Candrakīrti

← Encyclopedia

Candrakīrti was an Indian Buddhist philosopher of the Madhyamaka, or “Middle Way,” school, active around the first half of the seventh century. Little is known of his life with any security; the conventional dates, roughly 600 to 650, are inferences from the texts he answered and the texts that answered him rather than from records of the man. He stands at one of the densest junctions in the Indian philosophical record — close enough to the logicians Dignāga and Dharmakīrti to argue against their account of valid cognition, late enough that Nāgārjuna’s root verses had already been commented on for two centuries before he took them up — and yet the biography itself is almost entirely a later construction. Tibetan accounts surround him with the marvels owed to a master of his rank: that he drew milk from a painted cow to feed the monastery, that he bested a brahmin debater so completely the man’s tradition could not recover the point. These are the tradition’s own portrait of what such a mind was, told in the register of wonder rather than chronicle. The one fixed institutional claim the later sources make is that he taught at Nālandā, the great monastic university of northeastern India whose colleges set the curriculum for Mahāyāna learning across the Buddhist world — and even this rests on Tibetan biography written long after, not on any document from his century.

The two pillars

His standing rests on two works. The Prasannapadā — “Clear Words” — is a commentary on the Mūlamadhyamakakārikā, the founding verses in which Nāgārjuna, some four centuries earlier, had argued that all things are empty of any independent, self-standing nature. The accident of its survival has made it indispensable. The Prasannapadā is the only commentary on the Root Verses that comes down in the original Sanskrit, and because Candrakīrti embeds the verses he glosses, his commentary is the principal witness to Nāgārjuna’s own text in the language it was composed in; the rest of the early Madhyamaka commentarial library survives only in Tibetan and Chinese translation. To read the Mūlamadhyamakakārikā in Sanskrit at all is, in large measure, to read it through Candrakīrti’s transcription of it.

The Madhyamakāvatāra, “Entrance to the Middle Way,” sets out the same view as a graded path. It is the earlier of the two — the Prasannapadā refers the reader back to it repeatedly for the full version of arguments it only sketches — and it is cast not as line-by-line commentary but as an independent treatise in verse, to which Candrakīrti supplied his own prose auto-commentary, the bhāṣya. Where the Root Verses dismantle one category after another in no settled order, the Madhyamakāvatāra hangs the whole analysis on the bodhisattva’s ascent through the ten stages, the bhūmis, each correlated with one of the perfections. Nine of the ten stages are treated briefly. The sixth, the stage of the perfection of wisdom, swells to the size of all the others combined — well over two hundred verses — and it is here that the emptiness analysis proper is conducted: the refutation of arising from self, from other, from both, and from neither; the dismantling of the self through the figure of the chariot, which is neither identical with its parts nor different from them nor their mere collection nor their shape, and so is found, on inspection, to be a name resting on nothing self-standing at all. The Madhyamakāvatāra survives complete only in Tibetan translation, and it became, in that language, the single most studied textbook of the school. Across both works Candrakīrti read Nāgārjuna in the strictest possible terms: emptiness is not a refined something behind appearances but the simple absence of own-being — svabhāva — in anything whatever, the self included. The full doctrinal anatomy of that claim belongs to the account of śūnyatā; what is Candrakīrti’s own is the refusal of any exception to it, the insistence that the analysis stops nowhere short of total.

The method that became his name

The position carried a method, and the method became his signature. The dispute was not new with him. A century earlier the commentator Buddhapālita had read the Root Verses as pure reduction to absurdity — the Mādhyamika does nothing but draw the unwanted consequences latent in an opponent’s own commitments, and advances no claim that could be pinned to him. Bhāviveka had objected that bare consequences prove nothing unless the Mādhyamika is willing to mount arguments of his own: formal, autonomous inferences with premises he himself endorses. Candrakīrti answered Bhāviveka on Buddhapālita’s behalf, and pressed the point to its limit. To advance an autonomous inference, he argued, is already to grant that its terms exist in the shared, determinate way that establishes a thesis for both parties — and that is exactly the way of existing the Mādhyamika denies. A consistent Mādhyamika can therefore own no positive thesis of his own, not even the thesis that things are empty, except as the internal undoing of what the opponent has already conceded. The school should only draw out the unwanted consequences hidden in an opponent’s claims, asserting nothing on its own account.

The same severity governs his quarrel with the epistemologists. Dignāga had built a rigorous system of valid cognition — perception and inference as the two instruments by which truth is fixed — and Bhāviveka had imported its machinery into Madhyamaka argument. Candrakīrti refused the import. To grant the instruments of knowledge a self-standing authority, fixed independently of what they know, is once more to smuggle own-being back in through the side door; the conventional means of knowing are real only conventionally, mutually dependent with their objects, and carry no privilege the analysis must respect. By the same logic he turned on the Yogācāra of Vasubandhu and his heirs, the mind-only reading that resolved the world into consciousness: to make consciousness the one real thing that appearances depend on is to leave one essence standing where the analysis requires that none remain. Tibetan doxographers later named Candrakīrti’s stance Prāsaṅgika, the consequentialist approach, in contrast to the Svātantrika, the autonomist approach of Bhāviveka — a distinction those philosophers did not use of themselves, and one modern scholarship treats as a retrospective sorting imposed on a more tangled Indian record rather than a label they wore.

