Pre-Greek languages

[Stanley Friesen]

At 04:21 PM 10/19/99 -0400, Sean Crist wrote:

>On Fri, 15 Oct 1999, petegray wrote:

>When we have a family of related languages, on what grounds do we decide
>what the internal structure is for the family tree?  In other words, how
>do we decide which of these languages are more closely related than others
>in the family? Nearly everybody agrees that we should do so on the basis
>of shared characteristics of the languages which cannot reasonably be
>attributed to parallel innovation or to borrowing.  The differences of
>opinion arise partly from differing interpretations of the
>characteristics,

And in identifying such characteristics.

For instance, do shared new vocabulary items count?  How do you isolate
vocabulary shared due to ancestry from vocabulary shared due to borrowing?

For that matter, how does one rule out borrowing or influence for
morphological or phonetic features?

>If you believe that there is a grouping of "NW IE" languages (and I put
>quotes around it because I'm not clear on exactly which branches are to be
>included in this clade- Celtic? Germanic? Balto-Slavic?), this amounts to
>a claim that these branches are characterised by a set of shared
>attributes of the sort which I just mentioned.

There are a large number of shared vocabulary items between various pairs
Italic, Celtic and Germanic that are not found outside of Europe.  There
are also a large number of vocabulary items uniquely shared between
Germanic and Balto-Slavic.

This weekend at the library I discovered (or re-discovered) a booklet with
detailed stats on the distributions of roots from Pokorny and Mann.  I
intend to use the correlation tables at the end to apply one of the
grouping algorithms to get an estimated tree.

[I have some issue with the way they handled some of the minor IE
languages: for instance they group Phrygian with Dacian unconditionally,
possibly obscuring the relationships].

>(I earlier listed some of the evidence on whose basis Ringe et. al.
>produced the tree which I quoted; I can repeat both the tree and the
>evidence, if there's interest, but I hate to clutter the list with what's
>already been said.)

Or you could e-mail the evidence to me privately.  (Is this the one of the
U-Penn trees, or another one?)

--------------
May the peace of God be with you.         sarima at ix.netcom.com