The Question of the Importance of Samadhi

In Modern and Classical Advaita Vedanta.

Michael Comans, PhD

Philosophy East & West

Vol.43 No 1


Jan. 1993

Copyright by University of Hawaii Press


The word SAMADHI [1] became a part of the vocabulary of a number of Western intellectuals toward the end of the first half of this century. Two well-known writers, Aldous Huxley and Christopher Isherwood, were impressed by Eastern and specifically by Indian thought. Huxley made a popular anthology of Eastern and Western mystical literature under the title "The Perennial Philosophy" (1946), and in his last novel, "Island" (1962), words such as moksa and samadhi occur untranslated. In both these works, Huxley uses the words "false samadhi, " implying that the reader was already conversant with what samadhi actually is. Conveniently glossing over any emphisis regarding the Hindu goddess Kali-Ma, Isherwood wrote an account of the life of the nineteenth-century Bengali mystic Sri Ramakrishna Paramahamsa and his disciples (1959), and he published as the second part of his autobiographical trilogy an account of the years he spent with his own guru, Swami Prabhavananda of the Ramakrsna Order, in "My Guru and His Disciple" (1980). Why these writers were drawn toward Eastern spiritual thought, and to the Vedanta teachings in particular, is not the subject for discussion here. But perhaps one significant reason is that with the decline in organized religion after World War I, these writers found in the Vedanta, as presented to them by the followers of Sri Ramakrsna and his disciple Swami Vivekananda, a spirituality which emphasized the authority of firsthand experience as the only way to verify what was presented as the Truth. The Vedanta, as they saw it, was a "minimum working hypothesis," which could be validated through cultivating a certain type of experience, and that experience was seen to be a mystical, super-conscious state of awareness called samadhi.

Isherwood edited a book of articles titled "Vedanta for the Western World" (1948). In his introduction he emphasizes the centrality of having a direct, personal experience of Reality, which, he says, the Christian writers call "mystic union" and Vedantists call "samadhi." Isherwood raises the question as to how Reality can be experienced if it is beyond sense perception, and he answers the question in terms of samadhi experience:

Samadhi is said to be a fourth kind of consciousness: it is beyond the states of waking, dreaming and dreamless sleep. Those who have witnessed it as an external phenomenon report that the experience appeared to have fallen into a kind of trance. The hair on the head and body stood erect. The half-closed eyes became fixed. Sometimes there was an astonishing loss of weight, or even levitation of the body from the ground. But these are mere symptoms, and tell us nothing. There is only one way to find out what samadhi is like: you must have it yourself?

Huxley and Isherwood did not find Indian spirituality by journeying to India--rather it was India which found them; and the variety of Indian spirituality with which these Englishmen came into contact in California in the late 1930s was that of the Vedanta Society, founded by Swami Vivekananda and his followers, who were monks of the recently established (1886) Order of Ramakrishna. If we seek to locate the source of the orientation of spiritual life around the cultivation of samadhi experience, which has become one of the principal characteristics of modern Vedanta, it must be traced to Sri Ramakrsna himself. Ramakrsna was not a Vedantin in the orthodox sense of one who has received instruction centered on the exegesis of the sacred texts (sastra), which are generally in Sanskrit, from a teacher (acarya), and who then consciously locates himself within that specific body of received teachings (sampradaya). Ramakrsna, as is well known, affirmed that a variety of diverse disciplines and traditions within Hinduism, and even outside of Hinduism, were valid in that they were all efficacious means toward the same spiritual goal. However, as has been pointed out, it would be most correct to locate. Ramakrsna's teachings within a Tantric paradigm. [3] (The time spent under the direction of Totapuri, who was said to be an Advaitin, was much shorter than the time spent studying Tantra, and the information available on Totapuri is very meager, so it is difficult to be sure whether he was actually an Advaitin rather than a follower of yoga) Tantra is an expressly experience-oriented discipline and it relies upon yoga techniques, particularly those of Hatha Yoga, [4] to bring about a samadhi experience. Ramakrsna frequently underwent trance-like states, which are referred to in The Gospel of Sri Ramakrishna as samadhi experiences. A typical description in the Gospel would be the following passage:

At the mere mention of Krishna and Arjuna the Master went into samadhi. In the twinkling of an eye his body became motionless and his eyeballs transfixed, while his breathing could scarcely be noticed. [5]

Ramakrsna has himself linked the occurrence of samadhi with Kundalini Yoga, which is referred to in the treatises on Hatha Yoga and is fundamental to Tantra soteriology. For example, Ramakrsna is recorded as having remarked:

A man's spiritual consciousness is not awakened unless his Kundalini is aroused. The Kundalini dwells in the Muladhara. When it is aroused, it passes along the Sushumna nerve, goes through the centres of Svadhisthana, Manipura, and so on, and at last reaches the head. This is called the movement of the Mahavayu, the Spiritual Current. It culminates in samadhi. [6]

From the above we should be able to see the importance that the samadhi experience had in the life and teachings of Sri Ramakrsna. Such an experience-oriented view of spirituality was a legacy which passed from Ramakrsna to Vivekananda. Vivekananda was receptive to this view, for it seemed to agree with what he had studied of the British empiricist philosophers and the positivist Auguste Comte, insofar as they by University of Hawaii Press had stressed the centrality of empirical experience. Vivekananda extended the empiricist epistemology that all knowledge is derived from sense experience into the domain of metaphysics, for he thought that since experience is the basis of all knowledge, then if a metaphysical Reality exists, it, too, ought to be available for direct experience? And from his association with Ramakrsna he gathered that samadhi was the experience required in order to know God. In his writings he placed much emphasis on the necessity of attaining samadhi. He loosely translated samadhi as "super-consciousness,"[8] and he stated in his work Raja-Yoga, a commentary in English on the Yogasutras of Patanjali, that samadhi experience was the acme of spiritual life:

Samadhi is the property of every human being--nay, every animal. From the lowest animal to the highest angel, some time or other, each one will have to come to that state, and then, and then alone, will real religion begin for him. Until then we only struggle towards that stage. There is no difference now between us and those who have no religion, because we have no experience. What is concentration good for, save to bring us to that experience? Each one of the steps to attain samadhi has been reasoned out, properly adjusted, scientifically organized, and, when faithfully practised, will surely lead us to the desired end. Then all sorrows cease, all miseries vanish; the seeds of actions will be burnt, and the soul will be free for ever.[9]

