1The Century Punjab Spent Being a Wall
Part Five ended with Punjab becoming a province of a foreign state. Part Six is about what that province was then used for — and the answer is: it was used as a shield, and shields get hit.
There are two halves to this part and they are completely different in character.
The first half, from 1175 to about 1210, is the conquest of northern India. A new dynasty out of the Afghan
hills takes Multan, Uch and Lahore, destroys the Ghaznavids, loses a battle to Prithviraj Chauhan at Tarain, comes
back the following year and wins it — and out of that victory comes the Delhi Sultanate, the state that will
govern most of northern India for three hundred years.
The second half is about survival. From 1221 to 1306, the Mongols — the most successful military system in
human history, the people who destroyed Baghdad and China and Persia and reached Hungary — come down into
the Punjab again and again and again. They sack Lahore. They besiege Delhi. They are stopped, eventually, on the
Ravi.
India is one of the very few settled civilisations the Mongols attacked repeatedly and did not conquer.
That is a genuinely remarkable fact, and it is a Punjab fact, because the reason it happened is in this part.
Four things to hold on to
- Punjab becomes a military zone rather than a province. Under the Sultanate, the country between the Indus and the Sutlej is not really governed — it is garrisoned. Forts, marcher lords, cavalry, and a population that lives with raids the way other places live with weather.
- Whoever commands the Punjab frontier ends up ruling Delhi. This is the structural rule of the whole part, and you should watch for it. It happens in 1320 at the end of this chapter, and it keeps happening for four hundred years afterwards.
- The slave-soldier system from Part Five produces a state. Muhammad of Ghor bought and trained Turkish slave officers exactly as the Ghaznavids had. When he died, those slaves inherited his Indian empire and fought each other for it. The Delhi Sultanate is a slave dynasty, literally and by name.
- The sources improve enormously and stay one-sided. From here on we have detailed, dated, named Persian chronicles — Juzjani writing in the 1260s, Barani in the 1350s, and the poet Amir Khusrau, who was present at several of these events. All of them worked for the Sultanate. There is one contemporary Indian account of Tarain and it breaks off before the battle that mattered.
CERTAIN means contemporary evidence — a writer who was present or in the next room, an inscription, a coin.
LIKELY means good sources writing later, generally accepted. STORY means tradition, often written
centuries afterwards.
Part Six has a specific hazard that you should be warned about at the start. The most famous story in this entire
chapter — Prithviraj Chauhan blinded, taken to Ghazni, and killing Muhammad of Ghor with an arrow
aimed by sound — is not true. It comes from a poem written roughly four hundred years later, and it is contradicted
by a date. This part will show you exactly how that story works, because it is the clearest example in the whole book of
how national memory gets made.
2The Ghurids Take Multan — And Lose to a Queen LIKELY
The man who would conquer northern India began by taking two cities easily and then walking into the worst defeat of his life, inflicted by a regent ruling for a small boy.
When: 1175 Multan; 1176 Uch; 1178 the Gujarat disaster.
Where: Southern Punjab and Sindh, then across the Thar desert into Gujarat.
Who: Mu‘izz al-Din Muhammad of Ghor — “Muhammad of Ghor” — against the Ismaili rulers of Multan and then the Chaulukyas of Gujarat.
The evidence: Persian chronicles, above all Juzjani, writing within living memory; and Gujarati inscriptions and chronicles for the defeat.
Who the Ghurids were
Ghor is a mountainous district in central Afghanistan. Its rulers had been minor vassals of the Ghaznavids for
generations — the same Ghaznavids you watched conquer Punjab in Part Five. In the twelfth century they turned
on their overlords with startling violence.
In 1151 a Ghurid ruler named Ala al-Din Husayn took Ghazna and burned it for seven days. He is known to
history by the nickname he earned there: Jahan-suz, the World-Burner. The Ghaznavid dynasty survived
the destruction of its own capital by retreating to the province it had conquered a century earlier — and so
Lahore, from 1163, became the last capital of the Ghaznavid empire.
By the 1170s two Ghurid brothers ruled between them: Ghiyath al-Din at Firuzkuh, and his younger brother
Mu‘izz al-Din Muhammad at Ghazna. The elder took the west. The younger was given India.
What happened
1. Multan, 1175. Ghor took the city from its Ismaili rulers. This is the same Multan of Part Five — the sun-temple
- city, the Frontier House of Gold — and it changes hands here for the third time in this book.
- Uch, 1176. The fortress town at the confluence of the Chenab and the Sutlej, taken the following year. With Multan and Uch, Ghor held the southern Punjab and the road into Sindh. 3. Then, in 1178, he tried something ambitious and got it badly wrong. Rather than go north into the Punjab plain, he marched south-east across the Thar desert towards Gujarat, one of the wealthiest regions in India.
- He was met at Kayadara, near Mount Abu, and destroyed. The Chaulukya kingdom was at that moment ruled by a child, Mularaja II, with his mother Naikidevi as regent. Gujarati tradition has her riding out with her son and commanding in person.
- The retreat across the desert was worse than the battle. An army that has been beaten in the field and then has to recross the Thar to get home does not arrive intact.
Ghor never tried Gujarat again. After 1178 every campaign he launched went the other way — north into the Punjab
plain, along the road through Lahore and Sirhind towards Delhi.
That is the road that runs through Tarain. A queen regent in Gujarat, defending for a child, redirected the
conquest of northern India onto the route it actually took. If Kayadara had gone the other way, the great
confrontation of this part might have been fought somewhere in Rajasthan against a completely different opponent.
It is also worth noticing what the Persian chroniclers do with this defeat, which is: as little as possible. Juzjani records it
briefly. It is the Indian sources that remember who beat him.
- The desert did half of it. An army that crosses the Thar arrives thirsty, strung out and short of horses. The Chaulukyas fought it at the moment of maximum weakness.
- They chose the ground. Broken country near the hills at Mount Abu, not open plain — the standard answer to cavalry, and one Punjab keeps forgetting.
- They were a functioning state. Gujarat was rich, unified and had a real army, unlike the divided north-west.
- They did not wait to be invaded. They met him at the edge rather than letting him into the country.
- He picked the hardest possible approach. Attacking the richest kingdom in western India across a desert, from a base he had held for two years.
- He had no local ally and no local knowledge. Multan and Uch gave him a foothold, not intelligence about Gujarat.
- He underestimated a child’s regency. A kingdom with a boy on the throne looks weak from outside and often is not, because the nobility closes ranks around the child.
- And he learned from it, which is the genuinely dangerous part. Every later campaign in this part shows a man who does not repeat mistakes.
Muhammad of Ghor is one of the most consequential figures in Indian history and one of the least
romantic. He was not a brilliant tactician. He was not, by the accounts of people who worked for him,
especially charismatic. He had no son and no dynasty to leave anything to.
What he had was an almost inhuman refusal to treat a defeat as final.
He lost at Kayadara in 1178, and it ended his ambitions in western India permanently. He would lose again
at Tarain in 1191, and that one nearly killed him. Both times, the pattern is identical: he is beaten, he goes
home, he works out precisely what went wrong, and he comes back having fixed it.
Compare that with the men he fought. Prithviraj Chauhan beat him at Tarain and did not follow up.
Jayachandra of Kanauj watched Prithviraj fall and did not act. The Ghaznavids at Lahore lost
Peshawar and Sialkot and did nothing but wait for Lahore. Everybody in this war except Ghor treated
each battle as an event. He treated the whole thing as a project.
And he was not even the one who profited. He died childless in 1206 on a riverbank in the Punjab, and
within two years his Indian conquests belonged to a man he had bought as a slave.
The most successful invader of India in five hundred years was a stubborn, unlucky, unloved
administrator from a mountain district in Afghanistan whose whole method was to keep coming
back. Remember that when you read the next three wars, because the temptation is to explain them with
brilliance or destiny, and the honest explanation is persistence.
3Lahore, 1186 — The End of the Ghaznavids LIKELY
The dynasty that conquered Punjab in Part Five was finished off in Punjab, by a former vassal, in a city that had become its last home.
When: 1179 Peshawar; 1185 Sialkot; 1186 Lahore.
Where: The whole northern Punjab, ending at Lahore.
Who: Ghor against Khusrau Malik, the last Ghaznavid ruler.
The evidence: Juzjani and the Persian tradition. No Ghaznavid account of their own end survives.
Why it took seven years
Because Ghor did it the slow, correct way — the way almost nobody in this book does it.
1. Peshawar, 1179. Take the pass first. The doorway that Part Five spent forty pages on.
2. Raids into the plain, through the early 1180s, wearing down the Ghaznavid position without committing to
- a siege.
- Sialkot, 1185. He built or seized a fort at Sialkot and garrisoned it. This is the decisive move and it is not a battle. Sialkot sits north-east of Lahore; a garrison there cuts the Ghaznavid capital off from the hills and sits permanently on its flank.
- Khusrau Malik counter-attacked Sialkot and could not take it — and here is the detail that belongs in this book: he attacked it with Khokhar allies. The Salt Range people from Part Five are back, and this time they are fighting on the Ghaznavid side.
- Lahore, 1186. Ghor came in force, and rather than storm the city he negotiated. Khusrau Malik was induced to come out on terms — sources differ on exactly what was promised. He was seized, sent into captivity, and later killed with his family.
