In the Himalayas of Indian thought, three mountain ranges tower above the rest: Sāṅkhya-Yoga, Buddhism, and Vedānta. Rather than argue that this is so, let us consider why it is so. What is it that causes these three systems (or sets of systems) to stand out as the most important? This short chapter answers that question by demonstrating their relationship. What is special about these three is that they elaborate the three primary responses to the epistemological problem of the subject–object relation — an issue that is fundamental to any metaphysical system and is especially crucial for any philosophy that purports to explain the experience of enlightenment.
Sāṅkhya-Yoga presents the most radical dualism possible by completely sundering subject and object. The separation between the two is so extreme that, as is generally accepted, the system fails because there can be no communication or cooperation between them.
Early Buddhism conflates subject into object. Consciousness is something conditioned, arising only when certain conditions exist. The self is merely an illusion created by the interaction of the five aggregates. The self shrinks to nothing and there is only a void, but the Void is not a thing: it expresses the fact that there is absolutely nothing, no-thing at all, which can be identified as the self.280
Advaita Vedānta conflates object into subject. There is nothing external to Brahman, the One without a second. Since Brahman is a nondual consciousness, consciousness may be said to expand and encompass the entire universe, which is but the appearance of Brahman. Everything is the Self. One important consequence of this is that we all have (or rather are) the same Self.
Of these three, only Advaita Vedānta is obviously an attempt to describe the experience of subject–object nonduality. With Buddhism one must be more careful about such a generalization: it seems true for Mahāyāna, but not for Pāli Buddhism, at least not explicitly (an issue we return to). In the case of Sāṅkhya-Yoga, which is unequivocably dualistic, there seems to be no ground whatsoever for claiming that it is an attempt to describe the nondual experience. But chapter 7 suggests that Sāṅkhya-Yoga may be viewed as such an attempt, although an unsatisfactory one. Here it is necessary to summarize the claims of Sāṅkhya metaphysics, which will enable us to see why it is inadequate.
Sāṅkhya is dualistic because it explains the subject–object relation by postulating two basic substances: puruṣa, pure unchanging consciousness, and prakṛti, the natural world that encompasses everything else. It is significant that this is not a Cartesian dualism: prakṛti includes all mental as well as physical phenomena; what we experience as our mind (buddhi) and all its mental phenomena are evolutes of prakṛti too. Anything that can be experienced is prakṛti. Thus puruṣa is reduced to a pure “seer” which actually does nothing, although its presence is necessary not only for there to be awareness but also to act as a catalyst for the evolutes of prakṛti. In our usual deluded condition we are not able to distinguish between these two substances. Pure consciousness mistakenly identifies itself with its reflections; that is, I cling to “my” mental panorama and “my” body and its possessions. The puruṣa is so attenuated that it is not even able to realize the distinction between itself and prakṛti. As in Advaita, it is actually the buddhi, the most rarified part of the prakṛti, that realizes the distinction, whereupon it abdicates by itself and the puruṣa is established in its own true nature as solitary and independent, indifferently observing the natural world.
The main problem is that the polarity between puruṣa and prakṛti is so great that they are unable to cooperate. Puruṣa is so indifferent and prakṛti so mechanical that they cannot function together. The common simile to explain their interaction is that of a blind man of good foot carrying a cripple of good eye; but this is not a good analogy because in order to interact both men must have intelligence, whereas prakṛti does not. The simile would fit better if the cripple has no desire to go anywhere and so says and does nothing, while the blind man literally has no mind at all. Clearly, in such a case they would not cooperate, yet Sāṅkhya-Yoga claims that the whole universe evolved out of the interaction arising from the introduction of puruṣa to prakṛti.
Whereas Sāṅkhya is a metaphysical system, Yoga deals mainly with the yogic path which one follows in order to attain kaivalya, the “liberated isolation” of the puruṣa. It is significant that there is nothing within the eight limbs of yoga practice antithetical to Vedānta. In fact the yogic path seems to fit an Advaitic metaphysics better than a Sāṅkhya one. In samādhi, the eighth and highest limb, the mind loses ego-awareness and becomes one with its object of meditation, but this nondualistic experience can be described only as “as if one” in Yoga, since the ultimate goal is understood to be the discrimination of pure consciousness from all those objects it usually identifies with. But of course this nondual experience accords very well with the Advaitic aim of “realizing the whole universe as the Self.”
