From: Piotr Gasiorowski
Message: 18527
Date: 2003-02-07
----- Original Message -----
From: <a96_aeu@...>
To: <cybalist@yahoogroups.com>
Sent: Friday, February 07, 2003 2:31 PM
Subject: [tied] alb. gji (breast)
> Does anybody know if there is any relation between Albanian 'gji' (bosom, breast) and Latin 'die' (suckle) ?
Latin what? There is no such thing as Latin *die 'suckle'. Perhaps you mean Norwegian die 'suck'. Anyway, there is no connection.
The origin of <gji> is a nice problem because the solution is rather unexpected. You can use it as an illustration of how etymological analysis works. The def. form <gjiri> and the pl. <gjinj> show that the word once ended in *-n- (rhotacised between vowels in Tosk), and the initial <gj-> makes us expect pre-Albanian *s- unless the word is a recent loan with palatalised /g/ (but no source suggests itself readily). The vowel <i> may continue (most likely) pre-Albanian *i, *i: or *ei. The prototype might therefore be of the form *sin-, *si:n- or *sein-. Additionally, it may be significant that the word also means 'gulf' in Albanian.
This should immediately ring a bell. Lat. sinus has exactly the right form and meaning. Apart from 'bent surface, curve, fold' it means BOTH 'bosom, lap' AND 'bay, gulf'. Since these are derived, rather than inherited, meanings in Latin (e.g. 'bosom' from 'the hanging fold of a toga'), and the word has no obvious connections outside Latin (as far as I know), it seems obvious that the Albanian word is a loan rather than a cognate. There is one difficulty with this. <gji> must be older than the loans in which Lat. s- gives Alb. sh- before a vowel, and it forces us to assume that the initial voicing of old *s- in pre-Albanian (> Mod. Alb. gj-) took place _during_ Latin-Albanian contacts, not _before_ them. Any ideas how to test this hypothesis?
Piotr