The dynasty that Khusrau Malik ended was founded by a Turkish slave who seized a mountain town called Ghazna in
962. That is War Five of Part Five, two hundred and twenty-four years earlier.
In between, the Ghaznavids had destroyed the Hindu Shahis, annexed the Punjab, made Lahore a capital, ruled from
Iran to the Ganges, and then been driven out of Ghazna itself by the World-Burner in 1151 — surviving for a further
thirty-five years only because they still held the province they had conquered.
There is a hard symmetry in it. Punjab was the first thing the Ghaznavids took and the last thing they had. They
came into this book as invaders and they went out of it as a local dynasty being removed by a bigger one.
- He took the passes before the plain. Peshawar in 1179, seven years before Lahore. Nobody in this book does this and it works every time it is tried.
- He built a fort instead of fighting a battle. Sialkot did more to end the Ghaznavid state than any engagement.
- He was patient. Seven years for one city, from a man who had lost a whole army in a desert eight years earlier.
- He negotiated rather than stormed. Cheaper, faster, and it delivered the enemy king alive.
- They had already lost their homeland. A dynasty ruling its last province from a city it once treated as a colony has no depth left.
- They defended the capital and nothing else. Peshawar went, Sialkot went, and the response each time was to hold Lahore.
- Their allies were not enough. The Khokhars came, and a hill people’s war-band cannot substitute for a field army.
- And they trusted a negotiation. Which was, on the evidence of what happened to Khusrau Malik, a mistake.
There is no famous battle for Lahore in 1186. There is no siege worth painting, no last stand, no defiant
speech recorded by anybody. The Ghaznavid empire ended in a negotiation and a journey into captivity.
The thing that actually killed it was a fort at Sialkot, built or taken the year before, and garrisoned.
This is worth sitting with, because it is the least dramatic lesson in the entire book and one of the most
reliable. Wars in Punjab are usually decided by who is sitting on which piece of ground when the
campaigning season opens, not by what happens on the day two armies meet.
Sialkot cut Lahore off from the hill country and the Chenab. It gave Ghor a base inside Punjab that he did
not have to march to from Ghazna every year. It meant the next campaign started three days from the
enemy capital instead of three weeks. And when Khusrau Malik attacked it with everything he had,
including his Khokhar allies, he could not take it back — which told everybody watching, including his
own nobles, exactly how the war was going to end.
You will see this again immediately. The war that follows this one is fought at Tarain, and Tarain is
where it is fought because of a fort at Bhatinda — and the man who wins the first battle there will then
spend thirteen months besieging that fort instead of finishing his enemy.
4Tarain, 1191 — Won by Prithviraj LIKELY
An Indian army beat Muhammad of Ghor in open battle, and the man who won it made one decision afterwards that cost him everything.
When: 1191 CE.
Where: Tarain — Taraori, near Thanesar, on the plain north of Delhi. The Kurukshetra country.
Who: Prithviraj III Chauhan of Ajmer and Delhi, against Muhammad of Ghor.
The evidence: Persian chronicles for the battle; the Sanskrit Prithviraja Vijaya, written at Prithviraj’s own court, for the Indian side.
Why it happened where it did
Ghor took the fort of Tabarhindah — usually identified with Bhatinda — which sits on the road between the
Punjab plain and the Chauhan territories. That was a direct threat to Prithviraj’s frontier, and he marched.
The two armies met at Tarain, on the flat country near Thanesar. This is the invasion corridor. Every army
coming down the Punjab road towards Delhi passes through here, and the ground around Panipat and Thanesar
will host decisive battles in Parts Seven, Eight and Twelve. You are looking at the most fought-over patch of
farmland in India.
What happened
1. The Indian army was larger and better suited to the ground. Rajput heavy cavalry and elephants against
- a Ghurid force built around mounted archers.
- The Ghurid wings broke. The Persian accounts do not dwell on it, which is itself informative.
- Ghor charged the centre in person and fought Govind Rai, Prithviraj’s brother and the ruler of Delhi. The accounts describe Ghor striking him in the mouth with a lance and knocking out his teeth — and Govind Rai, in return, driving a javelin into Ghor’s arm.
- Ghor was carried off the field. Wounded and slipping from the saddle, he was rescued by a young horseman of his own army, who held him on his horse and got him clear. The chroniclers do not agree on the man’s name; they agree that the campaign ended in that moment.
- The Ghurid army broke and ran, and was pursued for a considerable distance.
- And then Prithviraj stopped. He did not pursue into the Punjab. He turned back and laid siege to Tabarhindah — which held out for thirteen months.
There is a Sanskrit epic called the Prithviraja Vijaya, written by a court poet named Jayanaka at Prithviraj’s own court,
while he was alive. It is the only substantial contemporary Indian account of any of this, and it survives in a single
damaged birch-bark manuscript found in Kashmir.
It celebrates the defeat of the Ghurid invader — this battle, the victory of 1191.
And then it breaks off. The manuscript does not cover what happened the following year.
That is not a conspiracy; it is a damaged manuscript and a court poem that had no reason to continue after its patron
died. But the effect on the historical record is total. The only Indian voice we have from inside this war is
describing a victory, and it falls silent before the defeat. Everything you will read about Tarain in 1192 was written
by the winners.
- Numbers and weight. A larger army with heavy cavalry and elephants, on ground where it could close with the enemy.
- He came to the enemy. He did not wait for Ghor to consolidate around Bhatinda; he marched immediately.
- He broke the wings first. Mounted archers are only dangerous while they have room. Crowd them off the flanks and they cannot work.
- He nearly killed the enemy commander. In an army built around one man, that is the whole battle — and Govind Rai, at the cost of his own teeth, very nearly did it.
- He did not pursue. A beaten army that gets home is next year’s army. This is the single most expensive decision in the chapter.
- He spent thirteen months on a fort. Thirteen months is enough time to raise, equip and train an entirely new invasion — and that is exactly what it was used for.
- He did not use the year to build a coalition. The alliance he brought to Tarain the second time was assembled in an emergency, not during the year he had.
- He treated it as a border war. Ghor treated it as a project. That difference is the whole of the next section.
Between the first battle of Tarain and the second, there is a gap of about a year. Both men had it. Look at
what each did with it.
Muhammad of Ghor went back to Ghazna. According to the Persian tradition he disgraced the officers
whose contingents had broken — one account has them made to eat from nosebags in public. He raised a
new army on a different principle: overwhelmingly mounted archers, drilled to work as separate mobile
divisions rather than one mass. He spent the winter preparing. He came back in the campaigning season
with a force built specifically to beat the army that had beaten him.
Prithviraj Chauhan besieged a fort.
That is not a joke at his expense, and it is worth being fair about it. Tabarhindah was a real strategic
problem: a hostile garrison sitting on his frontier road. Reducing it was a legitimate objective, and medieval
sieges were slow. He was also fighting other wars — the Chauhans had been in conflict with the Chandelas
and were on bad terms with Kanauj — and a king does not get to fight only the enemy that history says was
important.
But the arithmetic is brutal and it does not care about fairness. Thirteen months went into a fort. The
man he had beaten used the same months to redesign an army.
This is the third pattern of the whole book appearing in its purest form. Punjab and its neighbours produce,
again and again, commanders who win the battle and lose the war — because a victory that is not followed
up is a delay, and the invader in this book always, always comes back.
5Tarain, 1192 — The Battle That Made India Muslim-Ruled LIKELY
The same ground, the same two men, one year later — and one of them had spent the year thinking.
When: 1192 CE.
Where: Tarain again — the same plain near Thanesar.
Who: Ghor against Prithviraj at the head of a large confederacy of Rajput rulers.
The numbers claimed: Ferishta gives Ghor 120,000 horse and the Indian side several hundred thousand. Both are advertising.
What Ghor did differently
1. He built an army of one type. The new force was overwhelmingly mounted archers — light, fast, and
- capable of shooting from the saddle at a distance no Indian formation could answer.
- He divided it. Juzjani describes four divisions of ten thousand horse each, sent to harass the Indian army from all four sides, with a reserve of about twelve thousand heavy cavalry kept back under his own hand.
- He lied about wanting peace. Prithviraj sent a demand that he withdraw. Ghor replied that he was only a brother and a general and must await instructions from Ghiyath al-Din — implying a truce was being considered. It was a delaying message and it worked.
- He attacked at dawn. The Indian army, expecting negotiation and following the ordinary routine of a camp — morning washing, morning prayers, horses not yet saddled — was struck before it was formed up.
- Then he refused to close all day. The four divisions rode in, shot, and withdrew. Any Indian charge found nothing to hit. This went on for hours, in armour, in the heat, with casualties accumulating and nothing to strike back at.
- In the evening he sent in the reserve. Twelve thousand fresh heavy horse against an army that had been standing, sweating and dying since dawn. It broke. 7. Govind Rai was killed. Prithviraj was captured. Both certainties. What happened to him afterwards is the subject of the story below.
Ferishta’s figures — 120,000 Ghurid horse, and Indian totals running to three hundred thousand men and thousands of
elephants — are from 1600 and are not usable. Feeding 120,000 horses on a single plain is not possible.
What is solid: the Indian army was substantially larger; it was a confederacy of many rulers rather than one king’s
force; and the Ghurid army was smaller, homogeneous and under a single command. Those three facts explain the
result without needing any number at all.
