Re: numbers as letters
From: | Alex Fink <a4pq1injbok_0@...> |
Date: | Tuesday, May 8, 2007, 23:28 |
On Tue, 8 May 2007 17:43:26 +0100, R A Brown <ray@...> wrote:
>A more interesting system IMO was proposed by G. de Kolvorat in 1927 in
>which all the numbers from 00 through to 99 are represented by a single
>CV syllable, beginning with _ba_ = 00 and ending with _zu_ = 99. To
>express any number you just break it up into groups of two (prefixing a
>leading zero if necessary) and spell it out, e.g.
>164750 --> 16-47-50 = femina
>50462 --> 05-04-62 = caburi
I've always thought that this is the way the Japanese kana input method for
mobile phones should have worked. Using the digit<-->consonant mappings
that are currently in place (i.e. 1-based, instead of 0-based), we could have
a i u e o <11 12 13 14 15>
ka ki ku ke ko <21 22 23 24 25>
sa si ... <31 32 ... >
and then kana with dakuten could take the other five second digits:
ga gi gu ge go <26 27 28 29 20>
za zi ... <36 37 ... >
You'd have to decide on some arbitrary place to stick syllabic n and the
p-series and small tsu, but overall it just fits so nice and closely: 10
first digits = 10 rows in the kana table, 10 second digits = 5 vowels * 2
voicing states. And it would free up both * and # for features related to
kanji or kana-set conversion and all that jazz.
The method that's actually used, I think, is the completely uninspired, and
I'd guess far less efficient,
a i u e o <1 11 111 1111 11111>
ka ki ku ke ko <2 22 222 2222 22222>
etc., with (han)dakuten on *.
Alex