Asriel Wrote:Of course producing sentences doesn't improve your ability to understand the language any more than you already do. But outputting gets you used to coming up with content on your own.

And that's exactly what you shouldn't be doing if we're following the input method (no not just Khatz's blog, but actual linguistic research that supports it).

What the input method states is that in natural native speech almost nothing is actual original content. Sentences are simply slices of phrases taken from other contexts that the speaker already knows. Research has shown that most people have a set number of speech patterns that they just repeat all the time. Sure the vocabulary in the phrases might be different, but the speech patterns hardly ever change.

By getting enough input you'll eventually just KNOW what speech pattern to use, not because you know the "rules" as to why, but simply because you've heard that exact situation countless times. This shouldn't be surprising considering it is actually how native speakers process speech. They didn't get good at speaking from practicing "creating" speech. They got good a speech because they copied other people speaking.

Your post just then reminded me of something I've been thinking about for a while.

To the Japanese natives frequenting our forum who follow AEATT you should SRS the posts made by Tycho from Penny Arcade http://www.penny-arcade.com/

That dude's writing is, in a word, awesome. I want everyone to sound like that.

blackmacros Wrote:Your post just then reminded me of something I've been thinking about for a while.

To the Japanese natives frequenting our forum who follow AEATT you should SRS the posts made by Tycho from Penny Arcade http://www.penny-arcade.com/

That dude's writing is, in a word, awesome. I want everyone to sound like that.

Hey I know my English is bad but I'm not exactly doing AEATT so you can't blame it. What is that dude's background? Did he SRS?

masaman Wrote:

blackmacros Wrote:Your post just then reminded me of something I've been thinking about for a while.

To the Japanese natives frequenting our forum who follow AEATT you should SRS the posts made by Tycho from Penny Arcade http://www.penny-arcade.com/

That dude's writing is, in a word, awesome. I want everyone to sound like that.

Hey I know my English is bad but I'm not exactly doing AEATT so you can't blame it. What is that dude's background? Did he SRS?

No no, he's a native English speaker. But in my opinion his writing is bizarre, and full of character and incredibly awesome. activeaero was talking about how we basically copy our input when we output. My desire is to for you to read his writing, SRS it and become awesome mini-Tycho machines. Then I could die happy.

blackmacros Wrote:

masaman Wrote:

blackmacros Wrote:Your post just then reminded me of something I've been thinking about for a while.

To the Japanese natives frequenting our forum who follow AEATT you should SRS the posts made by Tycho from Penny Arcade http://www.penny-arcade.com/

That dude's writing is, in a word, awesome. I want everyone to sound like that.

Hey I know my English is bad but I'm not exactly doing AEATT so you can't blame it. What is that dude's background? Did he SRS?

No no, he's a native English speaker. But in my opinion his writing is bizarre, and full of character and incredibly awesome. activeaero was talking about how we basically copy our input when we output. My desire is to for you to read his writing and become awesome. Spontaneously. Or perhaps after SRS'ing it. Either will do, really.

mmm I always thought doing something too flavorful is a big no no until I'm totally comfortable with the basic stuffs, which I mess up all the time, but may be Magamo or Aijin can try it. I'll check his posts though. Thanks for the link!

thermal Wrote:Yeah pron is difficult. Pron includes not only mouth movements, but body movements as well. Only when one gets enough pron input can one hope to replicate these movements, so we need to get as much pron input and pron practice as we can.

I think I just output.

QuackingShoe Wrote:

thermal Wrote:Yeah pron is difficult. Pron includes not only mouth movements, but body movements as well. Only when one gets enough pron input can one hope to replicate these movements, so we need to get as much pron input and pron practice as we can.

I think I just output.

I think it's best to input and output. Wait, that felt like I said something different from what I just said.

IceCream Wrote:ok. its 3 months 16 days and a few hours since i started learning japanese properly. It's 2.7 months since i started SRSing.
i can pick up easy manga and read chunks without a dictionary. i can watch easy shows and get whats going on and what they're talking about. I can repeat audio ok.
but i don't feel like my ability to speak has improved at all in the time i have been SRSing. It takes me aaages to remember how to say/think things.
Am i some kind of speech retard or is this normal? I'm quite quiet in english too in general.
I was planning on going to Japan at the end of September, but it looks like that's gone a bit wrong. for the nice winter snowboarding. but for more 24hr learning time.

So, if i say, in 4-6 months to go to Japan, what are the very best things i can do to get better at speaking (along with understanding more), in that time?

How long did it take anyone else to start speaking and thinking in japanese? What are the best and most helpful things you've done that have improved this?

