Reply-To: elfling@egroups.com
Date: Sun, 31 Oct 1999 23:27:12 -0800
From: johnq@s...
To: elfling@e...
Message-ID: <7vjfcg$<96bj@e...>
In-Reply-To:  <381C4F36.429F@p...>
User-Agent: eGroups-EW/0.78
Mime-Version: 1.0
Subject: Re: Modern Language Similarities to Quenya
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

candon clannach lethgla-@pacbell.net wrote: 
original article:http://www.egroups.com/group/elfling/?start=3D1576
> Andreas Johansson wrote:
> 
> > A book I have tells me that all known languages have one (or
> > more) vowel we'd recognize as an "a". Is this correct?
> 
> Well, Proto-Indo-European, supposedly, did not have an /a/.
> 
> Candon

____________________

George L. Campbell, in his Compendium of the World's Languages, cites
the 1980 work of O. Szemer=E9nyi as being a "recent, authoritative
reconstruction of the [Proto-]Indo-European phonological system".  It
lists the vocalic phonemic inventory of Proto Indo-European as /i, e,
a, o, u/ plus their long equivalents, plus schwa.

As for languages without /a/ the only claims I know of are for certain
Northwest Caucasian languages.  These include W.S. Allen's 1956 claim
for Abaza (sometimes considered a dialect of Abkhaz) for a single vowel
phoneme of schwa (with several allophonic variants including [a]),
although most authorities give Abaza a phonemic inventory of /a/ and
/I/.  In the early 60's, A.H. Kuipers published a controversial article
arguing that Kabardian's vowels were all phonetically predictable and
therefore could not be classified as phonemes.  He eliminated schwa by
claiming it was a product of consonantal juncture phenomena, and
eliminated /a/ by positing it as the phonetic manifestation of a
consonantal distinctive feature of "openness", i.e.[+ open].  His claim
was generally assailed by most Caucasian language experts, and Kuipers
withdrew the claim in a subsequent article.  (Generally, the phonemic
vowel inventory for Kabardian is today accepted as /a/ and schwa.)