Tibetan Deity Cults As Political Barometers

Tibetan Renaissance Seminar > Student Generated Content > Interpretative Topical Essays > Tibetan Deity Cults as Political Barometers

Tibetan Deity Cults as Political Barometers

by Christopher Bell


Most of Tibetan history has been plagued by political instability, with a few brief moments of order and centralization. Because of this, many centuries are characterized by limited hegemonic control occurring either simultaneously in different areas of Tibet or subjugating and replacing one another in tumultuous shifts. After the fall of the Tibetan dynasty in the ninth century, the remnants of the imperial line split between west and central Tibet, both establishing much weaker political domains. Constant in-fighting and political splits kept either from growing in size or power to ever equal the previous Imperium.1 Once Buddhism was reestablished in Tibet, a number of monastic hegemonies took root and vied with each other for political control. In central Tibet, these included the hegemonies of the Sakyapa (c. 1249-1354), the Phakmo Drupa (c. 1354-1478), the Zhamarpa (c. 1478-1565), the Karmapa (c. 1565-1642), and the Gelukpa (c. 1642-1705). Beyond these were the several brief occupational hegemonies set up by foreign powers, such as various Mongol tribes and finally China.2

With these numerous shifts in power, Tibet has never completely been a unified land. Rather, it has been more akin to a ‘galactic polity,’ a conglomeration of regional administrations, at times central in relation to others and at times subordinate. This model was applied to Southeast Asia by Stanley Tambiah and to Tibet by Geoffrey Samuel.3 Samuel likewise draws a parallel between this model and that of the maṇḍala, a cosmologic and architectonic symbol that has so captivated the Tibetan psyche.4 As different Tibetan clans and castes fill the landscapes of the many Tibetan domains, so too do a variety of divine and demonic species fill the landscapes of the numerous maṇḍalas. This similarity is not superficial; the societies of gods and demons elaborated in Tibetan myths and legends are clearly modeled on Tibetan social forms (or rather is it the other way around?), and include everything from family feuds and kidnapping to inter-tribal conflict and scandal. The origin of many Tibetan clans usually entail a root family beginning on earth after a convoluted conflict ends in the sky realms, a god exiled or fleeing down a sacred mountain in order to begin a new life among mortals. Furthermore, the activities of gods and humans tend to overlap a great deal in the Tibetan cultural sphere, so much so that it is no surprise to find a change in the human realm coinciding with a change in the divine realm; indeed, it would seem that in many instances, one is a catalyst for the other.

Thus, it is reasonable to say that changes in the divine community can shed light on political and cultural shifts in Tibetan history and vice versa. As such, a number of notable intersections between the divine (or demonic) and the earthly will be explored that illustrate the complex relationship between the two. First, I will examine the popular narrative surrounding the establishment of Buddhism in Tibet and the subjugation of the indigenous gods and demons. Second, I will discuss the cults of Vairocana and Avalokiteśvara, their involvements with the Tibetan dynasty, and how their characteristics reveal the structure of Tibetan society in their respective periods of propagation. Then I will single out two protector deity cults that exemplify more localized hegemonic shift. Finally, I will explore the power of charisma and charismatic leaders in the popularization of deity cults.

Converting the Gods

While there are innumerable conflicting narratives active in Tibetan Buddhist culture, a number remain consistent with only minor alterations in details. One in particular is the establishment of Buddhism in Tibet in the seventh century and its subsequent public endorsement, as well as the establishment of its monastic tradition, in the eighth century. This story begins with the king of Tibet at the time, Trisong Detsen (Khri srong lde btsan; 742-796), who adopted Buddhism and desired to establish a monastic institution in Tibet.5 The account of what follows occurs in many sources, though I provide here Snellgrove’s retelling for its detail and brevity:

Khri Srong-lde-brtsan’s interest in Indian Buddhism was expressed most clearly by the foundation of the great monastery of bSam-yas [Samyé], completed in 779. The instigator of this great event seems to have been a certain gSal-snang of the sBa family, who had been governor of the southern district of Mang-yul, which bordered on the Himalayas, and who had visited Buddhist centers in India and Nepal, making the acquaintance of the Indian teacher Śāntarakṣita and commending him to his royal master. gSal-snang had already received consecrations from Śāntarakṣita in Mang-yul, where they appear to have founded temples together; he received the religious name of Jñānendra (Ye-shes dbang-po), meaning “Lord of Knowledge,” by which name we shall refer to him hereafter. Some interesting details are given in the later accounts suggesting that the Chinese connection already had a considerable hold on the king. Thus it is recounted that he continued to sends scholars to China in quest of Buddhist works, including even Jñānendra himself according to some sources, although this seems unlikely. Also before inviting Śāntarakṣita he is said to have sent representatives to examine him on the nature of his Buddhist teachings. He arrived with a Kashmiri interpreter, but there was opposition to his presence and he soon had to leave, since his arrival unfortunately coincided with a number of natural calamities, indicating the displeasure of local divinities at his presence. He then recommended to the king the name of Padmasambhava as one who would be able to deal with such obstructions, and carrying the invitation personally he returned to Nepal, where he met Padmasambhava who promptly set out on his special mission. Only when he had successfully quelled the gods and demons of Tibet did Śāntarakṣita, who thus appears in the role of his partner, return to preside over the foundation of bSam-yas Monastery and the first official ordination of six or seven Tibetan monks, among whom were Jñānendra, Vairocana (a disciple of Padmasambhava), a renowned translator rMa Rin-chen-mchog (a disciple of Vimalamitra), Sang-shi Ratna of sBa, and probably also Śrighoṣa (dPal-dbyangs).6

While they are referred to as gods and demons, neither word properly defines Tibetan deities, who have attributes characterized by both terms, hence their synonymous use. These divinities can be at times beneficent or pernicious, creating fortune or calamity, depending on their attitude, which is quite capricious. And so, the native Tibetan deities were too fierce for Śāntarakṣita to handle, but Padmasambhava was able to subjugate them with his tantric might and thus allow for the successful construction of Samyé monastery. Fuller accounts detail the constant trouble encountered during the construction of Samyé and the necessity to perform binding rituals against the local divinities that were causing the obstructions. Regardless of the details, the significant element in this story is the subjugation of local gods by the practitioner of a foreign religion, specifically Buddhism. This subjugation involved converting the deities to Buddhism and making many of them protectors of the Buddhist teachings. Also, many of them were placed in servile positions within the entourage of the newly imported and more powerful divinities of Indian Mahāyāna Buddhism, the bodhisattvas.

Thus, it could be said that the conversion of Tibetan deities to Buddhism—a precursor to the conversion of the Tibetan population—represents the first institutional hegemonic shift in Tibetan Buddhist history. The adoption of Buddhism in Tibet was accompanied by conflicts with indigenous forces, with the foreign religion eventually usurping the place of the previous religious tradition of Tibet. Certainly such traditions were acting contrary to the adoption of Buddhism in the imperial period, especially within the court itself, where the purveyors of the pre-Buddhist royal religion were actively working against Buddhism.7

This subjugation of pre-Buddhist Tibetan forces occurs alongside the reimagining of Tibet along new cosmological lines, where the Tibetan landscape is thought of in terms of the architecture of a maṇḍala. This is not only indicated by the architecture of Samyé, which was designed to look like a maṇḍala, but also by a mythic narrative involving the first Buddhist king of Tibet, Songtsen Gampo (Srong btsan sgam po; 617-650). In this narrative, Tibet is envisioned as the body of a giant demoness in repose and representing the wild and unpredictable environs of the lands. King Songtsen Gampo subjugates her by pinning her down with thirteen Buddhist temples, the most important of which is the Jokhang temple in Lhasa, which was built over a filled-in lake that represented the demoness’s life blood. With this dual element, of subjugation and maṇḍalification, Tibet begins a process of transforming into a great Buddhist maṇḍala.

The indigenous gods have now been tamed and converted to Buddhism and placed under the control of bodhisattvas. These bodhisattvas will start to actively populate the Tibetan landscape, transforming it into a Buddhist domain, and becoming the central figures in popular new deity cults. Yet, despite their power to reshape and even create worlds, these powerful foreign divinities will not entirely replace the lesser gods now under their control, for the latter are still propitiated and worshipped. However, for the next two discussions, we will turn our attention to those bodhisattvas who will come to represent royal power in various contexts and who will be involved in the fuller process of the maṇḍalification of Tibet.

The Divine King

The construction and consecration of the Samyé monastic complex represents the official endorsement of Buddhism by the growing Tibetan empire in the late eighth century. Its architectural design is the Buddhist maṇḍala, a symbol rife with cosmological and political overtones.8 The local deities having been subjugated, the next step was to propagate the esoteric Buddhism adopted by the dynasty. Yet the texts and practices of Buddhism did not become free to all; rather the court closely monitored and controlled the transmission of the new religion so that its institution remained commensurate with the goals and agendas of the empire. These goals involved increased centralization, cultural cohesion, political expansion, and royal power. In the cosmology of Mahāyāna Buddhism within the powerful structure of Vajrayāna, the bodhisattva who embodied these attributes, and thus was the ideal central divinity for the burgeoning empire, was Vairocana. Kapstein succinctly discusses Vairocana’s importance as the royal divine figure of the empire:

Unfortunately, we do not have precise contemporaneous evidence regarding the foundation of Samye and its plan, but the Testament of Ba’s account of its architecture and design is remarkably detailed and probably in large part authentic. According to this account—and here the later Tibetan histories and the plan of the later restorations of Samye all concur—Samye was designed in a manner that gave special prominence to Vairocana. It was Vairocana who was installed as the central divinity in the second storey, while the four-faced Sarvavid-Vairocana, surrounded by the eight foremost bodhisattvas and other deities, making up a maṇḍala of forty-two, occupied the uppermost shrine. The central divinity on the lowest storey was Śākyamuni, who is in this context also the Nirmāṇakāya, the emanational embodiment, of Vairocana. Of course, the plan of Samye included many other deities who do not seem to have been so closely related to the Vairocana traditions, but it is difficult to resist the impression that esoteric Buddhism was represented at Samye first and foremost by Vairocana. As Richardson [1998] has shown, this orientation was recapitulated in several other Central Tibetan temples of the late imperial and early postimperial period, and, indeed, the association of Vairocana with royal cult appears to have endured in West Tibet well into the second millenium.

