LINGUIST List 9.899

Fri Jun 19 1998

Disc: Schwa in Romance

Editor for this issue: Brett Churchill <brettlinguistlist.org>

Directory

1. recasens, schwa in Catalan and Occitan

MESSAGE

LANGUAGES/DIALECTS HAVING A STRESSED SCHWA IN
THEIR VOWEL INVENTORY

(a) I am interested in knowing about languages where
schwa [i.e., a (mid) central vowel] appears 
more or less systematically in stressed position.
A first issue of interest is whether
those languages/dialects do allow schwa to occur
in unstressed position as well.
Old Catalan and Old Occitan
appear to have had both a stressed schwa and an unstressed
schwa (in fact this is still the situation in two
dialectal varieties of those languages, namely,
Majorcan Catalan and Gascon Noir).

(b) A second question deals with the origin of stressed schwa
vis-a-vis the origin of unstressed schwa. In order to account for the
presence of stressed schwa in Old Catalan and Old Provencal, some
scholars have argued that it may have originated through assimilation
to unstressed schwa. This argument is consistent with the observation
that many languages allow schwa to appear in unstressed position but
do not so in a stressed syllable (which is in accordance with
unstressed vowels undergoing vowel reduction quite naturally).

Other scholars believe that stressed schwa preceded unstressed schwa
in Catalan/Occitan and thus that the latter vowel variety arose
through assimilation to the former one.


(c) A third question is about the quality of the
stressed vowel giving rise to stressed schwa. 
In case a language/dialect has schwa in stressed position,
what vowel did this schwa arise from?.
In Old Catalan the vowel giving rise to schwa
was stressed mid high front /e/. Is this a common
source for stressed schwa in other languages/dialects?.
What about Bulgarian where stressed schwa is also found?.

(d) Another issue concerns the quality of the unstressed
vowels giving rise to unstressed schwa.
In Old Catalan both unstressed /e/ and /a/ became schwa
historically
which explains why Modern Catalan has a phonological
rule reducing
mid front and low vowels to schwa systematically.
My particular concern is whether unstressed
/a/ is more prone to reduce to schwa than unstressed /e/; in other 
words, in languages/dialects where both /e/ and /a/
reduce to schwa in unstressed position, the
prediction could be that /a/ will reduce in the first place.
There is evidence in support of this hypothesis
both from Old Catalan and 
from modern Catalan dialects where vowel reduction is an ongoing 
process.

(e) A last observation is that prestressed vowels reduce
to schwa more easily 
than postressed vowels. Vowels in absolute word final position
are quite resistant to the process of interest, i.e., they do not become schwa too
easily. Again this piece of evidence is taken from Old Catalan
as well as from Modern
Catalan.

I would be very grateful if other linguists could provide
some information regarding those issues.

Daniel Recasens
Departament de Filologia Catalana
Universitat Autonoma de Barcelona
Bellaterra, Barcelona