Grzegorz Jagodzinski wrote:
> Of course it may be an old loan or an onomatopoeia. Of course Latin habe:-
> and Germanic habe:- may be unrelated.
They _are_ morphologically parallel, but the one contains *kap- (the
Germanic *b is due to Verner's Law) and the other *gHebH-. The
similarity of these roots may point to some kind of genuine relatedness
(but not of the regular type and therefore perhaps beyond our reach), or
may be a matter of phonaesthetic convergence; finally it may be just
an illusion produced by coincidence. How can we know? I've seen stranger
and closer coincidences than this.
> Of course Latin capere and Germanic
> kap-, ko:p- (two IE plain voiced stops in one root!) may be unrelated as
> well. This all can be just a result of coincidence. And I have right not to
> believe in the coincidence here. Have I? Which is more, I have a firm basis
> for such a belief.
You are absolutely free to believe in anything. But the intensity of
your belief is no argument in a discussion. Most amateurs find it hard
to believe that Gk. tHeos and Lat. deus are unrelated. Same meaning,
similar form. So what? The similarity is accidental none the less. No
matter how similar pre-English *ko:p-jan may be to <capio:>, we have no
right to regard them as related until the irregular correspondence
receives a plausible explanation. What you have argued is that if we
only squeeze enough laryngeals into the reconstructed root any
correspondence becomes possible as long as the stops in question have
the same place of articulation. I can't consider this sound methodology;
it's just an excuse for ignoring the normal constraints on reconstructions.
> See the list of more such "coincidences" on my page
> http://www.aries.com.pl/grzegorzj/lingwen/iesem2.html (near the bottom of
> the page). And notice that verbs meaning "have" change irregularily because
> of frequency and it is a LAW, not a coincidence. E.g. Polish miec' instead
> of imiec', Italian ho < habeo, Eng. has, had with no -v- etc.
I'm sure you know that <miec'> instead of <imiec'> is just one of
several cases of i-/zero doublets in Polish in words beginning with
*jI-. We also have <gra ~ igra> 'game' (the latter variant obsolescent
in Mod. Polish) and <skra ~ iskra> 'spark'. This means that there's
nothing exceptional about <miec'>. I agree that frequently used function
words often undergo irregular phonetic attrition, but was PIE *káp-je-ti
a function verb? The meaning of the root wasn't even 'have', strictly
speaking, but something like 'take in hand, grasp, catch'.
Piotr