Biblical Hebrew Forum

["Yitzhak Sapir" <yitzhaksapir AT gmail.com>]

On 10/2/06, Peter Kirk wrote:
> On 02/10/2006 20:56, Yitzhak Sapir wrote:

> >> Yitzhak, are you really claiming that all cases of qamats in Masoretic
> >> Hebrew were always pronounced as "o" and not "a" until after the time of
> >> the Masoretes?
> >>
> >
> > In the Massoretic tradition, yes.  In fact, it is not only what I
> > claim, it is what
> > modern linguists claim.  For example, Geoffrey Khan, in "Tiberian Hebrew
> > Phonology" in Phonologies of Asia and Africa (1997) identifies qamets as
> > being phonetically realized as "back, half-open rounded", and then 
> identifies
> > it as Unicode +0254 (A sort of upside down, backwards e, without the bar).
> > He later spells the pronounciation of "ship" )NYH as "?enij'je:", where e
> > represents Unicode 254.
> >
> >
> Well, this is of course hatef qamats, but I take your point that it was
> pronounced [ɔ], which is actually typographically an inverted "c" (yes,
> this letter was first used in African languages and was printed by using
> an inverted "c" block).

It is not just hatef qamets.  It is also qamets.  The last syllable of the 
word
has a qamets.  It is a long "ɔ:" because the Massoretes had a phonotactic
rule that said all stressed vowels were lengthened.    He distinguishes
quite a few varieties of lengths, but for the most part, identifies three,
short, half-long, and long.  In the word "ship", the first qamets is 
half-long.
Many of his claims are based on Judeo-Arabic texts with Tiberian
vocalization from Medieval times.  Since this is also the time from which
we have Tiberian-vocalized manuscripts of the Bible, it seems to be pretty
good evidence for such.  He makes various interesting points such as that
schwa, both vocalic and silent, were both allophones of zero, because
open short syllables could not close a word.  In the Massoretes terms',
the vocalic schwa didn't event count as syllables.  Khan calls them
dependent syllables.  He suggests that this is the reason that the
Massoretes themselves didn't mark the difference between them.

> I would suggest that in fact qamats was pronounced more like [ɒ]
> (typographically an inverted script "a" or alpha), which is more open
> than [ɔ] and to western ears sounds like a cross between "a" and "o".

Again, Khan's evidence is based on Judeo-Arabic manuscripts and
he explicitly at one footnote notes how various words in Arabic were
written with Massoretic marks.  P. Kyle McCarter's article on Hebrew
in the Cambridge Encyclopedia of the World's Ancient Languages
also uses a [ɔ] as the probable phonetic realization of qamets.  Fox's
article on Proto-Semitic isolated nouns apparently also views schwas
as allophones of zero.  I would really suggest reading Khan's article
in the above mentioned book.

> >> This is an improbably claim, as is surely demonstrated by
> >> transliterations as "a" into Greek, Latin and other languages, from say
> >> 3rd century BCE (LXX) to 4th century CE (Jerome).

> > That could be explained, and probably is to be explained, as dialect
> > differences.  We know that the Greek/Latin transcriptions do represent
> > a dialectical variant from Tiberian Hebrew.

> OK. But if qamats was pronounced more like "a" in some dialects, that
> undermines your claim that the "a" pronunciation was lost until
> reintroduced under Arabic influence. It seems more probable that if
> there were dialect variations they continued, and the Sephardi
> pronunciation descends from an earlier dialect with an "a" qamats.

Well, I think that because Arabic manuscripts show the qamets was
pronounced as [ɔ], the two are to be explained as similar but
different phenomena.  The Sefardi from late influence of Arabic on
Hebrew.  The Greek/Latin transcriptions, perhaps from influence from
Aramaic on Hebrew, or from speakers of the Israelite variant of Hebrew
that underwent some different sound changes early on but were now
adapting to read Biblical Hebrew which is mainly a Judaean variant.

Yitzhak Sapir