The sBa bzhed

As far as possible, this thesis limits itself to eighth to twelfth-century historical sources. The first few centuries of Tibet's assimilation of Buddhism witnessed the creation of several important historiographical traditions that are still not adequately understood. As recently as 1996, van der Kuijp produced the following sober assessment:

[T]he earliest Tibetan historiographical materials are extremely diverse and, regrettably, to a large extent still unpublished. Investigations into the literary sources used by authors of those texts that are available to us are also in their infancy, as is, consequently, research into the particular ways in which they have made use of them. This renders it particularly difficult to determine the original contributions made by these early authors in terms of how they interpreted them when they were not simply incorporating large portions of their sources into their own work. Although scholars of Tibetan historiography have made good progress since 1996, many works that Tibetan tradition and previous scholars ascribed to the eighth to twelfth centuries are in fact later redactions of the originals, or simply misattributions. Because such mistaken ascription causes great confusion, I must begin by ascertaining what texts can be included in my analysis.

Imperial and post-imperial sources fall within my period. I discuss the precise dates of these texts when I analyse their descriptions of Khri Srong lde brtsan in future chapters. If they are not chiselled into stone, then they are found for the most part in the so-called “library cave” near Dunhuang. These manuscripts date from different periods between the Tibetan occupation of Dunhuang in the late eighth century and the closing of the cave at the beginning of the eleventh century. Foremost among this group are the year-by-year royal records that constitute the Old Tibetan Annals, and the poetic account of imperial Tibet known as the Old Tibetan Chronicle. With the exception of the Old Tibetan Annals, I treat Dunhuang documents as post-imperial depictions of Khri Srong lde brtsan.

In exceptional circumstances, I accept quotations of imperial documents in much later sources, including the bSam yas edicts in dPa' bo gtsug lag phreng ba's sixteenth-century mKhas pa'i dga' ston (KGT). I treat documents from the Tibetan canon that claim to be the words of Khri Srong lde brtsan or of his contemporaries with great caution. Even if they were genuine once, the scope for later redaction is huge. As a result I have noted these texts in the appropriate places below, but never allow them to dictate my understanding of the imperial representation of Khri Srong lde brtsan.

Eleventh and twelfth-century histories are far more problematic. The extant versions of these texts are all later copies of the originals. As a rule, they show evidence of subsequent alteration, either as a result of their transmission or through conscious redaction by later generations of scholars. This chapter outlines the process of recension of an important history on Khri Srong lde brtsan, the sBa bzhed. Chapters Two and Three then investigate the antiquity of those histories attributed to Nyang ral Nyi ma 'od zer (1124-1196). It is important to date the notes, additions or entire redactions of these histories so as not to confuse a later depiction of Khri Srong lde brtsan with an earlier one.

The dBa' bzhed

The dBa' bzhed is a composite work of generations of redactors beginning in the ninth century. The manuscript of the dBa' bzhed dates to perhaps the fourteenth or fifteenth century, but the main narrative stems from the eleventh or early twelfth century. This dBa' bzhed itself represents a later redaction of the earliest entire version of the d/s/rBa(') bzhed history of Khri Srong lde brtsan's reign. Some ‘proto-version' of this history was perhaps originally composed during the late imperial period, and was then expanded. As a result, the dBa' bzhed manuscript might contain within it narrative strata from six centuries of changing Tibetan historiography. The oldest narrative strata contained in the dBa' bzhed describe Khri Srong lde brtsan in ways similar to Dunhuang texts. The interpolations and interlinear notes depict him from several, slightly different, perspectives—each reflecting the time of their addition.

The Narrative

The principal narrative of the dBa' bzhed reflects an eleventh or twelfth-century view of Khri Srong lde brtsan. Since its depiction falls within my timeframe, I shall therefore provide a brief precis of its content. The dBa' bzhed's full title is: ‘dBa' bzhed, the royal narrative (bka' mchid) concerning how the Buddha's Dharma arose in Tibet.' This description encapsulates the frame narrative of the text, and perhaps indicates the antiquity of its core depiction of Khri Srong lde brtsan. The dBa' bzhed first recounts the reigns of four major Tibetan Buddhist btsan pos, dividing the narrative into four parts with Khri Srong lde brtsan taking the lion's share.