Vivekananda was attracted to Ramakrsna for reasons somewhat similar to those that initially attracted Huxley and Isherwood to the Vedanta taught by the followers of Vivekananda: they all sought some direct, experiential verification of the propositions of religious metaphysics, and they all came to believe that the key to such verification lay in the attainment of a samadhi or "super-conscious" experience. This legacy of Ramakrsna, the search for an extra-ordinary experience in order to validate spiritual life, not only extended to the West via the Ramakrsna Order of monks that Vivekananda helped to found, but it also become a dominant view within the Western-educated Indian middle class through the spread of Ramakrsna-Vivekananda literature. The modern Indian philosopher, Sarvepalli Radhakrishnan, an eloquent advocate of the importance of experience in religion, has described samadhi in the following manner: "In samadhi or enstatic consciousness we have a sense of immediate contact with ultimate reality.. It is a state of pure apprehension.. "[10]

At this point the reader may wonder whether we are not stating the obvious, for is it not precisely because samadhi is so important that modern Vedantins such as Vivekananda and Radhakrishnan gave it such emphasis? It is certainly important to modern Vedanta, but the question can be legitimately raised as to what importance it has in the Upanisads, the very source of the Vedanta, and in the classical Vedanta such as in the works of Sankara, the most famous of all the Vedanta teachers. That is the topic which we shall now address.

The first point to be noted is that the word samadhi does not occur in the ten major Upanisads upon which Sankara has commented.[11] This is not a matter to be lightly passed over, for if the attainment of samadhi is central to the experiential verification of the Vedanta, as we can gather it is, judging by the statements of some modern Vedantins such as those cited above, then one would legitimately expect the term to appear in the major Upanisads which are the very source of the Vedanta. Yet the word does not occur. The closest approximation to the word samadhi in the early Upanisads is the past passive participle samahita in the Chandogya and Brhadaranyaka Upanisads.[12] In both texts the word samahita is not used in the technical meaning of samadhi ,that is, in the sense of a meditative absorption or enstasis ,although the closest approximation to this sense occurs in the Brhadaranyaka. In the first reference (BU 4.2.1) , Yajnavalkya tells Janaka: "You have fully equipped your mind (samahitatma) with so many secret names [of Brahman, that is, Upanisads]."[13] Here the word samahita should be translated as "concentrated, collected, brought together, or composed."

In the second occurrence (BU 4.4.23), Yajnavalkya tells Janaka that a knower of Brahman becomes "calm (santa), controlled (danta), withdrawn from sense pleasures (uparati), forbearing (titiksu), and collected in mind (samahita). This reference to samahita is the closest approximation in the Upanisads to the term samadhi, which is well known in the later yoga literature. However, the two terms are not synonyms, for in the Upanisad the word samahita means "collectedness of mind," and there is no reference to a meditation practice leading to the suspension of the faculties such as we find in the literature dealing with yoga. The five mental qualities mentioned in BU 4.4.3 later formed, with the addition of faith (sraddha), a list of six qualifications required of a Vedantic student, and they are frequently to be found at the beginning of Vedantic texts.[14] In these texts, the past participles used in the Upanisads are regularly changed into nominal forms: santa becomes sama, danta becomes dama, and samahita becomes samadhana, but not the cognate noun samadhi. It would thus appear that, while Vedanta authors understood samahita and samadhana as equivalent terms, they did not wish to equate them with the word samadhi; otherwise there would have been no reason why that term could not have been used instead of samadhana. But it seems to have been deliberately avoided, except in the case of the later Vedanta work, Vedantasara, to which we shall have occasion to refer. Thus we would suggest that, in the Vedanta texts, samadhana does not have the same meaning that the word samadhi has in yoga texts. This is borne out when we look at how Vedanta authors describe the terms samahita and samadhana. Sankara, in BU 4.2.1, glosses samahitatma as samyuktama, "well equipped or connected." In BU 4.4.23, he explains the term samahita as "becoming one-pointed (aikagrya) through dissociation from the movements of the sense-organs and the mind."[15] The term occurs again in the Katha Upanisad 1.2.24 in the negative form asamahita, which Sankara glosses as "one whose mind is not one-pointed (anekagra), whose mind is scattered."[16] In introductory Vedanta manuals, samadhana is also explained by the term "one-pointed" (ekagra).[17] The word samadhana can thus be understood as having the meaning of "one-pointed" (ekagra). In the Yogasutra, "one-pointed" (ekagra) is used to define concentration (dharana),[18] which is the sixth of the eight limbs of Yoga and a preliminary discipline to dhyana and samadhi. We may see, then, that the Vedantic samadhana means "one-pointedness" and would be equivalent to the yoga dharana, but it is not equivalent to the yoga samadhi.

The word samadhi first appears in the Hindu scriptures in the Maitrayni Upanisad (6.18, 34), a text which does not belong to the strata of the early Upanisads[19] and which mentions five of the eight limbs of classical Yoga. The word also occurs in some of the Yoga and Sannyasa Upanisads of the Atharvaveda.[20] Samadhi would thus seem to be a part of yogic practice which has entered into the later Upanisadic literature through such texts as the Yoga Upanisads as a result of what Eliade calls "the constant osmosis between the Upanisadic and yogic milieus."[21] The diverse teachings of yoga were systematized in Patanjali's Yogasutras, where it is explained that the goal of yoga is to restrain completely all mental fluctuations (vrtti) so as to bring about the state of samadhi. Samadhi itself has two stages, samprajana samadhi, or an enstasis where there is still object-consciousness, and asamprajatasamadhi or nirbijasamadhi, where there is no longer any object-consciousness. Asamprajnatasamadhi became known in later Vedanta circles as Nirvikalpa Samadhi.[22] The point to be noted about yoga is that its whole soteriology is based upon the suppression of mental fluctuations so as to pass firstly into samprajnatasamadhi and from there, through the complete suppression of all mental fluctuations, into asamprajnatasamadhi, in which state the Self remains solely in and as itself without being hidden by external, conditioning factors imposed by the mind (citta).

