Chapter Two　The Translation Bureau: The Politics of the Translation Movement¶

Borrowing a shell solved identity; next came the question of where scriptures came from. Chinese Buddhism had an institution Indian Buddhism never possessed: the translation bureau. It was a state project, not a study. Dao'an directed translation in Fu Jian's Chang'an; Kumārajīva opened a bureau of a thousand people at Yao Xing's Xiaoyao Garden, with distinct roles for lead translator, amanuensis, doctrinal verifier, and stylistic editor. By Xuanzang's time, full support from the courts of Taizong and Gaozong made the bureau a government office.¹ Volume One, Essay Ten has already recorded the periods of translation history and its great figures. This chapter asks only one question: when translation becomes a state project, what is produced along with power?

A contrary case lets us weigh the importance of the state project: Paramārtha. He had scholarship of the same rank, but no stable bureau. After Hou Jing's rebellion, he spent his life wandering through the Southern Dynasties, translating in fits and starts and repeatedly wishing to return west by sea. His translation of the Mahāyānasaṃgraha had to wait decades before northern scholarly lineages took it up to establish a school. What was translated, how much, and whether a translation counted never depended on the translator's ability alone. First came the question of whether a stable court stood behind him.²

On the royal side, there was more than one motive, and their weight differed in ordered and disordered times. Here we count those of ordered times. At the top stood cultural governance. Official translation bureaus belonged to the same series as writing history and compiling books: cultural projects that rulers in peaceful times regularly undertook. Their product was less a particular sutra than the genre of ‘the state-authorized edition.’ The second level was the wheel-turning monarch template. During the Renshou era, Emperor Wen of Sui issued three edicts distributing relics and building stupas in over a hundred prefectures, a conscious imitation of Aśoka's eighty-four thousand stupas. With it, the Dharma-protecting king model of Essay Three, Chapter Two was formally established in China. Translation was the cultural component of the same model.³ The third level was the most concrete: taking control. Since popular scripture-making could not be stamped out, the state took control of the source. The customs system and scripture catalogs below belong to this level. The audiences must also be distinguished. Aristocratic tastes shaped the book lists of the private translation era in the Eastern Jin and Southern Dynasties: Prajñāpāramitā, Vimalakīrti, and Nirvāṇa suited pure conversation. By the official translation era of the Sui and Tang, such metaphysical conversation had declined. The bureau's real audience was the court itself and the community of monastic doctrinal scholars. Popular preferences bypassed the bureau entirely, relying on popular copying and vegetarian assemblies. Remember, then: the sutras translated by the bureau were not cited in pure conversation or recited in villages. The six hundred fascicles of the Great Prajñāpāramitā and eighty of the Avataṃsaka were throne texts. Merit was counted by the fascicle; bulk meant dignity. The extreme case of political scripture use came under Wu Zhou. The Commentary on the Great Cloud Sutra invoked Maitreya's descent to legitimate a female ruler. In 693, Bodhiruci and others translated the Jewel Rain Sutra, directly inserting a passage in which the divine prince Sun-and-Moon Light appears as a woman in Mahācīna to rule. Royal preference rewrote a sutra, and this time left physical evidence.⁴

The first product is one this volume has long awaited: control over the sources of scripture. Essay Four, Chapter One recorded that India judged the Buddha's word by checking content, and that the check was loose. Anyone could submit texts to the corpus, and no institution had authority to close it. Essay Seven, Chapter Three recorded Nālandā's response: cataloging without closing the canon. China took another route. Dao'an's Comprehensive Catalog of the Scriptures had a separate register of suspect scriptures. Its judgments turned on whether provenance was clear, not whether doctrine agreed: no translator, no provenance for a Sanskrit original, and a text entered the suspect category. Sengyou's Collected Records Concerning the Tripiṭaka continued this, establishing a ‘Newly Collected Miscellaneous Register of Suspect Scriptures and Spurious Compositions.’ This was the first time in Buddhist history that the provenance of texts themselves was traced: origins identified and translators named, like the transmission chains Islam later gave hadith.⁵ India controlled people, not texts (Essay Four, Chapter One); China supplied control over texts. The bureau monopolized legitimate text production, and the catalog governed standing in the canon. With the pair in place, ‘the Buddha's word’ had customs in China for the first time.

