The Birth of Magic: Ancient Sorcerers and Shamanic Roots

by

Updated 24 August 2026 · 9 min read

The Birth of Magic: Ancient Sorcerers and Shamanic Roots

Nobody can tell you when magic began. Anyone who says otherwise is guessing at cave paintings.

What we can do is something narrower and much more interesting: read the earliest magic that was written down, and find out who was performing it, what they were paid, what they were afraid of, and where the record stops. That takes us back roughly four thousand years, and what it shows is not a world of mystics communing with the infinite. It shows professionals with a syllabus.

Mesopotamia: the exorcist had a reading list

First, a correction that matters, because it is repeated almost everywhere including in the earlier version of this article.

The Mesopotamian magic specialist is the āšipu, and the word is Akkadian, not Sumerian. Sumerian scribes wrote the role with the logogram MAŠ.MAŠ. This is not pedantry. Sumerian and Akkadian are unrelated languages, separated by centuries and by an entire linguistic family, and running them together is the sort of thing that makes a reader stop trusting everything else on the page.

What the āšipu actually was: a trained specialist in āšipūtu, the craft of ritual diagnosis and counter-ritual, attached to temples and to royal courts. The profession had a curriculum. A tablet catalogue known to scholars as the Exorcist’s Manual lists the texts a practitioner was expected to have mastered: incantation series, ritual series, omen collections, the lot. You could no more improvise your way into the job than you could improvise your way into being a scribe.

The comparison is not loose. Scribal training was the other great Mesopotamian curriculum, and it had a god of its own: Nabu, who holds the stylus that records what happens to people. Writing and ritual were two branches of one literate profession.

He was not the only medical option. The asû worked with drugs, bandages and procedures. The relationship between the two was collegial rather than competitive, and both are found treating the same patient. A modern reader wants one of them to be the doctor and the other the witch-doctor. Neither Mesopotamian nor the evidence supports that split.

And the āšipu’s biggest job was defence. The ritual series Maqlû, “Burning”, runs across nine tablets and was performed over the course of a single night: identify the witch attacking the patient, compel her to appear, burn her in effigy, and reverse what she has done. Figurines were made and destroyed. Incantations were recited in sequence. It is elaborate, expensive and precisely scripted, and it exists because witchcraft was believed to be happening constantly and to be actionable.

The gods involved are the ones you would expect. Shamash, the sun, sees everything and is therefore the judge of hidden crimes. Ishtar and Inanna are invoked for power. Marduk inherits the incantation-craft from his father Enki, the god of the sweet waters, of cunning, and of the spoken formula that changes a situation. In the Enūma Eliš it is Enki who subdues Apsu with a spell before Marduk ever faces Tiamat.

The threats have names too, and they are the reason amulets existed: Pazuzu, the wind-demon whose head was worn to keep worse things away, and Lamashtu, who took infants and against whom Pazuzu was the standard countermeasure. This is the practical end of Mesopotamian mythology, not cosmology, but what you hang around a baby’s neck.

Egypt: magic was the mechanism, not the alternative

Egypt’s word for the force is heka, and here is a second correction, because the earlier version of this article called it hekau. Heka is the power. Hekau is the plural, used for the spells and utterances themselves, and by extension for those who wield them. Heka is also a god in his own right, personifying the force.

The distinction matters because of what it implies. In Egypt, magic was not an alternative to religion or a folk practice running underneath it. Heka is the energy by which creation was performed and is sustained. The gods use it. The king uses it. Priests use it in temple ritual. Ordinary people use it on amulets and in household spells against scorpions.

What it defends is maat, the order of things, which is truth, justice and cosmic regularity in one concept, and which requires constant maintenance because Apep attacks the sun-barque every single night without ever winning and without ever stopping.

The literature is enormous and it is startlingly practical:

  • The Pyramid Texts, carved inside Old Kingdom royal pyramids from around the twenty-fourth century BCE, are the oldest substantial body of religious writing we have from anywhere. They exist to get a dead king safely through.
  • The Coffin Texts democratise the same equipment down to anyone who could afford a decent coffin.
  • The Book of Going Forth by Day, which is the Egyptian title and much better than the modern label “Book of the Dead”, is a spell manual for the afterlife, including the interrogation before Osiris and the weighing of the heart, with Anubis at the scales and Thoth recording the result.

Isis is the model practitioner. Her defining act in the mythology is not devotion but technique: she extracts Ra‘s secret name by poisoning him and refusing the cure, because knowing the true name is knowing the operating instructions. It is the same principle that runs through the whole of Egyptian mythology, and through half the magic in this series.

