Maybe this one has been beaten to death, but....
Following up yet again on our perennial discussion about keitai favoring
Japanese over English in both input and screen real estate: take a look at
http://www.robson.org/gary/writing/caption-cart0101.html
C-print is a transcription system for English that might be adapted to
numeric keypad input. It certainly saves screen space and keypresses.
Think of it! With just a little training, u 2 cn add 30% 2 yr typng spd!!
Getting soberly linguistic for a moment:
The c-print examples illustrate a contrast easily noticed in listening and
speaking the two languages: if Japanese is a stream of consonant-flavored
vowels, then English is a stream of vowel-flavored consonants. More so than
Japanese, however, English retains most of its flavor when you remove the
main seasoning.
A good c-print "eigo henkan" for numeric keypads might therefore be more
oriented around the frequency of consonants than vowels. Despite the
relatively high frequency of vowel-symbols. they are truly the overpacked
bags of English text.
By one estimate of letter frequency, the alphabet looks more like this:
" etaonrishdlfcmugywpvbkxjqz"
Yes, of the ten most frequent letters in English text, 4 are used to denote
vowel sounds that can often be guessed when omitted. (The supposedly-silent
word-ending 'e' is nettlesome, since it often signals to the first vowel to
its right to lengthen--e.g., 'quit' vs. 'quite'..)
Some words are so common they shouldn't contribute to frequency
analysis-based input.
For example, 't' and 'h' are significantly more common than they would be if
English had a symbol for the word 'the'. ('e' would be pretty high no
matter what.)
Throwing linguistic sobriety to the winds:
In fact, English almost did get such a 'the' symbol. In those cutesy-quaint
shopsigns saying 'ye olde Inn' or whatever, that 'y' used to be pronounced
'th'. Its name as a letter was 'thorn'. 'Y' looks like a thorn, after all,
and maybe the tongue-tip-pointing-through-the-teeth suggested the image of a
thorn to those sharp-tonged Anglo-saxon of thore, I mean yore, I mean way
back then. (Dost thou get my meaning?)
Writers of dialect often render 'the' as "th'". They probably shouldn't,
but that's beside the point.
So, getting Anglo-rootsy-cutesy here, how about just the letter 'y' for
'the'?
More practical: maybe just 'd', which is pretty common to start with. This
would certainly align it with some Empire City dialects. Hmm, on second
thought, maybe not. Pretty soon we might all start talking like we're from
d'Bronx, y'know? Ya don' wan' dat, do ya?
There are still input optimizations to be had after shorthand, however.
I coin the term 'eigo henkan' ("Eigo"="English", "henkan"= "transform", from
"kanji henkan") in order sidle up to a point raised, I believe, by Curt
Sampson: those irritating, wasteful spaces between words. They aren't
useless, but they are nevertheless annoying. I put an initial space in my
revised alphabet above for a good reason. The blank is the most common
"character" in English text.
Taking a cue from henkan, how about *always* offering a menu of possible
translations of the last space-separated word, whenever the user inputs two
spaces? You're already flicking that little stunted-joystick knobby thing
to the right anyway, it's not big deal to do it one more time as long as
you're there in the first place. (Also a good opportunity: upon entering a
punctuation mark.)
If it's done after a fully-spelled-out word, the menu can offer possible
abbreviations. I think that's important, because it would go a long way
toward getting people up the learning curve of 'standard' (c-print)
abbreviations. This is two weeks, according the article linked above.
Nobody's going to sit still for that. You need to get people going from day
one, somehow, up a more gentle, months-long learning curve.
If this double-space is done after a likely abbreviation, the menu can offer
both the full spelling for confirmation and some other possibilities, a la
word-processor spellcheck. This would also help people compose more
formal-looking messages when the context required it. (Yeah, OK, like
schoolkids who are trying to pass spelling tests. Where's my ruler....)
Users would, of course, have the option of keeping the message entirely in
c-print, for SMS and mobile e-mail to economize on screen space. Or just 2
B cute.
Finally, getting back to Space, the Final Frontier: how about varying or
alternating word color or intensity, perhaps in ways that convey meaning, so
that you could more comfortably leave spaces out entirely? And how about
playing with character width while you're at it? English is a stress-timed
language, so cramming all strings of unstressed words into about the same
amount of space might save a lot of space without degrading readability very
much.
ya ya mybe ths is stpd. bt mybe not, so rmmbr, u hrd it hr 1st, ok?
-michael turner
leap@gol.com
Received on Wed Dec 19 06:12:37 2001