The Significance of Comecrudan

[bwald]

----------------------------Original message----------------------------
AMR ventures:
"I don't know what "classical methods" here refers to, unless it is
again the imaginary consensus view.  The point I am making is that
the relatedness of Bantu (together with its most obvious relations)
with (at least some of) the Kordofanian languages, based on the congruence
of the nominal class prefix markers, is not something that, to my
somewhat limited knowledge, is questioned by any sane person.
If I am wrong, please provide me with the relevant references."

[does the comment suggest that some insane person has questioned whatever
is being referred to above?]

OK, for me, the consensus view and "classical methods" are
those that virtually EVERYBODY accepts as demonstrating relationship -- and,
as I said, marginalising skeptics (but not necessarily calling them
"insane"; that is
superfluous).  What is wrong with calling that the consensus view?

[I thought consensus is the point of departure from which further progress
must be
made.  That the intent of the program is to try to push consensus further,
NOT to trivialise
what consensus already exists.  However,  the rhetoric of the reply
suggests that trivialising
 it  is indeed intended -- presumably, because those methods fail to
"establish" more distant relationships -- even though, ON THE BASIS OF
THOSE METHODS HAVE ACHIEVED,
we infer that there ARE more distant relationships]

As for the rest,  I don't know what "congruence" means in this context.
Does it mean
that the singular and plural class prefix pairs are formally unrelated, as
is typical
of NC.   Or does it also mean that  an etymological connection between
Kordofanian class
prefixes and those of the rest of NC is "accepted" (or "established"?)?

For the general picture, carefully REread (I assume)  Thilo Schadeberg's
article on
Kordofanian in "The Niger-Congo Languages", ed. John Bendor-Samuel (1989).
Note the importance (I think) of Schadeberg's comment p.72 with regard to the
class system:
"...similarities between the Niger-Congo and Kordofanian noun class systems are
not only typological but can be extended to proper sound-meaning correspondences
as well."

So he is well within what I called "classical methods"; not typology, but
sound-meaning correspondences.  [And those sound-meaning correspondences
are "accepted" as likely,
not "established";  we don't dismiss them from further use just because
they are accepted;
an attempt will continue to establish them, because the attempt itself
might lead to an
unexpected insight, and the responsibility to establish them is taken
seriously.]

Next, the Kadugli group (Greenberg's Kadugli-Krongo branch of Korodofanian)
ALSO has
languages with similar class prefixes and/or class concord.  However, Thilo
states, p.74:
"...I maintain that it has not been shown that Kadugli is part of
Kordofanian, nor that it
should be classified as Niger-Congo"

and he refers to his 1981 paper in a volume dedicated to Nilo-Saharan for
further discussion.
Therefore, whatever "congruent" noun class system means, it was
insufficient for him to
include the Kadugli group in Korodofanian, or even NC.  And yet no one has
accused Thilo
of insanity, and no "sane" person would.  For his latest thoughts, get in
touch with him at
the Linguistics Dept at the University of Leiden.  (He has much more data
than he has had
the opportunity to exhaustively analyse.)

BTW, noun class systems typologically similar to NC also occur in some
branches of Khoisan.
Yet no one connects Khoisan with NC (or anything else for that matter --
well,  I vaguely
recall some early volume on a library shelf that prematurely? connects the
Khoi branch with Indo-European).  Aren't we all waiting to find out how the
"clicks" arose from more widespread
consonant types?  [Again, I vaguely recall hearing about somethin in the
phonetic literature, how
they could have arisen, maybe in Ladefoged & Maddieson, but the Khoisan
click  systems have
not yet been shown to have non-click origins.  ANYBODY GOT ANY FURTHER
INFORMATION ON THIS?]

"I have discussed in print, in easily accessible journals, why
this is NOT the consensus view and why it IS arbitrary and invented.
Am I to assume that my work is not being read?"

In my case the assumption is correct.  The more you are interested in
motivating readers to actually read your detailed published discussions,
the more you will be specific about what you have in mind -- in LIST
discussion.

P.S.  Can I guess that the proposal has something to do with the suggestion
that some very
specific features of compared languages are unlikely to be due to chance
because of their
specificity and/or general rarity among languages, and the question arises;
how can such features
be enumerated, justified and measured with respect to significance beyond
sampling error?  The
typological and even etymological fact of a prefixal noun class system in
NC has not been deemed
sufficient in itself.  These are indeed examples of the kinds of things
that need to be discussed.