Xerxes' Invasion of Greece (480–479 BCE)

Everything of consequence about the campaign passes through Greek hands. The earliest witness is Aeschylus' tragedy The Persians, produced at Athens in 472 BCE by a dramatist who had himself borne arms against the invasion, which dramatises the defeat from the Persian court's side and fastens on the bridging of the Hellespont as the very emblem of a mortal's hybris against the gods. The fullest is Herodotus, writing a generation later, whose seventh, eighth and ninth books narrate the invasion at length and shape it throughout as the punishment of overreaching pride and the vindication of Greek freedom. Later and far less reliable are Ctesias, Diodorus, Plutarch and others. Against this rich, adverse literary tradition the Persian side sets almost nothing: no royal inscription mentions Greece, and the campaign leaves its trace in the empire's own record only obliquely, in the Persepolis administrative tablets that document the labour and provisioning behind such an enterprise, and in the archaeology of the palaces, which shows no disruption at all through Xerxes' reign. Amélie Kuhrt cautions against the distortion this invites:

Such pieces of evidence help to balance our image of Xerxes' reign, reminding us of the dangers of letting a setback suffered in one campaign along the empire's north-western fringe define an entire reign of twenty-one years.

A. Kuhrt, The Persian Empire, p. 239^[1]

The methodological rule follows: Herodotus is indispensable and often superbly informed, but his is a hostile account, to be read as an argument with a shape, not as a neutral record.

The received explanation, largely Herodotus', is that Xerxes marched to punish Athens for its part in the Ionian Revolt and the burning of Sardis, and to complete the unfinished business of his father, whose own expedition had been checked at Marathon in 490. Punishment and the extension of the empire westward were certainly proclaimed aims, and Herodotus makes the king announce them in the accents of the Persian habit of conquest:

We never yet have kept quiet since we received this supremacy in succession to the Medes, when Cyrus overthrew Astyages; but God thus leads us, and for ourselves tends to good that we are busied about many things… I carefully considered always how in this honourable place I might not fall short of those who have been before me, nor add less power to the dominion of the Persians.

Herodotus 7.8, trans. Macaulay^[2]

The most searching modern reading, John Hyland's, takes this further and inward. For Hyland the decision for a royal campaign, led by the king in person rather than delegated to a general as the earlier Aegean wars had been, had less to do with punitive obligation or strategic fear than with the demonstration of effective kingship through action, in the wake of Xerxes' suppression of the revolts of Egypt and Babylon that had greeted his accession. By retracing and enlarging the marches of Cyrus and Darius, and by crossing more water than his father had crossed, the king positioned himself as worthy of their legacy and capable of exceeding it; the expedition was "both a demonstration of restored imperial unity and world rule... and an act of comparative performance meant to establish his own legacy as the pinnacle of Persian royal achievement."^[3] On this account the older idea of a defensive or purely retaliatory Persia, reacting to an Aegean arms race, sits awkwardly with the royal inscriptions' confident claim of universal rule, and misreads a Persian worldview in which the empire was already the centre of the world and a campaign extended, rather than completed, its edges.^[3]

Herodotus does not simply report the decision to invade; he stages it, in a sequence of court scenes that are among the most elaborate set-pieces in his work and among the least reliable as history. The young king resolves on war, is argued out of it by his uncle Artabanus, is driven back to it by a menacing dream, and at last carries the Persians with him. Around this frame move the recurring types of the Herodotean court: Mardonius the reckless favourite who urges the war for his own advancement, Artabanus the prudent elder who counsels against it, and the exiled Spartan king Demaratus, who tells Xerxes the truth about an enemy the king holds cheap. The design is tragic rather than documentary, Herodotus' shaping rather than a record of the court. Artabanus puts the Greek law of overreach into a Persian mouth:

Thou seest how God strikes with thunderbolts the creatures which stand above the rest and suffers them not to make a proud show; while those which are small do not provoke him to jealousy: thou seest also how he hurls his darts ever at those buildings which are the highest and those trees likewise; for God is wont to cut short all those things which stand out above the rest.

Herodotus 7.10, trans. Macaulay^[2]

Xerxes answers with the venturer's creed, that a boldness which risks half the feared evils is better than a caution that suffers none^[2], and the dream decides the matter. The pattern returns at the Hellespont, where the king surveys the host that hides the water and blackens the shores, "pronounced himself a happy man, and after that he fell to weeping" at the thought that of all that multitude not one would be alive in a hundred years, and Artabanus answers that death is the one refuge from a life so full of grief.^[2] The scenes announce the fall before the march has begun.

