Fixing Heisig: helpers
James R Hester
The conclusions that I draw from the recent postings about Heisig is
that most people are agreed that the method is sound, but disagree as
to how bad the mistakes in implementation are. My own opinion is that
a one--off compilation of the poor keywords would fix the most major
criticism, to wit, that some keywords are wrong. Another important
flaw (for those interested in only a subset of the whole lot) could
also be fixed; those kanji which are not part of the teaching kanji,
and not used as parts of teaching kanji later on, could be marked as
such. In that case, the only remaining criticism of Heisig would be
that some of his stories are pretty hopeless. I don't think this is
particularly important, as you are free to make up your own anyway,
and for most of the book do just that.
Now I'm not one for quixotic projects, so if there exists a book with
a similar method to Heisig, but without the above--mentioned
deficiencies, tell me.
Otherwise, let's compile a list. I suggest the following method:
where Heisig and Halpern differ as to the primary keyword for a kanji,
the Halpern keyword should replace the Heisig keyword. In the process
of checking Halpern, the status of the kanji (teaching/non--teaching)
could be checked as well. The final list will contain only those
kanji which have erroneous meanings in Heisig or are not teaching
kanji.
Sample list entry:
----------------------
Heisig#: heisig keyword
Halpern keyword: <- if different
Non-teaching kanji symbol (eg +symbol) <- if necessary
(optional entry)
If non teaching kanji, not used as part of other kanji signal
(eg ++ instead of +)
----------------------
I estimate the amount of labour involved is about 3 minutes per kanji,
at a guess, including typing it in to the computer. Would anyone like
to help with this? You do *not* have to be a kanji expert, or even,
come to think of it, know any japanese. You *do* have to have access
to Halpern and Heisig. And that's all. Hell, I know very little
Japanese and am by no means a kanji legend, but I know that I can tell
two english words apart, and isn't that what this is all about?
Any interested voluteers can send me mail, saying how many they would
like to do (10 or more would be fine) and I'll send them the numbers
to do. I'll collate the results and it can become a document that is
freely available, and will hopefully be very useful to a lot of
people.
If/when I get enough volunteers to make it to 2,000, I'll post again.
James.
--
_______________________________________________________________________
James Hester
Postdoctoral Researcher
National Institute for Research Into Inorganic Materials
e-mail: j...@nirim.go.jp NIRIM, Namiki 1-1
Phone/Fax: +81 298 55 4394 Tsukuba, Ibaraki 305
Japan
________________________________________________________________________
Ben Bullock
> Otherwise, let's compile a list. I suggest the following method:
> where Heisig and Halpern differ as to the primary keyword for a kanji,
> the Halpern keyword should replace the Heisig keyword. In the process
> of checking Halpern, the status of the kanji (teaching/non--teaching)
> could be checked as well. The final list will contain only those
> kanji which have erroneous meanings in Heisig or are not teaching
> kanji.
I have two comments about this:
1) it is extremely easy to get a list of Heisig vs. grade by using kanjidic
and some kind of simple program to extract and sort the information.
2) Halpern's keywords may be copyrighted, so it may be necessary for
him to grant permission before making a public list based on his
book's keywords.
--
Ben Bullock @ KEK (National Laboratory for High Energy Physics) / address:
1-1 Oho, Tsukuba, Ibaraki 305, JAPAN / TEL: 0298-64-5403 / FAX: 0298-64-7831 /
in japanese: ベン ブロック ＠ 高エネルギー研究所 茨城県つくば市大穂1ー1
MoebiusW
( Windows) app teaching kana. When he got finished with kana, I extended
it to kanji: I scanned in a bunch of kanji from Halpern. It would be a
snap to extend this, I already have the Halpern #'s at hand, and I could
title the kana dialog box with the Heisig summary. I wonder if there is a
textfile containing some xref of Heisig-to-anything? ( I could use
Bullock-san's flames as a starting point < ;-] )
I went to Asahiya but couldn't find a copy of Heisig. Could someone give
me an ISBN for this book?
Jim Breen
>> I wonder if there is a
>> textfile containing some xref of Heisig-to-anything?
>However, As Ben pointed out yesterday, there may be copyright problems
>if Halpern's keywords are reproduced in a public forum
>(unfortunately). If anyone from Kenkyusha wants to clarify this, feel
>free. So I've got cold feet on the Halpern front.
>Thus I suspect that the best variant is to use the keywords in
>J. Breen's `edict'. This has the added advantage that it is on line,
>If we use `edict', I think that the Halpern numbers are likely to be
>superfluous, as we are now completely ignoring Halpern's keywords. In
>any case, a crossreference from Heisig's numbers to anything would be
>useful.
