Chapter 3: Kumārajīva, the Three Treatises, and Sengzhao¶

The era of matching concepts truly ended with the arrival of one man: Kumārajīva (344–413) entered Chang’an in 401. He gave Chinese readers their first direct translations from high-quality Sanskrit manuscripts, accompanied by oral instruction capable of explaining the texts. Barely a year after Kumārajīva’s death, his disciple Sengzhao composed the first Chinese essays in the history of Chinese philosophy to think through emptiness in a genuinely Madhyamaka manner. This chapter examines the two men and the tradition they established together: the Three Treatise school.

I. Kumārajīva: From Kucha to Chang’an¶

Origins and early study

Kumārajīva, whose name was transliterated in full as Jiumoluoshipo and abbreviated in Chinese to Luoshi, was born in Kucha, now Kuqa in Xinjiang, in 344. He was a product of a region where Central Asian Kushan and Indian cultures met. His father Kumārāyana was a monk from India; his mother was a sister of the king of Kucha. His background gave him the conditions to master two languages and cultural traditions, one prerequisite for his later excellence as a translator.

After entering Kucha, Kumārāyana left the monastic order and married the king’s sister. When Kumārajīva was seven, however, his mother decided to ordain and took the young boy on a journey through Indian kingdoms in search of the Dharma. In Jibin, in the region of present-day Kashmir, he studied Sarvāstivāda Abhidharma under Bandhudatta. During a stay in Kashgar, he studied the Prajñāpāramitā scriptures and Nāgārjuna’s Madhyamaka under Sūryasoma, a prince of Yarkand and a Mahāyāna master, and thereby turned from Sarvāstivāda to Mahāyāna. He was a scholar rigorously trained in multiple traditions, with a foundation in Abhidharma as well as an education in Mahāyāna Madhyamaka.²

Captivity and exile

Around 382, Fu Jian of the Former Qin conquered Kucha under the command of Lü Guang. Kumārajīva was taken as war booty and sent east, but before he could reach Chang’an, Fu Jian’s regime swiftly collapsed after the Battle of Fei River in 383. Lü Guang, who later founded the Later Liang, detained Kumārajīva in Liangzhou, present-day Wuwei in Gansu. He spent about seventeen years there in exile, approximately 384–401.

During these years, Kumārajīva continued studying Chinese and gradually mastered a language previously wholly unfamiliar to him, accumulating the linguistic resources for his later work of translation.

Arrival in Chang’an

In 401, Yao Xing, ruler of the Later Qin, sent an army against Liangzhou and escorted Kumārajīva to Chang’an, honoring him as “Teacher of the State.” From 401 until his death in 413, Kumārajīva spent the rest of his life in Chang’an directing a large-scale translation enterprise. Those twelve years formed one of the most influential periods of translation in the history of Chinese Buddhism.

Accounts of his private life: historical sources and the narrative logic of monastic biography

Historical records and monastic biographies preserve several accounts of Kumārajīva’s private life. They are rare cases in Chinese Buddhist history that directly expose the tension between monastic discipline and actual conduct and therefore warrant demythologizing treatment here.

The first incident was Lü Guang’s forced marriage. According to Huijiao’s Biographies of Eminent Monks, while Lü Guang was taking Kumārajīva east, he ordered the “barbarian monk” to marry a daughter of the king of Kucha—other accounts say a cousin—in order to humiliate him. Kumārajīva adamantly refused, and Lü Guang made him drunk and compelled the union. Of the three incidents, this one has the strongest historical credibility. Multiple records confirm Lü Guang’s contemptuous treatment of Kumārajīva, and humiliating a captive through forced marriage fits contemporary practice.¹

The second was Yao Xing’s gift of palace women. After Kumārajīva reached Chang’an, Yao Xing treated him with the honors due a Teacher of the State, while at the same time giving him several palace women as consorts for a rather unusual reason. According to the Biographies of Eminent Monks, Yao Xing said: “The master’s intelligence and penetrating understanding are without equal under heaven. If he should die without descendants, how could the seed of the Dharma be allowed to end completely?” The logic is typical of secular power intervening in religious affairs: in the name of “transmitting intellectual genes,” a worldly patron imposed his protection upon a religious figure. Kumārajīva reportedly accepted and thereafter lived outside the monastic compound governed by the Vinaya. The source is again the Biographies of Eminent Monks, and the account has relatively high credibility.

