Biblical Hebrew Forum

[Yitzhak Sapir <yitzhaksapir AT gmail.com>]

Dear Gabe,

If you had any worries about a lull in list activity, your post can sure
generate some.  You would hear many voices and opinions here, and
some of them are based on no evidence at all.  Kevin Riley's post is
a good starting point based on actual evidence and studies (this topic
has been studied at length).

It becomes problematic to study the changes prior to the manuscripts
that are vocalized by the Masoretes.  Such study must be done
carefully.

Regarding the reading of waw as [v] rather than [w], its counterpart
in standard Semitic languages and also in many Oriental communities,
the reading [v] probably took place in the early centuries CE.  At that
point we see confusion, for example, between writing ybnh and ywwnh
for Yabneh.

The above is also example that by this time the b was pronounced
like a [v].

The linguistic process by which the above took place is called
spirantization.

The earliest evidence for spirantization comes from the 8th century
BCE Aramaic.  In order to understand the evidence it is necessary to
understand that Semitists reconstruct more sounds in Semitic than
the 22 letter alphabet.  Leaving aside the question of why the 22
letter alphabet was adopted, when more sounds were present in a
given language, a good basic idea to keep in mind is that writing is
just a poor representation of speech, and it is what people speak that
defines language.  It is still possible to use the 22 letter alphabet by
using similar sounding letters for multiple sounds.  Thus, for example,
shin and sin sounded similar enough to be written as one letter.  By
now, the pronunciation of both these letters, as well as samekh, has
changed, so the similarity is not as apparent as it used to be.  Shin
was also originally used to write another sound, [th] (like in 'thin'),
in languages that had it.  Although it no longer survives in Hebrew, it
and similar additional sounds are the key to understanding the
earliest evidence for spirantization.  In the Sefire treaty, c. 740 BCE,
the word yr$ 'he will inherit' is spelled yrt.  The root 'inherit' yr$
actually originally had the additional [th] sound of Semitic rather than
the [sh] sound.  Again, in Hebrew, these sounds eventually coalesced
so that the roots seem to be the same, but Semitists would treat y-r-sh
and y-r-th as as completely different sounds and roots.  The choice of
the author of the Sefire treaty to use the letter taw rather than shin
appears to indicate that in post vocalic conditions, t was pronounced
similar to th.

The Sefire treaty discussed above was written in Old Aramaic.  In
the next stage of Aramaic, Official Aramaic, various sounds were
spelled differently.  Thus, instead of shin for the additional th sound,
taw is now used.  In fact, this accounts for all the spelling differences
between Official and Old Aramaic.  This seems to indicate that
Official Aramaic also had spirantization.  The process of spelling
changes was one of diffusion and studies in the Elephantine papyri
show how a [dh] (like in that) is spelled with a zayin and sometimes
with a daleth.

Official Aramaic became standard with the rise and its adoption by
the Assyrian empire.  It remained the official language until the fall
of the Persian empire in the late 4th century.  At that time, many
people were speaking Aramaic, but there was no longer
communication between various Aramaic communities in Aramaic.
(Instead, the conversation took place in Greek).  As a result, the
language of the various Aramaic communities began to diverge
from each other.  In Judea, this led to the rise of Judean Aramaic.
All Aramaic communities have the spirantization, so it is clear that
spirantization took place before the fall of the Persian empire.

In light of the above, linguists generally date the process of
spirantization to the time of Official Aramaic, approx, 7th - 4th
centuries BCE.  The standard explanation is that with the
sound shift, various sounds were confused and this led to the
spelling shift noticeable in that period.

Hebrew and Aramaic are not the only languages that have
spirantization.  There is some evidence from Latin transcriptions
of Punic (a later stage of Phoenician) that Punic also had it.  In
such cases we see Latin use p or ph for pe in one case and f in
another.

The real problem is how to explain the adoption of spirantization
by Hebrew (and possibly Punic) speakers.  My personal conclusion
is that this requires us to reconstruct spirantization for the earliest
stage that included Hebrew, Phoenician, and Aramaic, a stage called
Northwest Semitic.  The reason we don't see it in writing for so long
is because, as mentioned before, writing is just a poor representation
of speech.  Additionally, the original spirantized variants may have
been slightly different than the additional th and dh sounds so they
were not confused.  However, the general consensus today is that
Hebrew speakers somehow adopted the spirantization from their
speech of multiple languages and applied it to Hebrew as well.
Various words that were coined in the last centuries BCE and
the early centuries BCE (in Mishnaic Hebrew), no longer have
spirantization.  This suggests that spirantization stopped being an
active sound rule in Hebrew at that time.  Linguists believe that it was
originally an active sound rule because of the great uniformity of its
application in Biblical Hebrew.

In light of all the above, it is sufficient to say that spirantization
was a rather early process and was active during Biblical times.
In German diaspora communities, th underwent the common
German change to s, leading to the Ashkenazic pronunciation
of [s] where [th] is used by other communities.  European
communities in general seem to continue the Israeli pronunciation
of w as [v].

The Tiberian Biblical vocalization system does not allow you to
determine syllable structure in a complete and unambiguous way.
You would therefore not be able to determine whether to pronounce
a bgdkpt letter hard or soft just from reading the vocalization without
the dagesh (or the rafe).  Additionally, the dagesh is used to
indicate doubling which in general in Semitic is not represented in
writing (again, writing is a poor representation of speech).  All this
means is that you have to know ahead of time how to pronounce
the word.  The vocalization ends up being just a guide, a very
detailed guide, but a guide.  If you depend purely on the vocalization,
you'd not know syllable structure, you'd not know the complete
application of bgdkpt as soft/hard and where a dagesh is hard or
soft.  However, you can fake it in most cases.  For example, in
the word Shabbath, you'd see that the dagesh in the b occurs in
post-vocalic position.  According to the rule, the spirantized variants
are used in post-vocalic positions.  Therefore, because the dagesh
here indicates the non-spirantized variant, this dagesh has nothing
to do with bgdkpt and is a doubling (hard) dagesh.  The rules only
get you so far, though.  Due to some minor changes in pronunciation,
including new additional vowels in what became the Tiberian
pronunciation after bgdkpt stopped being an active sound rule, there
are still exceptions.  Without previous knowledge of pronunciation to
fall back on, you'd not know how to deal with the exceptions.

As far as doubling goes, doubling is viewed as length of the
consonant.  The Masoretic literature records that t is present in three
variants of length.  So in that case, there is also a "one and a half"
length for the doubling of the t.  According to Ben Asher, it happens
in the word btym "house" in Deut 6:11 and 1 Chr 28:11 (when the
word has both qadma and azla on it).  According to Ben Naphtali,
it happens in all cases of the word btym that have a double accent.

Rabbi Aqiva would have called himself Aqiva, whenever he called
himself (or when his colleagues referred to him).

Yitzhak Sapir