1The Four-Hundred-Year Habit
Part Two ended with Alexander’s men walking out of Punjab. Part Three is about what walked in afterwards — and about the strangest pattern in the whole book.
Here is the shape of the next four hundred years, and it is worth having in your head before the details start.
Punjab is conquered from the north-west four separate times by four different peoples: Greeks, then
Sakas, then Parthians, then Kushans. Each arrives with cavalry, foreign gods and an unfamiliar language. Each
takes Taxila, takes the plain, and sets up a court.
And then, every single time, the same thing happens. Within two or three generations the conqueror is minting
coins in Indian scripts, putting Indian gods on the reverse, taking Indian titles, marrying into Indian families and
endowing Indian religious buildings. A Greek ambassador sets up a pillar in central India to the god Vishnu
and calls himself a devotee. A Kushan emperor puts the Buddha on his gold coinage in Greek lettering. A
Saka warlord’s family becomes the most important patrons of Buddhism in the north-west.
This is the four-hundred-year habit, and it is the single most important thing in this part. Punjab in these
centuries is not a place that resists foreigners. It is a place that digests them.
Four things that make this part different
- India acquires an empire, and Punjab is what it is built on. Chandragupta Maurya starts here, in the north-west, with the garrisons Alexander left behind. The first great Indian state is assembled from the outside in.
- Coins become the main evidence, and they are strange evidence. Something like thirty Greek kings ruled in this region whose names appear on no written source anywhere in the world. We know they existed only because they made money. You will meet several of them.
- Punjab becomes the most cosmopolitan place on earth. Greek sculptors, Buddhist monks, Persian administrators, Chinese silk, Roman gold and Central Asian cavalry, all in the same cities at the same time. This is the world that produces Gandharan art.
- And it ends with a genuine Punjabi victory. The last war in this part is fought by the armed republics of the plain, on their own account, against the greatest empire in Asia — and they win it.
CERTAIN means contemporary evidence — an inscription cut at the time, a coin, an eyewitness. LIKELY
means good sources writing later, generally accepted. STORY means legend or tradition, which may have a real
war underneath it.
Part Three has a particular problem. Most of it comes from four kinds of source that never talk to each other: Greek
and Roman historians writing centuries later and at second hand; Buddhist legend, which cares about lessons
rather than dates; Chinese official histories, which are excellent but only look at India sideways; and coins, which
are contemporary and truthful and cannot tell you a single thing about what happened.
Where those four disagree, this part says so.
2Chandragupta Clears Out the Macedonians LIKELY
The first Indian empire began as a rebellion against a garrison, in the north-west, by a young man nobody can identify with confidence.
When: Roughly 322–317 BCE, in the years right after Alexander’s death in 323.
Where: The Punjab plain, Gandhara and the Indus — the satrapies Alexander had left behind him.
Who: Chandragupta Maurya and his adviser Chanakya, against the Macedonian governors and their troops.
The evidence: Two short passages in Roman-era historians, and a large amount of Indian tradition written down much later.
What the Greeks say happened
The most important sentence in this war is from Justin, a Roman writer summarising a lost earlier history. India,
he says, after the death of Alexander, had shaken the yoke of servitude from its neck and put his governors to death
— and the author of that liberation was Sandrocottus, which is the Greek form of Chandragupta.
Then Justin adds a line that is easy to skip and worth stopping on. Having freed India, Sandrocottus turned that
freedom into servitude of his own.
Orosius, later, says much the same: that after Alexander’s death the Indians killed his officers and rose in revolt
seeking their freedom under the leadership of a certain Androcottus.
So the Greek tradition is clear on the shape of it. This was a war of liberation in the north-west, and it
produced a conqueror.
What actually happened to the Macedonian garrisons
Here is the honest and slightly deflating part, and it matters.
1. They were already falling apart. Philip, the satrap of the Indus country, was assassinated by his own
- mercenaries in about 325 BCE — before Alexander was even dead. Nicanor, another commander, was killed in the north. Alexander’s arrangements in Punjab were unravelling within months of his leaving. 2. The best Greek troops were pulled out to fight each other. When Alexander died, his generals began a forty-year war over his empire. Eudemus, the Greek commander in India, treacherously killed Porus — the same Porus who fought Alexander on the Jhelum in Part Two — seized his elephants, and in about 317 BCE marched west out of India with a hundred and twenty of them to join the fighting. He was defeated and executed shortly afterwards.
- Peithon withdrew to Kandahar in 316. The Macedonian presence in India simply drained away westward.
- Chandragupta pushed at a door that was already swinging. He did fight, and he did take the north-west, and the Greek sources credit him with killing governors. But the empire he pushed over was mostly holding itself up by then.
Justin says Chandragupta gathered a force of latrones — a Latin word that early translators rendered as “robbers”,
which produced a century of writing about a bandit chief who became an emperor.
The word is not that simple. It can mean a mercenary soldier, a hunter, or an irregular fighter, and Indian historians
including Raychaudhuri argued that the early translators had chosen the most insulting available option. The
alternative reading is far more probable: Chandragupta recruited mercenaries, hill fighters and the armed
republics of the Punjab — the same peoples who had fought Alexander in Part Two and had every reason to want the
garrisons gone.
If that is right, then the army that founded India’s first empire was substantially a Punjabi army, and the
Mauryan state was built with north-western manpower before it ever touched the Ganges.
- He struck at the moment of maximum confusion. Between Alexander’s death and the Greek generals sorting out their war, the north-west had no functioning authority for years.
- He recruited the people who were already angry. The republics and hill peoples of Punjab had been fighting garrisons on their own account since 325.
- He had a strategist. Whatever the truth of the Chanakya legend, the campaign shows planning: the north-west first, the Ganges second, in that order and not the other way round.
- Nobody was coming to help the Greeks. Every Macedonian with ambition was three thousand kilometres west, fighting for a bigger prize.
- Alexander’s settlement was never real. He left local kings in place under Macedonian supervisors and marched away. That arrangement needed his personal authority to hold, and he died.
- They were hated. The Greek sources themselves record that the occupying troops had provoked deep hostility, and that priests and local leaders encouraged revolt.
- Their garrisons were tiny and isolated. A few thousand men spread across a country of millions, with no reinforcement route that anybody was using.
- Their commanders left. The single biggest factor. India was a sideshow to the war over Alexander’s empire, and the professionals went where the prize was.
Chandragupta Maurya founded the largest empire India would have for two thousand years. We do not
know where he was born, who his parents were, or what community he came from.
Buddhist sources make him a member of a warrior clan. Jain sources have a different account again.
Brahminical tradition, in the play Mudrarakshasa, makes him low-born — and the Greek writers, who had
no stake in the Indian argument, also suggest humble origins. Later tradition attached him to Punjab, to
Bihar, to Nepal, and to at least three different clans. Every one of these traditions was written down
centuries after he died, by people with a reason to prefer their version.
Justin gives him the origin story of a folk hero rather than a king. A wild elephant of great size, he says,
came up to Chandragupta of its own accord and, as if tamed, took him on its back. In another passage a lion
approaches him while he sleeps, licks the sweat from him, and walks away without harming him.
Those are omen stories, and they are the sort of thing that gets told about a man because nobody knows
where he came from. You do not need a miracle to explain the ancestry of a prince.
The founder of India’s first empire is, biographically, almost a blank — and that blank is the
strongest single piece of evidence that he really did rise from nowhere. Everyone who came after
him needed him to have been one of theirs, which is exactly what happens to a man who was nobody’s.
3The Nanda War — Eating From the Edges STORY
The most famous strategy in Indian history was supposedly learned from a mother scolding her child over a hot dinner.
When: Roughly 322–320 BCE, overlapping with the war in the north-west.
Where: From the Punjab eastward down the Ganges to Pataliputra — modern Patna.
Who: Chandragupta against Dhana Nanda, the last Nanda king of Magadha.
The evidence: Buddhist and Jain traditions and a much later Sanskrit play. Almost nothing contemporary.
Why the Nandas were beatable
Remember from Part Two what Alexander’s men were told at the Beas: that the kingdom ahead of them could field
two hundred thousand infantry and thousands of elephants. It was the largest military power in India and it
terrified an army that had never been stopped.
They were also told the other half, and the Greek writers record it clearly. The Nanda king was despised by his
own people. Curtius and Diodorus both pass on the same information — that he was of low birth, said in one
version to be a barber’s son, and was hated for it. Alexander was standing on a riverbank being told
simultaneously that the enemy was enormous and that the enemy was unpopular. He acted on the first
half.
Chandragupta acted on the second, and it took him about five years.
The lesson at the doorway
Here is the famous story, and its label is Story for good reason.
In the tradition, Chandragupta and Chanakya attack the Nanda heartland directly, marching on the capital first —
and are beaten badly. Fleeing afterwards in disguise, Chanakya overhears a mother scolding a small boy. The child
has grabbed a hot cake and burned his fingers because he went straight for the middle. “You are as foolish as
Chandragupta,” the mother says. “He attacked the centre of the kingdom instead of starting from the
edges, where it is cool.”
Chanakya takes the point. They begin again: secure the frontier provinces first, take the outlying territory, build
allies and strength on the periphery, and only then close on the capital.
1. Secure the north-west first. This is why War One and War Two belong together. The Punjab was not a
- sideshow to the Nanda war; it was the base for it.
- Build an alliance. The tradition gives Chandragupta a partner, a mountain king called Parvataka, whom many scholars have tried to identify with Porus or his family. If there is anything in it, the invasion of Magadha was launched by a Punjab–hill coalition.
- Take the provinces, then the capital. Pataliputra falls last, not first.
- Dhana Nanda is destroyed, and the Mauryan empire begins from a throne in Bihar that was won using an army raised in the north-west.
The hot-cake story appears in Buddhist and Jain traditions written centuries later, and it has the exact structure of a
teaching tale: a wise man humbled, a lesson from an ordinary woman, a strategy learned from a kitchen. It is almost
certainly not something that happened.
