1The Century Punjab Learns to Fight for Itself
Every part of this book so far has described somebody else’s army fighting on Punjabi ground. This one is different, and it is different in a way that changes the rest of the series.
Part Eight ended in 1604 with a book being installed in a shrine at Amritsar, and with an emperor at Lahore
who left the people who put it there alone.
He died in October 1605. Within eight months his successor had executed the man who compiled that book.
What follows is a century of intermittent, escalating conflict between the Mughal state and the Sikh community,
running through five Gurus, three emperors and a great many other things that had nothing to do with either. It
ends in 1708 with the Punjab possessing something it has not had since the Shahis in Part Five: an army
raised from its own population, fighting for its own reasons.
Four things to hold on to
- This is not the only war in the part, and this book is not going to pretend it is. The Mughal empire in these years also lost armies in Central Asia and Afghanistan, failed three times at Kandahar, fought a savage war of succession, and was humiliated repeatedly on the north-west frontier — all of it staged through the Punjab and paid for out of Punjabi revenue. Those wars are in here too.
- The fighting moves into the hills, and stays there. Part Eight’s rule — the hills preserve you and the hills imprison you — becomes the central military fact of this part. Anandpur, Bhangani, Nadaun and Chamkaur are all in the Siwalik foothills, and so is almost everything that matters.
- The most important thing here is not a battle. It is what happens at Anandpur in 1699, and then what happens at Chamkaur in 1704 and at Nanded in 1708 — three moments in which authority is deliberately moved out of one man and into an institution. Nothing in the previous eight parts does this. Every army in this book so far has dissolved when its commander went down.
- And for the first time we can hear the other side. Jahangir wrote about Guru Arjan in his own memoir. Guru Gobind Singh wrote about his own battles, and wrote a letter to Aurangzeb that survives. This part has both emperors and their opponents describing the same events in the first person, and that has not been true anywhere in this book before.
The conflict in these pages is between a Muslim empire and a religious community that was not Muslim, and it is used
today, by people on several sides, to argue about things that have nothing to do with the seventeenth century. This book
is going to try to be useful rather than comfortable, and that means three commitments.
The rule about sources applies evenly. Jahangir’s memoir is a ruler explaining himself and is read as such. The Sikh
devotional literature of the eighteenth and nineteenth centuries is written to edify and is read as such. Where a claim
rests on one and not the other, this part says so.
The empire is not treated as one thing. Akbar left this community alone; Jahangir executed a Guru and later
travelled with the next one; Shah Jahan’s officers fought four battles against a Guru and then left him in peace for a
decade; Aurangzeb’s policy was different again. Four emperors, four different postures. A story that flattens them
into a single hostile block is not describing this century.
And the labels are used honestly. CERTAIN, LIKELY, STORY — including where a beloved tradition
falls into the third category. That is not disrespect. It is the same standard applied to Alexander in Part Two and to
Timur in Part Seven.
2Lahore, 1606 — The Execution of Guru Arjan CERTAIN
The emperor wrote down what he had done and why, in his own memoir, and the reasons he gives are the beginning of everything in this part.
When: May or June 1606, a few months into the new reign.
Where: Lahore, in custody.
Who: Guru Arjan, fifth Guru, compiler of the Adi Granth, on the orders of the emperor Jahangir.
The evidence: Jahangir’s own Tuzuk-i-Jahangiri; a Jesuit letter written from Lahore the same year; and Sikh tradition.
What Jahangir says
This is one of the very few events in this book where the man who ordered a killing explains it himself, in writing,
without being asked. His memoir is unusually frank, and the passage on Guru Arjan is worth reading closely
because everything about it is revealing.
He describes a Hindu at Goindwal on the Beas named Arjun, living in the manner of a religious teacher, who had
built up a following over three or four generations. He complains that too many people believed in him —
and specifically that Muslims as well as Hindus were among them. He calls it a shop of falsehood and says
he had long intended either to end it or to bring the man into Islam.
Then he gives the immediate occasion. His son Khusrau had rebelled and fled towards the Punjab, and on the way
had met the Guru, who — Jahangir says — treated him with favour and made a mark on his forehead.
He orders the Guru brought in, his property confiscated, and his execution.
First, it is a confession, not an accusation by enemies. This is not a hostile chronicler or a later tradition. It is the
emperor’s own account, and it is why the fact of the execution is labelled CERTAIN rather than argued about.
Second, the reasons he gives are mixed, and the mixture is the point. He names the Guru’s growing following —
including Muslim followers — as a standing grievance he had held for some time. He also names the meeting with the
rebel prince as the trigger. Historians have argued for a century about which mattered more, and the honest
answer is that Jahangir presents both and does not rank them.
Some historians add a third factor: a wealthy community with towns, a central shrine and an income from offerings is a
fiscal and political fact as well as a religious one, and new emperors take an interest in such things.
Third, there is independent corroboration. A Jesuit priest at Lahore, Jerome Xavier, wrote a letter in 1606
describing the death of a Hindu holy man in the city under the new emperor. A European clergyman with no stake
in the argument, writing from the same city in the same year, is about as good as seventeenth-century
evidence gets.
The detailed traditions about the manner of the death, and about the role of a Lahore official named Chandu Shah,
come from later Sikh sources and are not in the contemporary record. This book reports the fact and does not dwell on
the details, which are contested and which no purpose is served by rehearsing.
- A new reign needs to look decisive. He had been emperor for months, his son had just rebelled, and severity was the standard answer.
- The rebel connection made it defensible. Whatever happened between the Guru and Khusrau, it gave the order a political justification rather than a purely religious one.
- He had inherited a grievance. His own words say he had wanted to act for some time and had not.
- And Akbar was dead. The policy of leaving this community alone was Akbar’s personal policy, and personal policies do not survive their author. Part Seven’s lesson, in a new setting.
- It made the succession the point. The community did not scatter. It installed an eleven-year-old and kept going.
- It converted a religious movement into a political one. Before 1606 this community had no reason to arm. Afterwards it had a very specific one.
- It created a category that did not exist before. The concept of martyrdom in defence of the faith becomes central to Sikh identity from this moment, and it is the emperor who supplies it.
- And it did not remove anything. The book was already compiled and installed. The shrine was built. The organisation was in place. He killed the man and left the institution standing — which is exactly the mistake this book has watched empires make since Part Three.
Sikh tradition records that before his arrest, Guru Arjan sent word to his son about how the succession was
to be conducted — and that the instruction was that he should sit fully armed upon the throne, and
maintain an army to the best of his ability.
The wording comes from later Sikh sources rather than from any contemporary document, and this book
labels it accordingly: STORY. Whether those words were said in 1606 cannot be established.
What can be established is that this is precisely what happened.
Guru Hargobind was eleven years old. At his installation he did not take up only the items his predecessors
had used. He put on two swords, and gave them names — miri and piri, the temporal and the spiritual —
and the meaning was not subtle and was not meant to be. He began keeping horses. He accepted weapons
and riders as offerings where his predecessors had accepted money. Within a few years there was a fort at
Amritsar and a building facing the Harmandir Sahib from which secular decisions would be taken.
So the tradition may be a later community explaining its own turning point to itself in the form of a dying
instruction. Traditions often are. But the turn is real, it is dated, and its cause is not in dispute: it is a
direct response to what happened at Lahore in 1606.
Jahangir’s memoir records the removal of a religious teacher who had attracted too many followers.
What the order actually produced was an eleven-year-old with two swords and a hundred and
two more years of this.
3The Two Swords — Guru Hargobind Builds an Army, 1606–1620s LIKELY
A shrine had been built to be walked down into. Facing it, they now built a platform to be sat upon — and the difference between those two buildings is the whole of this part.
When: 1606 onwards; the imprisonment at Gwalior falls somewhere between about 1609 and 1619.
Where: Amritsar, and the fortress of Gwalior.
Who: Guru Hargobind, sixth Guru, installed at eleven.
The evidence: The buildings, later Sikh sources, and the fact that Mughal officers were fighting his men by the 1620s or 1630s.
What was built
1. The Akal Takht. Facing the Harmandir Sahib across the tank, a raised platform and building was constructed
- — the Akal Takht, the Throne of the Timeless One. The architecture is an argument. The shrine is set below ground level so that you walk down to it, in humility. The Akal Takht is raised, so that authority is exercised from it. Worship in one; decisions in the other.
- The two swords. Miri and piri — worldly authority and spiritual authority, worn together. The doctrine that the two cannot be separated dates from here, and it is the theological foundation for everything that follows in this part.
- A fort. Lohgarh — the iron fort — was built at Amritsar. A religious centre with a fortification is a different kind of place.
- Horses and men. Followers were asked for horses and weapons rather than only for money, and a body of mounted retainers was maintained. The numbers in later sources should be treated with caution; the existence of the force should not.
- And the drums. A war-drum was installed and beaten at the Akal Takht — a public announcement in the most literal available form.
