There's one thing that really confuses me about Heisig's method. When you encounter a kanji, it will have a certain keyword, but then he goes on to say that when this kanji is used as a primitive, use this OTHER keyword.

So basically, for a lot of kanji, you have to associate TWO keywords with them. This seems rather counterproductive to me, especially since the keywords aren't always related to the actual meaning of the kanji.

So... why is it done this way?

I think generally because the keyword is abstract and will be hard to build stories with. Take 大 for example. "Big" is abstract and you can't visualize it in and of itself without having some object that is big. In contrast, a St. Bernard is a "big" dog that can be easily visualized and put into a story. In my opinion, it is well worth learning the extra label in the long run.

You don't always have to use his primitive keywords, either. Just use whatever works best for you. And make sure you look up any keywords you want to change in the back, to make sure you don't pick one he'll use later.

But it's easier to use "crotch" than "or again" in a story.

In my group, our experience was that concrete keywords like "crotch" or "dog" are more amenable to use in mnemonic tags (or "stories" as Heisig calls them) than abstract concepts like "again" or "big". Well "big" is often OK, but "dog" is generally more useful.

When we encountered kanji on of whose parts was another kanji with a very abstract meaning, we sometimes had to resort to breaking it up into its pieces, or we invented our own concrete primitive. For example, for 辰 we used "martial artist" (a dragon fighter is one who adapts his style to best defeat his opponent) because "sign of the dragon" was too vague.

Beyond the fact that certain images are easier to build into concrete imagery ... some of the primitives that look identical in the modern writings of the characters REALLY DO have various origins that meant very different things. The four drops at the bottom of 魚　that we imagine as "fire," for example, actually began as a completely different thing altogether. The writing and look is simply identical to our fire primitive, so that's what we used. Something similar can be seen with "heaven" as it appears in 笑　... I'm not sure on that one, but I think the primitive on the bottom is actually not related to 天　at all.

In some cases I've noticed, a primitive seems to have gotten a different keyword because its origins as a sub-component in other kanji are actually completely different than the origins of the kanji itself. They may have been completely different characters at one point, but for the sake of simplicity, they evolved to identical stroke orders and appearance over time. Thus, the keyword Heisig put to the primitive itself can actually be more accurate as a meaning than the meaning of the primitive-as-a-kanji (in which case, using the more accurate and relevant definition is more helpful for creating a good story).

I agree; I felt it would be a lot simpler if there were only one keyword per kanji/primitive so for a while in the middle of the book I decided I wouldn't use any more special "primitive-only" keywords but just use the single kanji keywords in my stories. I gave up on this method after a few chapters because it really is much harder to come up & remember stories with abstract keywords -- it's just not worth it. I will say, however, that I think Heisig should have avoided using multiple keywords for the same primitive, or primitive keywords that duplicate kanji keywords.
Other than that, remembering more keywords doesn't seem to be a problem for me; the only real drawback to the approach is that for I sometimes forget the "real" meaning of kanji like 示 or 又 since they're so much more common in primitive form.

I guess my problem with it, is I hear lots and lots of complaints that the keywords Heisig chose often aren't a very good "meaning" for a particular kanji. So in cases like that, what's the point of even learning his primary keyword, instead of just focusing on the word he gives for the primitive?

Zarxrax Wrote:I guess my problem with it, is I hear lots and lots of complaints that the keywords Heisig chose often aren't a very good "meaning" for a particular kanji. So in cases like that, what's the point of even learning his primary keyword, instead of just focusing on the word he gives for the primitive?

That would be true if you found a kanji which both (a) was also a single primitive with a different keyword and (b) had a misleading primary keyword. All those complaints just mean that category (b) is non-empty -- they don't tell you whether it overlaps with (a) at all.

(Also, the thing about there being so many kanji is that even if the ones with poor or compromised keyword choices are a small fraction of the total there are still enough that people will grumble.)

Zarxrax Wrote:I guess my problem with it, is I hear lots and lots of complaints that the keywords Heisig chose often aren't a very good "meaning" for a particular kanji. So in cases like that, what's the point of even learning his primary keyword, instead of just focusing on the word he gives for the primitive?

In my view, the people making these complaints are making mountains out of molehills.

For the vast majority of cases, the keywords are reasonably close to at least one meaning/use of the Kanji. Heisig is trying to keep the keywords from bleeding together, so sometimes it is better to choose a keyword that represents a more obscure usage to avoid having two keywords that are very similar.

For example, he chose "rouse" for 起. That is reasonably close to one of the meanings of the Kanji, but he could have chosen "cause" or other words as well. In choosing, he had to take into account 因, which he has labeled "cause." All things considered, I think he did a pretty good job, and it is easier to work with his system than to change it.

howdycowdy Wrote:For example, he chose "rouse" for 起. That is reasonably close to one of the meanings of the Kanji, but he could have chosen "cause" or other words as well. In choosing, he had to take into account 因, which he has labeled "cause." All things considered, I think he did a pretty good job, and it is easier to work with his system than to change it.

