Re: Opinions here of the "classic" Nelson?

Bart Mathias <mathias@xxxxxxxxxx> wrote in
_LEQf.9907$CI6.2975@trnddc07:">news:_LEQf.9907$CI6.2975@trnddc07:

Jack Crane wrote:
Thanks to you all for taking the trouble to reply to my post.

I thought one of the strengths of the classic Nelson was that it
rationalized the traditional kanwajiten lookup system. I took a
vacation from Japanese for a decade, beginning with my move back to
the United States from Tokyo; I've only recently begun to get back to
my old hobby. But I seem to remember using a kanwajiten and running
into the problem, a Catch 22 if you will, of needing in some cases to
know the traditionally assigned radical of a kanji before you could
look it up. I thought Nelson's algorithm for kanji lookup retained
the the use of the 214 radicals, but corrected the irrationality of
the traditional system. (An example was given in that review I
quoted: In the traditional system, the radical of heiwa no wa is the
right half (kuchi) instead of the left half (nogi-hen)). If I'm wrong
about this, could you all correct both my memory and understanding?

I don't see much to correct, except the notion that the original
Nelson "corrected the irrationality of the traditional system" when he
moved the "wa" in question to the kuchi section.

No. The classic Nelson has "wa" in the nogi section. In the traditional
system "wa" is in the kuchi section. Isn't it a Catch 22 that in the
traditional system you have to already know that "wa" is in the kuchi
section in order to look it up?

Since posting last I discovered I still have a kanwajiten, Shougakkan's
"Shinsen Kanwajiten", where of all the 30 kanji in the kuchi section that
can be split into a left part and a right part, "wa" is the only one with
kuchi on the right. It seems to me that the classification of "wa" is
irrational. No?

If there is any irrationality here, the correction would be to move
the "kuchi" to the left side of the character. But there is no rule
that classifiers have to be on the left, or on the top, or anything
like that.

Exactly. That's the fault of the traditional "system"; there is no hard-
and-fast rule. In the system Nelson developed for the "classic Nelson",
his algorithm for kanji lookup has 12 steps. 1. All; 2. Lone; 3.
Enclosure; 4. Left; 5. Right; 6. Top; 7. Bottom; 8. NW; 9. NE; 10. SE;
11. SW; 12. High. Since in "wa" the left side is a radical (nogi hen),
"wa" is found in the nogi section.

Blame the original kanji creators, not the people who make
dictionaries.

The classifier "mouth" of "wa" was chosen for the word's meaning,
which may have been something like "talk things over; come into
harmony on a matter." (I'm assuming the "kuchi" was "mouth," but it
might have just been a circle, which could arguably be a good icon for
"harmony.") The "nogi" hinted at the pronunciation of the word. (But
why they chose to put it on the left I can't explain.)

A couple of books I have (including the above kanwajiten) shows that a
very old form of "wa" did have something like kuchi on the left and
something like nogi on the right. At some point they got switched, but
the traditional system is stuck with what is now a classificatory
anomaly. (To be fair, I see that my kanwajiten also has "wa" in the nogi
section, but only with a pointer to the kuchi section--requiring an extra
step to find out anything more about it.)

Fortunately most kanji dictionaries have a "reading" index for those
of us who hate to count strokes. I think both Nelsons have one.

A reading index is of little value if one knows nothing about the kanji
one is looking up, right?

Jack
.