When we examine the works of Sankara, however, we find a very sparing use of the word samadhi.[23] In the Brahmasutrabhasya he makes three references to samadhi as a condition of absorption or enstasis.[24] In the first (2.1.9) , he implicitly refutes the idea that samadhi is, of itself, the means for liberation, for he says:

Though there is the natural eradication of difference in deep sleep and in samadhi etc., because false knowledge has not been removed, differences occur once again upon waking just like before.[25]

What Sankara says is that duality, such as the fundamental distinction between subject and object, is obliterated in deep sleep and in samadhi, as well as in other conditions such as fainting, but duality is only temporarily obliterated for it reappears when one awakes from sleep or regains consciousness after fainting, and it also reappears when the yoga arises from samadhi. The reason why duality persists is because false knowledge (mithyajana) has not been removed. It is evident from this brief statement that Sankara does not consider the attainment of samadhi to be a sufficient cause to eradicate false knowledge, and, according to Sankara, since false knowledge is the cause of bondage, samadhi cannot therefore be the cause of liberation. The only other significant reference to samadhi in the Brahmasatrabhasya occurs in the context of a discussion as to whether agentship is an essential property of the self. According to Sankara's interpretation, sutras 2.3.33-39 accept agentship as a property of the self, but sutra 2.3.40 presents the definitive view that agentship is not an intrinsic property of the Self but is a superimposition. The word samadhi occurs in 2.3.39 (samadhy-abhavacca), and here Sankara briefly comments, "samadhi, whose purpose is the ascertainment of the Self known from the Upanisads, is taught in the Vedanta texts such as: 'The Self, my dear, should be seen; it should be heard about, thought about and meditated upon'" (BU 2.4.5) .[26] Sankara shows by the phrase atmapratipattiprayojana ("whose purpose is the ascertainment of the Self") that he acknowledges that the practice of samadhi has a role in Vedanta. However, these two references do not in themselves present a conclusive picture of Sankara's thought, for in the first reference it is evident that he does not consider samadhi to be a sufficient means for liberation, while in the second he has clearly given it a more positive place as a means for liberation. This second reference, however, has to be treated with some circumspection as it forms the comment upon a sutra which Sankara does not consider to present the definitive view. Another reference to samadhi, where it again seems to have a more positive value, occurs in the commentary upon the Mandukyakarika of Gaudapada, where in verse 3.37 the word samadhi is given as a synonym for the Self. Sankara glosses the word samadhi in two different ways, and in the first he says "samadhi = because [the Self] can be known through the wisdom arising from samadhi."[27] Thus we can see that, according to Sankara, samadhi has a role to play in Vedanta, but yet the first reference (2.1.9) indicates that this role is perhaps more circumscribed than the modern exponents of Vedanta would have us believe. We will attempt to resolve the matter through a wider examination of Sankara's thought, particularly in regard to his use of yoga.

The first specific mention of yoga is in the Katha Upanisad, and there is a verse in this Upanisad which details a type of yoga meditation:

The discriminating person should restrain speech in the mind, he should restrain the mind in the cognizing self, he should restrain the cognizing self in the 'great self' and restrain that 'great self' in the peaceful Self.[26]

Sankara introduces this verse with the comment that the Upanisad here presents "a means for the ascertainment of that [Self]."[29] In his commentary upon Brahmasutra 1.4.1, Sankara refers to this Katha verse with the remark that the sruti"shows yoga as the means for the apprehension of the Self."[30] In his commentary upon Brahmasutra 3.3.15, he again refers to this verse when he says that it is "just for the sake of the clear understanding of the Self that the sruti enjoins meditation, viz. 'the discriminating person should restrain speech in the mind.... "[31] It is therefore evident that Sankara considers the verse above to present a method of yoga meditation leading to Self-knowledge. As to his understanding of this Katha verse, he has explained it succinctly in his commentary on Brahmasutra 1.4.1:

This is what is said. 'He should restrain speech in the mind' means that by giving up the operations of the extemal senses such as the organ of speech and so forth he should remain only as the mind. And since the mind is inclined towards conjecturing about things, he should, by way of seeing the defect involved in conjecturing restrain it in the intellect whose characteristic consists in determining and which is said here by the word 'cognizing self'. Then bringing about an increase in subtlety, he should restrain that intellect in the 'great self', i.e. the experience, or the one-pointed intellect. And he should establish the 'great self' in the peaceful Self, i.e. in that supreme Purusa who is the topic under consideration, who is the 'highest goal'.[32]

Aranyaka Upanisad 2.4.11, which forms part of the well known Yajanavalkya-Maitreyi dialogue, Sankara briefly describes a method of contemplation which is similar to the one mentioned in the Katha 1.3.13. It is as follows:

[text] the skin is the one goal of all kinds of touch [commentary] such as soft or hard, rough or smooth.... By the word 'skin', touch in general that is perceived by the skin, is meant; in it different kinds of touch are merged, like different kinds of water in the ocean, and become nonentities without it, for they were merely its modifications. Similarly, that touch in general, denoted by the word 'skin', is merged in the deliberation of the Manas [mind], that is to say, in a general consideration by it, just as different kinds of touch are included in touch in general perceived by the skin; without this consideration by the Manas it becomes a non-entity. The consideration by the Manas also is merged in a general cognition by the intellect, and becomes non-existent without it. Becoming mere consciousness, it is merged in Pure Intelligence, the Supreme Brahman, like different kinds of water in the ocean. When, through these successive steps, sound and the rest, together with their receiving organs, are merged in Pure Intelligence, there are no more limiting adjuncts, and only Brahman, which is Pure Intelligence, comparable to a lump of salt, homogeneous, infinite, boundless and without a break, remains. Therefore the Self alone must be regarded as one without a second.[33]

