BUDDHISM, SCHOOLS OF: MAHĀYĀNA PHILOSOPHICAL SCHOOLS OF BUDDHISM

Mahāyāna Buddhists in India developed numerous theories on a wide range of topics, and according to Buddhist thinkers, all such theories must relate in principle to reaching nirvāṇa, the highest goal toward which Buddhists are meant to strive. Usually a theory prescribes a specific contemplative practice that will lead the Buddhist to that highest goal, and it is understood that the practice will lead to nirvāṇa only when guided by the theory that recommends it. Indian Mahāyāna Buddhists who accepted the soteriological importance of theory thus faced some critical interpretive tasks: select, defend, and articulate the correct (i.e., soteriologically efficacious) theory among competing theories.

As Mahāyāna Buddhist thought develops, these interpretive tasks focus on philosophical texts (śāstra s) that become, for one reason or another, the inviolable sources of a theoretical system. The main Mahāyāna texts of this kind were composed by three Buddhist thinkers: Nāgārjuna (c. 150 ce), Asaṅga (c. 325 ce), and Vasubandhu (c. 325 ce). Although other thinkers' works also received considerable attention, the works of these three thinkers form the core of the Mahāyāna philosophical schools. Their texts and ideas are studied and interpreted again and again by each generation of Indian Mahāyāna thinkers until the virtual disappearance of Mahāyāna philosophy in India (c. 1400 ce).

As each generation's commentators take up the study of these foundational figures, they invariably employ a kind of dialectical method: arguments are couched as discussions between the proponent and the opponent of a particular notion. These dialectical arguments respond to a wide range of views, both Buddhist and non-Buddhist. Hence, formulating its own retorts and criticisms, each generation develops new and often multiple interpretations of seminal texts and concepts.

By the sixth century, the rapid accumulation of competing interpretations leads Mahāyāna thinkers such as Bhāvaviveka (c. 525 ce) to systematically employ terms for what are often called "philosophical schools." One obvious circumstance underlies this move: commentators fall into well-defined camps that adopt a particular moniker to identify their position. Followers of Nāgārjuna, for example, speak of their philosophy as Madhyamaka (the "Middle Way"), a term that Nāgārjuna himself coins. Subsequent generations of Madhyamaka thinkers are always concerned to defend and articulate the works of Nāgārjuna and, to some degree, the commentaries on Nāgārjuna composed by their predecessors. One thus encounters an unambiguous cohesion within Madhyamaka texts, as evinced by their intertextuality, the continuity of their ideas, their appeal to the same authorities, and so on. Hence, in using a single moniker to refer to many thinkers and their texts, Bhāvaviveka is simply reflecting the obvious cohesiveness of their textual traditions.

In describing Buddhist thought as "schools," however, another motivation is the confusion caused by the multiplicity of views that develop over generations. By Bhāvaviveka's time, Buddhist thought exhibits many variations, and if one adds the opinions of non-Buddhist philosophers, one arrives at a tangled knot of theory. Parsing authors and texts into specific schools allows Buddhist thinkers to unravel that knot and present its various strands in a straightforward fashion. Instead of unpacking endless arguments between individual authors, one instead interprets them as debates between philosophical traditions.

The need to sort and classify the bewildering variety of Mahāyāna philosophical views becomes especially acute when Buddhism spreads to other regions, such as China and Tibet. Indeed, in Tibet an entire literature develops around the term school, and most Tibetan monastic libraries hold several dozen texts that are devoted to minutely parsing and classifying Indian philosophical systems. This literature, called doxography, has heavily influenced the academic study of Indian Mahāyāna thought. The success and influence of Tibetan doxographies stems in part from their ability to elaborate a general classification found in Indian doxographical texts. Enumerated in these terms, all Buddhist thought falls into the hierarchy of the "four schools": the lowest two, the Vaibhāṣika and the Sautrāntika, are not part of the Mahāyāna, but they provide the foundation for the higher schools; the latter two, Yogācāra and Madhyamaka, are considered Mahāyāna schools. This schema, along with its numerous subcategories, has become standard in the academic study of Mahāyāna philosophy.

Since it is so prominent, the model of the four schools will guide the discussion presented below. It will be useful, however, to begin with the problems inherent in the notion of "philosophical schools." An overview of the first two schools will then provide the overall context of Mahāyāna thought. After a concise historical synopsis, this entry focuses on the earliest historical forms of the two Mahāyāna schools, namely Madhyamaka and Yogācāra, along with a brief consideration of later developments.

"Schools" and "Philosophy"

The notion of a philosophical "school" is a difficult one, even in its English usage. In rough terms, a "school" is a voluntary association of various thinkers who articulate and defend a particular set of theories that are deliberately traced through a series of commentators to one or more original thinkers. A main concern in speaking of schools is the need to distinguish the mere avowal of a position from the systematic articulation and defense of that position within a tradition. To refer clearly to that type of systematic articulation, Indian Buddhist authors eventually settle on the term siddhānta, literally, the "conclusion" or outcome of one's theoretical arguments. It is this term that is often translated as "school."

