Where We Left Off
A short recap for anyone starting here, a reminder of what the coloured boxes mean, and a note on what changes in this part — because the evidence problem turns completely around.
Part 3 was built on one source. The Rigveda — 1,028 hymns, memorised for centuries before anyone wrote them down, composed by paid priests speaking to gods in front of the patrons who employed them. Not one line of it was meant to record history. Everything historical in it fell in around the edges of the asking.
From those edges we got a great deal. The rivers of Punjab under their older names. A world of herders, chiefs and priests, with cattle for wealth and horses for glory. A fight on the Ravi between a chief called Sudas and an alliance of ten. The first appearance, in the latest layer of the book, of the four-fold order that hardened over the next thousand years into caste.
And an argument. Part 3 laid out three positions on where the Sanskrit speakers came from — invasion, migration, indigenous origin — and gave each its strongest case. It ended where the evidence ends: pointing towards a gradual, small-scale, male-biased movement in the second millennium BCE, without proving it, and with the ancient DNA missing from the one region that matters most.
Part 3 also ended with the centre of gravity leaving Punjab. Iron made the wet forest of the Ganga plain worth clearing. Kingdoms formed out there. The land of the seven rivers, which had been the whole world of the hymns, became the north-western edge of somebody else's map.
And then, around 518 BCE, an empire arrived from the other direction and wrote our name in stone.
What changes in this part
Everything about the evidence.
For three parts you have been reading history assembled from things that were not trying to tell you anything — rock strata, brick foundations, broken pots, tooth enamel, hymns about cattle. The problem was silence.
From here the problem is the opposite. Suddenly there are documents. Persian kings carve lists of their provinces into cliffs. A Greek writer four thousand kilometres away describes what Indians eat and how they die. Officers in a Macedonian army keep journals. An ambassador lives at an Indian court for years and writes a book about it. And at the end of this part an Indian emperor covers his own empire in inscriptions written in the ordinary language of the people, telling them what he thinks and what he regrets.
This looks like an enormous improvement, and in most ways it is. But it brings a new problem, and it is the one to hold on to through this whole part.
All of these documents were written by somebody with a purpose. The Persian inscriptions are royal propaganda carved forty metres up a cliff where nobody could read them. The Greek accounts of Alexander were written centuries after he died, from lost memoirs by men who had served under him and had reputations to protect. The Indian emperor's rocks are the emperor's own account of himself.
Silence made us guess. Documents make us trust. Guessing at least feels like guessing.
What the boxes mean
Six kinds of coloured box run through this series. Each does one job. If you read nothing else, read the black ones.
- Word Box (slate blue) — explains a hard word the moment it first appears. You never have to already know something.
- In Real Terms (moss green) — turns a number or an idea you cannot picture into something ordinary.
- How We Actually Know This (sepia) — shows the evidence itself. The inscription, the coin, the excavated wall. And what it can and cannot prove.
- The Argument (ochre) — where the experts disagree. Each side gets its strongest case, not a weak version I can knock over. Then a plain statement of where things stand and what would settle it.
- The Hidden Assumption (plum) — digs out something everybody takes for granted without proving. This is the part of the series I am most attached to.
- Remember This (dark) — closes every chapter with the key points in the plainest words I can find.
A note on the famous bit
Most people come to this period for one thing: Alexander. He is in every school textbook, every list of great commanders, every documentary about the ancient world.
He gets two chapters here, and they are honest ones. The battle on the Jhelum was real, it was hard, and the man who fought him was a Punjabi king who came out of it with his kingdom enlarged.
But I am going to keep asking a question that the textbooks do not ask, and it will be uncomfortable in both directions. Why do we know so much about Alexander in India, when no Indian source from any period so much as mentions him?
The answer is not that Indians were careless. It is that importance is not a property of events. It is assigned, afterwards, by whoever kept the records — and for two thousand years the people keeping the records about this were Greeks, Romans, and then Europeans reading Greeks and Romans.
Meanwhile the man who actually mattered most to Punjab in this period — who took the whole north, threw the Macedonians out of it, and built the first empire that ran from Afghanistan to Bengal — is a footnote in the same textbooks, under a name most people cannot place.
1The First Time Anyone Wrote Us Down
Around 518 BCE, Punjab appears in a written record for the first time. It is not our record, it is not in our language, and it was not put there for us to read.
1.1 — From silence to documents
Here is the position at the end of Part 3. We have hymns that were composed here and were not written down for a thousand years. We have pottery. We have some bones and a very small number of ancient genomes. We do not have a single sentence written by anyone living in Punjab, about anything, ever.
Then a Persian king carves a list of his provinces into a cliff, and one of the names on the list is ours.
That is the moment this part starts. Not because anything changed on the ground — the villages and the herds and the fields carried on exactly as before — but because from here we can read about it.
It is worth being precise about how thin the improvement is at first. What we get from Persia is a name on a list, a tribute figure, and a few passing mentions in a Greek book written by a man who never came here. That is the whole documentary record of Punjab for the first two centuries of this part.
Compared with nothing, it is enormous. Compared with what we would want, it is almost nothing.
1.2 — What an inscription is for
The list is on a cliff face at Behistun, in western Iran, carved on the order of Darius the Great around 520 BCE. It runs to over a thousand lines in three languages, tells the story of how Darius took the throne, and names the lands he ruled.
Now here is the detail that tells you what kind of document this is.
The Behistun inscription is carved about a hundred metres up a sheer rock face. That is roughly the height of a thirty-storey building.
You cannot read it from the ground. You could not read it from the ground in 520 BCE either. When Darius's stonemasons finished, they cut away the ledge they had been standing on, so that nobody could climb up to it.
The first European to copy it properly, in the 1830s and 40s, had to be lowered down the cliff on ropes and at one point hired a local boy to swing across a gap.
So ask the obvious question. Who was it for?
Not readers. Almost nobody in the empire could read at all, and the ones who could were not going to be dangling on a rope. It was for the gods, for posterity, and for the fact of it — a permanent statement that this was true, placed somewhere it could never be altered or removed.
A royal inscription is not a report. It is a monument that happens to have words on it.
Two words in that paragraph need explaining before we go on.
Achaemenid: the family name of the Persian dynasty founded by Cyrus the Great. Their empire — the first of the really enormous ones — ran from the Aegean to the Indus for about two hundred years.
Satrapy: a province of that empire. Satrap: the governor of one. The word passed into many languages and still carries a whiff of "somebody else's man in charge of your place".
Inscription: writing cut into a hard surface — rock, stone, metal. The study of them is called epigraphy. Inscriptions are the backbone of ancient history for a simple reason: they survive. Everything written on leaf, bark, cloth or paper in this climate is gone.
Cuneiform: the wedge-shaped writing of Mesopotamia and Persia, made by pressing a cut reed into wet clay or chiselling the same shapes into stone.
1.3 — Herodotus, and the trouble with him
The other source for this period is a Greek writer named Herodotus, working around 430 BCE. He is usually called the father of history. He has also been called the father of lies, and both titles are deserved.
What he did was genuinely new. He set out to find out why the Greeks and Persians had fought, travelled widely, asked people questions, wrote down what they said, and — crucially — often told you where he got it and whether he believed it.
He also reports, with a straight face, that in the deserts of India there are ants bigger than foxes that dig gold out of the sand, and that Indians collect the gold by riding in on camels while the ants are asleep.
Herodotus never came to India. He is at least two or three people away from anything he says about it — a Greek who talked to Persians who had talked to somebody in the east.
So how do you use a source like that?
What is probably reliable: administrative facts he could have got from Persian records or officials. Which provinces existed. What they paid. That Indian troops served in the Persian army. These are dull, checkable, and there is no obvious reason to invent them.
What is probably not: anything colourful. Customs, monsters, marvels. These are exactly the things that improve with each retelling, and they are what a traveller's audience wants.
And the giant ants? There is a real suggestion that they are marmots — burrowing animals of the high plateaus north of Kashmir that throw up soil which does, in some places, carry gold dust. A traveller's tale about a large digging animal and gold-bearing earth needs only a few retellings to become an ant the size of a fox.
That is the honest position on Herodotus. Not a liar. A man faithfully writing down what he was told, by people who were often telling him rubbish.
1.4 — The name they gave us
Chapter 2 of Part 3 followed a word. Here it is again, one step further along.
The river the hymns called Sindhu reached Persian mouths and became Hindush, because Old Persian regularly turns an initial s into an h. That is the name on Darius's lists — the land at and beyond that river.
The Greeks took the Persian word, dropped the h as their language required, and got Indos for the river and India for the country beyond it.
Everything follows from those two mispronunciations. India. Indus. Indian. Later, through Persian and Arabic, Hindustan, and eventually Hindu and Hindi. A whole vocabulary of identity for a fifth of humanity, descended from foreigners struggling with the name of a Punjabi river.
And notice the shape of it. We did not name ourselves. The oldest surviving name for this country as a whole is a foreign administrator's label for the far edge of somebody else's empire.
That being written about is the same as being known.
The moment documents appear, historians relax. The dark ages end. We say a region "enters history" — as though it had been waiting outside.
But look at what actually entered history here. A place-name on a tribute list. A figure in gold dust. A note that soldiers from there fought in somebody's army. Some nonsense about ants.
Nothing about who lived here, how they were governed, what they grew, what they believed, what they called themselves, or what they thought about being a province. On all of that, the arrival of writing has told us precisely nothing, because the people doing the writing had no interest in it.
What we have gained is not knowledge of Punjab. It is knowledge of what Punjab looked like from Persepolis. Those are very different things, and the difference gets forgotten the moment a document exists, because a document feels like an answer.
This runs through the whole part, and it gets worse before it gets better. The Greek accounts of Alexander tell you a great deal about what the Macedonian officer corps thought was worth recording. The Persian lists tell you what an imperial accountant needed. Only at the very end of this part — with Ashoka's rocks — does anybody standing in this land write something down and leave it where ordinary people could see it.
A region does not enter history. It enters somebody else's filing system.
Around 518 BCE Punjab appears in writing for the first time — as a province name on a Persian king's list. Not our writing, not our language, not for us.
A royal inscription is a monument, not a report. The great Behistun text is carved a hundred metres up a cliff and the masons cut away the ledge behind them. It was never meant to be read.
Herodotus is the other source. He never came here. Use his dull administrative facts and discount his marvels — the giant gold-digging ants are probably marmots, three retellings later.
The Persians turned Sindhu into Hindush. The Greeks turned that into India. We did not name ourselves; the oldest name for this country is a foreign clerk's label for the far edge of an empire.
