Animal-Headed Beings at the Threshold

On the Tramenma of the bardo, the Zhitro deities, Anubis, Nana, Durga’s lion, and the mind’s own instinct at the edge of death

There's a moment that stops most careful readers of the Bardo Thödol cold: the eight Tramenma, the animal-headed dakinis, arriving in procession to greet — or rather, to test — the wandering consciousness of the newly dead. A lion-faced dakini roaring the discursive mind into silence. A tiger-headed, a bear-headed, a wolf-headed figure, each one a category of ordinary perception turned luminous and terrible. And anyone who has also stood before an Egyptian funerary papyrus knows the same jolt of recognition: Anubis, jackal-headed, weighing the heart against the feather of Ma'at while the beasts of judgment look on.

Two entirely separate civilizations, unconnected by any obvious road, both staring at death and both reaching, instinctively, for the same grammar: the human body, crowned with the head of an animal, standing at the threshold.

Widen the frame even slightly and the pattern multiplies almost embarrassingly fast. A Siberian shaman dons antlers to walk the spirit-road. An Aztec priest wears the flayed skin and skull-mask of the death-god Mictlantecuhtli. A Greek psychopomp, Hermes, sometimes travels in the company of the dog, the animal of thresholds and boundary-stones. The Sphinx guards the road to Thebes with a lion's body and a human face — the reverse configuration, but the same threshold logic. Ganesha, elephant-headed, stands at the door of every Hindu ritual not as a death-figure but as a gatekeeper of beginnings, which is its own kind of threshold. It's a question worth taking seriously rather than either swallowing whole or waving away — because the honest answer turns out to be more interesting, and considerably more useful to a contemplative, than either “it's all connected" or “it's all coincidence."

The seduction of the connection — and why it deserves real scrutiny

The diffusionist impulse isn't foolish. Trade routes were real, and the ancient world was far more interconnected than the popular imagination of isolated, hermetically sealed civilizations suggests. Roman glass has been found in Han-dynasty tombs. Buddhist monks and Greek philosophers engaged in discussions in the same Bactrian marketplaces, and their debates are recorded, astonishingly, in an early Pali text — the Milindapañha, the questions of the Indo-Greek king Menander to the monk Nagasena. Egyptian faience beads turn up in Iron Age British graves. A Roman coin hoard sits in a Vietnamese port town. The ancient world was not a set of sealed rooms; it was a set of rooms with more doors between them than most textbooks admit.

So the instinct to ask “could this have travelled?" is exactly the instinct intuitive historians have. The question is never whether to ask it, but whether you follow it all the way to the evidence, and stop precisely where the evidence stops — not one step further, in either direction. Refusing to ask the question is bad history. Answering it with wishful cartography is also bad history. What's needed is a method, applied evenly, that can tell the difference between an image that moved and an idea that was invented twice.

Let's actually apply that method. Let's take the claim seriously enough to test it against two real cases from the material at hand — one where the trail runs hot and holds up under pressure, and one where, despite every temptation to connect the dots, it goes cold.

Case One: Where the evidence actually holds — Durga's lion

Here's a genuinely startling fact, and it isn't speculation — it's numismatics, the kind of evidence that doesn't care what story you'd prefer to tell. The earliest images of the goddess who would become Durga, slaying the buffalo-demon Mahishasura, come from Rajasthan and date to roughly the 1st century BCE through the 1st century CE. She fights bare-handed. She has no lion. For the following three centuries — through the entire Kushan period — the lion simply does not appear beside her in the art of Mathura, the great sculptural workshop of the age. Anyone who has walked the sculpture galleries of the Mathura Museum can check this for themselves: Mahishasuramardini after Mahishasuramardini, arms flexed, foot planted on the buffalo's neck, and no lion anywhere in sight.

