Me:
> No. The shortening of /o(w)e~/ to *o: wouldn't
> cause circumflex.
Miguel argues:
> Yes it would.
Me again:
> The cases of circumflex involve an intermediary *h1
Miguel stubbornly argues some more:
> No. There is no need for *h1 at all.
Apparently you aren't paying attention. According to
my viewpoint, *h1 _is_ necessary to disambiguate
cases of circumflex from those that don't cause
circumflex. It also preserves the rules of
phonotactics that disallow the sequence of two vowels
one after the other in IE.
Honestly, languages that allow two vowels one after
the other are far rarer than those that disallow this.
So... Occam's Razor is on my side.
Nothing you've said so far proves against the presence
of *h1 here. There is also nothing showing that a
sequence of VV was ever possible in IE.
The onus is on you to prove your point of view.
> And there is no laryngeal in *h2akmõ, or i-stem
> loc.sg. *-e~i, etc.
What the hell is *h2akmõ? It's not IE. At any
rate, I don't know why you think there should be
an *h1 here anyways, based on what I've said. How
does any of this relate?
> The circumflex is the result of vowel contraction,
> or loss of a resonant.
Yes, duh. This is because IE *Vh1V > post-IE *VV (with
circumflex).
Most other people can see how pre-IE *-omi >
/-o(w)e~/ > IE *o: does not conflict with the above
at all.
So clearly, no, the circumflex is not a necessary
byproduct of the change theorized for the 1ps ending
and doesn't conflict with anything else.
= gLeN
______________________________________________________
Click here to donate to the Hurricane Katrina relief effort.
http://store.yahoo.com/redcross-donate3/