But customs had gaps, and the gaps were revealing. Native suspect and apocryphal scriptures continued to appear and to circulate widely: the Sutra on the Profound Kindness of Parents on filial duty, the Ten Kings family on judgment in hell, and texts on prolonging life and averting calamity. All were scriptures written to meet China's own needs.⁶ Put back into Essay Four, this is the Mahāyāna sutra movement in miniature. A catalog can govern standing, not need; when need is pressing enough, people make their own scriptures. The only difference was that China's catalogs kept these texts outside the main canon and judged them suspect or spurious. They could circulate only among the people, not enter doctrinal learning. The same need rewrote the canon under India's system and circulated outside the canon under China's. Customs did not eliminate scripture-making; it determined only where new scriptures circulated.

Customs' results must be read from two sides, because two suspect scriptures did not remain in popular circulation. They passed customs, entered the main canon, and later became among the most important scriptures of Chinese Buddhism. The first was the Brahmā's Net Sutra. No Sanskrit original has survived. Its earliest provenance is an anonymous ‘Postscript to the Bodhisattva Prātimokṣa,’ claiming that Kumārajīva recited it. Sengyou preserved the postscript without believing it and listed this precept text as translator unknown. Fajing's catalog placed it among suspect works. Three years later, the Record of the Three Treasures through Successive Dynasties cited the postscript as evidence and explicitly recorded translation by Kumārajīva at Caotang Monastery in the eighth year of Hongshi, giving it an official provenance at a stroke. Later catalogs copied Fei Changfang's catalog, and the Kaiyuan Record of Buddhist Teachings finally admitted it to the canon. The Sengzhao preface at its head was a still later fabrication. Modern dating places composition in the Southern Dynasties during the middle to late fifth century; the precept rules in its second fascicle draw in part on Dharmakṣema's translation of the Bodhisattva Stages Sutra.⁷ The second was the Śūraṅgama Sutra of the Great Buddha's Crown. The origin it claims is itself a story of passing a border: an Indian king forbade the sutra's export; an Indian monk cut open his arm to conceal it and sailed east. A book that entered the Chinese canon without passing Chinese scrutiny described its own origin as an evasion of India's border ban. The translation record bypassed the official bureaus of both capitals: a private monastery in Guangzhou in the first year of Shenlong (705), an Indian monk reciting, the demoted chief minister Fang Rong taking it down, and the translator returning home immediately afterward to accept punishment. Anyone seeking to trace its provenance would find the people involved long gone. By the time Zhisheng compiled the Kaiyuan Record of Buddhist Teachings, the sutra had circulated for twenty-five years and captivated court and public with its prose. Zhisheng's own two entries contradicted each other: one said Huaidi met an unnamed Indian monk in Guangzhou and translated with him; the other named Pāramiti and Fang Rong. The two records did not match, but it was admitted nonetheless.⁸

Read the two cases together, and the logic of customs' scrutiny becomes clear. First, scrutinizing provenance teaches scripture-makers to fabricate provenance. Once provenance and translator are the criteria, the target of forgery is completely specified. Ordinary suspect sutras lack credentials; these two had more complete credentials than genuine scriptures. Second, both exploited weak points in scrutiny. The Brahmā's Net used a weakness in time: the early days of cataloging. The least rigorous catalog accepted it first; later, the strictest catalogs repeated that record. The Śūraṅgama used a weakness in space: Guangzhou, at the empire's southern edge. Privately translated and already widely circulating, it was a settled fact by the time it came under scrutiny. No successful apocryphal sutra came in through the official bureaus. Third, and weightiest, need was the strongest credential. Both sutras met needs that scriptures translated from India could not. The Brahmā's Net supplied a bodhisattva precept text spoken by the Buddha himself, received by monastic and laypeople alike, and writing loyalty and filial duty into its provisions: filial duty is called a precept, rulers must not be slandered, meat is prohibited, and vegetarianism required. India did not produce such a scripture because it had no such need, while emperors taking bodhisattva precepts (Essay Four, Chapter Two) lacked precisely such a precept text. The Śūraṅgama supplied a comprehensive collection: tathāgatagarbha, dhyāna, divine spells, precept rules, and the fifty demons of the aggregates, with its own sequence from breaking delusion to distinguishing demonic states. Throughout, it was prose written by Tang people for Tang people; its style alone made it easy to accept. Customs could stop foreign forgeries, not scriptures its own people needed. Fourth, the ratchet of Essay Four, Chapter One: once in the Kaiyuan catalog, only entry, no exit. By the time later doubters produced evidence, these had long been scriptures in everyday use in Chinese Buddhism. One had become the bodhisattva precept text of Chinese ordination platforms for a thousand years; the other, the opening divine spell of monastic morning services. The subsequent dispute over authenticity ran from an allegation of forgery by the Tang layman Faxiang (法详) less than a century after the reported translation—the Japanese monk Deqing (德清) heard it in Tang China and carried it to Japan; Japan had already, by imperial order, assembled two schools to examine and certify the sutra as authentic, and Gen'ei recorded the sequence—through Zhu Xi's naming of Fang Rong, to the modern Hundred Falsities of the Śūraṅgama. Evidence piled higher, and the standing of the two sutras did not move at all. That modern debate belongs in Essay Fourteen.

