Illustration: Tomb of the Unknown Samaya-Breaker
People who enter Tibetan Buddhist tantra are frequently warned about samaya, the sacred commitments created through tantric empowerment. They may be told that breaking samaya with the vajra master can lead to terrible karmic consequences, illness, obstacles, or rebirth in Vajra Hell. What they are much less likely to be told is that Tibetan tantric literature also contains another side of the samaya system: traditions in which killing itself can become a tantric commitment, and in which samaya-breakers are explicitly identified as targets of wrathful ritual.
This is not an accusation constructed from hostile accounts of Tibetan Buddhism. It appears in Buddhist tantric texts, Tibetan biographies, ritual manuals, and the work of contemporary scholars of Tibetan religion.
One of the clearest examples comes from the life of the eleventh-century translator and tantric master Rwa Lotsawa Dorje Drak, famed for his mastery of Vajrabhairava, a wrathful form associated with Yamāntaka.
According to Rwa Lotsawa’s Tibetan biography, an abbot named Skyo ’Dul ’dzin publicly accused him of behavior unbecoming a Buddhist teacher. Among the accusations was the charge that Rwa used black magic to kill people.
Rwa did not simply deny it. He reportedly answered:
If I kill people, I do it only to those who are samaya-breakers. This is a tantric commitment.”
Rwa then declared that the abbot himself had become a samaya-breaker because of his wrong views and predicted that he would “depart” after seven days. According to the biography, seven days later the abbot became paralyzed and died. [1]
This story is preserved in the Tibetan biographical tradition and discussed by scholar Aleksandra Wenta in her study of Yamāntaka and ritual conflict. Wenta traces the episode both to the biography of Rwa Lotsawa written by his grandnephew, Rwa Yeshé Sengé, and to a later seventeenth-century history of the Vajrabhairava-Yamāntaka cycle. [2]
It would be easy to dismiss such a story as medieval legend. The problem is that Rwa’s statement about killing as a “tantric commitment” has a much larger textual context.
When not killing becomes the transgression
Jacob Dalton, a leading scholar of Tibetan tantric history, documents one of the clearest examples in The Taming of the Demons: Violence and Liberation in Tibetan Buddhism.
Dalton discusses a tantric confession text attributed to King Indrabhūti called the Confession for the Root and Branch Downfalls of the Vajrayāna. The practitioner does not confess to having killed someone. He confesses to having failed to do so.
Dalton translates the confession:
I confess that I have not annihilated and been loving toward those persons who do not believe in, or who despise and deprecate, the lama, the three jewels, and the Vajrayāna of Secret Mantra.”
Dalton explicitly describes this as a confession by practitioners who had neglected their vows to kill. [3]
The wording deserves close attention. The practitioner confesses that he has failed to “annihilate and been loving toward” those who reject or disparage the lama, the Three Jewels, and Secret Mantra Vajrayāna. The words “annihilate” and “loving” appear together.
This exposes one of the most troubling features of the tantric justification for ritual violence. Destruction can be recast as benevolence toward the person being destroyed. Within that moral logic, killing need not be understood as hatred of an enemy. It can instead be presented as a compassionate or loving tantric act.
Once violence is defined as compassion, the ordinary moral objection to it has been effectively reversed. The practitioner can believe that destroying an opponent is compatible with benevolent intention, while refusing to perform the required destructive act can become something requiring confession.
Dalton’s evidence raises a more serious issue than whether tantric descriptions of killing are “symbolic” or “literal.” It shows a religious framework capable of redefining destructive action as spiritually beneficial action. The person performing it may understand himself not as violating Buddhist ethics, but as fulfilling a higher tantric obligation.
For someone designated a samaya-breaker, that distinction would offer little reassurance. The very category that ordinarily protects another person from violence–compassion–has been incorporated into the justification for the violence itself. This is extraordinary when you think about it. It is also difficult to reconcile with the common modern presentation of tantric violence as nothing more than symbolic imagery representing the destruction of ego, anger, ignorance, or other internal obstacles.
Dalton is careful about the historical evidence. Scholars cannot know how often practitioners enacted such prescriptions literally. Tantric commentators sometimes interpreted violent commands metaphorically, and Dalton considers it likely that both approaches existed: many understood them figuratively, while some practitioners performed violent rites literally. [4]
But the textual point remains. The confession presupposes circumstances in which failing to perform destructive action constitutes a tantric failure. That is far removed from a simple prohibition against killing.
