Introduction

1

Buddhists, as is the case with adherents of many other religions, establish and stabilize their collective identity among other things through the (re)production of a particular body of literature deemed authoritative. Across the Buddhist world, this body of literature is variously designated as “The Pāli canon,” “the Chinese canon (dazang jing 大藏經),” “the Tibetan canon (i.e., Kanjur and Tanjur),” etc. Carsten Colpe advanced the proposition¹ that the Buddhist ‘Three Baskets’ (Sanskrit tripiṭaka, Pāli tipiṭaka) and the Hebrew Bible (acronym tanakh) represented the two independent forms of canonization in human history which became a model for all other processes of canon formation, bringing forth the Christian Bible, the Daoist canon, the Islamic Qur’an, etc. To what extent this claim does justice to the historical complexities of the ‘Three Baskets’² must remain open to discussion. But it alludes to the significance of the issue of canonicity that persists throughout the history of Buddhism and has great implications for a diversity of Buddhists in different cultural spheres.³ The present essay is a preliminary attempt to tackle the complexities through a case study, i.e., the reception of Mahāyāna sūtras in fourth- and fifth-century Central Asia.

2

Before delving into the case study, a few remarks on the concept of ‘canon’ are in order. Scholars of Religious Studies have long been working with the theoretical twofold typology⁴ of an “open canon,” i.e., a collection of authoritative texts in the general sense which does not exclude other texts from canonicity, and to which other texts of equal importance may be added at any time; and a “closed canon,” i.e., an exclusive collection of authoritative texts, to which only scriptural authority is assigned (and no others!). The borderline between the two kinds of canon is not ironclad and stable, but is porous and dynamic. An open canon can be ‘closed’ in some sense by separating the canonical texts from the apocryphal and by calling a halt to the addition of new texts into the collection. The act of closure, which forms “[t]he most important step toward canonization” (Assmann 2011, 78), does not, as it were, draw a line in the sand because its binding force is not permanent and its consequences are not irreversible. On the other hand, even if no new texts may be added to the body of literature, this does not necessarily imply the closure of the canon on the interpretative level, insofar as innovation de facto continues by dint of interpretative text production (e.g. translations, commentaries, etc.; Silk 2015, 6). The further the process of interpretation advances, the more difficult it becomes to standardize or harmonize the texts thereby produced.

3

Therefore, the utility of canonization as a “contra-present”⁵ bulwark against the tide of innovation should not be taken literally. Viewing the history of Buddhism in the longue durée, we observe that an open canon is the norm, while a closed canon merely occurs at one or two times and places, contingent on specific historical and socio-religious circumstances of a given milieu which make its closure desirable.⁶ On the surface of it, there seems to be an asymmetrical relation between these two kinds of canon: The vast majority of Buddhist canons exhibit a greater or lesser degree of openness, whereas closed canons stricto sensu are few and far between. This asymmetry, however, does not mean that the gravitation towards the closure of a canon is incapable of acting as a counterweight to its opening up. On the contrary, a closed canon remains an attractive option even in a Buddhist milieu whose canon is by and large open, and where attempts are made to seal off the body of authoritative literature in some sense. This is all the more the case when the very milieu is in a transitional phase of its history which entails redistribution of religious authority, as will be explained in detail below.

Khotanese Shift to the Mahāyāna

4

The emergence of a group of authoritative texts designated ex post facto as ‘Mahāyāna sūtras’ around the turn of the Common Era⁷ is a historical phenomenon which still evades any conclusive explanation. Despite their heterogeneity that renders any attempt at monothetic definition futile,⁸ Mahāyāna sūtras, especially those belonging to the earlier strata of this body of literature, are more likely to be subjected to skeptical scrutiny as regards their canonical status compared to the sūtras transmitted by the Mainstream⁹ schools. Disputations about their authenticity were initiated early on by the followers of the Mainstream tradition,¹⁰ to whom the texts were unheard of in the Dharma that had come down to them. For this reason and others, the historical argument serving as the basis for the criterion of authenticity or canonicity was not in favor of Mahāyāna sūtras of later historical provenance, and was thus utterly rejected by Mahāyāna scholastics such as Vasubandhu (fl. fourth century CE) (Cabezón 1992, 228). The early advocates of these sūtras were, in all likelihood, educated monks, or rather communities of such monks, who constituted “a number of differentially marginalized minority groups” (Schopen 2000, 24) struggling for recognition. Their struggle, to our knowledge, did not succeed in Middle-Period India to any significant extent.¹¹

5

The marginalized status of the Mahāyāna in a highly competitive environment might have been one of the motivations for an overland exodus from India.¹² It indeed happened. In the late second century, a number of Mahāyāna sūtras surfaced in Central China in the person of an Indo-Scythian missionary.¹³ This earliest known instance of cross-cultural transfer of Mahāyāna sūtras is probably the result of “long-distance transmission” rather than “contact expansion,” as Erik Zürcher plausibly argued (Zürcher 1990, 158–82). In other words, their mode of diffusion is not reliant on residential monasteries established near prosperous regions or supported by high-level patronage, but consists in incidental and intermittent nodes of communication which are connected through the agency of itinerant monks via transit zones over great distances. In the case of the earliest Mahāyāna sūtras in China, the Tarim Basin seems to have served as such a transit zone, which was not yet capable of affording monastic Buddhism in the late second or early third century.¹⁴ The absence of established monasticism also implies that there was no institutional establishment of any Mainstream school. This vacuum created unprecedentedly favorable circumstances under which Mahāyāna sūtras could take root among recent converts to Buddhism in local society and jockey with their Mainstream counterparts for canonical authority—a privilege they had never enjoyed in India.