Gwalior
At some point in Jahangir’s reign — the dating is genuinely uncertain and the sources disagree by as much as a
decade — Guru Hargobind was imprisoned in the fortress of Gwalior, which the Mughals used for political
detainees.
He was released. Sikh tradition holds that he refused to leave unless the rajas imprisoned with him were freed too,
that a condition was set allowing out only those who could hold on to his cloak, and that a garment with many
tails was made so that all of them could. The number given is fifty-two. The release is commemorated as Bandi
Chhor Divas, the day of the release of prisoners, and coincides with Diwali.
The cloak story is STORY. It appears in later Sikh sources, not in contemporary record, and the number fifty-two
has the shape of a remembered figure rather than a counted one.
The imprisonment itself is accepted; the reason for it is not recorded anywhere reliable. Suggestions range
from unpaid fines levied on his father’s estate to official unease about the fort and the armed following. Nobody knows.
And here is the part that complicates every simple version of this century. After the release, the relationship
between Jahangir and Guru Hargobind was not hostile. The sources describe them on reasonable terms, with the
Guru accompanying the emperor on travels including to Kashmir. There were no battles in Jahangir’s reign after the
release.
The same emperor executed the fifth Guru and travelled with the sixth. Any account of this century that cannot
hold both those facts is selling you something.
- It had an organisation already. Part Eight built the districts, the representatives and the system of offerings. Turning offerings into horses is an administrative change, not a revolution.
- It had a defensible location. Amritsar was its own town, built by the community, not a quarter of somebody else’s city.
- Its population were farmers. The Jat peasantry of the doabs joining the community in growing numbers brought with them exactly the horses, the physical hardiness and the tradition of local self-defence that an army needs.
- And the state was distracted. Jahangir’s reign was consumed by the Deccan, by Kandahar and by his own succession.
- It made the community a military question. A religious movement can be ignored; a fort cannot.
- It divided opinion inside the community. Not everyone thought the two swords followed from Guru Nanak, and rival claimants to the succession made exactly that argument.
- It put them on a collision course with the local administration, which is War Three.
- And it could not match the empire. A few thousand horsemen against the largest army in the world is not a strategy. It is a way of surviving until the hills.
Stand at Amritsar and look at what was built there in eighty years, because the whole argument of this part
is visible in the stonework.
The Harmandir Sahib was completed under Guru Arjan by about 1589. It sits in the middle of a tank, and
it is built below the level of the surrounding ground, so that a visitor descends to reach it. It has doors on all
four sides. Both features are deliberate and both are statements: you come down to it rather than climbing
up, and it is open from every direction.
The Akal Takht was built facing it from about 1606, and it is raised. It is a throne — the word is in the
name. From it, decisions binding on the community are announced. A war-drum was beaten there.
Fifty yards apart, across water: a building you go down into, and a building you look up at.
That is the doctrine of the two swords rendered in brick, and it was put up within a couple of years of
the execution at Lahore. It says that a community which will not defend itself will not survive to worship,
and that the people who make the second decision must be the same people who make the first.
Every ruler in the previous eight parts of this book combined temporal and spiritual authority as a matter
of course — Ashoka carving moral instruction on pillars, Mahmud collecting titles from the caliph, Akbar
making himself the final arbiter in religious disputes. What is new here is not the combination. It is
that the combination is being asserted by people with no state, no throne and no army worth the
name, against an empire that had both.
Within thirty years, Mughal detachments were fighting them. Within a hundred, the Punjab had an army
of its own.
4The Four Battles of Guru Hargobind, 1628–1635 LIKELY
Four engagements over about seven years, all of them small, all of them won — and then the Guru moved into the hills and the fighting stopped for sixty years.
When: Between about 1628 and 1635. The dates are disputed and some sources place the first as early as 1628 and others as late as 1634.
Where: Amritsar; Lahira near Mehraj in the Malwa country; Kartarpur in the Jalandhar doab; and Palahi near Phagwara.
Who: Guru Hargobind against local Mughal commanders under Shah Jahan.
The evidence: Later Sikh sources, chiefly. Contemporary Mughal records say very little, which is itself informative.
What happened
1. Amritsar. The traditional account has the fighting begin over a hunting hawk — a bird belonging to an imperial hunting party seized by the Guru’s men, or the reverse, depending on the version. A force under a
- commander the sources name as Mukhlis Khan was sent and was beaten.
- Lahira, near Mehraj. The occasion in the tradition is horses: two fine animals bound for the imperial stables, taken and then recovered by a follower named Bidhi Chand, whose exploits are among the best-loved stories in the community. A Mughal force under Lalla Beg was defeated in the Malwa country south of the Sutlej.
- Kartarpur. Fought against a force that included Painde Khan — a Pathan who had been raised in the Guru’s own household and had gone over to the other side. He was killed in the fighting, and the traditional accounts of his death are notably not triumphant.
- Palahi, near Phagwara. The last of them, fought on the withdrawal.
- And then he left. Guru Hargobind moved north-east to Kiratpur, in the Siwalik foothills, in territory belonging to a hill raja rather than to the empire. There were no further battles in his lifetime, and none for the next sixty years.
Almost everything we know about them comes from Sikh sources written considerably later — the Gurbilas literature
and the nineteenth-century compilations — which are devotional works composed to inspire. They give the hawk and
the horses, the named commanders, the numbers and the heroics.
Contemporary Mughal chronicles barely mention any of it. That silence is not evidence that nothing happened.
Court histories record what an emperor’s officials wanted recorded, and a local commander beaten by a religious
leader’s retainers in a Punjab village is not a thing anybody writes to Delhi about with enthusiasm. The Punjab
produced dozens of such affrays a decade and almost none of them are in the imperial record.
So the honest position is this. That armed clashes took place between Guru Hargobind’s followers and Mughal
detachments in this period is accepted by historians and is consistent with everything else we know — the
fort, the horsemen, the move to the hills afterwards. The details, the causes and the dates come from a tradition
rather than from a record, and they are held here at arm’s length in exactly the way this book has held Kalhana in
Part Four and the Prithviraj Raso in Part Six.
- These were small engagements against local forces. A provincial commander with a few hundred men, not an imperial army.
- They fought on their own ground. Their own town, their own villages, their own countryside, with the population on their side.
- Motivation is not evenly distributed. Troops sent to deal with an irritant fight differently from men defending their own place.
- And they had cavalry. The Jat peasantry of the doabs supplied horses and riders who did not need training in how to sit on one.
- The empire could always send more. Four victories against local detachments do not change the arithmetic against a state with a million men under arms.
- Amritsar could not be held. A town on an open plain, forty miles from Lahore. The oldest lesson in this book.
- So he went to the hills. Kiratpur was outside the imperial administration, in a raja’s territory, in country that cavalry cannot operate in — which bought sixty years.
- And the hills imprison as well as preserve. Part Eight’s rule. From Kiratpur you are safe and you are also irrelevant, and everything that happens for the next two generations happens because of that trade.
Of the four battles, the one that the tradition dwells on is Kartarpur, and it is not because of the tactics.
Painde Khan was a Pathan. By the accounts, he had been brought up in Guru Hargobind’s own household,
trained there, treated as family, and had fought on the Guru’s side in earlier engagements. He was, in the
versions that survive, an enormous and formidable man of whom the Guru was fond.
He quarrelled, left, and came back at the head of the other side’s troops.
The tradition has them meet in the fighting. Painde Khan is killed. And then — this is the part that has
kept the story alive for four hundred years — the Guru is described as shading the dying man’s face
from the sun with his shield, and speaking to him.
Take the scene as devotional literature, because that is what it is; nobody was taking notes.
But notice what a community chooses to remember about its own victories, because that is a
historical fact even when the scene is not. Of four battles won, the one preserved in most detail is the one
where the enemy was somebody they had loved, and the remembered image is not the killing blow but the
shield held up against the sun.
This book has spent eight parts reading victory literature — Persian court histories counting the dead,
Sanskrit inscriptions cataloguing subdued kings, Timur’s memoirs explaining a massacre as
administration. The genre is not usually like this.
5Balkh and Kandahar, 1646–1653 — The Wars the Punjab Paid For LIKELY
Shah Jahan spent seven years and an enormous fortune trying to take countries beyond the passes, and got none of them. Every rupee and every soldier went through the Punjab.
When: 1646–1647 Balkh and Badakhshan; 1649, 1652 and 1653 the three sieges of Kandahar.
Where: Beyond the Hindu Kush, and at Kandahar — staged from Lahore and Kabul the whole way.
Who: Shah Jahan, with his sons Murad, Aurangzeb and Dara Shukoh commanding in turn.
The evidence: The Mughal court histories, which record the expenditure with painful precision.
Why they went
Because the Mughals never stopped thinking of themselves as a Central Asian family in exile. Babur had lost
Samarkand three times and taken India as second best — Part Seven made the point, and it did not go away.
Shah Jahan’s ancestral homeland was Transoxiana, and when Uzbek politics fell into disorder in the 1640s the
chance to recover it looked real.