The older Mahāyoga samayas
Sam van Schaik finds the same reversal in early Tibetan Mahāyoga literature preserved among the Dunhuang manuscripts.
In “The Limits of Transgression: The Samaya Vows of Mahāyoga,” van Schaik examines early systems of tantric commitments in which older Buddhist prohibitions were transformed. Some Mahāyoga samaya systems contained positive injunctions concerning actions normally prohibited by Buddhist morality. Killing was among them. [5]
Van Schaik’s research is especially important because these are not late Tibetan legends about eccentric or “crazy wisdom” yogis. They are early manuscript witnesses to the formation of Tibetan tantric practice.
The samaya system established what a tantric practitioner was obligated to maintain. Under particular ritual and doctrinal conditions, practices normally classified as grave wrongdoing could become components of tantric discipline itself.
Dalton describes the same transformation in Mahāyoga. He notes that ritualized sex and violence became central elements in this literature and points to the Guhyasamāja Tantra’s deliberate inversion of ordinary Buddhist rules. [6]
The moral logic had been reversed. Under particular tantric conditions, killing could cease to be treated as the violation of a vow. Failing to perform it could itself become the violation.
“Liberation” really does mean killing in these instances
Much turns on the Tibetan word sgrol ba, usually translated as “liberation.” In Buddhist usage, “liberation” usually refers to enlightenment or freedom from samsaric suffering. In some wrathful rites, however, the same word functions as a euphemism for killing.
Modern explanations can create the impression that these texts are speaking only metaphorically, but scholars working directly with the ritual literature have been explicit about what the term can signify.
At the 2025 Congress of the International Association of Buddhist Studies, Sonam Choden presented research on the tantric categories known as the “seven transgressors” and the “ten fields of liberation.” The published conference abstract describes “liberation” in this context as “a euphemism for killing.” [7]
At the same conference, Bruno Galasek-Hul presented research on ritual violence in the canonical Yamāntaka corpus. According to his abstract, Rwa Lotsawa states in his commentary on the Vajramahābhairava Tantra that killing, bsad pa in Tibetan, is “the chief practice of this tantra and its deity-yoga.” Galasek-Hul also argues that Tibetan traditions understood at least some liberation rites as actual lethal rites, rather than solely as allegories for overcoming internal afflictions. [8]
Rwa Lotsawa’s statement that killing samaya-breakers was “a tantric commitment” therefore did not arise in a vacuum. The Vajrabhairava tradition he transmitted contained rituals of killing, and Rwa himself described killing as central to the tantra’s ritual activity.
Galasek-Hul is a scholar and translator of Sanskrit and Classical Tibetan Buddhist literature who has translated Yamāntaka material preserved in the Tibetan Kangyur for 84000. [9]
Who may be “liberated” or tantrically killed?
Tantric literature did not leave the potential targets entirely undefined. Tibetan traditions developed categories known as the “ten fields of liberation,” classifications of beings against whom destructive tantric rites could be justified.
Different versions of the list exist. They include enemies of the Three Jewels, enemies of the spiritual master, enemies of tantric practitioners, people who harm sentient beings, and people who have violated sacred commitments. One scholarly discussion translates one category as a person “who has depraved his pledges.” Another study, discussing Tāranātha’s account, describes the ten fields as categories of people who could be considered proper targets of rites intended to kill for a claimed greater good. [10]
The tradition thus developed a moral taxonomy for determining who could become the object of destructive tantric activity. That should be deeply troubling in any religious system in which a human guru may exercise authority in determining who has broken samaya and thus who can be ritually killed.
A person can first be judged to have violated the sacred relationship. That classification can then place the person within a ritual framework in which wrathful action against them becomes religiously intelligible, even necessary. The potential for abuse is obvious.
