The Germanic substrate - knives?

[Rick Mc Callister]

	You make a very interesting point regarding <knife>
	I think of knife in the sense of "cut, tear, strike, kill" but I
can see how "flesher, scraper" can work

	Watkins (1985: 19) lists knife under *gen- "hypothetical
Indo-European base of a range of Germanic words referring to compact,
knobby bodies and projections, sharp blows". 11 derivative forms are
offered.
	He cites Pokorny *gen. Unfortunately, I don't have access to
Pokorny and don't know if he also included knife.
	I found Watkins's wording as an expression of doubt.
	I also realize that Watkins only includes IE roots found in English
	BTW: does anyone know roughly what percentage of all IE roots these
include?

	Looking through *KVn- in Watkins, the closest other root to knife
is *ghwen- "to strike, kill [1985: 25]
	Do your sources offer any IE roots for the words you cite?

	Initial /*gwh-/ would have yielded Germanic /b, g/ --IF followed by
a vowel. Someone more knowledgeable will have to fill us in on what
/*ghwn-V-/ would have produced.
	Regarding initial (non-clustered) consonants, following Watkins,
	Greek initial (non-clustered) /kh-/ (also /th-/) corresponds to
Lithuanian /g/ < IE *ghw-.
	Lithuanian initial (non-clustered) /g-/ can also come from IE *gw-;
which gives Greek /b, d, g/
	Greek initial (non-clustered) /k-/ can come from IE *k (Lithuanian
/s-/) & *kw (Lithuanian /k, c, c^/)
	Again, I don't know what *Kn- would yield in Greek & Lithuanian

	The closest roots for "scraping" in Watkins are *gerbh-, *gher-,
*ghrei-, *ghreu-, *grat-. I've seen instances of /kn-/ > /kr-/ but never
the other way. BUT I'm not a linguist, so maybe someone with more
experience can fill me in.

	Keep in mind that the list is essentially a compilation of claims
(and possibilities) rather than my own work. I agree that a fair percentage
of the items in the list are probably from IE and I'm convinced that a
small percentage definitely are.

>Rick Mc Callister, on his very informative web site on the Germanic substrate
>(http://www.muw.edu/~rmccalli/subsGerIntro.html), mentions the reasons that
>words are included in this category:

>"....most evidence suggesting that a given word is non-Indo-European is
>essentially negative :
>*   the word does not exist in other branches of Indo-European
>*   its phonology does not conform to the norms of Indo-European
>*   the evolution of the word does not correspond to known phonological laws
>of the branch in question"

[ moderator snip ]

Rick Mc Callister
W-1634
Mississippi University for Women
Columbus MS 39701