SOURCE: "The Vedas as a Pramana" in Accomplishing the Accomplished: The Vedas as a Source of Valid Knowledge in Sankara, University of Hawaii Press, 1991, pp. 31-54.

[In the following essay, Rambachan analyzes Sankara's belief in sruti as the essential and true source of knowledge of brahman.]

Sabda can be seen as a pramana for our knowledge of the empirical world as well as ultimate reality. Advaita, however, is not primarily concerned with sabda-pramana as a vehicle of secular knowledge. As such a medium, sabda cannot lay claim to any particular uniqueness, for the knowledge which it conveys is, in most cases, available through other sources.¹ As a pramana of the empirical world, it does not have a sphere which is exclusively its own and which, by nature, it alone is capable of transmitting.² The special nature of sabda for Advaita, therefore, lies in its function as a means of knowledge for ultimate reality. In this capacity, sabda-pramana is synonymous with the Vedas or Sruti.³Advaita seeks to justify the view that, because of the very nature of ultimate reality, the Vedas alone can transmit accurate knowledge. All of the theories about sabda-pramana have emerged as a result of this central concern and the need to defend it against the criticisms of other Indian schools. We hope to show in the discussion below that Sankara posits no alternative to the Vedas for our knowledge of brahman. His acceptance of the function of the avatara as a teacher and the world, in general, as a revelation of brahman does not contradict the necessity for the Vedas.

2.1 The Avatara as a Teacher and the World as a Revelation of Brahman

Sankara's views on the nature of the avatara and His role as revealer emerge from his commentary on the Bhagavadgita. In his introduction to the text, Sankara states that the Lord, after creating the world and the ancestors of humankind (prajapatis), imparted to them the twofold paths of Works (pravrtti dharma) and Renunciation (nivrtti dharma), meant respectively for worldly prosperity and liberation. By this twofold path, He intended to secure order in the universe. When, however, as a result of unrestrained desire, human discrimination was overwhelmed and religion declined, the Lord incarnated Himself as Krsna, for the purpose of reintroducing and strengthening the Vedic religion.⁴

Whenever there is a decay of Dharma, 0 Bharata, and an ascendance of Adharma, then I manifest Myself.

For the protection of the good, for the destruction of evildoers, for the firm establishment of Dharma, I am born in every age.⁵

Sankara emphasizes that the Lord is by nature eternal, pure, intelligent and free. Through the unimpeded control of His creative power (maya), He appears as though embodied. Whereas the individual is born under the control of maya, the Lord incarnates through the mastery of maya. It is a voluntary, self-conscious descent, in full awareness of His majesty.

Though I am unborn, of imperishable nature, and though I am the Lord of all beings, yet ruling over My own nature, I am born by my own Maya.⁶

In the Bhagavadgita, the avatara is not limited by embodiment. Arjuna is limited in knowledge, but Krsna is conscious of all previous births of them both.⁷ Past, present, and future are all spanned in His awareness.⁸ The descent of the avatara is not for any personal need or desire born out of a sense of limitation. He is an already fully accomplished being whose actions are not characterized by any motive of personal achievement.

I have nothing whatsoever to achieve in the three worlds, 0 son of Pritha, nor is there anything unattained that should be attained; yet I engage in action.⁹

His actions are directed to world welfare and preservation.¹⁰ With reference to His instruction of Arjuna, Sankara writes,

Without any interest of His own, but with the sole intention of helping His creatures, He taught to Arjuna, who was deeply plunged in the ocean of grief and delusion, the twofold Vedic Religion, evidently thinking that the Religion would widely spread when accepted and practiced by men of high character.¹¹

The cardinal point to be noted in Sankara's discussion of the avatara as revealer is that He revivifies and reiterates the doctrines of the Vedas. The Bhagavadgita, according to Sankara, "is an epitome of the essentials of the whole Vedic teaching," and is not at variance with it.¹² The instructions of the avatara are in the form of a restatement and do not in any way supersede the primacy of the Vedic revelation. Krsna does not claim to be instituting a new path to freedom but emphasizing His traditional links and the conventional approach.

I taught this imperishable Yoga to Vivasvat; Vivasvat taught it to Manu; Manu taught it to Ikshvaku.

This, handed down thus in succession, the King-sages learnt. This Yoga, by long lapse of time, has been lost here, 0 harasser of foes.

That same ancient Yoga has been today taught to thee by Me, seeing that thou art My devotee and friend; for, this is the Supreme Secret.¹³

It is clear, therefore, that even in relation to the role of the avatara, the Vedas remain the original and authoritative pramana of brahman.

There are several passages in the commentaries of Sankara in which he suggests that brahman is directly and self-evidently revealed in the world and in human beings. In response to an objection that if brahman is a completely unknown entity it cannot become the subject of inquiry, Sankara replies "that the existence of Brahman is well known from the fact of Its being the Self of all; for everyone feels that his Self exists, and he never feels "I do not exist." Had there been no general recognition of the existence of the Self, everyone would have felt "I do not exist. And that Self is Brahman.¹⁴ Earlier on, also, in replying to a query that an unperceived Self cannot become the locus of superimposition, he contends that the Self is well known in the world as an immediately perceived entity. It is nothing but the content of the concept "I."¹⁵ Elsewhere, he remarks that in all cognition, brahman as absolute existence is cognized.¹⁶ Each cognition involves a twofold consciousness of the real and unreal. In a cognition such as "This is a post," the object is limited and finite. But the consciousness of existence or "is-ness," which is the persisting substratum of the object, is eternal. From the standpoint of Sankara, the entire universe can be seen as a name and form revealing of brahman, with whom it is identical.

The problem with these general forms of revelation, if they can be so termed, is that, through them, we are not made aware of the distinctive nature of brahman. Sankara develops his argument about the self-evident manifestation of the atman as the content of the "I" notion, by pointing out that, in spite of this knowledge, the unique nature of the Self remains unknown. As evidence of this, he cites the divergent and mutually contradictory views which different systems hold about the nature of the Self.¹⁷ The point, therefore, is that even though we are not completely debarred from all awareness of ultimate reality, we do not recognize its existence and our understanding is incomplete. What is needed is a valid source of knowledge through which we can apprehend accurately the unique nature of the Self. The Vedas, Sankara contends, is just such a pramana.

2.2 The Eternity of the Vedic Revelation

It is an important contention of both Advaita Vedanta and Purva-Mimamsa that the Vedas are eternal, uncreated, and authorless (apauruseya). The claim for the infallibility of these texts follows directly from this contention. If a personal author is ascribed to the Vedas, they will suffer from the limitations of authorship, and their status as a defect-free source of knowledge will be under doubt. Against the eternity of the Vedas, it may be argued that words originate along with or only after the objects which they signify have come into existence. They are as time-bound as their objects and can in no sense, therefore, be eternal. The assumption here is that the Vedas can be eternal only if the words of which they are composed are established to be so, and words are as transient as the particular things which they signify. This objection, Advaita claims, will be valid if the Vedic words did indeed primarily signify the particular ephemeral objects of the world. To meet this objection and in order to justify the eternity of the Vedas, Advaita argues that the primary significance of words are universals and not individuals. As universals are eternal, the connection between the word and the object signified is an eternal one.¹⁸ A universal, as conceived by Advaita, indicates the essential common characteristics existing in a group of particulars. It signifies both the generic shape (akrti) and also the generic nature (jati).'¹⁹

The inquiry about whether a word primarily signifies a particular (vyakti) or a universal (iati) has elicited a variety of views among the different schools of Indian thought, and the Advaita position is better high-lighted in relation to some of these views.²⁰ According to the Sankhya, a word signifies a particular only, for it is with individuals alone that we deal in everyday usage and activity. If the primary meaning of a word is its essential common characteristics, how can it be applicable to an individual? When one says "The horse is in the field," one is invariably referring to a particular member of the species. Besides this, there are proper names which are singular and can never refer to a group or class.

The main Advaita argument against this view is that if the primary significance of a word is a particular, the word cannot be used more than once. Each individual is unique, and the particular as known at one moment does not persist as the same in the next moment. The fact that the same word can be used repeatedly indicates that its primary significance is not an individual. Sankara adds that objects are limitless in number, and if the primary denotation of a word is an individual, it could not indicate all of them.²¹ In isolation, the word "horse" does not indicate any particular animal but rather the essential characteristics of the species. It is the knowledge of the universal which leads to the recognition of the particular. To recognize a particular animal as a horse, one has to first apprehend the universal characteristics of the species, and these are the primary denotations of a word. This fact does not make it impossible for a word to denote a particular object, for the knowledge of the particular is subsumed under the universal. According to Dharmaraja, "the same cognition that comprehends a generic attribute also comprehends the individuals."²² One may put the argument differently by saying that the individual significance is indirect or implicit, while its universal significance is direct or explicit. It is thus possible, according to Advaita, to conceive of words as being prior to all individuals and eternal.

