At 6:22:20 PM on Saturday, November 8, 2008, tgpedersen
wrote:
[...]
> It is also known that no Germanic word could have begun a
> p-, or be of the form TVT where T is any unvoiced stop (p,
> t, k, kW) and V is a vowel.
Ringe offers two possible counterexamples to that last
assertion. One is PGmc. *kWikWaz 'alive' (ON <kvikr>, OE
<cwic>) < *kWikwós < PIE *gWih3wós by hardening of *h3 to *k
by Cowgill's law (*h23 > *k/R_w). The other is PGmc. *te:k-
~ *tak- (Goth. <tekan> 'to touch', ON <taka> 'to take'),
which nicely matches Toch. B <täk-> 'touch, feel with the
hand; fetch, procure'). In _From Proto-Indo-European to
Proto-Germanic_ he derives these from post-PIE *deh1g- ~
*dh1g- 'touch'; from Adams' _Dictionary of Tocharian B_ I
gather that Ringe has previously attributed the Gmc.-Toch.
match to borrowing in a direction that can't be determined.
> Words of that form, of which some but not all are common
> to both West and North Germanic are therefore words of a
> substratum in North and West Germanic.
Or early (but post-Grimm) borrowings from Latin (probably
*punda 'pound', *katilaz 'kettle', the *kaup- family), or
the pre-Grimm borrowing *paido: 'cloak' (cf. Gk. <baíte:> 'a
shepherd's cloak') from some eastern language.
Brian