The Zangs gling ma

The bKa' thang Zangs gling ma is the earliest full-length Tibetan hagiography of Padmasambhava. It therefore contains important depictions of Khri Srong lde brtsan, his motives for inviting the Indian master and their relationship while Padmasambhava was in Tibet. Nyang ral Nyi ma 'od zer (1124-1196) wrote it, though he claimed to have merely discovered it, buried for him by Padmasambhava in the eighth century. Such finds are known in Tibet

as gter ma, “discovered” treasure. They are more deeply revered than real treasure among certain religious communities there. Many Tibetans believe that Padmasambhava himself created almost all gter ma, whether religious objects, ritual texts or sacred histories, and buried them around the

Tibetan landscape for future masters to discover. Tradition calls these masters gter stons, “treasure discoverers.” Most Tibetologists, in contrast, doubt that Padmasambhava authored the gter ma biographies of his life. The texts themselves show signs of their gter stons' authorship rather than Padmasambhava's. I prefer to say that Nyi ma 'od zer, not Padmasambhava, wrote the Zangs gling ma.

mNga' bdag Nyang ral Nyi ma 'od zer (henceforth Nyang ral) was a member of the Nyang/Myang clan. His ancestry stretched back to the beginning of the empire under gNam ri sLong mtshan in the seventh century. A faithful monk-minister of Khri Srong lde btsan, called Ban de Myang Ting nge 'dzin, was praised in the ninth-century Zhwa-temple pillar inscription. The Myang clan therefore enjoyed a high status and practiced Buddhism from the early centuries of its introduction into Tibet. Since the Myang family received royal favour, the clan increased its geographical area of influence even after the fall of the dynasty. Starting from their power base in Myang, they then shared the sBrang mkhar sbre can area of g.Yo ru with another clan. Later they spread south to the lHo brag region of g.Yo ru, where Nyang ral was born.

Nyang ral revived the imperial legacy of both his family and religious lineage using the direct dynastic link offered by gter ma. His biography, the Dri med, describes him meeting the “mad man” sMyon pa don ldan, who declares:

Now there are before me many learned professors of Dharma and practitioners who are accepted as realized yogins. But [[[Nyang ral]]] your coming is like the rising of the sun in the sky, shining for the welfare of beings... Listen up, all you fortunate ones assembled here! The previous king of this border country is these days Nyang the younger, with retreat hair piled on his head (ral pa can ).

I doubt that this was how Nyang ral first discovered he was a reincarnation of a former dynastic king. Perhaps it constitutes a later justification of Nyang ral's claim to that status. Nevertheless, the benefit of this identification is obvious. The Dri med legitimises his religious practices by claiming that he possesses not only rediscovered dynastic texts but also the authority of a reincarnated dynastic king. Nyang ral becomes a flesh-

and-blood personification of the dynastic golden age. Furthermore, as the Dri med shows, he used his unique access to the Buddhist lineages held by Khri Srong lde btsan to find gter ma that the king had buried for his future reincarnation to find. Nyang ral's personal fame and that of his gter ma revelations propelled him to primary position among the greatest ever treasure discoverers, the five ‘kings (of) treasure' (gter rgyal).

The Narrative

Nyang ral wrote two similar historical depictions of Khri Srong lde brtsan. His gter ma hagiography of Padmasambhava is called the bKa’ thang Zangs gling ma (henceforth ZL), since Nyang ral claims to have discovered it in the copper temple (zangs gling) at bSam yas. ZL was perhaps influenced by the no longer extant, and probably quite short, chos ’byung by Rong zom pa Chos kyi bzang po (eleventh century), and Zhang ston bKra shis rdo rje's (1097-1167) more specific history of the Seminal Heart (snying thig) tradition, that purportedly included Myang Ting nge 'dzin.

