Linguistic Origins of Hebrew: 1800 BCE?!

Henry Churchyard churchh at
Thu Sep 21 05:18:27 EDT 2000

> From: Ian Hutchesson <mc2499 at>
>> From: Henry Churchyard
>>> From: Ian Hutchesson <mc2499 at>
>>>> From: Henry Churchyard

>>>> I haven't offered any opinion as to when in the second millennium
>>>> B.C. Hebrew might have been become a distinct language, because I
>>>> don't have any strong opinions on the matter -- and in fact think
>>>> that the question is not precisely answerable, given the scanty
>>>> linguistic data available on "Canaanite" dialects/languages in
>>>> the second millennium.  If you pressed me to name a date, I might
>>>> throw out 1200 B.C., since that's the approximate period when the
>>>> sound change of the loss of word-final short vowels occurred -- a
>>>> change which caused a restructuring of the Hebrew stress system,
>>>> marks of which are still very visible in the Tiberian masoretic
>>>> patterns of placement of orthographic accent symbols roughly
>>>> 2,000 years later.  However, I don't in fact have any hard
>>>> evidence that the restructuring of the stress system was involved
>>>> in the differentiation of Hebrew from related languages.

> On what grounds can one date a linguistic change without having
> (what I consider to be) the only available indications of our
> knowledge of ancient languages (ie texts)?

First you reconstruct relative chronologies of sound changes (tied to
absolute dates in those cases where we have explicit external
evidence, but chronologically free-floating where there is no
available external evidence about exact exact dates of changes), and
then you extrapolate between the chronologically fixed points to
derive rough estimates as to dates of intermediate reconstructed
changes.  Of course, here it's a lot more complicated than that,
because our only available evidence for fixed chronological points
often applies to dialects other than the one we're interested in.  For
example, the Amarna letters are interesting, but it's not clear that
any of them (even those written from Jerusalem) provide evidence about
a dialect which is a direct ancestor of later attested biblical

Not quite sure why you're making so much of this, since I freely
admitted from the beginning that I only had a rather approximate date
to offer for word-final vowel loss in Hebrew, and that I was not in
fact very sure that word-final vowel loss was directly connected with
the differentiation of Hebrew from related languages.  You urged me to
throw out some kind of hypothesis, and that was the one which occurred
to me -- and I never claimed that it was anything other than a very
tentative one (though even so, it's still less problematic than the
idea that Hebrew first differentiated from Phoenician under Aramaic
influence in the 9th century B.C.).

> Is the Gezer calendar in some way involved in the construction of
> that date which is "accepted without much controversy"?

Nope -- the Gezer calendar hardly even indicates long vowels (much
less short vowels).

>> in some cases the sound changes themselves can be reconstructed
>> pretty firmly (without much doubt), yet the evidence on the
>> absolute dating of such sound changes can remain rather vague;
>> that's just the way the game works...

> In our particular case how can one date any of the sound changes
> talked about in relation to Hebrew when the only "sure" early Hebrew
> texts we have are from the period of Hezekiah?

Languages are not texts, remember?

In any case, for a number of the sound changes that I've been
interested in (having to do with vowels and stress), not even the
indisputably Hebrew inscriptions of the 1st. millennium B.C. are very
relevant, since such Hebrew inscriptions only transcribe a restricted
subset of long vowels (with no information about short vowels or
stress).  Therefore such phonological work basically involves directly
correlating the masoretic Tiberian orthography with the early
Canaanite vowel system (derived from the proto-Semitic three-vowel
two-quantity system by the change of most long [a] to long [o] and the
monophthongization of at least some [ay] and [aw] to long [e] and [o]
vowels), and reconstructing the necessary intermediate stages, as
discussed at length in section 1.4 of my dissertation (available

>>> The only thing that remains is why you haven't read more than one
>>> article of Garbini's, given that he has written quite a lot in
>>> your field

>> Actually, in the Hebrew/Semitic area, my field is historical
>> phonology.  I don't think that Garbini would describe himself as
>> being primarily a phonologist

> He has however written on most aspects of Semitic linguistics, many
> of which by necessity touch phonology.  At other times he has
> written specifically on phonological concerns or concerns with
> phonological importance.

I just looked up Garbini in the electronic MLA bibliography, and it
turned up a total of nine items by him published since 1962 (which is
about as far back as the bibliography goes), of which only one, the
article I mentioned earlier ("The Phonetic Shift of Sibilants in
Northwestern Semitic in the First Millennium B.C.") is obviously
phonological in focus based on the title or the supplied subject
headings.  Of course, the MLA may be spotty on material published in
Italian (but unfortunately, Italian is not really one of my languages

>>>>>> Two geographically-separate speech communities don't need any
>>>>>> influence from yet a third language to diverge from each other
>>>>>> linguistically;

> But of course influence from a third if present will have its impact
> on the divergence, will it not?

Certainly -- if the third language influences one of the two more
heavily than the other (and so acts as a differentiating factor rather
than a common factor).  But it's not a necessary condition, so that in
most cases of two related languages a historical linguist will feel no
particular need to seek out yet a third language in order to provide an
"explanation" for the cause of their differentiation.

>>>>> On what grounds do you assume that we have two "geographically-
>>>>> separate speech communities"?  Were not Phoenician products
>>>>> found throughout the uplands before the arrival of the
>>>>> Assyrians, then again in the Persian period?

