Re: Alphabets and diagraphs

On Fri, 2 Oct 1998 12:47:37 -0500, Carlos Thompson
<cthompso@...> wrote:

>When crating languages, I've see many of you using their owns scripts,
>including sillabaries and logographs but most of the times are letters.
>Many times are related to Latin letters.  When they are not, many ways =
of
>giving those extra sounds are given, like using diagraphs, supersigns, =
and
>other.
>
>I would ask which ways you prefered for latinising.
I've gone through a number of different systems for different languages,
but in the last few years I've mainly used a system I call Kolagian
Orthography. Most of the consonants are similar to their IPA values, =
except
that "y" is [j], "c", "j", and "x" are variable, and "h" has no sound of
its own. Other sounds are spelled by using digraphs, such as "dh" [D] and
"wr" (a velar approximant). The system is currently undergoing revision,
and probably the biggest change in the consonants is the use of capital
letters for the retroflex sounds. (I never used the "standard" method of
using the number 3 to mark retroflex sounds.)

=46or the vowel sounds, I used to use accented vowels such as =E9 [e] vs.=
 =EA
[E], but this presented problems for tonal languages with large numbers =
of
vowels. I addressed this problem in Mega-Eklektu by representing the most
common vowel sounds with unaccented letters (a e i o u) and digraphs: ea
[E], oa [O], ue [y], etc. (instead of =EA, =F4, =FC) and reserving the =
diacritics
for tones. (Using ea for [E] and oa for [O] takes some getting used to, =
but
think of the English words "head" "dead", "weather", and "broad".) Since
Jarrda's vowel system is similar to the one in Mega-Eklektu, and since
email often has problems with diacritics, I adopted the Mega-Eklektu
digraphs into the Jarrda romanization, and they are also going to be a =
part
of the new Kolagian Orthography (the first revision since 1995). There
aren't enough digraphs to represent all of the IPA vowel sounds (in
particular, the distinction between slashed o and o-e ligature is lost),
but few languages require that level of detail.