[ファブリス]

I wanted to discuss the Japanese keywords after a recent post from wrightak (from this thread) :

wrightak Wrote:

ファブリス Wrote:Practically, I'm not sure how you can evolve away from the english keywords while still reviewing characters on their own.

What I'm doing is just using a Japanese word instead of an English one. eg. おおきい　→　大きい、　ちいさい　→　小さい、　いく　→　行く、　うらやましい　→　羨ましい etc. I find it makes a big difference. Also, I don't discard the imagery from my story, I try to remember it but associate it to a Japanese word instead.

Have to say after all the discussions on custom keywords, this finally starts making sense to me. Perhaps that was my misunderstanding, it seemed to me that the emphasis on the previous discussions was on "custom keywords" and ability to choose any keyword you like, which I felt is extraneous to the method.

I also remember commenting on another discussion how I felt that someone was trying to do two things with their flashcard : learn vocabulary, and review the characters. I still think that for beginners in Japanese it's better to stick to the original, english keywords from RtK1.

However, for those of us who have completed RtK and have been reviewing for a while wrightak's example above looks really interesting.

We have to keep reviewing because we are not seeing or using some of the characters enough yet, at that stage, the effort spent in retaining the english keywords for the purpose of reviewing, and the sense in which they were learnt (eg. "spring" is water and not "jump"), might as well be spent in remembering a Japanese word.

So wrightak got me thinking. If we can agree to a list for all the RtK1 (and possibly RtK3) kanji, I would be happy to add it to the site.

Some points that came to mind :

a- each keyword should use Japanese reading (kun yomi) if possible, for consistency

b- this point might not agree with everyone, but I feel that the keywords should stick to the english meaning that Heisig chose, in order to allow most users to switch progressively to the Japanese keywords, I think a good example is 懐, one would think naturally of 懐かしい （なつかしい） which is common, but since Heisig went for the "pocket" meaning, I feel that we should use simply 懐 【ふところ】 for example in the sentence : 懐に優しい (budget-pleasing). Even though this may give fewer choices and result in a less common Japanese word, I think the tradeoff is worth it for people doing the transition because at the end of the day we're still learning a perfectly valid Japanese word, right?

I suggest point b, because then perhaps I could add a flag to the DB for each flashcard, and people could switch to the Japanese keyword only, whenever they want. Beginners can use 大きい for "large", but keep the english ones for less common words until they've done plenty of reviewing.

Another reason is that I think people learning new vocab this way will go from Japanese keyword to english translation, to the character. Perhaps this will help brush off the english keyword faster, but I'm thinking the english keyword will remain for a while and if the Japanese keyword uses another meaning, it will make transition difficult, and possibly not worth the extra time?

c- if keyword has more than one kanji (no other could be found), then other characters are written in hiragan, regardless. I also feel thats important to stick to the RtK method, and focus on one character per flashcard.

In my opinion, this is improvement on Heisig's method, not replacement. I don't think this would be best for complete Japanese beginners as I was, trying to learn Japanese vocab at the same time, however for all of us who've been reviewing for a while, this could make reviewing more rewarding.

Thing is, I'm busy with the vocab/sentence review area. But it would be great if we can put together a common list.

I would be very much in favour of this. As I have literally lost the will to study from English keywords recently. I have forgotten a lot of the English keywords for a lot of the Kanji and see them just in Japanese now. To get a list of agreed on Japanese keywords would be ace.

I think we should do it through Google documents, as this can be edited by anyone and we can all work on it together.

you have my support on this fabrise.

I'll add my vote as well. After a couple months of reviewing entirely in Japanese (and neglecting my RTK), I've been finding it difficult to go back to English keywords as well.

ファブリス Wrote:c- if keyword has more than one kanji (no other could be found), then other characters are written in hiragan, regardless.

I don't understand what this means.

This sounds good to me too. Since we're frequently going for the kun-yomi equivalent of the english RTK keyword, it seems that a lot of them are going to be certain without a lot of searching.

shaydwyrm Wrote:

ファブリス Wrote:c- if keyword has more than one kanji (no other could be found), then other characters are written in hiragan, regardless.

I don't understand what this means.

If the character only exists in compounds and not by itself+okurigana, then the other kanji of the compound will be in kana.

A possible issue is kanji that do not exist in any words, only an obscure historical person or place's name. That is no reason to hold back though.