Kandahar was a different matter and a more practical one. It sits on the road between Persia and India and had
changed hands between Mughals and Safavids repeatedly. It was lost in 1649 and Shah Jahan wanted it back.
What happened
1. Balkh and Badakhshan were taken in 1646 with relatively little difficulty, and Aurangzeb was installed at
- Balkh.
- And then it became clear that they could not be held. The country was hostile, the Uzbeks would not fight a decisive battle and would not stop raiding, and supplying an army across the Hindu Kush from India is a proposition that has defeated everybody who has attempted it.
- The withdrawal in 1647 was the disaster. The army came back over the passes late in the season and was caught by winter in the mountains. The accounts describe men and animals dying in the snow in large numbers. Everything gained was abandoned within about two years, and the cost is recorded in the court histories in figures that made contemporaries wince.
- Kandahar was besieged three times — in 1649 under Aurangzeb, in 1652 under Aurangzeb again, and in 1653 under Dara Shukoh with an even larger army and a siege train.
- All three failed. The fortress was strong, the Persians were competent, and Mughal siege artillery could not reduce it before the supply situation forced a withdrawal.
Because of what an army is, logistically, before it is anything else.
Every one of these expeditions was assembled, supplied and paid for through the Punjab. The troops
concentrated at Lahore. The grain came from the Punjab doabs. The transport animals were bought here. The revenue
that funded seven years of failure was drawn substantially from these districts, because the Punjab was one of the
wealthiest and best-assessed provinces the Mughals had — Part Eight’s Ain-i-Akbari tables show exactly how much.
And when armies came back broken, they came back through here.
This is the cost that never appears in a battle account. A province that stages an empire’s wars pays for them in
requisitioned grain, in commandeered carts and bullocks, in men taken for the transport train, and in the ordinary
inflation that follows a large army sitting in your district for a season.
Nobody wrote it down, because nobody has ever written that down. But when Part Ten describes a Punjab that
collapses with startling speed in the eighteenth century, the answer starts here — with seventy years of an empire
spending this province on wars in the Deccan, in Central Asia and at Kandahar, and getting nothing back.
- Supply across the Hindu Kush is close to impossible. Every army in this book that has tried it has been destroyed by geography rather than by an enemy.
- The Uzbeks refused a decisive battle. There was nothing to defeat, only country to occupy — the problem the Salt Range people posed to Sher Shah in Part Eight, at national scale.
- Kandahar was too strong for the artillery of the day, and sieges have a clock: the food runs out before the walls come down.
- And they were fighting for sentiment. Balkh was an ancestral homeland, not a strategic necessity. Wars fought for feelings tend to be expensive.
- The treasury. Contemporaries and modern historians agree the sums were enormous and that nothing was gained.
- The army’s reputation. Three failures at one fortress in five years is noticed by everybody on the frontier, and remembered.
- The frontier settlement. The Pashtun tribes drew their own conclusions about Mughal invincibility, and War Six is what came of it.
- And the succession. Dara Shukoh’s failure at Kandahar in 1653 damaged him badly at court, and Aurangzeb made sure it was remembered.
The Balkh campaign is barely known outside specialist history and it deserves better, because it is the
clearest demonstration in this book of a thing this whole series keeps circling.
In 1646 a Mughal army crossed the Hindu Kush and took Balkh — the ancient Bactria of Part Three, the
country the Kushans came from, the province Seleucus signed away in 303 BCE. Aurangzeb, then a prince in
his late twenties, was installed as its governor. On paper, the family had recovered its homeland.
He was there for a few months, and what he learned in them shaped a fifty-year reign. The revenue did not
cover the garrison. The Uzbeks would not fight and would not leave. The country was not conquered so
much as stood upon.
By late 1647 the decision was taken to withdraw, and the withdrawal was made too late in the year.
The army came back over the passes in winter. The court histories, which had every reason to
minimise it, describe men and horses lost in the snow and the survivors arriving at Kabul in a condition
nobody wanted to record in detail.
Now put it beside the other returns over those passes in this book. Genghis Khan turned back from
India in Part Six partly because of the climate. Timur’s grandson lost his cavalry to the Punjab
monsoon in Part Seven. A British army retreating from Kabul in 1842 will be destroyed in the
same country.
The mountains between the Punjab and Central Asia have killed more soldiers than any general
in these sixteen parts. They funnel every invasion of India down one road, and they destroy every army
that tries to go back up it at the wrong time of year. Nobody has ever solved them, and the empires that
keep trying keep paying for it out of the province at their foot.
6The War of Succession, 1657–1659 — Dara Shukoh’s Last Road LIKELY
The heir to the Mughal empire lost it in two battles and then spent a year running through the Punjab and Sindh, and the road he took is the road every fugitive in this book takes.
Aurangzeb — the last Mughal succession decided cleanly, and the most WON BY
destructive
- Shah Jahan falls ill in September 1657; Dara Shukoh is executed in August 1659.
- Samugarh and Deorai — and then Lahore, Multan, Bhakkar and the road to the Bolan.
- Dara Shukoh, the eldest son and designated heir, against his brother Aurangzeb.
- Mughal chronicles, and the account of François Bernier, a French physician who was in India and travelled with Dara’s camp.
What happened
1. Shah Jahan fell ill in 1657 and the four brothers — Dara, Shuja, Aurangzeb and Murad — went to war before
- anybody knew whether he would live. He did, and watched the rest of it from confinement.
- Samugarh, 1658. Aurangzeb and Murad destroyed Dara’s army near Agra. Dara fled.
- He came to the Punjab. Lahore was his province and its revenue and troops were his best remaining asset. He held it briefly, raised what he could, and could not hold it — Aurangzeb’s forces came up the Grand Trunk Road.
- Then Multan, then Sindh. Down the same road that every beaten claimant in this book has taken since Humayun in Part Eight: Lahore, Multan, the Indus, and then out.
- He tried Gujarat and Rajasthan, raised another army, and was beaten again at Deorai near Ajmer in 1659.
- And then he was betrayed. Fleeing towards Persia through the Bolan country, exhausted and with his family, he was taken in by an Afghan chief named Malik Jiwan whose life he had once saved, and who handed him over. 7. He was paraded through Delhi on a filthy elephant — Bernier describes the crowds and their reaction — condemned on a charge of apostasy, and executed in August 1659.
He was a scholar as much as a prince. He studied Hindu philosophy seriously, associated with Sufis and with the Hindu
ascetics of Benares, and commissioned and worked on a Persian translation of the Upanishads which he called
Sirr-i-Akbar, the Greatest Secret. He wrote a book arguing that Hindu and Muslim mystical traditions were describing
the same reality, and gave it a title that means the meeting-place of the two oceans.
That translation had a strange afterlife. It reached Europe in the eighteenth century by way of a Latin version, and was
read by German philosophers who had no other access to the Upanishads at all.
The charge on which he was condemned was apostasy, and the intellectual work was cited against him.
Sikh tradition also records a connection: that when Dara was ill, Guru Har Rai — the seventh Guru, at Kiratpur —
supplied a rare medicine from his stores. Whether it happened is not documented outside the tradition. What is
documented is that Aurangzeb afterwards summoned the Guru to account for himself, and that the Guru sent his elder
son Ram Rai instead — who altered a line of scripture to make it inoffensive to the emperor, and was disowned by his
father for it.
That disinheritance matters more than it looks. It established, at the highest possible cost, that the words in the
book were not negotiable to please a ruler — and the rival line it created would trouble the community for a century.
- He was the better soldier by a wide margin. Years of command in the Deccan and at Kandahar against Dara’s court career.
- He made alliances and discarded them. Murad was used to win Samugarh and imprisoned afterwards.
- He moved fast and did not stop. Two years from his father’s illness to his brother’s execution.
- And he had a simpler case. Orthodoxy against a prince who had translated the Upanishads was an argument that wrote itself at court.
- He had never commanded successfully. Kandahar in 1653 was his one great independent command and it failed publicly.
- He kept the throne and lost the army. Being at court with his father meant patronage; it did not mean loyal regiments.
- He fled instead of consolidating. After Samugarh he had Lahore, a province, a treasury and troops — and left rather than fight for it.
- And he trusted the wrong man at the end, which is how nearly every fugitive in this book finishes.
Trace the route of a defeated claimant in this book and it is always, without exception, the same route.
Humayun in 1540 loses at Kannauj, falls back on Lahore, argues with his brothers, and goes down through
Multan into Sindh and out. Dara Shukoh in 1658 loses at Samugarh, falls back on Lahore, cannot hold it,
and goes down through Multan into Sindh and out. A hundred and eighteen years apart, the same four
place-names in the same order.
And it is older than either of them. Khusrau Malik, the last Ghaznavid, was cornered at Lahore in
Part Six. Daulat Khan Lodi was run down in the Punjab in Part Seven. Sikandar Shah Suri went
into the Siwaliks in Part Eight.