We can see that the type of yoga which Sankara presents here is a method of merging, as it were, the particular (visesa) into the general (samanya). For example, diverse sounds are merged in the sense of hearing, which has greater generality insofar as the sense of hearing is the locus of all sounds. The sense of hearing is merged into the mind, whose nature consists of thinking about things, and the mind is in turn merged into the intellect, which Sankara then says is made into 'mere cognition' (vijanamatra); that is, all particular cognitions resolve into their universal, which is cognition as such, thought without any particular object. And that in turn is merged into its universal, mere Consciousness (prajnana-ghana), upon which everything previously referred to ultimately depends. There are two points which ought to be noted concerning Sankara's presentation of yoga which differ from the model we find in Patanjali's Yogasutra. The first concerns method. Sankara does not say that all thought forms must be restrained in the manner of the cittavrttinirodha of the Yogasutras. While in other places Sankara has mentioned that meditation involves the withdrawal of the mind from sense objects,[34] he has also made it clear that control of the mind (cittavrttinirodha) is "not known as a means of liberation."[35] Rather, Sankara's method involves thinking, although it is thinking of a certain type, leading from the involvement in particulars to a contemplation of what is more general and finally to the contemplation of what is most general, that is, Consciousness. Thus Sankara's method of yoga is a meditative exercise of withdrawal from the particular and identification with the universal, leading to contemplation of oneself as the most universal, namely, Consciousness. This approach is different from the classical Yoga of complete thought suppression.

The second point is one of approach, for nowhere does Sankara present the Atman-Brahman as a goal to be reached. On the contrary, his approach is that the Atman-Brahman is not something to be acquired since it is one's own nature, and one's own nature is not something that can be attained. This approach has its corollary in his method of negation: the removal of superimpositions in order to discover what is already there, although concealed as it were by all sorts of false identifications based ultimately upon the ignorance of who we really are. Such an approach is different from that of the classical Yoga of the Yogasutras, where a goal is presented in terms of nirvikalpasamadhi, which one has to achieve in order to gain liberation. That Sankara's method is one of negation in order to "reveal the ever revealed" is evident throughout his whole discussion of the role of action in the matter of liberation. In Brahmasutra 1.1.4, an opponent argues that the role of scripture is injunctive--it is to enjoin a person either to do something or to refrain from doing something--and the role of the Upanisads, too, after presenting the nature of Brahman, is to enjoin meditation upon Brahman as a means of release.[36] Sankara replies that if liberation is to be gained as a result of an action, then liberation must be impermanent. He specifies that actions can only be of four kinds: an action can produce something, or it can modify a thing, or it can be used to obtain something or to purify it.[37] He takes up each action in turn and argues that liberation is not something that can be either produced, attained, modified, or purified by any action whether physical, oral, or mental. His main argument is that if liberation is an effect of some kind of action, then liberation would have a beginning and would be time-bound and hence noneternal, and that such a consequence would go against the whole tradition that teaches that liberation is eternal. Sankara's view is that liberation is nothing but being Brahman, and that is one's inherent condition, although it is obscured by ignorance. He says that the whole purpose of the Upanisads is just to remove duality, which is a construct of ignorance;[38] there is no further need to produce oneness with Brahman, because that already exists. Sankara's frequent use of the phrase "na heya naupadeya" (cannot be rejected or accepted)[39] along with the word Atman indicates that the Self cannot be made the object of any kind of action whatsoever. Sankara has summarized all this in his commentary on the Brhadaranyaka:

... liberation is not something that can be brought into being. For liberation is just the destruction of bondage, it is not the result of an action. And we have already said that bondage is ignorance and it is not possible that ignorance can be destroyed by action. And action has its capacity in some visible sphere. Action has its capacity in the sphere of production, attainment, modification and purification. Action is able to produce, to make one attain, to modify or to purify. The capacity of an action has no other scope than this, for in the world it is not known to have any other capacity. And liberation is not one of these. We have already said that it is hidden merely by ignorance.[40]

Thus we can see that the perspective of Sankara is fundamentally different from that of the yoga tradition where, although the purusa is presented as not something to be acquired, liberation is nonetheless a real goal to be attained through a process of mental discipline, which necessitates the complete suppression of all mental activity.

That there is a certain ambivalence toward yoga on the part of the followers of Vedanta can be seen in Brahmasutra 2.1.3, "Thereby the Yoga is refuted," which offers a rejection of yoga following upon the rejection of Sankhya philosophy. The problem as Sankara sees it is that yoga practices are found in the Upanisads themselves, so the question arises as to what it is about yoga that needs to be rejected. Sankara says that the refutation of yoga has to do with its claim to be a means of liberation independent from the Vedic revelation. He says, "... the sruti rejects the view that there is another means for liberation apart from the knowledge of the oneness of the Self which is revealed in the Veda."[41] He then makes the point that "the followers of Sankhya and Yoga are dualists, they do not see the oneness of the Self."[42] The point that "the followers of Yoga are dualists" is an interesting one, for if the yogins are dualists even while they are exponents of asamprajnatasamadhi (nirvi-kalpasamadhi), then such samadhi does not of itself give rise to the knowledge of oneness as the modem exponents of Vedanta would have us believe. For if it did, then it would not have been possible for the yogins to be considered dualists. Clearly the modem Vedantins, in their expectation that samadhi is the key to the liberating oneness, have revalued the word and have given it a meaning which it does not bear in the yoga texts. And, we suggest, they have given it an importance which it does not possess in the classical Vedanta, as we are able to discerm it in the writings of Sankara.

The matter to be decided is what place samadhi, and yoga in general, holds in Sankara's thought. We suggest that his commentary upon the Bhagavadgita contains certain programmatic statements that are of general assistance in determining his views on the place of samadhim and yoga in the Advaita scheme of liberation. In the Gita, Sankara very frequently glosses the word yoga when it occurs in a verse by the word samadhi, thereby indicating that on many occasions he understands yoga to mean the practice of a certain discipline wherein samadhi is the key factor, as in verse 6.19, "...for one who engages in yoga concerning the Self" (yunjato yogam atmanah), which Sankara glosses as "practices samadhi concerning the Self" (atmanah samadhim anutisthatah).[43] It is evident that he considers samadhi as a state wherein normal distinctions are obliterated, as is evident from his statement in 18.66, "the evils of agent-ship and enjoyership etc. are not apprehended in deep sleep or in samadhi etc. where there is discontinuation of the flow of the erroneous idea that the Self is identical to the body."[44] Here, as in his commentary upon Brahmasutra 2.1.9, Sankara links deep sleep and samadhi, and it is evident that he recognizes samadhi to be a state wherein distinctions are temporarily resolved, as they are in deep sleep.