If we're gonna get a standardized list of keywords together for the site, the sooner we push on the better.

wrightak how about posting your master list somewhere that we can all take a look? Seeing as you are already working on this perhaps you could decide on the division of labour...

I reckon this could be a pretty useful addition to the site.

Jarvik7 Wrote:

shaydwyrm Wrote:

ファブリス Wrote:c- if keyword has more than one kanji (no other could be found), then other characters are written in hiragan, regardless.

I don't understand what this means.

If the character only exists in compounds and not by itself+okurigana, then the other kanji of the compound will be in kana.

Oh, I get it. I was confusing this with another potential issue with a few characters - for example　昇る　and 登る - which share 訓読み. I think this might be best dealt with by using a short phrase, for example 日がのぼる, to differentiate from 山をのぼる.

My biggest question is what good does it do to swap out a keyword on a card scheduled, say, six months in the future? Or would switching out the keyword reset the interval for that card? That would kinda hurt mentally if someone has to effectively fail all the cards just to get a Japanese keyword.

Second, I understand that this is supposed to be used by a student ready to take "the next step", but I see it causing a lot of overzealous newcomers to get discouraged by breaking the "divide and conquer" approach that helped so many of us.

And third, (admittedly I am really digging here) lets say that you do make these keywords following your guidelines. Since the image to keyword relationship is so strong (the whole motivation for changing them) couldn't it potentially cause learning speed bumps when someone is trying to go from ふところ to なつかしい?

I don't know really. When I started writing this I thought, "maybe it would be cool". The more I think about it, the more this all just starts to look like a way to turn the RTK method into a way to study vocabulary out of context. I think that Heisig may have envisioned a day when you stopped reviewing your cards and only dug up ones that you forgot from time to time just to refresh your memory. In this situation, assuming you are already reading as much as possible, having the keyword in Japanese would seem to be of questionable merit. I think it would be more productive and easier to add a card for things like 懐に優しい and 懐かしい歌 to your SRS of choice or the future phrase/sentence section of this site.

As someone who has already finished RTK1, what I'd like to see instead, for a writing/yomi exercise, is the most common word for each yomi. So most characters would have at least two words.

Flashcards would be kana/kanji, and be drilled both ways.
kana to kanji: write the word
kanji to kana: read the word

To be the most effective, one should learn the meaning of the words beforehand.
________________________________________

That was my original idea. But since there are something like 6000 common yomi (correct me if I'm wrong), it seems much more efficient just to learn all the JLPT1 vocabulary. There's no need to re-do keywords, as they will drop away. For example, after reading 御飯 a few times, you will think of 飯 as 御飯の飯 instead of "meal".　

Having keywords in Japanese would be awesome! Actually that was something I was thinking about too. I'm willing to help someway, even though my Japanese is probably not good enough yet.

Well, it's fantastic to see that everyone's really interested in this. I think my posts in the "Eliminating the need for English Keywords - help needed" thread were a bit too long winded and didn't sell it to you.

I've been working for a while on the list and I've got another two lists from two other users on the forum. When I get home, I'll post a link to a published google spreadsheet and what I suggest, like yorkii, is that whoever wants to contribute can be added as a collaborator to the spreadsheet.

I think the real debate comes when deciding on what criteria should be used in determining the appropriate Japanese keyword. When I started, I advocated point b of Fabrice's post but after working on it for a while (mainly for my own sake), it made more sense to choose words that I knew really well, whether they're close to Heisig's chosen word or not.

My second suggestion is that we can compile a list of candidate keywords for each kanji, and then we can vote on which one people think is the best, or develop a system where people can choose. For me, simplicity is always the highest priority since I'm always trying to concentrate on remembering the story and the writing of the character and I don't want to get bogged down with any other details.

Another element that I think is necessary, which shaydwyrm touched on, is an example sentence for most Japanese keywords. This is necessary in two circumstances. 1. For two words with the same hiragana (昇る⇔登る　菊⇔聞く）. 2. For words that you don't know yet. After all, there's no point in learning your Japanese keyword if you have no idea about how to put it in a sentence.

EDIT: wrong kanji, doh!

