Newsgroups: sci.lang
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!newsfeed.pitt.edu!godot.cc.duq.edu!newsgate.duke.edu!news.mathworks.com!newsfeed.internetmci.com!howland.reston.ans.net!ixnews1.ix.netcom.com!netcom.com!netcom7!alderson
From: alderson@netcom7.netcom.com (Richard M. Alderson III)
Subject: Re: Nostratic
In-Reply-To: Harry Bowman's message of Mon, 17 Jun 1996 15:28:30 -0400
Message-ID: <ALDERSON.96Jun19115203@netcom7.netcom.com>
Sender: alderson@netcom7.netcom.com
Reply-To: alderson@netcom.com
Organization: NETCOM On-line services
References: <31C5B1DE.41C6@cornell.edu>
Date: Wed, 19 Jun 1996 18:52:02 GMT
Lines: 106

In article <31C5B1DE.41C6@cornell.edu> Harry Bowman <hlb6@cornell.edu> writes:

>Well, I see that the expected invective has arrived.

Invective?  Hardly.  Criticism?  Assuredly.

>In fact, the first language family to be recognized was the fairly small
>Arabo-Canaanite subgroup of the Semitic family of Afro-Asiatic.  Medieval
>scholars noted the similarity of Hebrew, Aramaic, and Arabic.

Medieval scholars also noted the similarity of the Romance languages.  Unless
you mean to say that a study setting out similarities and differences among
Hebrew, Aramaic, and Arabic, and claiming that all were thereby shown to be
related, I will continue to claim primacy for the monograph on Hungarian and
Finnish.

>Some inaccuracy may have occurred in my glottochronology figures.  Being an
>astronomer, I am used to imprecise numbers.  I assumed that the standard 18%
>loss per millenium was an estimate, which I rounded to 20%.  I then calculated
>0.8^7 and 0.8^10 in my head and gave those figures.  (Try using 18% and look
>at the difference.  It is not insignificant).

I was not criticising the imprecision of your numbers.  My criticism is for
glottochronology as a technique.  I don't care whether you take the 81%
retention rate claimed for the '200-word list' or the 86% retention rate for
the '100-word list':  Both rates are intellectually dishonest, hand waving of
the worst sort.  They ignore the 100% retention rate between Old Icelandic and
Modern Icelandic, the retention rate of approximately 40% for Greenlandic, and
the wide range of retention rates demonstrated for languages all over the world
by historical linguists who really *would* like a tool that worked.  The
published rates urged by the proponents of glottochronology (really lexicostat-
istics in this context) are based on carefully chosen languages in order to fit
the preconceived notion that this *should* work.

>Remember guys, LINGUISTICS IS JUST A MODEL.  The real world may be different,
>but it doesn't mean that the model isn't useful within certain contexts.

I don't understand what you mean by this.  Linguistics is not a model (as I
understand the term from cosmological astronomy or high-energy physics), it is
a collection of disciplines all of which concern themselves with human language
in one or more aspects.  Most of these disciplines do involve models as I
understand the term:  the acoustic model of the human vocal tract, Chomsky's
"Language Acquisition Device", Stampe's process model of phonology, the "wave
theory" of spread of innovations from a central point to outlying areas.  But
no single one of these is "linguistics".

So please, I invite you to expand upon your capitalized statement above.  What
do you mean by "linguistics"?  What do you mean by "model"?

>The most amusing piece of the reply I have recieved is a statement that it is
>NOT true that some words are more stable than others.  This seemed to be
>stated rather emphatically.

At least you seem to have gotten the emphasis; it is sas that you were amused
by it, rather than being urged to think about the statement you had made.

>I don't know what planet that came from, but here on Earth, it is obvious that
>some terms are highly unstable...

As you yourself noted, what is obvious is not necessarily correct.

The problem for this hypothesis is that does not bear up under investigation.
(You know, the scientific method, and all that.)  Rather than gathering data
from hundreds or thousands of languages, and looking for items that appear to
be resistant to replacement, an _a priori_ list of fifteen items is postulated,
and then those fifteen items are looked for.

Fine.  An acceptable experimental technique, if adequate controls are used.

But they aren't used.  Roget would have allowed fewer synonyms, metaphors,
related meanings, and the like.  If one spreads the semantic net wide enough,
*anything* can be proven, and thereby *nothing*.

>I may also add that I think that the suggestions about large scale language
>groupings need to be considered carefully.  We cannot expect these families to
>be demonstrated with the sort of rigor found in Indo-European.  After all,
>people have been working for at least 200 years on the subject.

Oh, yes, we can.  Look at Bloomfield's work on Algonkian, or Haas' on Muskogean
(and on that family's connections with Algonkian).  Rigour does not equate to
length of time under investigation--Indo-European could be considered to be
fully proven after Bopp, in 1816 (just 30 years after Jones), and he did not
have the panoply of tools available to modern linguists.

>Also, one simply wouldn't expect the reconstruction to be as good on a 12,000
>year old grouping than one which is 7000 years old.  And I certainly think any
>talk of "Proto-World" is at least premature.  A lot can happen in 100,000
>years.  But I try to keep an open mind on these things.

It may come as a surprise to you, but I agree that the reconstruction of, for
example, Nostratic may not be as "good" as that of Proto-Indo-European.
However, the linguists who are working on it are trying to make it so.  As I
noted in my followup to your previous post, I think a limit of just enough in
the past to allow for PIE but nothing else is just too convenient.

You see, you have mistaken me.  I think some form of the Nostratic hypothesis
is correct, and I know several of the researchers in the field personally, so I
know something of the level of effort involved in the reconstruction.

I admire Greenberg's results in Africa--but there, he had the advantage of a
command of the data, which he does not have in the Americas, or in Asia.
-- 
Rich Alderson   You know the sort of thing that you can find in any dictionary
                of a strange language, and which so excites the amateur philo-
                logists, itching to derive one tongue from another that they
                know better: a word that is nearly the same in form and meaning
                as the corresponding word in English, or Latin, or Hebrew, or
                what not.
                                                --J. R. R. Tolkien,
alderson@netcom.com                               _The Notion Club Papers_
