Arabic and IE (response to Dr. Georg)

[Alexis Manaster-Ramer]

----------------------------Original message----------------------------


On Mon, 1 Feb 1999, Ralf-Stefan Georg wrote:

> But, Alexis, while you know me to admire Doerfer even more then you do (;-)
> ?), I don't see a reason for mentioning his name in the context of AA. It
> is true that he said those things
[sc. that Semitic is not derived from the same source as Cushitic--AMR]
> But is also true that he has never
> worked on it (he knows Arabic well, but that's it),

I pointed that out myself.

> he just voiced his
> general doubts on AA in some footnotes or footnote-like passages, trying to
> fight critique against his anti-Altaistic positions off

I dispute this too.  These are central points in his work, e.g.,
his famous paper on Japanese and Altaic, and they are not
responses to criticism but rather a crucial part of HIS
attack on Altaic.  His arguments against Altaic take the form
of certain universal claims about what lg families MUST look
like (e.g., that in a REAL lg family the word for '2' through
'5' must be cognate throughout the family) and the further
claim that Altaic does not satisfy the universal, therefore
is not a real family.  For these arguments to work, AA,
Uralic, IE (and some other lg families which even Sally
Thomason or Larry Trask would probably accept as real but
which Doerfer does not discuss) would HAVE to be spurious
because they also do not satisfy his universal claims.

Of course, the reality is that AA, IE, Uralic, etc.
(AND Altaic) are real families. It is Doerfer's universals
that are nonsense.  Same thing re various arguments of
Serebrennikov, Shcherbak, Clauson, and others against
Altaic.  And in my view the same thing applies to the
methodological claims of Lyle Campbell, Ives Goddard,
Sally Thomason, Donald Ringe, Johanna Nichols, and some
others.  They make up various methodological rules for
comparative linguistics, then claim (sometimes correctlty,
often not) that these rules were flouted in the setting
of proposed language families they do not like (e.g.,
Nostratic or Pakawan), and hence think they have an
argument that these families are spurious.

But, of course, as I (and others) have shown time and
again, if you took these rules seriously, then IE
and AA and Niger-Kordofanian and many other lg families
we all accept would have to be rejected too.  Of course,
it is particularly nice when we find that one and the
same person has him/herself done work which would have
to be thrown out if these methodological principles
were really valid, as in the case of Goddard clearly
and Hamp almost as clearly.  But that is not essential./

It is essential that we learn that a proposed linguistic
universal (as in Doerfer's work) OR a proposed methodological
principle (as in Campbell's, Goddard's, etc.) is violated by
some of the most widely accepted lg families such as IE,
AA, Uto-Aztecan, etc., etc.  The logic is the same.

> He may have been
> ill-advised to do this, but his claims to this effect *do not play a single
> role* in AA studies. I take it that, while every Altaicist - pro or con -
> knows Doerfer, there may be excellent AA'ists who have never heard his
> name. And they don't have to, for G.D. is simply not one of them and never
> wished to be. Actually, though I hate to say this in public about a scholar
> whom I regard in many respects as a (semi-) teacher of mine, what Doerfer
> said on AA was simply uninformed nonsense. Why quote it ?

You are right about the premises but not the conclusion.
It IS uninformed nonsense.
But you only feel free to say that because Doerfer is not
a well-known American linguist active on this list.  But
it is clear from what I have said in this and earlier
postings (and in print) that the claims of Thomason, Campbell,
Goddard, Hamp, Nichols, Trask, and others are no better informed
and no more sensible.  And THAT of course is why I mention
Doerfer.  Because people are blinded by sociological factors
such as prestige, familiarity, etc., and so accept statements
from people like these which really should be actionable
in a court of law.

So I am trying to wake you and others up by a cheap rhetorical
trick. If I can get you to accept that what Doerfer says is
uninformed nonsense (you words, not mine), then show you
that what he says is no better than what yourself hold or
what Thomason, Trask, et al. propound, then perhaps you
will draw the inevitable conclusion.

For example, Ringe has no more right to say ANYTHING
about probability than Doerfer does about Afro-Asiatic
and what he does say (to the applause of the opinion
makers in the field) is no more correct.  Nichols has
no more right to say ANYTHING about Altaic than D about
AA and again what she does say (as you and I jointly
show in a paper with Michalove and Sidwell in JL) is no
more correct.  Trask SURELY has no right to say ANYTHING
about the many lg families he does hold forth on than
Doerfer on AA, and again what he does say is no more
correct.

But I do not hear saying in public that they all (and
sometimes you) are talking uninformed nonsense. But
fair is fair.  And my cheap trick WILL work, and one
day soon you will (unless you are censored).

> AA is really, to
> the best of my knowledge, uncontroversial today, i.e. with those people who
> know enough about the issue to have a say in it.

If you mean by people whom I would consider competent.  But
some (maybe many) Semiticist think they are compoetent and they
do question or even openly reject AA.  And this is relevant because
there is a widespread feeling that someone who knows a lg has
therefore some insight into this lg's genetic connections.
So that an IEnist is assumed to be competent to speak about
Nostratic, a Turkologist about Altaic, and so on.  Larry
Trask just told us on another list that people he calls
Caucanists are competent to speak as to whether East
Caucasian and West Caucasian families are related to each
other.  But he includes people who work on the phonetics
or syntax of Caucasian languages, without inquring as to
whether they do any comparative linguistics.