The reason is geography and it is completely mechanical. If you lose in the Ganges plain, the Punjab is
the only direction with a road, a friendly province and a way out of India at the end of it. So
everyone goes there. And the Punjab, which had nothing to do with the quarrel, gets the losing army, the
pursuing army, and the requisitions of both.
Only one man in this book ever comes back up that road: Humayun, after fifteen years, and it takes a
Persian army to do it.
For everyone else, the road through Lahore and Multan is where their story ends — and the
province spends four centuries being the corridor along which other people’s defeats are conducted.
7The Khapash Pass, 1672 — The Frontier Burns Again LIKELY
A Mughal army was destroyed in a defile north-west of the Punjab for the second time in a century, and the emperor came to Hasan Abdal and spent a year and a half trying to fix it.
When: 1672, with the Yusufzai revolt of 1667 before it and Aurangzeb at Hasan Abdal in 1674–1675.
Where: The Khyber country and the passes — and the imperial headquarters at Hasan Abdal in the Punjab.
Who: The Afridi and allied tribes against Mughal armies under Muhammad Amin Khan and others.
The evidence: Mughal court histories, which do not enjoy this subject.
What happened
1. The frontier had been simmering since the 1660s. The Yusufzai rose in 1667 and were put down with
- difficulty.
- In 1672 the Afridis rose, under a leader the sources call Akmal Khan, and closed the Khyber.
- A large Mughal force under the governor of Kabul went in to clear it — and was caught in a defile the sources call the Khapash Pass, with the tribes above it.
- It was destroyed. The figures given run to many thousands killed and enormous numbers of prisoners and animals taken. The governor escaped with a fraction of his men.
- A second force met a similar fate the following year.
- So the emperor came in person. In 1674 Aurangzeb moved to Hasan Abdal, in the Punjab north-west of Rawalpindi, and stayed about a year and a half directing the campaign.
- And what worked was not a battle. It was roads, forts, garrisons, negotiation and, above all, subsidies — payments to tribal leaders to keep the passes open. The frontier was quieted by being bought.
Part Eight described the destruction of Raja Birbal’s column in the Karakar Pass in 1586 and said that the same thing
happens to every regular army that enters this country. Eighty-six years later, it happened again, to a bigger army, in the
same way, for the same reasons.
A column strung out on a mountain track. Tribes on the heights. No way to form, deploy or manoeuvre.
Destruction along the whole length of the line.
Akbar’s answer after 1586 had been roads, forts and payments. Aurangzeb’s answer after 1672 was roads, forts and
payments. The British answer after their own disasters in the same passes would be roads, forts and
payments.
Three empires, three centuries, one conclusion, arrived at independently each time after losing an army to learn it.
And notice where the emperor sat while he worked it out. Hasan Abdal is in the Punjab. For a year and a half in the
1670s, the Mughal empire was again being run from this province, for exactly the same reason it had been run from
Lahore in the 1580s: because the frontier was on fire and the only way to manage it was to be near it.
- They chose the ground every time. A defile removes every advantage a regular army has.
- There was nothing to capture. No capital, no army, no treasury — the same problem the Gakkhars posed to Sher Shah and the Khokhars posed to everybody.
- They only had to close the passes. The empire had to keep them open, which is far harder.
- And they could be bought, which is not the same as being beaten.
- Two armies and several years. At a moment when the Deccan war was already consuming everything.
- The emperor’s presence. Eighteen months at Hasan Abdal is eighteen months not spent anywhere else, and the Deccan and the Punjab both drifted while he was there.
- A permanent subsidy. The payments to the tribes became a standing charge that no later government has ever managed to stop paying.
- And it emptied the north. With the field army committed to the frontier and then to the Deccan, the Punjab’s own policing was left to provincial officers — which matters a great deal for everything after 1699.
In 1674 and 1675, Aurangzeb was at Hasan Abdal in the Punjab, running a frontier war that had already
eaten two armies.
He was not a young man; he had been on the throne for sixteen years. He had a rebellion in the north-west,
an unfinished war in the Deccan that would eventually consume the last twenty-five years of his life, and
an empire that on paper was the richest state on earth and in practice was being spent faster than it could
be collected.
And in November 1675, while he was still dealing with the aftermath, a religious teacher from the
Punjab hills was executed at Delhi on his orders.
That is War Seven, and the point of putting them side by side is this. From inside the imperial
administration, the second event was very much smaller than the first. The frontier had destroyed
armies. The Deccan was a war without end. A Guru in Delhi was a provincial matter, disposed of in a
paragraph, if that.
Three hundred and fifty years later, almost nobody outside specialist history has heard of the Khapash Pass.
The execution at Delhi is commemorated by a gurdwara at the site, by another where the body
was cremated, and by millions of people every year.
This book has said it in every part and it is never truer than here: the winner is whoever gets written
about — but the corollary is that the winner does not get to decide which of his own decisions turns out to
have been the important one.
8Delhi, 1675 — The Execution of Guru Tegh Bahadur LIKELY
The ninth Guru was executed in a public square in Delhi, and the reason recorded in the Sikh tradition is one that has almost no parallel anywhere in this book.
When: November 1675.
Where: Chandni Chowk, Delhi. The Guru had come from Anandpur in the Punjab hills.
Who: Guru Tegh Bahadur, ninth Guru, and three companions executed with him.
The evidence: Sikh sources, chiefly — including the Bachittar Natak, attributed to his son and written within a generation.
What the Sikh account says
Guru Tegh Bahadur had founded a town in the Siwalik foothills in 1665 on land bought from the raja of the local
hill state. He called it Chak Nanaki, after his mother; it became Anandpur, and it is the centre of everything for
the rest of this part.
The account preserved in the tradition is this. A delegation of Kashmiri Brahmins came to him at Anandpur,
reporting forced conversions under the governor of Kashmir, and asked for help. He undertook their cause — and
did so on the specific ground that people are entitled to their own faith, including a faith that was not his own.
He was arrested, taken to Delhi, and after some months given the choice of conversion or death. Three
companions — Bhai Mati Das, Bhai Sati Das and Bhai Dyala — were executed before him. He was
beheaded in Chandni Chowk.
The execution is not in doubt. Two gurdwaras in Delhi mark the place of the death and the place of the cremation, and
the event is fixed in Sikh memory with a precision that leaves no room for invention about whether it happened.
The cause is where the sources thin out. Mughal court chronicles of Aurangzeb’s later reign are sparse on the whole
subject — he had restricted official history-writing — and what survives says very little.
The nearest thing to a contemporary Sikh statement is the Bachittar Natak, a composition attributed to Guru
Gobind Singh and written within a generation. It describes his father’s death explicitly as a sacrifice made for the sake
of others’ right to their own religious marks and observances, and says that he gave his head and not his resolve. That
is a near-contemporary source from the closest possible witness, and it is why the Kashmiri Pandit account
is accepted by most historians.
Some historians add political factors — the size of the following at Anandpur, its revenue, official unease about a
fortified centre in the hills, the rival claims of the disowned Ram Rai. Those may all have contributed. They do not
displace what the closest source says the reason was.
What can be said without qualification is the shape of the thing, and the shape is genuinely unusual. A religious
leader of one community was executed after taking up the cause of a different community’s right to practise
its own faith. This book has covered five thousand years and it does not have another case quite like it.
- Aurangzeb’s religious policy had hardened. The jizya was reimposed in 1679; temple demolitions were ordered in several provinces; the direction of travel was clear well before that.
- A fortified centre with a following is a political fact. Anandpur was in hill territory outside direct administration, and its growth was noticed.
- The empire was overstretched and nervous. A frontier war just finished, a Deccan war beginning, and revolts elsewhere.
- And a public execution is meant to be seen. Chandni Chowk is the busiest place in Delhi. That was the point.
- The succession held again. A nine-year-old was installed at Anandpur, exactly as an eleven-year-old had been installed in 1606.
- It confirmed the argument. The doctrine of the two swords holds that spiritual life cannot be maintained without the power to defend it. This was the demonstration.
- It made the cause bigger than the community. A death on behalf of another religion’s freedom is not a sectarian grievance and cannot be dismissed as one.
- And it produced Guru Gobind Singh, who was nine years old when his father’s head was brought back to the hills.
There is a line of reasoning in the Bachittar Natak about this death which is worth setting out plainly,
because it is the intellectual foundation of everything in the rest of this part and it is usually skipped over
in favour of the drama.
The composition describes Guru Tegh Bahadur as having acted for the sake of people’s right to their own
religious observances — the sacred thread and the forehead mark, both specifically Hindu practices that
had nothing to do with his own community’s beliefs. He is described as breaking his own body rather
than his resolve, and giving his head without a sound of complaint.
Read what that claims. It is not that his own faith was under attack, though it was. It is that the right of
anybody to practise their own religion is worth dying for, including when it is not your religion.
Now measure that against the previous eight parts of this book.
Ashoka in Part Three renounced conquest and asked his descendants not to make war — the closest earlier
parallel, and it is an emperor legislating from a position of total power. Akbar in Part Eight held debates
between faiths and abolished the tax on non-Muslims — also an emperor, also from strength, and his
policy died with him within months.
