I noticed a discussion in the Yahoo Group archives several months back
regarding the nasal niggahiita and the wandering dot.
From my arm-chair research, the niggahiita seems weakly defined as
muddhaja, cerebral, retroflex. If this is the case, it seems logical
that the m would have a dot below (.m) rather than above (^m) in line
with other cerebrals (.d .t .n .l).
Perhaps someone can be more definitive.
Devanagari has two nasal markers. As I understand it, a dot above
(anusvara) indicates a nasal consonant (hense ^n). However, a cresent
with a dot above (anunasika or candrabindu [moon dot]) indicates a
nasal vowel. Confusingly enough, when a devanagari character already
has a diacritic above, a single dot is often used simply to minimize
clutter.
The brahmi script (according to the only chart recycled all over the
internet) and some southeast asian scripts (such as Thai and Lao) have
a dot (or small circle) above which produces something like (if not
precisely) a niggahita. For example: daM (black) gaM (hold).
Unicode provides us with many diacritics to play with (breve, tildes,
dots above and below). The a,A even has a combination breve (cresent)
above and a dot below, but this is not available with most other vowels.
However, instead of an M with a dot above or below, why do we use an M
at all? It is never by itself. It can not be pronounced without a
vowel, and confuses first-time readers. Perhaps, it is more logical to
drop the M and place a dot below the vowel (.a .i .u). Conveniently,
these vowels are all found in Unicode Latin Extended Additional range
U+1E00 - U+1EFF.
ah.a keva.t.tagaamasm.i ah.u keva.t.tadaarako.
vs.
aha.m keva.t.tagaamasmi.m ahu.m keva.t.tadaarako.