Hello Piotr,
I have to come back on your 'Bessa' explanations, after some days
(meanhile I read additional stuff)...
My conclusion : I didn't find SUFFICIENT your explanations
regarding "besa" etymology:
My Arguments Related to your basic explanation:
-------------------------------------------------------------------
I. "it can be derived without the slightest difficulty from
*bHéndH-tah2 (*-twah2 would also work" are:
-------------------------------------------------------------------
Argument 1: if it can be derived "without the slightest
difficulty" why Demiraj didn't derived Albanian
'besa' "without the slightest difficulty" from the PIE *bhendh ?
Here is Demiraj derivation for besa:
Albanian form: bese' [f] (tg)
Meaning: faithfulness, faith
Proto-Albanian: biTs/ia: {2}
Other Alb. forms: besoj [verb] (tg) {3} `to believe'
Page in Demiraj AE: 96
IE reconstruction: bh(e)idhi- {1}
Meaning of the IE root: persuasion, oath
Greek: `persuasion'
Notes: {1} Based on the genitive. {2} Phonetically, a full
grade form PAlb. *beiTs/ia: is also possible. {3} Denominative
formation in -o- of Alb. origin.
As you can see the derivation is from "bh(e)idhi-" (persuasion,
oath) and not from the PIE "bhendh".
If it is so obvious as you said, why Demiraj "couldn't find" the
PIE "bhendh" in this case?
Argument 2: What I found also strange in your derivation is the
fact that the meaning of the word "besa"
is COMPLETELY IGNORED : PIE *bhend -> alb. "besa".
There isn't an 'obvious word' related to "faith" in any other IE
languages that have derived words from
the PIE *bhendh.
As you can see Demiraj TRIED TO TAKE INTO ACCOUNT THE MEANING OF
THE WORD TOO:
He proposed 'bh(e)idhi' --> "Meaning of the IE root:
persuasion, oath"
My question is : did you ignored the meaning of the word in
your derivation? Seems yes...
Argument 3: The derivation of pese: : indeed has no difficulty
as result Demiraj 'indicated very easy'
IE : *penkwe (as you did too).
But there is a problem with "bese:" and from there
some 'obvious' workarounds :
a) your workaround : "tah2 (*-twah2 would also work; both
suffixes form nomina actionis)".
b) Demiraj workaround : '{1} Based on the genitive' etc...
A "real" IE root for 'besa' have to contains a "kw" or
something related in the stem...as in "pese:"
My question is : where is the PIE root that has a "kw" or
something related for bese:?
Argument 4: What happens with original "dh" from your
PIE "bhendh" until you arrive to 'besa' ?
It dissapears "by contraction"? It becomes albanian "t" or "ts"
(the (s) after t is there or not)? Or you completely ignored it
because you need ONLY the 't' fron 'thah2'?
In any case, here is a well known example where IE "dh" simply
became alb. "d"
IE : *dhwer-, *dhwor- ----> Albanian derë
Meanings: a doorway, a door, a gate ----> Meaning : a
gate.
So the main transformation is : IE.dh -> Alb.d .
Could you give us similar transformation of IE "dh" -> alb. "t"
or "ts" ? Could we have 2 transformation quite different where 'dh' -
> t,ts and dh-> d? Or 'dh' in 'bhendh' dissapears and the 'magic
word' 'tha2' appears again to explain the inexplicable...
Argument 5: Why a 'nomina actionis' form for this derivation?
Why not a verbal form?
Only to 'can arrive' to derive 'besa'?
(at least both have the same probability (we have also the
verb 'besoj'...).
Viewing this my conclusion is the following :
a) the PIE root for 'besa' cannot be easy explained by IE
root 'bhendh'
b) even more, there is not easy to derive albanian "besa" from a
known IE PIE
(see Piotr's 'tha2' or Demiraj 'genitive form').
As result : <alb. besa is an ancient loan (why not from 'bessi'
attested around 500BC by Herodot (485 - 425 BC), quite old isn't it)
and because it is an ancient loan could be very well viewed
as 'inherited' compared with quite recently 'slavic loans' in
albanian.
II. Now to come back more closer to our days : from the "IE times"
to "proto-albanian" or why not "pre-albanian" times, and take a look
on the problem of Albanian "c^" :
Here are the "loans" from Romanian to Albanian regarding the alb.
and rom. fricatives :
(better to say "supposed to be from Albanian to Romanian" ->you
need to make a demonstration here too
regarding the direction of these loans (see below)) ) of the
alb. and rom. fricatives :
#-----------------------------------------------------------
Alb. , Rom. ,Alb. <-> Rom.Words, Alb. sounds
#-----------------------------------------------------------
ç , ci ,çok <-> cioc , alb. ç as ch in church
# , ,çukë <-> ciucă ,
# , ,çufkë <-> ciuf ,
# , ,çupis <-> ciupi ,
#-----------------------------------------------------------
c , ţ ,cap <-> ţap , alb. c as ts in cats
# , ,cjap <-> ţap ,
#-----------------------------------------------------------
q , ce ,qafë <-> ceafă ,alb. q as ch in chair
#-----------------------------------------------------------
s , ci ,sorrë <-> cioară ,alb. s as s in see
, ,kësulë<-> căciulă ,
#---------------------------------------------------------
th , ţ ,thep <-> ţeapă ,alb. th as th in three =
# ,-------,---------------------------,-------------
th , ci ,thump -> ciump ,alb. th as th in three
# ,-------,---------,---------,-----------------------
th.........
if you go on with your assumption "loans from Albanian to
Romanian" as you can see the situation of these loans is quite
complex and your afirmation "ancient albanian ci kept by romanian"
became "today albanian s" ...is at least not complete:
It generates the following doubts.
a) If all these words are "loans from Albanian to Romanian" the
albanian "ancient" "ci" (c^) ("borrowed and kept by romanian"
as "ci") could have as correspondants sounds in today Albanian
not ONLY an s but also a th or a ç :
1. ci <-> th (thump <-> ciump)
2. ci <-> s (kësulë <-> căciulă)
3. ci <-> ç (çok <-> cioc)
As you can see from this examples an 'ancient' albanian 'ci'
could very well remained today "ci" as from "cioc->çok" or could
evolved to a "th" as in 'ciump->thump'...or (as you supposed) evolved
to an s as in "căciulă->kësulë" or in "cioară->sorrë".
The "Babylonie" of such conclusion results in my opinion due to the
first assumption :
'all these are loans from Romanian to Albanian'==> that is either
completely false or in any case
if it is not false you should clearly indicate the TIMEFRAMES for
each of these changes...if you want to give
a serious explanation on 'ci' <-> 's' transformation.
b) As regarding the slavic earlier loans (if they are really
slavic loans, I have doubts here too) in albanian the s remain s as
in :
1. <rom. sitã , <eng. sieve, grill , <sl. sito , <alb. sitë
2. <rom. coasã , <eng. scythe , <sl. kosa , <alb. kosë
so the albanian transformation (so s->sh seems finished at that
moment, by the way you have to indicate the
TIMEFRAME(->start,->end) of Albanian "s->sh" transformation too)
My doubts are :
1. You have to indicate why the Albanian-Romanian common words
you considered Romanian loans from Albanian.
2. You have to indicate with arguments why and especially WHEN
the "ci<->s" appears in Albanian
( related also to "ci<->th" and "ci<->ç" timeframes)
Best Regards,
marius