Re: Latin -idus as from dH- too

On 2008-03-13 23:33, Miguel Carrasquer Vidal wrote:

> Since *-tor, *-ti and *-tu are all unstressed suffixes, and
> Olsen's rules *do* apply to *-tr-ó- (*-tl-ó-) [a thematized,
> and stressed, derivative of *-tor], maybe that could solve
> some of the issues. *-tó- is of course inherently stressed
> (and so is *-áh2-, I think, but not *-eh1-).

A brilliant idea. Your own? Of course one can expect a lot of analogical
interference if a phonetic process affects a very productive and highly
transparent formation. Examples like Skt. pr.tHá- 'the palm of the hand'
< *pl.h2-tó- (cf. Gk. paláme:, Hitt. palhi- 'broad') can only be
explained if one assumes *-h2tó- > -tHó- without the usual compensatory
effect of the laryngeal. Cf. also *tl.h2-tláh2 > *tl.tHláh2 > *t&tHláh2

> Pre-Lat. *taðla: > tabula. The less transparent, the better. But we

seem to have the analogical restoration of the laryngeal in Ved.
ti:rtHá- < *tl.h2-tó-, Lat. su:bula 'awl' < *sjuh1-tláh2, etc.

Piotr