Quote:2. For words that you don't know yet. After all, there's no point in learning your Japanese keyword if you have no idea about how to put it in a sentence.

there is probably going to be some overlap with words that you know and don't know // can use, can't yet use as you get to some of the more obscure Kanji. Also, everyone's knowledge of the language is different even if you studied on he same course. I don7t think we should particularly cut corners in finding a simple word when a more specific word would be suitable. Changing to Japanese keywords seems like it should be considered a more "advanced" thing to do anyway. which, I think the majority of those expressing a positive opinion towards this are.

but yes, sentences should be used as from this you get a much clearer image of the word that is beind used and also, they're hella popular on this board since the AJATT site has been discussed (almost) to death

Far beyond my level at the moment to offer any suggestions to particular keywords, but here's my take on the idea overall:

Offer varying languages keywords selectable in options, not just English and Japanese keywords. It is OH SO SLIGHTLY possible that native Japanese would use this site to improve their own Kanji ability if that were the case. Why not offer the same to Germans, French and Spanish (are there any other languages Heisig translated to?).

Yes, it'll create more non-English stories which may mean shared stories can be tagged in their parent keyword. Someone like me can select to view only the English keyword shared stories.

Dilando: If you use the Japanese keyword, it's likely you'll start from scratch with a new account. Most likely you'll blaze through it though.

Nukemarine Wrote:Offer varying languages keywords selectable in options, not just English and Japanese keywords. It is OH SO SLIGHTLY possible that native Japanese would use this site to improve their own Kanji ability if that were the case. Why not offer the same to Germans, French and Spanish (are there any other languages Heisig translated to?).

I have nothing against including German, French and Spanish keywords on the site, but I think that they should be put in the same bracket as English keywords. Changing to Japanese keywords is very different and should be considered a different project all together. I think the probability of a native Japanese learner using the Heisig method as it currently stands is so negligible that the possibility can be dismissed.

dilandau23 Wrote:My biggest question is what good does it do to swap out a keyword on a card scheduled, say, six months in the future?

Assuming I understand what Fabrice is proposing, the next time you come to review the card you would then have the option to review using either the English or Japanese keyword.

dilandau23 Wrote:I see it causing a lot of overzealous newcomers to get discouraged by breaking the "divide and conquer" approach that helped so many of us.

It's like smoking, everyone knows it's bad for you but some people decide to press on regardless...

dilandau23 Wrote:I think it would be more productive and easier to add a card for things like 懐に優しい and 懐かしい歌 to your SRS of choice or the future phrase/sentence section of this site.

I agree with this, but changing the keywords to Japanese and having phrases in your SRS isn't necessarily the same thing.

yorkii Wrote:sentences should be used as from this you get a much clearer image of the word that is beind used

I agree that it's probably more beneficial for one's productive language skills to learn collocations and expressions. But isn't that why we're using an SRS?

ファブリス Wrote:懐 【ふところ】 for example in the sentence : 懐に優しい

I'm not sure if Fabrice is suggesting using just the word or the phrase. However while there is obvious merit in learning phrases, often the rest of the phrase will help with the kanji we want to produce. So personally for this exercise, I would be in favor, as far as possible, in having Japanese keywords as opposed to a 'key phrase'...

[ファブリス]

I'll post more when I have time but please keep in mind that it's not an exercise in learning on yomi; the goal is to help moving away from english keywords, help distinguish nuances of the kanji by using directly an associated Japanese word.

leosmith exercises for writing and on reading can be included in the next section. I already have actually a working review mode with a random kanji in the compound written in kana and underlined, and you have to write down the character, effectively a purely "active writing" test.

Most keywords would be kun reading, all if possible I think that would be best to avoid confusions for learners moving on to on readings afterwards.

synewave Wrote:It's like smoking, everyone knows it's bad for you but some people decide to press on regardless...

I don't follow. What's bad for you in terms of Japanese keywords?

To dilandau: I think the question you need to ask yourself is are you going to continue reviewing the RTK kanji? If so, are you going to continue reviewing from 'big' or 'おおきい'?

If you think you can forget about all of your RTK imagery and go straight to inputting words and phrases into your SRS then that's fair enough. But if you'd like to keep reviewing using the RTK imagery then you can either continually associate the kanji with English words like nightbreak, wink etc. or you can change to Japanese words.

I don't see why the divide and conquer approach needs to disappear just because your prompt is Japanese rather than English. The key elements remain: one label for each kanji and vivid imagery in your mind to help you remember the kanji's shape. Yes, you need to learn a single word for each kanji but you'll have to do that anyway at some point.

wrightak Wrote:I don't see why the divide and conquer approach needs to disappear just because your prompt is Japanese rather than English. The key elements remain: one label for each kanji and vivid imagery in your mind to help you remember the kanji's shape. Yes, you need to learn a single word for each kanji but you'll have to do that anyway at some point.