But the strategy it describes is real, and it is visible in what Chandragupta actually did. He did secure the periphery
before the centre. His empire was assembled outside-in. And the same principle turns up in the Arthashastra, the
manual of statecraft traditionally attributed to Chanakya, in the form of the mandala theory — the idea that your
neighbour is your enemy and your neighbour’s neighbour is your friend, and that a kingdom should be approached as a
set of rings from the outside.
The anecdote is fiction. The doctrine underneath it was taught, written down, and used.
- He attacked the right target for the right reason. He went after an unpopular king, not a weak one — a completely different calculation, and the correct one.
- He started from the edges. Whatever the source of the idea, the method is sound and is still taught.
- He had a frontier base with a fighting population. Punjab gave him manpower that was already mobilised and already hostile to the existing order.
- He lost first and adjusted. The tradition insists on an early defeat, and the fact that a hero-legend keeps an embarrassing failure in it is a point in its favour.
- Size is not the same as support. The largest army in India belonged to a king his own people would not fight for.
- They had no answer to a war on the rim. A centralised state that concentrates everything in the capital cannot respond to an enemy eating the provinces one by one.
- They had made themselves the point of the state. Kill the dynasty and there is nothing underneath — the same failure you will see again with Mihirakula in Part Four.
- Their reputation had already been used against them. The information that beat them was circulating on the Beas in 326 BCE. It just took an Indian to act on it.
Stand on the bank of the Beas in the monsoon of 326 BCE with the Macedonian army, as they were in Part
Two. They will not cross. They have been told what is on the other side: two hundred thousand foot, three
thousand elephants, the greatest power in India. Alexander argues for days and then gives up, builds twelve
enormous altars, and turns around. It is the only time in his life he is stopped.
Now move forward five or six years to the same riverbank.
The kingdom that frightened the best army in the world has been overthrown by a young man from the
same region, using soldiers recruited from the very peoples Alexander had just finished fighting — and the
reason it worked is the piece of information that Alexander’s own scouts had handed him at that riverbank
and that he had not known what to do with. The Nandas were hated. That was the whole weakness,
and it was in the briefing.
There is a line in Plutarch, written centuries later, that makes this almost unbearable. He says that
Androcottus — Chandragupta — had seen Alexander himself when he was a young man, and afterwards
used to say that Alexander narrowly missed making himself master of the country, because the
reigning king was hated and despised for the baseness of his origin.
We cannot verify a word of it. It is chronologically possible and completely unprovable, and Part Two
already warned you about the famous version of this story that turned out to rest on a printer’s error.
But keep the shape of it, because the shape is true whether the meeting happened or not. The empire on
the far bank could be taken. Alexander was told so, at the river, by people who lived there. He
turned back — and an Indian who was listening did it instead.
4Seleucus Comes Back — And Sells Half of Afghanistan for Elephants LIKELY
The only war in this book that ends with the loser getting the better long-term deal, and both sides knowing it.
When: 305–303 BCE.
Where: Somewhere east of the Indus. No location is recorded anywhere. Nobody knows.
Who: Seleucus I Nicator, Alexander’s former officer and now ruler of Asia, against Chandragupta.
The evidence: A handful of lines in Appian, Strabo, Plutarch and Justin. No campaign account survives.
Why it happened
Seleucus had spent twenty years fighting for a share of Alexander’s empire and had ended up with the largest
piece of it — Syria, Mesopotamia, Iran, and everything east. In his own mind, that inheritance included India,
because Alexander had taken it.
So around 305 BCE he came east to collect. Appian says simply that he crossed the Indus and made war on
Sandrocottus.
What happened next, which is remarkable
1. No battle is described. Not one ancient writer names an engagement, a place or a date. Appian goes straight
- from “waged war” to “until they came to an understanding with each other.”
- Seleucus gave up four provinces. Strabo lists what changed hands: Arachosia (Kandahar), Gedrosia, the Paropamisadae (the Kabul valley and the Hindu Kush foothills), and Aria (Herat). That is, roughly, the eastern half of modern Afghanistan and the country down to the Makran coast.
- He received five hundred war elephants. Strabo says the provinces were given to Sandrocottus upon terms of intermarriage and of receiving in exchange five hundred elephants. 4. There was a marriage agreement. The Greek word is epigamia, and what exactly it meant is argued about — a dynastic marriage, or a general right of intermarriage between the two peoples.
- And then the two states stayed friendly for generations. Seleucus sent an ambassador, Megasthenes, who lived at Pataliputra and wrote the Indika, the first substantial Greek description of India. His successor sent another. The border did not move again.
This is a live argument. The military historian John D. Grainger puts it about as honestly as it can be put: it is
unknown whether there was a pitched battle at all — and yet the outcome was a decisive Indian victory. Those two
statements sit together perfectly well. A commander who arrives, assesses the situation and negotiates away
four provinces has lost, whether or not anybody died.
A minority view, argued by Hartmut Scharfe in 1971, held that Seleucus actually kept overlordship and Chandragupta
became a subordinate. Essentially no other scholar has accepted it, and the terms make no sense on that reading.
Nobody hands over four provinces as the winner.
What tipped it was almost certainly arithmetic. Chandragupta could put an enormous army into the field close to home.
Seleucus was at the end of a very long march, with an unfinished war waiting for him in the west — and he knew what
he needed more than territory.
- Numbers, close to home. Plutarch claims six hundred thousand men, which is a fantasy figure, but the underlying point stands: the Mauryan state could field more soldiers in Punjab than Seleucus could bring across Iran.
- He was fighting on ground he had just conquered and organised. The north-west was his base, not his frontier.
- He had the one commodity his enemy urgently wanted. Elephants were the strategic weapon of the age and India was the only supplier that mattered.
- He knew Seleucus had a war behind him. Time was worth more to Seleucus than land, and Chandragupta priced accordingly.
- His real enemy was in Turkey, not India. Antigonus, the most dangerous of Alexander’s successors, was preparing for a war that would decide who owned Asia.
- The four provinces were expensive to hold and worth little. Mountain country, thin revenue, hostile populations, enormous distance.
- An eastern frontier that stays quiet is worth a great deal. He converted a permanent military problem into a permanent friendship, and it held for a century.
- And he was right. Which is the point of the story below.
Follow the five hundred elephants, because almost nobody does, and their journey is the reason this
obscure treaty belongs in world history.
They were handed over somewhere in the north-west around 303 BCE. Then they were walked — across
Afghanistan, across Iran, across Mesopotamia, and up into Anatolia. It is a distance of something like four
thousand kilometres, and every one of those animals had to be fed on the way. Feeding a single working
elephant takes a very large amount of fodder every day. Five hundred of them is a marching town.
In 301 BCE they arrived at a place in central Turkey called Ipsus, where Seleucus and his allies met
Antigonus in the largest battle of the age — the one that would finally settle who inherited Alexander’s
empire.
The elephants decided it. Antigonus’s cavalry, under his son Demetrius, charged and drove the enemy horse
off the field, and then found that they could not get back: the elephant line had been moved across
behind them, and horses will not face elephants. The best striking force in the army was locked out of
its own battle. Antigonus, in his eighties, was killed where he stood.
Seleucus took the largest share of the spoils and founded the dynasty that would rule the Middle East for
two hundred and fifty years.
The Hellenistic world was shaped at Ipsus, and Ipsus was won with animals bought from Punjab.
Chandragupta gained an empire from the deal; Seleucus gained one too. It is the rarest thing in this book —
a war that ended with both sides better off, and the only reason it worked is that each of them wanted
something the other did not value.
5The Two Revolts of Taxila STORY
The empire’s richest western city rebelled twice in one lifetime, and both times the complaint was the same: not the emperor, the officials.
When: The first around c. 272 BCE under Bindusara; the second later in the same reign, and possibly a third under Ashoka.
Where: Taxila — Takshashila — the great city of the north-west, and its province.
Who: The citizens of Taxila against the imperial administration; the princes Ashoka and Sushima sent to deal with it.
The evidence: The Ashokavadana, a Buddhist legend collection written down centuries later — plus one piece of unexpected corroboration.
What the legend says
1. Taxila revolts. The emperor Bindusara — Chandragupta’s son — sends his son Ashoka to put it down.
2. He sends him with nothing. In the story, the emperor refuses to provide weapons or chariots for the
- expedition. Ashoka was not the favourite son, and the tale is clear that this was meant as a humiliation. Gods then supply the missing equipment, which is how you can tell you are reading a legend.
- The city opens its gates. This is the part worth keeping. When Ashoka arrives, the citizens come out and explain that they are not opposed to the emperor or to the prince — they are opposed to Bindusara’s oppressive ministers. Ashoka enters the city without a fight.
- It revolts again. Before Bindusara’s death Taxila rises a second time. This time the eldest son, Sushima, is sent — and fails.
- And there is a third story, later and darker. Under Ashoka, his son Kunala is sent to Taxila as viceroy. In the legend, a queen alters a letter from the emperor by adding a single dot to a word, turning an order that the prince should study into an order that he should be blinded — and Kunala, believing it to be his father’s command, obeys it.
The Ashokavadana is Buddhist devotional literature. Its purpose is to show Ashoka’s transformation, and every episode
is shaped to that end. The dot-in-the-letter story is a folk motif. The gods handing out weapons is not history.
But look at the grievance the legend puts in the citizens’ mouths, and then look at what Ashoka himself carved on rock.
In his Separate Rock Edicts, addressed to his officials in the provinces, Ashoka worries about exactly this. He tells his
officers that all his subjects are his children; he warns them against acting out of jealousy, anger, cruelty, haste or
laziness; and he orders regular inspection tours specifically to check that officials are not mistreating people.
An emperor does not build a formal inspection system to solve an imaginary problem.
The legend and the inscriptions agree on the diagnosis even though only one of them is history: the Mauryan
empire’s weak point in its distant provinces was its own officials.
- It sent princes, not just armies. A member of the royal family arriving in person makes negotiation possible in a way that a general cannot.
- It could distinguish a tax revolt from a secession. Taxila was not trying to leave the empire. Recognising that is what let the first revolt end without a siege.
- The province was worth the effort. Taxila sat on the trade routes to Central Asia and was one of the wealthiest cities the empire owned.
- And it built machinery to prevent a repeat. The inspection tours in the edicts are a genuine administrative answer to a genuine problem.
- Distance. Taxila to Pataliputra is a very long way. Officials that far from the throne can behave as they like for years before anyone notices.
- The city had its own identity. It had been a Persian provincial capital, a Greek prize and an independent kingdom within living memory. It did not think of itself as a district.
- Wealthy cities have leverage. A place that generates a large share of the imperial revenue knows it is expensive to punish.
- And nothing was fixed structurally. Within fifty years of Ashoka’s death, the north-west was gone.
Almost every revolt in this book is about who rules. This one is not, and that is why it deserves its own page.
The people of Taxila did not march out to say that the Mauryas were foreigners, or that they wanted their
old kings back, or that they refused to be part of an Indian empire. According to the tradition, they came
out of the gates and said, in effect: we have no quarrel with the emperor and none with you. Your
ministers are robbing us.
Whether or not those words were said, the story survived because everyone recognised it. Taxila is months
away from Pataliputra by road, and in that gap an imperial officer is the whole government.
And the response is what makes it remarkable. The prince did not sack the city. He listened and went in
without a fight — and, if the tradition’s sequence means anything, that same man later built the only
empire in the ancient world with a formal system for inspecting its own officials.
The most sophisticated administrative idea of the ancient Indian state may have started as a
complaint shouted at a prince outside the walls of a Punjabi city.
Hold on to Taxila. You will see it again in every part of this book until the sixth century CE, when it stops
appearing — because by then it is a field of broken foundations, burned in the Hun invasions of Part Four
and never rebuilt.
6The War That Ended in Writing CERTAIN
The only war in this book whose most important consequence was carved on rocks in Punjab, in two scripts, by a man announcing that he would not do it again.
When: The Kalinga war around 261 BCE; the edicts carved from roughly 257 BCE onward.
Where: The war was in Odisha. The declaration is in Punjab and Gandhara — Shahbazgarhi, Mansehra, and Kandahar.
Who: Ashoka, third Mauryan emperor, and the north-western population who had to be told what he had decided.
The evidence: His own inscriptions, cut into rock in his lifetime, on his orders. As certain as evidence gets in this book.
Why a war in Odisha belongs in a book about Punjab
Because of where he chose to announce it.
The Kalinga campaign was fought on the east coast, about two thousand kilometres from the five rivers, and this
book will not pretend otherwise. What happened afterwards, though, is a Punjab story, and here is why. Ashoka
took the most personal statement any ancient ruler ever made about war, and had it cut into rock faces
on the north-western frontier, in the local script, for the people there to read.
At Shahbazgarhi and Mansehra, in the Gandhara country, his major edicts are carved in Kharoshthi — the
right-to-left script of the north-west that Part Two described as the Persian empire’s lasting gift to Punjab.
Elsewhere in the empire, the same edicts are in Brahmi. Further west, at Kandahar, a version was cut in Greek
and Aramaic, for populations who read neither Indian script.
This is an emperor translating himself. The Mauryan state had worked out that a message carved in a
language nobody local can read is not a message.
What the thirteenth edict actually says
Rock Edict XIII is the famous one, and it is worth being precise about its contents, because it is quoted constantly
and read carelessly.
1. It gives numbers, and terrible ones. A hundred and fifty thousand people deported, a hundred thousand
- killed, and many times that number dead afterwards. An emperor publishing his own casualty figures is close to unique in the ancient world.
- It expresses remorse in the first person. The text says the Beloved of the Gods felt remorse at the conquest of Kalinga, because conquering a country that has not been conquered before means killing, death and deportation, and that this is deeply painful to him.
- It says the suffering of the innocent is the worst part — that even those who escape are hurt by the injury done to those they love.
- It renounces conquest as policy. It states that the only true conquest is conquest by dhamma — by right conduct — and asks his sons and grandsons not to think of new conquests.
- And it names foreign kings. It lists the Greek king Antiochus and, beyond him, four more rulers — the kings we know as Ptolemy, Antigonos, Magas and Alexander. This is the single most useful passage in Indian history for fixing dates, because those five men can be dated from Greek sources.
Rock Edict XIII, with all its remorse and its casualty figures, is carved across the empire — including at Shahbazgarhi
in the north-west.
It is not carved in Kalinga. At the two sites in the conquered country itself, Dhauli and Jaugada, that edict is left out
— along with two others — and two different texts are put in its place: the ones addressed to his officials about treating
people justly.
Historians have argued about why for a century, and there is no settled answer. The obvious reading is uncomfortable
and hard to dismiss: the emperor’s public grief about the massacre was published everywhere except to the
people who had been massacred.
None of this makes the edicts fraudulent. Ashoka did stop conquering, and there is no evidence of another war of
expansion in a reign of nearly forty years, which is an extraordinary record. But this book’s rule applies to him as to
everyone: always ask who is writing, and where they chose to put it.
- He wrote to the frontier in the frontier’s own script. Kharoshthi in Gandhara, Greek and Aramaic at Kandahar. The empire understood that Punjab and the north-west were a different world and addressed them as one.
- He put policy on stone rather than in an archive. Paper burns and officials lie. A rock face beside a road is a promise the government cannot quietly withdraw.
- He made his officials the subject. The edicts are largely about the behaviour of the administration — the problem Taxila had revolted over twice.
- And he actually stopped. Whatever the motive, no further war of conquest is recorded in his reign, and that is not nothing.
- An empire that renounces war still needs an army. Ashoka did not disband his; his successors could not use it.
- Moral authority does not transfer. The system depended on the personal reputation of one emperor, and it did not outlive him by fifty years.
- The frontier remained a governor’s problem. Punjab was still administered from two thousand kilometres away by men whose conduct required inspection tours to control.
- And the passes were still open. The message carved at Shahbazgarhi faced west, towards the road that Demetrius would come down in ninety years.
Somewhere around 257 BCE, a man in Kandahar was employed to carve an Indian emperor’s religious and
moral programme into rock in fluent, idiomatic Greek. It was found in 1958, and it is not a rough
translation — it is competent Greek prose, using Greek philosophical vocabulary to render Indian ideas.
Beside it, the same message in Aramaic, the old administrative language of the Persian empire, still in use
across the region two generations after that empire had ceased to exist.
Think about what that document required. The Mauryan court had to know that a Greek-speaking
population existed out there. It had to employ somebody who could think in both languages well enough to
find Greek words for concepts that had no Greek words. It had to decide that those people’s opinion was
worth the cost of the stonemason.
And it had to want to be understood. An empire that only wants obedience does not need translators.
It needs tax collectors.
That is the thing about these inscriptions that survives every argument about Ashoka’s sincerity. All across
the north-west, in three scripts and two languages, a government carved messages into rock explaining
itself to its own frontier — not proclaiming victories, not listing conquests, but explaining what it thought a
ruler was for.
Nothing like it exists again on this soil for a very long time. The next several hundred years of Punjab’s
history, as the rest of this part will show, are recorded almost entirely on coins.
7185 BCE — The Empire Ends at a Parade LIKELY
India’s first empire was not conquered. It was killed by its own commander-in-chief during a review of the troops.
When: c. 185 BCE, about forty-seven years after Ashoka’s death.
Where: Pataliputra — but the consequence lands immediately on the north-west frontier.
Who: Pushyamitra, the Mauryan senani or commander-in-chief, against Brihadratha, the last Mauryan emperor.
The evidence: The Puranas, and a much-quoted passage in Bana’s Harshacharita written eight centuries later.
What happened
The account that everyone quotes comes from Bana, the seventh-century writer you will meet again in Part Four
as the biographer of Harsha. He describes it in a single vicious sentence: that the commander Pushyamitra, on the
pretext of showing him the whole army on parade, crushed his master Brihadratha and took the
kingdom.
That is essentially all we have for the event itself. The Mauryan empire — the largest state in Indian history to
that date, the empire of Chandragupta and Ashoka — ends in one line, at a military review, at the hands of the
man whose job was to command its soldiers.
Why the empire was already finished
1. The succession had fractured. After Ashoka the sources become confused and contradictory, with several
- claimants and an empire that may already have been governed in halves.
- The provinces had drifted. Kalinga went its own way; the Deccan slipped; and crucially, the north-west was no longer under real control.
- The economy that carried it was enormous and expensive. A state that runs on a vast bureaucracy and a standing army needs its revenue to keep flowing from provinces two thousand kilometres away, and by 200 BCE that was not happening. 4. The army was the last functioning institution, which is precisely why its commander was in a position to take the throne.
- And the frontier was undefended. Within a very few years of the coup, Greek armies came back through the passes — this time from the other direction.
Pushyamitra is accused in Buddhist sources of persecuting Buddhism. The Ashokavadana tradition says he offered a
bounty for the head of a Buddhist monk — a hundred coins each — and it names Sakala, modern Sialkot, as the place
he proclaimed it. Another text says he burned scriptures and destroyed hundreds of monasteries; the sixteenth-century
Tibetan writer Taranatha says the doctrine was wiped from the north within five years.
This is contested ground, and it is contested for modern political reasons as well as historical ones, so here is the state of
the evidence plainly. Étienne Lamotte found the Buddhist accounts inconsistent with each other on basic points.
Romila Thapar points to the absence of archaeological evidence for widespread destruction in this period. H. C.
Raychaudhuri argued the coup cannot be read as a religious uprising at all. Others, including K. P. Jayaswal,
suggested that if there was repression it was aimed at Buddhist communities that were politically aligned with the
Greeks — a security measure rather than a theological one.
There is also a dating problem inside the accusation itself. The bounty in the story is quoted in dinaras — a word that
comes from the Roman denarius and does not enter Indian usage until centuries after 185 BCE. The coin in the story
had not been invented yet. That does not prove the underlying memory false, but it does show the text being written
a very long way downstream of the events.
Against that: the accusations are old, repeated, and specific about places. Something is being remembered.