This theory, however, raised a problem for both Purva-Mimamsa and Advaita. The Vedas make mention of gods such as Vasu and Indra, and, according to this view, words cannot primarily indicate an individual. In this case, they will become noneternal, having no referent before the individual's existence and after the individual's destruction. Sankara's view is that Vedic words such as Vasu signify distinctive general characteristics, for even gods have forms and species. These general characteristics can be gathered from the Vedas. On the other hand, words such as Indra indicate certain ranks, like "commander." The conclusion, therefore, that Vedic words are universal in significance is not refuted. The names of gods are not interpreted to be proper names.²³

The eternal word, according to Sankara, is not of the nature of sphota.²⁴ He holds that "the letters themselves constitute the words."²⁵ Is it not a fact, however, that letters are noneternal, for they possess no reality before and after their utterance? This is not tenable in Sankara's view, for the letters are recognized to be the same in each new utterance. This recognition is not because of any similarity with a previous utterance, nor is it contradicted by any other means of knowledge. He also rejects the view that this recognition is produced by the fact of the letters belonging to the same species (jati).

The recognition would have been caused by the species if the letters were cognized as separate entities like individual cows at the time of each fresh utterance. But this is not so, for it is the letters themselves that are cognized to be the same at each fresh utterance, the recognition taking the form, "The word cow is uttered twice," but not "There are two words cow."²⁶

The letters, in other words, are not individuals which constitute a class, as cows comprise a species. He admits a variation in the apprehension of syllables, but ascribes this to differences in pronunciation due to peculiarities of the vocal organs, not to the intrinsic nature of the letters. Differences can also be attributed to variations of tone (dhvani).²⁷ Sankara's conclusion so far seems to be that since letters are recognized to be the same in each new utterance, they are eternal, and the words which they constitute share this nature. In addition to this, words are eternally connected to their referents, which, being universals, are also eternal.²⁸

As further evidence of the beginninglessness of the Vedas, Sankara remarks that no independent author of the Vedas is remembered.²⁹ This is an argument which was also adduced by Kumarila Bhatta. He contends that if there was an author of the Vedas, he should have been remembered in the long traditional succession of teachers and students, as in the case of, for example, the Buddha. There is no possibility of such an author being forgotten, since religious performances and their effectiveness would be founded solely upon his authority. In reality, however, there are no clear ideas of any composer and some vaguely attribute authorship to "God," Hiranyagarbha, or Prajapati.³⁰ In replying to the argument that the Vedas must have human authors because some sections are named after certain men,³¹ Jaimini explains that this can be accounted for by the fact that such men were foremost in the study and expounding of those portions.³²

Another important argument introduced by Sankara is the origin of the world from Vedic words. According to him, it is a matter of common experience that when one is intent on creating a desirable object, one first recollects the word signifying it and then produces it. The sequence is the same in the case of the creation of the world. Vedic words occur in the mind of Prajapati when He is intent on creation, and, corresponding to them, He creates the universe.³³ He creates the earth, for example, after the word bhuh occurs in His mind. This view does not contradict the Advaita doctrine of brahman as the material cause of creation, for it means simply that "when there is first a word without a beginning and bearing a meaning with which it has an eternal connection, then only is there a possibility of an individual cropping up which can be fit to be referred to by that word. In that sense, it is said to originate from a word."³⁴ Sankara states that the creation of the world from Vedic words is well known from sruti and smrti and he cites several references to support his claim.³⁵ Sankara would appear to be implying that the universe is cyclically created in conformity with ideas or universals which are eternally present in the creator. One wonders, therefore, whether by word, he is really suggesting idea rather than the uttered sound or linguistic symbol. A distinction between the two would have been very useful here. The word sabda is used to denote both idea and sound symbol.

Sankara considers another objection to the eternity of the Vedas. The objector accepts that the doctrine is maintained if one understands names such as Indra to connote eternal species rather than individuals. He contends, however, that the doctrine cannot be argued "in the face of the statements in the Vedas and smrtis that the whole creation, consisting of the three worlds, loses its names and forms and gets dissolved without a trace, and it emerges again as a fresh entity."³⁶ Sankara's reply is that the cyclical creation and dissolution does not refute the eternity of the Vedas because the names and forms of each creation are the same as those of the preceding world that was dissolved.³⁷ The analogy can be drawn between the creation and dissolution of the world and the individual states of deep sleep and waking. In both cases, there is a connection and continuity of activity with earlier states. But is this an appropriate analogy? The individual can easily recollect earlier behavior and activity after emerging from deep sleep, but is such a recall possible after all behavior is eradicated in cosmic dissolution? Sankara concedes that all empirical activity ceases at the time of dissolution (mahapralaya) but argues that because of God's grace, gods like Hiranyagarbha can recall names and forms of earlier cycles, including the Vedas.

From the fact that ordinary creatures are not seen to recollect their past lives, it does not follow that the fact must be the same in the case of divine beings as well. It is noticed that although as living creatures all are the same, counting from men to a clump of grass, still the obstruction to the manifestation of knowledge, glory, etc. increases successively all through the series at every stage; similarly when it is mentioned more than once in the Vedas and Smrtis that knowledge, glory, etc., become increasingly more manifest at each successive stage counting from men themselves up to Hiranyagarbha, it cannot be brushed aside as nonexistent. From this, it logically follows, on the analogy of a man risen up from sleep, that the recollection of the behavior in the past cycle is possible for beings like Hiranyagarbha, who had undertaken meditation and work in a superexcellent way in a past cycle, who have emerged at the beginning of the present cycle (as a result of past achievement), and who have been vouchsafed the grace of God.³⁸

Another reason justifying the identity of names and forms in successive creations is tendered by Sankara. Each new creation is impelled by the necessity of beings to experience the results of virtuous and unmeritorious acts of the past. It is also a field for the expression of likes and dislikes. It is not a causeless or accidental event and must, therefore, conform to earlier patterns. The new creation is potential in the past ones. It is not possible, according to Sankara, to conceive, for example, a different relation between senses and sense objects in each creation.

It is reasonable to conclude from this discussion that Sankara conceives the eternity of the Vedas in the sense of an identical but eternal flow (pravaha nityata). The eternity of brahman, on the other hand, is of an absolutely unchanging kind (kutastha nityata). It would appear also that Sankara ascribes to the Vedas the same empirical (vyavaharika) level of reality as the world, for he admits that the Vedas, like the world, are negated in the knowledge of nondual brahman.³⁹

2.3 Isvara as Revealer of the Vedas

Purva-Mimamsa uncompromisingly rejects the view that the Vedas were ever composed by anyone. Nyaya, on the other hand, ascribes authorship to isvara, whose existence they seek to establish inferentially. Sankara occupies a position between these two views.⁴⁰ Like the Mimamsa, but unlike Nyaya, Sankara admits the Vedas to be authorless (apauruseya).⁴¹ He seems, however, to understand this concept very differently, even though he does not offer us a definition anywhere. Commenting on Brahmasutra 1.1.3, Sankara explains why brahman alone can be the source of the Vedas.⁴² It is a well-known fact, he asserts, that the author of a text on any subject is more informed than the text itself. The grammar of Panini, for instance, represents only a part of the subject known to him. It is obvious, therefore, that the source of texts like the Vedas, divided into many branches and illuminating lamp-like a variety of subjects, must be omniscient and omnipotent. This is even more apparent from the effortlessness with which they emerge from God. The Vedas compare their own emergence with the ease of breathing.⁴³ Sankara states very clearly, however, that it is the eternally composed and already existent Vedas that are manifest like a person's breath.⁴⁴ Elsewhere, he explains that the projection of the Vedas should only be understood in the sense of the initiation of a cycle of transmission through a line of teachers and students, for no other kind of projection is possible for a text without beginning and end.⁴⁵Isvara, then, does not produce the Vedas but reveals or manifests them as they were in the previous creation. He offers various suggestions to this effect.⁴⁶ The general idea is that the Vedas are revealed in the same linguistic form at the beginning of each creation to qualified seers.

Further clarification of Sankara's understanding of the concept of apaurauseya is provided by Dharmaraja and Vacaspati. According to Dharmaraja, the Vedas are not eternal because they are produced by brahman. They are not, however, pauruseya because they depend on utterance of the same kind. A sentence can be described as pauruseya only if it is original and not the reproduction of an earlier utterance.

For instance, in the beginning of the cosmic projection, the Lord produced the Vedas having a sequence of words similar to that which had already existed in the Vedas in the previous cosmic projection, and not Vedas of a different type. Hence the Vedas, not being the object of utterance that is independent of any utterance of the same kind, are not connected with a person. The utterance of the Mahabharata etc., however, is not at all dependent on any utterance of the same kind. Thus two kinds of verbal testimony have been determined, viz., that which is connected with a person and that which is not.⁴⁷

Vacaspati argues along similar lines.⁴⁸Purva-Mimam sakas, he says, who do not believe in a creation or destruction, advocate a beginningless and unbroken sequence of Vedic study. Advaita, however, although differing from them in accepting the supreme Self to be the creator of the eternal Vedas, does not understand Him to be entirely free in respect of them, since He creates their sequence in conformity to the previous ones.

