Biblical Hebrew Forum

[Peter Kirk <peter AT qaya.org>]

On 12/02/2006 00:56, Yitzhak Sapir wrote:
> On 2/11/06, Peter Kirk wrote:
>
>   
> > I must say this sounds unlikely. More probable is some confusion,
> > because people write -ah on the basis that written final h is normally
> > silent in Hebrew, Arabic and English. This tends to obscure the cases,
> > rare in all these languages, where the final h should be pronounced. As
> > I see it, an actual phonetic shift t > h > 0 in both languages is much
> > less likely than a direct shift t > 0, with the h being written purely
> > as a mater lectionis. The dual shift theory makes it hard to explain why
> > final h was not lost in a few words like Allah/Eloah and others with
> > mappiq in Hebrew. Anyway, as I mentioned before, the Arabic orthography
> > in such matters is probably rather late, later than times when we more
> > or less know that Hebrew final h was not usually pronounced.
> >     
>
> He is writing in Hebrew and there is no confusion.  He quotes Bergstrasser,
> Hebraische Grammatik, amongst various points: "the final annihiliation of the
> h of the feminine ending -ah."  Then he says at the end after mentioning
> various other points, "things of this sort are mentioned by various other
> linguists such as ..."  This Grammar he quotes, is from the early 20th
> century -- pretty old.  To see what the current opinion is, the place to look
> is in various recent discussion of Semitic studies.  I don't think I'll be 
> able
> to do that soon, though. ...

Unfortunately I am also unable to do so. But linguistics and comparative 
Semitics have come a long way since Bergstrasser's day.

> ... You suggest that it makes it hard to explain why
> the h was not lost )eloah.  Well, the "ah" in )eloah is a furtive patah, and
> developed relatively late.  Supposedly, the word was pronounced for a long
> time as ")eloh" or ")iloh".  The -o- represents the development of long a > o
> in Canaanite.  Now, the final -a in feminine words was a short a.  So either
> we suppose that -h was used to spell final short -a's or we suppose that
> somehow final short -a's lengthened. ...

Well, if it is indeed the case that final -h was lost in words ending 
-ah but not in words ending -oh, which became -oah, that would be a 
possible explanation. But you would then have to explain why final -h is 
lost in the name Shlomo(h) but not in his other name Yediyah (with 
mappiq), among other words ending in a full "a" vowel (not just furtive 
patah) and he with mappiq. They may be rare, but even rarities need 
explanation.

