m/anusvAra & s/visarga

[Jacob Baltuch]

Dominique Thillaud wrote:

>>        I'm perhaps stupid but 'h' is a standard notation of breath and it
>>seems to me the real h and the aspirated occlusives are always followed by
>>a vowel, and the visarga never. Hence, why don't use simply 'h' for the
>>visarga ?
>>
>>        IF nextCh IS space OR IN consonants THEN visarga ELSE h END
>
>Hmm, excellent point, thanks.
>
>To be pedantic it is not strictly, stricly true (cf gRhNAti, azvAbhyAm,
>where h is followed by N or bh by y) but it looks like your point still
>stands that I will always be able to restaure visarga from the context.
>The algorithm will be a bit more complicated though.

I think it would work like this:

 'If the previous character is not k, g, T, D, t, d, p, b and if the next
  character is k, kh, p, ph or space then visarga else h'

I checked a standard list of conjuncts (only an indication and not
necessarily a full proof, maybe) and the only ones starting with h are:
hNa, hna, hma, hya, hra, hla, hva, and furthermore h cannot appear
at the end of a word (except if that word is a stem used metalinguis-
tically, e.g. if someone writes _about_ a stem such as 'madhulih' for
example; but I won't worry about that since if it is clear that one's
talking about a stem, it's got to be h, no such thing as a stem in H).

Furthermore visarga inside a word can only appear before k, kh, p, ph.

In other words, as you were implying, h is h and H is H and never the
twain shall meet. Which means I can use h for both.

Unless of course, for some perverse reason, there was a word out there
with a hk, hkh, hp or hph sequence hiding inside!

Despite the absence of conjuncts one can always use virAma for evil
things like this, after all, if one really wants to, but let's not get
paranoid... I hope there isn't, right?

Otherwise, for visarga I should be all set. Thanks. Good suggestion.