PIOTR WROTE:
<<The initial *gW- was still pronounced as a labiovelar in early historical
times but became /b/ in Classical Greek. In pre-Classical times the
(Proto-)Greeks had *gWigWro:sko:, not <bibro:sko:>.>>
I REPLIED:
<<Can you tell me precisely when the first Greek started using <bibro:sko>
instead of *gWigWro:sko:. For all you know it was 1000 years before Homer.
So this argument is empty. If you had "direct" evidence of when Greeks truly
last used *gW-, you wouldn't need to put an asterisk by it.>>
PIOTR REPLIED:
<<I put the asterisk there because that particular word is not attested early
enough. Actually, the tranformation of labiovelars in Greek can be dated much
more precisely than you suggest. /gW/ and /kW/ existed as phonemes distinct
from /b/ and /p/ and never confused with them at the time of the Linear B
inscriptions (mid to late 13th century BC).>>
First, just for the record, *gWigWro:sko: -- in that form -- is unattested
ever. Am I right?
Second, nothing you have said tells me how you know when "the first Greek
started using <bibro:sko> instead of *gWigWro:sko:." So that there is no
definite time line by which you can eliminate <bibro-> as the source of the
trade word for beaver.
To say that *gW- is found in Mycenaean really tells us nothing in that regard.
Mycenaean is NOT proto-Greek. By most conventional accounts, it is a
separate, "sister" branch and not ancestor to any of the branches of 1st
millenium Greek that yield all the texts you mention (Attic, Aeolian, Ionian,
Doric). So that the appearance of -gW in Mycenaean does NOT tell us whether
those other languages did not already have <b-> much earlier. Or when
exactly they had it.
When it was figured out that Mycenaean was not proto-Greek, the -gW in
Mycenaean created more problems than it solved for "regular development."
That's because it meant that -gW > b could not have occurred in the ancestor
proto-Greek and been passed on to the daughters by descent, but had to have
been adopted separately in the various different branches. How is it that
different Greek languages went in the direction gW > b on their own, when
there were a lot of other (perhaps less surprising) ways that other,
non-Greek languages went with gW- >? And I'm not aware of any direct
evidence of how and -- more importantly -- when this happened in those other
Greek languages.
For the purpose of the beaver, there's no way to know that early Attic or
other Greeks were not already using b- in words like basileus and bibro:-,
even before the Mycenaeans were writing in Linear B script. Attic/Ionic or
other Greek speakers may have innovated it centuries before it spread to
Mycenaean, if it ever did. And therefore -- yes -- <bibro-> equals beaver
could have been in effect a thousand years before <kastor> is first attested
as meaning beaver. (And once again -- all convention says that <kastor> is a
substituted word in Greek that replaced an earlier word -- whether that
earlier word was from PIE or a Greek innovation.)
PIOTR ALSO WRITES:
<<For example, Classical <basileus> was spelt <qa-si-re-u> /gWasileus/. The
change took place between that time and Homer's, not a thousand years before
him. The very end of the second [millenium] BC of the very beginning of the
first.>>>>
Homer did not write in Mycenaean. The change to <basileus> in the
language(s) in which Homer wrote could have happened even before Mycenaean
appeared in writing. I'm not even aware of any evidence that the change EVER
occurred in Mycenaean.
PIOTR ALSO WRITES:
<<In Slavic, Baltic, Iranian and Celtic, PIE *bH > b in both positions; the
further development of medial *-b- in Avestan is also regular.>>
I REPLIED:
> Of course, b > b and b_b > b_b are simpler explanation. Or to quote you,
"You posit arbitrary irregularity where the assumption of regularity works
better." On that basis we have to reject the unlikely bh > b in the case of
the beaver.>>
PIOTR REPLIED:
<<Steve, *bH > *b is not arbitrary irregularity but the _regular_ development
of *bH in those groups. >>
By what peculiar logic is *bh > b "more regular" than b > b? My point was
that you too are positing something that is not the simplest explanation.
There is nothing more regular than b > b, because it is no change at all.
Compared to b > b, any change is both less regular and more arbitrary.
PIOTR WROTE:
<<Actually, Balto-Slavic distinguished between medial *bH and *b in a subtle
way: the preceding vowel was lengthened before unaspirated *b, but not before
*bH. there is no such lengthening [before the medial stop in] the 'beaver'
word.>>
But of course the question proper to addressing the hypothesis becomes how
did it work not in native but borrowed words? And did the treatment of
borrowed words with <b> or <bh> change with time and with the proximate --
rather than the original -- source of the word?
Steve
PS - BTW - I think I saw in Ventris - Chadwick "basileus" is <ga-si-re-u>.
And I think among the dissenters on Aegeanet it is <pa-si-re-u>.