Newsgroups: sci.lang
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!udel!gatech!howland.reston.ans.net!ix.netcom.com!netcom.com!alderson
From: alderson@netcom.com (Richard M. Alderson III)
Subject: Re: IndoEuropean 'r' and laryngeals
In-Reply-To: Timothy Miller's message of Tue, 21 Mar 1995 21:15:08 -0500
Message-ID: <aldersonD5yM4q.MGx@netcom.com>
Reply-To: alderson@netcom.com
Fcc: /u52/alderson/postings
Organization: NETCOM On-line Communication Services (408 261-4700 guest)
References: <Pine.SUN.3.91.950321210317.16136K-100000@grad>
Date: Fri, 24 Mar 1995 19:29:14 GMT
Lines: 85
Sender: alderson@netcom20.netcom.com

In article <Pine.SUN.3.91.950321210317.16136K-100000@grad> Timothy Miller
<millert@csee.usf.edu> writes:

>What was the Proto-IndoEuropean 'r'?

>The American Heritage dictionary says that it could act as a vowel, and as a
>liquid.  How was it a liquid?  Something like the American 'r' seems unlikely.
>What was it really?

In Proto-Indo-European, we reconstruct a class of resonants, of which *r is one
member, based on a series of correspondences among the attested languages (as
with all other reconstructions).

*As a class*, the resonants pattern differently from the obstruents or from the
vowel, for example. in the phonotactics of root formation.

One important fact about the resonants is that all of them can provide syllabic
nuclei, that is to say, can "act as vowels" or "be vowels."  Thus PIE *r in a
syllabic position would be very much like the syllabic nucleus of (rhotic)
English _curb_ [krb].  (NB:  I do not require "syllabic r" to be a cluster of
shwa + r, either phonetically or phonemically.)  There has been a recent thread
in sci.lang on similar nuclei in Czech and other Slavic languages (sv. "words
without vowels").

The other PIE resonants are *l, *m, *n, *w, and *y.  One source of confusion
for those starting IE studies is that traditionally the syllabic forms of the
last two are written *u and *i; in fact, some early proponents of the resonant
analysis marked these with a diacritic meaning "non-syllabic" in positions
where that was needed.

>Could someone explain how schwa is related to a laryngeal?

Historically.

PIE *@ was posited based on a series of correspondences among short vowels
between Indo-Iranian on the one hand and the remaining languages of the family
on the other:

	I-I	Others
*a	 a	  a
*e	 a	  e
*o	 a	  o
*u	 u	  u
*i	 i	  i
*@	 i	  a

The last row is the problem:  If "sound change" is exceptionless, then how to
explain that non-Indo-Iranian *a sometimes corresponds to I-I *a and sometimes
to I-I *i?  Obviously, there was a different vowel there, and the convention
was adopted to represent it like the "obscure vowel shwa" of Hebrew.

There was also the question of long vowels:  In certain morphological contexts,
short vowels disappeared, long vowels--of any timbre--went to shwa, and short
vowel + consonantal resonant went to vocalic resonant.

In 1878 (although the paper is dated 1879, it appeared early), Ferdinand de
Saussure published a monumental monograph _Sur le syst\`eme primitif des
voyelles indoeurop\'eennes_ (_On the primitive system of the Indo-European
vowels_), in which he showed that shwa/vowel-length patterned like a resonant.
Further, he showed that there must be two *different* resonants, one which gave
a-coloring to the vowel, and one which gave o-coloring.

(Note:  There was another source of o-coloring, related to accent shift,
already known at the time; Saussure's discovery explained *o that did not
interchange with *e or *a in such contexts.)

Although disputed by other Indo-Europeanists at the time, this suggestion was
taken up by the Semiticist M{\o}ller, who thought Saussure's new resonants to
be similar to Semitic laryngeals.  The name was taken up by others at time
passed.

With Hrozn{\'y}'s decipherment of Hittite (which Knudtson had recognized as
Indo-European in 1902, though this was not accepted at the time), new data came
to light about the consonantal inventory of PIE.  This was summarized by Jerzy
Kury{\l}owicz in his 1927 (or 1925?) paper "h hittite et @ indo-europ\'een" (in
_Symbolae grammaticae in honorem Ioannis Rozwadowski_), in which he showed that
the Hittite consonant transcribed as <h with a breve underneath> (apparently a
pharyngeal or laryngeal continuant) corresponded to "consonantal shwa".

So that's how a shwa and a laryngeal are related in Indo-European.
-- 
Rich Alderson		[Tolkien quote temporarily removed in favour of
alderson@netcom.com	 proselytizing comment below --rma]

Please support the creation of the humanities hierarchy of newsgroups!
The second CFV for humanities.misc has been posted.  See news.groups.announce.
