Where We Left Off
A short recap for anyone starting here, and a reminder of what the coloured boxes mean.
Four parts have brought us here.
Part 1 built the land. Two continents crashed slowly, the Himalayas went up, rivers ground them into soil kilometres deep. Flat, fertile, watered all year — and sitting directly behind the only easy way into India. Then the first stone tools, and the arrival of farming.
Part 2 was Harappa. Over a thousand settlements, grid-planned cities, standardised bricks and weights, covered sewers. And four things missing: no king anyone can name, no temple anyone can identify, no writing anyone can read, and no horse. The cities emptied around 1900 BCE as the climate dried and the rivers moved.
Part 3 was the Vedic age. The Rigveda composed on this land, in the land of the seven rivers — a cattle-herding, horse-and-chariot, tribal, oral world with no cities at all. The Battle of the Ten Kings fought on the Ravi. And the long argument about where those people came from, which the evidence answers imperfectly and the politics answers loudly. Part 3 ended with the centre of Indian civilisation moving east, away from Punjab, pulled by iron and rice into the Ganga plain.
Part 4 brought Punjab back into recorded history. Persian rule on the eastern edge of an empire run from Iran. Alexander crossing the Indus, fighting Porus on the Jhelum, and turning back at the Beas when his army refused. Then Chandragupta Maurya assembling the largest empire India had yet seen, and his grandson Ashoka carving instructions to his people into rock — the first Indian writing anybody has been able to read since the Harappan seals went silent.
This part begins with that empire falling apart.
What the boxes mean
Same system throughout the series. Skip this if you know it.
Blue explains a hard word, right where it appears.
Example — Dynasty: a line of rulers from one family, passing power down through it. When historians say "the Kushan dynasty" they mean one family and its descendants holding power over a period.
Green turns something too big or too strange to picture into something ordinary.
Brown shows the evidence itself, and is honest about what it can and cannot prove. This part leans on it heavily, because much of what follows is reconstructed from coins.
Orange means serious people disagree. Both sides get their strongest case, then a plain statement of where things stand.
Purple digs out something everyone takes for granted without proving.
Dark boxes close each chapter with the key points in the plainest words I can find.
What makes this part different
Part 3 was an argument. Part 4 was a story with named people in it. This part is something else again.
For the next seven hundred years, almost every ruler of Punjab came from somewhere else. Greeks from Afghanistan. Scythian herders from the steppe. Parthians from Iran. A Central Asian people called the Yuezhi, who became the Kushans. Then, at the end, the Hunas.
That sounds like a grim story of a land repeatedly conquered. It is not what happened, and the difference is the most interesting thing in this part.
What actually happened is that Punjab absorbed them. Within two or three generations, a Greek king is asking a Buddhist monk about the nature of the self and the conversation is written down as a Buddhist scripture. A Central Asian emperor is putting Greek, Iranian and Indian gods on the same coin. And out of that mixture comes the single most influential piece of art ever made in this region — the first image of the Buddha with a human face, which travelled from Punjab to China, Korea and Japan and is still what a billion people picture when they think of him.
This is the part where Punjab is a crossroads rather than a battlefield. Chapter 9 asks what that actually did to the place.
1When a Big Empire Falls Apart
Ashoka's empire was the largest thing India had ever built. Fifty years after he died it was gone. Understanding why is more useful than memorising the dates, because the same thing happens again in Parts 6, 9 and 10.
1.1 — What happened
Ashoka died around 232 BCE. Within about fifty years the Mauryan empire had shrunk to a fraction of itself. Around 185 BCE, the last Mauryan ruler was killed by his own army commander, a man called Pushyamitra Shunga, who took the throne himself.
The Shungas ruled a much smaller territory in the Ganges plain. The north-west — Punjab, Gandhara, the Indus valley — was no longer theirs.
That is the bare fact. The interesting question is why.
1.2 — Why big empires come apart
This is worth spending a page on, because it is a pattern rather than an accident.
An ancient empire is not held together by nationality or by shared identity. Nobody in Punjab in 250 BCE thought of themselves as Mauryan. It is held together by four things, and if any of them fails the whole structure is in trouble.
One: the ability to move an army. If a province rebels, the centre must be able to get soldiers there before the rebellion sets. With no railways and no telegraph, that means weeks or months. The further away the province, the weaker this gets.
Two: the ability to collect tax. Armies eat. Officials must be paid. Tax collection needs officials who are loyal, literate and supervised, in every district. This is enormously difficult and expensive.
Three: local elites deciding it is worth going along with. No empire has enough soldiers to force every village. It rules through local men who calculate that cooperating pays better than resisting. Change that calculation and the empire evaporates without a battle.
Four: a clear rule about who rules next. This is the one that kills most empires. If the succession is disputed, the army splits, the provinces pick sides, and the empire spends its strength on itself.
Think of an empire as a business with branches across a continent, no phones, no email, and a three-month round trip to the furthest office.
It works only while the branch managers believe head office is strong, is watching, and will reward them. The moment they suspect otherwise — a weak new owner, a succession row, an army that lost a battle — they start behaving as owners of their own branch.
Nobody needs to rebel. They just stop sending the money and see what happens. When nothing happens, everyone else notices.
Ashoka's empire was probably too big for the technology of the time in the first place. And Punjab and the north-west were both the furthest from the capital at Pataliputra and the closest to the passes through which outsiders arrived. When head office weakened, this was always going to be the first region to go.
1.3 — What "the dark age" is really describing
You will sometimes see the period after the Mauryas described as a dark age or a period of decline and chaos. Be careful with that phrase.
Here is one that shapes how the whole of Indian history is usually taught.
The assumption is that a large unified empire is the normal, healthy state of affairs, and everything else is decline.
Watch how the standard story is arranged. Mauryas — good, big empire. Then a gap — dark age, foreign invasions, fragmentation. Then Guptas — golden age, big empire again. Then another gap. Then Delhi Sultanate, Mughals, British.
The high points are always the biggest empires. But who does a big empire benefit? Chiefly the people at the centre — the court, the capital, the tax-collecting class, and the historians they employ.
For an ordinary farmer in Punjab, a large distant empire mostly means somebody further away is taking your grain. There is little evidence that people in the so-called dark ages were poorer, sicker or less happy. Some of the periods labelled as decline were extremely productive — this one produced the finest art the region ever made.
Why the bias exists is simple: big empires generate records. Courts keep archives, kings put up inscriptions, capitals leave monuments. Fragmented periods generate far fewer written sources, so historians see less, and "we can see less" quietly becomes "less was happening".
That is Part 1's survivorship bias again — this time not in what survived, but in what was written down at all.
Test to carry through the rest of this series: when you meet the phrase "golden age", ask golden for whom? Chapter 7 applies exactly this test to the Guptas.
1.4 — What Punjab looked like in the gap
Not chaos. Something more interesting.
When the Mauryan structure withdrew, what was left was not emptiness. It was the older layer that had always been underneath — cities with their own governments, and tribal republics.
Punjab and its surroundings had a long tradition of these. Groups such as the Yaudheyas, the Malavas, the Arjunayanas and the Madras were not kingdoms in the usual sense. They were confederations, run by assemblies of leading men rather than by a single hereditary king. Several of them issued their own coins, which is how we know they existed and roughly where they were.
They survive a very long time. They are still there, and still being mentioned, seven hundred years later in the Gupta period — which is Chapter 7.
Gana-sangha: a Sanskrit term for a state run by an assembly rather than a single king. Often translated "republic", though that word carries modern baggage — these were not democracies. The assembly was made up of heads of leading families, not of everyone.
Numismatics: the study of coins. For long stretches of this part it is the main evidence we have, so the word is worth knowing.
Ashoka died around 232 BCE. By about 185 BCE the Mauryan empire was gone, ended by a coup.
Empires need four things: the ability to move an army, to collect tax, to keep local elites willing, and a clear rule for who rules next. Lose one and the rest unravel.
Punjab was the furthest from the capital and the nearest to the passes, so it went first.
What replaced the empire was not chaos but self-governing cities and tribal republics — the Yaudheyas, Malavas and others — who issued their own coins and lasted centuries.
And be suspicious of "dark age". Periods without big empires produce fewer records, not necessarily less life. Always ask: golden for whom?
