Hello everyone, I'm new here so please let me know if I'm doing this forum thing wrong!

I've been using Heisig's book as well as Anki for flashcards, but the deck I'm using shows the meaning on the front, and the kanji on the back. Is this the most useful way for me to do it? I find that I can conjure up the kanji in my head if I see the word, but if I then see the kanji, I can't always think what the meaning of it is. Would it be better for me to have the kanji symbol on the front of the card, and learn to extract the meaning from it that way round? Since that's how the reading will be done, when I can actually read Japanese, is it not logical to learn that way around?

Please correct me if I'm wrong, or give any advice which should help me along, because I'm rather struggling at the minute and I'm only 150 or so in!

Many thanks!

You still want Keyword to Kanji. Vocabulary starts to fill in the gaps if any. Finally, if you move onto sentence mining or doing subs2srs or some other type of reading comprehension study, that'll fill in even more. Stick to production style (keyword to kanji) as that's a foundation to help when you move on to vocabulary that uses kanji.

If you did have reverse cards (and Anki offers that option), you'll find they're really easy for the most part. Passive or recognition style review are known to be easier since you're calling on passive memory (sort of like you can read tens of thousands of words and know their meanings, but you'd be hard pressed to come up with 1000 unique words from memory). You're adding on to a bit of your review time doubling effort on to ensure you know that a Kanji happens to have a particular Heisig keyword.

There is one problem in that Kanji actually have more than one meaning. Heisig sort of cheated by creating Keywords, but it was necessary to create the production testing using one's native language. Plus, those meanings or flavors start to reveal themselves via vocabulary as stated above.

TL-DR: Stick to Keyword to Kanji. It's an important skill to build when you do learn vocabulary later.

I'm gonna have to disagree with Nukemarine here.

I switched to kanji -> keyword towards the end of RTK3 and never looked back, as the advantages were immediately obvious. Note that my current goal is reading only, so ymmv:

- Reviews take a fraction of the time.

- Readings/vocab becomes easier to learn as recalling kanji meanings faster gives you more mental 'hooks' to hang things on. I'm pretty sure this improved my reading comprehension.

- Heisig's own advice to review from keyword to kanji is IRRC a single throwaway sentence in the book. He goes on to claim that kanji -> keyword will "take care of itself", presumably through random exposure. The whole point of using SRS is that random exposure is inefficient.

- Reviewing kanji -> keyword (i.e. meaning) means you don't have to care about Heisig's ridiculous synonyms / obscure words chosen to keep the keywords unique. No more secret/secretive/secrecy nonsense, no more march/parade/fiesta insanity, no more block up/stuff up/plug up inanity.

- You can easily add kanji yourself without having to care about keyword conflicts. I'm currently up to ~3250. Irregular kanji are not a problem - their distinctiveness often makes them easier, in fact.

- "Because it's harder" is not a persuasive argument for doing something.

- You still get to use all the Heisig-style mental imagery / stories you want involving the components of the kanji, with the added bonus that your imagery is focused on the *meaning*, where it should be, rather than the components.

- AFAIK there is no real-world situation in which going keyword -> kanji is a useful thing to be able to do. When writing, *choice* of kanji is as much a problem as writing them, and Heisig keywords are not going to help with that (though I'm speculating here as I don't write freehand and probably never will).

- etc.

HTH

Coming up with a kanji after seeing a keyword or vocab item enables you to write the kanji. That's the goal of the Heisig books. Reading is another skill that you'll learn alongside automatically.

Now, if you don't feel you want to bother with writing, the other way around might work, as stated above me, but if you want to learn to write them all, you do it in the order Heisig mentioned.

It's not a throw-away sentence, it's a clear instruction to reach a clearly defined goal. Different goals don't make those instructions invalid, they just don't apply anymore, and in that case, you're not doing "Heisig" anymore. Your system might consist of the same techniques to some degree but overall, it's something different.

I agree with the above two posts, with a large caveat.
for most people there will be an ocean of time between when you start learning kanji, and when you might decide to leave a note in japanese on a table for a friend, or write a hand-written letter.
3, 5 maybe 10 years.
Going keyword to kanji the first time you learn RTK is great because it really cements the understanding, but continuing to rep those cards for years, especially into RTK3, is masochistic. the writing benefit is not worth it IMHO.
I plan on maybe doing rtk again, keyword to kanji, in a couple years maybe, once all the other tasks more important than handwriting have been ironed out.

it's so low on the list of priorities, seems like a waste of effort in the beginning when time and energy are really valuable.

Well I can actually put the writing knowledge to use, you know. I have to write reports by hand, and doing it in all Kana is tough for my colleagues. However, I'm still busy doing RTK, not in the followup-phase yet, where I'd fill in the holes with Japanese vocab. I may very well know the words and the Kanji and can't make the connection yet. For someone in my situation, the writing skill is extremely useful, and in my case, I think that production is the best way to do it for me.