Entering history is not the same as being understood. It means somebody else started keeping a file.
2A Province of Persia
For roughly two centuries, part of this land was the far eastern corner of the largest empire the world had yet seen. What that actually meant on the ground is much less clear than the maps suggest.
2.1 — Cyrus and Darius
The Persian empire was built in about thirty years, which is very fast for something that size.
Cyrus the Great took Media, then Lydia in Anatolia, then Babylon. By the time he died around 530 BCE the empire ran from the Aegean coast to somewhere in Central Asia. Greek writers say he campaigned in the east and came off badly in the deserts of what is now Balochistan. Whether he ever reached the Indus is not settled.
What is reasonably firm is that by the time of Darius, a generation later, the mountain country west of the Indus was Persian. The great inscription at Behistun, carved around 519 BCE, lists Gandara — Gandhara, the region around Peshawar and Taxila — among the lands Darius rules.
It does not list Hindush.
But the later inscriptions do — the ones at Persepolis and at Darius's tomb, cut in the years after. So somewhere between roughly 518 and 513 BCE, Darius added a province called Hindush to the empire, and that province is on our side of the river.
We have no account of the campaign. No battle, no general, no date, no resistance recorded. A name appears on a later list that was not on the earlier one. That is the entire evidence for the conquest of Punjab by the largest empire on earth.
2.2 — What Hindush was
Here is the first serious problem, and it runs through this chapter.
We do not know where Hindush was.
The name means, roughly, "the Indus land". That could be the lower Indus — modern Sindh. It could be the west bank. It could be the whole Punjab plain. Persian inscriptions list provinces; they do not draw boundaries, because a Persian king thought in terms of peoples who paid him, not lines on a map.
Most historians put Hindush somewhere in the Indus valley and the western Punjab, with Gandhara separately to the north-west. Beyond that the honest answer is that the map in your textbook, with a neat shaded province and a firm eastern edge, is a modern invention. Nobody in 500 BCE had that map or wanted one.
2.3 — The heaviest tribute in the empire
Herodotus gives a list of what each province paid. The Indian province is at the end, and it is the largest figure on the list: three hundred and sixty talents of gold dust, when everyone else paid in silver.
A talent was a weight, roughly 26 kilograms. So three hundred and sixty talents of gold is somewhere around nine thousand kilograms — nine tonnes of gold, every year.
Gold was worth about thirteen times its weight in silver. Convert it and the Indian province's tribute comes out at roughly a third of the entire empire's income, from one province out of twenty.
To picture nine tonnes: it is about the weight of two full-grown male elephants, in gold, handed over annually.
And now the caution, which matters more than the number. We have no way to check this. Herodotus is a Greek writing a century later. The Persian administrative archives that survive — thousands of clay tablets from Persepolis — do not record it. Ancient tribute figures are routinely inflated, by the writers who report them and often by the kings who claimed them.
What the figure reliably tells us is not how much gold moved. It is that Greeks in the fifth century BCE believed India was fabulously rich. That belief is itself a historical fact, and it has consequences later in this part.
2.4 — The voyage down the Indus
One episode from this period is oddly vivid, and it is the clearest sign that Persian interest here was serious.
Darius, according to Herodotus, wanted to know where the Indus went. So he sent a Greek captain called Scylax with a fleet to sail down it, find the sea, and then come home the long way — west along the coast of Iran and Arabia and up to Egypt.
They did it. The voyage took thirty months.
Think about what that means as a piece of state behaviour. This is not raiding. This is an empire commissioning a survey of a river system at the far end of its territory, and then mapping the sea route back to its own centre. It is the act of a government that intends to stay.
It also matters for what comes next. Two centuries later Alexander sent his own fleet down the Indus and along that same coast, and he knew it could be done — because a Persian king had already had it done.
2.5 — Punjabi soldiers in Greece
The other firm thing is that men from here went west to fight.
When Xerxes invaded Greece in 480 BCE, Herodotus lists the contingents in his army, and there are Indians in it. He describes them wearing clothes made of "wool that grows on trees" — which is a Greek seeing cotton for the first time and doing his best. They carried cane bows and arrows tipped with iron.
Indian troops were still there the following year at the battle of Plataea, on the losing side.
Tribute: a regular payment from a subject region to an imperial centre. Not quite tax — tax implies you get government in return. Tribute is closer to protection money with a ceremonial name.
Aramaic: a Semitic language, unrelated to Persian, which the Persian empire used for its everyday administration across all its provinces — because it was already the common written language of the Near East and it used an alphabet rather than several hundred cuneiform signs.
Kharoshthi: a script used in Gandhara and north-west India from roughly the fourth century BCE. It is written right to left, and it is descended from the Aramaic alphabet the Persians brought. This is the single most durable thing Persia left here.
2.6 — What Persia left behind
So what did two centuries of being a Persian province actually do to Punjab?
The honest answer is: less than you would expect, with one large exception.
There is no Persian architecture here. No palace, no fire temple, no provincial capital in the Persian style, no garrison that anyone has excavated. There are no Persian administrative documents from this region — and the Persians were prolific record-keepers who left tens of thousands of tablets elsewhere. Compare that with what Rome left in Britain, which was a smaller province held for a shorter time, and the difference is stark.
The exception is writing.
The Persian empire ran on Aramaic, and Aramaic used an alphabet. Somewhere in the north-west, in the centuries after the Persians arrived, that alphabet was adapted to write Indian languages. The result is Kharoshthi — written right to left, like its parent, and used across Gandhara and Punjab for the next five hundred years.
Hold that against Part 3. The Rigveda has no word for writing at all. By the end of this part an emperor is having his personal opinions carved into rock in Kharoshthi, in the north-west, for ordinary people to read. Some of the machinery that made that possible came in through a Persian filing system.
Every map of the Achaemenid empire shades in a chunk of north-west India. What does that shading mean?
The name is on royal inscriptions in three languages, repeatedly, over decades. A tribute figure exists and is the largest on the list. Troops from here served in imperial armies as far away as Greece. Darius commissioned a naval survey of the Indus. Persepolis reliefs show a delegation identified as Indian bringing goods. And the administrative language left a permanent mark in the form of a script. Empires do not casually do those things to territory they do not hold.
Not one Persian building, garrison, coin hoard or administrative tablet has been excavated in this region. For an empire that left mountains of paperwork in Iran and Babylonia, that silence is deafening. The tribute figure comes from a Greek who never came here and cannot be checked. "Hindush" may mean only the lower Indus, not Punjab at all. And by the time Alexander arrived, whatever Persian authority existed had evaporated so completely that he had to negotiate with local kings one at a time, as if the empire had never been here.
What would settle it: Persian administrative material from the region. A single archive of tablets or Aramaic documents from Taxila or the Indus valley would change this overnight. Excavation in the relevant areas has been limited, and the levels of the right date at Taxila are thin.
Why it matters beyond this chapter: because the same question comes back for every empire in this series. The Mauryas, the Mughals, the British and the modern Indian state have all shaded Punjab in on their maps. Shading is cheap.
Underneath that whole disagreement sits something neither side usually states.
That an empire's map is its control.
Look at any historical atlas. Empires are solid blocks of colour with hard edges. Persia is purple, the Mauryas are orange, and the line between them is a line.
No ancient empire worked like that, and most modern ones did not either.
What an ancient empire actually consisted of was a set of relationships, of wildly varying strength, radiating out from a centre. Near the capital: real administration, garrisons, tax collectors, courts. Further out: a governor who mattered and a road that worked. Further out still: a local king who sent something every year and was otherwise left alone. At the edge: a name on a list because somebody's grandfather had once agreed to send something.
Punjab, for Persia, was at the far end of that scale. The purple on the map and the purple around Persepolis are the same colour and mean completely different things.
This is not pedantry. It changes the story. If you think Punjab was ruled by Persia for two centuries, then Alexander's arrival is a transfer of power between two foreign empires and the locals are furniture. If you think Punjab was a patchwork of local kingdoms that sometimes sent tribute west, then what Alexander walked into was a crowded, functioning political landscape of its own — which is exactly what the Greek sources then describe, without ever noticing that it contradicts the map.
Watch for this every time a map appears in the rest of this series. Ask what the colour is doing.
Between about 518 and 513 BCE Darius added a province called Hindush to the Persian empire. We have no account of the conquest — only a name that appears on later lists and not on earlier ones.
Herodotus says it paid the heaviest tribute in the empire, in gold dust. The figure cannot be checked and is probably inflated. What it reliably proves is that Greeks believed India was rich — which matters later.
Darius had the Indus surveyed by sea, a thirty-month voyage. Punjabi troops fought in the Persian army in Greece.
Persia left almost nothing physical here — no buildings, no garrisons, no archives. It left one enormous thing: the Aramaic alphabet, which became Kharoshthi, the first script used to write the languages of this region.
The purple block on the map is not control. It is a relationship, and at this distance from Persepolis it was a thin one.
3Taxila
One city in northern Punjab sat where three worlds met. It is also the site of the most repeated claim in Indian educational history — and that claim needs handling carefully.
3.1 — Where the roads met
Look at where Taxila is. About thirty kilometres north-west of modern Rawalpindi, on the Potohar plateau, just east of the Indus.
That position is the entire explanation for the city.
Three routes converge there. The road west through the Khyber and on to Kabul, Balkh and eventually Persia. The road north into Kashmir and over towards Central Asia. And the great road east and south-east, down through Punjab, across the rivers, and on to the Ganga plain — the route that later Indian texts call the Uttarapatha, the northern road, and that the Mauryas and the Mughals and the British all rebuilt in turn under different names.
A city at a junction like that does not need to produce anything. It gets rich on other people's traffic, and it collects other people's ideas as a side effect.
3.2 — What has actually been dug up
Taxila is not one city. It is three, in a row, built over about eight hundred years.
The earliest is a low mound now called Bhir Mound, occupied from roughly the sixth century BCE. This is the Taxila of this chapter — the city the Persians taxed, that Panini may have known, and that Alexander walked into.
Taxila was excavated over twenty years, from 1913, by John Marshall — the same archaeologist who was in charge when Harappa and Mohenjo-daro were identified.
What Bhir Mound looks like. Winding, irregular lanes. Houses of rough rubble stone, built up against each other with no shared plan. Soak pits for waste rather than street drains. Rooms added on as families needed them.
Now put that next to Part 2. Fifteen hundred years earlier, at Harappa, three hundred kilometres away, streets ran on a grid, bricks were made to standard proportions, and sewage ran under the road in covered channels.