The killing language survives into modern Tibetan liturgy
This vocabulary did not disappear with medieval Tibet. A Vajrakīlaya liberation rite by the twentieth-century Nyingma master Khenpo Jigme Phuntsok addresses “enemies and samaya-breakers,” calls upon spiritual powers to judge them, and requests “the most severe punishment” for malevolent foes. [11]
A Vajrakīlaya practice composed by Dudjom Rinpoche similarly invokes victory over “negativity, enemies and samaya-breakers.” Jamyang Khyentse Chökyi Lodrö describes oath-bound Dharma guardians assembled and “eager to act and liberate samaya-breakers, enemies and obstructors.” [12]
Mipham Rinpoche’s ritual poetry can be considerably more graphic. In one invocation, a protector is called the “stealer of the lives of all samaya-breakers.” The text describes piercing their hearts, cutting the main artery of an enemy’s life-force, and offering mountains of flesh, oceans of blood, and banks of bones from samaya-breaking enemies. [13]
It is possible to read such liturgies symbolically like some modern scholars and translators do, but it is not reasonable to pretend that the literal language of killing is absent. Is it willful ignorance on the part of scholars who won’t admit that these practices are real, have been perpetrated against human beings, and continue into the present day? Is it because they have too much invested to look at the language honestly?
The oath-bound protectors
The practitioner is not the only being described as being bound by samaya. Wrathful Dharma protectors are also portrayed as oath-bound. They are ritually commanded to protect the teachings, the lama, and tantric practitioners, and to act against enemies and those who violate samaya.
Within the tantric worldview, these protectors are treated as spiritual beings capable of inflicting pain and harm, rather than merely as psychological symbols. This creates a reciprocal structure of sacred obligation: the disciple is bound to the tantric master and mandala, while protectors are bound to defend that system.
The person identified as a samaya-breaker can then be positioned as the object against whom those protectors act. Phrases such as “liberate samaya-breakers” are not random poetic decorations. They occur within a religious system of vows, offenses, protectors, and ritual consequences.
Early Tibetan samaya material could be extremely explicit about those consequences. One Dunhuang text translated by Dalton warns that if the tantric bond is transgressed, buddhas, bodhisattvas, vajra protectors, gods, humans, siddhas, and yogins will become enraged and draw blood from the offender’s corrupted heart. [14]
This doesn’t sound like peaceful Buddhist language, does it? There are probably many modern practitioners of tantric Buddhism who think that the historical Buddha’s injunctions against harming oneself or others will protect them from tantric retribution. Sadly, this just isn’t so.
It should be emphasized these texts make clear that the tantric vows to “kill” or destroy samaya breakers supercede the moral injunctions against killing found in early Buddhism.
The threat of Vajra Hell is also very real
The consequences attributed to broken samaya extend beyond destructive rituals performed by human practitioners. Traditional Tibetan teachings repeatedly threaten the samaya-breaker with Vajra Hell.
Longchenpa quotes tantric sources warning that someone who disparages the master will experience suffering in this life through sickness, poison, harm from ḍākinīs, and violent forces, followed by death and descent into hell. [15]
Chögyam Trungpa used even starker language when teaching Western students about samaya. He compared it to a highly electrified fence. Attempting to escape it, he said, leads to destruction and Vajra Hell, a condition in which there is “no remedy and no freedom at all.” [16]
A teaching presented as Padmasambhava’s advice to Yeshe Tsogyal warns that breaking the vajra master’s command leads to “Unceasing Vajra Hell” with “no chance for liberation.” [17]
These statements should be placed beside the more inviting descriptions of Vajrayāna empowerment routinely presented to Western students. Whatever one believes about the literal existence of Vajra Hell, the threat itself is part of the tradition.
What happens to the person who is “liberated” (tantrically killed)?
Here is where I disagree with the reassuring explanation sometimes offered for these annihilation rituals. One interpretation I was given by a Tibetan lama many years ago says that an exceptionally accomplished tantric master kills only out of compassion and then transfers the consciousness of the victim to a better rebirth or an enlightened state.
Buddhist sources genuinely employ this logic. Dalton documents traditions in which killing is explained as compassionate liberation. He also describes the Rudra myth, in which Rudra is destroyed, purified, and subsequently reconstituted as a protector. [18]
But that explanation does not account for every text discussed here. The confession translated by Dalton speaks of “annihilating” those who reject or disparage the lama, Three Jewels, and Vajrayāna. Rwa Lotsawa’s biography describes the death of a critic after Rwa identifies him as a samaya-breaker. The ten fields of liberation define classes of people who may become targets of destructive rites.
