And Rosta wrote:
> One can afford to leave 'vowel' as a nontechnical term, it being useful 
> to have nontechnical terms for use in nontechnical descriptions. Then 
> its technical phonetic substitutes would be 'vocoid', or 'approximant' 
> and 'resonant', and its technical phonological substitute would be, say, 
> 'nucleus' (or 'voceme', even).
> Linguistic typology, involving the impressionistic comparison of 
> different languages, has need of nontechnical terms, such as 'vowel', to 
> capture the powerful resemblances among different languages' mechanics; 
> so I'm not suggesting that 'nontechnical' means 'not of value to 
> linguistics'. (But I am suggesting it means 'not of value to a 
> structural analysis of a single language'.)

As this thread continues, I'm coming to very similar conclusions to 
And's above.

The thread started, if you recall, with Livni's question about what he 
called a 'rhotic vowel'. I have no problem at all in referring to [r=] 
(or any other similar sound) as a (phonological) vowel in any particular 

In Sanskrit [r\`=], [r\`=:], [l`=] and [l`=:] are all classified as 
vowels and each has its own representation in the Devanagari abugida.

In 1980 edition of TY Serbo-Croat I read: "trilled _r_, as a consonant 
or vowel."

There is clearly a difference in usage of the term 'vowel' in phonetics 
and phonology, otherwise Kenneth Pike wouldn't have introduced the term 
_vocoid_ to denote a 'phonetic vowel' as distinct from a phonological vowel.

However, I concede that I may be interpreting Crystal's statement about 
phonological vowels being syllabic nuclei perhaps too loosely. While I 
have no problem in accepting that in language X or Y, sounds such as 
[r=], [l=], [n=] may constitute phonological vowels, I am having some 
trouble extending that to _voiceless_ sounds such as the [s=] in my 
pronunciation of _pst!_ [ps=t].

Interjections, as we have been rightly reminded, often contain sounds 
outside of the normal phonology of a language. Do some languages 
extensively use sounds such as [s=], [f=] and [x=] as syllable nuclei?
How common is this?

I assume Tristan got his Nuxálk examples & notion of syllable-less 
language from:

In the article, it is stated that [ʦ’ktskʷʦ’] "he has arrived" has been 
parsed into 0, 2, 3, 5, or 6 syllables depending which analysis is used!

However, the section does not make the case for phonological words 
having no syllables. Nor does it explain how these other analyses can 
arrive at differing numbers of syllables.

One would ideally like to have had recordings of these so-called 
'syllable-less words'. Certainly IMO one needs *much more* detailed 
phonological information to make any intelligent comment. I would wish 
to read the arguments given for 0 syllables, 6 syllables and at least 
one of the intermediary number.

Also spectrograms of at least some of the examples would have helped, 
for as Mark J. Reed wrote:
 > As far as I know, syllable nuclei are readily identifiable from
 > spectrograms of fluid speech.  You can also mostly tell whether such
 > nuclei are actually vowels or some other form of sonorant, although
 > the distinction lies along a spectrum of degrees of friction with no
 > clear line in the middle.
 > For the most part it's the boundaries between syllables that are
 > imprecise and dependent on phonological interpretation.


It would be interesting to see if spectrograms of fluid spoken Nuxálk 
would confirm the existence of syllable-less phonological words.

 > I don't know anything about this [sxs] thing, though.  The
 > transcription certainly doesn't give me enough to figure out how to
 > pronounce it myself.

No it does not. I can find four ways to pronounce the thing - they are 
different. Tristan himself observed that [lrn] could be interpreted as 
disyllabic [l=rn=] or monosyllabic [lr=n]. _Without more data_, there is 
no way one can figure out its correct pronunciation.

 >I'd love to hear a recording and/or see a
 > spectrogram of someone pronouncing it.


 >Was there a link to one that I missed?

If there is, I've missed it as well.

"Ein Kopf, der auf seine eigene Kosten denkt,
wird immer Eingriffe in die Sprache thun."
[J.G. Hamann, 1760]
"A mind that thinks at its own expense
will always interfere with language".