Do Aghoris really eat human flesh?

There is a figure the guidebooks to Varanasi warn you about. He sits where the cremation fires never go out, smeared with the grey ash of the dead, and he may raise to his lips a bowl that was once a human skull. The documentary that filmed him gave the segment a one-word title, and the word does most of the work before you have seen anything. Cannibal.

The word is doing too much. It carries a crime inside it, the killing of a person for food, and that crime is exactly the part the record does not support. What the scholarship documents is narrower and stranger, and to see it clearly you have to pull apart three things the single word has fused into one: a crime that is real but is not theirs, a rite that is reported of them and is not a crime, and an argument about God that explains why any of it is thinkable at all.

The word smuggles in a crime

Start with the word, because the confusion is built into it. Cannibalism means the eating of human flesh by a human being. It does not mean, and has never meant, that the eater did the killing. Anthropology sorts it by who was eaten and why, and several of its recognised forms involve people nobody murdered: the dead of one’s own group eaten at a funeral, the dead eaten in famine, human tissue taken as medicine. So the precise answer to the question in the title is yes, in a narrow and rare sense, and the work only begins after that yes.

What the word carries in ordinary use, and what the documentary meant by it, is something else. It carries a killing. On that charge the honest answer is short. There is no evidence for it.

The lurid version of the Aghori is older than the Aghori. It is inherited from the Kapalikas, the “skull-bearers,” the medieval Shaiva ascetics the Aghor tradition is usually described as succeeding, and from the way their neighbours wrote about them. How direct that succession is remains genuinely open, and our longer study of the Aghoris sets out the gap in the record rather than papering over it. Our most vivid pictures of the Kapalikas are not their own scriptures. They left none that survive, and the historian David Lorenzen, whose 1972 study The Kapalikas and Kalamukhas remains the standard account, had to reconstruct them almost entirely from the words of their enemies. Indian scholarship reaches them by a steadier road: R. G. Bhandarkar’s 1913 survey of the minor Shaiva systems, and V. S. Pathak’s History of Saiva Cults in Northern India, built in 1960 out of inscriptions, which lets a reader see what the epigraphic record attests as against what hostile Sanskrit drama alleges. Those words are Sanskrit plays: villains garlanded in human bone who abduct a heroine to sacrifice her at a cremation-ground shrine. The plays are real evidence, but of a particular kind. They tell us what the skull-bearer looked like to a frightened public. They are not a record of what he did. The Encyclopaedia Britannica, summarising the specialist consensus on the Kapalikas, states it flatly: the esoteric rites are said to have included human sacrifice, “though there is no evidence for the latter.”

So set the killing aside. It is not there.

What the record actually shows

Now the harder half, the part that is true, and the part that needs the most care.

The earliest sustained Western account is a colonial one. In 1893, H. W. Barrow read a long paper, “On the Aghoris and Aghorapanthis,” to the Anthropological Society of Bombay, and it remains the baseline that later scholars cite. It is a Victorian document, built on hearsay and a reformer’s horror, and it should be read as proof that the reputation was already fixed by the nineteenth century, not as proof of what was routinely done. Look at the shape of the colonial testimony and the point makes itself, though not quite in the way a hostile reader would expect. Henry Balfour, exhibiting an Aghori’s skull-bowl to the Anthropological Institute in 1897, printed a description sent to him by a Dr Drake Brockman, who had the life-history from the man himself and who says plainly that the rest of it came from asking pundits: the Aghori will eat “offal and filth of every description,” and “often human flesh when obtainable.” Russell’s 1916 volume retells a story he takes from Barrow, of a landowner at Allahabad who refused a piece of human flesh an Aghori offered him and was pelted with it, and who afterwards grew so rich he could not store his wealth. That one is a boon-legend, and the flesh in it is refused. Crooke, in 1896, passes on a report in which someone “saw the Faqir cut off and eat a piece of the flesh of a corpse.” Three items and three different kinds of thing: a characterisation of a sect gathered by enquiry, a miracle story, and a single claim of sight. Not one of them is the author’s own observation. That distinction, between a reputation and a practice, turns out to be the whole of the matter.

The closest and most careful study is modern. The anthropologist Jonathan Parry spent years at the burning grounds of Banaras, and his essay “Sacrificial Death and the Necrophagous Ascetic” and his book Death in Banaras are where the subject is taken seriously. Parry treats the necrophagy as real enough to think with. He does not claim to have watched it. Writing in 2008, introducing another scholar’s book, he set the catalogue down himself with one careful word: the practices “supposedly include coprophagy and necrophagy and the consumption of other highly polluting substances.”

