[b-hebrew] 'Vocal shwa' had no true/phonemic quality at the time of LXX.

Yitzhak Sapir yitzhaksapir at gmail.com
Wed Jun 30 18:14:17 EDT 2010

On Wed, Jun 30, 2010 at 9:30 AM, Garth Grenache wrote:
> Dear all,
> Is it not so that the shwa represents a reduced vowel which has lost
> its original quality?

No.  The schewa is commonly represented and discussed this way but
this is not how the Masoretes themselves viewed the schewa.

> The shwa fills what once was a short syllable, but has now become
> a kind of presyllabic consonantal glide.

No.  The (vocal) schewa from the Masoretic point of view was a short
open syllable.  This is compared to a short closed syllable or long
syllables that were not represented with a schewa.  Not always did a
short open syllable lose its original vowel quality.  In the word Elohim,
for example, the hataf seghol is a reflex of the original [i] sound.  In
any case the various rules in Masoretic times meant that short open
syllables were generally dependent for their quality on the following

The rules are involved and differ on details.  But the main points
they make are:

Normally, the schewa was pronounced as hataf-patah.
However, if the following syllable started with yodh, the schewa would
be pronounced as a short hirik.
If the following syllable started with a guttural, the schewa would be
pronounced with the quality of the guttural.
But if the following syllable started with a guttural and had a qamats,
the schewa would still be pronounced as a patah.

These rules are explicitly stated in the works of the Masoretes

> The same shwa is used in Hebrew Sh:lomoh as in Aramaic Sh:lama,
> and let me assure you that Aramaic Sh:lama is not pronounced
> "Shelama" or "Shalama", but simply "Shlama".  The same shwa
> points are used after consonants that closed syllables, representing
> NO vowel.

It is interesting that you bring up Sh:lomoh.  In Masoretic vocalization,
according to the above rules, this would be pronounced as Shalomo.
However, if we were to apply the Masoretic "guttural rule" as described
above, we would get Solomo.  We get the same thing with עמורה.  The
Masoretic vocalization is (amor@ (@ = qamats).  But if applying the
"guttural rule" we would get (omor at .  In both these cases we get
vocalizations very similar to the LXX vocalization for these words.

I think this is evidence for the existence of this rule in ancient times.
There are no doubt many cases where the "guttural rule" did not apply
in ancient times.  However, it appears from the LXX transcriptions,
that the guttural rule also applied in many cases that were not just
gutturals.  The gutturals preserved the rule for the Masoretes but
it appears that originally it was more extensive.

Another reason to consider the antiquity of the guttural rule is that
it was no longer productive at the time of the Masoretes.  Around
the 5th century CE, long a: shifted to qamats.  There were other
cases that had qamats before this.  The great majority of the
qamats vowels, however, are a result of this shift.  After this shift, it
appears that the rule that the vocal schewa before a guttural takes
the following vowel did not accordingly change with the shift.  In
other words, a schewa before a guttural with qamats did not take
a qamats now.  It still took a patah.  This shows us that the
"guttural rule" predates the long a: > qamats shift.

That the schewa was also used in the end of syllables shows that the
Masoretic approach to the vocalization was based on the syllable,
not on the sound of the individual consonant.  If the consonant
marked the onset of a heavy syllable it received a full vowel (a
"king").  Otherwise it received a schewa or hataf if it was sounded.

This is a simple rule but it is wrong to take it now and reconsider it
as if the meaning of the schewa was a phonetic marker for a zero
vowel, especially since the Masoretes themselves record that this
is not the case.

> Therefore it should be recognised that traditions which attach a
> certain pronunciation to shwa are developments.

No, these are explicit statements in the works of the Masoretes
themselves, not "developments."

> So whilst it is true that Masoretes would often give the shwa the
> quality of the vowel in the following syllable, it can hardly be
> asserted that this is what shwa means.

Again, the Masoretes asserted this is what it means (in the case
of gutturals).

> Likewise shwa doesn't mean an 'e' as in Modern Hebrew.  Shwa
> means a reduced vowel with no remaining phonemic value.
> Whenever it is pronounced, the quality of it is not meaningful,
> and it's 'proper pronunciation' according to this or that tradition
> is not indicative of the original value, which becomes irrelevant
> in the Tiberian tradition.

Not only do I think some of the above statements are unfounded,
and inconsistent with explicit statements by the Masoretes, I'm not
sure I completely understand the whole paragraph.

> The state of tradition at the time of the LXX was likely the state
> of Aramaic's reduced vowels,

I'm not sure where you get this "likely".  Could you provide the
evidence for this?

As for Aramaic's reduced vowels, I'm not sure that (as you wrote
above) the pronunciation of שלמא was Shlama in Biblical times.
Comparative evidence shows that the original word was sala:m,
like in Arabic.  In NWS (Hebrew and Aramaic) s became shin, and
in Hebrew, a: became o:.  So this is how we get $@lo:m in
Hebrew.   The qamats is a reflex of the original a vowel.  In
Biblical Aramaic, the vocalization $al at m@ may reflect the
state of pronunciation before the initial a vowel dropped to give
$l at m@.

> As even the post-exilic consonantal texts of the Tnakh have
> theophoric forms with YO- (yod, waw, no he) I find it a difficult
> assertion to say that Greek LXX forms with IW- represent Yoho-.

I addressed cases where the vocalization in the LXX comes in
in parallel to the MT's YHW.  I agree that reading yo: rather than
yoho: is a possible alternative, but all I pointed out is that there
was an exact match between the LXX and the Masoretic

> Rather Greek IW- is used to transliterate both Yo- and Yho-,

What is the evidence for this?

> and the omega (W) therein represents the long o: in Yo-.
> Likewise in Greek IOUDA,

It might.  It could also represent the Masoretic vocalization.

the OU represents u: before the

> What am I suggesting?  That the vowels of the LXX can be
> explained with the full vowels of the Hebrew words,

Perhaps, although this is much less likely for יהודה (which is
not spelled יודה in Biblical times), and the comparative
evidence in the cases of Solomon and Gomorra supports
the contrary position.

Yitzhak Sapir

More information about the b-hebrew mailing list