Because the council is so plainly literary, the historical decision has to be reconstructed around it, and the scholarship stages the question in three overlapping readings. The first takes Herodotus' proclaimed motive, vengeance, at close to face value: the debt owed for the burning of Sardis and the check at Marathon. Revenge is the reason Herodotus supplies for almost every Persian campaign, and it has been read as a Greek gloss, the motive his audience would most readily grant; but Mesopotamian royal ideology did treat the avenging of insult as a genuine and sufficient ground for war, so the vengeance theme may record a real strand of Persian justification rather than only a Greek one.^[4] The second reading makes the war a requirement of kingship: a king newly come to the throne, fresh from suppressing the revolts of Egypt and Babylon, had to prove himself by conquest, and Greece was the nearest unconquered ground. In this form, argued by J. M. Cook, George Cawkwell and others, Xerxes "had to conquer something", and the choice of Greece followed from the logic of a monarchy that legitimised each king through his victories.^[4] The third reading distrusts the scene itself: Briant and others treat the council debate, and with it the tidy division of motive between a warlike Mardonius and a cautious Artabanus, as a Herodotean construction that cannot recover anything actually said at the Persian court.^[5][4]

The readings are less opposed than they appear. Xerxes' own speech in Herodotus already runs the motives together: the Persian custom of ceaseless conquest, the will of the god, the duty to match his ancestors, and a blow against Athens cast as pre-emptive. Those are the same justifications the royal inscriptions and their Assyrian models deploy, so a Greek text and the Persian ideology point the same way. The balance now rests there: the invasion was less the product of a single grievance than the expression of a kingship expected to enlarge the empire and to be seen doing so, of which Hyland's account of the campaign as royal spectacle is the fullest modern statement. What no source provides is a Persian record of a strategic calculation; the "grand strategy" sometimes credited to Xerxes rests entirely on the Greek narrative it claims to correct.^[4]

What impressed the Greeks most, and what they made the material of the king's impiety, were two great works of engineering carried out in the years of preparation. To spare the fleet the passage round the stormy foot of Mount Athos, where an earlier Persian fleet had wrecked in 492, a navigable channel was cut clean through the neck of the Athos peninsula, wide enough for two triremes to row abreast; modern excavation has confirmed its course near the towns of Sane and Acanthus. Herodotus, who reports that the work took three years, names its purpose exactly, and in doing so states the campaign's governing logic:

It seems to me, making conjecture of this work, that Xerxes when he ordered this to be dug was moved by a love of magnificence and by a desire to make a display of his power and to leave a memorial behind him; for though they might have drawn the ships across the isthmus with no great labour, he bade them dig a channel for the sea of such breadth that two triremes might sail through, propelled side by side.

Herodotus 7.24, trans. Macaulay^[2]

The canal was one part of a much larger programme, spread across Thrace and Anatolia, of broadened roads, stockpiled depots and river bridges, drawing on the same imperial labour dynamics visible in the Persepolis archives; a Treasury tablet records payments to Egyptian, Levantine and Greek workers on palace construction in exactly these years.^[6] The second work was the crossing itself. At the narrowest reach of the Hellespont, between Abydos and the far shore, Phoenician and Egyptian engineers flung two bridges of boats, lashed with flax and papyrus cables strained on windlasses, some 674 vessels in all moored head to stream and floored over with timber, brushwood and stamped earth, with screens along the sides so the animals would not shy at the sea.^[2] When a storm broke the first attempt, Herodotus reports the act that became the byword for the king's arrogance:

Then when Xerxes heard it he was exceedingly enraged, and bade them scourge the Hellespont with three hundred strokes of the lash and let down into the sea a pair of fetters… 'Thou bitter water, thy master lays upon thee this penalty, because thou didst wrong him not having suffered any wrong from him.'

