Re: Additional diacritics (was: Phonological equivalent of...)

Mark J. Reed skrev:
 > Ok, this confuses me:
 >
 > Tristan> Everybody agrees B is the best choice for the
 > Tristan> voiced bilabial
 > fricative, and then the voiceless
 > Tristan> one *should* be U. Then it's only natural you
 > Tristan> slip and type F
 > sometimes!
 >
 > How the heck do you get U? It seems obvious to me that if
 > the voiced foobitybop is symbolized by B, then the
 > unvoiced foobitybop should be symbolized by P, regardless
 > of the particular value of foobitybop.

That was me and it was a typo for P! My motor control of my
right hand is real crappy (I have cerebral palsy, as I have
mentioned before on this list, I'm sure) and I'm quite
likely to hit the wrong key with the right finger or even to
hit the wrong key with the wrong finger as seems to be the
case here, which I do quite frequently. The reason you don't
usually see the results of my acrobatics is that I use a
spellchecker, however the spellchecker was set not to mark
single-letter words. I have changed that setting now! :-)

In case you wonder why I don't train myself to type only
with my left hand the answer is that my better hand -- or at
least more nimble, as it isn't very strong -- sits on my
worse arm. and vice versa.

Drustanus scripsit:

 > On 09/02/07, Benct Philip Jonsson
 > <conlang@...> wrote:
 >
 >> I *read* p\ correctly, but when I want to *type* a
 >> voiceless bilabial fricative I always type P or F first
 >> -- each of them about 50% of the time, then I think "s**t
 >> its p\!", or I don't notice it at all. An unnecessary
 >> detour IMHO. Everybody agrees B is the best choice for
 >> the voiced bilabial fricative, and then the voiceless one
 >> *should* be
 >> U. Then it's only natural you slip and type F soometimes!
 >
 > Perhaps then P could be allowed as either p\ or v\, and
 > simply require the author to specify if it's unclear.
 > ("P=phi" is easier to remember than "CXS 2.3".) In either
 > case, p\ and v\ should be the recommended CXS for the two
 > symbols.

The most fool-proof solution would be

* p\ for the voiceless bilabial fricative
* v\ for the labiodental approximant
* m\ for the labiodental nasal
* b\ as alternative to B
* P and F as alternatives for any of p\ v\ m\

That's not my preference, since I find it wasteful to have
three good symbols bound up as doubles for other symbols
like this indefinitely, but a solution I could live with,
and while we're at fool-proofing, why not introduce a\ for
æ and 9\ for Œ, to remove the main source of confusion
between CXS and X-SAMPA?

BTW the voiceless bilabial fricative used to be a small
capital F in very ancient IPA!

 >
 >> > although I always use v\ for ʋ.
 >>
 >> Me too, and m\ for the labiodental nasal, but I'm
 >> repeating myself.
 >>
 >> Moreover I think b\ and p\ should be (re)assigned to ȸ
 >> U+0238 LATIN SMALL LETTER DB DIGRAPH and ȹ U+0239 LATIN
 >> SMALL LETTER QP DIGRAPH
 >
 > Now that idea I am wholly against. It's one thing to
 > redefine an unrecommended symbol when there's perceived to
 > be a problem. It's another thing to redefine a default
 > symbol for something to a character that isn't even
 > standard IPA. ...

I see no reason to get religious over either IPA or X-SAMPA.

I'm not at all reverent towards IPA as a symbol system --
ever tried to write IPA by hand, or even do it *fast*? -- I
only see it as useful because it is reasonably familiar to
others. Neither do I see any advantage other than mnemonic
with CXS being a transliteration of IPA rather than
consistent in its inner logic. FWIW I suspect X-SAMPA would
have been a better mnemonic if it had not been grafted on
SAMPA which in turn was grafted on IPA.

IPA is not set in stone, and has been expanded considerably
in recent revisions. As for the db and qp digraphs the self-
styled Oldfogey and guardian of IPA Peter Ladefoged used
them himself, so you can be pretty sure they'll get in in
the next revision. Much the same goes for the alveopalatal
symbols. They have been in de facto use by linguists working
on East Asian languages for decades. It's no coincidence
that these symbols have been included into Unicode even
before they have been admitted into IPA! Unicode has
embraced a variety of non-IPA phonetic characters which are
in common use among linguists, and I see no reason why CXS
can't do likewise at least in cases where there is a Unicode
character. Surely one can use p_d and b_d for the
labiodental stops, but good-looking they are not, and how is
one to devise a consistent symbolization for alveopalatals
*and* keep all of s\ z\ l\ in their present functions?!

 >
 >> > (BTW: Isn't it about time CXS was deprecated anyway?
 >> > Can we send messages in Unicode happily enough on this
 >> > list nowadays?[*] If so, my view is that CXS shouldn't
 >> > change at all, and anyone unhappy with it should switch
 >> > to Unicode. Is there anyone who can't view/enter
 >> > IPA/Unicode, and can't reconfigure their computers to
 >> > allow it?)
 >>
 >> As others have said it is one thing to be able to read
 >> Unicode and another to be able to type them. In fact I
 >> use type my wiki and HTML pages using CXS and then
 >> convert them with Henrik's Perl module.
 >
 > True; I suppose I'm just strange to have almost all IPA
 > chars at my fingertips... Still, I think it's the way we
 > should be going even if we can't get there just yet...

Agreed.

 >
 >> (*) Rather than mending TIPA someone should develop a
 >> version of TeX which (a) uses UTF-8 natively (b) allows
 >> the user to define any escape sequence they damn well
 >> please for any Unicode character and (c) allows the user
 >> to switch between sets of such escapes at will.
 >
 > Take a look at XeTeX or Omega; the latter's OTPs sound
 > almost exactly like what you want. For standard TeX you
 > can enter Unicode chars with the utf-8 inputenc mode and
 > the UCS package, but this just converts utf-8 characters
 > to commands defined in other packages to produce that.
 > There's also work going on integrating Omega with pdfTex
 > as well as the Lua scripting language into LuaTeX (by the
 > people who brought us pdfTeX & ConTeXt mostly), which will
 > probably be able to do almost everything anyone ever wants
 > TeX-like systems to do...

Actually I haven't felt any need to use TeX for linguistic
stuff since I got access to Unicode, and since I don't do
math I have no use for TeX anymore, but that's only me of
course! I guess what I really want is something with a real
edge over the Microsoft Keyboard Layout Creator -- i.e.
easily definable key maps and escape sequences in Windows,
if only inside a single text editor!

Ho de Isaak egraphe:

 > Anyway, it seems that the CXS custodians disagree about
 > further amendents. I am sure the stability is the main
 > positive feature of CXS. Last additions like [i\] for [1],
 > [;] for palatalization etc. were made by extensive usage.
 > So, if one wants a symbol to be used, let him use it, and
 > specify what it means, untill the community sees the
 > advantages or at least start understanding what you mean.
 > At the same time, apply common sense and avoid radicalism.
 > Stability is the best guarantee that we will be able to
 > understand one another. Remember the tower of Babel ! :)

We're all collectively the custodians of CXS, only some of
us are are also recorders. Otherwise I agree.
--

/BP 8^)
--
   B.Philip Jonsson mailto:melrochX@melroch.se (delete X)
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
"Truth, Sir, is a cow which will give [skeptics] no more milk,
and so they are gone to milk the bull."
                                     -- Sam. Johnson (no rel. ;)

Reply