And note the shape of it, because you have seen it four times already. A coalition of many kings against a single
command. Ten kings on the Ravi in Part One. The Malavas and Kshudrakas in Part Two. Yashodharman’s alliance that
dissolved on winning in Part Four. Anandapala’s confederacy at Waihind in Part Five. It is the most reliable losing
formation in Indian military history and it will lose again at Panipat, twice.
- He refused the battle his enemy wanted. The whole Indian system depended on closing with the enemy. He spent a day making sure that never happened.
- Mobility beat weight. Horse archers in loose divisions against heavy cavalry and elephants on an open plain is a solved problem, and he had solved it over the winter.
- He used deception without embarrassment. The false negotiation bought him the dawn, and the dawn bought him the battle.
- He kept a reserve and used it late. The commonest failure in this book is committing everything at once. He committed twelve thousand men after ten hours.
- He believed a message. A commander facing an enemy who has already been beaten once has no business relaxing his camp routine on the strength of a letter.
- His army could not be commanded as one thing. Dozens of rulers, each with his own contingent, own honour and own opinion, and no mechanism for a single decision under pressure.
- His equipment was wrong for the fight he got. Heavy armour and elephants are decisive at contact and a liability across ten hours of being shot at.
- He had given the enemy a year. Everything above is downstream of that.
Here is the version almost every Indian schoolchild knows.
Prithviraj is captured and taken to Ghazni, where he is blinded. His loyal court poet Chand Bardai comes
to him in captivity, and together they plan a revenge. Chand Bardai persuades Muhammad of Ghor that the
blind king can still shoot an arrow guided by sound alone. A demonstration is arranged. At the moment of
the shot, the poet recites a couplet that tells Prithviraj — in verse, without the court understanding —
exactly where the sultan is sitting. Prithviraj looses the arrow at the sound and kills him. The two friends
then kill each other rather than be taken.
It is a magnificent story. It comes from the Prithviraj Raso, a Braj-language epic whose surviving form dates
from roughly the sixteenth century — about four hundred years after the events.
And it is contradicted by a date. Muhammad of Ghor died in 1206, fourteen years after Tarain, on a
riverbank in the Punjab, and that is the subject of the very next war in this book. Whatever
happened to Prithviraj, he did not shoot the man who beat him.
So what did happen? The Persian sources say he was executed. There is also a striking piece of physical
evidence: coins survive carrying Prithviraj’s name on one side and Muhammad of Ghor’s on the
other, which suggests he was kept alive for a short time as a tributary ruler at Ajmer before being put to
death — the standard account being that he was executed for conspiring against his new overlords.
Now the part that matters, and it is not really about Prithviraj at all.
The Raso is not history and it was never trying to be. It is what a defeated tradition does with an
unbearable memory: it gives the loss a meaning, gives the dead king a last act, and lets the villain die by the
hero’s hand. Every culture in this book does this. The Greeks did it to Cyrus in Part Two. Kashmir did it to
Mihirakula in Part Four.
You are allowed to love the story. You are not allowed to teach it as fact — and the reason is in the next four
pages, where Ghor is still alive, still campaigning, and about to be killed in a much less satisfying way by
people nobody has written a poem about.
6Damyak, 1206 — The Khokhars Kill a Sultan LIKELY
The conqueror of northern India was killed in a tent on the bank of the Jhelum by people from the Salt Range whose names nobody wrote down.
When: The rising in 1204–05; the assassination on 15 March 1206.
Where: The Salt Range and the Jhelum; the killing at Damyak, near Sohawa in Jhelum district.
Who: Muhammad of Ghor against the Khokhars, with Qutb al-Din Aibak coming up from Lahore.
The evidence: Juzjani and the Persian chroniclers, who disagree with each other about who did the killing.
Why the Khokhars rose
Because Ghor lost a battle two thousand kilometres away, and news travels.
In 1204, at Andkhud in what is now northern Afghanistan, Muhammad of Ghor was defeated by the
Khwarezmian empire and its allies. It was a serious defeat, and a rumour ran ahead of the truth: that the sultan
was dead.
The Punjab took the rumour at face value and rose. The rising was led by the Khokhars, the people of the Salt
Range and the Jhelum country whom you met in Part Five, charging bare-headed into the Ghaznavid line at
Waihind in 1008 and nearly breaking it. They cut the road between Lahore and Ghazna — which meant that the
Ghurid empire in India was severed from the Ghurid empire in Afghanistan.
What happened
1. Ghor came back in person, in 1205–06. He would not delegate this one, which tells you how dangerous he
- thought it was.
- A hard campaign in the Jhelum country. The accounts describe heavy fighting and heavy losses, in the broken salt-and-ravine country that Part Five described as the best defensive ground in Punjab. 3. Aibak came up from Lahore and caught the Khokhars between two forces. The rising was crushed with considerable slaughter.
- Ghor turned for home, having reopened the road, and camped on the Jhelum at Damyak.
- On the night of 15 March 1206, assassins entered his tent and killed him. He was about sixty-two and had no son.
The Persian sources do not agree, and it is worth setting the options out honestly rather than picking the exciting one.
Version one: Khokhars. Survivors of the campaign he had just finished, taking revenge in the obvious way, on their
own ground, two weeks after he had massacred them. This is the version most commonly repeated and it has the merit
of motive and geography.
Version two: Ismaili assassins. The fidais of the Nizari Ismailis, whose speciality was exactly this — infiltrating a
camp and killing a ruler in his tent. Ghor had made himself a target: he had taken Multan from an Ismaili dynasty in
1175, and Ismaili communities had been persecuted under his rule.
Version three: his own people. A minority reading points at the court, since a childless sultan whose empire had just
been cut in half is a problem with an obvious solution.
This book cannot tell you which. What it can tell you is that the Khokhar version is the one that stuck, and that it
stuck partly because it makes narrative sense — the wronged hill people avenging themselves on the tyrant. Be a little
suspicious of a story that satisfying, even when it is probably true.
- He came himself, immediately. A revolt that cuts your empire in half is not a job for a governor.
- Two armies from two directions. Ghazna and Lahore closing on the same country is a trap that a regional rising cannot escape.
- He had a lieutenant he could trust. Aibak had been running India for thirteen years and did exactly what was needed.
- They acted on a rumour. The sultan was not dead, and rising on bad intelligence is how a rebellion becomes a massacre.
- They were a people, not a state. No fortress line, no field army, no ally, no way to hold ground against two converging armies.
- But they could not be finished. This is the point. He beat them, and two weeks later he was dead in his tent, and Khokhar rebellions go on troubling Delhi for another two hundred years.
- And they had already cost him the empire. The road he came to reopen closed again the moment he died.
Muhammad of Ghor conquered more of India than anybody since the Mauryas. He ended the Ghaznavid
dynasty, destroyed the Chauhans, took Delhi, Ajmer, Kanauj and Benares, and founded — indirectly, by
dying without an heir — the state that would govern northern India for three centuries.
He is buried in a small, plain tomb beside a road in Jhelum district, in the Pakistani Punjab, a few
kilometres from where he was killed.
It is not a monument on the scale of what he did. It is a modest building in a village, and for most of its
existence it was in poor repair, known mainly to the people who lived around it.
Think about the geography of that for a moment. The man who took Delhi lies in the Salt Range, in the
country of the people who probably killed him, on the bank of the river where Alexander fought
Porus in Part Two and where Ranjit Singh’s empire will be built in Part Thirteen. Everything in this
book keeps happening on the same forty kilometres of ground.
And there is one more thing about that grave, which is the real point of this war. Ghor lost every
important thing he had at the moment he died. No son. No dynasty. His empire split in three within
two years. His Indian conquests went to a man he had bought in a slave market.
The Delhi Sultanate — the thing that actually came out of Tarain — was not founded by the conqueror. It
was founded by his property.
7The War of the Slaves, 1206–1228 LIKELY
Three men who had all been bought as boys fought each other for twenty years over the Punjab, and the winner founded a dynasty called, without embarrassment, the Slave Dynasty.
When: 1206–1228, from Ghor’s death to the drowning of the last rival.
Where: The whole Punjab — Lahore, Multan, Uch, Sirhind, and Tarain again.
Who: Qutb al-Din Aibak, then Iltutmish at Delhi; Taj al-Din Yildiz at Ghazna; Nasir al-Din Qabacha at Multan.
The evidence: Juzjani, who served this dynasty and knew several of these men personally.
What a slave dynasty actually means
Part Five explained the system and this is where it produces a country. Turkish boys were bought from the steppe,
trained as soldiers, educated, converted, and promoted on ability. A slave officer had no clan, no lands and no
family faction, so his loyalty ran to his master alone. It made him the ideal instrument — and, on his master’s
death, the only person holding the weapons.
Muhammad of Ghor was asked, according to a story the chroniclers liked, whether it grieved him to have no sons.
He is said to have replied that other kings had one son or two, and that he had thousands — meaning his
Turkish slaves, who would inherit his empire and keep his name in the Friday prayers. Treat it as a good
line rather than a transcript. It is also, as it turns out, exactly what happened.
What happened
1. The empire splits three ways. Aibak takes Delhi and Lahore; Yildiz takes Ghazna and claims to be the senior heir; Qabacha takes Multan, Uch and the Sindh country. All three were Ghor’s slaves. All three now had to fight.