What this proves: that Harappan city planning did not carry forward. It was not passed down and improved on. It was lost, and the cities that came afterwards in the same region were, by that measure, considerably less sophisticated.
What it does not prove: that Taxila was poor or backward. It was rich and it was important. It simply organised itself the way most ancient cities organised themselves — by accretion, not by plan. Harappa is the outlier in this story, not Taxila.
The two later cities on the site, Sirkap and Sirsukh, do have grid plans. Those come with the Greeks and the Kushans, in Part 5.
3.3 — The university question
You will have heard that Taxila was the world's oldest university. It is on tourist boards, in textbooks and in a great many speeches.
The claim needs unpicking, and unpicking it is not an attack on Taxila.
Everyone agrees Taxila was a famous centre of learning. The disagreement is about what kind of thing it was.
Buddhist story collections repeatedly describe young men travelling long distances — from Magadha, from Kosala, from Kashi — specifically to study at Takkasila. They name subjects: the Vedas, medicine, archery, law, accountancy, elephant lore. Tradition places several famous figures there, including a celebrated physician and, later, Chandragupta's adviser. Students paid fees or worked for their teachers. There was a recognised reputation that pulled people across a subcontinent. If a place attracts students from a thousand kilometres away for advanced instruction across many subjects, calling it a university describes what it did.
A university is an institution — it has a corporate existence apart from any individual, buildings held in common, admissions, a syllabus, examinations and qualifications that outlive its teachers. None of that has been found at Taxila, in the ground or in the texts. What the sources describe is many individual teachers, each taking private pupils in his own house, in a city that happened to be famous for having a lot of them. That is a scholarly quarter, not a university. And the comparison matters, because India does have an unambiguous ancient university — Nalanda, from around the fifth century CE, with excavated lecture halls, dormitories, a library, royal endowment and a formal admissions process. Calling Taxila a university flattens the difference and, oddly, sells Nalanda short.
What would settle it: excavation of a building complex at Bhir Mound identifiable as shared teaching space, or a text describing corporate governance. Neither exists.
Why the claim is so sticky: because "world's oldest university" is a much better sentence than "internationally renowned concentration of private teachers". It is not a lie. It is a translation of an ancient thing into a modern category it does not fit — which is the commonest way ancient history goes wrong, and it happens in every country's version of its own past.
3.4 — Panini
Somewhere in this region, probably in the fifth or fourth century BCE, a man from a village in Gandhara wrote a book. It is one of the most remarkable things ever produced by a human mind, and almost nobody outside a specialist field has read a page of it.
His name was Panini. The book describes the grammar of Sanskrit.
That sounds dull. It is not, and here is why.
Imagine trying to write down the complete rules of a language — not a phrasebook, but every rule by which any correct sentence can be built, with no exceptions left unexplained.
Panini did that for Sanskrit in under four thousand short rules, across eight chapters. The whole thing is roughly the length of a long magazine article.
He managed it by inventing techniques that nobody in the world used again until the twentieth century. He built an artificial notation — a set of technical symbols and abbreviations that mean nothing in ordinary Sanskrit — purely to compress the rules. He wrote rules that operate on other rules. He established a fixed order of application, so that when two rules conflict you know which one wins, without having to state the exception separately.
That is, in structure, how a computer language is specified. Panini was doing it around two and a half thousand years before there was anything to run it on.
The cost of that compression is that the text is unreadable without training. A rule can be two words long. Whole verbs are implied from a rule three chapters earlier. It has needed commentaries for over two thousand years, and the commentaries have needed commentaries.
Two things follow from Panini that matter for this book.
First, he is evidence of what was happening intellectually in the north-west in exactly this period. Not a court, not a temple, not an army — a man in a village near the Indus performing an act of abstract analysis at a level nobody would match for millennia.
Second, he tells you something about Sanskrit itself. By Panini's time Sanskrit had stopped being anybody's ordinary spoken language and was becoming a learned one, which is precisely why it needed fixing in place. The languages people actually spoke in Punjab by now were the Prakrits — the everyday descendants of the older speech. And it is the Prakrits, not Sanskrit, that Ashoka's inscriptions will be written in at the end of this part.
Uttarapatha: "the northern road". The great trunk route from the north-west across Punjab to the Ganga plain. Every empire in this series rebuilds it; the Grand Trunk Road is its descendant.
Sutra: literally "thread". A rule or statement compressed to the absolute minimum number of words, designed to be memorised. Indian technical works of this period are written as strings of them.
Prakrit: the ordinary spoken languages of northern India in this period, descended from the same ancestor as Sanskrit but changed by everyday use. Punjabi's ancestors are among them. Sanskrit was the fixed, learned form; Prakrit was what people actually said.
Taxila sat where the roads from Persia, Central Asia and the Ganga plain met. That junction is the whole explanation for the city.
The earliest city there, Bhir Mound, has winding lanes, rubble houses and soak pits — noticeably less planned than Harappa fifteen hundred years earlier. Harappan city planning was lost, not inherited.
It was a famous centre of learning, drawing students across the subcontinent. It was not a university — no institution, no buildings held in common, no syllabus. Nalanda, much later, is the real thing.
Panini, from a village in this region, wrote the complete grammar of Sanskrit in under four thousand rules, using formal techniques nobody matched for two thousand years.
By Panini's time Sanskrit already needed preserving. What people around him actually spoke were the Prakrits — and those are the ancestors of Punjabi.
4Alexander Arrives
In the spring of 326 BCE a Macedonian army crossed the Indus on a bridge of boats. Before we follow it, we have to be clear about who was writing all this down, and when.
4.1 — Who he was
Alexander became king of Macedon in 336 BCE, at twenty, when his father was murdered. Macedon was a hard, mountainous kingdom on the northern edge of the Greek world, which the southern Greeks regarded as half-barbarian and which had recently conquered them.
Two years later he crossed into Asia and attacked the Persian empire — the same empire that had held Gandhara and Hindush in Chapter 2. He beat it three times in pitched battle, burned Persepolis in 330 BCE, and by the time the last Persian king was murdered by his own men, Alexander was the ruler of everything from Greece to Afghanistan.
Then he kept going. Three brutal years in Bactria and Sogdiana — modern northern Afghanistan and Uzbekistan — fighting an insurgency rather than an army. And in 327 BCE, having inherited the Persian claim to Gandhara and Hindush, he came over the mountains to collect it.
He was twenty-nine.
4.2 — The sources, and why they are a problem
Everything that follows in the next three chapters comes from Greek and Roman writers. It is worth stopping to see how that chain works, because it is the weakest link in this part and almost nobody mentions it.
Not one eyewitness account of Alexander's campaign survives. Not one.
Several were written. His fleet commander wrote one. His official historian wrote one, until Alexander had him arrested and he died in custody. His engineer wrote one. And one of his generals wrote a memoir after becoming king of Egypt.
All of them are lost. We know they existed because later writers quote them.
The five accounts we actually have were written between three hundred and five hundred years after the events, by men who had never been to India, working from those lost books plus a great deal of accumulated story.
To feel the gap: it is like reconstructing the Anglo-Sikh Wars of the 1840s from books written today, when every first-hand account has been destroyed and all you have are quotations from them inside later works.
What follows from this. The best of the five is generally reckoned to be the one that leans hardest on the two eyewitness sources. But one of those eyewitnesses was a general who later became a king and had every reason to make his own role look good, and the other served in Alexander's court. The official historian was executed for insufficient enthusiasm, which tells you what kind of writing was expected.
What can still be trusted: the broad sequence — where the army went, in what order, roughly when. Multiple independent chains agree, and the geography checks out.
What cannot: numbers of any kind, speeches, motives, and above all casualty figures. Every one of those is a place where a writer had freedom and a reason to use it.
4.3 — Over the mountains
Alexander crossed the Hindu Kush and split his army in two.
Half went straight down the direct route towards the Indus — through the Khyber country — with orders to secure the road and build a bridge over the river.
Alexander took the other half north, into the hill country of Bajaur and Swat, and spent a winter reducing it. This was not a march. It was a series of sieges against hill peoples who had no intention of submitting to anybody, and the sources describe it as vicious on both sides.
One episode from it should be stated plainly rather than skipped, because the popular version of Alexander does not include it.
At a hill fort called Massaga, the defenders included several thousand Indian mercenary soldiers. After a hard siege they agreed terms and came out. According to more than one of our sources, Alexander then had them surrounded and killed — because he judged that if released they would simply fight him again further on. The figure given is around seven thousand.
Our sources are, remember, sympathetic to him. This is what got through a friendly filter.
4.4 — The rock
The set piece of the hill campaign is a mountain stronghold the Greeks called Aornos.
It was a flat-topped height above the Indus with steep sides, water on top and enough ground to grow food — meaning it could not be starved out and could not easily be climbed. Local tradition, the Greeks were told, held that a god had once tried to take it and failed.
That last detail is the point. Alexander built his whole public image on descent from Herakles and Achilles. Being told that Herakles had failed here made the place irresistible, and the sources say so openly.
He took it by engineering. His men filled a ravine with earth and timber to build a causeway across to the rock face, under fire, over several days, until his catapults were in range. The defenders slipped away at night.
It was a genuinely impressive operation and it had almost no strategic value. He wanted it because of the story attached to it.
Phalanx: the Macedonian infantry block. Men packed shoulder to shoulder in deep ranks, each carrying a pike about six metres long, so that five ranks of pike-points projected ahead of the front man. Almost unstoppable from the front on flat, firm ground. Clumsy and vulnerable everywhere else — which matters a great deal in the next chapter.
Companion cavalry: Alexander's heavy horse, and his own personal command. This was the weapon he actually won battles with: hold the enemy in place with the phalanx, then drive the cavalry into a gap.
Siege train: the catapults, towers and engineers that travelled with the army. This was the real Macedonian advantage over most opponents, and it is why hill forts that had never fallen before started falling.
4.5 — Taxila changes sides
In the spring of 326 BCE the army crossed the Indus on a bridge of boats and came down into the Punjab plain.
And there Alexander was met not by resistance but by an ally.
The king of Taxila — the Greeks called him Taxiles, from the city; his Indian name appears as Ambhi — came out to meet him, submitted, and supplied the army with grain, silver, cattle, elephants and several thousand troops.
This is usually written up as cowardice or treachery. It was neither. It was a local political calculation, and to see it you have to forget the map from Chapter 2 and look at the actual landscape.
Punjab in 326 BCE was not a country. It was a set of adjacent kingdoms and republics along and between the five rivers, most of them small, all of them wary of each other. Taxila's principal enemy was the kingdom across the Jhelum, ruled by the man the Greeks would call Porus.
