The Box from the Tsenker Confluence

A small dark wooden box with dovetailed corners lying open on a grey foam tray beside a folded square of faded red silk and a pair of steel tweezers
The box as it was photographed on the second day, with the silk lifted out and laid beside it. Larch, dovetailed, 138 by 96 by 71 millimetres; the corroded iron hasp is still shut on its staple, which means whatever opened this box did not go through the fastening.

Delgerkhaan sum, in Khentii, is three thousand nine hundred and eighty-six square kilometres of grass with two thousand and forty people on it at the last count anyone bothered to publish. Where the Kherlen — which rises on the south slopes of the Khentii below Burkhan Khaldun and runs 1,254 kilometres to a lake over the Chinese border — takes in the Tsenker, there is a low platform of trodden earth that teams have been reading since the 1990s. It is called Avraga, and in 2020 an Australian group put twenty separate radiocarbon determinations through livestock bone lifted off it. Every one came back thirteenth century. It is as close as archaeology has got to a place Genghis Khan is thought to have wintered, and the people who did the work were careful to say that it still puts nobody in particular on the ground.

Four hundred metres east of that platform, in the third week of September 2023, a crew cutting a drainage trench for a new fence line went through the lid of something at ninety centimetres. The man who stopped the work was D. Naranbaatar, a hydrologist who was there about the fence and not about the thirteenth century, and who photographed the object in place from four sides before he lifted it. That is more than most people manage, and it is the reason the find is worth discussing at all.

Nine sheets of pale fibrous paper laid out in three rows under glass on a grey table, each covered in narrow vertical columns of brushed marks, with gloved hands at the lower edge
The nine sheets under glass at Öndörkhaan, four days after lifting. Seven are written on one side only. The columns run top to bottom and left to right, which is correct for the script and correct for every printed primer of it.

Inside the box, wrapped twice in a square of silk gone the colour of weak tea, were nine sheets of paper written in vertical columns in the Mongolian script. Four are lists — of horses, of felt, of two names and a quantity of millet. Five are verse. Within a fortnight the box had a name, the Delgerkhaan box, and the name was doing most of the work that names do. By November the five poems were being described in three languages as the private notebook of Genghis Khan, and the fact that not one line of them says who wrote them had stopped being mentioned.

It is worth being plain about what would have to be true. Nothing survives that Genghis Khan is known to have written, or to have had written down at his dictation, in any hand or any script. His law code, the Yassa, exists only as descriptions by men who came after him — Juvayni, Rashid al-Din, al-Maqrizi — because no scroll or codex of it has ever been found. The one substantial Mongol account of his life, the Secret History, survives in full only because Ming officials spelled the Mongolian out phonetically in Chinese characters to train interpreters, and folded the result into an encyclopedia in 1408. There is no signature, no marginal note, no dictated letter. There is a great deal of him in other people’s writing and none of him in his own.

What Carbon Cannot Say

Two technicians in white coats at a stainless steel bench, one weighing a vial of pale powder on a balance, the other loading a rack of narrow sealed quartz tubes
Sample preparation for the dating. The tubes are sealed under vacuum and combusted; what reaches the accelerator is a bead of graphite the size of a pinhead. Fifty milligrams of paper went in and one date came out.

Two of the nine sheets were sampled. Accelerator mass spectrometry wants about fifty milligrams, counts the carbon-14 atoms directly instead of waiting for them to decay, and will work on anything from a century old to fifty thousand. The results were reported as calibrated ranges, because a radiocarbon result is never a year: the raw age has to be pushed through a calibration curve before it means anything in calendar terms, and even then a one-sigma figure is only 68 per cent confidence. Both sheets came back inside the thirteenth century. The press release said so. The press release also called this confirmation.