2The Greeks Who Stayed
Alexander left and his army went home. But not all of them. Two centuries later a Greek king was ruling Punjab from Sialkot, and a Buddhist scripture records him asking a monk whether the self exists.
2.1 — Where they came from
When Alexander died in 323 BCE his generals carved up his conquests. The eastern share went to Seleucus, whose empire ran from Turkey to Central Asia. Part 4 covered what happened next: Chandragupta Maurya pushed him out of India, and Seleucus settled for a treaty, some territory and, reportedly, five hundred war elephants.
But north of the Hindu Kush, in what is now northern Afghanistan and southern Uzbekistan, there was a province called Bactria — settled with Greek soldiers and colonists, with Greek cities, Greek gymnasia and Greek coinage, sitting on very rich farmland.
Around 250 BCE its governor, Diodotus, stopped taking orders from the Seleucids and declared himself king. Bactria became an independent Greek kingdom, two thousand kilometres from Greece, with a Central Asian population and a Greek ruling class.
And when the Mauryan empire weakened, that kingdom looked south.
2.2 — Into India
Around 180 BCE a Bactrian Greek king named Demetrius crossed the mountains into the Indus valley. Over the following decades Greek rulers held territory across Gandhara, Punjab and at times further east.
Historians call them the Indo-Greeks, to distinguish them from the Bactrian Greeks north of the mountains. They are, on any measure, one of the strangest political entities in world history — a Greek-speaking dynasty ruling Indian territory, with Greek gods on one side of their coins and Indian writing on the other.
2.3 — Menander
The greatest of them, and the one who matters most for Punjab, was Menander I, who ruled roughly 165 to 130 BCE.
His capital was Sagala. Sagala is modern Sialkot, in Punjab, Pakistan — about a hundred kilometres north of Lahore.
His coins are found over an enormous area, which suggests a long reign and a wide territory. Greek and Roman writers knew of him. And then there is the extraordinary thing.
2.4 — The king in a Buddhist scripture
There is a Buddhist text, preserved in Pali, called the Milindapanha — "The Questions of Milinda". Milinda is Menander.
It is a long dialogue between the king and a Buddhist monk named Nagasena. The king asks hard questions; the monk answers with arguments and analogies. They discuss what a person actually is, whether there is a permanent self, how rebirth can happen without a soul that carries over, and what makes an action good.
The most famous passage is about a chariot. Nagasena asks the king what a chariot is. Is it the wheels? The axle? The frame? The pole? None of these alone is the chariot, and yet remove them all and nothing is left. So "chariot" is a word applied to an arrangement of parts, not a thing existing on its own — and, the monk argues, a person is the same.
Stop and look at what this is.
A Greek-descended king, ruling from a city in Punjab, sitting down with an Indian Buddhist monk and arguing philosophy — and the argument being thought valuable enough that Buddhists preserved it as scripture and are still reading it two thousand years later.
This is not a conqueror imposing a culture. It is not a conquered people submitting to one. It is a conversation.
And notice the chariot. Part 3 spent a chapter on the chariot as the weapon that changed warfare. Here, three hundred years later, the same object turns up as a tool for thinking about the nature of the self.
The text says at the end that Menander was persuaded and became a Buddhist. That may be devotional framing rather than fact — texts like this often end with the important person converting. But something in it is real: Buddhists chose a Greek king as the questioner, which means his interest was plausible to their audience.
And there is a hint from another direction. Some of Menander's coins carry the eight-spoked wheel, a Buddhist symbol. A Greek and Roman source reports that when he died, his ashes were divided between cities that each built a monument over their share — which is precisely what Buddhist tradition describes being done with the Buddha's own remains.
2.5 — Reading history off coins
Now the awkward part, and it applies to the next two centuries of this story.
We know about the Indo-Greeks almost entirely from their coins. Something like thirty or forty kings are known. For most of them, the coins are all there is. No chronicle. No inscription. No tomb. Just money.
Coins are far better historical evidence than they sound, and it is worth understanding why.
They carry a name and a portrait. Indo-Greek coins name the king and often show his face — sometimes as a young man, sometimes older, which lets you order the issues of a single reign.
They carry titles and claims. "Saviour", "the Just", "the Victorious". A king who suddenly starts calling himself "the Invincible" has probably just won something.
They are bilingual. This is the crucial feature. Indo-Greek coins typically have Greek on one face and a Prakrit translation in the Kharosthi script on the other. That means the same name and title appear in two scripts — which is exactly how Kharosthi was deciphered in the nineteenth century, and it is why we can read these coins at all.
Where they are found tells you where power reached. Plot the find-spots of a king's coins and you get a rough map of his territory and trade.
The metal and weight tell you about the economy. A ruler debasing his silver is usually a ruler under financial strain.
And the style tells you who they were talking to. A Greek deity on the coin speaks to one audience; an Indian one, or a Buddhist wheel, speaks to another. A king who changes the images is changing who he needs to persuade.
Coins are official, dated in relative sequence, produced in large numbers, and hard to fake. They are contemporary — struck by the ruler himself, not written about him a century later by someone with an agenda. For a period with no surviving chronicle, they are as close to a primary state document as we get, and careful analysis of styles, monograms and overstrikes has reconstructed a workable sequence of kings.
A coin tells you a man claimed to be king and had a mint. It tells you nothing about whether he ruled for one year or thirty, whether he controlled a province or a single city, whether he was loved or hated, or what he actually did. The sequence of Indo-Greek kings is repeatedly revised because it rests on stylistic judgement, and different specialists produce different orderings. And coins are propaganda — the ruler chose every element. Reading them as neutral fact is like writing a country's history from its banknotes.
What would improve it: inscriptions and datable excavated layers. Every properly stratified dig at Taxila or in Gandhara that puts a coin in a datable context tightens the whole sequence.
The general lesson, which is worth more than the specific case: when a period is known from only one kind of evidence, the shape of that evidence quietly shapes the history. Known from coins, the Indo-Greeks come out as a list of kings, because that is what coins record. The farmers, weavers and traders who made up almost everyone leave no coins with their names on.
2.6 — Greeks becoming Indian
One piece of evidence deserves its own moment, because it shows the direction of travel.
At Besnagar in central India there is a stone pillar. Its inscription records that it was set up by a man named Heliodorus, who describes himself as a Greek, an ambassador sent from a Greek king to an Indian court.
The pillar is dedicated to Vasudeva — a form of Vishnu — and Heliodorus describes himself as a devotee.
A Greek diplomat, sent by a Greek king, put up a stone pillar in India to honour an Indian god, and had it inscribed in an Indian language.
Not as diplomacy. Not as politeness. He calls himself a follower.
Within roughly a century and a half of Alexander's army going home, the Greeks who stayed were worshipping Indian gods and writing in Indian script.
Hold onto this. It is the pattern for everything in this part — and Chapter 9 is about what it means.
A Greek kingdom survived in Bactria (northern Afghanistan) after Alexander, broke away around 250 BCE, and pushed into India around 180 BCE as the Mauryas weakened.
Menander was the greatest of these Indo-Greek kings, ruling from Sagala — modern Sialkot, in Punjab.
A Buddhist scripture, the Milindapanha, records him debating philosophy with the monk Nagasena — including the famous argument that a person, like a chariot, is a name for an arrangement of parts.
We know these kings almost entirely from coins. That gives us names and a rough map, and almost nothing about ordinary life.
And a Greek ambassador called Heliodorus put up a pillar in India to an Indian god and called himself a devotee. The absorption had already begun.
3The Long Domino Chain
A war on the Chinese frontier pushed one people west. They pushed the next. That one pushed the next. Four dominoes later, Punjab had new rulers — and the chain is one of the clearest demonstrations in history that this region was never isolated.
3.1 — The chain
In the second century BCE, on the northern frontier of China, a confederation of nomadic peoples called the Xiongnu became powerful. They fought their neighbours and won.
Among the losers were a people called the Yuezhi, who were driven west out of their pastures in what is now western China.
The Yuezhi, needing somewhere to go, moved into the lands of the Sakas — the people Greek writers called Scythians, herders of the Central Asian steppe.
