Nostratic et al. Part 2: The Challenge

[Alexis Manaster-Ramer]

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

Dorothy has been kind enough to suggest I do
a quick and brief (!) part 2, so I'll just say some things
and refer y'all to the literature for more (on
general issues, please see my paper with Michalove and
Georg in the Annual Review of Anthro, just published).
And I will only use less than third of the over 20K bytes
of part 1.

(1) Re Nostratic, please read Nostratic:
Evidence and Status, ed. by Brian Joseph and Joe Salmons,
out of Benjamins as well as my review of I-S's
dictionary in Studies in Lg in 1993.  In addition
I can email the draft of Hitchcock's and my reply to
Ringe's (highly acclaimed) voodoo-mathematical "refutation"
of Nostratic to anybody who asks.

As for my own views (which several people asked me
about) on Nostratic, it struck me that in the recent
Dolgopolsky book, which I do NOT recommend except if
you run out of firewood, the best etyma for kinship
terms are for in-laws, which, assuming exogamy,
strongly hints of borrowing.  For this and many
other reasons published by me, a significant number of the supposed
Nostratic etyma are old borrowings.

However, some Nostratic
comparisons are much less likely to be borrowings,
incl. my own comparison of the IE words for
FIVE, FINGER, and FIST (all themselves cognate
within IE, as per Meillet and Saussure, but
many IEnists dissent for reasons I cannot
comprehend) with Altaic and Uralic words
with various bodypart meanings but no numeral
meaning.  I see here a real Nostratic etymon:
*payngo or something close to that.
Perhaps because I discovered it, this (and some other
things, of course, which there is no room
for here) makes me ready now
(as I never have before) to say that Nostratic
(meaning IE, Uralic, and Altaic at least)
is extremely probable and close to being a fact.

(2) Re Altaic (the controversial language
family comprising Turkic, Mongolic, Tungusic,
Korean, and Japanese-Ryukyuan), all I'll say
say is (1) it is an accident of the sociology
and psychology of a few crucial people that it is not
as accepted as Uralic or IE, (2) for details,
see the in-print paper by Georg, Michalove,
Sidwell, and me "Telling General Linguists about
Altaic" in Jo. of Linguistics, and the
bibliography therein. and (3) it is unconscionable, as we
spell out there, that some well-known
general linguists with NO expertise in the
field have been publishing disinformation
about the debates within the Altaic field and that
such people write book reviews
and encyclopedia articles on comparative Altaic.

(3) Re Pakawan, this is a small family of extinct
Native American lgs of Texas
and N Mexico which I posited and which includes
Goddard's Comecrudan.   Pakawan =
Comecrudan + Coahuilteco + Cotoname.
Pakawan is related to Karankawa, forming
what I now call Pakawa-Karankawan.  Sapir and
Swanton (two leading classifiers and describers
of N. American lgs a long time ago) claimed that these lgs
were also related to Tonkawa and at times they
included Atakapa.  I exclude Tonkawa,
and am unsure about Atakapa (Pam Munro of UCLA
has some goodish arguments for Atakapa belonging
elsewhere).

The MAIN reason why these lgs are good to talk about
is that the amount of data we have on them,
while adequate, is manageable, and can be
accessed by anyone with some basic skills in
comparative linguistics (these being very similar
in turn to the skills taught in generative
phonology) and a week or so of free
time. The data are found in just two
or three places, which is convenient. All unpublished
data known to exist are available from me by email,
and will some day  be published.

