[tied] Re: The role of analogy, alliteration and sandhi in counting

From: Abdullah Konushevci
Message: 34689
Date: 2004-10-16

--- In cybalist@yahoogroups.com, "Richard Wordingham"
<richard.wordingham@...> wrote:
>
> --- In cybalist@yahoogroups.com, Pe»us Hrubi¹ <hrubisp@...>
wrote:
>
> I don't think a direct change pWe- > c^e- is necessary or
> impossible. I would assume the usual route appplied, so pWe > kWe
>
> ke > c^e. However, given the Latin to Romanian development kwe >
> c^e, the Greek development tw > ss, the hypothesised Albanian
> development k^w > c'w (ultimately > s) and the widespread
> development pj > tS, pWe > c^e does not seem far-fetched to me.
>
> Richard.
************
I think that under the hypothesised Albanian development *k^w > c'w
(ultimately > s) you have in mind *k^we:rna: > sorrë 'crow, rook'
(Rom. cioara 'id.'). Same derivation I find also in Alb.
<sor•k•adhe> from *k^er-w-o-s, attested in Latin <cervus> `deer'
from suffixed metathetic lengthened e-grade *k^we:r-k-.
I guess that /*w/ in many different environments yields zero:
*dhwo:ra: > derë 'door', *k^wn.-ti > kandi > qen 'dog', due to i-
Umlaut and *n. >an; *g^hwon-o- > zani/zëri `sound' (cf. Sl.
zvono `bell', zvoniti `to sound'); *g^hwa:g(h)/*g^hwe:g(h) > Alb.
zog `bird', pers. zag, arm. jag (Pokorny g'ha:gw(h) 409). It seems
that the rule of reduction of gw > g and zw > z is present also in
some old Slavic loans, like Sl. gvozd > gozhdë `nail, spike', but
Alb. <gjeth> from much older form <gath> *gw-es/gw-os (Pokorny 480).
As we have seen, *kW > p is attested in Illyrian PN Ulpiana <
Ulkiana (cf. Ulkinion, Ulka, Alb. ulk > ujk), but in sandhi we may
have also *g^h > d (usually preced by prefix <në> `in, on') and *k^,
g^ > b (usually preceded by prefix <më> `id.').
Question is did labiovelars and palatalovelars very early
synchronized in Albanian and which evolution is oldest *kW > s, *gW
> z or *k^w > s or *g^w > z, followed at most by front vowels.


Konushevci