At the beginning of his commentary upon the Gita, Sankara makes a significant statement concerning the relation of Sankhya to Yoga.[45] He says that Sankhya means ascertaining the truth about the Self as it really is and that Krsna has done this in his teaching from verses 2.11 up until 2.31. He says that sankhyabuddhi is the understanding which arises from ascertaining the meaning in its context, and it consists in the understanding that the Self is not an agent of action because the Self is free from the sixfold modifications beginning with coming into being. He states that those people to whom such an understanding becomes natural are called Sankhyas. He then says that Yoga is prior to the rise of the understanding above. Yoga consists of performing disciplines (sadhana) that lead to liberation; it presupposes the discrimination between virtue and its opposite, and it depends upon the idea that the Self is other than the body and that it is an agent and an enjoyer. Such an understanding is yogabuddhi, and the people who have such an understanding are called Yogins. From this it is clear that Sankara relegates Yoga to the sphere of ignorance (avidya) because the Yogins are those who, unlike the Sankhyas, take the Self to be an agent and an enjoyer while it is really neither. They are, therefore, in Sankara's eyes, not yet knowers of the truth.

Sankara again clearly demarcates Sankhya and Yoga in his comments on verse 2.39, where Krsna says, "O Partha, this understanding about Sankhya has been imparted to you. Now listen to this understanding about Yoga.... "According to Sankara, 'Sankhya' means the "discrimination concerning ultimate truth, " and the 'understanding' pertaining to Sankhya means a "knowledge which is the direct cause for the termination of the defect which brings about samsara consisting of sorrow and delusion and so forth." He then says that Yoga is the "means to that knowledge" (tatpraptyupaya) and that Yoga consists of both (a) karmayoga, that is, performing rites and duties as an offering to the Lord once there has been a relinquishment of opposites (such as like and dislike) through detachment, and (b) samadhiyoga.[46] In 4.38, Sankara again explains the word yoga occurring in the verse as referring to both karmayoga and samadhiyoga.[47] It is evident that Sankara understands the word yoga in the Gita to refer to both karmayoga and to the practice of meditation, that is, samadhiyoga. It is also evident that he considers yoga to be a means leading to Sankhya-knowledge but that it is not the same as Sankhya-knowledge. In 6.20, Sankara says that one apprehends the Self by means of a "mind which has been purified through samadhi."[48]

From the evidence of the above we suggest that according to Sankara the role of samadhi is supportive--or purifying--and is preliminary to, but not necessarily identical with, the rise of the liberating knowledge. As is well known, Sankara considers that knowledge alone, the insight concerning the truth of things, is what liberates. To this end he places great emphasis upon words, specifically the words of the Upanisads, as providing the necessary and even the sufficient means to engender this liberating knowledge. Sankara repeatedly emphasizes the importance of the role of the teacher (guru/acarya) and the sacred texts (sastra) in the matter of liberation. For example the compound sastracaryopadesa, "the instruction on the part of the teacher and the scriptures," occurs seven times in his commentary on the Gita alone, along with other variations such as vedantacaryopadesa, and it regularly occurs in his other works as well.[49] The modem Vedantin, on the other hand, has overlooked, possibly unknowingly, the importance which sacred language and instruction held in the classical Vedanta as a means of knowledge (pramana) and has had to compensate for this by increasing the importance of yogic samadhi which is then put forward to be the necessary and sufficient condition for liberation.

The contrast between the Vedanta of Sankara and some of its modem exponents is clear enough. But it should not be thought that the modem emphasis on yogic samadhi is without precedent. As we have mentioned, there is evidence of yoga techniques in the principal Upanisads themselves although it did not then have a dominant emphasis, and this is reflected in the approach of Sankara in his commentaries. However, in the centuries following Sankara, Advaitins have exhibited a gradual increase in their reliance upon yoga techniques. This can be shown by examining a few of the Advaita Prakaranagranthas, noncom-mentarial compositions by Advaita authors.

The only noncommentarial work that is widely accepted as the composition of Sankara is the Upadesasahasri. In this work the word samadhi rarely occurs. The word samahita is used in 13.25, and we have previously argued that samahita (concentrated) has a meaning equivalent to the word samadhana, one-pointedness of mind, but it does not have the same meaning as nirvikalpasamadhi.[50] Sankara mentions samadhi three times in the Upadeaasahasra,[51] but he does not extol it; on the contrary, speaking from the understanding that the Self is nirvikalpa by nature, he contrasts the Self and the mind and says:

As I have no restlessness (viksepa) I have hence no absorption (samadhi). Restlessness or absorption belong to the mind which is changeable.[52]

A similar view is expressed in 13.17 and 14.35. In 15.14 Sankara presents a critique of meditation as an essentially dualistically structured activity.[53] Furthermore, in 16.39-40, Sankara implicitly criticizes the Sankhya-Yoga view that liberation is dissociation from the association of purusa and prakrti,[54] when he says:

It is not at all reasonable that liberation is either a connection [with Brahman] or a dissociation [from prakrti]. For an association is non-eternal and the same is true for dissociation also.[55]

Thus it is evident from the above that Sankara implicitly rejects both the soteriology of yoga, namely, that liberation has to be accomplished through the real dissociation of the purusa from prakrti, and the pursuit towards that end, that is, the achievement of nirvikalpa or asamprajatasamadhi.

