On 2006-09-25 14:26, Piotr Gasiorowski wrote:
> Caland adjectives have compositional variants in *-i- -- that's the
> definition. They need not have -u-forms, but those that have them --
> an only those -- may also form -neu-factitives reconstructible all
> the way back to PIE.
By the way, I wonder if anybody has described the laryngeal-deleting
effect of the nasal infix -- I'm not aware of any studies of it. It
affects stem-internal laryngeals only (which would have become
interconsonantal after the insertion of *-n(e)-) and is reminiscent of
Saussure's effect in O-infixations. Thus, the verb derived from *tn.h2ú-
'thin, stretched' is not *tn.h2néu- but *tn.néu- (Skt. tanóti, mid.
tanuté, Hom.Gk. tánutai, from PIE *tn.néu-ti, *tn.nu-tór), and from
*wn.h1-ú- 'zealous' (Skt. vanú-) we get *wn.néu- (Skt. vanóti, vanuté).
See also another intriguing connection:
http://tech.groups.yahoo.com/group/cybalist/message/39236
It should be emphasised that the addition of a suffix to a *CRH- root in
the zero grade causes no laryngeal deletion, so the above consitutes
support for the infix analysis.
Piotr