Re: vowels: are they necessary?

On Dec 8, 2004, at 10:29 AM, Paul Roser wrote:

> On Tue, 7 Dec 2004 18:25:19 -0500, # 1 <salut_vous_autre@...>
> wrote:
>
>> I would like to know, </FONT></P>
>> When a consonant is fricative or trilled, it can be continued as long
>> we
>> want. Is there any languages that has some words that are only
>> consonants
>> without vowels? A little word that is only a rolled [r], a [s], a [v],
>> without the vowel releasing. It would be conceivable.
>
> A couple of languages that have been noted in the linguistic
> literature for
> their voweless words include the Berber languages, Tashilhayt &
> Tamazight
> (as elsewhere noted), Nez Perce, Bella Coola (AKA Nuxalk). The African
> language Lendu (AKA Balendru) also has words with syllable trills and
> sibilants, and there are numerous other examples of languages with
> syllabic
> consonants.
>
> Most widespread seem to be syllabic nasals, laterals and trills,
> followed
> by syllabic sibilants /s, z/, followed by other syllabic fricatives.
>
> I've read that Tamazight even has syllabic stops, but I've never heard
> the
> examples given, so I am suspicious at to whether they are released, in
> which case I suspect that there might be an ultra-short vowel or, if
> voiceless, an ultra-short voiceless vowel (similar to Shoshoni - or is
> it
> Comanche? I forget...)
Both languages have voiceless vowels (Comanche is what became of
Eastern varieties of Shoshoni after the introduction of the horse;
several Eastern Shoshoni bands rode off to the south plains in the
early 1600s). The debate in the 50s was whether voiceless vowels in
Comanche are phonemically voiceless or not; they appear to be
distinctive under strict Structuralist definitions of the phoneme, but
with a sufficiently abstract phonological analysis (historical or
generative), they can be made to be predictable.

Voiceless vowels in Shoshoni arise under the following conditions:

1.  they are short
2.  they are not part of a diphthong or in hiatus with another vowel
3.  they are unstressed

Word-finally, vowels may be optionally devoiced if the above conditions
hold:

/t1pa/ ['t1Ba] ~ ['t1p\a_0] 'pine nut'
/punku/ ['pu~Ngu] ~ ['pu~Nku_0] 'horse; pet'
/t1as1n/ ['t1as1] ~ ['t1as1_0] 'also'

Before /h/, vowels are always devoiced if the above conditions hold:

/haints1h/ ['haiJtS1_0] 'friend'
/haints1h-n1:n/ ['haiJtS1_0h"n1:~] 'friend-PL'
/haints1h-a/ ['haiJtS1_0"ha] 'friend-ACC'

The voiceless vowel in 'friend' remains voiceless regardless of its
position within the word; it is never voiced (another word with similar
properties that I can think of off the top of my head is /nattahsun/
['nat:a_0"s:u] 'medicine'.

The voiceless vowels in Shoshoni give the appearance of being the
release phase of preceding voiceless consonants since the voicelessness
of the vowel has an assimilatory force on preceding consonants, and in
fact there are often no measurable formants for voiceless /1/ when it
appears word-finally following a voiceless stop, as in /pikapp1h/
['piGap1_0] ~ ['piGap_h] 'buckskin', which gives the appearance of this
vowel being merely aspiration. Weakening of formants is most evident
with /1/ and sometimes with /a/. Mid and round vowels do not show the
same degree of formant weakening.

Other Numic languages (Chemehuevi, for example) have lost these final
voiceless vowels altogether. I suspect that if Shoshoni survives
another 50 years, its final voiceless vowels will likewise disappear.

So. More than you wanted to know about voiceless vowels in Shoshoni :-).

Dirk
--
Dirk Elzinga
Dirk_Elzinga@byu.edu

"I believe that phonology is superior to music. It is more variable and
its pecuniary possibilities are far greater." - Erik Satie

Reply