Alex wrote:
>accepting *a-ja: one gets in trouble with the masculine since there
>is an "l" ,see "ãla". as for Rom. "asta", Alb. "ato" it appears
hard >to believe there is no connection with Latin "iste" ,
Umbric "esto" >etc., where "*to"= "der da, he there" and "*es/*is",
thus I don't >think there is any "*a-k'jo-*tom (Acc. of *to)" but an
*es-tom/*es-ta
You are right Latin "iste" (Umbric "esto") are very well the
source (or at least part of) the source of these dem. pronoun shapes
in Romanian.
But we are talking here about a category very often ignored by the
linguists: We have very close forms between Proto-Albanian and Latin
for some words (of course due to the Indo-European common heritage).
I name this Latin-PAlb-CommonForms and to give you some other
examples we have:
Latin duo - PAlb *dwuai - Alb. dy - Rom doi (where Rom ua>o is
regular as in una >o ) -> nobody can told me here that the Albanoid
Subtratum that have been Romanized didn't influenced this word at
all (when the PIE was *duoi-jo). On the other hand to consider that
the Latin had no influence here is a big mistake too.
So this category Latin-PAlb-CommonForms should be introduced as a
concept that should include this type of cases.
( NOTE: Regarding the Romanian doi 'two' I think that is
unaccetable to explain like Rosetti did: that i in doi represents
the mark of the plural (Rosetti ILR I): ok, plural! but from where?
For sure not from duo but in this case from where? (a similar
explanation is given by Rosetti for the i in Rom. trei 'three')).
I. To come back now to our topic: and taking again the example of
Romanian form asta 'that one. fem': is impossible to obtain an
accented *'a in Romanian from a Latin i. Based on this, Densusianu
explained very well that the 'a' of 'asta' is not from 'i'
of 'ista'.
Next to can derived asta from ista: Densusianu suppose that *a- here
is from the previous word in the phrase like: luna esta -> luna-a-
esta -> luna aiesta -> luna aiasta (and by contraction) -> luna
asta. His explanation is a very clever one...
But the doubt still remain: the doubt is related to this *a that is
taken from the previous word in the phrase: because this assuption
remains only a supposition: that can never be proved.
We could also ask here: why a? Or better: why only a? Or better: if
true, this happened maybe under a known influence -> that could be
found very well in the Alb. form atë 'this one'
Because based on 'atë' we can easy obtained an accented *'a that
could be linked with the Romanian form : asta.
So the PRom forms could be the Latin forms but these forms were
reshaped/influenced based on the known local forms of Proto-Albanian.
Trying to make an organic overview of the Romanian Forms we saw in
addition that Rom. aia 'that one fem.' cannot be from Lat. illa that
could give only (and really have given) Rom. ea 'she'. Once again
a 'mysterious' a- appeared in front 'a-ia' and once again we have
the Alb. counter-part a-jo 'she but also this one fem.'. Based on
the 2 forms Rom. aja and Alb. ajo : we can logically consider a PAlb
form a-ja:...
(I didn't trust at all here Abdullah derivation -s- > -j-. He
followed Pedersen with the difference that Pedersen cosnidered -s- >
zero and treated -j- as a hiatus filler. Also seems that Abdullah
didn't want at all to integrate in his explanation the Romanian
form: aia)
So if we want to have an organic view and to include all these words
in a same schema -> We need to say that the Balkan Albanoid
Substratum have had a proclitic *a- used to construct the forms of
these demonstrative pronouns :
See Romanian : a-ia, ã-la, a-cesta, a-sta, a-ceeaSi, a-cea
and also Albanian a-të, a-ia, a-y, a-so, a-saj etc...
This PAlb proclitic *a- has as cognate the proclitic a- in Skt. See
Skt. a-sau 'that'.
