hm, I just came to Nr: 349, f.e.: tune (same in ger) - my dictionary says: ordnen, mildern, Lied, Melodie. I see why he didnt take "song" or "order", they come later, but isnt this a change of meaning to say "tune" or is my dictionary just too small to have all meanings? Besides, Im concerned if Im only learning unimportant meanings of kanji, instead of learning the most common ones. Maybe its all silly, but I cant help but doubt since some ppl in their stories seem to misinterpret as well, or warn of only using Heisig.

I see a lot of people saying that many keywords don't match their relative kanji. I am using modified, PDA versions of EDICT and KANJIDIC dictionaries to check the majority of the kanji meanings and their common uses (http://www.csse.monash.edu.au/~jwb/j_edict.html). Both the dictionaries are based on Jim Breen's stuff, so they should be accurate.

In terms of raw kanji meanings, I find that about 99% of the keywords match the first meaning, in case of multiple meaning most are first and second meanings. In terms of meaning, (not-raw kanji meanings) some use keywords that do not show up in common usage. For those I try to incorporate the more common meanings into my stories. I do this for abstract and difficult words as well.

for example, frame 878 - junior:
the junior #disciple# #gang# member is always GOING RUNNING

oregum wrote:

In terms of raw kanji meanings, I find that about 99% of the keywords match the first meaning

You can see why this is so by reading the history of KANJIDIC

(Jim Breen, author of KANJIDIC) wrote:

The first file was compiled initially from the file "kinfo.dat" supplied by Stephen Chung, who in turn compiled his file from a file prepared by Mike Erickson. I originally added about 1900 "meanings" by James Heisig keyed in by Kevin Moore from the book "Remembering The Kanji". I later added the meanings from Rik Smoody's files, compiled when he was working for Sony in Japan. These appear to have been based on Nelson.

Last edited by Katsuo (2008 December 28, 9:06 am)

Raichu wrote:

My goal is to learn the kanji, not to learn Heisig's keywords. They're just a means to an end.

Personally, I view them as a bit of an end in and of themselves.  RTK1 doesn't teach you the readings of the kanji, only the keywords and how to write them, so you're not going to be able to fluently read Japanese after you're done with it; you're still going to have to look up kanji and compounds often to determine their reading and meaning.  The Heisig keywords greatly facilitate this process, since you can search a dictionary by keyword rather than radical search or handwriting recognition.

You'll probably run into more compounds than single-kanji words anyway, so some lookup will be required.  And for determining the meaning of single-kanji words, what better way to do it than to recall the Heisig keyword?

Tobberoth wrote:

I don't know, I've completed RtK1 and only ONCE have I found reason to change the keywords: Heisig messed up village and town, they should be exchanged. Other than that, I consider them right on the spot.

Whoa. I've been so engrossed in keywords, I can't believe I missed that one. Even though I knew 町 is town, I must have associated 田 with village. I was thinking, rural 田舎（いなか）.

Raichu wrote:

Katsuo wrote:

(Jim Breen, author of KANJIDIC) wrote:

I originally added about 1900 "meanings" by James Heisig keyed in by Kevin Moore from the book "Remembering The Kanji".

Small world. Kevin and I went through RTK together.

So Jim Breen is the one responsible for Heisig's keywords in Kanjidic. That's one of the reasons I avoid relying on Kanjidic, since it doesn't tell you whether or not the keyword is a real meaning or an artificial one assigned by Heisig.

What do you mean by "real" meaning? There are no meanings more real than Heisigs ones. All you can do is check the words the kanji is used in and deduct a meaning, there's no official standard what a kanji means.

I've gotten pretty sick of this confusion. Well over half of my failed cards older than two reviews are caused by synonymous keywords. I just got department wrong for the fifth/sixth time because I think of bureau's kanji! (The FBI - Federal Bureau of Investigation is the governmental department in charge of federal crimes...　Also my EX-word shows 局 as きょく a bureau; a department) I'm sure if I could get them together at the same time I could put a bunch of extra effort into crafting stories that distinguish between all the different kinds of similar keywords... but I'm much to lazy for that!

Yesterday I installed GreaseMonkey and the keyword replacement script. Now I've started editing any confusing keywords as they come up. I leave the actual keyword, but after it I add something distinguishing in parentheses - either a common reading in hiragana, or "not whatever" where whatever is the word I get this keyword confused with (e.g. not bureau).

I haven't had time to see how effective this is (or if it even shows up in reviews), but I thought I would post it anyway. I can't see how adding disambiguating information could harm the learning process (you still need to remember the kanji, their story and how they are put together). Also adding readings you already know not only disambiguate, adding them also starts the transition to real Japanese without sacrificing English keywords (there have been a number of words I already knew, but that I only discovered the kanji while going through Heisig: 懐かしい).

I feel as though a large part of my time in Heisig was spent on creating ways to distinguish similar keywords. I wonder if that is justified; if RTK is meant to give you a general idea of each kanji, and given that many kanji have rather close meanings, does it make sense to try and distinguish them by giving them keywords with different nuances? Shouldn't overlap be ok?

zodiac wrote:

I feel as though a large part of my time in Heisig was spent on creating ways to distinguish similar keywords. I wonder if that is justified; if RTK is meant to give you a general idea of each kanji, and given that many kanji have rather close meanings, does it make sense to try and distinguish them by giving them keywords with different nuances? Shouldn't overlap be ok?

Well, you still need to know how to differentiate between them, just because two kanji have extremely similar meanings doesn't mean they are in any way interchangeable. If you have Japanese vocabulary to back you up, you should be fine with just assigning one Japanese word each simply to know which one is which. As for the underlying concept, I'd say it's fine if they share the same one.

lanval wrote:

By the way, are there more mistakes than the village and town confusion? Is there a list somewhere? I havent found one so far.
Its starting to get confusing for me too. Even with words that have no relation, but because the stories kind of link. Currently I hate this "state of mind" primitive. So damn abstract. It gets extra hard to make distinctions in stories who all have some abstract component...

Even if there was such a list, nobody would agree on it completely.  Nearly every time that somebody complains about a keyword being "incorrect", it usually stems from the ignorance of that particular learner.  The problem is that you can't make a good choice as far as an alternative keyword unless you know all or most of the compounds that a particular kanji is used in.

From the limited examples that people have provided from the German translation, it seems that the translator preferred keywords that represented the most common specific usage of the character, whereas Heisig himself seemed to prefer more ambiguous keywords that blanketed the often multi-faceted nature of kanji. Therefore, neither route is technically wrong, but instead it is usually opinion-based as to what it is better.

And so to answer your question, the town/village confusion is the only compelling evidence of an actual mistake I have seen so far.

Last edited by Dragg (2009 January 12, 6:25 pm)

Raichu wrote:

Yes many kanji don't have a specific sense, but I'd say more do than don't. Some have more than one (like 本). I know I have been annoyed more than once at kanjidic for giving me some sense for a kanji that later turned out to be spurious meaning assigned by Heisig.

I think kanjidic should have had a separate field for the RTK keyword, even if it's a repeat of the kanji's meaning.

More "spurious" than the meanings assigned by Nelson or the like? There are no standard meanings for any given character and even among Japanese dictionaries meaning-related clauses differ wildly.

There are quite a few keywords that are poorly/incorrectly chosen in my opinion.  I posted in the thread I made about it.