Chapter 1Pella

Foreword

The Achaemenid archive survives because Alexander burned Persepolis — not in the fire, but under what the fire brought down.

Thousands of clay tablets recording the administration of the Persian empire — rations, travel authorizations, the ordinary paperwork of a state — were stored in rooms inside the fortification wall. The fire did not bake them; the excavation found them unbaked, which is how they were made. What preserved them was the wall itself, whose upper courses came down and sealed them where they lay, and the burning is the likeliest reason the archaeological account gives for the wall coming down. Without that collapse they would have crumbled. Because of it they are readable.

That is the shape of this whole subject, and it is why the book is called what it is.

Almost everything we know about Alexander III of Macedon comes to us through the account of men on his payroll. Ptolemy and Aristobulus wrote the version that Arrian, four and a half centuries later, made canonical. What Arrian leaves out reaches us only through Cleitarchus, Curtius and Diodorus, who had reasons of their own.

So the argument of this book has to be made from the ground and from the tablets rather than from the narrative, and it is this.

Terror was not a lapse in Alexander's method. It was a component of it.

It was introduced deliberately at Thebes in 335, when a city was stormed, roughly six thousand people were killed and thirty thousand sold, and the temples and one poet's house were spared and nothing else was. Arrian records that an allied council voted the sentence, which is a real legal distinction and changes nothing about what was purchased, which was twelve years in which no Greek city followed Thebes' example.

It was industrialized at Tyre in 332, over seven months, against an island city half a mile offshore. Arrian gives eight thousand dead and thirty thousand enslaved. Diodorus and Curtius add two thousand crucified along the shore, and Arrian does not mention it at all.

It was carried into the Persian heartland at Persepolis in 330, and the fire itself is the one thing this book will not adjudicate — Arrian states it flatly as policy, the vulgate gives a drunken feast and an Athenian courtesan throwing the first torch — because what decides the question is not the fire but the months of systematic stripping that nobody disputes came first.

And then it was normalized, in Bactria and Sogdiana, into two years of population war that nobody writes about.

Three chapters of this book are the set pieces and three more carry the deaths the tradition does not count, because Sogdiana and the Makran desert cost more lives than the famous battles and stand on no monument.

This is a book with a ledger in it. City by city, dead and enslaved, with the range across the sources shown and the arithmetic on the page, and with every figure attributed to whoever supplied it.

I want to say plainly what the verdict is not.

It is not a choice between the conqueror and the exterminator. At this scale, in the fourth century BC, those were one job. The argument is that the method was in place before Thebes, that it never afterwards changed, and that his reputation survived it because the people who wrote the surviving account had served in his staff.

Against the staff account this book sets every other side that left a record.

The Babylonian astronomical diaries, written in cuneiform within days of Gaugamela, record a defeat, a panic and the price of barley. Another tablet fixes the date of his death — 10 or 11 June 323 — more precisely than any Greek source does. The Persepolis Fortification tablets show what an Achaemenid administration actually did. Egyptian and Middle Persian material remembers him not as the Great but as the accursed. And no Indian source records the campaign in the Swat valleys at all, which is a finding rather than a gap.

Several things in here are corrections rather than arguments.

The mole at Tyre, the most admired piece of engineering in his career, followed a submerged sand spit that was already there. Geoarchaeological work published in 2007 established it. Brilliance and luck, precisely disentangled.

Darius III was a competent king who lost to a better army, not the stage coward of the Greek tradition, and Chapter 7 and Chapter 10 treat him accordingly.

Every oration in Arrian, Curtius and Diodorus is a composition by the historian. Not one speech in this book is quoted as speech.

And the class list at Mieza — the boys who supposedly studied under Aristotle and then divided the world — is a later reconstruction. Only one of them is actually placed there by an ancient source.

The single chapter in which the documented fact is mercy is Chapter 7. After Issus, Darius' mother, wife and children were taken, and were treated with a correctness that even hostile sources concede.

The last chapter is the accounting, and beside it the other memories: a body hijacked in 321 by one of his own officers and taken to Egypt to found a dynasty on a corpse, a tomb that has never been found, a romance that ran through more than two dozen languages, a Roman weeping at a statue, a Corsican in Egypt, and two modern states still arguing over whose he is.

One convention, stated once here. Macedonian in this book means what it meant in the fourth century BC. The modern naming dispute belongs to the last chapter and is dealt with there.

He was dead at thirty-two, eleven days into a fever, in Babylon, and nobody knows what killed him.

