I am not going to try and convince anyone that my idea is the best way, but from what I gather it seems like a logical progression of the idea behind RTK.

Sure, vocab are easy to remember.  Great!  In fact the examples I put above give you 17 words in about 5 minutes.  So why not go ahead and learn the words?  It will take you much less time than RTK (like I said, a few days to a few weeks if your memory is honed), and then you don't need to spend countless hours NOT enjoying your input and instead looking up stuff in the dictionary repeatedly.

I think that memory techniques are efficient, especially when applied broadly like I suggested and even if words are easy to learn one at a time, they are even easier to learn if you think about them in a systematic way.  I think we all have stuff to learn from the people who really master their field, not that we may ever achieve what they have, but why not follow their lead?

After three months of Heisig, I would absolutely lose my mind if I tried to keep learning things in isolation.  With Heisig, the only thing that kept me going until the end was a vague sense that the payoff would come later, and a large -- very large -- amount of spite.  I couldn't let it win, you see.

Now that I'm to the learning with sentences part, it finally feels like I'm making some sort of progress in Japanese, since every single word and grammar point that I pick up is accompanied by a satisfying "Hey, I can read that!" moment.  It's very soothing on my tattered nerves, and it makes learning new material enjoyable.

If you can learn the words in context just as fast then of course, go with context.  I, however, propose that if your memory is honed, that learning words alone is much faster because you can easily apply memory techniques (much like RTK).  Furthermore, I think if you already know the words, the context is much easier to learn as well.  I don't know how many times I have come across a sentence with a single word that I didn't understand, and the whole sentence became incomprehensible.

I am actually using kanji odyssey right now.  I think it is great!  One reason is that it hardly introduces a sentence with yet-to-be-studied words in it.  Real life input is not like that in my opinion.

Mcjon01,
I agree and empathize with what you are saying.  However, I spent a year without RTK and AJATT trying to learn Japanese only to be frustrated with how little progress I was making.  If I had known what a huge payoff RTK would have been, I definitely would have started with it, and if I had known how having a large vocabulary would have helped my studies in a similar way, I would have gladly gone right into that as well!  To think of trying to understand a language without knowing the words just doesn't make any sense, as well as trying to read a sentence full of kanji that you don't know!

I think I could be in the same place I am now (two years of study) in just 4-5 months time by following my outline, time I would gladly have given up to know that the next 1 1/2 years would be much more comprehensible input.

Are there any people at all who see benefit in what I am talking about?

I think you should take a detour by this site:  how-to-learn-any-language.com.
Check out the forums. Contests regularly take place to see just how much can be done in a time frame of X.

Try what you proposed, you might find the results quite surprising ! Just for fun I did one month of german speed-learning. That month was spent entirely reading literature and a dictionary (admittedly, I did go through assimil the first two weeks or so). I can now read wikipedia articles for fun and get the gist of most articles.

But regarding Ajatt, I believe khatz's advice is to get in contact with real texts. I'm pretty certain he used the *HECK* out of that wordtank dictionary. If that's not learning words, I don't what is. I don't know.. Khat's advice of going with texts is good because at some point something /will/ pick your curiosity and make you look up a few words. Also, my experience with german and english have made me realize knowing words is not *that* important. Seriously, how many times have you seen a foreigner string up a sentence with all the correct words but that people just can't get any ounce of meaning from ?

In any case, many words just do not make much sense in isolation so khat's advice is not totally unfounded. So yeah -- a good way to go I believe is to do both concurently.
Generally, knowing 4 out of 5 words in a syntactic group is enough to get the meaning of the fifth, so no need to rush.

Hope that made sense


Edit: Holly freaking yellow mutant frog! This thread had no replies when I started typing

Last edited by pitwo (2008 August 02, 6:58 pm)

Zarxrax,
I agree, using sources that have all known vocab is a great idea, and I have used those exact same graded readers successfully.  So what do you propose after that?  How do you find sources where all the vocab is easy to understand?  Why not get a lot of vocab in your head in the first place, and open up the rest of literature to within reach?

I am gathering that NO ONE thinks learning words is a good idea.  Out of all the people who have successfully applied RTK, no one thinks that this idea could be extended?  Also, I think not knowing just one word is enough to make you start to lose interest in something.  What about manga, where often there is one hard word and the rest is stuff like "da to omou n desu kedo...".

What about if someone laid out a list for you and gave you mnemonics and methods to memorize them.  Would you use it?

Last edited by danieldesu (2008 August 02, 7:11 pm)

Another thought: having learned a kanji previously makes it familiar when you see it in the future.  It may be easy to learn a word while you are reading a sentence, but what if you already knew the word for "philosophy" for example.  When you see it in a sentence, you may be able to learn the word from scratch, but if you knew it already, then you are already familiar with it and you can say "oh, so that is how you use the word 'philosophy'" instead of "ok, here is this new word 'philosophy', and I am seeing it is used in this way."  I think the former will be much more comfortable seeing and using that word in the future.

danieldesu wrote:

Zarxrax,
I agree, using sources that have all known vocab is a great idea, and I have used those exact same graded readers successfully.  So what do you propose after that?

Have you tried "Read Real Japanese Fiction"? ( http://www.amazon.com/Read-Real-Japanes … 4770030584 ) I've had this sitting on my desk for a few weeks now, and am planning to start into it next week when I go on vacation. From what I've looked at so far, it looks great. You have Japanese text on one page, and notes explaining it on the opposite page. There is also a dictionary in the back containing the words that the stories use, so you probably don't have to keep an electronic dictionary handy.

