On 2008-01-30 23:34, fournet.arnaud wrote:
> I suppose -H- in the structure *gh_l-H-to is explained by tonal features
> in balto-slavic
Not just that; *g^Hélh3-to- corresponds exactly to Skt. hárita-, a
'yellow/green/fallow' colour term. Of course also the pan-Indo-Iranian
word for 'gold', Skt. híran.ya-, Av. zaraniia-, OPer. daraniya, Oss.
zærin-, is related and requires a laryngeal, as does the Skt.
compositional form <híri-> 'golden' (*g^Hl.h3i-)
> but
> which H3 particularly ?
You mean, why *h3? First and foremost, Gk. kHlo:rós < *g^Hl.h3-ró-.
Secondly, the otherwise very rare *-wo- suffix (as in Lat. helvus 'bay',
Lith. z^e~lvas 'greenish', Gmc. *Gelwa- 'yellow') is for some reason
frequent after the third laryngeal. There seems to be some poorly
understood allomorphy rule at work in those *-wo- derivatives, but they
strongly suggest a root-final *h3 nevertheless.
Piotr