Chapter 1 · Pella

Macedon in 356 BC was a frontier kingdom, and the Greeks to the south called it barbarian. The word did its work at the edges of things — of language, of manners, of whether a people belonged inside the conversation or outside it — and the Macedonians were placed outside. They were a monarchy in a world of city-states, ruled by a king with a nobility on horseback, holding ground that ran up against Illyria in one direction and Epirus in another. To an Athenian this was the wrong side of a border in every sense that mattered. It is worth saying at the start that the kingdom's northern neighbors were not an audience for its ambitions but a standing military problem, and that a Macedonian king's first job, before Persia was ever a thought, was to survive his own frontier.

A note on a word, made once and then left alone. Macedon and Macedonian are used in this book in their ancient sense: the fourth-century BC kingdom, its people, its army, its dynasty. They are not used to advance or deny any modern national claim. There is a live argument about whose inheritance this is, and it is a real argument with real politics inside it, but it belongs at the end of this book rather than in a chapter about a child. Here the words mean what they meant to the men who carried the pikes.

The pikes are where to begin, because the pikes were there before he was. Philip II rebuilt the Macedonian army around the sarissa — an infantry pike far longer than the spear it would meet, held in both hands, which meant the man carrying it could not also carry a hoplite's shield and could not fight as a Greek citizen-soldier fought. That single change cascaded through everything. It required a different shield, a different drill, a different relationship between the ranks, and above all it required men who trained. A phalanx of sarissas is not a militia formation. It is a machine with a maintenance schedule, and its parts have to be paid, fed, and kept together through the seasons when farmers go home. Whatever else Philip did, he made the Macedonian infantryman into a professional, and a professional is expensive.

Alongside it he built the second element: the Companion cavalry, deployed in a wedge. A wedge is not a decorative formation. A line of horse pushes; a wedge enters. Shaped to a point and led from that point, it is designed to go into a gap in an enemy formation rather than to grind along its front, which means it is useless without someone watching for the gap and willing to commit before it closes. The Macedonian army's decisive arm was therefore an arm that depended on judgment exercised at speed by whoever rode at the tip. That is a fact about equipment and formation, and it is also, as things turned out, a fact about a personality. The instrument was built to be used by a man who would go first.

All of this had to be paid for, and the payment is the part of Philip's achievement that gets least attention because it is the least dramatic. A phalanx that trains is a phalanx that eats without producing food. Engineers are specialists and specialists are hired. Cavalry is expensive in a way infantry is not, because horses have to be bred, fed, and replaced. Where the money came from, in what amounts, and against what security, is not something the surviving record itemizes, and this book will not pretend otherwise. What can be said is that the question does not go away — it follows the army east, and it becomes acute at the moment the Macedonians walk into the treasuries of the Persian empire and find bullion reserves on a scale that makes the ancient figures stop being credible. Where does an army's money come from, and what does that money do to the army: it is the right question to ask about Philip's creation, and the chapters at Babylon and Susa are where it can finally be asked with evidence. Here it can only be noted that the machine was expensive, that somebody solved that problem, and that the somebody was not Alexander.

The third element was the siege train, and it is the one the Greeks noticed last and should have noticed first. Philip could field engineering that nobody else in Greece could field. What that meant in practice was that a wall stopped being an argument. For most of Greek history a city that shut its gates had bought itself a season, possibly a war, certainly a negotiation; a besieger without engines could starve a place or bribe it, and both took time he usually did not have. Against a siege train the calculation reverses. The gates become a delay rather than a defense, and the defenders' choice narrows to terms or destruction. Every one of this book's set-piece horrors is downstream of that reversal. The apparatus that would produce them was in the field, functioning, and in his father's hands before Alexander could walk.

He was born at Pella in July of 356 BC. That sentence is close to the limit of what the birth can be made to yield.

It is worth holding it against the other end of the life for a moment. The death is dated to 10 or 11 June 323 BC, in Babylon, and it is dated that precisely because a cuneiform tablet written in Babylon says so — not because any Greek recorded it. The beginning has no such witness. There is no day, only a month, and the month arrives through writers working centuries downstream from a house that left no archive. The birth is fixed no more closely than that, and the reason is banal: at the time, it was the birth of a son to a king of a middling northern kingdom, and nobody had any cause to write it down.

This is the point at which a biography of Alexander has to be honest about its own foundations, and it may as well be honest now rather than in a note somewhere. There is no contemporary narrative of Alexander. Not a diminished one, not a fragmentary one — none. Every surviving continuous account of his life was written between roughly 300 and 500 years after he died, by men working from earlier books that no longer exist. The five surviving accounts are collectively called the Alexander historians, and the best of them, Arrian, was writing in the second century AD, which is to say that the most reliable narrative of the campaign reaches us filtered through Arrian four and a half centuries later. Plutarch, also second century AD, is the single source for several of the most quoted moments in the life. Diodorus wrote in the first century BC, Curtius in the first century AD. Justin summarized a summary.