A debt to Nāgārjuna, paid in full

What Candrakīrti added to Nāgārjuna was not new doctrine but a refusal to soften. Where later readers had looked for a place to set the analysis down — a level of discourse where the Mādhyamika might at last say something in his own voice, a positive thesis salvaged from the wreck of every other — Candrakīrti closed each exit in turn. He preserved the catuṣkoṭi, the four-cornered exhaustion that denies a thesis, its negation, both together, and neither, so that no corner of the square remains to rest on; and he extended it without remainder to the self, to the means of knowledge, and to emptiness itself. The two truths he kept intact — the conventional, in which things function ordinarily, and the ultimate, in which no thing is found to have an essence — but he denied that the conventional could ever be propped up by some quiet residue of own-being to make it stable. It is stable enough as it is, dependently, and it needs no more. His Mahāyāna was a Buddhism in which the dialectic does not arrive anywhere it can stop, because there is nowhere self-standing to arrive.

A modest hearing, a decisive afterlife

In India his influence appears to have been modest in his own time and grew afterward. Through the seventh, eighth, ninth, and tenth centuries he is little cited; the synthesis of Madhyamaka with Yogācāra worked out by Śāntarakṣita and his successors — the line Candrakīrti’s epistemology had been built to refute — was the reading that carried into the first Tibetan transmission. His recovery came late and from outside the centers that had ignored him. From roughly the eleventh century, as his texts were translated and retranslated into Tibetan and studied in Kashmir, his star rose; in Tibet it did not stop rising. There his texts were studied as the summit of philosophical Buddhism, and the Gelug tradition founded by Tsongkhapa in the fourteenth and fifteenth centuries took Prāsaṅgika-Madhyamaka, in Candrakīrti’s reading, as the final and correct account of how things are — placing the Madhyamakāvatāra at the center of the monastic curriculum, where it remained the standard textbook on emptiness for the colleges of Tibetan Buddhism. The analytic temper he sharpened in the courtyard did not stay there; it passed, through the yogic lineages that run back to the Indian adepts such as Tilopa, into a setting where the same view of the empty self was not only argued but rehearsed. What had been one voice in a seventh-century Indian debate became, for much of Tibetan Buddhism, the standard against which every lesser view was measured.

Texts, editions, and scholarship

A point of caution governs the bibliography from the start: a second Candrakīrti exists in the record, a tantric author of the Pradīpoddyotana, a commentary on the Guhyasamāja Tantra belonging to the Ārya tradition of esoteric exegesis. He is a different man of the same name, and the Madhyamaka philosopher should never be read through his work or credited with it.

The Prasannapadā was first edited in Sanskrit by Louis de La Vallée Poussin in the Bibliotheca Buddhica series, Mūlamadhyamakakārikās de Nāgārjuna avec la Prasannapadā commentaire de Candrakīrti (St. Petersburg, 1903–1913) — the edition that put Nāgārjuna’s verses, as Candrakīrti transmits them, before modern scholarship for the first time. La Vallée Poussin also edited the Madhyamakāvatāra from the Tibetan in the same series (1907–1912) and rendered parts of it into French. A century on, the philological standard for the Prasannapadā has been reset by Anne MacDonald’s critical edition of its first chapter, In Clear Words: The Prasannapadā, Chapter One (Vienna: Austrian Academy of Sciences Press, 2015), which works from the full manuscript tradition and corrects the older text at many points. The complete Madhyamakāvatāra was first put into English by C. W. Huntington with Geshe Namgyal Wangchen in The Emptiness of Emptiness: An Introduction to Early Indian Mādhyamika (Honolulu: University of Hawaii Press, 1989), which prints the treatise with extensive notes; the bhāṣya and the text together also circulate in the Padmakara Translation Group’s Introduction to the Middle Way (Shambhala, 2002), set against the nineteenth-century Tibetan commentary of Jamgön Ju Mipham.

For the seventeenth-century reader of Nāgārjuna in the original, the verses embedded in the Prasannapadā are the foundation of every modern edition; a freely readable English rendering of all twenty-seven chapters of the Mūlamadhyamakakārikā is hosted at wisdomlib. On the method-dispute that organizes the later reception of the whole school, the essential collection is Georges Dreyfus and Sara McClintock’s The Svātantrika-Prāsaṅgika Distinction: What Difference Does a Difference Make? (Boston: Wisdom, 2003), which establishes that the categories are a Tibetan ordering device laid over a more fluid Indian situation rather than two self-conscious Indian parties. The Stanford Encyclopedia of Philosophy’s surveys of Madhyamaka and of the two truths in India place Candrakīrti’s reading within the wider field with full bibliographies. Across all of this, one fact recurs and shapes the rest: the man who fixed the most uncompromising reading of emptiness left almost no mark on the India that produced him, and became, four centuries and a thousand miles away, the measure of the view itself.

→ Related: Nagarjuna · Buddhist Madhyamaka · Madhyamaka Sunyata · Tsongkhapa · Tibetan Buddhism · Buddhism · Mahayana · Vasubandhu