It is furthermore important to recall in this connection that the Tibetan court appears to have been very restrained in its commitment to the esoteric traditions of tantric Buddhism. The circulation of the tantras was restricted by order of the court, and permission to translate and to transmit them strictly controlled. It is against this background that the apparent prominence of Vairocana becomes particularly significant. In eighth- and ninth-century Tibet, the tantric free-for-all that comes to characterize aspects of Tibetan Buddhism during the eleventh century had not yet occurred, so that the accentuation of particular esoteric traditions in the otherwise intentionally limited field of imperial-period Buddhism gains considerable gravity.9

Kapstein goes on to discuss the Mahāvairocanābhisambodhi tantra, which was “central to the officially sanctioned esoteric Buddhism of the Tibetan empire,”10 as well as further architectural, sculptural, and iconographic evidence to illustrate the prevalence of the Vairocana cult in imperial-period Tibet. What is significant is that the bodhisattva that was popularized, Vairocana, was chosen by his imperial supporters because he embodied imperial motivations. Vairocana is a king bodhisattva sitting at the center of his maṇḍala, his kingdom, expanding outward and overtaking the world with Buddhist Truth. All of these elements parallel the imperial project at this time:

[T]he cult of Vairocana was widely promulgated with imperial support, and…it expressed a significant homology between, on the one hand, emperor and empire, and on the other, Vairocana and his maṇḍala or realm. …Trhi Songdetsen and his successors sought not merely to present to their dominions conceptual analogies and symbols, but rather to make use of these in a thoroughgoing “maṇḍalification” of the kingdom that surely also involved the promotion throughout Tibet of temples, teachers, book copying, ritual practices, and much else besides. The conversion of Tibet, therefore, was from this perspective much more than the adoption of an alien religion, as if it were a question of the application of a mere patina or veneer; it was to be the wholesale conversion, the fundamental transformation, of a human domain into a Buddha-realm, an empire governed by superhuman insight, power, and law. For this indeed was the imperial ideal, already latent in indigenous Tibetan conceptions of the Tsenpo’s divinity, and Buddhism provided an exceptional vehicle for the expression of it.11

With the fall of the empire, the cult of Vairocana lost its central power, focus, and patronage, though certainly Vairocana is still propitiated in several monastic contexts. Samyé monastery fell into disrepair and remained empty until it was reconsecrated in the eleventh century by Ra Lotsāwa Dorjé Drak (Rwa lo tsā ba Rdo rje grags; 1016-1128/1198).12 This time period marks the start of the Tibetan renaissance, a period of intense religious ferment and social upheaval occurring after a century and a half of scarring political, cultural, and economic disintegration. The Tibetan dynasty, its bodhisattva, and its religious strictures are gone, allowing for new opportunities of religious expression. Nonetheless, the memory of the once glorious dynasty has remained within the collective Tibetan conscience and created a longing to constantly engage with that past. This desire helps give rise to a new deity cult, one centered on another more powerful bodhisattva, and this deity cult will be involved in a fuller reimagining of Tibet as a pure Buddha realm.

The Divine Kingdom

Tibetan Buddhist history is divided into three great periods as defined by a common Tibetan scheme. The seventh to the ninth centuries represent the first transmission of Buddhism (snga dar) in Tibet, from its adoption by the court during Songtsen Gampo’s reign to its end during the reign of the last dynastic king, Lang Darma (Glang dar ma; 803-842). The ninth to tenth centuries represent the dark age, or period of fragmentation (bsil bu’i dus), that occurred in the wake of the empire’s disintegration. The eleventh century onward represents the second transmission of Buddhism (phyi dar) in Tibet. This latter period is the beginning of the Tibetan renaissance, an era of great social and religious activity, where monastic Buddhism is reintroduced, religious institutions, temples, and monasteries are reestablished or founded, and influential religious leaders come to the fore of Tibetan history as instigators of new religious movements. This era is roughly from the eleventh to the early fourteenth century, after which such activity has cooled and those movements that were most resilient have solidified into religious sects and lineages.13

Before such sectarian designations, though, all of Tibet was actively involved in the construction of a new religious and cultural identity aided by these new religious movements and their charismatic leaders. This developing identity partly involved taking Buddhism out of the monasteries and elite spheres and popularizing it on the lay level. The new religious group most successful at doing this at this time was the Kadampa (bka’ gdams pa), a reform movement begun by the famous Indian master Atīśa (b. 972/982). It was especially through the efforts of three charismatic figures that the Kadampas were able to cultivate a strong popular base—a necessity for any strong and lasting institution. These figures, the “three brothers,” were Puchungwa Zhönu Gyeltsen (Phu chung ba Gzhon nu rgyal mtshan; 1031-1106), Chenga Tsültrimbar (Spyan snga Tshul khrims ’bar; 1038-1103), and Potopa Rinchensel (Po to pa Rin chen gsal; 1027-1105). Being transmission holders of Atīśa’s practices and doctrine, these and other Kadampa leaders popularized Buddhism through a number of successful means.14 For our purposes, it is important to note that the central bodhisattva propagated by the Kadampas, and thus given a great deal of attention by the growing Tibetan laity, was Avalokiteśvara. This is understandable since Atīśa himself actively promoted practices dedicated to Avalokiteśvara while in Tibet. Ronald Davidson succinctly describes the proliferation of this new cult:

With the Kadampa emphasis on popular religion, Kadampa preachers like the three brothers turned Avalokiteśvara and Tārā into the religious ancestor/ancestress of the Tibetan people, so that the eventual ideology of Songtsen Gampo and his queens as the emanations of these divinities was made possible by Kadampa missionary activity after Atiśa’s demise. The eleven-headed Avalokiteśvara practices, Atiśa’s legendary conversations with the goddess Tārā, the difficult law-and-order situation in Tibet, and the emphasis on these two bodhisattvas brought by other Indian masters during the late eleventh century all assisted the focus on the two deities who save devotees from the eight great dangers. This movement eventually spawned such Kadampa mythic and meditative practices as the “doctrine of the sixteen spheres” (thig le bcu drug gi bstan pa) and made popular a lay-oriented, Avalokiteśvara-focused fasting program (smyung gnas), whose propagation was closely associated with the Kadampa monks. In this surge of popular religiosity, the position of the Jokhang in Lhasa became central. Unlike such big monasteries as Samyé, the Jokhang[’s]…main purpose was the intersection between the Tibetan people and the Buddhist divinities.15

In the eleventh and twelfth centuries, the cult of Avalokiteśvara grew exponentially in part due to the popularization of a legendary narrative tying Avalokiteśvara back to the first Buddhist king of Tibet, Songtsen Gampo, who comes to be considered an emanation of Avalokiteśvara. Alongside this is the re-creation of the traditional Tibetan myth, in which the progenitors of the Tibetan people, a monkey and a rock ogress, are identified with Avalokiteśvara and Tārā, as discussed above by Davidson. Therefore, the Tibetan people themselves are believed to be the children of Avalokiteśvara, who is the bodhisattva of compassion, in contrast to Vairocana, the Bodhisattva of royalty. Likewise, the cult focus shifts from Samyé to Lhasa; while the former was the religious center and the latter the administrative center during the imperial period, Lhasa came to strongly incorporate both elements with the revivification of religious activity at the Jokhang during the renaissance. Now the laity as well as the monastics could engage with the divine, an encounter reserved only for the elite at Samyé in the eighth and ninth centuries. Such a compassionate divinity is certainly more popular in a time when royalty is non-existent and turmoil is everywhere. Nonetheless, Avalokiteśvara takes on royal attributes as the source from which the bodhisattva king Songtsen Gampo emanates; through Avalokiteśvara's compassion he is believed to take active agency in Tibet whenever there is social upheaval and a need for order. These elements all tie into the grander social project wherein Tibet is envisioned to be his Buddha field of activity, created out of the bodhisattva’s compassion and destined to be a great realm of Buddhist Truth. The king may be gone, but the kingdom continues through the active participation of this new bodhisattva and his cult, which, along with Tārā’s, has come to be the most popular and enduring in Tibetan history.16

This shift from the cult of Vairocana to that of Avalokiteśvara also correlates with the process of social conversion that Kapstein explores in this era:

[W]hen we think of conversion, it is individual conversion that we have in mind. Following James, we sometimes think of this as a sudden and dramatic reorientation of consciousness, marked by profound changes of sentiment and of faith. By contrast…when it is conversion of a nation that is at issue, the gradual transformation of cosmological frameworks, of ritual, intellectual, and bureaucratic practices, and of the historical and mythic narratives through which the national identity is constituted are among the key themes to which we must attend. Moreover, in the case of Tibet, it is now clear that there were in fact two conversions that can perhaps be roughly correlated, but by no means identified, with the so-called earlier and later propagations of the Buddha’s teaching in that land. In the first instance, there was the imperial adoption of Buddhism, which corresponded with the expansion of the old Tibetan empire, the formation within it of a literate administration, and the need to represent Tibetan imperial power both within and beyond the frontiers of Tibet. In the second, there was a conversion of the conversion narrative itself, ensuring that the Buddhist conquest of Tibet would endure long after the conquered empire had vanished.17

To summarize, the first national conversion was the imperial adoption of Buddhism, where Buddhism was used to validate the empire, as embodied by the cult of Vairocana and the maṇḍala of Samyé. The second conversion was the transformation of the first conversion narrative to Buddhist concerns, where the utopian vision of the empire was instead used to validate Buddhism, as embodied by the cult of Avalokiteśvara and the maṇḍala of Tibet.

Demonic Hegemonies

The Divine Hierarchy

At this point, our concern has primarily been with bodhisattvas, with a brief look at local gods and demons. Since we are now moving away from Buddhist bodhisattvas and shifting our focus to converted local gods and hybrid deities, it would be helpful to examine the distinctions between the various types of divinities and exactly how they can be classified before proceeding. Geoffrey Samuel’s classification scheme, discussed below, is most helpful toward this end because it incorporates the two-part system explored in René de Nebesky-Wojkowitz’s work with the taxonomy of deities provided by Giuseppe Tucci in his monolithic opus, Tibetan Painted Scrolls.18

This scheme involves four major divisions. First, there are the symbolic tutelary deities (Tib. yi dam; Skt. iṣṭadevatā) and bodhisattvas of Vajrayāna; these deities are encountered at the highest levels of Buddhist monastic ritual practice. Second are the transcendental deities (’jig rten las ’das pa’i srung ma), who consist of the heavenly gods; such beings are not concerned with mundane worldly affairs. Third are the worldly deities (’jig rten pa’i srung ma) who inhabit the intermediate spaces, are associated with geographical features like mountains, lakes, and forests, and are subject to the laws of karma; these deities constantly interact with humans. Fourth, there are the numerous malevolent spirits and ghosts who constantly bring harm to human beings through illness, poor fortunate, and calamity.