May I suggest following James Heisig's advice on the matter? Since he's the guy who wrote the book?

Obviously, once you learn them well enough, you can review the Kanji in whatever order you like. Probably not a bad idea to switch to recognition (Kanji->keyword), so that you get used to reading Kanji. I made that switch about a month after I finished RtK.

But, while you're learning, going Kanji->keyword will be useless. You'll be relying on visual memory instead of following the Heisig method, and hopeless confusion will set in by the time you're 6-700 in. Producing the Kanji, from the keyword, is what's needed to help you make a habit out of following the method and relying on your stories rather than visual memory.

Thanks a lot for all replies, I've read them all and considered all the ideas carefully and decided to do what the majority suggested and do Kanji to keyword(s), just for being able to read it (I'll work on the writing part later and possibly regret my decision but that's for future me to be angry at past me about) as it seems more convenient and appropriate at the time. I do understand the long-term benefits but they seem less useful for my particular goals than what the short term benefits will allow me to accomplish.

Thanks a thousand times over to everyone who answered though, all of you helped me out a lot


Last question, does anybody know what I'd search for in Anki to find a 'flipped' heisig deck to learn it kanji to keyword? Thanks!

Hi Sming, you're welcome

Easiest way is to edit the card template.

Select the deck, hit 'browse', then 'cards...'. My settings are below. The colour is pretty revolting but easy on the eyes ...

Code:

Front Template:

<span style="font-size: 250px; ">{{Kanji}}</span><br>


Styling:

.card {
font-family: arial;
font-size: 50px;
text-align: center;
color: black;
background-color: tan;
}


Back Template:

{{FrontSide}}
<hr>
<a href="http://kanji.koohii.com/study/?framenum={{text:Heisig number}}">{{Keyword}}</a>
<br>
<br>
画数: {{Stroke count}}, Nr: {{Heisig number}}

Like a few other people here, I'd have to also agree that if you ever want to write kanji, kanji->keyword isn't going to cut it. Recognition leaves you with a sort of passive, surface ability to recognize characters and will frequently leave you confused when confronting similar kanji. If I were doing recognition only, I wouldn't even bother with Heisig.

If the general goal is some level of fluency, I see no reason why being able to write should be dumped on the side of the road. Even worse, many people on these forums dump even speaking the language they study as well. Active recall plays a huge role in both, and it's not as if these skills all exist independently of each other. Active recall of both vocab and kanji makes recognition easier by miles. Yes, it's more work, but it's also a deeper, more well-rounded understanding that enables you to do more with the language than read manga. Call me crazy, but it seems to me that the massive investment involved in learning Japanese ought to have some payoff aside from picking up a book.

I second that. See, I was a bit lazy lately and did my reviews very quickly, was satisfied when I'd flip the cards and the kanji were not totally alien and I remembered the stories, figuring I'd be able to remember them when I had to.

I then reverted back to my old style:
See the keyword -> write it with pen on paper -> flip card to check

And I realized I had only a very hazy memory of the last maybe 100 I learned, so I'm currently going through a very harsh "fail everything you can't write properly" schedule, because let's face it, learning in a half-assed way is just not worth it. Might as well skip the kanji completely.

I'm now back to my old learning style, and it works again. It's solid. It's tough, but it's totally worth it in my opinion.

I know them SO much better from doing production work, keyword->kanji, better than I ever did the other way around.

I'm a believer ^_^

Sming Wrote:Thanks a lot for all replies, I've read them all and considered all the ideas carefully and decided to do what the majority suggested and do Kanji to keyword

Going with the majority can't fail...or can it?

At least read the book's instructions before you embark on this journey. You might find that Heisig understands his own method a little better than the majority of people on the Internet.

There is already a good discussion here, but I just had to toss in my own two cents. Let me preface this by agreeing with the OP that I have not found Heisig's claim (that if I just do keyword to kanji then the reverse will just work itself out) to be true. I'm saying this after 2.5 years and over 1700 kanji (I've been interrupted a lot). I'll try to be concise in addressing some of the arguments that each side is bringing up.

First, anotherjohn's argument that, "'because it's harder' is not a persuasive argument for doing something": I agree. But NOT doing something BECAUSE it is hard is also known as "cutting corners," which most people agree is a bad plan when learning. Be careful you aren't conflating "difficult" and "inefficient" in the word "hard." Writing kanji out hundreds of times is a difficult *and* inefficient way to study. Heisig's keyword-to-kanji is kind of hard sometimes, yes, but also pretty efficient. I, for one, don't get a great sense of accomplishment from doing things that are too easy. Recall Kennedy's call for landing on the moon: "We do so not because it is easy, but because it is hard."

