Compassionate Killing or Conflict Resolution? The Murder of King Langdarma according to Tibetan Buddhist Sources

Jens Schlieter
Science of Religion & Center for Global Studies
University of Berne


Is it justifiable to kill a tyrant? And just, who counts as a tyrant? Any ethical discussion about the legitimacy of tyrannicide or, to put it in more modern terms, the killing of dictators, faces particular difficulties. Obviously whenever this question arose in philosophical discussion it was intimately related to the vital interests of competing religious groups or organizations that were striving for political power. Indeed, it was mostly religious groups or thinkers who shaped and defined the concept of “tyrant.” According to Salisbury’s classical definition, a tyrant is a ruler who does not bow to the laws of God kept by the priests.² A king turns into a tyrant if he refuses to obey the divine laws of nature and instead makes use of his power for his own purposes. Violence against him is fully justified if, in the end, he “raises his sword.” Whereas the true emperor merely by acting conformably proves to be a legitimate king, he by violating the laws of nature turns into Lucifer and should thus be killed.³


In the European context the question of tyrannicide has been discussed from two main perspectives: first, in regard to the innocent victims suffering from the abuse of power, and with a second focus on the ruler’s lack of divine legitimacy. To use the terminology of the Christian Middle Ages, a distinction was made between a “tyrant by virtue of his actions” (tyrannus ex parte exercitii) and a “tyrant by virtue of illegitimate claims” (tyrannus ex defectu tituli).⁴ The latter, by the way, was in the eyes of most Christian thinkers the only sufficient justification for an attempt on a tyrant’s life. Never did the spiritual welfare, the salvation of the tyrant, play an important role in Western ethical debates on tyrannicide.

Rarely, then, was a position held like the one of Diogenes, cited by the Cynic and Stoic philosopher Dion Chrysostomos: a tyrant has no friends, his existence is marked by the constant fear that someone might make an attempt to assassinate him—so, in the end, he is happy only in the short moment when he realizes that he has been hit by a (deadly) attack, “because only then he feels released from the greatest evil.”⁵ But even in this text, the misfortune that goes along with being a tyrant is only of secondary concern. Indeed, a systematic teaching culminating in the justification of tyrannicide as a way of “liberating” a wrongdoer from his miserable existence did not develop in the European tradition.

In contrast to the European discussion of the legitimacy of tyrannicide, some Mahāyāna Buddhist texts propagate the idea that it is justifiable to kill a tyrant—defined as a militant enemy of Buddhism—out of compassion. It has been argued that an evil person, if killed without hatred, can be “liberated” (Tib. sgrol ba) from future bad karma that otherwise would have forced him to experience unlimited suffering in lower rebirths. These elements and stratagems underlying a Buddhist justification of tyrannicide will be analyzed in the following.

More precisely, the relationship between the idea of “liberation-killing” motivated by compassion and the political need of conflict resolution will be analyzed on the basis of one of the most famous “historical” examples: the murder of the last emperor of the early Tibetan dynasty, Langdarma (gLang Dar ma), by the Buddhist monk dPal gyi rdo rje. A legendary description of this important event is an essential part of nearly every book on Tibetan history.⁶ Yet, from the historiographer’s critical point of view, the assassination of Langdarma seems to consist of a possibly historical core overlaid with mysterious accounts. And though enigmatic, this disputable event had enormous effects on Tibetan history and historiography. The regicide may have taken place around the year 842. It brought the Yarlung (Yar klungs) dynasty, the whole of the early Tibetan empire, to an end. The following 70–150 years of the so-called “dark era” of Tibet saw little Buddhist activity, until a “second propagation” (phyi dar) of Buddhism led to a millennium of Buddhist dominion in Tibet.

The importance of the assassination that, according to the sources, was carried out by dPal gyi rdo rje, a Buddhist monk of lHa lung, can be gauged by the fact that he is revered by some major Buddhist schools as the precursor of Buddhism in Tibet.⁷ At least one monastery identifies him as a former reincarnation, or even traces its lineage of abbots back to him.⁸ Moreover, some contemporary Tibetan Buddhists have used his deed as an example of and justification for fighting against oppression—in their case the Chinese occupation of Tibet (since 1959).⁹

One last example of the sustained references to and perception of this meaningful event may be mentioned. Up to the present day the figure of Langdarma forms an essential part of certain annual mask dances. These culminate in a ritual in which all negative spirits are transferred into him as symbolized by a sacrificial substitute. An effigy of him, after absorbing these spirits, will be “killed” ritually. This can be seen as a significant indication of the fact that at least the more traditional schools felt no need to distance themselves from this act of tyrannicide. The act of symbolically repeating a killing in the form of a drama or mystery play is, of course, a well-known form of coping with the “great atrocity” (Freud) of regicide or tyrannicide. Without question, this event is deeply inscribed in the “collective memory” of the Tibetan community.

Much more astonishing, though, are the recent results of research on the early Tibetan historical treatment of Langdarma. Some have come to the conclusion that the portrayal of Langdarma found in Buddhist historical accounts that depicts him as determined to destroy Buddhism, is most probably a distorted picture. Not only that, but even the assassination of Langdarma may be fictitious.¹⁰ Non-Buddhist historical texts of more or less the same age found in the caves of Dunhuang do not mention the slaying of Langdarma. On the contrary, it is stated in one of these texts that Buddhism flourished at the time of Langdarma’s reign,¹¹ although others do speak of confusing turmoil. Rather than assessing and re-assessing these scanty and conflicting sources in order to find a clue if the assassination of Langdarma really happened,¹² it may be more useful to seek an answer to the following questions: Why has the report of this event (which will be analyzed below), found its way into Buddhist historical accounts? What functions and objectives do such reporting serve?

Obviously, the sources represent different perspectives in their accounts, as well as setting differing accents. A more elaborate examination of every single narrative might uncover further individual characteristics. Positions range from the cautious account in the Blue Annals of ’Gos lo-tsā ba gZhon nu dpal, whose description is scarce and in plain words, to the highly mythological narrations by gSal snang or Bu ston.¹³ As I would like to focus on the strategies for justifying violence presented in these texts, I will limit myself to some general trends that are part of nearly every account of the episode in question.

One of the general concerns of Tibetan Buddhist historians is to specify the unusual circumstances surrounding such an action. As a result, the story tends to give primarily not an account of a past event but an outline of or model for possible conflict resolution. In their descriptions, the historiographers offer some broad orientation about situations that, for them, “require tyrannicide.” But they also deal with the question how it happened that the Buddhist succession, and even institutionalized Buddhism as such, nearly disappeared in Central Tibet in the ninth century. All the historians dealing with the events of the ninth century lived and composed their works after the “second propagation” (phyi dar) of Buddhism in Tibet (beginning in the first half of the eleventh century).¹⁴ In other words, they had to look for sufficient reasons to explain the disintegration of the monastic order. The possibility should be considered that the temporary decline of Buddhism was projected onto one convenient enemy, the “anti-Buddhist” figure of Langdarma. His killing, then, could be seen as a necessary sacrifice, an act of purification of a community that, in one way or another, was already engaging in misguided actions. If this interpretation can be supported, the repetition of the deed in mystical plays would fit well into the scheme of a ritual intended to “‘purify’ violence; that is, to ‘trick’ violence into spending itself on victims whose death will provoke no reprisals,”¹⁵ as René Girard put it. This interpretation as a symbolic, “sacrificial” killing is, for the moment, our interpretation. It needs to be confirmed that texts of the indigenous tradition provide further evidence for an interpretation of the act as a “sacrificial offering by liberation-killing” (bsgral ba’i mchod pa). Yet, this should not be concluded solely on the basis that later rituals present an alleged killing of Langdarma.