As Samuel explains, the last two categories of worldly deities and malevolent spirits are not wholly distinct, there is a degree of fluidity between them.19 However, a common distinction that is made is that worldly deities were malevolent spirits that have since become tamed and who now serve the Buddhist teachings. In turn, those deities still classified as malign are placed within the retinues of worldly deities. This fluidity exists in all four divisions and there appears to be a degree of mobility between them.

A further “demonic taxonomy” exists that overlaps the above scheme, indicating several different species of ambivalent spirits. The term “demon” has become popular in referring to these deities, given their initial penchant for pernicious activity; de Nebesky-Wojkowitz’s own book title illustrates this usage. This is unfortunate given the stigma and strong malicious nature associated with the word in the West. The term also suggests the linguistic difficulty inherent in translating these various spirit classes into English. While English has one overarching term, “demon,” there are several kinds that exist in Tibet that are indicative of vastly different attributes and qualities both beneficent and malevolent. Nonetheless, because most of these beings are in the habit of being angry, violent, and harmful, I have decided to continue the convention of the term. However, I have also developed a system for demon classification that assigns equivalent English terms in order to recognize their etymological distinctions. Below I have provided some of the more important demon classes commonly encountered in Tibetan culture. This system is imperfect, but it is nonetheless an attempt at remedying the inadequacy of past translations:20

  1. Violence Demons (Tib. gnod sbyin; Skt. yakṣa): These deities were initially malicious beings that caused disease epidemics. As a way to signify their later conversion to guardians of the Buddhist teachings, they have come to be associated with the ancient Indian spirits called yakṣas, beings popularly found in Hinduism, Jainism, and Indian Buddhism.
  2. Obstacle Demons (Tib. bdud; Skt. māra): Obstacle demons are openly malevolent spirits endowed with the nature of their namesake. They create obstacles, usually to prevent the successful completion of rituals or to prevent success on the path to enlightenment. They were opposed to the Buddhist teachings in their past lives, are said to be black in color, and eat human flesh. The Sanskrit word associated with them refers to the kind of behavior exhibited by the god Māra, the personification of saṃsāra, who attempted—and ultimately failed—to hinder the enlightenment of the historical Buddha Śākyamuni.
  3. Flesh-eating Demons (Tib. srin po; Skt. rākṣasa): As with the obstacle demons, the most common attribute of flesh-eating demons is implied by their name. They are associated with the Indian demons called rākṣasa, the most popular demons encountered in Indian epics and lore.
  4. Serpent Demons (Tib. klu; Skt. nāga): These entities are serpentine deities who generally abide in lakes, rivers, and subterranean realms. They are known to pollute water and hinder the construction of dykes, irrigation works, and temples. If angered, they can cause diseases such as leprosy. When assigned as protectors of the Dharma, they are commonly employed as guardians of treasure texts (gter ma).21 Their name implies the shape they commonly hold, though iconographic representations also depict serpent deities with human upper bodies and a snake tail instead of legs.
  5. Mighty Demons (Tib. btsan): These are indigenous Tibetan deities who are known to be war-like and wrathful. They are red in color and inhabit mountains and rocks. They are generally believed to be the spirits of past monks who have rejected their Buddhist vows. Once they are tamed, mighty demons are assigned as the protectors of temples, sanctuaries, and monasteries.22
  6. Conqueror Demons (Tib. rgyal po): Conqueror demons are the spirits of evil kings or high lamas who have broken their vows. They are generally white in color.
  7. Violator Demons (Tib. dam sri): These are spirits who have violated their vows in past lives and who can inspire individuals to do the same. They are part of a larger class of ambiguous demons called sri.
  8. Sky Demons (Tib. the’u rang): These are ancient Tibetan deities of ambiguous nature but who are primarily associated with the sky. They are also harmful, causing death and disease. Erik Ḥaarh explains that these deities were originally the spirits of Pre-Buddhist ancestors found in the heavenly spheres.23

Given the overlapping natures of these demons, many of these distinct class titles have been used synonymously with each other. Individual demons can also be associated with multiple categories. For instance, the demon Tsiu Marpo (Tsi’u dmar po), who will be discussed below, traditionally belongs to the mighty demon (btsan) class but is considered the king of the violence demons (gnod sbyin); many of his appellations also refer to him as a violence demon. Also, despite the numerous Sanskrit terms associated with each demon, they are not necessarily indicative of the origin of these deities in India. Most of these deities are indigenous and have been later assimilated into Indian tantric classification systems, yet they still retain many of their Tibetan attributes. However, it certainly appears that the Tibetan divinity system has mimicked if not inherited numerous classes of demonic entities found in the Indian system.24 In any event, these classification schemes are quite arbitrary and seem to have been later creations in an attempt to order the startling complexity and heterogeneity of the numerous divinities encountered in the Tibetan religious arena. Nonetheless, it is important to be aware of this larger divine community as a foundation for discussing the careers of two specific demons important in the political clime of sixteenth- and seventeenth-century Tibet.

Tsiu Marpo and the Sixteenth Century

While bodhisattvas, at least in the examples we have explored, tend to be involved in grand nation-building processes, the lesser gods usually begin from and remain within much more local venues. These deities and their cults generally have their starts within families as clan protectors or progenitors, and continue to be tied to local monasteries and institutions once subdued and converted into protectors of the Buddhist teachings (chos skyong).25 However, despite the more regional character of these gods and demons, a handful take on greater celebrity by being involved with clans and institutions that grow to greater political prominence. In this regard, two figures come to mind, both of whom are intricately involved with one another: Tsiu Marpo and Pehar (Pe har).

Both Tsiu Marpo and Pehar have acted, at various moments in their careers as protector deities, as the main guardian of Samyé monastery. Pehar was there first, and Tsiu Marpo took over the position after the former left. Popular myth states that both were subjugated by Padmasambhava in the eighth century and assigned as guardians of Samyé. However, the earliest textual mention of Tsiu Marpo, as Todd Gibson explains, is in a late-fifteenth-century treasure text discovered by Padma Lingpa (Padma gling pa; 1450-1521).26 Gibson goes on to point out that “neither the twelfth-century chos ’byung of Nyang-ral Nyima ’Od-zer, nor the Sba bzhed (date uncertain) mentions any name similar to his among the many chos skyong who were appointed to guard various parts of Bsam yas, although in both lists numerous gnod sbyin occur.27 This strongly indicates that despite the mythic rhetoric of subjugation Tsiu Marpo did not exist, at least as an assigned protector of Samyé, until much later.

The earliest text to be focused exclusively on Tsiu Marpo is his root tantra and ritual manual the Warlord’s Tantra with Accompanying Sādhanas (dmag dpon gyi rgyud sgrub thabs dang bcas pa bzhugs so), found in The Great Treasury of Termas (Rin chen gter mdzod chen mo), and written by the Nyingma master Ngari Penchen Padma Wangyel (Mnga’ ris paṇ chen Padma dbang rgyal; 1487-1542).28 This text details Tsiu Marpo’s mythic origins and subjugation as well as the ritual manuals used to propitiate him. It is in the sixteenth century, then, that Tsiu Marpo not only gets his start as a distinct deity, but also becomes the focus of a particular ritual dance to be discussed shortly.

It should be noted that in contrast to the primarily peaceful bodhisattvas, protector deities like Tsiu Marpo are generally characterized as being ferocious and blood-thirsty.29 They serve the much more violent and pragmatic purpose of attacking the enemies of Buddhism; in specific ritual instances this usually means the enemy of an individual lay patron or a particular monastic institution, which in many cases can mean another monastic institution. An example of such a request is as follows:

You, the yakṣa, king of the btsan, residing in the middle of blazing chain-lightnings, who open wide your horrifying mouth and stare with red eyes—together with your hordes, come surging forward, soar quickly atop of the enemies, send the painful illness of the btsan into the upper part of the bodies of your foes, and cast colic into the lower part of their bodies. Having gathered big black clouds on the sky, cast thunder, lightning, thunderbolts, and great pieces of hail, order the the’u rang demons to cause fits of madness, and command the ma mo [demonesses] to cause fits of fainting to your foes. Bind tightly your enemies by means of your red snare, strike them hard with the great stick. With the flaming sword, cut them into huge pieces. With the flaming iron hook, tear out their heart-veins; having chewed the bodies of the foes with your flaming teeth, swallow them down into your stomach which is like a ‘rākṣasa-hole’. Send various kinds of illusions to deceive your enemies. Send suddenly the illnesses called gag nad and gzer nad. Destroy seven generations of the multitude of obstacle-creating demons and enemies.30

This description overflows with the kind of violent imagery commonly found in texts concerning protector deities. While such beings have been converted to Buddhism, their vicious natures have not been wholly expunged. They have to be constantly reminded of their vows to protect the Buddhist teachings and they are given baser offerings, such as beer, in the course of their propitiation. Yet it’s this violent nature that is nonetheless offensively utilized against one’s enemies.