The third incident was swallowing needles before an audience. Of the three, it is the most miraculous and most characteristically illustrates the narrative conventions of monastic biography. According to the account, Kumārajīva’s disciples considered imitating their teacher after he broke the precepts. He brought out a bowl full of iron needles, swallowed them one by one in public without changing expression, and then said: “If you can do as I have done, you may imitate this as well. Otherwise, apply yourselves to keeping the precepts and do not imitate me.” Astonished, the disciples never again entertained the idea.

The story is physiologically impossible: a person cannot swallow a large quantity of iron needles unharmed. It belongs to the literary type of a saint’s miracle. Its function is not to record an event but to solve a narrative problem: how can an eminent monk visibly conflict with the precepts yet retain religious authority? The answer is to reestablish that authority through an even more awe-inspiring supernatural power. What ordinary people cannot do further proves that the saint has transcended their level, so that the ordinary framework of discipline applies to him differently. This is a standard method by which hagiography handles a “stain upon the saint,” and parallel structures occur across religious traditions.

Tibetan tantra supplies a structurally symmetrical example pointing in the reverse direction. Instead of reconstructing authority afterward through a miracle, it makes miraculous demonstration a prior threshold: a practitioner must prove the corresponding supernatural attainments before qualifying to practice sexual yoga (karmamudrā). Across Tibetan traditions, the accepted qualification is profound realization of the completion stage. Intuitive descriptions of being “qualified” often mention the ability to reattach a fallen fruit to its branch, fly through the air, or rejoin a severed head. The Fourteenth Dalai Lama stated explicitly that in the Tibetan tradition a practitioner must be able to display such miraculous powers as “flying through the air” before qualifying. The Gelug nun Thubten Chodron has cited her teacher’s example of “making an apple fall from a branch and then reattaching it.” The narrative logic is identical to the needle-swallowing story: proof of supernormal capacity establishes that the person has passed beyond the scope of the ordinary disciplinary framework. The Biographies of Eminent Monks recognizes this retrospectively; tantra posits it in advance.⁴

A prediction at the threshold of death. The Biographies of Eminent Monks also reports that shortly before his death, Kumārajīva said: “If the scriptures and teachings I have transmitted contain no errors, my tongue shall not decay after my body is burned.” After his cremation, the story says, everything became ashes except his tongue. This too is a standard hagiographic conclusion: an extraordinary bodily relic gives final sacred endorsement to his life’s work of translation.

This book’s treatment is as follows. The first two incidents—forced marriage by Lü Guang and the gift of women by Yao Xing—have relatively high credibility as historical facts and reflect Kumārajīva’s actual situation within political power relations. He remained a “guest official” and strategic resource in the hands of rulers and never occupied a position of complete autonomy. The swallowing of needles and incorrupt tongue are necessary components of saintly narrative and are not accepted as historical fact. Separating these levels helps us understand how the Biographies of Eminent Monks works as a literary-religious genre: while recording people, it simultaneously preserves historical facts and protects religious authority, two purposes that sometimes generate tension within the narrative.

II. The Chang’an Translation Bureau: Scale and Method¶

Scale

Kumārajīva’s translation bureau in Chang’an was one of the most important Buddhist translation institutions in Chinese history. Yao Xing supplied state support. Work took place at Ximing Pavilion and Caotang Monastery in Chang’an, with hundreds of monks participating. Translation by individuals and small groups in Dao’an’s age became a major scholarly project supported by the state.

A change in translation method

Kumārajīva’s method was more systematic than earlier collaborative translation. In fluent Chinese, he orally explained the meaning of the Sanskrit source while comparing the manuscript before him with older translations by An Shigao and others and identifying their errors and shortcomings. A group of Chinese monks deeply learned in the Dharma—including Sengzhao, Daosheng, Huiguan, and others—polished his explanation into a written text, after which Kumārajīva personally checked it sentence by sentence against the source.