One last aspect that will guide our interpretation should be mentioned here. In the murder of Langdarma, we can identify one of the very few cases where the act of killing a tyrant out of compassion has been reported in relation to a “historical” person. No clues are given in the texts at hand that it may have been only a “killing” of a symbolic nature. Some Tibetan Buddhist Tantrics subscribed to a specific teaching of “liberation-killing,” which has been understood at least by some as a merely symbolic way of killing hypostasized “counter-forces” or “poisons,” even within the perpetrator himself. Surely there have been commentators who asserted this kind of reading; on the other hand it has to be stressed—and this will be shown below—that other texts unequivocally speak about real killings. Moreover, it remains necessary to investigate to what extent this teaching really applies to the slaying of Langdarma—indeed whether it represents a strategy to justify a killing that occurred for other reasons.

As the following thesis posits, the texts possibly followed a complex strategy in their depictions of regicide: first of all, they delivered a dogmatically or philosophically grounded reason why this particular deed was necessary in order to deter the destruction of one’s own religion; secondly, they explained why a period of decline in the clerical order followed upon (or even: because of) it; and thirdly, they provided a script for a mythical scheme that allowed for the community-revitalizing repetition of ritualized killing in a dramatic mystical play.

The Description of Langdarma

Buddhist writings have bred the following picture of Langdarma. Around 838, Langdarma was brought to power by supporters of the noble party, who espoused pre-Buddhist religious beliefs. Before his inauguration, the nobles murdered King Ral pa can, who had been recently endowed with gifts and privileges, which he passed on to the Buddhist clergy.¹⁶ According to many sources, Langdarma (who governed from ca. 838 to 842) was the brother of Ral pa can. The relationship is noteworthy, for the new king could claim a certain legitimacy for his succession to power. In Buddhist texts, this right to rule is indeed granted to him. There is broad consensus in textual sources that he began his opposition to the institutionalized Buddhist doctrine after about six months in power. Langdarma did not, however, wage a campaign against the Buddhists purely on his own. Several texts mention that he was supported by a number of ministers who were “ill-disposed toward the Dharma” (blon po chos la gnag pa, MTPH 14a6). Some texts cite as one reason for this animosity the fact that in certain circles the introduction of the Buddhist doctrine was held responsible for the widespread storms and rampant famine of the time (see BZ 78,5–6, MTPH 14b1). Furthermore, he supposedly became possessed by a “demon” (gdon),¹⁷ (BZ 80, 1–2; MTPH 14b4, DTM 32a), who then drove him to forbid the display of religious practice of all Buddhist monks. Those who disobeyed his new rules were armed and forced to go hunting. This would, of course, have been a humiliation to Buddhist monks, who had to submit to the training rule to abstain from killing animals.¹⁸ Those monks who refused the king’s order were allegedly “killed” (bsad do, BZ 79, 8–10; MTPH 14b2; CB 145b4). According to Bu ston’s Chos ’byung (CB 145b), the king put a halt to the translation work and destroyed the monastery schools. Afterwards he commenced with the persecution of believers, and had those who resisted executed. Interestingly, in the Deb ther dmar po no passage explicitly mentions the killing of believers, though it does state that the monks could no longer practice their clerical profession or pass on religious texts and transmissions (DTM 31a). Several texts portray famines and epidemics exploited by the king as false accusations or pretense (de la snyad byas, GML 235, 10) to advance his major ambitions—to destroy the religion.¹⁹

Noteworthy is the king’s epithet, “ox” (glang).²⁰ There seems to be unanimous concurrence that this nickname was conferred later in his life. The description in one particular text illustrates how it was used as a justification for the slaying of the king. One might even conclude that it was created precisely for this purpose. The DTM recounts that King (Khri) Dar ma (’U dum btsan) was first nicknamed Ox after the Buddhist teachings began to dwindle, religious practice was suspended and no new ordinations took place. The report further tells of a traditional prophecy that a “king with the name of an animal will be born,” who is depicted here (and elsewhere) as a satanic “manifestation of Māra” (bDud kyi sprul, DTM 32a). Thus the killing of the “ox” would have been a fitting response to the prophecy. This account can be viewed as a substrategy for justifying the act of violence.

The verbal equation of the “tyrant” with an animal is yet another aspect. Unvirtuous people can even sink “below” the animals. Several dogmatic works of the Sarvāstivāda school, for instance, pose the question whether it is worse to kill animals or virtueless people bereft of the basis for salutary and good behavior (samucchinna-kuśalamūla).²¹ As Samten G. Karmay has shown, some portraits of the king suggest that his activities were not directed against Buddhism itself, but rather against a certain “monastic establishment,” which demanded the support of a considerable number of monks, construction projects and the like in difficult economic and military circumstances.²² This establishment was the one threatened by the king’s deeds, whereas non-monastic Tantric circles continued to practice unhindered at this time, and especially after the end of the Yarlung dynasty. Given the fact that the name dPal gyi rdo rje appears as the ninth abbot in a list of abbots from Samye (bSam yas), Karmay concludes that the murderer was actually an abbot from this renowned monastery. In addition, he was able as a result of this murder, to defend his disputed claim to power.²³ However, this noteworthy thesis is based on sparse textual evidence and therefore can neither be verified nor falsified.

Nevertheless, the picture is not complete without a description of the actual killing, as the texts portray it.

The Killing of Langdarma

The hermit dPal gyi rdo rje, who lived secluded in a cave near Lhasa and who, according to some sources, was a bodhisattva (i.e. byang chub sems dpa’, MTPH 14b5), resolved to put an end to the unsuitable behavior of the demon-king. Some versions tell that the patron goddess of Tibet, dPal ldan lha mo (Śrīdevī),²⁴ appeared to him in a dream (rmi lam) and informed him that the time had come to “kill the sinister king” (BZ 81/10: sdig rgyal gsad; CBMT 438/16; MTPH 14b6; GML236/18).²⁵ Some accounts also emphasize that dPal gyi rdo rje originally acts after he rouses “great compassion for the king” (btsan po la snying rje chen po skyes, CBMT 439/1–2).²⁶ The GML states that he put his own life and interests in abeyance, took the teachings of the Buddha to heart, and made this courageous decision (spobs pa, GML 236/20). dPalgyi rdo rje pondered a suitable “means” (BZ 81/12: thabs, Skt. upāya).

He took a reversible cloak that possessed both a white and a black side. Using coal he colored a white horse black. In the robe, with the black side turned out, he hid a bow and placed three arrows in the wide-cutsleeves. Armed in this way, he rode to Lhasa. According to a number of texts, he bowed before the king, who was reading the inscription on an obelisk.²⁷ As he started to return to an upright position after this ostensible display of respect, he cocked the hidden bow so that he could shoot the deadly arrow at the king.

The story of this assassination, disguised as homage, may, however, have a deeper meaning. At the precise moment when the king reads the stone inscription or turns to the stranger who has come in reverence, his mind is most likely absent of hate or other insidious thoughts. To kill him now would mean to facilitate a better rebirth for him, because the spiritual condition at the time of death is regarded as particularly significant.²⁸ The deed is accompanied by the words: “I am the (black demon) Ya bzher nag po. A king fraught with sin must be killed this way” (MTPH 14a2). Furthermore, it has been reported (i.e., BZ 82, CBMT 440) that the king’s last words expressed that the killing had come either three years too early or three years too late.²⁹

In any case, dPal gyi rdo rje manages to flee after the successful assault. He turned the robe inside out (now displaying the white side) and rode his horse through a river, claiming, “I am the (white heavenly ghost) gNam The’u dkar po” (CB 146a3),³⁰ so that his pursuers could not recognize him.³¹ According to Bu ston’s report, dPal gyi rdo rje flees to Khams with three important writings in order to preserve the Buddhist teachings. The renewed dissemination of Buddhism thus began in East Tibet,³² but without direct participation from dPal gyi rdo rje, who lost the rank of an ordained monk due to the assassination (MTPH 147b). Evidently, it proved to be difficult for historiographers to give an account of dPal gyi rdo rje’s responsible actions without breaking with the ruling that his ordination was no longer valid. The propagation, however, did take place after dPal gyi rdo rje’s flight to Khams, where he and two other monks carried out the consecration of the first novices (dge tshul, Skt. śrāmaṇera). Most texts assert that the other two took over the necessary functions of “teacher” and “instructor” (mkhan po, Skt. upādhyāya; slob dpon, Skt. ācārya). dPalgyi rdo rje was thus still involved in the reinstallation of Buddhism, but he no longer took a leading role in it.³³ In several versions (e.g. BZ85/3–4) he explains that in killing the sinful king he was in effect renouncing his vows and therefore could not carry out ordinations anymore.

