Subject: Re: Vowel Reduction and the Definite Article
Date: Sat, 2 Sep 2000 19:06:43 -0500 (CDT)
>>> From: Peter Kirk
>>> I would actually lean towards the phonological argument I mention
>>> there. I suggest that the original was baaree'shiyt (absolute
>>> with article), with three long vowels. The Samaritan
>>> pronunciation apparently supports this. But the rather anomalous
>>> qamets two syllables before the stress (a position where it is
>>> usually replaced by a schwa) became shortened, in the Jewish
>>> pronunciation tradition of this obviously well known word, to a
>>> schwa.
>> From: "George Athas" <gathas AT globalfreeway.com.au>
>> But is the definite article reducible. I know that qamets is
>> reducible in propretonic position, but does this also encompass the
>> definite article when it has been lengthened to qamets? Or, put
>> another way, do we reduce vowels first and then add the definite
>> article, or do we add the definite article and then reduce vowels?
> From: "Peter Kirk" <Peter_Kirk AT sil.org>
> Probably normally we reduce vowels first then add the definite
> article. But when we do add a definite article and end up with a
> propretonic qamets (quite a rare situation), we end up with
> something which would have sounded rather oddly heavy to Hebrew
> ears, and it would not be surprising if reduction crept in here,
> especially in such a prominent position. But I would value input
> from an expert on Hebrew phonology. Henry, where are you?
Well, to have unreduced long _aa_ in an open syllable which is
separated by at least one unreduced syllable from the word's main
stress is somewhat infrequent within words without prefixed clitics --
but such unreduced _aa_ vowels are in fact extremely common when
clitics such as the definite article (or the waC- of wayyiqtol) occur
before a guttural consonant or _r_.
A reduction of b- + (h)aC- + reeshiit > b@reeshiit (where haC-, with
"C"=assimilating consonant, is the apparent underlying/reconstructed
phonological shape of the definite article) could not be done by
normal strictly-synchronic phonological processes; if such a thing
happened, it would have to be a sporadic irregular diachronic change.
I couldn't really say beforehand exactly how unusual such long _aa_
vowels in an open "propretonic" syllable really are in non-clitic
contexts, so I decided to do a search.
The list of cases I found is attached below. Excluded are forms where
such _qames._ occurs only in a clitic prefix (definite article or
conjunction), forms of _qodashim_ and _shorashim_ etc. where the
_qames._ under the first root consonant is probably a _qames. qat.an_
(short o) rather than long aa, and finally w@qaataltaa/w@qaataltii
forms with shifted stress (where the _qames._ is in the syllable
directly preceding the unshifted stress position, and so is not really
"propretonic" in the relevant sense):
===========================================
Forms that I didn't notice to be proper names, or to have _qames._ in
certain highly specialized phonological/morphological environments:
Ezr 7. 6 b:aq:ASAto1w;
Est 5. 7 uwbaq:ASAti1y;
Est 7. 3 b:.baq:ASAti1y;
Est 5. 8 'Et-b:aq:ASAti2y
Est 9.12 uwmah-b:aq:ASAte6k
Est 5. 3 uwmah-b:aq:ASAte4k
Est 7. 2 uwmah-b:aq:ASAte4k
Est 5. 6 uwmah-b:aq:ASAte4k
Jos 3.17 b:EHA6rAbA3h
Jos 3.17 b:EHA7rAbA4h
Eze30.12 HA8rAbA3h
Gen 7.22 b:EHA8rAbA3h
2Kg 2. 8 b:EHArAbA1h;
Jos 4.18 hEHArAbA2h
Hag 2. 6 w.'Et-hEHArAbA1h;
Exo14.21 lEHArAbA2h
Isa 3. 4 $AreyhE2m
Hos 7.16 $AreyhE3m
Hos 9.15 $AreyhE3m
Jer 2.26 $A8reyhE3m
Jer32.32 $A8reyhE3m
Jer17.25 w.$A6reyhE3m
Jud 5.29 $ArowtE3yhA
Isa49.23 w.$Aro8wteyhEm_4
Lev 6. 8 'az.k:ArAtA3h:
Lev 2. 2 'Et-'az.k:ArAtAh:_4
Lev 2.16 'Et-'az.k:ArAtA4h:
Lev 5.12 'Et-'az.k:ArAtAh_4
Lev 2. 9 'Et-'az.k:A6rAtA3h:
Num 5.26 'Et-'az.k:A6rAtA3h:
Forms for which I happened to notice that _qames._ was in a
phonologically-special environment (elided yodh, elided 'aleph, or
preceding an _r_ middle radical in Pi``el verb form):