"centum"/"satem" "exceptions" [was Re: Northwest IE attributes]

[petegray]

>> Bird's summary of Pokorny's roots, ....2050 core roots.
>> Of these exactly 775 have neither e nor e: anywhere.

> 38%.  Is that including a(:) and o(:)?

I simply got the computer to check - so it includes absolutely everything
listed in Bird without e or e:.

> Benveniste treats roots with CeRC / CRC ablaut as CeRC roots.

Then it is no wonder that so few (comparatively) roots appear with CRC -
whether the R is /i/ or /u/ or anything else.

Is it possible that the apparent imbalance in numbers of CeC roots and
CiC/CuC is a chimaeara - merely the result of the way we record the root in
question?
And further, since CeC includes CeR, a comparison would not be about -e-
and -i-/-u- at all, but about the number of roots without  a third consonant
compared with the number of roots with third consonant, with medial R, but
without full grade.    Is anything meaningful being compared there?   It
does not seem to be a comparison of like with like.

Peter