And then, in that very same Kushan world, something else was already circulating: coins and seals depicting Nana — a goddess descended in an unbroken line from Sumerian Inanna and Babylonian Ishtar, carried westward and eastward through Iran and into Bactria, where the Kushan emperors worshipped her by name, minted her image on gold dinars, seated her upon a lion, and placed a lion-headed scepter in her hand. This is not a reconstruction from fragments. Kanishka's own coinage names her — NANA — in Greek script, lion beneath her throne, for anyone with a magnifying glass and a museum case to see.

Only after this — from the Gupta period onward, fourth to sixth century CE — does Durga acquire her lion mount. The chronology fits with almost suspicious neatness. The contact zone is not hypothetical; it's the Kushan Empire itself, a genuine crossroads where Hellenistic, Iranian, Mesopotamian, and Indian cults sat in the same temples and on the same coinage for centuries, sometimes on the very same coin — a Kushan emperor might mint Shiva on one face and a Hellenistic sun-god on the other, with no sense of contradiction. This is about as clean a case of documented religious borrowing as comparative mythology gets: named deity, dated artefact, plausible route, correct sequence, and a political context — imperial patronage across a genuinely cosmopolitan multi-ethnic realm — that explains exactly how and why the borrowing would have happened.

It's worth being precise about what actually traveled, though, because precision here is the whole discipline. Durga's many arms are older than this contact and owe nothing to it — the multi-armed convention was already native to Indian sacred art, visible in yakshi and nature-spirit imagery centuries before the Kushans arrived. Her buffalo-slaying myth is indigenous, possibly rooted in an old agrarian and pastoral symbolism of the wild bull subdued. What appears to have arrived from outside is one specific, isolable element: the lion beneath her feet. And the donor, notably, is Mesopotamian and Iranian by way of Bactria, not Egyptian — a different pantheon, a different trade geography (overland through Bactria and the Hindu Kush, not by sea through the Red Sea as with Egypt), and a different mechanism of transmission (imperial coinage and a shared multi-ethnic capital, not itinerant trade-goods carried in a merchant's satchel).

So: real evidence exists for some western influence flowing into the early Indosphere — through Mesopotamia's Iranian afterlife, via the Kushans, into Gupta-era Hindu iconography. That much should be said plainly and without defensiveness. A tradition confident in its own depth does not need to deny every foreign thread in its weave; it only needs to know which threads are its own and which were borrowed, and Durga's lion is a case where the historical record actually lets us know.

Case Two: Where the trail goes cold — the Zhitro’s Tramenma, Simhamukha, and the Bardo deities

Now hold that same standard of evidence up to the animal-headed deities of the Guhyagarbha-Tantra’s Zhitro cycle, and something different happens: the trail doesn't lead west at all. It leads, cleanly and continuously, south — into India's own tantric soil, and the further back you push, the less any Egyptian or Mesopotamian thread is needed to explain what you find.

Simhamukha's roots sit inside the yoginitantras that flourished across India roughly 800–1000 CE — the same milieu that produced Vajrayogini and a whole pantheon of wrathful, theriomorphic female deities, often organised in circles or mandalas of sixty-four, each one presiding over a direction, a bodily energy, and a category of transformed affliction. You can go and stand inside the physical evidence for this: the ninth-century Hirapur Yogini temple in Odisha, open to the sky, its circular colonnade lined with sixty-four goddesses, many of them animal-headed — buffalo, boar, horse, elephant, lion, jackal, vulture, owl — centuries before any Tibetan lama codified the Tramenma. There are sister temples at Ranipur-Jharial and Bheraghat, part of a genuine, dateable, geographically continuous tantric current that runs through medieval India like a river beneath the more visible temples of Vishnu and Shiva.

When Secret Mantra carried its tantric transmissions north into Tibet in the eighth century — Padmasambhava's own biography is inseparable from the charnel-ground yoginis — and when the terma revealers of the eleventh century onward brought forth the texts we now call the Bardo Thödol, they were drawing on a fully-formed, indigenous, documented Indian tantric inheritance. No missing link to Egypt is required. No caravan needs to have carried a jackal-god's silhouette across two continents and eight centuries for Simhamukha to end up with a lion's head. The genealogy is already whole, already local, already sufficient.