Although clearly useful for Buddhist exegetes and academic interpreters, the concept of a siddhānta or school holds several problems. For example, if a taxonomy of schools is to be of any use, one must sort each thinker and his works into one school or another. In this sorting process, however, the way in which a thinker may resist or reinterpret his own school is all too easily lost. The sorting of thinkers into this or that school may also lead one to ignore noteworthy differences and create false boundaries. Candrakīrti (c. 625 ce) and Śāntideva (c. 650 ce), for example, diverge significantly in their views, but since they are lumped together into the same school, their important differences may be ignored. Likewise, Dharmakīrti (c. 650 ce) and Śāntarakṣita (c. 750 ce) are sorted into distinct schools, but their thought may converge in ways that are not apparent in terms of their schools' definitions. Along these same lines, the taxonomy of schools does not fare well when confronted with liminal cases, where a thinker's allegiances are difficult to discern.

The English term school may also suggest a type of institutional coherence that does not apply to these philosophical traditions. It appears that only men wrote Buddhist philosophical texts, and nearly all were monks. As such, they received their sustenance through a monastic institution, and they held property in common with that institution. The practical circumstances of a monk's life and the norms that regulated his behavior were also guided by the rules passed to him upon ordination. If issues such as sustenance and behavioral norms lie at the core of a monk's institutional identity, then one must identify nearly every Buddhist philosopher first and foremost as a monk from a particular monastery regulated by the rules of a particular monastic tradition. Monastic traditions, moreover, were not distinguished by their philosophies; instead, each tradition was set apart primarily by its regional origin and the often mundane details of its rules. Hence, in institutional terms, an Indian Buddhist philosopher is first distinguished not by a particular philosophy, but rather by the regional affiliation and rules of his monastery's code. To put it another way, in some cases the color of a monk's robe indicated unambiguously the monastic tradition that he followed, but no such visible cue ever marked the philosophical school that he upheld.

Nevertheless, it would be a mistake to suppose that allegiance to a Mahāyāna philosophical school had no impact on a monk's life. Certainly, that allegiance located the philosopher within an intellectual community, one that extended across many monasteries in many monastic traditions; and it is clear that these intellectual communities engaged in both censure and approval of a thinker's works. A thinker's commitment to a Mahāyāna philosophical school also located him within a wider discourse on philosophy conducted by many traditions, both Buddhist and non-Buddhist. That is, although many Indian Buddhist thinkers did not compose their works in Sanskrit, Mahāyāna philosophers did use Sanskrit, and they thus shared a literary language with a wide range of non-Buddhist theorists. This may be one reason for the remarkable number of extended debates between Mahāyāna philosophy and these other, non-Buddhist traditions. In those debates, Mahāyāna Buddhist thinkers acted as intellectual defenders of the faith against philosophical critiques originating outside the Buddhist community, and this role probably impacted their lives in terms of patronage and prestige.

Another practical impact of allegiance to a school was the Mahāyāna notion of philosophy itself. Mahāyāna philosophical texts cover many of the same topics and use many of the same methods found in Euro-American philosophical traditions. And, as in classical Greece, philosophy here must not be interpreted as dry theory, but rather as systematic thought that is meant to explain, guide, and sustain contemplative practices. This does not mean that every argument correlates straightforwardly with a contemplative practice—consciously or not, Indian Buddhist thinkers often made philosophical decisions that have no obvious relation to such practices. Nevertheless, in a fundamental way Buddhist thinkers link their arguments to specific contemplative practices, such as meditations that analyze personal identity. This linkage reflects the avowed soteriological context of all Buddhist thought, namely, the cultivation of meditative experiences that allegedly eliminate suffering and lead to nirvāṇa. Indeed, from the traditional view, the Buddhist thinker's philosophical work was itself a kind of spiritual practice that moved the thinker closer to these final goals. Philosophy is therefore called "seeing" (darśana ), a metaphor that evokes a central goal of Buddhist contemplative practice: an experience (anubhava ) in which one sees things as they truly are (yathābhūtadarśana ). Thus, in speaking of Mahāyāna "philosophy," one must recall that it is implicated deeply in this type of contemplative goal.

Despite the problems noted above, the schema of the four philosophical schools remains useful. The usual procedure is to begin by discussing the first two schools: the Vaibhāṣika and the Sautrāntika. Doxographers maintain that these schools do not embody any Mahāyāna philosophy, and it is precisely for this reason that they are presented first. In doing so, doxographers are able to present the shared, foundational notions that run through all Buddhist thought.

Foundational Theories

From its earliest period, Buddhist thought rested on a straightforward set of claims about human goals and the means to achieve those goals. In brief, the main human goal is the elimination of suffering, and the means to that end is the elimination of the causes of suffering. The strands of early Buddhist thought that develop into Mahāyāna philosophy specify that suffering's cause is a type of "ignorance" (avidyā ), a distorted way of seeing the world that stems especially from misconceptions about personal identity. Ignorance creates and sustains mental dispositions that motivate and guide actions, and since those dispositions are rooted in a fundamental error, the actions guided by them are doomed to failure. Ignorance, moreover, permeates the minds of all unenlightened beings; hence, all of their actions—including those aimed at their highest goal of eliminating suffering—end in frustration.

