Madhyamaka

← Encyclopedia

Madhyamaka is the Buddhist philosophical school that holds all phenomena to be empty of inherent existence — śūnyatā, emptiness — and that takes this emptiness, rather than any positive doctrine of its own, as the heart of what the Buddha taught. The name means “the school of the middle,” and the middle in question runs between two errors: the claim that things truly exist on their own terms, and the claim that they do not exist at all. It is a strand of the Mahāyāna, the “great vehicle” of Buddhism that took form in the early centuries of the Common Era, and within that movement it became the philosophical instrument by which the new Prajñāpāramitā (“Perfection of Wisdom”) literature, with its insistence that all things lack a fixed essence, was given rigorous argumentative form.

The founder and the founding text

The school traces itself to Nāgārjuna, a thinker active in southern India around the second century CE, about whom little can be established beyond the texts attributed to him; the life and the legend are treated under his own name, and need not be retold here. The chief of these texts, the Mūlamadhyamakakārikā — “Root Verses on the Middle Way” — proceeds not by asserting a system but by dismantling them. In twenty-seven chapters of terse Sanskrit verse it takes up causation, motion, the agent and the act, the senses, the elements, desire, time, the bound and the freed, the constituents of the person, the Buddha himself, the four noble truths, and nirvāṇa — the whole inventory of categories that earlier Buddhist analysis had used to take the world apart — and argues in each case that the thing examined cannot bear the independent, self-standing existence ordinarily ascribed to it. Under pressure every such notion dissolves into contradiction.

The method is dialectical and deliberately negative. It refutes positions rather than replacing them, on the reasoning that emptiness is not one more thesis but the absence of any thesis to cling to. Nāgārjuna says as much in a famous formula of the Mūlamadhyamakakārikā: those for whom emptiness is itself a view, a position to be held, he declares incurable. The technical edge of the procedure lies in the kind of negation it employs. A negation can be implicative — denying one thing in order to affirm another, as “this is not white” suggests some other color — or it can be non-implicative, denying without smuggling in a hidden positive in its place. Madhyamaka trades almost wholly in the second. When it says a thing has no own-being, it is not pointing past the thing to a truer entity behind it; it is removing a misdescription and leaving nothing in its place. The same discipline governs the recurring four-cornered exhaustion, the catuṣkoṭi, in which a thesis, its negation, their conjunction, and their joint denial are each shown untenable, so that the mind is left with no corner of the square to come to rest on. The school’s central explication of emptiness — what śūnyatā and svabhāva mean, and why their denial is liberating rather than nihilistic — is developed in the companion entry on Madhyamaka and śūnyatā; what follows here concerns the school as a historical body: its dialectic, its divisions, and its travels.

The two truths

Holding the dialectic together is the doctrine of two truths. Conventionally — saṃvṛti-satya, the truth “of the everyday” or “of concealment” — the world is real enough; things arise, function, and pass away by dependent origination (pratītyasamutpāda), and to deny this is the nihilism the school rejects. Ultimately — paramārtha-satya — nothing is found to possess the fixed own-being that language and common sense impute to it. The two are not rival accounts but one reality described at two levels, and the relation between them is exact: because things arise only in dependence on causes, conditions, and the parts and concepts by which they are picked out, they are empty of self-nature; and because they are empty of self-nature, they are free to arise in dependence. Emptiness and dependent origination name the same fact from two sides. Nāgārjuna draws the sharpest possible conclusion from this: without reliance on the conventional, he writes, the ultimate cannot be taught, and without grasping the ultimate, nirvāṇa is not reached. The everyday is not a ladder to be kicked away. It is the only medium in which the ultimate can be communicated at all.

Crucially, emptiness itself is held to be empty. It is not a hidden absolute behind appearances, not a Ground or a One or an ineffable plenum that the negations clear a path toward. To treat emptiness as a final reality — the trap Nāgārjuna’s incurable patient falls into — would be to reintroduce, at the last moment, exactly the self-standing essence the analysis was meant to dissolve. The emptiness of emptiness is the school’s guarantee that it has built no metaphysics of its own. There is no thing that emptiness is; there is only the way things are, which is to say the absence in them of any way they have to be.

