On Fri, 7 Mar 2008 11:04:06 +0100, "fournet.arnaud"
<
fournet.arnaud@...> wrote:
>----- Original Message -----
>From: Miguel Carrasquer Vidal
>
>I do.
>
>A beautiful application of the law by Guus Kroonen can be
>found here:
>
>http://website.leidenuniv.nl/~kroonengj/montreal.pdf
>
>http://website.leidenuniv.nl/~kroonengj/iwoba3.ppt
>
>===========
>
>I find it hard to swallow
>that a "weak" consonant such as *n
>could provoke gemination of stops.
Gemination before a resonant is quite common. A good example
is Catalan <poble>, <segle>, pronounced /pobble/, /seggle/
(or /pople/, /sekle/). Whether through gemination or direct
devoicing, we also have Dutch d,g,v,z > t,k,f,s before
suffixes starting with n or l, in words like (gebeurd-nis >)
gebeurtenis, (koning-lijk >) koninklijk, (begraav-nis >)
begrafenis, etc.
The exact course followed in Proto-Germanic remains a matter
of dispute. Kluge himself assumed assimilation Tn > dd,
followed by standard Grimm devoicing, Lühr suggests
gemination before /n/ (Tn > ddn) with subsequent loss of the
nasal and Grimm devoicing (ddn > dd > tt). Given the Dutch
and Catalan examples above, another route could be Tn >
ddn/tn > tt.
>especially an assimilation of that kind.
>C-n > -CC- (left to right)
>when PIE phonotaxis is (right to left).
You mean like Bartholomae's law (*bhudh-tó- > buddhá-)...
>stake is obviously st_H2-k.
>I suppose nobody contests
>that *st_h2 is the root.
Guus Kroonen's analysis, with which I agree, is:
PGmc. N. *steko:, G. *stukkaz [or *stakkaz --mcv] < PIE
*stégh-o:n, G. *stgh-n-ós
This explains the Germanic variants:
*stekan-: ON ljós-stjaki m. ‘torch’, OHG stehho m. ‘pole,
club’
*stekkan-: OHG stecko m., OE stecca m. ‘stick’
*stakan-: ON ljós-staki m. ‘torch’, OE staca m. ‘stake’
*stakkon-: ON stakka f. ‘stump’
*stukka-: ON stokkr m. ‘stick, trunk’, OHG stoch m. ‘stick’
=======================
Miguel Carrasquer Vidal
miguelc@...