However such a view became blurred in the writings of post-Sankara Advaitins. This can be briefly shown by examining some later Advaita prakarana texts. For example, in the popular fourteenth-century text Pancadasi, we find a mixture of Vedantic and Yogic ideas. Towards the conclusion of the first chapter on the "Discrimination of the Real" (tattvaviveka) , the author explains the Upanisad terms sravana, manana, and nididhyasana (vv. 53-54), and then proceeds to describe the cultivation of samadhi as the means whereby the mediate verbal knowledge derived from the Upanisads is turned into immediate experience (vv. 5962). However, in chapter nine, "The Lamp of Meditation" (dhyanadipa), meditation is prescribed for those who do not have the intellectual acuteness to undertake the Self-inquiry; and in chapter seven (v. 265), the author repeats the verse of Sankara from the Upadesasahasri("As I have no restlessness"), which was cited above. Therefore it would appear that the Pancadasi is an early example of a Vedantic text which is consciously making room for classical Yoga but which has not lost sight of Sankara's perspective.[56]

The Vivekacudamani is a popular text in contemporary Vedanta circles and is ascribed to Sankara. However, it is highly unlikely that it is a genuine work of Sankara, for the fact that there are no Sanskrit commentaries on this work by any of the well-known commentators on the works of Sankara would indicate that the Vivekacudamani is either a late composition or that it was not regarded as a work of Sankara by the earlier Advaitins.[57] In this text, samadhi comes in for considerable praise; for example:

Reflection should be considered a hundred times superior to hearing, and meditation a hundred thousand times superior even to reflection, but the Nirvikalpaka Samadhi is infinite in its results.[58]

We can observe in this text how samadhi is treated as the indispensable requirement for liberation, and we can see in the following verse that samadhi is advocated for the same reason as is given in Yogasutra 1.1.4: "at other times [the Self] takes the same form as the mental modifications (vrttisarupyamitaratra)":

By the Nirvikalpaka Samadhi the truth of Brahman is clearly and definitely realized, but not otherwise, for then the mind, being unstable by nature, is apt to be mixed up with other perceptions.[59]

As a final example of the use of samadhi in this work we cite the following verse:

Through the diversity of the supervening conditions (Upadhis), a man is apt to think of himself as also full of diversity; but with the removal of these he is again his own Self, the immutable. Therefore the wise man should ever devote himself to the practice of Nirvikalpa Samadhi for the dissolution of the Upadhis.[60]

If we compare the idea contained in this verse with the ideas of the Upadesasahasri, we find that nowhere in the Upadesasahasri does San-kara advocate the dissolution of the upadhi: On the contrary, his attitude throughout the Upadesasahasri is to show that an upadhi is to be negated merely through the knowledge that it is an object, for as an object it cannot be identical with the perceiver; and because an upadhi is essentially unreal (mithya), it cannot negate the nondual truth, and therefore no additional effort need be expended for its removal.

As a final example of the increasing tendency to identify Vedanta and Yoga, we refer to a late Vedanta text, the Vedantasara of Sadananda (fifteenth century A.D.). He, like the author of the Pancadasi, has added samadhi to the triad of sravana, manana, and nididhyasana. What is of interest here is that he has reinterpreted samadhi to make it conform to Advaitic ideas; for example, nirvikalpa samadhi is said to be the state where the mind is without the distinctions of knower, knowledge, and object of knowledge and has become totally merged in the "nondual reality."[61] Furthermore, this text lists the eight limbs of Yoga practice mentioned by Patanjali (Yogasutra 2.29), suitably reinterpreted to conform to the Vedanta. There are other, later Vedanta texts which also do this.[62] Thus we see that through the centuries Vedanta has increasingly accommodated itself to Yoga, leading to the almost complete absence of a distinction between the two in modem times.


Although the importance of concentration is evident from the early Upanisads (BU 4.4.23), a form of yoga practice leading to the absorptive state of samadhi is only in evidence in the later texts. We have seen that Sankara does speak of a type of concentration upon the Self which is akin to yoga insofar as there is the withdrawal of the mind from sense objects, but he does not advocate more than that and he does not put forward the view that we find in classical Yoga about the necessity of total thought suppression. We have seen that he has used the word samadhi very sparingly, and when he has used it, it was not always in an unambiguously favorable context. It should be clear that Sankara does not set up nirvikalpasamadhi as a spiritual goal. For if he had thought it to be an indispensable requirement for liberation, then he would have said so. But he has not said so. Contemplation on the Self is obviously a part of Sankara's teaching, but his contemplation is directed toward seeing the ever present Self as free from all conditionings rather than toward the attainment of nirvikalpasamadhi. This is in significant contrast to many modem Advaitins for whom all of the Vedanta amounts to "theory" which has its experimental counterpart in yoga "practice." I suggest that their view of Vedanta is a departure from Sankara's own position. The modem Advaitins, however, are not without their forerunners, and I have tried to indicate that there has been a gradual increase in samadhi-oriented practice in the centuries after Sankara, as we can judge from the later Advaita texts.








Michael Comans, Ph.D. began a serious study of Advaita Vedanta as a resident student of Swami Dayananda Saraswati. After returning to Australia he studied for his doctorate under the guidance of the renowned Indologist, Professor J.W. de Jong, at the Australian National University. He also has the opportunity to study from Senior Pandit Srinivasa Shastri of Deccan College, Pune, and Dr. R. Krishnamurti Shastri of Madras Sanskrti College. Dr. Comans taught at the University of Sydney (Centre for Indian Studies) for a number of years. He now spends his time both in Sydney, where he teaches Vedanta and Sanskrit, and in India.


Abbreviations are used in the notes below as follows:

BSBh   Brahmasutra-Sankarabhasyam    with  the   Commentaries
       Bhasyaratnaprabha    of   Govindananda,   Bhamati   of
       Vacaspatimisra and Nyaya-Nirnaya of Anandagiri. Edited
       by J. L. Sastri. Delhi: Motilal Banarsidass, 1980.

BU     Brhadaranyakopanisad.

ChU    Chandogyopanisad.

US     Upadesasahasri of Sankaracharya, A Thousand Teachings:
       in Two Parts--Prose  and  Poetry.  Translated by Swami
       Jagadananda. Madras: Sri Ramakrishna Math, 1979.