This is not that. This is a man with a small following in a hill town, with no army capable of anything, no
state, no leverage and no protection, going to Delhi on behalf of people who were not his own and who could
do nothing for him.
Whether every detail of the tradition is exact cannot be established. What is established is that the
closest source to the event, written within a generation by the man who had the most reason to
know, says that was the reason — and that a community which could have chosen any
explanation chose that one and has never chosen another.
9Bhangani, 1688 — The First Battle LIKELY
The tenth Guru’s first battle was fought at twenty-one against a coalition of Hindu hill rajas, and it was a Muslim holy man from Sadhaura who came to fight on his side.
When: September 1688.
Where: Bhangani, near Paonta on the Yamuna, in the Sirmur hills.
Who: Guru Gobind Rai — later Guru Gobind Singh — against a coalition of hill rajas under Fateh Shah of Garhwal.
The evidence: The Bachittar Natak, a first-person account attributed to the Guru himself. Unusually good for this book.
Why the hill rajas fought him
This is the fact that most retellings of this century get wrong, and it needs saying at the start. Guru Gobind
Singh’s first two decades of fighting were mostly not against the Mughal empire. They were against the
Hindu rajas of the Siwalik hill states.
The reasons were about power, not religion, and they were entirely ordinary. Anandpur and Paonta were growing.
The Guru’s following was large, armed and expanding into country the rajas regarded as theirs. He accepted
followers from every caste, which was corrosive to the social order the hill states were built on. And his presence
disturbed the delicate arrangement by which those rajas paid tribute to the empire and were otherwise left alone.
He had moved to Paonta, on the Yamuna, under the protection of the Raja of Sirmur — who was in dispute with
his neighbour of Garhwal. That neighbour assembled a coalition.
What happened
1. A confederacy of hill rajas marched on Paonta, led by Fateh Shah of Garhwal and joined by others
- including Bhim Chand of Kahlur, in whose territory Anandpur stood.
- Part of the Guru’s force deserted. A body of Pathan mercenaries in his pay went over to the other side before the battle — the Bachittar Natak names them and does not soften it. 3. And a Muslim holy man came to fight for him. Pir Buddhu Shah of Sadhaura arrived with his sons and his followers and fought on the Guru’s side. Two of his sons were killed.
- The battle was fought at Bhangani and the coalition was broken.
- And the Guru wrote it down. The Bachittar Natak gives a first-person account — who fought, who ran, who died, and the confusion of it — in verse.
Stop and consider what this document is, against everything that has come before it in these nine parts.
It is a commander’s own written account of his own battle, composed within his lifetime, in a language of the
region, which survives.
Now count the alternatives this book has had to work with. Alexander’s battles reached us through Greeks writing four
hundred years later. Tarain was described by a Persian court historian in Delhi. Waihind was counted by al-Utbi for his
employer. Timur’s campaign comes through a text that may not be his. Mihirakula was written up by a Kashmiri poet
six centuries afterwards. In eight parts, the only comparable first-person military document is the Baburnama
— and Babur was the invader.
Here the man writing is defending his own ground, and he is writing in the Punjab about the Punjab.
It is not neutral — it is a devotional and autobiographical composition, in verse, by a religious leader, and its purpose is
not the purpose of a modern battle report. But it names the men who deserted him. It records that mercenaries
in his pay changed sides. It does not pretend the fight was easy. Court histories do not usually do those things.
- He was defending, on ground he knew. The hill country around Paonta, with the river behind the rajas’ approach.
- His people had a reason to fight. A coalition of rajas assembling to remove a community is a clarifying event for that community.
- The coalition was a coalition. Several rajas, several agendas, no single command — the losing formation this book has now documented seven times.
- And he had unexpected help. Pir Buddhu Shah’s men arrived when the Pathans left, and the arithmetic changed.
- They were fighting a movement, not a rival. Beating a raja settles a border; beating a community settles nothing.
- They underestimated a twenty-one-year-old. He had been Guru since the age of nine and had spent the intervening years preparing for exactly this.
- Their own subjects were joining him. Which is why they attacked, and also why they could not win.
- And they would try again. Repeatedly, for sixteen years, with the same result until they got the empire to do it for them.
Of everything at Bhangani, the detail that deserves to be better known is who turned up.
Pir Buddhu Shah was a Muslim divine of Sadhaura, a town in the plains below the hills, with a following
of his own. He had no obligation to the Guru, no shared religion, and nothing to gain. When the Pathan
mercenaries deserted, he came with his sons, his brother and his followers, and fought.
Two of his sons were killed in the battle.
The tradition records that afterwards, when the Guru asked what he might give him, the Pir asked for a
comb from the Guru’s hair with some hairs still in it, and for his turban. That part is devotional tradition
and this book marks it as such.
The rest is not in dispute, and it sits awkwardly across every simplified version of this century that anybody
wants to tell.
At the first battle of the tenth Sikh Guru, the enemy was a coalition of Hindu rajas, the deserters
were Muslim mercenaries in his own pay, and the ally who came unasked and lost two sons was a
Muslim holy man.
Hold on to that when Part Ten and Part Eleven get bloodier, because it stops being obvious. The wars in this
book divide along lines of power, land, revenue and fear far more often than along lines of faith — and
when they do divide along faith, it is usually because somebody in authority found it useful.
Pir Buddhu Shah’s own end is recorded as unhappy: the tradition holds that he was killed some years later
on the orders of the local Mughal administration, for having helped the Guru.
10Nadaun, 1691 — Fighting on the Rajas’ Side LIKELY
Three years after the hill rajas tried to destroy him, the Guru fought a Mughal army alongside them — and then they went back to being his enemies.
The hill rajas and the Guru together — and the alliance lasted about as long as WON BY
the battle
- 1691.
- Nadaun, on the Beas in the Kangra hills.
- Guru Gobind Rai and Bhim Chand of Kahlur, against a Mughal force under Alif Khan.
- The Bachittar Natak again — the Guru’s own account.
What happened
1. The hill rajas stopped paying tribute. The Mughal governor of Jammu sent a commander named Alif Khan
- to collect it.
- Bhim Chand of Kahlur asked the Guru for help — the same Bhim Chand who had marched against him at Bhangani three years earlier.
- And the Guru came. He fought at Nadaun on the Beas alongside the rajas, and the Bachittar Natak describes the action, including his own part in it, in detail.
- The Mughal force was defeated and withdrew.
- Then the rajas made terms with the empire anyway and resumed paying, which is what hill states do.
- And within a few years they were his enemies again, and would be for the rest of his life.
Nadaun is the clearest evidence in this part that the wars of the 1680s and 1690s in these hills were not a religious
conflict, and it is worth setting out the ledger.
1688 at Bhangani: Hindu rajas attack the Guru; Muslim mercenaries desert him; a Muslim pir fights for him and loses
two sons.
1691 at Nadaun: the Guru fights alongside the same Hindu rajas against a Mughal army.
1700 onwards at Anandpur: the same rajas besiege him, then ask the empire for troops, and Mughal and hill-state
forces attack him together.
Three configurations in twelve years, involving the same handful of people.
What actually drove it was land, tribute, precedence and fear — the ordinary business of small states with a large
neighbour and an inconvenient new power in the middle of them. Religion is present throughout, it matters to everyone
involved, and it is not what determines who is on which side in any given year.
That is not a comfortable finding for anyone’s preferred version, which is precisely why it belongs on the page.
- Hill country against a plains force, for the fourth time in this book. Nadaun sits on a river in broken ground.
- A tribute-collecting expedition is not a war. Alif Khan came to enforce a payment, not to conquer, and such forces withdraw when the cost rises.
- The Guru’s men were the best troops on the field. Trained, motivated and by now experienced.
- The rajas settled immediately afterwards. They wanted better terms, not independence, and a victory improved their bargaining position.
- The Guru gained a reputation he could not use. Being the best soldier in the hills made him more dangerous to his neighbours, not less.
- And it put him in the imperial record. A religious leader who beats a Mughal detachment in the field becomes an administrative problem with a name.
The Siwalik foothills between the Sutlej and the Ravi held a scatter of small kingdoms — Kahlur, Sirmur,
Garhwal, Guler, Nurpur, Mandi, Kangra, Bilaspur and a dozen more. They had existed for centuries, they had
their own dynasties, their own hill forts and some of the finest miniature painting produced anywhere in
India.
And in this part they do nothing but change sides.
They fight the Guru, then fight beside him, then besiege him, then call in the empire, then break their own
guarantees. They pay tribute to Delhi, stop paying, are made to pay again. Not one of them ever combines
with the others into anything that could resist either the empire or the community growing in the middle
of them.
This is the oldest failure in this entire book and it is now in its ninth appearance. Ten kings on the
Ravi in Part One. The Malavas and Kshudrakas in Part Two, who combined and then could not agree on a
commander. Nine republics submitting separately to Samudragupta in Part Four. Anandapala’s confederacy
in Part Five, which dissolved when one elephant bolted. Prithviraj’s coalition at Tarain in Part Six. The
Punjab powers of Part Seven — Khokhars, Langahs, Afghans — who never once made common cause.
