Biblical Hebrew Forum

[Yitzhak Sapir <yitzhaksapir AT gmail.com>]

On 9/21/05, Karl Randolph wrote:
> Yitzhak:
>
> I don't know why you are putting so much effort to trying to
> convince me that your reading of language development
> is correct, if not to validate to yourself because of your own
> self doubts. I am not trying to convince you that mine is
> correct, just claiming that the surviving data is too sparse
> to say.

You can rest assured that I have no self doubts.  That is, I
can recognize that Dr. Steiner's method is a much more
systematic and measured one -- having to reconstruct exact
minimal pairs in order to prove not just that the two phonemes
existed but that they were serving independently in the
language and the independence is what minimal pairs is out
to prove.  But the difference between reconstruction of g' and (
versus reconstruction of th, $, t, &, and s, is different (But I
guess 20 minimal pairs have to be produced for these, altogether).
In any case, you've stated in the past, in the context of the
XSD thread that very few people reacted when you suggested
one of the related theories and you took that as a sign that
you're right.  So, I have a feeling that since every time people
stop responding to your claims on this issue, you think this
issue is some kind of "undecidable from evidence" claim,
and it's not.  It's just that you are too stubborn to even
look at the evidence.

> Take for example, when did Aramaic and Hebrew split? If
> we take Biblical dates, that was when Jacob left his father-
> in-law Laban at about 1800 BC, give or take a century. In
> other words, from then to Biblical Aramaic, we have a
> separation of about a thousand years, plenty of time for
> extensive language shift in both Hebrew and Aramaic.

Why, why, why do you keep trying to drive the discussion
back to this issue?  Dr. Steiner respects the bible no less
than you and yet proceeds to use normative linguistic
theory for his reconstructions.  In any case, if you read
the Bible literally, it would suggest that the split happened
before this period.  But that a split of some kind happened
before this period doesn't mean that the different phonemes
didn't survive for a long time.  From evidence in New Kingdom
Egyptian loan words, there appears to be a constant
differentiation between th and $.  In the Execration Texts,
dated to the 20th-18th centuries, Ashqelon [from th-q-l,
weigh] appears with an s', and Ashterot [from (-th-t-r] appears
with an s, while Beth-Shemesh and Jerusalem, appear with
an $.  This type of s/s' - $ difference is pretty consistent also in
other roots that are traceable to th - $ respectively.  The Torah
itself contains clearly only Ashterot, but also mentions Shalem
which is traditionally taken to be Jerusalem.  Joshua mentions
all of these, and also Ak$ap.

> Further, when was the alphabet invented and used for
> Semitic languages? The earliest examples I have seen
> claimed in the literature is from about the same time
> period.

What does it matter what you have seen when you choose
to selectively look at the evidence?

> Alphabetic use could have been in existence for
> centuries, or was new. Do we have indications from that
> period as to how many letters it had at that time?

Epigraphic South Arabic uses 29 letters that are an
independent and earlier divergence from Proto-Canaanite
than the Hebrew/Phoenician/Aramaic one.  Some of the
additional Proto-Canaanite alphabet letters can be clearly
related to the ones in ESA.   So yes.

> How many extensive writings from that period remain that
> are recognizably Aramaic or Hebrew?

If you could prove to me that the Torah is written in Hebrew,
and not in Aramaic or Phoenician, then we can discuss this.

> Can we build a time
> line from that period to the Biblical usage from that era?

Generally, yes.  It really depends on what you mean by time
line and what you mean by Biblical usage.

> If the answers to this paragraph are "Don't know" "Don't
> know" "None" and "No", who has the hubris to claim to be
> able to "prove" language development from that period?

Perhaps you should acquaint yourself with evidence before
firing questions at random.  Various evidence at different
times generally allows us to place in relative context when
various phonemes merged.

> Now assuming that ancient Hebrew, Aramaic and other
> Semitic languages acted the same as do more modern
> languages, we find even more reasons not to be able to
> state categorically how ancient Aramaic and Hebrew
> developed.

You tend to assume a lot, and you know what that brings
about.  If you assume, at least do it *after* you've examined
the evidence.

