On prana.s.ta versus pra.na.s.ta

[Madhav Deshpande]

Sorry, my statement about the Paninian rules needs a correction.  P.8.4.36
(naze.h .zaantasya) prohibits the change of -n- to -.na- when the root
ends in -.s-, as in the form -prana.s.ta-.  Thus, Panini allows
pra.nazyati, but not pra.na.s.ta.  I remembered this exception right after
hitting the 'send' command on my previous message.  Sorry for the
confusion.
                        Madhav Deshpande

On Sat, 27 Feb 1999, Madhav Deshpande wrote:

> Hello,
>
>         Belvalkar himself seems to have changed his mind between the
> critical edition and the Banaras edition.  In BG verse 18.72, the critical
> edition gives the reading -prana.s.ta-, while noting the reading
> -pra.na.s.ta- as coming from some unspecified mss.  The Banaras edition
> onthe other hand gives the reading -pra.na.s.ta-, without noting the other
> reading.
>         A question was raised as to what would be sanctioned by the
> grammarians.  Panini lists this root as -.naza- in his Dhaatupaa.tha.
> Then rule P.8.4.14 (upasargaad asamaase 'pi .nopadezasya) says that the
> -n- of a root, which is listed with -.n- in the Dhatupaa.tha, changes to
> -.n- after the r/.r/.s in the Upasarga, even when there is no compounding
> (asamaase 'pi).  The last condition would then allow forms such as
> -pra.na.s.ta- (where there is traditionally said to be compounding between
> pra and na.s.ta) and pra.nazyati (where there is traditionally not
> supposed to be any compounding).  The form -pra.nazyati- occurs in BG
> 2.63, and neither the critical edition, nor Belvalkar's Banaras edition
> cite the variant -pranazyati-.  Thus, from a purely theoretical point of
> view of Paninian grammar, the form would be -pra.na.s.ta-.  This is,
> however, not to say anything about the forms of the Ur-Mbh.
>         Best,
>                                         Madhav Deshpande
>