Cladistic language concepts

[Scott DeLancey]

----------------------------Original message----------------------------
On Wed, 12 Aug 1998, Ghiselin, Michael wrote:
 
>                In spite of that the preliminary results are very
>           interesting.  Not only has a cladistic language concept been
>           generally presupposed but, as you say, linguists do not even
>           consider the topic particularly interesting.  Why should
>           this be?
 
Isn't it the case that biological taxonomy started simply as a way of
classifying biological phenomena in the world, and the idea that the
categories discovered reflect common descent is a separate (and later)
notion?  Comparative historical linguistics is quite different; our origin
myth traces it to William Jones observation that similarities among
several languages which we since recognize as Indo-European can be
explained only by recognizing them as descended from "some common source",
and discovering and demonstrating descent groups has always been seen as
what comparative linguistics is about.
 
>           they need.  Linguists do not have the elaborate system of
>           categories, such as phylum, class, order etc., that we
>           zoologists do.  And unless I am mistaken (please correct me
>           if I am) they do not believe that there are important
>           differences that need to be expressed by giving a taxon a
>           higher rank, as when our own species has been put in a
>           separate order or even kingdom.
 
No, I don't think that's true, at least in principle.  We certainly
recognize families, branches, subbranches (think of orders, families,
genera), although there's no standardized terminology for it.  I
think the reason why we don't have the articulated taxonomic framework
that biology has is that we don't have any very clear idea of how to
measure degrees of relationship--i.e. a statement like "the
degree of relationship of Klamath and Sahaptian is roughly equivalent
to that of English and German" is purely impressionistic, with no
way of operationalizing a measure of relatedness that would guarantee
that two linguists would agree on an answer.  (Nevertheless, unfortunately
in my view, one hears statemments of this sort all the time).
     Linguists might disagree about the need for higher-order categories--
extreme polygeneticists, I suppose, could argue that every currently-
recognized linguistic lineage represents descent from a distinct
invention of language.  I don't think anyone really would, though--
again, the reason why we don't have terminology for higher-order
groupings is that we don't really know how to recognize them.  (The
term "phylum" has been used a bit in linguistics, but without any
strict definition--often it simply means "a relatively deep postulated
genetic unit which is still controversial").
 
>                Linguists must have methodological problems with
>           respect to paraphyly, parallelism and convergence.  But so
>           far as I can tell, they treat these as problems to be
>           overcome in reaching a strictly genealogical arrangement.
 
Unfortunately this has historically been true.  As Larry Trask (I
think it was?) mentioned, this is changing a bit--we may be on the
verge of developing a framework for investigating contact phenomena
sufficiently rigorous and explanatory to have a legitimate place in
the same discipline as comparative linguistics.  We really do need
a science of areal linguistics, as we have for genetic linguistics,
but we don't yet have one.
 
Scott DeLancey
Department of Linguistics
University of Oregon
Eugene, OR 97403, USA
 
delancey at darkwing.uoregon.edu
http://www.uoregon.edu/~delancey/prohp.html