Re: [tied] the fascination of illV

From: Miguel Carrasquer
Message: 30687
Date: 2004-02-05

On Wed, 04 Feb 2004 23:52:33 +0100, altamix <alxmoeller@...> wrote:

>I guess is high time to put the point on the "i" and to show how fascinating
>in the Romanistic world the word "illV" can be.
>It-s reflexes in Romanian are considered to be as follow:
>el ( he) < Lat. illum
>ea (she) < Lat illa
>ãla ( that he one) < Lat illum
>aia ( that she one) < Lat. illa
>ãia ( these (masc)) < Lat. illum
>alea( these (fem)) < Lat illa
>al ( of) < Lat. illum
>a( of) < Latin illa
>ai (of masc. pl) < Latin illi
>ale ( of fem. pl) < Latin ille

Apart from the errors and omissions in the above (the demonstrative forms
have affixed -a, so that <ãla> is not from illum, but from (probably)
<illum hac>; the plurals are not from illum/illa but of course from
illi/illae; the forms al, ai, etc. are just forms of the preposition <ad>
with suffixed definite article).

>There are more of these *grosso modo said* , I don't remember right now
>about all them and there is not really a need to. To summ-up, the reflexes
>of Latin "illV" should be as follow:
>el, ea, ãla, aia, ãia, alea, al, a, ai, ale

The only relevant forms are:

3rd. person pronoun:

nom. el /jel/ ea /ja/
acc. (â)l o
dat. lui ei /jei/
encl. (â)i (same as masc.)

nom. ei (/jej/) ele /jele/
acc. (â)i le
dat. lor (same as masc.)
encl. le/li (same as masc.)

definite article

nom. -l, -le -a, -ua
dat -lui -ei

nom. -i -le
dat. -lor -lor

These forms are regularly derived from the (Vulgar) Latin paradigm of
<ille>, which was:

rect. ille, illum > élle, éllu illa > élla
encl. illum, ille > llu, lle illam > lla
obl. *illui > llúi *illaei > llÉi
encl. illi: > lli illi: > lli
rect. illi: > élli illae > élle
encl. illi: > lli illae > lle
obl. illo:rum > llóru illorum > llóru
encl. *illis > lle *illis > lle
rect. illo:s > éllos illa:s > éllas
encl. illo:s > llos illa:s > llas
obl. illo:rum > llóru illo:rum > llóru
encl. *illis > lles *illis > lles

In Romanian, the relevant soundlaws are those for the development of /ll/,
which are:

stressed unstressed
*lla *wá *wã
*lle lé le
*llE *l^é --
*lli *l^í *l^(i)
*llo lo --
*llu lu l(u)

The forms thus become:

rect. *élu *éwã
encl. *lu, *le *wã
obl. *lúi *l^éi
encl. *l^i *l^i
rect. *él^i *éle
encl. *l^i *le
obl. *lóru *lóru
encl. *le *le

The final steps to get the standard Romanian forms are simple:
- loss of final -u
- l^ > j
- loss of final -i
- j-prosthesis on initial e-

The only complication is the development of *wã. In isolation, /wã/
develops to /o/. After vowel, w > 0, then if the result is ãã, or aã,
these become /a/. In the sequence wãwã (stella lla > *steawãwã), -wãwã >
-wãã > -ua. Arromanian has wã > o also after /(e)a/ (stella > *steawã > MR
steao), against DR steá, with loss of -w- [*steawã > *steaã > stea].

The final result is:

rect. jél jéã > já
encl. -l, -le, (â)l o, -a, -ua
obl. lúj, -luj jéj, -(j)ej
encl. (â)j (â)j
rect. jéj jéle
encl. -j, (â)j -le, le
obl. -lor, lór -lor, lór
encl. le (li) le (li)

>Even phonologicaly the "die:s" does not fit. It should have yelded *dees.

No, that's just your ignorance. It yielded *die. Latin /i/ is preserved
when stressed in hiatus, and unstressed -Vs becomes -V.

>Since it was so several times assumed that final "s" yelded "i" in
>BalkanoRomanace ( what a pitty!!!), the word should have been *deei.

No, that's just your ignorance. -Vs only yielded -Vj if the preceding
vowel was stressed.

>Dear Miguel, if "Diana" was already "Djana" in Latin it should have entered
>Romanian as "Djana" and it should have yeleded "Ziana".

No, that's just your ignorance. It yielded zânã.

Miguel Carrasquer Vidal