2. Aibak is enthroned at Lahore in 1206. Note the address. The Delhi Sultanate begins in Punjab.
3. Aibak dies in 1210 falling from a horse while playing chaugan — polo — at Lahore. He is buried there,
- and his tomb is still in the city. 4. Iltutmish takes over. He had been Aibak’s own slave, then his son-in-law: a Turk of good family sold into slavery by jealous brothers, bought in Bukhara, bought again in Delhi, and now sultan. He spent the next twenty-five years turning a war-band into a state.
- Tarain, 1216. Iltutmish met Yildiz on the same plain that had already decided this book twice in twenty-five years, beat him, and had him executed. Three battles at Tarain in one generation.
- Qabacha holds out at Multan until 1228, and then Iltutmish came for him. Driven from Uch and cornered on the Indus at Bhakkar, Qabacha drowned in the river — whether by accident while fleeing or by his own choice, the sources are not sure.
- And by 1229 Iltutmish held it all, with formal recognition from the caliph in Baghdad, which mattered enormously at the time.
In 1221 a fugitive prince came to him asking for help. That prince was Jalal al-Din Mingburnu, the last Khwarezmian
ruler, and the army chasing him belonged to Genghis Khan.
Iltutmish said no. Politely, at length, and with a plausible excuse — the sources have him replying that the climate of
the region he could offer would not suit an honoured guest. It was a refusal dressed as hospitality, and it may have
saved the Sultanate.
Sheltering Jalal al-Din would have made Delhi a Mongol target at a moment when the Mongols were
destroying every state that annoyed them. Iltutmish had been sultan for barely a decade, his rivals were still alive
in Multan and Ghazna, and his army had never faced anything like the force sitting on the Indus.
It is one of the coldest and most correct decisions in this book, and it is the reason the next war is a Mongol campaign
in the Punjab rather than the end of the Delhi Sultanate.
- He held the middle. Delhi and the upper Punjab sat between Ghazna and Multan. He could fight either rival without both at once, and he did exactly that — Yildiz in 1216, Qabacha in 1228.
- He built institutions, not just an army. A coinage, a land-assignment system, a corps of loyal officers, and recognition from the caliph. His rivals ran military governments; he ran a state.
- He knew which wars not to fight. The refusal to Jalal al-Din is the clearest example, and there are others.
- He outlived them. Twenty-six years on the throne, in an age when most claimants got five.
- Yildiz was facing the wrong way. Based at Ghazna, he claimed India while the Khwarezmians and then the Mongols came at him from behind. He lost Ghazna and arrived at Tarain as a refugee with a claim.
- Qabacha ruled the poorest share. Multan and Sindh could not out-produce the Ganges plain, and twenty years of holding on is not the same as winning.
- Neither of them combined. Two rivals who both faced the same stronger enemy and never allied. The oldest failure in this book, now being committed by the conquerors.
- And the Mongols were behind them both, which meant they were fighting for the Punjab with an axe hanging over their heads.
Iltutmish was born into a Turkish family of the Ilbari tribe on the steppe, and by the account he told his
own chroniclers, he was a boy of striking looks and ability whose brothers were jealous enough to take
him to a horse fair and sell him.
He passed through several owners, was bought in Bukhara, and eventually reached Delhi, where Qutb al-
Din Aibak — himself a slave, himself bought as a boy — purchased him. He rose through the ranks,
married Aibak’s daughter, was made governor of Badaun, and in 1211 took the throne.
He then ruled for twenty-six years, defeated two rival sultans, refused Genghis Khan’s enemy, obtained
recognition from the caliph in Baghdad, issued the silver tanka that became the standard coin of northern
India for centuries, finished the Qutb Minar, and died in his bed in 1236 having handed on a functioning
state.
And then he did one more thing that almost nobody in this book does. He named his daughter as his
successor. Passing over his surviving sons, he designated Raziyya, on the grounds — as Juzjani reports the
argument — that she was more capable than any of them. She ruled Delhi for three and a half years before
the Turkish nobility destroyed her.
Put the whole life in one line and it stops sounding like medieval history and starts sounding like an
argument. A boy sold by his own family became the founder of the most durable state in medieval
India, and used his last political act to try to hand it to a woman.
The system that produced him was slavery. That is not a contradiction to be smoothed over — it is the
actual mechanism. The Delhi Sultanate promoted purely on ability, because it had removed birth
from the equation by buying its ruling class as children. It is one of the strangest facts in Indian
history and it is why this dynasty, and not Ghor’s, is what Tarain really produced.
8The Indus, 1221 — Genghis Khan Looks at India CERTAIN
The most dangerous man in the world stood on the bank of the Indus, watched an enemy escape by jumping a horse into the river, and then decided not to come any further.
When: 24 November 1221 for the battle, with Mongol operations in the region continuing into 1223.
Where: The Indus, somewhere in the Attock–Kalabagh country, and then the western Punjab and Sindh.
Who: Genghis Khan against Jalal al-Din Mingburnu, the last Khwarezmian prince.
The evidence: Juvayni and Rashid al-Din on the Mongol side; Nasawi, Jalal al-Din’s own secretary, on the other. Unusually good.
How the Mongols got here
In 1218 the Khwarezmian empire — which ruled Iran and Central Asia — massacred a Mongol trade caravan and
then executed Genghis Khan’s envoys. What followed is one of the most complete destructions of a state in history.
Bukhara, Samarkand, Merv, Nishapur, Herat: taken and, in several cases, effectively erased.
The Khwarezmian shah fled and died on an island in the Caspian. His son Jalal al-Din did not run. He raised what
troops he could in Afghanistan and, at Parwan, beat a Mongol army in the field — one of very few commanders
anywhere to manage it.
Genghis Khan came after him personally. He caught him on the Indus in November 1221.
What happened
1. Jalal al-Din was pinned against the river with no boats — the trap this book has described at Peshawar in
- 1001 and will describe again at Sobraon in Part Fourteen.
- He fought well anyway. The accounts describe him breaking the Mongol centre before being enveloped. Genghis extended his wings and closed them.
- And then he jumped. With the battle lost, Jalal al-Din turned his horse, rode to the bank — which the sources describe as a high one — and put the animal over the edge into the water. He swam the Indus with his sword and shield, and came out on the far side, in the Punjab.
- Genghis Khan watched him do it. Juvayni reports the Khan putting his hand to his mouth in astonishment and telling his sons that a father should be fortunate to have such a son.
- He then ordered the pursuit. Mongol detachments crossed into the Punjab and Sindh and hunted Jalal al-Din for months, raiding Multan and the surrounding country.
- Jalal al-Din survived, asked Iltutmish for shelter, was politely refused, and eventually left India westward. He was killed years later in Anatolia.
- And Genghis Khan turned around.
Genghis Khan spent months on the Indus with an intact army and northern India in front of him, weakly held by a new
sultanate with unresolved civil wars. He considered marching through India and returning to Mongolia the long way
round. Then he went back the way he came.
The reasons given in the sources are a mixture of the practical and the marvellous. Heat and disease — Mongol
armies and Mongol horses did badly in the Indian climate, and this is the explanation most historians find sufficient on
its own. Bad water and bad going in the country ahead. And an omen: a famous story in which a strange beast
appears to the army at the river and speaks, and the Khan’s Khitan adviser Yelü Chucai interprets it as heaven’s
warning to turn back.
The omen story is STORY and is usually read as a wise minister dressing up practical advice in a form a warlord
would accept. The climate is almost certainly the real answer, and it is worth stating plainly what it means:
northern India was saved in 1222 by its own weather.
- Every man was mounted, and had spare horses. An army that moves at that speed chooses every engagement it fights.
- Discipline by unit, not by heroics. Decimal organisation, signal flags, drilled feigned retreats. It was the most professional army in the world by a distance.
- They enveloped rather than charged. Exactly what happened to Jalal al-Din: hold the centre, extend the wings, close.
- They fought a war of extermination and advertised it, so that most cities surrendered before the army arrived.
- Climate. The plains of India in summer are hostile to a horse-based army from the steppe in a way no defender could have arranged.
- It was a pursuit, not an invasion. Genghis came here chasing one man, not to take a country. Objectives matter.
- Iltutmish refused to give him a reason. The single best decision made by anyone on the Indian side in this century.
- And the empire turned east. Genghis died in 1227 and Mongol strategic attention went to China and the west. India was a flank.
There is something worth pausing on in that scene at the river, and it is not the jump.
Genghis Khan had spent three years destroying Jalal al-Din’s family, his cities and his empire with a
thoroughness that is still visible in the archaeology of Central Asia. He had come to the Indus specifically to
kill this man. He had him surrounded, beaten and against a river.
And when the man escaped by doing something spectacular, the Khan reportedly stopped, watched, and
told his own sons to look at him — because that was what a son should be.
The remark is preserved by Juvayni, a Persian historian writing for the Mongol administration a generation
later, whose own family had suffered in the conquest and who is not, as a rule, generous about his
employers.
It is worth keeping for two reasons. The first is that it is a rare, credible glimpse of a man who is otherwise a
statistic in most people’s heads — a commander who could admire the enemy he had just failed to kill.
The second is the geography, and this book keeps insisting on geography. On one bank stands the man
who conquered more of the world than anyone before or since. On the other bank, dripping and
alive, stands a prince who has just crossed into the Punjab.
