There is a pattern in the ethnographic and historical record that is real, well attested, and very easy to sentimentalise.
The pattern: in a striking number of societies, the person who deals with the spirits is not straightforwardly a man or straightforwardly a woman. Sometimes they are a third recognised category. Sometimes they are a man who takes on a woman’s dress and speech, or a woman who takes on a man’s role and weapons. Sometimes the office is simply held by women in a world where public authority otherwise is not.
The sentimental version says these societies revered gender-variant people and that we have lost something. The record is harder than that, and better. Some of these roles carried enormous prestige. Some carried prestige and danger at once. Several were destroyed, deliberately and within documented living memory, by colonial administrations and religious authorities who understood exactly what they were dismantling.
This article deals in named traditions rather than a general idea of the sacred outsider, and it tries to say what happened to each of them.
Bissu: A Fifth Gender With a Priesthood
Among the Bugis people of South Sulawesi in Indonesia, five genders are recognised, not two. The anthropologist Sharyn Graham Davies has documented them: oroane, broadly men; makkunrai, broadly women; calalai, assigned female and living in a masculine role; calabai, assigned male and living in a feminine role; and bissu.
The bissu are the fifth, and they are not simply a further variation. They are understood to combine all the others, and that combination is what qualifies them for the work. Bissu are ritual specialists. Historically they were attached to the courts of the Bugis kingdoms, guarded the regalia, blessed rice harvests and pilgrims, and were the custodians of La Galigo, one of the longest works of literature in the world. Their language, basa bissu, is a ritual register distinct from ordinary Bugis speech.
The reasoning is explicit rather than metaphorical: a person who spans the human genders can address gods, because they are not confined to one side of the human condition. In Bugis cosmology this is a qualification, in the way that literacy was a qualification for an Egyptian priest.
What is usually left out is what happened next, and it is recent. Bissu were persecuted during the Darul Islam rebellion of the 1950s and 1960s, when the movement’s campaign against practices deemed un-Islamic targeted them directly. They were persecuted again under Operasi Toba in 1966, in which bissu were killed, forced to cut their hair, and made to present as men. Numbers collapsed. A small number of bissu continue to practise in South Sulawesi today, and the tradition remains under pressure. Any account that describes the bissu purely in the present tense as a celebrated institution is describing a survival and calling it a status quo.
Babaylan and Asog: The Role the Colonisers Recognised First
The Philippines offers the clearest documented case of a magical office being destroyed on purpose, because the people doing the destroying wrote down why.
The site’s babaylan entry describes them as the spirit mediums, healers and oral historians of the precolonial Philippines, most of them women. That last clause is the crucial one and the entry is right to lead with it. The babaylan, called catalonan in Tagalog areas and other names elsewhere, held the ritual authority of the community. They conducted the pag-anito, negotiated with the spirits, delivered healing, and carried the genealogies and the epics. In a society where the datu held political authority, the babaylan held spiritual authority, and the two were not ranked.
Alongside them, and named in the Spanish sources, were the asog or bayoguin: people assigned male at birth who lived as women and performed the babaylan role. The Spanish chroniclers found them intolerable and said so at length, which is why we have the documentation. Their accounts are simultaneously the best evidence that the role existed and a record of the campaign to end it.
That campaign was successful in the lowlands. Spanish missionaries specifically targeted the babaylan, because the office was the institutional competitor to the priest. The practice survived in the highlands and in the Visayas in altered form, and babaylan-led revolts run through Philippine colonial history as a repeated pattern. The word has since been reclaimed by Filipino and Filipino-American writers and activists, which is a real continuity even though it is not an unbroken one.
Mesopotamia: The Priests Who Were Neither
The oldest written evidence for gender-variant ritual specialists comes from Sumer and Akkad, and it is old enough to be startling.
The cult of Inanna, who becomes Ishtar in Akkadian, included personnel described in the texts as gala, kurgarru and assinnu. The details are contested and the terms do not map neatly onto any modern category, which is exactly why the sources are worth quoting rather than summarising. The gala were lamentation singers who performed in eme-sal, a dialect of Sumerian otherwise used by women in literary texts. Some texts describe the kurgarru and assinnu as neither male nor female, or as having been transformed by the goddess. A well-known line in a hymn to Inanna says she turns a man into a woman and a woman into a man, and lists this among her powers.
That is a claim being made about a goddess by her own worshippers, in her own liturgy, more than three thousand years ago. Whatever else it means, it means the association between gender crossing and divine power is not a modern reading imposed on ancient material.
