--- In
cybalist@yahoogroups.com, "alexandru_mg3" wrote:
>> Your said that :
>> 1. " Since /*s/ > /sh/ happened for sure at some historical
>> moment in Albanian and since all Latin loanwords in Albanian
>> exhibit this feature, the most likely assumption is of course
>> that the change took place afterwards."
>
> This is only one possibility. And it isn't 'the most likely' in
> this situation (viewing the moment of the slavic loans).
Of course it is the most likely.
> The other one is that /*s/ > /sh/ was ALREADY active on all the
> period of Latin loans, and that 'strunga','sterp','brusture' etc...
> are older than this moment.
It depends what do you mean by "active".
Anyway, let's denote by t1 the historical moment of the last Latin
loanwords in Albanian and by t2 the historical moment of the first
Slavic loanwords in Albanian. Since the feature F:= {/s/ > /sh/} is
exhibited by all Latin words, the most likely hypothesis -- by all
means the one which makes less assumptions -- has to be that F
occured in (proto-)Albanian for all words concerned at some tF such
as t1 < tF < t2, explaining thus everything. If one supposes that
tF < t1, then at t1 Albanian had no longer the phoneme /s/, so the
supplementary assumption that all Latin /s/ from loanwords mapped
with perfect regularity in Albanian /sh/ (just by chance similar to
what happened earlier at tF) has to be done. That makes two instead
of one, plus the coincidence. It is _highly_ unlikely.
Cheers,
Marius Iacomi