From: Piotr Gasiorowski
Message: 27167
Date: 2003-11-14
----- Original Message -----
From: "alex" <alxmoeller@...>
To: <cybalist@yahoogroups.com>
Sent: Friday, November 14, 2003 6:31 AM
Subject: [tied] Albanian "dy"
> I would like to ask if Alb."dy" (two) changes its form when it is used
> for feminine (or in other cases) as Rom. does .
> masc "doi", fem "doua"
In those Albanian dialects that have preserved vowel-length distinctions,
the masculine is /dy/ and the feminine is /dy:/. The etymology of <dy> is
somewhat difficult. It can't go back to *dw�: (which would have given *d(w)o
> *de) or to *dw�i(h1) (> *d(w)ai > *de). The <y> can't be a reflex of old
*-u: (which would have been unrounded word-finally, ending up as *i). Hamp
believes that <y> is a special development of PIE dual *-o:u,
monophthongised to *u: approximately in Roman times (_after_ the unrounding
of inherited *-�:), and then fronted to <y>. I think <dy> more likely
reflects the "Lindemann form" *duwo: with initial stress: *d�wo: > *du: (the
contraction taking place after the delabialisation of inherited *u:) > dy
(as in Hamp's account). Feminine /dy:/ represents a later contraction of
*dy-� (with -� < *-a:s).
Piotr