The solution is to eliminate the fundamental misconceptions about one's personal identity that fuel ignorance, and one does so by demonstrating that their object, an essentially real and immutable "self" or ātman, does not exist. The procedure is to engage in a type of reductive analysis whereby, with the aid of contemplative practices, one searches through the constituents of body and mind in order to determine whether any of them—singly or in combination—could be such a self. Having seen that there is no such self to be found, one uses meditation to deepen that experience and explore all its implications. Eventually one becomes free of the misconceptions that create suffering; hence, one attains nirvāṇa, utter freedom from suffering.

This basic theory, which also lies at the core of Mahāyāna philosophy, is the main concern of the abhidharma, a style of Buddhist thought presented in great detail by the Vaibhāṣika school. The most basic of the four schools according to Indian and Tibetan doxographers, the Vaibhā-sikas derive their name from the Mahāvibāṣa (Great commentary) that is their inspiration. Their principal task is to articulate an elaborate taxonomy of all the truly real constituents of the body and mind in order to facilitate an exhaustive search for the self. These psychophysical constituents, called dharma s, are discovered through analysis to be the irreducibly real building blocks of the universe, and when one knows them as such, one is seeing mind and body as they truly are (yathābhūtadarśana ). Since a person is nothing other than those irreducible constituents of mind and body, and since no essential self or immutable identity is numbered among those constituents, one concludes that this alleged essential self (ātman ) is not truly real.

This theory of "no-self" (anātman ) is meant to demonstrate that no fixed essence lies at the core of personal identity, but it does not deny that in a contingent way, one can speak intelligibly of persons or selves. A traditional example is a chariot: when one performs the Vaibhāṣikas' reductive analysis of a chariot, one finds only the parts, such as the wheels, axle, and so on. At the same time, one knows that there is no chariot separate from those parts; if there were, it would absurdly follow that the chariot would still exist even after its parts were removed. Hence, even though it may seem that a chariot exists, if one accepts irreducibility as a criterion of true existence, one must admit that no such chariot truly exists. Nevertheless, one is still able to use the word "chariot" intelligibly when engaged in the practical task of, for example, driving the chariot. Thus, in terms of practical actions and use of language, a chariot does exist.

Codifying these two ways of existing, the Vaibhāṣika refers to another key concept for the Mahāyāna: the "two truths" or "two realities," namely the "ultimate" (paramārtha ) and the "conventional" (saṃvṛtti ). According to the Vaibhāṣikas, if one wishes to know whether an entity exists ultimately, then one employs their analytical techniques; if, at the end of that reductive analysis, the entity in question has not been reduced to some more fundamental constituents, one concludes that it is ultimately real. On the other hand, even if the entity is reducible to more fundamental constituents, one may decide that from a practical or linguistic point of view, it still appears to be existent. In such a case, the entity will be considered conventionally real because, although it does not withstand analysis, it does conform to the conventions that govern the use of language and practical actions. Thus, since it can be reduced to more fundamental constituents, a chariot is not ultimately real. Nevertheless, in terms of the conventions that govern the use of the word chariot, it appears to be real for practical purposes; hence, a chariot is conventionally real. Likewise, since a person can be reduced to more fundamental constituents, no person is ultimately real; nevertheless, in practical and linguistic terms, one can speak of a person as conventionally real.

In order for the schema of the two realities to make sense, the Vaibhāṣika must explain precisely what it means for one to know that an entity exists ultimately. Their view amounts to a kind of taxonomic atomism: an ultimately real entity is irreducible, and one has full knowledge of this fact when one sees that the entity, due to its essence or nature (svabhāva ), belongs to one or another of the irreducible categories that exhaustively account for all the stuff of the universe. In other words, the endpoint of the Vaibhāṣika analysis is not just that the thing in question cannot be broken down further, but also that one knows in an affirmative sense what it truly is by virtue of its nature; and one arrives at this knowledge by correctly categorizing the irreducible thing in question.

In emphasizing this taxonomic approach, the Vaibhāṣikas' method betrays a realist attitude toward categories. This realism attracts the criticism of the second non-Mahāyāna school, the Sautrāntikas, who critique it by pointing to its naïve assumptions. One such assumption is the belief that categories—or more generally, words and concepts—refer in some direct and straightforward way to real entities in the world, such that the things expressed by a particular word or concept are understood to be the same. For example, when one uses the word or concept blue, one appears to be referring to a thing that is somehow, by its nature, the same as all other blue things. In fact, say the Sautrāntikas, words and concepts do not refer in this way to real things. Thus, the seeming sameness of each thing called "blue" is an illusion; in actuality, each thing is utterly unique, and its unique identity or nature cannot be fully expressed through words or concepts.