Yet what is the most interesting is that the puruṣa, like the jīva of Jainism and the ātman of Vedānta and Nyāya-Vaiśeṣika, is eternal and omnipresent; it has no particular locus but is ubiquitous, pervading everywhere. So the puruṣa is very different from consciousness as we normally understand it, an “internal world” counterposed to the external world. Clearly Sāṅkhya is far from capturing commonsense duality. That the puruṣa is so emptied of any function — it has almost nothing to do except to be unvarying consciousness — is also significant. In this respect it is similar to the Nirguṇa Brahman of Vedānta, which is also devoid of any attributes in itself, or one might characterize puruṣa with the Buddhist term śūnya, as empty. But that the puruṣa is all-pervading leads to problems for Sāṇkhya, since there is supposedly an infinity (or at least a very large number) of completely distinct, unrelated puruṣas. How can they all occupy the same infinite space without affecting each other in some way? Given that they are all devoid of any attributes, how are they to be distinguished from each other? A corollary problem is that each undifferentiated puruṣa has a relationship with only one particular buddhi. Furthermore, each liberated puruṣa, being ubiquitous, must coexist with all of prakṛti, yet be completely unaffected by it.
For these and other reasons, this most extreme dualism between subject and object fails. The failure of Sāṅkhya-Yoga, it should be noted, is not incidental but is due to a basic inadequacy: the duality is so radical that it precludes any cooperation between the two categories. In accord with this, there are two ways one might try to resolve the problem. One could conceive of all puruṣas as various reflections of one unified consciousness and subsume prakṛti as another aspect or manifestation of this consciousness. Or, given the functionlessness of the puruṣa, one could eliminate it altogether and incorporate consciousness into prakṛti. Either solution, of course, transforms Sāṅkhya into a completely different system, because the root dualism has been abandoned. The first alternative makes Sāṅkhya into Vedāntic monism, and the second makes it into the anātman pluralism of Pāli Buddhism. If, as some scholars believe, Sāṅkhya is the oldest Indian metaphysical system, this may well have been what happened historically: when its dualism came to be recognized as an unsatisfactory description of the enlightenment experience, Indian philosophy developed the diametrically opposed alternatives of Vedānta and Buddhism. The conflict between these two alternatives is our main concern, but I refer to Sāṅkhya-Yoga again when we return in chapter 7 to the question of whether all these systems might be responding to the same nondual experience.
The nature of nirvana is the greatest problem of Buddhist philosophy, probably because the Buddha himself refused to speculate on it. His attitude was, in effect, that if you want to know what nirvana is, then you must attain it. But clearly nirvana does not involve the isolation of a Sāṅkhya-like pure consciousness, because there is no such thing in Buddhism. The unique feature of Buddhism is that there is no self at all, and never was; there are only five skandhas, aggregates or “heaps” of elements which constantly interact. These skandhas do not constitute a self; the sense of a self is merely an illusion created by their interaction. The Buddha emphasized that we should not identify anything as the self.
Thus nirvana is probably best characterized as the realization that there is no self: although this by itself is not much help because what that means — what there is that realizes this — is still unclear. The Buddha compounded the mystery by emphasizing that nirvana is neither annihilation nor eternal life. Clearly this is necessary since there never was a self to be destroyed or to live eternally, but it is confusing in so far as our thinking naturally tends to dichotomize into one or the other.