> Take just the example of Latin, a language for
> which we have historical records: its spelling became
> frozen so that even when pronunciation changes occurred
> they did not show up in the written record until later, and
> then as new spellings in Spanish, French, Italian and
> other Romance languages. Yet even then, medieval to
> modern pronunciation of Latin is more based on French
> and Spanish pronunciation than the original Roman one.
> In particular, the letter "c", which used to be consistently a
> "k" sound, now is given a "s" pronunciation before "e" and
> "i", though there are other examples as well. But when the
> spelling is allowed to change, then letters tend not to
> change their values, as when the German Stratte became
> Strasse or the Greek thallattes became thallasses.

This is essentially how Dr. Steiner suggests g' might have
evolved, by proximity to the trill r.  However, the shift of c
to s did not create a new phoneme.  The /s/ phoneme
already existed.

> Again assuming that ancient Hebrew acted the same as
> more modern languages, as languages first adopt an
> alphabet, as long as the writings remain short on the order
> of graffiti, an ill fitting match is often tolerated. But once
> legal documents are written in the language, such as
> business contracts or extensive religious texts, people will
> either use a well attested other language (as did the
> Chinese (whose legal language no one spoke before the
> adoption of Mandarin as the legal language), medieval
> Europeans Latin or the example you linked to of the early
> Arabs using Aramaic) or to avoid confusion they adjust the
> alphabet by subtracting and adding graphemes to match
> the language.

So if someone was to show that as late as 100 BCE,
when legal texts and contracts and religious texts were
written in Hebrew, the phonemes of h_ and h. were still
differentiated, would that prove it to you?

Also, Are you calling the inscription from Tell Fekherye a
"grafitti"? It is both a legal and religious document.

> Therefore, unless you can prove that the
> early Hebrews were using an ill fitting alphabet when
> Moses wrote Torah in the 15th century BC, you cannot
> rule out the historically attested to pattern from other
> languages indicating that the 22 graphemes represented
> the 22 consonantal lexemes recognized by the ancient
> Hebrews.

Don't the above examples regarding place names in the
Execration Texts prove this?  Would place names in
Thutmose III's list be sufficient? He's usually dated to the
early 15th century.  What would prove it to you?

> One area where we take almost the opposite tacks is in
> the use of cognate languages: I take the view that while
> cognate languages can sometimes, in rare cases, clear
> up something that is otherwise unclear in a language,
> most of the time a study of cognate languages will add
> nothing to the understanding of a language, many times
> actually confuse the issue and/or give misleading
> information, therefore the use of data from cognate
> languages should be done with extreme caution;

You don't "take the view."  You make the assumption.  You
don't view the facts unbiased and let the facts point you
towards which view is correct.  You don't even want to look
at the facts.  That's what this paragraph of yours is all about.

> you, OTOH, seem to take the position that if something is
> attested to in a cognate language, it must also be found in
> Biblical Hebrew as well,

No, I don't.

> If we had native speakers to
> interview, we could ignore the cognate languages entirely
> as the cognates would be extra noise to filter out. The
> main use I see for cognate languages is where
> documents written in them explain concepts or lexemes
> rarely used in Hebrew and where the use in Hebrew does
> not indicate its meaning. Hence I tend to ignore the
> cognate languages as irrelevant (most of the time), while
> you consider their study indispensable.
>
> Seeing as I have a major philosophic difference from you
> as to the validity of cognate language study, unless I
> changed my philosophy, even an extensive study of the
> materials you insist on will not change my mind.

First, you shouldn't mock raising people from the dead.
That belief is no less important than Sinaitic authorship
of the Torah, which you seem to hold highly.  But while
we may not be able to raise people from the dead to
check our assumptions, we can check the very last
assumption you made.  And it's all up to you.  And you're
still alive, I hope!  So look at the evidence, and see if you
change your mind!

> The only
> thing that will make me accept your theory is for you to
> produce extensive alphabetic documents dating from
> about the time of Jacob and Laban, showing the
> development of the alphabet as well as both Aramaic and
> Hebrew from then to Moses,  Since you, by your own
> admission, start with the DSS,.......

I don't start with the DSS.  That's misreading what I said.
What I said is that I don't ignore any part of the evidence
from any time period.  You ignore a large portion of evidence
from the Massoretic period and completely refuse to look
at much earlier evidence.

> Look at the validity of data from cognate languages: if that
> data is to be used only sparingly and with extreme
> caution, willy nilly throwing in such data is GI.