6

It is in this historical-geographical setting that Khotan, an oasis kingdom situated on the southern rim of the Tarim Basin (Fig. 1), comes into focus. The Iranian ruling élite of the kingdom was so eagerly in pursuit of Indian identity that the kings adopted an Indic honorary epithet (Sanskrit Vijaya, Khotanese Viśa’),¹⁵ and that the legendary foundation of the kingdom was anchored in the legend of the Mauryan king Aśoka (Yamazaki 1990, 55–80; Mayer 1990, 37–65). Although multifarious ties with India for long-distance trade and cultural exchanges should render the introduction of Buddhism a matter of course, we know next to nothing as to how Buddhism began in Khotan. As a matter of fact, no archeological evidence for the presence of residential monasteries in Khotan before the late third century has so far come to light.¹⁶ To be sure, absence of evidence is not evidence of absence; but one may at least infer from the evidential vacuum that monastic orders affiliated with Mainstream schools, even if founded in Khotan at that point, did not gain any great social prestige or visibility.¹⁷ This inference, on the other hand, implies that Mahāyāna monks were provided with a golden opportunity to make forays into the religiously virgin soil of Khotan, where the authenticity of Mahāyāna sūtras, however, may not have gone uncontested.

7

Zhu Shixing 朱仕行 (203–282),¹⁸ a Chinese monk aspiring to the Mahāyāna, travelled westward in search of Mahāyāna sūtras. Around 260, he procured at Khotan an Indic manuscript of the Pañcaviṃśatisāhasrikā Prajñāpāramitā, one of the most influential Mahāyāna sūtras of all time. As he was about to have the manuscript sent back to China, he met great opposition from local monks who were adherents of Mainstream Buddhism. The dispute over the fate of the manuscript was resolved in a highly dramatic fashion, according to the Mingxiang ji 冥祥記 ‘Signs from the Unseen Realm’ by Wang Yan 王琰 (late fifth century):¹⁹

8

Most of the monks and laymen of the Western Regions practiced the Lesser Vehicle, and when they heard that Shixing sought the Mahāyāna sūtras, they all thought it strange and did not give him the texts. They said, “You do not know the correct Dharma, and these will lead you astray.” Shixing responded, “The sūtras say that after a thousand years the Dharma will spread eastward. If you doubt that this was the Buddha’s saying, then let us test it with the utmost sincerity.” With that he set afire a pile of wood and poured oil over it. When the smoke and flames were at their peak, Shixing picked up the sūtras and, weeping and bowing his head, uttered the vow: “If these sūtras emerged from the golden mouth (i.e., spoken by the Buddha), they should be disseminated and spread across the land of Han (i.e., China). Let all the Buddhas and Bodhisattvas bear witness!” With that he threw them onto the fire, causing it to flare up brightly. When the smoke had cleared, it became evident that the words of the texts were all intact, and the birch-bark leaves were as before. The entire nation reacted with joy and reverence. So he stayed behind to become a worthy recipient of offerings. (Campany 2012, 75–76, with modifications)

9

This episode is intriguing in many respects. It is perhaps for the first time in the history of Buddhism that the ordeal by fire had the final say on the issue of a text’s controversial authenticity.²⁰ More remarkably, the disputation allegedly occurred in a sphere of patrons or faith community, which showed no favor to any specific strand of Buddhism, whether Mahāyāna or Mainstream. This appears to be in tune with the aforementioned inference drawn from archeological data. It remains unclear how much weight should be attached to this episode, the historicity of which was already questioned by Zürcher (Zürcher 2007, 63). The possibility that it was related, in China, with exaggerated unction and thus contains fictitious elements cannot be excluded, insofar as Mahāyāna polemic against Mainstream opponents was a beloved literary trope in early Medieval Chinese Buddhism which often has no basis in historical fact.²¹

10

Be that as it may, there is circumstantial evidence suggesting that the episode is based on some source of greater antiquity.²² Even if we are dealing with an example of the Chinese imaginaire,²³ not everything of the imaginaire is a figment of collective imagination: At least the conviction that Mahāyāna sūtras had not yet caught on and the fact that their authenticity was still subject to doubt in Khotan by the late third century may not be entirely unfounded in reality. This stands, however, in stark contrast to what Faxian 法顯 (d. ca. 420),²⁴ one of the most renowned Chinese pilgrim-monks, claimed to have witnessed in Khotan at the very beginning of the fifth century:²⁵