The defensible position is that Pushyamitra was a Brahminical restorationist who revived the old royal
sacrifices, that Buddhist tradition remembered his reign as hostile, and that the claims of mass destruction
are not supported by anything in the ground. Anyone who states more than that in either direction is arguing
about the present.
- He already had the army. The shortest route to a throne in a failing state is to be the person the soldiers already obey.
- He restored the old rituals. He performed the ashvamedha, the horse sacrifice — the ancient ceremony by which a king claims supremacy — which was both a religious act and a public announcement that the Mauryan experiment was over.
- He fought the Greeks and could say so. Kalidasa’s play Malavikagnimitra, written centuries later, has his grandson Vasumitra defending the sacrificial horse against Yavana cavalry on the bank of a river called the Sindhu. Late evidence, but it shows what the dynasty was remembered for.
- The north-west, immediately. The Shungas were a Ganges power. Punjab was not theirs and would not be Indian-ruled again for a long time.
- The administrative machine. The inspection tours, the edicts, the standardised provincial government — all of it goes, and nothing replaces it.
- The idea of a single Indian state. It does not seriously return until the Guptas, five hundred years later, and even then it does not cover Punjab.
- And it created the vacuum that the rest of this part fills. Four foreign dynasties walk through the door that opened in 185 BCE.
Two men stand on a parade ground in about 185 BCE.
One is the great-grandson — or thereabouts — of the emperor who published his own casualty figures and
asked his descendants never to conquer anything again. He is reviewing an army he has inherited and does
not command in any meaningful sense.
The other is the man who does command it.
There is a hard question underneath the scene, and this book is not going to pretend to answer it, because
nobody honestly can. Did the Mauryan empire fall partly because it had talked itself out of the
business of force?
It is an old argument. Nineteenth- and twentieth-century historians, some of them with obvious agendas,
blamed Ashoka’s pacifism for the collapse. Others have pointed out, correctly, that the empire survived him
by nearly fifty years, that its actual problems were succession, distance and money, and that no ancient
state that size lasted much longer anyway. Ashoka never dissolved his army — Rock Edict XIII warns forest
peoples that the emperor still has power and expects them to behave.
What is not in doubt is the sequence, and the sequence is the story. An emperor announces that
conquest is not worth what it costs. Two generations later, his family is removed by the officer in
charge of the army, at a parade, and within a few years foreign cavalry is back in the Punjab.
Put those facts next to each other and resist the urge to draw the obvious moral. The obvious moral is the
reason this period gets argued about so badly. But the facts are the facts, and they are in that order.
8The Greeks Come Back — Demetrius Through the Passes LIKELY
A hundred and forty years after Alexander’s men refused to go on, their descendants came back down the same road — and this time they stayed for two centuries.
When: Roughly 200–180 BCE, accelerating after the Mauryan collapse in 185.
Where: The Kabul valley, Arachosia, Gandhara, Taxila, the Indus and the western Punjab.
Who: Demetrius I, king of Bactria, son of Euthydemus, against whatever authority was left in the north-west.
The evidence: Overwhelmingly coins. A few lines in Strabo, Polybius and Justin. No battle account at all.
Who these Greeks were
Bactria — the country around Balkh in northern Afghanistan — had been settled with Greek colonists by
Alexander and his successors. Around 250 BCE its governors broke away from the Seleucid empire and set up as
independent kings. Strabo says the Greeks who made Bactria revolt grew so powerful, because of the
fertility of the country, that they became masters not only of Ariana but of India.
These were not visitors. They had been in Central Asia for three or four generations, they spoke Greek and minted
Greek coins and built Greek cities, and they were also thoroughly of that region. When Antiochus III came east in
about 206 BCE and besieged Bactria, the settlement he reached included recognising the local dynasty and
betrothing his daughter to a young prince, described by Polybius as a neaniskos — a youth.
That young man was Demetrius, and about fifteen years later he came through the passes into India.
What happened
1. He takes the Kabul valley and Arachosia, the very provinces Seleucus had signed over to Chandragupta in
- War Three. The 303 BCE settlement is undone.
- He comes down into Gandhara and the Indus country. Taxila is taken; the coin finds there are extensive. 3. The Mauryan collapse is the trigger. Ancient and modern accounts agree that the destruction of the Mauryan dynasty by Pushyamitra is what opened the door. India’s first empire fell in 185 and Greek armies were in the Punjab within a few years.
- A separate Indo-Greek kingdom emerges. Within about twenty years there are Greek kings ruling from India rather than from Bactria — the first clearly being Apollodotus I, whose coins cluster in Gandhara and the western Punjab.
- And they start behaving like Indian kings almost immediately, which is the theme of the rest of this part.
Something like thirty to forty Greek kings and queens ruled in Bactria and north-western India across two
centuries, and the great majority of them are attested by nothing whatsoever except their coins. No history mentions
them. No inscription names them. We know their faces, their titles and roughly when and where they ruled, and
nothing else at all.
The chronology of the whole period has been reconstructed largely by numismatists — Osmund Bopearachchi’s work
is the standard — from coin styles, mint marks, overstrikes and hoard distributions. It is a genuine intellectual
achievement and it is also, unavoidably, a scaffolding.
This means any date in this section may move. Where this part gives a year for an Indo-Greek king, read it as an
estimate with a decade of give either side.
- The door was open. The Mauryan frontier administration had gone and the Shungas were a Ganges power with no interest in the Indus.
- They already held the far side. Bactria is on the strong side of the passes — exactly the position that Part Five will show decides everything.
- Cavalry and siegecraft. Greco-Bactrian armies were technically sophisticated and had the Hellenistic world’s engineering behind them.
- They came as a kingdom, not a raid. They founded cities, appointed governors and issued a currency. This was annexation from the start.
- There was no north-western state left to stop them. Taxila had been an imperial province for a century and had no independent army of its own.
- The Shungas had just fought a civil war. A new dynasty consolidating in Bihar was not going to campaign on the Indus.
- The republics were not yet reorganised. The Yaudheyas and their neighbours are a factor again later in this part, but not in 185 BCE.
- And the passes were unguarded. This will be the last time this book says it politely.
Pick up a silver coin of Demetrius I and look at the portrait, because it is one of the most deliberate images
in ancient history.
He is shown wearing an elephant’s scalp as a headdress — the animal’s skin drawn over his head, with
the trunk and tusks rising above his forehead and the ears framing his face.
This was not a fashion. It was a quotation, and every literate person in the Hellenistic world would have
caught it instantly. Alexander had been portrayed exactly this way after his Indian campaign, and
the elephant scalp meant one thing: the conqueror of India.
So here is a Greek king, born in Central Asia, four or five generations after Alexander, putting on the
costume of a dead Macedonian to announce that he has done the thing Alexander’s army refused to finish.
It is the ancient equivalent of a man being photographed in someone else’s uniform.
And the coins tell you the rest of the story too, if you keep looking. Within a generation or two, the coinage
of his successors carries Greek on one side and Kharoshthi on the other — Indian script, Indian titles,
Indian weight standards — and Indian gods start appearing on the reverse. One of Demetrius’s own issues
may already carry the goddess Lakshmi.
He arrived dressed as Alexander and his grandchildren were minting money for Indians to use, in
the local script, with local gods on it. That is the four-hundred-year habit, visible on a piece of metal you
could hold in one hand.
9Menander of Sialkot Marches on Pataliputra LIKELY
A Greek king ruling from Sialkot got further into India than Alexander ever did — and is remembered in India not as a conqueror but as a man who asked a monk what a chariot was.
When: Reigned roughly 165/155–130 BCE. Even the start of his reign is disputed by a decade.
Where: Capital at Sagala — Sialkot. Territory from the Kabul river to the Ravi, and campaigns far to the east.
Who: Menander, known in Indian sources as Milinda, against the Shungas and whoever else stood in the way.
The evidence: An enormous coinage, two Indian literary references, a prophecy, and a Buddhist dialogue.
The march east
Three separate Indian sources describe Greek armies pushing deep into the Ganges plain in this period, and they
are worth taking one at a time because they are very different kinds of evidence.
1. Patanjali, the grammarian, writing around 150 BCE. In his Mahabhashya he needs examples of a particular past tense — the one used for recent events within the speaker’s own lifetime — and he reaches for
- two: “arunad Yavanah Saketam” and “arunad Yavano Madhyamikam”. “The Yavana was besieging Saketa.” “The Yavana was besieging Madhyamika.” This is a grammar textbook, which makes it almost uniquely valuable: nobody lies in a grammar example. He picked those sentences because his readers would recognise them as things that had just happened.
- The Yuga Purana, a strange short apocalyptic text. It says that after conquering Saketa, the country of the Panchalas and the Mathuras, the Yavanas — wicked and valiant — will reach Kusumadhvaja, which is Pataliputra; that on reaching its thick mud fortifications all the provinces will fall into disorder; and that a great battle will follow, fought with tree-like engines, meaning siege machinery.
- Strabo, quoting Apollodorus, says that more tribes were subdued by the Bactrian Greeks than by Alexander — by Menander in particular — if he really crossed the Hypanis, the Beas, and advanced eastward.
Notice that last conditional. Even the Greek source is hedging. But put the three together and the picture is
consistent: in the middle of the second century BCE, an army from Punjab under a Greek king went east
and reached the gates of the Mauryan capital. Alexander’s men had turned back at the Beas. Menander’s did
not.
Why he did not stay
The Yuga Purana answers this, and it is one of the most quietly devastating passages in Indian literature. The
Yavanas, it says, intoxicated with fighting, will not remain in the Middle Country; there will be a civil war
among them, arising in their own land, and there will be a terrible and ferocious war.
That is exactly what happened, and it is the subject of the next war in this book. The Greek advance into India
was stopped by Greeks.
The numismatist Joe Cribb has argued that there is no coin evidence for Menander east of Taxila, and that the Indian
literary references to a march on Pataliputra may belong to somebody else entirely — possibly the Kushans, two
centuries later, whose reach into the Ganges plain is much better attested.