On THIS list Trask said something about how Goddard
and Campbell may somehow because of their knowledge of
American Indian languages in general be able to see why
Comecrudan is a valid family even though he Trask does
not.  But how exactly does Lyle's work on Uto-Aztecanm
or Mayan or Ives's on Algonquioan give them an insight
into Comecrudan?  Just because it is the same continent?

Does someone who knows Hungarian very well thereby
become an expert on the classification of the languages
of Eurasia?  Does someone who knows Arabic become
an expert on the languages of Africa?

Does Chomsky, who I happen to know from personal
experience knows English and Hebrew, have a right to
speak as to whether IE and Semitic are related?
Of course, not.  And he does not pretend to.

So why do we applaud people who speak of probability
who do not understand things taught in Statistics 101,
people who speak about the methods of linguistic
classification who have never done any work on
classifying anything, people who tell us that
IE is unrelated to Afro-Asiatic who have never
read a page of Illich-Svitych and/or use their
power to prevent the review and discussion of his
work?  Why? Why? Why?

Because of the banality of error (the title of
a book I will finish when I recover from my current
illness, so please dont nobody steal it).  It is
all too easy to acquiesce in error which is all
around us and supported by prestige and familiarity.

THAT is why I mention Doerfer and Afro-Asiatic.
He is not as prestigious OR familiar to THIS
audience (although IN HIS OWN BAILIWICK he is
a truly great historical linguist) and so it is
easier to see and admit that he is talking
"uninformed nonsense".  And that then makes it
possible to see that he is not alone...

> of dissent may exist somewhere - well-informed or not - , to be called
> "controversial".

The equality of women, the right of the people of Timor to independence,
the evolution of species, these all ARE controversial.
We cannot unfortunately SAY they are not and let the forces
of darkness take over. I am sure S. J. Gould has better things
to do than fight the creationists, but he has no choice, and
he certainly cannot make them disappear by saying they do not
exist.

> What then isn't ? Is the proposal of a Basque - Armenian
> connection (to the exclusion of IE) "controversial" ? At least one person
> out there holds it (and I think only one, and this was on an e-mail list,
> but these things *do* sometimes get published !).

If that person were running the official journal of the Linguistic
Society or writing a widely used textbook or encyclopedia of
lx, you would HAVE to deal with it, as Gould deals with creationism
and as I try to deal with, say, Ringe's "probabilities".
As it is, you don't have to worry about him (her?) TOO much.
If it DOES get published, well, I have an easy criterion:
if it is published ina journal which refuses to publish
or review work on Nostratic, then you DO have a responsibility
to fight.  Otherwise, you have no moral responsibility
as far as I can see.

> If I write in one of my
> next papers that I don't really understand what makes people believe in the
> validity of such an abominable thing as "Algonquian", while everybody knows
> that Cree is Siouan, Menomini Na-Dene and Fox Sino-Tibetan,  will you call
> Alg. "controversial" and point people to what I've written ? No, you'll
> drop me a note and ask me whether I'm feeling OK and (hopefully) forget
> about the issue, assuming I was temporarily out of my mind (if I'm lucky !).

Of course.  But I have written you such notes re your opposition
to Altaic, I have tried patiently to explain about Nostratic
and the Nostraticists to Sally Thomason.  I have tried patiently to
teach Ringe some elementary mathematics.  I have tried sending
notes to Larry Trask and I have spent hours with Eric Hamp.
It does not work, though.  But discussions like the present one
MAY.

> Is Sino-Tibetan controversial  (Roy A. Miller fights it ruthlessly, but he
> seems to be the only one) ?

I seem to recall Sagart attacking ST as well. Certainly the
state of ST lx is not satisfactory.

 Is Uralic (Doerfer in his younger days was
> quite skeptical about it, but again without having actually worked on it) ?
> Is Indo-European (remember Trubetzkoy !) ?
>
I mentioned both of these myself.

> Let's stop this futile discussion, if I don't think that two or more
> languages are related, I say so, assuming that everyone listening to me is
> educated and intelligent enough to know that I'm a human being only and by
> this virtue - fallible.

Eric Hamp has noted in print that you should NOT say that.
I defer to him on this point.

> Everybody who asks around will find that there are
> actually people who do believe that, say, Altaic is valid grouping. I
> happen not to. And when I say "No, I don't think it is", this is an
> accurate information. About what I think, that is. Even saying "It isn't"
> is not more than that. Every intelligent and truly interested person will
> then go on asking me what makes me so positive about it, and I'll have no
> choice but to mention all those names, including yours, who hold different
> views. If I'm not asked further, so what, then the asker doesn't deserve
> better.
> What's the problem with that ?
>
None, IF the asker knows that you are just expressing YOUR
opinion and that the issue is controversial.  But then
why would (s)he ask?  Certainly the person who asked whether
Arabic might be related to IE did not know that the
people to tried to misinform him were merely expressing
THEIR opinion and that the question was controversial.
I am not sure they themselves knew that either.  That's
why I posted my long posting about Nostratic.

Alexis