From Ambhi's chair, a Macedonian army arriving from the west was not an invasion of India. India was not a thing anybody was defending. It was an enormous, well-equipped, temporary force that could be pointed at his neighbour.
He was not wrong about the immediate result, and he was not the last ruler in this series to make that calculation. Watch for it again in Part 6, and again in Part 10, and most consequentially of all in Part 11.
Alexander came to India in 327 BCE as the conqueror of Persia, collecting what had been Persian provinces. He was twenty-nine.
Not one eyewitness account survives. Everything we have was written three to five hundred years later, from lost memoirs by men with reputations to protect. Trust the sequence and the geography. Do not trust the numbers, the speeches or the casualty figures.
The hill campaign in Swat was brutal, and included the killing of several thousand Indian mercenaries after they had surrendered — reported by sources friendly to him.
He took the rock of Aornos by building a causeway across a ravine, for a place of almost no strategic value, because he had been told a god once failed to take it.
Taxila joined him without a fight. Not out of cowardice — because Punjab was a set of rival kingdoms, and a foreign army was a weapon you could aim at your neighbour.
5Porus at the Jhelum
Part 1 promised you this battle in full. It was fought in the monsoon on the Vitasta, it was the hardest fight of Alexander's career, and the man who lost it kept his kingdom.
5.1 — The river in flood
The Greeks called it the Hydaspes. The Rigveda called it the Vitasta. We call it the Jhelum.
In May 326 BCE, Alexander's army came up to its western bank and stopped, because on the other side there was an army waiting and the river was rising.
The timing was the worst possible. The monsoon was arriving. A Punjab river in flood is not the river you see in the dry season — it widens, the current strengthens, and the banks turn to deep mud. Every crossing point was watched.
And the man watching was the king of the country between the Jhelum and the Chenab. The Greeks wrote his name as Poros. It is very likely a rendering of Paurava — a man of the Puru lineage.
If that is right, and most scholars think it is, then this is a thread running straight out of Part 3. The Purus were one of the peoples who fought against Sudas at the Battle of the Ten Kings, on the Ravi, eleven centuries earlier. Their descendants were still here, still holding land between two of the same rivers, when a Macedonian army walked into it.
5.2 — The night crossing
Alexander could not force a crossing in front of Porus's army. Elephants terrify horses, and men climbing wet out of a river into a waiting line do not last.
So he cheated, and the way he did it is the most impressive thing he did in India.
For several weeks he made his army noisy and unpredictable. Cavalry rode up and down the bank at night. Camps were struck and rebuilt. Boats moved about. Grain was stockpiled openly, as though he intended to sit there until the monsoon passed. Porus shadowed each movement from the other side, and then — as the pattern repeated and nothing happened — stopped shadowing every one of them.
Then, on a night of heavy rain and thunder, Alexander took a picked force some twenty-five kilometres upstream, to a wooded bend where an island screened the far bank, and crossed. The storm covered the noise.
The rest of the army stayed in the main camp, in full view, and made a great deal of noise as if preparing to cross where they were.
By the time Porus understood what had happened, there was a Macedonian force on his side of the river and behind his flank, and he had to decide in minutes whether to face it or hold his position against the main camp.
5.3 — Elephants in the rain
Porus sent a detachment under his son to deal with the crossing. It was too small and it was destroyed. His son was killed in it.
Then he turned his main force to face Alexander, and put his elephants in the front line at intervals, with infantry between and behind them.
What happened next is the only part of this battle everybody agrees was hard.
To understand the fight you need to understand what an elephant does to a Macedonian phalanx.
The phalanx works because a wall of six-metre pikes is impossible to reach past. It requires two things: flat firm ground, and an enemy that comes at you from the front.
An elephant is four tonnes of animal that walks into the pikes and keeps walking. The points do not stop it. Worse, horses will not go near one — they panic at the smell — so the cavalry that normally wins the battle cannot be used against them.
Now add the weather. The bank was churned mud after weeks of rain. Porus's chariots bogged down and were useless. But so did the phalanx's footing, and a pike block that cannot hold its line is just a crowd of men holding long sticks.
The Macedonians eventually broke the elephants by going round them — axes and heavy curved swords at the legs and trunks, javelins at the drivers. It worked, and it produced the worst part of the day for the Indian side: wounded elephants panicked and turned back into their own infantry, which was packed in behind them.
More Indians were probably killed by their own elephants than by Macedonians.
The fight went on for hours. Porus stayed on the field, on an elephant, and was wounded in the shoulder. He was among the last to leave. Even the sources written to flatter Alexander say so, and they did not have to.
5.4 — How close it ran
The written casualty figures are worthless and it is easy to show why. One account gives the Indian dead as twenty-three thousand and the Macedonian dead as about three hundred and ten. That is not a battle. That is a massacre, and it does not fit a fight the same source describes as lasting eight hours against a general who held his position to the end.
But there is a second kind of evidence, and it is much better, because it was not written by a historian.
Alexander struck coins about it. Large silver pieces have survived showing, on one side, a horseman charging a retreating elephant carrying two figures, and on the other a standing figure with a thunderbolt. Numismatists connect them directly to this campaign.
What that proves: the battle mattered enough to be commemorated in metal, and the thing worth commemorating was beating an elephant. Nobody issues a coin about an easy win.
There are two further pieces of physical logic. Alexander founded two cities on the spot, one of them named for the horse that died there — you do not do that for a routine engagement. And within a few weeks his army refused to go any further east. Armies that have just won cheaply do not usually stop.
What it does not prove: that the result was in doubt. Alexander won, clearly, and captured Porus. "Harder than expected" is not the same as "nearly lost".
5.5 — "As a king"
Porus was brought before Alexander wounded, and — the sources make a point of it — still an enormous man, taller than anyone around him.
Alexander asked how he wished to be treated.
Alexander then did something that had nothing to do with generosity and everything to do with strategy. He gave Porus his kingdom back, and added territory to it.
He had no choice worth the name. He had one army, thousands of kilometres from home, and no administration. A hostile province behind him was a knife at his supply line. A local king who owed him his throne was the cheapest possible garrison. Every empire in this series works out the same thing eventually.
But it also happens to have made a very good story, and the story is why Porus is the one Indian name most people can produce from this whole period. He is remembered because he lost well to somebody famous — which is a thin kind of immortality, and it is worth noticing who hands it out.
The horse died here too. Alexander had ridden Bucephalus since he was a boy, and named a city for him on the bank of the Jhelum. Nobody has ever found it.
That the enemy in a victory story is a real portrait.
Everything you know about Porus comes from people writing to make Alexander look good. Everything. There is no Indian source about him at all — not a text, not an inscription, not a coin of his own.
Now notice which Porus those writers produced. Huge. Brave. Dignified in defeat. Answers a conqueror with one perfect line. Fights to the last and is admired for it.
Ask what that portrait is doing. A hero is only as good as the man he beat. If Porus is a coward, Alexander has beaten a coward. If Porus is a magnificent warrior-king with a noble reply ready, then Alexander has defeated a worthy enemy and then been generous to him — which makes Alexander twice the man.
The noble defeated enemy is one of the oldest devices in war writing, in every literature that has one. It is a compliment paid to the winner in the currency of the loser.
So what does that leave? Not nothing. Two things survive the discount, because they are not the kind of detail a flatterer needs to invent: that the battle was long and difficult, and that Alexander put Porus back on his throne. The second is administrative and can be checked against what happens afterwards.
The rest — the height, the dignity, the line — may be true. We have no way of knowing, and we never will. What we are actually reading is what Greeks needed an Indian king to be like.
This applies far beyond Porus. Every noble savage, every worthy adversary, every honourable enemy in any conquest narrative anywhere is doing the same job. When you meet one, ask whose reputation is being built.
The battle was fought on the Jhelum — the Vitasta of Part 3 — in May 326 BCE, in the monsoon. "Porus" is almost certainly Paurava, a man of the Puru lineage, whose people fought against Sudas on the Ravi eleven centuries earlier.
Alexander won by deception: weeks of false alarms until Porus stopped reacting, then a night crossing twenty-five kilometres upstream under cover of a thunderstorm.
Elephants nearly did the job. Horses will not face them and pikes do not stop them. What broke the Indian line was wounded elephants turning back into their own packed infantry.
The written casualty figures are worthless. The coins Alexander struck about beating an elephant are better evidence, and so is the fact that his army refused to go any further a few weeks later.
Porus got his kingdom back — because a grateful local king is the cheapest garrison there is. Everything else you know about him was written by people who needed Alexander to have beaten someone worth beating.
6The Refusal at the Beas
Three rivers further east, the army that had not lost a battle in eight years stopped walking. What turned Alexander back was not an enemy. It was a rumour, and his own men.
6.1 — Three more rivers
After the Jhelum, Alexander went on east. He crossed the Chenab. He crossed the Ravi — the Parushni of Chapter 8 of Part 3, in the same country as the Battle of the Ten Kings.
It was not a triumphal procession. At a place the Greeks called Sangala, somewhere in central Punjab, a people they name as the Cathaeans fought a hard defensive battle from behind a barricade of wagons and had to be taken by siege. Our sources — friendly ones — record heavy Macedonian losses.
Then the army reached the Beas, the Vipas, and stopped.
6.2 — What they had heard was on the other side
The reason they stopped was information, and it came from the local kings.
Beyond the Beas, they were told, the country ran on across the Ganga plain to a kingdom called Magadha, ruled by the Nanda dynasty from a capital far to the east. And the Nandas could put an army into the field that made anything in Punjab look small.
The figures the Macedonians were given vary by source, but the smallest version is roughly: two hundred thousand infantry, twenty thousand cavalry, two thousand chariots, and between three and four thousand elephants. The largest version is more than double that.
Discount them as heavily as you like. Cut them by three quarters and they are still catastrophic.
The elephants are the part that matters. At the Jhelum, two hundred elephants had come close to breaking the Macedonian line and had left the army badly shaken. Now they were being told about numbers ten or twenty times greater.
And consider what these men had already done. They had crossed from Europe into Asia eight years earlier. They had walked something over seventeen thousand kilometres. Many had married, had children in garrison towns, and had not seen home since their twenties.
The monsoon had been falling on them for weeks. Their leather was rotting, their iron was rusting, and there were snakes in the camp.
They were not cowards. They were finished.
6.3 — Coenus speaks
Alexander called an assembly of his officers and made the case for going on. The speech we have is a later writer's composition, but the situation is not in doubt.