The link to magic is direct. In Mesopotamian practice the incantation-priest and the lamentation-singer were technical professions, trained and paid, and the gala belonged to that world of scribal and ritual specialists alongside gods of writing like Nabu. This was not a fringe. It was staff.
Something similar appears further west. The galli, castrated priests of Cybele, served the Phrygian mother goddess of mountains and wild animals, and their processions through Roman cities were famous, popular and objects of Roman elite contempt in roughly equal measure. Roman writers who despised them still did not question that the goddess required them.
The Norse Case: Prestige and Shame in the Same Practice
Norse magic supplies the most useful counter-example in this article, because it refuses to be a straightforward story of honoured outsiders.
Seidr was a form of magic concerned with prophecy, with influencing minds and luck, and with harm. It was practised primarily by women. The volva, the travelling seeress named for her staff, prophesied from a raised seat before an assembled household, and the Saga of Erik the Red gives a detailed account of one such session: the food she is served, the clothing she wears, the assistant who must sing the songs.
Freyja is said in Ynglinga saga to have brought seidr to the Aesir. Odin learned it. And here the tradition turns on itself: the same source says that this magic brought with it ergi, so great a shame that men were held not to practise it without disgrace. Loki throws exactly this accusation at Odin in the Lokasenna, and Odin throws it back, because Loki has borne children.
So the Norse material holds two things at once. The magic is real and powerful and the chief god of the pantheon wanted it badly enough to take it. And a man who practised it was shamed for it, in a legally and socially specific way that the word ergi carried. Prestige and stigma attached to the same act, and which one applied depended on who was doing it.
This is a healthier model for the whole subject than the celebratory one. Gender-crossing did not confer honour automatically anywhere. It conferred access, and what a society did with someone who had that access varied enormously.
Greek myth arrives at something adjacent from another direction. Tiresias is the blind Theban prophet who lived seven years as a woman, was asked by Zeus and Hera to settle which sex enjoys sex more, answered, and was blinded for the answer and given prophecy in compensation. He is the greatest seer in Greek tradition, and the myth is explicit that his authority comes from having been both. He keeps his wits among the dead, in a place where everyone else is a shade.
Two-Spirit: A Modern Word for Many Older Things
The term two-spirit was proposed in 1990, at an intertribal Native American and First Nations gathering in Winnipeg, by Myra Laramee, and adopted as a pan-Indigenous English umbrella term. It was created for a specific purpose: to replace berdache, an ethnographic label derived from a French and ultimately Persian word for a kept boy, which Indigenous people had never chosen and found offensive.
Two-spirit is therefore not an ancient term and was never claimed to be. It is a contemporary, political, self-chosen word, and treating it as the name of a historical institution gets both halves wrong.
The historical roles it replaced the old label for are many and are not one thing. The Diné nadleehi, the Lakota winkte, the Zuni lhamana and dozens of others are distinct roles in distinct societies with distinct duties, and the degree of ritual involvement varied widely. Some carried specific ceremonial functions. Some did not. We’wha, a Zuni lhamana who travelled to Washington in 1886 and was received in Washington society, is the best-documented individual, and the documentation exists largely because Matilda Coxe Stevenson wrote about her.
Two cautions belong here, and they are the reason this section is short.
First, the popular claim that such people were universally revered across Native American societies flattens hundreds of distinct nations into one and is not supported. Treatment ranged from high ritual status to ordinary acceptance to hostility, depending on the nation and the period. A generalisation that sounds respectful is still a generalisation.
Second, several of these traditions are living and their communities are specific about who may describe them and how. Where that is the case, the appropriate move is to name the tradition, point at what its own members have published, and stop. Sue-Ellen Jacobs, Wesley Thomas and Sabine Lang’s edited volume Two-Spirit People and Brian Joseph Gilley’s Becoming Two-Spirit are the places to go, and both include Indigenous voices directly rather than about.
Women in the Western Magical Tradition
One correction to the earlier version of this article is worth making explicitly, because the original section on female alchemists rested on a claim that does not survive contact with a date.
Mary the Jewess is real and important. She worked in Alexandria, probably in the first to third century CE, and everything we have of her comes through quotation in Zosimos of Panopolis. She is credited with the tribikos, a three-armed still, and the kerotakis, a sealed vessel for exposing metals to vapour. The balneum Mariae, the water bath still used in kitchens as the bain-marie, carries her name. None of her own writing survives. The later identification of her with Miriam, sister of Moses, is a medieval accretion and is not part of the original record.