That word is doing honest work, and Ron Barrett, a medical anthropologist who spent his own years with the Aghor lineage, says Parry should have done more of it. Barrett’s complaint is worth quoting exactly, because it is the complaint this article exists to answer. Parry, he writes, “could have more clearly distinguished contemporary from historical accounts as well as the testimony of nondevotee observers from those of practicing disciples.” Barrett adds, fairly, that none of Parry’s data are inaccurate in themselves. The trouble is not that anyone was told something false. It is that four different kinds of knowing were laid side by side without labels.

So hold the two halves apart from here on. What is eaten is the already dead. What is not in evidence, from any of these sources, is the killing.

The distinction is not a technicality. Banaras does not cremate everyone. Certain bodies are held too pure or too impure for the fire and are given instead to the river: renunciates, small children, victims of smallpox and leprosy, the very poor whose families cannot afford wood. They are weighted and set adrift, or left at the ghat. The corpse in these accounts is one of these, unclaimed and ownerless, already surrendered by everyone else. What is described would revolt almost anyone. What is never described, in any of it, is the making of the corpse.

There is a logic even inside the horror, which is the surest sign that what is being described is a ritual and not an appetite. The Aghori is said to prize a particular kind of skull: the skull of someone who died an untimely death, a young man killed by snakebite being the example that recurs, and who therefore, in the tradition’s reckoning, never received kapal-kriya, the ceremonial cracking of the skull on the pyre that releases the pran, the breath of life. In such a skull, the belief runs, the power is still sealed inside, and can be turned to uses that include countering the effects of poisons, which fits the snakebite neatly. Three clinicians set this reasoning out in 2013 in a short cover piece for a paediatric neurosurgery journal, and it is worth saying plainly that the piece is a literature review rather than fieldwork, and that its only ethnographic source is Parry, whom it cites for all of it. So that is one source, not two. What it describes is a system rather than a frenzy, and that is all it establishes. These are claims about what a skull holds, made by the people who hold them. They are not findings about skulls.

Here is the point the documentaries never reach. Almost none of this is witnessed. The acts are claimed, feared, whispered about, and very rarely seen by anyone who then writes it down. That is not a caveat to add at the end. It is the condition of the entire archive. Anthropology has a long argument about this problem, and William Arens pushed it hardest: reports of cannibalism have always travelled further and faster than cannibalism. Arens took it to a conclusion most anthropologists reject, that the practice is never adequately documented anywhere, and the funerary cannibalism described at the top of this piece is among the cases that refute him. The narrower observation survives him all the same. The Encyclopaedia Britannica makes the same observation about the neighbouring charge, that accusations of human sacrifice have been far more widespread than the practice ever was, and notes that the early Christians were themselves accused of eating human flesh by people who had misunderstood the Eucharist.

How often the corpse-eating happens, whether it is once in a life or something rarer or something less rare, the record does not tell us, and this piece will not pretend otherwise. What the record does show is that the talk is doing work. The Aghori’s power over ordinary people, his ability to draw alms and awe and a wide berth, runs on the reputation. The talk was always doing work.

The crime that is not this

There is one confusion that has to be cleared without hedging, because it is where real harm is done.

Open an Indian newspaper often enough and you will find a story of “human sacrifice”: a child taken, a killing staged for a promised fortune or a cure. They are real, and they are monstrous, and they are murders, prosecuted as murders. They are also not this.

The counting deserves a moment, because the number gets thrown about. India has no offence called ritual killing. The National Crime Records Bureau records human sacrifice only as a motive, one row in the table of motives for murder in its annual Crime in India. Read that row and the national figure is small. The Deccan Herald, compiling it, counted 103 such killings in the whole country between 2014 and 2021, single digits to the mid-twenties in a year against roughly thirty thousand murders. It is also, almost certainly, an undercount. The Maharashtra Andhashraddha Nirmoolan Samiti, the rationalist organisation Dabholkar founded, counted at least seven such murders in that state between late 2013 and 2016, against the two the Bureau recorded.

The men prosecuted for these killings are, on the record of the cases themselves, self-styled tantriks and black-magic men selling wealth and sons and cures to the desperate. Whether any one of them ever stood in some relation to some Aghor lineage is not a thing this article can know, and it does not need to know it. The claim worth making is the one the evidence carries and no more. There is no reliable evidence that human sacrifice is a rite of the Aghor tradition. No text of the tradition prescribes it. No ethnographer who has lived with the lineage reports it. A crime committed by a man who calls himself something is not a practice of the thing he calls himself.