Not in any creation is brahminicide the cause of good nor the horse-sacrifice the cause of evil, any more than fire can wet or water can bum. Just as, in this creation, the study of the Vedas in the settled sequence is the cause of prosperity and beatitude, and (studied) otherwise is the cause of evil even as a verbal thunderbolt, even so does it happen in another creation; hence, the creator, who, though omnipotent and omniscient, creates the Vedas in accordance with what they were in earlier creations, has not a free hand.⁴⁹

2.4 The Necessity and Justification of the Vedas as a Pramana

The general justification of Sankara for a special means of knowledge like the Vedas is that it provides the knowledge of those things which cannot be known through any of the other available sources of knowledge. More specifically, it informs us of the means of attaining good and avoiding evil, in so far as these cannot be known through perception and inference and are the two ends naturally pursued by us. The Vedas are not concerned to provide information about these dual objectives to the extent that they are within the range of human experience. Such knowledge is easily available from perception and inference.⁵⁰ One imagines, for example, that a scripture is not necessary for instruction about road safety.

According to Sankara, the two categories of knowledge inaccessible to all other pramanas and attainable exclusively through the Vedas are dharma and brahman.⁵¹ We are afforded a clear statement of Sankara's view on the knowledge of dharma in his commentary on Brahma-sutra 3.1.25.⁵² Here he is responding to the objector's claim that the slaying of animals in sacrifices might be responsible for the soul's birth as a plant. He answers that the knowledge of merit (dharma) and demerit (adharma) is derived solely from the scriptures. From the Vedas alone can we know which acts are virtuous and which are not. The reason is that these are supersensuous realities, beyond the capacity of the senses. In addition to this, dharma and adharma vary with time and place. An act that may be sanctioned at a certain time and place and under some circumstances may not be approved with a change of these factors. It is impossible therefore, to learn of dharma from any other source.⁵³

It is necessary, however, for the individual to be made aware of the persisting existence of the Self in a future life if one is to be motivated to attain what is good in that life. The materialists (Carvakas), for example, who deny all future existence, do not show any such concern. Sruti, therefore, informs us of this future existence and of the particular means of attaining good and avoiding evil in that life.⁵⁴ In a typical discussion which illustrates very well his procedure for legitimizing the Vedas as a pramana, Sankara shows why this knowledge of a future existence is not otherwise attainable.

After a series of Upanisad quotations to show support for the doctrine, a question is tendered.

Objection: Is it not a matter of perception?

Reply: No, for we see the divergence of opinion among different schools. Were the existence of the Self in a future body a matter of percepton, the materialists and Buddhists would not stand opposed to us, saying that there is no Self. For nobody disputes regarding an object of perception such as a jar, saying it does not exist.

Objection: You are wrong, since a stump, for instance is looked upon as a man and so on.

Reply: No, for it vanishes when the truth is known. There are no more contradictory views when the stump, for instance, has been definitely known as such through perception. The Buddhists, however, in spite of the fact that there is the ego-consciousness, persistently deny the existence of the Self other than the subtle body. Therefore, being different from objects of perception, the existence of the Self cannot be proved by this means. Similarly, inference too is powerless.

Objection: No, since the sruti points out certain grounds of inference for the existence of the Self, and these depend on perception, (these two are also efficient means of the knowledge of the Self).

Reply: Not so, for the Self cannot be perceived as having any relation to another life. But when its existence has been known from the sruti and from certain empirical grounds of inference cited by it, the Mimamsaks and logicians, who follow in its footsteps, fancy that those Vedic grounds of inference such as the ego-consciousness are the products of their own mind, and declare that the Self is knowable through perception and inference.⁵⁵

This knoweldge of dharma and adharma is derived from the ceremonial portion (karmakanda) of the Vedas.⁵⁶ This does not, however, exhaust the authoritative subject matter of the Vedas. The karmakanda, authoritative as it is, is not accepted by Sankara as providing a solution to the fundamental problem. It accepts the desires for the enjoyment of the results of various actions but does not question the origin or legitimacy of these desires. This propensity, as maintained by Sankara, is born out of a basic self-ignorance, the perception of oneself as a limited being. As long as this false notion is not removed by the knowledge of one's already-accomplished identity with brahman, one continues to search for fullness through the results of limited actions. Actions, however, produce inescapable results, and we are trapped in a futile quest through successive births and deaths (samsara). Our ever-accomplished freedom and unlimitedness, the real end of all our actions, perpetually eludes us. The removal of this ignorance (avidya) is the authoritative aim and concern of the jnanakanda (knowledge section) of the Vedas.⁵⁷Sruti eliminates this ignorance by teaching about the true nature of the Self. It is the intention of all the Upnisads, Sankara says, to establish the identity between atman and brahman.⁵⁸

We hold that it is the definite conclusion of all the Upanisads that we are nothing but the Atman, the Brahman that is always the same, homogeneous, one without a second, unchanging, birthless, undecaying, immortal, deathless and free from fear.⁵⁹

Sankara is equally emphatic on the absolute inapplicability of all pramanas, except sruti, to the knowledge of brahman.⁶⁰ He is tireless in explaining the incompetence of sense perception in apprehending brahman. Sankara refuses to accept that because brahman is an existent entity; like all such realities, it must be the object of other sources of valid knowledge. The senses are naturally capable of grasping and revealing their appropriate objects. Brahman, however, remains unapproachable through any of them because of its uniqueness.⁶¹ The organs can only grasp a differentiated object within their range.⁶² We have already considered the nature and evolution of the five sense organs.⁶³ Each organ evolves out of a particular element which enables it to apprehend a quality proper to that element. The eyes, for example, evolve out of the subtle sattva aspect of fire and are the organs for perceiving the quality of form which is unique to fire.⁶⁴ It is the special relationship, therefore, between sense organ and element which empowers each one to cognize an appropriate quality. Sound, sensation, form, taste, and scent are their respective spheres of functioning. Brahman, however, has neither sound, touch, form, taste, or smell. It is without qualities (nirguna) and outside the domain of the sense organs.⁶⁵Brahman is limitless, and to become an object of sense knowledge is to be finite and delimited, to be one object among many objects. A brahman that is sense apprehended is, therefore, a contradiction. However perfect or magnified the capacity of a sense organ is imagined to be, it will function only in a limited sphere of activity. Sankara refutes the allegation that there is any contradiction in the Bhagavadgita 's denial of brahman as both sat and asat, by interpreting these terms with reference to the nonavailability of brahman as an object of sense knowledge.

Objection: Every state of consciousness involves either the consciousness of existence or that of nonexistence. Such being the case, the Knowable should be comprehended either by a state of consciousness accompanied with the consciousness of existence, or by a state of consciousness accompanied with the consciousness of nonexistence.

Answer: No; for being beyond the reach of the senses, it is not an object of consciousness accompanied with the idea of either (existence or nonexistence). That thing, indeed, which can be perceived by the senses, such as a pot, can be an object of consciousness acompanied with the idea of existence, or an object of consciousness accompanied by the idea of nonexistence. Since, on the other hand, the Knowable is beyond the reach of the senses and as such can be known solely through that instrument of knowledge which is called Sabda, It cannot be, like a pot, etc., an object of consciousness accompanied with the idea of either (existence or nonexistence) and is therefore not said to be sat or asat.⁶⁶

In addition to the inherent limitations of the sense organs and the absence in brahman of any quality that can be apprehended by any one of them, there is the impossibility of objectifying brahman. The process of empirical knowledge involves a distinction between the subject and object, the knower and known. We know things by making them the objects of our awareness, and in this way they are available for our scrutiny and analysis. Knowledge of an object presupposes the subject, the knower. Brahman, however, is the eternal subject. As awareness, it illumines everything, and the entire universe, including mind, body and sense organs, is its object.⁶⁷ It is impossible for the unchanging knower to be made an object of knowledge, like a pot or a thought. It is absurd to conceive of the subject as an object, for in its absence there is no subject to know the subject as an object. It is the light even of lights.⁶⁸

Even in the state of ignorance, when one sees something, through what instrument should one know that owing to which all this is known? For that instrument of knowledge itself falls under the category of objects. The knower may desire to know, not about itself, but about objects. As fire does not burn itself, so the Self does not know itself, and the knower can have no knowledge of a thing that is not its object. Therefore through what instrument should one know the knower owing to which this universe is known and who else should know it?⁶⁹

It is not possible to circumvent this difficulty by positing that the Self can be both subject and object. This might have been tenable if the subject and object were complementary and not opposed. By nature, however, the subject and object are absolutely opposed and such contradictory qualities cannot be posited of the same entity.⁷⁰ No division of any kind can be made in the case of the atman.⁷¹

If perception is unfitted for furnishing us with the knowledge of brahman, are any of the other four pramanas (inference, comparison, postulation, and non-cognition) more competent? The general view of Sankara is that these sources are more or less dependent on perception for their data and can have no access to areas from which it is debarred. We have already, for instance, considered the nature of inference as a pramana.⁷² Inferential knowledge is derived from a knowledge of the invariable relation (vyapti) between a thing inferred (sadhya) and the ground from which the inference is made (hetu). Brahman, however, has no apprehensible or differentiating qualities with which it has an invariable relation and which can form the ground of an inference.⁷³ It is impossible, therefore, to infer the existence of brahman.

There is no hint, however, of the skeptic in Sankara. He is unwavering in his position that brahman is knowable and that sabda-pramana is the only valid means. In a discussion in which he is concerned to establish that brahman is changeless and indivisible in spite of being the material cause of the creation, his views are unequivocal.