Nyang ral's extensive history of how Dharma arose in Tibet is called the Chos ’byung Me tog snying po sbrang rtsi’i bcud (MTN). I suspect that Nyang ral wrote the history based in part on the hagiography, although MTN omits Padmasambhava's activities in India, before coming to Tibet. It also includes some other episodes in the life of Khri Srong lde brtsan, not present in ZL (see below). Here is a brief precis of their shared content:

Bodhisattva Manjusri incarnates as Khri Srong lde brtsan in order to establish Buddhism in Tibet and tame the savage Tibetan people. Light emanates from his heart and enters the womb of Khri lDe gtsug brtsan's wife, Queen Jincheng Gongzhu. Khri Srong lde brtsan is born with the marks of a ‘child of the gods' (lha’i bu). At twenty-one years of age, he aspires to practice the Dharma by building bSam yas Monastery. He invites Abbot Santaraksita

to tame the subterranean spirits around the construction site; but the Indian master fails to bring them under control. Santaraksita recommends Padamasmabhava from Uddiyana, a siddha who is more skilled in the Mantrayana. Padmasambhava arrives in Tibet and performs many miracles, eventually binding every Tibetan deity to an oath to protect Buddhism. The spirits help to build bSam yas, as do the wives of Khri Srong lde brtsan. Padmasambhava and Santaraksita consecrate the main shrine, and then ask to return home. The king implores them to stay and teach Buddhism for the good of Tibet. The two masters agree.

Tibetan and Indian scholars then work together to translate all the Buddhist scriptures from Sanskrit. Khri Srong lde brtsan sends two Tibetans, Nam mkha’i snying po and Vairocana, to search out new teachings in India and return with them to Tibet. Khri Srong lde brtsan rewards both for their success, but some jealous ministers conspire to remove them from court. Nam mkha’i snying po remains, but Vairocana is banished to Tsha ba rong, much

to the dismay of Khri Srong lde brtsan. The king invites another great abbot, Vimalamitra, who displays his disinterest in worldly status in a similar fashion to Padmasambhava. He refuses to prostrate to the king, who instead bows to him. Khri Srong lde brtsan then invites all the scholars and masters to bSam yas where they translate many more texts.

Finally, the narrative returns to Padmasambhava. He initiates Khri Srong lde brtsan into several mahayoga lineages and prophesies that the king will discover the teachings again in a future life. He extends the lifespan of Khri Srong lde brtsan, but eventually the king dies and Padmasambhava leaves Tibet to tame another country’s demons. His lengthy farewell speech includes advice for all members of society, from the new king, Mu tig btsan po, downwards. It especially recommends practicing the mantra of Avalokitesvara, om mani padme hum. Mu tig btsan po, his retinue and the disciples of Padmasambhava are anguished at his departure, but vow to practice Buddhism and govern the country in the manner of the master and Khri Srong lde brtsan respectively.

I believe that I have discovered the earliest recension of ZL. This text underwent a process of interpolation akin to that which transformed sBa bzhed 1 into sBa bzhed 2 (see Chapter One, above). The ZL version included in the Rin chen gter mdzod is longer than the original ZL, as is the version that Blondeau studied in the late 1970's. There is good evidence to suggest that different redactors edited and augmented these versions at

some time after Nyang ral wrote or dictated the original ZL. However, several versions of ZL that I have collected recently show no signs of major additions or omissions. In fact, their contents best fit what information it is possible to gain of the original ZL from the evidence of MTN. This discovery is vital for my assessment of Nyang ral's depiction of Khri Srong lde brtsan in ZL, since a number of the episodes of his life in the Rin chen gter mdzod and Blondeau versions are probably later interpolations.

The published editions of the ZL text (ZLa, b and d), in 41 chapters, are all very similar. As a result, it appeared to Tibetologists that all versions of ZL contained a single narrative of the Indian Buddhist master, Padmasambhava—beginning with his father's search for a wish-fulfilling jewel and a son and ending with his farewell to Tibet after the death of King Khri Srong lde brtsan. In 1936, the pioneer of Tibetan historiographical study, A.I. Vostrikov, identified ZL as one of Nyang ral's gter ma discoveries:

the so-called “Will of Padma” (Padma-bka'-chems) ... is supposed to have been found by the Tibetan king Nan-ral.