The Sautrāntika critique resorts to complex and technical arguments, but to appreciate its relevance to the development of Mahāyāna thought, one need only attend to a main outcome: namely, that the Sautrāntika view moves away from the notion that all things are fixed in categorical identities. For the Vaibhāṣikas, the universe is composed of irreducible elements, each of which belongs by its very nature to a particular category. But according to the Sautrāntikas, the nature of a thing cannot be fully captured by a categorical identity. This leaves open the explicit possibility that any given thing is susceptible to multiple interpretations at the level of words and concepts.

General Trends and Problems in MahĀyĀna Thought

Examined through the traditional schema of the four schools, the first two schools—Vaibhāṣika and Sautrāntika—are usually discussed in the somewhat contrived and ahistorical manner presented just above. Such an approach scarcely does justice to these two "lower" schools, but it does capture an important facet of Mahāyāna thought: namely, that it is explicitly rooted in non-Mahāyāna Buddhism. Mahāyāna thinkers accept all of the elements discussed above: namely, that the elimination of suffering is a main spiritual goal; that ignorance is the primary cause of suffering; that ignorance is eliminated by knowing things as they truly are; that on the theory of no-self, persons are not ultimately real; and that an entity that is not ultimately real may nevertheless be considered conventionally real. Rather than rejecting these basic theories, Mahāyāna thinkers modify them in a way that creates a conceptual transition—not a radical discontinuity—from the non-Mahāyāna to the Mahāyāna.

A key element in this conceptual transition is a fundamental change in the notion of nirvāṇa. In non-Mahāyāna thought, nirvāṇa, the state in which suffering has utterly ceased, stands in strict opposition to saṃsāra, the world of suffering. Saṃsāra, moreover, is literally created by ignorance, and on most accounts, this means that everything within saṃsāra is tainted by ignorance. For the Vaibhāṣika, this taint is an irreversible and indisputable fact about the dharma s or fundamental building blocks that constitute saṃsāra. Nirvāṇa, on the other hand, is utterly free not only of suffering, but also of the ignorance that causes suffering. Hence, on the Vaibhāṣika view, saṃsāra and nirvāṇa must be entirely distinct.

In both literature and philosophy, Mahāyāna moves away from this strict distinction between saṃsāra and nirvāṇa. In a literary work such as the Vimalakīrtinirdeśa Sūtra, one learns that those close to true nirvāṇa are capable of seeing this world as a blissful paradise called a buddha-field, while those farther from that state still see it as a world of suffering. In the systematic texts of the first Mahāyāna philosopher, Nāgārjuna, one reads that in ontological terms, there is no difference whatsoever between saṃsāra and nirvāṇa. And in other systematic Mahāyāna works, nirvāṇa is redefined as "unlocated" (apratiṣṭhita) in that it is situated neither within the world of suffering that is saṃsāra, nor in a quietistic nirvāṇa that is diametrically opposed to that world. This new, nondualistic paradigm for nirvāṇa accompanies a redefinition of the highest goal for Buddhists. In short, for the Mahāyāna, the proper and highest goal of a Buddhist is not only the elimination of one's own suffering, but rather the attainment of buddhahood: a state of perfect bliss in which, while still active in a world that appears to be one of suffering, one is maximally efficient at leading other beings to nirvāṇa. Buddhahood is the goal that guides the bodhisattva ideal, the Mahāyāna's central ethical motif, which is based on a strong sense of compassion for all beings.

The Mahāyāna's new paradigm emphasizes the nonduality of saṃsāra and nirvāṇa, and to make good philosophical sense, it must be accompanied by a revision of the Vaibhāṣika ontology. It can no longer be the case that the fundamental building blocks of reality are fixed by their very nature in immutable identities; that is, it can no longer be the case that the stuff of the world of suffering must always remain what it has always been, namely, the direct or indirect product of ignorance. Instead, it must be the case that the world appears as it does—as wracked with suffering or as a blissful buddha-field—not because of some fixed and essential nature of things, but instead due to the minds of the beings that are experiencing that world. Here is the relevance of the Sautrāntika's critique of Vaibhāṣika thought: the nature of a thing cannot be fully captured by a categorical identity, and multiple interpretations of its identity are possible. It is crucial for Mahāyāna theory that the Sautrāntika critique be correct.

From a doxographical perspective, the Mahāyāna's ontological revision is facilitated by extending the doctrine of no-self. The doctrine of no-self rejects the notion that persons have a fixed, essential identity: one may seem to be an ultimately real person, but in fact, one is not truly or ultimately a person because one is reducible to the real, fundamental elements of which one is composed. For the Mahāyāna, a similar critique applies to the fundamental elements or dharma s that supposedly make up the person: an infinitesimal particle of matter, for example, seems to be an infinitesimal particle, but it is not truly or ultimately an infinitesimal particle. Indeed, according to the Mahāyāna, all of the Vaibhāṣikas' allegedly fundamental elements of the universe lack any fixed, essential identity as elements. All things are therefore completely mutable, and the world of suffering that is saṃsāra is not fixed in its nature: it can be the very locus of nirvāṇa.

