On Tue, 07 Oct 2003 10:01:11 +0200, "Jean-Paul G. POTET"
<potetjp@...> wrote:

>"But the schwa isn't an independent phoneme, is it? Isn't [r&] the
>_phonetic_ realization only, of the
>Czech _phoneme_ syllabic "r"?" David WORDINGHAM, ENGLAND
>
>[So [&] is schwa? OK.]
>
>I was necessarily talking about phonetic realizations since I started from
>what I heard. As I know next to nothing about Czech, I'm unable for the
>moment to determine its phoneme chart, and say whether [&] is the
>realization of the phoneme /&/.

There is no phoneme /&/ in Czech.

>All I wanted to say is that, for me, the
>so-called "syllabic [r] of Czech" does not exist, and that it is the
>spelling that mislead some linguists to posit its existence.

So where did the spelling come from? It's the pronunciation that led the
Czechs to spell syllabic r that way.

>Maybe I could provisionally propose the rule: in Czech, whenever <r> [r] is
>found in a consonantic cluster without a written vowel after it, it is
>followed by the default vowel [&]. <prst> [pr&st] CRVC not *[prst] *CRCC.

What you hear as [&] is simply an integral part of the phoneme /r/.

Compare, for instance, the following descriptions of Italian initial /r-/
and Spanish post-/pre-consontal /Cr-/, /-rC/:

"For speaker A [..], the trill in rana 'frog' contains two contacts with an
open phase between them, but there are two other features to notice. First
the contacts are preceded by a short approximant or vowel-like sound of
about 50 ms duration. Secondly, after the contacts there is another
approximant interval, lasting over 50 ms, with a similar formant structure
to that seen in the open phase. This is part of the consonant duration, as
the tongue does not move away from the consonantal position until it ends.
The end of the consonant is indicated by an abrupt upward transition of the
third formant, as well as a significant upward shift in F1."
[Ladefoged & Maddieson, p. 219]

"Cuando la r vibrante va al lado de otra consonante, como en prado, parte,
etc., se intercala entre la momentánea oclusión de la r y la consonante que
la precede o sigue un pequeño elemento vocálico de timbre análogo al de la
vocal de la misma sílaba a que la r pertenece. La intercalación de dicho
elemento es espontáneo e inconsciente. Su duración, aunque en muchos casos
iguala y aun supera a la de la misma r, siempre es relativamente menor que
la de una vocal breve [..] Dada la naturaleza de dicho elemento, puede
decirse que la oclusión de la r vibrante es siempre intervocálica. Aun en
los casos en que la r se halla en posición final, ante pausa, su oclusión
va también seguida de elemento vocálico. El carácter vibrante de la r
aparece en realidad como resultado de la momentánea interrupción de un
sonido vocálico, producida por una rápida oclusión ápicoalveolar."
[Navarro Tomás, pp. 113-114]

>Besides, schwa also occurs in other consonantic clusters: e.g. <kdy> [g&
>'di] "when" CV-'CV.

That violates Czech accentuation rules: the stress must fall on the first
vowel.

=======================
Miguel Carrasquer Vidal
mcv@...