Apparently, an apologetic strategy is being pursued here. It transforms the possibly historical core of this event into a mythical scene, one that places both the victim, and the assailant, firm in his decision, in a constellation that categorically rules out all potential alternatives to assassination. Accordingly, in donning a black cloak, dPal gyi rdo rje takes on a “dark” existence, which he distances himself from after the act by turning the white side of the cloak outwards. He dyes his horse black, so that it becomes a mystical vehicle that advances the successful symbolic transformation at the time of the incident. Furthermore, because a divine being gives the order, the assassination acquires an even higher legitimacy to the point of becoming unavoidable. To the extent that the victim Langdarma displays animalistic or demonic features, he is successfully stripped of his human attributes. This characterization becomes important, since each pre-modern attempt to justify the act of tyrannicide poses a particular dilemma: The continuity of the hallowed rank of the king, or the institution of kingship, shall, nevertheless, be maintained. Thus the exceptional nature of tyrannicide is stressed in order to prevent this form of “conflict resolution” from escalating out of control. Therefore, the described circumstances must demonstrate an infallible urgency. This kind of urgency, however, can be evoked most effectively through inner motives that compel action. In contrast, even a stark portrayal of external circumstances that call for action can be challenged, since peaceful alternatives, whether for conflict resolution or at least conflict avoidance (e.g., escape, total refusal), are always given.

However, let us now turn to the more specific “religious” justification of the “liberation killing” of Langdarma.

Elements of a “Religious” Justification for Violence: Killing out of “Compassion” (snying rje) and as “Liberation” (sgrol)

Before addressing the notion of killing as a method of liberation, it is necessary to consider the following observation. Many analysts presume that the accounts of the killing of Langdarma are dealing with an instance of ritual “liberation.” Although this interpretation seems to be obvious, one needs to examine to what degree the historiographic texts themselves reflect to this particular understanding. It is notable that the relevant term “liberation” (sgrol, or bsgral [pf.]; Skt. mokṣa) was not used in early texts to describe the killing of Langdarma, but rather terms that merely express the actual act of killing such as Tibetan gsod pa, “to kill.” This is the word dPal ldan lha mo uses to convey her request concerning the wicked king when appearing in the dream (see BZ 81/10; MTPH 14b6). Only a few texts state explicitly that the tyrant-king “must be liberated.”³⁴ Therefore, the question arises whether one can conclude that the killing was interpreted in the sense of a “liberation” after the fact. Alternatively, one may posit that the references to the “great compassion” of the assailant and the “demonic” obsession (gdon) of the victim suffice to support the above interpretation.

Of the various threads of “liberation killing” that can be traced either as formative forerunners or as stages of its Buddhist development, we may confine ourselves to the three most important. Firstly, the prefiguration of the notion of a “liberation killing” among the non-Buddhist Saṃsāramocaka and some Śaiva schools; secondly, early Mahāyāna thoughts on the bodhisattva’s justification of killing “mass murderers”; and thirdly, the Tantrist-ritualistic strand of “liberation killing.”

As a first step, it appears worthwhile to emphasize the original formative elements of these tenets, namely the contexts of karma and reincarnation. It is a general precondition of this teaching (in contrast to the Western background), that someone gathers an excessive amount of bad karma through his acts, which will lead him to a worse rebirth. A further supposition is the idea that the one “to be liberated” is not able to correct his fatal behavior due to his karmic delusion.³⁵

As mentioned above, in the early Indian context the notion of a violent liberation can, among other things, be traced back to the non-Buddhist school called “those who liberate [others] from the cycle of rebirth” (’khor ba sgrol byed pa, Skt. saṃsāramocaka). Their teachings, however, are known only through the polemics of other schools. According to those sources, this school expounded the doctrine that the murder (Skt. hiṃsā) of beings which are trapped in the cycle of rebirthis sanctified for the perpetrator as well as the victim.³⁶ Yet, influential thinkers of the Buddhist tradition dismissed this teaching as a “barbarian” practice.³⁷

The notion that someone should forcefully be “liberated” against his or her own will seems to go beyond at least the early Pāli Buddhist teachings. According to Rupert Gethin’s analysis of the Theravāda-Buddhist concept of “compassion” (Skt. karuṇā), it seems to be clear that in these sources any act of “compassionate killing” is not compatible with the idea of compassion at all. Someone who develops a compassionate intention is indeed no longer able to engage in an act of killing, since the unwholesome roots of hate and delusion which alone enable the intent to kill to arise are no longer present.³⁸ In early strands of the Vinaya, even active participation in voluntary euthanasia has been excluded from permitted behavior.³⁹

In contrast, the virtue of active compassion of the bodhisattva in Mahāyāna Buddhism seems to lend itself more readily to the doctrine in question. A well-known, yet not very influential text of Indian Mahāyāna, where this idea of “liberation killing” is expounded, is the Skill-in-Means Sūtra (Skt. Upāyakauśalya-sūtra). One of the bodhisattva’s deeds described in this sūtra consists of the following simile:⁴⁰ a homicidal thief, who has hidden himself among 499 traders on a voyage by ship, is waiting for the opportunity to kill everyone on board and take their belongings. However, on board of the ship is the “captain (or: leader of the traders) Great Compassionate” (ded dpon snying rje chen po, Skt. *sārthavāho mahākāruṇikaḥ). He is informed about the evil man and his plan—again as a dream appearance!—by gods of the sea, who tell him, moreover, that the murder of the 500 traders would result in massive negative karma for the culprit, since all of these traders are on the path to awakening. After seven days of contemplation, the captain becomes convinced that he must kill the indicated assailant in order to protect him from the karmic effects of his own misdeed. He cannot, in line with the earlier premonition, inform the traders because they—in contrast to him, the captain—would kill the thief with wrathful emotions and thoughts. Fully aware of the eons he will have to spend in hell for his deed, he thus kills “with great compassion” the thief, who as the text asserts will now be reborn in a heavenly sphere. Lastly, it is ascertained that the bodhisattva-captain’s actions do not derive from bad karma, but can be regarded rather as a “skilful means,” as a bodhisattva’s act of virtue. One of the historiographic texts on the killing of Langdarma even refers directly to the simile of the bodhisattva-captain. bSod nams grags pa comments on Langdarma’s murder as follows: “It was the act of a fearless bodhisattva-hero; it resembles the deed of the captain ‘great compassionate,’ who killed the dark man with the short spear” (DTM 32a).⁴¹ It may be mentioned parenthetically that also Tenzin Gyatso (bsTan ’dzin rGya mtsho), the current Fourteenth Dalai Lama, has referred to this simile,⁴² in order to justify killing in exceptional cases:

If someone has resolved to commit a certain crime that would create negative karma, and if there exists no other choice for hindering this person from the crime and thus the highly negative karma that would result for all his future lives, then a pure motivation of compassion would theoretically justify the killing of this person.⁴³

Yet the Dalai Lama emphasized that it is generally “better not to use this method,” since “killing out of mercy” is a principle that “always bears the risk of negative reasons and feelings.”⁴⁴

In contrast to the aforementioned passages⁴⁵ mainly setting forth guidelines for the bodhisattva’s course of action, which may encompass killing tyrants, some Vajrayāna Buddhist sources that deal with “liberation killing” concentrate entirely on the liberation of the victim. Moreover, this exceptional event in turn fosters a progression on the path to spiritual awakening of the slayer.⁴⁶

Schmithausen points out that in some sources, for example the Nayatrayapradīpa, the Tantric practice of “liberation (killing)” (sgrol) results in the victim’s rebirth in a more favorable existence and his subsequent entrance upon the way to buddhahood; whereas another, more traditional source, allows him to transfer to a paradisiacal sphere, and finally, to the realm of buddhahood.⁴⁷

It is difficult to pinpoint exactly when the Tantric elements of a ritual practice of liberation emerged, since they were passed on among exclusive circles. Yet a great deal suggests that they were already being practiced in Langdarma’s time. In any case, the myth of the Indian deity Rudra (Śiva) and his suppression by the Buddhist divine being Vajrapāṇi plays a significant role in this belief.⁴⁸ The associated ritual practice reenacts the transformation of Rudra, formerly a non-Buddhist deity, into a protective deity of Buddhism.