That cognate languages can allow us to reconstruct phonemes
that merged in later languages is an observable fact in modern
languages today. But you cannot claim I'm "willy nilly throwing
in data" until you're able to examine the data yourself to be
able to know, yourself, what is cautious use of the data.

> > Arabic isn't a form of North Semitic.  If you can't verify facts from the
> > wikipedia yourself, don't quote randomly from it.  Wikipedia is a handy
> > resource, but because it is not edited nor monitored by those with
> > expertise on the subject, you should verify everything you read there.
> >
> That's handy of you, first you link to an article, but the
> moment that it disagrees with your theories, you diss
> it. What consistency.

Wikipedia is a constantly dynamic resource.  It's great for getting facts
that are true up to the last minute.  It's also great for getting the latest
rumors and hoaxes that are believed up to the last minute.  To create a
hard link to the current version is something that is rather complex to
do, and so if I give you a link, it can be changed one minute after.  No
one stands behind the information, no one stands behind checking
and verifying the information, no one stands behind maintaining the
information.  I know nothing of the people who edited the article and
what they have contributed before to the subject and what they know
of the subject.  So long as there's no one who stands behind information
then definitely it's you who has to do the checking to make sure the
information is dependable.  It is a handy online resource for repository
of facts.  But you must check them yourself.  I do.  It gives me many
times a great place to start in looking up topics related to some
subject.  But I check each fact.  You should too.

> > Yes, Arabic is not North Semitic.  "North Semitic people"
> > (whatever that means) can speak Arabic.  But that doesn't
> > make the language North Semitic.
> >
> Again, according to the article, the reason that Arabic
> today is not north Semitic is the result of migration.
> Therefore the similarities between it and Ugaritic can be
> attributed to a common origin not necessarily shared by
> Hebrew or even Aramaic.

The reason that Arabic today is not North Semitic is because
Arabic is classified as South Semitic.  That classification is
not a result of migration of people.  It's a result of the relations
between the languages.  Migration of people is a suggestion
that comes after the classification is made and which may be
independently supported.

> > > > You're the one assuming that each root must be uniquely
> > > > represented graphically.
> > >
> > > Excuse me, where did I make that claime?
> >
> > Sorry, I meant the other way around.  Each unique set of
> > graphemes that make up a root uniquely represents one
> > root.  That is an assumption of yours.
>
> Excuse me, and where did I make that claim?
>
> If you can't even reproduce my position accurately, how can
> you critique it?

https://lists.ibiblio.org/sympa/arc/b-hebrew/attachments/20050825/2b0b6f49/attachment.bat
"I started with the assumption that I would be able to find one and one
only distinct meaning per lexeme".  I'm not sure if you've defined
lexemes before but you've treated a set of graphemes that
represents a root as a lexeme.  The above "represents one root"
is equivalent (and refers) to your "one and only distinct meaning".

> > The vocalization information in the MT is as hard data as the
> > consonants themselves.  If you trust the graphemes of
> > consonants to represent phonemic ("speech") values properly,
> > you should similarly trust the graphemes of vowels to
> > represent phonemic values properly.
> >
> If those graphemes representing vowels were from native
> speakers, your claim would be valid. But seeing as the
> last native speaker of Biblical Hebrew died about a
> millennium previous to the invention of the vowel
> graphemes, there is no way that they can accurately
> represent Biblical pronunciation. Accurate memory is
> often lost in less than a generation, let alone a millennium.

Again, so long as you refuse to look at vocalization evidence which
native speakers did provide, you are in no place to look at non-
vocalization information that native speakers provided as well.

> > So are the Massoretic consonants.  If you want to go earlier
> > than the Massoretic consonants you must involve yourself
> > in the entire body of studies around the DSS, the Hexapla,
> > and early Hebrew inscriptions.  So long as you are unable to
> > fully compare the Massoretic Text to the results of such
> > study yourself, you're stuck with the Massoretic Text, for all
> > that it contains.
> >
> Wrong.

Which part is wrong?

> Notice, I did not answer all your points, not by a long shot.
> It is not because I don't have answers, it is just that I've run
> out of time. But neither did you answer all my points.

So long as you refuse to look at the data, we are in no equal
position.  In fact, it was mostly my comments about your
refusal to look at the data, in various forms, that you edited
out.  Stop being selective about data and accepting only
what you assume is proper.  Take all the evidence and make
conclusions from the evidence.

Yitzhak Sapir