Pennsylvania tree for IE

[Larry Trask]

On Fri, 16 Apr 1999, Sheila Watts wrote:

[on my reference to the Penn tree]

> Well-known, but not to me, alas: could we have a reference?

Sorry.  It's this:

Tandy Warnow, Donald Ringe and Ann Taylor. 1995. Reconstructing the
evolutionary history of natural languages. Institute for Research in
Cognitive Science, IRCS Report 95-16. Philadelphia: University of
Pennsylvania.

The tree itself is reproduced on page 369 of my textbook:

R. L. Trask. 1996. Historical Linguistics. London: Arnold.

The tree was constructed by selecting a sizeable number of "characters"
found in some but not all IE languages, and then asking a computer
program to construct a "best tree".  The authors report that most
branches of the family fall pretty naturally into a tree, but that
Germanic is rather perverse: morphologically, it clusters with
Balto-Slavic, but, in its lexicon, it falls between Italic and Celtic.
Their conclusion is that Germanic started off as an eastern branch of IE
but that its speakers then migrated westward, adopting large numbers of
western lexical items as a result.

Incidentally, I was in error the other day when I reported that the
published tree puts Germanic into the western group: it shows Germanic
in the eastern group, though the authors acknowledge the awkwardness of
this decision.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk