Re: [tied] trzymac'

On 2006-05-28 16:02, Sergejus Tarasovas wrote:

> From earlier *tur- (>*tUr-) + a predecessor of *jIma"ti (c), *tur-
> being a reflex of a root noun ('grasp, grip'), thus along Miguel's
> lines? The formation type would be that of *bolgodari"ti 'thank' etc.

English offers an even neater parallel: "get/grasp/catch hold of sth."
The compound would have to be old enough for the morphological seams to
have been obscured, despite its seemingly "Lekhitic" status. I would
expect something like *tvIr- rather than *tUr- as the Slavic reflex of
*//twr-//, but this should be no problem: we may well be dealing with
*t(v)r^ima- < *tvIrima-.

> The reflexes of *tur(H)- are well known in Slavic (*tvori"ti
> (b) 'make', *tvI``rdU (c) 'hard' -- with a strange violation of Hirt's
> Law etc.).

I wonder if there are any traces of non-iterative *tvIr-eNti
(*tvIr-Im-/o-) with matching semantics. Any such form would count as
clinching evidence in favour of this analysis.

Piotr