Actually, the file is `kanjidic', not `edict' (which exists, but does not
contain kanji-specific information.)
The kanjidic file *does* contain xrefs between Heisig nos and all sorts of other
indices (Halpern, Nelson, Gakken, etc.), and I hope before the end of '95
to have a complete set of S&H indices and O'Neill indices too.
The "meaning" fields in kanjidic have come from all sorts or sources, and you
will find words in there that also crop up in Heisig, Nelson, Halpern, etc.
>I don't see any easy way to write a program to speed up the
>construction of this list. It's a case of using `xjdic' to look up
>each one.
It is fairly easy to pull a index -> index list out of kanjidic. A few
lines of perl or C.
--
Jim Breen [ジム・ブリーン@モナシュ大学]
Department of Robotics & Digital Technology. Monash University.
Clayton VIC 3168 Australia (p) +61 3 905 3298 (f) +61 3 905 3574
j.b...@rdt.monash.edu.au [http://www.rdt.monash.edu.au/‾jwb/]
James R Hester
> I have been working on a project for my kid, a little PowerBuilder and C++
> ( Windows) app teaching kana. When he got finished with kana, I extended
> it to kanji: I scanned in a bunch of kanji from Halpern. It would be a
> snap to extend this, I already have the Halpern #'s at hand, and I could
> title the kana dialog box with the Heisig summary. I wonder if there is a
> textfile containing some xref of Heisig-to-anything? ( I could use
> Bullock-san's flames as a starting point < ;-] )
Sounds like a nifty program. Ideally, it would give the Heisig
keyword unless the Heisig word was incorrect, when it would give the
Halpern keyword....ie it would give the Halpern keyword all the time.
However, As Ben pointed out yesterday, there may be copyright problems
if Halpern's keywords are reproduced in a public forum
(unfortunately). If anyone from Kenkyusha wants to clarify this, feel
free. So I've got cold feet on the Halpern front.
Thus I suspect that the best variant is to use the keywords in
J. Breen's `edict'. This has the added advantage that it is on line,
If we use `edict', I think that the Halpern numbers are likely to be
superfluous, as we are now completely ignoring Halpern's keywords. In
any case, a crossreference from Heisig's numbers to anything would be
useful.
I don't see any easy way to write a program to speed up the
construction of this list. It's a case of using `xjdic' to look up
each one.
> I went to Asahiya but couldn't find a copy of Heisig. Could someone give
> me an ISBN for this book?
ISBN 0-87040-739-2; but you may not want to buy the book if you don't
plan to use his method of memorisation, the book's rather expensive
and as Ben has repeatedly said, it's no good as a reference.
James R Hester
I just said:
> Thus I suspect that the best variant is to use the keywords in
> J. Breen's `edict'. This has the added advantage that it is on line,
> If we use `edict', I think that the Halpern numbers are likely to be
> superfluous, as we are now completely ignoring Halpern's keywords. In
> any case, a crossreference from Heisig's numbers to anything would be
> useful.
I've just checked the file 'kanjidic.doc', and guess what? The
original kanji meanings in kanjidic (the online japanese dictionary
compiled by J. Breen) were taken from Heisig, and later augmented with
meanings from Nelson. A quick check showed that the first word in the
`kanjidic' english meaning field appears to be the Heisig keyword, and
where this keyword is wrong in Heisig, it is unchanged in kanjidic.
> I don't see any easy way to write a program to speed up the
> construction of this list. It's a case of using `xjdic' to look up
> each one.
So this isn't an option either. My present plan is now to indicate
which Heisig keywords are misleading, together with the Halpern number
to facilitate lookup. I'm still looking for volunteers.
Jim Breen
>I've just checked the file 'kanjidic.doc', and guess what? The
>original kanji meanings in kanjidic (the online japanese dictionary
>compiled by J. Breen) were taken from Heisig, and later augmented with
>meanings from Nelson. A quick check showed that the first word in the
>`kanjidic' english meaning field appears to be the Heisig keyword, and
>where this keyword is wrong in Heisig, it is unchanged in kanjidic.
Blast. I though I hade flushed out the erroneous ones (bear in mind that
only about 1800 kanji had those ex-Heisig originals.) If anyone can assist
by pointing out the obviously incorrect "meaning" fields, I will fix
the KANJIDIC file.
(Of course, the "meaning" of a kanji is a point for endless discussion,
much of it futile.)