The “liberation” (sgrol ba) belongs to one of a group of four Tantric ritual acts, that of forceful subjugation (mngon spyod, Skt. abhicār[uk]a), and is often treated as one of the practices of “sexual unification” (sbyor ba). The main difficulty in adequately interpreting the initial teaching of “liberation killing” is the status of the teaching itself. Was it meant as a symbolic practice or an instructional guide for real killings? It can be shown that in some texts⁴⁹ that prefer a symbolic reading, opponents of this teaching changed its core message and replaced any allusion to real killing by a symbolic subjugation of counteracting spiritual forces.⁵⁰ Nonetheless, a number of texts state unambiguously that certain people should be killed, for instance, opponents of (Mahāyāna) Buddhist teachings and those who commit patricide or matricide.⁵¹ As part of the ritual, the fatal “liberation” (bsgral) of the adversary is carried out with the “thunderbolt dagger” (phur pa, Skt. kīla) against a surrogate victim in human form (ling ga), from which the evil forces have been previously banished.⁵² According to the theory, the body of the opposing adversary (conceived as Rudra⁵³ or Māra) is transformed into a sacred space and thereby reveals its obscured buddha-nature.⁵⁴ A more meticulous description of the ritual practice must here be dispensed with.⁵⁵

In the end, the decisive point is that the practitioner of the liberation ritual has to make sure that the ‘consciousness’ of the freed person is transferred to a heavenly realm. This can be assured only when the practitioner possesses an absolutely pure, compassionate state of mind.⁵⁶ This situation given, Tantric sources claim that the person’s own power and spiritual capacity may grow considerably. Interestingly, in regard to dPal gyi rdo rje the texts say nothing about a strengthening or spiritual progress as a result of his endeavor. On the contrary, considerably more attention is paid to the fact that he may no longer consecrate or sanctify.

Hence the question remains whether there are clear indications that the killing of Langdarma was a Tantric sacrificial ritual. Seyfort Ruegg argues that the most apparent sign lies in dPal gyi rdo rje’s “supreme” or “great compassion” (snying rje khyad par can), which he cultivates in order to prepare for the assault.⁵⁷ A further hint is the fact that the king is called “Māra.”⁵⁸ Supreme compassion, though, is already mentioned in the early Mahāyāna sūtras, and thus is not a distinctive characteristic of Tantric practices. The comparison to Māra suggests a somewhat clearer precondition, but by itself does not suffice. Another indicator is certainly the manifestation of dPal ldan lha mo. Yet she calls merely for the slaying, not the liberation. Moreover, the Upāyakauśalya-sūtra, among other texts, had already dealt with a deity who appears in a dream to reveal a dangerous individual and to order his killing. Indeed, the concept of “liberating” (sgrol) rarely turns up in early historiographic descriptions of the Langdarma-episode. Moreover, some Tantric texts emphazise that “liberation” (sgrol ba) should not be equated with “killing” (bsad pa), the latter being the word used most in descriptions of the slaying of Langdarma.⁵⁹

To interpret the act of killing based on descriptions of it in the early texts, the presumption of a sacrificial ritual of liberation cannot be dismissed, but it is not very probable either. The use of themes of early Mahāyāna texts seeking to justify the slaying of violent aggressors evinces a much clearer link. The singularity of dPal gyi rdo rje’s attack, the extraordinary circumstances of his deed, makes much more sense in the context of an attempt to validate or justify tyrannicide.

The assumption of numerous analysts that the event analyzed above is a ritual “liberation killing” may have arisen due to the way the account is structured. It comes very close to the classic scheme of ritual phases: following the phase of detached preparation, dPal gyi rdo rje undergoes a phase of liminal transition, in which the transformation of the participants takes place, and the final phase of reintegration ensues.⁶⁰ This transformative scheme can in fact be observed in ritual killing.⁶¹ Yet in the case of Langdarma’s killing the attack is committed by a single person without any other participants. This particular act can thus hardly be viewed initially as a community-building act. To be sure, it is not inconceivable that a “liberation killing” ritual could develop certain community-building functions. It seems to be clear that the killing of Langdarma could be accepted in these terms, as is illustrated by the annual mask dances.

The community-building function, then, is obviously present in the ceremonial repetition of the killing, but not yet in the initial killing itself. Through the communal re-enactment of the original “sacrifice”—that is, the killing of Langdarma, as it is present in cultural memory—the order based on this act is reinforced.⁶² This may be most strongly represented in annual ritual sacrificial festivals, but the function also endures through the repeated depictions of the killing of Langdarma, and the increasing embellishments of them, in historiographic texts.


The core of the previously discussed Buddhist validation of violence against violent wrongdoers reveals that in contrast to the Western discussion the suffering caused to the victims of a tyrant’s rule is by itself not a sufficient justification for tyrannicide.⁶³ The legitimacy of violence rather is linked to the spiritual welfare of the wrongdoer. The Buddhist argumentation relates primarily to the doer of deeds, who accumulates merit or guilt depending on his actions and intentions. The reasoning therefore revolves around the perpetrators, who risk their spiritual destiny. The slaying implies, albeit indirectly, that further intentional acts of violence will be prevented.

Nonetheless, the depiction of the killing of Langdarma involved various aspirations, and even different strategies for justifying it. Some strategies indeed openly strive to justify violence. Other accounts of the event, on the other hand, set out to explain the temporary turmoil and setbacks in the establishment of the Buddhist monastic order in Tibet by projecting these conditions onto two offenders—the “tyrant” and the “liberator.” Some of these accounts present an ambivalent picture of dPal gyi rdo rje, since it was he and his deed that played an important role in the subsequent decline; yet most descriptions are very sympathetic with the assassinator.

Under yet another angle, the regicide became the basis for a myth among a ritual community that modified this incident a posteriori into a (symbolic) “liberation killing,” namely the plot of a sacrificial drama.

But given that the argumentation above has some support, it seems in the end highly unlikely that Langdarma was originally killed with the motive of “freeing” him in a Tantric-ritualistic sense.

All things considered, the purpose of the legend is hardly questionable, namely to present and legitimize a violent mode of conflict resolution. Indeed, it may be presenting itself as a possible model for resolving conflicts yet to come.  ■


¹ I would like to express my gratitude for helpful comments on an earlier draft of this paper to Karénina Kollmar-Paulenz, Michael Zimmermann, and Helmut Eimer.

² See Policraticus III.15; Katharine S. B. Keats-Rohan, ed., Ioannis Saresberiensis, Policraticus I–IV (Turnholt, Belgium: Typographi Brepols Editores Pontificii, 1993), 229–30; trans. in Cary J. Nederman, John of Salisbury: Policraticus (Cambridge: Cambridge University Press, 1990), 32–4.

³ Policraticus VIII.17; see Clemens C. I. Webb, ed., Ioannis Saresberiensis episcopi carnotensis Policratici: Sive de nugis curialium et vestigiis philosophorum libri VIII, 2 vols. (London, Oxford: Clarendon, 1909), 2:160–6; cf. Robert A. Lauer, Tyrannicide and Drama (Stuttgart: Franz Steiner, 1987), 34. Salisbury even justifies the use of flattery and fraud in the killing of tyrants (Policraticus VIII. 18, 20). Does this seem necessary for him to demonstrate the possibility of committing tyrannicide despite the military superiority of the ruler?