Look again across the most widely circulated suspect works: Awakening of Faith, Perfect Awakening, Śūraṅgama, Brahmā's Net. Their degrees of suspect authenticity differ, but all belong to the tradition of the true and eternal. Scriptures made to meet native needs mostly matched what native intuitions most urgently wanted; Essay Six has explained this level. The conclusion of the customs section therefore needs one more sentence. Customs stopped hundreds upon thousands of crude works and admitted two carefully made ones; those two shaped Chinese Buddhism more than all the suspect scriptures it barred. Customs determined where new scriptures circulated; need determined which apocryphon would enter the main canon.

The translators’ and the court’s lists represent two different sets of preferences. The central sequence sketches the state translation process, from a Sanskrit manuscript through the lead translator, recorder, doctrinal verifier, and stylistic editor to an imperial preface and catalog review. It is not a claim that every translation had an imperial preface or a secure court-sponsored workshop: Chapter 2 sets Paramārtha’s unstable working conditions against that model. The royal list remains a source of demand, not the destination of the popular branch; the English image places that branch in its own box.

The provenance gate is simplified. The chapter gives the Fanwang jing and the Śūraṅgama Sutra as two major exceptions that entered the canon with fabricated or conflicting credentials. “Translator unknown,” “suspect,” and “spurious” remain different catalog judgments in that account. The final line about customs determining where new scriptures circulated must be read with those two successful admissions and the chapter’s account of their subsequent authority. The two lists retain the chapter’s contrast between translators’ scholastic traditions and the court’s preferences for length, state protection, and commissioned content.

The second product: terminology, and the legislative power behind it. Dao'an formulated ‘five losses of the original, three difficulties,’ first raising translation from a craft into an undertaking with a conscious method. Kumārajīva's new translations replaced the old vocabulary of the concept-matching age in batches and fixed the basic translation terms for the next millennium. Xuanzang established ‘five kinds not to translate,’ launched a new-translation movement, and publicly judged the old translations ‘erroneous.’⁹ No retranslation was merely academic. By then, old translations were the schools' basis; the doctrinal scholars of the Northern and Southern Dynasties based all their schools on them. Retranslating a term shook a school's grounds. Whoever set the glossary defined what could be said in doctrine. In standardizing terminology, the bureau was also deciding who qualified to establish a school.

One night in the new-translation movement deserves a separate entry. Translating the Great Prajñāpāramitā, Xuanzang found its six hundred fascicles cumbersome and wanted to follow Kumārajīva's precedent in abridging it. That night he had nightmares: climbing dangerous cliffs, wrestling fierce beasts. Next morning, he announced that he would heed the dream's warning and delete not a word; the following night brought auspicious dreams. The account comes from a disciple and is read as hagiography at half value. But the psychological fact it records is clear: by the seventh century, the corpus itself had become too sacred for a translator to abridge. The ratchet of Essay Four, Chapter One worked in another civilization just as before: in, never out.¹⁰

