PIE *gn- - know/ken

[Glen Gordon]

MIGUEL:
 The most acceptable solution from my point of view is that PIE
 did not have any voiced stops at all.

JENS E.R.:
 Not wanting to open the whole can of worms again, let me just ask
 this: Is a change from lenis to voiced stop natural and frequently
 seen?

So far I prefer the "opposite" to Miguel - that *d is really the fortis
stop and the rest are lenis. The fortis would be derivative from the
earlier Pre-IE ejective as Gamr. proposed for Common IE *d in the first
place.

JENS E.R.:
 [...] why derive this from a basically voiceless protoform?

Well, I think Miguel has many reasons for his idea but I'll speak for
myself on why I believe the same thing. The voiceless aspirates *ph,
*kh, *th are not supportable in common IE and seem to be very much
isolated to IIr. Yet, if this is true, we of course run into the problem
of why *bh, *gh and *dh exist without voiceless contrasts (big no-no).
On top of this problem, *b does not exist as well in IE and there has
never been acceptable evidence of this phoneme. These two main things
violate our understanding of how world languages are supposed to
operate.

Hence, since the problems in traditional IE phonology are
voice-oriented, the simpler solution is to accept that IE did not have
voiced stops and then we don't run into problems at all. Germanic then
is seen to hold some phonological archaicisms. Voiceless systems can
often develop voicing contrasts from such a system (Sumerian for one)
and this voicing would have to exist AFTER IE had spread out enough and
the archaic Germanic (with IE *d = *t) had lost some contact. Perhaps
while some IE dialects were still in contact with each other, an
isogloss spread across a certain region, making the voiceless *d a
voiced plain. Note however Greek voiceless th, ph and kh in all
environments, thus we can't say the same thing for *dh.

The voicing of *dh, *gh, and *bh would have been even later than this.
When voicing finally occured, sometimes it merged with the now "voiced
plain". It should come to no surprise that the language that has
voiceless ph, th and kh in contrast with bh, dh, and gh (Sanskrit) is
the same one that DIDN'T merge the phonemes together and maintained the
aspiration. This caused a necessity for the development of voiceless
counterparts. Voiceless th, ph and kh are still found in Sanskrit after
the voiceless mobile *s-. Laryngeals, voiceless as well, apparently
create other examples of these phonemes.

This would appear to be a simpler solution since we finally don't have
to apologize for an imbalanced phonological system in IE or create extra
"band-aid" phonemes that lack proper evidence.

--------------------------------------------
Glen Gordon
glengordon01 at hotmail.com

Kisses and Hugs