> ==============
> PIE *gw can only be PAA *q
>
> Whatever that hatchek root really is.
> I suppose it should better be read *ts
> What are the PAA data supporting #516 ?
>
> Arnaud
> ================
I have written this a couple of times but for your benefit I will again
provide the link:
http://geocities.com/proto-language/c-AFRASIAN-3_table.htm
P.R
==============
It's quite obvious I disagree with your phonetic fairytales.
I won't benefit from this link.
Arnaud
================
> PIE *gwem does not exist.
> the root is *gw
> with two extensions : gw_H2 and gw_m.
> And Meillet does not reject gw_n- for LAtin.
> A third one.
>
> Arnaud
==============
That is where you are wrong.
============
I disagree with you.
I could say the same.
Arnaud
================
No root in PIE consists of a single consonant; the very shortest are *CV,
e.g. *me.
============
I disagree.
there is no *me "root".
*me and (He)gh(-o) are the same root.
me stems out of ngga?u-aH1
where instrumental -aH1 attracted stress
hence caused a syllabic implosion
and nggw-eH1 > me(:).
(He)-gho stems out of ngga?u
Arnaud
===========
If you want to write *gW_H2, be my guest but no PIE root has no *V.
P.R
===========
It's no use writing -e-.
Only the skeletton of consonants makes sense.
C_C is the same thing as CeC.
Try to think half a millimeter above the surface of things.
Arnaud
=================
In this case, the PPIE root was *gWa- to which *H (*?) was added,
lengthening it to *a: and preserving its central quality.
Without lengthening, *gWa- became *gWé- when combined with other formants:
*gWéw-, *gWéy-.
What is your basis for saying *gWem- does not exist?
=============
The root is *gw_
which can be incremented into : *gw_H2 ; *gw_m ; *gw_n
Arnaud
============
Actually, I think it likely that it should be emended to *gWyém-, an
extension of *gWéy- rather than *gWé-.
But it exists in Arabic as <Sym> and in Egyptian as <S[j]m>.
Patrick
=========
I not only disagree
but this time
I think you are hopelessly *wrong*.
Arnaud
===============