I just finished studying Heisig's "Remembering the Kanji". I would often look up Kanji in order to figure out their meaning when the English keyword was ambiguous. To my surprise, many of the keywords seem to not capture (what I think is) the meaning of the Kanji very well.
For instance, some keywords seem to have been chosen to match compounds containing the Kanji:
- 発 "discharge" as part of 発射 "shooting/discharge".
The actual meaning of 発 is probably closer to "emit/depart".
- 組 "association" as part of 組合 "association/union/guild".
The actual meaning of 組 is probably closer to "set/group".
- 旗 "national flag" as part of 国旗 "national flag".
The actual meaning of 旗 is probably closer to "flag".
I imagine that this might be on purpose, as the book does the same with primitive meanings. For instance, the primitive 艮 "silver" is derived from the Kanji 銀 "silver". However, I feel that both on the primitive and on the Kanji level, this leads to more confusion than it does good, at least for me.
Some keywords seem to have been chosen based on a fairly obscure meaning of the Kanji:
- 安 "relax" seems to more commonly mean "cheap"
- 印 "stamp" seems to more commonly mean "mark"
- 況 "but of course" seems to more commonly mean "condition"
Finally, there are some keyword choices which I simply do not understand at all:
- RtK is the only source claiming that 氷 means "icicle".
Instead, it seems to mean "ice" with the word for "icicle" being 氷柱, which would make sense.
- RtK is the only source claiming that 般 means "carrier".
Instead, it seems to mean "sort/kind".
- RtK is the only source claiming that 采 means "grab".
Instead, it seems to mean "dice" or "form".
- RtK claims that 娠 means "with child".
It only appears as part of the compound 妊娠 "conception/pregnancy", so it is hard to tell.
I know that RtK does not set out to teach words or capture all of the possible meanings of a Kanji with a single English keyword. However, I feel like for many Kanji, there would have been better choices than the ones in RtK and that this would help with learning words later. I found the examples above after looking up just a small fraction of the Kanji, so I imagine that there are many more like this. I haven't heard many people complain about this aspect of RtK, so maybe I am thinking about all of this in the wrong way?
Trying to come up with better keywords, I have found it to be incredibly difficult to determine the range of meanings of a Kanji, and even harder to find a principal meaning. Nearly all online dictionaries are based on KANJIDIC, which in turn builds on the RtK keywords. I found this very surprising, given that RtK does not set out or claim to provide definitive meanings for Kanji, so it seemed very unfitting for a general-purpose dictionary. Due to this, the strangely chosen keywords for Kanji like 氷 and 般 are presented as their primary meanings in many dictionaries. Other sources like Duolingo, Wiktionary or the Kyōiku and Jōyō lists on Wikipedia often disagree with no way to tell which is correct. Is there a definitive resource for looking up the meaning of Kanji?
Finally, I have tried to determine the meaning of specific Kanji by looking up all the words written with them on jisho. However, I am not sure if this is a sound approach, as the Japanese word might have existed before the writing, with the Kanji being a merely adequate but not perfect fit for the meaning of the word. For instance, I have found that in the case of 旗, many sources agree on the meaning of the Kanji being "national flag", while its usage in words suggests that it means "flag".
Which brings me to my final question: What determines the meaning of a Kanji? Is it the Japanese words that are written with the Kanji? Is it the Chinese word that the character represents? How do Japanese people know which Kanji to use for which words? Or are all of these issues and questions just a consequence of my misconceptions?