Whatever else Punjab is in these sixteen parts — corridor, prize, battlefield, shield — on that November
afternoon it was the far bank. The place you get to if you are lucky. For the next eighty-five years, its job
is to be the bank the Mongols do not cross.
9The Sack of Lahore, 1241 LIKELY
The great city of Punjab was stormed and emptied in a single winter, and the Sultanate that was supposed to defend it effectively wrote it off.
When: December 1241.
Where: Lahore — capital of the Ghaznavids in Part Five, and of Aibak in War Six.
Who: A Mongol army under a commander the sources call Tair, against the Sultanate garrison.
The evidence: Juzjani, who was in Delhi at the time and had every reason to know.
What happened
1. The Sultanate was in chaos. Iltutmish had died in 1236. Raziyya had been overthrown and killed in 1240.
- Delhi was on its fourth ruler in five years and the Turkish nobility was fighting over the throne.
- The Mongols came down through the passes in strength and invested Lahore in December.
- The governor fled. Malik Ikhtiyar al-Din Qaraqash abandoned the city and made for Delhi. Juzjani, who does not usually editorialise, is clear about what this meant.
- The citizens fought on without him, and the walls were breached.
- The city was sacked. The population was slaughtered or enslaved on a scale that the chroniclers do not attempt to quantify.
- Delhi did not come. No relief force was sent. This is the fact that shapes the next sixty years.
Look at what the Mongol empire was doing in the same twelve months, because it puts Lahore in a frame that Indian
accounts rarely provide.
In April 1241, Mongol armies destroyed a Polish and German force at Legnica and, two days later, annihilated the
Hungarian army at Mohi. Central Europe lay open. In December 1241 they stormed Lahore. And in that same
December the Great Khan Ögedei died in Mongolia, and the armies in Europe turned around and went home to attend
to the succession.
In one year, the same empire was sacking a city on the Ravi and a city on the Danube. Nothing in this book
gives a better sense of the scale of what Punjab was up against — or of how little the people of Lahore could possibly
have known about why it was happening to them.
- They came at the perfect moment. Five years of succession chaos in Delhi, and a frontier province with a governor who did not want the job.
- Lahore was exposed. It sits close to the Ravi and to the road from the passes, with no natural defence in front of it.
- They had siegecraft. The Mongols had absorbed Chinese and Persian engineers and were, by this date, the best besiegers in the world.
- And nobody relieved it. A city can hold if it believes an army is coming. Lahore knew none was.
- It was fighting itself. The nobility that had destroyed Raziyya was busy consuming the next several sultans.
- It had made a strategic judgement, and it was brutal. Delhi decided it could not hold the far Punjab against the Mongols and would defend the Sutlej line instead.
- Its governors had no incentive to die there. A frontier command with no relief promised is a posting to survive, not a duty to hold.
- And it worked, from Delhi’s point of view. That is the uncomfortable part of the story below.
Here is the strategy the Delhi Sultanate adopted after 1241, stated as plainly as the chroniclers never quite
state it.
Punjab would be defended as far as was cheap, and no further. The Sultanate held the line of the Beas
and Sutlej with forts and garrisons, kept its field army for the doab, and treated the country between the
Indus and the Ravi as a buffer — a zone where Mongol armies could be allowed to arrive, spend a season,
exhaust their supplies and go home, without ever threatening the taxable heartland around Delhi.
Judged as policy it was rational and it succeeded. The Delhi Sultanate is one of the very few states in Eurasia
that faced repeated Mongol invasion and was still standing at the end of it. The Ganges plain was never
systematically devastated. The revenue kept coming. The dynasty survived.
Judged from Lahore it looks different. For roughly sixty years, the great city of the Punjab was a place
the empire had decided not to defend — raided, occupied at intervals, its trade broken, its population
reduced, and its governor a man whose realistic options were flight or death.
This is what it means to be a frontier rather than a heartland, and it is the same lesson as the Shahis in Part
Five with the polarity reversed. There, Punjab fought alone because nobody would come. Here, Punjab
bleeds because a rational government in Delhi has done the arithmetic and concluded that it is cheaper to
let it.
Every part of this book so far has asked why nobody defends Punjab. This is the first part where
the answer is a policy rather than a failure.
10Balban’s Iron Frontier, 1266–1287 LIKELY
A sultan who had also been bought as a slave decided that Punjab would be turned into a fortified military district — and it cost him his son.
When: 1266–1287, Balban’s reign as sultan; he had effectively run the state for twenty years before that.
Where: The line of forts from Bhatinda and Sunam to Samana and Dipalpur, and the Multan frontier.
Who: Ghiyath al-Din Balban against the Mongols of the Chagatai realm, and against everybody else.
The evidence: Barani, writing seventy years later, and the poet Amir Khusrau, who was there and was captured.
What he did
Balban is not a likeable figure and this book will not pretend he is. He came to power over the bodies of rivals,
ruled by terror, executed on suspicion, and invented an elaborate court ceremonial designed to make himself seem
more than human. He also, more than any other individual, is the reason the Delhi Sultanate survived the
thirteenth century.
1. He refused to campaign anywhere else. No southern adventures, no eastern conquests. For twenty years the
- field army stayed within reach of the frontier. Given what every other ruler in this book does with an army, that restraint is remarkable.
- He rebuilt and garrisoned a fort line. Bhatinda, Bhatnair, Sunam, Samana and the forward posts around Dipalpur and Lahore were repaired, stocked and manned as a permanent system rather than as individual strongpoints.
- He cleared the forests of the doab and settled soldiers along the roads, because the raiders were not only Mongol — the country between Delhi and the Punjab had become unsafe on its own account.
- He put his own family on the frontier. His eldest son Muhammad, the heir, was made governor at Multan and Lahore. This is the crucial decision and it is the one that destroys him. 5. And he mostly did not fight pitched battles. The system was designed to absorb raids, deny supplies, and make a Mongol campaign into the Punjab unprofitable rather than impossible.
In 1285 a Mongol force came down into the Multan country. Prince Muhammad — called Khan-i-Shahid, the martyr
prince, afterwards — went out to meet it and was killed, apparently caught by a detachment while his own force was
dispersed.
The poet Amir Khusrau was in his service and was taken prisoner in the same action. He later wrote about it, and his
account of the prince’s death and of his own captivity is one of the very few pieces of first-hand testimony about the
Mongol wars written by somebody who was actually caught in one.
Barani says Balban never recovered. He was already an old man; he had built a system, held a frontier for twenty years,
and lost his heir to the thing he had built it against. He died in 1287, and within three years his dynasty was gone.
- Depth instead of a wall. Not one line but a belt of forts, so a raid that got past the first post was still inside a defended zone with nothing to eat.
- Permanent garrisons, permanently paid. The most boring and most decisive item on this list. Part Five’s Shahis lost their last king because his troops were not paid.
- Strategic patience. Balban did not chase glory. Twenty years of not invading anybody is what made the frontier affordable.
- He accepted the loss of the far Punjab. Ugly, and it worked.
- The heir. A frontier command is a real command, and real commands kill people.
- The dynasty. With Muhammad dead and Balban old, the succession collapsed and the Khalji faction took the throne in 1290.
- The far Punjab. Lahore stayed a damaged, half-held city throughout.
- And it did not stop the Mongols coming. It made them unprofitable, which is a different thing — as the next three wars show.
Almost every voice in this part belongs to a chronicler writing in a library in Delhi years afterwards. Amir
Khusrau is different: he was on the field, he was captured, and he came back and wrote.
He is one of the great figures of Indian culture — a poet in Persian and, by tradition, in the early Hindavi
that stands behind modern Hindi and Urdu; a musician credited by tradition with instruments and forms
still played today; a man who served seven sultans and outlived nearly all of them. He is buried next to his
teacher, the Sufi saint Nizamuddin Auliya, in Delhi.
And in 1285 he was a court poet attached to a prince on the Multan frontier, and the Mongols came, and his
patron was killed, and he was taken.
What he wrote afterwards is not a battle report. It is grief, and it is a description of what the enemy looked
like from close up — faces, smell, the sound of the language, the ordinary terror of being a captive in a
moving camp. He got away and made it home.
Every other war in this part is described by men who heard about it. Juzjani in Delhi. Barani seventy
years later. Ferishta from another century entirely. Khusrau is the one who was there, and the thing he
chose to record is not the manoeuvre or the casualty count but what it was like.
That is worth more than a dozen numbers, and it is the closest this book gets, in the whole medieval
section, to a Punjabi war seen from inside it.
11Jaran-Manjur, 1298 — The Frontier Fights Back LIKELY
For seventy years the policy had been to absorb the raids. Then Delhi got a sultan who decided to meet them at the river and destroy them.
When: 1298 CE.
Where: Jaran-Manjur, on the Sutlej.
Who: Ulugh Khan and Zafar Khan, generals of Alauddin Khalji, against a Mongol force under Kadar.
The evidence: Barani and Amir Khusrau. Khusrau was writing for this court and celebrating these victories.
What had changed in Delhi
Alauddin Khalji took the throne in 1296 by murdering his uncle, and proceeded to run the most aggressively
organised government medieval India had seen. Price controls in the markets, a standing army paid in cash rather
than land, a systematic revenue assessment, a network of informers, and a habit of executing anybody who looked
like a rival.
Almost all of it was built to pay for one thing. He needed a large, permanent, cash-paid army, and he needed
it because the Mongols were coming every other year. The famous price controls in Delhi’s markets exist so
that a soldier’s fixed salary would keep buying the same amount of grain.