> ... This makes it hard to explain the
> second person pronoun -ta, also probably originally short.  It also requires
> us to assume a certain position on the issue of case endings -- that they
> were lost -- because then we have more short -a's to deal with.
>
> At Siloam we have various feminine words - hnqbh, )mh, zdh -- but the verb
> hyt, apparently "hayta."  Why is there no -h for hyt, but there is for the
> following word zdh?
>
>   
Even in biblical Hebrew final (long) -a is not always written with a he 
mater lectionis, especially in certain verbal endings. Perhaps this rule 
applied to more verbal endings in inscriptional Hebrew, or there was 
less consistency in such matters than has been imposed in biblical Hebrew.
> > > All the forms of final -h being a m.l. can be traced, then, to
> > > some consonantal -h with varying degrees of confidence:
> > > 1) -h for -o in 3rd person masculine, originally -ahu
> > > 2) -h for -a in feminine words, originally -ah
> > > 3) -h for -e in final -h verbs (bnh, etc), where the h, being
> > > part of the verbal root suggests it would have been
> > > originally consonantal.
> > > I guess both #2 and #3 are controversial, though.
> > >       
>
>   
> > Certainly! For #2 see above. #3 is almost certainly wrong, for these
> > "lamed-he" verbs have cognates in Arabic etc with a final Y or W vowel,
> > and there are traces of this Y or W in some forms in Hebrew. It seems
> > clear that the final he is simply a marker of a final vowel.
> >     
>
> Take the Balaam Inscription.  This inscription uses -h for final feminine
> markers -- (lmh, mlkh, )nph, (nyh, khnh -- it has "mh" for "what?" but has
> "lm" for "why?" and "$)ltk" for "asked of you."  Various final -h verbs are
> denoted with h -- $hh, ybkh, )th.  At Siloam, we have "hyh" as a final -h
> verb.  The original consonant in this verbs was probably -y, and in those
> cases it shifted sometimes to -w.  Perhaps in other cases it shifted to -h.
> The assumption that -y and -w are traces and -h is not is not clear to me
> at all.  On the other hand, we have in Siloam "wzh" (and this) which
> suggests a use for -h as final -e, if we assume that the -h is not some
> development due to what seems to me a probable original glottal stop
> in this pronoun, which was preserved in the feminine.
>
>   
Well, it is by no means an "assumption" that -h in these lamed-he verb 
forms is only a mater lectionis, for this is the Masoretic pronunciation 
of these forms. It is possible to hypothesise that between a -y or -w 
ending to such verb forms and the Masoretic and modern Hebrew vowel only 
ending there was an intermediate stage with an -h ending. But, given 
that there is no way of telling from inscriptional Hebrew which final 
he's are pronounced and which are silent, there is as far as I know no 
evidence available to support this hypothesis. That doesn't mean that 
the hypothesis is wrong, just that it is unprovable.
> > And there are other cases of final silent he in biblical Hebrew which
> > are not covered above, e.g. names like Shilo(h) and Shlomo(h), the
> > directional he suffix, nouns like LAYLA(h) which are masculine and have
> > an unstressed he ending, masculine names ending in -ya(h) contracted
> > from -yahu, particles like ZE(h) and MA(h), etc etc. You would find it
> > hard to claim that all of these were originally pronounced "h". And you
> > still have to answer the point of why 99% of final h sounds were lost
> > but 1% survived.
> >     
>
> The directional he suffix is present in Ugaritic, which suggests it was
> originally with consonantal force. ...

Why? Is it certain that matres lectionis were never in Ugaritic?

> ... In any case, you need to use
> inscriptional evidence.  Furthermore, the above are inconsistent as
> regards patah vs qamatz, generally taken to reflect a difference in
> length.  Thus layla(h) has a qamatz, but ma(h) has a patah, as does
> )eloa(h).
>
>   
> > > The -h spelling for 3rd person possessive letter is
> > > attested in various places, including Arad, Lachish,
> > > Mesad Hashavyahu, and near Jerusalem.  It is hard
> > > therefore to see this as a dialectical variation or different
> > > spelling convention and more reasonable to see the -w
> > > as an alternative spelling convention that began to
> > > replace the -h in the late 7th or 6th century.
> > >       
> > Agreed, if we can agree that this is an alternative spelling convention
> > which does not necessarily reflect a change of pronunciation.
> >     
>
> You may want to argue it is not an immediate change of
> pronunciation but a change that had taken long before.  But
> given that there was originally an h in the pronoun, and that
> -w- came to signify -o- through consonantal -aw- -> -o- it is
> hard to say that the change was not due to pronunciation
> at all.
>   

My argument is that the change from -ahu to -aw or -o had very likely 
taken place so long before Hebrew was widely written down that the fact 
that the same final he could represent both pronunciations is coincidental.
>   
> > OK. But it is by no means proven that the -h in pre-exilic spelling ever
> > represented -ahu rather than being a mater lectionis for -o. We need to
> > look elsewhere for evidence of the date of the change from -ahu to -o,
> > and to confirm that this actually did take place.
> >     
>
> Well, can you explain then why they would use a -h for -o?
> The suggestion that -h was an historical spelling that originally
> represented a consonant, which, after having developed into
> -o, meant that -h was taken to mean a m.l. for -o makes sense.
> The idea that they would use a -h for -o out of the blue does
> not.  Especially since -aw- is closer to the -o- sound and if
> someone had to pick a letter out of the blue when the script
> is consonantal, they'd probably use -w and not -h.
>   

