--- In
cybalist@yahoogroups.com, Piotr Gasiorowski <gpiotr@...> wrote:
>
> On 2007-05-16 17:20, tgpedersen wrote:
>
> > Ahem. You have made your derivation more parsimonious by dodging
your
> > own argument from pompe < *kWonkWe < *kWenkWe < *penkWe, which is
> > therefore also a three-step process.
>
> But if *kWénkWe is accepted for common Italo-Celtic (whether a true
> synapomorphy or an areal innovation), we get a neat scheme for both
> branches, with *pompe and <qui:nqe> derived from the same PIt.
form. The
> overall picture _is_ parsimonious.
> Piotr
>
Piotr, *kWenKWe CANNOT BE a Common Italo-Celtic form:
Let's follow your idea of p..kW > kW..kW Common to Italo-Celtic...
There are three words having p...kW
1. penkWe 'five'
3. pekW- 'to cook'
2. perkW- 'name of a tree' ->*perkWunyo > Proto-Celtic *perkunia:
(attested Hercynia (Caesar), loaned with initial p- in Gothic
fairguni 'mountain') <-> and on the other hand: Q-Italic *kWerkWus
(Latin quercus 'oak') => so you can see as me that there is no common
p...kW > kW ...kW
Because :
EITHER
1. the delabialisation in Proto-Celtic *perkWus (-> *perkWunia:)
happened BEFORE p..kW>kW...kW in Proto-Celtic
but
2. the delabialisation kW/u > ku in Proto-Italic *perkWus (Latin
quercus) happened AFTER the p..kW > kW..kW in Proto-Italic
so BASED on 1..2 There is no Common Proto-Italo-Celtic p...kW >
kW...kW...
(The point was raised by Watkins 1966)
OR
even more drastically (and this is my Point):
3. THERE IS NOT AT ALL p...kW > kW...kW in Proto-Celtic (see
Hercynia < *perkunia < *perkWunia )
versus
PIE *perkW- > Q-Italic *kWerkus > Latin quercus =>
so p...kW > kW...kW Transformation is available ONLY for Q-Italic >
Latin ...
And I say this based on kW/u > ku Timeframe:
As you can see the delabialisation kW/u > ku (perkunia/quercus <
*perkWunia/*perkWus) happened in both Languages, so based on what to
supposed different timeframes for it? Only to keep alive the p..kW >
kW...kW for Proto-Celtic?
In fact kW/u > ku : is the REAL COMMON Transformation (belonging to
Dialectal PIE times: you agree (I hope) that this kind of
delabialisations kW/u > ku are very old, really belonging to PIE (or
PIE-dialectal) times
So this story regarding different kW/u > ku timeframes in Proto-
Italic and Proto-Celtic Times , Only to keep the fix idea of a common
p..kW > kW...kW should be dropped: the delabialisation : kW/u > ku is
really Common To a PIE-Dialect (maybe common to the Italo-Celtic-PIE-
dialect if this PIE dialectal really existed)
As result =>
the P-Celtic form *pempe < *penkWe shows that there are ADDITIONAL
CONTEXTS at least for P-Celtic (but pt-, ps-) WHERE initial p- is not
lost (because:
3. if there is no *p..kW > *kW...kW in Proto-Celtic and
4. and PIE *penkWe gave P-Celtic *pempe,
=> 3. and 4. shows us an initial PIE *p- 'well and safe' that
is 'still there'...in p-Celtic *pempe
Also, I don't like these artificial equations (that seems that you
like a lot):
*p > *kW > *p (see Romanian apa too :) )
when *p > *p is the simplest way to arrive to *p
why not to try a more complicate solution:
*p > *kW > *p > *kW > p ...etc...
Marius
P.S.:
1. My point regarding the existence of PIE P/Q dialects and the
existence of *p-forms in Dialectal PIE : keeps also in the equation
the Proto-Germanic *pempe too...your supposition p..kW > kW..kW for
Celtic, ignore any reasonable explanation regarding the Germanic p-
forms (where my explanation says that -> 'the Germanic p-Forms are
inherited from the PIE Dialectal P-forms')...
2. Timeframes:
1. p-forms in PIE : *penpe, *xWlpos, *petwores, *xWap-eh2
2. p...kW > kW..kW , xW..p > xW...kW p..tw>kW..tw in PIE-Q-Dialectal
3. mixtures of (some) PIE P-Q dialects
4. kW/u > ku in merged Q-P-PIE-Dialects
5. PIE big split kept the trace of original PIE P/Q dialects
3. These timeframes has sense also in Relation with *h2ap-eh2 'water'
word too (if one of *h3 or *h2 => xW)
The original PIE form was apa: (as in Romanian :))
based on the large spreading area of this form
NEXT in the PIE-Q-Dialect : NOT only p...kW > kW..kW but also kW...p
> kW..kW and also xW..p > xW..kW took place...
Now if *xW was *h3 or *h2 doesn't matter here, lets say *xW > *h2
PIE-P-Dialect *h2ap-eh2 > *xWap-eh2 > *ap-eh2 > apa:
PIE-Q-Dialect *h2ap-eh2 > *xWap-eh2 > *xWakW-eh2 > akWa:
4. SO PIE-Q-Dialect has transformed:
*penkWe '5' > [p...kW > kW..kW] > *kWenkWe
*petwores '4' > [p...tW > kW..tW] > *kWetwores
*xWlpos 'wolf' > [xW..p > xW..kW] > *xWlkWos
*xWap-eh2 'water' > [xW..p > xW..kW] > *xWakW-eh2 > akWa: