'grey' in PIE

A detailed refutation of Lubotsky (1989) etymology of *k^as-o- is
perhaps in order. He reconstructs the PIE colour stem as *k^h1ei-,
assuming it to be an extension of *k^eh1- 'grey', for which the only
evidence given in the article is Skt. *s'a:rá- 'motley, spotted' and Gk.
ke^rulos, a fabulous sea-bird (no precise description available). The
evidence is dubious, to say the least, and so it's hard to accept the
reality of *k^eh1- in the first place. As for the extended *k^h1ei-, he
needs the *h1 next to the initial stop to account for Slavic *s^- < *x-
(BTW, the theory that *h1 causes the aspiration of a preceding stop
anywhere in IE is Lubotsky's personal and controversial opinion, not
necessarily shared by other IEists). However, the Slavic adjectives
*s^e^rU and *s^e^dU can both be loans from early Germanic (from,
respectively, *xaira- 'grey' and *xaida- 'bright, shining'). Note that
the latter (and possibly the former as well) is a likely cognate of Skt.
ketú- 'brightness' (no aspiration!) and thus has nothing to do with the
'brown, grey, dark' root *k^jeh1- (sic!) as visible in Skt. s'yá:va-
and, among others, Slavic *sivU and *sinU < *k^ih1-wo-, *k^ih1-no-. Even
assuming a secondary (anti-samprasarana) full grade *k^eih1- to account
for Celtic *keiro- 'dark brown' we don't get anything that could be used
to support Lubotsky's 'hare' etymology. Laryngeal metathesis doesn't
happen right-to-left, and there's no chance that *k^jeh1- could be an
extension of **k^eh1- rather than **k^ei-. I rest my case, m'lud.

Piotr