By extending the critique of essential identity from persons to all the elements of the universe, Mahāyāna thinkers encounter three main issues. First, they must present a new style of critique that is not simply reductive; in other words, the claim that irreducible entities are not ultimately real cannot be supported by reducing them once more, since one will just arrive at the same problem. Instead, some other kind of analysis must be brought to bear. Second, Mahāyāna thinkers must specify what it means for one to see the true identity of things—to "see them as they truly are" (yathā-bhūtadarśana )— when that new analysis reaches its culmination. In other words, the Vaibhāṣika analysis leads to a straightforward and even intuitive conclusion: when one is looking at a chair, in fact what one is seeing is a bunch of irreducible particles of matter; the notion of a "chair" is just a convenient fiction. But if, as Mahāyāna thinkers maintain, even those irreducible elements are not truly real, what then is left for one to be seeing? This problem relates to the third issue: namely, that if even the fundamental building blocks of the world are not truly real, how then does one give an account of conventional reality? For the Vaibhāṣika, an entity such as a chair is not ultimately real because it can be reduced to its more fundamental parts. Nevertheless, in conventional terms one may speak of a "chair" as real, and one can do so because the term chair actually refers to those irreducible parts that are functioning together in a particular way. Thus, for the Vaibhāṣika, the warrant for claiming that a chair or a person is conventionally real is precisely the fact that one can point to the ultimately real elements of which it is composed. Mahāyāna thinkers, however, deny the ultimately reality even of those elements. Of what, then is conventional reality constructed?

These three issues—the need for a new style of analysis, a new account of knowing things as they are, and a new approach to conventional reality—all raise another issue: namely, that Mahāyāna thinkers seem to be arguing that, to at least some degree, the Vaibhāṣikas and other non-Mahāyāna philosophers are just plain wrong. Not only do they seem to argue that many Buddhists are wrong, but since Mahāyāna thinkers accept the Vaibhāṣikas' claim that their theories come from words of the Buddha, they seem to say that the Buddha was wrong too. To deal with this problem, followers of the Mahāyāna do not reject most of the previous canonical texts, perhaps in part because causing such a schism was considered as heinous as matricide. Instead, Mahāyāna thinkers sought a method to reconcile their innovations with the long established Buddhist community in which they were embedded. They settled on the notion of "skill in means" (upāyakauśalya).

Strictly speaking, "skill in means" may not be a philosophical concept, but it certainly functions as a philosophical method. In its most basic form, it amounts to this: the teaching must be tailored to the audience. That is, one presents theories and arguments at a level that the audience is capable of understanding, and if the audience cannot understand (or will inevitably reject) the highest level of one's philosophy, one uses a lower level of analysis that will prepare the audience to understand or accept the higher level. In part this means that arguments must be couched in such a way that they fit into a hierarchy of levels, and as Mahāyāna thought develops in India, this attention to levels of analysis becomes the central motif of late Mahāyāna thought.

Historical Synopsis

Considered in historical terms, the two Mahāyāna schools—Madhyamaka and Yogācāra—develop through the following stages: Nāgārjuna (c. 150 ce) composes the early Madhyamaka texts; Asaṅga and Vasubandhu (both c. 350 ce) compose the early Yogācāra texts; Dignāga (c. 450 ce) and Dharmakīrti (c. 625 ce), with the help of Bhāvaviveka (c. 500 ce), integrate a set of theories known as "Buddhist epistemology"into Mahāyāna thought; from the seventh century onward, Madhyamaka and Yogācāra subschools develop in reaction to the developments of Buddhist epistemology, and later Madhyamaka thinkers such as Śāntarakṣita and Kamalaśīla (both c. 750 ce) create a synthesis that explicitly employs a hierarchy of schools representing levels of analysis. The discussion below focuses especially on early Madhyamaka and Yogācāra, and it concludes with a brief examination of their later subschools.

Early Madhyamaka: Nāgārjuna

The first systematic Mahāyāna thinker was Nāgārjuna, and his historical primacy is matched by his philosophical importance. As noted earlier, to move beyond early Buddhist thought Mahāyāna thinkers confront three main needs: a new style of analysis that moves beyond reductionism, a new account of knowing things as they are, and a new approach to the definition of conventional reality. Nāgārjuna's approach to these issues sets the stage for all subsequent Mahāyāna thought.

To formulate a new style of analysis, Nāgārjuna must critique the claim that through a strictly reductive analysis, one comes upon things that are ultimately real. The early Buddhist style of reductive analysis is straightforward: one analyzes an entity by attempting to break it into its component parts, and if it cannot be broken down further, the entity is ultimately real. A chair, for example, is not ultimately real because it can be broken down into more fundamental parts; and when the analytical process is brought to its conclusion, one eventually arrives at irreducible, partless particles that are the basic stuff of the chair.