11

This country (i.e., Khotan) is prosperous and happy; its people are well-to-do […] The [monks] number several tens of thousands, most of them belonging to the Mahāyāna. They all obtain their food from a public fund […] The ruler of the country lodged Faxian and his companions comfortably in a monastery, called Gomatī, which belonged to the Mahāyāna. At the sound of a gong, three thousand priests assemble to eat […] The monks of the Gomatī monastery belong to the Mahāyāna, which is deeply venerated by the king; and they take the first place in the procession of images. (Giles 1923, 4–5, with modifications and omissions)

12

The monastery of Gomatī (aka. Gomatīra),²⁶ consisting exclusively of Mahāyāna monks, represents a new order of Khotanese Buddhism, as distinguished from Indian monasteries affiliated with the Mainstream schools. The foundation of this monastery cannot be historicized with certitude,²⁷ but its continued presence in Khotan as the foremost monastery under royal patronage up to the end of the tenth century is borne out by Khotanese documents from Dunhuang.²⁸ In such a kingdom as depicted in Faxian’s eyewitness account, it would be utterly unacceptable to subject any Mahāyāna sūtra to a fire-ordeal. Furthermore, it is conceivable that knowledge about Mahāyāna sūtras, and a fortiori manuscripts of Mahāyāna sūtras, must have become valued cultural capital such that most Buddhists in that kingdom were eager to avail themselves of. The rigid demand for Mahāyāna sūtras naturally triggered the proliferation of their translations in the local language, as will be shown below. On balance, there seems to have been a historic transition between the late third century and the fifth century, in which the kingdom of Khotan, especially its ruling élite, shifted to the Mahāyāna.

Beginning and End

13

A not altogether speculative theory on the religio-historical landscape conducive to the Khotanese shift to the Mahāyāna is beyond our reach,²⁹ since we are relatively ill-informed about the time period in question, which is, for the most part, shrouded in darkness. In the course of the fourth century, Khotan receded almost entirely from the vision of Chinese historiographers, since little is known apart from sporadic records of tributary envoys dispatched by the Khotanese kings.³⁰ The kingdom, it seems, remained a vassal state pledging its allegiance to various rulers, Chinese and Proto-Tibetan alike, who in turn wielded hegemony over the Hexi corridor and governed the Tarim Basin on a loose reign. The southward relocation of the kingdom of Nu(h)ava³¹ in the late fourth through the fifth centuries, probably due to an advancement of the desert, marked “a dividing point in Central Asian history” (Brough 1965, 611). This resulted in the desertion of the major towns of Caḍ̱ota and Kroraina and the breakdown of the southern Silk Route, which, though not entirely going out of use, never recovered its former vitality and was superseded from the fifth century onward by the northern route (Vaissière 2005, 123). In other words, the kingdom of Khotan lost its eastern boundary as well as a long-standing shortcut to Dunhuang and northern China. What consequences the changing geopolitical circumstances had in the socio-religious domain remains to be plumbed.

14

It may not be simply fortuitous that the kingdom of Khotan began developing a local literacy along with the rise of the Mahāyāna at about the same time as the kingdom of Nu(h)ava was in decline, where a different type of Buddhism had been prevalent. Some documents written in a Gāndhārī-based chancellery language, which fell into disuse in the aftermath of the kingdom’s leaving the Tarim Basin, convey a noteworthy image: Buddhists in third- and fourth-century Caḍ̱ota “worshipped stūpas and bathed Buddha images but recorded few, if any, texts” (Hansen 2004, 306), and local clerics “lived at home with their wives and children, owned property, and donned Buddhist vestments only for occasional ceremonies” (Hansen 2004, 279).³² Although two Nu(h)avan dignitaries had claimed to have “set out in the Mahāyāna” (mahāyānasamprasthita),³³ they hardly seem to have done anything beyond paying plain lip service to their Mahāyāna devotion. As Faxian, the aforementioned pilgrim, sojourned at Kroraina in 399, he saw the end of a debased form of the Śrāvakayāna, which stood in stark contrast to the state of affairs in coeval Khotan:³⁴

15

The king of this country (i.e., Nu[h]ava) has received the Dharma, and there may be some four thousand and more [monks], all belonging to the Lesser Vehicle. The common people of these countries as well as the clergy practice the Dharma of India, but to a greater or lesser degree. (Giles 1923, 2–3, with modifications)

16

There was no Mahāyāna institution whatsoever, at least not that Faxian was aware of.³⁵ The Buddhist cult entrenched in Nu(h)ava was in many respects different from the Mahāyāna in Khotan, but one of the most salient distinctions between the two Buddhist cultures was highlighted by text-centeredness, namely, the significance of texts in the midst of the faith community.³⁶ In Nu(h)ava, religious authority did not hinge on expertise in authoritative texts, but was rooted in the clergy’s ordained roles in rituals and cultic activities: For the failure to attend communal ceremonies or to put on proper vestments on such occasions, fines (in bolts of silk) were stipulated;³⁷ but there was virtually no trace of any normative statement as to the literary learning of a monk-priest, whose life as a householder made it rather difficult, if not impossible, to cultivate textual expertise. In Khotan, however, it was the other way around: According to Faxian quoted above, Mahāyāna monks in this “ideal civitas buddhica” (Deeg 2005, 86) enjoyed high-level patronage since no later than the early fifth century. Hence there are good reasons to believe that they, under the auspices of “a public fund,” had the time and leisure to seriously engage in textual study and scholastic discussions, which were as important as, if not more so than, their ritual obligations.