(4) Some other controversial linguistic groupings
which it would be nice, and not too difficult,
to work on are:

(a) Vovin's tentative linkage of Ainu to
Austroasiatic (which I am sure is right)
and Bengtson's proposal to link THAT to Nahali (which
appeals me),

(b) Diakonov's tentative(?) linkage of Sumerian
to Austroasiatic (which also appeals to me),

(c) Sapir's linkage of Haida to the rest of Nadene,
endorsed by Pinnow and Greenberg and recently ably
defended by someone with the same name as me (Anthropological
Linguitstics 1996),

(d) Eric Hamp's linkage of Hattic to IE (which I strongly doubt),

(e) Swadesh and Hamp's linkage of Chukchee-Kamchatkan with Eskimo-Aleut
(unless Michael Fortescue has already done it),

(f) the competing North-Caucasian and IE linkages
proposed for Etruscan,

(g) Bengtson's proposals re Basque, North Caucasian,
and Burushaski and Starostin's re North Caucasian
and Yeniseyan (I leave out Sino-Tibetan cause then
it is NOT manageable for mere mortals like me), and

(h) Austronesian + Austroasiatic = Austric, as argued
by various people every 20 years or so, Pater Wilhelm
Schmidt being the first and recently L.V. Hays being
of great importance.

That's my challenge to y'all.  Work one or more
of these instead of perpetrating or merely
repeating or even merely allowing others to
perpetrate or teach, unchallenged, the
pseudomethodological fabrications we
keep hearing instead of real work, and
perhaps as an excuse for not doing
real work.

I can't think of any other reasonably
straightforward problems just now, but
there are some I am forgetting. I do think
it is a scandal that virtually no
one is doing this yet, while so many are doing
so much to malign the whole field of language classification
and to do away with the teaching of comparative linguistics
in universities, instead.

(5) Of course, there are many other classification issues that need
work, but most are not easily accessible.
There ARE some simple mathematical issues that have not
been solved and anyone who knows elementary probability
theory and loves linguistics is invited.

(6) Finally, I say that comparative linguistics
of the OTHER kind, i.e., the one that deals not with
classification but with
the (pre)history of language families only
minimally controversial like Indo-European,
Dravidian, Uto-Aztecan, Kartvelian, Uralic, etc.,
is also full of manageable problems which remain
unsolved because (oh no, don't say it !) quite simply
most of the work being done in THIS field suffers from
precisely the methodological problems which are laid by
Trask, Thomason, et al. (usually incorrectly) at the door of
the kind of linguistics that does deal with classification.
The shoe is on the other foot, as I am prepared
to document in detail if asked (and
already have in various articles, notably in
International Jo. of Dravidian Linguistics, IJAL, Georgica,
JIES, etc.).

If we can have the dravidian etymological
dictionary in successive editions not
give a single reconstruction, state
a set of correspondence which account
for a fraction of the comparisons
that are then made, and freely mix
borrowed and inherited and accidentally
resemblent forms, there something wrong
ESPECIALLY if at the same time it
is widely procalimed that Dravidian
is a model language family, and the
Dravidianists models of modern comp.
linguists.  The most beuatiful example,
which I discuss in print, is how
the DED manages to get the word
for menstruation in one Dr. lg
derived from the etymon for house
simply because they look somewhat
alike, sound laws do notr count among
friends, and the word  happens to
occur in the phrase 'menstruation
hut' (but they forgot which is
hut and which menstruation, no
matter the rigid order of such
expressions in Dravidian).
Kartvelian is little better,
esp. in the Klimov version.  Much of
Native American work ditto (e.g.,
in the family I have worked on for
more than a decade, Uto-Aztecan).
IE is FAR FAR better, partly because
so much EASIER (thaks to Sanskrit
and other old data being available)
but even here we find a LOT of this
kind of stuff.

And yet we all know where the critics
of Nostratic, Amerind, etc., come from.

This then is theother challenge: let us
educate a new generation to clean u[p
the historical linguistics of well-established
language families even as we try to test
and refine theories that posit more far-flung
families.

Of course, systematically taking positions in
comparative linguistics (even IE) in major
universities and cannibalizing them for yet
more positions in generative or sociolinguistics
so that most departments that did just recently
no longer can pretend to teach comp. ling. AT ALL
is not the way to do it.  Nor is the policy of
certain journals to effectively ban comp lx
as no longer of "theoretical" interest.

Alexis Manaster Ramer
Professor of Computer Science
Wayne State University