There is no violation of the texts about partlessness, since partlessness is accepted on account of its very "mention in the Upanisads," and the Upanisads are the only authority about it, but not so are the senses etc. Hence it has to be accepted just as it is presented by the Upanisads. The Upanisads prove both the facts for Brahman—the nontransformation of Brahman as a whole and partlessness. Even the things of this world like gems, incantations, herbs, and so on, are seen to possess many powers capable of producing incompatible effects under a variety of space (environment), time, and cause. And even these powers can be known not from mere reasoning but from such instructions as "Such a thing has such kinds of potency with the aid of such things, on such things, and for such purposes." So what need has one to argue that the nature of Brahman, whose power is beyond all thought, cannot be ascertained unless it be through the Vedas? So also it has been said by an author of the Purana, "Do not bring those things within the range of argumentation which are beyond thought. The nature of a thing beyond thought consists in its being other than the things within Nature." Hence a supersensuous thing is truly known from the Vedic source alone.⁷⁴

It is not possible, according to Sankara, to even guess about brahman without the assistance of the Vedas.⁷⁵ He leaves no room for any doubt about this conclusion. Brahma-sutra 1.1.2, for instance, reads "That (is Brahman) from which (are derived) the birth etc. of this (universe)." The following sutra (1.1.3) reads, "Because of being the source of the scriptures" (sastrayonitvat). Sankara sees another possibility in the Sanskrit compound of the latter and reads it also as "Since the scriptures are its valid means."⁷⁶ He justifies this reading on the ground that since sutra 1.1.2 made no explicit mention of the scriptures, one might construe that an inferential argument is being presented for establishing brahman as the source of the world. Any such doubt ought to be removed, and it must be made clear that brahman is known as the source of the universe from the scriptures alone. They are the only valid means of this knowledge.⁷⁷ We can briefly note that Sankara also dismisses independent reasoning as a suitable means of arriving at accurate knowledge of brahman.⁷⁸

One has to be extremely cautious in examining Sankara's exegesis of Upanisad verses treating the unknow-ability of the Self. He never accepts any of these passages literally, and there is no basis for concluding, as some have done, that in Sankara's view, none of the pramanas can give us knowledge of brahman.⁷⁹ There are basically two ways in which Sankara interprets these statements. First of all, the Self is unknowable in the sense and manner of an object. The knowing process generally involves the knowledge of an object different from oneself. As the eternal knower, the witness of every cognition, brahman can never be known in this manner.

Teacher: If you think "I have known Brahman well enough," then you have known only the very little expression that It has in the human body and the little expression It has among the gods. Therefore Brahman is still to be deliberated on by you.

Disciple: I think Brahman is known. I do not think I know Brahman well enough; (i.e. I consider) not that I do not know: I know, and I do not know as well.

Teacher: He among us who understands that utterance, "Not that I do not know: I know and I do not know as well," knows that Brahman. It is known to him to whom It is unknown; he does not know to whom It is known. It is unknown to those who know well, and known to those who do not know."⁸⁰

Secondly, brahman is unknown in the sense of being undisclosed through any other pramana but sabda-pramana. Brhadaranyaka Upanisad 3.6.1, for example, consists of a discussion between Gargi and Yajnavalkya. Beginning with earth and ending with the world of Hiranyagarbha, Gargi questions him about the successive pervasiveness of each factor. According to Sankara, the inference suggested here is that "whatever is an effect, limited and gross, is respectively pervaded by that which is the cause, unlimited and subtle, as earth is pervaded by water."⁸¹ When Gargi, however, asks, "By what is the world of Hiranyagarbha pervaded?," Yajnavalkya refuses to proceed with the discussion.

"Do not, 0 Gargi, push your inquiry too far, lest your head should fall off. You are questioning about a deity that should not be reasoned about. Do not, 0 Gargi, push your inquiry too far." Thereupon Gargi, the daughter of Vacaknu, kept silent.

Sankara does not construe Yajnavalkya's refusal as an indication of the impossibility of any further knowledge. On the contrary, he charges Gargi with disregarding the proper method of inquiry. Yajnavalkya terminates the discussion, according to Sankara, because of Gargi's attempt to establish brahman inferentially, whereas it is to be known only from the Vedas. The idea is that brahman is not unascertainable but must be approached through the apposite pramana.⁸²Kena Upanisad 1.3 is a classic declaration of the predicament of conceptualization and instruction about brahman.

The eye does not go there, nor speech, nor mind. We do not know (Brahman to be such and such); hence we are not aware of any process of instructing about It.

Sankara does not reiterate this sense of perplexity and impotence. Concluding his remarks on this verse and introducing the following one, he writes

The contingency of the total denial of any process of instruction having arisen from the text, "We do not know Brahman, and hence we are not aware of any process of instructing about It," an exception to this is being stated in the next verse. True it is that one cannot impart knowledge of the highest with the help of such means of valid knowledge as the evidence of the senses; but the knowledge can be produced with the help of traditional authority. Therefore traditional authority [agama] is being quoted for the sake of imparting instruction about it.⁸³

It is palpable, therefore, that Sankara presents an argued and developed rationale for sabda-pramana as the only source of brahmajnana. Radhakrishnan's view that it is difficult to find support in Sankara for the claim that inquiry into the Vedas is the only avenue to the knowledge of brahman is absolutely without basis.⁸⁴ Unsubstantial also is Devaraja's argument that Sankara's reliance on sruti is an expression of his "ultra-orthodox mood."⁸⁵ It is not at all possible to dismiss Sankara's affirmation of the sruti as simply an attempt to clothe his views with a sanction of a traditional authority. One may perhaps dispute his exegesis of particular scriptural passages but not his endorsement of sruti as the only credible pramana of brahmajnana.

2.5 The Authority and Infallibility of the Vedic Revelation

The authority of the sruti within the sphere of its own subject matter is, according to Sankara, independent and self-evident. Its function in relation to the revelation of brahman is comparable to the perception of an object through the eye.⁸⁶ This independent authoritativeness is underscored by his analogy with the sun. The Vedas, he says, are as trustworthy with respect to their own subject matter as the sun is with regard to the objects which it illumines.⁸⁷ The suggestion here seems to be that while objects depend for their revelation on the light of the sun, the sun itself is self-illuminating. Similarly, the authoritativeness of the sruti is independent and self-evident. The validity of the Upanisad does not await inferential verification.⁸⁸ There are no misgivings about sruti's infallibility with respect to its subject matter. "Knowledge of Reality springs from the Upanisad texts alone."⁸⁹ "Truth is the fact of being in accordance with the scriptures."⁹⁰ Vedic statements, he affirms, unlike those of men, are not delusive, equivocal, or deceptive about their theme.⁹¹ They admit of no doubt and are productive of accurate knowledge.⁹² In this respect, there is no difference between ritualistic texts and those informing us of the nature of brahman.

That rites like the new and full moon sacrifices produce such and such results, and have to be performed in a certain definite way, with their parts following each other in a particular order, is a supersensuous matter beyond the range of our perception and inference, which we nevertheless understand as true solely from the words of the Vedas. Similarly it stands to reason that entities like the Supreme Self, God, the deities, etc., of which we learn also from the words of the Vedas as being characterized by the absence of grossness, etc., being beyond hunger and thirst and the like, and so on, must be true, for they are equally supersensuous matters. There is no difference between texts relating to knowledge and those relating to rites as regards producing an impression. Nor is the impression conveyed by the Vedas regarding the Supreme Self and other such entities indefinite or contrary to fact.⁹³

In the main, however, Sankara's principal justification of the reliability and authoritativeness of the Vedas is an epistemological one. The sruti fulfills the criteria of being a pramana. It has the capacity to generate certain and fruitful knowledge.

Is or is not certain and fruitful knowledge generated by passages setting forth the nature of the Self, and if so, how can they lose their authority? Do you not see the result of knowledge in the removal of the evils which are the root of transmigration, such as ignorance, grief, delusion and fear? Or do you not hear those hundreds of Upanisad texts such as, "Then what delusion, and what grief can there be for one who sees unity?" (IS.U. 7).⁹⁴

In addition to its fruitfulness, this knowledge can neither be produced nor nullified by any other pramana, for there is none superior to the Vedic texts.⁹⁵ Suresvara suggests four reasons when a pramana may be disregarded:⁹⁶ (i) if it reveals something already revealed by another authoritative source of knowledge; (ii) if its revelations are contradicted by another source of knowledge; (iii) if it reveals ambiguous or doubtful knowledge, (iv) if it reveals nothing. The Vedas, however, according to him, reveal brahman, which is beyond the scope of all other pramanas. Their revelations are neither ambiguous nor contradicted by any other pramanas. In addition, they are productive of fruitful knowledge.⁹⁷ A similar view has been tendered by Vacaspati. He argues that the authoritativeness of a pramana consists in generating knowledge which is unsublated, not already understood, and indubitable. This capacity is an intrinsic one and not dependent on any other pramana.⁹⁸

