> Hans Kuhn: Das letzte Indogermanisch
> has the following Germanic doublets:
> "
> kriechen und nd. krupen,
> [Pok. *gr-eu- + -g / + -b, Balt., Germ.]
> streichen und streifen,
> [Pok. *str-ei- + -g / + -b, Celt., Balt. Germ.]
> tauchen und taufen,
> [Pok. ?, *dheu-b-, Gr., Celt., Balt., Germ.]
> nd. Siek und Siepen "feuchte Bodensenke",
> [Pok. *seikW- *seip-, Skt., Ital. Celt. Balt. Germ.]
> engl. shrink und hd. schrumpfen,
> [Pok. ?, *skerb(h)-, *skreb(h)-, Gr., Ital., Balt., Germ.)
> Strunk und Strumpf,
> [Pok. *strenk *streng-, ?, Gr., Celt. Germ.]
> got. *auhns/ altschw. ugn und dt. Ofen,
> an. ylgr "Wölfin" und ulfr ,Wolf,
> dt. leihen und bleiben
> "
>
> Odd geographical distribution, if it's caused by a PIE morpheme and
> not substrate.
Schrijver's 'language of geminates'
(Schrijver: Lost languages of Northern Europe, in
Early Contacts between
Uralic and Indo-European:
Linguistic and Archaeological
Considerations)
has
"
Proto-Germanic *du:B-:
Old Icelandic du:fa 'to immerse'
Proto-Germanic *duff-:
Faeroese duffa 'to bob up and down (of a ship)'
Proto-Germanic *dubb-:
Norwegian dubba 'to stoop', Middle Dutch dubben 'to immerse'
Proto-Germanic *dup:
Dutch duypen 'to hang one's head',
Proto-Germanic *dupp-: German düppen, Norwegian duppa 'to dive'
Proto-Germanic *dump-:
Norwegian, English, Danish dump 'hole, pit, pond',
East Frisian dumpen 'dive'
Cognates:
Lithuanian dubùs 'deep', dum~blas 'mud in water, marsh';
Old Irish domain,
Welsh dwfn 'deep' < *dubni- and others.
"
He doesn't mention the many variants of *dukk- *deuk- "dive" but adds
*sup- *supp- vs. *sug- *sukk-
*dup- and *duk- are strong verbs, *dupp- and *dukk- are not.
Similarly for *sup- *sug- vs *supp- *sukk-
*gr-eu-b- and *gr-eu-g- are strong in Germanic
*str-ei-b- and *str-ei-g- are strong in Germanic,
*str-i-bb- and *str-i-gg- are not.
Did the substrate language of geminates already have ablaut, in spite
of not being IE?
Torsten