Yet there are a few passages in the Pāli Canon that apparently contradict the usual Theravada interpretation. In the Kevada Sutta (Dīgha-Nikāya), for example, the Buddha says that name-and-form are wholly destroyed “where consciousness is signless, boundless, and all-luminous.” Aṅguttara-Nikāya 1.6 claims that “This mind (citta) is luminous, but is defiled by adventitious defilements.” (See note 281.) This distinction between our usual conditioned consciousness and an all-luminous consciousness seems inconsistent with the common view in early Buddhism that consciousness is the result of conditions and does not arise without those conditions. Needless to say, it accords very well with the Vedantic position regarding “self-luminous” Brahman. In the Brahmanimantanika Sutta (Majjhima-Nikāya) the Buddha criticizes the idea of an omnipotent Brahma, but it is significant that within the Pāli Canon “there is no expressed contradiction or even recognition of the Vedānta theory of an ātman or brahman as the one ultimate reality.”282
It is also significant that much the same controversy between early Buddhism and Vedānta is found internally within Buddhism. Abhidharma, the philosophical branch of early Buddhism, analyzed reality into a set of discrete dharmas whose interaction creates the illusion of a self. Nirvana in Pāli Buddhism seems to have been understood as the cessation of cooperation among these various dharmas, leading to their quiescent isolation from each other.283 Since consciousness is conditioned, existing only as a result of their interaction, this would seem to be the cessation of all consciousness as well. But insofar as these dharmas are believed to exist objectively, such a view may be criticized as ontologically lopsided; although the self has been analyzed away, the reality of the world as objective has been left largely unaffected. Yet the elimination of consciousness requires a redefinition of what it means for something to exist. What remains must somehow incorporate consciousness (or what we have understood as consciousness) within itself. Our usual dualistic understanding of self-confronting-object may be likened to a scale balancing two weights; one weight cannot be removed without affecting the other side of the balance. A basic principle of this book is that we cannot change one pole of any duality without transforming the other just as much. It is not possible to deconstruct one half of the consciousness/matter duality by simply absorbing it into the other, undeconstructed half. If we deny mind as an ontological category then we must redefine matter as other than inert and find what we have understood as mind within it.
Mahāyāna accepted the theory of dharmas — an important point often overlooked — but not their objective reality. It expanded the denial of the self (pudgalanairātmya) into a denial of the reality of dharmas (dharmanairātmya) because all dharmas are relative and hence śūnya. There is a higher, absolute truth (paramārtha), which cannot be described (according to Mādhyamika) but which (according to Yogācāra) comes close to the nondual attributeless consciousness of Advaita Vedānta. The relation between Mādhyamika and Yogācāra, the two main philosophical systems of Mahāyāna, is of considerable relevance to this work. Significant and unresolvable differences between them would constitute a challenge to our defense of a core doctrine of nonduality. So it is noteworthy that the weight of scholarly opinion favors the view that they complement much more than they contradict each other. On this matter, it is worth our while to quote three of the twentieth century’s greatest Western scholars of Buddhism. First, Guiseppe Tucci, perhaps the foremost student of Tibetan Buddhism:
It is generally said that Mahāyāna may be divided into two fundamental schools, viz., Mādhyamika and Yogācāra. This statement must not be taken literally. First of all it is not exact to affirm that these two tendencies were always opposed to each other. Moreover . . . the antagonism between the Mādhyamika and the first expounders of the idealist school such as Maitreya, Asaṅga and even Vasubandhu is not so marked as it appears at first sight. . . . The fact is that both Nāgārjuna as well as Maitreya, along with their immediate disciples, acknowledged the same fundamental tenets, and their work was determined by the same ideals, though holding quite different views in many a detail.284
The translator and expounder of the Prajñāpāramitā, Edward Conze, develops this view by explaining their difference of perspective:
Mādhyamikas and Yogācārins supplement one another. They come into conflict only very rarely, and the powerful school of the Mādhyamika-Yogācārins demonstrated that their ideas could exist in harmony. They differ in that they approach salvation by two different roads. To the Mādhyamikas “wisdom” is everything and they have very little to say about dhyāna, whereas the Yogācārins give more weight to the experience of “trance.” The first annihilate the world by a ruthless analysis which develops from the Abhidharma tradition. The second effect an equally ruthless withdrawal from everything by the traditional method of trance.285