The Indian divisions

The first systematic commentators on Nāgārjuna’s verses included Buddhapālita and Bhāvaviveka in the fifth and sixth centuries, and from their disagreement the school’s lasting methodological fault line opened. Buddhapālita read the Root Verses as pure reductio: the Mādhyamika does nothing but draw the unwanted consequences (prasaṅga) latent in an opponent’s own commitments, and advances no claim that could be pinned to him. Bhāvaviveka objected that mere consequences prove nothing unless the Mādhyamika is willing to mount arguments of his own — formal, autonomous (svatantra) inferences with premises he himself endorses. Candrakīrti, in the seventh century, answered Bhāvaviveka on Buddhapālita’s behalf, arguing that to advance an autonomous inference is already to concede that its terms exist in the shared, determinate way Madhyamaka denies; the consistent Mādhyamika can therefore own no positive thesis whatever, not even the thesis that things are empty, except as the internal undoing of what the opponent has granted.

Tibetan doxographers later named the two stances Prāsaṅgika (Candrakīrti’s, “consequentialist”) and Svātantrika (Bhāvaviveka’s, “autonomist”). The labels are a retrospective sorting: the Indian philosophers did not wear them, and modern scholarship treats the pair as a Tibetan ordering device imposed on a more tangled Indian record rather than as two self-conscious parties. The distinction nonetheless carried real philosophical weight — it turns on whether emptiness can be demonstrated or only elicited, and on how much of ordinary epistemic machinery the Mādhyamika may borrow without betraying the position — and once sharpened it became one of the most consequential questions in the later tradition.

Madhyamaka did not develop in isolation. Its great counterpart within the Mahāyāna was the Yogācāra or “mind-only” school of Asaṅga and Vasubandhu, which analyzed experience in terms of consciousness rather than emptiness; later Indian thinkers such as Śāntarakṣita worked to synthesize the two, producing the Yogācāra-Madhyamaka that the earliest Tibetan transmission received. Against the older Theravāda and the broader Abhidharma analysis of reality into irreducible momentary factors (dharmas), Madhyamaka pressed its most pointed challenge: it turned the Abhidharma’s own reductive method back on the dharmas themselves, arguing that the ultimate building blocks were no more self-standing than the wholes they were meant to explain.

The Tibetan reception

Carried north, Madhyamaka was received in Tibet as the summit of Buddhist philosophy, the highest of the graded “tenet systems” through which the curriculum ascended. There Candrakīrti’s reading came to dominate. From roughly the eleventh century his Prasannapadā and Madhyamakāvatāra were studied as the definitive account of Nāgārjuna, and the Prāsaṅgika position was generally ranked the final and correct view. The most influential systematization came from Tsongkhapa in the fourteenth and fifteenth centuries, whose Geluk school took Prāsaṅgika-Madhyamaka as definitive while insisting, against quietist readings, that emptiness leaves conventional reality fully intact and ethically binding — that to see through own-being is precisely not to dismiss the world but to understand how it works. Other lineages of Tibetan Buddhism read the same texts toward a more positive “emptiness of other” (gzhan stong), in which the ultimate is empty of everything extrinsic to it but not empty of its own luminous nature — a reading Tsongkhapa’s heirs rejected as a relapse into the very absolutism Madhyamaka was built to refute. The dispute over the precise sense in which things are “empty” has remained alive among the Tibetan schools ever since, and the analytic machinery of the Root Verses passed, through the yogic lineages of figures such as Milarepa and the Indian mahāsiddhas before them, into a tradition where the dialectic was not only debated in the courtyard but rehearsed in meditation.

The soteriological point

What the analysis was for was never merely theoretical. Seeing through the apparent solidity of self and world was held to loosen the grasping that the tradition diagnoses as the root of suffering. The reasoning is austere and direct: one clings to what one takes to be real in its own right, and fears its loss in proportion to that taking; show that nothing — least of all the self — was ever self-standing to begin with, and the grip has nothing left to close on. Emptiness, in this telling, is liberating because it removes what there is to fear losing. This is the sense in which Madhyamaka belongs not to metaphysics as a spectator sport but to the long Buddhist project of release, and the sense in which its insight overlaps with the broader vocabulary of liberating knowledge — the gnosis by which a tradition holds that seeing rightly is itself the freedom, not a means to it. The negations do their work on the one who performs them.

