Re: C-IPA underlying principles and methods

En réponse à Roger Mills <romilly@...>:

>
> I think I understand that, but don't see the virtue of it. Why have
> two
> symbols to represent the same thing?
So that people can choose whatever they prefer, when they want to show a
phonetic process, or just for aesthetics, without risking to be misunderstood
nor making unofficial variants like the X-SAMPA we use here. Also, nothing
obliges you to use all variants. They are there only because of the power of
the system.

> As, IIRC your [s+] = [T]; but an [s] moved forward wouldn't in my view
> turn
> into [T], but would still be some kind of s-like sound.
Didn't I explain that + is purely a graphical diacritic of C-IPA, not of the
IPA itself? It allows you to move in the IPA chart, not to "move" sounds. I
constantly repeat that they move *characters*, not sounds.

>
> If [i}] is [e] etc., what's the point of a modified symbol to
> represent
> something that already has a symbol of its own???
The point is that you're allowed to do it. Nothing forces you to. But } has
uses to mark characters that don't have a simple symbol for them (like [@}], X-
SAMPA [6], or [O}], X-SAMPA [Q]). I was just making an example, nobody forces
you to use } with i to mark e, since there is already a symbol for it. The
point is: it is *not* forbidden, and it is unambiguous. My point was to
demonstrate how the C-IPA diacritics were used, not where they should be used
or not.

  I can see using
> diacritics to represent sounds that have no official IPA/SAMPA symbol;
But that's not an IPA transcription then, and I have no interest in "repairing"
the IPA, as I said.

>
> Lordy, IPA/SAMPA is complicated enough, without introducing the
> possibility
> of representing the _same_ sound with two distinct writings....!
> :-))))
>
But this possibility comes from the *simplicity* and power of the system, and
nobody forces you to use twenty different writings. I was just demonstrating
how the rules worked, nothing more. The C-IPA diacritics are mainly to mark
symbols that don't have an obvious translation in ASCII, like [N+] (or [g~], if
you prefer, both are unambiguous anyway) for the velar nasal (because, as I
said earlier, [N] is the uvular nasal in C-IPA. Don't blame me, blame the IPA
for using the small capital N for the uvular nasal. And I'm only trying to
transliterate the IPA, nothing more. If the IPA is counter-intuitive in some
places, I don't have to "correct" that. I wouldn't know how to do it since
people disagree on what's best already).

As I said, C-IPA is a system of *rules*, rather than a simple transliteration
done with vague ideas. The goal being that once you know the rules and the
basic characters and C-IPA diacritics, you can build yourself the rest of the
IPA without having to learn it by heart, and everybody will understand even if
you don't build the remaining characters exactly like them, because they will
know the rules to and can apply them in reverse. I replace learning each
character separately by learning much less characters and a few *productive*
rules. That and the ability to mark "exotic" sounds in an aesthetically
attractive way are the two main principles behind C-IPA.

C-IPA has to be taken as a *system* (which is why I said that I'm not sure on
the actual implementation yet. Plenty other implementations are possible. I'm
just trying to find one which allows for mnemonics to help learning the rules),
not just as a series of "IPA character <-> ASCII transliteration" lines. The
fact that many IPA characters can be represented in various ways is just an
interesting byproduct of the productivity of the rules, not a goal per se. So
arguing about symbols is useless here. C-IPA has to be taken as a full system
of rules to provide a transcription of IPA to ASCII, not as a chart of
polygraphs. And if you consider it this way, your critiques don't actually
apply.

Christophe.

http://rainbow.conlang.free.fr

Take your life as a movie: do not let anybody else play the leading role.