When the thirteen-year-old Khri Srong lde brtsan takes over the governance of the empire, the narrative shifts from the btsan po to a small group of Tibetan ministers and their conspiracies against the Dharma. The dBa' bzhed's principal protagonist is the Buddhist minister dBa' gSal snang. Despite the rival ministers' destruction of all that previous Buddhist kings had achieved and their interdiction against its future practice, dBa' gSal snang goes in search of the Dharma to India and Nepal where he worships at Buddhist pilgrimage and monastic sites.

dBa’ gSal gnang convinces the btsan po to invite the Indian abbot Santaraksita to Tibet. Santaraksita in turn recommends the tantric master Padmasambhava to tame the land in order to build bSam yas Monastery. However, Khri Srong lde brtsan grows suspicious of the siddha's power and asks Padmasambhava to leave Tibet half way through the narrative. The btsan po instead appoints dBa' gSal snang to ‘the highest religious authority (chos kyi bla) as head [at his] right side (sa g.yas kyi tshugs dpon).' For a while thereafter, though, Santaraksita continues to play a more prominent role than dBa' gSal snang, for instance in debate with the Bon pos or digging out the site of bSam yas with Khri Srong lde brtsan. Santaraksita's religious preeminence reflects his role in earlier documents as the first abbot of bSam yas.

When the abbot dies, dBa' gSal snang is ordained as Ye shes dbang po and becomes the main moral goad of Khri Srong lde brtsan. Under his guidance, the btsan po hosts and decides the outcome of a debate between the “gradual” and “sudden” paths of Buddhism. His sanction of the gradual path enables the establishment of monastic colleges and a major translation effort. For the dBa' bzhed, his reign marked an unprecedented high-point in the rise of Buddhism in Tibet: Where the Dharma did not get established during the reign of the five previous kings, Lha sras Khri Srong lde btsan, Acarya Bodhisattva, dBa' Ye shes dbang po and ’Ba’ Sang shi, these four, established the shrines of the Three Jewels. The dBa’ bzhed’s Process of Recension Some parts of the dBa' bzhed narrative date back to the ninth or tenth century. Recently van Schaik and Iwao discovered two Tibetan fragments, among the Chinese-language Dunhuang collection (British Library), that may once have formed part of a tenth-century version of the s/dBa(') bzhed. They contain the story of the arrival of Santaraksita at the court of Khri Srong lde brtsan. The king does not trust the abbot, and therefore dispatches a minister, rGyal sgra legs [gzigs], to find another Indian translator who can verify Santaraksita's exposition of the Dharma. rGyal sgra legs gzigs finds a Kashmiri brahmin named Ananda, versed in Indian lore, who testifies to Santaraksita's orthodoxy. Van Schaik and Iwao translate the fragmentary text thus:

He was uncertain whether [.] was placed in the Ra sa Bang khang. To the minister rGyal sgra legs [.] he ordered: ‘Investigate whether there are any foreigner evil spirits or black magic.' [.] there was no interpreter [.] interpreters of Kashmir and Yang li at the six market-places [.] Three interpreters were found: the two [.]e sbyin brothers and A-nan-ta from Kashmir [.] the son of the brahmin sKyes bzang who was a serious convicted criminal in Kashmir [.] sacred scriptures and the tradition of ritual exposition [.]

This vignette is also included in the dBa' bzhed, part of which reads:

The [the btsan po] ordered Zhang blon chen po sBrang rGya sbra (sgra) legs gzigs,

Seng 'go lHa lung gzigs and 'Ba' Sang shi, ‘You three ministers, go to Ra sa Vihara

(pe har) to meet Acarya Bodhisattva and prostrate in front of him. Then investigate whether I need to suspect the presence of black magic and evil spirits from lHo bal or not.' The three arrived at Ra sa Vihara. There was no translator.

Van Schaik and Iwao critically compare the versions and come to three main conclusions:

1. ‘The extent of the correspondence clearly indicates that these Dunhuang fragments must be equivalent to one of the original sources of the Dba' bzhed (or at least this part of the Dba' bzhed).'

2. ‘.and also confirms the status of the Dba' bzhed as the earliest extant Tibetan Buddhist history.'

3. There exist variations between the two texts, which suggest that the dBa' bzhed has added extra detail to, and even changed, the narrative it received:

a. Santaraksita is no longer ‘placed’ (bzhags) in the Jo khang, but rather ‘requested to stay' (bzhugs su gsol). This may denote his increased status in Tibetan historical literature.

b. The name of the Jo khang is changed from Ra sa Beng khang to Pe har (vihara).

c. A description of Khri Srong lde brtsan's motivation for caution is added before his speech.

d. Two new ministers may have been added to the text to give them some role in this narrative (and perhaps thereby increase their status in history).

e. More orthodox brahminical learning is ascribed to Ananda, which would make him more fit to judge the veracity of Santaraksita.