One man answered him: Coenus, a senior commander who had led a decisive wing at the Jhelum and could not be accused of shirking.
The reply attributed to him is, in substance, that the men had done everything asked of them, that most of those who set out were dead or scattered in garrisons, that those left were worn out, and that a good commander knows when to stop. Then he said the thing that actually ends arguments: he asked Alexander to go home and, if he wished, to come back later with a fresh army.
The assembly was silent. Alexander went into his tent for two days, in the hope that the mood would turn.
It did not. He came out, announced that the omens for crossing were bad, and gave the order to turn back — which allowed everybody to pretend the gods had decided rather than the infantry.
Coenus died of illness within weeks. Alexander gave him a grand funeral. Some of the sources note the coincidence and let the reader draw the conclusion.
Before leaving, Alexander had twelve enormous altars built on the bank of the Beas, one for each of the Olympian gods, to mark the limit of his advance. They were described for centuries afterwards.
Nobody has ever found them.
6.4 — The road home and what it cost
The return was worse than anything in the advance.
The army went south down the Jhelum and the Chenab into the Indus, with a fleet built in Punjab carrying part of it. On the way they fought their way through the peoples of the lower Punjab — and at one town of the Malli, almost certainly on the site of Multan, Alexander went over the wall ahead of his men, was cut off inside with a handful of companions, and took an arrow through the chest that went into his lung.
He nearly died. The reprisal against the Malli afterwards was, even in our friendly sources, a slaughter.
At the mouth of the Indus the force split. The fleet felt its way west along the Makran coast — a genuine feat of navigation. Alexander marched the rest of the army through the Gedrosian desert behind it, and that march killed more of his people than every battle of the entire campaign put together. Most of the dead were the camp followers: the wives, children, servants and traders who moved with an ancient army and are almost never counted.
He reached Babylon and died there in June 323 BCE, at thirty-two, of a fever, or of drink, or of poison, depending on which of our late sources you prefer.
He had been in India for about two years.
6.5 — Why India does not remember him
That importance is a property of events.
Here is a fact that should be much better known than it is.
No Indian source of any kind mentions Alexander. Not a Sanskrit text, not a Prakrit one, not a Buddhist or Jain chronicle, not a Puranic king-list, not an inscription, not a coin. Nothing. For two thousand years, the single most famous event in this period of Indian history left no trace whatsoever in Indian memory.
Everything you know about it, including this chapter, comes from Greek and Roman writers and from Europeans reading them.
The near-miss is almost funny. Ashoka's inscriptions, at the end of this part, do name five Greek kings — he was writing to them. One of the names looks like Alexander. It is a different Alexander, a minor contemporary king in north-western Greece. The great one is not there.
So which is right? Was Alexander's invasion a turning point in Indian history, or a two-year raid that nobody bothered to write down?
The answer is that "important" is not a fact about an event. It is a decision made later, by whoever is keeping the records.
From a Greek desk, this is the moment the known world reached its eastern limit — the climax of the greatest story in their history. From an Indian desk, it is a foreign army that came over the mountains, fought a hard battle with one Punjabi king, was stopped by its own men at the fourth river, went home the long way and lost most of its people in a desert. Neither desk is lying. They are ranking.
And notice the machinery that made one ranking win. The Greeks wrote it down; Rome inherited it; Europe inherited Rome; and Europe wrote the history textbooks that were then used in Indian schools. That is a chain of custody, not a verdict.
Then ask the question the other way round, because it is the useful one. What has been left out of your education by the same process? The next three chapters are about a man who built an empire from Afghanistan to Bengal, and an emperor who covered a subcontinent in writing. Both are better documented than Alexander in India. Neither is as famous.
At the Beas in 326 BCE, Alexander's army refused to go further east. Not defeated — finished. Eight years, seventeen thousand kilometres, and weeks of monsoon.
What stopped them was a report: the Nanda empire of Magadha, with elephants numbered in the thousands. Two hundred had nearly broken them on the Jhelum.
Coenus spoke for the men and Alexander gave way, announcing that the omens were bad so that the gods could take the blame. Coenus died within weeks.
The retreat cost far more than the advance. Alexander took an arrow through the lung at Multan, and the march through the Gedrosian desert killed more of his people than all his battles combined. He died at Babylon in 323 BCE, aged thirty-two.
No Indian source mentions him. Not one, ever. His fame here is entirely inherited from the people who wrote it down — which is what fame usually is.
7What He Actually Left
Within about ten years of Alexander's death there was no Macedonian authority left east of the Indus at all. But something was left, and it was not what he intended.
7.1 — The garrisons collapse
Alexander did what conquerors do. He appointed governors, left garrisons of Greek and Macedonian troops, founded cities named after himself, and marched away.
It came apart almost immediately, and the sequence is worth following because it tells you how thin the conquest was.
The governor he left in charge of the country between the Indus and the Jhelum was murdered by his own mercenaries while Alexander was still alive. Another senior officer was killed in fighting shortly after. The Greek troops settled far to the north, in Bactria, mutinied twice and tried to walk home.
Then, around 318 BCE, the Greek commander left in India murdered Porus — the king Alexander had personally restored — seized his war elephants, and marched them west to sell his services in the wars between Alexander's generals.
That is the end of the Macedonian presence in Punjab. Not a defeat. A commander stripping the province for assets and leaving.
Of the cities Alexander founded here, not one has been securely identified on the ground. The city named for his horse, on the Jhelum, is lost. The altars on the Beas are lost. Two years of campaigning, and the physical trace in Punjab is effectively zero.
7.2 — What did survive
Four things outlasted him, and only one of them was on purpose.
A word. Indian languages acquired yavana as the term for Greeks, and then for westerners generally. It comes, through Persian, from "Ionian" — the Greeks of the Anatolian coast the Persians had dealt with. Panini already knew a form of it before Alexander arrived, so the word came the same way the alphabet did, through the Persian empire. It is still there in modern Indian languages.
A sea route. Nearchus's fleet proved that the coast from the Indus to the Persian Gulf could be navigated by a large force. That knowledge fed, over the following centuries, into a serious maritime trade between India and the Mediterranean world.
Elephants. The Macedonians had been badly frightened by them on the Jhelum, and the lesson took. For the next century every Hellenistic king wanted elephants, and the only reliable supply was India. That has a direct consequence in the next chapter.
A Greek population in the north. The settlers Alexander planted in Bactria did not go home. Their descendants set up their own kingdom, and a century later some of them came back over the mountains into Punjab as rulers in their own right. That is Part 5, and it is where the genuine Greek influence on this region actually happens — long after Alexander, and nothing to do with him.
Within a decade of Alexander's death, a man from Magadha had taken the whole of north India including Punjab. Historians have argued about the connection for two hundred years.
Alexander removed, killed or destabilised the local powers of Punjab one by one, then left a vacuum garrisoned by demoralised foreign troops who murdered their own officers. The region's kings had been beaten, its confederations broken and its manpower drained. A conqueror coming from the east a few years later walked into something far softer than the crowded, well-armed landscape of 327 BCE. Greek sources also say the future emperor met Alexander as a young man and drew a lesson from it.
The Nanda kingdom was the largest state in India before Alexander arrived and was exactly what stopped him at the Beas. Its wealth, its standing army and its elephants were built on the iron and rice agriculture of the Ganga plain that Part 3 described — a process that had been under way for four hundred years and owed nothing to anybody Greek. Chandragupta's real achievement was taking Magadha itself, in the east, which Alexander never came within eight hundred kilometres of. Punjab was a mopping-up operation afterwards.
What would settle it: almost nothing, because it is a question about what would have happened otherwise, and there is no evidence for things that did not happen. Be suspicious of confident answers in either direction.
Why the question keeps being asked: because European historiography needed Alexander to have mattered, and Indian nationalist historiography needed him not to have. Both had an answer before they had a method.
7.3 — The trap of the turning point
Let me be plain about the shape of these two chapters, because I have been hard on a man most readers admire.
Alexander was an extraordinary soldier. The Jhelum crossing was brilliant. The Aornos causeway was brilliant. He asked things of an army that almost no commander has ever got, and got them for eight years.
None of that is the same as mattering here.
The habit this period should break you of is the turning point — the idea that history proceeds by dramatic moments in which a famous person changes everything. It is the shape stories have, so it is the shape we impose.
The things that actually changed Punjab in this part were slower and duller and had no author. An alphabet arriving through an imperial filing system. Iron and rice building a state in a river plain eight hundred kilometres east until it was strong enough to reach back. A road across the passes that made one city rich. A grammar written in a village. None of those has a battle in it, and all of them outlasted every army in this chapter.
Macedonian authority in Punjab lasted about ten years and ended not in battle but in a commander murdering Porus, seizing his elephants and marching them west to sell.
Not one of the cities Alexander founded here has been securely identified. The altars on the Beas are gone. The physical trace in Punjab is effectively nothing.
Four things survived: the word yavana, the sea route to the Gulf, a Hellenistic appetite for elephants, and a Greek population in Bactria whose descendants return in Part 5. Only the last is a real inheritance, and it happened a century later.
Whether he made the Mauryan empire possible is disputed. Most historians say the engine was in Magadha and had been running for centuries — but Punjab was certainly easier to take in 317 BCE than in 327.
Beware the turning point. What actually changed this land was an alphabet, a road, a grammar and a crop — none of which has a hero in it.
8Chandragupta
A man whose birth nobody can agree on took the largest kingdom in India, threw the Macedonians out of Punjab, and built the first empire that ran from Afghanistan to Bengal. He did it in about twenty years.
8.1 — A man of uncertain birth
We do not know where Chandragupta came from, and the disagreement is instructive.
Buddhist tradition makes him a member of a recognised warrior clan called the Moriyas. Jain tradition makes him the son of a village headman whose people kept peacocks. The Brahmanical texts and a much later Sanskrit play make him low-born and say so with some contempt. A Roman writer, summarising a Greek one, says simply that he was of humble origin.
Notice the pattern. Each tradition gives him the origin that suits its own view of who ought to be allowed to rule. The Buddhists and Jains, who did not accept birth as the basis of authority, are relaxed about a modest start. The Brahmanical sources, which did, treat it as a scandal.
Nobody in any of these sources is reporting. Everybody is arguing.
What survives the discount is thin but useful: he was not from an established royal house, and everybody agrees on that even while disagreeing about what it means.
8.2 — Chanakya, and a book that may not be his
Attached to Chandragupta in every version of the story is an adviser: a Brahmin from Taxila, variously called Chanakya, Kautilya and Vishnugupta.