17

The text-centeredness of the Mahāyāna community in Khotan is otherwise corroborated by chance-finds of manuscripts written in a local variety of the Brāhmī script, probably dating back to the fifth and sixth centuries (i.e., Early Turkestan Brāhmī; after Sander 2005, 137–38). Many of these manuscripts are copies of Mahāyāna sūtras in Sanskrit,³⁸ bearing witness to the wide range of knowledge on Mahāyāna literature which was accessible to educated monks at home in Sanskrit. A heptad of Mahāyāna sūtras appears to have been particularly well-received, as is evinced in fragments of their Khotanese translations which can be ascribed to the same time period on paleographic grounds (Skjærvø 2012, 118–19): the Anantamukhanirhāradhāraṇī (Loukota 2014, 13–27, 57–59), the Bhaiṣajyagurusūtra, the Ratnakūṭa (aka. Kāśyapaparivarta; Skjærvø (2003), 409–420; Maggi (2015), 101–143), the Saṅghāṭasūtra,³⁹ the Śūraṃgamasamādhisūtra (Emmerick 1970), the Suvarṇabhāsottamasūtra,⁴⁰ and the Vimalakīrtinirdeśa (Skjærvø 1986, 229–60). Six out of the seven Khotanese texts were translated from Sanskrit Vorlagen, with the sole exception of the Bhaiṣajyagurusūtra, the sūtra of the Healing Buddha, which is proven to have a certain affinity to the fifth-century Chinese version (T.1331)⁴¹ and thus is probably of non-Indian provenance (Loukota 2019, 67–90). Apparently, the Khotanese reception of Mahāyāna literature, especially at its incipient stage, was by no means a one-way street, and India was not the only source of authority.

18

That being said, the lion’s share of Mahāyāna sūtras circulating in fifth- and sixth-century Khotan was Sanskrit (or Middle Indic) in origin. There was thus a gap between the language of the authoritative texts and that of the faith community, which was eastern Middle Iranian in speech. As long as members of the community were aware of the gap, the regulatory mechanisms controlling the translation of those texts became essential. This raises the question, above all, of whether the Buddha’s Word, in its ideal form, should be translated at all. Oskar von Hinüber (2014, 147–48) has pointed out an intriguing phenomenon that some Mahāyāna sūtras of the utmost importance, e.g. the Saddharmapuṇḍarīkasūtra (viz. the Lotus Sūtra), seem to have never been translated into Khotanese. While this curiosity still remains to be fully accounted for in religio-historical perspective, it points to a defining feature of a conservative stratum of the local Buddhist community, namely, the overarching emphasis on ‘looking after the words’ (Textpflege) as the foremost “custodian of the tradition,”⁴² which takes priority over ‘looking after the meaning’ (Sinnpflege). In a Khotanese context, the priority of the former finds expression in the reluctance, if not deliberate refusal, to translate a text so as to maintain its original form in Sanskrit, a language that was incomprehensible to everybody in Khotan except educated monks.

19

The institution of looking after the words marks the first step towards canonization, according to Aleida and Jan Assmann (1987). Texts, which become fixed in wording, are thereby not only made taboo and ritualized,⁴³ but also displaced and increasingly distanced from everyday life. The words are consecrated at the expense of the meaning. This process is thus to be complemented by expertise in the exegetic and applicative interpretations, lest the decay of the meaning (Sinnverfall) be inevitable. That is the reason for the rise of the expert-interpreter, the specialist in looking after the meaning (Assmann and Assmann 1987, 12–13). The Assmanns’ theory holds mutatis mutandis for the Khotanese institution of looking after the words by way of non-translation: The Buddha’s Word, in its Indic form, was foreign to many Khotanese Buddhists,⁴⁴ who were speaking in tongues in the ritualized recitation. This was deemed a problem as well as an opportunity by a man of letters, whose chef-d’œuvre is hailed as a milestone in the history of Khotanese literature.