Thus, when a reductive thinker such as a Vaibhāṣika completes the analysis of a chair, he concludes that a chair is actually just many particles. Hence, in ultimate terms, a chair exists as something other than itself: what seems to be a chair is not ultimately a chair; instead, it is actually irreducible particles. But, for these reductive thinkers, an irreducible entity such as a particle does ultimately exist as itself because it cannot be reduced to anything more fundamental. As such, that entity has svabhāva, literally, "own-existence." To speak of a thing's svabhāva, therefore, is to speak of what a thing is in and of itself; in other words, it is to speak of its "essence," the best translation of svabhāva.

To move beyond reductive analysis, Nāgārjuna focuses on this notion of essence. He accepts that, for an entity to exist ultimately, it must have an essence (svabhāva ), but for him, to have an essence is not just a matter of being irreducible. Instead, he maintains that the notion of an essence is a way of indicating that the entity's identity is utterly devoid of any dependence on other entities. In short, he understands the notion of essence as independent or nonrelational existence. Hence, in lieu of reduction, his analysis examines the ways in which an entity might be dependent on other entities. If the entity is found to be dependent, then one must conclude that it lacks essence (svabhāva ) and is thus not ultimately real.

For Nāgārjuna, dependence comes in various forms. For example, he begins his most influential work by arguing that causally produced entities cannot have essences because they depend on their causes. His analysis of causality, however, is only part of a larger strategy: namely, the analysis of relations. Even entities that are not causally produced are susceptible to this analysis. Perhaps the most radical example is nirvāṇa itself, which reductionists such as the Vaibhāṣika consider to be an ultimately real element that is free of any causal conditioning. Nāgārjuna, however, maintains that it is conditioned in another sense: one cannot give an account of what nirvāṇa is in itself without referring to its opposite, the world of suffering that is saṃsāra. In other words, nirvāṇa has no meaning without saṃsāra, just as "long" is meaningless without "short." Nāgārjuna thus concludes that "nirvāṇa is not at all different from saṃsāra " (Mūlama-dhyamaka-kārikā 25.19).

When Nāgārjuna radically rejects any distinction between saṃsāra and nirvāṇa, he is not espousing some type of monism. Instead, he is drawing a consequence from a more fundamental point, namely, that saṃsāra and nirvāṇa lack essence. That is, in order to draw a distinction between them in ultimate terms, one must do so in terms of their essences—what each is in itself without depending on anything else. Saṃsāra and nirvāṇa, however, both lack essence because the identity of each is dependent on the other. Hence, any attempt to draw any ultimate distinction between them must fail.

More important is another conclusion of Nāgārjuna's analysis: since only an entity with an essence can be ultimately real, saṃsāra and nirvāṇa are not ultimately real. Nāgārjuna goes on to extend this analysis not just to saṃsāra and nirvāṇa, but to all things, and the upshot of his critique is that they all lack essence. In other words, to have an essence is to have some fixed, nonrelational identity, and no entity can fulfill this requirement. Moreover, since only a nonrelational entity—that is, one with an essence—could be ultimately existent, Nāgārjuna maintains that no entities whatsoever exist ultimately. To know all things as they are truly or ultimately is therefore to recognize that none exist ultimately.

Here one encounters the second issue that all Mahāyāna thinkers must face, namely, the need for a new account of "seeing things as they truly are." As with the Vaibhāṣika, Nāgārjuna accepts that suffering can only be stopped by eliminating ignorance, and that to eliminate ignorance one must see things as they truly are. For the Vaibhāṣika, to see things as they truly are is to experience what is ultimately real, namely, the foundational elements of the universe. In doing so, one can eliminate ignorance: the confused belief that somewhere among those elements one will find one's absolute, fixed identity or self (ātman ). For Nāgārjuna, however, ignorance is not just a confusion about one's personal identity; instead, it is the deeply ingrained cognitive habit that makes beings see all things as if they had some fixed, absolute identity or essence (svabhāva ). Thus, to eliminate ignorance one must realize that that no entity has any such essence, and this means that one must realize that no entity is ultimately real. But if no entity is ultimately real, what does it mean to see things as they truly are? At the end of the analysis, what is left that one could see?

To answer this question, Nāgārjuna employs a metaphor that runs throughout Mahāyāna thought. Inasmuch as no entity can have a nonrelational identity, every entity lacks essence, and Nāgārjuna speaks of this lack of essence as "emptiness" (śūnyatā ). Thus, to know an entity in ultimate terms is to know its emptiness, which is a metaphor for its utter lack of essence. The danger, however, is that one will construe this as some kind of absolute nothingness at every entity's core. In that case, Nāgārjuna would be wrong to say that all things lack essence because they would have an essence, namely, that absolute nothingness. Responding to the danger of this type of nihilistic interpretation, Nāgārjuna points out that even emptiness lacks essence and is thus ultimately unreal. Thus, just as a person is empty of really being a person, emptiness is empty of really being emptiness. By understanding this "emptiness of emptiness" (śūnyatā-śūnyatā ), one avoids nihilism.

