shankara's vivaraNa on vyAsa bhAShya of yoga sUtra-s

Vidyasankar Sundaresan vidya at CCO.CALTECH.EDU
Fri Oct 9 01:00:06 CDT 1998

Ramakrishnan Balasubramanian wrote:

> BTW, I forgot to mention in my previous post. From Sri Subbaramiya's
> quotes it should be clear that Sh thought that yoga was a part of
> sharvaNa, manana abd nididhyAsana. As per advaitins only karma is not
> enjoined in the upanishhads. As per Su yoga seems to follow sannyAsa, at
> least in the normal cases. It was a point of discussion among later
> advaitins whether the triad of shravaNa, manana and nididhyAsana was
> enjoined or merely a suggestion in the upanishhads.

You are right in saying that both camps in this debate quoted Sankara in

However, if one reads Sankara and Suresvara's works without any
preconceptions in mind, what they are saying is a very nuanced view. I'm
afraid this discussion will get too much into technical pUrva mImAMsA, but
inasmuch as it is necessary to understand some of this other school before
one can appreciate the advaita position, here goes.

The one thing to appreciate with any talk of vidhi is that it requires an
agent, who satisfies criteria of eligibility. Another important thing to
remember is that a vidhi can only enjoin actions, it cannot enjoin
knowledge. Finally, the result of such enjoined action has to be something
that comes into existence, it cannot be something that already exists.

On all these counts, vidhi has no place with respect to the Atman. It
always exists, therefore it is not the result of any action that is to be
performed. AtmavidyA is knowledge, not action, therefore it cannot be
enjoined. Finally, there is the question of eligibility and agency. Part
of knowing the true Atman is the knowledge that it is not an agent, and
that it is the Self of all, including oneself. Hence there is no question
of eligibility and when the true Atman is known, there is no agent either.

Readers might recall a question I asked a few weeks ago, which Anand Hudli
helped to clarify, regarding different kinds of vidhis. The reason for my
asking that question was precisely the above issue. Both Sankara and
Suresvara explicitly argue against the enjoining of Self-knowledge, as
briefly discussed above. However, they offer strong arguments agains the
apUrva vidhi and adhikAra vidhi positions, but allow an interpretation of
niyama and parisaMkhyA vidhis. These arguments are to be found in the
commentary on bRhadAraNyaka upanishad, especially at 1. 4. 7, the
corresponding vArttika, naishkarmyasiddhi 1. 45-89 and upadeSasAhasrI 18.
9ff (verse part). The upshot of this is that meditation on the Self can be
seen as a niyama vidhi, because after the rise of jnAna, manana and
nididhyAsana help in countering the force of prArabdha karma. Or it can be
seen as a parisaMkhyA vidhi, so that meditation on the not-Self, e.g. a
symbol, is excluded. In practical terms, these two interpretations
coalesce, and I believe Suresvara himself thinks so. For example, in the
vArttika, (4. 4. 20-22) and in naishkarmyasiddhi (1. 85-89), he allows for
both the niyama and parisaMkhyA vidhi interpretations, but strongly denies
the apUrva and adhikAra vidhi interpretations.

There is one argument of Suresvara that I still don't understand well,
namely, the statement that, "if a vidhi were only niyama or parisaMkhyA in
nature, and not apUrva and not adhikAra, then there is no eligible agent
on whom anything is enjoined." I take this to mean that even when agency
is ruled out, and eligibility is similarly ruled out, so that it is
impossible to say, "Self-knowledge is enjoined," there is still room to
say that SravaNa, manana and nididhyAsana fall under the category of
niyama/parisaMkhyA vidhi. To understand the logic behind this argument
offered by Suresvara requires a more involved knowledge of mImAMsA than I
have at this point. Nor does it seem appropriate for this list, so I will
leave it at that.

In any case, this complex argument is what leaves room for some later
authors to say that upanishads enjoin it, and for others to say that the
upanishads do not enjoin anything. And both sides can quote Sankara
selectively. From what I know of the later authors, it seems to me that
the bhAmatI school authors deviate more from Sankara's thought in many
important ways than the vivaraNa authors. This is probably because
vAcaspati miSra, the first author in that school, was an independent
thinker whose views were not identical with those of Sankara.

In the midst of all this hair-splitting mImAMsA argumentation lies the
role of yoga. Sankara explicitly says that a steady recollection of
Self-knowledge (Atma-vijnAna-smRti-saMtAna) is the only sure means (ananya
sAdhana) to attain citta vRtti nirodha. It is important to remember that
yogasUtra 1. 2 defines Yoga as citta vRtti nirodha. Concluding the same
argument, Sankara says that Atma-vijnAna-tat-smRti-saMtati is what is
meant by manana and nididhyAsana, which may be seen as a niyama vidhi.

Obviously, recollection or memory (smRti) presumes that what is being
recalled is something that has already been known/experienced, so that
this is *after* the Self is known. In other words, the ideal jIvanmukta is
effectively the best Yogin, and nididhyAsana results in complete citta
vRtti nirodha. If this is so after the rise of AtmajnAna, yogAbhyAsa is
important as a preparatory step before AtmajnAna too, as Rama pointed out
from the naishkarmyasiddhi quotation. This yogAbhyAsa is also *after* the
renunciation of all karma, so that what this means is that the mumukshu is
advised to practise yoga, in order to steady the mind and to develop the
qualities of viveka, vairAgya etc.

Clearly, it would be foolhardy to think that there is a great unbridgeable
gap between yoga and advaita vedAnta. Maybe in terms of the doctrine
presented in the yogasUtra bhAshya, there is room for difference of
opinion, but in terms of the practice of yoga, there is ample room for
agreement. That is why Sankara explicitly says (in brahmasUtra bhAshya 2.
1. 3) that there are many doctrines in sAMkhya and yoga that are identical
with those of vedAnta, but there are other doctrines that are different.
It is to prevent confusion with respect to these other different doctrines
that Sankara argues against these schools. Therefore, there can be partial
acceptance of those doctrines that are similar/identical or those that do
not contradict vedAntic teaching.


"bhava shankara deshikame sharaNam"
List archives :

More information about the Advaita-l mailing list