Balban had absorbed the raids. Alauddin decided to fight them.
What happened
1. A Mongol army under Kadar crossed into the Punjab in 1297–98 and raided down towards the Sutlej.
2. Alauddin sent his brother Ulugh Khan and his best field commander Zafar Khan to intercept — not to
- shadow, not to contain, but to bring it to battle.
- They caught it at Jaran-Manjur on the Sutlej. A raiding army encumbered with plunder and captives, on the wrong side of a river, is a very different proposition from a Mongol army manoeuvring freely. 4. It was destroyed. Barani gives the prisoners as around twenty thousand. Treat the figure as a court number and the outcome as real.
- The prisoners were taken to Delhi and killed. Alauddin made a public spectacle of it. This is deliberate policy, and it is discussed below.
Alauddin’s treatment of captured Mongols was systematic and it was appalling. Prisoners were executed in large
numbers at Delhi; the sources describe them being trampled by elephants and their heads built into towers. In 1298 he
also had the settled Mongol converts living near Delhi — the so-called New Muslims — massacred after a mutiny,
along with, by Barani’s account, their families.
This book is not going to launder that by calling it deterrence and moving on. It was mass killing of prisoners and of
a resident community, and Barani, who admired Alauddin, records it without much discomfort.
It is also true that it was a policy with an objective, and that the objective was achieved: after 1306, Mongol invasions
of India essentially stop for twenty years. Both those statements belong in the same paragraph. A history that gives
you only the second one is propaganda, and one that gives you only the first is not explaining why it happened.
- They caught a loaded army. Raiders carrying plunder and driving captives cannot manoeuvre, and will not abandon the loot.
- A river at the enemy’s back. The oldest trick in this book, used for once by the Indian side.
- A standing, paid, professional army. Alauddin could send a field force at short notice because he had one sitting ready.
- Intelligence. The informer network that made Alauddin hated at home also meant he generally knew where a Mongol column was.
- They were raiding, not campaigning. The objective was plunder, and plunder is heavy.
- They had grown used to a passive enemy. Seventy years of Balban-style absorption had taught them that nobody came out to fight.
- They crossed a river with a hostile field army in range. A mistake they had not needed to worry about before.
- And the Chagatai realm was divided, sending forces on the initiative of individual princes rather than as a single strategy.
Here is a chain of cause and effect that almost nobody puts together, and it is the most interesting thing
about this reign.
A Mongol army is riding towards the Sutlej. To stop it, Delhi needs a large cavalry force that exists all the
time — not a feudal levy summoned when trouble appears, but professional soldiers on a permanent
payroll.
Paying them in land grants does not work, because a soldier with land becomes a landholder with opinions.
So Alauddin paid them in cash, at a fixed salary.
But a fixed salary is only as good as what it buys. If the price of grain rises, the soldier is effectively taking a
pay cut, and either the salary must rise — which the treasury cannot afford at that scale — or the army
starts to dissolve.
So Alauddin fixed the prices. Grain, cloth, horses, cattle, slaves — all set by decree, with state
granaries to break hoarding, licensed dealers, registered merchants, inspectors in the markets
and savage punishments for short weight. It is the most thorough system of state price control in the
medieval world, and Barani describes it in obsessive detail.
It was not economic theory and it was not benevolence. It was a war measure, and the war was on the
Sutlej.
Keep this in mind for the rest of the book, because it is one of the quiet themes. What happens on the
Punjab frontier reaches into the price of bread five hundred kilometres away. A raid on the Sutlej
changes what a weaver in Delhi pays for grain, because of a decision made by a frightened government
about how to keep enough horsemen in the field.
12Kili, 1299 — The Year the Mongols Reached Delhi LIKELY
This one was not a raid. It was an invasion aimed at taking the country, and it got to the outskirts of the capital.
When: 1299 CE.
Where: Kili, near Delhi — after the Mongols had marched the length of the Punjab without stopping to plunder.
Who: Qutlugh Khwaja, son of the Chagatai ruler Duwa, against Alauddin Khalji in person.
The evidence: Barani, and Amir Khusrau writing close to the event.
Why this one was different
Every Mongol force in this part so far had come to take things and leave. Qutlugh Khwaja came to conquer.
Barani is explicit about the change in intent, and the behaviour of the army confirms it: it came down through the
Punjab without stopping to plunder, which for a Mongol raiding force is close to unnatural, and made straight for
the capital.
Barani gives the size as two hundred thousand. That is not a usable number. What is usable is the reaction in
Delhi, which the sources describe as close to panic: refugees crowding into the city, the roads cut, and a serious
argument in the sultan’s council about whether to fight at all.
Alauddin was advised not to give battle. He was told to hold the fortifications and wait. He refused, and
marched out to Kili with everything he had.
What happened
1. The armies met at Kili. Alauddin took the centre himself.
2. Zafar Khan, on the right, charged — and broke the Mongol wing in front of him.
3. Then he kept going. He pursued the fleeing Mongols far beyond the field, which is exactly what a feigned retreat is designed to produce.
4. He was surrounded and cut off. The accounts describe his force fighting until it was destroyed, and Zafar
- Khan being killed with it.
- Alauddin did not move to support him. This is where Barani becomes interesting, and it is the subject of the story below.
- The main armies never fully engaged. Qutlugh Khwaja had been wounded, his losses in the fighting with Zafar Khan were heavy, and he withdrew during the night.
- He died on the way home. The invasion collapsed with him.
Barani hints — without stating it outright, in the way of a chronicler who knows what he is doing — that Alauddin
was not sorry to see Zafar Khan destroyed, and that the reinforcement which could have saved him was
deliberately withheld.
The motive suggested is jealousy. Zafar Khan was the most admired soldier in India, the victor of Jaran-Manjur, and a
man whose reputation was becoming its own power base. Alauddin had taken the throne by murdering his own uncle
and spent his reign eliminating anyone who might do the same to him.
It cannot be proved and it cannot be dismissed. Barani wrote seventy years later, admired Alauddin’s governance,
and had no obvious reason to invent this. Against that, a commander who over-pursues into an envelopment is beyond
saving by the time anyone realises, and “he could have been supported” is the easiest accusation in military history to
make and the hardest to test.
Keep both readings. The nastier one is not required to explain the battle, and this reign gives you no grounds
to rule it out.
- Alauddin came out and fought. Against advice. A capital that shuts its gates against a Mongol army invites a siege it may not survive.
- Zafar Khan’s charge cost them dearly. Whatever happened afterwards, the fight that killed him inflicted losses the invaders had not budgeted for.
- Their commander was wounded. A Mongol army built around one prince has the same single point of failure as everyone else’s.
- Delhi had a real army in being. Ten years earlier there would have been nothing to march out with.
- They had come too far to be reinforced. The length of the Punjab lay between them and home, and it was now hostile.
- Conquest needs sieges. Taking a fortified capital is slow, and slow is the one thing this army could not be.
- They lost their prince. Wounded at Kili, dead on the retreat.
- And they never tried it this way again. Later invasions in this part go back to raiding.
Barani tells a story about the aftermath of Kili that is one of the strangest compliments ever paid to a dead
man.
For years afterwards, he says, when Mongol horsemen watered their horses and an animal shied or refused
to drink, the riders would ask it mockingly whether it had seen Zafar Khan.
It became a saying — a thing you said to a nervous horse. The man who had broken their wing at Kili and
then died surrounded, still being invoked on the steppe as the standard explanation for why an animal was
afraid of something it could not see.
Take it as folklore rather than reportage; Barani is writing seventy years later and it is exactly the kind of
detail a chronicler polishes. But it is a real thing to have been remembered for.
And notice the shape of it. Zafar Khan is the best soldier in this part of the book. He wins the first
Mongol defeat in Punjab in seventy years. He then dies in a trap, possibly abandoned by the
sultan he served. He gets no monument, no dynasty, no chapter in the chronicles under his own name.
What he gets is a joke that Mongol cavalrymen told each other about their horses, written down by a
Persian historian in Delhi two generations later.
The winner is whoever gets written about — and sometimes the only person who writes you down is
the enemy, laughing, at a watering place a thousand kilometres away.
13Targhi at the Gates, 1303 LIKELY
The closest Delhi ever came to falling to the Mongols — and it happened because the sultan was five hundred kilometres away besieging a fort in Rajasthan.
When: 1303 CE, for about two months.
Where: Outside Delhi, with the Sultanate’s field army shut inside the unfinished fortress of Siri.
Who: The Mongol commander Targhi against Alauddin Khalji.
The evidence: Barani, who treats it as the most frightening episode of the reign.
How it happened
1. Alauddin was at Chittor. He had spent months besieging the great Rajput fortress and had taken it in August
- 1303, at heavy cost, with his army worn down.
- Targhi came down the Punjab road fast and light — no siege train, no baggage worth the name, moving at the speed a Mongol column could actually move.
- He arrived before Alauddin could reassemble. The sultan got back to Delhi with a tired army and shut himself inside Siri, the new fortified camp he was building, which was not finished.
- The Mongols blockaded the city for about two months. They cut the roads, stopped the grain coming in, and camped on the Jumna. Barani describes prices in Delhi going through the roof and the population in real distress.