Wrightak, I think this is a great idea but also think it will be most useful for learners who have finished RtK1 or have come to the book later in their Japanese study than the norm (and thus have a larger Japanese vocabulary to work with). Heisig somewhere wrote that he started on this process after he had learned the most elementary Japanese. A person in that position would find it daunting to learn Japanese vocabulary AND the kanji at the same time: for such a learner, learning the Japanese key word at the same time as the kanji WOULD be adding an extra layer to the process and run contrary to the 'divide and conquer' approach.

On the other hand, such a system would be great for Japanese returnees and Japanese school kids (in addition to the intermediate and advanced non-native learners that you have in mind).

A bit off topic: reading this branch of the forum from beginning to end has demonstrated how great a community has assembled around this site. One member threw up a bright idea and then everyone pitched in with questions, ideas, and refinements. It is exciting to watch.

wrightak perhaps the smoking analogy wasn't the best, however my post was in response to what dilandau23 was saying about people perhaps going straight to Japanese keywords as opposed to following Heisig's 'Divide and conquer' maxim.

Don't get me wrong, I'm all for Japanese keywords, but someone coming to Heisig with no Japanese experience would most likely have an easier time of it using keywords in their own language.

If Heisig intended us to use Japanese keywords from the start, he would have written his book using them! However as many of us are 'graduates' moving over to Japanese keywords seems appropriate.

Post your lists and let's take a look

EDIT: Just read PParisi's post properly and he said pretty much what I was trying to say!

[ファブリス]

wrightak Wrote:For words that you don't know yet. After all, there's no point in learning your Japanese keyword if you have no idea about how to put it in a sentence.

I understand, but that's already going further than what I was considering... and towards dilandau's reservations.

In RtK Vol.2 Heisig says you have to learn the words at one point or another anyway, in regard to the words appearing out of context. Except for words with identical readings, I would stick to a simple keyword.

I still have some reservations about this. But I think the discussion and the list can be fruitful either way : perhaps as dilandau points out, this would be best handled as part of the sentence/vocab review update.

If so, then it could be a starter "vocabulary list" available to all, and interestingly.. if you reviewed this vocab, you would be reviewing the kanji at the same time, hence you might move away completely from the current RtK1 review area, at the same time continuing to review the characters.

In fact, looking at the end of RtKVol2, Heisig mentions possible ways to review with the flashcard layout depicted in the book :
1) is from meaning to writing, which is what we're doing on this website
2) is from exemplary compound to on-yomi ( i.e. 弾丸 -> flip card -> ダンガン )
3) is from inflexted kanji to kunyomi ( eg. 弾く -> flip card -> ひく )

But he says : after working your way through RtK1 you may feel you no longer need to do the drill (1), in which case he recommends reviewing directly from the hiragana kunyomi to the writing : ie. たま -> flip card -> 弾

Which, is exactly what this list would be!

How does that sound ?

So in summary, if I add this list to the next vocab review area, you will be able to review with Japanese keywords. In addition as it would be a regular vocab list, you would be able to go and change some of the entries to your liking (swap it with any other word from the dictionary).

[ファブリス]

I think this would solve the problem of review intervals that dilandau mentionned. Vocabulary is new information, so the reviews would start from scratch I'm afraid, but at the same time they would replace the RtK1 reviews if the list is correctly designed to include one Japanese word for each kanji.

昇る and 登る is a problem that makes me think that a special review mode might be needed for this, because the question side of the flashcard would be identical if the hiragana of the word alone is shown. Vocab lists, as opposed to sentences, are only compounds.

So.. maybe what we really need is a kun-yomi review mode to supplant the RtK1 review, but something separate from the RtK1 review. RtK1 newcomers review with english keywords. Once they have completed the first volume they can move on to the kunyomi review and learn 2000 words at the same time. That would keep most learners busy for a while, although a good exercise would be then to lookup sentences with each word, and add those to the sentences section.

