[tied] Re: Germanic Scythians?

--- In cybalist@yahoogroups.com, "Glen Gordon" <glengordon01@...>
wrote:

>
> Piotr to Torsten:
> >I'm strongly against such paper linguistics (quite apart from the

fact that

> >I reject the glottalic solution except as a possible model of pre-

PIE

> >phonology).
>
> I wonder though if we could get away with saying that *k/*g/*gH,
> originally being largely pronounced as *[kH]/*[k]/*[g] in Mid IE,
> became either *[k]/*[kg]/*[g] (as in pre-Hellenic/pre-IndoIranian)
> or the preserved *[kH]/*[k]/*[g] (as in pre-Germanic). In other
> words, we might think of multiple phonological systems at once
> for IE, dependent on dialect area.
>
> At least that way, we wouldn't have the need to theorize a *kH to
> even out the lop-sided system since *gH would be an areal
> innovation in a post-IE stage involving dialects like Hellenic and
> IndoIranian.
>
> Just a thought, is all.
>
>
> - gLeN
>

An even looser thought:
What do you do when history or speech habit changes dumps a voiced
stop in your system? You have a nice old [k]/[g] system, with an
extra stop, [g'], glottalised or whatever. Suddenly the extra stop
loses its distinguishing feature and starts crowding your [k]/[g]
space. Solution: exaggerate the k-ness of your [k], making it [kh],
so you end up with

[kh], [k], [g]

or exaggerate the g-ness of your [g], making it [gh], so you end up
with

[k], [g], [gh]

(And, since the latter is lopsided, it decays quickly (with a half-
life of ?) into ... etc)

Torsten