A Book to Remember

20

That milestone is the so-called Book of Zambasta. It is a voluminous, metrical compendium on various aspects of Mahāyāna Buddhism, consisting of twenty-four cantos in total. The vast majority of this book has come down to us in an almost complete main manuscript (St. Petersburg, SI P 6) which can be dated to the seventh or eighth century on paleographic grounds. The original title of this book is unfortunately lost to history, while Zambasta is but the name of a magistrate (pharṣavata)⁴⁵ in the kingdom of Khotan, who, together with his son and family, commissioned the book. The floruit of the magistrate named Zambasta is unknown, but the Book of Zambasta must have circulated in the Tarim Basin at least two centuries before the production of the main manuscript, as a fragment (Berlin, T III Š 16) written in Early Turkestan Brāhmī (fifth/sixth century) has been identified as part of this book.⁴⁶ The Book of Zambasta, therefore, was probably in the making during the fourth and fifth centuries, i.e., the aforesaid ‘Dark Ages’ of Khotanese history.

21

As for the man⁴⁷ who brought the Book of Zambasta into being, no biography is forthcoming. The poet was probably well-read in Mahāyāna literature, since some cantos of the book are adapted from Mahāyāna sūtras, such as the Bhadramāyākaravyākaraṇa (canto 2; Régamey 1938, 5–6), the *Maitrībhāvanāprakaraṇa (canto 3; Duan 2007, 39–58), and the *Tathāgatapratibimbapratiṣṭhānuśaṃsā (canto 23),⁴⁸ while sourced quotations from other Mahāyāna sūtras are found throughout the book.⁴⁹ He was also familiar with certain established clusters of Mahāyāna sūtras, such as the Buddhāvataṃsaka and the Mahāsaṃnipāta, which seem to have gained currency in Khotan at that time (Emmerick 1968, 187). Despite his erudition, the poet seems to have had a somewhat different idea of what he was doing. By modern standards, this book was an indigenous Khotanese composition; but nowhere did the man speak of his own contribution as authorial. Instead, he used such verbs as ‘to translate,’ ‘to recite,’ and ‘to extract’ wherever reference is made to the activity he performed.⁵⁰

22

However the verbs are to be construed, on no account would the poet have put in a claim to authorship, which would have been tantamount to taking the credit due to Buddhas. He rather considered himself something of a messenger conveying the Buddha’s Word to his fellow countrymen, whose mindset towards authoritative texts he trenchantly critiqued:

23

The Khotanese do not value the Dharma at all in Khotanese. They understand it badly in the Indian language. In Khotanese it does not seem to them to be the Dharma. For the Chinese the Dharma is in Chinese – in the Kashmirian language [the Dharma] is such [as] the Kashmirian sweetened wine⁵¹ – but they so learn it (i.e., in Chinese) that they also understand the meaning of it. To the Khotanese that seems to be the Dharma whose meaning they do not understand at all. When they hear it together with the meaning, it seems to them thus a different Dharma. (verses 23.4–6; Emmerick 1968, 343, 345, with modifications)

24

This oft-quoted passage is no doubt by far the most celebrated part of the Book of Zambasta. It has long been disputed what language “Kashmirian” was and whether the differentiation between “Indian” and “Kashmirian” was historical.⁵² But the purport of this passage has not been sufficiently explicated in its own right, and becomes clearer only if one takes into account the immediately following verses, which are often omitted from quotations:

25

Even an ordinary being would not utter a speech which has no meaning. How much less would the all-knowing Buddha be likely to utter meaningless words! In words the essential thing is the meaning. The meaning is indeed so much the essential thing that you should look on it in such a way that the Dharma is preached with that meaning. […] The meaning being unperceived, no one would escape from woes in saṃsāra. (verses 23.7–8, 11; Emmerick 1968, 345)

26

Apparently, the main point is that the Buddha did not utter meaningless words, and that a proper understanding of the Dharma’s meaning is prerequisite for its soteriological efficacy. The poet thus addressed not so much an issue of church language⁵³ as of the priority of looking after the words, which is part and parcel of a seminal mindset traceable to a possibly pre-fourth-century Buddhist milieu in Khotan. In that milieu, attempts were made to canonize and perpetuate authoritative texts of Indian origin primarily by precluding Khotanese Buddhists from translating the texts into their own language. By that means, religious authority was monopolized by itinerant monks who brought along Indic texts and by ritual specialists who gained exclusive access to this sacrosanct body of literature.

27

The Buddha never spoke Khotanese, to be sure. But it is one thing to cherish Indian texts as valued sources of Buddhist teachings, and quite another to isolate the Buddha’s Word from the rest of the Khotanese-speaking world, illiterate in the Indian language. This conservative mindset, as is argued above, would naturally result in the decay of the meaning and, what is worse, a lingering loss of vitality in the roles played by those texts in the everyday life of ordinary Buddhists. These repercussions loom especially large in such a milieu as fourth- and fifth-century Khotan, where the rise of residential monasteries prepared the ground for a more durable locus of the interactions between clergy and laity. The poet of the Book of Zambasta thus responded, as it were, to the call of the Zeitgeist with alacrity. By restoring the centrality of the institution of looking after the meaning, he vindicated his decision to preach the Dharma in Khotanese not as expedient means, but as the sole approach that holds out the prospects of reenacting the Dharma’s soteriological efficacy in such a ‘borderland’ as Khotan, which was overshadowed by the perfection of the Indian ideal.⁵⁴ This extraordinary man thus took on the herculean task of making Buddhas speak to his fellow countrymen, and his ambitious undertaking, as is evinced in the long-lasting impacts⁵⁵ of the Book of Zambasta, was crowned with great success.