Thanks

Have you heard of the "silent period" during 2nd language acquisition? I think you're still firmly in it:

http://www.everythingesl.net/inservices/..._93415.php

You've only been learning for 3 months, your brain is still taking in all this new information. Just keep listening and SRSing and give it time.

activeaero Wrote:What the input method states is that in natural native speech almost nothing is actual original content.

Depends on how you define original content. Outputting is a creative action. Are we just splicing together fractions of sentences without creating anything of our own? Yeah, possibly. It's still not whole sentence. We don't mimic sentences we've heard before, we make them up as we go. Someone just using memorized sentences would sound like a robot.

The proper way to connect the various grammatical pieces takes a LOT of training. You can watch Japanase 24 hours a day for a year, you will still not hear every sentence you will need to use for even a week in Japan.

activeaero Wrote:What the input method states is that in natural native speech almost nothing is actual original content. Sentences are simply slices of phrases taken from other contexts that the speaker already knows. Research has shown that most people have a set number of speech patterns that they just repeat all the time. Sure the vocabulary in the phrases might be different, but the speech patterns hardly ever change.

By getting enough input you'll eventually just KNOW what speech pattern to use, not because you know the "rules" as to why, but simply because you've heard that exact situation countless times. This shouldn't be surprising considering it is actually how native speakers process speech. They didn't get good at speaking from practicing "creating" speech. They got good a speech because they copied other people speaking.

This is what I'm talking about. I'm talking about creating the CONTENT, not the speech itself. Of course you'll be incomprehensible if you make up everything you say.

The point of output is to use those "slices of phrases" that you "already know." The difference is that you're using "different vocabulary" in "different phrases and contexts." You don't try and learn anything new. You reinforce what you "already know."

And of course, if you're doing it right, you're outputting regardless of whether you know the "rules" or not. The point is to use it enough that it becomes second nature so the "rules" are meaningless.

And by "output," I'm not talking about just sitting there and writing "This is a pen." "This is an apple." "He is a boy."
I'm talking about actually conversing with actual native speakers (which may count as both input AND output *gasp*)

Let's face it...you're learning the language to converse with people who speak the language, right? If you want to get better at conversing, then converse!

When I find out I'll tell you!

For me, even thought pronunciation is super easy, my speech ability is still very weak. But I do I feel it improves a bit everyday, even without virtually no effort, so I just keep doing what I always did.

When I'm closed in my mind I can speak what I want, but when I face a real opportunity I simply can't maintain the pace. I can't come up with words at the required speed. Like an iceberg, the words are there, but very little shows up on the surface. I'm afraid of doing mistakes or I feel ashamed of what I might say.

I feel what I need now is not really output training, it is social interaction.

I find that after doing a lot of steady reading, or steady listening(like an audiobook, tv has less of an effect), I think in Japanese more and it's more fluid than other times.

Asriel Wrote:This is what I'm talking about. I'm talking about creating the CONTENT, not the speech itself. Of course you'll be incomprehensible if you make up everything you say.

The point of output is to use those "slices of phrases" that you "already know." The difference is that you're using "different vocabulary" in "different phrases and contexts." You don't try and learn anything new. You reinforce what you "already know."

Well then it looks like we are on the same page.

IceCream Wrote:...

In case you missed it, you might find this interesting: http://forum.koohii.com/showthread.php?p...2#pid64392

I think the call-and-response format has merit even without the extended audio I mention there, though I think it's best to roleplay one side of a full discourse rather than have individually shuffled cards (I'm still going to do those, but mostly to prime myself for the extended audio). You still get the benefit of context, esp. if it's from a memorable scene. One thing I've learned is that context really stays with you from when you first create/study a card. That applies to just about every subject I've studied where I extracted the content myself.

You might also like resolve's examples of active recall (scroll down to "If we move the target language to the... ") here and references to Pimsleur's incorporation of context/production: http://web.archive.org/web/2008020205435...nlearning/

The goal, SRS-wise, is to make these cards so that they can be memorized in a very robust way so they can be focused on, and later used in a variety of situations. So you'd want to try and think of the context in ways that are universally applicable, with the kinds of phrases that you'd want to repeat. I imagine progressively remixing these bits and pieces to create almost fan-fiction like dialogues so that you're maintaining a sense of context.