- Alauddin would not come out. Having marched out and won at Kili four years earlier, he now did the opposite — because this time his army was not fit to fight.
- And then Targhi left. No storm, no assault, no battle. He had no way to take a fortified position, his supplies ran down, and he went home.
He treated 1303 as a system failure and rebuilt the system, and the measures are worth listing because they are the
reason the last two invasions in this part fail.
He finished Siri and made it his capital and garrison.
He repaired every fort on the road from the Punjab to Delhi and stocked them with grain, weapons and
permanent garrisons.
He put a proper commander on the frontier — Ghazi Malik, at Dipalpur, with orders to keep the marches and a
free hand to do it.
He raised more cavalry, which is what the price controls were paying for.
And he stopped leaving. No more long personal campaigns to distant fortresses while the north was open.
Every one of those is a lesson taken from two frightening months, and together they close the Punjab
corridor. It is the most competent response to a near-disaster by any ruler in this book.
- He moved when the army was elsewhere. Chittor had taken months and had left the field force exhausted and far away.
- He travelled light and fast. No plunder, no siege train, nothing to slow him. The mistake of Jaran-Manjur, corrected.
- He went for the capital, not the countryside. Blockading Delhi threatened the whole state at once.
- The forts on the road were not yet what they became. He was not stopped anywhere between the Sutlej and the Jumna.
- Light and fast means no siege capability. The same choice that got him there stopped him taking anything.
- Alauddin refused the battle. Correctly. A commander who knows his army is unfit and sits behind walls anyway is doing the hard thing.
- A blockade eats the besieger too. Two months camped outside a large city in a country that has been stripped is not sustainable.
- And there was no second army coming. The Chagatai princes did not coordinate; there was no follow-up force to turn a blockade into a siege.
Barani was a courtier and a moralist and he mostly writes about sultans. In his account of 1303 he does
something he rarely does: he writes about the city.
He describes the roads to Delhi cut, so that nothing came in from the country. He describes grain becoming
scarce and then unaffordable. He describes people who had fled the Punjab ahead of the Mongols crowded
into a capital that could not feed them. He describes the sultan and his army inside a half-built fortress,
visible to everyone, and not coming out.
No battle is fought. Nobody is a hero. The whole episode consists of two months of hunger and fear, ended
by an enemy who got bored.
This is what almost every war in this book actually looked like to the people it happened to, and it
is nearly always the part that the chronicles leave out. Tarain gets a poem. Kili gets a joke about horses. The
two months in 1303 when Delhi went hungry get a few sour paragraphs from a writer who happened to
think grain prices were worth mentioning.
Remember it in Part Seven, when Timur comes down this same road in 1398 and does to Delhi what Targhi
could not. The difference between a frightening season and a catastrophe is, very often, only
whether the besieger brought engineers.
14Amroha and the Ravi, 1305– 1306 — The Last Invasions LIKELY
Two campaigns in two years, both destroyed — and then the Mongols stop coming, and nobody at the time could have known it was over.
When: 1305 Amroha; 1306 the Ravi.
Where: Amroha in the Ganges doab; then the crossings of the Ravi in the Punjab.
Who: Ali Beg and Tartaq in 1305; Kebek in 1306. Against Malik Nayak, and then Malik Kafur and Ghazi Malik.
The evidence: Barani and Amir Khusrau, both writing for the winning court.
Amroha, 1305
1. Ali Beg and Tartaq came down into the doab and, instead of going for Delhi, spread out to plunder the
- country east of the capital.
- Alauddin did not defend the villages. He sent a field force under Malik Nayak to find the raiding columns and destroy them.
- They were caught near Amroha, dispersed, loaded and out of position.
- The Mongol commanders were taken alive and brought to Delhi, where they were killed — trampled by elephants, according to the accounts.
The Ravi, 1306
1. A larger force under Kebek came the following year, aiming at the Punjab and the country beyond.
2. This time Delhi did not wait for them to arrive. The frontier command under Ghazi Malik at Dipalpur,
- with Malik Kafur commanding the field army, moved to intercept in the Punjab itself.
- They were caught at the Ravi. The accounts describe the Mongol force being broken at the river crossings — the worst possible place to be attacked, and by now the Sultanate’s standard method. 4. The slaughter and the prisoner-taking were on a very large scale, and the executions at Delhi that followed were the largest yet.
- And that is the end. There is no serious Mongol invasion of India for the next twenty years.
It is tempting to give Alauddin the whole credit, and he deserves a great deal of it. The fort line was real, the standing
army was real, the frontier command at Dipalpur was real, and four invasions in nine years were beaten.
But the other half of the answer is in Central Asia. The Chagatai realm — the Mongol successor state that these
invasions came from — was falling into internal conflict. Duwa, who had launched most of them, died around 1306.
What followed was a long period of succession disputes and civil war, and a state fighting itself does not launch
expeditions across the Hindu Kush.
So: India stopped being worth invading at roughly the same moment that the invaders stopped being able to
invade. Both are true and the second one was not arranged in Delhi.
You have seen this exact pairing before. Part Five ended with Punjab being freed from the Huns by a battle near
Bukhara that had nothing to do with India. The frontier is never entirely in charge of its own fate.
- It fought them in Punjab, not in the doab. The 1306 interception is the whole change: the war was moved back onto the frontier.
- Rivers, again. Both battles caught an enemy at a crossing or dispersed with plunder. The Sultanate had learned exactly one trick and it kept working.
- A permanent frontier command. Ghazi Malik at Dipalpur was not waiting for orders from Delhi; he was there, with troops, all the time.
- Terror as declared policy. Ugly, effective, and honestly recorded by the people who did it.
- They were still raiding a country that had stopped behaving like a victim. The tactics that worked in 1241 were suicidal by 1306.
- Plunder slows an army and it will not drop it. Both defeats turn on this.
- They had no base this side of the mountains. Every campaign started from Afghanistan and had to get home.
- And their own state was coming apart behind them.
Count it out, because the number is the story.
From Genghis Khan on the Indus in 1221 to the Ravi in 1306 is eighty-five years. In that time the Mongols
destroyed the Khwarezmian empire, the Abbasid Caliphate and the city of Baghdad, the Song dynasty of
China, the Kievan Rus, the Assassins of Alamut, and the armies of Poland and Hungary. They ruled from
Korea to the Danube. It is the largest contiguous land empire in history.
They came into the Punjab repeatedly across those eighty-five years, and they took Lahore, and they
blockaded Delhi, and they did not take India.
The reasons in this part are unglamorous and they are all here: the climate; the distance from their base;
the fact that most of these were raids rather than conquests; a Chagatai realm that fought itself; and —
genuinely, substantially — a fortified Punjab frontier, a paid standing army, and two sultans, one cruel and
one crueller, who understood exactly what they were dealing with.
And the cost was paid in one place. Lahore sacked. Multan raided. The whole country between the
Indus and the Sutlej treated for sixty years as a zone that Delhi would not defend, and then for
another forty as a garrison. The Ganges plain, the treasury, the capital, the dynasty, the culture that
produced Amir Khusrau — all of it went on, protected.
This is the plainest possible statement of what Punjab is in this book, and Part Six is where it is truest.
India survived the Mongols. Punjab absorbed them.
15From Dipalpur to Delhi, 1320 LIKELY
The man who had spent fifteen years holding the Punjab frontier marched down the road he had been defending and took the throne. It becomes a habit.
When: 1320 CE.
Where: From Dipalpur in the Punjab to Delhi, along the invasion road in reverse.
Who: Ghazi Malik and his son against Khusrau Khan, who had seized the throne months earlier.
The evidence: Barani, and Amir Khusrau — who, characteristically, was in Delhi for it.
What happened
1. Alauddin died in 1316 and the Khalji dynasty collapsed into murder within four years — a favourite, then a
- coup, then another coup.
- Khusrau Khan took the throne in 1320, having killed the last Khalji sultan. His legitimacy was thin and his support was narrow.
- Ghazi Malik was at Dipalpur, where he had been warden of the marches since 1305 — the man Alauddin had installed after the fright of 1303, and by tradition the veteran of a very large number of engagements against the Mongols.
- He refused to recognise the new sultan, raised the frontier troops, and marched on Delhi.
- He won, and Khusrau Khan was executed. Ghazi Malik took the throne as Ghiyath al-Din Tughluq, founding the dynasty that governs Part Seven.
Watch the mechanism, because it now repeats for four hundred years.
The Punjab frontier command is the only place in the empire where a general has a large, permanent,
battle-hardened army under his personal control, far from the capital, with a legitimate reason to have it.
Every other force in the state is either small, or dispersed, or under the sultan’s eye.
So when Delhi becomes weak, the army that takes it comes from Punjab. It happens here with the Tughluqs in 1320. It
happens in Part Seven, when Bahlul Lodi comes down from Sirhind and takes Delhi in 1451. It happens in Part Seven
again, when Babur takes Lahore first and Delhi second. It happens in Part Eight with the Suris, and in Part Twelve the
same road is used by the Marathas and the Afghans.
Punjab in this book is almost never the prize. It is the springboard. And the reason is not glory or spirit or
martial character — it is that this is where the empire keeps its soldiers, because this is where the enemy comes from.
- It has the only real army. Fifteen years of continuous Mongol warfare produces troops that a palace coup cannot match.
- It has a legitimate grievance. A usurper in Delhi lets a frontier general march as a defender of order rather than as a rebel.