1. When the word samadhi is used in this article, it refers only to the higher stage of samadhi known as
Nirvikalpa Samadhi, which is an "enstasis without thought constructions." See footnote 22 below. 2. Vedanta for the Western World, ed. C. Isherwood (London: Unwin Books, 1975), p. 15. 3. The three years of continuous Tantric sadhana under the direction of the Bhairava Brahmani was his longest and most significant training. See W. Neevel, "The Transformation of Sri Ramakrishna," in Hindu-ism.' New Essays in the History of Religions, ed. B. Smith (Leiden: E. J. Brill, 1976). The time spent under the direction of Totapuri, who was said to be an Advaitin, was much shorter than the time spent studying Tantra, and the information available on Totapuri is very meager, so it is difficult to be sure whether he was actually an Advaitin rather than a follower of yoga. 4. M. Eliade, Yoga: Immortality and Freedom, Bollingen Series, no. 56 (New York: Princeton University Press, 1973) , pp. 227 ff., and The Hathayogapradapika of Svatmarama (Madras: Adyar Library, 1984), p. 125. 5. Ramakrishna, The Gospel of Sri Ramakrishna, trans. Swami Nikhi-lananda (Madras: Sri Ramakrishna Math, 1974), p. 195. 6. Ibid., p. 814. Also cf. pp. 310, 576. 7. Cf. The Complete Works of Swami Vivekananda (Calcutta: Advaita Ashrama, 1970), vol. 1, p. 470, "If there is a God, you ought to be able to see Him. If not, let Him go." Also cf. his introduction to Raja-Yoga, pp. 125 ff., and vol. 2, p. 220, "Knowledge can only be got in one way, the way of experience; there is no other way to know." 8. Ibid., vol. 1, pp. 137, 180, 181,212, and vol. 5, p. 300. 9. Ibid., vol. 1, p. 188. 10. S. Radhakrishnan, Eastern Religions and Western Thought (London: Allen and Unwin, 1940), p. 51. 11. G. A. Jacob, A Concordance to the Principal Upanisads and Bhagavadgita (Delhi: Motilal Banarsidass, 1971); G. M. Kurulkar, Sasandarbhanighantusahita Dasopanisadah (Pune: Tilak Maharastra Vidyapitha, 1973). 12. ChU 8.1.3, 4, 5; BU 4.2.1, 4.4.23. 13. The Brhadaranyaka Upanisad, with the Commentary of Sankaracarya, trans. Swami Madhavananda (Calcutta: Advaita Ashrama, 1975), p. 410. 14. Cf. BSBh, p. 36; Vivekacudamani of Sri Sankaracarya, trans. Swami Madhavananda (Calcutta: Advaita Ashrama, 1974), vv. 19-27; Aparoksanubhati or Self-realization of Sri Sankaracarya, trans. Swami Vimuktananda (Calcutta: Advaita Ashrama, 1977), vv. 3-8. 15. Ten Principal Upanisads with Sankarabhasya, Works of Sankaracarya in Original Sanskrit, vol. 1 (Delhi: Motilal Banarsidass, 1978) , p. 937, " samahitah indriyantahkaranacalanarupad vyavrtya aikagryarupena samahito bhutva." (Hereafter, all Upanisad references containing Sankara's commentary will be to this work). 16. Ibid., p. 78, "asamahitah-anekagramana viksiptacittah." 17. Tattva Bodha of Sankaracharya (Bombay: Central Chinmaya Mission Trust, n.d.), p. 7; Aparoksanubhuti (cited n. 14 above), v. 8. 18. Georg Feuerstein, The Philosophy of Classical Yoga (Manchester: Manchester University Press, 1980), p. 84. 19. Paul Deussen, The Philosophy of the Upanishads (New York: Dover, 1966), pp. 23-26. Also, see Winternitz quoted in S. Dasgupta, A History of Indian Philosophy (Delhi: Motilal Banarsidass, 1975), vol. 1, p. 39. Eliade considers the Maitrayani to belong to the same period as the Bhagavadgita, i.e., between the second century B.C. and the second century A.D. (Eliade, Yoga, p. 124). 20. Amrtabindu 6, 16; Aruneya 2. It also occurs in the Bhagavadgita at 2.44, 53, 54. 21. Eliade, Yoga, p. 114, remarks: "It is true that the Upanisads remain in the line of metaphysics and contemplation, whereas yoga employs asceticism and a technique of meditation. But this is not enough to halt the constant osmosis between the Upanisadic and yogic milieus." 22. I do not know why later Vedantins used the word nir-vikalpa to characterize what is essentially the yogic asamprajnata samadhi. Perhaps they wished to distinguish their practice from that of classical Yoga. The word nirvikalpaka was first introduced into the astika ("orthodox") tradition by Kumarila Bhatta, who used it in his explanation of perception, under the influence of the Buddhist philosopher Dignaga. See D. N. Shastri, The Philosophy of Nyaya-Vaisesika and Its Conflict with the Buddhist Dignaga School (Delhi: Bharatiya Vidya Prakashan, 1976), p. 438. See also footnote 1 above. 23. I am assuming that Sankara is not the author of the Yogasutrabhas-yavivarana, as this issue has not yet been settled. See W. Halbfass, Tradition and Reflection: Explorations in Indian Thought (New York: State University of New York Press, 1991), chap. 6. 24. BSBh 2.1.9 (p. 365, line 6), 2.3.39 (p. 545, line 10), 2.3.40 (p. 551, line 2) ; Word Index to the Brahma-Sutra-Bhasya of Sankara, T.M.P. Mahadevan, general ed., 2 vols. (Madras: University of Madras, 1973). 25. BSBh 2.1.9 (p. 365, line 6). 26. Ibid., 2.3.39 (p. 545, line 10). 27. Mandukya 3.37 (p. 224, line 3). 28. Katha 1.3.13. Cf. J. Bader, Meditation in Sankara's Vedanta (Delhi: Aditya Prakashan, 1990), chap. 3. 29. Katha 1.3.13 (p. 83, line 11). 30. BSBh 1.4.1 (p. 295, line 10). 31. Ibid., 3.3.15 (p. 694, line 12). 32. Ibid., 1.4.1 (p. 295, lines 12 ff.). 