I think the people who take exception to those kinds of keywords are indeed as you say making mountains out of molehills. However, it seems the majority of complaints are about things such as renaming 卆 to baseball. There are a fair number of keywords that are renamed to totally ambiguous things for no apparent reason.

A lot of people seem to be treating RTK almost as a religion. Heisig is not a god and he has made a fair number of mistakes and bad decisions (村 vs 町 anyone?). That is only understandable as he wrote the book as he was just starting to learn kanji and had not yet started to learn Japanese itself.

For a good concrete example, even in the 4th edition's last printing (which was released something like 30 years after the first printing) there are various major and obvious mistakes in regards to stroke order and how the kanji are split into primitives (according to the errata). Yes there is a 5th edition which fixed those that were known, but the point is that the errors survived for ~30 years.

Just because RTK is the best book out for learning kanji doesn't mean it's not open to criticism. The success of RTK owes more to the method than to the data in the book.

Oh yes, dukelexon's right. I forgot that some primitives are derived from different kanji to the ones they resemble today.

And yep, I couldn't agree more that on the whole Heisig did a remarkable job. Finding 2000+ keywords that are unique and close enough to each kanji would have been quite a task.

One point I might mention though is that I think sometimes when a kanji has more than one meaning, he chose a keyword because a story with that sense came to mind. Sometimes, however, a different sense or perfectly valid different shade of meaning might suggest a story to someone. I remember sometimes in our group when we were stumped for a story, we'd look up other meanings and usages of the character and from time to time something would suggest itself.

Also I think I might have mentioned once before, a small number of his primitives are culturally biased and might not fit people who are not American. In those cases I try and find another suitable keyword for that primitive. For example, I prefer "emu" to "turkey" (I don't think I've even seen a live turkey). But in any case, if you poke around people's stories, people do generally seem quite happy to change the meaning of a primitive here or there because it suits their background and experience.

Jarvik7 Wrote:

howdycowdy Wrote:For example, he chose "rouse" for 起. That is reasonably close to one of the meanings of the Kanji, but he could have chosen "cause" or other words as well. In choosing, he had to take into account 因, which he has labeled "cause." All things considered, I think he did a pretty good job, and it is easier to work with his system than to change it.

(snip)

A lot of people seem to be treating RTK almost as a religion. Heisig is not a god and he has made a fair number of mistakes and bad decisions (村 vs 町 anyone?). That is only understandable as he wrote the book as he was just starting to learn kanji and had not yet started to learn Japanese itself.

(snip)

Just because RTK is the best book out for learning kanji doesn't mean it's not open to criticism. The success of RTK owes more to the method than to the data in the book.

In fact, a rather interesting thread developed when the poster proferred that RTK was above criticism: [http://forum.koohii.com/showthread.php?tid=1020] where I and others posted well reasoned problems with RTK. And yes, I too believed it was his methods that definitely were easier to defend that how he implemented those methods. That said, he offers unique words for all 2042 kanji he gives and provides stories for a large number of those.

I liken it to this: Heisig provided a well defined path. We're now just busy trimming trees, sweeping sidewalks and planting bushes along that path (God, could I get more metaphoric?).

Nukemarine Wrote:In fact, a rather interesting thread developed when the poster proferred that RTK was above criticism: [http://forum.koohii.com/showthread.php?tid=1020]where I and others posted well reasoned problems with RTK.

Erm, Nuke ... that "poster" was ME. I've posted in this thread, too ... you see those paragraphs, posted a few inches above yours, days ago, with at least one honorable mention and reply in the very same thread (thanks, Raichu)? Yeah, hi. Hello there. (waves)

I feel kind of like I'm invisible, now ... but, meh. I guess nobody really pays attention to the actual human source of words on the Internet. I don't blame you for that -- we all do it, and I'm guilty of the same.


More to the point ... regarding that thread, I kind of feel like you've misrepresented my words, or at the very least misinterpreted them. It wasn't really my intent to suggest that RTK was "ABOVE" criticism.

Rather, I was asking a wide-sweeping question. I wanted to promote a discussion about how and why people that have never tried Heisig generally seem to be the ones that are so vehemently against it, and everything it represents. Let's not pretend, while we all rush to prove to each other how "not married" to RTK we really are, that the vast majority of the press Heisig gets with the Japanese-learning community on the Internet isn't severely negative.

It's slowly beginning to turn around, but still ... if you look at what is said about James Heisig and Remembering the Kanji in general out there, well -- it's still mostly in the other camp. You know, the camp that says he sucks, the method sucks, and only stupidly delusional people would ever invest the slightest bit of time in it. What's more, almost all of the harshest barbs seem to come from people who've only ever vaguely "heard" of it all, rather than actually put it to the test.

I was merely pointing out that much of the ACTUAL, applicable evidence ... or, rather, as no real studies have been done ... personally communicated results from those that have used it seem to indicate that it is superior as a method to the sorts of things most other learners use ... namely the traditional tedious, agonizing, rote-learing that the Japanese Ministry of Education itself favors. My point was simply that bearing this fact in mind (those that have actually used it have excelled with it), most of the vitriol and spite that is flung at Heisig-learners is generally baseless.

