Culture · Essay
The Kalash: a religion of about four thousand people, and the effigies they stopped carving
In three valleys of the Hindu Kush a few thousand people keep a religion older than Islam in the region. On the gandau effigies raised for the notable dead, why they are gone from the valleys, and what it takes to hold a tradition at this scale.Written to last.
By Confinity Editorial · 2026-04-23 · 6-minute readQuiet tools, kept out of the way.
There are three valleys in the Hindu Kush, in Chitral district in northern Pakistan, where about four thousand people keep a religion that is not Islam, not Hinduism and not Buddhism, and that appears to predate all three in the region.
Everything about Kalash remembrance has to be read against that number. Four thousand people is not a minority. It is roughly a large secondary school, holding a complete cosmology, a calendar, a language, and a body of funerary practice with no institutional support anywhere.
The most striking Kalash funerary object is the one you can no longer see in the valleys.
Gandau, sometimes gandao, were wooden effigies carved for people of standing: someone who had given the great feasts of merit, who had accumulated honour, whose death warranted it. The figure was raised near the graveyard, often mounted on a horse, and it stood in the open.
They are angular, frontal and unmistakable, and they weathered in the mountain air over decades. The graveyard held the coffins; the effigies held the reputations.
There are almost none left in place. They were collected out of the valleys through the twentieth century, and Kalash gandau now sit in museums in Pakistan and in Europe and in private hands. The practice of carving them has effectively stopped.
That is the same trajectory as the Toraja tau tau, taken further because the community is far smaller and had far less power to object. An object made so that a person would remain visible to their own village is now visible to strangers in another country, and not to the village.
Kalash burial did not put people in the earth.
The dead were placed in wooden coffins set on the surface of the ground in the communal graveyard, not buried. The graveyards are open, and the coffins weathered and eventually broke apart, so that older ones were visible in a way that outsiders found shocking and Kalash did not.
Underground burial has become general over recent decades, partly under pressure from neighbours and partly from tourism: the open graveyards attracted visitors, photography and disturbance, and the community's response was to close the practice.
A funerary tradition changed here for a reason that had nothing to do with belief. Being looked at became intolerable. It is the same argument the Australian protocols make about names and images, arrived at from the other direction.
Kalash death observances are not quiet, and this is the part most reporting gets wrong by omission.
A death, particularly of an older person who lived well, is marked with days of drumming, dancing and singing at the bashali ground or in the village square, with food distributed at the family's expense. The scale reflects the person's standing, and it costs the family a great deal.
The logic is the one that appears in Toraja and in Ghana: a funeral is a public statement about who this person was, and the community's attendance is the record. Songs composed for the dead are sung, and those songs are the durable part, because they are the only part with words in them.
The Kalash year runs on its own festival cycle, and the largest is Chaumos, held around the winter solstice, running for roughly two weeks.
It is a purification and renewal festival: the valleys are closed, outsiders are restricted, goats are sacrificed, torches are carried, and there is continuous singing and dancing. The ancestors are addressed as part of it.
The other festivals, Joshi in spring and Uchau in late summer, mark the movement of livestock and the harvest. The whole calendar is agricultural, and the remembrance is embedded in it rather than separate, which is what happens everywhere before a religion acquires an institution to schedule things for it.
The Kalash are frequently written about in a tone of imminent extinction, which is both unhelpful and not quite accurate. What is accurate:
The community is very small and conversion to Islam is a continuous fact, driven by marriage, by economic advantage and historically by considerable pressure. Kalash who convert leave the religious community. The population has been roughly stable in recent decades, which given the outflow means the underlying birth rate is doing a lot of work.
Tourism is a genuinely double-edged thing. It brings money and outside interest that has translated into some legal protection, and it also brings people who treat festivals as spectacle, photograph what they should not, and misrepresent the community in ways that have caused real harm.
The Kalasha language is spoken by essentially everyone in the community and by almost nobody outside it, and it carries the songs and the prayers. It has been written down only recently and partially.
The point of setting all this out is not to be mournful about it. It is that the Kalash system has no redundancy at all.
There is no book. There is no institution outside the valleys. There is no clergy in the sense of a body that trains successors and holds records. The knowledge sits in a small number of people, the songs sit in a language with no substantial written corpus, and the objects that carried individual reputations are in museums in other countries.
Every other tradition in this library that survived a catastrophe survived it because something was written down, or because the practice was distributed across millions of households. The Kalash have neither.
Which makes the recent documentation work, meaning dictionaries, recordings and ethnography done with the community rather than at it, the single most consequential thing happening to Kalash memory. Not the festivals, which are in reasonable health. The written backup, which did not exist.