Re: Re[4]: Fw: [tied] Pferd

From: fournet.arnaud
Message: 49473
Date: 2007-08-09

Germanic f is a "decent" pointer to *p but it is not decisive,
if ones looks at number four from *kwettwer and wolf from wlkw-os.
*p can arise from *kw, and this happens in Germanic and also LAtin.
lupus also is from *wlkwos.
 
So if I try to summarize my idea more "consistently",
I start with root *kw_r "tree",
this root can give :
*kw_r-kw hence both p_rkw and kw_rkw "oak"
*kw_r-esno "tree"
*kw_r+ XX "other tree names"
 
Note that in Sanscrit parkata (supposedly *perkw-) is a Ficus
But karibha is also a Ficus (< *kwer- )
If you accept parkata, take karibha too.
 
Besides,
how do you explain foraha with -o- in front of quercus with -e- ?
 
To be continued, I presume !
 
 
 
----- Original Message -----
From: Brian M. Scott
To: fournet.arnaud
Sent: Thursday, August 09, 2007 8:35 AM
Subject: Re[4]: Fw: [tied] Pferd

At 8:16:12 AM on Tuesday, August 7, 2007, fournet.arnaud
wrote:

> From: Brian M. Scott

>> At 3:00:32 PM on Monday, August 6, 2007, fournet.arnaud
>> wrote:

>>> The rook *kwr- for "oak" also exists in LAtin : Quercus.

>> Except that <quercus> is from *perkWu-.

> A.F : Quercus comes from supposedly *perkwu if one is to
> accept that This *perkw root can link Latin quercus
> oak-tree with German faraha fir-tree.

Which of course I do. And that should be OHG <foraha>.

> Neither Germanic nor LAtin helps determine if the root
> starts with *p or *kw.

Note Skt. <parkati:-> 'heilger Feigenbaum', with /p-/. And
the Gmc. *initial* /f-/ is a pretty decent pointer to *p-.

[...]

> Personnally, I prefer considering a dialectal origin of a
> LAtin word to accepting a remote semantic relationship
> between a fir-tree and an oak-tree.

Remote?! The semantic relationship is very close.

> As for hyrst, it does not mean oak or tree but grove.

It's beside the point here: it belongs with OIr <crann>, not
with OHG <foraha> and Latin <quercus>.

[...]

>>> Note that Irish has crann "tree" and Breton" has prenn
>>> "tree" from *kwr-enn.

>> From PCelt. *kWresno- 'wood, tree', according to
>> Matasovic, from PIE *kWres-; OE <hyrst> is cognate, and
>> <quercus> is unrelated.

>>> Note that this root *kwr- also gives :
>>> - portuguese car_v_alho
>>> - other names of tree : charme, charmille, etc in French

>> <Charme> is from Latin <carpinum>. After loss of the /i/,
>> the group /rpn/, like most groups of three consonants not
>> ending in /r/ or /l/, was reduced by loss of the middle
>> consonant, but not before the nasal was labialized; the
>> result was /rm/.

> A.F : so !?

So I see no obvious reason to object to Pokorny's assignment
of <carpinus> to an extension of *(s)ker- 'to cut'. You now
appear to be conflating at least *three* PIE roots.

> These phonetic details confirm the idea that *kw_r/k_r is
> an obvious root for oak and some other tree names in
> Celtic and Italic languages.

You're inconsistent: on the one hand you object to the idea
that a single root might refer to fir and oak in different
daughter languages, but on the other hand you're now happy
to have 'an obvious root for oak _and_some_other_ tree_names_ '
(emphasis added).

[...]

Brian