The tale is a good one. He is humiliated at the Nanda court, swears revenge, finds a promising boy, trains him, and engineers the destruction of a dynasty through patience and manipulation. It has been told in India for two thousand years and is still being told.
Attached to that adviser, in turn, is a book: the Arthashastra, a manual of statecraft. And here we need to be careful.
The Arthashastra was lost. For centuries it was known only as a title mentioned in other works. Then in 1905 a manuscript on palm leaf was handed to a librarian in Mysore, who identified it and published it. That is how recently this text re-entered the world.
What it contains. A staggeringly detailed system of government. How to set tax rates by crop and by district. How to run an intelligence service, including the use of merchants, ascetics and prostitutes as informants. How to manage irrigation, mines, forests and brothels. How to test the loyalty of a minister by having an agent try to corrupt him. How to poison a rival and how to avoid being poisoned. It is unsentimental to the point of coldness and there is nothing else like it from the ancient world.
What the dating actually shows. The text names its author as Kautilya. It does not name Chandragupta. A statistical study of its language in 1971 concluded it was the work of more than one hand, and most specialists now hold that the version we have was compiled and finalised several centuries after the Mauryas — possibly around the second century CE.
What that means: the Arthashastra cannot be used as a description of how Chandragupta actually governed. It is not a memoir and it is not a record. Some of its material may go back to Mauryan practice; there is no way to tell which parts.
And what it does prove: that somebody in ancient India worked out, in extraordinary detail, how a bureaucratic state should function. That is a real fact about Indian political thought, and it does not need to be tied to a famous name to be impressive.
Which raises a problem far bigger than one book.
That a text named after a person was written by that person.
This is one of the great quiet distorters of ancient history, and it is not remotely confined to India.
In the ancient world, attaching a famous name to a work was normal, respectable behaviour. It was how you signalled a tradition. A pupil writing within a master's school might sign the master's name. Later editors compiling generations of accumulated material would credit the founder. This is why we have medical works "by Hippocrates" written centuries after he died, and legal texts "of Manu" assembled over hundreds of years.
The assumption does two kinds of damage.
It compresses time. A body of thought developed over four hundred years by many people gets collapsed into one genius in one lifetime — which flatters the individual and erases everyone else.
And it lends borrowed authority. "This is how the Mauryan empire was run" is a much stronger claim than "this is how some later thinkers believed a state ought to be run". The name does that upgrade for free, and nobody notices it happening.
Apply the test as we go. When you meet a text tied to a famous name in this series — and you will, repeatedly, in Parts 6 through 9 — ask two questions. Does the text itself claim that authorship, or did somebody else attach it later? And how much of the argument depends on the name being right?
In this case: Chanakya may well have existed. The Arthashastra is a genuine masterpiece. The link between them is tradition, not evidence, and almost every popular account of the Mauryan empire treats it as fact.
8.3 — Taking Magadha, and taking Punjab
Around 321 BCE Chandragupta overthrew the Nanda dynasty and took Magadha — the kingdom whose reported army had turned Alexander's men back at the Beas five years earlier.
That was the hard part, and it is the part Greek sources know least about. He was taking on the strongest state in India with its own standing army, and he won.
Then he turned north-west. By about 317 BCE the Macedonian governors were dead or gone and Punjab was his. Taxila became the seat of a viceroy — a post held, in the following generation, by royal princes. For the next century and a half Punjab was not an independent set of kingdoms and not a distant province of a foreign empire. It was the north-western wing of an Indian one, governed from a city inside it.
8.4 — Five hundred elephants
Around 305 BCE the man who had inherited the eastern half of Alexander's empire — Seleucus — came east to take India back.
We have no account of the fighting, which usually means it did not go well for the side whose historians we are reading. What we have is the settlement, and the settlement is unambiguous.
Seleucus gave up the eastern provinces: Kandahar, the Kabul valley, Balochistan. In return he received five hundred war elephants and, apparently, a marriage alliance.
Look at the shape of that deal.
Seleucus handed over territory — several provinces of what is now Afghanistan and Pakistan, permanently — and received animals.
In any negotiation, the side that gives up land and takes movable goods is the side that lost. This was not a partnership of equals. Chandragupta pushed the Indian frontier several hundred kilometres west of where any Indian ruler had held it, and it stayed there.
And Seleucus knew exactly what he was buying. Four years later, at a battle in Anatolia against a rival for Alexander's inheritance, those Indian elephants were used to cut the enemy cavalry off from its own infantry. Seleucus won, and with it he won the largest share of Alexander's empire.
So: Punjab's elephants decided who inherited Alexander. That is a far more concrete Indian contribution to Greek history than anything Alexander contributed to ours.
8.5 — Megasthenes at Pataliputra
Seleucus then sent an ambassador, Megasthenes, who lived at the Mauryan capital for years and wrote a book about India.
That book is lost too. What survives are quotations of it inside later Greek and Roman authors — and, as with everything in this part, the quoters had their own purposes.
But some of it can be checked, which is rare and worth savouring.
Megasthenes describes the capital, Pataliputra, on the Ganga: a very long, narrow city defended by a wooden palisade with hundreds of towers and dozens of gates, behind a wide moat. For a long time this was treated as traveller's exaggeration.
Then excavation at Patna found massive wooden posts in the ground, in a line, of the right period.
He is right about other things too — the road network with distance markers, the scale of the administration, the emperor's personal security. He is wrong or garbled about plenty: he divides Indian society into seven rigid classes, which matches nothing in any Indian source, and he states flatly that there is no slavery in India, which is contradicted by Indian texts.
A foreign observer at a court, then: excellent on what he could see, unreliable on what had to be explained to him. Which is what foreign observers usually are, in every century of this series.
8.6 — The end at the rock
Chandragupta's death is the strangest thing about him.
Jain tradition holds that towards the end of his life he gave up the throne to his son, became a Jain monk, travelled south with his teacher, and died at Shravanabelagola in Karnataka by sallekhana — the Jain practice of fasting to death when the body's usefulness is over.
There are inscriptions at that site associating the place with his teacher and with a disciple named Chandragupta, though they were cut many centuries later. No non-Jain source mentions any of it.
So: unconfirmed, not impossible, and told by the tradition that would most want it to be true. I include it because it is part of how he is remembered, and I flag it because it is not evidence.
What is firm is that he ruled for roughly twenty-four years, that his son succeeded him without the empire falling apart, and that his grandson inherited something that ran from Kandahar to Bengal.
Nobody agrees on Chandragupta's birth, and each tradition gives him the origin that suits its own view of who may rule. All agree he was not from an established royal house.
The Arthashastra is a genuine masterpiece of political thought, lost until 1905. It is not a description of Mauryan government — the surviving version is composite and centuries later. Its link to Chanakya, and Chanakya's to Chandragupta, is tradition rather than evidence.
Around 321 BCE he took Magadha — the kingdom that had frightened Alexander's army off at the Beas. By 317 BCE Punjab was his, governed from Taxila.
Around 303 BCE Seleucus gave up Kandahar, Kabul and Balochistan in exchange for five hundred elephants. Land for animals is what losing looks like. Those elephants then won Seleucus the largest share of Alexander's empire.
The most consequential figure in this part is the one you were taught least about.
9Ashoka's Rocks
For the first time in this series, somebody standing in this land writes something down and leaves it where ordinary people can read it. He is the emperor, and what he writes about is his own remorse.
9.1 — Kalinga
Chandragupta's grandson came to the throne around 268 BCE. His personal name was Ashoka. In his own inscriptions he almost never uses it; he calls himself Devanampiya Piyadasi, "Beloved of the Gods, of gracious mien".
About eight years in, he attacked Kalinga — roughly modern Odisha, on the east coast — which was the last substantial region of the subcontinent outside Mauryan control.
He won. And then he did something no ruler in the ancient world had done before or, as far as I know, has done since in quite this way.
He had the casualty figures carved in stone, across his empire, along with a statement that he was ashamed.
This is the best-evidenced thing in this entire part, and it is worth being clear about why.
Everything else in Part 4 is second-hand: Greeks quoting lost Greeks, chronicles written centuries later, traditions with an interest. Ashoka's edicts are the thing itself. Carved in his lifetime, on his order, in his own voice, in dozens of places from Afghanistan to Karnataka, and still legible.
The thirteenth Major Rock Edict gives the Kalinga figures directly: a hundred and fifty thousand deported, a hundred thousand killed, and many times that number dead by other causes. Then it says the Beloved of the Gods felt remorse, that the slaughter and deportation weigh heavily on his mind, and that he now considers the only true conquest to be conquest by dhamma.
What this proves: that a Mauryan emperor caused these words to be cut into rock in multiple provinces. That is a fact about the eleventh year of his reign, and it is unusually secure.
What it does not prove: that the figures are accurate — they are the state's own numbers about its own war — or that the feeling behind them was what the words say. An inscription is evidence of what was said, not of what was meant.
The edicts also give us something rarer: a date anchor. One of them names five Greek kings Ashoka says he sent messengers to. All five can be independently dated from Mediterranean sources, and all five were on their thrones at the same time in the 250s BCE. That single sentence fixes the whole of early Indian chronology to within a few years — the only such anchor we have for the period.
9.2 — What the edicts actually say
Most people know the story of the conversion and almost nobody has read the edicts. They are short, plain and not at all what you would expect.
The first surprise: they are mostly not about Buddhism.
Ashoka's own Buddhist commitment is stated — one of the minor edicts says plainly that he became a lay follower and then drew closer to the monastic order. But the great public edicts, the ones cut into rock faces on roads, are addressed to everybody, and the thing they preach is called dhamma, which in these texts means something closer to decency than to doctrine.
What he actually asks for:
- Obedience to parents and respect for elders and teachers.
- Generosity to friends, to Brahmins and to wandering ascetics — named separately and equally.
- Restraint towards living beings; less killing of animals.
- Truthfulness, and restraint in speech.
- Decent treatment of servants and slaves.
- Moderation in spending, and scepticism about expensive ceremonies — he says bluntly that women's rituals of this kind are of little value.
- And, repeatedly, tolerance between religious sects.
That last one deserves quoting, because it is the most remarkable sentence any ancient ruler wrote about religion. The twelfth edict says, in substance: honour other sects as well as your own; whoever praises his own sect and disparages another's, out of devotion to his own, damages his own sect more than anyone else's.
Written in the third century BCE, by an emperor, about his own religion, and put on public display.
The edicts also record practical works. Wells dug at intervals along the roads. Shade trees and mango groves planted. Rest houses built. Medical treatment provided — and he specifies, twice, for animals as well as people. Officials appointed with a specific brief to look after the welfare of the population and to check on the treatment of prisoners.
