from early on in his studies, so that was probably a driving force.
That's interesting. Do you base that 100 hours as believable on some kind of experience in outputting?
I believe he was in Japan for 6 months (he cut it short because he wasn't enjoying it), and he said he barely left his room and that his level was too low to join in with conversations, or to even understand them.
Yes but I wonder if that is an exaggeration. He went to school there. A short conversation here and there, it quickly stacks up.
I don't really base that 100 hours on much. I learned French when I was in school. At the end I could speak rather well but my knowledge of grammar was quite limited and my vocab even more so. Then I didn't really speak any French again for 11 years.
Around the time I started lingq, I also took around thirty hours of Italki lessons over a period of +/- 3 months. After those 30 hours I was again able to speak comfortably. Although still with limited vocab, I was able to manage a conversation without too much thinking/struggling on any topic - not perfect and definitely with some mistakes but good enough that some people would likely say I was/am conversationally fluent.
So I would think that if someone had extensive input, had some output experience in the past and then did more than 3 times what I did as described above, that they would have a good level.
Don't take anything any YouTuber tells you at face value.
I don't, but I think I agree with his ideas about language learning in general (very similar to Steve) so I'm just curious about what the end result of that has been for those who've done it.
The oversall math of the input does workout. My guess he was listening to very specific information that provided a great deal of space repitition. Three or four hours a day of input spaced throughout the day would enable someone to understand and even speak a language after 2 years. Also he mostly lilky was internally repeating what he was listening to thus building an internal phrame work of the language.
My suggestion is be happy for his journey and don't compare his work too the the work that you will do now or in the future. Two people can put in the same amount of time but the results will be different. Results are based upon how we as learners learn.
I think you're right. I originally started because I was intrigued (but somewhat skeptical) by what Benny Lewis said. He made the bald claim that you can do any language in 3 months if you just start speaking (combined with memorization). I hypothesized that 6 months should be good enough to *understand* any language if you put in enough input.
I doubted him because of the somewhat disheartening number of hours the FSI says it takes.
We were both wrong. The FSI is right in broad sweep. Some languages take more effort than others. That said, if all you are looking for is conversational fluency and listening comprehension and no reading or writing, then the FSI's numbers I believe are off by a factor of at least 2 (for the distant from English languages).