This process represented a fundamental advance. Kumārajīva knew both Sanskrit and Chinese and could establish a direct connection between them. Translation no longer depended upon a foreign monk reciting in broken Chinese followed by indiscriminate rearrangement from people who knew only Chinese.

Style and limitations

Kumārajīva’s translations are celebrated for their literary beauty. He once compared the difficulty of translation to chewing food and feeding it to another person: not only does the food change shape, its flavor also dissipates. Translation inevitably loses something of the original. Yet to Chinese readers his own translations possess great aesthetic beauty. His versions of the Lotus Sūtra, Vimalakīrti Sūtra, Diamond Sūtra, and Amitābha Sūtra remain the versions most often recited in Chinese Buddhism.

At times, however, beauty came at the expense of precision. His version of the Mūlamadhyamakakārikā followed Piṅgala’s commentary, which had itself already attenuated Nāgārjuna’s meaning. Kumārajīva’s translation was therefore the product of a source first passing through commentary and then through translation, already two steps removed from the Sanskrit original.

III. Principal Translations¶

Kumārajīva translated a vast number of works. Those most influential in Chinese Buddhism fall mainly into the following categories.

Madhyamaka treatises, the foundation of the Three Treatise school

Mahāyāna scriptures

IV. Foundations of the Three Treatise School¶

The Three Treatise school took as its foundation three treatises translated by Kumārajīva: the Treatise on the Middle, Treatise in Twelve Gates, and Hundred Treatise. It was the Chinese form in which Nāgārjuna’s Madhyamaka philosophy was transmitted. Part V has already treated the philosophical content of these works in Chapters 5.1 and 5.2; here the perspective is the history of transmission in China.

The Three Treatise school was not yet established as a school during Kumārajīva’s lifetime. It emerged through several generations after his death and crystallized formally only in the sixth and seventh centuries. Kumārajīva’s disciples—Daosheng, Sengzhao, Huiguan, Daorong, and others collectively known as the “Ten Philosophers of Kumārajīva’s Gate”—formed the first generation. Several masters of the Southern Dynasties followed, and during the Sui and Tang periods Jizang (549–623) completed the integration of the Three Treatise school as a systematic philosophical tradition.

V. Sengzhao: The First Chinese Thinker in a Madhyamaka Mode¶

Life

Sengzhao (374–414), whose lay surname was Zhang, came from the Guanzhong region in what is now Shaanxi. Born into poverty, he initially made his living copying documents and thus encountered a large body of writings by Laozi and Zhuangzi, which he greatly admired. Legend says that after reading the Laozi, he exclaimed that it expressed the ultimate teaching, yet felt that it “could be grasped spiritually but was difficult to seek through ordinary feeling.” Only when he encountered an older translation of the Vimalakīrti Sūtra did he suddenly understand, resolving to become a monk. After Kumārajīva arrived in Guanzhong, Sengzhao became his attendant disciple and participated in the principal translations. He was one of Kumārajīva’s most trusted Chinese students.

Sengzhao died in 413 or 414 at only about thirty-nine years of age. He was one of the most important thinkers to die young in the history of Chinese Buddhism. His principal works are four essays later compiled as the Treatises of Zhao.

VI. The Treatises of Zhao: The Philosophy of Four Essays¶

The collection generally contains four essays: “Things Do Not Shift,” “Emptiness of the Unreal,” “Prajñā Is Not Knowing,” and “Nibbāna Is Nameless.” They were the first Chinese philosophical writings in the history of Chinese Buddhism to think emptiness in a Madhyamaka rather than matched-concept mode. Their literary quality is extremely high, as is their philosophical density.³

“Things Do Not Shift”

The title literally argues that “things,” the myriad phenomena, “do not shift,” move, or change.