Tsiu Marpo makes his debut in Tibetan ritual history first with his cult being liturgically established by Ngari Penchen, then ritually established in the dance performance (’cham) called Offering to the Basket of the Sūtra (mdo sde mchod pa). This ritual masked dance was established by the Sakya lama Künga Rinchen (Kun dga’ rin chen; 1517-1584) in 1570 and its central deities were Tsiu Marpo and Pehar.31 Pehar may be involved with this ritual, but mythically he has left Samyé at this point; deities may be primarily associated with one particular monastery or location, but their presence tends to be omnipresent, being available for ritual engagement wherever their statues, paintings, or ritual texts are active. Tsiu Marpo comes to the fore as the protector of Samyé for both the Nyingma and Sakyapa in the sixteenth century, the latter of which have powerful political ties at this time:

The princes of Rin-sprungs [Rinpungpa], masters of a large part of Tibet until about 1565, and the princes of Gtsang [Tsang] who progressively eliminated them from 1565 to 1642, had for chaplains the Bka’-rgyud-pa, and above all the Karmapa, the Jo-nang-pa, and the Sakyapa, like Kun-dga’ rin-chen, who was initially the religious master of the Rin-sprungs-pa, and later the princes of Gtsang. Until the seizure of power by the Dalai Lamas, it had thus not been Pehar who served as a model for the other chos-skyong, but the gnod-sbyin, the btsan, the eater of breath, flesh, and blood, Tsi’u dmar-po of Samye. Along with the Gtsug-lag-khang of Lhasa [the Jokhang], Samye is the oldest religious center and the most venerated in Tibet. It is Tsi’u dmar-po—who became the protector of Samye after Pehar’s transfer to ’Tshal and from ’Tshal to Gnas-chung—that the other chos-skyong model themselves after, including Pehar—who Tsi’u dmar-po eclipsed after having succeeded him.32

It is clear that in the sixteenth century, before the political rise of the Gelukpas, the Sakya and Kagyü were powerful sects with ties to the hegemonic rulers of the day. As such, the Sakya control of Samyé at this time coincides with the creation and rise of Tsiu Marpo’s cult. Ritual performances like the Offering to the Basket of the Sūtra, headed by Tsiu Marpo, has greater political overtones, wherein it represents Sakya power and their status as the religious purveyors of the Rinpungpa and Tsang administrations. Through the centuries, Samyé continues to embody strong religious symbolism, since it is the first Buddhist monastery of Tibet and it was founded during the great Tibetan dynasty. That Tsiu Marpo would come to be its protector, and that his cult would be bound up with the religious and political powers in the sixteenth century, marks the height of this demon’s influence, at least in central Tibet; but it was not to last.

Pehar and the Seventeenth Century

Pehar is mythically much older than Tsiu Marpo. Ariane Macdonald attributes his creation to the famous twelfth-century treasure revealer (gter ston) Nyangrel Nyima Özer (Nyang ral Nyi ma ’od zer; 1136-1204),33 and a great body of myth illustrates his subjugation by Padmasambhava and assigning to Samyé monastery as its protector—specifically as the protector of its riches. Pehar is also far more engaged with Tibetan history than Tsiu Marpo. After spending seven centuries at Samyé, he moved to the Nyingma monastery of Tselgungtang (Tshal gung thang) on the bank of the Kyichu (Skyid chu) River not far from Lhasa sometime in the sixteenth century.34 Having spent almost a century at this location, Pehar began to fight with a high lama of the monastery named Zhangtselwa (Zhang tshal ba). When a new monastery was being built, Zhangtselwa demanded that the artisans and painters create no images of Pehar. This angered Pehar and he retaliated by burning the monastery down. In response, the infuriated Zhangtselwa subdued Pehar through ritual means and imprisoned him in a thread-cross (mdos). A thread-cross is a ritual implement made of crossed sticks surrounded by a web of multicolored thread. Thread-cross rituals draw demons or deities toward the cross and trap them within the web.35 Once Pehar was trapped within the thread-cross, Zhangtselwa locked it in a box and threw the box into the Kyichu River. As the box was floating down the river, it was spotted by the fifth Dalai Lama, who happened to be at Drepung (’Bras spung) monastery that day. He ordered an abbot of one of the monastic colleges to retrieve the box but not to open it. The abbot went and fetched the box, but as he was walking back to Drepung the box grew heavier until he could not hold it any longer. The abbot became curious and opened the box, at which time a white bird flew from it and landed on a nearby tree.36 The Dalai Lama ordered that a shrine be built around this tree for Pehar and eventually the monastery of Nechung (Gnas chung)—literally “small place”—grew up around it. To this day, the tree on which Pehar landed as a bird is viewable within the central shrine room. Pehar then began to possess the body of a monk at Nechung, and this oracle lineage was formally established as a state-recognized office by the fifth Dalai Lama.37

The historical connection between protector deities overall and the fifth Dalai Lama is worthy of further research. Considering the Nyingma sympathies and the ecumenical approach of the fifth Dalai Lama, it is no doubt through his efforts that many deity cults grew in popularity. Not only did oracle lineages—such as Nechung and, quite possibly, the Samyé oracle of Tsiu Marpo—become state sanctioned institutions during his reign, but he was also responsible for the composition of many ritual texts dedicated to protector deities, texts that he either composed personally or commissioned.38

Dan Martin’s research also shows that Pehar has some degree of involvement in eleventh- and twelfth-century lay religious movements, discussing specifically a group of popular lay practitioners called “the four children of Pehar.” These figures were known to practice magic and subvert the Buddhist teachings; some were speculated to be incarnations of various demonic deities. As Martin explains, the identities of these figures are, in most instances, obscured by the excessively negative nature of the accounts that discuss them, which were primarily written by Buddhist monastics. Despite the excessive discussion of Pehar, with whom these individuals are associated, Martin speculates in his conclusions that Pehar may not have had any connection at all with these figures initially. Rather, he may have come to be associated with them anachronistically in the works of thirteenth-century writers who disparaged such movements and made a connection with Pehar out of their distaste for this trickster-like deity.39

As Macdonald illustrates, with the rise of the Gelukpa to power in the seventeenth century, Tsiu Marpo, the deity whose cult was propagated by practically all of the Gelukpa’s enemies, falls from centrality and Pehar comes to the fore. Not only does Pehar have a mythic history connected with Samyé that’s older than Tsiu Marpo’s, but he has migrated from Samyé to Drepung in Lhasa, the very heart of the fifth Dalai Lama’s new administration. This all ties in well with the obsession that the fifth Dalai Lama and his government had with antiquity as a validating force. Pehar, being the first protector of Samyé, associates the Gelukpa government with the era of Padmasambhava and the founding of Samyé. Furthermore, the fifth Dalai Lama comes to be seen as an emanation of Avalokiteśvara. This quality is retroactively attached to his previous incarnations and allows his new kingdom to tap into the narratives surrounding King Songtsen Gampo and the powerful cult of Avalokiteśvara discussed above, which has continued to be popular and active in this era. No doubt the motivating force behind connecting the nascent Gelukpa government to the past was in an effort to authenticate its existence and symbolically compare it to the now idealized legend of the once great Tibetan dynasty.

Furthermore, while Tsiu Marpo was involved with one of a few state-sanctioned oracle traditions in Tibet, the state oracle of Nechung, an emanation of Pehar, far eclipses him in importance, the latter being the primary oracle consulted by the Dalai Lama’s government from this time onward. The oracle signifies a much more dynamic role that deities can have within Tibetan history and society. The direct communion between the god speaking through the oracle and the audience that attends such sessions shows a palpable relationship between the divine and the members of the community who engage in this process. Specifically, a Tibetan oracle is a member of a community, or in this instance a political institution, who becomes possessed by a specific deity in order to provide communal advice or prophetic pronouncements. Oracles are further expected to resolve conflicts between monastic communities as well as between individual patrons of the lay community.40

Here we have an explicit hegemonic shift in the divine dominion. Tsiu Marpo was the main connection to Samyé, a conduit to the past, which the various sectarian powers in Tsang harnessed for their political consolidation of power in the sixteenth century; however, his importance was eclipsed in the next century by Pehar as part of a very successful process of ancient authentication utilized by the fifth Dalai Lama’s new government to justify its existence. This turning back to past power and glory is a constant leitmotif found throughout the Tibetan cultural and historical landscape.41

The Charisma of Cults

By way of a final exploration of deity cults in Tibetan political history, it should be emphasized that charisma plays a large part in the development of most if not all cults. With the indigenous deities first converted to Buddhism, it is the figure of Padmasambhava who is central. With the bodhisattvas Vairocana and Avalokiteśvara, it is the legendary Buddhist kings of Trisong Detsen and Sontsen Gampo who root these divinities in the world. Furthermore, we know that Atīśa, as well as Kadampa figures like the three brothers, were instrumental in popularizing the cult of Avalokiteśvara to an incredibly successful degree. With Tsiu Marpo and Pehar, their cults began simply with their liturgical propagation by such figures as Nyangrel Nyima Özer and Ngari Penchen, as well as Künga Rinchen, before being later implimented for political solidarity by administrative officials and even the new Tibetan government. In these latter cases, oracles too represent arenas of charisma, given the concrete engagement with the divine that they provide. In all these instances, the Weberian ideal occurs quite often, where such charismatic and personalized beginnings are routinized into institutional forms and maintained by monastic and sectarian lineages.

A relatively modern example of this surrounds the cult of Dorjé Shukden (Rdo rje shugs ldan). Dorjé Shukden is considered by many today to be a controversial deity. This is in part due to the contemporary conflict between the fourteenth Dalai Lama and the New Kadampa Movement, a group that specifically endorses the cult of Dorjé Shukden. There’s a complex history, both old and new, behind this, but mainly it involves the uncomfortable relationship that Shukden has had with the lineage of the Dalai Lamas. The common account is that Dorjé Shukden is a wrathful spirit who was a rival of the fifth Dalai Lama in his past life. This rivalry in part surrounds concerns for orthodoxy; whereas the fifth Dalai—and the fourteenth Dalai Lama currently—was traditionally perceived as ecumenical and particularly conciliatory with the Nyingma’s, Shukden represents an exclusivistic movement within the Gelukpa wherein all non-Gelukpa’s are considered perverters of the Buddhist teachings, and Gelukpa accommodation to the other sects is just as bad. This continues to be the purview of the New Kadampa Movement today. Dorjé Shukden thus upholds these ideals by attacking the enemies of his cult, which seem to be more often Gelukpas friendly toward the other sects rather than members of those sects.

