Lexical Retention

[JoatSimeon at aol.com]

>Patrick C. Ryan

>In any case, you have it rather backwards. If the root and all manifestions
>of it (like, perhaps, in 'bear') were absent in Germanic, what would, as it
>has, make possible our reconstructing an IE root would be its presence in,
>at least, three (nominally) other branches.

-- quite true, if it _exists_  in at least three branches.  Some roots,
obviously, do.  Some, equally obviously, don't -- we don't have a complete
vocabulary of PIE, to put it mildly.  And in the nature of things we never
can.

Some parts of PIE have been completely and irrevocably lost because they
didn't survive into extant or recorded languages.

Obviously, this portion will be larger the further back we go in time and the
less we have in the way of writtten records of forms no longer extant -- a
reconstructed proto-Romance, even if we had no records of Latin, would be
much more complete than PIE.

The greater the temporal distance, the more information lost to entropy.