Re: [tied] Grimm and Verner

Hi, Anders,

Be careful when you draw information from web pages whose authors' credentials are unknown. If you need a reliable description of Grimm's and Verner's Laws, it is more prudent to consult a handbook. The page you refer contains an oddly confused misrepresentation of both celebrated laws; the author evidently has little idea of what really happened in Proto-Germanic. Here's what happened (apart from very minor complications):

(1) GRIMM'S LAW

(1a) PIE *bH, *dH, *g(^)H, *gWH > PGmc. *v, *ð, *G, *Gw (voiced breathy stops > voiced fricatives)

(1b) PIE *b, *d, *g(^), *gW > PGmc. *p, *t, *k, *kw (voiced stops > voiceless stops)

(1c) PIE *p, *t, *k(^), *kW > PGmc. *f, *þ, *x, *xw (voiceless stops > voiceless fricatives)

NOTES:

(i) *v, *ð, *G were primarily frictive, but developed stop allophones ([b, d, g]) in some positions, e.g. after nasals; many linguists, including Yours Truly, prefer to symbolise these PGmc. phonemes as *b, *d, *g for the sake of orthographic simplicity, which is OK as long as you remember that they were not _always_ realised as stops, e.g. phonemic /d/ = phonetic [ð/d]. This kind of voiced stop/voiced fricative allophony is found in many languages, e.g. Spanish.

(ii) Dorsal *x tended to be weakened to [h] in syllable onsets, hence the variant symbol *h for the same phoneme.

(iii) Grimm's Law did not affect the second element of voiceless obstruent clusters:

PIE *sp, *st, *sk, *skW > PGmc. *sp, *st, *sk, *skw

PIE *pt, *kt > PGmc. *ft, *xt

(2) VERNER'S LAW

A voiceless fricative (*f, *þ, *x or *s) remains voiceless if it is initial or immediately preceded by a stressed vowel, or if it forms a cluster with another voiceless obstruent (e.g. *xs, *fs, *ft); otherwise it becomes voiced:

*f, *þ, *x, *s > *v, *ð, *G, *z (or, as some prefer to write them, *b, *d, *g, *z) after an unstressed syllable if they are flanked by voiced segments or word-final.

NOTE: In North and West Germanic dialects a later process known as _rhotacism_ led to the merger of *z and inherited *r as /r/. Rhotacism applied also in environments like PIE *-sdH- [-zdH-] > PGmc. *-zð- > N & W Gmc. -rd-. There was no rhotacism in Gothic, which preserved /z/ and /r/ as distinct phonemes; /z/ was not yet rhotacised in Old Runic either.

It is not uncommon for writing systems to use the same spelling for voiced/voiceless pairs (Modern English does not distinguish the two "th" sounds and usually has the spelling <s> for /z/; the ancient Romans had <C> for both /k/ and /g/). Proto-Germanic *p was a rare sound; its frequency rose in the historically attested languages mainly thanks to borrowing. In some early versions of the Runic script a special p-rune may have been found superfluous.

Piotr

Piotr

----- Original Message -----
From: malmqvist52

To: cybalist@yahoogroups.com

Sent: Saturday, December 15, 2001 6:01 PM

Subject: [tied] Rune-essay Mads Peder Nordbo

... I think I've developed a hypothesis how this would fit with the p/f/b- part of Grimms and Verner's law's For simplicity I copied a presetation of Verner's law from a web page: http://www.cord.edu/faculty/sprunger/e315/grimm.htm *****' Verner's Law Karl Verner, 1875 Danish linguist wondered why not every I-E stop changed in the same way. His formulation established that Grimm's Law was consistent and could account for all known cognate evolution Intermediate step in Stage 1 shift: (1)All voiceless stops changed once: ph --> f th --> theta kh --> h sh --> s z (2)If the sound was in an initial position or immediately after a stressed verb, it changed no further. (3)Those in other positions changed to voiced spirants (b, d, g)

...