Biblical Hebrew Forum

[Ratson Naharadama <yahoo-arch AT heplist.com>]

Regarding the question of how we know Akkadian, and other dead semitic 
languages, relied on 3 basic vowels (usually labeled a, i, and u):  This 
is done through analyzing the words, how they changed through time and 
region, and their reflections in other languages that we do know the 
pronunciations for. Because of the set structure of the human mouth, 
from which we can fairly confidently assume that these languages were 
being articulated, the mutation of the words can be reasonably 
accurately be predicted and "back predicted". Such things as the 
y-consonant frequently degrading into an i-vowel, the w-consonant 
frequently degrading into a u-vowel are the first big hints to what the 
value of those vowels are. That these vowels are regularly interchanged 
with different logograms and phonograms as if they all represent the 
same vowel is also a big hint, after the process of elimination, that 
there were 3 basic vowels from which these languages structured their 
words.  How these vowels affected neighboring consonants, and how 
consonants affected neighboring vowels reïnforces these big-hints.  That 
these words are transcribed and sometimes borrowed into other languages 
which retain the basic vowels (sometimes coloured through the known 
processes that would colour them between these languages) reïnforces it 
further (especially when a word gets borrowed into several widely 
different languages directly from the source language, and each of these 
borrowings are consistent with each other).

This same kind of analysis has been used on many different language 
families (most particularly the descendants of the PIE and Proto-Semitic 
languages), and new written discoveries over the decades has shown that 
the predictions made by these linguists have been fairly accurate.  With 
the analysis — which was briefly given a few examples in the first 
paragraph — there comes a point where one stops saying "it is possibly 
that" and starts saying "it is highly probable that" or even "we 
confidently say, with very little chance of error, that."

It is because of the fixed structure of the mouth, and the regular way 
in which certain utterances are perceived by the hearer's ear (one of 
you is very fond of "perception") that the predictability of this works 
so well.

Might I recommend to listers that they read or listen to some of the 
work of Seth Lerer and John McWhorter.  I choose them in particular in 
this case because you can get lectures by each of them from The Teaching 
Company (TTC) so you don't even have to actively read any of their 
books, just lazily lay back and listen. :) One is Seth Lerer's "History 
of the English Language" (second edition), and the other is John 
McWhorter's "Story of Human Language."  Both of these lectures touches 
on how linguists do what they do, both give plenty of spoken examples, 
and McWhorter's lecture does go into the Semitic languages as well.  
Both lecturers work consists of 36 half hour lectures (so there is 
plenty of information).  A quick search has shown me that you can 
download these using torrentz.com, but it looks like these may be 
pirated copies, so be aware that doing so would be illegal (which is not 
recommended).  Instead, you may want to go to teach12.com which is TTC's 
website.  Of course, you can always get their books (and other people's 
books) on the topic, as well. :)

--
Ratson Naharädama
Denver, Colorado