Edward J. Thomas agrees with Conze:
While the school of Nāgārjuna started from the standpoint of logic, and showed the impossibility of making any statement free from contradictions, the Laṇkāvatāra [according to the writer “the chief canonical text for the doctrine of subjective idealism”] started from a psychological standpoint, and found a positive basis in actual experience.286
Last but hardly least is the historical testimony of Buddhism itself, in particular the fact alluded to by Conze that the debate between Mādhyamika and Yogācāra eventually led to their synthesis in the Mādhyamika-Yogācāra school of Śāntarakṣita and Kamalaśīla. If we remember that both systems were not just philosophies but had primarily a soteriological function — that is, were meant to be paths of liberation — then the contrast between the Mādhyamika and Yogācāra viewpoints becomes comprehensible as differences in perspective. Mādhyamika emphasizes that reality is śūnya in the sense of “empty of predication.” One can say nothing about reality because (as argued in chapter 2) that would be to superimpose concept on percept. This does not amount to an assertion of nonduality because in limiting itself to a negative critique of all dualities, Mādhyamika makes no positive claims. The important point is that this seems to be done in order to clear the way for the experience we have been describing as nondual. This experience must be distinguished from any claims whatsoever — ontological, epistemological, or otherwise — made on the basis of it, for from the Mādhyamika perspective any such claim would be a savikalpa attempt to determine the bare nirvikalpa percept.
The conflict with Yogācāra arises when Yogācārins call that percept “mind” (or “consciousness”: vijñāna). This does not mean that Yogācāra is “subjective idealism,” as Thomas and others (including Śaṅkara) have understood it to be. Rather than the world being the projection of a subjective ego, the apparent distinction between subject and object is one that arises (or seems to arise) within the transcendental mind (vijñāptimātratā) — a view consistent with our core doctrine of nonduality. Subjective idealism is not to be found anywhere within Buddhism, nor do I see how it could be there, given the common acceptance by all Buddhist schools of anātman, which denies any ontological self. Mādhyamika naturally criticizes Yogācāra for trying thus to put a label on reality (“mind”) and dialectically criticizes the term by showing that it is relative and hence delusive. But all that does not refute the Yogācāra claim, which is simple nonduality. When delusions fall away and I experience reality, the consciousness that is aware of the world and the world itself are not distinguishable. At that moment all of what is experienced is myself, for which reason it may be called “my mind.”
Here the difference between the logical and the psychological standpoint is crucial, as Conze and Thomas have pointed out. Although the goal of Mādhyamika is to transcend the intellect, its path is still intellectual. The conceptualizing mind is surmounted by exhausting it — that is, by negating all conceptual possibilities. Only by that could the leap to prajñā occur. Yogācāra, as its name suggests, is a more meditative approach. The aim of the system is not to prove the existence of transcendental mind but to initiate the practitioner into experience of it, which occurred in samādhi. Hence Yogācārins had no need to fear the concept of “mind,” for their meditative practice kept them from confusing such labels with reality itself; whereas Mādhyamika, the logical path of exhausting all concepts, could not tolerate it. So the difference between these two philosophical perspectives ultimately derives from their different soteriological approaches and does not extend to the nondual experience at which they both point.
Chapter 6 also supports the compatibility of the two systems by pointing out that the Mādhyamika reconciliation of impermanence with immutability, and of all-conditionality with no-causality, is equivalent to the Yogācāra relation between the paratantra and pariniṣpanna natures. Further discussion of the historical debate between Mādhyamika and Yogācāra is beyond the scope of this work.287
Śaṅkara’s Advaita Vedānta is generally regarded as having best developed and systematized the main strand of Upaniṣadic thought, which stresses the identity of Ātman and Brahman. Brahman is an infinite, self-luminous consciousness that transcends the subject–object duality. As “the Witness” (sākṣin), it is that which cannot be made into an object. Unqualified and all-inclusive, perhaps its most significant feature is that it is “one without a second,” for there is nothing outside it. Hence Ātman — the true Self, what each of us really is — is one with this Brahman. Tat tvam asi: “That thou art.” This is “All-Selfness”:
. . . there is nothing else but the Self.
To realize the whole universe as the Self is the means of getting rid of bondage.
