[b-hebrew] yhwh pronunciation

Yitzhak Sapir yitzhaksapir at gmail.com
Mon Jul 25 22:13:16 EDT 2005

Dear Rolf,

You have written:

> > Could it be that the pointing of the masoretes only show the pronunciation 
> > of the period? And that thus we cannot look to them as a good source for
> > archaic vowel pronunciation? And that consequently the older LLX 
> > transliterations give a better source for understanding archaic vowel 
> > pronunciation? (within their limitations)

> The answer is clearly No. We cannot learn anything regarding the old
> pronunciation of Hebrew from the LXX.  The Masoretes were extremely careful
> when they copied the text, and we have every reason to believe that their
> vocalization represented the pronunciation that was used in B.C.E.


> The Masoretes would not dream of changing anything or invent anything new.
> Their aim was to reproduce the consonantal text accurately and to point and
> vocalize the text exactly as they heard it read in the Synagogue. However,
> in one respect they had a problem which can be illustrated with modern Greek
> (I used this example at my defence).  In Greek there are several letters
> that are pronounced as the English "e" is pronounced, in exactly the same
> way. Many errors are the result of this situation when people write down
> what they hear others say, because they must all the time make choices of
> which letters to write.
> A similar situation existed when the Masorets worked on the text. On the
> basis of transcriptions made by Josephus and Origen we see that the Hebrew
> vowels were consistently transcribed except patah and shewa, and to some
> extent segol. The vowels patah and shewa were both pronounced as an
> "a"-sound in Masoretic times, and by hearing the text recited a distinction
> between the two could have been problematic.

And now:

> >  I interpret this to mean, "Even if the first vowel [of the Babylonian transcription 
> > of initial theophoric element Yahwistic names] is an 'a' sound, [and we assume 
> > that Yahwistic names use a pronounciation close to the pronounciation of yhwh], 
> > that does not necessarily argue in favor of a patah as the first vowel of yhwh, 
> > [because] the names Gedalyahu and Gemaryahu both [have] the sign 
> > representing GA as their first syllable, [and yet their MT pronounciation has a 
> > schwa] and these names have no theophoric elements at the beginning [that 
> > might otherwise cause the Massoretes to change the pronounciation for fear of 
> > profaning the sacred.]"  
> You have understood my arguments correctly. Moreover, the vowels "a" and "e"
> can be more open or more closed, to the point where the sounds can resemble
> one another.  The Babylonian scribes would naturally choose  the syllable
> with the vowel they would use to pronounce the Jewish names, and this may
> have been more open or more closed that the vowel used by the Jews.
> Please also remember Zadok`s words about the shift from "a" to "o," which
> probably had not yet occurred. So there are many uncertain factors in the
> Babylonian writing.

We have essentially three or four languages here: Babylonian, Hebrew, and
Greek and possibly Latin.  We also have different stages to each language.
Evidently, the Greek of the original LXX must have been somewhat
different from the
Greek of several centuries later.  Late Babylonian must be considered
different from
Neo Babylonian.  And First Temple Hebrew must be considered different from 
Massoretic Hebrew.  To this end of determining vocalization, we have Babylonian
transcription of names, Greek and Latin transliterations, and also vocalization 
information as preserved at Qumran, apart from the Massoretic
rendering.  Even the
Massoretic rendering cannot be considered independently as there are several
traditions, including Babylonian and Israeli vocalization traditions. 
There may be
differences of pronounciation between sects, ie, Tiberian may represent Qaraite
while Babylonian represents Pharisaic pronounciation.  The Greek and Latin
transliterations as found in Origen and others show transliterations
which may be
considered "later", ie, the 2nd person suffix is written *-ak as in
Mishnaic Hebrew
and not *-ka, but also transcriptions which may be considered earlier
as in "l:(ebed"
(with seghols), which is transliterated "laabd" in Ps 36:1, which
shows no helper
vowel around the guttural or between the /b/ and the /d/.  These would
suggest that
the transliteration is of a variant tradition than the Massoretic.  In
some cases, the
Massoretic is later; in others, the variant tradition behind Origen is
later.  So the
statement that the MT (Tiberian) tradition definitely repesents the
in the years BCE is inaccurate or misleading.  In some cases, it does.
 It is not
necessarily true and actually probably not true that in all cases it
does.  And we
can learn something of the old pronounciation of Hebrew from the LXX and Origen,
just like we can learn something from the MT Tiberian and other traditions.  For
example, we may conclude that a pronounciation absent of the helper vowels 
around gutturals or in the case of two ending consonants survived into
the years CE
and therefore spans the entire "Biblical period" (no matter how long
one wants to date
it).  It also seems all linguists believe that the development of
these helper vowels
occured during the Second Temple or Mishnaic period, and therefore in the First
Temple period, these vowels were absent in all
traditions/sects/dialects of Hebrew.