And the aggressive end was equally official. Execration texts were red pots and bound-captive figurines inscribed with the names of enemies, then deliberately smashed and buried. The state did this. Sekhmet, who is plague and its cure in the same goddess, had priests who were specifically plague-ritualists. Magic here is not the opposite of order. It is how order is enforced.

Siberia: the word is doing more work than it can bear

Now the part that requires care.

“Shaman” is a specific word from a specific place. It comes from Evenki, a Tungusic language of Siberia, and it entered European languages in the seventeenth century through Russian accounts of the region. It described the ritual specialists of particular Siberian peoples: the practitioner who drums, enters an altered state, and is understood to travel to other worlds to retrieve a lost soul, learn a cause, or negotiate with spirits.

All of that is well documented, and the Siberian material is genuinely extraordinary. The costume alone is a whole cosmology worn on the body: iron pendants, antlers, and the drum treated as a mount rather than an instrument.

What happened next is the problem. Over the twentieth century, and above all through Mircea Eliade’s Shamanism: Archaic Techniques of Ecstasy in 1951, the word was stretched to cover ritual specialists from Amazonia to the Arctic to Australia, on the argument that they all shared an underlying technique of ecstasy. It became a universal category.

Ronald Hutton’s Shamans (2001) is the clearest account of why that is a problem, and it is worth stating plainly because this series will use the word again: calling a Yoruba babaláwo, a Korean mudang, a Mapuche machi and an Evenki practitioner all “shamans” tells you almost nothing about any of them, and quietly imports a Siberian model into places it does not fit. The machi enters trance and climbs a carved pole toward the sky world in southern Chile. The resemblance to a Siberian soul-journey is real, and it is not evidence of a shared origin. Where this article can name the specific tradition, it will.

The Indus Valley: where the evidence runs out

The Mature Harappan period, roughly 2600 to 1900 BCE, produced one of the largest and best-planned urban civilisations of the ancient world. It also produced thousands of small carved seals, many showing animals and human or divine figures.

One of them, seal M-304, shows a seated figure with a horned headdress, surrounded by animals. In 1931 John Marshall identified it as a “proto-Śiva”, an early form of Shiva in his aspect as Paśupati, lord of animals, and by extension a link back to Rudra in the Vedic material.

The identification has been contested more or less ever since. Doris Srinivasan and others have argued the figure’s features do not support the reading, that the “horns” and the posture have other explanations, and that the whole interpretation runs backwards from what we already knew about later Hindu mythology rather than forwards from the object.

The honest position is that we do not know, and the reason we do not know is that the Indus script has never been deciphered. We have somewhere in the region of four thousand inscribed objects, and the inscriptions are short, often only a handful of signs. Without a bilingual text, without agreement on the underlying language, the seals cannot be made to speak.

This section exists because it is the useful one. The temptation with the Indus Valley is to fill the silence with whatever the reader already believes. An earlier version of this article did exactly that, suggesting the seals showed “individuals performing rituals” and “attire associated with magical rites”. They may. Nothing in the evidence establishes it.

What the four have in common is a problem, not an origin

It is tempting to close by saying these traditions share a root. They do not, in any way anyone can demonstrate. Mesopotamia and Egypt were in contact and influenced each other; Siberia and the Indus Valley belong to entirely separate stories.

What they share is a situation. In every one of these societies, bad things happened for no visible reason, and someone had to be able to say why, and to do something about it. The answers were built independently and they do not resemble each other much in detail. But the job description is nearly identical, and four thousand years later it has not changed as much as we like to think.

The next article in this series looks at what happened when those jobs acquired names, and at how much damage a badly translated word can do.

Further Reading

  • Tzvi Abusch, The Magical Ceremony Maqlû: A Critical Edition (2016), and Mesopotamian Witchcraft (2002). The standard work on the ritual described here.
  • JoAnn Scurlock, Magico-Medical Means of Treating Ghost-Induced Illnesses in Ancient Mesopotamia (2006). On the āšipu and asû working side by side.
  • Robert K. Ritner, The Mechanics of Ancient Egyptian Magical Practice (1993). The definitive account of heka as a practical technology.
  • Geraldine Pinch, Magic in Ancient Egypt (1994). The accessible way into the same material.
  • Ronald Hutton, Shamans: Siberian Spirituality and the Western Imagination (2001). On the Siberian evidence, and on what the word has been made to carry.
  • Mircea Eliade, Shamanism: Archaic Techniques of Ecstasy (1951). Read it, but read Hutton alongside it.
  • Asko Parpola, The Roots of Hinduism: The Early Aryans and the Indus Civilization (2015). The most serious attempt at the Indus problem, and honest about its limits.

Explore More on Gods and Monsters