Early Tibetologists identified ZL with the version that opens the nineteenth-century Rin chen gter mdzod. Giuseppe Tucci is the only exception, since he had his own version of ZL and was aware of its divergences from the Rin chen gter mdzod edition. Yoshiro Imaeda then found and passd on to Anne-Marie Blondeau another version in the Kathmandu National Archives, in 45 chapters, which is different from any of the published versions. Blondeau made several studies in French of a conspicuous addition to the beginning of this ZL version—the story of the stupa (reliquary) in Nepal that

Tibetans call Bya rung kha shor. Blondeau also pioneered the study of Padmasambhava's biographies as literature, codifying them into two emic Tibetan categories: the so-called “Indian tradition” that records that he was born naturally from his mother's womb; and the “Tibetan tradition” that relates that he was born supernaturally, from a lotus. ZL thus conforms to the second category. In the same article, Blondeau notes a reference in KGT to a ZL in 48 chapters. She comments in English:

We therefore have further evidence here of the revisions undergone by the gter-ma since their ‘discovery', which makes any hypothesis based on their chronology so difficult.

Erik Pema Kunsang then translated ZL into English on the basis of ‘the first text in the ... Rinchen Terdzo.’ His translation is in most respects very good. Yet his overall dependence on the Rin chen gter mdzod ZL further perpetuates the myth of its authoritative status as the twelfth-century ZL.

Instead, this chapter categorises all 11 versions of ZL currently available to scholars into three distinct recensions—ZL1 (versions a-e), ZL2 (f-g) and ZL3 (h-k). I first catalogue the material contained in ZL1 and ZL2 but not included in ZL3. Second, I identify borrowings from ZL in later Tibetan Buddhist literature: Nyang ral's own MTN and two fourteenth-century biographies, the Padma bka' thang and gSer phreng. MTN follows ZL3, so

that recension has the greatest claim to a twelfth-century origin. The Padma bka’ thang also follows ZL3 in part, further strengthening this claim. The later gSer phreng, surprisingly, appears to synthesise the Padma bka’ thang with ZL1. This witness is of too late a date to ensure a twelfth-century origin for ZL1. No later history follows the ZL2 redaction, which may be only a few centuries old. I conclude that ZL3 was not only the oldest, but also the most widespread and influential ZL recension, before 'Jam mgon Kong sprul legitimised ZL1 by including it as the first text of his popular Rin chen gter mdzod.

ZL1

This recension represents the shared content of ZLa-e. Leaving aside their transmissional differences, which I discuss in the appendix, versions a-d are strikingly similar to each other. They share the same order of episodes, scenes constituting those episodes and wording of almost every element of those scenes. On average, for every folio of ZLa (the Rin chen gter mdzod version), it will diverge in only one minor spelling variant, omission or addition from ZLb, c or d. This variation is really rather miniscule. Tucci's ZLe, in contrast, contains many transmissional errors, on almost on every line. However, its content (mistaken omissions apart) entirely agrees with ZLa-d against ZLf- g (recension 2) and ZLh-k (recenson 3). Therefore ZLa-e together constitute a single recension, ZL1.

Comparing this recension with ZL2 and 3 reveals many elements and episodes contained only in ZL1. These include: Chapter 1—More detail in the episode where King Indrabhuti gains a wish-fulfilling jewel from the castle of precious substances (ZLa 3a5-4a3). Chapter 2—The entire episode in which Padmasambhava is forced to flee from the charnel ground called bSil ba'i tshal because of the evil king, Cakraraja (ZLa 7a6- 7b6).

Chapter 3—The episode where Padmasambhava visits rDo rje ldan (Bodhgaya) in India and two Buddhist monks meet and pacify him (ZLa 8b1-9a1). Chapter 4—The episode in which pandits at rDo rje ldan call on Padmasambhava in

a ceremony to help them defeat four heretics in debate (ZLa 13a4-14b4).

Chapter 9—the episode where Padmasambhava finds water for his royal escorts at sTod lung by driving his staff into rocky ground (ZLa 22b3-5). Chapter 13—In the episode in which Khri Srong lde brtsan sends Nam mkha'i snying po to India, the additional element which makes the idea Padmasambhava's rather than the king's (ZLa 33a2-6).