The Sakas, displaced in turn, moved south and west. Some went into Iran. Some came down through Afghanistan and into the Indus valley and Punjab.
And the Yuezhi followed them, eventually settling in Bactria — where they ended the Greek kingdom there — and later moving into India themselves as the Kushans, who are Chapter 4.
Line the chain up and look at the distance.
A war on the border of China set off a chain of displacements that changed who ruled Punjab. Four peoples, roughly four thousand kilometres, across perhaps a century.
Nobody involved planned it. No one in the Chinese frontier wars had any idea that Punjab existed, let alone that they were about to change its rulers. Each group simply moved because the group behind them had arrived and there was nowhere else to go.
Part 1 said Punjab sits behind the doors into India. This is what that means in practice. Push hard enough anywhere on the Eurasian steppe, and eventually somebody comes through the Khyber.
3.2 — The Sakas in Punjab
The first Saka ruler we can name in the north-west is Maues, around the early first century BCE, based around Taxila. After him the most prominent is Azes, whose coins are found in enormous numbers across Punjab and Gandhara.
Historians call them Indo-Scythians. Their coinage keeps the Indo-Greek pattern — Greek on one side, Kharosthi on the other — which tells you they inherited a working administration and did not smash it.
Alongside and after them came the Indo-Parthians, from Iran. Their best-known king is Gondophares, ruling around the middle of the first century CE from the Taxila region.
Gondophares has an odd afterlife. In early Christian tradition, the apostle Thomas is said to have travelled to India and worked at the court of a king whose name appears in those texts in a form matching Gondophares.
Whether Thomas came to India is genuinely disputed and is not settled by this. But the detail is telling in another way: a king ruling from Taxila was known by name in the eastern Mediterranean.
That is not an isolated corner of the world. That is somewhere people had heard of.
3.3 — A script from the west
One thing these rulers used, and Punjab used for centuries, deserves a section of its own.
The Kharosthi script was the writing system of the north-west — Gandhara, Punjab and the Indus valley — from roughly the fourth century BCE to the third or fourth century CE.
It descends from Aramaic, the administrative script of the Persian empire, which arrived when Persia ruled this region in the period Part 4 covered. And it kept an Aramaic habit that no other Indian script has: it is written from right to left.
Kharosthi: the script of ancient Gandhara and Punjab. Derived from Aramaic, written right to left. Used for Gandhari, a Prakrit language.
Brahmi: the other ancient Indian script, written left to right, used by Ashoka across most of his empire. Almost every modern Indian script — Devanagari, Gurmukhi, Bengali, Tamil, Telugu — descends from Brahmi.
Prakrit: the everyday spoken languages of ancient India, as opposed to formal Sanskrit. Gandhari was the Prakrit of the north-west.
Ashoka, in Part 4, used both: Brahmi across most of his empire, Kharosthi in the north-west, because that is what people there read.
And then Kharosthi died. It went out of use in India by around the fourth century CE, survived a while longer along the Silk Road in Central Asia, and then vanished completely. It has no descendants. Every modern Indian script comes from Brahmi instead.
Punjab had its own writing system for seven hundred years, and it left no children. That is worth sitting with — it is the second writing system this land has lost, after the Harappan one.
3.4 — A calendar that never stopped
One thing from this period is still running.
An era of year-counting began in 78 CE and became known as the Saka era. Its origin is disputed — it is traditionally attributed to Kanishka, though the dating of his reign now sits somewhat later, and other explanations are offered.
What is not disputed is what happened to it. The Saka era spread across India, was used for dating inscriptions and manuscripts for the next two thousand years, and in 1957 was adopted as the official national calendar of the Republic of India.
The Government of India Gazette is dated in an era that started in 78 CE, in the north-west, in a period when Punjab was ruled by people whose ancestors had been pushed off the steppe by a war on the Chinese border.
Almost nobody using that date knows this. It is one of the longest unbroken threads in this entire series — a number that started counting during the events of this chapter and has not stopped since.
Here is one that quietly distorts how this whole period is taught.
The assumption is that peoples move because they want to conquer. The standard phrasing gives it away — waves of invaders, successive conquests, foreign incursions. It makes movement sound like ambition.
Look at the actual chain in this chapter. The Yuezhi did not set out to conquer anything. They lost a war and were driven out. The Sakas moved because the Yuezhi arrived in their pastures. Each group was pushed, not pulled.
Pastoral herders move because their grazing fails, because the climate turns, or because somebody stronger takes their land. Conquest is often the consequence of migration, not its purpose — you arrive somewhere with your family and your animals, and you fight because you now need this land and somebody else is on it.
Why this matters: "invader" implies a choice and a motive. It builds a story with villains. "Displaced people who arrived armed" describes the same event and produces a very different feeling about it.
Both descriptions can be accurate about the violence that resulted. But one of them tells you what caused it.
A war on the Chinese frontier pushed the Yuezhi west. They pushed the Sakas. The Sakas came down into Punjab. The Yuezhi followed and became the Kushans.
Four peoples, four thousand kilometres, and nobody planned any of it.
The Sakas and Parthians kept the existing system running — same bilingual coinage, same administration. This was not destruction.
Punjab's script was Kharosthi — descended from Aramaic, written right to left. It died out completely and left no descendants. Every modern Indian script comes from Brahmi instead.
And the calendar that started in 78 CE is still the official national calendar of India today.
4The Kushans
A herding people driven off the Chinese frontier built an empire that stretched from Central Asia to the Ganges, sat astride the Silk Road, put every god they had ever met on the same coinage, and sent Buddhism to China.
4.1 — From refugees to emperors
The Yuezhi arrived in Bactria as a displaced people in five separate groups. In the first century CE, a leader of one of those groups — Kujula Kadphises — brought the five under one authority and started expanding south.
His successors pushed into Gandhara, Punjab and down into the Ganges plain. By the early second century CE the Kushan empire ran from the edge of Central Asia to Mathura and beyond.
Their principal cities were Purushapura — modern Peshawar — and Mathura, with Bactrian centres to the north. Punjab sat squarely in the middle of it, on the main road between the two.
4.2 — Kanishka
The most famous Kushan ruler is Kanishka, whose reign began around 127 CE.
Under him the empire reached its greatest extent and its greatest wealth. He is remembered in Buddhist tradition as a great patron — the ruler who called a major council, built enormous monuments, and helped Buddhism spread outwards.
For a long time the Kushan king-list was a mess, assembled from coins and later legends, with dates argued over by a century or more.
Then in 1993, at a site in northern Afghanistan called Rabatak, an inscribed stone slab was found. It is written in the Bactrian language — an Iranian language — using the Greek alphabet, which itself tells you a great deal about this world.
The text names Kanishka, lists the kings before him in order, and describes the extent of his territory.
Why this mattered so much: it settled the royal genealogy, which coins alone could not do. Coins tell you a man was king; they do not tell you whose son he was or who came before. One inscription did in a paragraph what a century of numismatics had struggled with.
And it is a reminder about this whole part. The evidence is thin enough that a single new find can rearrange the chapter. Some of what is written here rests on material discovered within living memory, and more may yet turn up.
4.3 — The Silk Road
The Kushan empire's real significance is where it sat.
The Silk Road is not a road. It is a name for the network of routes carrying goods between China, Central Asia, India, Iran and the Mediterranean. Nobody travelled the whole length; goods were handed on from trader to trader, changing hands and gaining price at every stage.
The Kushans controlled the junction where the routes from China, the routes from Iran and the routes into India all met. Everything moving between those worlds passed through their territory, and they taxed it.
Imagine owning the only bridge on a road that everybody has to use.
You do not need to produce anything. You do not need to conquer anybody further. You simply sit on the crossing and take a share of everything that passes.
That is what the Kushans had, and it is why they were rich enough to issue large quantities of gold coinage — the first Indian dynasty to do so on a serious scale.
And Punjab is on that route. Goods coming through the passes from Central Asia went down through Gandhara and across Punjab towards the Ganges plain and the ports. Every caravan crossed the five rivers.
What moved: silk and lacquer from China. Spices, cotton, ivory and gemstones from India. Horses from Central Asia. Glass, wine and coin-silver from the Roman world. And travelling with the goods, as always, ideas.
4.4 — Every god on one coinage
The single most revealing thing the Kushans left is their coins, and specifically which gods are on them.
Across Kushan coinage you find deities from at least four traditions:
- Greek — the sun and moon, Herakles, and others inherited from the Bactrian Greek tradition.
- Iranian — a large group, including a sun god and a widely worshipped goddess of the region.
- Indian — a figure usually identified with an early form of Shiva, shown with attributes that later become standard.
- Buddhist — and this is the famous one. Some of Kanishka's coins show a standing Buddha, with his name written beside him in Greek letters.
Picture that coin. An emperor descended from Central Asian herders, ruling from Peshawar, issues money showing the Buddha, labelled in the Greek alphabet, in an empire where the administrative language is Iranian.
Four civilisations on one small disc of metal.
This is what a genuine crossroads looks like. Not a place where cultures politely coexist in separate boxes, but a place where they are melted together on the currency people used to buy bread.
Buddhist tradition is emphatic. He is remembered as a great patron who convened a major council, built a huge relic monument at Peshawar that later Chinese pilgrims described with awe, and supported monasteries across his empire. He put the Buddha on his coinage — a serious public statement. Buddhism expanded dramatically during and after his reign, and it is hard to explain that without royal backing.
The Buddha coins are a small minority of Kushan issues. The great majority show Iranian deities, and the Rabatak inscription itself opens by invoking Iranian gods. If Kanishka had converted in the sense the tradition implies, the coinage would look very different. And the accounts of his patronage come from Buddhist sources written centuries later, which had every reason to claim a great emperor as one of their own — exactly the pattern seen with Menander in Chapter 2.
Why this is worth understanding: the question assumes a person belongs to exactly one religion, which is a much later way of thinking. For a Kushan emperor ruling Greeks, Iranians, Buddhists and Hindus, honouring all of their gods was not confusion or hedging. It was policy — and probably also sincere.
Watch for this assumption elsewhere in the series. It causes constant misreadings of pre-modern rulers.
4.5 — How Buddhism reached China
This is the Kushans' largest effect on world history, and it happened through Punjab and Gandhara.
The mechanism was simple and slow. Traders moved along the Silk Road. Monks travelled with the caravans, because caravans were the only safe way to travel. Monasteries were built along the routes, partly as religious institutions and partly as the ancient equivalent of roadside inns, offering shelter to travellers.
Each monastery was a node. Monks taught, texts were copied, local people converted, and the next monastery was founded further along the road.
Following that chain out of Gandhara and across Central Asia, Buddhism reached China in the first centuries CE. From China it went to Korea, and from Korea to Japan.
There are today hundreds of millions of Buddhists in East Asia.
The road their religion travelled to get there ran out of the Ganges plain, across Punjab, through Gandhara, over the passes, and along the caravan routes of Central Asia.
And it travelled with merchants, in the space between one trading post and the next, because that was the only way anything moved.
This region was not a backwater that outsiders occasionally invaded. For several centuries it was one of the most consequential places on Earth.
The Kushans were the Yuezhi from Chapter 3, settled and organised. Their empire ran from Central Asia to the Ganges, with capitals at Peshawar and Mathura and Punjab in the middle.
Kanishka, from around 127 CE, was the greatest of them. The Rabatak inscription, found in 1993, settled the royal line that coins alone could not.
They controlled the junction of the Silk Road and grew rich taxing it — rich enough to strike large amounts of gold.
Their coins carry Greek, Iranian, Indian and Buddhist gods — including the Buddha, labelled in Greek letters. Kanishka was probably a patron of everything rather than a convert to one thing.
And Buddhism travelled from here, with merchants, up the Silk Road to China, Korea and Japan.
5Gandhara: The Face of the Buddha
For five hundred years after his death, nobody made an image of the Buddha as a man. Then, in this region, somebody did — and gave him wavy hair, a heavy robe with deep folds, and the features of a Mediterranean god. That image conquered half the world.
5.1 — The centuries with no Buddha
The Buddha died around the fifth or fourth century BCE. For roughly the next five hundred years, Buddhist art did not show him.
This is not a gap in the record. There is plenty of early Buddhist art — carved railings and gateways at great monuments, crowded with figures, animals, trees and scenes from the Buddha's life.
He is in those scenes. But he is not shown.
Instead, artists marked his presence with a symbol:
- A pair of footprints, where he stood.
- An empty throne, where he sat.
- A wheel, for his teaching.
- A tree, for the moment of his awakening.
- A parasol held over an empty space, marking where an important person was.
You can see scenes where a crowd kneels, gestures and looks — at nothing. The figure they are worshipping is simply absent.
Aniconic: not showing a figure in human form. From Greek — an (not) plus eikon (image). Early Buddhist art is aniconic in this sense. Several other religious traditions have gone through similar phases.
Why? Nobody is certain. Proposals include: that a being who had escaped rebirth could not properly be shown in a body; that his absence was itself the point; or simply that it was not the custom and nobody had thought to break it.
5.2 — And then a face
Around the first century CE, in the region of Gandhara — roughly the Peshawar valley, the Swat valley and the Taxila region, sitting immediately north-west of Punjab — sculptors began making the Buddha as a man.
And the man they made looks, to anybody familiar with classical sculpture, startlingly Mediterranean.
- Wavy hair, gathered into a topknot, carved in the manner of Greek and Roman statuary — quite unlike the tight curls of later Indian images.
- A heavy robe with deep, naturalistic folds, draped over one shoulder and falling in the way a Greek or Roman garment falls.
- Facial features — the straight nose, the modelled brow, the calm proportion — following classical conventions.
- Standing poses with the weight resting on one leg, a classical device for making a figure look relaxed rather than rigid.
The material is usually a grey-blue stone called schist, quarried locally, and later stucco.
Put a Gandharan Buddha next to a Roman statue and the family resemblance is unmistakable. Put it next to earlier Indian sculpture and it looks like a different tradition.
Now consider how it got there. Alexander came and went, three hundred years earlier. Greek settlers stayed in Bactria. Greek kings ruled Punjab. Greek craft traditions carried on being taught. Roman trade brought more objects and more images along the sea and land routes.
Three centuries of accumulated contact, and then somebody in Gandhara — a Buddhist, working for Buddhist patrons — reached for those tools when a new job came up.
Nobody decided this. It happened the way most cultural change happens: a craftsman with a particular training was asked to make something nobody had made before, and used the skills he had.
5.3 — But there was a second workshop
At the same time, roughly, images of the Buddha also began appearing at Mathura, far to the south-east, on the Yamuna — also inside the Kushan empire.
The Mathura Buddha looks quite different. Carved in a distinctive mottled red sandstone. Broader, heavier, more powerfully built. The robe thin and clinging rather than heavily draped. The face rounder, the expression more direct. It grows visibly out of earlier Indian traditions of carving nature-spirit figures.
Two schools. Two very different-looking Buddhas. Both under the same empire, both starting at around the same time.
This has been argued for over a century, and the argument has never been purely about art.
The Gandharan images are unmistakably classical in technique. Contact with Greek and Roman art in this region was continuous for three centuries. Making a god look like a man was completely normal in the Greek tradition and unprecedented in the Buddhist one. The obvious explanation is that the classical tradition supplied both the idea and the skill, and Gandhara — where that tradition actually lived — did it first.
Indian sculpture already had a long tradition of carving powerful human and semi-divine figures. The step to depicting the Buddha is smaller than it looks and did not need an outside push. The Mathura images grow visibly and traceably out of local predecessors, whereas the Gandharan ones appear with less local ancestry. And the classical resemblance may be style rather than origin — a Gandharan sculptor would carve a Buddha in the manner he was trained in, whoever had the idea first.
Why it was fought over so bitterly: the argument ran hot in the colonial period, when "Greek influence explains Indian art" was a claim with obvious political weight, and the response to it had equal weight. The scholarship on both sides was real; so were the stakes outside it.
What is certain: both schools existed under the Kushans, both were producing Buddha images by the early centuries CE, and the Gandharan version is the one that travelled.
The whole Gandhara-versus-Mathura argument rests on a picture of how culture works that does not survive contact with the evidence.
The assumption is that influence flows one way, from a source to a receiver. One culture "influences"; the other is "influenced". One gives; the other takes.
Look at what a Gandharan Buddha actually is. An Indian religious idea, given a Greek-derived body, carved in local stone by a local craftsman, paid for by a Central Asian empire, for worshippers in a Buddhist monastery.
Who influenced whom? The question has no answer, because that is not what happened. Nothing flowed one way. A new thing was made out of everything that was available in that place at that time.
And notice why the one-way picture is so attractive: it lets you assign credit. If influence flows, someone is the source and gets to be the origin. That is exactly why this argument became political — both sides wanted the arrow pointing their way.
The honest description is less satisfying and more accurate: Gandharan art has no single parent. It is what happened when several traditions were in the same room for three hundred years.
This is Part 3's warning in another form. There, peoples were not billiard balls. Here, cultures are not either.
5.4 — Why this is Punjab's largest export
The Gandharan Buddha did not stay in Gandhara.
It travelled the route Chapter 4 described — out through the passes, along the Silk Road, into Central Asia, on to China, then Korea, then Japan. At each stage local artists adapted it. The features changed. The proportions changed. The robe changed.
But the fundamental decision — that the Buddha is shown as a seated or standing human figure, calm, robed, with a topknot and elongated earlobes — held all the way.
Picture a Buddha statue. Any Buddha statue — in a temple in Thailand, a garden in California, a museum in Tokyo, a shop in Delhi.
The basic form you are picturing was worked out in and around the north-west of this region, in the first centuries CE, by sculptors using Greek techniques for an Indian religion under a Central Asian empire.
It is, by an enormous margin, the most widely reproduced image ever created in this part of the world. More people have seen a descendant of the Gandharan Buddha than have seen almost any other image in human history.
For about five hundred years after the Buddha's death, Buddhist art did not show him. It used footprints, an empty throne, a wheel or a tree instead.
Around the first century CE, sculptors in Gandhara — next door to Punjab — began carving him as a man, using Greek techniques: wavy hair, deep robe folds, classical features.
At the same time, Mathura produced a different, more Indian-looking Buddha. Which came first is unresolved and is probably the wrong question.
Influence does not flow one way. A Gandharan Buddha is an Indian idea in a Greek-derived body, carved in local stone, paid for by a Central Asian empire.
And that image travelled to China, Korea and Japan. It is the most reproduced image this region has ever produced.
6How People Actually Lived
Kings and coins are the easy part. Here is what can be reconstructed about ordinary life — the cities, the trade, the languages, and the oldest Buddhist books that survive anywhere in the world, written on bark.
6.1 — Taxila
The best-excavated site of this whole period sits at the north-western edge of Punjab: Taxila, near modern Rawalpindi.
Taxila was not one city but a sequence of them, built near each other over centuries as rulers changed. Excavation has exposed successive settlements, each with its own layout.
One of them, usually called Sirkap, was laid out on a grid — a main street running the length of the city with side streets crossing it — in the manner of Greek town planning. Within it, archaeologists found houses, workshops, shrines and a Buddhist monastery, along with jewellery, coins, tools and imported objects.
Stratigraphy is the reason Taxila matters so much. It is a simple idea with enormous power.
When people live in one place for a long time, rubbish, collapsed buildings and new floors pile up. Deeper is older. Dig down carefully, recording exactly which layer each object comes from, and you get a sequence.
Now add coins. A coin names a king. Find a coin of a known king sealed in a layer, and everything in that layer is roughly of his time or later, and everything below it is earlier.
This is how the loose coin-based king-lists of Chapter 2 get anchored to real time. Coins alone give you a floating sequence. Coins in datable layers give you a calendar.
The honest limitation: much of the early excavation at Taxila was done to standards that would not be accepted today, and a great deal of contextual information was lost in the process. Some of what we would most want to know was dug away before anyone thought to record it.
6.2 — Trade
Two directions, and both left evidence.
Westward, to Rome. A Greek merchant's handbook from the first century CE, written for traders working the Indian Ocean, describes ports, routes, sailing seasons and — usefully for us — what to buy and sell at each place. It names a port at the mouth of the Indus, and lists goods coming down from the interior: turquoise, lapis lazuli, cotton cloth, silk and indigo.
That is a description of the Punjab and Gandhara trade reaching the sea.
Northward and eastward, along the Silk Road. Chinese silk moving west, horses moving south, and the whole exchange described in Chapter 4.
A merchant in Alexandria could buy cloth that had been woven in Punjab. A monastery in western China could hold a text copied in Gandhara. A Kushan noble could drink from Roman glass.
None of these people travelled the whole distance. The goods did, hand to hand, over months and years, each trader taking a cut.
Punjab was not on the edge of this system. It was a section of the main line.
6.3 — What language did people speak?
Several at once, depending on who you were and what you were doing.
- Gandhari, a Prakrit — the everyday spoken and written language of the north-west, written in the Kharosthi script.
- Greek, in administration and on coinage, for centuries after any Greek ruler had gone. It survived as a script and an official language long after it stopped being anybody's mother tongue.
- Bactrian, an Iranian language written in Greek letters, used by the Kushans — the language of the Rabatak inscription.
- Sanskrit, increasingly, as a formal and religious language, gaining ground through this period.
A merchant in Taxila might well deal in three of these in a working day. Multilingualism was not remarkable; it was the ordinary condition of a trading crossroads.
6.4 — The oldest Buddhist books in the world
This is the most extraordinary survival of the whole period.
In the 1990s, a collection of ancient manuscripts came to light — scrolls of birch bark, written in Gandhari in the Kharosthi script, dating to around the first century CE. Further finds followed.
They are Buddhist texts: sermons, commentaries, verse collections, monastic material.
They are the oldest surviving Buddhist manuscripts anywhere in the world. Older than any surviving manuscript from Sri Lanka, Tibet, China or Southeast Asia.
Buddhism is a world religion with a vast literature. The physically oldest copies of any of it are birch-bark scrolls from the north-west of this region, written in the local language of Punjab and Gandhara, in a script derived from Aramaic, that no longer exists.
They survived because they were placed in clay pots and buried — probably as a reverent way of disposing of worn-out sacred texts, the way a damaged book might be respectfully retired rather than thrown away.
Somebody, roughly two thousand years ago, decided a worn-out scroll deserved better than a rubbish heap. That decision is why we can read it.
They are also extremely fragile — dry, brittle bark that can crumble when unrolled — and the work of conserving and reading them has taken decades.
6.5 — Religion on the ground
Buddhism was strong in the north-west in this period — the monasteries, the art, the manuscripts all say so, and it clearly had elite and royal support.
But it was never the only thing. Shiva-like figures on Kushan coins, the Heliodorus pillar from Chapter 2, Iranian deities in the Kushan pantheon, and continuing local worship all point to a landscape of overlapping practice rather than a single official religion.
We instinctively ask "what religion were the people of Punjab in 150 CE?" and expect one answer.
The assumption is that religion is exclusive — that belonging to one means not belonging to another.
That is a real way of thinking about religion, but it is not the only one, and it was not obviously how this world worked. A merchant might give to a Buddhist monastery because the monks sheltered his caravan, honour a local deity because that is what his family had always done, and use coins showing Iranian gods without giving it a thought.
Asking which one he "really" was may be importing a question he never asked.
Be careful in both directions, though. The opposite error — imagining a serene pluralism with no conflict — is also unsupported. Chapter 8 covers a period when things went badly wrong. Religious groups in this period competed for patronage, argued sharply in their texts, and sometimes did each other real harm.
The accurate picture is neither harmony nor exclusivity, but something messier: overlapping practice, real competition for royal money, and boundaries that were much softer than the categories we use today.
Taxila, at Punjab's north-western edge, is the best-excavated site of the period. One of its cities was laid out on a Greek-style grid.
Stratigraphy plus coins is what turns a floating list of kings into a real chronology — deeper is older, and a datable coin in a layer dates the layer.
Trade ran west to Rome by sea and north along the Silk Road. Punjab was on the main line, not the edge.
People used Gandhari, Greek, Bactrian and Sanskrit — often several at once.
The oldest Buddhist manuscripts in the world are birch-bark scrolls from this region, written in Gandhari in a script that no longer exists.
7The Guptas, and Why Punjab Was Barely In Them
The Gupta period is taught as India's golden age. It was centred six hundred kilometres east of Punjab, and Punjab was never really part of it. Both of those facts are worth examining.
7.1 — The empire in the east
Around 320 CE, a ruler called Chandragupta I established a kingdom in Magadha, in the eastern Ganges plain — the same region the Mauryas had ruled from.
His son Samudragupta expanded it enormously. His grandson Chandragupta II presided over its height. The empire lasted, in strength, until the middle of the fifth century.
This is the period associated with a great deal of what is now considered classical Indian civilisation — Sanskrit literature and drama, mathematics and astronomy, temple architecture, and sculpture of extraordinary quality.
7.2 — Punjab's actual relationship to it
Here is the part usually skipped.
Our best single source for the extent of Gupta power is a long inscription carved onto a stone pillar — a pillar that Ashoka had already used, six hundred years earlier, for his own edicts. Samudragupta's court poet added a record of his conquests to it.
What matters is how the inscription describes different territories, because it does not treat them all the same.
- Some kingdoms are described as conquered and absorbed — mostly in the Ganges plain, the empire's core.
- Some southern kingdoms are described as defeated and then restored to their rulers, who acknowledged Gupta supremacy.
- And then there is a third category: frontier kingdoms and tribal republics who are described as submitting, paying tribute and obeying commands — but who are plainly not being annexed.
Punjab and its neighbours are in that third category. The peoples named include the Yaudheyas, the Malavas, the Arjunayanas and the Madrakas — the same self-governing confederations from Chapter 1, still there six hundred years later.
Read what that inscription is actually admitting, underneath the boasting.
The greatest Gupta emperor, at the height of his power, describes Punjab's peoples as having submitted and paid tribute. He does not describe them as having been conquered and governed.
In the language of empire, that is a polite way of saying: they agreed to be left alone in exchange for a payment and a form of words.
So when a textbook shows a map with the Gupta empire covering northern India in one solid colour, and Punjab inside it — be careful. That colour is doing an enormous amount of work.
Meanwhile the far north-west — Gandhara and beyond — was under the control of successor rulers of the Kushans and their overlords, and was not Gupta territory at all.
7.3 — Golden for whom?
The intellectual output is not a nationalist invention. Mathematical and astronomical work from this period includes a decimal place-value system with a symbol for zero — arguably the single most consequential idea India has given the world. Sanskrit literature and drama reach a peak that later ages treated as the standard. Temple architecture and sculpture set forms used for the next thousand years. A Chinese Buddhist pilgrim who travelled through around 400 CE described a prosperous, orderly and well-governed land. Long stretches of peace in the core allowed all of it.
Every item on that list is a court and elite achievement. Sanskrit drama was written for and performed before an educated elite; most people did not speak Sanskrit. The mathematics was produced by a tiny scholarly class. The temples were built with the labour of people whose lives we know nothing about.
Meanwhile several historians argue the period saw the hardening of caste, a decline in the legal and social position of women compared with earlier centuries, and the beginning of large-scale land grants to religious institutions and officials — a process that tied cultivators more firmly to the land and shifted power to local landholders.
And the visiting pilgrim, whose account is often quoted as proof of prosperity, also described a group of people forced to live outside settlements and to strike a piece of wood to announce their approach so that others could avoid contact with them.
The fair summary: it was a period of extraordinary elite cultural production, resting on a society that was becoming more rigidly stratified. Both halves are true. Most popular accounts keep only the first.
And for Punjab specifically, the question is largely academic — whatever the Gupta age was, Punjab was mostly outside it.
Chapter 1 introduced this and here is where it pays off.
The assumption is that the history of a region is the history of the largest state that claimed it.
It follows almost automatically from where the evidence comes from. Big states produce inscriptions, court poetry, coins and monuments. So the history gets written from their point of view, and everyone inside the borders on the map gets treated as part of the story.
But for most of the seven centuries in this part, Punjab was governed by people who left far fewer records — republics with assemblies, cities running their own affairs, local rulers, tribal confederations. They issued coins, which is nearly all we have of them. They wrote no court poetry and built no monuments to themselves.
So they are nearly invisible, and the visible thing — the big eastern empire — gets to be the period's name.
The Yaudheyas appear in the record as a people the Mauryas dealt with, and are still there when Samudragupta lists his tributaries six hundred years later. Six centuries of continuous existence, and they get a line in somebody else's victory inscription.
That is not because they did not matter. It is because they did not employ a poet.
The Gupta empire, from about 320 CE, was centred in the eastern Ganges plain — six hundred kilometres from Punjab.
Samudragupta's own inscription puts Punjab's peoples — the Yaudheyas, Malavas, Arjunayanas and Madrakas — in the category of tributary frontier peoples, not conquered provinces. Be sceptical of maps showing Punjab in solid Gupta colour.
The achievements were real: the decimal system with zero, Sanskrit literature, temple architecture. The phrase "golden age" is disputed, because the period also saw caste harden and land grants tie cultivators down.
Always ask: golden for whom? And remember that the history of a region usually gets written by whoever employed a poet.
8The Hunas
The last arrivals of this part came through the passes in the fifth century, ruled from Sialkot, and are remembered in Indian and Chinese sources as destroyers. Whether they deserve that reputation is genuinely unclear — and the reason it is unclear is instructive.
8.1 — Who they were
In the fifth century CE a new group came down through Afghanistan into Gandhara and Punjab. Indian sources call them Hunas. In European and Chinese sources related groups appear under other names, often translated as White Huns or Hephthalites.
Their relationship to the Huns who attacked the Roman empire at roughly the same time is debated — they may be connected, or the name may simply have been applied broadly to steppe peoples. What is agreed is that they came from Central Asia, they were mounted and militarily formidable, and they arrived in force.
8.2 — What happened
Around the middle of the fifth century they pressed on the Gupta frontier. The emperor Skandagupta claims in an inscription to have defeated them, and for a time the frontier held.
But the effort was costly, and Gupta strength declined afterwards. By around 490 CE a Huna ruler named Toramana was in control of a substantial territory in north-western and central India.
His son Mihirakula followed him, from around 515 CE.
And Mihirakula's capital was Sagala — Sialkot. The same city that had been Menander's capital seven hundred years earlier, at the start of this part.
Seven centuries apart, two rulers from entirely different worlds — a Greek from Bactria and a Huna from Central Asia — chose the same Punjab city as their seat.
That is not a coincidence. It reflects geography. Sialkot sits in the upper Punjab plain, near the routes coming down from the north-west, in well-watered country between the Chenab and the Ravi.
Rulers keep choosing the same places because the land keeps offering the same advantages. Part 1 said geography sets the board. This is the board being set the same way twice, seven hundred years apart.
Around 528 CE, a ruler from Malwa named Yashodharman defeated Mihirakula, and recorded the victory on stone pillars. Huna power in India broke soon after.
But the Guptas did not recover. The empire fragmented over the following decades, and northern India entered a long period of regional kingdoms.
8.3 — The reputation
Mihirakula has an exceptionally bad reputation. Later Indian texts describe him as cruel. A Chinese Buddhist pilgrim who travelled through a century afterwards recorded that he had persecuted Buddhists, destroyed monasteries and killed monks. Another Chinese envoy who passed through earlier also reported destruction.
By the seventh century, Gandhara's monasteries were largely ruined and its Buddhist civilisation was finished.
Multiple independent sources, Indian and Chinese, describe Huna destruction and specifically Mihirakula's hostility to Buddhism. The archaeological record does show destruction and abandonment at monastic sites in the north-west in roughly this period. The Gandharan tradition, which had flourished for five hundred years, ends. Something ended it, and the sources name a culprit.
The sources are late and partisan. The Chinese pilgrim accounts were written a century or more after the events, by devout Buddhists collecting local stories, in a landscape already in ruins. Ruins invite an explanation, and a remembered villain supplies one.
Attributing archaeological destruction to a named person is very hard. A burnt layer tells you a building burned. It does not tell you who lit it or why.
And there were other causes at work. Gandharan Buddhism depended heavily on royal and merchant patronage. When the Kushan and Gupta orders came apart and Silk Road trade was disrupted, that funding dried up. A monastery with no donors empties whether or not anyone attacks it.
