Re: CPA - An ASCII-based phonetic alphabet

Thank you all for the many replies on my post.  I expected to find some
in the list after my return from Berlin (which was the reason for my
absence from CONLANG); but I was overwhelmed by the amount of discussion
that followed.

Some people seem to like it, others less so; some pointed out things
I could have done better in their opinion.

Well, CPA certainly isn't "perfect"; it is not immutably final, either.

Christian Thalmann <cinga@...> made the following suggestion:

> 1) Use the asterisk for impl and ejec stops rather than the [`].
The problem with this is that the asterisk is already used as a
prefixed modifier character and thus cannot be used as a suffixed
diacritic.  With your suggestions, sequences like [k*i] would be
ambiguous: is it a glottalized [k], or a centralized [i]?

One could, however, sacrifice one of the wild cards, or use something
else
like [\], instead of [`], though it would look more ugly.

> 2) Use the now free [`] symbol for retroflexion, as it is currently used
>    in SAMPA.
> 3) Use the now free [.] symbol for the syllable break, as it is currently
>    used in SAMPA.
> 4) Use the now free [-] symbol for linking (absence of a break).
> 5) Use the now free [=] symbol for syllabic consonants, as it is currently
>    used in SAMPA.
Yes, these suggestions are well worth considering, whichever character
is used for retroflex consonants.  Especially the use of [.] for
syllable
break and [-] for linking greatly improves the system.  I'll adopt these
two in the next revision, as soon as I have settled on how to mark
glottalization and retroflexion.  There's nothing wrong with CPA
agreeing
with SAMPA in a number of things; indeed, it surely makes things
easier.
There are quite a number of symbols that are the same in both CPA and
SAMPA
anyway; the differences are mainly where SAMPA does something hideous.

> 6) Allow [V] as an alternate notation for ["E].
Yes, I have already realized that this would be a good thing, and it
will
certainly be in the next revision.

Jörg.