It is not confirmation of anything except the paper. What carbon dates is the death of the plant the fibre was beaten out of — not the day a sheet was made, and emphatically not the day somebody wrote on it. Blank medieval paper is not rare, and it is not expensive. This is exactly the trap the Vinland map was built out of: the parchment dated honestly to between 1423 and 1445, and the ink turned out to be full of anatase, a titanium dioxide manufactured for pale pigments only since the 1920s. Yale gave the thing up as a forgery on 1 September 2021, forty-nine years after Walter McCrone first said so, and three years after its own conservation scientist had put the modern pigment through the map’s ink lines in front of a symposium.

The ink on the Delgerkhaan sheets is the second problem. It is iron gall — oak galls, iron sulfate and gum arabic — which is an acid, and which over centuries browns the fibre around every stroke and can in the end eat clean through it. On these sheets it has done none of that. There is no haloing, no feathering into the fibre, no perforation: a clean, sharp, entirely undamaged mark sitting on paper eight hundred years old. Set against that, the tradition the Mongol chancery actually worked in ground a stick of soot on a stone, and carbon ink does not harm paper at all: the particles are chemically stable and never threaten the sheet’s strength. It sits on the fibre rather than in it. The wrong ink is on these sheets, and it is behaving as the right one would.

A tall grey granite slab standing alone in a glass museum case, its face carrying long weathered incised columns running from top to bottom, lit from the left
The replica in Ulaanbaatar of the stone found near Khirkhira in 1802. The original — 202 cm of granite — went to Saint Petersburg between 1829 and 1832, took damage on the road, and has been in the Hermitage since 1936.

Then there is the hand itself. The Mongolian script was adapted from the Old Uyghur alphabet about 1204, by a scribe called Tata-tonga who had been taken from the Naiman and was then invited to teach the letters to the court, the khan’s sons among them. The earliest inscription in it that anybody can date securely is a granite slab cut about 1225, whose five lines record that the khan’s nephew Yisüngge shot 335 alds — some 536 metres — at an assembly after the Khwarazm campaign. That is the whole of the securely dated corpus from the lifetime in question: five lines about archery. Its letterforms are awkward, and several of them are ambiguous enough that scholars still disagree over what parts of it say.

The Delgerkhaan hand is not ambiguous anywhere. It is regular, evenly spaced, consistent from sheet to sheet, and every letter is distinct from every other letter. That is not what a chancery hand of the 1220s looks like. It is what a script looks like after four hundred years of standardisation, printed in a primer, copied by somebody working slowly with the primer open beside him.

Sovereign of Silence

The first poem is the longest, the best of the five, and the one that gives the game away in its last word. A Mongol ruler had no crown to be unfettered by. What he had was the paiza, a tablet an envoy carried that let him demand horses, rations and a bed at any relay station on the road; one surviving pass says only that by the power of eternal heaven this is an order of the Emperor, and whoever does not show respect will be guilty of an offence. That is the register. Crowns, kings and sovereigns are the translator’s furniture, not the khan’s.

One thing has to be said before the text. All five poems are given here in the English of the Öndörkhaan hand-out, but the lineation and the punctuation have been regularised in the setting — our doing, not the translator’s, and worth saying out loud in a piece that is about to count what falls at the ends of lines. Nothing was added, cut or reordered; the words below are the words that were circulated.

Upon the endless steppe, under the eternal sky,
I stand alone amid the whispering grass.
The wind tells tales of days gone by,
And I listen, my heart as open as the land.

The stars, my silent guides, watch over me,
Their light a reminder of those I have lost
And those I have yet to find. In this hour
I am both king and wanderer,
Seeking not conquest, but understanding.

The world, vast and unyielding,
Echoes with the sound of my horse’s hooves.
Yet within me there is a quiet,
A space where my soul dwells,
Unfettered by the burden of my crown.

Here, in the solitude of the steppe,
I find my truth, my peace.
For though I may command armies,
It is in the silence that I am sovereign.

A lone rider standing beside a shaggy dun horse on open grassland at night beneath a dense field of stars

Night on the Kherlen grasslands, some twenty kilometres from the trench. The pony is the local working sort, thirteen hands and left out all winter; the silent guides overhead are the only thing in the poem that has not changed since 1227.

