SankarAcArya's bhagavad gItA bhAshya: 2. 19-20

Vidyasankar Sundaresan vsundaresan at HOTMAIL.COM
Tue Jun 8 03:44:27 CDT 1999

My earlier response to this seems to have vanished into cyber-limbo for the
time being, so I apologize if two copies of this appear on this list.

Charles Wikner <WIKNER at NACDH4.NAC.AC.ZA> wrote:

>A few points on the second pAda:
>(1) While Warrier has an avagraha negating bhavitA, four other
>    editions to hand, as well as Jacob's Concordance, show a
>    space (i.e. bhavitA without the negation).  Which form is
>    generally accepted as correct?

The practice of putting avagraha in the sandhi between bhUtvA and abhavitA
seems to be a modern one, a sign of how academic transliteration into Roman
script has affected nAgarI script in reverse. My impression is that
manuscripts generally use avagraha only to show the elided akAra after the
visarga. In the absence of this avagraha, both readings would be possible,
depending upon how one splits the words in the verse.

According to SankarAcArya's commentary, the reading should be abhavitA,
which is the one I've adopted. He explains bhUtvA = bhavanakriyAM anubhUya,
paScAd abhavitA = abhAvaM gantA. This reading also seems to make more sense
to me, while the word bhavanakriyA prompted the translation "come into
being." And the words "vA na" capture the meaning of abhUtvA bhavitA, in
addition to bhUtvA abhavitA.

If the reading were bhavitA, instead of abhavitA, one has to split this pada
as follows - bhUtvA na bhavitA vA bhUyaH. This seems very problematic to me,
as it would imply "or, having once been, it does not come into being
again." The "or" would indicate an alternative to the statement in the first
pada of the verse. This rules out rebirth, but leaves room for the first
"birth," whereas the point is that the Atman is never born.

>(2) I usually consider dhAtu /bhU as punctual and dhAtu /as as
>    durative/stative (that's personal, and may be wrong!), so I
>    have difficulty associating /bhU with the Immutable.  Since

I usually do the same too. Within the spectrum of being and becoming, one
has to choose!

>    /as does not have all lakAra and /bhU fills in the gaps, can
>    one treat bhUtvA and bhavitA as derived from /as rather than
>    /bhU, to replace "come into being" (which jars) with "been"?
>    (e.g....having [always] existed, It will not cease to exist.)

I would have done so too, except that the text and the commentary seem to
approach this from a slightly different angle. This ties in with the comment
on "vA na" below. I must also draw attention to verse 2. 26 (... enaM
nityajAtaM nityaM vA manyase mRtam, etc.). It seems to have gone generally
unnoticed that after spending so much effort on describing the Atman as
unborn, immutable etc., the gItA still shows Krishna as leaving room for
Arjuna to think otherwise.

Therefore, the only way to interpret 2. 20, is to take it that the popular,
uninvestigated notions of "birth" and "death" are assumed and that the Atman
is being described accordingly. The general confusion about it all arises
from mistaking the Atman to be dead when the body dies, or born when a body
is born. Thus, if one were to assume that the Atman is born, i.e. comes into
being, apparently from non-being, then it is denied that the Atman ever goes
back to the state of non-being. Conversely, if one considers the Atman to
die, i.e. to pass into non-being, from a state of being, then it is denied
that the Atman will ever come back to a state of being. {This interpretation
also properly takes into account the word "again" (bhUyaH) in the verse. If
there were no assumption that the Atman once was not, but now is (bhUtvA),
the word bhUyaH is not justified.} Taken together, the listener is forced to
see that there are only two possibilities - either the Atman always is or
always is not.

Now, the commentary on 2. 18 has already pointed out that the Atman is self-
established (svataH siddha), without granting which there can be no talk of
anything. Therefore, the Atman cannot possibly be always non-existent, which
leads to the descriptions aja, nitya etc. Thus, the argument attacks the
assumption that something could die or be born and still be considered
"existent." And this goes back to the argument about sat (real) and asat
(unreal) in verse 2. 16, so that all that is seemingly born and seemingly
dies is shown to be ultimately unreal. This seems to be a somewhat
roundabout way of going about it, but the purpose is to point out that all
becoming needs being as the basic ground. Although this argument about a
substratum (adhishThAna) is not explicitly made here, it seems to be
implicit in the structure of the argument. I haven't still stopped
marvelling at the wonderful word-play in this verse, and the meaning that
the commentator has extracted out of it - AScaryavat paSyAmi.

>(3) My understanding of Sankara's comment on "vA na" is that the
>    converse also holds true, i.e. having [never] existed, it
>    will not come into existence.  This seems to echo verse 16.
>    (Possibly translate "vA na" as "and vice versa".)

Great suggestion. Now, why didn't I think of that?!! A. Mahadeva Sastri
translates as "nor the reverse," but "vice versa" does sound better.

>This verse (and a few others in the gItA) are in anuSTubh meter:
>is there any significance or explanation for this?

Isn't this meter trishTubh?

In this instance, the meter change could simply be due to the quotation of
part of the kaTha text. The other notable instance where the change occurs
is in the 11th chapter, describing the viSvarUpa darSana. The Sloka meter is
resumed only after Krishna resumes the human form that is familiar to
Arjuna. The poetic effect of changing the meter is to convey Arjuna's sense
of grandeur, tinged with fright. The changes of meter do mark off the
moments of critical transition and the effect is something to be heard when
recited out loud.

>I am thoroughly enjoying this series of posts - thanks for sharing them!

Thanks! I look forward to more of your comments in the future, as I know
they are based upon a keen understanding of Sanskrit grammar. The
translation could only benefit by them. It is certainly a daunting task, and
I hope my effort does not hinder understanding, even if it does not further


More information about the Advaita-l mailing list