This is also interesting and applies I think to the idea of using the feedback loop of studying cards with their imaginary contexts and then applying them to oneself: http://www.scientificamerican.com/articl...increase-c

sorry to revive the thread but this was interesting to read. I was thinking the same thing as ice cream. How to improve output, I'm seriously lacking in that department. I think I should divide that into sections, basic,conversational,functional,sem-fluent,fluent,native-level. In terms of speaking. Obviously I want to get to functional level of speaking before anything else at the moment. I'm thinking by the end of this summer to at least get to conversational level. Hopefully...
Understanding what's being said isn't the problem all though I should definitely follow conversation podcasts,etc for more in-depth understanding of random conversations. Example would be like about weather,school,jobs,relationships,etc,etc any-topic.

Why does the question of output keep coming up so often? Seriously all you have to do is start speaking Japanese. You have to learn it by speaking it. Here's the way that I have learned how to do it (and my speaking ability is the best compared to reading and writing).

1. Find some kind of audio where there are two speakers having a conversation. Then when one person says a sentence/fragment the other person will reply. That's what you want to practice, the back and forth of the conversation. So for example.

person A: 行きますか
person B: ええ、今日は行きますよ。

Bam, that's all you need. But what do you do with it?

Get your audio on something that you can easily stop and start again, rewind and fast forward, like Winamp or something. Then do these 4 steps.

Step one. Translate.

Make sure you know exactly what is being said so that you don't learn something wrong, because you are going to be repeating these conversations a few times.

Step two. Repeat after the speaker.

Play the first sentence then stop the tape. Repeat after the person EXACTLY as they do. Hit play, listen to the second utternace and then stop and repeat EXACTLY as you hear it. Do this again and again. If it is too hard for you to sound exactly like what you heard then try harder. I would suggest at least 3 or 4 times per sentence. We are talking output here so you need lots of time to get your mouth working.

Step three. Listen to the first sentence and then RESPOND to it/ Say the first sentence and listen to the response

So listen to person A and then instead of repeating person A, stop the tape after he/she finishes and OUTPUT what person B would have said. So let's look at my example

PERSON A: 行きますか。
YOU, NOT PERSON B: ええ、今日は行きます。

Step 4. Shadow the conversation.

Basically you speak at the exact same time as the tape. This makes you hear where you are screwing up your pronunciation or tell you if you are speaking too fast or too slow. Do this once or twice at the end just to make sure you go it.

Find another exchange and do it again. After spending time doing this you will be able to speak fluently.

The one downside to this is that it takes time to repeat the same things over and over again so expect to spend some time on it. However I promise you you will become a fluent speaker this way.

Edit, I'm saying tape because I'm oldschool like at. Its probably going to be a dvd, mp3, or some downloaded movie you have.

There's an addition to this that Antimoon suggests. Like Bizarrojosh talks about, get some conversations from somewhere. I use subs2srs though there may be some useful phrases from the other sources (tae kim, core 2k/6k) but not as much since they're not used for a conversation for the most part.

So you have a sentence that's the reply or question to another person's sentence. Before you repeat it, put yourself into that situation or apply the sentence to your situation. Change things to refer to you (gender terms) and the person(s) you're likely to say this to (name of spouse, child, parent, boss, co-workers, bartender, etc). Picture yourself in a location you're likely to say this at (work, home, bar, etc.). Now say it as if the above were happening to you. What I do when it's subs2srs cards is read the sentence as it's written, then I try to repeat it without reading. It's during the repeating part that I might try a little role-playing like this.

It takes a little bit more time, and there's no need to do it with every sentence. Still, you should make it a common habit and do it with sentences you think can be put into a situation you might experience. Subs2srs are great sources those are situations you have seen and have been put in the middle of emotionally just by watching it. Now you're doing it again on a sentence by sentence basis. The core 2k/6k are good because the voice actors put emotion into what they're saying. Got your sentences, now act with them.

I have an interesting experience with output. I never really studied chinese at all but sort of picked up some from my wife. I know probably < 500 words but I know them extremely well and can output them like nobody's business. Of course I can't read them and never actively studied them but can use them because I just kept speaking with my wife. Those 500 words can get me pretty far. So you can definitely get good at basic output with 0 reading input. I guess that's how babies do it too.

As far as Japanese I've been studying for about 1 1/2 yrs and working on jlpt2. I can't speak worth a darn but my reading is pretty good and know tons more words than chinese. It's amazing how much more work I had to do to get to a good reading level vs. just simple output.

We went to Japan recently and it's amazing that you don't really need a lot of words for day to day speaking. Checkin/Checkout from hotels, train reservations, ordering food. I kept wanting to practice more complicated things but other than watching tv and reading, in day to day life complicated words don't come up too much in daily conv. I think my JP output will eventually surpass my chinese but it's been much harder to get there for some reason.

Recently I've been watching 24 in japanese and doing a lot of role play so hopefully that will improve my output.

