Well, don't write it off yet. My opinion is
that Lat. digitus would have given Albanian *diit- > *di:t (cf.
kujtoj <-- cogito:) or *digjet (cf. shëgjetë <-- sagitta) depending on the
date of the borrowing, while hypothetical *dictus (with early syncope) would
have ended up as *dift (cf. luftë < lucta). I can't see how either the
initial <d-> or the final <-sht> can be explained if we start with
<digitus>. As for the absence of palatalisation, the reconstruction
*gwist- assumes an initial cluster (*g + *w), NOT a labiovelar, and the expected
reflex of *gwi- in Albanian is <gi->.
As far as I'm concerned, the bottom line
seems to be that Rom. de$t derives from <digitus> while Alb. gisht,
despite the superficial similarity, is a different word, probably inherited and
possibly related to <kvistr>.
Piotr
----- Original Message -----
Sent: Wednesday, September 18, 2002 8:35 PM
Subject: [tied] Re: *gwistis
I had expected to find PIE gwistis > Albanian gisht.
I suppose it was too simple to be true. (The apparent resistance to
palatalisation had surprised me.) If, as George has just said, it too has
nothing to do with PIE *gwistis, e.g. Old Norse kvistr, this should definitely
go on the list of pseudo-cognates.