One edict is about his own kitchen, and it is the most human thing he wrote.
It says that formerly, in the royal kitchen, many hundreds of thousands of animals were killed every day for curry. Now, it says, only three are killed: two peacocks and a deer — and the deer not always.
Then, as though he has just heard himself, he adds: even these three will not be killed in future.
Discount the number at the top; state kitchens exaggerate. What is interesting is the shape of the sentence. It is a man setting a rule, noticing it is not good enough, and tightening it in the same breath — in a text he knows is going to be carved into rock and be there for ever.
Set beside the hundred thousand dead at Kalinga, the two peacocks are almost unbearable. Both are in his own words, on his own orders, within a few years of each other.
9.3 — The one edict he left out
Now the detail that stops this chapter from becoming a hagiography, and it is a good one.
The great Kalinga remorse edict — the one with the figures and the shame — was inscribed at sites across the empire.
It was not inscribed in Kalinga.
At the two edict sites in Kalinga itself, that text is missing. In its place are two separate edicts found nowhere else, addressed to his officials, instructing them to treat the people fairly, to avoid arbitrary imprisonment, and to make the population understand that the king regards them as his children.
So the conquered province got reassurance and administrative instruction. The rest of the empire got the confession.
There is a second detail in the same direction. The remorse edict also contains a warning: the forest peoples of the empire are told that although the Beloved of the Gods feels regret, he still has power, and that they should repent, lest they be killed.
He did not disband the army. He did not release Kalinga. He did not abolish the death penalty, though he granted condemned prisoners three days' grace.
This has been argued about since the edicts were first read, and it will not be settled.
Nothing compelled him to publish it. No ancient ruler admits to a hundred thousand dead by his own act; every other king in the ancient world carved his victories. The reversal is total and expensive: the state stops sacrificing animals, funds medical care and public works, appoints welfare officials, and the emperor writes about his own regret in the first person. And it lasted for the remaining twenty-five years of his reign, which is a very long time to sustain a pose.
Kalinga was the last conquest available; there was nothing left to conquer that was worth the cost. A doctrine of non-aggression is convenient for a ruler who has finished aggressing. Dhamma as he defines it — obey your parents, respect your betters, do not quarrel between sects, be generous to holy men — is an ideology of social order, and an empire of that size held by a small administration needs one badly. Note also what he did not give up: the throne, the army, the death penalty, or the province he had just taken. And the remorse edict was carefully not shown to the people he was remorseful about.
Rulers are not sincere or insincere in the way individuals are. A policy can be a real change of heart and a shrewd move at once, and the second does not cancel the first. Asking whether he "meant it" imports a modern notion of private feeling into a public document that was never about private feeling.
What would settle it: nothing. There is no evidence for what a man felt, and there never will be. Everything we have is what he chose to publish.
What is not in dispute: that the policies existed, that they were expensive, and that they ran for twenty-five years.
9.4 — The rocks near Punjab
Two of the fourteen great rock edicts sit in the north-west, at Shahbazgarhi and Mansehra in what is now Pakistan's Khyber Pakhtunkhwa, on the roads leading down from the passes towards Taxila and Punjab.
They are cut in Kharoshthi — the script that came out of the Persian filing system in Chapter 2. Everywhere else in the empire the edicts are in Brahmi, and at Kandahar there is a version in Greek and Aramaic together.
Stop and look at what that means. An Indian emperor, wanting to be understood in the north-west, had his words cut in the local script; wanting to be understood in Afghanistan, had them cut in Greek and Aramaic. He was not publishing in a sacred language. He was publishing in whatever the reader used — and the language throughout is Prakrit, the ordinary spoken speech, not Sanskrit.
Then there are the pillars: polished sandstone columns, some over twelve metres and weighing dozens of tonnes, quarried in one place and moved hundreds of kilometres. One of them stood at Topra, in what is now Haryana and was long part of greater Punjab.
It is not there now. In 1356 a Delhi sultan admired it so much that he had it dug out, wrapped in silk cotton, floated down the Yamuna on a barge and re-erected in his new capital. He could not read a word of it. Nobody could, by then. It stands in Delhi today.
That story belongs to Part 6, and I mention it here because it is the neatest illustration of the next section.
9.5 — Forgotten, and found
Ashoka was lost.
Not the rocks — the rocks were always there, in plain sight, all over India, covered in writing. What was lost was the ability to read them, and with it the man.
Buddhist chronicles kept in Sri Lanka remembered a great emperor called Piyadassi. Indian tradition remembered a Mauryan king called Ashoka. Nobody connected them to each other or to the inscriptions, because for something like fifteen hundred years nobody in the world could read the script.
In 1837 an official of the East India Company's mint in Calcutta, James Prinsep, worked out Brahmi. He got there partly by comparing short donor inscriptions at Sanchi, where the same two-word ending kept repeating and turned out to mean "the gift of".
Suddenly the rocks spoke. They spoke in the voice of a king called Beloved of the Gods, and the Sri Lankan chronicles used the same title for Ashoka. The identification was made almost at once — but it was not proved until 1915, when an edict was found at Maski that names him outright.
That a ruler's own inscriptions describe his reign.
This chapter has better evidence than any other in this part, and that is exactly why the assumption is dangerous here.
The edicts are genuine, contemporary and in his own voice. All true. Now notice what follows from that and what does not.
They tell us what he wanted publicly known. They do not tell us what the empire was like. Not one edict mentions a tax rate, a harvest, a famine, a rebellion, an official removed, a war after Kalinga, or a single thing that went wrong. That is not because nothing went wrong in a twenty-five-year reign over most of a subcontinent. It is because rulers do not carve their failures.
There is a second layer, subtler and more interesting. The edicts repeatedly complain that they are not being obeyed. Ashoka keeps ordering that they be read aloud on fixed days. He keeps appointing new officials to check on dhamma. He keeps saying people are not doing enough.
Read that as a historian rather than as an admirer and it inverts the picture. A ruler who has to keep repeating an instruction is a ruler whose instruction is not being followed. The edicts are usually read as a portrait of a moral empire. They can just as well be read as the frustrated correspondence of a man whose subjects were carrying on exactly as before.
And here is the sting. Everything you have ever read about Ashoka's greatness comes, in the end, from Ashoka. The Buddhist chronicles that fill in the rest were written by an institution he had endowed. There is no independent account of his reign by anybody. He is the best-documented figure in ancient India and almost the entire documentation is his own press.
The most reliable source in this part is also the most interested one.
After conquering Kalinga around 261 BCE, Ashoka had the casualty figures — a hundred thousand killed, a hundred and fifty thousand deported — carved into rock across his empire, along with a statement of his own remorse. No other ancient ruler did anything like it.
The edicts are mostly not about Buddhism. They ask for respect for parents, generosity to holy men of every kind, restraint towards animals, honest speech, and above all tolerance between sects.
The remorse edict was not inscribed in Kalinga itself. He kept the army, the province and the death penalty, and warned the forest peoples that he still had power.
He wrote in Prakrit, the ordinary spoken language, in Kharoshthi in the north-west and in Greek and Aramaic in Afghanistan — always in whatever the reader could actually use.
He was then forgotten for fifteen hundred years, until James Prinsep deciphered Brahmi in 1837. His lions are on India's state emblem and his wheel is on its flag — recovered by an accountant in Calcutta reading a script nobody had read since the Guptas.
10An Honest List of What We Do Not Know
Two lists, as always. This part had more documents than any before it, and the first list is still long — because documents answer the questions their authors cared about.
Genuinely unknown
These are open. Where a document exists but cannot be trusted, that counts as open too.
- Where Hindush was. The Persian province is a name on a list. Whether it covered Punjab, or only the lower Indus, or something in between, is not established.
- How Persia took it, and how deeply it held it. No campaign account, no garrison, no archive, no building. A name appears on a later inscription that was absent from an earlier one. That is all.
- Whether the tribute figure is real. Nine tonnes of gold a year, on the word of a Greek who never came here, uncorroborated by the Persian records that survive.
- Almost every number in the Alexander campaign. Army sizes, casualties, elephants, dead at Massaga, dead at the Jhelum. All from sources three to five centuries later, all with reasons to inflate or deflate.
- Where the battle on the Jhelum was fought. Several sites are proposed. The river has moved. Nothing has been excavated that settles it.
- Where Alexander's cities in Punjab were. Bucephala, Nicaea, the twelve altars on the Beas. All described, none found.
- Anything about Porus that is not Greek. No Indian text, inscription or coin. Everything we have was written to make his conqueror look good.
- Chandragupta's origins. Four traditions, four different answers, each shaped by that tradition's view of who ought to rule.
- Whether Chanakya existed, and what he wrote. The Arthashastra names Kautilya and never names Chandragupta. The surviving version is composite and centuries later. The link is tradition.
- How the Mauryan empire was actually administered. We have a manual that may not describe it, an ambassador's account that survives only in quotation, and an emperor's edicts that never mention a tax rate.
- What Chandragupta's treaty with Seleucus actually contained. Whether the marriage was dynastic, what the exact territorial terms were, whether there was a battle first.
- How Chandragupta died. The Jain account is plausible and uncorroborated by anyone outside that tradition.
- Whether Ashoka's remorse was what it says it was. No evidence exists for what any human being felt, and none ever will.
- Whether the edicts were obeyed. Ashoka's own repeated complaints suggest not much, but that is an inference from tone.
- Why the Mauryan empire fell apart within fifty years of his death. Overextension, succession disputes, cost, provincial breakaway — all proposed, none established.
- What ordinary life in Punjab was like through any of this. Two hundred years of documents, and we still cannot say what a farmer near the Ravi paid, ate, believed or feared.
Solid
These are firm enough to build on.
- Punjab was written about for the first time around 518 BCE, as a Persian province name. The word Hindush comes from Sindhu, and India comes from Hindush.
- Indian troops served in the Persian army in Greece in 480 and 479 BCE.
- The Aramaic alphabet came in through Persian administration and became Kharoshthi, the first script used to write the languages of this region.
- Taxila was a major city and a famous centre of learning. Its earliest level is less planned than Harappa, fifteen hundred years earlier.
- Panini wrote a complete formal grammar of Sanskrit in this region and period, using techniques not matched for two millennia.
- Alexander crossed the Indus in 326 BCE, defeated a king on the Jhelum, restored him to his throne, was refused by his own army at the Beas, and left. He was here about two years.
- Not one eyewitness account of that campaign survives.
- No Indian source of any period mentions Alexander.
