[sci.lang] International Character Codes/Sets

alan@pdn.UUCP (Alan Lovejoy) (08/27/87)

It should be pointed out that there are numerous phones (sounds) in
English that do not have their own letter (in English spelling):
"th" (voiced as in "the" and unvoiced as in "thin"), "sh", "zh" (which
is the voiced form of "sh") and "ch" (although some insist this is
just "tsh").  Also, many vowel sounds have to share letters with
other sounds: "A" has at least 4 (allo)phones ("at any all ale"). 
"I" has at least 3 phones ("Cindi I.."). "E" has 4 ("ten meter eclair"),
one of which is also expressed by "ee" and "i", and two others which 
are also expressed by "a". "O" has 4 ("or do go on"), one of which it
shares with "a".  "U" has 4 ("Ur-Lord hunts unique nudes"). "Ou" is either
"u", "o" or a third unique sound ("would"), which can also be written
"oo" ("wood")...and I can go on, but I've made my point: there are more
vowel and consonant sounds in English than our current alphabet can handle,
it you desire to have one letter exclusively for each allophone.

Whether the general public ever changes English spelling is irrelevant.
We need to be able to use computers to express the sounds of English
and other languages (linguists need this capability, at least).  There
should be standard mechanisms to allow for this on general-purpose
computer systems.  

Of course, there already is a standard alphabet used by linguists
that can express any human speech sound with more than adequate precision:
The Internation Phonetic Alphabet.  We need ISCPT: Internation Standard
Code for Phonetic Transcription (or whater you want to call it).

--Alan "kaen yu rid dhihs?" Lovejoy