Small states in this region do not federate. They bargain individually with whoever is strongest,
and are eaten in sequence.
The hill states were absorbed by Ranjit Singh in Part Thirteen, one at a time, without much difficulty. By
then they had been failing to combine for a very long time.
11Anandpur, 1699 — The Khalsa CERTAIN
Not a battle. The most consequential thing in this part, and one of the most consequential in the whole book.
Punjab acquires an army that is not a dynasty, a clan or a mercenary WHAT CHANGED
force
- Vaisakhi, 1699.
- Anandpur, in the Siwalik foothills above the Sutlej.
- Guru Gobind Singh and the community summoned to the town.
- Sikh sources, and the observable fact that everything afterwards is different.
What was done
The Guru called the community to Anandpur for the spring festival and, before the assembly, asked for a head.
The account is that he asked for someone willing to give his life, took the man who came forward into a tent, and
returned with a bloodied sword. He asked again. Five times in all, five men came forward, and after the fifth he
brought all five out alive.
These five were the Panj Pyare, the Five Beloved. What they had in common was that they had volunteered. What
they did not have in common is the point: they came from different castes and different parts of India,
including from groups that Indian society at the time treated as low.
He then administered an initiation — amrit, prepared with water and sugar and stirred with a double-edged
sword, with recitation — and established the order:
1. The initiated took the name Singh, and women Kaur. One name for everybody, replacing the surname that
- announced your caste. This is not a symbolic gesture; a shared name in a caste society is a demolition.
- They kept the five articles of faith — uncut hair, comb, steel bracelet, prescribed undergarment and a blade — worn by everyone, which makes the member of the order visible and impossible to hide.
- They were bound by a code of conduct that applied to all equally. 4. And then the Guru asked the five to initiate him. He received the rite from the men he had just created, on the same terms as anybody else.
The founding of the Khalsa in 1699 at Anandpur, the initiation, the names, the articles of faith and the code are
established beyond dispute — they are the constitution of a community that has existed continuously since, and its
practice is the evidence.
The dramatic details of the day — the tent, the sword, the five calls — come from Sikh sources rather than
from any external record, and this book flags that in the same way it has flagged everything else. Nobody outside the
community was writing it down.
What can be checked from outside is the effect. Within a year, Mughal and hill-state forces were treating
Anandpur as a military problem of a different order. Within eleven years, an army raised from this order had
taken Sirhind and was administering territory. Within a century it ruled the Punjab.
Something happened at Anandpur in 1699 that turned a religious community into a standing military force. The
mechanism is documented in its own constitution, and the results are visible in everybody else’s records.
- It removed caste from recruitment. Every other Indian army in this book drew its fighting men from restricted groups. This one could recruit from the entire population, and did — overwhelmingly from the Jat peasantry, but formally from anybody.
- Initiation is a commitment mechanism. A man who has taken a rite, changed his name and put on visible articles he cannot remove has burned his boats. Deserting is not possible in the ordinary way.
- It is voluntary. Nobody is conscripted, levied or bought. Part Five’s Ghaznavids ran on slave-soldiers; Part Seven’s sultans on assignments of land revenue. This runs on people who chose.
- And authority was distributed. The Guru taking initiation from the five he created put the order above the man — which is the whole of War Twelve.
- Visibility. The articles of faith make a member identifiable at a glance, which is an asset in a battle and a death sentence in a persecution. Part Ten is about that.
- It split opinion inside the community. Not everybody accepted the new order, and the divisions of the eighteenth century begin here.
- It alarmed everybody outside it. The hill rajas and the imperial administration both read it correctly as the creation of an army.
- And the war started within a year.
Go back through the previous nine parts and look for a founder who subordinated himself to the thing he
founded.
Chandragupta built an empire and made it his family’s. Ashoka carved his conscience on pillars and it died
with him. Every Greek king in Part Three, every sultan in Parts Six and Seven, every Mughal in Part Eight —
the state was the man, and the man was the state, and when he died it was fought over by his sons. Part
Eight counted five commanders who were winning a battle and lost it because they personally
went down. That is what happens when authority lives in one body.
At Anandpur in 1699, the Guru created an order, and then knelt and received initiation into it from five
men who had been ordinary members of the community that morning.
Whatever else it was, that is a deliberate act of constitutional engineering. It says that the order is not his
property. It says that its authority does not descend from him but is held collectively. And it establishes that
the five, acting together, can bind him.
Five years later at Chamkaur, that provision was used, and it saved his life against his own wishes.
Nine years later at Nanded he took it to its conclusion and ended personal Guruship altogether, vesting it in
the book compiled in Part Eight.
This book has watched states, dynasties, empires and armies dissolve on the death of one man for three
thousand years — the Nandas, Menander’s kingdom, Harsha’s empire, Yashodharman’s alliance, Hemu’s
army, Jasrat’s Punjab. It is the single most reliable failure mode in these sixteen parts.
In 1699, in a hill town on the Sutlej, somebody finally solved it.
12The Sieges of Anandpur, 1700–1704 LIKELY
Four years of intermittent siege ended with a guarantee of safe passage, given on oath, and broken within hours.
When: Fighting from 1700; the final siege and evacuation in December 1704.
Where: Anandpur and its ring of small forts in the Siwaliks.
Who: Guru Gobind Singh against the hill rajas and imperial forces under the faujdar of Sirhind.
The evidence: Sikh sources; some Mughal administrative record; the Zafarnama, the Guru’s own letter to Aurangzeb.
What happened
1. The hill rajas attacked first, from 1700, and were repeatedly beaten in the engagements around Anandpur
- and Nirmohgarh.
- So they asked the empire in. Unable to remove him themselves, the rajas petitioned for imperial troops, and the governors of Sirhind and Lahore supplied them.
- The final siege closed in 1704. Anandpur was invested by a combined force and cut off from supply. It lasted months.
- Conditions inside became desperate. The accounts describe the food gone and the animals eaten. A number of the Guru’s followers left, and some formally renounced him in writing — a document called a bedava. Remember them; they come back in War Thirteen.
- Terms were offered. The besiegers guaranteed safe passage if the town were evacuated — and, according to the Sikh accounts and to the Guru’s own later letter, swore to it on the Quran.
- The evacuation began on a December night in rain, with the community strung out on the road.
- And the guarantee was broken almost at once. The column was attacked on the march.
- At the Sarsa river, in flood, the family was separated in the dark. The Guru and his two elder sons went one way with a small party. His mother and his two younger sons went another. Nothing after that is recoverable.
The breaking of the safe-conduct is central to how these events are remembered, and it is not only a later tradition. The
Guru wrote to Aurangzeb about it himself.
The Zafarnama — a letter in Persian verse, sent about a year later and discussed in full in War Thirteen — accuses the
emperor directly of an oath sworn on the Quran and broken, and asks what a man’s word is worth if his sacred oath is
not.
That is not a chronicler’s accusation. It is a letter from the injured party to the person he holds responsible,
and it survives.
Whether Aurangzeb personally authorised the guarantee, or whether it was given by commanders in the field who
exceeded or ignored their orders, is genuinely disputed and cannot be settled here. Imperial armies in this period were
commanded by provincial officers with wide discretion, and the emperor was two thousand kilometres away in the
Deccan.
What is not disputed is that a guarantee was given, that the column was attacked on the road, and that the
man it was given to wrote to the emperor and said so.
- They stopped fighting and started waiting. Four years of failed assaults, and then a blockade — the same lesson Akbar learned at Mankot in Part Eight.
- They combined. Hill rajas plus imperial troops, for once acting together, which the rajas alone had never managed.
- A fort needs a field army behind it. Part Five said this at Nandana in 1014 and it has not stopped being true. There was no relief and there was never going to be.
- And the promise did the last of it. A garrison that will not surrender to force may leave on terms — which is exactly why the terms were offered.
- It was a town, not a fortress. A religious centre with defences, holding a civilian population including families.
- The hills feed nobody. Part Eight’s rule again — you cannot be finished there and you cannot supply an army there.
- No allies, anywhere. The rajas were the besiegers and the empire was behind them.
- And the community inside was not an army. It was a population, and populations starve first.
The Sarsa is not a large river. In December, in an ordinary year, it is not much of an obstacle at all.
On the night the evacuation reached it, it was in flood, it was raining, it was dark, and there was a Mughal
force behind them on the road.
What happened at that crossing is the hinge of the whole part, and it happened in about an hour, in
confusion, with nobody in control of it.
The community was cut into pieces at the water. Some got across, some did not, some drowned.
Baggage was lost — including, by the accounts, manuscripts. The Guru crossed with a small party and his
two elder sons, Ajit Singh, seventeen, and Jujhar Singh, fourteen. His mother, Mata Gujri, crossed
separately with the two younger boys, Zorawar Singh and Fateh Singh, who were about nine and six.