At this stage, we can underline a conclusion which was only hinted at earlier. Advaita does not attempt to establish the authority or infallibility of the Vedas from the fact of isvara's omniscience. The reason is that Advaita finds it impossible to demonstrate the existence of God by any kind of independent reasoning. In the absence of such a proof, all arguments become helplessly circular, "omniscience being proved from the authority of the scriptures and the (authority of the) scriptures being proved from the knowledge of the omniscience of the author."⁹⁹ In Indian philosophy, Nyaya champions the rational theology and seeks to establish God's existence by a syllogistic inference. This argument, as we have seen, is based on a knowledge of the invariable relation (vyapti) between the object perceived (hetu) and the object inferred (sadhya). The Nyaya argument takes the following form: All created or produced objects, for example, pots, have sentient beings as their makers, who are aware of the material cause and purpose of creation. The universe is a created object because it is a compound of insentient parts which could not have assembled themselves. From this fact, it is inferred that the world has a creator. In brief, the Vedas are authoritative because they are derived from God, who is reliable and trustworthy.¹⁰⁰ Sankara accepts that the world is an effect but argues that it cannot be certified by inference that brahman is the cause. While the universe is an object of perception, brahman is not, and an invariable relation (vyapti) cannot be established between them.¹⁰¹ Sankara also advances other strong arguments against the conclusions of a purely rational theology.¹⁰² It is difficult, he contends, to explain the inequalities of creation unless we ascribe to God the possession of likes and dislikes. If in order to avert this charge,. one argues that God is impelled by the merits and demerits of beings, the defect of a circular argument arises. God acts in accordance with karman and karman produces results when impelled by God. To suggest that this mutual dependence is beginningless does not avert this difficulty. Moreover, Nyaya themselves admit that the impulse to act is an indication of the defect of likes and dislikes. The Yoga concept of God as a special indifferent purusa does not help. Nyaya maintains that God is distinct from matter and individual souls. How then does God control them? God, matter, and souls being omnipresent and partless, neither can be related by conjunction nor inherence. Those who resort to inference argue that God molds matter (pradhana) even as a potter with clay. But this is not possible because pradhana is conceived of as being formless and beyond the range of perception. How is it possible to work upon such a material? We are thrown into enormous difficulties if perceptual experience is used as the basis for inferences about God. We will be forced to conclude that God possesses a body like us and is consequently subject to all of our limitations. Finally, Nyaya argues that God, matter, and soul are eternal and infinite. In this case, Sankara says, God will be unable to measure the limits of all three, and God ceases to be omniscient. On the other hand, if God knows the limits of all three, they cannot be infinite and will come to an end, depriving God of rulership.

Unlike the rationalists, however, the Advaitin is not constrained into dependence upon observed facts for the knowledge of God. Sruti is the source for ascertaining the nature of the cause. For this reason, one has no difficulties in accepting brahman to be both efficient and material cause, although we find no such analogy in experience. Conscious agents are not generally material causes.¹⁰³ To the argument that in conformity with experience, it is not admissible for brahman to create without organs, Sankara rejoins,

This supreme and sublime Brahman, is to be known from the Vedas alone, but not from reasoning. Moreover, there cannot be any such rule that since somebody is seen to have some power in some way, another should also have it in the same way. Moreover, this also has been stated that even though all distinctions are denied in Brahman, still It can have accession of all powers owing to the presence of a variety of aspects conjured up by ignorance. In support of this is the scripture, "He moves and grasps even though he is without feet and hands, he sees without eyes and hears without ears" (SV.U. 3.19), which shows the possession of all kinds of power by Brahman, even though It is devoid of organs.¹⁰⁴

One should not conclude from the above argument that Sankara finds no use for inferential arguments about God's existence. The problem with these kinds of arguments is that they merely suggest possibilities; they are not conclusive. Once, however, the reality and nature of God are ascertained from the sruti, he attempts as far as possible to show that the conclusions of sruti conform to reason. In this attempt, he unhesitatingly uses inferential arguments and analogies.¹⁰⁵

Before concluding this section of our discussion, it is necessary, on the basis of our ascertained conclusion, to refute some of the widely accepted interpretations of Sankara's orientation to the authority of the Vedas.¹⁰⁶ Having seen that he does not try to establish authoritativeness on the basis of an inference from God's omniscience, there is no basis whatsoever for the view that sruti is acceptable to him because it embodies the records of the religious experiences of ancient mystics. The uniqueness of sruti is that its authority is not personal or derived. It is, as we have seen, apauruseya. We cannot emphasize strongly enough the purely conjectural character of the view that the Vedas are merely meant for inferior aspirants who are incapable of directly discovering revealed assertions, or that these assertions are discoverable through some other source. Sankara has not left his matter open to speculation, and such conclusions are entirely indefensible. The overwhelming evidence of his major commentaries affirms that he saw sabda-pramana as the only definitive source of brahmajnana. His unambiguous justification of this pramana is the impossibility of knowing brahman otherwise. The view that the Vedas "contain truths which man could, by the exercise of his own faculties, discover" is entirely irreconcilable with Sankara's vindication of their authority.¹⁰⁷ The nature and detail of his justification of the Vedas as the only pramana of brahman do not lend any support to the view that his aim was merely to seek the approval of their authority for his conclusions.

2.6 The Limitations of Sruti and Its Relation to Smrti

The word smrti is derived from the root smr (to remember). It is generally used to indicate authoritative texts other than the Vedas.¹⁰⁸Smrtis are also a form of sabda-pramana, but unlike the sruti, they are of human origin (pauruseya) and, therefore, less authoritative. Sankara uses the analogy of perception and inference to describe the relationship between sruti and smrti. Smrti is dependent on sruti even as inference is reliant on perception for its data. Smrtis are not, therefore, independently authoritative.¹⁰⁹ Compared to the direct and independent validity of the Vedas, the authority of smrti is remote because it "depends on some other source of knowledge and since the memory of the speaker intervenes."¹¹⁰

Smrtis are authoritative only when they conform to Vedic texts. They are to be discarded in those cases where they directly contradict the sense of the Vedas. This is the method, according to Sankara, of reconciling and deciding between mutually opposed smrti texts.¹¹¹ One need not reject an entire smrti text because some parts are opposed to Vedic doctrines. Advaita, for example, shares some doctrines in common with Sankhya and Yoga. Although they are both dualists, Sankhya subscribes to the qualityless nature of the Self and Yoga emphasizes the value of detachment. These views are compatible with and acceptable to Advaita.¹¹² Sankara also acknowledges the authority of Kapila and his followers with respect to the nature, functions, and products of the gunas.¹¹³

How should we view smrti texts which do not contradict Vedic ones, but for which we can find no corroboration in the Vedas? In such cases, according to Sankara, we are to infer the existence of the sruti text upon which the smrti is based."¹¹⁴

The sruti shares with all other pramanas the characteristic of having a circumscribed concern and sphere of authority. As we noted earlier, it is intended for the revelation of dharma and brahman, both of which are incapable of being known through any other pramana. Its purpose is not to disclose matters within the range of human experience, ascertainable through any of our ordinary means of knowledge. If a sruti statement contradicts a well-established fact of our everyday experience, it cannot be considered authoritative because such a matter would be outside its authority.

Sruti is an authority only in matters not perceived by means of ordinary instruments of knowledge such as pratyaksha or immediate perception;—i.e., it is an authority as to the mutual relation of things as means to ends, but not in matters lying within the range of pratyaksha; indeed, sruti is intended as an authority only for knowing what lies beyond the range of human knowledge.… A hundred srutis may declare that fire is cold or that it is dark; still they possess no authority in the matter.¹¹⁵

If, however, sruti did describe fire as being cold or dark, we should construe its meaning figuratively.¹¹⁶ Vedic texts are not meant for creating things anew or reversing the nature of anything. They are revelatory and are concerned with simply expressing things as they are. They do not misrepresent facts.¹¹⁷ In order to accomplish its purpose, the sruti uses conventional words and meanings and cites examples from our everyday world. By these examples, "the scriptures seek to tell us about some other thing which does not contradict them. They would not cite an example from life if they wanted to convey an idea of something contradictory to it. Even if they did, it would be to no purpose, for the example would be different from the thing to be explained.""¹¹⁸

As conceived by Sankara, one pramana does not contradict another. Each pramana only reveals knowledge that cannot be obtained by another.¹¹⁹ Clarification has been provided on this point by Suresvara.¹²⁰ According to him, two pramanas whose spheres are entirely different cannot be contradictory. The eye, which perceive forms, and the ears, which apprehend sounds, are not opposed. It is only when two pramanas deal with the same object and are contradictory that they are opposed.

If a thing is perceived by the senses it cannot be revealed by the Veda; if a thing is genuinely revealed by the Veda it cannot be an object of sense-perception. A perception (purporting to bear on a revealed subject is only) a semblance of a perception; and a revealed text (bearing on what is subject to perception is only) a mere semblance of a revelation.¹²¹

Suresvara goes on to add that pramanas do not have to cooperate with each other to produce knowledge as the various members of a syllogism do. Each is authoritative within its own sphere and independently capable of giving rise to valid knowledge.