The lost books underneath them are the real evidence, and we have those only as quotation and paraphrase inside the surviving five. Ptolemy son of Lagus, who was on the campaign and later made himself a king, wrote one. Aristobulus of Cassandreia, an engineer who was also there, wrote another. Arrian preferred both of them, which is a preference for two men with reasons. Callisthenes of Olynthus was the official campaign historian — the propaganda baseline, and his work is gone. The reputation that came down to us was assembled, in the first instance, by men on the payroll of the man it describes, and then edited by Greek and Roman writers with agendas of their own several centuries downstream. That is not a caveat to be got out of the way. It is the subject.

There is a second channel, and the reader will need it later, so it should be named now. Running parallel to Arrian's staff tradition is what scholars call the vulgate, descending from Cleitarchus of Alexandria — probably not an eyewitness — and reaching us through Diodorus, Curtius, and much of Plutarch. The two channels do not merely differ in tone. They differ in what they contain. Arrian, preferring Ptolemy and Aristobulus, inherits their silences, and the killings he passes over are precisely the ones the vulgate reports. This has a consequence that governs the entire book: when Arrian is quiet about a killing, that silence is itself evidence. It also has a cost that must be admitted in the same breath. Curtius, who supplies a great deal of the worst of it, was a Roman rhetorician moralizing about tyranny, and quite possibly writing about Caligula with Alexander's name on it. The prosecution case against Alexander rests disproportionately on him. That needs saying out loud, at the front, rather than being discovered by a reader who notices halfway through that the atrocities keep arriving from one direction.

Against both channels this book sets a third kind of evidence, which is not Greek at all and is the reason it can claim to do something the standard biographies did not. Cuneiform tablets from Babylon — the astronomical diaries, recording observation, weather, commodity prices and political events, gathered in the Sachs and Hunger edition published by the Austrian Academy of Sciences between 1988 and 2014 — were written contemporaneously, in the city, as events happened. They are the only strictly contemporary written testimony about Alexander that survives anywhere. And at Persepolis, tens of thousands of Elamite and Aramaic administrative tablets from the Achaemenid fortification archive survive for a reason with no equal in the study of any other conqueror: his fire did not touch them, and what buried them was the collapse of the wall they were stored in, which the fire is thought to have caused. The empire's own paperwork was preserved by the burning of the empire. When this book reaches Gaugamela, Babylon, Persepolis and the deathbed, it will tell those events twice — once from the Macedonian staff account, and once from people who were writing while it happened and were not on his payroll.

Which brings us to the omens, because the childhood of Alexander as it has been transmitted is not a childhood. It is a portfolio of signs.

The temple of Artemis at Ephesus, the Artemision, burned. The tradition places the fire on the night Alexander was born, and the coincidence is the whole of its purpose: a great temple in Asia destroying itself at the moment the man who would take Asia drew breath. Nobody who tells that story is in a position to check the synchronism, and nobody telling it wanted to. The fire may well have happened; the calendar joining it to a delivery room in Pella is a later weld. Alongside it in the tradition sit Plutarch's snake, a thunderbolt, and the horse Bucephalas. The specific content of these stories matters less than their shape, and their shape is identical. Each attaches to an infant, in advance, a quality the adult turned out to possess.

Every one of these was composed after the conquest, to explain the conquest. That is the mechanism, and it is worth being precise about how it works, because it is not lying exactly. A world that has just watched one man dismantle the largest empire it had ever built needs an account of how such a thing is possible. The available account is that it was foretold. So the signs are written backward from the outcome, each one shaped to fit what is already known to have happened, and they accumulate in the tradition until the childhood consists of nothing else. The modern historian's term for this material is retrospective omen literature, and the honest summary of the birth and the childhood at Pella is that the depth of evidence is poor and the density of legend is total. Strip the omens out and there is very little left in the box.

What is left is Olympias, and she has been handled almost as badly as the omens, in the opposite direction.

She was a Molossian princess, from the royal house of Epirus, and her marriage to Philip was a treaty with legs. This is not a diminishment of her; it is the ordinary grammar of dynastic marriage in that world, and it cuts both ways. Philip married repeatedly and married politically — his wives included Audata the Illyrian, daughter of a defeated neighbor — and the court at Pella was polygamous. That word is usually deployed to make the Macedonian court sound exotic. It is more useful to see what it does structurally. Each wife was a relationship with a border. Each son by each wife was a claim. The king's household was therefore not a family in the sense a modern reader will supply but a standing arrangement of rival lines, in which a queen's position depended on her own kin, her husband's current need for them, and the survival and prospects of her son.

