>Its derivation from Latin <bestia>, even semantically seems more
>plausible, phonetically is very hard to be explained. Indeed, e > i
>is regular before a consonantal cluster, but reduction of the
>cluster st > sh is very unexpected in Albanian (cf. Latin crista >
>Alb. kreshtë), for its distribution in Albanian is very high,
I agree that the output of PAlb st is Alb sht.
However in Lat. bestia we have tj-a : a hiatus that gave ts /c/ in
Roman Times that next passed to s in Albanian :
Ex: Lat puteus <-> PRom. putsu ; PAlb putsu > Rom. puts ; Alb pus
So Lat. tj-V gave PAlb ts-V > Alb. s-V an din this case we have:
Lat. bestia > [ti-a > tsa] bestsa > [s>sh ts>s] beshsë > [e>j shs
>sh] bishë
Only The Bests,
Marius
P.S.:
Albanian Timeframes:
---------------------
What I want to point out is that inherited PAlb *ty don't passed to s
via c^ > c > s (as Piotr, and not only him, suggested) but it passed
to s only via :
PAlb *ty > PAlb c > Alb s
The timeframe of this transformation should be located in Roman Times
(see Latin ti-V > PAlb ts > Alb s that is also affected). At that
moment of time PAlb c^ was already c.
So PAlb c^ (PIE k'w,kw/+) was passed already to c /ts/ before Roman
Times and also PAlb c /ts/ (PIE k') passed to PAlb *th also before
Roman Times : with these timeframes there is no need to invent a
transition at each 50 years in Roman Times in order to obtain Alb s
both from c^ as from ty and to avoid also the superposition with
original PAlb c (PIE *k').
But the main argument that sustains these timeframes is not the
previous one (that is a valid one too), the main argument is the
treatments of Latin ci,ce,gi,ge in Albanian that gave q, gj
indicating No Traces of c^, g^ in Albanian during Roman Times.
To Resume the Albanian timeframes:
Before Roman Times (but not long before)
-----------------------------------------
PIE k' > PAlb ts
PIE g' > PAlb dz
PIE k'w, kw/+ > PAlb c^
PIE g'w, gw/+ > PAlb g^
PIE ty > PAlb ty
PIE dy > PAlb dy
Note also that in this period:
PAlb tsj > PAlb c^
PAlb dzj > PAlb g^
During Roman Period :
---------------------------
PIE k' > PAlb ts > PAlb th
PIE g' > PAlb dz > PAlb dh
PIE k'w,kw/+ > PAlb c^ > PAlb ts
PIE g'w,gw/+ > PAlb g^ > PAlb dz
PIE ty > PAlb ts
PIE dy > PAlb dz
PIE k/+ > PAlb q
PIE g/+ > PAlb gj
IMPORTANT NOTE: We don't have any PAlb c^ and g^ during Roman Period
otherwise sooner or later Latin ci,gi,ce,ge would have done in PAlb
c^,g^ (as in PRomanian) > Alb s,z that is not at all the case (the
output is Alb. q, gj).
So there is no PAlb c^, g^ in Roman times as Piotr suggested in PAlb
Albanian. 1,2,3
Before First Slavic Loans the Albanian Phonetism was:
-----------------------------------------------------
PIE k' > PAlb th
PIE g' > PAlb dh
PIE k'w, kw/+ > PAlb c^ > PAlb ts > s
PIE g'w, gw/+ > PAlb g^ > PAlb dz > z
PIE ty > PAlb ts > s
PIE dy > PAlb dz > z
PIE k/+ > PAlb q
PIE g/+ > PAlb gj
NOTE: the timeframes indicated above explains all the Latin loans in
Albanian and also all the Romanian-Albanian Common Words that
contains the clusters above.
Note: Also there is no Sl. c^ or c that gave Alb s as Piotr indicated
in Albanian 1,2,3.
The examples that were given as arguments are related to some Slavic
clusters having:
A) c^k that was normal to gave sk in Albanian at that moment in
time:
Sl. bic^&k& > Alb. bisk 'branch'
Sl. bric^&k& > Alb. brisk 'razor' Rom. brishcã
(please note: ORom c^k > Rom shk)
or B) nc^ (or a similar context o^c^)
Sl. poro^c^iti > Alb. porositi Rom. porunc^i 'to command'
These clusters gave Normally *s in Albanian based on the phonetic
status in Albanian at that moment of time (see above rules) but this
Doesn't Means At All that they represents Sl. c^ > Alb. s.
For a similar example: as you can see above ORom. c^k gave Rom.
shk in Romanian but this doesn't mean At All that Rom c^ passed to sh.
Note also that later Slavic loans in Albanian shows Alb c for
Slavic c^ (this clearly indicates that Alb /ts/ reappers at that
moment of time in Alb. but Alb. c^ still wasn't there again).
In Conclusion:
We don't have any Slavic example showing a simple c^ or c that
gives s in Albanian, as Piotr suggested.
Only the Bests,
Marius