Cute, not really useful however. Keywords are meant for giving you a key to remembering the kanji for testing, you're not actually suppose to review and remember the keywords.

When learning new compounds, remembering the keywords can be of benefit if there is an obvious result of combining the two keywords:
微笑　びしょう    (smile)　     "delicate laugh"
引力　いんりょく (attraction)　"pulling power"
余暇　よか        (leisure)　   "too much spare time".

I would agree with Tobberoth and expand on it a little bit.

I think it's better to use the keywords as just a key to help you write the Kanji.  You really need to expand on each kanji's "keyword" definition and add other meanings, usages, nuance.  Becoming married to the keywords, which I think this manner of review just strongly reinforces them, will keep.  What I personally liked to do was use an anki deck that tested keyword -> Kanji (writing) ability, then on the answer card had lists of all the common words that kanji was used in (and their English translations).  This gives you a very broad sense of usage for each kanji.  Nowadays, when I look at a Kanji and am trying to recall its meaning, I usually come up with a list of associated words rather than just one single keyword.  And since your writing ability eventually passes from even relying on the keywords to be able to write it (and just pure recollection / motor memory), the keywords usefulness is rather short lived. Eventually the kanji becomes much less a "keyword" and just a kind of general feeling/idea/concept. 

Not only that, but Heisig also mentions that reviewing in this manner (Kanji -> Keyword) isn't so useful as this information is generally gained automatically through normal review (Keyword -> Kanji).  While reviewing like this might speed up your recognition ability a bit, I think your time would be better spent reviewing how to write them.  Also, I would be worried about spending a lot of my review time typing answers in English.


I do like what bodhisamaya said though...  it is pretty useful to remember compounds like that.  But I think you gain the ability to do that without going kanji -> keyword.

I think Tobberoth was just trying to be nice and save someone tens/hundreds of hours of review time on a method he doesn't see as very useful.  If I posted a method where I suggested learning vast amounts of Japanese using romaji before learning anything about written Japanese, I would be scathingly criticized.     I think it's a similar situation (criticism of what he sees as a failed methodology), though his comments weren't even scathing - just a little blunt.

To the original poster - the app is pretty cool, though.   I like how it brings up the wrong kanji in red if you make a mistake and confuse the keywords for two kanji.   You should also give the answer button a shortcut key and keep the focus on the text box, to keep people from having to click in the text box, type an answer, click "Answer," then click back in the text box.

Last edited by Ben_Nielson (2009 March 01, 7:23 pm)

smujohnson wrote:

It's good.

Tobberoth, quit being such a stick in the mud.  It's definitely not the first time that you are anti-something against something someone has written.  Why don't you start a thread about how the Kanji poster is all wrong, saying it should show you the keywords instead, so you can practice writing the kanji?

My guess is the only reason you are saying this is because Heisig said in the intro that you should practice with the keywords first only.  I suppose if you are going to take the attitude that is whatever Heisig says is law, I suppose you will be too closed minded to benefit from a simple practice that this guy has written up, after having mastered this sites keyword -> kanji practice?  I have been looking for a way myself to go from the kanji to keywords, as for some kanjis, I can't go the opposite way whereas I can get them from the keyword.

Oh I'm sorry, I thought this was a discussion forum and that someone posted his idea to get feedback on it. I didn't say you aren't allowed to review kanji -> keyword, I just said it's a waste of time since it's way more important to know Japanese keywords than English keywords. I guess I shouldn't assume people are interested in learning Japanese just because they learn kanji, eh?

"The only reason I'm saying it" it because I know it's true. When you see a word with 耗, you CAN think of decrease. Or, you could think of 消耗. Not only will thinking of "shoumou" help you understand the meaning, it will help you understand the reading as well. And, it will help you differentiate between 消 and 耗 which have similar meanings. It will also be an extra word for your vocabulary instead of a connection between an english word and a japanese kanji.