Herodotus 7.35, trans. Macaulay^[2]

The anecdote should be read as it is meant, as a Greek fable of hybris, the yoking of the sacred water like a slave. From the Persian side, as Hyland shows, the bridging and the canal were victories over nature in the old Mesopotamian idiom of royal display, mastery of sea and mountain as a proof of world-rule; the same king who is made to whip the sea is the imperial pathfinder leading his lesser kings across waters no Persian fleet had sailed.^[7]

The host gathered in Cappadocia and moved by way of Sardis to the Hellespont; the crossing of the bridges, Herodotus says, took seven days and nights without a break. On the European side, at Doriscus in Thrace, the king held the great review and census that furnishes Herodotus with his famous catalogue of the nations, contingent by contingent, each in its own dress and weapons under a Persian commander. The scene has usually been read as a muster of a fighting army; both Pierre Briant and Hyland read it, more persuasively, as a review of the empire. Most of the forty-seven ethnic contingents Herodotus parades never appear again in his battle narratives, which name only Persians, Medes, Elamites, Saka, Bactrians and Indians; the exotic others, in their skins and feathers, are there for display. Briant draws the conclusion:

Xerxes was not interested in his military forces but in the ethnic and cultural diversity of the peoples of his empire; rather than a review of the troops, this was a review of the Empire.

P. Briant, From Cyrus to Alexander, p. 197^[5]

The review embodied in the flesh the lists of subject peoples carved in the royal inscriptions, and its most pointed element, that command of every contingent had passed from native chiefs to Persians, declared that the empire of the Great King was a Persian empire. The mobilisation was thus, in Briant's phrase, a response "to motivations that were more ideological than strategic."^[5] Hyland adds the practical corollary: the recruitment burden fell heaviest on Iran and the east for the army and on the Mediterranean seaboard for the fleet, while token delegations from remoter provinces served the needs of symbolic heterogeneity, perhaps drawn from the escorts of the tribute-caravans that converged on the court each year.^[8]

Herodotus gives the expedition figures that have astonished and misled readers ever since. At Doriscus the land army is tallied at 1,700,000 by the device of penning ten thousand men at a time inside a walled circle:

Of the whole land-army taken together the number proved to be one hundred and seventy myriads: and they numbered them throughout in the following manner: they gathered together in one place a body of ten thousand men, and packing them together as closely as they could, they drew a circle round outside.

Herodotus 7.60, trans. Macaulay^[2]

He reserves his climactic arithmetic for the eve of the first storm, adding the fleets and the reinforcements and doubling the combatants to account for the servants, transport-crews and camp-followers:

Thus five hundred and twenty-eight myriads three thousand two hundred and twenty was the number of men whom Xerxes son of Dareios led as far as Sepias and Thermopylai.

Herodotus 7.186, trans. Macaulay^[2]

That is 5,283,220 souls, with 1,207 triremes and 3,000 smaller vessels. No modern scholar defends these numbers, and the reasons are not a matter of taste. The correction of the scale is the firmest single result of the modern scholarship. Sean Manning's study of Achaemenid armed force sets the figures against the whole documentary record of the ancient Near East and finds them impossible: both the royal inscriptions and the administrative documents show states that dealt in tens of thousands of soldiers, and neither the Assyrians nor the Persians ever claimed to bring hundreds of thousands, let alone millions, to a battle. The largest army credited to any Near Eastern king in a literary source is the 120,000 that Shalmaneser III boasted of, and few Assyriologists trust even that.^[9] Comparison with better-documented periods is decisive:

[C]omparison with later periods suggests that the vast majority of numbers in literary sources are of no value in understanding the size of the armies involved.

S. Manning, Armed Force in the Teispid-Achaemenid Empire, p. 326^[10]

Manning points out that solid evidence of gatherings of more than 100,000 soldiers is extremely scarce before the Napoleonic Wars, while the claim of 100,000 or more is everywhere, and that the figures classical writers assign to "barbarians" are estimates made from outside, by men who did not share a language with the enemy's clerks.^[10] Even the popular rescue, that Herodotus or his source confused the words for "thousand" and "ten-thousand" and so multiplied the army tenfold, still leaves an army twice as big as the largest known Macedonian and Roman armies.^[10]

The credible scale is recoverable, if only within a range. The most careful scholarship converges on a land army well under 100,000 combatants, and probably far smaller. Hyland, working from the Persepolis ration texts, which fix the daily flour allowance per man and so let the documentary record speak to the numbers, offers a lowest reasonable estimate:

In summary, the lowest reasonable estimate for Xerxes's land army would add up to 50,000 soldiers, 20,000 noncombatants, and at least 30,000 animals.