The confusion is not helped by the law itself. Maharashtra’s anti-superstition statute carries the official name “The Maharashtra Prevention and Eradication of Human Sacrifice and other Inhuman, Evil and Aghori Practices and Black Magic Act.” The rationalist campaigner Narendra Dabholkar had pushed for that law for eighteen years and could not get it passed. He was shot dead on a morning walk in Pune on 20 August 2013. The cabinet cleared the measure the following day, and the ordinance was promulgated six days after his murder.

The word aghori sits there in the title, used as everyday Hindi and Marathi use it, as a common noun for the ghoulish and the occult. That common noun is not the proper noun, and the statute proves the point better than any argument could. Section 2 defines “human sacrifice and other inhuman, evil and aghori practices and black magic” as meaning, precisely and exhaustively, the commission of any act described in the Schedule. The Schedule then lists twelve acts: assaulting a person under pretext of expelling a ghost, defrauding by display of so-called miracles, preventing a snakebite victim from reaching medical treatment, sexual exploitation by a man claiming supernatural power, and so on down the list. Every item is an act. Not one is a belief, an affiliation or an order. Section 12 goes further and expressly protects religious observance, down to pradakshina and kirtan and the piercing of a child’s ears, and the only religious groups the law names anywhere, the Varkaris and the Jains, are named there, in the clause that exempts them. Dabholkar answered the charge that the bill was anti-religious by pointing out that it does not mention any god or religion and targets only fraudulent practices. So the phrase in the title is a label for a list of crimes. The Aghor Panth, the ascetic order seated at Kina Ram’s math in Varanasi, is a lineage one of whose branches runs a free leprosy ashram. Reading the one for the other is the commonest way the tradition is libelled.

Why a corpse at all

If it is not appetite and not murder, what is it for? The answer is a piece of theology, and it is the same one that explains why the Aghori lives in the cremation ground and drinks from a skull in the first place.

Begin with the name, because the name is the argument in miniature. Aghora is built from a-, “not,” and ghora, “terrible.” It means “the non-terrible,” and it is not the sect’s invention. It is one of the five faces of Sadashiva, the southern face of dissolution and death, and the paradox is deliberate: the most fearsome aspect of God is given the name that says, do not be afraid. The Vedic liturgy bows to exactly this, and it does so in a mantra addressed to that same southern face. It sits in the tenth book of the Taittiriya Aranyaka, inside the block of five Brahma-mantras, one for each face, which is also transmitted as the Mahanarayana Upanishad.

अघोरेभ्योऽथ घोरेभ्यो घोरघोरतरेभ्यः । सर्वतः शर्व सर्वेभ्यो नमस्ते अस्तु रुद्ररूपेभ्यः ॥

aghorebhyo’tha ghorebhyo ghoraghoratarebhyaḥ · sarvataḥ śarva sarvebhyo namas te astu rudrarūpebhyaḥ

“Now, O Śarva, my salutations be at all times and all places to Thy Rudra forms, benign, terrific, more terrific and destructive.” (tr. Swami Vimalananda)

Taittirīya Āraṇyaka 10.45 · the same mantra is numbered 19.1 in the eighty-anuvāka Āndhra division printed as the Mahānārāyaṇa Upaniṣad, and 45 in the sixty-four-anuvāka Drāviḍa division that Sāyaṇa comments on; a variant second line reads sarvebhyaḥ sarva śarvebhyo #

It is worth saying what this mantra is not. It is no part of the Shri Rudram, which is Taittiriya Samhita 4.5 and 4.7, and which it is often taken for. Sayana, glossing the line, takes the three terms as the three strands of nature. The aghora forms are the calm ones, the ghora the fierce, the ghoratara fiercer still, and the salutation goes to all of them at once, in all places and at all times. The tradition had, long before the Aghori, a name for the discovery that the fearful and the gracious are one thing. The Aghori worships Shiva in that discovery, most often as Bhairava, the fierce form whose name the tradition glosses as the destroyer of fear.

