The Battle of Plataea (479 BCE)

In the summer of 479 BCE, on the plain of the shallow Asopos river in Boeotia, below the ruined town of Plataea, the army that Xerxes had left in Greece under his kinsman Mardonius was destroyed, and Mardonius with it. It was the last and the decisive battle of the great invasion. After the naval defeat at Salamis the previous autumn the King had withdrawn to Sardis, leaving a picked force to winter in Thessaly and finish the conquest in the spring; Plataea is where that plan failed. The Persian army broke and fled, the survivors were penned in their fortified camp and killed, and what remained withdrew from Europe. On the same reckoning, and by Herodotus' account on the very same day, a second Persian force was beaten at Mycale on the Ionian coast. With those two defeats the attempt to annex the Greek mainland ended, and the western frontier became a zone of intermittent war rather than of new provinces.

Plataea has a particular importance for this compendium, and a particular difficulty. It is the one full-scale land battle of the whole war, and so the clearest test we have of the empire's army against the massed Greek hoplite; Salamis and Artemisium were fought by the subject navies, and at Marathon a decade before, the Persian narrative is even thinner. Yet Plataea too comes to us almost entirely from Herodotus, writing around the 440s, from the winners' side, and with a design that shapes every page. His Mardonius is a marked man, a reckless deceiver who rejects wise counsel and drags the army to its doom; his Persian numbers are impossible; and he attributes the defeat to a supposed inferiority of Persian arms and discipline that the material evidence does not bear out. To read Plataea from the Persian side is not to deny the defeat, which was real and heavy, but to strip away the freight the victory carried and ask the harder questions: how large the force actually was, what it was made of, how Mardonius fought the campaign, and why, having very nearly won it, he lost the battle.

The one source, and its design

Everything of consequence about Plataea is in Herodotus, Book 9, and there is no counter-current. The royal inscriptions are not annals and say nothing of Greece; the later Greek writers who touch the battle (Diodorus from the fourth-century Ephorus, Plutarch in his Aristeides, Ctesias) descend from the same stream and add little that is independent. The Companion's survey of the Persian Wars states the resulting problem squarely: modern accounts of these campaigns are for the most part "rationalized reproductions of Herodotus' account without raising serious doubts about the quality and perspective of the source," and the Persian view of these events "is entirely disproportional compared with the Greek and western perspective."^[1] The whole war, in that reading, is a Greek master-narrative of Asian hybris and divinely sanctioned Greek freedom, and Plataea is its climax.

Herodotus' handling of Mardonius is the clearest case of design. As John Hyland shows, the historian "marks Mardonios as a villain from the start, crediting him with qualities of deception and love of trouble-making antithetical to the Achaemenid ideological values of truth and order."^[2] After Xerxes withdraws, Mardonius is made a kind of substitute king whose death is foreshadowed at every turn, a figure who, like Xerxes and Croesus before him, rejects the "wise warner" and misreads the oracles he consults. Much of this is retrospective: the blame heaped on Mardonius in Herodotus "is rooted in subsequent efforts to blame him for the outcome," a scapegoating shared by the Persian court and the medizing Greeks alike.^[2] The facts Herodotus preserves are often reliable; the frame around them is a moral fable, and the two have to be pried apart.

After Salamis: the King's plan, not a rout

The popular picture, taken from the Greek sources, is of a King fleeing in terror and abandoning a doomed rump of an army. That is not what happened. As Pierre Briant stresses, "despite the resounding defeat, the military outcome" of Salamis "was not catastrophic. The Persian army was practically intact," and the decision that followed was a considered division of force, not a panic: "Mardonius was given the job of pursuing the offensive in Greece, with the army. Xerxes returned to Sardis, along with the navy."^[3] From Sardis the King remained "in constant communication with Mardonius," and "continued to oversee the entire operation." Delegating the completion of a war to a general while the King began his homeward progress had ample Achaemenid precedent, and would, had Mardonius won, have added to the King's own credit exactly as the victories of Darius' lieutenants had at Behistun. The retreat, in short, was a strategy: leaving Mardonius in Greece kept the Greek fleet pinned in home waters and ensured that "the decisive battle would take place on Greek soil."^[3]

Mardonius wintered in Thessaly on the support of Greek clients, from the dynasts of Larissa to the Theban oligarch Attaginus. The feat of feeding an army through a Greek winter without the imperial infrastructure that had supplied the previous winter's camp at Sardis is itself a mark, Hyland notes, of Persian logistical skill and of the sacrifices required of the new Greek subjects.^[4] It is worth pausing on how far this was, by the standards of the ancient Near East, an unprecedented thing to do. As the Companion observes, the Persians wintering in Greece to renew the campaign is "the first example in the history of Ancient Near Eastern empires, as far as we know, that their armies did not return home after an annual campaign but stayed on the 'peripheries' to continue their advance in the next year. This does not look as if the Persians suffered serious setbacks in 480."^[1]

The diplomacy: the offer to Athens

Before he marched, Mardonius tried to win the war without a battle. In the spring of 479 he sent Alexander I of Macedon, the King's client and a formal guest-friend of Athens, to carry a royal peace offer: autonomy, help rebuilding the burned temples, and additional territory of Athens' choosing, in return for submission. Briant preserves the terms as Herodotus gives them, in the King's own instructions to Mardonius:

"The king's orders to him are, first, to restore to Athens her territory, and, secondly, to allow her to choose in addition whatever other territory she wishes, and to enjoy her liberty. Let Athens but come to terms with the king, and he has his instructions to rebuild the temples which have been destroyed by fire." (Herodotus 8.140, trans. after Rawlinson, in Briant 2002)^[3]

The offer was not weakness. As Hyland argues, Mardonius' ability to threaten Athens again if it refused undercut any reading of the overture as a sign of Persian exhaustion; the aim was a diplomatic wedge, to "raise disagreement between Athenian hardliners and adherents of compromise" and to sow mutual suspicion between Athens and its Spartan and Aeginetan allies.^[4] Had Athens accepted, the shame of Salamis would have been washed away and replaced by "the perception of a fearsome enemy's pacification through a belated recognition of royal might and mercy." Athens proved more united than the Persians wished and refused twice, stoning a councillor who so much as proposed putting the offer to the assembly. Mardonius then retook and razed the evacuated city, signalling the recapture to Xerxes at Sardis by fire-beacons, before withdrawing to fight on ground of his own choosing.