J. Hyland, Persia's Greek Campaigns, p. 166^[8]

Of the fighting men perhaps 14,000 were royal guards and the rest provincial soldiers and conscripts, mostly Iranian, with the cavalry a small fraction of the whole.^[8] Kuhrt reaches a comparable figure by the opposite route, reasoning back from the fighting at Plataea:

While the precise numbers of Xerxes' invasion army must remain unknown, it is clear from the only real land battle fought in Greece (Plataea) that the effective fighting force of the Persian army was around 60,000 men.

A. Kuhrt, The Persian Empire, p. 240^[1]

The fleet is somewhat less certain, because Aeschylus was an eyewitness at Salamis; Hyland's region-by-region reconstruction, cross-checked against attested later ship-numbers, works from a minimum estimate of about 650 ships rather than Herodotus' 1,207, still large enough to have outnumbered the Greeks.^[11] The upshot is a serious correction of scale but not of significance: an army of some tens of thousands, with a fleet of a few hundred ships, was still an enormous undertaking to move and feed a thousand miles from home. The "countless horde" is a literary construction; the real expedition was formidable, finite, and stretched to the edge of what the ancient world could sustain.

The true measure of the enterprise was set less by how many men Persia could raise than by how many it could feed across a march of a thousand miles, and the preparation the Greeks remembered as engineering was, at root, a problem of supply. In the years before 480 magazines of grain and flour were laid down along the whole line of advance. Herodotus names the depots and Briant lists them: Leuce Acte on the Thracian coast, Tyrodiza in the country of Perinthus, Doriscus, Eion on the Strymon, and Macedonia, each a stocked staging-post on the road the army would take.^[5] On the march itself the burden fell on the subjects along the route, who were required to feast the king and his host as it passed, a levy Herodotus reports as the ruin of the towns that bore it. Christopher Hassan, surveying the army's logistics, reads this pairing of pre-positioned depots and requisition as the genuine Achaemenid achievement of the campaign, on a par with the supply systems that moved later armies across comparable distances; the fleet, on his account, was partly there to carry provisions by sea in step with the army on land.^[12] The daily ration is also what lets the modern head-counts be checked: Hyland's estimate of the army's size is built from the flour allowance per man in the Persepolis texts, so that a question of numbers becomes a question of tonnage.^[8]

The ancient writers grasped the constraint as clearly as the modern ones. In the council Artabanus warns the king that his two most dangerous enemies are not the Greeks but the land and the sea: no country could feed so vast a host for long, and no harbour was large enough to shelter so great a fleet from a storm.^[2] The verdict on the campaign has come round to his. What turned Xerxes back was less any one defeat than the impossibility of keeping an army of that size fed through a second and a third season on the far side of the Aegean. The logistics that made the expedition possible also fixed its ceiling.

Within this frame the famous battles fall into place as episodes of a single westward drive that very nearly succeeded politically before it failed militarily. At the pass of Thermopylae in high summer 480 a Greek force under the Spartan Leonidas held the coastal defile for some days until the position was turned by a mountain path, and was destroyed; the simultaneous naval clashes off Artemisium, indecisive in themselves, cost both fleets and were broken off when the land position fell. The road to central Greece lay open. Here the Persian achievement was as much diplomatic as military: as Kuhrt stresses, the Persians repeatedly and successfully exploited the divisions among the Greek states, so that virtually all of central Greece went over to them, the Peloponnesians retired behind the Isthmus, and Athens was evacuated and its temples burned.^[1] The turn came at sea. Drawn into the narrows of Salamis, where their numbers could not tell, the Persian fleet was beaten by the Greek ships while the king watched from a throne on the shore; among his captains that day was Artemisia of Halicarnassus, the one woman to command a squadron in the fleet. Xerxes withdrew to Sardis to direct affairs from there, leaving Mardonius with a picked force, essentially Iranian, to hold Greece over the winter and renew the offensive in 479. When the Athenians, returned to their city after Salamis, refused his offers of alliance, Mardonius marched south again in the summer of 479, occupied the evacuated town a second time, and on withdrawing into Boeotia burned and pulled down what the first sack had spared, so that Athens was destroyed twice within two years.^[2] Later that summer Mardonius was defeated and killed at Plataea, the war's one pitched land battle, and on the same reckoning a Persian army was destroyed at Mycale on the Ionian coast. With that the bid to annex Greece was over, and the western frontier settled into intermittent warfare rather than becoming a belt of subject provinces. The received story of the return, a starving and terrified host fleeing through hostile country, Kuhrt notes, is contradicted by the loyalty of Thessaly, Paeonia and Macedon and undermined by Herodotus' own report of an alternative route: it is one of several dramatised endings for the vanquished foe.^[1]