The doctrine underneath is a severe non-dualism. If there is one reality and that reality is Shiva, then it is present without exception, in the temple and the corpse alike, in the flower and the ash, in what the world calls pure and in what the world calls filth. The human sorting of the world into clean and unclean is, on this view, provisional, useful for daily life and false at the deepest level. To feel genuine disgust, genuine dread, is to still believe in your gut that the world is really divided, that some things are God and some are not. Kaula tantra speaks of pasha, the nooses that bind the soul, and the Kularnava Tantra counts eight of them, naming the things one would least expect a scripture to call fetters: revulsion, doubt, fear, shame, aversion, and then family, propriety and birth (13.90). This is a different use of the same word from the pasha of Shaiva Siddhanta, which means metaphysical impurity and karma and maya. Here the nooses are the decencies themselves. None of the Aghor texts available in translation invokes that list, and it would be tidy to pretend otherwise; what the ethnographers report is the same idea in the tradition’s own vocabulary, which Barrett, whose chapter on it is called “Death and Nondiscrimination,” renders as non-discrimination: the position that the world’s sortings into clean and unclean are surface, and that the practice is aimed at the ascetic’s own habit of sorting. The severe practice is aimed at cutting those. The ascetic goes to the shmashana, the cremation ground, and handles the one set of things his culture has loaded with the maximum charge of horror, precisely so that his perception will be forced to catch up with his metaphysics.

How much of this is meant literally is a real question, and the honest answer keeps two things in tension. Some of it is bodily and extreme, done to the flesh, as both Parry and the colonial record attest. But a strand of the practice is inward. In the rite the traditions call shava-sadhana, the discipline upon a corpse, the body the seeker finally has to master is his own, and the corpse he sits on becomes a mirror held up to his own dying ego. Aghori practice shares its deepest features with the vamachara, the left-hand way of Tantra, which takes the very things orthodoxy forbids and turns them into instruments. The skull-bowl and the cremation ground are its oldest signs, and they reach far beyond this one sect: the same skull-cup passes into Tibetan Buddhist ritual, where it is a token of impermanence and no licence to eat anything. The point, everywhere, is the confrontation with death, and the end the Aghori is after is moksha, release. The full theology, the descent from the Kapalikas, and the history of the living lineage are laid out in our longer study of the Aghoris.

That is the frame the corpse-eating belongs to. It is the most literal possible enactment of a single claim: that nothing is truly impure. It is not a horror. It is an argument carried to its most confronting conclusion.

From the corpse to the leprosy ward

The last thing worth knowing is the part the documentaries never reach, because it is not frightening, and it is where the tradition itself would ask you to look.

Over the twentieth century one Aghor lineage, the Shaiva tradition anchored in Baba Kinaram at Varanasi, turned decisively away from the charnel ground. The reformer was Aghoreshwar Baba Bhagwan Ram, who founded the Sri Sarveshwari Samooh on 21 September 1961 and, the following year, the Kushta Seva Ashram at Parao, near Varanasi, a free ashram for people with leprosy. It is worth being exact here, because “the Aghoris run hospitals” is as loose as “the Aghoris eat corpses.” One named reformer, one named organisation, one named institution. Ron Barrett, whose 2008 study Aghor Medicine is the fullest account in English, reads the whole shift as one movement: away from the embrace of ritually polluted substances, and towards the healing of the diseases that carry the heaviest stigma. Jishnu Shankar, working from the Samooh’s own Hindi and Bhojpuri records rather than from outside them, arrives at the same turn by the other road, and shows it as a deliberate programme taken up against leprosy stigma and dowry and illiteracy together. Leprosy is among the most shunned of all conditions, a disease that renders its sufferers, in the eyes of society, as untouchable as a corpse.

Look at the logic, and on Barrett’s reading it has not changed at all. The old Aghori sought out the most polluted thing, the dead body, to prove that nothing is truly impure. The new Aghori seeks out the most polluted person, the leper, the outcast, the one the world treats as filth, and heals him, on the identical principle: that Shiva is present in the untouchable as fully as in the temple. There is an old radicalism in this. The medieval Kapalikas, as the historian David Lorenzen reconstructs them, stood entirely outside the caste order, and their initiation was held to place a man beyond the reckoning of birth, opening a door the orthodox world kept locked. The modern ashram inverts that same hierarchy from the other end. It does not climb the ladder of purity. It walks to the very bottom and calls that God. The transgression became compassion without altering its metaphysics by a single degree.

The traveller will always see the ghoul at the burning ghat, and the word will always arrive before the man. The eating of the dead is reported, rarely, and mostly at second hand. The killing is not there at all. The murders done in the name are another set of men entirely.

What the fear was hiding was never an appetite. It was an argument about God, and, at the end of it, a hand held out to the people no one else will touch.