In Sankara's view, then, the knowledge of the Vedas is not opposed to fact. He denies, for example, that there is any conflict between sruti and perception with regard to the nature of the Self. The claim of the sruti that the Self is free from all limitations is not opposed to our perceptual experience. The latter has for its object the Self as identified with various limiting adjuncts (upadhis). Sruti, however, points to a Self free from all erroneous identification.¹²² Similarly, Sankara denies that there is any contradiction between perception and the unity of brahman.¹²³ We should remind ourselves, however, of Sankara's position that not everything revealed by sruti is explicable on the analogy of our everyday experience. The nature of brahman as both efficient and material cause cannot be inferred from any of our experiences of creation.¹²⁴ If, after the meaning of a sruti text is well ascertained, a conflict arises with any other pramana, sruti must be accorded primacy.¹²⁵

Sankara mentions some very specific topics which it is not the function of the sruti to reveal. It is not the purpose of the sruti to inform us of the details and order of the creation of the world. We neither observe, nor are told by the texts that the welfare of humanity depends upon this kind of knowledge. In fact, when the texts are properly analyzed, we find that such passages are intended for instruction about brahman. They are not independent passages but are subservient and linked to those discussing brahman. Accounts of creation, which involve analogies of clay, iron, sparks, etc., are only meant for showing the nondifference of effects from cause and upholding the unity of brahman.¹²⁶

Similarly, sruti is not concerned to provide information about the individual self (jivatman).

The individual soul, present in everybody as the agent and experienced in association with such limiting adjuncts as the intellect, is known from common experience itself, and so it is not mentioned in the Upanisads for its own sake. But as God is not thus familiarly known from common experience, He is intended to be declared in the Upanisad for His own sake. Hence it is not proper to say that any mention of Him is uncalled for.¹²⁷

Finally, not only is sruti limited with reference to its content, but its injunctions have a limited applicability. The one who has gained brahmajnana stands outside the pale of injunctions. Directives to act or to refrain from action are relevant to one who is in search of appropriate means for gaining some desirable object or avoiding an undesirable one. Injunctions somehow appear superfluous to the brahmajnani who has no unfulfilled personal wants.

That man, verily, who rejoices only in the Self, who is satisfied with the Self, who is content in the Self alone,—for him there is nothing to do.

For him, there is here no interest whatever in what is done or what is not done. Nor is there in all beings any one he should resort to for any object.¹²⁸

The point seems to be that the brahmajnani, having shed self-centered wants, spontaneously becomes a source and example of right action. The directives which aim at bringing about this effortless ideal are redundant once it is discovered. It is sruti's own tribute to her ideal. As Sankara remarks,

If man who has realised the identity of the Self and Brahman has still to bow down to injunctions, even though he is beyond all mandates, then there will remain none who is outside the pale of scriptural direction; and so all actions will become fit to be undertaken by all and sundry at all times. But that is undesirable. Nor can he be directed by anybody, for even the scriptures emanate from him. Not that anyone can be impelled by any sentence issuing out of his own wisdom. Nor is a well-informed master commanded by an ignorant servant.¹²⁹

The conclusions I have reached in this discussion about Sankara's understanding of the nature of the authority of the Sruti differ radically from some of the opinions I summarized in our Introduction. I terminate the discussion here by reflecting on these.

It is indisputable that there is a profound epistemological basis for Sankara's dependence on sruti as the only authoritative source of brahmajnana. There is no dearth of evidence to support the view that he saw the sruti as the only valid source of this knowledge. His way of justifying the necessity for a pramana in the form of words completely belies the argument that his recourse to sruti was motivated merely by the wish to gain the support of an authoritative tradition for his personal views. Sabda-pramana, contrary to the view of current opinion, is perceived by him as a unique source of knowledge about brahman, justified by the fact that, as human beings, we cannot otherwise know brahman. Sruti would not satisfy the criterion of novelty if the knowledge which it provides could be obtained from any other course. I have also highlighted his argument that the sruti, like all other valid sources of knowledge, does not need the confirmation or verification of any other pramana. It is a self-valid source of fruitful knowledge.

I have not found any evidence in the commentaries of Sankara to support the conclusion that he accepted the sruti as authoritative and infallible because it embodied the self-certifying experiences of ancient mystics. The grounds of his argument for sruti's infallibility are very different. When Sankara does not even seek to establish the authority of the sruti on the basis on isvara 's omniscience, it is difficult to conceive that he would derive it from human authority. In this matter, his views are closely allied with those of Purva-Mimamsa. I have noted the connection between his arguments for the uncreated, eternal, and impersonal (apauruseya) nature of the sruti and its infallibility.

While Sankara advances various arguments for the validity of the sruti, it appears to me that he ultimately falls back upon the claim that the sruti fulfills the criteria of being a pramana. It provides a knowledge which is not available through any other source and which is not contradicted by another valid pramana. In addition, this knowledge is seen to be fruitful in the elimination of samsara and its attendant—evils such as grief, fear, and delusion. It would seem that the onus is rather thrust upon the one who does not accept sruti as a pramana to disprove its validity.…

Notes

¹ This does not contradict my previous argument that the nature and method of sabda-pramana distinguishes it as an independent source of knowledge.

² Some Advaitins, however, argue that even in the empirical world, sabda-pramana alone can inform us of another person's thoughts and emotions.

³ The Vedas are collectively referred to as sruti (literally, "that which is heard"). This term suggests the oral transmission of knowledge in a succession of teachers and students. It is suggested that the reason for the oral transmission of Vedic knowledge was the absence of a written script at the time when the Vedas were composed. It appears that even long after writing was introduced, there was a clear preference for the oral transmission of scripture, and religious learning through the written word was looked down upon. Perhaps it was felt that the oral transmission of a tradition was a far better way of ensuring its living continuity. Vedic words had to be handed down exactly as they had been heard, and correct sounds and pronunciations became all-important. Continuous repetition became the mode of learning, and sravana (listening) became the first procedure in assimilating knowledge. See William Cenker, "The Pandit: The Embodiment of Oral Tradition," Journal of Dharma, 5 (1980), pp. 237-251.

⁴ B.G.B., Intro., pp. 2-3.

⁵ B.G. 4:7-8.

⁶ Ibid., 4:6.

⁷ Ibid., 4:5.

⁸ Ibid., 7:26.

⁹ Ibid., 3:22.

¹⁰ Ibid., 3:23-24.

¹¹ B.G.B., Intro., p. 4.

¹² Ibid. "Know this by long prostration, by inquiry, by service, those men of wisdom who have realized the truth will teach thee wisdom" (ibid., 4:34).

¹³ B.G. 4:1-3.

¹⁴ B.S.B. 1.1.1, p. 12.

¹⁵ Ibid., Intro., p. 3.

¹⁶ B.G.B. 2:16, pp. 34-37.

¹⁷ B.S.B. 1.1.1, p. 12.

¹⁸ Ibid., 1.3.28, p. 209.

¹⁹ This contrasts with the Nyaya view of universals as real, eternal, and independent of their respective particulars, to which they are related by inherence.

²⁰ For a summary of these views, see Datta, Six Ways of Knowing, pp. 259-273.

²¹ B.S.B. 1.3.28, p. 209.

²² VP, Ch. 4, p. 94.

²³ B.S.B. 1.3.28, p. 209. Purva-Mimamsa deals with this objection by arguing that such words as Vasu refer to unique individuals who are eternal, and thus the connection between them and the words signifying them is eternal. They also contend that gods are birthless, deathless, and unembodied. This theory, however, is not acceptable to Sankara. On the evidence of various texts, he argues that gods are also embodied. It is only through extraordinary merit that this status is achieved, and it is, in fact, lost when this merit is exhausted. See ibid., 1.3.26-27, pp. 204-207.

²⁴ Ibid., 1.3.28, p. 213. The sphota doctrine was postulated by Indian grammarians to explain how the individual and separate syllables which comprise any word can reveal its meaning. The syllables are presented separately to the hearer. Corresponding to each word and sentence, there is a latent and indivisible sphota which conveys its meaning. As the sounds of a word are sequentially uttered, the corresponding sphota is progresively revealed. Since the sphota is conceived by its advocates to be eternal and indivisible, there is no question of a time sequence, and it can be grasped as a whole. The theory of sphota is rejected by Sankara.

²⁵ Ibid.

²⁶ Ibid.

²⁷ Ibid., p. 214.

²⁸ The doctrine of the eternity of the word and its connection with its referent is taken over by Sankara from the Purva-Mimamsa school. See M.S.J. 1.1.5-23, pp. 8-16. Sutras 6-11 detail several arguments against the eternity of the word. In sutras 12-17, Jaimini attempts to refute each argument individually and follows this, in sutras 18-23, with independent arguments for the eternal word. According to Jaimini, the momentary nature of the word is not due to its noneternity but is the result of the function of its manifesting agency. Utterance only manifests what is already existing. When we speak of the production of a word, we only indicate its utterance and not its creation anew. Changes in pronunciation are only indicative of changes in tone and not in the word itself. If the uttered word was transient, it would vanish immediately, and there would be no possibility of comprehension. We should be reminded that the intention behind the Mimamsa view is the justification of the Vedas as a defect-free source of knowledge. The view here seems to be that if the relationship between words and their meanings is fixed by human convention, like everything human it will be liable to error. The argument, therefore, is that this relationship is natural, eternal, and free from error.

²⁹ B.S.B. 1.3.29, p. 216.

³⁰ See Ganganatha Jha, The Purva-Mimamsa in Its Sources (Benares: Benares Hindu University, 1942), p. 153.

