Re: Latin longus < *dlh1-ongh-o

--- In cybalist@yahoogroups.com, "Abdullah Konushevci"
<akonushevci@...> wrote:

> ************
> For me this is all the question of phonotactics: a set of
> constrains, language characteristic, on how sequences of segment
> pattern. Because -dl- has constrains on this clsuter of segment
> modeling, it is natural this cluster to be treated as simple /l/,

as

> do, for example, also cluster -sl-. This cluster was replaced in
> Albanian and Greek with -gl- that has no constrains in both
> languages (cf. Alb. i glatë 'long' > Gheg <i gatë> and Tosk <i
> gjatë> or Greek *dl.ku > glukus 'sweet'). This cluster in Slavic
> languages has no constrains: dlaka 'hair', dlan 'palm' etc. So, as
> Germanic languages, as Latin, regarding cluster -dl- has

constrains,

> therefor, *dlongho- as possibly extended (*dlon-gho) has as

regular

> outcome OHG lang and Latin longus.
>
> Konushevci

You are right here. Only the Lithuanian ilgas seems not to be
regular: ilgas < *(d)lh1-gh-o-

The PAlb/Dacian? form was *glagta: < [gl < dl; al < lh1] < PIE *dlh1-
gh-to (reshaped in -eh2) (see Albanian-Tosk (i) gjatë)

Note also here another good example of PIE *lh1 > PAlb/Dacian? *al

So the problems are mainly 'at the end' of the word : like how to
unify: *dolh1-i-gh-o- with *dlh1-ongh-o etc... The meaning and the
forms are 'so similar' so hardly we can imagine different roots.

(see also the other messages and the proposal of a possible n^).

Marius