Indian perfume set.

[Koontz John E]

On Mon, 6 Oct 2003, Rory M Larson wrote:
> I should add that this fact makes it quite possible
> that the IOM /pyubraN/ - OP /nubdhaN/ set is indeed
> from proto-MVS as John originally suggested, rather
> than a later garbled loan from IO to OP as I
> hypothesized yesterday.

I'm a little bothered by the chronology, I have to admit.

ProtoMVS  hpi opraN
          (h)pyopraN

PostMVS   pyobraN         RobraN

Modern    (?) pyubraN     nubdhaN

The OP works out, assuming I've been right all along (more or less
following in the footsteps of the CSD editors) in asserting that *py > *pr
> *R, the last in word initial position only, in Dhegiha.  The
developments *R > n and *o > u are across the board in OP.  But in IO, as
I undestand it, *o > u is initial only.  I had thought it was perhaps
something more like "in pre-stem inflectional syllables,"  which would
work, but if it's strictly initial, then this wouldn't work unless the
sequence is new in IO or has remained essentially analyzable in parts
since ProtoMVS.

Of course, what I was actually assuming here was a certain degree of
fuzziness or reanalysis in the IO transcription.  I know the sources here
are quite varied and not all entirely reliable.  So maybe it's really
pronounced pyobraN?  Or may be in the context u and o don't contrast?

Another alternative is that the OP form represents a really old instance
of *hp(i)=opraN, where as the IO one is fairly recent, but accidentally
parallel.  It does seem to me that the IO analysis is a very plausible
explanation of the OP form.

As far as the nu '(wild) potato' analysis of nu in OP nubdhaN, I rather
doubt that.  Mint does spread by rhyzomes, very agressive rhyzomes,
somewhat fleshy, but thin, but it's not usually the rhyzomes that interest
someone who encounters it, unless they happen to plant it in their garden.
Note that nu is from *Ro < *pro, another case of *pr > *r, though without
preceding *py.

JEK