Second is the, "But I don't foresee ever having to actually write these," argument: Compelling enough, and I (generally) agree. In the age of computers, even many actual Japanese adults can get away with almost never needing to write them. This method was developed in the 70's, when writing was absolutely essential, and the cover of the book explicitly states that it is for remembering the "meaning and writing of Japanese characters." Others in this thread have nicely summarized why you can never be sure when you may have to be able to write them out (for instance, living in Japan I had to write out my address several times on forms at the bank, when applying for a video rental card (obsolete now, perhaps), and when special ordering a book at the bookstore).

And of course, as others have argued, doing it the "hard way," improves recall. Anotherjohn said he switched to kanji-to-keyword at the end of RtK 3. I think that's probably a prime time to do so, but just because it worked out great then doesn't mean it will work out great on frame 173. He already had TONS of study "the right way" to build up his understanding. The OP does not.

As I said at the top, I find that many characters are still very hard for me to attach to keywords when seeing them in context, and this is very frustrating. However, that just means I need to keep at it. And it probably also means it's time to start working kanji-to-keyword as well. For various reasons I won't get into here, I make paper cards for every character, so I sometimes use those to study in reverse, but I also recently decided I should use an Anki deck that way as well, while still learning new cards (and reviewing old ones) here on RevTK in the traditional sense.

Sorry this was so long. I know the OP may have already made up his mind, and that's his prerogative, but I hoped this might add something useful to the conversation for any others who might come across this thread in the future.

Stansfield123 Wrote:At least read the book's instructions before you embark on this journey. You might find that Heisig understands his own method a little better than the majority of people on the Internet.

That's not necessarily the case. As Heisig has pointed out himself, he has no pedagogical training nor has he done any sort of study to show the effectiveness of RTK vs. other methods, or depending on how you use it. Heisig developed a method for learning kanji that worked for himself, and then on the urging of others he published it as a book. His introductions to both books are questionable in a number of places.

But the goal is not to write English keyword to Kanji. Like Nukemarine wrote, vocab will fill the gaps sooner or later. Heisig is only a stepping stone.
If keyword to Kanji is more effective for recalling the Kanji and the goal is to recall the Kanji, I see no reason to do it the other way around. You'll start to confuse them sooner or later, and forget some completely, because they have no meaning in a Japanese context. That's why you do it the hard way and then learn vocab.

There's no meaning to learning English keywords.　Learning to read Kanji and get the keywords is not what you're trying to learn. In the end, what you will do, is reading Japanese in Japanese, and RTK only prepares you for that.

Reading them in English is an unneeded skill that in the end might even be counterproductive.

How about keyword to kanji. Then after that, you can do the core deck for your vocab using recognition(JPN -> ENG). That way, you might be able to recall the meaning of the kanji by using the english word/s to integrate it into the kanji and kana's meaning, whilst at the same time having no difficulty in writing it.

Personally I'm not able to have a clear image of what a kanji looks like until I learn how to write it. I've also found production always sticks to my memory far, far better than recognition even though recognition takes less time to review (basically because I'm writing the kanji as I review them - if I didn't then production would just as fast.)

You're supposed to switch to vocabulary and eventually stop reviewing kanji at some point, though (and again, I think that with vocabulary production is a better choice because you always want to be able to find the word you're looking for, and simply because for me it sticks better that way, but ymmv as always).

You need to do both. Recognition and production are two completely different skills, and production has the additional difficulty of requiring fine motor control and muscle memory. In other words, if you want to be able to write, you have to practice writing. There's just no way around it.

The key to this controversy is in the first paragraph of the introduction to the first volume. The first sentence talks about "correlating the writing and the meaning of Japanese characters..." Twice more in that paragraph Heisig refers to writing, but never to reading. The first paragraph is his thesis statement: this book is about learning to associate writing with meaning. Hence, keyword to kanji, because that is the order of writing. But if your goal is to read Japanese, then keyword to kanji does not make sense because that is not the order of reading; kanji to keyword is. Take your pick. And there's no rule saying you can't study them both ways.

I have found that the best method for me is...

...neither.

What I do is I go through each kanji in the book and on the koohii page, and get my favorite mnemonic into my head.  Then I go to jisho.org and look up common words using that kanji and its onyomi, then add at least two of those words into my anki deck.  I then do recognition reviews on those vocabulary words rather than reviews on RTK kanji to keyword or vice versa.  This ensures that I am learning both useful vocabulary and the kanji's reading for when I encounter it in future words, rather than learning an abstract keyword that may or may not always correspond to the kanji's actual usage.

That isn't to say that RTK is not useful, quite the opposite, the only reason I am able to have decent retention is because I have the mnemonics to back me up if I forget what a kanji is supposed to imply.  I just have the mnemonic lead me to a general concept rather than necessarily the specific RTK keyword.

Since it's my method, I'm biased, of course, but I feel this is the best way to gain and retain kanji information, at least for reading.  The only exception IMO would be is if you want to do handwriting (as many do), in which case naturally keyword to kanji is necessary.  Even then I feel there are various improvements that can be made upon the system.  I think Japanese learning is going to look a lot different in five years.