- It knows the road. The Punjab-to-Delhi corridor is the ground it has spent its career on.
- And its commander has a reputation. Men follow a man who has beaten the Mongols.
- Palace power is not military power. A throne taken by assassination is held by a faction, not an army.
- The capital is downstream of the frontier. Geography does the rest.
- The Sultanate never solved its succession. Not once, in three hundred years, does the throne pass peacefully more than a few times running.
- And the system was built this way on purpose. A strong frontier army was the answer to the Mongols. It was also, permanently, the answer to whoever was sitting in Delhi.
Ghiyath al-Din Tughluq ruled for five years, and the way he died is one of the great unresolved arguments of
Indian history — so it belongs here, at the end of the part, as a warning about what the next one contains.
In 1325 he was returning to Delhi from a campaign in Bengal. His son Muhammad had a wooden reception
pavilion built for him at Afghanpur, near Delhi. The sultan entered it with his younger son and his retinue.
The structure collapsed and killed him.
Ibn Battuta, the Moroccan traveller, who arrived in India a few years later and served the son as a judge,
reports the version that was circulating in Delhi: that the pavilion had been designed to fall, that the
elephants were made to strike a particular pillar, and that the son arranged it. He also reports that the
rescue was oddly slow.
Modern historians are split, and reasonably so. Ibn Battuta was writing years later from court gossip in a
city that had many reasons to hate Muhammad bin Tughluq. Wooden buildings do collapse. And a man
who wanted his father dead had easier methods available.
This book cannot tell you which it was, and it is not going to pretend. What it can tell you is what
came next, because that is Part Seven: Muhammad bin Tughluq, who moved the capital of India from Delhi
to the Deccan and marched the population there and then marched them back; who issued a token
currency that the country forged into worthlessness; and under whom the Khokhars of the Salt Range —
the people who have been in every part of this book since 1008 — rose again.
The frontier general who took the throne in 1320 founded a dynasty that would nearly destroy the
state it inherited. And in 1398, down the same Punjab road, comes Timur.
16The Whole Chapter on One Page
Five patterns to carry into Part Seven
1. Whoever commands the Punjab frontier eventually rules Delhi. This is the new rule of the book and it
- starts here. The frontier is where the empire keeps its real army, because the frontier is where the enemy is — which means it is also, permanently, the place from which the capital can be taken. Ghazi Malik in 1320 is the first. Bahlul Lodi in 1451 and Babur in 1526 are both in Part Seven.
- Punjab is spent so that the doab can be saved. After 1241 this is explicit policy, not failure. Delhi decided the far Punjab could not be held and defended the Sutlej instead. It worked, the Sultanate survived the Mongols, and the bill was paid entirely by one province. Watch who pays in 1398, when Timur comes down the same road.
- The victory that is not followed up is only a delay. Prithviraj beat Muhammad of Ghor and then besieged a fort for thirteen months while his enemy redesigned an army. It is the same failure as Yashodharman’s alliance in Part Four dissolving on the day it won. In this book, the invader always comes back, and the defender almost never does.
- A coalition of kings loses to a single command. Ten kings on the Ravi in Part One; the Malavas and Kshudrakas in Part Two; Anandapala’s confederacy in Part Five; Prithviraj’s confederacy at Tarain here. Six for six. It will be seven at Panipat in Part Seven and eight at Panipat again in Part Eight.
- The men who could not be finished were never a kingdom. The Khokhars have now been in three parts of this book. They nearly broke Mahmud’s line in 1008, they killed a sultan on the Jhelum in 1206, and they rise again under the Tughluqs. They have no capital, no dynasty, no chronicle and no monument — and no empire in this book manages to remove them. The states of Punjab lose with grim regularity; its unmanageable people do not.
Tughlaqs, Timur, Khokhars and Lodis, 1320–1526. The dynasty founded on the Punjab frontier nearly destroys the
state it inherited. The Khokhars rise again under Jasrat, who fights the Sayyids of Delhi to a standstill for twenty years
from the Salt Range. Multan becomes an independent sultanate under the Langahs. And in 1398 Timur comes down
the Punjab road and does to Delhi what Targhi could not, leaving northern India in ruins behind him. Out of that
wreckage comes Bahlul Lodi, an Afghan chief from Sirhind who takes the throne in 1451 — and, at the end of it all, five
raids out of Kabul, the capture of Lahore in 1524, and the First Battle of Panipat in 1526.
17Words Used in This Part
The dynasty from Ghor in central Afghanistan that destroyed the Ghaznavids and conquered northern India between 1175 and 1206.
“Muhammad of Ghor.” Also called Shihab al-Din in some sources, which causes endless confusion. The conqueror of Tarain.
A slave-soldier, bought young and trained for service. Because he had no clan or land, his loyalty ran to his master alone. This system produces the Delhi Sultanate.
The first dynasty of the Delhi Sultanate (1206–1290), so called because Aibak, Iltutmish and Balban had all been slaves.
Taraori, near Thanesar in Haryana. Three battles were fought here in twenty-five years — 1191, 1192 and 1216. It sits on the road every invader uses.
The Rajput dynasty of Ajmer and Delhi. Prithviraj III was its last significant ruler.
A people of the Salt Range and the Jhelum country. They fought at Waihind in 1008 in Part Five, killed Muhammad of Ghor in 1206, and rebel again in Part Seven.
The Mongol successor state in Central Asia. Almost every Mongol invasion in this part comes from it.
A Mongol commander of high rank — roughly, a general or a lord.
An assignment of land revenue given to an officer in return for military service. The Sultanate’s basic method of paying an army — which is exactly what Alauddin Khalji moved away from.
The silver coin introduced by Iltutmish, the standard currency of northern India for centuries. The ancestor of the modern word taka.
Polo. Qutb al-Din Aibak died playing it at Lahore in 1210.
The frontier commander — the officer holding the Punjab against the Mongols from a base such as Dipalpur. In this part the post produces a dynasty.
A town in the Punjab, south-west of Lahore, used as the Sultanate’s forward frontier command under Alauddin Khalji.
The fortified new city Alauddin Khalji built at Delhi, unfinished when Targhi blockaded it in 1303 and completed immediately afterwards.
Juzjani’s history, completed around 1260. The nearest thing to a contemporary account of the conquest and the early Sultanate.
Barani’s history, written around 1357. Our main source for the Khalji Mongol wars — vivid, moralising, and seventy years after the events.
A Sanskrit epic written at Prithviraj’s own court by Jayanaka. The only contemporary Indian account, surviving in one damaged manuscript, and it breaks off before 1192.
A Braj epic of roughly the sixteenth century. The source of the blinding-and-arrow legend. Magnificent literature and not history.
Sources used for this part
- Minhaj-i-Siraj Juzjani, Tabaqat-i-Nasiri, in Raverty’s translation — for the Ghurid conquest, the slave sultans and the sack of
- Lahore
- Ziauddin Barani, Tarikh-i-Firuz Shahi — for Balban, Alauddin Khalji and the Mongol wars
- Amir Khusrau, Khaza’in ul-Futuh and his other works — contemporary, and the only participant’s voice in this part
- Hasan Nizami, Taj-ul-Maasir, for the immediate aftermath of Tarain
- Ferishta, Tarikh-i Ferishta, in Briggs’s translation — late, and the source of the largest numbers
- Ibn Battuta, Rihla, for the death of Ghiyath al-Din Tughluq at Afghanpur
- Juvayni, Tarikh-i Jahangusha; Rashid al-Din, Jami al-Tawarikh; and Nasawi’s life of Jalal al-Din Mingburnu — for the Indus in 1221
- Jayanaka, Prithviraja Vijaya; the Prithviraj Raso; and the Hammira Mahakavya — for the Indian tradition and its later reshaping
- Chaulukya inscriptions and Gujarati chronicles for the defeat at Kayadara
- Coins of Prithviraja and Muhammad bin Sam, for the period between the battle and the execution
- Elliot and Dowson, The History of India as Told by Its Own Historians, for the translated Persian material
- Standard modern accounts of the Delhi Sultanate, the Ghurid conquest and the Mongol invasions of India
The open questions here include: who actually killed Muhammad of Ghor at Damyak; what happened to Prithviraj after
his capture, and how long he survived; whether Alauddin Khalji deliberately allowed Zafar Khan to be destroyed at Kili;
whether Muhammad bin Tughluq engineered the collapse at Afghanpur; and every army size and casualty figure in the
part without exception.
One caution deserves to be repeated from the front of this part. The most famous story here — Prithviraj blinded,
taken to Ghazni, and killing Muhammad of Ghor with an arrow aimed by sound — is from a poem written roughly four
hundred years later and is contradicted by the date of Ghor’s death. It is treated in this part as literature, because that is
what it is, and because how a defeated tradition rebuilds its own memory is genuinely more interesting than the legend
itself.
And the standing warning applies with full force. Every battle in Part Six is described by a Persian chronicler
working for the Delhi Sultanate. Juzjani served it, Barani’s family served it, Amir Khusrau was its court poet. There is
one contemporary Indian account of any of this, and it is a court epic celebrating a victory, in a single damaged
manuscript, which stops before the defeat. Nobody in the Punjab — not the Khokhars who killed a sultan, not the
people of Lahore in 1241, not the villagers of the Sutlej country who lived through eighty-five years of it — left a word.