Comparison and its limits

Western readers have repeatedly compared the negative dialectic to strands of skepticism, to apophatic theology, and to the via negativa of mystical writing — the unsaying that runs from Pseudo-Dionysius the Areopagite through Meister Eckhart and the Christian heirs of Neoplatonism. The comparisons are suggestive and have been pursued at length. They are not equivalences. The apophatic theologian unsays predicate after predicate of a reality believed to exist supremely, clearing the names away so that what exceeds them may be approached in its fullness; the negation is in service of a presence too great for speech. Madhyamaka unsays inherent existence itself, and declines the further step of locating any absolute behind the negations — its emptiness is emphatically not a name for God, nor for the One, nor for a Ground of being. The grammar of denial rhymes across the traditions; what each tradition means by the denial does not. The contrast is instructive precisely where it is exact: two great enterprises of saying-by-unsaying, arriving at opposite destinations by similar roads, one toward a plenitude beyond predication and the other toward the dissolution of the very assumption that there is anything, anywhere, that stands on its own.

The same independence of mind that set Madhyamaka apart from its Western cousins also distinguishes it within the Indian field. The nondual Shaiva metaphysics of Abhinavagupta reads the play of appearances back to a single supreme consciousness that is utterly real; the dualist Vedānta of Madhva reads its scriptural inheritance toward an enduring plurality of real substances ranged under a sovereign God. Madhyamaka refuses both terminations at once: there is no hidden unity to dissolve into and no irreducible manifold to come to rest on, only dependent arising, empty of own-being all the way down.

Texts, editions, and scholarship

The Mūlamadhyamakakārikā survives in Sanskrit chiefly through the verses embedded in Candrakīrti’s commentary, the Prasannapadā (“Clear Words”), and in Tibetan and Chinese translation; the kārikās are conventionally counted as some four-hundred-odd stanzas across the twenty-seven chapters. A freely readable English rendering of all twenty-seven chapters is available through the Mūlamadhyamaka-kārikā translation hosted at wisdomlib. The standard scholarly translation into English, made from the Tibetan together with a verse-by-verse commentary reflecting the tradition through which the text was transmitted, is Jay L. Garfield’s The Fundamental Wisdom of the Middle Way (Oxford, 1995). For a sustained philosophical reconstruction — treating Nāgārjuna’s denial of svabhāva as a coherent argument with consequences for metaphysics, epistemology, the self, language, and truth — the reference work is Jan Westerhoff’s Nāgārjuna’s Madhyamaka: A Philosophical Introduction (Oxford, 2009); Westerhoff’s analysis of the three senses of own-being is set out in his study of the Madhyamaka concept of svabhāva. Paul Williams’s Mahāyāna Buddhism: The Doctrinal Foundations (2nd ed., Routledge, 2009) places the school within the wider Mahāyāna and remains the standard survey. On the Tibetan division that came to organize the whole later reading of the school, the essential collection is Georges Dreyfus and Sara McClintock’s The Svātantrika-Prāsaṅgika Distinction (Wisdom, 2003), which establishes that the categories are a Tibetan ordering of a more fluid Indian record. The Stanford Encyclopedia of Philosophy’s entries on Madhyamaka and on Nāgārjuna give current scholarly overviews with full bibliographies. Across all of these, one fault line of interpretation recurs: whether the Root Verses are best read as a skepticism that suspends all theses, as a deconstruction of metaphysics that clears the field of essences, or as a mysticism reached by argument rather than by vision — disagreements as old as the text, which it sustains because its conclusions are stated as denials rather than as a position to be defended.

The school’s enduring puzzle is how a philosophy that refuses every position could be a position at all. It is a question the Mūlamadhyamakakārikā raises against itself, in the verse where Nāgārjuna allows that if he had any thesis he would be at fault — and then adds that he has none. The answer the tradition gives is that the refusal, too, lets go of itself: the emptiness that empties all things does not stop short at its own door, and the hand that opens to release the world does not close again around the act of opening.

→ Related: Buddhism Theravada · Milarepa · Gnosis · Nagarjuna · Candrakirti · Madhyamaka Sunyata · Tibetan Buddhism · Tsongkhapa · Vasubandhu · Apophatic Theology