--- In
cybalist@yahoogroups.com, "fournet.arnaud"
<fournet.arnaud@...> wrote:
>
> > It isn't. I quote from Johnny Cheung's 'Studies in the Historical
> > Development of the Ossetic Vocalism':
> >
> > "The suffix -tæ is the common plural ending. Plural suff. -tæ
> > is clearly connected with Sogd. (Bud.) -t', -th, -t, (Man.) -t'
> > (after n), Chr. -t', (obl.) -ty, Khz. -c, Yagn. -t, (obl.) -ti,
> > W. ºt (in the pl. suff. -i:s^t). It is also frequently found
> > in North Iranian (i.e. Scythian, Sarmatian, vel sim.) tribal
> > names mentioned in classical sources: Massage-tai, Sauroma-tai,
> > Thussage-tai, Auxa-tai, Zaka-tai, etc.
> >
> > The voiceless nature of the dental in Oss. -tæ poses
> > a problem though.
>
> Not so clear.
It's completely clear, and you edited out the explanation.
The Avestan and Vedic cognates of Ossetic's plural marker
prove that it originated as an independent word or enclitic
in Proto-Indo-Iranian whose reflexes continued to function
as such long afterwards. Even Ossetic betrays the marker's
original independence, having forms in which it can appear
_after_ the case markers instead of before them.
As such, we should expect its initial *t to undergo all of
the same sound changes of any other word-initial *t, which
in Ossetic regularly result in 't'.
> The morpheme is a phonetic problem and is frequent in North
> Iranian, precisely where Uralic is. Uralic morpheme -Ht-
> "plural" clearly has a devoicing feature (clearly exhibited
> in Moksha). You have not proved it's not a LW from Uralic.
I don't have to prove it. A Proto-Indo-Iranian origin for
a North-East Iranian plural suffix is the natural default
until proven otherwise. It's you, rather, who's obligated
to prove it is a loan from Uralic, if you want to convince
others of that.
Why, when so few loans from Uralic appear in Iranic, would
we expect the latter to borrow something so close to a
language's core as its plural marker?
The problem here, as in several of your other weird claims,
is that you're motivated on the basis of solitary similarity
when much more is necessary to build cases for relationship
or loan. If you can't let go of coincidental similarities
and insist on trying to build a novel theory on the basis of
each and every coincidence you come across, then you'll end
up chasing a lot of wild geese.
Please see the article 'How likely are chance resemblances
between languages?' at
http://www.zompist.com/chance.htm .
David