The Legend of Shonitapur: Thankot and Its Kirat Origins

Shonitapur is a legend, not a documented settlement. The name belongs to Puranic literature, where Shonitapura (Sanskrit shonita, blood, and pura, city) is the fortress capital of the asura king Banasura. Nothing in the Nepali epigraphic record places that city at Thankot, and we know of no published excavation that would. What Thankot has is a tradition, repeated here for generations, that the ridge above the town and the gorge below it are the ground the story describes.

The story is set out in the Bhagavata Purana (canto ten, chapters sixty-one to sixty-three) and elaborated in the Harivamsa and the Shiva Purana. Banasura, son of Bali and great-grandson of Prahlada, is a devotee of Shiva so complete that Shiva grants him a thousand arms and, in the fuller tellings, a promise to guard his city in person. His daughter Usha dreams of a young man she has never met. Her companion Chitralekha, who can draw any face in the world, sketches until Usha recognises him: Aniruddha, the grandson of Krishna. Chitralekha then carries the sleeping prince from Dwarka to Shonitapura by magic.

Banasura finds Aniruddha inside the women's quarters and imprisons him. Krishna, Balarama and Pradyumna march on the city. Because of the promise he made, Shiva fights on the defenders' side, and the episode becomes one of the rare moments in Puranic narrative where two of the great gods stand against each other. Krishna cuts away all but four of Banasura's thousand arms, spares his life at Shiva's request, and Usha and Aniruddha are married. In the telling that names the place, the blood of the fallen soaks the ground, and the city is Shonitapura from then on.

Read it as what it is: a story about desire, a father's pride and the limits of a god's promise. It is a good story. It is not a municipal record.

This is where most writing about Thankot's ancient name goes wrong, and where being honest is also more interesting. The identification most widely made across the subcontinent puts Shonitapura at Tezpur in Assam, whose modern district is named Sonitpur and whose heritage sites, Agnigarh above all, are presented locally as the fortress where Aniruddha was held. Other claimants appear in the Uttarakhand hills and in the Himalayan foothills more generally.

A Puranic city name attaching itself to several real places is normal rather than suspicious. The Puranas were compiled and recompiled over many centuries by people who expected their own landscape to appear in them. Reading a sacred geography onto the hill behind your town is an ordinary thing for a community to do, and it usually tells you more about the community than about the hill.

So the defensible statement about Thankot is a narrow one. There is a local tradition that identifies the western approach to the Kathmandu Valley with Shonitapura. That tradition is old enough and widespread enough locally to be worth recording, and it is part of how people here describe their own place. It is not evidence that a city of that name stood on this ground, and no writer should present it as if it were.

The Kirat are a real part of the valley's historical memory, but the memory is chronicle memory. The Gopalarajavamsavali (the fourteenth-century dynastic chronicle that is the oldest surviving Nepali king list) gives a sequence of Kirat rulers before the Licchavis. Later vamshavali (genealogical chronicles) compiled under the Mallas and the Shahs expand that list, add reign lengths, and add colour. None of these are contemporaneous with the period they describe. They are later compilations working from earlier compilations.

The oldest securely dated written record from the Kathmandu Valley is the pillar inscription of Manadeva at Changu Narayan, conventionally read as 464 CE. Everything before that date, in the valley as a whole and at Thankot in particular, rests on chronicle tradition rather than on a document written at the time.

Nothing yet published fixes a date on the ground at Thankot. Dry-stone terrace walls run up the lower slopes of Chandragiri, and older residents will point out stretches of masonry described locally as pre-Licchavi. We are not aware of a published stratigraphic report from the Department of Archaeology that would let anyone say how old any of it is. Terrace walls in the middle hills are rebuilt constantly, often on the same line for centuries. A wall that looks ancient may have been relaid in living memory on a foundation nobody has dated.

What can be said with confidence is geographical rather than archaeological, and it is the stronger claim anyway. The saddle above Thankot, where the Nagdhunga crossing now runs, is the lowest and easiest way over the western rim of the valley. Anyone walking into the basin from the middle hills, from the Trisuli, or from the plains by way of Chitlang arrives here. A route that obvious does not wait for a dynasty to be recorded before people begin using it, and the settlement at the foot of it is old for reasons that have nothing to do with Banasura.

The modern name is usually parsed into two elements. Than in Nepali indicates a shrine or the seat of a deity, and here it is read as the seat of Mahalaxmi, whose temple stands in the old bazaar. Kot is the ordinary Nepali word for a fort or fortified post, the same element that appears in Nuwakot, in Makwanpur Gadhi's neighbouring hill forts and in dozens of strongholds across the middle hills. Thankot, on this reading, is the fort at the shrine.

This is folk etymology in the technical sense. It is a plausible reading of the parts, offered by local people, rather than a derivation traced through dated texts. It is worth stating because both halves match what is on the ground: there is a shrine of real consequence, and there was a post. It may still be wrong. Newar place names in the rim towns frequently differ from their Nepali equivalents, and the Newar form of a name often preserves an older layer that the Nepali form has flattened.

Three things keep the Shonitapur story alive at Thankot rather than letting it drift somewhere else.

The first is the ground itself. The approach from the west is a defile. You climb, you cross a saddle, and the whole valley opens beneath you at once. Places with that shape collect stories about sieges and last stands, because the geography is already telling one.

The second is the town's own festival. During the Mahalaxmi Jatra, Thankot is covered in sindoor (red vermilion powder) thrown by the handful over the chariots and over everybody present. Residents themselves make the connection between the red of the jatra and the city of blood. That connection is a modern reading rather than an origin, and it should be described as one, but it is a real part of how the town holds the story.

The third is the word kot. A place with fort in its name invites a story about what the fort was for. The documented answer, that Thankot was a customs and control post on the valley's western gate, is covered in our chronicle on Thankot as a customs post and the western gate. The Puranic answer is Banasura. Both have been in circulation here for a long time, and only one of them has paperwork.

There is no ruin at Thankot labelled Shonitapur, and anyone who tells you otherwise is selling something. What there is instead is a landscape you can read.

Walk the old bazaar first, then take the trail toward the Chandragiri ridge and look back. The relationship between the town, the saddle and the valley floor becomes obvious within twenty minutes of climbing, and it is the single thing that explains why every chronicle on this site keeps returning to the same twelve kilometres of hillside. The Chandragiri cable car takes you to 2,551 m at the summit in about nine minutes if you would rather not walk it.

For the Kirat and Licchavi layer, the honest itinerary is elsewhere in the valley: Changu Narayan for Manadeva's inscription, and the National Museum at Chhauni for the sculpture sequence. At Thankot, the visible heritage begins with the Mahalaxmi temple and its jatra and the surviving fabric of the Newar bazaar.

Three categories of claim run through everything written about early Thankot, and keeping them apart is most of the work.

Legend is Puranic and mythological material: Banasura, Usha, the thousand arms, the river running red. It is culturally real and historically unverifiable. We label it as legend every time.

Chronicle tradition is the vamshavali material: the Kirat king lists, the reign lengths, the founding stories of temples. It is a later record of earlier events, sometimes preserving genuine memory, sometimes tidying it. We attribute it to the chronicle tradition rather than stating it flat.

Documented history is what can be tied to an inscription, a dated order, a survey or a photograph. At Thankot that record begins seriously with the Malla and Gorkhali periods and becomes dense in the nineteenth and twentieth centuries. The Gorkhali campaign against the valley is documented history in this sense, with dates that can be checked.

Nothing on this page loses its value by being called a legend. Shonitapur tells you what this town thinks it is: a gate, guarded, worth fighting over. Every later chronicle on this site says the same thing in prose you can date.