The popularity of this cult, however, is less than a century old. Its growth is the result of a monastic figure who began a revivalist movement at the turn of the century that propagated these ideals as well as the deity at their center. This figure is Pabongkhapa Dechen Nyingpo (Pha bong kha pa Bde chen snying po; 1878-1941), whose charismatic power and oratory prowess are especially noted in the accounts of his activities. Georges Dreyfus notes these qualities, as well as his promulgation of Shukden’s practice, in his analysis of this figure:

In promoting Shuk-den as the protector of his charismatic movement, Pa-bong-ka did not invent the practice of this deity, which he received from Ta-bu Pe-ma Baz-ra, but he transformed a marginal practice into a central element of the Ge-luk tradition. This transformation is illustrated by the epithets used to refer to Shuk-den. Instead of being just “The Spirit from Döl” (dol rgyal), or even “The Great Magical Spirit Endowed with the Adamantine Force” (rgyal chen rdo rje śugs ldan rtsal), he is described now by Pa-bong-ka and his disciples as “The Protector of the Tradition of the Victorious Lord Mañjuśrī (i.e., Dzong-ka-ba)” (’jam mgon rgyal ba’i bstan sruṅ) and “The Supreme Protective Deity of the Ge-den (i.e., Ge-luk) Tradition” (dge ldan bstan bsruṅ ba’i lha mchog).42

There are two notable features here. The first involves the promotion of the movement as an active and even violent protection of the Geluk sectarian identity, which Dreyfus explores in further detail. The second feature, which is more important for our purposes, is how a little known deity practice can grow in prominence and political notoriety through the efforts of one charismatic individual. Such efforts did not stop with Pabongkha, however, as his disciple Trijang Rinpoche (Khri byang Rin po che; 1901-1983) was equally charismatic and “exercised considerable influence over the Lhasa higher classes and the monastic elites of the three main Ge-luk monasteries around Lhasa.”43 Regardless of how charismatic a leader is, the lineage and practices they propagate will not continue to flourish without the continued maintenance of equally effective disciples. Indeed, it seems that before the liturgies, oracles, and state endorsements, charismatic and devoted individuals like Pabongkha are required to give deity cults their start; through a combination of rigorous transmission and political opportunity they then can be poised to burgeon into a new divine hegemony either replacing a previous one or, as in this case, coexisting uncomfortably with the current rule. Despite their grand myths of subjugation, legendary agency in history, and roles in shifting political institutions, gods have to start somewhere.


These complicated alliances between divine and human agents are similar to those forged between human institutions. Indeed, this parallel is made all the stronger by illustrating the social models found in the divine and demonic communities, which understandably mimic familiar worldly forms. Great bodhisattvas tend to be contrasted with kings, a common motif found among buddhas and bodhisattvas overall. Such kings are at the center of their maṇḍala, ruling over their Buddhist kingdom populated by other bodhisattvas and lesser gods. In this case, propitiating or otherwise engaging with these world-producers and world-rulers is akin to fostering a relationship with a worldly king, and the process uses very similar ritual structures.44 Beyond this, the powerful deities who populate this new maṇḍalic cosmology are also commonly approached in establishing allegiances. Given the fierce nature of these deities, many of them are given military identities, such as generals and soldiers—recall that Tsiu Marpo’s root tantra is called the Warlord’s Tantra. As such, they represent a very real military presence to Tibetans, who can be coaxed to defend monasteries and administrations and attack enemies. With these demons brought under the banner of Buddhism, their abilities can be used toward its maintenance and propagation. Thus, given the cosmological framework made available in Tibet with the advent of Buddhism, it is reasonable to assume that part of institution-building involves developing a relationship with preexisting powerful and divine agents and institutions, which, it must be stated, Tibetans accept as a reality. Such relationships must also begin somewhere, and it appears that it is through the activities and ministrations of charismatic individuals—who are working with preexisting templates of divinity within the milieu of the Tibetan worldview—that deity cults even begin and are able to flourish. Once available, these cults can then be utilized to propagate the ideologies and enforce the political endeavors of those institutions to which these charismatic individuals are tied. Through such efforts, deities can become inextricably bound to political institutions, rising and falling with the tide of hegemonic (in)stability. It should also be noted that all of these cults still exist today, in various degress of influence and transmission; regardless of the current political situation, Tibet never seems to completely discard what was once a useful source of religious identity and cultural cohesion. Ultimately, the history of Tibetan hegemonies can shed light as much on the popularity and transmission of deity cults as the latter can provide insight into the political situation in any given era. Tibetan deity cults are powerful forces and indicators of historical change, acting as political barometers from the earliest historical period of Tibet to the present.


My two primary sources regarding the political significance of Tsiu Marpo and Pehar in the sixteenth and seventeenth centuries are two French articles written by Ariane Macdonald (Macdonald 1978a and 1978b). I provide their translations here.

"Political Rivalries and Religious Centers at Samyé Monastery in the 16th Century:
The Spiritual Lineage of the 5th Dalai Lama in Literature, History, and Art.
by Ariane Macdonald

Translated by Kathryn Fuss and Christopher Bell

The return of an audience has changed the perspectives and themes approached in the conferences this year. They were centered on the places that link oral tradition and the religious past of ancient Tibet. Indeed, it is necessary to reiterate the legendary account of the origin of the pre-Buddhist god Bar-tshigs brag-btsan, and its development in the festival of the Mdo-sde mchod-pa, based on research, such as what was provided to us by Lcog-rten Ta lama in Mysore, from 1975-76 (see the Directories of the past two years). However, we analyzed it by systematically confronting each of the details of six drawings in the Wise collection which refer to the festival—though Lcog-rten Ta lama considered them very inaccurate, as reported in his account—as well as the photographs of Samye, which were used to support the reconstruction of the performed dances in the courtyard of the temple of the protector gods of Samye. These dances, which took place on the 13th, 14th, and 16th of the fifth month, were represented, in an incomplete form and with some errors, in two drawings that bear the classifications 3021 and 3025, although there is nothing specific regarding the location and date in the legends of the Wise series. In addition, only two of the three events that mark the day appear there, none are represented as a whole and separately in these drawings. Those which carry the classifications 3020 and 3022 to 3024 refer—with some inaccuracies reported in a sometimes contradictory manner by Lcog-rten Ta lama and Gnas-chung rinpoche—to the masked dances dedicated to Padmasambhava who, according to the Wise legends, stood before the principal temple of Samye on the 15th of the fifth month, and according to our informers, in reality the 10th of the first month. One of three photographs was taken at Samye, in front of the temple of the protector gods on the 16th of the fifth month, at the moment when the procession, which constitutes the third event of the day and which is of crucial importance, leaves the temple. In addition, with the indications provided by Lcog-rten Ta lama, we have connected the account of the visit of the 13th Dalai Lama to Samye in 1900, which contains some bits of information on the Mdo-sde mchod-pa dances, drawn from his biography which, to the eyes of the Tibetans themselves, was not written with great clarity. Although, the confluence of this disparate material—the details of which will be given in a pending publication and supporting documents—emphasized some of the hypotheses on the method to use and the ways to follow, only the account provided by oral tradition gives a continuation and a direction to the contemporary photographic, iconographic, and written documentation. Not to stress the fact, but significantly Bar-tshigs-brag-btsan doesn’t appear there. However, it is completely exceptional that the existence and autonomy of a pre-Buddhist god had been preserved at Samye and that we know it. In some special cases, only one god plays the role of protector of the Law and god of the soil, subjugated by Buddhism.

This observation made us understand the importance of oral tradition and its possible use to attempt—with the data acquired through research carried out on the pre-Buddhist religion of Tibet after reading some of the Touen-houang documents (8th-9th centuries) and those of the past two years—to reconstruct the components of the festival in the time of the Tibetan monarchy and their significance in the religious context of ancient Tibet. It thus seemed important to us, on the one hand, to emphasize the gaps of our oral information, and on the other hand, to research, under the terms used to render some of the Buddhist concepts, the form and direction that they had before the introduction of Buddhism to Tibet. We started from the principle that Buddhists have used, substituting Buddhist notions for indigenous religious concepts, the notions and key terms of this religion, by deforming them and appropriating their sense. Here are some attached examples that use the words Mdo/Mdos, Ri-rab/Meru, ’khrungs-lha/sku-bla, taken from among others, which control the general interpretation of the festival.

The research that we have carried out has established several points on the origin and development of the mythology and the repeated worship of Pehar and Tsi’u dmar-po, who themselves succeeded as Buddhist guardians of the temple of riches at Samye—the only temple named Dkor-mdzod gling, Pehar lcog, or Tsi’u dmar lcog (a mysteriously applied term, principally for temples of protector gods). Permit us to point out that which appears essential: these points were neither known, nor allowed, because they are contrary to fixed Tibetan tradition, to government echelons, under its definitive form, in the 18th century. The legends that attach themselves to the origins of Pehar (there are many), from his nomination as Protector of the temple of riches at Samye by Padmasambhava, to the end of the 8th century, and even his name, Pehar, did not exist prior to the second half of the 7th century. Their inventor, the ‘treasure discoverer’ (gter-ston) Myang-ral Nyi-ma ’od-zer (1124-1136—1192-1294), was the first reincarnation of King Khri-srong lde-bstan. We will return to this essential feature in time regarding the motivations for the creation of Pehar and the theory of reincarnations that make Tibetan Buddhism so distinctive. Myang-ral is also a former incarnation of the Fifth Dalai-lama, author of a commentary on a ritual of Pehar and his army leaders, ru-’dren, which form his following, uncovered by this very same gter-ston. This has been used in the liturgy of the god since he became the state oracle at Gnas-chung in 1642. And at Samye, since around 1570, Tsi’u dmar-po, incarnated in his medium, demanded Sa-skya-pa Kun-dge’ rin-chen to institute, at the time of the Mdo-sde, a “dance of the four army leaders” (ru-’dren sde-bzhi’i gar-’cham). The dance of the “seven blazing brothers,” ’Bar-ba spun-bdun, of which Tsi’u dmar-po is the chief, and the dance of the army leaders of Pehar (ru-’dren gyi ’cham)—which were performed in the courtyard of the temple (lcog) of these two gods at Samye on the 13th, 14th, and 16th of the 5th month—cannot however under this form have been part of the celebrations of the festival in the 8th century or at the beginning of the 9th century, the date of its foundation according to later tradition.