The claim appears to contradict intuition: everything plainly changes, so why say that things “do not shift”? Sengzhao’s argument is approximately this. What we ordinarily call “change” presupposes that a thing continuously exists and passes from one state into another. Yet if every phenomenon arises and ceases from moment to moment and lacks intrinsic nature, no continuously existing “thing” moves from one place to another. The subject presupposed by “movement” does not exist. Thus, in the strict Madhyamaka sense, “things do not shift” does not mean that things are static and unchanging; it means that the “continuous subject” presupposed by the concept of change does not exist.

“Emptiness of the Unreal”

The title literally means that because things are “unreal”—not genuinely or substantially existent—they are “empty.”

Sengzhao’s point is that things are not “real,” possessing no intrinsic substantial existence, and are therefore empty. He calls their existence a “provisional designation”: things exist as temporary designations in language, not as substantial realities. At the same time, he criticizes two erroneous extremes. One treats emptiness as sheer absence, the annihilationist view that emptiness means nothing exists. The other treats conventional existence as “real being,” the eternalist view that things possess substantial intrinsic natures. The correct understanding is that things are “neither existent nor nonexistent.” Conventionally they do exist, because convention establishes them; ultimately they are empty, because they lack intrinsic nature. This is a Chinese expression of Madhyamaka’s two truths.

“Prajñā Is Not Knowing”

The title literally argues that prajñā, wisdom, “does not know” in the ordinary sense.

Its central claim is that genuine prajñā does not follow the model of a “knower” knowing an object. That is cognition through discrimination, a cognitive mode opposing subject and object. Genuine prajñā is nondiscriminative and does not objectify. It is identical with emptiness itself and does not posit a separate “knower” outside emptiness to know emptiness. Through the apparently paradoxical phrase “not knowing,” Sengzhao expressed this nondual condition of cognition. The position is very close to Madhyamaka’s “nonconceptual wisdom” (nirvikalpa-jñāna), discussed in Chapter 5.5.

“Nibbāna Is Nameless”

The title literally argues that Nibbāna cannot be named.

Its central claim is that Nibbāna is not a state or “place” that can be named or described. Once it can be named, some substantial object is implied as the thing being described, and conceptual proliferation has already begun. Nibbāna is where the path of speech ends and mental activity ceases. Any name used to refer to it inappropriately reifies it.

Together, these four essays constituted China’s first systematic Chinese-language exposition of Madhyamaka’s central themes: absence of intrinsic nature, establishment through conventional designation, nondiscriminative wisdom, and Nibbāna as nonsubstantial.

VII. Sengzhao’s Distance from Matching Concepts and His Limitations¶

Transcending matching concepts

Compared with the “six schools and seven traditions” of the matched-concept age, Sengzhao’s advance was real. He abandoned “original nonbeing” in its cosmological sense as an explanation of emptiness and directly employed “unreality,” the absence of substantiality. He did not interpret prajñā as Daoist nonbeing, but as non-objectifying, nondiscriminative wisdom. Nor did he substitute the “mind’s nonattachment” for an ontological claim about emptiness, but directly addressed the ontological proposition that “the mode of existence of things is empty.” He was the first author in Chinese Buddhist history to think about emptiness at the proper epistemological and ontological levels.

Remaining Chinese coloring

Yet Part V, Chapter 5.2, has already noted a subtle shift. The argument in “Things Do Not Shift” differs structurally from Nāgārjuna’s proof of emptiness. Nāgārjuna’s reductio derives a contradiction from “if something had intrinsic nature, it could not depend on conditions,” thereby eliminating intrinsic nature. Sengzhao’s “non-shifting” resembles a metaphysical reflection on “the temporality of things.” It reaches the conclusion that “no enduring subject exists,” but follows a different route from Nāgārjuna’s inferential chain from dependent origination to emptiness.

One further point is even more important: Sengzhao’s language retains a powerful Zhuangzian atmosphere. Parts of “Things Do Not Shift” read remarkably like the Zhuangzi and are filled with paradoxical rhetoric. In one sense, this was deliberate. Sengzhao used modes of thought familiar to Chinese readers to guide them toward a Madhyamaka understanding. The strategy also meant, however, that Chinese readers still encountered Madhyamaka through a “Zhuangzi filter.”