Dawn’s Respite

Here the forger reaches for a pen, and the pen is the tell. There were no pens on the Kherlen; there was a brush, a stone and a stick of soot. The sword is wrong too, and wrong in a more interesting way. The Mongol arm was the bow — bamboo core, horn glued along the belly, sinew along the back, the whole thing left to cure — drawn off the thumb behind a ring of leather, bone, horn or silver, and carried in pairs, a long one for distance and a short one for horseback. A man of that century laying down his weapon for the evening would not have laid down a sword.

In the hush of dawn, before the world stirs,
I stand among the remnants of the night.
The chill of the morning air, a cloak about my shoulders,
Speaks of battles fought, of victories and defeats.

But in this tranquil hour my thoughts drift to the quiet,
To the plain beauty of the earth beneath.
The dew on the grass, a mirror to the fading stars,
Whispers of resilience, of strength found in gentleness.

“Let the sword rest,” I murmur to the coming day,
“For now I seek the solace of the pen.
In words I find my haven, my escape,
A world where the warrior can mend.”

A pointed grey canvas bell tent and a small open fire with a blackened kettle standing over it at first light, a padded coat thrown across a wooden saddle, a pale green sky with two stars still showing

First light over a camp in Delgerkhaan sum. The saddle is wooden and high at both ends and the coat is a deel — but the tent is a canvas bell of the sort a survey crew carries, not the felt ger a herder sleeps in, and the fire is burning split billets on a plain where the fuel is dried dung, because there is no wood to burn.

The Conqueror’s Soliloquy

This is the poem that has been quoted most, and it is the one that contradicts the record most directly. It shows a man writing at night, alone, at length, for himself. Set that beside the Yassa, which was the most consequential thing he ever ordered into existence and which survives in not one line of its own: everything anybody knows about it comes through Juvayni, through Rashid al-Din’s compendium of 1306–11, through al-Maqrizi writing in the 1440s. A man who did not commit his own law to a page he touched is an odd candidate for a private diary in verse.

Beneath the moon’s watchful eye, I pen my thoughts,
A solitary figure on the canvas of the night.
The ink flows, a river of dreams and doubts,
Charting a course through the landscape of my soul.

Around me the camp lies silent, the warriors at rest,
Unaware what turmoil stirs within their leader’s breast.
For though I wield power and send legions to fight,
It is in the silence of the night I face my test.

“To conquer oneself,” I write beneath the stars’ glow,
“Is the greatest victory, the hardest one can know.
For the heart is a wild thing, untamed and free,
And to master it, a man must first learn to let go.”

A seated man in a padded robe at a low table under a tent by lamplight, a shallow black stone dish and a bamboo brush beside a blank sheet of plain paper, a buckled leather satchel hanging from the tent pole behind him

A reconstruction made for the Öndörkhaan display. The four objects on the table are the whole of a thirteenth-century Mongol chancery scribe's kit — clay lamp, inkstone, bamboo brush, unsized sheet — and the stone is the point: it is where a stick of soot is ground with water. Nothing on this table would make iron gall ink.

Whisper of the Steppe

This is the poem that comes nearest to something real, and it does so by accident. Steppe poetry was not filed; it was sung, and it is sung still. The long song will spend four minutes on ten words and stretch one piece to thirty-two stanzas and three hours, over the two strings of a morin khuur — 130 hairs from a stallion’s tail on one and 105 from a mare’s on the other. The Jangar runs to twenty-five or twenty-six chapters, and past a hundred in some tellings, and is carried by singers called jangarchi. And the verse itself works by head-rhyme: in Mongol alliterative verse each line is a phrase, the stress falls on its first word, and lines are paired so that they alliterate with one another. Not one line of the five Delgerkhaan poems does that anywhere. What two of them do instead is rhyme at the end, in couplets, and they do it in exactly the same shape: eight lines, the first pair unrhymed, the remaining three pairs rhymed. This one runs sky/high, song/long, embrace/space. Shadows and Light runs soul/toll, light/night, beam/dream. The other three are nothing like that. The Conqueror’s Soliloquy rhymes in two of its three stanzas and does it on an a-a-b-a plan rather than in pairs. Sovereign of Silence, the longest and the best of them, gets one rhyme in eighteen lines and puts it between the first line and the third. Dawn’s Respite, twelve lines, has none at all. That the collection is not uniform is the point: two of the five sound like English verse, three do not, and the two that do are rhymed in the English — in a translation. End-rhyme does not survive being carried from one language to another; it has to be built again at the far end by somebody who wants it there, and somebody wanted it in two poems out of five. The steppe kept its poetry in people. It did not keep it in a box.

