anusvara, anunAsika and chandrabindu,bindu and nasal mutes

[Charles Wikner]

On Sun, 28 Feb 1999, Harry Spier <harryspier at HOTMAIL.COM> wrote:

> To sum up can anyone clarify the relationship between anusvAra,
> anunAsika, nasal mutes and the signs bindu and chandrabindu?

Strictly speaking the terms anusvAra and anunAsika refer to the
_spoken_ sound and not the _written_ form (as do Panini's rules).

anunAsika: This refers to the five nasal stops, as well as any
=========  vowel or three of the semi-vowels (y,l,v) when they
           are pronounced through nose and mouth together (at
           the same time) (P.1.1.8)

anusvAra:  refers to the nasal _after_-sound appended to a vowel,
========   and normally followed by a consonant.  There are two
           types of anusvAra:
           (a) "true" anusvAra before a sibilant, h, or r (which
               do not have equivalent nasal sounds).  This is
               generally pronounced as G-like in northern India
               and m-like further south.
           (b) "substitute" anusvAra before a stop or before the
               three semivowels y, l, and v.  These may be sounded
               as the class nasal before a stop, or as a nasalized
               semivowel before a semivowel.  (In English the "n"
               functions as an anusvAra when similarly sandwiched
               between vowel and consonant, e.g. wink, winch, wind.)

In transliteration, Whitney and Monier-Willams distinguish between
the true and substitute anusvAra-s: I see no benefit in doing so,
nor is it current practice.

The anusvAra mainly arises as a substitute for a pada-final "m"
before a following consonant (P.8.3.23).  Note that an upasarga
is also considered to be a pada, thus sam+dhi --> saMdhi.

bindu: the dot above the line indicates either type of anusvAra.
=====

candrabindu: (a) above the line (where a bindu would be positioned)
===========      indicates an anunAsika vowel.
             (b) above the line, over a semivowel (y,l,v), indicates
                 an anunAsika semivowel.
             (c) to the right of a syllable and with a virAma below
                 it, is used in some texts to indicate the "true"
                 anusvAra (i.e before sibilant, h, r) and may be
                 pronounced "gna".  (There are also other forms
                 that indicate this function.)

> Is it just as correct to write the class nasal for "m" before mutes
> or nasals as it is to use the anusvAra sign the bindu. Is this merely
> a matter of presentation, equivalent and equally correct?

No.  The replacement of the substitute anusvAra is optional at the
end of a pada (P.8.4.58-59).  To cater for the variety of traditions
in pronunciation, this anusvAra should remain as such in the written
form in both devanAgarI and transliteration.  E.g. Monier-Williams
dictionary makes the substitute (in sound) before a stop but not
before a semivowel, and thus his dictionary order for words beginning
with "sam" (e.g. saMkalpa is found where you would expect saGkalpa,
and after saMyoga which he treats as a true anusvAra in sound).
(Interestingly, M-W never indicates an anusvAra before an oSThya
stop, e.g. sambanda never saMbanda.  Anyone know why?)

Retaining the notation for the substitute anusvAra also distinguishes
between saMga (sam+gam, company) and saGga (<-- saJj, attachment).
The former is not used in the gItA, but the latter is, e.g. 2:62.
(Aside: Bhaja Govindam 9 (Chinmaya edition) begins satsaGngatve
nissaGgatvaM... and then gives the padAni with anusvAra for both G's.
Should that be satsaMgatve nisaGgatvaM... ?)

However, you will find people wrongly using the substitute anusvAra
_within_ a pada (as short-hand, due to font limitations, to improve
legibility), e.g. *aMgya, *aMktvA -- in these examples rather use
halanta-G (yech!) in print.  That said, don't be confused when the
anusvAra is used in compounds: ahaMkRtaH (BG 18:27) is correct.

I hope that more knowledgeable grammarians will confirm or correct
this understanding.

Regards, Charles.