Modern protector liturgies call for the lives of samaya-breakers to be taken, while traditional samaya teachings threaten violators with Vajra Hell. It therefore distorts the range of evidence to reduce the entire tradition to a benevolent method for giving someone a better rebirth.
My own understanding is also shaped by personal experience. After many years within Tibetan Vajrayāna and extensive involvement in its advanced tantric practices, I do not experience the language of “liberation” in this context as an innocent synonym for compassion. I view the activity surrounding broken samaya, wrathful deities, and spiritual punishment as threatening destruction, not rescue.
That is my testimony and interpretation. Historical scholarship cannot independently prove what happens metaphysically to a person’s consciousness after death, nor can it establish that consciousness can literally be annihilated forever. But it does establish something significant: the tradition contains an extensive ritual vocabulary in which the death of a human adversary can be described as “liberation,” the killing of certain people can be morally justified, failure to kill can appear as a tantric fault, and the person accused of broken samaya can become an object of wrathful activity.
Rwa Lotsawa was not an isolated aberration
The temptation will always be to turn Rwa Lotsawa into an embarrassing exception, but the evidence does not permit that easy solution.
Wenta’s research places Yamāntaka within a much larger history of tantric ritual directed against enemies and used in political or religious conflict. It also shows that violent tantric practices produced controversy within Tibetan Buddhist society and prompted efforts to regulate their use.[19]
Cathy Cantwell and Robert Mayer’s Early Tibetan Documents on Phur pa from Dunhuang demonstrates that violent Vajrakīlaya ritual material existed at an early stage of Tibetan tantra. Their work compares Dunhuang manuscripts with the later received Nyingma tradition and finds significant continuity between them. [20]
Dalton’s study similarly demonstrates that violent themes were deeply involved in Tibet’s assimilation of tantric Buddhism during the period following the Tibetan Empire. His book includes translations of early Dunhuang liberation rites. [21]
David Gray’s research on the Cakrasaṃvara Tantra places another major Highest Yoga Tantra tradition within the antinomian setting of medieval Indian tantra and examines the transgressive practices associated with that textual and ritual world. [22]
Ritual killing therefore occupied a genuine place within tantric literature and religious imagination. Tibetan authors debated its legitimate use, and some celebrated masters were remembered precisely for having exercised such powers. This does not establish how often ritual prescriptions resulted in physical killings. It does establish that the violent material cannot be dismissed as an isolated aberration or a purely modern misreading.
What should students be told before empowerments and other tantric commitments?
This history creates a serious question of informed consent. People receiving Highest Yoga Tantra or Mahāyoga empowerments are entering practices in which samaya is fundamental to the relationship established among disciple, guru, deity, and mandala.
Students may hear that samaya is sacred and that devotion to the guru is essential. They may hear that Vajrayāna is dangerous if practiced incorrectly. But how many are told before empowerment that Tibetan sources threaten samaya-breakers literally with ritual death and Vajra Hell?
How many are shown ritual texts calling upon protectors to take the lives of samaya-breakers? How many are told that the tradition developed categories of people considered legitimate “fields of liberation,” meaning targets of tantric killing?
When I began practicing Tibetan Buddhism, the wrathful protector texts were not allowed to be translated into English. The violent material was kept secret or phrased symbolically.
People cannot meaningfully consent to religious obligations when important dimensions of those obligations are concealed, minimized, or explained only after the commitments have already been undertaken. They may learn what the system says and does about broken commitments only after they are judged to have broken them. By then, it is far too late.
The wrathful material exists within the tradition, has existed for centuries, and concerns the consequences of entering and breaking precisely the commitments created through tantric initiation. Students have a right to know that before they enter the mandala in the first place.
Notes
David B. Gray, The Cakrasaṃvara Tantra: The Discourse of Śrī Heruka: A Study and Annotated Translation (New York: American Institute of Buddhist Studies at Columbia University, 2007); David B. Gray, “The Cakrasamvara Tantra: Its History, Interpretation, and Practice in India and Tibet,” Religion Compass 1, no. 6 (2007): 695–710, https://doi.org/10.1111/j.1749-8171.2007.00046.x. Journal record.
Rwa Yeshé Sengé, The All-Pervading Melodious Drumbeat: The Life of Ra Lotsawa, trans. Bryan J. Cuevas (New York: Penguin Classics, 2015), 241–42. Publisher’s page.