One Persian text is regularly dragged into the campaign, and it does not belong there without heavy qualification (see Xerxes I). The trilingual proclamation of Xerxes from Persepolis, the so-called daiva inscription, XPh, records that the king put down the worship of the daivas, the false gods, and established the worship of Ahura Mazdā in their place:

And among those countries there were (some) where formerly the daivas had been worshipped. Afterwards by the favour of Auramazda I destroyed that place of the daivas.

Xerxes, XPh §4b, trans. Kuhrt^[13]

It has sometimes been read as a report of a specific cultic purge in a rebel land, and by extension linked to the burning of the Athenian Acropolis or to the Babylonian revolts. Both links fail. The text names no shrine and no place; it is a commemorative statement of royal piety, and its own logic, as Kuhrt shows, elaborates Darius' teaching that rebellion against the king is rebellion against the god's order, not a chronicle of any single event.^[1] Crucially, the notion that the daiva-sanctuary was a Babylonian temple cannot stand, because it is now established that there is no evidence Xerxes destroyed any Babylonian cult-centre at all.^[1][14] The inscription is genuine and important Persian evidence for the ideology within which the reign, and so the campaign, was conceived; it is not evidence for what happened in Greece, and serves only as context for it.

The campaign is the sharpest test in all Achaemenid history of writing about an empire through its enemies. Herodotus' virtues are real: his catalogue of Persian commanders gives linguistically authentic names, many matching individuals known from the Persepolis archives; his account of the preparations squares with the labour and supply system the tablets document; his geography and his sense of Persian court practice are often sound. His governing purpose, though, is to make the Persian defeat the exemplary punishment of pride, and that purpose bends everything: the numbers swell to a scale of hubris designed "to melt away with the reversals of Xerxes's fortunes,"^[8] the king's every triumph is undercut, the Persian military achievement in overcoming Greek resistance and taking Athens is obscured, and divine favour is credited only to the Greek side.^[7] Aeschylus, staging the same events eight years after Salamis, gives the pattern its mythic form, the sacrilegious yoking of the Hellespont bringing ruin on the yoker. In the play the ghost of Darius names his son's offence:

He conceived the hope that he could by shackles, as if it were a slave, restrain the current of the sacred Hellespont, the Bosporus, a stream divine; set himself to fashion a roadway of a new order, and, by casting upon it hammer-wrought fetters, made a spacious causeway for his mighty host. Mortal though he was, he thought in his folly that he would gain the mastery over all the gods, aye even over Poseidon.

Aeschylus, Persae 745–750, trans. Smyth^[15]

That the two greatest sources are a patriotic history and a victory tragedy is the fact the reader must hold at every point. The Persian evidence cannot supply an alternative narrative, but it can and does contradict specific Greek claims, the millions, the destroyed Babylonian temples, the humiliating rout, and it reframes the whole: not the near-conquest of the world halted by providence, but a large frontier campaign, mounted to display and extend a kingship already claiming universal rule, that reached the limit of its logistics and was turned back at the edge of the map.

On the Persian side the campaign left almost no mark. No royal inscription names Greece, Athens, Salamis or the Hellespont, and the daiva text, the one document sometimes attached to the reign's western wars, in the way of all royal propaganda admits no reverse of any kind; a "Persian version" still in circulation centuries later, reported by Dio Chrysostom, went so far as to deny that Xerxes had ever been defeated.^[4] The episode that for the Greeks became the hinge of their history was, in the empire's own memory, a frontier operation that ended without conquest and left the centre untouched. The war looms so large in the record because the side that lost the campaign wrote its history at length, while the side that kept its empire did not think the reverse worth commemorating. That silence is a caution against reading a long and otherwise secure reign, as the Greek tradition did, through its single western defeat.

Set in the empire's own scale, the invasion of 480–479 was neither the apocalyptic collision of the Greek imagination nor the ruin of Persia. It cost the empire nothing of its heartland, and its administrative and monumental record shows the state running in good order for a further twenty years. The old verdict, inherited whole from the Greeks, that Salamis began the long Persian decadence depends on the very sources whose bias the modern reassessment has exposed.