Darwin-L Message Log 5:71 (January 1994)

Academic Discussion on the History and Theory of the Historical Sciences

This is one message from the Archives of Darwin-L (1993–1997), a professional discussion group on the history and theory of the historical sciences.

Note: Additional publications on evolution and the historical sciences by the Darwin-L list owner are available on SSRN.

---

<5:71>From delancey@darkwing.uoregon.edu  Thu Jan 13 16:32:15 1994

Date: Thu, 13 Jan 1994 14:16:05 -0800 (PST)
From: Scott C DeLancey <delancey@darkwing.uoregon.edu>
Subject: Re: Systematics and linguistics
To: darwin-l@ukanaix.cc.ukans.edu

Sally Thomason suggests (Tue, 11 Jan 1994) that, given what we know
about interlinguistic influence (the various phenomena traditionally
rather imprecisely referred to as "borrowing"), we wouldn't expect
it to confuse the issue of genetic relationships among languages:

> rare.  That is: slight to moderate linguistic borrowing -- not
> just words, but also sounds, syntax, and even some word structure
> -- doesn't obliterate the main lines of descent of a language;
> and when borrowing becomes so extreme that the main lines of
> descent are seriously obscured, there are usually clues in the
> structure of the language.

In theory, this is true; as Sally (who, after all, wrote the book
on the subject) puts it:

> As in biology, linguistic evolution is, as far
> as I can tell, mostly non-reticulate...as long as you're
> dealing with completely separate languages and not dialects of
> the same language, and as long as you are looking at languages
> as wholes rather than at individual
> linguistic features taken separately.

In practice, particularly in the Greenbergian context
with which this thread started, I don't think it is, i.e. I disagree
that:

>    So I don't think reticulation, to use the biological terminology
> Kent Holsinger was using, is too likely to be a stumbling block --
> at least not often -- in the effort to establish relationships
> among languages.

Empirically, it has freaquently been, and continues to be, a stumbling
block.  For some celebrated cases, e.g. the relationships or lack of
them among Chinese and various Southeast Asian languages (Vietnamese
and languages of the Hmong-Mien and Tai-Kadai families), further work
has largely succeeded in unravelling the problem (though respectable
and otherwise sensible scholars still try and reopen the issue every
now and then).  Others, e.g. the problem of the parentage of Japanese,
or of the relationship between Indo-European and Afro-Asiatic, remain open
and controversial.  The problem which recurs again and again is that
we find two languages with some substantial amount of what appears
to be common vocabulary, that, if they are indeed related, are related
so distantly that more conclusive evidence of relationship has been
obscured over time.  When, as (at least arguably) in the case of
Japanese, there is evidence of this sort linking the language to
two distinct genetic stocks, there is indeed a stumbling block in
the effort to establish relationship.  I don't think that, at least
in the case of animals, there could be any biological parallel to
this situation.
     And precisely this argument is prominent in discussions of
Greenberg's work.  Much of his evidence can be discarded on various
grounds--bad data, erroneously transcribed data, misanalyzed forms,
etc.  And undoubtedly a considerable proportion of what's left
represents chance resemblance.  Greenberg's answer is that even
so, he has enough data that there will still be enough left to
prove his claims.  The problem is that it is still not possible to
eliminate borrowing as an explanation for at least some of these
data--so the possibility of borrowing again represents a stumbling
block.

Scott DeLancey                          delancey@darkwing.uoregon.edu
Department of Linguistics
University of Oregon
Eugene, OR 97403

---

Your Amazon purchases help support this website. Thank you!