Chapter 14—A similar element assigning agency to Padmasambhava for sending Lo tsa ba Vairocana to India (ZLa 67b3-5); the episode where Vairocana is taught the doctrine from under a copper vessel and through a copper pipe (ZLa 38a6-38b2); the inclusion of the jealous king of rDo rje ldan in the Indian narrative (esp. ZLa 41b2-6); and the first and last parts of the episode in which Vairocana is banished to Tsha ba rong (ZLa 42b5-43a1 and 43a5-43b3).

Chapter 16—The parts of the story of Vimalamitra in India, in which he takes ordination from Sri Simha, goes to live at VikramasTla Monastery (ZLa 47b4-6) and reflects that he must travel to Tibet (ZLa 48b2-4); and the long episode in which Vairocana sends g.Yu sgra snying po to bSam yas, meets with Vimalamitra and teaches Buddhism to Khri Srong lde brtsan (ZLa 50b2-52b1).

Chapter 18—The episode where Padmasambhava is foiled in his attempt to give Khri Srong lde brtsan the nectar of immortality from the vase of life (ZLa 55a2-56a4); and the part of Padmasambhava's prophecy concerning Lha lung dPal gyi rdo rje (56b3- 4).

Chapter 19—The episode in which Padmasambhava takes Ye shes mtsho rgyal as his consort (ZLa 58b2-4); more detail in the description of the allotment of bKa' brgyad bDe gshegs 'dus pa deities to Khri Srong lde brtsan and eighth-century Tibetan masters (ZLa 60b6-61b2).

Chapter 20—The episode where the nagas fill Khri Srong lde brtsan's treasuries with wealth (ZLa 62a6-64a3).

Chapter 21—The precis of Khri Srong lde brtsan's life, and the description of his death and dissolution into the heart of Manjusri (ZLa 71a6-71b5). Chapter 22—The record of Mu tig btsan po's ascension to the throne and the death of Santaraksita (ZLa 71b5-72a4); and Padmasambhava's allusion to those events (ZLa 72b2-3).

There are too many additions in this list to be explained as accidental omission or even missing folios in versions ZL2 and 3. ZL1's additions are also internally consistent with other passages unique to that recension. This is why I posit a recensional difference between these versions of ZL, rather than a transmissional difference.

ZL2

Blondeau’s ZLf and my ZLg are so similar to each other that they share unique transmissional errors against all the other versions of ZL. They constitute a different recension from ZL1 and 3 also because of a number of large additions present in both ZLf and ZLg and nowhere else: Chapters 1 and 2 are unique in ZL2 and added before the other versions even begin. As a result, no chapter numbers in ZL2 align properly with ZL1 and

ZL3.

—Chapter 1 contains an extended episode (e.g. ZLf 1b1-7a2) telling the story of the previous lives of the main ZL protagonists in Nepal, building and worshipping the Bya rung kha shor Stupa. Then a precis of Padmasambhava's eight incarnations (ZLf 7a-12a4). —Chapter 2 (ZLf 12a4-13b2) tells some of the stories, set in Amitabha's pure land, etc. that are usually associated with later rDzogs chen and especially Snying thig histories.

Chapter 3 [[[Chapter]] 1 in ZL1 and 3] provides more detail (of a different sort to ZL1, this time with 8 white and 8 gold goddesses, etc.) in the episode where King Indrabhuti gains a wish-fulfilling jewel from the castle of precious substances. Chapter 10 [8] includes an episode in which Padmasambhava tames demons and spirits in Nepal on his way to meet the messengers at Mang yul (ZLf 34b2-39a4).

Chapter 11 [9] records more taming of local forces in Mang yul (ZLe 40b3-41a5).

Chapter 12 [10] narrates more taming of Tibetan deities around Has po ri (ZLf 47a5- 47b3 and 48a5-48b1).

ZL2's additional episodes suggest an interest in Nepal, in rDzogs chen and in taming narratives.