Tradent: Words and Deeds⁵⁶

28

As is mentioned above, nowhere did the poet himself claim to be the ‘author’ of the Book of Zambasta, some cantos of which he allegedly ‘translated,’ ‘recited,’ or ‘extracted’ from scriptural sources. All the three verbs should be taken cum grano salis. For instance, canto 2, which he claimed to have ‘translated,’ is, by modern standards, no translation at all but a recasting, if not a recomposition from scratch (Régamey 1938, 5–6). Hence each of the three concepts (i.e., ‘translator,’ ‘bard,’ and ‘epitomist’) at best captures one of the multiple and intertwined dimensions of what it meant for him to produce this magnum opus, but none of them do full justice to his self-identity. Admittedly, any attempt at encapsulating the poet’s multifaceted activity in a single term is nothing short of curtailing him on a Procrustean bed. Despite its potential risks, such an attempt can be made on an ad hoc basis, as long as it identifies an apt substitute for ‘author’ such that offers an increased potential for comparative analysis. To my mind, a candidate for the term of that character is ‘tradent.’

29

The term ‘tradent’ has long been used in the study of Jewish rabbinic literature to highlight the ways in which rabbinic sages themselves understand their role in the making of this body of materials. As the de facto creators of rabbinic literature, they deny any creative role (and any innovative intent) in their own efforts, but only take responsibility for “preserving the integrity of the received version as received from an authoritative teacher” (Jaffee 2007, 22). In other words, the tradent, while producing the text, claims not to accomplish any work of originality but merely to pass on ancient teachings. Robert Mayer (2015), to the best of my knowledge, makes the first attempt at adopting this term into the field of Buddhist Studies.⁵⁷ His intention is to shed new light on the idiosyncratic role played by Treasure revealers (gter ston) in the formation of Tibetan Treasure literature (gter ma). The Tibetan tradents share such conservative concerns of rabbinic sages as “they safely co[rral] individualistic flourishes within the safe bounds of the stock repertoire of established and accepted ritual modules” (Mayer 2015, 233). Although the genre of literature discussed by Mayer differs from the Book of Zambasta in significant aspects, they have at least one characteristic in common, namely that their genesis cannot be adequately accounted for through the assumption of an author of originality.

30

Sten Konow was struck by an ostensible lack of originality in the Book of Zambasta, which he attributed to “a learned collector [but] not an original poet” (Konow 1939, 32). A principal factor in this impression is the poet’s reluctance to claim any authorial credit for himself, as is mentioned above. He sought to be seen as a conservative tradent faithful to the tradition, and as such he gave voice to his apprehensions about possible mistakes that he could have committed in performing his duties as a tradent:

31

Since I have translated this teaching, however extremely small [and] poor my knowledge, I seek pardon from all the divine Buddhas, for whatever meaning I have spoiled here. (verse 1.189; Emmerick 1968, 9; modified after Maggi 2009a, 157)

32

Whatever there may be here which the Buddha has not spoken in a sūtra one should not accept. That is all my fault. (verse 8.48; Emmerick 1968, 141)

33

First-person statements of this kind, at first glance, appear to resemble the usual disclaimers in scholarly publications. It is customary for scholars to include, in acknowledgments of their publications, statements to the effect that all remaining mistakes are their own. If the parallelism could be taken for granted, it would follow that the poet of the Book of Zambasta, like every scholar, made every effort to steer clear of mistakes, and that despite his best efforts, he was aware of the existence of possible, undetected mistakes which could be pointed out by a learned reading public.⁵⁸ It remains to be examined whether this is the case.

34

The mistakes, as is quoted above, are basically twofold: subtraction and addition. The former consists of misrepresentations of Buddhist teachings whose meaning is thereby (partially) lost in ‘translation,’ while the latter results in the contamination of scriptural sources with non-scriptural ones. Both concern meaning rather than the words, in accord with the aforementioned emphasis on the primacy of looking after the meaning, which the poet vehemently championed. The statements thus presuppose the semantic integrity of a closed canon of the Mahāyāna, from which nothing should be taken away and to which nothing should be added. This presupposition is reminiscent of the famous canon formula (e.g. in Deuteronomy 13.1: “The entire word that I command you shall you take care to perform; you must neither add to it nor take away from it!”), which is deeply rooted in the Biblical and Greek traditions.⁵⁹ But in the Buddhist world, there is no precedent for the statements in the Book of Zambasta, while a closed and fixed Buddhist canon was not entrenched elsewhere than in Sri Lanka before the fifth century CE (Collins 1990, 89–126).