33. BU 2.4.11 (p. 764, lines 11 ff.). See also Madhavananda, trans., Brha-daranyaka Upanisad (cited n. 13 above), pp. 253 ff. I have cited Madhavananda's translation here as I cannot make any significant improvement on it. 34. Cf. commentary on Katha 1.2.12 and Bhagavadgita 16.1. 35. BU 1.4.7 (p. 663, line 9). 36. BSBh 1.1.4 (p. 69, line 6). 37. Ibid., 1.1.4 (p. 79, lines 7 ff.). Also, for the reference to action as consisting of four types, cf. BU 3.3.1 (p. 798, lines 22 ff., and p. 801, lines 1 ff.), 4.4.22 (p. 933, lines 21 ff.); Mundaka 1.2.12 (p. 152, lines 25 ff.) ; US 17.50; Shri Shankaracharya's Upadeshasahasri with the Gloss Padayojanika, ed. D. V. Gokhale (Bombay: The Gujarati Printing Press, 1917); Shri Shankarabhagavatpada's Upadeshasahasri with the Tika of Shri Anandagiri Acharya, ed. S. Subramanyasastri (Varanasi: Mahesh Research Institute, 1978). 38. BSBh 1.1.4 (p. 79, line 1); also BU 2.1.20 (p. 739, lines 20 and 24). 39. BSBh 1.1.4 (p. 64, lines 2 and 4; p. 84, lines 3 ff.; p. 85, lines 1 ff.; p. 87, lines 4 ff.). 40. BU 3.3.1 (p. 798, lines 19 ff.). 41. BSBh 2.1.3 (p. 354, lines 1 ff.). 42. Ibid., 2.1.3 (p. 354, line 3). 43. Bhagavadgita with Sankarabhasya, Works of Sankaracarya in Original Sanskrit, vol. 11 (Delhi: Motilal Banarsidass, 1978) 6.19 (p. 107, lines 9 ff.), and also 5.21, 6.4, 8.10, 12.6, 13.10, 18.33. 44. Ibid., 18.66 (p. 296, lines 6 ff.). 45. Ibid., introd., 2.11 (p. 9, lines 14 ff.). 46. Ibid., 2.39 (p. 27, lines 13 ff.). 47. Ibid., 4. 38 (p. 80, line 18). 48. Ibid., 6.20 (p. 107, line 16 [my emphasis]). 49. Ibid., 2.21 (p. 20, line 12), 2.63 (p. 36, line 12), 8.8 (p. 128, line 16), 13.30 (p. 215, line 23), 13.34 (p. 217, line 19), 18.16 (p. 263, line 19), 18.17 (p. 264, line 4), 18.50 (p. 281, line 7), 18.55 (p. 284, line 9); Word-Index to Sankara's Gitabhasya, ed. Francis X. D'Sa (Pune: Institute for the Study of Religion, 1985). Also cf. BU 2.1.20 (p. 744, line 23), 2.4.2 (p. 767, line 5), 2.5.15 (p. 776, line 12); ChU 6.15.2 (p. 537, line 12), 8.1.6 (p. 571, line 2); Katha 1.5.12 (p. 96, line 1); Mundaka 1.2.12 (p. 153, line 5), 2.2.7 (p. 162, line 22); US 17.51-52. In an otherwise interesting and insightful article, "The Path of No-path: Sankara and Dogen on the Paradox of Practice" (Philosophy East and West 38, no. 2 [April 1988]), David Loy has come to an erroneous conclusion (p. 133) that "there can be no means--not even sruti--to realize Brahman.... "But if that were the case, it would not be possible to explain Sankara's concerted effort in meticulously commenting on sruti; and such a statement also overlooks the numerous references where he states that the sruti is the means of knowledge for Brahman. It is precisely because Sankara sees no other way to arrive at the knowledge of the unconditioned Absolute that he resorts to the sacred words of the Upanisads as the means to dispel the ignorance of the ever present Self. Among Western scholars, Sankara's views on sruti have been well articulated by W. Halbfass in his discussion of the role of sruti in Sankara's thought; see his Tradition and Reflection (cited n. 23 above), chap. 5. 50. Samadhana is mentioned in US 17.23-24. Cf. Tattvabodha (cited n. 17 above) , p. 7: "samadhanam kim? cittaikagrata." 51. US 13.14, 17 and 14.35. 52. Ibid., 13.14. 53. Ibid., 15.14. 54. The Sankhyakarika of Isvara Krsna, ed. and trans. S.S. Suryanarayana Sastri (Madras: University of Madras, 1973), vv. 20, 21, 66, 68. 55. US 16.39-40. 56. Cf. Pancadsi of Sri Vidyaranya Swami, trans. Swami Swahananda (Madras: Sri Ramakrishna Math, 1975). 57. There are two commentaries on the Vivekacudamani: one is by a little known writer, Harinathabatta, and the other is a recent commentary by Sri Chandrasekhara Bharati, who was the Sankaracarya of Sriagiri Matha from 1912 to 1954. See R. Thangaswami, Advaita-Vedanta Literature: A Bibliographical Survey (Madras: University of Madras, 1980), p. 218; Advaita Grantha Kosa, prepared by a disciple of Sri Ista Siddhindra Saraswati Swami of the Upanisad Brahmendra Mutt (Kancheepuram: n.n., n.d.), p. 67. Perhaps the Vivekacudamani is itself a work of one of the Sringiri Sankaracaryas? 58. Vivekacudamani (cited n. 14 above), v. 364. 59. Ibid., v. 365. 60. Ibid., v. 357. 61. Vedantasara or the Essence of Vedanta of Sadananda Yogindra, trans. Swami Nikhilananda (Calcutta: Advaita Ashrama, 1974), p. 110. 62. The Aparoksanubhuti has been ascribed to Sankara but is unlikely to be a genuine work. See Encyclopedia of Indian Philosophies, ed. Karl Potter, vol. 3, Advaita Vedanta up to Samkara and His Pupils (Delhi: Motilal Banarsidass, 1981), p. 320. The final forty-four verses (out of 144) describe yoga. Here, however, yoga is consciously reinterpreted within a Vedantic manner: "The complete forgetfulness of all thought by first making it changeless and then identifying it with Brahman is called Samadhi known as knowledge" (Vimuktananda's trans., cited n.14 above, v.124) . The Sarvavedantasiddhantasarasangraha is another work which is most likely not a work of Sankara. See Thangaswami, Advaita-Vedanta Literature (cited n. 57 above), p. 220; Potter, Advaita Vedanta up to Samkara, p. 339; Advaita Grantha Kosa (cited n. 57 above), p. 68. In this work we again find the grafting of yogic nirvikalpasamadhi onto Vedanta teachings; see The Quintessence of Vedanta, trans. Swami Tattwananda (Emakulam: Sri Rama-krishna Advaita Ashrama, 1960), pp. 171 ff.