- Macedonian authority east of the Indus lasted roughly ten years and ended with a Greek commander murdering Porus and taking his elephants west.
- Chandragupta took Magadha around 321 BCE and Punjab by about 317 BCE, and Punjab was thereafter governed from Taxila as part of an Indian empire.
- Around 303 BCE Seleucus ceded Kandahar, Kabul and Balochistan for five hundred war elephants, and those elephants won him his share of Alexander's empire four years later.
- Megasthenes lived at the Mauryan capital as an ambassador, and his description of a wooden palisade at Pataliputra has been confirmed by excavation.
- Ashoka conquered Kalinga around 261 BCE and published the casualty figures and his own remorse, in his own name, across his empire.
- The edicts exist, in Prakrit, in Brahmi, Kharoshthi, Greek and Aramaic, and they are the earliest surviving Indian writing that can be read.
- The five Greek kings named in the thirteenth edict fix early Indian chronology to within a few years. It is the only anchor of its kind for the period.
- The remorse edict was omitted at the two sites in Kalinga.
- Brahmi was deciphered by James Prinsep in 1837, and Ashoka's identity was confirmed by the Maski edict in 1915.
A note on how this part felt different
Parts 1, 2 and 3 were about not having enough. This part is the first where the problem is having the wrong things.
We have thousands of words about a two-year foreign campaign and almost nothing about the two centuries either side of it. We have a detailed manual of ideal government and no idea whether any of it was practised. We have an emperor's inner life, published by the emperor, and no independent word about his empire.
That imbalance is not an accident and it is not really about India. It is what happens whenever the surviving record was produced by particular people for particular purposes — which is to say, always.
The correction is not to distrust everything. It is to keep asking, of every document in the rest of this series, the three questions from Part 3: who is speaking, to whom, and what do they want?
This part has far more evidence than the three before it, and the list of unknowns is just as long — because documents answer only the questions their authors cared about.
What is firm: the Persian province and the name it gave this country, Kharoshthi, Taxila, Panini, Alexander's two years, Chandragupta's empire, the elephant treaty, and Ashoka's rocks.
What is not: almost every number, almost everything about Porus, the origins of Chandragupta, the authorship of the Arthashastra, and what any of it was like to live through.
Having documents is not the same as having answers. It means somebody else has chosen the questions.
Sources & further reading — Part 4
Timeline of Part 4
Dates in this part are firmer than in any before it, because of the five Greek kings named in Ashoka's thirteenth edict. Where a date is still argued over, the table says so.
| When | What |
|---|---|
| c. 559–530 BCE | Cyrus the Great builds the Persian empire. Greek writers say he campaigned in the east and came off badly. Whether he reached the Indus is unsettled. |
| c. 519 BCE | The Behistun inscription is cut. It lists Gandara among Darius's lands. It does not list Hindush. |
| c. 518–513 BCE | Hindush appears on Darius's later inscriptions. Punjab enters the written record. No account of the conquest survives. |
| c. 515 BCE | Scylax sails down the Indus for Darius and back to Egypt by sea. Thirty months. |
| 480–479 BCE | Indian troops serve in Xerxes' army in Greece, and are at Plataea on the losing side. |
| c. 500–400 BCE | Bhir Mound, the earliest city at Taxila, at its height. Aramaic in administrative use; Kharoshthi begins to take shape from it. |
| c. 500–350 BCE | Panini composes the Ashtadhyayi in Gandhara. Dating debated across roughly two centuries. |
| c. 430 BCE | Herodotus writes. India is the twentieth satrapy, pays the heaviest tribute, and has gold-digging ants. |
| 336 BCE | Alexander becomes king of Macedon at twenty. |
| 330 BCE | Persepolis burned. The Achaemenid empire ends. |
| 327 BCE | Alexander crosses the Hindu Kush. The Swat campaign: Massaga, and the killing of surrendered Indian mercenaries. The rock of Aornos taken by causeway. |
| Spring 326 BCE | The army crosses the Indus on a bridge of boats. Taxila submits and supplies it. |
| May 326 BCE | The battle on the Jhelum. Night crossing in a thunderstorm; elephants; Porus defeated, wounded, and restored to his kingdom. Bucephalus dies. |
| Summer 326 BCE | Sangala taken with heavy losses. At the Beas the army refuses to go on. Twelve altars raised; the retreat begins. |
| 326–325 BCE | Down the Jhelum and the Indus. Alexander takes an arrow through the lung at a Malli town, almost certainly Multan. |
| 325 BCE | Nearchus takes the fleet along the Makran coast. Alexander marches through the Gedrosian desert and loses more people than in all his battles combined. |
| 323 BCE | Alexander dies at Babylon, aged thirty-two. |
| c. 321 BCE | Chandragupta overthrows the Nandas and takes Magadha. |
| c. 318 BCE | The Greek commander in India murders Porus, seizes his elephants and marches them west. |
| c. 317 BCE | Macedonian authority east of the Indus ends. Punjab becomes Mauryan, governed from Taxila. |
| c. 303 BCE | Treaty with Seleucus: Kandahar, Kabul and Balochistan for five hundred war elephants, and a marriage alliance. |
| 301 BCE | At Ipsus in Anatolia, those elephants win Seleucus the largest share of Alexander's empire. |
| c. 302–288 BCE | Megasthenes serves as ambassador at Pataliputra and writes the Indika. |
| c. 297 BCE | Chandragupta dies. Jain tradition says by fasting, in Karnataka. Bindusara succeeds. |
| c. 268 BCE | Ashoka is consecrated. |
| c. 261 BCE | Kalinga. A hundred thousand killed, a hundred and fifty thousand deported, by the state's own count. |
| c. 258–232 BCE | The edicts, on rocks and pillars from Kandahar to Karnataka, in Prakrit, in four scripts. Shahbazgarhi and Mansehra serve the north-west. |
| 232 BCE | Ashoka dies. The empire begins to come apart within a generation. |
| 185 BCE | The last Mauryan king is killed by his own commander-in-chief, Pushyamitra Shunga. |
| c. 185 BCE | Part 5 starts here. |
| 1837 CE | James Prinsep deciphers Brahmi. Ashoka can be read again after fifteen hundred years. |
Glossary
Every hard word used in Part 4, in plain English.
| Word | What it means |
|---|---|
| Achaemenid | The Persian dynasty founded by Cyrus the Great. Their empire ran from the Aegean to the Indus for about two hundred years. |
| Aornos | The Greek name for a mountain stronghold above the Indus, taken by Alexander by building a causeway across a ravine. Its exact site is disputed. |
| Aramaic | The everyday written language of the Persian empire. Unrelated to Persian, but it used an alphabet, which made it useful. Kharoshthi descends from its script. |
| Arthashastra | A manual of statecraft attributed to Kautilya. Lost for centuries and rediscovered in 1905. The surviving version is composite and centuries later than the Mauryas. |
| Bhir Mound | The earliest of the three cities at Taxila, occupied from about the sixth century BCE. Winding lanes, rubble houses, soak pits. |
| Brahmi | The script most of Ashoka's edicts are cut in, and the ancestor of almost every Indian script including Gurmukhi and Devanagari. Deciphered in 1837. |
| Companion cavalry | Alexander's heavy horse and personal command. The weapon he actually won battles with. |
| Cuneiform | The wedge-shaped writing of Mesopotamia and Persia, pressed into clay or chiselled into stone. |
| Devanampiya Piyadasi | "Beloved of the Gods, of gracious mien." The title Ashoka uses for himself in almost all his inscriptions. It is why nobody could identify him for so long. |
| Dhamma | In Ashoka's edicts, a public ethic rather than a doctrine: respect for parents and elders, generosity to holy men of all kinds, restraint towards living beings, honesty, and tolerance between sects. |
| Edict | A public order or announcement from a ruler. Ashoka's were cut into rock faces and stone pillars across the empire. |
| Epigraphy | The study of inscriptions. The backbone of ancient Indian history, because inscriptions survive and manuscripts in this climate do not. |
| Gandhara | The region around Peshawar and Taxila, on the routes from the passes into Punjab. A Persian province, then Mauryan, and later the centre of a famous school of sculpture. |
| Hindush | The Old Persian name for the province at and beyond the Indus, from Sanskrit Sindhu. The Greeks turned it into India. |
| Hydaspes | The Greek name for the Jhelum — the Vitasta of the Rigveda. Site of the battle with Porus. |
| Kalinga | Roughly modern Odisha. Ashoka's last conquest, and the war he published his remorse about. |
| Kharoshthi | The script of the north-west, descended from the Aramaic alphabet, written right to left. Used for the edicts at Shahbazgarhi and Mansehra. |
| Magadha | The kingdom on the Ganga plain, with its capital at Pataliputra, that became the base of the Mauryan empire. The reported size of its army is what turned Alexander's men back. |
| Maurya | Chandragupta's dynasty, c. 321–185 BCE. The first empire to run from Afghanistan to Bengal. |
| Megasthenes | Seleucus's ambassador at Pataliputra. His book on India is lost and survives only in quotations by later writers. |
| Nanda | The dynasty that ruled Magadha before the Mauryas, and that Chandragupta overthrew. |
| Pataliputra | The Mauryan capital, on the Ganga at modern Patna. Megasthenes described its wooden palisade; excavation has found the posts. |
| Paurava | The likely Indian form behind the Greek "Poros" — a man of the Puru lineage. The Purus appear in the Rigveda in Part 3. |
| Phalanx | The Macedonian infantry block: deep ranks of men with six-metre pikes. Unstoppable from the front on firm flat ground, and vulnerable everywhere else. |
| Prakrit | The ordinary spoken languages of northern India in this period, as opposed to the fixed learned Sanskrit. Ashoka's edicts are in Prakrit. Punjabi's ancestors are among them. |
| Sallekhana | The Jain practice of fasting to death at the end of life. Jain tradition says Chandragupta died this way. |
| Satrapy | A province of the Persian empire. A satrap was its governor. |
| Sutra | "Thread." A rule compressed to the smallest possible number of words, designed to be memorised. Panini's grammar is written entirely in them. |
| Talent | An ancient unit of weight, roughly twenty-six kilograms. Herodotus's tribute figures are given in talents. |
| Tribute | A regular payment from a subject region to an imperial centre. Closer to protection money than to tax. |
| Uttarapatha | "The northern road." The trunk route from the north-west across Punjab to the Ganga plain. The Grand Trunk Road is its descendant. |
| Yavana | The Indian word for Greeks, and later for westerners generally. From "Ionian", by way of Persian. Panini already knew a form of it. |