To see what was translated, consider two final preference lists. The translators' list is orderly: lead translators translate their own scholarly lineages. Kumārajīva translated the whole Nāgārjuna family; Paramārtha the Saṃgraha and Kośa; Xuanzang traveled west for the Yogācārabhūmi and translated the full Yogācāra library on his return; Yijing specifically sought the Sarvāstivāda Vinaya; Amoghavajra translated the Vajraśekhara esoteric family. Essay Five, Chapter Two said that the Indian Buddhism known in the Chinese and Tibetan canons was the list of texts Nālandā sent abroad. The Chinese side reveals finer detail: the Indian Buddhism China saw was a composite of a few lead translators' own lists. Among translators, there was also a dispute over literary form and fidelity to substance. Dao'an formulated the five losses to warn of damage; Kumārajīva lamented that translating sutras was ‘like chewing rice and giving it to others: not merely losing the flavor, but making them retch.’¹¹ The royal list had three items: bulk, for dignity and the measurement of merit; state protection—the Humane Kings and Golden Light were retranslated across dynasties without interruption, with Amoghavajra adding esoteric spells in his retranslation of the Humane Kings; and commissions, as in the Wu Zhou case above. There are two records of the lists colliding, one mild, one fierce. The mild case: Taizong twice urged Xuanzang to return to lay life and assist in government. The court wanted talent, not Dharma; Xuanzang used translation as his shield and negotiated throughout his life.¹² The fierce case was outside the bureau. In the Western Jin, Wang Fu made the Sutra of Laozi's Conversion of the Barbarians, co-opting the Buddha as a product of Laozi's westward conversion mission. Buddhism answered with the apocryphal Sutra on Pure Dharma Practice, appointing Confucius as Bodhisattva Rutong, Yan Hui as Bodhisattva Pure Light, and Laozi as a manifestation of Mahākāśyapa. This was exclusion by absorption replayed on Chinese soil. Both sides used the same kind of pseudepigraphy as in India: Viṣṇu co-opted the Buddha through the doctrine of avatars, Laozi through the conversion-of-the-barbarians sutra, and Buddhists replied with the three sages' conversion mission to the east. This dispute of apocryphal scriptures continued intermittently for nearly a thousand years, ending only with the burning of scriptures under the Yuan's founding emperor.¹³

Hence this chapter's conjecture, explicitly marked as the rules require: every retranslation is a doctrinal renegotiation. Xuanzang's new-translation movement was half scholarship and half replacement of the authority of older translations. Circumstantial evidence lies in the outcome. Xuanzang's translations are generally acknowledged to be more precise, yet the Faxiang Yogācāra school founded on them declined within decades. Tiantai, Huayan, and Chan, built on old translations, lasted for a millennium. China ultimately chose a more usable basis, not more accurate texts. The bureau could standardize terms, not preferences. What kind of Buddhism China wanted had already been settled in the disordered times of the next chapter.

Another question hangs behind the entire translation movement. China was not religiously vacant; the Daoist line was already there. Why did people turn to Buddhism anyway, and so thoroughly? First look at what the rival offered: the fangxiandao tradition's quest for immortality—serving the personal longevity of the First Emperor of Qin and Emperor Wu of Han—then the late Han Way of Great Peace and Way of the Celestial Masters, with talismans, confession of faults, and a system of libationers; then Ge Hong's theory of the golden elixir, ‘My life is in my hands, not Heaven's,’ Shangqing visualization, and Lingbao fasting rites. Theory and practice were both present. Structurally, however, there were three fatal points. First, the promises fell due at different times. Longevity's promise had to be fulfilled in this world, and emperors died one after another, piling up failures. Buddhism's promise would be fulfilled only after death and could not be falsified. The quest for immortality was originally an extravagance of orderly times; the First Emperor and Emperor Wu both sought it in flourishing ages. In disorder, talk of longevity lost its appeal, while rites for the dead were something everyone needed (the next chapter). Second, a gap between systems. Buddhism arrived with a complete system accumulated over a thousand years: canon, doctrine, renunciant institutions, and Vinaya. Daoism then had no monastic system, no scripture catalog, and no complete doctrinal system. It subsequently filled out its system under Buddhism's pressure. Lingbao absorbed wholesale the frameworks of eons, rebirth, and deliverance; Lu Xiujing modeled his Three Caverns bibliography on Buddhist catalogs; Daoist establishments modeled themselves on monasteries, and precept transmission on full ordination. Indian Buddhism borrowed Śaivism's rites (Essay Seven, Chapter Two); Chinese Daoism borrowed Buddhism's institutions. Third, the Yellow Turban shadow: in court memory, organized Daoism meant rebellion. Buddhism entered as an expatriate religion and initially looked harmless. The limits of the shift must also be drawn. Daoism retained imperial bloodline narratives—the Tang Li family's claim of Laozi as ancestor, and the dispute over Daoist precedence—arts of nourishing life, and native identity. What Buddhism fully took over was sophisticated doctrine and the rites for the dead: precisely the top and bottom ends.¹⁴

The four questions can be brief here; the motives and audiences have been counted above. Add only the weightiest native condition. Paper lowered manuscript copying costs one more step. Scripture-copying workshops and copying for merit made the promise of merit for copying, described in Essay Four, Chapter One, flourish in China. Paper, customs, and standardized terminology: with all three in place, Buddhism's basic conditions in China were ready. Scriptures could now be made and translated in quantity; a mass of believers in need would come only with disorder. The next chapter concerns the market of disordered times.

References¶