In the silence of the steppe I hear a whisper,
A voice carried on the wind, ancient and deep.
It speaks of the earth, of roots and sky,
Of the fleeting moment when eagles soar high.

In the whisper is wisdom, a sage’s song,
A reminder that even the mighty can long
For the simple grace of the world’s embrace,
And find in the wild a sacred space.

A golden eagle high above a great sweep of wind-flattened grassland, a low stone cairn hung with pale rags on the ridge below

An ovoo on the ridge above the Tsenker, with the rags that are added a handful at a time by anyone passing. The eagle is a genuine coincidence and the photographer says so.

Shadows and Light

Eight lines, and the weakest of the five: a moon, a shadow, a toll, a dawn. It is also the poem the papers reached for when they wanted to say the collection had been found near the grave, which it was not, because nobody has found the grave. Ikh Khorig, the Great Taboo, is 240 square kilometres around Burkhan Khaldun that were closed to outsiders from 1227 until 1924 by the Darkhad, who were empowered to kill anyone found inside, and then kept closed by the Mongolian People’s Republic on its own account. A Mongolian–Japanese team got in with ultrasound in 1989 and logged about 1,380 possible noble graves. A National Geographic project then put ultra-high-resolution satellite imagery in front of more than ten thousand volunteers, who returned 2.3 million tags across 6,000 square kilometres; the team walked the hundred best and, reporting in 2014, confirmed fifty-five archaeological sites — Bronze Age mounds, deer stones, city walls, ovoos, Mongol-period burials, from ten-metre rock piles to two-hundred-metre walls. Not one of them was the tomb.

Beneath the moon’s gentle glow,
I stand, a shadow on the land.
The night’s own quiet speaks to my soul,
Of battles won and the unforgiving toll.

Yet in the darkness there is also light,
A promise of dawn beyond the night.
It is in this dance of shadow and beam
I find my hope, my dream.

A standing figure in a long coat and a brimmed hat on bare frost-covered ground, a long dark shadow lying across the frost, a line of dark hills along the horizon and a full moon in a deep blue sky

Moonrise over frozen ground in Khentii in late October. Burkhan Khaldun, well to the north of this line of hills, stands 2,340 metres, and its landscape has been a World Heritage Site since 4 July 2015 across 443,739 hectares.

Where the Matter Rests

Nothing above says the box is worthless, and it would be a poor reading of the evidence to conclude that. A larch box with its hasp still shut, a square of silk, four working lists of horses and felt and millet and a name, all of it in the ground four hundred metres from the best-dated Mongol camp in the country: that is a genuinely interesting object, and the four lists may in the end be the part worth publishing. It is the five poems that are the intrusion, and they are an intrusion of a familiar kind. Somebody wanted the man in the ground to have had an interior life, and supplied one, in English metre, in European ink, in a hand off a printing press.

The sheets are in Ulaanbaatar now, where the Chinggis Khaan National Museum opened in October 2022 across nine floors with more than ten thousand objects, of which the museum states over 85 per cent are originals. The Delgerkhaan sheets are not on display. Asked what would settle the matter, the conservator handling them said that a milligram of the ink would do it in a fortnight, and that she had been waiting since January for permission to take it.

She was asked, finally, what she thought the poems were worth. She said they were worth what any wish is worth, and that she would rather have the millet list.