Helena Blavatsky was not an alchemist, and the claim made for her in the earlier version was false. That version said she was “one of the first to promote the idea of magic as a bridge between material and spiritual realms”. That idea is older than writing, and in its specifically Western form it is Renaissance Hermeticism, three and a half centuries before her. Blavatsky’s actual significance is different and does not need inflating: she co-founded the Theosophical Society in 1875, and Theosophy is the channel through which a great deal of South Asian religious vocabulary entered Western occultism, and through which much of modern esotericism descends. The 1885 Society for Psychical Research investigation concluded that her phenomena were fraudulent, a finding the Society partially revisited a century later. She belongs in the history of modern occultism, prominently, and not in a section about alchemy.
The larger point the original was reaching for is sound and can be made without either error. Where the Western magical tradition was institutional, it excluded women by rule, and where it was domestic and practical it did not. That is why the record shows so few women in academic alchemy and so many in cunning-craft, herbal healing and divination: not a difference in aptitude or interest, but a difference in which doors had guards on them.
What the Pattern Actually Is
Reading these traditions together, the association between gender variance and magical authority is real but it is not mystical. It is structural, and three mechanisms do most of the work.
- Liminality is a qualification. A specialist who mediates between worlds is easier to imagine, and easier to accept, if they already stand between categories. The bissu explanation says this outright.
- Spirit work is one of the few offices that admits outsiders. Where inheritance, land and public authority are closed, the ritual role is often the one door open. That is a fact about what else is shut, not about spirituality.
- And an outsider is a safer holder of dangerous knowledge. Someone outside the lineage system has fewer interests in the quarrels the magic will be used to settle.
The last thing worth saying is the thing the celebratory version leaves out. Almost every tradition in this article was attacked, and mostly not long ago. The babaylan were suppressed by missionaries. The bissu were hunted in the 1960s. Berdache was the label under which two-spirit people were catalogued, and residential schooling and missionary pressure did the rest. These are not ancient losses. They are recent policy, and in several cases the people who carried out the policy left records explaining that the office was the target because it was the alternative authority.
That is a more useful thing to know than that some cultures honoured the sacred outsider. It says what happened, and it says why so little of it is left.
Further Reading
- Sharyn Graham Davies, Gender Diversity in Indonesia: Sexuality, Islam and Queer Selves. The standard English account of the Bugis five genders and the bissu.
- Carolyn Brewer, Shamanism, Catholicism and Gender Relations in Colonial Philippines. On the babaylan and what was done to the office.
- J. Neil C. Garcia, Philippine Gay Culture, for the asog and bayoguin in the Spanish sources.
- Sue-Ellen Jacobs, Wesley Thomas and Sabine Lang, eds., Two-Spirit People: Native American Gender Identity, Sexuality, and Spirituality.
- Brian Joseph Gilley, Becoming Two-Spirit: Gay Identity and Social Acceptance in Indian Country.
- Will Roscoe, The Zuni Man-Woman. On We’wha, and the best-documented individual case.
- Neil Price, The Viking Way: Magic and Mind in Late Iron Age Scandinavia. The full treatment of seidr, the volva and ergi.
- Carol Clover, “Regardless of Sex: Men, Women, and Power in Early Northern Europe”. Short, and the clearest statement of how Norse gender categories actually worked.
- Julia M. Asher-Greve and Joan Goodnick Westenholz, Goddesses in Context, for Inanna’s cult personnel with the philological caution the subject needs.
- Raphael Patai, The Jewish Alchemists, for Mary the Jewess in her actual context.
- Meredith K. Ray, Daughters of Alchemy: Women and Scientific Culture in Early Modern Italy. Women in the alchemical record, with archives behind it.
Explore More on Gods and Monsters
Southeast Asia: Babaylan · Philippine Mythology · Indonesian Mythology · Southeast Asian Mythology
Mesopotamia and the ancient Mediterranean: Inanna · Ishtar · Ereshkigal · Nabu · Cybele · Mesopotamian Mythology · Sumerian Mythology · Roman Mythology
Norse magic and its costs: Volva · Freyja · Odin · Loki · Skadi · Norse Mythology
Greek seers and boundary-crossers: Tiresias · Zeus · Hera · Circe · Medea · Hecate · Greek Mythology
Elsewhere in this series’ territory: Witch · Warlock · Machi · Native American Mythology