Re: [tied] Re: Lambdacism

What you're missing is the conditioning environment. In Ossetic-type lambdacism, only palatalised /r/ (followed by /i/ or /j/) changes into /l/. More precisely, *ri > li and *rj > l . *arba- or whatever underlies "Arpo-" would not have been affected, as opposed to *ripa-. Ossetic has ruxs 'light' < *rauxs-, fyr 'much' < *paru-, ruvas 'fox' < *raupasa- (by the way, in these examples PIr *r < PIE *l), zärond 'old' < *zarant- (cf. Gk. geront-), etc.

Piotr

----- Original Message -----
From: Alexander Stolbov

To: cybalist@yahoogroups.com

Sent: Monday, January 07, 2002 10:23 PM

Subject: [tied] Re: The Scythian Brothers

Is it possible to have simultaneously L(e)ipoxais < *ripa-xs^ayah and Arpoxais ?

If we have lambdacism here we should have Lipoxais and *Alpoxais,

if we don't have lambdacism then we should have *Ripoxais and Arpoxais. Right?

Or I don't understand something here?