Last edited by Tobberoth (2009 March 02, 3:21 am)

From my experience, just my own, it takes only one or two times kanji > keyword before the keyword sticks. From this I gather that the time being "wasted" is minimal. If it takes longer to get the kanji to stick then more time is being wasted and I move on.

tobberoth wrote:

Cute, not really useful however. Keywords are meant for giving you a key to remembering the kanji for testing, you're not actually suppose to review and remember the keywords.

There is a difference between providing useful feedback and responses like this. You can be "right" without this sort of tone, Mr. Tobberoth.

Oh I'm sorry, I thought this was a discussion forum and that someone posted his idea to get feedback on it. I didn't say you aren't allowed to review kanji -> keyword, I just said it's a waste of time since it's way more important to know Japanese keywords than English keywords.

There are many paths up the mountain, no?

TS wrote:

I'm going to make a wild guess that the people bashing Tobberoth don't have the amount of experience in Japanese that he does.  Studying from the keyword to the kanji is not an efficient use of your time.  That time is better spent studying actual Japanese.

It matters more how you say something than what is being said.

Example:

Read your quote [TS] in this post then read his[MrT].

Tobberoth wrote:

I guess I shouldn't assume people are interested in learning Japanese just because they learn kanji, eh?

You'd be surprised... or not... people are straaaaaaange... lol.

Tobberoth, in my opinion, you tend to use absolute terms in your statements. This creates a mood of "any other option is pointless". To me, everything you say applies only to you and might apply to others including myself. However, you say in a way that it must apply to everyone. Hence, the stick in the mud moniker. If people get tired of reading what you write, then you cease communicating, even when you provide viable wisdom.

On topic

Well, Tobberoth, when you say "waste your time" do you forget about what an SRS does? If "keyword to kanji" is so easy that it's a waste of time, then wouldn't an SRS space it out so far that it becomes a moot point fast?

Now, in time you're adding Japanese words. For those, should you create recognition type cards, you recreate this effect in the best way. Plus you get the all important On and Kun reading reinforcement. Your post says the same.

But here's a problem: I'm not going to add a word or two for EVERY kanji in RTK or RTK3. I'm adding vocabulary via frequency, so the rarer kanji will not get covered with my controlled study. However, they will be appearing as I read and watch tv (with subtitles) alot. This all important learning by osmosis will lose punch if I don't recognize a kanji.

So yeah, keyword to kanji is moot should you get 3 or 4 vocabulary words to reinforce on and kun. But for rarer kanji that just will not come up in formal study outside a 20,000 word vocabulary list or lots and lots of names, it just is "more useful". I'm going to recognize a majority of the kanji without keyword to kanji review. There will be those that slip through the cracks, which becomes apparent as you look at kanji from random media.

Anyway, if anyone is really interested in Keyword to Kanji review, it's best to use Anki. Yeah, I like RevTK so much that I put off doing that (sorry, the ease of using stories is WAY, WAY to great, we love you Fabrice). But the option is there.

Tobberoth, I know everything you say is your opinion. I'm saying why it gives the impression of "stick in the mud" to people. Granted, if it is opinion, why should the other party even bother to prove you wrong?

Your second paragraph brings zero sum fault up again. It's not one way or another. Learning Kanji via Heisig is not learning Japanese. However, I'm not wasting my time better spent on learning Japanese by using RevTK. The little investment of time on RevTK improves the time I spend on Japanese. I am capable of doing both.

The other problem, well, depends on a point of view. I'm for kanji RECOGNITION. Now, for me, I take that as being Kanji to Concept. If I see a kanji and don't know what it means, something is wrong. This has happened quite a few times few, where a mature kanji (stack 4 or higher in RevTK) is easy should you see the keyword but seeing the kanji draws a blank. Others have also stated this. That is a definite problem is my book.

Recognition is the easier of the two. However, there will be some kanji where the recognition is harder for whatever reason. Now, how do you know which is which? The beauty of an SRS is it would catch them both if you tested both ways. The easy ones space out quickly, and your left with items giving you issues. That's not a waste of time.