Nāgārjuna may avoid nihilism, but many questions remain concerning the realization of things as they truly are. Here one should recall that such a realization comes not only through Nāgārjuna's arguments, but also through their integration into a contemplative practice. But what sort of practice would it be? What kind of meditative experience would the arguments help to induce? It should already be clear that the meditation on things as they truly are—that is, the meditation on emptiness—cannot be an experience of some absolute nothingness or any other negative content. It would also seem problematic to hold that the meditation has positive content, such as an object. That is, the meditative experience of emptiness is an experience of any entity's ultimate reality, and if that experience is of some object, then one might conclude that the object experienced was the fixed, ultimate essence of that entity. This would seem to contradict Nāgārjuna's notion that all things lack essence. Hence, the meditative experience of emptiness apparently can be neither of something, nor of nothing.

This conundrum of emptiness clearly vexes subsequent Mahāyāna thinkers, and it leads to many developments in Mahāyāna thought. It also points to problems in the third issue that Nāgārjuna faces: an account of the conventional. As noted above, on Nāgārjuna's view, if one seeks the fixed, nonrelational essence that would constitute the ultimate identity of an entity, one fails to find any such essence. And to exist ultimately, a thing must have such an essence; hence, one concludes that no entity exists ultimately. But as with the Vaibhāṣika, Nāgārjuna maintains that an entity that does not exist ultimately may nevertheless exist conventionally. Hence, even though he denies the ultimate reality of all things, including the Buddhist path, he does not at all mean to deny that many such things, most especially the Buddhist path, are real and valuable in a conventional sense.

Concerning the conventional, the Vaibhāṣikas are straightforward: it is just a matter of recognizing that words such as "chair" are convenient fictions that allow us to speak easily of what is really there, namely, many irreducible particles. Thus, conventionally real things are composed of the irreducible, ultimately real stuff of the universe. But for Nāgārjuna, there is no such stuff, nor does one find anything else that is "really there" in the case of a chair or anything else. How then does one make sense of conventional reality?

To answer this question, Nāgārjuna must redefine the notion of conventionality. For the Vaibhāṣika, a conventional entity depends on the ultimate because it is made from ultimately real stuff, but for Nāgārjuna the conventional and the ultimate define and depend upon each other through their mutual exclusion, as in other dyads such as "long" and "short" or saṃsāra and nirvāṇa. And since an ultimately real entity has an utterly independent or nonrelational identity, a conventionally real entity must be its antithesis: its identity is utterly dependent or relational. As Nāgārjuna puts it, "We say that emptiness is that which is interdependence" (Mūlamadhyamaka-kārikā 8.24). In other words, when one sees that all entities fail the test of ultimacy because they are all empty of any nonrelational identity, one should also realize that if they have identities even conventionally, those identities must be rooted in the radical relationality that is "interdependence" (pratītya-samutpāda ).

In seeing the conventional as interdependence, Nāgārjuna sets a theme for all subsequent Mahāyāna thought, but as with the notion of emptiness, he leaves many questions unanswered. For example, he has not addressed the Vaibhāṣikas' basic intuition that the conventional is made up from ultimate stuff. In other words, the modality of conventional reality may indeed be interdependence, but does not such a concept presuppose that there are things standing in the relation of interdependence? A relation cannot exist without relata, so how could it make sense to speak of the relation that is interdependence if there are not really any entities to be related? These questions, along with the problems of knowing emptiness, create fertile ground for the growth of Mahāyāna thought.

Early Yogācāra: Asaṅga and Vasubandhu

Not long after Nāgārjuna, Asaṅga and Vasubandhu develop the other major strand of Mahāyāna thought, the Yogācāra (literally, "Practice of Yoga"). Also called citta-mātra ("mind only"), this school emphasizes mind (citta ) or consciousness (vijñāna ) in its responses to the three issues mentioned above, namely, style of analysis, seeing things as they truly are, and an account of the conventional.

In terms of analysis, Asaṅga and Vasubandhu largely follow Nāgārjuna's lead. Asaṅga's main style of analysis is also relational, but his focus is different. He begins with the assumption that the notion of an ultimately real personal identity is rooted in one's sense of subjectivity. His analysis is relational because on his view, the reality of that subjectivity is tied to the reality of the objects that it allegedly knows. Hence, in order to follow previous Buddhist thinkers in demonstrating that there is no such fixed, ultimately real personal identity, he must show that all the objects allegedly perceived by this subjectivity are ultimately unreal. Therefore, the subjectivity must also be ultimately unreal because it can truly be a subjectivity only if it perceives objects.

Focusing on the relation between subject and object in this way, Asaṅga follows Nāgārjuna in extending the critique of essence beyond persons to the fundamental elements of the universe, but unlike Nāgārjuna, his approach creates a more obvious bridge between non-Mahāyāna and Mahāyāna thought. That is, as with non-Mahāyāna thought, the main goal is still to refute one's notion of an ultimately real self (ātman ), but Asaṅga suggests that this goal is best reached by critiquing the self's alleged objects. The intuition here is that the false impression of ultimately real selfhood is rooted in one's sense of subjectivity as a perceiver of objects. Thus, in showing that there are no ultimately real elements that could serve as objects, Asaṅga is just offering a more profound and effective rejection of any fixed, absolute personal identity.

