Re: Uralic Loanwords in Germanic

--- In cybalist@yahoogroups.com, Piotr Gasiorowski <gpiotr@...> wrote:

>
> W dniu 2010-02-14 21:56, Brian M. Scott pisze:
>
> > Ringe doesn't think so, though he notes that owing to
> > remodelling in all of the daughters, the PGmc. nom. sing.
> > isn't recoverable. For instance, he conjectures a nom.-acc.
> > sing. *segaz.
>
> The most common remodellings include simple thematisation (*siGiz- ->
> *siGiza-) and morphological truncation yielding an i-stem (*siGiz- ->
> *siGi-). The type od OE lamb/lambru, OHG lamb/lembir may reflect
> *lamBaz, but ON lamb is just "plain thematic" *lamBa-, with an early
> loss of the suffix (in all case forms). The full -e/os-stem paradigm is
> not attested anywhere in Germanic; we can only see its partial remnants.

For this particular noun W. Streitberg, "Zur Geschichte der /es/-Staemme" (PBB 15:504-6, 1891) suggested metaplasm by extraction from compound names, at least in West Germanic. He observed that Segimerus, Segimundus, and Segestes came from the same family. The forms with Segi-m- against Seges- can be explained by assuming assimilation of */sm/ to */mm/, degemination of */mm/ after an unstressed syllable, and closing of unstressed /e/ toward /i/ before the resulting /m/. For the first and second processes W.S. has other examples. The two names in Segim-, later *Sigim-, were presumably popular enough to allow the extraction of *sigi-, once the declension of the /s/-stem had become problematic for speakers.

In East Germanic the opposite process apparently restored *Sigisme:r-, *Sigismund- from the phonetically regular forms in *Sigim-, since we have them as names of Goths and Burgundians, and Italian Sismondi (from Langobardic?). That is, the early EGmc /s/-declension was still sufficiently coherent for speakers to recognize *sigis- as the combinative stem-form of this noun.

Forgive me if this matter was already covered before I joined the group!

DGK