The UPenn IE Tree (web site)

[X99Lynx at aol.com]

In a message dated 9/23/99 4:04:49 PM, BMScott at stratos.net wrote:

<<The algorithm does not root the tree; that was done afterward on the
basis of linguistic considerations.

...Do you mean historically, or in the construction of the tree?
Historically it is certainly not a point event, and in the tree it is
simply a consequence of what character values are shared by which
languages.>>

I'm having a little trouble following that and this might be helpful to
anyone else on the list who takes a look.

I've read the report "Mathematical Approaches to Comparative
Linguistics" (mathcompling.ps) especially looking for what was done
specifically to "root" the tree.  I found this: "Because rooted trees are
desirable, directionality constraints implied by some of the linguistic data
were encoded as characters, using techniques already in use by systematic
biologists, and these characters were included in the dataset."

It's obvious those "directionality constraints" can't rest too much on dates
of attestation alone, since some of important specific characters they
mention would have to have been dated from before those dates.

It also says here that <<The Indo-Hittite hypothesis is supported by only one
character, but it is difficult to impugn that character. Should that
character be impugned, a subgrouping of Hittite and Tocharian is possible,
but moving the root below the Italo-Celtic sub-group seems less likely than
the present rooting due to geographic constraints.>>

So I presume that geography as well as pure "linguistic" considerations were
also used in the rooting.  The following seems to suggest some substantial
geographic data was being used:

<<The second placement for Germanic which produces a reasonable result is
just outside the Satem Core. This placement avoids the need to posit an early
geographic move for Germanic, but does not provide a simple explanation for
all the incompatible characters.>>

Do you know if they address any specifics about the rooting procedure?   Do
you know if they address at all the fact that they are using *PIE to
establish the 'cognates' and reconstructed forms they are using as
'characters'?  It would seem because these aren't 'raw properties,' they
might be seen as "rooting" the tree from the start.

Part of my problem may be that I haven't been able to get Ringe's text -
"95-16.ps.Z" open.  Did you open it?  Distiller keeps telling me it is not a
ps file.

Appreciate any help you might have.

Regards,
Steve Long