⁴ See Mario Turchetti, Tyrannie et tyrannicide de l'Antiquité à nos jours (Paris: Presses Universitaire de France, 2001).

⁵ See Gernot Krapinger, Dion Chrysostomos Oratio 6. Text, Übersetzung, Einleitung und Kommentar (Graz: Verlag für die Technische Universität Graz, 1996), 29/45. It is clear from the context (see ibid., 31/53–4), that Gr. plegê, “stroke,” refers to an assassination.

⁶ See, for example, Tsepon Wangchuk Dedhen Shakabpa, Tibet: A Political History (New Haven: Yale University Press, 1984), 48–53.

⁷ A description of the event without commentary or assessment can be found in, among others, Jikdrel Yeshe Dorje Dudjom Rinpoche, The Nyingma School of Tibetan Buddhism: Its Fundamentals and History, trans. and ed. Gyurme Dorje and Matthew Kapstein (Boston: Wisdom Publications, 1991), 1:523–4, 536, 612, and in the works of the Fourteenth Dalai Lama, as in My Land and My People (New York: McGraw Hill Book Company, 1962).

⁸ For instance he is thought to have reincarnated as ’Khrul zhig Rin po che of Rong phug Monastery, according to Robert A. Paul, The Tibetan Symbolic World: Psychoanalytic Explorations (Chicago: University of Chicago Press, 1982), 1–3.

⁹ See Jane Ardley, Violent Compassion: Buddhism and Resistance in Tibet, Political Studies Association, 50th Annual Conference, April 10–13, 2000, online: (accessed October 20, 2003), 25–6. It should be added that violent solutions are dismissed by a very large number of Tibetan Buddhists today. Critics of the idea of a “compassionate liberation-killing” can be found among leading Buddhists as early as lHa Bla ma Ye shes ’od, Atiśa, or bTsong kha pa.

¹⁰ Zuiho Yamaguchi, “The Fiction of King Dar ma’s Persecution of Buddhism,” in Du Dunhuang au Japon: études chinoises et bouddhiques offertes à Michel Soymié, ed. Jean-Pierre Drège (Geneva: Droz, 1996).

¹¹ See fragment PT 840, translated and discussed by Samten Gyaltsen Karmay, The Arrow and the Spindle: Studies in History, Myths, Rituals and Beliefs in Tibet (Kathmandu: Mandala Book Point, 1998), 79, 91–2; cf. Yamaguchi, “The Fiction,” 241–2. In the mKhas pa’i dga’ ston of dPa’ bo gTsug lag phreng ba (finished 1564) the life of dPal gyi rdo rje is portrayed, with extensive use of ancient and probably lost texts (e.g., the Yer pa’i dkar chag) but, oddly, without a word said of the assassination of Langdarma. See also Tsultrim Kelsang Khangkar, “The Assassinations of Tri Ralpachan and Lang Darma,” Tibet Journal 18-2 (1993): 21–2.

¹² Regicides surely happened quite often in the early dynasty. Haarh, in his classical study of early Tibetan Yarlung kings, has produced evidence that at least six kings of the last eleven emperors were slain: Erik Haarh, The Yar-lung Dynasty: A Study with Particular Regard to the Contribution by Myths and Legends to the History of Ancient Tibet and the Origin and Nature of Its Kings (Copenhagen: Gad, 1969), 328. All in all, regicide is seen to have been a most common practice by which ruling kings were removed, not only in Tibet. Was therefore the killing of Langdarma an instance of ritual regicide rather than a tyrannicide? This, I believe, can be discounted because it would then have to be proven that this “Buddhist” case involved installing a new king. Ritual regicide earns its legitimacy through the support of the office as such. Ideally, a suitable candidate for succession to the throne is always available if the regicide succeeds. In our case, though, the slaying helped cause the collapse of the institution of kingship. Even if this historical fact cannot be portrayed as the intention of the assailant, it is still noteworthy that the texts, even those that report about this collapse, hardly care to analyze the end of monarchy. This clearly contradicts the nature of archaic regicide as envisioning a smooth transfer of power.

¹³ ’Gos lo-tsā-ba gZhon nu dpal persistently maintains in the “Blue Annals” (Deb ther sngon po, written 1476–1478) his intention to report only what he actually believes to be credible (personal communication with Helmut Eimer). He limits himself, for instance, simply to relaying the message of the killing by dPal gyi rdo rje (see DTNG 1:53). Also the Fifth Dalai Lama attempts to put the event into perspective. The “higher motives” are mentioned as passing remarks (see BDTH 45b). In not heroizing this feat, the Fifth Dalai Lama may have been motivated by the fact that any justification for tyrannicide could be turned around and used against his own claim to power, which itself was not achieved peacefully.

¹⁴ See Paul’s highly hypothetical reconstruction, which sees the enduring triumph of Buddhism in Tibet as an outcome of the constitutive regicide as an example of patricide (The Tibetan Symbolic World, 288–92). Langdarma appears as an “infanticidal father,” who is killed by the “junior male” and avenger dPal gyi rdo rje.

¹⁵ René Girard, Violence and the Sacred, 2nd ed. (Baltimore: Johns Hopkins University Press, 1992), 36.

¹⁶ There is mainly unanimity on the murder of Ral pa can in historiographic sources. A number of texts though, including Chinese chronicles, report a natural death.

¹⁷ For gdon see René de Nebesky-Wojkowitz, Oracles and Demons of Tibet: The Cult and Iconography of the Tibetan Protective Deities (Graz: Akademische Druck- und Verlagsanstalt, 1975), 310–7.

¹⁸ Furthermore, it represented a social degradation, given their rejection of the professional slaughtering of animals, as documented by Lambert Schmithausen. See his “Zum Problem der Gewalt im Buddhismus,” in Krieg und Gewalt in den Weltreligionen: Fakten und Hintergründe, ed. Adel Theodor Khoury, Ekkehard Grundmann and Hans-Peter Müller (Freiburg, Basel, Vienna: Herder, 2003), 85–6.

¹⁹ In addition, Langdarma is said to have been delivering “anti-Chinese” diatribes: the Chinese Princess, in fact a female yakṣa, has brought the divinity of yakṣas, Śākyamuni, to Tibet. See Samten Gyaltsen Karmay, The Great Perfection (rDzogs chen): A Philosophical and Meditative Teaching of Tibetan Buddhism (Leiden and New York: Brill, 1988), 8–9.

²⁰ The MTPH 6b2 even attributes to him an ape’s head and a sheep’s rear.

²¹ See Lambert Schmithausen and Mudagamuwe Maithrimurthi, “Tier und Mensch im Buddhismus,” in Tiere und Menschen: Geschichte eines prekären Verhältnisses, ed. Paul Münch and Rainer Walz (Paderborn, Munich, Vienna, Zurich: Ferdinand Schöningh, 1998), 211 (with a reference to the Peking Tanjur).

²² Karmay, The Arrow and the Spindle, 97–8.

²³ See with reference to Dunhuang document no. III (IOL 682/2): Karmay, The Great Perfection, 77–8.

²⁴ According to legend, this patron deity, who is significant and not the least squeamish, rides on a mule with a saddle made from her own son, whom she killed herself (cf. Nebesky-Wojkowitz, Oracles and Demons, 22–37).

²⁵ This is part of the strategy of justification and has the following elements (BZ 81–2, GML 236): dPal ldan lha mo “says” more precisely to him, (a) there is no other siddha in Tibet except for him, (b) the teachings of the Buddha are threatened, (c) the time has come to kill the king, (d) she will help him, and (e) he should not despair or doubt.

²⁶ Compare CB 14b6, BZ 81/11, DTM 32a, MTPH 14b6.

²⁷ BZ 81/15; CBMT 439/14–6; MTPH 14b7. The description of the king reading the inscription possibly illustrates the request that he may return to the “ancestral writings” (yab mes kyi yi ge, MTPH 14b4) found in several sources. In the GML too, the Buddhist ministers urge the fickle King Langdarma to respect the customs and writings (yig tshang) of the forefathers. It is nevertheless peculiar that he is murdered at exactly the moment of this retrospective contemplation; or does it demonstrate, that his mental condition is temporarily “wholesome,” thus again indicating a convenient moment to kill him?