However, Myang-ral isn’t only the inventor of a ritual dedicated to a Buddhist Pehar, preserved in volume ti of the Rin-chen gter-mdzod; he is also the inventor of a ritual completely dissociated from the first, preserved in volume pi of the same collection, devoted to the construction and offering of a Mdos to the rgyal-po pre-Buddhist god Pe-dkar or Pe-kar, which, we have remarked, is related to Pehar among others, and which is from this name. The reading of the ritual of Myang-ral, entitled Rgyal-po dkar-po drug-mdos, is always in force at Samye and Gnas-chung, but we are unaware of which time of the year is it; this shows that the bonds between Pe-kar and Pehar are much deeper than we thought. To simplify in the extreme, let us say that the Mdos, the pre-Buddhist character of which is admitted by all, is a construction, stable yet temporary (that is the case for what is written here) in multiple stages, surmounted with a wire cross. According to Myang-ral’s text, it is the palace in which the guardian of riches (dkor-srung), the king of Samye, Pe-dkar, resides. It is built on an “excellent mountain,” ri-rab, a term that translates to Meru in Buddhist literature. It is the central cosmic mountain in which tradition assimilates the main temple of Samye, which would have been constructed from a cosmological model. Myang-ral describes the castle of Pe-dkar as a bastion with three floors flanked by four turrets at an angle, lcog. King Pe-dkar resides there with his ministers, his army leaders, and his spouses, represented by dough at the base of the Mdos. They carry the name of the two queens of King Khri-srong lde-btsan. The offering of the mdos to Pe-dkar, the king master of life (srog-bdag rgyal-po Pe-dkar), is a ritual of “ransoming the life,” srog-glud. The offering of the Mdos, the palace of the demon rgyal-po of Samye, identified with the palace of the king, the king’s spouses, his ministers, and his army leaders, is for preserving the life of the king; this encourages one to reflect on the significance of the Mdo-sde mchod-pa in the royal era, and on the same name of the festival. This “offering to the basket of the sūtra” (mdo is the translation of this Sanskrit term) is indeed included in a group of three of four mchod-pa, festivals dedicated to three or four categories of Buddhist writing, by several Tibetan informants and by Tibetan literature dating to the 14th century. All of these festivals have in common the propitiation of the master of the soil, and their association with some Buddhist texts seems artificial. On the other hand, in each of the six drawings devoted to the festival of Samye in Wise the spelling is Mdos-sde and not Mdo-sde. Our dge-lugs-pa informants consider it very much an error, but a master rnying-ma-pa assured us that it was a Mdos that was at the center of the festival, although none are represented in the drawings (Bar-tshigs brag-btsan is also absent). It seemed indispensable to ask the witnesses of the festival some questions on the presence of a mdos at the time of the Mdo-sde mchod-pa. The reading of Myang-ral’s ritual forces us to revise the architectural and symbolic design of the temple of Samye, and we have devoted several conferences to the question.

We have also sought to specify the relationships which, before the introduction of Buddhism to Tibet, could contrast the ’khrungs-lha Bar-tshigs brag-btsan with some divinities dangerous to the king, like the rgyal-po (meaning ‘king’) Pe-dkar.

However, just as the rituals dedicated to Pe-dkar and to Pehar have a single inventor, the gter-ston who likewise fixed the liturgy of Tsi’u dmar-po in the 15th century, Mnga’-ris pan-chen—also the reincarnation of Khri-srong lde-btsan and former incarnation of the fifth Dalai Lama—evoked within a short cryptic text, the title of which is ambiguous, the origins of Bar-tshigs brag-btsan. Described as the chief of a group of “seven blazing brothers,” it is certain that he served as a model for the creation of the mythology of Tsi’u dmar-po, who belongs to the same group and, at different moments of the ceremonies of the Mdo-sde mchod-pa, who is confused with or opposed to the indigenous god. The really essential fact is that this ritual dedicated to Bar-tshigs brag-btsan defines him in the colophon not as a ’khrungs-lha—a term which was applied to him by Lcog-rten Ta lama and the author of the Guide to the Restorations of Samye in 1854, p.212 (chos-rgyal gyi ’khrungs-lha Bar-tshigs brag-btsan)—but as the “sku-lha of the king” (rje’i sku-lha). The hypothesis that we had formulated, in another context, on the similarity established between the ’khrungs-lha dbyar-gsol and the sku-bla dbyar-gsol—a pre-Buddhist festival dedicated to the god who is the “depositor of the king’s heart,”—appeared to us to find confirmation here. The most significant moment, if one searches for the most ancient remnants of the festival, would then be the third event, which closes the ceremonies on the 16th day, “the procession of the rgyal-po surrounded by a crowd of his subjects,” a loose translation of the expression rgyal-po khrom-phebs. This is also glossed by Lcog-rten Ta lama: “The king (rgyal-po) is the god Pehar, who circles the crowd (khrom) of the Tibetan gods of the soil, the lha-’dre, who are all reunited at Samye that day.” This definition is completely parallel to that which the same informant gave us for the festival of ’Dzam-gling spyi-bsang, which takes place on the 15th day of the 5th month in Lhasa, during which the Tibetan gods of the soil take possession of their mediums. A bsang (fumigation ritual) is offered to the lha-’dre all over Tibet, the chief of which is the king of Gnas-chung, Pehar. This is, another informant said, to commemorate the subjugation of the lha-’dre by Padmasambhava at Bsam-yas, which is celebrated on ’Dzam-gling spyi-bsang in Lhasa.

It would thus seem that the rgyal-po khrom-phebs, the precedent on which the festival taking place in Lhasa was modeled—originally the 15th day of the 5th month—was a date marked at Samye by a masked dance performed at the main temple. Information on its name is very contradictory but, in the oldest text that mentions it (15th century), it is considered a dance “destined to subjugate the gods of the soil,” sa-’dul gyi ’cham.

However, in the procession that circles the temple of Samye, it is not Pehar incarnated in the medium of Gnas-chung, who enters into trances, as in Lhasa. A great mannequin of Pehar, preserved in one of the chapels of the temple Pehar lcog at Samye, is carried at the head of the procession. It is followed by the medium of Bar-tshigs brag-btsan and his serving horses. We are unaware of whether he is accompanied or not by local gods; it is one of the important gaps in our information. However, that the pre-Buddhist god takes the lead of the gods of the soil subjugated by Padmasambhava—of which Pehar is considered the representative—seems important. One could deduce that one of the central elements of the pre-Buddhist sku-bla gsol-ba at Samye consisted of a procession directed by the sku-bla Bar-tshigs brag-btsan, imposing his authority on the other pre-Budhhist divinities, who are of ambivalent nature and are potentially dangerous to the life of the Tibetan king, like the rgyal-po Pe-dkar. This procession would have followed a ransom rite for the royal person, his spouses, and his palace, symbolized in the offering of the Mdos to Pe-dkar.

"The Cult of Pehar and Tsiu Marpo in Written and Oral Tradition:
The History of Nechung Monastery and its Mediums (A Continuation)"
by Ariane Macdonald

Translated by Kathryn Fuss and Christopher Bell

The subject of study this year follows the investigation breached last year on the festival of the "mdo-sde mchod-pa" at Samye. We retraced our development through an account—collected from the end of 1975 to the start of 1976, in Mysore—by one of the witnesses of this very rare ceremony, almost as little known to Tibetans as to Tibetologists, Sku-ngo Dalama.

We noted the essential roles played in the two masked dances ('cham), and in the procession that follows, by two protector deities of Buddhism, Pehar and Tsi'u dmar-po, who conflict with the land god of pre-Buddhist origin, Bar-tshigs brag-bstan. We have stumbled over several difficulties, the location of Tsi'u dmar-po's temple, for example, though we do not have to deal with the same problem in determining the date of this festival in its current form. We attribute the festival's foundation to an 18th-century Tibetan prince, but Sku-ngo Dalama has indicated to us that the 'cham had been introduced into the preexisting ensemble by the great Lama Sakyapa in the past. This was likely, since it is the Sakyapa monks of the principle temple of Samye who are responsible, these days, for the performance of one of the two masked dances.

The questions to resolve relate to several points, the first of which concerns the origin and development of the mythology and cult surrounding Pehar, as well as on the creation, mythological development, and liturgies devoted to Tsiu Marpo. Because of the displacement of the god Pehar to 'Tshal Gung-thang and 'Tshal Gnas-chung, it was necessary to find him a successor at Samye, and it appears impossible for us to treat the problems separately. They illuminate each other and they can only be understood according to the political circumstances that often explain the changes in religious ideas. Thus, in general, the history of the cult of Pehar, the fact that he became the protector deity of the Dalai Lama's government in 1642, and that his medium of Gnas-chung has since then become the oracle of the state, gives this god an importance and an influence that he had never had before.

We decided to begin the history of Pehar from his beginnings to the moment where we understood that it was the Sakyapa lama Kun-dga' rin-chen (1517-1584) who established in his great lineage the development of the different stages of the Mdo-sde mchod-pa. This is thanks to reading the Guide to the Restorations of Samye (Byam-yas dkar-chag), written in 1854, and the History of the Sakyapa Family (Sa-skya gdung-rabs), printed in 1630. This text and its complete biography—a very rare text that we borrowed from our colleague and friend at Delhi, E. Gene Smith, with his habitual generosity—proves that he (Kun-dga' rin-chen) founded, around 1570, the congregation of the Sakyapa monks, the rab-byung grva-tshang, who are charged with the organization of the Mdo-sde mchod-pa. They prayed to the protector god of Samye at that time, Tsi'u dmar-po, speaking through the mouth of his medium, to take ownership of the ensemble of the temples, to install himself there permanently, and to reconstruct and augment his own seat, the temple of riches where, currently, the second masked dance of the Mdo-sde mchod-pa takes place.

