Re: [tied] *ekwos and friends

--- In cybalist@..., dusty_ceratos@... wrote:

> Sorry, Piotr - and what if the way from *ek^wo- to <kos'-> was
> completly different:
> 1) satemization succeeds and give *esv-
> 2) *-v- dissapears (it doesn't, as you say in (3). Maybe k'w was
> monofonemic and merges with k', like k and kw)

There was no *k^W phoneme in PIE, and if you propose one entirely ad
hoc, it only weakens the proposal instead of strengthening it.

> 3) initial e- > o- (maybe e- > a- > o- - in my native dialect I
> hear form kas'a, kas'achka - (little) horse, nursery word)
> 4) originates proteic g: os'- > gos'- (like (g)usenica)
> 5) gos'- > kos', maybe by analogy of kon'.
> Is this (or similar) development possible?

Well, this has become quite tortuous again. The truth is that <kos'>
and *ek^wos cannot be related in any straightforward and
methodologically sound way without a host of extra assumptions. The
real question is, do we really need so many conjuring tricks to
explain something that looks like a fairly recent expressive
formation? What's wrong with <kos'> being (for example) a truncated
pet variant of <konik>?

Piotr