That deserves to be taken seriously. Coins usually follow armies, and their absence is a real argument. Against it:
Patanjali is contemporary, and a raid or a campaign does not always leave a mint behind it.
There is a further complication worth knowing. Some numismatists now think there were two different Indo-Greek
kings called Menander, separated by decades, and that older accounts have merged them into one very long and very
impressive reign.
The safe version: Greek armies from Punjab campaigned deep into northern India in the second century
BCE; Menander was the most powerful Indo-Greek king; the association is likely but not proven.
- He ruled from inside India, not from Bactria. Sagala is a Punjabi city on the Punjabi plain. His power base, his revenue and his recruits were all local.
- He had the best coinage in the region, and more of it than any other Indo-Greek king. A ruler who can pay reliably can campaign for decades.
- He fought a Shunga state past its best. Pushyamitra’s dynasty had never recovered the reach of the Mauryas.
- He was accepted by his Indian subjects. His coins carry Kharoshthi legends and Indian royal titles, and Buddhist tradition adopted him outright. A foreign king who is not resented is a foreign king who can take his army far from home.
- Civil war at home. The Greek kingdoms spent the later second century BCE destroying each other, which is War Nine.
- The line was too long. Sialkot to Patna, held with a cavalry army, against a country that has depth. Everybody who tries this in this book fails at it.
- No institution behind the man. His kingdom fragmented on his death into a scatter of small realms, several ruled simultaneously.
- And something worse was moving. While the Greeks fought each other, the Sakas were coming off the steppe behind them.
Menander is one of very few foreign rulers in Indian history whom Indian tradition adopted rather than
resented, and the reason is a book.
The Milindapanha — “The Questions of Milinda” — is a Buddhist text in Pali in which King Milinda, that is
Menander, debates with a monk named Nagasena. It is a real classic of Buddhist literature, and its most
famous passage is an argument about a vehicle.
The monk asks the king how he came. In a chariot, says Menander. So Nagasena asks him to point to the
chariot. Is the axle the chariot? No. The wheels? No. The frame, the yoke, the reins, the goad? No, none of
those. Then is the chariot something separate from all of them? No. Then, says Nagasena, there is no
chariot — “chariot” is only a word we use for a collection of parts assembled in a certain way.
And then he turns it around, because the argument was never about transport. In exactly the same way,
he says, there is no permanent self called “Nagasena” — only a body, feelings, perceptions and
thoughts, working together, and a name we give the arrangement.
Menander, in the story, concedes the point and becomes a supporter of the faith.
Consider where this is set. Sagala — Sialkot — in the second century BCE. A Greek king, heir to the
philosophical tradition of Athens, sitting in a Punjabi city arguing about the nature of the self with an
Indian monk, in a debate that both sides thought worth writing down.
Plutarch adds the ending, and it is startling. When Menander died in camp, he says, the cities of his
kingdom quarrelled over his remains, and the dispute was settled by dividing his ashes equally so that
each could carry a share away and raise a monument over it.
That is precisely what Buddhist tradition says was done with the relics of the Buddha. A Greek king from
Punjab was buried like an Indian saint, and the man who recorded it was a Greek writer in
Roman Boeotia who almost certainly had no idea what he was describing.
10The Greek Civil Wars — And the Chariot Through the Blood LIKELY
Nobody defeated the Greeks in India. They spent sixty years defeating each other, and the last act was a son driving a chariot over his father’s body.
When: Roughly 170–130 BCE for the main phase, with fragmentation continuing for a century after.
Where: Bactria, the Kabul valley, Gandhara and the Punjab — Greek against Greek, throughout.
Who: Eucratides I against the house of Euthydemus and Demetrius; then everyone against everyone.
The evidence: One lurid passage in Justin, and a vast confusing mass of coins.
What happened
1. Eucratides seizes Bactria, around 171 BCE, overthrowing the dynasty that Demetrius belonged to. He is
- competent, ruthless and successful, and he calls himself Eucratides the Great on his own coinage, which tells you something about him.
- He then invades the Indo-Greek kingdom. The Greeks in India and the Greeks in Bactria — the same people, two generations apart — go to war with each other over the Kabul valley and Gandhara.
- He is besieged and survives it in a way that is barely credible. Justin says that a king called Demetrius, ruler of the Indians, penned him in with sixty thousand men, and that Eucratides, with a garrison of only three hundred soldiers, held out for four months by continual sallies and escaped. Treat both numbers as literary. Treat the siege as real.
- He pushes into India as far as the Indus, and holds ground there for a time before Menander takes much of it back.
- And then his own son murders him. Returning from India, Justin says, he was killed on the march by the son with whom he had shared his throne — a son who did not conceal the murder but drove his chariot through his father’s blood and ordered the body to be thrown out unburied. 6. The Greek position never recovers. Bactria falls to nomads within about fifteen years, and in India the Greek kingdoms fragment into a scatter of small realms ruling simultaneously in different valleys.
Justin is a Roman epitomiser writing centuries later, summarising a lost work, and the passage is exactly the kind of
thing a summariser keeps because it is memorable. The chariot detail is unverifiable.
What is verifiable is the outcome. Eucratides’ coinage stops abruptly around 145 BCE, his successors are short-lived
and confused, and the whole Greco-Bactrian kingdom collapses within about fifteen years of his death. Something
violent happened at the top, and the numismatic record and the story agree on the date.
- No Indian power could touch them for sixty years. The Shungas could not reach the Indus, and the north-west had no state of its own.
- Their military system was excellent. Combined arms, good cavalry, real siege engineering, and disciplined infantry.
- And every one of those advantages was turned inward. The best army in the region spent two generations fighting the second-best army in the region, which was made of the same people.
- No rule for the succession. Hellenistic kingship went to whoever could take it. That is survivable in a large empire and lethal in a small one.
- Two capitals, one people. Bactria and India pulled in opposite directions and there was never a mechanism to decide between them.
- They were a ruling class, not a population. A few tens of thousands of Greeks governing millions of Indians and Bactrians. Split that class and there is no reserve.
- They stripped the northern frontier to fight in the south. Which is exactly when the nomads arrived — and that is War Ten.
Here is the strangest fact in this entire book, and it belongs to the Indo-Greeks.
Somewhere between thirty and forty Greek monarchs ruled in Bactria and north-western India between
about 250 BCE and the first century CE. We have their names. We have their portraits — realistic,
individual portraits, in high-quality silver, showing men and at least one woman ageing across a reign, jaw
by jaw and wrinkle by wrinkle. We know their royal titles: the Saviour, the Just, the Unconquered, the
Victorious.
And for most of them, that is the entire historical record. Not one written source anywhere in the
world mentions them. No chronicle, no inscription, no letter, no poem, no legend. They ruled Punjab and
Gandhara, they raised armies, they fought, they lost, they died — and the only evidence that any of it
happened is coins turned up by ploughs.
Whole reigns have been reconstructed from the mint marks on a few hundred pieces of silver. Wars are
inferred from overstrikes — a coin of one king hammered flat and restruck with another king’s face, which
is what a conqueror does with the captured treasury of the man he has just beaten. You can read the
collapse of a dynasty in the fact that one man’s portrait is showing faintly through the metal
underneath another man’s.
Think about what that means for everything else in this book. These kings were not obscure. They were the
government. They controlled Taxila and Sialkot and the Kabul valley in an age of literate, well-organised
states — and they vanished so completely from the written record that we do not know what a single one of
them did on a single day.
The winner is whoever gets written about. Menander is remembered because a monk wrote down an
argument with him. The other thirty had no monk, and so they have faces and nothing else.
11The Sakas Take Taxila LIKELY
A migration two thousand miles away in the grasslands of Central Asia pushed one people into another until, several collisions later, Punjab had new masters.
When: Roughly 90–20 BCE, the conquest spreading over two generations.
Where: Gandhara, Taxila, the Punjab plain, and eventually down into Sindh and Gujarat.
Who: Maues, then Azes I, against the fragmented Indo-Greek kingdoms.
The evidence: Coins, a dated copper plate from Taxila, and the archaeology of the city itself.
Why they came
This is the first time in this book that a war in Punjab is caused by something happening in Mongolia, and it will
not be the last.
Somewhere in the second century BCE, a people the Chinese called the Yuezhi were driven west out of the frontier
country of China. They collided with the Sakas — the Central Asian people the Greeks called Scythians — and
pushed them off their pastures. The Sakas moved south and west, into Bactria, and in about 130 BCE they finished
off what the Greek civil wars had left of the Greco-Bactrian kingdom. The last Greek king in Bactria, Heliocles,
disappears around then.
Having taken Bactria, they came down the passes.
Note the sequence carefully, because it is the mechanism behind half of this book. Nomads move on the steppe.
Two thousand miles away, a farmer in the Punjab has a new landlord. The same chain of events will
produce the Kushans in War Twelve and the Huns in Part Four.
What happened
1. Maues takes Gandhara and Taxila, some time around 90–80 BCE. He is the first Saka ruler in India whose
- coinage is substantial, and he adopts a title the Greeks would recognise and the Persians used: King of Kings, the Great. His coins carry it in Greek on one side and Kharoshthi on the other. 2. We have a dated document. A copper plate from Taxila, recording a Buddhist relic dedication by a man named Patika, is dated in the reign of the great king Moga — which is Maues. It is the sort of unglamorous administrative object that is worth more than a chronicle.
- Azes I finishes the job, around the middle of the first century BCE, extinguishing most of what was left of Greek rule in the Punjab. An era beginning in 58/57 BCE is associated with him, and its relationship to the later Vikrama era of Indian tradition has been argued over for a hundred and fifty years.
- They govern through satraps. The Sakas brought with them the Persian provincial system, and the words come with it: kshatrapa and mahakshatrapa, satrap and great satrap. These titles enter Indian political vocabulary here and stay for centuries.
- And they patronise Buddhism heavily, almost immediately. Saka officials and their families appear again and again in the north-west as donors to monasteries and stupas.
Taxila is not one city but several, built beside each other over centuries, and the excavations let you walk from one
conquest to the next.
