Re: *-s > *-z > Norse -r

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

> In a nutshell:
>
> PIE inflectional *-s became voiced in PGmc., giving *-z (= early
> Runic <-R>). Its further development in Scandinavian was -r.
> Rhotacism, or the lenition *z > r, is a fairly common process, cf.
> Latin intervocalic *-s- > *-z- > -r- (tempus : temporis, flo:s :
> flo:ris).
>
> PIE *gHostis > PGmc. *gastiz (Runic gastiR) > ON gestr 'guest'
> PIE *gHostejes > PGmc. *gastijiz > *gasti:z > ON gestir 'guests'
>
> PGmc. *gaizaz > ON geirr 'spear'
>
> (Note the shortening or loss of unstressed vowels.)
>
> The further fates of the final -z (-R) were pretty complicated and
> differed from one Old Norse dialect to another. It was often
> assimilated to the preceding consonant after the intervening
> unstressed vowel had been lost:
>
> *stainaz > *stainR > steinn 'stone'
> *sto:laz > *sto:lR > stóll 'chair'
> *i:saz > *i:sR > íss 'ice'
>
> There were some characteristic dialectal developments:
>
> Proto-Norse *mannR > OIc. maðr (acc. mann) 'man'
>
> In Old Swedish an epenthetic reduced vowel (written <e>) appeared
> before a stem-final consonant and unassimilated -r. A similar
> epenthesis has taken place in Modern Icelandic and Faroese (where

the

> spelling is <u>):
>
> *xringaz > OSw. ringer; OIc. hringr > ModIc. hringur 'ring'
> *fiskaz > OSw. fisker; OIc. fiskr > ModIc. fiskur 'fish'
>
> In the later history of the Scandinavian languages (except

Icelandic

> and Faroese) the declensional system was radically simplified. In
> particular, the nom. and acc.sg. of masculine nouns were levelled

out

> and the nom.sg. -r ending (OSw. -er) was dropped. This was a
> morphological, not a phonetic proces, so -(e)r of other origin was
> unaffected.
>
> Piotr

I thought *z > *R > *r had been dropped in favor of *z > *r (which
means the rune for <R> should be reinterpreted to mean <z>)? Is <R>
needed?

Torsten