(keitai-l) The economy of Japanese text?? Kill me now. (Or kill me later.)

From: Michael Turner <leap_at_gol.com>
Date: 12/04/01
Message-ID: <002501c17ca0$e79381a0$454fd8cb@phobos>
 It seems we've got another thread going about how wonderfully suited the
Japanese writing system is to the small screen.

 This argument may have merit.

 But not because Japanese text is so much more economical.  Because it
isn't, from what I can see.

 Everybody here is always so interested in my opinions, so I feel duty-bound
to take this up again, and settle it once and for all.  (Guards, block the
exits, please.)

 The question is: economical with what?  Bytes?  Pixels?  Or--total area
consumed by equivalent text of equivalent legibility, in an interface
optimized for the specific writing system?  Which of these measures really
matter?

 Bytes hardly count after about 4800 baud.  You can't really read that fast,
you liar.

 And pixels count for less and less.  Resolution has improved dramatically.
You can waste pixels with wild abandon.  I know I do.

 So that leaves space. You remember space, don't you?  Space, that takes up
actual space?  (OK, OK--"physical dimensions".  Happy now?)

 Space matters.  It is even set to increase, for mobile text.  Because we're
all getting older and--sorry to break this to the younger segment of the
audience--our eyes are NOT getting better.  In fact, if my deteriorating
eyesight is any indication, in about 10 years we're all going to need a
wearable computer just do image-enhancement on *real* reality (forget
"virtual" reality) as we blunder around in our own kitchens and struggle
down the sidewalk trying not to bump into people.

 So space matters.  It matters so much that communications
interfaces--including writing systems themselves--are optimized for it, as
they have been for long before computers, yielding an approximately equal
space requirement, on average, for literate cultures.

 Or maybe this is just me.  I should present a shred or two of evidence,
right?

 OK, here is my laughably unscientific, one-measly-data-point
"demonstration":

 Among the readily-available parallel corpora, I selected both the Japanese
and the English editions of Bob Glickstein's "Writing GNU Emacs Extensions"
from my bookshelf.  After blowing dust off the tops, that is.

 Heh.  I knew they'd come in handy someday.

 (Yes, I know Emacs is the One True Editor and all, but after years of
half-hearted attempts to learn it, I find I'd rather chisel characters into
polished granite with a plastic coffee-stirrer.  I'm sorry if I just
insulted your religion.  Hey, careful with that 747, now....)

 Taking any pure-Japanese/pure-English paragraph-pair at random, I find that
the Japanese and the English consume almost exactly the same amount of space
on the page.  Without applying a millimeter-resolution ruler, I couldn't see
a difference.

 Yes, my eyes are bad, but they're not THAT bad.

 Same publisher, same audience, same page size, same purpose.  Have I
controlled enough variables for you?

 In fact, the Japanese edition is LONGER by about 20 pages.  In fairness,
however, this difference must be attributed to the exigencies of laying out
a Japanese book with a lot of romaji. The mandatory-roman portions take up
more space than they do in the English edition, because the roman characters
are bigger, to be stylistically consistent with the chosen Japanese
typeface.

 But that fact does not detract from my argument--it lends weight to it.
Optimize a textual interface for Japanese, and any equivalent roman
character content it's forced to carry will tend to take up more space than
if you were optimizing for typical roman character text.

 Now, a keitai screen serves up an interface optimized for presentation of
Japanese.  From the above discussion, you can guess what this will mean.
You're not guaranteed proportional spacing (much less such niceties as
kerning) so you might see unnecessary increases along the horizontal axis.
And lines will be vertically sized and spaced for kanji legibility, not
English, which can get by with less in that axis.  Generally, a character
cells sized for kanji will simply be bigger.  Roman characters will jangle
loosely around in them, glancing longingly at each other.

 In short, you just won't be able to fit the same amount of information on
that screen if it's written in the equivalent English.  But not because
English, and its writing system, are intrinsically less economical than
Japanese and its writing system.  German, now there you'll see sprawl no
matter what.  But English and Japanese seem to be about the same in ways
that matter.

 I trust I've made my point.

 Or maybe not.  I guess there's no satisfying some of you people.  (Guards,
unblock the exits.)

 I'd offer more supporting evidence for you holdouts, but the Emacs
Mujahedin just charged in, and are now demanding that I cut off my left
pinkie for blaspheming.  I think they were about to apply the Kalashnikov
treatment, but warmed to me slightly when they discovered that I am as
unkempt and irritable as they are, and smell about as bad.  I might settle
with some cash, ammo, and maybe a goat.  Or maybe I can delay them until the
Xemacs Militia of Total Digital Liberation arrive, get distracted by the
presence of their rivals, and start shooting.....

 Like we don't have enough holy wars already.

-michael turner
leap@gol.com



[ Need archives? How to unsubscribe? http://www.appelsiini.net/keitai-l/ ]
Received on Tue Dec 4 11:13:40 2001