The above view, that we can learn nothing from Greek/Latin is apparently held by
you also in regards to Babylonian transliteration.  In the case of Babylonian 
transcription of names, it is true that there may have been
differences in pronounciation
that affected vowels.  Two factors are present here - one, the
perception of the vowel,
(or consonant) which may have been perceived differently in different
languages, and
two, the encoding.  By perception I mean that a native speaker of Hebrew would 
speak his name (say "Jehu"), and then a Babylonian native speaker
would "perceive" it.
Hypothetically, even if "patah" was realized the same, the Babylonian
cuneiform may
not have allowed for example a "ia" (yod with patah) and therefore a
"ie" (yod with
seghol) had to be used.  This is what I mean by "encoding."  However,
the Massoretic
Tiberian vocalization is also quite  deficient with respect to First
Temple period names,
since it was put down to writing for the first time over a thousand
years after the First
Temple period ended.  One can choose any language to see that vowels change 
drastically over a thousand year period,  to understand that the Babylonian 
transliteration, with all its short-comings may be more useful than
the MT for the
purposes of determining pronounciation of vowels in the First Temple period.

Lastly, you seem to be making very little distinction between the
various vowels.
Tiberian distinguishes various vowels which you call "a" (patah vs qamats), and 
quite a few vowels which you call "e" (seghol, tsere, schwa,
hataf-seghol).  While
in many cases the names of the symbols apparently represent their shape
("seghol" = cluster, "ravia(" = diamond/square), in the case of vowels
they generally
seem to represent their quality: "patah" = open, "tsere" = narrow, "schwa" = 
empty, "hataf" = stolen.  It seems to me a little bit odd to claim
that "patah" (open)
could have been "more closed."  In any case, I have consistently argued for a
"qamats" under the y of yhwh, but you have downgraded it to a patah and then
apparently argued that it could have been a schwa that was mistaken by
for a patah.  You have also seemed to compared "Jehu"' to names beginning with
schwa despite the fact that Jehu has a tsere - a long vowel generally
transliterated as
"ae" and schwa is very very short.

I think the initial vowel in yhwh is most likely a long "a".  This is
based on the
MT independent component "Yah", the MT theophoric element -yahu, the MT
theophoric element -yah, all with qamats, the various Greek
transcriptions that all
give "a" at that location, and the similar Babylonian transcriptions
that also seem
to give "a" (which you say in the case of IA could also represent IE).
 While it may be
that the Babylonian cuneiform used for "ia" could actually also
represent "ie" (and
here I find it odd that if so, Jehu's name, was not transcribed as
ie-u'-wa), I still
find it needing of proof that a schwa would have been transcribed by Babylonians
as "ie".  If the vowel was a schwa, why did the Babylonians transcribe
"G:" as "ga"?
It rather seems to me that in the case of Gedalyahu and Gemaryahu rather than 
arguing for differences of vowel quality, the Babylonian transcription
shows that at
this time initial vowels had not yet reduced to a schwa, as they did
in other words
as well such as m:lakim ("kings") (from "*malkim", here with a patah).
 In any case,
while initial vowels may be less stable, the fact that the theophoric
endings -yahu
and -yah have a qamats would seem to argue that the vowel in the thephoric
prefix y:ho- was also initially qamats, as does the prefix yo- (< *yaw).

Yitzhak Sapir

More information about the b-hebrew mailing list