Set inside that, Olympias' conduct stops looking like the pathology the tradition assigns her. The received Olympias is a figure of hysteria and snakes and vindictive scheming — a woman characterized by ancient male writers using the vocabulary those writers had for powerful women, which was small and unkind. Elizabeth Carney's work on her, and on women and monarchy in Macedonia more broadly, stands as the corrective to that caricature. What can be said flatly here is narrower and firmer: she was a foreign princess in a court of rival wives, whose leverage ran through her brother's kingdom and her son's prospects, and who behaved throughout in ways consistent with understanding both of those things exactly. Whether she also did the specific things the tradition accuses her of is a question this book takes up where the evidence for it arises, which is not yet.

We cannot know what Olympias thought about her son in the years at Pella. We can know some of what she would have known. She would have known that a Macedonian king's succession was not a matter of birth order administered by a clerk, and that a half-brother is a competitor before he is a relative. She would have known that her own standing at Pella was not permanent, because nobody's was. And she would have known what the army in the fields outside was worth, because everyone in that kingdom knew — it was the thing that had turned Macedon from a place that got raided into a place that raided. The inference is marked as inference. But a mother in that position with a son in that position was not, on any reasonable reading, a spectator.

The boy himself is nearly invisible, and this is where the discipline has to hold. The record for the years the sources bracket as 356 to 343 does not preserve a personality. It preserves anecdotes attached to a personality already famous. There is no letter home from Pella. There is no household account, no schoolmaster's complaint, no remark of his that reaches us through anyone who was in the room and wrote it down at the time.

This absence deserves a word, because the method of this book depends on it. Military history at its best is built out of the letters and diaries of the men who were there — the soldier writing home, the officer's private journal that contradicts his official report, the survivor's testimony taken down while it was still raw. For Alexander that material does not exist, on any side. Not the Macedonian phalangite's, not the Persian officer's, and later, when the killing starts in earnest, not the Sogdian villager's. The honest response is to say whose voices are missing and why, and then to refuse to supply them. This book will not invent a representative soldier to carry a scene, and it will not put a thought into Alexander's head that the evidence has not put there. Where it infers, it will say that it is inferring. Where the record stops, it will say that the record stops. It stops here, at the beginning, more completely than anywhere else in the life.

So what can actually be evidenced about the childhood at Pella? That he was the son of the king by a Molossian wife, in a polygamous royal household in which his standing rested on his mother's kin and his father's intentions and on nothing more permanent than either. That the kingdom he would inherit was on the way up and had just been rebuilt from the ground by the man at the head of it. That the rebuilding was military, expensive, professional, and recent enough that its newness would have been visible to a child — the drill, the money, the specialists, the trains of engineering equipment going out in the campaigning season and coming back. And that in about 343, while he was still a boy, a philosopher was engaged to teach him, which is itself evidence of something: a king does not commission that kind of education for a son he does not intend to hand a kingdom to.

And one more thing, which is the argument of this chapter and, in Ian Worthington's persistent formulation across his work on Philip and Alexander, the argument that hangs over the whole reign: the instrument was Philip's. The sarissa phalanx was Philip's. The Companion wedge was Philip's. The siege train was Philip's, and so were the revenues that paid for all of it and the officer corps that ran it. Worthington's case is that Philip built the instrument and Alexander squandered the empire. One does not have to accept the second half to see that the first half is simply true, and that it is the correct starting position for anyone trying to understand what happened next.

Because what happened next was extraordinary by any measure, and the measures are worth putting on the table before the omens get another chance at them. In eleven campaigning years he destroyed the largest empire the world had yet built. He marched an army roughly twenty thousand miles, from the Danube to the Punjab. He was never defeated in a pitched battle. He died at thirty-two without naming an heir. Those are the outcomes the birth legends were reverse-engineered to explain, and they are strange enough that one can see why a Greek world casting about for a cause reached for a snake and a burning temple.

This book proposes a duller and more disturbing cause. The machine existed first. It was designed to be driven from the point of the wedge by someone who would commit before the gap closed, and it came with an engineering capability that made walls negotiable and terms optional. A boy grew up next to it, was raised to inherit it, and inherited it. What he then did with it — including the parts of it that the men on his payroll left out of their accounts, and that only their enemies and the tablets and the burn layers preserve — is the ledger this book intends to keep, city by city, with the arithmetic shown and the sources named.

At Pella in 356 the ledger is empty. It is the last time it will be.