Thanks for the help guys. Sounds like If I just go for the simple stuff first, then build my way up,I can definitely get far. My understanding and reading are getting pretty so I'm pretty good with those.

So for output start small, one person speaking and other person talking back,etc. Get some audios,practice with that, get some written text as well. I've specifically made a deck for just this thing

I've always found a correlation between how fluidly one can write and how fluidly one can speak. Now, of course, one will be able to speak much faster than one can write (usually), but why not use that as a method of getting better at speaking as well? Use lang-8 to post journal entries. In each entry try to use a new grammar point or word (if it's a word, I recommend using two or more new ones) and watch how it enters your active vocabulary. I've been writing in Japanese for a bit now and I've noticed that even when I don't speak Japanese for a week or so, as long as I listen to it and write in it, my output while speaking still increases every time as more and more new things enter my active vocabulary.

Also, something I found to work nicely for both speed of output (in regards to getting my mouth used to forming the words) as well as pronunciation practice is to find something simple (reading passages from textbooks work well) and read them into a recording device. I just use my macbook's speaker + Quicktime 7, but a normal recorder would work I'm sure. Anyway, just record yourself reading the passage (2 to 3 times is best) and review it. Notice what words sound weird to you, or if you are not at that level try to get a native to check it over for you. Compare word pronunciation with native material and see what needs work. Also, make a strong attempt to read it as fast as you can, as if it were your own native language. I found that often, by the third recording, I was able to read it as fast as I would be able to in English.

I don't think the whole "think completely in Japanese without forcing myself" aspect happened until I was five years into studying it. And I talk to myself a lot too, in the languages I study. Lately it's been a weird mix of French, Russian, Spanish, and Mandarin.

Ryuujin27 Wrote:I've always found a correlation between how fluidly one can write and how fluidly one can speak. Now, of course, one will be able to speak much faster than one can write (usually), but why not use that as a method of getting better at speaking as well? Use lang-8 to post journal entries. In each entry try to use a new grammar point or word (if it's a word, I recommend using two or more new ones) and watch how it enters your active vocabulary. I've been writing in Japanese for a bit now and I've noticed that even when I don't speak Japanese for a week or so, as long as I listen to it and write in it, my output while speaking still increases every time as more and more new things enter my active vocabulary.

I think text chat is a good in-between. The language used has a lot more in common with speech and you have to deliver it fairly quickly. You don't get to practice many of the things that count towards spoken fluency when writing compositions. It's still good general practice though, of course.

I'm not talking about writing compositions, per say. When I write on lang-8 or when I just write something down in Japanese, I think of something to write about and then I just type. Whatever I think comes out and when I'm done I don't go over it I just hit submit (or just delete it if I was doing it for practice). That way it's more like verbal communication in that you can't really take anything back.

Ryuujin27 Wrote:I've always found a correlation between how fluidly one can write and how fluidly one can speak. Now, of course, one will be able to speak much faster than one can write (usually), but why not use that as a method of getting better at speaking as well? Use lang-8 to post journal entries. In each entry try to use a new grammar point or word (if it's a word, I recommend using two or more new ones) and watch how it enters your active vocabulary. I've been writing in Japanese for a bit now and I've noticed that even when I don't speak Japanese for a week or so, as long as I listen to it and write in it, my output while speaking still increases every time as more and more new things enter my active vocabulary.

I'm going to have to completely disagree with you on many levels here. Maybe, as you've said, you've found a correlation between writing and speaking in your experiences, and I'm not trying to say that your experiences are wrong. I am going to say that if you talk with 1st graders, they won't know how to write jack-shit and if they can speak, this proves that they can speak better and more fluently than write their native language. In other words, a young childs writing skill does NOT determine his or her speaking skill at all. It's probably the complete opposite (but I don't know that, just a guess). However, with that said, a 1st grader still doesn't understand the world as deep or broadly IN TERMS OF LANGUAGE as a person who can read the "classics" of that language (in English we have things like Shakespeare, etc etc). Thus a person who can read and write effortlessly probably has a larger command of the language and can express him or herself much much better than a person who can't read or write. But, again, I disagree completely. Just because you can write fast doesn't mean you can speak fast, and just because you can speak fast it doesn't mean that you can write fast.

I'm still a language noob, but for other skill-type things, I've often found a correlation between how well I do something and how often I attempt to do it.

My plan for building oral skills is to face a mirror with a strategically placed mustache and pair of glasses and ramble about whatever I know enough Japanese to string a coherent sentence into while a pre-recording of timed aizuchi responses plays from behind it.