On 2006-09-23 11:00, tgpedersen wrote:
> How do you know those adjectives are not part of a Caland system
> and thus not particularly loyal to their *-u?
"Being part of a Caland system" means, essentially, that if an adjective
in *-ró- occurs in a compound as its the first member, *-i- is
substituted for *-ró-, as in YAv. xruui-drau- (*kruh2-i-) 'having a
bloody club' vs. xru:ra- 'bloody' (*kruh2-ró-) or Gk. argi-kéraunos
'with swift/vivid lightning' vs. argós < *h2r.g^-(r)ó- (Skt. r.jrá-)
'swift, shining'. There are good reasons to believe that this
"substitution" has a purely phonetic explanation. The alternation of
*-ró- and *-ú- if free-standing adjectives seems to be due to the
preferential survival -- dependent on phonetic factors -- of doublet
forms of ultimately the same origin (sometimes both survive, in cases
like *h2r.g^-ró- ~ *h2r.g^-ú-, *bHr.g^H-ró- ~ *bHr.g^H-ú-, each pair
sharing one and the same Caland alternant). These are principled
alternations. What they certainly don't mean is that you are free to go
ahead and ignore the suffix if you so please. Derivatives with a nasal
infix were only formed from athematic bases, which in the case of
adjectives means the *-u- variant.
Apart from *-neu- factitives there's also good comparative evidence for
other types of *CR-ne-C- presents of *CeRC- roots. For example, *demh2-
'conquer, tame' had a nasal present *dm.néh2-ti whose direct reflexes
are finely preserved in Greek and Celtic.
Piotr