Re: [tied] Polish sibilants [was: Slavic palatalistions: why /c^/,

On Mon, 24 Oct 2005 17:36:51 +0200, Grzegorz Jagodzinski
<grzegorj2000@...> wrote:

>Piotr Gasiorowski wrote:
>
>> A few have already been shown. What else do you need? More information
>> about the significance of tongue shape? Any good handbook of
>> articulatory phonetics can give you the details. For example, Polish
>> <sz> is apical, English <sh> is laminal and bunched up ("domed, to use
>> Ladefoged's terminology"), while Polish <s'> is predorsal (also domed,
>> but without a laminal or apical component). The acoustic effect is
>> different in each case, although the place of articulation is the same
>> for all three (postalveolar). If <sz> and <s'> are distinguished in
>> Polish, it means that they lie on alternative paths crossing the same
>> place of articulation.
>>
>> Piotr
>
>All this is true except the claiming of the same place of articulation. Can
>you cite any sources? I am particularily interested in this as one of my
>webpages is devoted to Polish sibilants
>(http://www.aries.com.pl/grzegorzj/gram/uni/sibilants.html). I have
>collected quotations from 12 sources there, none of them mentions the same
>place of articulation (for <sz> and <s'>). The more frequent terms are
>"fore-tongued alveolar" for <sz> and "middle-tongued-duropalatal" for <s'>
>(literary translations from Polish).
>
>Ladefoged, for complete unknown reason, sticks to the opinion that it is <s>
>which is alveolar (none Polish source agree with this, all say of dental
>articulation here; of course the English /s/ _is_ alveolar but not the
>Polish one).

In "The Sounds of the World's Languages", Ladefoged &
Maddieson clearly state that Polish /s/ and /z/ are dental
(p. 154).

>As a consequence, he terms <sz> postalveolar (while it is
>termed alveolar in Polish literature).


>In the earlier version of his www pages he spoke of retroflex articulation
>of <sz>. I wrote him a nice piece of e-mail once and after changing ideas
>finally he agreed that <sz> is not retroflex. Unfortunately, IPA has not
>proper symbols for such a sounds. That is why Ladefoged decided finally to
>introduce a special symbol, a low-dotted /s/ for the Polish <sz> (see
>http://phonetics.ucla.edu/appendix/languages/polish/polish.html). I do not
>understand why IPA constructors hate the Slavic transcription - but the
>haceked <s^> would be better solution. Of course using [S] (integral-like)
>for the Polish <sz> is incorrect as it is easily acoustically
>distinguishable from the English sh [S] (palato-alveolar). Ladefoged terms
>the Polish sound "flat postalveolar" contrary to truly retroflex fricatives
>which he terms "apical post-alveolar".

The true retroflex fricatives are "sub-laminal palatal".

>Thus, I also _doubt_ that <sz> is
>_apical_ (the Polish literature - like Dl/uska - term it _coronal_ rather,
>see the descriptions collected on my page, the address above, however
>Sawicka uses the term "apical").

Coronal is apical and/or laminal. In the case of <sz>,
"coronal" seems adequate, as the tongue blade and tip are in
a more or less horizontal ("flat") line. Whether the main
constriction is between the apex and the alveolar ridge, or
between the blade and the postalveolar region is not
entirely clear to me. In any case, I suppose Ladefoged
prefers the term "post-alveolar" because the main
constriction is not between the alveolar ridge and the upper
teeth, as is in the case with laminal /s/, or at the
alveolar ridge, as is the case with apical /s/ (both types
of /s/ also lack a sublingual cavity and/or lip-rounding,
which contributes to the acoustic difference).

Note also that "alveolar" is defined by Ladefoged &
Maddieson (p. 14) as referring to the front part of the
alveolar ridge. The ridge itself is the dividing line
between alveolar and post-alveolar.

>And finally <s'> is alveopalatal, not alveolar and even not postalveolar

I prefer postalveolar to the confusing terms alveolopalatal
and palatoalveolar. <sz> involves the post-alveolar and
alveolar regions, and the tip and blade of the tongue, <s'>
involves the palatal and postalveolar regions, and the
predorsum and blade of the tongue.

>(and even if somebody uses this term, places of articulation of both <sz>
>and <s'> are never claimed to be the same because <sz> is generally termed
>alveolar, not postalveolar). Ladefoged writes "alveolopalatal are
>palatalized postalveolar" but also "Polish contrasts six sibilants at three
>places of articulation: alveolar, post-alveolar and alveolopalatal". If
>three places, <sz> and <s'> could not have the same place of articulation.

"Place of articulation" is used, for historical reasons, by
Ladefoged & Maddieson in a loose way for the more accurate
"articulatory gesture". See the table on p. 15, and the
discussion in that chapter as a whole.

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