³¹ M.S.J. 1.1.27, p. 19.

³² Ibid., 1.1.30, p. 20.

³³ B.S.B. 1.3.28, p. 211.

³⁴ Ibid., pp. 209-10.

³⁵ Ibid., pp. 210-11.

³⁶ Ibid., 1.3.30, p. 217.

³⁷ Ibid., pp. 217-21.

³⁸ Ibid., pp. 218-19.

³⁹ Ibid., 4.1.4, p. 821. Sankara cites BR.U. 4.3.22: "The Vedas are no Vedas." See also V.P., Ch. 4, p. 113.

⁴⁰ The attitude of early Mimamsa writers about God is a matter of some controversy. It is sometimes argued that they did not discuss God because they were primarily concerned with establishing the Vedas as a self-evident, eternal source of knowledge and inquiring into ritual. Jaimini says nothing about God. Kumarila Bhatta, on the other hand, severely criticized theistic arguments for God and an omniscient person and seems to find the concept absurd. It is of interest to note a few of his arguments. It is not proper, he argues, to attribute the creation of a world that is full of pain and suffering to God. Suffering cannot be traced back to merit and demerit, which are nonexistent at the beginning of creation. Compassion cannot be the motive for creation, since there are no beings to whom compassion can be shown. Besides, on this view, the world would be made entirely happy, for there is nothing that could deter the compassionate activity. If it is deterred, God could not be omnipotent. Why should God create? If God's activity is purposeless, God is not an intelligent person. If God creates because of God's desire for sport (lila), God cannot be regarded as one who is complete. If the theist is concerned about finding a cause to explain the world process, karman can be regarded as a sufficient cause. Against the notion of an omniscient person, Kumarila argues that whether a person knows all can only be verified by someone who is omniscient. Logically, therefore, there should' be many omniscient persons. See Jha, The Purva-Mimamsa in Its Sources, pp. 47-52.

⁴¹ B.S.B. 1.2.2, p. 111.

⁴² Ibid., 1.1.3, pp. 18-19.

⁴³ "Those that are called the Rg-Veda (Yajur-Veda, etc.) are but the exhalation of this great Being" (BR.U. 2.4.10).

⁴⁴ BR.U.B. 2.4.10, p. 251.

⁴⁵ B.S.B. 1.3.28, p. 210. See also B.G.B. 15:15, p. 409. Here also he interprets authorship in the sense of initiating the regular succession of teaching.

⁴⁶ "The sacrificers, having acquired fitness to receive the Veda as a result of the earlier performance of good deeds, received it as it had already existed among the rsis" (Rg-Veda mantra, quoted in B.S.B. 1.3.29, p. 217). Or, "In the days of yore, the great rsis received through austerities, with the permission of the self-born One, the Veda, together with the anecdotes, that had remained withdrawn during dissolution" (ibid). See also 1.3.30, p. 219.

⁴⁷ V.P., Ch. 4, pp. 115-116.

⁴⁸ Vacaspati Misra flourished in the first half of the 9th century A.D. He occupies a very important place in the history of Advaita thought. His two most important works are Bhamati and Tattvasmiksa. Bhamati is a commentary on a portion of Sankara's commentary on the Brahmasutra, while Tattvasmiksa is a commentary on the Brahma-siddhi of Mandanamisra. The first work is supposed to have been named in honor of his wife.

⁴⁹Bhamati, 1.1.2, pp. 141-42.

⁵⁰ BR.U.B., Intro., 1.1, pp. 1-5.

⁵¹ B.S.B. 2.1.6, p. 314.

⁵² Ibid., 3.1.25, pp. 585-586.

⁵³Mimamsa is in full agreement with Sankara on this point, even though they do not agree, as we shall see, that the Vedas are also a pramana for brahman. The Mimamsa argument is that although dharma is an object of knowledge, it is not amenable to senseperception. Perception can only apprehend objects which are in existence at the time and are in contact with the organs. Dharma, however, is not in existence at the time of perception and has to be brought into being by certain acts. In addition, it has no external or tangible form and cannot be in contact with any of the sense organs. The other pramanas such as inference, presumption, etc., are more or less dependent on perception and are not, therefore, applicable. See M.S.J. 1.1.4, pp. 6-7. Also, Jha, The Purva-Mimamsa in Its Sources, pp. 175-176.

⁵⁴ BR.U.B., Intro., 1.1, pp. 1-2.

⁵⁵ Ibid., pp. 2-3.

⁵⁶ It is very important to note the specific sense in which Sankara is using the concepts of dharma and adharma in this discussion. Here it implies punya (merit) and papa (demerit) accruing particularly from the performance and nonperformance of recommended ritual activities. Actions are understood as having a twofold result: seen (drsta) and unseen (adrsta). Sankara's contention is that the unique relation between any action and its unseen result can be known only from the Vedas. The adrsta result is conceived of as a subtle, persisting impression that has the potency of bearing good or evil in the course of time.

⁵⁷ BR.U.B., Intro., 1.1, pp. 3-5. Sankara's attribution of an independent authoritative aim to the jnanakanda of the Vedas is perhaps the most important exegetical divergence from Purva-Mimamsa, whose views he adopts on so many other matters. Purva-Mimamsa considers only the injunctive statements of the Vedas to be authoritative. All other passages serve as auxiliaries to injunctions. We shall consider the details of this interesting controversy later. For further statements of Sankara on the limitations of the ritual portions of the Vedas, see B.G.B. 2:42-44, p. 61; MU.U.B. 1.2.12, pp. 109-111.

⁵⁸ BR.U.B. 2.3.6, p. 236.

⁵⁹ Ibid., 4.4.6, p. 504. "The ultimate aim of the Upanisads is to teach Self-knowledge" (BR.U.B. 3.5.1, p. 336).

⁶⁰ The terms brahman and atman are interchangeable here because of their identity.

⁶¹ B.S.B. 1.1.2, p. 17; 1.1.4, p. 22.

⁶² BR.U.B. 3.9.26, pp. 388-389.

⁶³ See above, Ch. 1.4.

⁶⁴ The other elements, the particular qualities which they manifest, and the organs evolved out of them are as follows:

Space	Sound	Ear
Air	Touch	Skin
Water	Taste	Tongue
Smell	Nose	Earth

⁶⁵ "One becomes freed from the jaws of death by knowing that which is soundless, touchless, colourless, undiminishing and also tasteless, eternal, odourless, without beginning and without end, distinct from Mahat, and ever constant" (KA.U. 1.3.15).

⁶⁶ B.G.B. 13:12, pp. 345-346. Also 2:25, p. 51.

⁶⁷ "This body, 0 son of Kunti, is called the Field (Kshetra); that which knoweth it is called the Knower of the Field (Kshetrajna)" (B.G. 13:1).

⁶⁸ Ibid., 13:17.

⁶⁹ BR.U.B. 2.4.14, p. 261. Also, KE.U.B. 2.1, pp. 59-60.

⁷⁰ "It being an established fact that the object and the subject, that are fit to be the contents on the concepts "I" and "it" (respectively) and are by nature contradictory as light and darkness, cannot logically have any identity, it follows that their attributes can have it still less. Accordingly, the superimposition of the object, referable through the concept "it" and its attributes on the subject that is conscious by nature and is referable through the concept "I" (should be impossible), and contrariwise the superimposition of the subject and its attributes on the object should be impossible" (B.S.B., Intro., p. 1).

⁷¹ AI.U.B. 2.1, pp. 48-49.

⁷² See above, Ch. 1.4.

⁷³ B.S.B. 2.1.6, p. 314.

⁷⁴ Ibid., 2.1.27, p. 355.

⁷⁵ Ibid., 2.1.11, p. 322.

⁷⁶ It should be pointed out that there is no twisting of the text here, for the compound does indeed offer both possibilities of meaning.

⁷⁷ B.S.B. 1.1.3, pp. 18-20.

⁷⁸ The possibilities and limitations of reason in relation to the acquisition of brahmajnana will be considered later. It must be mentioned that although Sankara dismisses perception, etc., as valid primary sources of the knowledge of brahman, this does not imply that they have absolutely no role in the process of gaining this knowledge. The subsidiary functions which they are assigned will become apparent as we proceed.

⁷⁹ See, for example, S.K. Mukherjee, "Sankara on the Limits of Empirical Knowledge," in Annals of the Bhandarkar Oriental Research Institute, 12 (1930-31), 68.

⁸⁰ KE.U. 2.1-3. For one of Sankara's finest discussions on this paradox, see his full commentary on Part 2 of this Upanisad.

⁸¹ BR.U.B. 3.6.1, p. 343.

⁸² Ibid., p. 344.

⁸³ KE.U.B. 1.3, p. 49. The word agama literally means "traditional knowledge." See also B.G.B. 18:50, p. 487.

⁸⁴ Radhakrishnan, Indian Philosophy, 2, 617.

⁸⁵ N.K. Devaraja, An Introduction to Sankara's Theory of Knowledge, p. 66.

⁸⁶ B.S.B. 1.1.1, p. 9. These views of Sankara are to be understood with reference to the theory of svatah-pramanya-vada. See above, Ch. 1.2.

⁸⁷ B.S.B. 2.1.1, p. 304.