Some Huna coins carry Shaiva imagery, which has been read as showing Mihirakula favoured Shiva — but favouring one cult is not the same as a campaign of destruction, and rulers across this period routinely patronised several traditions at once.
The likely fuller answer: Gandharan Buddhism died of several causes at once — the collapse of the patronage networks that funded it, the disruption of the trade that enriched them, political fragmentation, and violence. The Hunas were part of it, not the whole of it.
What would help: more carefully dated excavation at Gandharan monastic sites, establishing when each was abandoned and whether destruction layers are actually present.
This argument is a good example of a habit that runs through a great deal of history writing.
The assumption is that large outcomes need a single cause, and preferably a person.
An entire civilisation ends. Why? Because a bad king destroyed it. That is a satisfying answer. It has an agent, a motive and a moral.
The alternative — that a complex institution collapsed because the money stopped, the trade routes shifted, the political order fragmented and violence happened, over several decades, with no one deciding it — is harder to hold in the mind and impossible to put in a sentence.
So the villain wins, in the record and in the retelling.
Notice this is the same shape as the invasion theory in Part 2. Cities emptied; the explanation offered was a conqueror. Careful work showed a slower answer: climate, water, trade, cost. Both times, a person was blamed for a process.
That does not mean villains never exist or that Mihirakula was harmless. It means: when you meet an explanation that is one bad man, check whether anyone has looked for the slower answer underneath.
The Hunas came through the passes in the fifth century CE. Skandagupta held them off, at a cost the Guptas never recovered from.
Toramana ruled from around 490 CE, then his son Mihirakula — from Sialkot, the same city Menander had ruled from seven hundred years earlier.
Yashodharman defeated Mihirakula around 528 CE. Huna power broke, but the Gupta empire fragmented anyway.
Mihirakula is blamed for destroying Buddhism in the north-west. The sources are late, partisan and written among ruins. Real disruption, probably; sole cause, probably not.
Gandharan Buddhism most likely died of several things at once — lost patronage, disrupted trade, political collapse and violence. Be suspicious when a whole civilisation's end is pinned on one man.
9What Seven Centuries of Outsiders Did to Punjab
Greeks, Scythians, Parthians, Kushans, Hunas. On paper, seven hundred years of foreign rule. In practice, something much stranger and more interesting happened — and it says something about what Punjab is.
9.1 — The pattern
Line up what happened to each group after they arrived.
The Greeks. Within a century and a half, a Greek ambassador is putting up a pillar to an Indian god and calling himself a devotee. A Greek king is the questioner in a Buddhist scripture. Their coins carry Indian script, Indian titles and Buddhist symbols.
The Sakas and Parthians. They keep the existing bilingual coinage and administration. They take Indian titles. They patronise Buddhist institutions. Within generations they are indistinguishable in the record from local rulers except by name.
The Kushans. They put Indian gods on their money alongside Iranian and Greek ones. They fund Buddhist monasteries on an enormous scale. They adopt Indian royal titles. The art made under them becomes the defining art of Indian Buddhism.
The Hunas. Even here — Huna coins follow the existing regional coin types, and Huna rulers appear as patrons of local cults. Later Indian genealogies absorb Huna groups into existing social categories rather than treating them as permanently outside.
The pattern is the same every time, and it takes about two or three generations.
A group arrives armed. It takes power. It discovers that ruling requires the existing officials, the existing tax system, the existing scribes and the existing coinage, because building new ones from nothing is impossible. It keeps them.
Its rulers marry locally, because that is who is there. Their children grow up speaking the local language alongside their parents'. Their grandchildren fund the local temples and monasteries, because that is how a ruler here demonstrates legitimacy.
Nobody surrenders. Nobody decides to assimilate. It simply happens, through the ordinary business of governing a place and living in it.
9.2 — Why this happens
Three reasons, and none of them is about anybody's virtue.
One: numbers. Every arriving group was vastly outnumbered by the people already here. A ruling class of some thousands over a population of millions cannot impose its language and customs even if it wants to.
Two: the machinery of government is inherited. Tax registers, weights, scribes, local headmen, courts — these take generations to build. A conqueror who destroys them destroys his own income. So he keeps them, and with them he keeps the language and the categories they run on.
Three: legitimacy is local. To be accepted as a ruler you have to look like one to the people you rule. That means funding their institutions, honouring their gods, using their titles. A king who refuses is a permanent occupier, which is expensive; a king who complies becomes, in time, simply the king.
Underneath the whole way this period is usually described sits a word that does far more work than it can support: foreign.
Ask when a Kushan family stopped being foreign. After one generation in Punjab? Three? Ten? When they started speaking the local language? When they funded a local monastery? When they married in?
There is no answer, because "foreign" is not a property of people. It is a judgement made by somebody, at some moment, for some purpose.
And notice the double standard that runs through nearly every account of this period. The Kushans are described as foreign rulers. The Mauryas are not — though Chandragupta's origins are obscure and his empire was assembled by force. The Vedic people of Part 3 are argued over endlessly. The Guptas are simply Indian.
The label attaches to some groups and not others, and the deciding factor is usually whether their descendants stayed and were absorbed into the categories used later — not anything about the arrival itself.
This is Part 3's assumption in a different costume. There it was: ancestry decides belonging. Here it is: arrival decides foreignness. Both treat a permanent identity as something you can read off an origin. Neither survives contact with what actually happens to people over three generations.
9.3 — What Punjab actually gained
Not nothing, and not simply "foreign influence".
- A permanent position on the world's main trade route, and the wealth, cities and cosmopolitanism that came with it.
- Gandharan art, and through it the image of the Buddha that half the world uses.
- A role in transmitting Buddhism to China, Korea and Japan.
- A gold coinage and a monetised economy at a scale not seen here before.
- A calendar that is still India's official national era.
- A tradition of religious plurality visible on the coins themselves.
9.4 — And what it permanently lost
This needs saying too, because the losses were real and final.
The Kharosthi script. Punjab's own writing system for seven hundred years, gone completely by around the fourth century CE, with no descendants. This is the second writing system this land has lost.
Gandhari. The regional language in which the oldest Buddhist books in the world were written. Extinct.
Gandharan Buddhism. An entire religious civilisation — hundreds of monasteries, a great artistic tradition, a distinct scriptural language — reduced to ruins by the seventh century and never rebuilt.
Stand at Taxila or Takht-i-Bahi today and you are looking at the physical remains of something that was, for five hundred years, one of the great religious and artistic centres of the world.
Almost nobody who lives in the region now is Buddhist. The script the monks wrote in cannot be read without specialist training. The language is dead.
Punjab has lost more than most places, and it has lost it more than once. The Harappan script went silent. Kharosthi went silent. Both times, an entire way of writing the world simply stopped, and what replaced it had no memory of it.
Part 13 will describe a third such rupture, within living memory.
Five groups arrived and ruled: Greeks, Sakas, Parthians, Kushans, Hunas. Every single one was absorbed within two or three generations.
Not because anyone was noble about it, but because they were outnumbered, because they had to inherit the machinery of government, and because legitimacy had to be earned locally.
"Foreign" is not a property of people. It is a judgement, applied unevenly, usually decided by whether the descendants stayed.
Punjab gained a place on the Silk Road, the Buddha's face, a role in sending Buddhism to East Asia, a gold economy, and a calendar still in use.
And it permanently lost the Kharosthi script, the Gandhari language, and the entire Buddhist civilisation of the north-west. This land has now lost two complete writing systems.
10An Honest List of What We Do Not Know
This part rests on coins, a handful of inscriptions, some excavated cities, and religious texts written by people with a case to make. Here is where that leaves us.
Genuinely unknown
- The order and dates of most Indo-Greek kings. The list comes from coins and stylistic judgement, and different specialists produce different sequences.
- How long any of them ruled, or how much territory they actually held. Coin distribution is a rough guide, not a map.
- Whether Menander became a Buddhist, or whether the text simply ends the way such texts usually end.
- Whether Kanishka converted, or patronised everything.
- Which came first, the Gandharan or the Mathuran Buddha image — and whether the question is even well-formed.
- Why Buddhist art was aniconic for five centuries, and what changed.
