Biblical Hebrew Forum

[Yitzhak Sapir <yitzhaksapir AT gmail.com>]

Dear Laurent,

Just because you encounter here ideas which seem to corroborate your
own feelings, does not mean that your feelings are correct.  Sometimes,
you and those whose opinions you read here are correct.

It may be that both are you are wrong and are guided by the same
mistaken assumptions.

That is the case here.

The hataf patah is present on all kinds of letters.  It is only their common
presence on gutturals that makes you think they should *only* be on
gutturals.  But this is not the case.  Hataf means "short," and in Masoretic
literature it refers to all short vowels, even in closed syllables.  On non-
gutturals it is often marked to suggest the schewa should be vocal.

Gen 1:18 has a hataf patah.  In case you thought it was a mistake in
the Leningrad codex, Yaaqov Sapir (no known relation) of Jerusalem
sent a letter to the rabbis of Aleppo asking to consult the Aleppo codex
regarding this word, asking if it was "meteg and schewa only", and the
reply was "not so -- it has a hataf patah on the lamed."  For his questions
see here:
http://www.aleppocodex.org/9.html#label14
http://www.daat.ac.il/daat/vl/tohen.asp?id=278
A similar word is Gen 2:12: [uu-zahav]

Isaac lists different words with some hataf patah and some without,
regarding double letters, including modern pronunciation.  But the
modern pronunciation is not the Tiberian Masoretic pronunciation.
One can't compare the two as they are two different systems.
In the Tiberian system, long vowels followed by long consonants
could create a situation where the long consonant would be
shortened:
Jonah 2:4,6 [ya,so:-va've:-ni:]

In the above case, we have the sequence [so:vve:].  In this case,
the long vowel (o:) and long consonant (v:) ended up in an overly
long syllable (so:v), which is three morae.  This could lead to a
situation where the consonant would be shortened - [so:ve:].
To prevent this, the Masoretes pronounced the schewa in these
cases (preceded by a long vowel and followed by an identical
consonant) as vocal ([so:vave:]).  This was not necessary when
the vowel was short [hinni:].  In such cases, there was no overly
long syllable (hin is two morae), and so there was no need to
make the schewa vocal in this case.

In some of these cases where the schewa was vocal, the
Masoretes marked the vocal schewa with a hatef patah, since
that is all a hatef patah is - a short patah, and a schewa is
a (generally) short vowel usually pronounced as patah.  So,
the Horayat HaQore writes: "If one arguest that the dalet of
Mordochai (and other letters in other words) has hatef
qames tell him, 'but this is only a device used by some
scribes to warn that the consonants should be pronounced
fully, and not slurred over'.  Hatef qames is written in some
texts.  It is not used in others, but the reader nevertheless
pronounces the word in the same way when he comes to
read it."

In two cases listed by Isaac for a hatef schewa under a non
guttural have a long vowel preceding it and an identical
consonant following: Ex 15:10, and Gen 29:8.  We even know
there was a hataf patah in the Aleppo codex because they
were also questions of Yaaqov Sapir in the same list
mentioned above.

In the three other cases, there is a long vowel but no identical
consonant.  These were also pronounced as vocal schewa's
because they fit in one of the following categories:
1) under the kaf of )kl 'to eat' when the lamed has seghol
2) under the lamed of hlk 'to go' where the accent is
retracted to the first syllable and the following word has
conjunctive dagesh.
2a) Saadya Gaon identifies rule (2) as a specific case of a
more general rule: when the vowel preceding the schewa
has stress and the second vowel following the schewa
also has stress and a dagesh.

Thus Isaac's Num 18:10 is an instance of rule 1.
Isaac's Gen 18:21 is an instance of rule 2a.

There are additional rules listed by the Masoretes for cases
where the schewa is to be vocalized.  In general, these

I do not have a ready rule to answer the case of Jer 22:28.
Perhaps, this is a case where the loss of the vowel (the
hatef in this case represents an original vowel of the word,
compare the singular hut'al) did not go into effect due to
the slightly different intonation of the word that is given
to it in its function as a question.

Since Isaac and I figure so prominently in this discussion,
I thought it might be interesting to point out another word
that has a hatef patah in some manuscripts.  The next
to last word in Gen 21:6.  This is very close to rule 2a
except instead of a dagesh we have a maqqef that joins
two letters.  I am sure it is there in Isaac's school bible
because it appears that specifically the Mikraot Gdolot
version has it and Isaac's school bible version appears
based on the Mikraot Gdolot.  But then, the name Isaac
from the Septuagint transliteration of my name shows
you that there was originally a vowel there: yisaxaq =
isaac (the guttural was not transliterated, but the two
a's show the vowels before and after the guttural).

Yitzhak Sapir