I never said that RTK is not without flaws ... as a graduate of RTK1, as I progress through my studies, I can easily see many aspects of the process which is open to criticism.

Here's a post of mine from that very thread, where I elaborate on that:

Quote:I'll grant that there are certain elements of Heisig's advice regarding how to apply the method that might be improved upon, but the vast majority of the criticism against Heisig lies with the idea of the method itself.

Thus, let me state it a little differently ... the speed and efficiency at which one learns using Heisig is often something that is marginalized into obilivion by those who would favor rote memorization, or an all-or-nothing approach. There are those that seem viciously determined to put forth the idea that learning using the METHOD itself is not only inadvisable ... it's outright detrimental to your studies. The counter-argument I posited was in asking those critics to explain the detriment in learning the writing and meaning of 50 kanji in a few hours.

It's that mindset specifically that has me so frustrated. There are "dodgy" keywords that Heisig himself picked (yeah, I wrinkled my nose when I got to "decameron," like everyone ... and how we get "drown" out of something more like "sinking" is beyond me) ... but I think that's something else entirely.

Here is what I do maintain, and what I DO claim: there's really no other kanji-memorization method (widely known and recognized, at least) that I know of that will produce such rapid, tangible results. If anyone can show me anything else that matches the speed and ease of a Heisig-based mnemonic/imagery approach, I'll back down from this point.

It really is honest to goodness, just downright BETTER than what most people do when they're learning the kanji. No, it's not a religion. It's just something I feel is becoming increasingly evident. In fact, I have a suspicion that if we ever get to the point in which a real, actual, honest-to-goodness study is done ... Heisig's method (or at least, one similar to or based on it) will be as demonstrably superior to traditional methods as is the fact that teaching people to read English using phonetic theory (the way it's done now) as opposed to simple whole-word recognition (the way it was done decades ago) is far better for the vast majority of learners.

hi everybody,
iam just a lurker and a beginner of nihongo studies.in my 5o's from the boondocks .ima not very tech savvy either .
i agree with the above mentioned views completely. here in india havn't met/heard of anybody studying by heisig method. my teachers seemed not interested . i just started since two weeks and am still collecting material from the net.and am already impressed by my almost 300 kanji easy recgnition and recall. otherwise hav been trying to learn jap since two years when i didn't progress beyond 150 . even which i forgot after sometime!!

actually i find it a bit difficult to navigate thru all this and usually just by chance i get to see and do interesting things.
so please excuse any breach of netetiquette.

thanks for all the support and sharing on the net.
sha

dukelexon Wrote:Here is what I do maintain, and what I DO claim: there's really no other kanji-memorization method (widely known and recognized, at least) that I know of that will produce such rapid, tangible results. If anyone can show me anything else that matches the speed and ease of a Heisig-based mnemonic/imagery approach, I'll back down from this point.

You are absolutely correct. That isn't the argument though. The anti-heisig crowd maintain that learning 3000 english keywords is not beneficial, as you will still be completely unable to actually read anything in Japanese.

I think Heisig can be beneficial to someone trying to get to grips with Japanese writing, but I'd also say its intellectually dishonest to teach people "If you complete RTK, you will be able to read 3000 kanji".

NightSky Wrote:I'd also say its intellectually dishonest to teach people "If you complete RTK, you will be able to read 3000 kanji".

And if you could actually find someone who's argued for Heisig by making that claim, I'd be more inclined to agree that it's a huge problem. People will get loose and say that they "know" 2000 kanji, which may be a bit of an overstatement, but I've never really heard anyone go ahead and claim that they can "read" them. Rather, I find that this is what anti-Heisig people SAY or THINK that Heisig-learners say. Once again, criticisms over imagined issues.

One point I would make is that after spending a few months with Heisig's method, the readings will come much, much, MUCH faster than were you to try swallowing everything about a kanji all at once. This is the way I learned my first 200 kanji, and I wouldn't recommend it to anyone. I'd spend a few hours trying to learn a bare handful of five or ten, agonizing over each seemingly random stroke, and I'd have to keep it in a daily flashcard cycle for the next week or two just to have a CHANCE at recalling the information long-term.


Post-Heisig: At the moment, I know the kanji for nearly every Japanese word I knew before putting everything aside for RTK1, and what's more ... I don't learn new vocabulary anymore without kanji. In other words, for most of the things I could read (and understand) in kana before I started ... I can now read in kanji. There are still some gaps I'm filling in, but still ... it was the best use of time I've invested in Japanese study to date.

Anything new to me is now learned from kanji down; before, it was from kana up, and the kanji step didn't come until months and months (sometimes a year) later. It makes so much sense, and assists so much in the understanding of vocabulary, that I can't believe I ever did it any other way.

This is solely because I took a few weeks out of my "main" study to complete RTK. The readings fall into place like dominoes, now. The combined time of learning reading+writing in separate missions is less than doing so all at once, and involves far less headache and frustration.

I'd also say that even with the kanji that I know from nothing but Heisig ... well, in actual study, I find it far more valuable overall to be able to at least "recognize" 2000 than was my "absolute" knowledge of 150 in the past. It took me almost a year for that first 200; a little over 2 months to complete RTK1.