Regarding the kakali nishada in Begada. The varnam Inta chalamu is notated
with a dns ris in the anupallavi. Veena Dhanammal plays it with a kaisiki
nishada, not kakali. But I've heard another paThAnthara with pdp sns,
which also utilizes kaisiki nishada. At least to my ears, the second one
sounds closer to how Begada "should" sound. Furthermore, the dhaivata in
Begada is rarely handled as a straight note; even in phrases like ni da pa
(e.g. Nee pada pankaja), the da is higher in its value than the straight
note. Thus, even in the first paThAnthara, nns ris is how it really
sounds. I would venture to say that "dns" as written down is more a
limitation of the notation scheme used, than a real problem in the raga
itself. Note that this is not an appeal for standardization, even in the
written notation! :-). Rather, musicians should follow their musical
sensibilities about what sounds good here, rather than be worried too much
about the dns notation. Whether something sounds good or not, is of course
a subjective opinion.
There are many phrases in Begada where the arohana krama of pdps is
violated. Like - gmpd nsdp mpgr sa, pdpd rsni da pa, dnrs ni dp, etc. In
all these cases, the ni is kaisiki. The kakali nishada, when it is used,
is very close to the shadja. I guess you could call it the chyuta shadja
nishada i.e. the nishada "touching" the shadja. In many places, it sounds
almost like two sa's in succession rather than like sa and ni. But I would
still call it a nishada if for nothing else but the sake of consistency. A
similar usage (janta sa, the second sa sounding almost like a ni) in
Mohanam is frowned down upon, and is usually called an admixture of
Kalyani! Though of course Maharajapuram Viswanatha Iyer got away with it.
:-).
S. Vidyasankar