The Indo-Greek city — Sirkap — is laid out on a grid, with a main street and regular blocks, in the standard Hellenistic
town-planning pattern used from Sicily to Afghanistan. The Sakas moved in and lived in it. They did not raze it and
start again; they repaired it, extended it, and put up their own religious buildings inside somebody else’s street plan.
In one part of Sirkap stands a shrine known as the double-headed eagle stupa, and its decoration is the single best
summary of this period in one object. The niches on it are framed in three completely different architectural styles at
once: a Greek pediment, a Persian arch and an Indian gateway, side by side on the same wall, with a Central
Asian bird on top.
Nobody designed that as a statement about cultural fusion. It is just what a builder in Taxila had in his
pattern book.
- They arrived while the Greeks were fighting each other. Sixty years of civil war had left a dozen small kingdoms where there had been one power.
- Horse archers, in numbers. The steppe military package — mobility, missile fire, and the ability to refuse battle — against static kingdoms.
- They took the north side first. Bactria before Gandhara, Gandhara before the plain. The same correct sequence Persia used in Part Two.
- They kept the existing machinery. Greek legends on the coins, existing cities, existing mints, existing officials. Continuity is cheaper than conquest.
- They had already spent themselves. War Nine is the reason War Ten was easy.
- They lost the reservoir. Bactria was where Greek manpower came from. Once it was gone, the Indo-Greek kingdoms could not replace losses.
- They were fighting the wrong direction. The kingdoms in India were oriented east and south, towards each other and towards the Ganges, when the danger was behind them.
- And they did not disappear so much as dissolve. A handful of Greek kings hang on in eastern Punjab for decades afterwards; then the names stop appearing and the population simply becomes part of the north-west.
Trace the chain backwards, because when you lay it out end to end it is genuinely astonishing.
A farmer somewhere near the Jhelum in about 80 BCE pays his taxes to a new official with an unfamiliar
name and an unfamiliar accent. That is his whole experience of the event.
Behind that official is a Saka king who has taken Taxila. Behind the Saka king is the loss of the Bactrian
pastures. Behind that is the arrival of the Yuezhi, pushed west off their own lands. And behind the Yuezhi is
a war they lost, on the Chinese frontier, against a confederation called the Xiongnu — a conflict fought
roughly four thousand kilometres away by people who had never heard of the Punjab and would
not have been able to place it on a map.
The Chinese histories record the start of that chain in detail, because for them it was a border war. They
have no idea what happened at the far end of it.
This is the thing to carry out of this war and into the rest of the book. Punjab’s history is not only made
in Punjab. The five rivers sit at the end of a very long corridor, and pressure applied anywhere along it
eventually arrives here as cavalry.
You will see the mechanism again with the Kushans in two wars’ time, and with the Huns in Part Four, and
with the Mongols in Part Six. Every one of those invasions starts as somebody else’s problem,
thousands of kilometres away, and ends as a change of government on the Jhelum.
12Gondophares and the Parthians LIKELY
The third foreign dynasty in a century — and the one that ended up, improbably, in the Christian Bible stories.
When: Roughly 20–50 CE, with the dynasty lingering in pockets afterwards.
Where: Sistan and Arachosia first, then Gandhara, Taxila and the western Punjab.
Who: Gondophares, against the late Saka rulers of the north-west.
The evidence: Coins, one dated stone inscription, and a Christian apocryphal text written in Syriac.
What happened
This is the shortest of the four conquests and the least consequential, so this part will not stretch it.
1. Parthian nobles from eastern Iran move east. The Parthian empire that ruled Iran had a loose structure
- with powerful regional families, and one of them built a kingdom of its own in Sistan and Arachosia.
- Gondophares takes Gandhara and Taxila from the Sakas, some time in the first quarter of the first century CE, and rules from there.
- We can date him, unusually precisely. An inscription from Takht-i-Bahi, in the Peshawar country, is dated to the twenty-sixth year of Gondophares and to a year of a longer era, and this pair of numbers is one of the anchor points for the whole chronology of the region.
- The kingdom does not survive him by much. Within a generation the Kushans are coming down from Bactria, and the Indo-Parthian territories are absorbed.
Gondophares matters for this book for three reasons that have nothing to do with the size of his kingdom.
First, he shows how fast the north-west changed hands — Greek, Saka, Parthian and then Kushan inside about a
hundred and fifty years, all ruling from the same cities and using the same mints.
Second, his dated inscription is one of the few fixed points in a period where almost every date is an estimate.
Third, and most unexpectedly, his name travelled further than his kingdom ever did.
- The Saka kingdoms had fragmented into competing satrapies within two generations of conquering the place.
- They came from the near side of the mountains, from Sistan and Kandahar — a short campaign into a divided country.
- They inherited rather than rebuilt. Same cities, same coinage system, same administrative titles. The satrap system the Sakas introduced carried straight over.
- It was one man’s achievement. The kingdom is essentially Gondophares, and it comes apart into local rulers after him.
- It faced the wrong way. Its centre of gravity was in Afghanistan, while the danger was gathering in Bactria to the north.
- The Kushans were a different order of thing. A unified confederation with the manpower of the whole Yuezhi migration behind it.
There is a Christian text called the Acts of Thomas, written in Syriac probably in the third century CE, and
its opening is set in India.
In it, the apostles divide up the world and India falls to Thomas, who does not want to go. He is effectively
sold as a slave to a merchant acting for an Indian king, and taken east. The king’s name, in the Syriac text, is
Gundaphar.
The king wants a palace built and gives Thomas the money for it. Thomas, being an apostle rather than an
architect, gives the money away to the poor over a period of months, and reports at intervals that the work
is progressing well.
Eventually the king comes to inspect his palace and finds an empty field. Thomas tells him that the palace
has been built, but that he will not be able to see it in this life — that it is waiting for him in heaven,
paid for with what he gave away.
The king imprisons him. Then, in the story, the king’s brother dies, is shown the palace in the next world,
and comes back to report it — and Gundaphar is converted.
Now the reason this is in a book about the wars of Punjab. For centuries, Gundaphar was assumed to be
pure invention, a made-up name in a made-up story. Then, in the nineteenth century, coins began turning
up in the Punjab and Afghanistan bearing the name Gondophares, and an inscription was found at Takht-
i-Bahi dated in his reign.
The king in the legend was a real king, ruling in the right region at roughly the right decade.
That does not make the story true. Nobody can show that Thomas went to India, and the palace is a
parable that exists in other forms elsewhere. What it shows is something else, and it is the theme of this
entire part: first-century Punjab was connected to the wider world closely enough that its king’s
name was circulating in Syriac in the Middle East, attached to a story that people wanted to tell.
Taxila in this century traded with Rome, took Chinese silk, used Greek script, prayed in Persian-arched
Buddhist shrines and was ruled by an Iranian dynasty. Somebody, somewhere along that road, heard the
name of the king of the Indians. It stuck.
13The Kushans Take Everything CERTAIN
The fourth and largest conquest — and the one that turned Punjab and Gandhara into the crossroads of the ancient world.
When: Roughly 30–150 CE, from Kujula Kadphises to Kanishka.
Where: From Bactria and Kabul across Gandhara and the Punjab and deep into the Ganges plain.
Who: Kujula Kadphises, then Vima Kadphises, then Kanishka — the greatest of them.
The evidence: A vast coinage, the Chinese Hou Hanshu, and one inscription found in 1993 that rewrote the family tree.
Who they were
They were the Yuezhi — the people whose westward migration in War Ten had knocked the Sakas into India in
the first place. Having taken Bactria, they settled there and split into five chiefdoms.
The Chinese Hou Hanshu tells the next part with admirable brevity. The prince of the Guishuang — that is, the
Kushan section — named Qiujiuque, attacked and destroyed the other four chiefs and made himself
king. Qiujiuque is Kujula Kadphises. The same passage says he then invaded Parthian territory, took the Kabul
region, and died at over eighty years old.
That is one paragraph in a Chinese court history covering the birth of an empire that would rule from Central Asia
to Bihar.
What happened
1. Kujula unifies the Yuezhi and takes Kabul, pushing the Indo-Parthians out of Gandhara.
2. His successors take the Punjab and push east. Vima Kadphises introduces a substantial gold coinage —
- the first large-scale gold currency India had seen, minted from the enormous flow of Roman gold coming in through the sea trade. 3. Kanishka, from about 127 CE, is the peak. He rules from Purushapura — Peshawar, with a second capital at Mathura, and his reach into India is far greater than any predecessor’s in this part.
- And we know that because of the Rabatak inscription. Found in Afghanistan in 1993, written in Bactrian in Greek script, it lists the cities and regions that submitted to Kanishka — including Saketa, Kausambi, Pataliputra and Champa. It also gives the Kushan royal genealogy in order, which settled arguments that scholars had been having for a century.
- Punjab is at the centre of it, not the edge. This is the crucial point. Under the Kushans, the Peshawar valley and the Punjab plain are not a frontier province of a Ganges empire or a distant outpost of an Iranian one. They are the imperial heartland.
The Kushan empire sat astride the Silk Road at the point where four worlds met, and it taxed the traffic. The
consequences show up everywhere in the archaeology of the north-west.
Roman gold coins arrive in quantity and are melted down for Kushan issues. Chinese silk passes through. Buddhist
monasteries multiply across Gandhara on an enormous scale. And in the workshops of Gandhara, sculptors trained in a
Greek tradition of drapery and anatomy begin carving Indian religious subjects — including, for the first time, the
Buddha as a human figure rather than as a symbol. Be careful with that last claim, because it is one of the most
argued-over questions in Indian art history: the Kushan-era workshops at Mathura have an equally strong claim to
have got there first, and which school invented the Buddha image is not settled.
Kanishka’s coinage is the clearest single window into the period. On the reverse, in Greek lettering, appear Iranian gods,
Greek gods, Hindu gods — and the Buddha, labelled BODDO. A Central Asian emperor ruling from Peshawar,
minting coins in the Greek alphabet, in the Bactrian language, showing an Indian teacher, for a market that
ran from Rome to China.
