Re: Internal reconstruction and PIE

What does Glen mean by 'Old IE'?

There is something of a lexical gap when speaking of 'PIE', in that we
have to distinguish between the post-Anatolic state of the
proto-language, and what it was in unity with Anatolic. With full
unity, we then travel back through a deep, but not limitless time of
development as an independent language family, until 'unity' with some
other language family can be hypothesized ('Indo-Uralic', perhaps).

Assigning times is a dangerous exercise. I for one have difficulty
believing anything beyond 7500 years ago is particularly recoverable.
My own inclination for the last date of post-Anatolic unity is rather
late -- 2800 BCE, even 2500 BCE.