So yeah, if you force the discussion to be SOLELY kanji to keyword, I have to agree with you. Such is the strength of strawman arguments. However, kanji recognition (kanji to concept) is not a bad thing to train. The keyword is great when doing production via flashcards, it is not great for recognition.

But hey, you can waste your time adding a vocabulary word for every kanji, even though it won't be a word you'll use or see, and learn stuff out of context. That's your time being wasted, not mine.

Last edited by Nukemarine (2009 March 02, 1:04 pm)

Tobberoth wrote:

kazelee wrote:

It matters more how you say something than what is being said.

Yeah, I'd definitely disagree with you there.

Analysts wouldn't. Research doesn't. The meaning of communication is the response you get, much like Nukemarine is trying to explain to you.

Past RTK, we're all approaching our path to fluency along parallel but separate paths. I didn't post above, but I did try recognition with Japanese words (see Japanese keyword thread). Problem began to arise that I was not ready for some of the advanced words. Plus, some words were picked that did not work for me.

Remember, I didn't start RTK already knowing Japanese, whereas you did. So for you, Japanese keywords would be perfect. For me, they would be a hindrance (not now, then). Parallel paths at work.

Plus, because of my personal study choices, I'm not adding vocabulary words to cover every kanji in RTK1 and 3. I study RTK1 and 3 so that knowledge is there when the words with them happen to pop up. Sometimes that's in formal studies (adding grammar or vocabulary sentences), sometimes in informal situations (reading, watching, walking around town when I lived in Japan).  Thing is, in informal situations, I do not rush to dictionaries or write down words I don't know. It's just not my thing.

Anyway, we do seem to agree: Kanji to keyword is not efficient. Now, if you did kanji to keyword and marked the kanji correct should you be close (similar word or concept), then nothing wrong with that in my opinion as I've been harping Kanji to Concept for a year now. However, I like it better with how Anki does double cards with spacing. So the problem with Kaiser is that you're doing two study methods for the same kanji. You're going to corrupt the benefits of SRS a little bit that way.

Togil, I'm pretty sure most people here will disagree with you on that. The ability to write the kanji by hand is a bonus, that's not why you learn it that way. The thing is, to be able to write a kanji by hand, you have to know it perfectly. That's the goal: Know the kanji so well it becomes second nature. Just going kanji -> keyword will never give you that skill, which is why students of traditional methods are usually very bad at kanji. They mix them up, they forget them, they can't write anything without a computer telling them the right kanji. I would know, that's how I learned kanji first.

The intricate knowledge of the various parts of the kanji and how they cooperate to create the symbol is what one needs to become fluent in Japanese, both writing and reading. Simply recognizing shapes won't ever make you good enough.

togil wrote:

Did I already mention that my Kanji Kaiser program can also be used as a dictionary? Just copy a kanji to the input box and press enter, then you see its Heisig meaning.

Or you could use a real dictionary. Tcharam!

Sorry buddy, your intentions are good, but your ideas are old and they don't work well.
You need to be able to produce kanji from memory to understand them. It's not something you get only by knowing some keywords, it comes from daily real usage.

We are aiming high here in this forum. For instance, JLPT1 is the bottom line, not the final goal.
But if you indeed met someone that reached a highlevel of japanese or chinese only by recognizing the kanji, tell us the story.

I tend to agree with Tobberoth, but I would add a caveat.

I'd rather spend the time learning at least one vocab word that uses the kanji, but as I'm only about 900 characters in, I've been holding off on really getting into vocabulary (I did study some in college, so I know a bit).

I wish there were a Kanji-Japanese Keyword "style" list.  I know people have done some full lists like this, but what I mean is something I've mentioned in other threads.

I'd love a purely recognition deck where you could un-suspend cards and they would only use keyword vocabulary up to the point you were at.  I'm not sure if it's practical, but for example, if you had 浴 in Heisig, an example vocab word could be 混浴 since you already learned 混.

I wish I had a list like this, but when I look up a kanji in Sanseido or something, it's hard to decide what's a good example.  Someone with more Japanese knowledge than I would need to do it, I think.  At least one word or compound per kanji would be very helpful to me, at least.