- They unified before they invaded. Kujula spent his career destroying his rivals among his own people first. Every other conqueror in this part invaded while still divided.
- They held both sides of the passes. Bactria and Gandhara under one crown, which is the combination Part Five will show decides everything.
- They took the trade instead of just the land. Their revenue came from taxing the traffic between empires, which is far more reliable than farming.
- They sponsored everybody. Buddhist, Iranian, Greek and Hindu cults all appear on their coins. A ruling house that endorses every god in its territory does not have a religious opposition.
- The north-west had been changing hands every forty years. No local power had had time to build anything.
- The Indo-Parthian kingdom was one man deep, and he was dead.
- The Ganges kingdoms were divided, as they had been since the Shungas faded.
- And the Kushans were not raiding. They were a state — dated inscriptions, a gold currency, cities, a religious policy. Nothing in India at that moment was in the same category.
Kanishka built a stupa at his capital, and for centuries afterwards travellers who saw it could not stop
talking about it.
Chinese pilgrims came through Peshawar over a period of five hundred years and each of them described it,
and they do not agree about the height, because the thing burned down and was rebuilt more than once.
The figures they give run into the hundreds of feet, and they describe a great tower on a stepped base
topped with tiers of gilded copper discs. Whatever the true dimensions, it was reckoned by people
who had travelled from China across the whole of Asia to be the most impressive building they
had seen anywhere on the journey.
It was excavated at the start of the twentieth century, at Shah-ji-ki-Dheri outside Peshawar. The base alone
is enormous. And in the ruins the excavators found a metal casket with an inscription naming Kanishka,
containing bone fragments — presented as relics of the Buddha, and later sent to Burma.
Treat the identification of the relics with the caution it deserves. Treat the rest as solid.
Here is the point. In the second century CE, the tallest and most famous building known to the
travellers of Asia stood at Peshawar, and it was paid for by a Central Asian emperor as a
monument to an Indian religion.
Nothing about that sentence would have seemed odd at the time. That is what the four-hundred-year habit
does when it is allowed to finish. The invader of 30 CE is, by 130 CE, the largest religious patron in India and
the man Buddhist tradition would remember alongside Ashoka.
And two hundred years later, in Part Four, the whole of it — the stupa, the monasteries, the workshops, the
city — is burned by the next people to come down the same road.
1490 CE — The War Over a Marriage Proposal LIKELY
Seventy thousand cavalry were sent across the highest mountains on earth because of a diplomatic insult, and they were beaten without a battle by a man who cut a supply line.
Ban Chao of Han China — with a fraction of the men and no pitched battle at WON BY
all
- 90 CE.
- The Tarim Basin, in what is now Xinjiang — reached from Gandhara across the Pamirs.
- The Kushan viceroy Xie against Ban Chao, the Han protector-general of the Western Regions.
- The Hou Hanshu, the official history of the Later Han. A single detailed account, from one side.
Why it happened
The Kushans had been useful allies. They had helped Han forces in Central Asia, and on that basis the Kushan king
made a request that in his own world was entirely normal between allied powers: he asked for a Han princess
in marriage.
Ban Chao refused, and detained the envoy who brought the request. In Chinese court terms, marrying a princess to
a barbarian ruler on demand was a humiliation and a precedent. In Kushan terms, the refusal — with the envoy
held — was a public insult from a partner.
The greatest empire in India and the greatest empire in East Asia went to war over a wedding that did
not happen.
What happened
1. The Kushans send an army over the Pamirs. The Hou Hanshu gives the figure as seventy thousand cavalry
- under the viceroy Xie. Even discounted heavily, this is one of the most difficult marches ever attempted — the passes are among the highest used by any army in history, and there is nothing to eat on the way.
- Ban Chao refuses to fight. Badly outnumbered, he ordered his troops to fortify, gather in supplies, and stay behind their defences. He had grasped the one fact that decided everything: an army that has crossed
- the Pamirs cannot feed itself where it lands.
- The Kushan army starves. Its supplies run down. It cannot storm the fortifications and it cannot forage a country that has been stripped.
- Xie sends for help. He despatches messengers to the kingdom of Kucha, asking for supplies and support.
- Ban Chao is waiting for them. He had anticipated the move, placed men on the route, and ambushed and killed the messengers.
- And then he makes his point. He sent the evidence back to the Kushan commander so that Xie would know his message had been intercepted and no help was coming.
- The Kushans ask for terms. Xie sent envoys to apologise and requested permission to withdraw. Ban Chao let the army go, and after this the Kushans resumed sending tribute missions to the Han court.
Seventy thousand cavalry across the Pamirs is not a plausible logistical proposition and should be read as “a
very large army” written by the winning side. Horses need fodder daily and there is none at that altitude.
And the account comes entirely from the Chinese record, written to celebrate one of the Han dynasty’s most admired
officers. There is no Kushan version. No Indian source mentions this war at all.
What supports it: the Hou Hanshu is a serious administrative history, not a legend collection, and it is precise about the
diplomatic sequence in a way that fabrications usually are not. The tribute missions afterwards are recorded too. The
war happened, the Kushans lost it without a battle, and the number is advertising.
- He refused the battle he was supposed to fight. Outnumbered many times over, he made the engagement about supply instead of about combat.
- He understood the enemy’s real problem. The mountains had already done most of the work before the Kushans arrived.
- He guarded the communications, not the front. The whole campaign turned on one ambush on a road.
- He let them go. A destroyed army makes a permanent enemy; a released one makes a tributary. He got the second.
- They fought for prestige, not for anything they needed. The war had no objective that could be captured.
- They crossed the worst terrain on the planet to reach the fight. Their army arrived already half beaten by the road.
- Cavalry cannot besiege. Faced with men behind walls who would not come out, the greatest mobile army in Asia had nothing to do.
- They had one line of communication and did not protect it. The single most avoidable mistake in the whole of this part.
In 90 CE, an army of the empire ruled from Gandhara and the Punjab fought an army of Han China. It is
one of the very few direct military contacts between the Indian and Chinese worlds in the whole of ancient
history, and almost nobody in either country has ever heard of it.
Think about what had to be true for it to happen. A Central Asian dynasty ruling from Peshawar had to be
powerful enough to consider itself an equal of the Han. Chinese officials in the Tarim Basin had to be close
enough to Gandhara to have an opinion about it. And both governments had to be running a diplomacy
sophisticated enough for a marriage refusal to count as a national insult.
Punjab in the first century CE was not on the edge of anything. It was a great power, in
correspondence with the other great powers, and it went to war with China the way France might
go to war with Spain.
And here is the part that ought to stay with you. This is the highest point the north-west reaches in this
entire book — the moment when the land of the five rivers is the centre of an empire that trades with
Rome, argues with China, mints gold, and produces the sculpture that will define Buddhist art for two
thousand years.
The defeat in 90 CE was not fatal. Kanishka’s reign, the peak of Kushan power, comes a generation after it.
But it is the first crack, and the next war is the second one.
From here, this part goes downhill, and the people who push it downhill are Punjabis.
15The Republics Throw the Kushans Out CERTAIN
After four hundred years of being conquered by everybody, the farmers of the plain organised themselves and removed the greatest empire in Asia.
When: Building through the second century CE, decisive in the third, complete by about 250 CE.
Where: The whole plain — Rohtak, the Sutlej–Yamuna country, Bahawalpur, the hills and the doabs.
Who: The Yaudheyas, Arjunayanas, Kunindas, Malavas and Audumbaras against the Kushan empire.
The evidence: Their own coins and seals — and, decisively, an inscription put up by an enemy.
Who these people were
They were the gana-sanghas — republics, run by assemblies of clan heads rather than kings, electing a war-leader
when they needed one. You met them in Part Two, when the Malloi and Kshudrakas nearly killed Alexander. You
met them again in War One of this part, supplying the manpower that made Chandragupta’s empire possible.
They had been there the entire time. Under the Mauryas, the Greeks, the Sakas, the Parthians and the
Kushans, the same communities kept farming the same land, kept their assemblies, and kept an armed population.
Empires came and went above them.
And in the second and third centuries CE, they stopped putting up with it.
The inscription that proves it
The single best piece of evidence comes from an enemy, which is what makes it so good.
Around 150 CE, the Saka ruler Rudradaman — a great satrap in Gujarat, a descendant of the Saka conquerors of
War Ten — had a long inscription carved on a rock at Junagadh. It is a boast, as these things are. And in the
middle of the boast, he claims to have destroyed the Yaudheyas by force, and explains why that was worth
mentioning: because they were unwilling to submit, and were made proud by having manifested their title
of heroes among all Kshatriyas.
Read that again, because it is unusual. This is a foreign ruler, on his own victory monument, pausing to
explain that the people he beat were famous for refusing to submit and were rightly proud of it. You do
not write that about a minor tribe. You write it to make your own victory sound larger.
And he did not destroy them, because their coins carry on afterwards for another two hundred years.
What happened to the Kushans
1. The empire was squeezed from the west. In Iran, the Sasanian dynasty overthrew the Parthians in 224 CE
- and pushed east almost immediately, taking Bactria and eventually Gandhara. Part Four opens with that war.
- Which cut the Indian half off from the Central Asian half. Kushan power depended on holding both ends of the trade route. Once the western end was gone, the eastern end was a rich province with no reinforcements.
- And the republics rose in the middle of it. Yaudheya coinage of this period is abundant and confident, and it carries a legend that is a political statement: Yaudheya-ganasya jaya — victory to the Yaudheya nation. A seal from Bijaygarh calls them the Yaudheyas who hold the spell of victory.
- Others rose with them. The Arjunayanas and the Malavas struck coins announcing their own victory in the same grammatical form. The Kunindas minted in the hills. Across the whole plain, communities were putting the word for victory on their money at the same time.
- By about 250 CE the Kushans are gone from the Punjab, and a rump dynasty lingers around Mathura, minting worse and worse gold, until it disappears in the following century.
- They waited for the right moment. Not in Kanishka’s reign. They moved when Persia had cut the empire in half.
