Re: laterals (was: Pharingials, /l/ vs. /r/ in Southeast Asia)

>> Another pair of flaws to add to the heap.
>
>Yes, but are there any languages that make a *phonemic* distinction
>between "lateral stop" and [t]?  If so, then it is a flaw.  Otherwise,
>however, it's not really a flaw, as a design principle in the IPA is
>only distinguishing sounds that are contrasted in at least some
>languages.  If, on the other hand, they only occur in the affricate
>[tK], then there's no need to make a special character, *especially* if
>the "normal" [t] does not occur.
Then it's a flaw of the design principle of
the IPA alphabet, because that alphabet is used
not only for phonemic transcriptions, but also
and very frequently for phonetic transcriptions
that are meant to tell the reader how a sound
actually sounds and not just what phoneme it
belongs to in a certain language.

Besides, what are the raising and lowering
diacritics doing in the IPA chart, then? Is there
a language that contrasts e.g. a raised [e] with
a lowered [e] and with [E], so that the raised and
the lowered ones cannot be transcribed phonemically
as /e/ and /E/? What about [r\] and [l\], is there
a language where all the repertoire of rhotics for
which the IPA provides a symbol are contrasted?
Why having symbols for the alveolopalatal
fricatives at all, is there a language where those
contrast with palatalized [S] and [Z]? What about
the symbol for simultaneous h and sh, [x\], is
there a language where it contrasts with [S]?
Swedish contrast [x\] with [s\] and sometimes
with [s`], but that could be rendered phonemically
simply as /S/ vs. /s\/ vs. /s`/, or as /S/ vs.
/S_j/ vs. /s`/ leaving aside the symbol [s\] too.
[x\] is "phonemically superfluous" because its
only purpose is to show that at the _phonetic_
level the "hsh" doesn't sound the same as a
"sh" like that of English.


>Actually, from hearing those Haida sounds, I suspect what you mean by
>"lateral stop" is "stop with *lateral release*", which is indicated in
>the IPA by a superscript l.
All laterals consist of a blocking of the central
area of the mouth (by means of raising the central
part of the tongue) while the airstream is released
laterally (by means of creating an opening at the
side(s) of the tongue). What distinguishes a lateral
approximant from a lateral fricative (and from a
lateral stop and a lateral affricate) is the
degree of _lateral_ closure, not the degree of
_central_ closure. If there is no central blocking
of the airstream during the production of the lateral
sound, be it at the alveolar, palatal, velar or
uvular area, and thus the air is allowed to flow
in any way (plosively, fricatively, approximantly
or affricately) through the central part of the
mouth as well as through the lateral part, the
sound is no longer a true lateral, but things
like a "stop with lateral release" (which is not
the same as a lateral stop).


>> Sorry, that was an error of mine, because
>> the normal laterals (I mean, the usual ones) are of
>> course the approximant ones; unlike the normal rhotics
>> which are plosive
>
>Are you counting taps/trills as "plosive"?  Because that's what most
>rhotics are.  And AFAIK, those aren't normally counted as
>plosives/stops.
Most taps/flaps and trills _are_ plosive, whether
they are 'counted' as "plosives" or not. Castilian
Spanish -rr- is nothing but a quick succession of
alveolar d's, that is, a quick succession of alveolar
stops. The alveolar tap is usually a quick alveolar
stop, though not unusually it is realized as
a quick fricative or even as a quick approximant.
But the word "plosive" is usually applied merely
to non-rhotic plosives, that is, to sounds where
the stop is neither very quick as in a tap/flap
nor quickly repeated in a row like in a trill
(which is simply a row of taps), and thus where
the feeling of 'vibration' or 'trembling' is not
produced. The word "trill" is also usually applied
merely to the plosive kind of multiple-pulse
rhotics, but here I'll broaden its meaning to
include any multiple-pulse rhotic be it plosive
or not. A non-plosive 'trill' is then e.g. Czech
r^, which is an alveolopalatal fricative-trill (or
rhoticized fricative if you like). English r is
usually an apico-postalveolar approximant-flap,
but when syllable final in rhotic dialects, the
rhotacism that it causes to the vowel with which
it usually merges phonetically (there isn't such
a long path from an approximant to a vowel) is
usually of the multiple-pulse kind.

Cheers,
Javier

Replies