[tied] Re: Sanskrit Rta... and related terms

--- david_russell_watson <liberty@...> wrote:

> --- In cybalist@yahoogroups.com, "stlatos"
> <stlatos@...> wrote:

> > all dental > retro after r. (dialect)
>
> Don't you count far far more instances of 'rt',

'rth', 'rd',

> or 'rdh' than 'rt.', 'rt.h', 'rd.', or 'rd.h'?

That's not the point. I said "dialect" not to
describe classical texts but the newer languages with
kat.a- from krta-, for example.

> Actually, can
> you name even one word in Sanskrit with one of the

latter

> clusters? I couldn't recall a single one myself and

so checked

> at http://webapps.uni-koeln.de/tamil/.
> I thought that there would be at least a few, out of

Proto-

> Indo-Iranian *-ržd(h)- at least, but could find not

a one.

What about mr.d.- "be gracious, pardon"; mr.d.i:ka-
"favor" from mld-dH3- "give softness/allowance/mercy"?

> > nasal drops as above unless following C has same

place of

> > articulation
>
> But from P.I.E. onward, if not before, a nasal

assimilated

> the position of the following stop, did it not?

Not in my rules; at morpheme boundaries only n
assimilates.

> Is this why
> you posit an otherwise unknown palatalized 'n': so

that you

> can get rid of it later with this rule?

No; there were other reasons (some already shown).
This is not the only word containing n^; it's not
gotten rid of by this rule.

> Why not just leave
> it out from the beginning?

The r wouldn't be long (also see previous).

> Not all of your rules are incorrect, but even when

not they

> have exceptions and ordering problems of which you

seem

> unaware.

I didn't write all my rules in that. I just put in
the relevant ones for my examples in that discussion.
In my derivations I've seen no problem in internal
consistency.

I would suggest you familiarize yourself with the

> mainstream theory, and then try to lend to the

improvement

> of that, where possible and desirable, rather than

try to

> reinvent the wheel, as that theory explains far

more, has

> far fewer counterexamples against it, and and

involves fewer

> optional rules than does yours. :^)

What counterexamples? I don't think there's
anything wrong with optional rules when there's lots
of free variation in and among IA languages.






__________________________________
Yahoo! Mail - PC Magazine Editors' Choice 2005
http://mail.yahoo.com