comment-Glottochronology.htm
COMMENT
from R. L. Trask, lecturer in linguistics at the University of Sussex
Subject: Re: Glottochronology
To: Patrick C. Ryan
"Glottochronology and its relative lexicostatistics do not "rule out" anything. The whole
point of these two techniques is that they can *only* be applied to languages whose genetic
relationship is already established, and between which cognates have already been identified.
Neither can possibly be used to establish either cognation or a genetic relationship not already
established; attempts to do this are nonsense and produce only nonsense.
To use a crude but serviceable analogy, consider what philologists do with multiple manuscript
copies of a single original text. By comparing these, they hope to establish which copies are
older and which younger, and to work out which manuscripts have been directly copied from
which other ones, and which ones must have been directly but independently copied from a
single earlier manuscript which no longer exists. This is a plausible exercise. But now suppose
some eager scholar were to try to do the same thing with manuscripts *which were not known to
be copies of the same original text*. The result would be gibberish, and the same is true of any
attempts to use lexicostatistics or glottochronology to establish historical relationships not
already known to exist.
What lex attempts to do is to provide a rough measure of the degree of linguistic distance
between languages already known to be related; what glotto tries to do, very dubiously, is to
convert the result into a time depth for separation. Neither is of the slightest use in establishing
historical relationships not already known to be correct."
May 22, 1996
=======================
"But perhaps I should add a clarification. I am here talking only about the traditional, and simple, version of lexicostatistics, in which words from two languages are compared only pairwise; for this case, I stand by my statements (though a few of my colleagues would take
issue with me).
However, if the term `lexicostatistics' is loosely extended, as it sometimes is, to more
sophisticated statistical approaches, then my statements are no longer necessarily valid."
June 12, 1997
=======================
Larry Trask
COGS
University of Sussex, Brighton, BN1 9QH, England
by the Author:
A Dictionary of Grammatical Terms in Linguistics (1993)
Language Change (1994)
Language: the Basics (1995)
A Dictionary of Phonetics and Phonology (1996)
The History of Basque (1997)
the latest revision of this document can be found at
HTTP://WWW.GEOCITIES.COM/Athens/Forum/2803/comment-Glottochronology.htm
Patrick C. Ryan * 9115 West 34th Street - Little Rock, AR 72204-4441 * (501)227-9947
PROTO-LANGUAGE@email.msn.com