35

It is not clear whether the Khotanese poet penned the lines by way of off-the-cuff remarks or drew inspiration from a trope that originated in other traditions. Nor is there any definitive evidence for a Khotanese canon of Mahāyāna sūtras, whether closed or not, before the emergence of the Book of Zambasta. The idea of the totality of Mahāyāna sūtras as valued objects of cultic reverence seems to have been gaining ground in *Cugopa(n),⁶⁰ a petty kingdom to the west of Khotan (present-day Karghalik), no later than the second half of the sixth century.⁶¹ It seems conceivable that the aspiration towards the demarcation, if not the closure, of a Mahāyāna canon, something which never occurred in India, had been in gestation for some time at the southwestern corner of the Tarim Basin, as the Khotanese poem saw the light of day. It may thus come as no surprise that the poet in Khotan conceived a similar idea.⁶² The contours of a Mahāyāna canon may be discernible in canto 6 of the Book of Zambasta, which, according to its introit (verse 6.1; Emmerick 1968, 117), contains fifty-nine verses, each from a different sūtra. If so, this canto would be a florilegium of Mahāyāna sūtras, which, as Mauro Maggi argues, constituted a Mahāyāna “canon of fifty-nine texts as recognized in Khotan” at that time (2009b, 347).

36

The claim in the introit is partially borne out by the recent identification of the sources of twenty-odd verses in this canto (R. Chen and Loukota 2018). Although a good half of the canto still remains unsourced, so far nothing speaks against the assumption that the poet did live up to his words by making precisely a verse out of each sūtra. If the fifty-nine Mahāyāna sūtras add up to something of a canon, they provide an advantageous lens through which to appraise the extent to which the poet delivered on his purported commitments as a tradent. Due to the limited space of this essay, we will content ourselves here with looking into a single verse, i.e. verse 6.12, being a quotation from the Tathāgataguhya(ka), a Mahāyāna sūtra which was first translated into Chinese in the late third century. The verse in question appears to be an abridgement of a lesser-known simile, in which Jīvaka, the king of physicians, is mentioned:⁶³

37

With herbs has Jīvaka prepared and adorned a girl, [thereby he] removes the diseases [of the world]. Just so does the Buddha through the body of dharmas [remove] all afflictions (kleśa) without effort. (R. Chen and Loukota 2018, 161)

38

A Sanskrit version of the same simile was quoted in its entirety by the seventh-/eighth-century Buddhist scholastic Śāntideva in the Śikṣāsamuccaya, an anthology containing numerous quotations from a variety of Mahāyāna sūtras. In that context, the simile, taught by Vajrapāṇi to Śāntamati, reads as follows:⁶⁴

39

Just as, Śāntamati, when the king of physicians Jīvaka collected all medicine, he made the form of a girl (composed of) a collection of medicinal herbs, which is agreeable, good-looking, well-made, well-completed, and well-prepared. She was going to and fro, standing, sitting down, and sleeping, without thinking or imagination. Thither came sick dignitaries: kings, vicegerents, guild-leaders, bankers, courtiers, and petty rulers. Jīvaka let them unite with the medicine-girl. Immediately after the union that they consummated, all their diseases were appeased, and they became free from illness, sound, and unimpaired. […] Just so, Śāntamati, is the Bodhisattva (i.e., the Buddha) essentially characterized by the body of dharmas. Whatever sentient beings – women, men, boys, girls – distressed by passion, hatred, and delusion, touch his body, all their afflictions (kleśa) are soothed as soon as they touch it, and they feel (their) body free from distress. (R. Chen and Loukota 2018, 162–63, with modifications)

40

Compared with the Sanskrit version, the Khotanese verse is so laconic that one can hardly make sense of it without looking up the original narrative context. It lays bare the unsettling fact that the tradent did take things away. That is to say, he condensed a meandering narrative into a verse of four lines, and, in doing so, reduced the source information to its skeleton. In consequence, the meaning was often veiled, if not entirely spoiled.

41

On the other hand, things are added to the simile, as is evinced by the phrase ‘without effort’ (anābhoga), which finds no counterpart in any other version of this sūtra. This phrase is probably an innovation by the tradent, who interpreted the Buddha’s salvific use of his body of dharmas as ‘effortless.’ This interpretation is in line with a seminal idea that all activities of the Buddha or a spiritually advanced Bodhisattva are carried out spontaneously, without volitional effort, for any practitioner from the eighth stage of the Bodhisattva path onwards abides in an impassive state devoid of superficial appearances.⁶⁵ Judging from this example, it seems suspect whether the tradent ever made efforts to refrain from ‘subtraction’ and ‘addition,’ as one may suppose; and even if he did, his efforts did not bear fruit to any significant degree. Nolens volens he made tremendous contributions to the diversity of the textual tradition of the Mahāyāna, keeping an eye not only on metrical constraints, but also on the latest scholastic trends. He seems to have had no guilty conscience at all about weaving together ideas of different provenances.