In any case, you sound like you might even be farther along in your studies than I am, so hey, do it however you think would work best for you! Everyone learns a little differently, and maybe your idea would be good for you (though i personally can't really see the benefit).

danieldesu wrote:

I am gathering that NO ONE thinks learning words is a good idea.  Out of all the people who have successfully applied RTK, no one thinks that this idea could be extended?  Also, I think not knowing just one word is enough to make you start to lose interest in something.  What about manga, where often there is one hard word and the rest is stuff like "da to omou n desu kedo...".

Actually I agree with you, and I do a similar thing. I keep a separate spreadsheet with J/E word pair lists, using mnemonics extensively I add about 50 new words a day easily. Most of these words come from sentences, but some just come from word lists eg. today I added a list of linguistics terms to help with reading the dictionary and Japanese grammar explanations.

I pretty much consider these pairs like Heisig keywords. Hooks to get the real meaning later. I also keep a sentence anki deck, and words with corresponding sentences are marked. Ideally, %90+ will eventually be marked. It's effective for me because when I get into a word learning zone I can create hooks for new words very very fast. It's effortless, practically. On the other hand, stopping constantly while reading manga is annoying and slow.

For what it's worth, I wish I had been doing this from the start. Of course I didn't have the intensive boot camp training in mnemonics that Heisig gives you then.

Last edited by shakkun (2008 August 02, 8:33 pm)

danieldesu wrote:

shakkun!
I think a person without an extraordinary memory can learn 200 words a day (and by learn, I mean much like we learn kanji with RTK... create hooks).  If we did this for only two weeks already we are at 2800 words.  Doesn't that sound appealing?

It's not how many words you can learn in a day, but rather keeping up on reviews. They pile up incredibly fast for words when you add that many.

But yes, I am all for learning words first then learning the usage details later from context. But, I'll admit I haven't given the sentence method a fair try.

Last edited by radical_tyro (2008 August 02, 8:48 pm)

danieldesu--

I see what you're getting at. I agree with those who said, "Give it a shot and see what happens," then report back with results.

One suggestion I do have though-- instead of creating a whole deck full of vocab, why not just do RTK2? That's full of about 2,300 real Japanese words, tied to kanji!

I'm actually (slooowly) working on a modified RTK2 deck that I'm going to use to eventually replace this site for my RTK1 reviews. I'm also going to replace some of the ... okay, no better way to say it... awful example words used in RTK2 with ones I might actually use someday.

Only downside to RTK2 is the lack of kun-yomi, but that can be gotten around later on. Just add them in as you need them.

Good luck with it and report back to us.

Funny thing, I just read Ray Jay's article where he mentioned that he had studied 1500 Vietnamese words in the first five days.

I did AJATT for awhile before I burnt out. My fault, I should have chosen more interesting and shorter sentences. Nevertheless, I changed my approach because of that. Kind of a mix between AJATT and your idea.
I mix sentences with single vocabulary items (a more detailed description of what I learn can be read on my language log: http://how-to-learn-any-language.com/fo … &TPN=2 )

I actually got some idead from Kato Lomb, a polyglot who taught herself 16 languages. She recommends learning from context (in her case novels), but she did mention that she looked up words to get a feeling for a language in the beginning. And she mentioned that there are unambiguous words like the names of cities and countries, chemical elements, a lot of scientific words, etc.
Do you need a context to learn Japanese names or cities? No, that why I add the names of a few cities and prefectures or celebrities into Anki everyday. And there are a lot of words that you can learn without context. Table, Electricity,telephone,telegram, train, car, train, bicycle, parking lot, student, elementary school, junior highschool, highschool, university, university student,...
And besides the word table, the other words are connected by similar kanji, which makes memorizing easier. I learnt most of these words from context, but I can see the advantage of learning these words in the beginning. If someone could provide a list of 1000-1500 unambiguous Japanese words in a good order  with mnemonics, one could easily learn this list in one to two weeks. And think that this could really accelarate the process of learning Japanese. I have acquired my current Japanese level mainly by context, so I won't say that it works. I just want to say that I see definite merits in your approach. Go for it and give it a try.

I might try it with another language in the near future. Stuart Ray Jay has inspired me to try this.

Yeah, the system in RTK2 is far from perfect, but some of the pure/semi-pure/not-too-pure groupings have their uses, I think. For the rest, it's a matter of brute-forcing, but that's what it sounds like you want to do anyway, so I don't see a problem with it. At least for some of them, you'll have a thin thread of sorts to make picking up the readings a little faster. And in the end, you'll have a firm grasp of on-yomi, you'll be done with English RTK1 keywords, and you'll have a springboard of sorts to work with. (Says the person who needs to stop posting on message boards and get his butt in gear. )

What I *wish* existed was a real frequency dictionary of Japanese. I've seen them for German, but never a proper one for Japanese. (Yes, I'm aware of the file that has frequency data for 5 years' worth of Japanese newspapers, but a proper frequency dictionary wouldn't just consider newspapers... it would cover a mix of media, both written and spoken, to get a better overall grasp of what words come up the most.)

The people who put out KO claim that they use some sort of frequency theory behind their vocab progression, but it's a claim without anything they've actually shown to back it up, so it's hard to guess how they picked these words as being some of the most frequent.