²⁸ In one version of the Mahāparinirvāṇa-sūtra [Kosho Yamamoto, The Mahayana Mahaparinirvana-sutra: A Complete Translation from the Classical Chinese Language in 3 Volumes. Annotated and with Full Glossary, Index, and Concordance, vol. 2 (Tokyo: Karinbunko, 1974), 393–4], the idea is expounded upon that the mental state of the killer who kills out of compassion should be similar to that of the one being killed. For the significance of the mental condition of the dying person in early Buddhism, see, for example, Saṃyutta Nikāya IV, 308–11 in Bhikkhu Bodhi, trans., The Connected Discourses of the Buddha, 2 vols. (Boston: Wisdom Publications, 2000), 2:1334–6.

²⁹ T.K. Khangkar, “The Assassinations …”, 19, includes as a possible interpretation: “had the king been assassinated earlier, the persecution of religion would not have taken place, or had he not been assassinated he would have then become a religious king”.

³⁰ It is noteworthy that the assailant identifies himself with a supreme deity from the pre-Buddhist pantheon who plays an important role in the mythological legitimation of earlier kings (see Nebesky-Wojkowitz, Oracles and Demons, 97–8).

³¹ In one variant (GML 237/19–238/4) the pursuers came nonetheless to his cave. One of them is said to have dared to enter the cave and thereafter brushed up against dPal gyi rdo rje. He recognized the murderer by his quick pulse; still he did not expose him, which this narrative uses as another element to express the faultlessness of dPal gyi rdo rje’s deed.

³² In the “Blue Annals” (DTNG/T 1:203) it is mentioned that after the murder of Langdarma, not one monk was ordained for more than seventy years in Central Tibetand Tsang; non-clerical Tantrics could continue to practice. This supports Karmay’saforementioned thesis of a certain “anti-monastic” attitude on the part of Langdarma.

³³ See BZ 84/14–5; MTPH adds a (up to now un-traced) quotation: “The Bhagavat spoke: ‘It is a light offence, to attach a sword to the trunk of a mad elephant and to let him kill other life forms,’ but it weighs heavier, ‘when one ordains somebody and does not take care of him.’” Cf. 35a1–2 (I am following Uebach’s translation).

³⁴ Tib. sgrol ba ’gyur (DTM 33/5; CBMT 439/7). The “liberation” of Langdarma is also cited in a later work by the contemporaneous hagiographer gNubs chen sangs rgyas, who did not need to expound upon the liberation ritual since Langdarma was already successfully “freed” (bsgral) (see Carmen Meinert’s article in this volume with regard to rDo rje brag gi rig ’dzin Pad ma ’phrin las, bKa’ ma mdo dbang gi bla ma brgyud pa’i rnam thar, vol. 37 (Leh: Smanrtisis shesrig spendzod, 1972), 173/4–6.

³⁵ This type of person is described as “completely delusional” (Skt. icchantika) in some texts. The killing of such person is legitimized in the Mahāparinirvāṇa-sūtra (see Yamamoto, The Mahayana Mahaparinirvana-sutra, 394–5); given the above discussed elements, it is hardly probable that there could be a parallel in the history of European thought to this teaching.

³⁶ See Wilhelm Halbfass, “Vedic Apologetics, Ritual Killing, and the Foundations of Ethics,” in Tradition and Reflection: Explorations in Indian Thought (Albany, NY: State University of New York Press, 1991); also Albrecht Wezler, “Zur Proklamation religiös-weltanschaulicher Toleranz bei dem indischen Philosophen Jayantabhaṭṭa,” Saeculum 27 (1976): 329–47, 335–6.

³⁷ For instance in the description in Bhavya’s Tarkajvālā. (Tibetan text in Halbfass, “Vedic Apologetics,” 127; translated by Ernst Steinkellner and Wilhelm Halbfass on pp. 109–10 of the same work: “When an ant has been killed in a golden vessel, being pierced with a golden needle, it is liberated from saṃsāra; and he, too, who kills it is supposed to have accumulated the seed of liberation.”) The teaching appears here, however, in a form that predates the emergence of this concept in Tantric Buddhism.

³⁸ See Rupert Gethin, “Can Killing a Living Being Ever Be an Act of Compassion? The Analysis of the Act of Killing in the Abhidhamma and Pali Commentaries,” Journal of Buddhist Ethics 11 (2004).

³⁹ See Damien Keown, “Attitudes to Euthanasia in the Vinaya and Commentary,” Journal of Buddhist Ethics 6 (1999): 260–70.

⁴⁰ See Mark Tatz, The Skill in Means (Upāyakauśalya) Sūtra (Delhi: Motilal Banarsidass, 1994, 70–1). For the various versions of this sūtra see ibid., 17–8. For a similar Tibetan and Chinese version (*Mahā-Upāyakauśalya; Ch. Da fangbian fo bao’en jing 大方便佛報恩經 , T 3, no. 156), in which the leader of a group of bandits who want to rob a trade caravan is killed compassionately, see Paul Demiéville, “Le Bouddhisme et la guerre,” reprinted in Choix d’études bouddhiques (1929–1970) (Leiden: Brill, 1973), 293 (referring to Da fangbian fo bao’en jing, T 3.161b–2a). Indeed, the narration should not be interpreted as an allegory, but rather as a simile, as Michael Zimmermann noted while commenting on an earlier version of this paper. See his treatment on the nine similes of the Tathāgatagarbha-sūtra in his A Buddha Within: The Tathāgatagarbhasūtra, The Earliest Exposition of the Buddha-Nature Teaching in India (Tokyo: The International Research Institute for Advanced Buddhology, Soka University, 2002), 34–9.

⁴¹ byang chub sems dpa’ dpa’ bo stobs ldan gyi mdzad pa ste/ ded dpon snying rje chen pos mi nag mdung thung can bsad pa bzhin no. Tucci misunderstood this sentence (see his translation: DTM/T 161).

⁴² He places the narrative, however, in a (unnamed) Jātaka, perhaps the Khurappajātaka (no. 265) whose framing story tells of the future Buddha who in his previous life as the leader of forest protectors “overcomes” [P. paharitvā; see Viggo Fausbøll, ed., The Jātaka, Together with its Commentary, vol. 2 (London: Pali Text Society, 1879), 335/23] a group of five hundred robbers in order to protect a caravan leader with five hundred wagons.

⁴³ “Wahrheit hat viele Bedeutungen” (Truth has Many Meanings), conversation with Hank Troemel, May 6, 1994, printed in Hank Troemel, ed., Theosophie und Buddhismus (Satteldorf: Adyar Verlag, 1994), 35 (translation from the German mine).

⁴⁴ Ibid., 35.

⁴⁵ In other important works, too, e.g., the Bodhisattvabhūmi, the killing of a tyrant is justified if the act occurs out of compassion or sympathy. See Lambert Schmithausen, “Aspects of the Buddhist Attitude towards War,” in Violence Denied: Violence, Non-Violence and the Rationalization of Violence in South Asian Cultural History, ed. Jan E. M. Houben and Karel R. van Kooij (Leiden: Brill, 1999), 59–60.

⁴⁶ The legitimacy of a new practice in Tantric Mahāyāna is measured by whether its innovation serves the awakening (bodhicitta); see Robert A. Mayer, A Scripture of the Ancient Tantra Collection: The Phur-pa bcu-gnyis (Oxford, Gartmore: Kiscadale, 1996), 39–41. According to Tripiṭakamāla’s Nayatrayapradīpa the killing (bsad pa) is to serve the function of placing the “deceased” in a position to realize “non-dual” meditation, and thus to embark on the path to liberation (see Peking Tanjur, rgyud ’grel, vol. 81 (Nu 6b4–26b1), Nu 6a5–7; quoted and translated in CB/T 198).