Neither group knew where the other had gone.
Within a week, everything that happens in Wars Twelve and Thirteen had happened, and by the end of
December 1704 all four of his sons and his mother were dead.
This book has described a great many disasters. What makes this one different is that it was not a battle
and nobody planned it. A river came up, in the dark, at the worst possible moment, and separated a
family that would never be in the same place again.
Every account of these events — devotional, scholarly, hostile — agrees on this much. And it is the reason
the winter of 1704 occupies the place it does in the memory of the Punjab: not because of the tactics, but
because of what an ordinary flooded stream did to a group of people who were already walking away from
everything they had.
13Chamkaur, December 1704 LIKELY
Forty men in a mud house against an army — and the most important thing that happened there was an order the Guru did not want to obey.
When: December 1704, within days of the Sarsa crossing.
Where: Chamkaur, on the plain below the hills, near the Sutlej.
Who: Guru Gobind Singh and about forty Sikhs, against a pursuing imperial force.
The evidence: Sikh sources, and the Zafarnama, which describes it in the Guru’s own words.
What happened
1. They took shelter in a mud-walled house — a garhi, a fortified compound of the ordinary Punjabi kind,
- belonging to a local family. It was not a fortress. It had walls and a gate.
- The pursuing force surrounded it. The numbers in the sources are very large and are the usual sort; what is not in question is that the disproportion was extreme.
- The Sikhs went out in small parties to fight. A few at a time, out of the gate, into the surrounding force. This is what you do when you cannot hold and cannot escape: you make the taking of the place expensive.
- Ajit Singh, seventeen, went out and was killed.
- Jujhar Singh, fourteen, asked to go, and went, and was killed.
- And then the surviving Sikhs ordered the Guru to leave.
- He went. With two or three companions, at night, through the encircling force, and escaped into the country to the west.
Read carefully what the accounts describe, because it is easy to skim past as a devotional detail and it is not one.
Five of the surviving Sikhs constituted themselves as the Panj Pyare — the five, acting collectively, in the form
established at Anandpur in 1699 — and issued a command to the Guru to leave the position.
He did not want to go. The accounts are consistent about that. He went because, under the constitution he had himself
created five years earlier, the five acting together could bind him, and he had accepted initiation from them on
those terms.
Now put that beside the previous nine parts of this book.
Porus’s elephants turn and his army is destroyed. Anandapala’s elephant bolts and eight centuries of Indian
coalition-building dissolve in ten minutes. Muhammad of Ghor is wounded at Tarain and his army runs.
Qutlugh Khwaja is hit at Kili and a full invasion goes home. Hemu takes an arrow at Panipat while winning,
and loses.
Five commanders, five armies, one failure mode: authority lived in one man’s body and could not survive it.
At Chamkaur, in a mud house, with his two elder sons dead outside the gate, the machinery built to solve that
problem was operated — and it worked. The institution overruled the leader, the leader complied, and both the
institution and the leader survived the night.
It is the single most consequential thing that happens in these hundred years, and it happens in a farmyard.
- Numbers, against a mud wall. There was never any doubt about the outcome of the siege itself.
- They had the pursuit right. Catching a scattered party days after the Sarsa was competent soldiering.
- And there was no relief. Nobody in the Punjab was coming.
- They came for one man and did not get him. Everything else was incidental to that objective, and it failed.
- They gave the community its founding memory. Forty against an army, in a mud house, is the kind of thing that gets remembered for three hundred years — and has been.
- The order survived the battle. The Khalsa was not at Chamkaur. It was scattered across the Punjab, and it still existed on the morning of the 23rd.
- And within six years an army raised from it took Sirhind. Which is Part Ten.
After Chamkaur the Guru was alone in open country in December, hunted, with an imperial administration
searching every road.
The tradition records that he was sheltered near Machhiwara — a name you have seen already in this
book, because Bairam Khan won a battle there in 1555 that refounded the Mughal empire — and that he
was then moved out of the district by two Pathan brothers, Ghani Khan and Nabi Khan, horse traders
who had known him.
They put him in a litter, dressed as a Muslim holy man, and carried him openly along the roads, telling
anyone who asked that they were escorting the Uch da Pir — the holy man of Uch. The accounts have them
stopped and questioned, and a local man vouching for them under oath.
The details are Sikh tradition and this book labels them so. What is not tradition is the shape of the thing,
which is corroborated by everything else about this year.
The Guru survived the winter of 1704 because Punjabi Muslims hid him from a Muslim emperor’s
officers.
Set it beside the rest of the ledger this part has been keeping. At Bhangani a Muslim pir fought for him and
lost two sons, while Muslim mercenaries in his pay deserted. At Nadaun he fought a Mughal army beside
Hindu rajas who had tried to kill him three years earlier and would try again. At Anandpur those same
Hindu rajas called in the imperial troops. And in the worst week of his life, two Pathan horse dealers
carried him out of a search cordon.
Nothing in this century sorts cleanly by religion, and the people who lived through it knew that
better than anybody arguing about it now.
14Sirhind, Muktsar and the Letter, 1704–1708 LIKELY
Two children died at Sirhind, forty deserters came back and died at a pool in the desert, and the man who had lost everything wrote a letter to the emperor telling him he had lost.
When: December 1704 at Sirhind; December 1705 at Khidrana; the letter written about 1705–06.
Where: Sirhind on the Delhi road; Khidrana — now Muktsar — in the dry Malwa country.
Who: Guru Gobind Singh; the faujdar of Sirhind, Wazir Khan; and forty Sikhs who had left him.
The evidence: Sikh sources, and the Zafarnama in the Guru’s own hand.
Sirhind
The party separated at the Sarsa — Mata Gujri and the two younger boys — was betrayed by a servant and handed
over to the authorities at Sirhind.
Zorawar Singh was about nine. Fateh Singh was about six.
They were held at Sirhind, brought before the faujdar Wazir Khan, and required to convert. They refused. They
were put to death in the last week of December 1704. Their grandmother died at Sirhind at the same time.
This book is not going to describe it further. The manner of the execution is recorded in the Sikh tradition and is
commemorated every December, and nothing is gained by setting it out here beyond what has been said. Two
children were killed by a provincial official because they would not change their religion, and the
memory of it is one of the load-bearing facts of Punjabi history.
It is worth stating plainly what these deaths did, because the consequences run through the next hundred years of this
book.
Sirhind stopped being a town and became a name. Six years later, in 1710, an army raised from the Khalsa under
Banda Singh Bahadur stormed it and destroyed it. Wazir Khan was killed, and the site was left in ruins so deliberately
that it was said to be under a curse. That is Part Ten, and it is the reason Part Ten opens the way it does.
And notice which of these events the empire recorded. Aurangzeb’s later reign restricted official history-writing,
and the Mughal record of the winter of 1704 is thin to the point of silence. A provincial faujdar disposing of the family
of a hill-country religious leader was, administratively, nothing at all.
It is the other side that wrote this one down, and that is possible because of what was built in Part Eight and formalised
in 1699. For the first time in this book, the losing side’s account is the one that survives in full.
Khidrana, December 1705
1. The forty who had left Anandpur during the siege — the ones who had signed the bedava renouncing him
- — were shamed at home, according to the tradition, by the women of their villages.
- A woman named Mai Bhago — Mata Bhag Kaur — is remembered as having rallied them and led them back.
- They found him at a pool at Khidrana in the dry country of the Malwa, with an imperial force in pursuit.
- They fought there and nearly all of them died, holding the ground long enough for the pursuit to be broken. Mai Bhago survived, badly wounded.
- And the Guru is said to have torn up the document. The forty are remembered as the Chali Mukte — the forty liberated — and the place was renamed Muktsar, the pool of liberation.
- It kept killing individuals and missing the institution. Guru Arjan in 1606, Guru Tegh Bahadur in 1675, four sons and a mother in 1704 — and the Khalsa was still there in 1708.
- The empire was two thousand kilometres away. Aurangzeb spent the last twenty-five years of his life in the Deccan, and the Punjab was run by provincial officers with local priorities.
- Persecution recruits. Sirhind is the clearest demonstration in this book: the killing of two children produced an army six years later.
- And the country was with them. A man hunted through the Punjab in 1704 and 1705 was fed, hidden and moved by ordinary people, including people of the emperor’s own religion.
- Anandpur, and everything in it. The town, the library, the accumulated work of forty years.
- Four sons and a mother, inside one month.
- The community scattered. The organisation built over a century was broken up and had to be reassembled from nothing.
- And a century of what follows. Part Ten and Part Eleven are the bill.
Some time in 1705 or 1706, in the Malwa country, having lost his town, his library, his four sons and his
mother inside a single winter, Guru Gobind Singh wrote a letter to the Emperor Aurangzeb.
It is in Persian, in verse, and it is called the Zafarnama — the Epistle of Victory.
It is not a plea and it is not a lament. It is an indictment, written by a man with nothing left, addressed to
the most powerful person in the world, and its argument runs roughly like this.