Asaṅga's strategy requires some means to refute the reality of any entities that could be construed as objects existing in distinction from the subjectivity that perceives them. For the Vaibhāṣikas, those objects basically consist of the irreducible elements that they take to be ultimately real; hence, Asaṅga's critique must demonstrate that all those elements are not ultimately real. Buddhist models of consciousness require that his analysis cover two general types of objects: mental objects, which are immaterial, and sense objects, which, being material, are allegedly composed of irreducible particles.

In terms of mental objects, Asaṅga develops a critique elaborated further by Vasubandhu. As with Nāgārjuna, the critique employs a relational analysis whereby an entity could be ultimately existent only if it is utterly free of dependence on other entities. Exploring an area not systematically addressed by Nāgārjuna, Asaṅga and Vasubandhu refute the ultimate reality of mental objects by demonstrating that their allegedly independent existence is contradicted by the linguistic and conceptual relationality that enables them to be mental objects in the first place. Their detailed arguments demonstrate that the referent of a word or concept is necessarily mind-dependent to at least some degree, and being dependent, that referent cannot be ultimately real.

The critique of sense objects, which thinkers such as the Vaibhāṣika take to be composed of irreducible particles, is developed especially by Vasubandhu. Here too the continuity with non-Mahāyāna thought is striking, for Vasubandhu chooses to employ a reductive analysis to demonstrate the ultimate irreality of such particles. But in contrast to reductionists such as the Vaibhāṣika, Vasubandhu is willing to reduce matter to the point where it no longer exists. In short, he demonstrates that irreducibility is incompatible with material existence: if material particles were irreducible, then they could have no size, but if they have no size, then how could an accumulation of them form gross objects such as jars or chairs? On the other hand, if they do have size, then they clearly are not irreducible, since they then must have parts, such as front, back, left, right, top, and bottom. The conclusion of this analysis is a philosophical idealism that totally denies the existence of matter. And although a few academic interpreters maintain that Vasubandhu does not mean to refute the reality of matter, such an interpretation ignores all subsequent Indian commentators and all Tibetan doxographers.

The critique of mental and physical objects developed by Asaṅga and Vasubandhu differs in an important respect from Nāgārjuna's approach. For Nāgārjuna, when one reaches the conclusion of the analysis, one's "seeing things as they truly are" is just seeing emptiness, which is not readily construed as seeing anything at all; indeed, later Indian Madhyamakas will speak of it as the "seeing that is non-seeing." Asaṅga and Vasubandhu, however, see their analysis as ending in a realization that has a far more affirmative content. The positive nature of that realization is shown by the way they redefine emptiness. For Nāgārjuna, emptiness is utter lack of essence, and since all things are empty, all things lack essence; hence, they are all ultimately unreal. For Asaṅga and Vasubandhu, emptiness is the absence of subject-object duality in the mind of the perceiver. Thus, although ultimately real objects do not stand in opposition to some ultimately real subject, there remains nevertheless the undeniable fact of consciousness itself.

Asaṅga and Vasubandhu explain their new notion of emptiness through the theory of the "three natures" (trisvabhāva ). The usual order of enumeration is: the constructed nature (parikalpitasvabhāva ), the dependent nature (paratrantrasvabhāva ), and the perfect nature (pariniṣ-pannasvabhāva ). It is helpful to place the dependent nature at the head of this list because the other two—the constructed and the perfect—are actually two different modes of the dependent nature. That is, the dependent nature is the causal flow of consciousness itself: it is the sequence of one moment of consciousness produced by its own previous moment and going on to produce its own subsequent moment. This ongoing stream of consciousness can appear in two different modes. For ordinary persons, it appears with a dizzying variety of sensory and mental objects, and each mental moment except for the deepest sleep is replete with such an object. Asaṅga's and Vasubandhu's analysis of mental and sensory objects, however, demonstrates that none of these objects is ultimately real. Nevertheless—and this is the key ontological claim—the conclusion that the objects are ultimately unreal does not adequately account for the fact that those objects are appearing. Instead, one must see that denying the ultimate, independent reality of those objects is the same as affirming their conventional, dependent existence within the mind itself. To put it another way, when one sees the color blue, the apparent existence of the blue object as an external, independent object is false. But the fact that it is appearing to consciousness is undeniable, and since the Yogācāra analysis shows that it could not be external and independent of the mind, it must be within the mind itself. Seeing that flow of mind in that way—namely, as devoid of the apparent subject-object duality—is to see the perfect nature. Thus all objects and all subjects are not distinct, but this is not to deny their reality altogether. Rather, the denial of subject-object duality still leaves intact the causal flow of mind in which all those apparently dualistic experiences are occurring.