- Almost everything about ordinary people. No personal letters, no household records, no accounts of a normal life. We have kings' coins and monks' books.
- How much of the destruction in the north-west the Hunas actually caused, as against patronage collapse and trade disruption.
- What the tribal republics were actually like inside — how their assemblies worked, who could speak, how decisions were made.
- The origin of the Saka era of 78 CE.
- What the population of this region was, at any point in these seven centuries.
Solid
- The Mauryan empire ended around 185 BCE and the north-west separated from eastern control.
- Greek kings from Bactria ruled parts of Punjab and Gandhara from around 180 BCE, and Menander's capital was Sagala — modern Sialkot.
- The Milindapanha exists, is a Buddhist text, and takes a Greek king as its questioner.
- Indo-Greek coinage is bilingual, Greek and Kharosthi, and that bilingualism is how Kharosthi was deciphered.
- Sakas, then Parthians, then Kushans ruled successively in the north-west, and each kept the existing administrative and monetary system.
- Kanishka's reign began around 127 CE. The Rabatak inscription gives the Kushan royal genealogy.
- Kushan coinage carries Greek, Iranian, Indian and Buddhist deities, including a Buddha labelled in Greek letters.
- Gandharan sculpture used classical techniques and produced the human Buddha image that spread across Asia.
- The oldest surviving Buddhist manuscripts in the world are Gandhari birch-bark scrolls from this region.
- Samudragupta's inscription lists Punjab's peoples as tributary frontier groups, not annexed provinces.
- The Hunas ruled in the north-west from around 490 CE, with Mihirakula seated at Sialkot, until Yashodharman's victory around 528 CE.
- Kharosthi and Gandhari both went extinct, leaving no descendants.
Look at the shape of what we have for this period and notice what it excludes.
Coins — made by rulers, about rulers. Inscriptions — commissioned by rulers and rich donors. Religious texts — written by monks, about religion, often centuries later. Excavated buildings — the ones sturdy enough to leave remains, which means temples, monasteries and city walls, not houses of mud and thatch.
Every one of those sources comes from someone with money, power or an institution behind them.
So the history of these seven centuries is a history of kings, monks and merchants — because those are the people who could afford to leave a trace.
The farmers who grew the grain that fed the monasteries, the women who wove the cloth that went to Rome, the labourers who carved the Buddhas — all of them are in this part only as an inference. We know they existed because the buildings stand and the coins were spent. We know nothing else about them at all.
That is not a failure of research. It is the nature of the evidence, and it is worth naming rather than quietly working around.
Sources & further reading — Part 5
Timeline of Part 5
Dates in this part are unusually uncertain. Where a date is disputed, the dispute is in the chapter.
| When | What happened |
|---|---|
| c. 232 BCE | Ashoka dies. Where Part 4 ended. |
| c. 250 BCE | Diodotus breaks Bactria away from the Seleucids. An independent Greek kingdom in Central Asia. |
| c. 185 BCE | Pushyamitra Shunga kills the last Mauryan ruler. The empire ends. The north-west separates. |
| c. 180 BCE | Demetrius leads the Bactrian Greeks across the mountains into the Indus valley. |
| c. 165 – 130 BCE | Menander rules from Sagala (Sialkot), in Punjab. The Milindapanha records him debating a Buddhist monk. |
| c. 176 – 160 BCE | The Xiongnu drive the Yuezhi west. The domino chain begins. |
| c. 110 BCE | Heliodorus, a Greek ambassador, erects a pillar in central India to an Indian god and calls himself a devotee. |
| c. 80 BCE onwards | Maues, then Azes: Saka (Indo-Scythian) rule around Taxila and Punjab. |
| c. 20 – 50 CE | Gondophares and the Indo-Parthians rule from the Taxila region. |
| 1st century CE | Kujula Kadphises unites the five Yuezhi groups. The Kushan empire begins. Gandhari birch-bark Buddhist scrolls are written. |
| c. 78 CE | The Saka era begins. Still India's official national calendar today. |
| 1st century CE | First human images of the Buddha appear, in Gandhara and at Mathura. |
| c. 127 CE | Kanishka accedes. The Kushan empire at its height, from Central Asia to the Ganges. Gold coinage; Buddha coins labelled in Greek letters. |
| 1st – 3rd centuries CE | Silk Road trade at its peak. Buddhism travels from Gandhara through Central Asia to China. |
| c. 320 CE | The Gupta empire begins in the eastern Ganges plain. Punjab remains outside it. |
| c. 350 CE | Samudragupta's inscription lists the Yaudheyas, Malavas, Arjunayanas and Madrakas of the Punjab region as tributary frontier peoples. |
| c. 3rd – 4th century CE | The Kharosthi script falls out of use in India. It leaves no descendants. |
| c. 399 – 414 CE | The Chinese pilgrim Fa-Xian travels through India and writes an account of it. |
| c. 455 – 467 CE | Skandagupta holds off the first Huna pressure, at heavy cost. |
| c. 490 – 515 CE | Toramana rules a large territory in north-western and central India. |
| c. 515 – 540 CE | Mihirakula rules from Sialkot — the same city as Menander, seven centuries earlier. |
| c. 528 CE | Yashodharman defeats Mihirakula. Huna power breaks. |
| c. 550 CE | The Gupta empire fragments. Northern India enters an age of regional kingdoms. Part 6 starts here. |
| 1993 CE | The Rabatak inscription is found, settling the Kushan royal genealogy. |
| 1990s CE | Gandhari birch-bark scrolls come to light — the oldest Buddhist manuscripts in the world. |
Glossary
Every hard word used in Part 5, in plain English.
| Word | What it means |
|---|---|
| Aniconic | Not showing a figure in human form. Early Buddhist art marked the Buddha with footprints, an empty throne, a wheel or a tree instead of a body. |
| Bactria | The region of northern Afghanistan and southern Uzbekistan. A Greek kingdom after Alexander; later the Kushan heartland. |
| Bactrian | An Iranian language written in the Greek alphabet. The Kushan administrative language. |
| Brahmi | The ancient Indian script written left to right. Ancestor of almost every modern Indian script, including Gurmukhi and Devanagari. |
| Dynasty | A line of rulers from one family, passing power down through it. |
| Gana-sangha | A state run by an assembly of leading men rather than a single king. Often translated "republic". Not a democracy. |
| Gandhara | The region around Peshawar, Swat and Taxila, immediately north-west of Punjab. Home of the Buddha image. |
| Gandhari | The Prakrit language of Gandhara and the north-west, written in Kharosthi. Extinct. |
| Indo-Greek | The Greek-descended kings who ruled parts of Punjab and Gandhara from around 180 BCE. |
| Kharosthi | Punjab's ancient script. Derived from Aramaic, written right to left. Died out around the fourth century CE with no descendants. |
| Kushans | The Yuezhi after they settled and organised. Rulers of an empire from Central Asia to the Ganges, first to third centuries CE. |
| Milindapanha | "The Questions of Milinda". A Buddhist text recording a dialogue between King Menander and the monk Nagasena. |
| Numismatics | The study of coins. The main evidence for much of this period. |
| Prakrit | The everyday spoken languages of ancient India, as distinct from formal Sanskrit. |
| Sagala | Ancient name of Sialkot, in Punjab. Capital of Menander and, seven centuries later, of Mihirakula. |
| Saka | The people Greek writers called Scythians — steppe herders who moved into the Indus valley and Punjab. Also called Indo-Scythians. |
| Saka era | A year-count beginning in 78 CE. Adopted as India's official national calendar in 1957. |
| Schist | The grey-blue stone most Gandharan sculpture is carved from. |
| Silk Road | Not a road but a network of trade routes linking China, Central Asia, India, Iran and the Mediterranean. Goods passed hand to hand. |
| Stratigraphy | Reading a site by its layers. Deeper is older. Combined with datable coins, it turns a floating king-list into a chronology. |
| Stucco | A plaster used for sculpture in later Gandharan work, cheaper and quicker than carving stone. |
| Xiongnu | A nomadic confederation on China's northern frontier whose wars set the domino chain of Chapter 3 in motion. |
| Yuezhi | The people driven west by the Xiongnu, who became the Kushans. |