>>>> Phoenicians traded quite extensively, but at any given time the
>>>> mass of speakers of southern/Judaean Hebrew (the predominant
>>>> language of the Hebrew Bible) were located 60 miles or more from
>>>> the mass of the speakers of Phoenician.  Contact between speakers
>>>> of related languages is one of the things that creates a "dialect
>>>> continuum", but it does not necessarily nullify basic
>>>> geographical separation.

>>> Did such a separation stop the hellenistic cultural invasion of
>>> Jerusalem? ... high priests called Jason and Menelaus?  Nor does
>>> that geographic separation guarantee the disinterest of people in
>>> one area wanting to immitate the mores of those in another.

>> Again, cultural influences only prove that there were contacts
>> between the speakers of the two languages; for you to prove that
>> the 60 miles distance was absolutely linguistically irrelevant, you
>> would basically have to prove that it was not the case that
>> speakers of Hebrew generally spent more time interacting with other
>> speakers of Hebrew than they did interacting with speakers of
>> Phoenician.  (Not sure why you pull up Jason and Menelaus, since
>> the Phoenician language was moribund in Phoenicia by their date,
>> and it wasn't really Phoenician culture that they were influenced
>> by...)

> It was an analogy: the Seleucids (in whose time Jews were using
> Greek names, texts began to reflect hellenistic philosophical
> concerns, and Jerusalemites were taken on various hellenistic
> cultural traits) were a lot further away than the Phoenicians who
> had more direct contact with Jerusalem through trade.  I asked "Did
> such a separation stop the hellenistic cultural invasion of
> Jerusalem?", meaning by "such a separation" as found in the 60 miles
> between Judah and the Phoenician coast.  The rhetorical question
> regarding the Seleucids requires a negative answer, just as I would
> think that the implied question about the Phoenicians might.

I don't think your analogy strictly proves anything, since it's not
necessarily the case that patterns of cultural diffusion in the late
1st. millennium B.C. were identical to what they had been in the early
1st. millennium B.C.  Anyway, there were Hellenistic cities located
closer than 60 miles to Jerusalem in Seleucid times -- and all this is
actually somewhat irrelevant, since the main criterion remains the
specifically-linguistic (not vaguely cultural) one I stated above:
"For you to prove that the 60 miles distance was absolutely
linguistically irrelevant, you would basically have to prove that it
was not the case that speakers of Hebrew generally spent more time
interacting with other speakers of Hebrew than they did interacting
with speakers of Phoenician."

>>> The Latin encapsulated in later church decrees would not have been
>>> considered the language of those people who lived in Rome at the
>>> time of Julius Caesar.

>> Not sure what you mean; are you denying that Biblical Hebrew as we
>> find it in the Bible is predominantly based on a Jerusalem-area
>> dialect (regardless of whatever absolute date we might or might not
>> choose to assign to it)?

> I really don't have a clue what the language of the bible is based
> on.  It is a form of Hebrew which doesn't reflect the DSS, which is
> an enormous sample of Hebrew, so enormous one has to conclude that
> the scrolls as a bulk could only have been produced in Jerusalem.
> There are at least four forms of Hebrew in the scrolls, a few
> examples of MH, as well as BH, Qumran scrolls specific Hebrew, and
> what seems to be a Hebrew related to Samaritan, so the origins of
> some of these forms are not transparent.  This leaves us still to
> understand how biblical Hebrew fits in.

Your perplexity seems to be mostly the result of your fixed _a priori_
preconception that these different varieties couldn't largely reflect
different historical/chronological stages (leaving Samaritan Hebrew
out of consideration).

>>>>>> In any case, if you look at the Gezer Calendar, what you seem
>>>>>> to find is a generic schoolboy exercise, containing a short
>>>>>> formulaic text without any complete sentences or
>>>>>> finite-inflected verbs, and written in the older orthography
>>>>>> which was rather parsimonious in the use of _matres lectionis_.

>>>>> Coincidentally parsimonious in the use of matres lectionis!

>>>> What's coincidental about it? Lack of _matres lectionis_ was the
>>>> original state of affair, and the use of _matres lectionis_
>>>> developed fairly slowly over the centuries (until you eventually
>>>> arrive at the Yiddish or late Punic situation).

>>> It is this lack which is part of the evidence.

>> Not really; no one has disputed that speakers of Hebrew (and
>> speakers of other languages, including Aramaic) were influenced by
>> Phoenician orthographic practices when they first adopted the
>> alphabet to be used in writing their languages.  The Hebrew and
>> Phoenician written traditions developed the use of _matres
>> lectionis_ at different rates, but their starting point was
>> basically the same, and the Gezer Calendar is fairly close to the
>> starting point.

> This seems to assume that you accept the Gezer calendar as Hebrew
> rather than Phoenician

Nope -- just that Phoenician orthographic practices were influential
when the 22-letter alphabet was adopted to write various
non-Phoenician languages in the early first millennium B.C. (such
orthographic influence doesn't have much to do with linguistic
relatedness, since Aramaic orthographic practices were even more
heavily influenced by Phoenician than Hebrew orthography was).  If
you're really interested in this subject, you could look up Blau's
"Pseudo-Corrections in Semitic" book...

Henry Churchyard   churchh at

More information about the b-hebrew mailing list