Re: [tied] Rune-essay Mads Peder Nordbo

From: Piotr Gasiorowski
Message: 11815
Date: 2001-12-16

That gets you into trouble if you try to analyse intervocalic -sC- clusters in the same way. There are sure examples of Germanic *-st- and *-sk- (< *-st-/*-sd- and *-sk-/*-sg-) contrasting with *-zd- and *-zg- (< *-sdH-, *-sgH-). I can't think of any instances of *-zb- < *-sbH-, but phonotactic symmetry requires that intervocalic *-sp- be analysed as /sp/ rather than /sb/.
 
Piotr
 
 
----- Original Message -----
From: Miguel Carrasquer Vidal
To: cybalist@yahoogroups.com
Sent: Sunday, December 16, 2001 4:02 AM
Subject: Re: [tied] Rune-essay Mads Peder Nordbo

The relevance of Grimm's (and Verner's) law is that it turns the PIE
*b hole (near absence of */b/ in Proto-Indo-European), into a
Proto-Germanic *p hole (near absence of */p/: PIE *p > PGmc. *f/*b,
PIE *bh > PGmc. *b, and no */b/ to turn into */p/.  The sequence *sp,
which is usually regarded as unaffected by Grimm's law, is in fact
better analyzed as Proto-Germanic */sb/).