And where a genuine Egyptian object actually did travel east, it's instructive to notice exactly what that borrowing consisted of and how little weight it can bear. The household protector-god Bes — squat, grinning, apotropaic, a guardian of childbirth and the home — rode the Red Sea and Indian Ocean trade routes as far as the Kushan world, turning up on ivory combs and decorative plaques at sites like Begram in Afghanistan, part of the same great Indian Ocean and overland network that moved Roman glass, Chinese lacquer, and Indian ivory across thousands of miles. It shows that Egyptian pictures could travel that far, carried as small portable decorative motifs, the ancient equivalent of a souvenir. It does not show that Egyptian theology traveled with them. Bes at Begram is unconnected to death, judgment, or the afterlife — he is a nursery guardian, not a psychopomp. An image riding along in a merchant's luggage, stripped of its ritual context and repurposed as decoration, and a cosmology transmitted through lineage, initiation, textual commentary, and generations of trained teachers, are two entirely different weights of evidence. It would be a mistake — the single most common mistake in amateur comparative mythology — to let the first stand in for the second.

The tell in the details: what each animal head is actually doing

There is also a structural tell here, if you look closely not just at that an animal head appears, but at what the animal head is actually doing inside its own system — because superficial resemblance and functional resemblance are not the same test, and conflating them is where most diffusionist arguments quietly go wrong.

Anubis has the head of a jackal, at least in part, because jackals, in the real ecology of the Nile valley, dug up shallow desert graves — his head encodes an observed, ecological fear, domesticated into a god precisely so that the danger could be placated and controlled. His role, correspondingly, is fixed inside a single, static judgment scene: the heart on the scale, weighed once, against a fixed and unchanging feather, in front of a tribunal that does not move. It is a courtroom image. Time stands still inside it.

Simhamukha has the head of a lion, at least in part, because a lion's roar is the sound that shatters discursive thought — her head encodes an auditory and psychological metaphor for wrathful wisdom cutting through conceptual proliferation, not an observed predator-fear translated into theology. Her role belongs to a moving sequence, a journey through successive days of the bardo in which different animal-headed figures represent different classes of ordinary consciousness dissolving one after another, wave after wave, each one a different flavor of grasping being burned through. It is a process image. Time is the entire point of it.

The surface — a beast's head on a human body — is identical. The interior logic generating that surface is not merely different; it is generating two entirely different kinds of religious object, one static and juridical, the other dynamic and psychological. That's usually the signature of independent innovation rather than borrowing: two cultures reaching for a similar shape of solution while solving entirely different problems with it, the way two engineers in different centuries might both arrive at the wheel without either one having seen the other's cart.

The wider gallery: how far this pattern actually extends

Once you start looking for this specific grammar — human body, animal head, positioned at a threshold — it becomes almost difficult to stop finding it, and the breadth of the gallery is itself part of the evidence.

Siberian and Central Asian shamanic traditions clothe their psychopomps in antlers, bird-feathers, and bear-claws, the shaman's ritual costume itself becoming a composite creature the moment the drum starts and the journey to the spirit-world begins. Mesoamerican death-gods — Mictlantecuhtli among the Aztec, the Maya lords of Xibalba — wear the fleshless jaw of an animal skull fused to a human frame, presiding over an underworld journey structured, like the bardo, as a sequence of trials rather than a single static scene.