To the seer, all things have verily become the Self.
Whoever has realized and intimately known the Self, all is his Self, and he, again, is indeed the Self of all.288
So the Ātman should not be understood as a distinct self that merges with Brahman. To realize Ātman is to realize Brahman because they are really the same thing. One may state, in answer to the Buddhists, that a consciousness of self is needed to organize experience, but that turns out to have been Brahman itself, once Brahman is realized — that is to say, when Brahman realizes its own true nature. The world of differences and change is māyā, illusion; there is nothing but the all-inclusive Self (which must somehow incorporate māyā). Yet this sounds awkward, since the concept of a self seems to presuppose an other, a nonself from which it is distinguished — a point to which we will return later. So perhaps the term Ātman should be rejected as superfluous, because it suggests another entity apart from Brahman. Yet the two terms do serve a function, since they emphasize different aspects of the Absolute: Brahman, that it is the ultimate reality which is the ground of all the universe; Ātman, that it is my true nature.
For Śaṅkara mokṣa, liberation, is the realization that I am and always have been Brahman. My individual ego-consciousness evaporates or is realized to be an illusion, but not the pure, nondual consciousness that it was always just a reflection of. It must be emphasized that I do not attain or merge with this Brahman; I merely realize that I have always been Brahman. Śaṅkara uses the analogy of the space within a closed jar: that space has always been one with all space; there is but the illusion of separateness. That there is really nothing to attain becomes even more significant when we remember that the same is true for Sāṅkhya-Yoga and Buddhism: however else it may be characterized, one’s true nature has always been pure and unstained. The Sāṅkhya puruṣa is an indifferent seer, which was always merely observing, unaffected by pain or pleasure. In Buddhism, there never was a self; it was always just an illusion.289
Yet, just as there are passages in the Pāli Canon which sound Vedāntic, so there are passages in the Upaniṣads which, at first encounter, seem Buddhistic. Perhaps the most famous is Yājñavalkya’s instruction to his wife Maitreyī in the Bṛhadāraṇyaka: “Arising out of these elements (bhūta), into them also one vanishes away. After death there is no consciousness (ne pretya samjñā ’sti).” Maitreyī is amazed by this, so Yājñavalkya explains it in the well-known passage on nonduality quoted in chapter 1:
For where there is a duality, as it were, there one sees another. . . . But when, verily, everything has become just one’s own self, then what could one see and through what? . . . Through what could one know that owing to which all this is known? So, through what could one understand the Understander? This Self . . . is imperceptible, for it is never perceived.290
In his commentary, Śaṅkara interprets this passage as meaning that when one realizes Brahman there is no more particular or dualistic consciousness. But perhaps there is the same problem with consciousness as with the self. Just as our concept of a self normally presupposes a nonself, so consciousness is usually understood to require an object. In fact it is very difficult to conceive of what consciousness could be without an object, a problem which is of course very much the heart of the issue. In English, for example, all the verbs for consciousness are normally intentional, requiring both subjects and objects (“I am conscious of . . .”, “you are aware of . . .”, “he knows that . . .”). Advaita does not deny that our normal “I-consciousness” is intentional: “There is no manifestation of the ‘I’ without a modification of the mind directed to the external” (Sureśvara).291 The claim of Advaita is rather that only the pure consciousness which is Brahman is self-luminous and nondual. But if there is nondual consciousness without an “I” that has it, and without an object that “I” am aware of, can that still be called consciousness? Perhaps either reply, yes or no, could be justified, which suggests that the difference between these opposed standpoints may be merely linguistic. We return to this crucial point in the next chapter.
The similarities between Mahāyāna Buddhism and Advaita Vedānta are so great that some commentators conceive of the two as not really distinct from each other.
Buddhism and Vedānta should not be viewed as two opposed systems but only as different stages in the development of the same central thought which starts with the Upaniṣads, finds its indirect support in Buddha, its elaboration in Mahāyāna Buddhism, its open revival in Guaḍapāda, which reaches its zenith in Śaṅkara and culminates in the post-Śaṅkarites.