Aleksandra Wenta, “Tantric Ritual and Conflict in Tibetan Buddhist Society: The Cult of Yamāntaka,” LEA: Lingue e Letterature d’Oriente e d’Occidente 11 (2022): 333–53, especially 341–43, https://doi.org/10.36253/lea-1824-484x-13637. Open-access article.
Jacob P. Dalton, The Taming of the Demons: Violence and Liberation in Tibetan Buddhism (New Haven, CT: Yale University Press, 2011), 25–26. Publisher’s page.
Dalton, Taming of the Demons, 26–31.
Sam van Schaik, “The Limits of Transgression: The Samaya Vows of Mahāyoga,” in Esoteric Buddhism at Dunhuang: Rites and Teachings for This Life and Beyond, ed. Matthew T. Kapstein and Sam van Schaik (Leiden: Brill, 2010), 61–80. Author-hosted text.
Dalton, Taming of the Demons, 21–31.
Sonam Choden, “Who Deserve to Be ‘Liberated’? The Buddhist Tantric Ideas of the ‘Seven [Kinds of] Transgression/Transgressor’ and ‘Ten Fields of Liberation,’” abstract for the XXth Congress of the International Association of Buddhist Studies, Leipzig University, 2025, in Abstracts, 121–22. Conference abstract volume.
Bruno Galasek-Hul, “Ritual Violence in the Yamāntaka Corpus of the Kangyur,” abstract for the XXth Congress of the International Association of Buddhist Studies, Leipzig University, 2025, in Abstracts, 122–23. Conference abstract volume.
Bruno Galasek-Hul, trans., The Tantra of Black Yamāri, Toh 467, 84000: Translating the Words of the Buddha, 2023. Translation and translator information.
Aleksandra Wenta, “Demons, Wicked Ones and Those Who Violate the Samayas: Dharma Against the Enemy in Tantric Buddhism,” in Dharma and Abhidharma, ed. C. Pieruccini and A. Rigopoulos (Firenze: Firenze University Press, 2022), and Wenta, “Tāranātha on the Emergence of the Tantric Cycle of Vajrabhairava-Yamāntaka: Writing a Tibetan Buddhist Historiography in Seventeenth-Century Tibet,” Revue d’Etudes Tibétaines 56 (2021): 118–65. University repository record; open-access RET article.
Khenpo Jigme Phuntsok, “Purba Gulkhukma Liberation Rite,” trans. Lotsawa House, 2020. English translation.
Dudjom Rinpoche, “The Essential Daily Practice of Vajrakīlaya,” trans. Lotsawa House, and Jamyang Khyentse Chökyi Lodrö, “Shower of Wisdom Blessings,” trans. Lotsawa House. Dudjom Rinpoche translation; Jamyang Khyentse Chökyi Lodrö translation.
Mipham Rinpoche, “Söllo Chenmo: The Great Invocation of Gesar,” trans. Lotsawa House. English translation.
Jacob P. Dalton, “How Dhāraṇīs WERE Proto-Tantric: Liturgies, Ritual Manuals, and the Origins of the Tantras,” in Dhāraṇī and Pratibhāna: Memory and Eloquence of the Bodhisattvas, ed. Claudio Cicuzza (Vienna: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien, 2022), 307–35. Open-access chapter.
Longchen Rabjam, “The Ocean of Siddhis,” trans. Adam Pearcey, Lotsawa House, 2018. English translation.
Chögyam Trungpa, “The Crazy-Wisdom Holder and the Student,” in The Lion’s Roar: An Introduction to Tantra, ed. Judith L. Lief (Boston: Shambhala, 1992), 151–56. Digital text.
Padmasambhava, Dakini Teachings: Padmasambhava’s Oral Instructions to Lady Tsogyal, trans. Erik Pema Kunsang (Boudhanath: Rangjung Yeshe Publications, 1989), section on samaya and guru yoga. Publisher’s page.
Dalton, Taming of the Demons, 3–20, 31–50.
Wenta, “Tantric Ritual and Conflict,” 333–53.
Cathy Cantwell and Robert Mayer, Early Tibetan Documents on Phur pa from Dunhuang (Vienna: Verlag der Österreichischen Akademie der Wissenschaften, 2008). Publisher-hosted text.
Dalton, Taming of the Demons, especially 51–103.