ZL2 also stands alone against the rest of the tradition in many minor details. ZLg diverges from ZLf, on average, 12 times per folio of ZLf, and ZLf stands closer to the general tradition than ZLg. Nevertheless, both versions' general similarity to each other and divergence from ZL1 and 3 outweigh such small differences. While ZL2's redactors altered only a minority of chapters, its transmission is more corrupted than that of ZL1. ZL3

ZLh-k are all new discoveries, either mine from the Kathmandu National Archives or Karma Phuntsho's from Bhutan. They contain all the same episodes as each other, and none of the additional material found in recensions ZL1 or 2. ZLh, ZLj and ZLk also share errors against ZLi, for example wrongly locating Uddiyana in the south. ZLi correctly reads ‘west' (2b1), because an editor or scribe apparently rectified the line, rather than because it belongs to one of the other recensions. Its editor also changes slob dpon to ghu ru throughout, so theoretically ZLi is its own recension. But it is a lone version against ZLh, ZLj and ZLk, its alterations postdate its agreement with them, and in all other respects it follows them. Therefore it is easier to collect ZLi with these versions against ZL1 and 2 for the purpose of debating their antiquity as recensions.

Comparing ZL with Later Quotations

It is evident that the 11 versions of ZL are classifiable under three recensions. But which recension most closely resembles Nyang ral's original ZL? It is important to judge their relative claims to authenticity, since I need to quote one or more versions as representative of a twelfth-century depiction of Khri Srong lde brtsan. Relying on a later redaction, or including an interpolated episode into my analysis below, will mislead my investigation of eighth to twelfth-century portrayals of Khri Srong lde brtsan. There are four options:

a. the unique details in either ZL1 or ZL2 are original and omitted in the other of the two and in ZL3

b. the unique details in ZL1 and ZL2 are later additions to the original recension, ZL3

c. some of the unique details in ZL1 and/or ZL2 are original, others are later additions

d. two or all three recensions constitute different, original recensions that Nyang ral wrote or dictated during his life.

In order to assess the antiquity of the extra episodes in ZL1 and ZL2, I examined later sources in the Tibetan historical tradition. The ZL narrative most influenced Nyang ral's own twelfth-century chos 'byung, MTN, and two fourteenth-century bKa' thangs, the Padma bka' thang and gSer phreng. These texts borrow many scenes and episodes, and even include some loose quotations, from one or another version of ZL. By comparing the three ZL recensions with these later texts, I have concluded that option b. is the most likely description of ZL's redactional history. This means that ZL3 holds the greatest claim to be the original ZL recension.

MTN is especially important in deciding the case in favour of ZL3. ZL3 contains no episodes that are not also in ZL1 or 2, while ZL1 and 2 each have their own additional episodes not present in ZL3. These additional elements are all missing from the same places in MTN.

The Me tog snying po MTN only covers the Tibetan section of ZL. Where it follows ZL, MTN seems to be based on ZL3, since it contains none of the extra episodes from ZL2 and only one of the extra episodes from ZL1, the “vase of life” episode. However, it includes that episode at a later point in its narrative than in ZL1. I believe that the inclusions of this episode in both ZL1 and MTN occured indepently of each other. It seems more likely that someone (possibly Nyang ral) added this episode to MTN because it was present in the sBa bzhed tradition, rather than because it had been previously included in ZL. MTN contains six other episodes from the life of Khri Srong lde brtsan not featured in any ZL, but instead resembling the sBa bzhed. Here I shall cite MTNd and sBa bzhed G, though these episodes exist in the other versions as well:

1. Khri Srong lde brtsan's speech aged five (MTNd 273.4-11 and sBa bzhed G 4-5).

2. The choice that he gives to his ministers regarding what to build in Tibet (MTNd 274.6-19 and sBa bzhed G 36).

3. His support for a heretic rite in order to fly to India in search of an exemplar for bSam yas Monastery (MTNd 274.20-75.14 and sBa bzhed G 40-42). 4. The king's dream that leads to finding bSam yas' statues of the gods on a hillside (MTNd 293.17-94.6 and sBa bzhed G 42-43).

5. The appearance of a huge lotus at the consecration of bSam yas (MTNd 302.20-03.12 and sBa bzhed G 57).

6. The bSam yas Debate, between Hwa shang Mahayana and Kamalasila, and its consequences (MTNd 395-411 and sBa bzhed G 62-78).

