Harald Hammarstrom wrote:
> I don't know of any particular reason but afaik in latin as well as
> vulgar latin it was as you say 11-17 as digit + ten and 18-19 had
> subtraction. No romance language continues the latin 17-19 but replace
> them by (continuations of) decem et digit or decem ac digit or
> decem digit asyndetically (e.g Catalan, modern (but not old) French).
> Spanish and Portuguese have long replaced their inherited seze with
> diecis'eis/dezasseis and there are also attested similar forms for e.g
> 12 (diz e dos), 13, 14 from the 13th century. Romanian has of course
> completely remodeled their 11-19 possibly by slavic influence or
> other influence that spurred both the slavic and the Romanian
> formations. Bearnese is unique among Romance lgs in having the form
> tr'es-ch'eys (three-six) for eighteen!
>
> Alll according to Price in _Indo-European Numerals_ ed. Gvozdanovic.
>
> mvh Harald
>
To me it seems there is one system three in Alb. Slavic and Rom. where
different is just the preposition which binds the words. For Slavic
should be "na"; for Alb. "mbë", and for Rom. "pe" and "spre"
For "ten" we have in Alb. "djetë" in Rom. "zece", in Slavic variants of
"desiat"
Now, let us see from 11 to 19:
Alb. "mbë" means in, on, after; I will like to ask Abdullah how here the
"mbë" is "felt" by native speakers. Is this felt as "on" or as "after"?.
I doubt about this prep. being felt as "in" in this case, but I can be
wrong.
Alb:
njëmbëdhjetë (një-mbë-dhjetë), one-on-ten
dymbëdhjetë (dy-mbë-dhjetë) , two-on-ten
trembedhjetë
katërmbëdhjetë
pesëmbëdhjetë
gjashtëmbëdhjetë
shtatëmbëdhjetë
tetëmbëdhjetë
nentëmbëdhjetë
In Rom. the things are a bit complicated due the short forms; the prep.
"spre" given as deriving from Latin "super"
means "toward", "to" ("english "to" is for me a bit too generalised and
thus I would advice to see "spre" as "toward")
Rom, literary:
unsprezece (un-spre-zece), one-toward-ten
doisprezece (doi-spre-zece), two-toward-ten
treisprezece
patrusprezece (*)
cincisprezece (*)
Saisprezece
Saptisprezece
Optisprezece
nouãsprezece
(*)there is no use of patrusprezece or cincisprezece. These forms are
weierd sincoped. For "patrusprezece" there is "paisprezece" and for
"cincisprezece" there is "cinsprezece"; if the syncope/assimilation of
second "ci" is relativ easy to explain, the elidation of the group
"-tru-" is not so easy, but for this see punct B.
A. The numerals used by folk's mouth are in the shorted form ( I wonder
if Alb. has too a such short form for them). Thus we have:
unSpe ( instead of unsprezece)
doiSpe ( instead of doisprezece)
treiSpe
paiSpe
cinSpe
SaiSpe
SaptiSpe
optiSpe
nouãSpe
I must say, I did not read anything about these shorted form and at the
first view one should say there are the shorted forms where people
abandoned ( meanwhile I love this word) the "ten":
unsprezece > unspre(zece)
The phnetic change is curious. Why "spre" > Spe (s^pe) ?
One will assume there is spre > spe (trough sincope of "r"): from this
*spe > s^pe, where "s^" is to obtain under the influence of the "e";
this change appears a bit unusual though.
That should be one hypothesis.The second one should be that we do not
have here to count with the preposition "spre" but with the prep "pe"
(on).In this case appears the question : where from is the "s" there? In
a construct as "unpezece" there should be no explanation for the "s".An
opinion about this should be that the counting system here was "one and
on ten" " unu Si pe zece":
unu-Si-pe-(zece) > unSipe > unSpe ( the "i" was absorbed in the "i" of
"s^")
B. The reduction of "tru" arrise an another problem with it. As there is
to see, the composition in the shorted manner but in the literary one
too has there an "i" which I suppose is considered to be analogical by
"four", "six", "seven" and "eight", forms "modeled" after "two" ,
"three" and "five"
The nominativ forms are unu, doi, trei, patru, cinci, sase, sapte, opt,
nouã. The forms for 12, 13, 15, appears to be the forms which made to
appear this parasitar "i":
12= doi-spre-zece / doi-S-pe
13= trei-spre-zece / trei-S-pe
15= cin(ci)-spre-zece /cin(ci)-S-pe
If one want to let the explanation be trough analogy, then we can
explain the aparition of "i" in the 14,16,17,18 trough influence of
12,13,15.
These won't explain tough the reduction of "patru" to "pai".
My questions here are:
-in fact how can be explained this reduction of "patru" to "pai"?
-why is this countig system considered a Slavic one since the Alb.
presents the same way to count?
If I am not too wrong in South Slavic there is too a curious reduction
of some numbers from 11-19: for instance there should be "adin-na-cat"
ehre "c"="ts" but "dva-cat", "tri-cat", "c^etr-na-cat" etc. ( for the
slavic examples I just qouted from memory, thus please verify these
exemples if there are some doubts regarding their correct form)
Alex