⁴⁷ See Lambert Schmithausen, “Buddhismus und Glaubenskriege,” in Glaubenskriege in Vergangenheit und Gegenwart: Referate, gehalten auf dem Symposium der Joachim-Jungius-Gesellschaft der Wissenschaften, Hamburg, am 28. und 29. Oktober 1994, ed. Peter Herrmann (Göttingen: Vandenhoeck & Ruprecht, 1996), 76.

⁴⁸ See Mayer, A Scripture of the Ancient Tantra Collection, 104–5.

⁴⁹ These are ritual handbooks or texts like the Phur ba bcu gnyis (Mahāyoga tradition); cf. Mayer, A Scripture of the Ancient Tantra Collection, 104–28, in particular 123.

⁵⁰ See René de Nebesky-Wojkowitz, Tibetan Religious Dances: Tibetan Text and Annotated Translation of the ’Chams yig [by Nag dbang blo bzang rgya mtsho, the Fifth Dalai Lama] (The Hague: Mouton, 1976), 184. The handbook for ritual dance ’Chams yig, 23b, mentions the liberative separation of the spirit from its evil form (see 34a, ibid., 222–4), which corresponds to the conception of liberation killing. The former however has been assimilated wholly symbolically.

⁵¹ See Pelliot 42, 68/4–72/4 (ninth century) in Ariane Macdonald and Yoshiro Imaeda, eds., Choix de documents tibétains conservés à la Bibliothèque Nationale complété par quelques manuscrits de l'India Office et du British Museum, 2 vols. (Paris: Bibliothèque Nationale, 1978), 1:48–61; translation and comments in Carmen Meinert’s contribution to this volume (119-21); see also Mayer, A Scripture of the Ancient Tantra Collection, 70–1 and Peter C. Verhagen, “Expressions of Violence in Buddhist Tantric Mantras,” in Violence Denied: Violence, Non-Violence and the Rationalization of Violence in South Asian Cultural History, ed. Jan E. M. Houben and Karel R. van Kooij (Leiden: Brill, 1999), 275–85.

⁵² See Cathy Cantwell, “To Meditate upon Consciousness as vajra: Ritual 'Killing and Liberation' in the rNying-ma-pa Tradition,” in Tibetan Studies: Proceedings of the 7th Seminar of the International Association for Tibetan Studies, Graz 1995, ed. Helmut Krasser et al. (Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 1997), 1:109; see also Rolf A. Stein, “Le Liṅga des danses masquées lamaïques et la théorie des âmes,” Sino-Indian Studies 5-4 (1957): 203.

⁵³ See Robert Mayer, “The Figure of Maheśvara/Rudra in the rÑiṅ-ma-pa Tantric Tradition,” Journal of the International Association of Buddhist Studies 21-2 (1998).

⁵⁴ See Mayer, A Scripture of the Ancient Tantra Collection, 106–7.

⁵⁵ The following two essays should be noted: Cantwell, “To Meditate upon Consciousness,” 112–5 and Carmen Meinert’s contribution to this volume.

⁵⁶ See Cantwell, “To Meditate upon Consciousness,” 111; Schmithausen, “Zum Problem der Gewalt,” 96; Robert Mayer and Cathy Cantwell, “A Dunhuang Manuscript on Vajrakīlaya,” Tibet Journal 18-2 (1993): 9–10.

⁵⁷ See (with reference to Bu ston’s ’Chos byung) David Seyfort Ruegg, “Problems in the Transmission of Vajrayāna Buddhism in the Western Himālaya about the Year 1000,” Acta Indologica 6 (1984): 129, and his “Deux problèmes d’exégèse et de pratiques tantriques,” in Tantric and Taoist Studies in Honour of R. A. Stein, ed. Michel Strickmann (Brussels: Institut Belge des Hautes Études Chinoises, 1981), 1:223.

⁵⁸ See Seyfort Ruegg, “Deux problèmes d’exégèse,” 1:223.

⁵⁹ See Rolf A. Stein, “La soumission de Rudra et autres contes tantriques,” Journal Asiatique 283-1 (1995): 142 (with references). Some texts, such as the Trailokyavijaya-mahākalpa-rāja, use “to kill” (bsad); ibid., 153.

⁶⁰ See Arnold van Gennep, The Rites of Passage, trans. Monika B. Vizedom and Gabrielle L. Caffee (Chicago: University of Chicago Press, 1960), 21–2; also Victor Turner, The Ritual Process: Structure and Anti-structure (Chicago: Aldine Publishing Co., 1969).

⁶¹ See Walter Burkert, Homo Necans, 2nd ed. (Berlin, New York: Walter deGruyter, 1997), 45, 49–50; see also Carmen Meinert’s contribution to this volume (107–10).

⁶² See Girard, Violence and the Sacred, 132–4.

⁶³ See Schmithausen, “Buddhismus und Glaubenskriege,” 76–7.


Note: The extension “/T” after an abbreviation signifies the page number of the indicated translation.

BDTH Fifth Dalai Lama (Ngag dbang blo bzang rgya mtsho) (1643), Bod kyi deb ther dpyid kyi rgyal mo’i glu dbyangs (Beijing: Mi rigs dpe skrun khang, 1988). Translated in Ahmad Zahiruddin, A History of Tibet, Oriental Series 7 (Bloomington: Indiana University, 1995).
BZ gSal snang. sBa bzhed (zhabs btags ma), (partly written before the ninth century, revised later). In Rolf A. Stein, ed., Une chronique ancienne de bSam-yas: sBa bžed (Paris: Institut des Hautes Études Chinoises, 1961).
CBMT Nyang Ral pa can Nyi ma’i ’od zer (ca. 1180). Chos ’byung Me tog snying po’i sbrang rtsi’i bcud (rNying ma’i chos ’byung chen mo). In Chab spel Tshe brtan phun tshogs, ed., Chos ’byung me tog snying po sbrang rtsi’i bcud (Lhasa: Bod ljons mi dmans dpe skrun khang, 1988); likewise: credited to the same author as CBMT (ca. 1200). Byan chub sems dpa’ sems dpa’ chen po chos rgyal mes dbon rnam gsum gyi rnam thar rin po che’i ’phreng ba. In Rin chen gter mdzod chen po’i rgyabs chos, vol. 7 (Paro: Ugyen Tempai Gyaltsen, 1980), 1b1–151a4.
CB Bu ston Rin chen grub (1322). bDe gshegs bstan pa’i gsal byed Chos kyi ’byung gnas. In János Szerb, ed., Bu ston’s History of Buddhism in Tibet. Critically Edited with a Comprehensive Index (Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 1990). Partial translation in Evgenij Obermiller, History of Buddhism (Chos-ḥbyung) by Bu-ston (Heidelberg: Harrassowitz, 1932).
CJG bSod nams rtse mo (1167). Chos la ’jug pa’i sgo. In Sa skya bka’ ’bum (sDe dge), Ga 263–317a.
DTM bSod nams grags pa (1478–1554). Deb ther dmar po gsar ma. Text and translation in Giuseppe Tucci, Deb t’er dmar po gsar ma (Rome: Istituto per il Medio ed Estremo Oriente, 1971).
DTNG ’Gos lo-tsā-ba gZhon nu dpal (1476–1478). Deb ther sngon po. Translation in George N. Roerich, The Blue Annals, 2 vols. (Calcutta: Royal Asiatic Society of Bengal, 1949/1953).
GML bSod nams rgyal mtshan (1368). rGyal rabs gsal ba’i me long. 2nd edition (Beijing: Mi rigs dpe skrun khang, 1988). Translated in Per K. Sørensen, Tibetan Buddhist Historiography: The Mirror Illuminating the Royal Genealogies. An Annotated Translation of the XIVth Century Tibetan Chronicle: rGyal-rabs gsal-ba’i me-long (Wiesbaden: Harrassowitz, 1994).
IOL Tibetan Dunhuang manuscripts from the former collection of the Indian Office Library. See Louis de la Vallée Poussin, Catalogue of the Tibetan Manuscripts from Tun-huang in the Indian Office Library (Oxford: Oxford University Press, 1962).
MTPH Nel-pa Paṇḍita (1283). sNgon gyi gtam Me-tog phreng ba. Text and translation in Helga Uebach, Nel-pa Paṇḍitas Chronik Me-tog phreṅ-ba (Munich: Bayerische Akademie der Wissenschaften, 1987).
T Takakusu Junjirō 高楠順次郎 , and Watanabe Kaigyoku 渡辺海旭, eds., Taishō shinshū daizōkyō 大正新修大藏經 , 100 vols. (Tokyo: Taishō issaikyō kankōkai, 1924–1932).