It is difficult to establish whether these fragments, together, constitute one side of a folio intended for an earlier version of the dBa' bzhed, or part of a story that the dBa' bzhed incorporated in its narrative. The overwhelming similarity between this version and the dBa' bzhed episode demonstrates that the latter contains narratives that stem from at least the ninth or tenth century. The subtle differences show that the narratives have been redacted since those times. The evidence of the Dunhuang fragments therefore suggests that the main part of the dBa' bzhed contains an eleventh¬century narrative.

Other parts of the extant dBa' bzhed narrative are clearly younger than the eleventh century. These probably include the major additions at the beginning and end of the dBa' bzhed—on the relations of Khri Srong lde brtsan's ancestors to Buddhism in Tibet (1a1-4a5) and the debate between Buddhism and Bon over who should perform the bTsan po's funeral rites (25b1-31b6) respectively. No sBa bzhed contains either the dBa' bzhed's opening folios or its last section. The sBa bzhed redactors had no reason to omit the beginning and end of the dBa' bzhed, if it was contained in their archetype. Hence I suspect that these parts were not present in the dBa' bzhed version that the sBa bzhed redactors used.

The last section of the dBa' bzhed, especially, stands out as distinct from the rest of the narrative. S0rensen suggests that it is a ninth-century narrative added as an appendix to the dBa’ bzhed. Other evidence confirms that this section is an interpolation. First, the main part of the dBa' bzhed ends with rdzogs so // at 25b1. Second, Santaraksita had said in the main body of the dBa’ bzhed that Buddhists would no longer debate with non-Buddhists, so this additional section on just such a debate contradicts his prophesy and would make the dBa’ bzhed internally inconsistent if

considered as a single text. Third, Vairocana, who plays no part in the main narrative, plays an important role here. Fourth, he is referred to as a slobs dpon (26b4), which is a late term used neither in early Tibetan documents nor in the main part of the dBa’ bzhed—where mkhan po is preferred. This latter point suggests that this section is perhaps an eleventh or twelfth-century addition, though this would contradict S0rensen’s assessment. Michael Willis recently conducted a study on the date of interpolations and accretions evident in dBa’ bzhed. He independently corroborates the

above arguments and offers new insights into its process of recension. Willis gives a date-range of 1000 to 1100 as a working hypothesis for the core dBa’ bzhed narrative, but notes its later amendments. He concludes that later redactors, over the thirteenth and fourteenth centuries, added narratives to the end of the main story. Several readers then annotated the text from the fourteenth century onwards. I shall therefore omit the initial and final section and the interlinear notes from my analysis. Instead I focus on the core narrative's depiction of Khri Srong lde brtsan, which probably dates to the eleventh or early twelfth century.

Later redactors expanded and transformed the dBa' bzhed narrative into new recensions. These are covered here by the term sBa bzhed, which encompasses sBa bzhed versions S, G, and P. Comparing the recently discovered dBa' bzhed with these three versions sheds light on the process of redacting this important narrative tradition. Pasang and Diemberger began this work, outlining the dBa' bzhed's relationship to later versions of the same narrative in notes to their translation of the text. Here I argue that sBa bzhed G represents an interpolated and in some cases ideologically altered redaction of the dBa' bzhed; and that sBa bzhed P closely follows sBa bzhed G. sBa bzhed S is an edited and reduced version of a source similar to sBa bzhed G, with the additions of an extra section on later history that has earned it the description zhabs btags ma.

The sBa bzhed narrative follows the dBa' bzhed's order of events, but also adds several episodes to the life-story of Khri Srong lde brtsan. The sBa bzhed begins by describing his father, Khri lDe gtsug brtsan (sBa bzhed G 1-3). His Chinese wife, Jincheng Gongzhu, gives birth to Khri Srong lde brtsan. The sNa nam queen, bZhi steng, then abducts the boy, who resists and claims Chinese uterine descent (4-5). After Khri Srong lde brtsan attains the throne (8) and reaches twenty-one years old (14), he gathers his ministers and persuades them to practice Buddhism (19).

