List of radicals by frequency?
Madison
strokes; what I haven't seen, however, it a list of radicals ordered
by THEIR frequency (not that of their related kanji). Does one exist?
Sorry if this happens to be a trivial question, but I haven't been
able to resolve it.
Jim Breen
It depends on definitions of radicals and frequency. I assume you
mean the 214 "classical" radicals, and not modifications such as
Nelson's reallocations or Spahn and Hadamitzky's 79.
Also, what do you mean by "frequency"? Do you mean how often do they
occur in the 1,945 常用漢字? in the 6,355 JIS X 0208 kanji? in all
50,000 in Morohashi? Or do you want to take into account how
common the kanji themselves are?
For a rough idea of how often they appear in the JIS X 0208 set, you
could look at the "radkfile" which has the kanji indexed by each
visual element. Not that these are not the same as the 214 classical
radicals, but most of not all of thr 214 are embedded in it.
You can see the file at:
http://ftp.monash.edu.au/pub/nihongo/radkfile.gz
--
Jim Breen http://www.csse.monash.edu.au/~jwb/
Clayton School of Information Technology,
Monash University, VIC 3800, Australia
ジム・ブリーン@モナシュ大学
Madison
to get the question out didn't stop and think long enough about what
exactly I was asking. So let me clarify, please.
First, in this particular case - the purpose of the exercise is making
it easier for a beginner/intermediate person to study Japanese (as
opposed, I guess to any other purpose for which this can be put...),
which, I believe, implies that the user is able to find an unknown
kanji as fast as possible.
Second, let me describe the two ideal tools I would like to see. The
first would be a page like yours here (http://www.csse.monash.edu.au/
~jwb/cgi-bin/wwwjdic.cgi?1R), but modified so the radicals are
arranged by, for example, this frequency list (http://infohost.nmt.edu/
~armiller/japanese/kanjifreq.htm) or this (http://www.tidraso.co.uk/
kanji_frequency.html) or any other frequency list which you would
consider meaningful or useful. In this tool, you can use 79, 100, 150
or 214 radicals or whatever; what counts is their placement in the
table.
The other tool I would REALLY love to see would look like the first
one, but utilizing instead visual elements (which, IMHO, are more
important and useful - to a learner, at least - than the radicals).
One final tweak (for the time being...) would be this - once the kanji
is identified, its information page should not only have the readings,
etc., but also (a feature which I personally found very useful) a link
to words beginning, ending or and/or containing the kanji (as in this
page - http://jisho.org/kanji/details/%E7%94%B2).
End of wish list....
Jim Breen
> First, in this particular case - the purpose of the exercise is making
> it easier for a beginner/intermediate person to study Japanese (as
> opposed, I guess to any other purpose for which this can be put...),
> which, I believe, implies that the user is able to find an unknown
> kanji as fast as possible.
A fair goal. A lot of the work many of us have been doing for
years has had that aim.
> Second, let me describe the two ideal tools I would like to see. The
> first would be a page like yours here (http://www.csse.monash.edu.au/
> ~jwb/cgi-bin/wwwjdic.cgi?1R), but modified so the radicals are
> arranged by, for example, this frequency list (http://infohost.nmt.edu/
> ~armiller/japanese/kanjifreq.htm) or this (http://www.tidraso.co.uk/
> kanji_frequency.html) or any other frequency list which you would
> consider meaningful or useful. i
I actually think that would be a waste of time. I can imagine the screams
and yells if I reordered the "multi-radical" table that way. People
looking up that table generally neither know nor care about the
relative frequency of a radical or shape - they just want to find
it in the table as quickly as possible. Putting them in some sort
of frequency order would require them to scan up to 250 looking for
the one they want, whereas by ordering in block by stroke count, they
can dive to the one they want.
> One final tweak (for the time being...) would be this - once the kanji
> is identified, its information page should not only have the readings,
> etc., but also (a feature which I personally found very useful) a link
> to words beginning, ending or and/or containing the kanji (as in this
> page - http://jisho.org/kanji/details/%E7%94%B2).
WWWJDIC has had the link form the kanji information to words beginning
with and containing a kanji for about 12 years. The "containing"
option presents the "ending" cases first by default.
Madison
> > Second, let me describe the two ideal tools I would like to see. The
> > first would be a page like yours here (http://www.csse.monash.edu.au/
> > ~jwb/cgi-bin/wwwjdic.cgi?1R), but modified so the radicals are
> > arranged by, for example, this frequency list (http://infohost.nmt.edu/
> > ~armiller/japanese/kanjifreq.htm) or this (http://www.tidraso.co.uk/
> > kanji_frequency.html) or any other frequency list which you would
> > consider meaningful or useful. i
>
> I actually think that would be a waste of time. I can imagine the screams
> and yells if I reordered the "multi-radical" table that way. People
> looking up that table generally neither know nor care about the
> relative frequency of a radical or shape - they just want to find
> it in the table as quickly as possible. Putting them in some sort
> of frequency order would require them to scan up to 250 looking for
> the one they want, whereas by ordering in block by stroke count, they
> can dive to the one they want.
You are probably right about the expected screams and yells :);
nevertheless, these two approaches are not mutually exclusive. Storage
space on the net is, after all, essentially unlimited and my proposed
search is just another form which can sit, side by side, with the
"radicals by stroke count" form. If a user gets frustrated with one
form, he or she can always switch to the other.
I guess that my main point - again, made as an adult spare-time
student, not a scholar - is that kanji are subject to the same law of
diminishing returns as are many other things in life. As has been
pointed out (http://www.sabotenweb.com/bookmarks/about/douglas.html),
the 500 most frequently used kanji accounts for 79.4% , and the 1000
kanji accounts for 93.9% for newspaper kanji. While a serious scholar
would probably aspire to learn as many kanji as the human brain can
contain, I think 80% for 500 kanji is not a bad deal at all! I would
guess (and I don't know if it's possible to confirm or refute this)
that, say, the 20 most frequently used radicals account for 40%-60% of
the 500 most frequently used kanji. If that, or something like it, is
the case - a search form with the 20 most frequently used radicals
listed in decreasing order of frequency use, would allow a student to
quickly identify a significant chunk of the most useful kanji. That, I
believe, is a really powerful tool if the purpose of the exercise is
rapid mastery of kanji and comprehension of kanji text.
And, again, if that doesn't work for the student - stroke-order
radical tables are always one mouse click away...