🚀 go-pugleaf

In article <168931214B.JAREA@UKCC.UKY.EDU>, JAREA@UKCC.UKY.EDU writes:
> In article <1992Oct31.194009.44286@kuhub.cc.ukans.edu>
> miner@kuhub.cc.ukans.edu writes:
>>
>>[discussion of sonority hierarchy, etc.]
>>
>>> are other problems to, like "table" in French: one syllable [tabl], but two
>>> sonority peaks [a] and [l]).
>>
>>A minor quibble, with a theoretical point: since the [l] is voiceless
>>phonetically (is it not?), how can it constitute a sonority peak?
>>Are we assuming that the voicing property of the /l/
>>in /tabl/ is not present at the point where syllabification occurs,
>>but its sonority is?  Or are French words ending in (phonetically)
>>voiceless liquids not amenable to this kind of treatment?
>>
>  
> Two minor worries come to mind:  first, is it agreed that French /tabl/
> contains only one syllable?

I assume the devoicing of the /l/ in this form (and of the /r/ in
similar forms ending in /r/) is predictable and therefore the result
of a rule or process, since I don't know of any languages which have
underlyingly voiceless lateral and rhotic approximants; if 
syllabification takes place after this devoicing rule applies then
yes, I would say French "table" has one syllable.
>  
> Second: no mention had heretofore been made of voicing except with regard
> to continuants, and that was in a post from hubey@pilot, concerning the
> sonority hierarchy, wherein he lists:
>  
>     .....
>     Second:  liquids (l, r)
>     Third:   nasals  (n, m)
>     Fourth:  Any other voiced continuant
>     Fifth:   Voiceless continuants
>     .....
>  
> I read this to imply that nasals are "voiced" continuants (and perhaps also
> liquids).  Do we all concur?  

Well, I do.  I take the liquids & nasals in the hierarchy to be voiced.
I don't know where voiceless approximants come in anybody's sonority
hierarchy.  I doubt that they come anywhere because their
voicelessness will almost surely be the result of a process and
syllabification can perhaps be posited to take place before the
process of devoicing applies, so that a sonority hierarchy would not
have to include them.

Further, I read it as merely citing "l, r" as
> examples of liquids, not as exhaustive; are the voiceless lateral affricates
> of, say, Welsh and Sassarese included as laterals at that level of the
> sonority hierarchy?

I don't know about Sassarese but the Welsh sound in question is not a
lateral voiceless approximant but a voiceless lateral *fricative* and
I assume it would occupy the same place in a sonority hierarchy as
other voiceless fricatives.
>  
> Perhaps a safer example of a voiceless final /l/ in French would be 'boucle'
> /bukl/. 

This is better because the problem with /tabl/ that I saw was that Zvi
suggested that the /l/ could be a sonority *peak* which would mean
that it is more sonorous than surrounding segments, in this case /b/,
but a voiceless approximant couldn't possibly be more sonorous than a
voiced stop.

If we were to claim that such words are monosyllabic, and justify
> this by excluding voiceless laterals (and nasals?), what do we do with
> voiceless final vowels, as in some central varieties of Italian (for example
> 'ragazzi' is often pronounced with a voiceless, i.e. whispered, vowel at the
> end.

As in Japanese, which devoices (and may even delete) high vowels in
the neighborhood of voiceless consonants, any rule or process refering
to syllables applies before the devoicing/deletion.  Dialectal "ragazzi" would
syllabify as three syllables before the devoicing of the final vowel.
This would be the general strategy in dealing with devoicing of
syllabics, and as I suggested to Zvi it ought to work for French,
although there may be languages that can't be handled this way.

[...]
-- 
-Ken <miner@kuhub.cc.ukans.edu> {Sometimes you have to look reality straight
 OPINIONS ARE MY OWN            {in the eye, and deny it. - G. Keillor