You swore on your Quran and you broke the oath. He returns to this again and again — not as an
insult but as a legal and religious point. A ruler whose sacred oath means nothing has forfeited the thing
that makes him a ruler.
Your commanders did not defeat me. He describes Chamkaur: a handful of men, an army against them,
and what it cost to take a mud house.
Killing my children changes nothing. One of the most quoted lines in Punjabi memory is his — that
what does it matter if four have been extinguished, when the coiled one still remains.
And there was no other way left. The most famous couplet in the letter states the principle plainly: when
all other means have been exhausted, it is lawful to take up the sword.
Now consider this document against the nine parts before it. This history has been assembled from Greeks
writing four centuries late, Chinese pilgrims describing ruins, Persian court secretaries counting the dead
for the men who paid them, victory pillars carved by winners, and a Kashmiri poet writing six hundred
years afterwards. In Part Seven a Punjabi finally wrote down what an invasion looked like. In Part Eight it
was put in a book.
Here, a Punjabi who has just lost everything writes directly to the emperor who took it, in the
emperor’s own literary language, and tells him he has lost.
Aurangzeb died in the Deccan in 1707. The Guru died at Nanded in October 1708, stabbed by an assassin,
having done two last things: he sent a follower north to the Punjab with authority to act — and he ended
personal Guruship for ever, declaring that after him the Guru would be the book.
The follower was Banda Singh Bahadur, and Part Ten begins with him on the road to Sirhind.
15The Whole Chapter on One Page
Five patterns to carry into Part Ten
1. The institution that outlives the man — finally. Nine parts of this book have watched states collapse when
- one person died: the Nandas, Menander’s kingdom, Harsha’s empire, Yashodharman’s alliance, Jasrat’s Punjab, Hemu’s army. Five commanders lost battles they were winning because they personally went down. At Anandpur in 1699 the problem was deliberately solved, at Chamkaur in 1704 the solution was used, and at Nanded in 1708 it was made permanent. Part Ten is the first part of this book where the leadership can be killed without the movement ending — and it is tested immediately.
- Killing the man leaves the machinery standing. Jahangir executed a Guru and got an eleven-year-old with two swords. Aurangzeb executed a Guru and got a nine-year-old who founded the Khalsa. A provincial officer killed two children at Sirhind and got an army at his gates in six years. This book has watched empires make this error since Part Three; here it is made three times in a century by people who should have learned.
- The hills preserve and the hills imprison. Kiratpur, Paonta, Anandpur, Bhangani, Nadaun — everything in this part happens in the Siwaliks, for the reasons Part Eight set out. You cannot be finished there and you cannot win from there, and Anandpur in 1704 is what the trade eventually costs.
- Nothing sorts cleanly by religion, and the people involved knew it. Hindu rajas besieged the Guru and fought beside him. A Muslim pir lost two sons for him and Muslim mercenaries deserted him. Muslim horse traders hid him from a Muslim emperor’s officers. A Guru died for the Hindu right to a sacred thread. The wars in this part divide by land, tribute, power and fear — and when they divide by faith it is usually because someone in authority found it useful.
- And the record is now ours. The Bachittar Natak and the Zafarnama are first-person Punjabi accounts of Punjabi wars, written by the man who fought them. For the first time in nine parts, the losing side’s version survives in full — and from here to the end of this book, it never stops.
Banda Singh Bahadur and the Years of Persecution, 1708–1748. A follower sent north from Nanded with a letter
of authority raises an army in the Malwa within a year. Samana in 1709; Sirhind in May 1710, where Wazir Khan is
killed and the town destroyed; the first Sikh administration, its own coinage struck in the Gurus’ name, and a land
settlement that abolishes the landlords. Then the empire’s full weight: Lohgarh, the retreat to the hills, the long siege of
Gurdas Nangal in 1715, and the executions at Delhi in 1716. What follows is thirty years in which the community is
hunted through the Punjab with a price on every head, survives in the forests and the deserts and across the Sutlej —
and comes out of it organised into fighting bands. And at the end, in 1739, Nadir Shah comes down the road from
Kabul, and the Mughal Punjab ends for good.
16Words Used in This Part
Temporal and spiritual authority. The two swords taken up by Guru Hargobind in 1606, and the doctrine that the two cannot be separated.
“Throne of the Timeless One.” The raised building facing the Harmandir Sahib at Amritsar, from which decisions binding on the community are taken.
“Iron fort.” The fortification built at Amritsar under Guru Hargobind; the name is used again later for Banda Singh Bahadur’s stronghold.
“The day of the release of prisoners” — commemorating Guru Hargobind’s release from Gwalior, and coinciding with Diwali.
A composition attributed to Guru Gobind Singh containing a first-person account of his early life and battles, including Bhangani and Nadaun. Near-contemporary and unusually valuable.
“Epistle of Victory.” The Guru’s letter in Persian verse to Aurangzeb, written after Chamkaur, accusing him of breaking an oath sworn on the Quran.
The order founded at Anandpur on Vaisakhi 1699 — initiated, bound by a common code, taking the names Singh and Kaur, and open to anyone regardless of caste.
“The Five Beloved” — the five who came forward at Anandpur in 1699. Also the standing institution of five initiated Sikhs acting collectively, whose decision can bind even the Guru.
The initiation rite of the Khalsa, prepared with water and sugar and stirred with a double-edged blade.
A Mughal district military commander. Wazir Khan, the faujdar of Sirhind, is the official responsible for the events of December 1704.
A small mud-walled fortified compound of the ordinary Punjabi kind. The one at Chamkaur was not a fortress.
The written renunciation signed by the forty who left Anandpur during the siege — and torn up by the Guru after Khidrana.
“The forty liberated” — the men who returned and died at Khidrana, which was renamed Muktsar, the pool of liberation.
The term used for Guru Gobind Singh’s four sons, all of whom died in December 1704.
Founded as Chak Nanaki in 1665 by Guru Tegh Bahadur in the Siwalik foothills. The centre of everything in the second half of this part.
A genre of later Sikh narrative literature about the lives of the Gurus. Devotional in purpose, and the source of most of the detail about the battles of Guru Hargobind.
Jahangir’s own memoir, in which he records and explains his order for the execution of Guru Arjan.
“The Greatest Secret” — Dara Shukoh’s Persian translation of the Upanishads, cited against him at his trial for apostasy.
Sources used for this part
- Jahangir, Tuzuk-i-Jahangiri, in Rogers and Beveridge’s translation — for the execution of Guru Arjan in the emperor’s own words
- The letter of the Jesuit Jerome Xavier from Lahore, 1606 — independent contemporary corroboration
- The Bachittar Natak and the Zafarnama, attributed to Guru Gobind Singh — first-person accounts from the Punjab
- The Gurbilas literature and later Sikh narrative sources for the battles of Guru Hargobind and the events of 1704–05, treated as
- devotional tradition and labelled as such
- Mughal court histories of Shah Jahan’s reign for Balkh, Badakhshan and Kandahar
- François Bernier, Travels in the Mogul Empire, for the war of succession and the fate of Dara Shukoh
- Mughal records of the north-west frontier campaigns of 1667–1675
- The buildings at Amritsar and Anandpur, and the gurdwaras marking the sites in Delhi, Sirhind, Chamkaur and Muktsar
- Standard scholarly histories of the later Mughal empire and of the Sikh community in the seventeenth century
The open questions here include: the relative weight of the political and religious motives in the execution of Guru
Arjan; the dating of the Gwalior imprisonment, which the sources place anywhere across a decade; the dates and details
of Guru Hargobind’s four battles, which rest on later Sikh sources and are barely mentioned in Mughal records; the
specific reasons for the execution of Guru Tegh Bahadur, where Mughal record is nearly silent; whether Aurangzeb
personally authorised the safe-conduct broken at Anandpur or whether it was given by commanders in the field; and
every troop figure in the part.
On method, three things should be said plainly.
First, this part applies the same test to Sikh sources that earlier parts applied to Persian, Greek and Sanskrit
ones. The Bachittar Natak and the Zafarnama are near-contemporary and are treated as strong evidence; the Gurbilas
literature is later and devotional and is treated as tradition. That is not scepticism about the community’s memory. It is
the same standard used on Kalhana in Part Four, the Prithviraj Raso in Part Six and the Malfuzat-i Timuri in Part Seven.
Second, the Mughal side is not treated as one thing. Akbar left this community alone; Jahangir executed one Guru
and later travelled with his successor; Shah Jahan’s officers fought four small battles and then let the matter drop for a
decade; Aurangzeb’s policy was harder again, and much of what was done in the Punjab in 1704 was done by provincial
officers while the emperor was two thousand kilometres away. Four emperors, four postures.
Third, this part has deliberately not described the manner of the deaths at Lahore in 1606 or at Sirhind in
1704. Both are recorded in the tradition, both are commemorated, and neither requires elaboration here. The historical
facts are stated. Anything beyond that would be for effect, and these particular events do not need any.