The fact that this *a- wasn't lost at the beginning of the Albanian
words (normal rule in Alb.) indicates that this *a- was a distinct
word in Dacian Times not yet included in the front of the next word
(In fact is the same idea that Densusianu have had to can explain
Rom. asta from Lat. ista by adding a new a, with the difference that
now this *a is explained in an organic way)
II. If we have concluded about the proclitic PAlb a- we need to
identified also the main form of P-Alb. dem. pronoun 'this/that'
(fem. form in our discussion)
This form was:
PIE *k'jo fem. *k'ja:j < PAlb *tsja: > Rom. cea <-> Alb. so
For arguments regarding the forms above see the following
correspondance and examples:
1. Alb. sonte 'this night' > PAlb *tsja: *nakta < PIE k'ja:i nokWti
2. similar as in Rom. 'cea mioriTã' (in Mioritza poem) 'that
sheep-(dim.)' where Rom. cea < PAlb *tsja:
To talk now about the words as they can be fpound in Romanian and
Albanian dictionaries we have:
3. Rom. cea /c^ja/ - Alb so < PAlb: *tsja: /cja:/
4. Rom. acea /a-c^a/ - Alb aso < PAlb: *a tsja: /a cja/
III. In addition we can identify for PAlb 'she' the PAlb form ja: <
PIE ia: (fem. sg. of PIE dem. jo) based on the folowing words:
5. Rom aia /a-ja/ - Alb ajo < PAlb. *a ja: /a ja:/
Note: All the above forms shows Rom a: <-> Alb o (PAlb a:)
indicating PAlb. common forms in Pre-Romanian and Pre-Albanian older
than Roman Arrival in Balkans (when Lat a: > Alb a) .
For the fonetism Rom c^ - Alb s : see Rom. 'cioara' /c^warra >
c^orra > c^oara/ <-> Alb. 'sorrë' < PAlb *c^wara
Note also that PAlb c^ < PAlb cj is regular - conform Alb vis <
PAlb weits-ja: < PIE *weik'-ja:
IV. The last element to talk about here is PIE *to 'that one there'
that gave PAlb. *ta (see Alb. atë)
Having now a final picture (regarding the fem.sg. form of pers. and
dem. pronouns) the Daco-Moesians could have been talked like this:
1. "a tsja: budza" 'that/this lip' (PIE *o k'ja:i budh-jo)
=> Alb. aso buzë <-> Rom. acea budzã
2. "tsja: budza" 'id. or closer' (PIE *o k'ja:i budh-jo) =>
Alb. so buzë <-> Rom. cea budzã
3. "a tsja: ta budza" 'id. or closer' (PIE *o k'ja:i to budh-jo)
=> Alb. atë buzë <-> Rom astã budzã
4. "budza a ja:" 'id. or closer' (PIE *budh-jo o ja:) =>
Alb. ajo budza <-> Rom. budza aja
So I think that based on this:
Rom. 'asta' is linked with the contracted form *atsta (Alb. atë) of
PAlb. *a tsja ta
(for ct>st in Rom. see neguTator > *neguTtor > negustor and for k't
> t in Albanian see tetë '8' < *atsto:(ti) < ok'to:(ti) )
Rom 'aceasta' is the uncontracted form of PAlb. '*a tsja: ta'.
Note also that Romanian forms were derived/re-shaped/influenced (We
will never know exactly ) by the Latin forms of the Latin
correspondant pronouns.
Rom asta < Lat. ista
Rom. aceasta < Lat. *ecce - ista
Note: On the idea that 'aceasta' and 'asta' are related forms see
also DEX that considered aceasta as a compound form based on 'asta'
< Latin *ecce ista.
So in conclusion, if the Proto-Albanians have talked like: '*a tsja:
ta' ,'*tsja:' and '*a-ja:' and the Romans have talked
saying: 'ista', '*ecce ista' and 'illa' :
PRom forms like *a+esta, *a-ts-ta (Rom. aista/asta) ; *a-tsja+sta
(Rom.aceasta) , *a-tsja (Rom. acea), *tsja (Rom. cea) , *a-ja (Rom
aja) or *ja (Rom. ea) could be well imagine...
Also Albanian forms like: a-jo, a-so, a-të, a-saj a well explained
too.
Best Regards,
Marius
P.S.
And this is not the single example regarding Balkan Latin Forms that
was influenced by a Local Balkan Language too:
In Romanian we say: 'Eu am' for 'I have'
In Albanian they say: "Unë kam" for 'I have'
Even for Rom verb 'a avea' that is for sure the Latin
verb 'habe:re', the form at pers.I.sg. ind. pres. 'am' reflects the
Proto-Albanian form 'kam' . Densusianu accepts this obvious
influence. Rosetti says that Rom.I.sg form was reshape based on the
Rom.I.pl form 'avem' and as result Romanian 'am' has no link with
the Albanian form 'kam' (sic!)
But each time when I remeber the Albanian expression: "nuk i kam"
meaning 'I don't have them' I will always say that this is the same
expression as the Romanian one "nu -i am" 'id.' (doesn't matter
what Rosetti said)
Now finally to say how much in an expression like the one above is
from Latin than from Proto-Albanian or inverse we can never know: is
better to say that is from both and is due to the long interaction
between an "Albanoid" Local Population and the Roman Empire,
sometimes due also to some closer PAlb / Latin forms in some cases.
Best Regards,
Marius