Mesopotamia, independently of its later export to India, gives us the lion-headed, bird-talloned being Lamashtu, who preys specifically on the threshold moment of childbirth — arrival, not departure, but a threshold all the same. Norse mythology gives us less a single hybrid figure than an entire class of them: the valkyries who choose the slain on the battlefield, sometimes described in the old poetry with birdlike, corvid associations, ravens themselves — Huginn and Muninn — serving Odin as psychopomp-messengers between worlds. Japanese folk religion offers the tengu, part bird, part human, guardians of mountain thresholds between the human world and something older. Egypt gives us not only Anubis but Sekhmet, lion-headed, goddess of both plague and its cure, standing at the threshold between destruction and healing.

None of these cultures needed to whisper the idea to one another across oceans and millennia they never crossed. Anthropologists studying rites of passage across totally unrelated societies — from Ndembu initiation rites in Zambia to Christian monastic vows — noticed the same thing structurally, and gave it a name that has since become indispensable to the whole field: liminality. Beings caught between one state and another are almost universally depicted as “neither this nor that," composite, monstrous in the old and non-pejorative sense of the word — a monstrum, a showing, a sign — because ordinary singular categories (fully human, fully animal) simply cannot hold a being who belongs to neither the world of the living nor the world of the dead, neither childhood nor adulthood, neither the sacred and the profane in their settled, daytime forms. The hybrid form isn't decoration laid on top of an otherwise ordinary figure. It is the most economical possible image for in-between-ness itself — a visual solution to a conceptual problem that has no other elegant solution, arrived at independently by minds separated by every ocean on earth.

Which means the convergence a careful reader notices — Anubis and Simhamukha, Sekhmet and the Tramenma, the tengu and the valkyrie — is real, but it isn't a trade-route fact waiting to be confirmed by some future archaeological find. It's something closer to a fact about the architecture of human consciousness when it turns, anywhere on earth, at any time, to actually look at death, or birth, or any other doorway the ordinary categories of mind cannot walk through unchanged.

For the Dzogchen reader, in particular

This is where the historical inquiry stops being merely a historical inquiry and becomes worth something more.

In Dzogchen, the peaceful and wrathful deities of the bardo are not taught as external beings who must be historically pedigreed, traced to a lineage of borrowed images, or authenticated by an unbroken chain of transmission from some older civilization in order to be functionally real. They are taught as the spontaneous display — the rölpa — of one's own awareness meeting the dissolution of ordinary reference points. They are called self-arising, rangjung: not fabricated by deliberate imagination the way a novelist invents a character, but appearing the way light appears from the sun, as the natural, uncontrived expressions of one’s own mind when the usual conceptual scaffolding of a self and a world falls away. The texts are explicit that these forms are not created by any external agency and are not, in the end, “somewhere else" — they are recognised as one's own appearances.

Seen from that angle, the fact that unconnected human beings — in the Nile valley, on the Deccan plateau, in the high plateaus of Tibet, in Siberian taiga, in Mesoamerican jungle, on the fjords of the North Atlantic — all independently arrived at the same class of image for the threshold of death is not a disappointing debunking of the tradition's uniqueness. If anything, it would be more troubling for Dzogchen's own claim if these images had all been shown to derive from one single ancient source, passed hand to hand like a technology — because that would suggest the forms were merely cultural artefacts, arbitrary conventions that happened to spread, rather than something arising lawfully from the nature of mind itself wherever a mind turns honestly toward its own dissolution.

Instead, what the evidence actually shows is closer to a confirmation of something the tradition already claims: that when mind meets the edge of ordinary categories, it doesn't invent arbitrarily, culture by culture, each one a private accident. It produces the same small handful of forms, lawfully, independently, the way water produces the same hexagonal crystal wherever on earth it freezes, regardless of whether the water in Tibet has ever heard of the water in Egypt. The animal-headed being at the threshold may not be, in its deepest layer, a historical artefact passed hand to hand along a caravan route at all — though sometimes, as with Durga's lion, a particular form genuinely was. It may be closer to what the tradition has called it all along: a natural pattern of mind itself, arising wherever the psyche is honest enough to look directly at its own impending dissolution, and finding, everywhere it has ever looked, on every continent, in every age, roughly the same face looking back.