⁸⁸ Ibid., 1.1.4, p. 23. The argument here is that the dependence of one pramana on another leads to infinite regression.

⁸⁹ Ibid., 2.1.4, p. 307. In support, he cites a well-known text, "I ask you of that Being who is to be known only from the Upanisads" (BR.U. 3.9.26).

⁹⁰ BR.U.B. 1.4.14, p. 123.

⁹¹ B.S.B. 2.3.6, p. 453.

⁹² B.G.B. 13:4, pp. 336-337.

⁹¹ BR.U.B. 1.3.1, p. 32. See also, B.S.B. 2.1.1, p. 302: "One cannot surmise the possibility of perceiving supersensuous things without the Vedas." "Vedic texts are the valid means to us in the matter of generating knowledge about supersensuous things" (B.S.B. 2.3.32, p. 945).

⁹⁴ BR.U.B. 1.4.7, p. 92.

⁹⁵ CH.U.B. 8.12.1, p. 475. This should be seen in the context of our discussion on the criterion of valid knowledge. See above, Ch. 1.1.

⁹⁶ Suresvara (ca. A.D. 800) is a direct disciple of Sankara and was installed by him in charge of the Math at Sringeri. Until very recently, though, that he was identical with Mandanamisra, the disciple of Kumarila Bhatta. It is now held, however, that Suresvara is closer in view to Sankara than Mandana. See E. Deutsch and J. A. B. van Buitenen, A Source Book of Advaita Vedanta (Honolulu: University Press of Hawaii, 1971), pp. 223-224. Suresvara's chief works are, Naiskarmya siddhi and Brhadaranyakopanisad-bhasya-varttika.

⁹⁷ See N.S. 3:34-38, pp. 168-170.

⁹⁸Bhamati, 1.1.4, pp. 157-160.

⁹⁹ B.S.B. 2.2.38, p. 436.

¹⁰⁰ For a summary of some of the Nyaya arguments about God, see Radhakrishnan, Indian Philosophy, 2, 165-173. Also, G. Chemparathy, An Indian Rational Theology (Vienna: De Nobili Research Library, 1972).

¹⁰¹ B.S.B. 1.1.2, p. 17. The Nyaya argument might have difficulty in proving that there is a single creator. On the analogy of common experience, one could argue that complex effects are generally produced by several agents acting in coordination.

¹⁰² See ibid., 2.2.37-41, pp. 434-438. We may notice here the parallels in argument with Kumarila Bhatta.

¹⁰³ Ibid., 2.2.38, pp. 435-436.

¹⁰⁴ Ibid., 2.1.31, pp. 359-360.

¹⁰⁵ Perhaps the best examples of this are to be found in those parts of his Brahma-sutra commentary where he sets out to refute the Sankhya doctrine of the evolution of the world from insentient matter. For example,

It is not seen in this world that any independent insentient thing that is not guided by some sentient being can produce modifications to serve some special purpose of man; for what is noticed in the world is that houses, palaces, beds, seats, recreation grounds, etc., are made by the intelligent engineers and others at the proper time and in a way suitable for ensuring or avoiding comfort or discomfort. So how can insentient Pradhana create this universe, which cannot even be mentally conceived of by the intelligent and most farfamed architects, which is seen in the external context to consist of the earth, etc., that are fit places for experiencing the results of various works, and in the context of the individual person, of the body and other things having different castes, etc., in which the limbs are arranged according to a regular design, and which are the seats for experiencing various fruits of actions? (B.S.B. 2.2.1, p. 369).

¹⁰⁶ See above, Introduction.

¹⁰⁷ Radhakrishnan, Indian Philosophy, 2, 518. Radhakrishnan's views are contrary to the main argument of Sankara for justifying the Vedas. According to Sankara, the Vedas as sabda-pramana are necessary because the knowledge which they afford is not available through any other means. If this knowledge was available through human faculties, Sankara's argument would not stand.

¹⁰⁸ The term includes texts like the Manu Smrti, the Bhagavadgita, the Puranas, and the Mahabharata. It is also used with reference to the works of other schools such as Nyaya, Vaisesika, and Yoga.

¹⁰⁹ B.S.B. 1.3.28, p. 210.

¹¹⁰ Ibid., 2.1.1, p. 304.

¹¹¹ Ibid., pp. 303-304.

¹¹² Ibid., 2.1.3, pp. 306-307.

¹¹³ B.G.B. 18:19, p. 461.

¹¹⁴ B.S.B. 1.2.25, p. 149. Sankara's conclusions on the respective authority of sruti and smrti are derived from Purva-Mimamsa writers who have discussed this matter in interesting detail. See M.S.J. 1.3.1-6, pp. 55-68. With regard to the smrtis composed by Manu and others, Kumarila has proposed five alternatives: (1) That the authors of these texts were entirely mistaken about what they wrote. (2) That their assertions were derived from personal observation. (3) That they learnt about what they wrote from others. (4) That they intentionally made wrong statements to mislead others. (5) That their assertions are based on Vedic injunctions. He advances various arguments for the rejection of all alternatives, except the last. For most of the smrti injunctions, Kumarila says, corroborative Vedic texts are easily found. In the case of texts for which no such corroboration can be found, we must presume that such Vedic texts were known to the smrti compilers but are now lost along with many others. The basis of this presumption is the fact that the compilers of the smrtis had also learned and studied the Vedas. Kumarila, however, does not accept all smrti literature to be equally authoritative. Only those parts of the smrtis which are concerned with dharma have their origin directly in the Vedas. Those that relate to pleasure and pain as experienced in the world are derived from direct perception. Stories, which are encountered from time to time, are meant for praising dharma and condemning adharma. See Jha, Purva-Mimamsa in Its Sources, pp. 214-218.

¹¹⁵ B.G.B. 18:66, p. 513.

¹¹⁶ Ibid. Also, B.S.B. 2.1.12, p. 324.

¹¹⁷ PR.U.B. 6.2, p. 490.

¹¹⁸ BR.U.B. 2.1.20, p. 209. Purva-Mimamsa writers have also considered the question of whether Vedic words and their denotations are the same as those in common use. They have concluded that the words must be the same if Vedic injunctions are to be understood and meaningful. See M.S.J. 1.3.30, p. 91. In fact, Purva-Mimamsa accepts that there are cases in which the meaning of the Vedic word may have to be sought among non-Aryan people. Such a situation arises if the word used by a non-Aryan is exactly the same as used in the Vedas but unknown to the Aryan vocabulary. See M.S.J. 1.3.10, p. 74.

¹¹⁹ BR.U.B. 2.1.20, p. 209, 217.

¹²⁰ See N.S. 3:84-86, pp. 207-208.

¹²¹ Ibid., p. 208.

¹²² BR.U.B. 1.4.7, pp. 81-82. Also N.S. 3:44-45, pp. 173-174.

¹²³ BR.U.B. 2.1.20, pp. 218-219. The general view of Advaita on this point is that the fields of perception and sruti are different. Perception is concerned with the empirical world while sruti discloses absolute reality. Sruti does not deny the empirical validity of perception.

¹²⁴ "As for the argument that creation after deliberation is seen in the world only in cases of such efficient causes as the potter and others, but not in the case of materials, that is being answered. Any argument from common sense is not applicable here; for this is not truth to be arrived at from the Vedas (alone); its meaning should conform to Vedic statements" (B.S.B. 1.4.27, p. 296).

¹²⁵ BR.U.B. 3.3.1, pp. 318-319.

¹²⁶ B.S.B. 1.4.14, pp. 272-273.

¹²⁷ Ibid., 1.3.7, p. 166.

¹²⁸ B.G. 3:17-18.

¹²⁹ AI.U.B., Intro., p. 8. Also, B.G.B. 2:46, 69. The fact that the brahmajnani transcends the necessity for the sruti does not in any way detract from its indispensability as a pramana of brahman. The point is that a pramana, having successfully given birth to knowledge, is no longer needed for that purpose. Its value is not thereby reduced, nor does it suggest that knowledge is otherwise attainable.

Abbreviations

AI.U: Aitareya Upanisad

B.G.: Bhagavadgita

Bhamati: Bhamati of Vacaspati Misra

B.S.: Brahma-sutra

BR.U.: Brhadaranyaka Upanisad

CH.U.: Chandogya Upanisad

IS.U.: Isa Upanisad

KA.U.: Katha Upanisad

KE.U.: Kena Upanisad

MA.U.: Mandukya Upanisad

MA.U.K.: Mandukya Upanisad Karika of Gaudapada

M.S.J.: Purva-Mimamsa Sutras of Jaimini

MU.U.: Mundaka Upanisad

N.S.: Naiskarmyasiddhi of Suresvara

N.Y.S.G.: Nyaya Sutras of Gotama

PR.U.: Prasna Upanisad

SV.U.: Svetasvatara Upanisad

TA.U.: Taittiriya Upanisad

T.B.: Tattvabodha of Sankra

V.P.: Vedanta Paribhasa of Dharmaraja

V.S.: Vedantasara of Sadananda

The letter B added to the abbreviation of any text (as BR.U.B.) indicates the commentary (bhasya) of Sankara on the said text. Thus, B.G.B. means Sankara's bhasya on the Bhagavadgita.