Following dBa’ gSal gnang’s recommendation, the king invites Santaraksita to Tibet (20), who then recommends Padmasambhava to tame the site of bSam yas Monastery (27). Padmasambahava partially binds the Tibetan spirits to defend Buddhism, but some ministers force him to leave Tibet before he completes the process (33). Khri Srong lde brtsan then tricks his ministers into agreeing to build bSam yas Monastery (36). Santaraksita performs a ‘heretical' rite to revivify a corpse. This enables the king to fly around the world seeking a fitting exemplar for the construction of bSam yas (40-42). The sBa bzhed includes a lengthy description of the construction and features of bSam yas (42-56), as well as depicting the miraculous appearance of a large lotus at the consecration of bSam yas (57).

Santaraksita ordains the first seven Tibetan monks (57), but this is followed by doctrinal divisions in the samgha and the bSam yas Debate (64-75). The Indian “gradual” school defeats the Chinese “sudden” school, and the victors set up translation centres with the generous patronage of Khri Srong lde brtsan (76). The Tibetan monk-minister and abbot of bSam yas, Ye shes dbang po (dBa' gSal gnang), dies (78). sBa bzhed G ends with a brief description of Mu ne btsan po's reign (79-81) and a discussion of the sBa bzhed's various names. sBa bzhed S contains an extra (zhabs btags) section on Khri gTsug lde brtsan (Ral pa can, r.815-841).

The sBa bzhed not only adds to, but also significantly alters, the dBa' bzhed's depiction of Khri Srong lde brtsan. It portrays the btsan po as a more faithful Buddhist, for instance by ascribing the doubts that he has about Santaraksita in the dBa' bzhed to certain Tibetan ministers. It also emphasises that the btsan po's conversion and the transmission of the Dharma from India to Tibet was predestined in a previous lifetime. Finally, the sBa bzhed shifts the blame for dismissing Padmasambhava from Khri Srong lde brtsan to jealous ministers. These ‘pious alterations' suggest the influence of growing Tibetan religiosity on historiography. I suspect that the sBa bzhed's redactors edited those parts of the dBa' bzhed narrative that depict the btsan po as a cautious, secular ruler, in order to create an equally positive image of Khri Srong lde brtsan as a respectful Buddhist king.

The sBa bzhed’s Process of Recension

Most of sBa bzhed G is created by expanding the narrative contained in the dBa’ bzhed. So, in many parts of the narrative, almost every word of the dBa’ bzhed is included while additions of various lengths are interspersed among them. These additions either clarify the grammar of the sentence or provide extra details to the story from earlier written sources or oral tradition. For example, in one episode the dBa’ bzhed describes Santaraksita prostrating to Khri Srong lde brtsan and then convincing the btsan po to welcome Padmasambhava to Tibet. The sBa bzhed expands this episode, perhaps in line with the growing popularity of religious themes in Tibetan historiography. With the dBa’ bzhed text in bold, it reads:

When Abbot [[[Santaraksita]]] arrived at the palace, he introduced (brda sbyar) [his presence] to bTsan po [[[Khri Srong lde brtsan]]], then prostrated to the btsan po. And the btsan po said: ‘[I] cannot [accept] the prostration of an ordained man.' Then Master [[[Padmasambhava]]] prostrated to a boulder, so that it split into pebbles (shags = shag [ma]). Since the btsan po prostrated, the master asked after his health. ... Acarya Bodhisattva

(Santaraksita) spoke to the btsan po: ‘Previously, when the Buddha dwelt in the world, there was no one among the divinities and demons who was not bound to his oath. Since there are many harmful [[[beings]]] in Tibet who are not bound under oath, the btsan po is prevented from practicing the Dharma. So, in this way, after inviting this the most able [[[master]]] in 'Dzam bu gling, named Padmasambhava, he resides [here]. During the reign of btsan po Srong btsan sgam po, [he] was prevented from practicing the Dharma; and the lakes and reservoirs broke their banks and a great river flooded into 'Phang thang [plain].'

Some of the additions are grammatical modernisations. They unpack sentences to show, for instance, who is speaking and to whom. However, the first major addition in this quote reverses the dBa' bzhed's message. It depicts Khri Srong lde brtsan as unwilling to be superior to Santaraksita and then frightened into prostrating to Padmasambhava by the latter's magic. In the dBa' bzhed, it is natural and proper that the foreigner at the btsan po's court should bow to his imperial authority. But the redactors of the sBa bzhed appear to believe that Khri Srong lde brtsan in fact bowed to the religious master.

