[tied] Re: A New language tree

--- In cybalist@yahoogroups.com, Piotr Gasiorowski <gpiotr@...> wrote:

> mkelkar2003 wrote:
>
> > Just liket that graded ablaut stuff, the actual 5 vowels are not
> > observed in any langauge.
>
> If you mean that /a/ /e/, /o/, /e:/ and /o:/ don't appear as different
> vowel phonemes in any IE language, Greek and Latin, for example,
> immediately falsify your claim. Both retain the original vowel system
> very faithfully.

That is the problem. The "orginal" system has been projected back in
time by looking at what exists in reality. According to Richard
Worthington:

"A problem with reconstruction is that shared changes are very easily
projected back."

In other words changes that are NOT shared are a matter faith. One has
to assume that the original language had them (archaisms?) that were
preserved in one brach but oh so conveniently were lost in the other.
Or they are "later" innovations. But there is a thrid possibility.
Those dialects do not genetically belong to a family at all. The
Nichols/Garret wave model is in action.

PIE *a, *e and *o fell together as *a in

> Proto-Indo-Iranian, but traces of the old distinctions are observable

Or what was together elready expanded later as Misra has shown in the
case of the Gypsy language. The Sanskrit system could be and inmho IS
the origina.

M. kelkar

> there, as you may have learnt on many previous occasions if you had

only

> bothered to follow the discussions in which you participated.
>
> Piotr