Sum: Sarich and historical linguistics

[Larry Trask]

----------------------------Original message----------------------------
Last week I posted a passage from an article by Vincent Sarich
attacking mainstream historical linguists and claiming a
surprising degree of success in finding cognates among living
IE languages.  I received responses from eight members of the
list, plus a brief one from Sarich himself, to whom somebody had
kindly forwarded my posting.

I asked particularly for comments on the following points.

>          *the use to which Sarich puts Buck's dictionary;

There weren't many comments on this point.

>          *the claim that any given living IE language retains about 60%
>          of the PIE lexicon in easily recognizable form;

This is the point I was particularly interested in.  But, as it
happens, most of my replies were from people who are specialists
in languages other than IE.  So I'm still not too sure about this.
Part of the problem is that Sarich does not make his procedure
at all clear, which makes it hard to evaluate his reported results.
My own attempts at comparing living IE languages from different
branches usually result in vastly smaller percentages even of
genuine cognates, let alone of obvious cognates.  But then I
require perfect semantic matching.  Sarich apparently doesn't,
but I await clarification.  I've asked him if he can clarify his
procedure, and he's promised to get back to me in a few days.

>          *the claim that genuine cognates among living IE languages are
>          overwhelmingly obvious and trivial to identify by inspection alone;

Two respondents, plus Sarich, queried my wording here.  If I've
misrepresented him, I apologize, but this was certainly my impression
after several readings.

>          *the claim that this result automatically generalizes to other
>          families, even to families which are as yet unrecognized.

To Sarich's reported results generally, but more specifically to this
point, responses were mixed.  They ranged from cautious sympathy
through marked skepticism to outright hostility, with a clear bias
toward the hostile end.  A couple of respondents specializing in
other families reported that obvious cognates, or even true cognates,
were generally much harder to find in their families than Sarich
reports for IE, and one respondent provided some wonderful data
demonstrating just how difficult it can be to recognize cognates
by inspection, and to distinguish these from non-cognates, even in
a moderately close-knit family.

So, I'm still not too sure what to think of Sarich's results, but
perhaps he'll be able to clarify matters on another occasion.

One final point, not relevant to my query.  Professor Sarich has
pointed out that, even though his work has commonly been taken
as providing one of the cornerstones of the out-of-Africa hypothesis,
he himself does not endorse that hypothesis.

But this takes me to another point, this time relevant.  If I
understand Sarich's position correctly -- and I hope I do -- he
wants to see all humans now alive as descended from a rather small
and homogeneous population which lived much more recently than
100,000 years ago.  Accordingly, he is happy to be told that there
may be evidence for deriving all known languages from a single
language spoken in the not-too-distant past -- or at least to be
told that comparative linguistics can reduce all known languages
to just a few large families.  If so, I find this worrying, since
I don't think linguistic results should be held hostage to any
non-linguistic beliefs about human prehistory.

My thanks to Ross Clark, Alice Harris, Konst Krasukhin, Steve Long, Marc
Picard, Robert Ratcliffe, Vincent Sarich, Theo Vennemann, and Benji Wald.


Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk

Tel: 01273-678693 (from UK); +44-1273-678693 (from abroad)
Fax: 01273-671320 (from UK); +44-1273-671320 (from abroad)