This book’s judgment is that Sengzhao was the historical turning point between matched-concept interpretation and genuine understanding of Madhyamaka. He accomplished what no other Chinese thinker of his time could. His understanding nonetheless bore the imprint of the Chinese intellectual setting, and the complete escape from the matched-concept framework was not a task he could finish. It required more texts, more debate, and more time.

VIII. Jizang: The Great Synthesizer of the Three Treatise School¶

The Three Treatise tradition continued developing after Sengzhao. During the Sui and Tang periods, Jizang (549–623) completed its most systematic synthesis. Jizang descended from a Sinicized family originally from Parthia, modern Iran, and was born in Jinling, present-day Nanjing. As a child he ordained under the Three Treatise monk Falang of Xinghuang Monastery (507–581), and he later traveled and taught widely. He became known as the Master of Jiaxiang.

Jizang’s principal contributions

Jizang wrote systematic commentaries on the three treatises and numerous works elaborating the school’s positions. The most important included the following:

The Profound Meaning of the Three Treatises. This was a general statement of the Three Treatise position. It systematically explained the school’s fundamental thought, its differences from other currents of Buddhist thought, and the Madhyamaka methodology of “refuting without establishing.”

The Meaning of the Two Truths. This was a detailed discussion of the two truths. Jizang’s theory has a distinctive feature: he understands the two truths as a “pedagogical instrument,” a gate of teaching, rather than two aspects of the world. The two are established to counteract the different attachments of sentient beings; they are not two independently existing levels of reality. This treatment comes relatively close to the spirit of Prāsaṅgika Madhyamaka’s “establishment through conventional designation,” though it retains its own Chinese interpretive coloring.

Jizang also resolutely defended the position of “establishing no dharma.” He criticized contemporary systems of doctrinal classification, such as Tiantai’s five periods and eight teachings, holding that even the proposition “this teaching is higher than that teaching” constitutes a new attachment. This carried the Three Treatise spirit of “refuting without establishing” through to its conclusion.

The historical fate of the Three Treatise school

After Jizang, the Three Treatise school did not endure as an independent institution. It did not establish an organized line of transmission like Chan or acquire a broad popular base like Pure Land. Its philosophy continued in dispersed form as it was absorbed within Tiantai, Huayan, Chan, and other schools. The fate of the Three Treatise school replicated the fate of Madhyamaka philosophy in China: it did not become an independent mainstream, but became instead a philosophical ground tone within other schools.

IX. Historical Influence¶

The longevity of the translations

Kumārajīva’s translations remain in use today. His versions of the Lotus Sūtra, Diamond Sūtra, Vimalakīrti Sūtra, and Amitābha Sūtra are still the texts most often recited in everyday Chinese Buddhist life. Their longevity exceeds that of the work of any later translator. Xuanzang’s versions are more precise, but never approached Kumārajīva’s literary popularity.

Influence on later schools

The intellectual framework established by Kumārajīva and Sengzhao directly influenced nearly every later Chinese Buddhist tradition.

Tiantai, discussed in Chapter 10.5, took Kumārajīva’s translation of the Lotus Sūtra as its fundamental scripture. Zhiyi’s Tiantai teaching and contemplation rested upon this text.

Chan, discussed in Chapter 10.8, drew most frequently on Kumārajīva’s translations of the Vimalakīrti Sūtra and Diamond Sūtra. It regarded Vimalakīrti’s silence as the highest mode of expression, while the Diamond Sūtra occasioned the sudden awakening of the Sixth Patriarch Huineng.

Pure Land, discussed in Chapter 10.9, counted Kumārajīva’s Amitābha Sūtra among its three principal scriptures; it became the scriptural basis most commonly recited in Chinese Pure Land devotion.

Influence on translation methodology

Kumārajīva inaugurated a tradition that prioritized free translation and literary beauty. Together with Xuanzang’s later tradition of precise, literal translation, it formed the two poles of Chinese Buddhist translation history and established the basic framework for later Chinese debates over translating religious texts.

