I forward you my message from Balkanika regarding this derivation:
===================================================================
From: "alexandru_mg3" <alexandru_mg3@...>
Date: Sat Oct 30, 2004 7:56 pm
Subject: Re: Deriving 'aluna'-'lajthi' from PIE *el-1 -> same
semantism as in 'hazel nut'
Hello,
I didn't say that 'aluna' and 'hazel nut' are derived from the same
roots.
I wanted only to say that 'hazel nut' is contructed based on the
same ideea as 'aluna' : 'kind of brown nut' BUT of course not from
the same PIE roots.
Until now I thought that we have only 'a kind of brown' ideea
in 'aluna' but now I think that we have also the 'nut' notion inside.
So in my opinion 'aluna'/ 'lajthi' is a compound word that at his
origin means 'hazel nut'. So based on this semantism the PIE form is
something like : *e:l + V + *knei-k^-ja
where:
*e:l is a long form of Pokorny *el-1 'red, brown in trees and
animals names'
and *kne-i is a i-bases extension of Pokorny PIE root *ken-1 (ken-1
is the PIE root that gives also Latin nux etc...)
see below From Pokorny:
" Root / lemma: ken-1
Meaning: to press, pinch, etc.
German meaning: as basis for extensions the meaning
`zusamendrücken, kneifen, zusamenknicken; Zusamengedrücktes,
Geballtes'
[...]
A i-basis in kneigʷh-, knei-b- `neigen'? (see there).
u-basis kneu- and extensions:
mir. cnû, Gen. cnô (*knûs, *knuu̯os), cymr. cneuen, Pl. cnau,
mcorn. knyfan, mbret. knoen `Nuß', gall. *knou̯â; derived mir.
cnuas
`Nüsse; harvest'; with d-suffix aisl. hnot, ags. hnutu, ahd. (h)nuz
`Nuß'; with k-suffix lat. nux, -cis `Nuß' (`Nuß' also actually
`Kügelchen, clots '); Lohmann ZceltPh. 19, 62 ff.
"
Note : Most probable the linked vowels V of *e:l + V + *knei-k^-ja
was a kind of 'wa' that
explains both 'la' in Albanian from 'lwa' > 'lla' > 'la'
and explains also 'u' in Romanian 'lwa'+'nj' > 'lua' + 'nj' > 'lu'
+ 'ni' (or was simple an -a- in Proto-Albanian with a dialectal form
-u- in a Northern Proto-Albanian Dialect)
Albanian Derivation gives:
--------------------------
PIE *e:l + V + *knei-k^-ja > Early PAlb. a:l + V['wa'] + knei+tsia >
PAlb. a:lwaknjtsia > llanjtsia > lanjtsia > lajthi
Romanian derivation gives:
--------------------------
PIE *e:l + V + *knei-k^-ja > Early PAlb. al + V['wa'] + knei+tsia >
PAlb. *alwaknitsia > allunitsa > aluniTa > aluna (regression
because -iTa is a diminutive in Romanian)
Note : no rothacism in Romanian due to 'll'
and explained also Albanian Dialectal form 'lakthi' indicated by Orel
---------------------------------------------------------------------
PIE *e:l + V + *knei-k^-ja > Early PAlb. a:l + V['wa'] + knei+tsia >
PAlb. *alwaknitsia > Dialectal Alb *alwak(n)tsia > llaktsia > lakthi
So in my opinion the Dacian pronunciation started with something
like :
*alwakneitsia = *alwa + *kneitsia = 'kind of brown' + 'nut')
and ended with something like :
*alwaknitsia = *alwa + *knitsia > *aluanitsia = *alua + *nitsia
Best Regards,
marius
--- In
cybalist@yahoogroups.com, Piotr Gasiorowski <gpiotr@...>
wrote:
> On 04-11-08 11:37, alexandru_mg3 wrote:
>
> > Could Albanian "lajthi" ('hazel nut') to belong to the same
root as
> > a compound word PIE *e:l + *V + *knei-k^-io: with the
meaning 'kind
> > of brown nut' ?
> >
> > Where e:l is a long form of Pokorny "el-1" 'red brown in the
name
> > of tree or animals'
>
> Pokorny's root is definitely too inclusive, in stark contrast to
the
> shaky reconstructibility of the underlying adjective *h1el-u-. I'm
not
> even sure if *h1élh1on- 'red deer' is derived from a PIE colour
term
> (though the EIEC essentially agrees with Pokorny, despite certain
> reservations): as far as I'm concerned, it looks like an old active
> participle derived from a stative verb -- *h1él-h1-on(t) (a Narten
> root?), although I can't identify the verb itself (well, perhaps it
_is_
> *h1eleh1- 'be reddish-brown' after all, but I'd like to see its
direct
> attestation).
>
> >
> > The Proto-Albanian form would be in this case : *al+wa+kneitsia
>
> > *allaknitsa ?
>
> I still don't see how you could get <lajthi> out of it. It seems
more
> profitable to compare it with Slavic *le^ska '(hazel) stick, hazel'
(Cz.
> líska, Blg. leska, Pol. laska, leszczyna) as a cognate. It's a
difficult
> etymon, especially if Baltic *lazda: is to be integrated with the
rest
> of the data. Pokorny's suggestion, *le:g^(H)- is hardly
satisfactory.
> Any thoughts, anybody?
>
> Pioitr