Re: persian and semitic

From: John Atkinson (johnacko_at_bigpond.com)
Date: 02/28/05 

Date: Mon, 28 Feb 2005 00:58:45 GMT

"Miguel Carrasquer" <mcv@wxs.nl> wrote in message
news:dlq0219217c79pufap9n8h52u12oaa1943@4ax.com...
> On Sat, 26 Feb 2005 03:12:08 GMT, "John Atkinson"
> <johnacko@bigpond.com> wrote:
>
> >Kabard and other Circassian languages (NW Caucasian) have three levels of
> >vowel height, but no contrast of front-back, rounded-unrounded, or
length.
> >(Some other NW Caucasian languages seem to have only two vowels.)
>
> /@/ and /a/. The abundance of palatalized and labialized
> and even labio-palatalized consonants in NW Caucasian
> suggests that we're dealing with either a "degenerate"
> 3-vowel system (*i and *u passed their front-/backness on to
> neighbouring consonants, and merged as /@/, while /a/
> remained unchanged) or a 6-vowel system (*i, *i: and *u, *u:
> labialized/palatalized neighbouring consonants, all short
> vowels merged to /@/, and all long ones to /a/).

With the Arandic languages Kaytetj and the western dialect of Anmatjirra
(which have only two vowels) this seems to be part (but not all) of the
motivation. Arandic languages don't have extra palatized and labialised
consonants compared with surrounding languages with the usual Australian
phonemes. (They *have* developed pre-stopped nasals, but this doesn't seem
to have interacted with the vowel changes.) The following are the main
diachronic changes which led to the original 3-vowel system /a, i, u/ to go
to /a,@/:

(1) -Ci > -yCi
(2) All unstressed V > @
(3) Drop initial C and shift stress to 2nd syllable
(4) uC- > @Cw-
(5) i- > a-

((1)-(3) in all Arandic languages, (4) and (5) only in some -- others retain
/i/ and /u/ with low frequency.)

Examples:
Nali (we inclusive) > Nayli > Nayl@ > ayl@
muNa (night) > muN@ > uN@ > @Nw@

> If the
> second analysis is correct, it would allow or the
> possibility of 1-vowel languages (*/a/ > /@/, */i/ > /@/ +
> pal., */u/ > /@/ +lab.).

Some workers used to analyse some NW Caucasian languages as having only one
vowel phoneme (or even none at all -- Kuipers). But I don't think this is
generally accepted.

John.

Relevant Pages

  • Re: India free education bill tabled
    ... I have seen that Ranjit takes pains to use upper and lower case letters to distinguish consonants and long/short vowels. ... Easier would be to adopt one "lipi" for all Indian languages - Malayalam covers the waterfront in terms of sounds :-) ... Hmmm, maybe Sanskrit should be language of the US too, or maybe residents here will work towards getting the US to adopt Devanagari to write American. ...
    (soc.culture.indian)
  • Re: Three conversions from belief here (due to science)
    ... to make sure your culture differs from Germanic as much as possible? ... I think the ratio of consonants to vowels in Finnish isn't really all ... that weird when compared to most Indoeuropean languages. ... Also in Czech language there are names without vowels. ...
    (talk.origins)
  • Re: India free education bill tabled
    ... to distinguish consonants and long/short vowels. ... Lot of Sanskrit got printed that way. ... languages in one sentence, ...
    (soc.culture.indian)
  • Re: If you were to design a language, how many vowels and consonants would you use?
    ... between people with different native languages. ... It would give you zero vowels and zero consonants. ... They get by fine using approximations that use sounds and sound combinations that ...
    (sci.lang)
  • Re: persian and semitic
    ... >(Some other NW Caucasian languages seem to have only two vowels.) ... and even labio-palatalized consonants in NW Caucasian ... labialized/palatalized neighbouring consonants, all short ... vowels merged to /@/, and all long ones to /a/). ...
    (sci.lang)