So far as the similarities between Buddhism and Vedānta are concerned, they are so many and so strong that by no stretch of the imagination can they be denied or explained otherwise. So far as the differences are concerned, that are few and mostly they are not vital. Most of them rest on a grave misunderstanding of Buddhis. (Chandradhar Sharma)292
Surendranath Dasgupta agrees in the conclusion of his study of Śaṅkara’s system:
His Brahman was very much like the śūnya of Nāgārjuna. It is difficult indeed to distinguish between pure being and pure nonbeing as a category. The debts of Śaṅkara to the self-luminosity of the Vijñānavāda Buddhism can hardly be over-estimated. . . . I am led to think that Śaṅkara’s philosophy is largely a compound of Vijñānavāda and Śūnyavāda Buddhism with the Upaniṣad notion of the permanence of the self superadded.293
Lalmani Joshi also argues for this:
In [Gauḍapāda’s] Āgamaśāstra we find an endeavour to synthesize and bring about a concord between Mahāyāna Buddhism and Advaita Vedānta. In this endeavour seem to have crept into Vedānta certain basic tenets of Mahāyāna philosophy, and the result was the nondualistic Vedānta of Saṅkara.
. . . The Advaita turn in Vedānta in and after Gauḍapāda can reasonably and satisfactorily be explained only by recognizing the debt of Gauḍapāda and Śaṅkara to the Mādhyamika and Vijñānavada systems of thought.294
The objectivity of this conclusion is further supported by the different sympathies of their proponents: Sharma is an Advaitin, Dasgupta a Hindu critical of Śaṅkara, and Joshi is sympathetic to Buddhism.
It is undeniable that Śaṅkara was much influenced by Mādhyamika dialectic, which he employed in his own criticisms of other systems. But the similarities go far deeper, to the extent that Śaṅkara’s rather shrill condemnation of Buddhism begins to sound like a family quarrel between two brothers — which arguments are often the most violent.
The Buddhist doctrines of non-origination (ajātivāda), of phenomenal world as illusion or mere appearance (māyāvāda), of twofold division of truth into ultimate (paramārtha) and temporal (vyavahāra), and of Reality (tattva) being without attributes (nirguṇa) and beyond fourfold description have become so completely Vedāntic that their origins have nearly been forgotten.295
Śaṅkara’s main criticism of Mādhyamika, that it espouses nihilism, certainly misses the point of Nāgārjuna’s negations, which is the nonconceptual jump to prajñā that occurs as a consequence of negating prapañca. As T. R. V. Murti puts it, Nāgārjuna does not deny Reality, he simply denies all views about Reality. The only difference is that Mādhyamika condemns even consciousness as unreal; but I have already argued that this is relative ego-consciousness — that is, dualistic consciousness apart from its object — and not what might be called nondual consciousness. Nor, according to Sharma, are there any significant differences between Vedānta and Yogācāra, which Śaṅkara admits profoundly influenced his teacher’s teacher Gauḍapāda.296
In conclusion, we have seen why Sāṅkhya-Yoga, Buddhism, and Advaita Vedānta are the preeminent systems in Indian philosophy: because they elaborate the three possible solutions to the problem posed by the subject–object relation. Sāṅkhya-Yoga is the most radical possible dualism. Buddhism denies the self completely by conflating it into the object, which is critically dissolved into dharma-elements. Conversely, Advaita denies the object completely, for “there is nothing else but the Self.” After refuting the extreme dualism of Sāṅkhya, we are left with Buddhism and Vedānta, whose solutions to the subject–object problem seem to be diametrically opposed. But we have also suggested their compatibility. We have noticed some Vedāntic elements in Buddhism and (notwithstanding the claims of many Indian scholars, who want to see Buddhism as an offshoot of Hinduism) the much stronger Buddhist influence on Vedānta. And we have cited the opinions of several prominent scholars who argue for their affinity, indeed sometimes their identity.
But this is not enough to resolve the ontological differences between Mahāyāna and Advaita. To stop here would be to neglect what will perhaps be the most fruitful area of our inquiry into subject–object nonduality. This issue will be continued in the following chapter.