From: Arnaud Fournet
Message: 61676
Date: 2008-11-16
----- Original Message -----
From: G&P
To: cybalist@yahoogroups.com
Sent: Sunday, November 16, 2008 9:50 AM
Subject: RE: [tied] Re: oldest places- and watername in Scandinavia
>How reliable are reconstructions of non-IE families?
We'd really need to hear from someone who works in that field. Trouble is,
some who work in that field are convinced by evidence that is unconvincing
to others. Nonetheless, some things are clear, particularly the patterns of
what belongs with what and what is still to be proved. Altaic appears to be
a disaster area. Was it a family? How does it relate to Uralic? Some
non-IE families, such as Polynesian, can be fairly well reconstructed. The
most exciting area is perhaps A-A, because of possible links with PIE. But
I gather that there are big problems there too. In general, no other family
is as well known or well understood as PIE, mostly because of lack of
sufficient diverse early evidence.
Peter
=======
Chamito-Semitic is known from much earlier sources than IE.
Look at Semitic (at least -2600) and Egyptian (-3300).
Egyptian is indeed the language whose lifespan is the longest known to us :
about 4800 years.
I'm glad you are leaving open the possibility of links between CS and IE.
I will just briefly say that I prefer the "old" word Chamito-Semitic (CS),
because my opinion about the word Afrasian or Afro-asiatic is that this word
was invented by Greenberg to sell the myth that the perimeter of CS was
known once and for good.
This word is basically a marketing operation by Greenberg mixed with some
kind of scientific putschism that previous works amount to nothing.
I don't want to give any credentials to that myth which is part of the
problem we now have in general when dealing with macro-comparative issues.
Rick once asked what was to be thought about Ehret's Toward Reconstructing
Proto-Afro-Asiatic, 1995,
I can give my own little point of view about this book,
I have not read the book thoroughly, but the reasons for that will appear in
my sort of review itself.
I have not tried to find a "real" review of it. I suppose someone did that
before.
I think this book is symptomatic of the current status of that field in a
way.
Before reading Ehret, I had read Bomhard & Kern 1994,
and I was expected something of that caliber, something Big,
and I've been hugely disappointed.
I don't agree with everything in Bomhard & Kern 1994,
but at least, this is serious work, when Ehret's book is not.
To put it very shortly, to those who have never opened that book,
most of the pages of the book are columns where supposed roots of CS are
exemplified by "roots" and words from the subfamilies of CS.
Some features are conspicuous and especially disturbing.
1. Berber languages are entirely wiped off,
I guess there is not a single word of Berber in the book.
Imagine Pokorny without Germanic !?
I suspect Ehret cannot read French and dodged that problem.
In spite of the complete disregard of Berber data, Ehret nevertheless makes
the claim Berber should be separated from Tchadic and grouped together with
Semitic and Egyptian.
This claim is based on nothing and I'm afraid it's definitely wrong (for a
certain number of reasons).
That idea is based (I'm afraid) on the deep lack of knowledge of Berber and
the superficial shared innovation that Arabic and Berber have emphasis
instead of Glottalization.
This argument is very shallow and proves nothing as Proto-Semitic clearly
had Glottalization.
And this is explained in Bomhard & Kern 1994 and I agree,
for the reasons explained in Bomhard & Kern 1994 and others that can be
added.
2. Semitic is "dealt with" by Arabic words,
This is clearly not enough as these words should at least be put back in the
context of (proto-)Semitic.
And a workable proto-Semitic system should at least be described and
explained,
3. Egyptian is "dealt with" thru the traditional graphemics of the
hieroglyphic latinization,
There is (about ?) never a word of Coptic nor an attempt to reconstruct what
Egyptian words could be like.
As least, a discussion of what the hieroglyphic latinization stands for as
phonemes should be done,
Here again, I suspect Ehret cannot do that by himself.
In other words, _half of CS_ is not dealt with in a satisfactory way if
dealt with at all.
To be unkind, I'm afraid "Boreafrasian" is not a subgroup of
Chamito-Semitic,
it more looks like the part of the family where Ehret's competence is low or
non existent.
It sounds very harsh but I'm afraid this is the real situation.
4. Other families (whose perimeter is unknown as I said before) are
"exemplified" by "proto-roots",
There is no possibility to check what these proto-roots are worth
as there is never a single word from a real language.
Imagine Pokorny without a single word from a real IE language !!
This is what this book actually is.
5. Ehret makes the claim he has avoided "over-imaginative" comparisons,
unfortunately, between 5 to 10% look so,
6. another disturbing feature is the tendency to regloss glosses of
proto-roots,
for example four items meaning "to cut with a knife" reglossed "to separate"
this (near fictitious) example creates the impression that thru "semantic
drift" of reglosses, many compared items do not match when looked at at the
elementary level of primary languages,
but this cannot be checked as (once again) there is no real lexical data on
real languages anywhere.
And to be frank, (I'm not far from thinking that) the absence of a clean
perimeter of languages and subgrouping and the absence of any real lexical
data must be (interpreted as) a kind of strategy to make the book somehow
immune to concrete criticism. I'm afraid this strategy of "fuzzying all
issues and datas" also achieves the goal of making the book useless either
as a reference or a working tool.
The next part of the title of the book is "vowels, tone, consonants, and
vocabulary"
It's never explained why we should hypothesize tones in CS,
and what they could have become in languages with no tones.
It's also clear that the reconstruction is not satisfactory as *p and *f are
distinguished, but on what ground ?
There are a certain number of syllabic monsters and things like *Caà and
other freakish phonological items.
We are left to imagine what the book could have been with a reasonably
ambitious target.
It seems it could have become a reference book on reconstructing Cushitic
and Omotic (and maybe Tchadic).
Unfortunately, it's nothing like that, and it's nearly nothing at all,
as the title is misleadingly ambitious and the contents is "strange".
All in all, it's little wonder this book was not published in a serious
house.
The book's title is in a way unvoluntarily humouristic.
Imagine you are a termit and you can build a five-meter high termit-hill,
now you can bravely name it "Reaching the Moon".
I think the distance between the book and "Reconstructing Proto-Afroasiatic"
is about that size.
Sorry for the harshness,
Best.
A.