The precise and particular circumstances of the establishment of the Sakyapa college at Samye and the reconstruction of the Dkor-mdzod gling showed us at which point our analysis of the nature of the god Pehar, over the past years, has been distorted by preconceived and inspired ideas, and by the role held by Pehar since 1642 as the principal protector of the government founded by the Fifth Dalai Lama, whose family is bound, in the legend, to a form of Pehar. This is why we have found the definition of Pehar—provided by the biography of the Third Dalai Lama (1543-1588), a contemporary of Sakyapa Kun-dga' rin-chen—incomprehensible and aberrant in comparison with the actual official definition. In effect, all of the religious orders classify Pehar in the category of rgyal-po gods (earth gods whose name signifies 'king'). When he speaks through a medium, he often qualifies himself as a rgyal-'gong demon, a hybrid semi-rgyal-po, semi-'gong-po, the latter being a group of pre-Buddhist demons. For his part, Tsi'u dmar-po is considered a gnod-sbyin and a btsan. Or, in the detailed description of a great painting of Pehar and of the Third Dalai Lama, made in 1558, the protector god of Gnas-chung gave him his instructions for the painter to represent him with the name 'Od-ldan dkar-po, Pe-dkar rygal-po, and qualified Yang-le-cel as the great general of the Gnod-sbyin. In addition, when, in the same text, the Third Dalai Lama returned to Samye in 1560, which coincided at that time with the Mdo-sde mchod-pa (the festival was thus celebrated before Kun-dga' rin-chen became abbot of Samye), he met the protector god of the temple of the riches of Pe-dkar (Pe-dkar dkor-mdzod gling). It could only be Tsi'u dmar-po, qualified as a 'protector god of magical transformations or emanations,' and of the great Gnod-sbyin. The identification of Protector of magical emanations, an epithet of Pehar, for the great Gnod-sbyin Tsi'u dmar-po and for the god who was himself incarnated as the medium of Gnas-chung in 1558, was definite. Moreover, it was confirmed for us by the real oracle of Gnas-chung with whom we read these passages in 1975. He did not explain more than that to us, and the great painting conserved at Gnas-chung could not help to resolve the problem, for it had never been unrolled at that time. It seems clear to us now that according to the reading of the biography of Kun-dga' rin-chen, the establishment of the Mdo-sde mchod-pa festival around 1570 and the description of Pe-dkar of 1558 must be placed within the historical and religious context of the 16th century; it is the Third Dalai Lama, Kun-dga’ rin-chen, Padma dkar-po, Zhig-po gling-pa and many others who attest that the guardian of the temple of riches at Samye, the one who is himself incarnated as a medium, was Tsi'u dmar-po and not Pehar. At this time, the lineage of the Dalai Lamas, who did not carry this title yet, had very limited political power. The princes of Rin-sprungs, masters of a large part of Tibet until about 1565, and the princes of Gtsang who progressively eliminated them from 1565 to 1642, had for chaplains the Bka'-rgyud-pa, and above all the Karmapa, the Jo-nang-pa, and the Sakyapa, like Kun-dga' rin-chen, who was initially the religious master of the Rin-sprungs-pa, and later the princes of Gtsang. Until the seizure of power by the Dalai Lamas, it had thus not been Pehar who served as a model for the other chos-skyong, but the gnod-sbyin, the btsan, the eater of breath, flesh, and blood, Tsi'u dmar-po of Samye. Along with the Gtsug-lag-khang of Lhasa, Samye is the oldest religious center and the most venerated in Tibet. It is Tsi'u dmar-po—who became the protector of Samye after Pehar's transfer to 'Tshal and from 'Tshal to Gnas-chung—that the other chos-skyong model themselves after, including Pehar—who Tsi'u dmar-po eclipsed after having succeeded him.

We have thus noted that the dominant trait of a composite god isn't fixed once and for all, but varies according to the attraction exercised by the protector god of the group in power at one given era, at least with regard to Pehar and Tsi'u dmar-po in the 16th and 17th centuries. Thus it appears useful to survey one then the other from the beginning of their story, through their eventual developments, and across the datable texts and categorized chronologies of the Indian or Tibetan travelers who, passing through Samye, visited the temple of riches and mentioned one or several gods who are the guardians of it. Of course, also keep presently in the mind the theories concerning Pehar and Tsi'u dmar-po in the 18th-century accounts, for example those of Klong-rdol bla-ma and Sle-jung bzhad-pa'i rdo-rje. We thought that it was necessary to accord more importance to the similar name of the temple, and to the variation Pe-har/Pe-dkar, which is perhaps not without significance.

The most ancient of these texts seems to be a biography of Atisha, the great Buddhist Bengali who arrived in Tibet in 1042. It rests on the evidence of contemporaries and announces, in a single sentence, that in 1047 Atisha visited Samye and took up residency in the north part, where one finds the riches of the temple of Pe-kar. The concise text isn't easy to translate: gzims-chung yang Pe-kar gling gi dkor-mdzod byang-dngos-mar mdzad. It attests well that the oldest name of the temple of riches was Pe-kar gling (gling here is a synonym for temple), but the sense of the word Pe-kar does not arise clearly from the context. What is important is that it is not a transcription of vihara ("temple" in Sanskrit), Pe-har or Bihar, which is used. Also, the name of Pe-dkar appears in the biography of a Tibetan contemporary of Atisha, Rin-chen bzang-po, who thwarts his magical feats like that of a demon (cho-'phrul). This Pe-dkar is not bound to Samye.

The evidence that follows was read at the same time in the three pages where the author of the Guide to Samye of 1854 briefly mentions the restorations carried out on Samye during the 12th to 14th centuries and in the summarized texts themselves when we had them. The confrontation is very instructive.

The first date of the restorations is in the work of Rva lotsava, a Tibetan master the chronology for whom is poorly fixed (he is said to have lived for 150 years) but who—after the parallels provided by the lives of his contemporaries, notably Dar-ma mdo-sde and Ras-chung-pa—seems to have been active in the 11th-12th centuries. In the water-dog year, according to his biography, folio 119a—corrected in 1476 by 'Gos lotsava to the fire-dog year, which seems right and corresponds to 1106 (and not to 986 like Roerich thought, p. 378 of the Blue Annals)—Rva lotsava reconstructed the temple, which had fallen into ruin. But first of all he chastised the guardians of the temple of riches, who were guilty of having left it in a state of collapse.

The author of the Guide to Samye resumes this passage of the biography as follows: "Two days before arriving at Samye, Rva lotsava entered into a 'burning' meditation to test the abilities of the guardian of the riches of Samye. The next day, on the eve of his arrival, Pe-har, the master of riches (dkor-dbag Pe-har gyis), declared to the attendant of the temple, 'A great danger threatens me, I can't stay here. ' And he escaped." However the biography of Rva lotsava, the source of the Guide, states that the guardians of Samye were five 'masters of the earth,' gzhi-bdag and not dkor-bdag, five indigenous gods in the category rgyal-po, the rgyal-po sku-lnga, accompanied by their spouses and their sons, the yum-sras. These are Bya-khri spyan gcig, Rdo-rje grags-ldan, Gnod-sbyin chen-po, Bya-rgod thang-nag, and Putra nag-po. Pehar isn't named. This is because Bshad-grva, the author of the Dkar-chag of Samye, knew all the subsequent literature that presents Pehar as the guardian of the riches of Samye, in which—to summarize and without a doubt also standardize the very interesting story of the five rgyal-po—he substituted the singular Pehar for them. But he is also missing from the rnam-thar of Rva lotsava and, according to us, for a completely peremptory reason: It is that he was not yet invented.

The inventor of Pehar—this quasi-homophone Buddhist with the name of a pre-Buddhist earth god, Pe-kar or Pe-dkar, who tried to oppose the construction of Samye before being subjugated by Padmasambhava—is the author of one of the oldest legendary biographies of Padmasambhava: Myang-ral Nyi-ma 'od-zer, born in 1124 or 1136, died in 1204. He is also the "discoverer" of the most ancient consecration ritual to the five rgyal-po, of which Pehar is the chief, who are the manifestations of the body, words, thoughts, deeds, and intelligence of the five Jinas. Subsequently, as emanations they do not have spouses and sons, according to Rva lotsava, but spouses and ministers. These ministers are the five rgyal-po in the biography of Rva lotsava: Pehar and his mythology were born.

But as Pehar is a name which one can apply to any guardian of the temple—so says the author of the History of the Sakyapa in an account entitled 'Khon-lugs pur-pa, composed in 1626—since it is the transcription of the sanskrit vihara which means "temple," in the era when Kun-dga' rin-chen rebuilt the temple of the guardian deity of Samye's riches, it is Tsi'u dmar-po who is named Pehar.