Ardley, Jane. Violent Compassion: Buddhism and Resistence in Tibet. Political Studies Association, 50th Annual Conference, April 10–13, 2000. Online: (accessed October 7, 2003).

Burkert, Walter. Homo Necans. 2nd ed. Berlin, New York: Walter de Gruyter, 1997.

Cantwell, Cathy. “To Meditate upon Consciousness as vajra: Ritual ‘Killing and Liberation’ in the rNying-ma-pa Tradition.” In Tibetan Studies: Proceedings of the 7th Seminar of the International Association for Tibetan Studies, Graz 1995, edited by Helmut Krasser et al., 1:107–18. Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 1997.

Fourteenth Dalai Lama (bsTan ’dzin rGya mtsho). My Land and My People [from the Tibetan]. New York: McGraw Hill Book Company, 1962.

Demiéville, Paul. “Le Bouddhisme et la guerre.” In Choix d’études bouddhiques (1929–1970), 261–99. Leiden: Brill, 1973.

Dudjom Rinpoche, Jikdrel Yeshe Dorje. The Nyingma School of Tibetan Buddhism: Its Fundamentals and History. Translated and edited by Gyurme Dorje and Matthew Kapstein. 2 vols. Boston: Wisdom Publications, 1991.

Gethin, Rupert. “Can Killing a Living Being Ever Be an Act of Compassion? The Analysis of the Act of Killing in the Abhidhamma and Pali Commentaries.” Journal of Buddhist Ethics 11 (2004): 167–202.

Girard, René. Violence and the Sacred. 2nd ed. Baltimore: Johns Hopkins University Press, 1992.

Haarh, Erik. The Yar-lung Dynasty: A Study with Particular Regard to the Contribution by Myths and Legends to the History of Ancient Tibet and the Origin and Nature of Its Kings. Copenhagen: Gad, 1969.

Halbfass, Wilhelm. “Vedic Apologetics, Ritual Killing, and the Foundations of Ethics.” In Tradition and Reflection: Explorations in Indian Thought, 87–129. Albany, NY: State University of New York Press, 1991.

Karmay, Samten Gyaltsen. The Great Perfection (rDzogs chen): A Philosophical and Meditative Teaching of Tibetan Buddhism. Leiden and New York: Brill, 1988.

———. The Arrow and the Spindle: Studies in History, Myths, Rituals and Beliefs in Tibet. Kathmandu: Mandala Book Point, 1998. Khangkar, Tsultrim Kelsang. “The Assassinations of Tri Ralpachan and Lang Darma.” Tibet Journal 18-2 (1993): 17–22.

Lauer, Robert A. Tyrannicide and Drama. Stuttgart: Franz Steiner, 1987.

Macdonald, Ariane, and Yoshiro Imaeda, eds. Choix de documents tibétains conservés à la Bibliothèque Nationale complété par quelques manuscrits de l'India Office et du British Museum. 2 vols. Paris: Bibliothèque Nationale, 1978/1979.

Mayer, Robert. A Scripture of the Ancient Tantra Collection: The Phur-pa bcu-gnyis. Oxford and Gartmore: Kiscadale, 1996.

———. “The Figure of Maheśvara/Rudra in the rÑiṅ-ma-pa Tantric Tradition.” Journal of the International Association of Buddhist Studies 21-2 (1998): 271–310.

Mayer, Robert, and Cathy Cantwell. “A Dunhuang Manuscript on Vajrakīlaya.” Tibet Journal 18-2 (1993): 2–16.

Nebesky-Wojkowitz, René de. Oracles and Demons of Tibet: The Cult and Iconography of the Tibetan Protective Deities. Graz: Akademische Druck- und Verlagsanstalt, 1975.

———. Tibetan Religious Dances: Tibetan Text and Annotated Translation of the ’Chams yig [by Nag dbang blo bzang rgya mtsho, the Fifth Dalai Lama]. The Hague: Mouton, 1976.

Paul, Robert A. The Tibetan Symbolic World: Psychoanalytic Explorations. Chicago: University of Chicago Press, 1982.

Seyfort Ruegg, David. “Deux problèmes d´éxegèse et de pratique tantriques.” In Tantric and Taoist Studies in Honour of R. A. Stein, edited by Michel Strickmann, 1:212–26. Brussels: Institut Belge des Hautes Études Chinoises, 1981.

———. “Problems in the Transmission of Vajrayāna Buddhism in the Western Himālaya about the Year 1000.” Acta Indologica 6 (1984): 369–81.

Schmithausen, Lambert. “Buddhismus und Glaubenskriege.” In Glaubenskriege in Vergangenheit und Gegenwart: Referate, gehalten auf dem Symposium der Joachim-Jungius-Gesellschaft der Wissenschaften, Hamburg, am 28. und 29. Oktober 1994, edited by Peter Herrmann, 63–92. Göttingen: Vandenhoeck & Ruprecht, 1996.

———. “Aspects of the Buddhist Attitude towards War.” In Violence Denied: Violence, Non-Violence and the Rationalization of Violence in South Asian Cultural History, edited by Jan E. M. Houben and Karel R. van Kooij, 45–67. Leiden: Brill, 1999.

———. “Zum Problem der Gewalt im Buddhismus.” In Krieg und Gewalt in den Weltreligionen: Fakten und Hintergründe, edited by Adel Theodor Khoury, Ekkehard Grundmann and Hans-Peter Müller, 83–98. Freiburg, Basel, Vienna: Herder, 2003.

Schmithausen, Lambert, and Mudagamuwe Maithrimurthi. “Tier und Mensch im Buddhismus.” In Tiere und Menschen: Geschichte und Aktualität eines prekären Verhältnisses, edited by Paul Münch and Rainer Walz, 179–224. Paderborn, Munich, Vienna, Zurich: Ferdinand Schöningh, 1998.

Shakabpa, Tsepon Wangchuk Dedhen. Tibet: A Political History. New Haven: Yale University Press, 1984.

Stein, Rolf A. “La soumission de Rudra et autres contes tantriques.” Journal Asiatique 283-1 (1995): 121–60.

Troemel, Hank, ed. Theosophie und Buddhismus. Satteldorf: Adyar Verlag, 1994.

Turchetti, Mario. Tyrannie et tyrannicide de l'Antiquité à nos jours. Paris: Presses Universitaire de France, 2001.

Turner, Victor. The Ritual Process: Structure and Anti-structure. Chicago: Aldine Publishing Co., 1969.

Yamaguchi, Zuiho. “The Fiction of King Dar ma’s Persecution of Buddhism.” In Du Dunhuang au Japon: études chinoises et bouddhiques offertes à Michel Soymié, edited by Jean-Pierre Drège, 231–58. Geneva: Droz, 1996.

Zimmermann, Michael. A Buddha Within: The Tathāgatagarbhasūtra, The Earliest Exposition of the Buddha-Nature Teaching in India. Tokyo: The International Research Institute for Advanced Buddhology, Soka University, 2002.


This paper was published in Buddhism and Violence / ed. by Michael Zimmermann with the assistance of Chiew Hui Ho and Philip Pierce. – Kathmandu : Lumbini International Research Institute, 2006, pages 131–157. ISBN 99946-933-1-X.

Header image: Tibetan king Lang Dharma (Drag Yerpa). © Elke Hessel


More Papers by Jens Schlieter
More Papers about Violence in (Tibetan) Buddhism
PDF of Jens Schlieter’s paper