Well, the standard scholarly theory, from what I understand, is that yod 
and vav were used for final -i and -u respectively but originally all 
other final vowels were represented by he. It seems a very subjective 
argument to insist that vav rather than he was used for final -o vowels, 
especially when it is well known to everyone, except apparently you, 
that he is used for final -a and -e vowels.
>   
> > > So all
> > > we may be able to say is that the h in the 3rd person
> > > possessive was dropped sometime during the First
> > > Temple period, but before the 6th century.
> > >       
>
>   
> > No, I don't think we can say this, as we have no evidence that the
> > change didn't take place much earlier - even if we accept that it did
> > take place at some time, which is presumably evidenced by
> > reconstructions early Semitic from a range of languages.
> >     
>
> The idea that it occurred during the 1st Millenium BCE is
> based on the concept that prior documents were written in
> purely consonantal script with no m.l.  The earliest document
> with m.l. is from 9th century Tell Fekherye.  This would mean
> that at the start of the 10th century, the script having been
> purely consonantal, a final -h represented an -h- sound. If
> it were -o, it would not have been written, as many scholars
> propose for Gezer.  The development of -h to -o through the
> loss of -h probably happened during the 1st Millenium BCE.
>   

Well, what is the evidence that no m.l. were used before the 9th 
century? It seems that there is circular reasoning here. The most you 
can say is that there are no entirely unambiguous m.l.'s before the 9th 
century/
>   
> > Agreed, it is not proven. But it is also not disproved. In fact the
> > survival of such spellings as Shilo(h) and Shlomo(h) for people and
> > places prominent in the first temple period (but not later) strongly
> > suggests to me that this was a first temple period spelling convention.
> >     
>
> You have no evidence of Shilo(h) and Shlomo(h) from pre-exilic
> inscriptions.  Given your statement that they are "prominent in the
> first Temple period," you'd have to actually explain why prominent
> names are not attested in pre-exilic inscriptions.
>   

The answer is obvious: the few pre-exilic inscriptions which survive 
almost all date from long after the period of Solomon and the 
destruction of Shiloh. So these names and their spellings would only 
have been preserved on perishable materials such as the precursors of 
Samuel and Kings.
> ...
>
> Huh?  Since when is the assumption that the spelling of Samuel and Kings
> -- as we have them today! -- are pre-exilic in nature?  Since when is the
> assumption that if a book was composed in pre-exilic times, its current
> spelling is pre-exilic?  This is what I call circular reasoning.  What we have
> are Biblical spelling, which almost all scholars concur is a later
> development than pre-exilic spelling.  The fact that the book is pre-exilic
> in nature, which is not agreed in any case for any of these amongst
> scholars, does not mean that the spelling is too.  What you may be able
> to show is that the spelling of the Torah is relatively earlier than
> the spelling
> of Samuel which is relatively earlier than the spelling of Chronicles.  But
> you may also be able to explain that in the sense of the times the
> spelling was fixed, or in the case of Samuel and Chronicles, the time of
> composition.  In any case, all this may tell you is how the spelling
> developed in the Second Temple period only.  You cannot conclude from
> this anything regarding the First Temple period.
>
>   
The study which I would like to do, and which is entirely non-circular 
and based on normal scholarly methods, is to compare the distribution of 
spellings of words which are spelled differently in different places in 
the entire corpus of ancient Hebrew, inscriptional, biblical and ideally 
DSS. I would expect to see evidence of a gradual increase in the use of 
matres lectionis: inscriptional < Torah ~ Samuel/Kings < post-exilic 
biblical < DSS. And I would suggest that certain names might prove to be 
partial exceptions to the general pattern, in that they are spelled in 
the earlier biblical books in ways which are more characteristic of the 
general spelling of inscriptional Hebrew. This, if it does actually 
happen at all regularly, would be evidence that the names in these books 
tend to be written with pre-exilic spellings, perhaps the original 
spellings, even while the general orthography of the books was updated. 
Perhaps I could get a PhD for this study - or has someone done it already?

--
Peter Kirk
peter AT qaya.org (personal)
peterkirk AT qaya.org (work)
http://www.qaya.org/