Re: [tied] Re: lat. nux, -cis - PIE?

On 04-11-10 01:59, alexandru_mg3 wrote:

>
>
> "If <lakthi> is an authentic form,
> then you seem to be right about the origin of the <-j->, but we should
> start with something like *lakVc- or *lagVc- in Proto-Albanian"
>
> But in this case you are very close from my proposed form:
> PAlb. *alwaknic-ja or *alaknic-ja that fit very well too.
> There is no issue to derive 'lajthi' from here.

Except that you artificially stretch the word ONLY in order to
accommodate all the segments needed to relate it to the Romanian word.
That's Procrustean linguistics.

> Also there is no issue to compare Alb. 'lajthi' with existing
> Rom. 'aluniTa':
> 1. the lost of an initial a in Alb
> (as in: Rom. amorTi / Alb. mërdhi)
> 2. th < c (Rom. Tarc / Alb. thark)
> 3. and j < nj
>
> all of them are 'normal' derivations rules in Albanian.
>
> Applying them we easy obtain PAlb *alanjc-(ja)/*alwanjc-(ja) that
> is quite the same with the current Romanian word: 'aluniTa' 'small
> hazel nut'.

*wan^ > uni in Romanian? How come?

> So there is no issue to compare the Romanian and the Albanian
> form...in contrary.
>
> Could we ignore this obvious match between Romanian and Albanian
> forms 'lajthi'/'aluniTa' especially when we have about three hundred
> old common words between Albanian and Romanian ?
>
> Why Latv. lagzda or Slavic *le^ska are considered closer (even
> today there are no common derivation available) to Albanian 'lajthi'
> but not the existing Romanian word : 'aluniTa', when there are no
> difficulties to derive them from a common PAlb form and the meaning
> is the same ?

There are serious formal difficulties with your derivation. It looks
like an attempt to construct a common term for 'cat' and 'dog' through
Grimm's Law and metathesis (dog > tak > kat). Creative but hardly
compelling. Of course if you stuff a reconstruction full of extra
segments that can be manipulated to achieve the desired effect and then
deleted, relating almost _any_ two words is easy.

Piotr