This is by no means the only example of such an addition. Identifying in each episode the process of recension that, as a whole, transformed the dBa' bzhed into the sBa bzhed highlights the increasingly Buddhist attitude of Tibetan historians towards the imperial period. The sBa bzhed redactors seek to explain the dBa' bzhed by adding extra details to each scene. Yet these are almost invariably religious (rather than political) details. Instead of explaining the dBa' bzhed, they sometimes even reverse its position (as in the case of Padmasambhava meeting Khri Srong lde brtsan) by making religious figures appear superior to royal ones.

sBa bzhed P follows an ancestor of sBa bzhed G very closely. However, it also contains its own interpolations and modernisations: After Abbot [[[Santaraksita]]] of Za hor arrived at the palace, [he] introduced [his presence] to bTsan po [[[Khri Srong lde brtsan]]]. The btsan po welcomed [him] and since [[[Khri Srong lde brtsan]]] prostrated Master [[[Padmasambhava]]] asked after his health. Master [[[Padmasambhava]]] stretched out the palm of his right hand to a boulder, so that it split into pebbles. . At that time Acarya Bodhisattva (Santaraksita) spoke to the btsan po: ‘Previously, when the Buddha dwelt in the world, there was no one among the divinities and demons who was not bound to his oath. But since there are many harmful deities (lha 'dre) in Tibet who are not bound under oath, the btsan po is prevented from practicing the Dharma. Here and now, after inviting this most able [[[master]]] in 'Dzam bu gling, named Padmasambhava, he resides [here]. Previously, to prevent [you] from practicing the Dharma, a great river flooded into 'Phang thang [plain].'

dPa' bo gTsug lag phreng ba evidently consulted an old version of the sBa bzhed. Yet the final text (sBa bzhed P) contained in his KGT is very much his own. He quotes from a sBa bzhed version within the same recension as the oldest (twelfth-century?) source for sBa bzhed G. However, he omits Santaraksita's prostration and has Khri Srong lde brtsan bow to Padmasambhava. He also adds his own comment on the scene, where he expressly denies any other histories' claims that a religious figure would prostrate to a king. This very thorough and thoughtful historian thus quotes the sBa bzhed and unconsciously contradicts its source, the dBa' bzhed.

sBa bzhed S omits parts of sBa bzhed G/P and the dBa’ bzhed from its narrative, which I shall show below suggests that it comes from a later recension than G and P. sBa bzhed S reads:

[[[Santaraksita]]] arrived at the palace and introduced [himself] to an ambassador in order to meet the btsan po. Then, he prostrated to the btsan po but the btsan po said ‘[I] cannot [accept] the prostration of an ordained man. Then [[[Padmasambhava]]?] prostrated to a boulder that split into pebbles. After the btsan po prostrated, the master asked after his health. ... Acarya Bodhisattva spoke to the btsan po: ‘Previously, when the Buddha dwelt in the world, there was no one among the divinities and demons who was not bound to his oath. But since there are many harmful [[[beings]]] in Tibet who are not bound under oath, the btsan po is prevented from practicing the Dharma. Here and now, after inviting this the most able [[[master]]] in 'Dzam bu gling, named Padmasambhava, he resides [here]. As for expelling, subduing and binding to oath all [those forces] that previously prevented the Tibetan btsan po from practicing the Dharma, this mantrin [[[Padmasambhava]]] will weed them out (sngo thog)'

Here sBa bzhed S omits Santaraksita's description of natural calamities, including flooding in 'Phang thang, that showed the untamed nature of Tibet's autochthonous spirits. But some of those descriptions were present in the dBa’ bzhed (marked in bold in the quote from sBa bzhed G). It is therefore more likely that sBa bzhed S omitted parts of the common ancestor of sBa bzhed G and P than that sBa bzhed G and P's ancestor added elements to sBa bzhed S. Such omissions are widespread in sBa bzhed S. They probably represent a purposeful condensation of the narrative, a new redaction.

Chapter Three shows that the fourteenth-century Mes dbon gsum rnam thar (MBNT) follows sBa bzhed S almost verbatim. I can only conclude that the omissions are recensional rather than transmissional, and that the sBa bzhed S recension existed in the fourteenth century when it was included into MBNT. Given that the sBa bzhed recension on which G and P are based dates roughly to the twelfth century, the sBa bzhed S recension dates to probably the thirteenth century. Yet I find it difficult to understand in any one case (for instance, the references to natural calamities) why the redactors of sBa bzhed S should decide to omit such details.