X. Translational Choices and Misunderstanding: A Structural Tension¶

That Kumārajīva’s translations are still recited today sufficiently demonstrates his achievement. Yet the achievement had a cost that should be identified in conclusion.

The lexical gravity of “emptiness”

Using the Chinese character kong, “empty,” to translate śūnyatā was the best available choice in the Chinese lexicon of the period. Yet kong already carried its own weight: the sky, empty space, and Daoist “nonbeing.” These associations pushed śūnyatā in a dangerous direction.

Nāgārjuna’s śūnyatā is a relational predicate: to say that X is empty means that X does not exist through intrinsic nature, not that X does not exist. It describes an absence within the mode of existence, not a state of nothingness. Yet the Chinese context of kong naturally points toward the latter: fundamentally, the myriad things are nothingness and come from nonbeing. This was the route taken by the matched-concept School of Original Nonbeing, which understood Buddhist emptiness through the Daoist cosmological priority of nonbeing over being. The misreading did not result from deficient comprehension. The lexical gravity of kong pulled the reader toward it.

Kumārajīva corrected the crudest form of the drift in his translations and oral teaching, but the associations of kong itself were not dissolved, only suppressed. They could reappear at any time in readers insufficiently familiar with the argumentative structure of Madhyamaka.

The dual-track predicament of “wisdom” and prajñā

The treatment of prajñā was more complex than that of “emptiness.” Chinese Buddhism ultimately adopted a dual-track solution: retaining the phonetic transcription bore as a technical term while using “wisdom” as its everyday explanation. The two terms coexist, explain one another, and interfere with one another.

In Chinese culture, “wisdom” means a cognitive and practical capacity accumulated through time, such as the sage’s discernment in Confucianism or the sage’s stillness in Daoism. It can be acquired, transmitted, and built gradually. The core of prajñā, by contrast, is direct and nonconceptual. It is not knowledge about emptiness but intuition within emptiness—Sengzhao’s “not knowing” in “Prajñā Is Not Knowing,” where the objectifying structure of the act of knowing dissolves. Once prajñā is explained as “wisdom,” it readily becomes “learning about emptiness” rather than “the dissolution of the subject-object distinction.”

The philosophical structure of X / not X / therefore called X in the Diamond Sūtra

The Diamond Sūtra repeatedly uses formulations such as “is not adornment; therefore it is called adornment.” This is not meaningless paradox but a compressed expression of the two truths: “is not X” indicates the absence of intrinsic nature at the ultimate level, while “therefore it is called X” reaffirms conventional establishment. The structure assumes that readers already know emptiness does not mean “nothing exists.” When that assumption holds, it is a precise philosophical formulation. When the cosmological associations of kong erode the assumption, it becomes a riddle, as though it claimed that “logic itself must be transcended,” rather than a concrete application of the two truths.

An internal cause: the contradiction of self-reference and transmission

This problem is structurally isomorphic with the problem of Mahāyāna’s rise, though it operates on a different level. In the latter, new content was grafted onto an old structure of authority through attribution to the “word of the Buddha,” and the mechanism of transmission itself changed the content. Here, a new philosophical framework was grafted onto an old network of associations by reusing existing vocabulary. Again, the mechanism of transmission itself produced the drift; it was not deliberate deception.

Yet a cause deeper than this analogy makes the problem impossible to solve completely. If the philosophical claim of śūnyatā is fully understood, it must be applied back to the concept of śūnyatā itself, the argument of “the emptiness of emptiness.” Any word used to transmit it, once fixed as a stable technical term, must itself be emptied by emptiness. This is not merely the failure of a particular translational choice but a structural feature of the philosophical claim itself: it is a rule that returns upon itself. The difficulty of translation is an amplified manifestation of this self-reference in transmission across languages. The native associations of kong add friction, but even without them the difficulty of transmission would remain, appearing in another form.

This is why the Three Treatise, Tiantai, and Chan schools each had to clarify kong anew. It was not that earlier translators had erred, but that translation is work that must repeatedly be done. The commentaries and debates of every generation continuously manage this tension; they do not resolve it once and for all.

References¶