The 昇る vs 登る problem could be solved in a couple different ways:

1) Present the kanji in a sentence like someone mentioned earlier. This can create a problem of introducing too much other information that the learner might not know (such as grammar, other words, etc)

2) Present a bit of additional information, for example it could be "のぼる (as in the sun)" and "のぼる (as in a mountain)" - just enough information to distinguish it from other similar characters. You could even include the english keyword as part of that information if need be. Remember that this isn't vocabulary drilling, it is still kanji drilling, just with Japanese keywords. There will still be the readings / vocabulary sections which we're all waiting for anxiously

3) Where possible choose a different common word that uses the kanji, possibly a compound. For example 登録 or 登頂 (I chose 登 to be the compounded one since it has better/more common compounds available than 昇, although you could argue to keep 登る since it's more common than 昇る, and use a less common compound for 昇). In this case we have no choice but to use a compound since there are no reasonable alternatives that are just kanji+okurigana (昇り＆登る wouldn't be much better would it?)... This option would obviously frequently change the Heisig keyword even in instances where it wouldn't be the best idea to (see comment below), as well as introduce many compounds which I think we can all agree are not a good idea.

Clearly (to me at least) #2 seems like the best option. Someone earlier mentioned that we should try to keep as close as possible in meaning to the Heisig keywords (eg pocket vs nostalgic). imo this is a lot of extra work for very little benefit. Many of the Heisig keywords do not accurately represent the common meaning of the characters, and to fulfill the requirement might mean bringing in more compound words than we'd like. Let's stick with the most common (preferably kunyomi) non-compound word.

The example we've been using is actually especially hard because they both have the exact same meaning, infact both of my J-J dictionaries list those two (along with 上る) under the same entry with a note at the end saying how they are USUALLY used, after several pages of shared definitions (my J-E lists them as separate words along the lines of their conventions, but J-E/E-J dictionaries are never good tools to learn nuance). They tend to be used differently conventionally, but they are in fact interchangeable. EDICT uses an identical entry for each of the three words without listing the conventions, which is one of the major reasons I dislike it

I think most of our matches will have more obviously different yet related meanings (早いvs速い) or totally unrelated meanings (暑いvs厚い). This kind would be suited nicely with option 2. ex: "あつい as in thickness" vs "あつい as in temperature"

Jarvik7 Wrote:I think most of our matches will have more obviously different yet related meanings (早いvs速い) or totally unrelated meanings (暑いvs厚い). This kind would be suited nicely with option 2. ex: "あつい as in thickness" vs "あつい as in temperature"

You're right. Something will have to be done with the identical 訓読み as well as kanji without 訓読み。I think your 'option 2' is most appropriate.

Perhaps "あつい as in book" or あついほん might work. Also "あつい as in temperature" might make one think of 熱い as well as 暑い。Perhaps あついなつ and あついコーヒー or similar would be better?

Part of the beauty of English keywords (whether you agree with them or not) is their standardized nature.

Moving over to Japanese keywords it seems like there will, out of necessity, be a variety of "keywords", 訓読み similar in meaning to Heisig's; 訓読み but with a different meaning; 訓読み with extra info; compounds.

Where possible using a 訓読み keyword faithful (or not) to the Heisig keyword is my preferred option.

by using appropriate sentences (not just things like "あついほん") like I suggested originally, one will know which kanji they need to write.

この用紙はあつい。
vs.
今日、あついっすね。

「うわ、温かい、このスープ」「温かいどころか、あついですよ！

for example.

What Fabrice said in his last post sounds like a much better way to do this. I really do understand everyone's desire to move away from English keywords, but all one needs to do is look at nac_est's post to see the perils involved. (Not attacking you nac_est, maybe you could take on Japanese keywords, but at the same time perhaps others would get overwhelmed). Also, a lot of people seem to have different ideas about how to go about doing it. There is a common American(I think) saying, "if it ain't broke don't fix it". I think it really applies here. All (or at least most) of us learned using keywords in a language very familiar to us. It worked. Why change something that worked? Just think about all the wrong ways some of you went about studying kanji until you found RTK.

Build something completely separate and new is fine, but don't tempt newcomers to take on something they might not be ready for. Yes, yes they can decide for themselves I know, but like Heisig, we need to aggressively warn them that the path they take is wrought with peril and wash our hands of it.

Jarvik: I don't agree that it is not vocabulary drilling. For common words like 大きい and 登る, I could see how you would feel like that but what about something like 冴える or even something like 且つ. And this brings me to one last issue I foresee, with a straight port to Japanese keywords. Unless you change the structure of the cards to take into account inflections, is it correct to think of 小 as 小さい? If you include the inflections then it is most surely vocabulary drilling.