Review of "New Japanese-English Character Dictionary" by Jack Halpern

Published by Kenkyusha, 1990, 4th printing, ISBN 4-7674-9040-5, 8,000 JPY
Published by NTC, 1993, 1st printing, ISBN 0-8442-8434-3, 49.95 USD

Review written by Ken Lunde (lunde@mv.us.adobe.com)


NOTE: This review contains a few Japanese characters. This file is being sent
      with the Japanese characters in JIS code (so set your terminal or code
      conversion software appropriately).


        In brief, the "New Japanese-English Character Dictionary" is an
excellent reference for *understanding* kanji. Each kanji entry provides
*core meanings*, and this is very useful to the student of Japanese. I wish I
had this dictionary when I first began studying Japanese back in 1985, and
I recommend it without reservation. Now for the details...
	Kanji dictionaries are one of the most useful language tools for the
student of Japanese. While there are many such works in print, most are
designed for the native speaker of Japanese. There are now three kanji
dictionaries for the English speaker:

o "The Modern Reader's Japanese-English Character Dictionary" by Andrew
  Nelson, 1974.

o "Japanese Character Dictionary With Compound Lookup via Any Kanji" by
  Mark Spahn and Wolfgang Hadamitzky, 1989.

o "New Japanese-English Character Dictionary" by Jack Halpern, 1990/1993. 

I am not reviewing the first two kanji dictionaries in this list -- their
primary focus is not necessarily on kanji information, but on providing kanji
compounds.
	Virtually all Japanese are familiar with the 214 classical radicals
and how to lookup kanji using them. For the non-native Japanese speaker,
this is a totally new concept, one which may take a long time to learn and
master. I consider myself one who has come to appreciate the system of 214
radicals, and make use of many for-the-native-speaker kanji dictionaries that
fall under that scheme, such as 大漢語林 (1992, Taishukan Shoten).
	Before I actually begin, I will list the features that I found to
be extremely useful in the "New Japanese-English Character Dictionary"
(henceforth to be referred to as "NJECD"):

o Focuses on understanding character meanings
o Very up-to-date
o Very accurate
o Fast method for indexing kanji
o Plenty of cross-references
o Traditional kanji indexing methods
o Assorted reference material

Now I will go over these points one at a time.
	One great feature of NJECD is that it helps the student of Japanese
to understand the *meaning* of kanji by providing *core meanings*. Also
included are cross-references to other kanji with related meanings. This
provides you with a clear sense of the kanji's true meaning.
	NJECD was produced by computers, and the compilation and checking
done with several hundred programs (the details of production are nicely
described in the beginning of the dictionary -- in both English and
Japanese!). Each kanji entry provides a "core" meaning (among other useful
information). This along with the selection of kanji compounds shows that
NJECD was well thought-out and executed.
	It was very pleasing to see that the use of computers in the
compilation of NJECD did not adversely affect the selection of proper
kanji shapes. There are many kanji in JIS X 0208-1990 that have the
wrong shape, and some dictionaries treat these as the standard forms.
Some examples include the following entries: (KUTEN code in parentheses
after the encoded kanji)

NJECD Index   Encoded Kanji
 690          掴 (3647)
 741          溌 (4014)
 996          焔 (1775)
1253          祷 (3788)
1645          醗 (4016)
1843          顛 (3731)
2902          繋 (2350)

If one looks at the forms used in JIS X 0208-1990, they are different (but
incorrect) -- the forms used in JIS C 6226-1978 (the 1978 version of JIS X
0208) are correct. To the kanji learner, this accuracy and attention to
detail is encouraging.
	NJECD introduces a totally new method for indexing kanji called
SKIP (System of Kanji Indexing by Patterns). Being a 214-radical person, I
have to admit that I was a bit skeptical at first, but after spending about
an hour learning the basic principles of SKIP, I found that I could look up
kanji much faster than before. I was impressed.
	It is not part of this review to go over the SKIP rules (although I
I provide a brief introduction later). The SKIP rules are conveniently
located right inside the front and back covers, and inside you find even more
detailed information on SKIP, especially examples.
	I found a multitude of cross-references, such as for incorrect
classifications (wrong stroke count or wrong SKIP code), variant forms,
and kanji synonyms. These enable one to find kanji in case the stroke
count can vary, or else if you happen to use a variant form. The kanji
synonyms at the end of each entry are quite useful. It shows just how
many other kanji have a related meaning.
	If you do not happen to like SKIP (not sure why anyone would not
like it, though), there are other indexing methods made available, such
as:

o Traditional 214-radical lookup
o Readings lookup
o Kanji synonyms lookup (Appendix 11)

NJECD lists other indexing methods, but the ones listed above (plus SKIP
itself) are the primary ones. There is also a compact index of kanji arranged
by SKIP code. This may be a very good way to find many kanji quickly (I used
NJECD in this way recently).
	NJECD also includes a lot of good reference material. The beginning
of the dictionary includes a good introduction to the Japanese writing
system (good for the beginner and advanced student). Detailed information
on SKIP is also included. One particular gem was a listing of the 1,945
Joyo Kanji along with the NJECD index number. The Gakushu Kanji subset
(1,006) is divided up by grade level.
        Now for that brief introduction to SKIP that I promised. A SKIP number
consists of three parts delimited by a hyphen. The first part represents the
basic pattern. SKIP identifies four basic patterns, and numbers them
accordingly:

1 = Left-right
2 = Top-bottom
3 = Enclosure
4 = Solid

        For the first three patterns, the second SKIP part represents the
number of strokes in the pattern (Pattern #1: left side; Pattern #2: top;
Pattern #3: the enclosure), and the third and final SKIP part represents the
remaining number of strokes. Examples include:

剣 = 1-8-2 (1st pattern, 8 strokes in pattern, 2 remaining strokes)
書 = 2-6-4 (2nd pattern, 6 strokes in pattern, 4 remaining strokes)
国 = 3-3-5 (3rd pattern, 3 strokes in pattern, 5 remaining strokes)

        The fourth pattern, Solid, has four subpatterns:

1 = Top line
2 = Bottom line
3 = Through line
4 = Others

The second SKIP part for the fourth pattern is the total number of strokes
in the kanji. The third SKIP part is the subpattern. An example is:

下 = 4-3-1 (4th pattern, 3 total strokes, 1st subpattern)

	One thing that I would like to see as a result of NJECD is an input
method that is driven by SKIP codes. Page 110a has a footnote that sort of
talks about this. This would be an excellent way to input kanji that could
not otherwise be input by readings.
	I am into JIS codes in a big way, and it would have been very
convenient (for me) to have the JIS/KUTEN codes for each kanji entry. However,
having JIS codes is not a mission of this dictionary.
	So, how can one get this fine piece of work? The Kenkyusha edition
costs 8,000 JPY in Japan, and someone told me that it can go for 120 USD in
the US (ouch!). Even worse, I was recently told that its price in Europe
can exceed 200 USD. The NTC edition, which was just published, goes for a
much more reasonable 49.95 USD.
---------------------------- END OF REVIEW ---------------------------------