- Every man was already a soldier. There was no army to raise. A republic of armed farmers does not need to mobilise; it only needs to decide.
- They could not be decapitated. There was no king to kill and no capital to take. This is the exact opposite of every other polity in this part, all of which died when one man did.
- They had outlasted five empires already. Persistence is a strategy, and it had been working for six hundred years.
- Two fronts, one army. Sasanian Persia in the west and rebellion in the east at the same time.
- Their empire was a trade route. Cut it in Bactria and the whole revenue model collapses.
- Their capital faced the wrong way. Peshawar is superb for an empire that spans the passes and fatal for one that has lost the far side.
- And nobody came to help them. They had become an Indian dynasty, and it did not save them — which is the last note this part will sound.
Four hundred years, and look at what has passed over this ground.
Macedonian garrisons. A Mauryan emperor with a bureaucracy and inspection tours. A Greek king in an
elephant-skin hat. Another Greek king arguing about chariots with a monk in Sialkot. Saka satraps. An
Iranian dynasty whose king’s name reached Syriac scripture. A Kushan emperor who put the Buddha on
gold coins and fought China.
Every one of them ruled from a capital and issued money with a face on it — his face, his title, his claim.
That is what a coin was for.
And then, at the end of this part, the coins in the soil of Rohtak and Ludhiana and Bahawalpur start saying
something different. No portrait that matters. No king. Six words of Brahmi: Yaudheya-ganasya jaya.
Victory to the Yaudheya nation. Not to a man. To the assembly, to the gana, to the people who voted.
We do not know the name of a single Yaudheya leader. Not one. There is no Yaudheya chronicle, no epic, no
capital we can point to with confidence, no monument. They were beaten by Rudradaman and carried on;
they were ruled by Kushans and carried on; and when the moment came, they threw out the largest empire
in Asia and went back to farming.
They are, on the evidence, one of the longest-lived political communities in Indian history — and their
entire surviving voice is one word, stamped in copper, hundreds of thousands of times, and dropped in the
dirt.
Every empire in this part wrote its name on something expensive. The people who outlasted all of
them wrote theirs on the cheapest object they had, and it is the one you can still pick up in a
ploughed field.
16The Whole Chapter on One Page
Five patterns to carry into Part Four
1. Punjab digests its conquerors. This is the defining pattern of Part Three and it does not survive Part Four,
- which is why it matters. Greeks, Sakas, Parthians and Kushans all arrive as foreigners and all become Indian within two or three generations — Indian scripts on their coins, Indian gods on the reverse, Indian titles, Indian religious patronage. The question to carry forward is why this stops working.
- Punjab’s wars are started somewhere else. A defeat on the Chinese frontier pushes the Yuezhi west, which pushes the Sakas south, which changes who collects tax on the Jhelum. Nothing in that chain has anything to do with Punjab. You will watch the identical mechanism produce the Kidarites and the Huns in Part Four, and the Mongols in Part Six.
- Whoever holds both sides of the passes holds everything. The Kushans are the great success of this part and they are the only power in it that ruled Bactria and Gandhara together. The moment Persia takes the far side, the Indian half falls within a generation. This is the single most reliable rule in the entire book and Part Five is built on it.
- Kill the king, end the state. The Nandas, the Mauryas, Eucratides’ kingdom, Gondophares’ kingdom — every monarchy in this part dies with one man. The one polity that cannot be decapitated is the republic, because there is no head. The Yaudheyas outlast every empire in these four hundred years by having nobody important in them.
- The winner is whoever gets written about. Thirty-odd Greek kings ruled here and left nothing but their faces on silver. Menander survives because a monk wrote down an argument. The Yaudheyas survive because they stamped one word on a great deal of copper. In Part Four this problem becomes total: not one contemporary Punjabi account survives of any war in that entire chapter.
The Guptas and the Huns, 250–650 CE. Sasanian Persia takes Gandhara and finishes the Kushans. For a hundred
years the republics of the plain govern themselves — the longest stretch of self-rule Punjab gets between Alexander and
Ranjit Singh — until Samudragupta arrives at the frontier and takes their submission on paper. Then the passes open
again, and keep opening: the Kidarites, then Toramana and Mihirakula of the Alchon Huns, ruling the Punjab from
Sialkot. Skandagupta stops them once and it breaks his empire. A king nobody remembers breaks them for good at
Sondani. And by the end of it, Taxila — the city that has been in every part of this book so far — is a field of mounds,
and stays one.
17Words Used in This Part
The Indian word for Greeks, from “Ionian”. It later broadens to mean any westerner. When a Sanskrit text says Yavana, ask which century before assuming it means Greek.
A republic — a state run by an assembly of clan heads rather than a hereditary king. The Yaudheyas, Malavas, Arjunayanas and Kunindas were all of this type.
Satrap and great satrap — provincial governor titles brought into India by the Sakas from the Persian system. Rudradaman was a mahakshatrapa.
The horse sacrifice. An ancient royal ritual in which a horse wanders for a year, escorted by an army, and any king who stops it must fight. Pushyamitra revived it.
The manual of statecraft traditionally attributed to Chanakya. Cold, practical and detailed. Its mandala theory is the “eat from the edges” principle written as doctrine.
A Sanskrit play about Chanakya’s political manoeuvring after Chandragupta’s victory. Written centuries later; the source of much of the popular Chanakya legend.
A Buddhist collection of legends about Ashoka, written down centuries after him. The source for the Taxila revolts and the blinding of Kunala. Devotional, not documentary.
The right-to-left script of the north-west, derived from Aramaic. Ashoka’s Gandhara edicts are in it, and so are the Indian sides of Indo-Greek and Saka coins.
The Prakrit form of dharma. In Ashoka’s edicts it means something like right conduct — a public ethic rather than a specific religion.
Ancient Sialkot. Capital of Menander, and later of the Hun king Mihirakula in Part Four.
Taxila. The great city of the north-west — Persian province, Greek prize, Mauryan trouble spot, and the place four dynasties in this part chose to rule from.
The Indo-Greek city at Taxila, laid out on a Hellenistic grid and then lived in by Sakas, Parthians and Kushans in turn.
“The Questions of Milinda.” The Pali dialogue between King Menander and the monk Nagasena. A classic of Buddhist philosophy set in Punjab.
A short apocalyptic Sanskrit text that describes the Yavana advance on Pataliputra as prophecy. Contested as a source, and the only Indian account of that campaign.
Patanjali’s great commentary on Sanskrit grammar, around 150 BCE. Its examples accidentally record the Greek sieges of Saketa and Madhyamika as recent events.
The Central Asian people, displaced from the Chinese frontier, whose migration pushed the Sakas into India — and who later became the Kushans themselves.
The official history of the Later Han dynasty. Our main source for Kujula Kadphises and for the Kushan–Han war of 90 CE.
A Bactrian-language inscription in Greek script, found in Afghanistan in 1993, giving Kanishka’s conquests and the Kushan royal genealogy. It settled arguments a century old.
A coin hammered flat and restruck with a new ruler’s design. Since the old image often shows through, overstrikes let numismatists reconstruct wars nobody wrote down.
Sources used for this part
- Justin, Epitome of Pompeius Trogus, Books XV and XLI — for Chandragupta, Seleucus and Eucratides
- Appian, Syrian Wars; Strabo, Geography, Books XI and XV; Plutarch, Life of Alexander; Orosius — for the Seleucid–Mauryan
- settlement
- Polybius, Book XI, for Antiochus III in Bactria and the young Demetrius
- Megasthenes, Indika, surviving in fragments quoted by later writers
- The rock edicts of Ashoka — Shahbazgarhi and Mansehra in Kharoshthi, Girnar, Dhauli and Jaugada, and the Greek and Aramaic
- inscriptions from Kandahar
- Patanjali, Mahabhashya; the Yuga Purana (John Mitchiner’s translation); the Hathigumpha inscription of Kharavela
- Milindapanha, in the Rhys Davids translation
- The Ashokavadana, in John S. Strong’s translation; Bana, Harshacharita; Kalidasa, Malavikagnimitra
- The Taxila copper plate of Patika; the Takht-i-Bahi inscription of Gondophares; the Rabatak inscription of Kanishka; the Junagadh rock inscription of Rudradaman (Kielhorn’s translation)
- The Hou Hanshu, chapter on the Western Regions, in John Hill’s translation
- The Acts of Thomas, for the Gundaphar legend
- John Marshall, Taxila (1951), for Sirkap and the archaeology of the north-west
- Osmund Bopearachchi on Indo-Greek and Bactrian coinage; Joe Cribb on Kushan and Indo-Greek numismatics
- John D. Grainger on the Seleucid–Mauryan war; Étienne Lamotte, Romila Thapar, H. C. Raychaudhuri and K. P. Jayaswal on
- Pushyamitra and the persecution question
The open questions in this part are unusually numerous. They include: whether Chandragupta’s latrones were bandits
or mercenaries; whether the Seleucid–Mauryan war involved a battle at all; almost every date in the Indo-Greek
sequence; whether there were one or two kings called Menander; whether Menander’s army really reached
Pataliputra; how much of the Ashokavadana preserves anything historical; the scale and reality of Pushyamitra’s
persecution of Buddhists; and the precise chronology of the Kushan kings, which the Rabatak inscription improved but
did not settle.
Two warnings in particular. First, the chronology of the Indo-Greeks is a numismatic reconstruction, not a record. It
is careful and impressive work, and it is still an inference. Any date given here for a Greek king may move by a decade in
either direction, and some of the kings may need to swap places.
Second, and more important: almost nothing in this part was written by anyone who lived in Punjab. The Greek
and Roman writers were working at second and third hand, centuries later, from lost books. The Chinese histories are
excellent and are looking at India from the side. The Buddhist texts are teaching, not reporting. The Sanskrit sources
that mention these events are mostly later, and one of the most important of them is a grammar textbook that recorded
a siege by accident.
What Punjab itself left behind from these four hundred years is coins. Hundreds of thousands of them, honest about
who was in charge and silent about everything else — and, at the very end, a copper coin with no king on it that says
victory to the nation.