42

These observations invite us to reconsider the aforesaid statements in rhetorical and pragmatic terms. A word-for-word rendition of the original was apparently not what the tradent actually aspired to. He owned up to his “faults” and pleaded with Buddha for leniency; but there is no indication whatsoever that he strove to steer clear of such “faults,” which occur on nearly every page of the Book of Zambasta. Therefore, to read those statements simply as a plea of mea culpa is to miss the point. The tradent was different from the scholar who adds the usual disclaimers to a publication before it goes to the learned reader, insofar as the target audience of the Book of Zambasta consisted of Khotanese believers who understood the Dharma “badly in the Indian language” (verse 23.4; Emmerick 1968, 343). They were not quite capable of reading Indic Buddhist texts, much less comparing the Khotanese poem with its (mostly unspecified) Indian sources. In this regard, the supposed concern about the ambiguity of responsibility for potential mistakes seems to have been at least excessive, and thus is unlikely to have motivated the tradent to add those statements.

43

The quest for the function of those statements entails a better understanding of the tradent’s role in the transmission process. By dint of those statements, the tradent was not primarily aimed at confessing his own “faults,” or admonishing others against such “faults.” His objective was, to my mind, rather to inculcate a sense of reverence and awe for Mahāyāna sūtras in Khotanese believers by underscoring the sacredness and integrity of this body of literature as the Buddha’s Word, which must therefore remain intact. It is beyond the shadow of a doubt that the tradent ran rings around his countrymen in terms of textual expertise. Both the sermon, to which the Book of Zambasta was probably tailored,⁶⁶ and the authority derived from this missionary role were precisely based on the tradent’s power to control the process of conveying the meaning of the Buddha’s Word to the Khotanese. Hence it is also plausible to read those statements as an emphatic asseveration of his mastery over this body of literature rather than a token of his ostensible concern about mistranslation etc.⁶⁷ A special role was accorded to the tradent in his capacity as expert-interpreter, who was thus entitled to change, update, and harmonize the sūtras according to certain criteria. On a par with those sūtras, his exegesis was canonized.

Concluding Remarks

44

History is more complex than what chance finds reveal. Khotanese Buddhists were not the homines unius libri (‘men of one book’), and neither was the Book of Zambasta their bible. It would be a methodological hazard to plumb the ethos of a particular era through a sole book, however informative it could be. Hence one must bear in mind the sampling nature of the present research, which at best represents a limited view of what actually happened in Khotan during the fourth and fifth centuries CE. Incomplete as it may be, the limited view does spotlight a deep-seated transformation of the structure and the source of authority in the local Buddhist milieu, which was caught in a transition from a ritual-oriented priesthood based on long-distance transmission to a text-centered monasticism under the supremacy of the Mahāyāna. If the Book of Zambasta is anything to go by, an essential aspect of this transformation was the canonization of Mahāyāna sūtras with special emphasis on the principle of looking after the meaning rather than the words, despite the high esteem in which the latter had been held theretofore. The closure of a Mahāyāna canon is likely to have taken place at least on the ideological level, setting in motion a paradoxical process: While exclusive sacredness was awarded to the sūtras, the focus was shifted to their interpretation, and authority was removed from the text-bearer and bestowed on the expert-interpreter, i.e., the tradent, whose exegesis was accorded quasi-canonical status and carried weight with Khotanese believers. Authority was thus redistributed.

45

The historical factors that triggered this transformation remain nebulous for the most part. The influence of the Chinese model is not impossible, but its likelihood is not to be overestimated either, inasmuch as Khotanese monks were confronted with different problems from their brethren in China. In addition, it merits special note that the kingdom of Khotan was forced to cut loose from its suzerain in northern China during the period in question, partly due to the aforementioned desertion of Caḍ̱ota and Kroraina. By the mid-fifth century, Chinese military power was no longer in a position to effectively shelter vassal states on the Silk Routes from external assaults⁶⁸ at a moment when the territorial expansion of the Avars (aka. Rouran, Ruanruan)⁶⁹ and the Hephthalites ushered in a reshuffle of regional power.⁷⁰ In the aftermath of the warfare against the Hephthalites (484–534), Hans Bakker observed “the dissolution of the Gupta empire and the rise of autonomous, regional states in northern India” (2017, 24). It is thus not unlikely that the disintegration of the Sino-centric tributary system in the Tarim Basin a few decades earlier would have compelled oasis states overshadowed by the Avars and later also by the Hephthalites, such as Khotan and *Cugopa(n), to seek autonomy while their diplomats were tactfully mediating between the powers to maintain a fragile independence (Rong 2018, 74–75). Against this historical background, it is possible to hypothesize that the ruling élite in Khotan or *Cugopa(n) readily shifted to the Mahāyāna and ardently endorsed the clerical pursuit of canon and authority in order to unite the people of the country, particularly at a time of political upheaval, under a localized identity of Mahāyāna Buddhism, a religion which distinguished themselves from not merely their near neighbors in the Tarim Basin (e.g. Kucha) but also their nomadic rivals. Admittedly, this hypothesis is speculative; but it might not be useless here to present a working hypothesis that will be tested and refined in case further evidence comes to light.