1. For a detailed history of the western kingdoms see, Vitali 1996; for a discussion of the central kingdom lineage, see Hazod 2000.
2. Periodization in Tibetan history is equally tumultuous. For a suggested periodization scheme, see Cuevas 2003.
3. See Tambiah 1985, pp. 252-286, and Samuel 1993, pp. 62-63.
4. For a fuller discussion of the maṇḍala, see this online essay: The Maṇḍala.
5. It should be noted that King Trisong Detsen’s adoption of Buddhism was not a simple matter; see Kapstein 2000, pp. 23-65.
6. Snellgrove 2002, pp. 429-430.
7. Again, see Kapstein 2000, pp. 23-65. For more on the unnamed pre-Buddhist royal religion of Tibet, see Ḥaarh 1969 and Davidson 2005, p. 223-224.
8. For discussions of Samyé’s architectural significance, see Chayet 1988 and Mémet 1988. For an exploration of the maṇḍala as a political model, see Davidson 2005, pp. 24-28.
9. Kapstein 2000, pp. 60-61.
10. Ibid.
11. Kapstein 2000, pp. 63-65.
12. For a detailed exposition on Ra Lotsāwa, see Davidson 2005, pp. 129-141.
13. Davidson 2005 concerns itself exclusively with this era.
14. For a discussion of the Kadampa popularization of Buddhism during the renaissance period, see Davidson 2005, pp. 249-257.
15. Davidson 2005, pp. 252-254. While Tārā is not discussed in this analysis, a detailed exploration of her and her cult can be found in Beyer 1978.
16. For a fuller discussion of the cult of Avalokiteśvara, particularly in relation to the core textual corpus of his cult, the Maṇi Kambum, see this online essay: The Maṇi Kambum.
17. Kapstein 2000, p. 65.
18. See de Nebesky-Wojkowitz 1998 and Tucci 1999.
19. See Samuel 1993, pp. 161-167. Samuel (1993, p. 167) relates the Tibetan pantheon to the Tibetan political system with its lack of centralization and formal bureaucratic structure.
20. The main sources used to compile this list are Samuel 1993, pp. 162-163; Tucci 1999, pp. 717-730; Beyer 1978, pp. 293-301; Kelényi 2003, pp. 28-44; and Blo bzang tshe ring 1982, pp. 384-387.
21. For more on Serpent Demons as guardians of treasure texts, see Davidson 2005, pp.217-218.
22. For a detailed study of mighty demons and their connection to the Tibetan dynastic kings, see Gibson 1991.
23. See Haarh 1969, pp. 216-219.
24. See Bhattacharyya 2000.
25. Two vivid examples of this are found in Hildegard Diemberger’s “The Horseman in Red. On Sacred Mountains of La stod lho (Southern Tibet),” which discusses the protector deity Gangmar (Sgang dmar) and his associations with the lords and lands of Latö (La stod), and Guntram Hazod’s “bKra shis ’od ’bar. On the History of the Religious Protector of the Bo dong pa,” both of which can be found in Blondeau 1998.
26. See Gibson 1991, p. 201. His variant name in this text is Tsi Mara (Tsi ma ra).
27. Gibson 1991, p. 201. For those references mentioned, see Nyang ral 1979, pp. 466-76, and Stein 1961, pp. 36-42.
28. Two other texts that also concern Tsiu Marpo, were written by Ngari Penchen, and which immediately follow this one in The Great Treasury of Termas are Gnod sbyin ya ba rkya bdun gyi sgrub thabs srog gtad kyi cho ga ’dod pa’i re skong zhes bya ba bzhugs so and Chos ’khor skyong ba’i bstan srung btsan rgod chen po’i phrin las che ba btsan dgyes pa’i rngam glu zhes bya ba bzhugs so. See ’Jam mgon 1976-1980, pp. 299-369.
29. Bodhisattvas do have wrathful emanations, however, such as Hayagrīva, who is a wrathful form of Avalokiteśvara and who usually keep the worldly gods in line.
30. de Nebesky-Wojkowitz 1998, p. 167; Blo bzang 18th century, fol. 6b.6-7b.1.
31. For detailed discussions of Tibetan ritual masked dances (’cham), see de Nebesky-Wojkowitz 1976 and Kohn 2001.
32. This is an excerpt translated from Macdonald 1978a, p. 1141-1142; see the Appendix. Both of Macdonald’s articles provide details on the Tsiu Marpo cult and its political connections in the sixteenth century, though I disagree with her final assessment that Tsiu Marpo and Pehar are one and the same; their mythic histories, iconography, and liturgical corpus strongly suggest two very distinct figures. The Kagyüpa, also powerful at this time, equally had a hand in the growth of Tsiu Marpo’s cult. It is significant that when Samyé was reconsecrated in the sixteenth century, it was done so by Ngari Penchen, his younger brother Lekden Dorjé (Legs ldan rdo rje; 1512-1625?), and their friend and great Kagyü master Rinchen Püntsok Chökyi Gyelpo (Rin chen phun tshogs chos kyi rgyal po; 1509-1557); for detailed biographies of these three figures, see Gu ru Bkra shis 1990, pp. 531-544.
33. See Macdonald 1978b, p. 1025 and Appendix.
34. See Peter 1978a, p. 329.
35. For more on thread-crosses, their construction, and ritual use, see Beyer 1978, pp. 310-359, and de Nebesky-Wojkowitz 1998, pp. 343-397.
36. Pehar has a penchant for transforming into birds, and indeed I noted several paintings of white birds along the walls of the main temple at Samyé. During my time there, an old Tibetan woman who was circumambulating the inner temple explained to me that the painting of the white bird that so perplexed me was a form of Pehar, thus reinforcing this mythic history.
37. See de Nebesky-Wojkowitz 1998, p. 104; Peter 1978b, p. 288; and Gibson 1991, p. 60.
38. See de Nebesky-Wojkowitz 1998, pp. 578-579 for a cycle of texts either composed or commissioned by the fifth Dalai Lama.
39. See Martin 1996a and 1996b.
40. A fair amount of material exists on Tibetan oracles; see the “Oracles in Tibetan Deity Cults” section of this bibliography: Tibetan Deity Cults.
41. For a detailed but incipient exploration of the cult of Tsiu Marpo and, to a lesser degree, Pehar, see my thesis, external link: Tsiu Marpo: The Career of a Tibetan Protector Deity.
42. Dreyfus 1998, pp. 246-247.
43. Ibid, p. 245. For discussions on Dorjé Shukden, see Dreyfus 1998 and de Nebesky-Wojkowitz 1998, pp. 134-144.
44. Again, for the political and royal symbolism of the maṇḍala, see Davidson 2005, pp. 24-28.


Beyer, Stephan. 1978. The Cult of Tara: Magic and Ritual in Tibet. Berkeley: University of California Press.

Bhattacharyya, Narendra. 2000. Indian Demonology: The Inverted Pantheon. New Delhi: Manohar Publishers.

Blo bzang dpal ldan ye shes. 18th century. Gnod sbyin dgra lha’i rgyal po tsi’u dmar po la gsol mchod rdzogs ldan dga’ ston.

Blo bzang tshe ring, ed. 1982. Bdud rtsi snying po yan lag brgyad pa gsang ba man ngag gi rgyud ces bya ba bzhugs so. Lhasa: Bod ljongs mi dmangs dpe skrun khang.

Blondeau, A. M., ed. 1998. Tibetan Mountain Deities, Their Cults and Representations. Wien: Verlag Der Österreichischen Akademie Der Wissenschaften.

Chayet, Anne. 1988. "Le monastère de bSam-yas: sources architecturales." In Arts Asiatiques 43, pp. 19-29.

Cuevas, Bryan. 2003. "Provisional Periodization Scheme for History of Central Tibet." The Tibetan and Himalayan Digital Library. external link: http://www.thdl.org/reference/time/cuevas_timeline.html

Davidson, Ronald. 2005. Tibetan Renaissance: Tantric Buddhism in the Rebirth of Tibetan Culture. New York: Columbia University Press.

de Nebesky-Wojkowitz, René. 1976. Tibetan Religious Dances: Tibetan Text and Annotated Translation of the ’Chams Yig. Christoph von Fürer-Haimendorf, ed. The Hague: Mouton.

—. [1956] 1998. Oracles and Demons of Tibet: The Cult and Iconography of the Tibetan Protective Deities. New Delhi: Paljor Publications.

Dreyfus, Georges. 1998. "The Shuk-den Affair: History and Nature of a Quarrel." In Journal of the International Association of Buddhist Studies 21(2), pp. 227-270.

Gibson, Todd Allen. 1991. "From btsanpo to btsan: The demonization of the Tibetan sacral kingship." Ph.D. dissertation, Indiana University.

Gu ru Bkra shis ngag dbang blo gros. 1990. Gu bkra’i chos ’byung. Beijing: Krung go’i bod kyi shes rig dpe skrun khang.

Ḥaarh, Erik. 1969. The Yar-luṅ Dynasty. København: G. E. C. Gad’s Forlag.

Hazod, Guntram. 2000. "The Yum-brtan Lineage." In Civilization at the Foot of Mount Sham-po: The Royal House of lHa Bug-pa-can and the History of g.Ya'-bzang. Tsering Gyalbo et al, eds. Vienna: Österreichischen Akademie der Wissenschaften, pp. 177-191.

’Jam mgon Kong sprul blo gros mtha’ yas. 1976-1980. Rin chen gter mdzod chen mo 62. Paro: Ngodrup and Sherab Drimay.

Kapstein, Matthew. 2000. The Tibetan Assimilation of Buddhism: Conversion, Contestation, and Memory. New York: Oxford University Press.

Kelényi, Béla, ed. 2003. Demons and Protectors: Folk Religion in Tibetan and Mongolian Buddhism. Budapest: Ferenc Hopp Museum of Eastern Asiatic Art.

Kohn, Richard. 2001. Lord of the Dance: The Mani Rimdu Festival in Tibet and Nepal. Albany: SUNY.

Macdonald, Ariane. 1978a. "Le culte de Pehar et de Ci'u dmar-po dans la tradition écrite et orale. Histoire du monastère de Gnas-chung et de ses médiums (suite)," "Histoire et philologie tibétaines" (conférences 1976-1977). In Annuaire de l'Ecole Pratique des Hautes Etudes, Paris, pp. 1139-1145.

—. 1978b. "Les rivalités politiques et religieuses centrées sur Samye au XVIe siècle. La lignée spirituelle du Ve Dalai-Lama dans la littérature, dans l'histoire, et dans l'art." "Histoire et philologie tibétaines" (conférences 1977-1978). In Annuaire de l'Ecole Pratique des Hautes Etudes, Paris, pp. 1023-1030.

Martin, Dan. 1996a. "The Star King and the Four Children of Pehar: Popular Religious Movements of 11th-to 12th-century Tibet." In Acta Orientalia Academiae Scientiarum Hungaricae 49(1-2), pp. 171-195.

—. 1996b. "Lay Religious Movements in 11th- and 12th-Century Tibet: A Survey of Sources." In Kailash 18(3-4), pp. 23-55.

Mémet, Sébastien. 1988. "Le monastère de bSam-yas: essai de restitution." In Arts Asiatiques 43, pp. 30-32.

Nyang ral Nyi ma 'od zer. 1979. Chos 'byung me tog snying po sbrang rtsi'i bcud. Paro: Ngodrup and Sherab Drimay.

Peter, Prince of Greece and Denmark. 1978a. "Tibetan Oracles in Dharamsala." In Proceedings of the Csoma de Körös Memorial Symposium. Louis Ligeti, ed. Budapest: Akadémiai Kiadó, pp. 327-334.

—. 1978b. "Tibetan Oracles." In Himalayan Anthropology: The Indo-Tibetan Interface. Hague: Moulon, pp. 287-298.

Richardson, Hugh. 1998. "The Cult of Vairocana in Early Tibet." In High Peaks, Pure Earth: Collected Writings on Tibetan History and Culture. London: Serindia Publications, pp. 177-181.

Samuel, Geoffrey. 1993. Civilized Shamans. Washington: Smithsonian Institution Press.

Snellgrove, David. [1987] 2002. Indo-Tibetan Buddhism: Indian Buddhists and Their Tibetan Successors. Boston: Shambhala.

Stein, R.A. 1961. Une chronique ancienne de bSam yas: Sba-bzhed. Paris: Publications de L'Institut des Hautes Études Chinoises.

Tambiah, Stanley. 1985. Culture, Thought, and Social Action: An Anthropological Perspective. Boston: Harvard University Press.

Tucci, Giuseppe. [1949] 1999. Tibetan Painted Scrolls, 3 vols. Bangkok: SDI Publications, pp. 713-742.

Vitali, Roberto. 1996. The Kingdoms of Gu.ge Pu.hrang: According to mNga'.ris rgyal.rabs by Gu.ge mkhan.chen Ngag.dbang grags.pa. Dharamsala: Tho ling gtsug lag khang lo gcig stong 'khor ba'i rjes dran mdzad sgo'i go sgrig tshogs chung.