At present I am not sure whether the additional episodes in MTN belong to the original twelfth-century version. Dan Hirshberg is collating the three known versions of MTN, as part of a PhD at Harvard, and further shall have to await his analysis. Nevertheless, it appears that the “vase of life” ZLf 83b2-6 reads: de nas mnga' bdag gis slob dpon rnams la bdag bsod nams yang rab tu gyur pas / bsam pa thams cad 'grub / dam pa'i chos nyi ma shar ba lta bur bsgyur nas bshad pas bka' drin che / da yang bdag la tshe yang rab tu gyur pa'i mchog cig zhu'o zhes tshogs kyi 'khor lo mdzad nas zhus pas / de la slob dpon

rgyal srid dar bar mchi / glang gi lo'i lo bar rgyal pos lnga bcu rtsa drug bzhes / de la rgyal po 'das dus yin /

MTN follows almost exactly the same narrative as ZL2 and 3 here (compare MTNd 335.18-36.5). Elsewhere, MTN (MTNd 347.1-351.12) includes a long, repetetive and garbled account of the vase of life narrative. At this point all versions of ZL merely begin a new chapter. Like the sBa bzhed account, MTN only records one attempted ritual, not two (as in ZL1). Also, MTN has the ministers describe the content of the vase as '"Mon water"'— whereas ZL1 says '"poison"'. Like ZL1, however, MTN begins the account with Khri Srong lde brtsan requesting the empowerment from Padmasambhava (though not in the same words) and ends it with Padmasambhava burying the vase as gter ma (along with others).

episode is included in MTN because it is in the sBa bzhed. Overall, the evidence of MTN suggests that ZL1's extra episodes were interpolated without reference to MTN. Perhaps they both include coincidental similarities because they both draw on the same bKa' thang/sBa bzhed tradition. There are better grounds for dismissing the idea that ZL2 influenced MTN. None of the episodes of ZL2's section on Padmasambhava in Tibet appear in

MTN, although MTN does include its own additional description of taming in Tibet (MTNd 278.18-80.9). This disparity provides a reason to doubt the existence of ZL2's extra episodes in the twelfth century. If the ZL2 additions had been available to Nyang ral while writing his MTN (or vice versa), surely he would have used them, rather than add other narratives in their place. Scholars do not yet understand Nyang ral's writing practices. Yet it does not make sense of the evidence to posit that ZL2's additions were written in the twelfth century.

MTN stands closest to ZL3. Despite adding a large amount of new material to its description of Khri Srong lde brtsan, MTN does not omit any narrative contained in the Tibet section of ZL3. Instead, none of the major additions in ZL1 chapters 9-22 or ZL2 chapters 10-12 are included in the same place in MTN's narrative, where it would have been appropriate for Nyang ral to position them. Most probably, Nyang ral wrote or dictated the MTN section

on Khri Srong lde brtsan using an ancestor of ZL3 as a template. He or some later redactor then added extra information that (s)he had read in, or heard read from, a text in the sBa bzhed tradition. If Nyang ral used a ZL3 ancestor for MTN, it does not necessarily mean that he did not create more than one ZL recension. However, the misalignment and disparity of both the vase of life episode and the taming narrative, between MTN and ZL1 and ZL2 respectively, means that there is reason to doubt that these additions existed in ZL before or at the time of MTN's creation. There are no such doubts concerning ZL3's antiquity.

The evidence of MTN suggests that ZL3 retains most faithfully the overall, unadorned narrative that Nyang ral wrote in the twelfth century. It is not merely the least accreted version, since the Tibet section of ZL3 contains no major elements missing from MTN. This fact indicates that ZL3 has not been augmented at all. It appears that Nyang ral included the Tibetan portion of that narrative in his MTN, probably after he wrote ZL. Then he (or a later redactor of MTN) added extra episodes from the sBa bzhed tradition, other sources and/or personal visions. MTN was a near contemporary of the original ZL and followed an independent transmissional/recensional trajectory from ZL down to the present. It thus constitutes my primary witness for the relative antiquity of ZL3 against ZL1 and ZL2.

However, I found it necessary to corroborate this witness with other texts that quote ZL. Nyang ral could only have used one recension as the basis for MTN. The fact that he follows ZL3 therefore does not necessarily mean that he did not write ZL1 and/or ZL2 (option d. above). It only indicates that he did not incorporate them into MTN. As I shall show, one later text agrees with ZL1, perhaps dating its extra episodes to the thirteenth or fourteenth century.