Why Boeotia: the cavalry campaign

Mardonius' choice of the Asopos plain in Boeotia was deliberate and shrewd, and it is the key to the whole campaign. He fell back from Attica, Herodotus says, because "Attica is not cavalry country," and withdrew "to Thebes, to fight near a friendly city and in country suited to cavalry."^[5] Boeotia offered ample grazing and room to manoeuvre, and, in the event of victory, narrow passes over which the beaten enemy would have to retreat, maximising the chance for mounted pursuit. There, in the territory of the Thebans who had joined him, he cleared the ground for a fortified camp:

"He proceeded to cut down the trees in the lands of the Thebans, although they were on the side of the Medes, moved not at all by enmity to them, but pressed by urgent necessity both to make a defence for his camp, and also he was making it for a refuge, in case that when he engaged battle things should not turn out for him as he desired." (Herodotus 9.15, trans. Macaulay)^[6]

When the Greek coalition under the Spartan regent Pausanias came down over Mount Kithairon and took up a strong position on the low hills opposite, the two armies settled into a standoff of nearly a fortnight. Throughout it Mardonius used the weapon in which he held a clear advantage. His cavalry, absent from the enemy order of battle, patrolled the open ground and harassed the Greeks with bow and javelin, denying them the river water and forcing them back on a single spring. Hyland's judgement reverses the usual picture of a general blundering toward disaster: "Mardonios's actions throughout the campaign suggest careful attention to the pre-conditions for battle... the Persian dominance in cavalry, absent from the enemy army, provided the capability to exploit Pausanias's more precarious supply situation. He appears to have been eager for a fight, but not until the enemy's logistical breakdown paved the way for a successful outcome."^[4] On the eighth day the cavalry intercepted a Greek supply column coming down from the pass; on the twelfth, in a decisive stroke, Persian riders fouled and spoiled the Gargaphia spring, the coalition's chief water source. It was this that forced Pausanias to order the night withdrawal from which the battle grew. On the eve of the fighting, Hyland concludes, "Mardonios's cavalry was on the verge of winning the campaign."^[4]

The cost of that advantage was one memorable loss. Early in the standoff the cavalry commander Masistius, second in esteem only to Mardonius, was thrown when his horse was shot, and killed by Athenian infantry before his men could rescue him; the whole army mourned him, Herodotus says, shaving their own heads and their horses' manes, so that the wailing "filled the whole of Boeotia."^[7] The episode was a warning that cavalry could harass but could not break massed hoplites in good order, a lesson Mardonius took and did not repeat.

The false problem of supply

A long tradition, following certain hints in Herodotus, holds that Mardonius was in fact running out of food, and gave battle prematurely out of desperation. Hyland subjects this to close scrutiny and rejects it. The passages usually cited (a claim that Alexander of Macedon secretly warned the Greeks that Mardonius was low on supplies, a vague remark in Thucydides, some Phocian raiding from Mount Parnassos) do not bear the weight. Against them stands the report, which Herodotus himself puts in the mouth of Artabazus, that there was "plenty enough food stored at Thebes to support a long-term encampment"; Boeotia, entering the campaign a month after its own harvest, could readily feed an extra fifty thousand mouths.^[8] Any anxiety about depletion, Hyland concludes, "entailed calculations of the future impact on his Boiotian partners rather than the troops themselves," and the famous capture of the Greek supply train was a windfall, not a lifeline.^[8] Mardonius' position was the exact opposite of Pausanias', whose army could draw nothing from the burned countryside and had to haul everything over Kithairon from the Peloponnese. The general kept the choice of when and how to fight.

The deeper reason he chose to fight when he did was not hunger but ideology and opportunity together. When Herodotus has Mardonius reject Artabazus' counsel of caution and prefer "to give battle according to the Persian custom," Hyland writes, "one thinks of Darius's pride in the nineteen battles won in a single year through the will of Ahuramazda."^[8] For a Persian noble, and for a campaign meant to redeem the King's honour, there was "no better way to reverse the impression of divine abandonment in the battle that had gone wrong in the royal presence than to reattempt the ordeal in hopes of Ahuramazda's renewed protection." When Pausanias' night retreat left the Greek army split into three separate groups, sleepless and short of water, the moment looked ripe. Mardonius ordered the pursuit.

The army and the medizing Greeks

Before the battle it is worth being plain about what the army was, because the received numbers are worthless. Herodotus gives Mardonius 300,000 men, besides the Greek levies. The figure belongs to the same literary habit, exposed by Sean Manning across the whole documentary record of the ancient Near East, that assigns "barbarians" armies of hundreds of thousands to millions where the states themselves dealt in tens of thousands.^[9] The credible order of magnitude is far lower. Hyland's reconstruction, worked from the ration requirements and the eventual order of battle, gives Mardonius "35,000 to 40,000 combat troops, including some or all of the Immortals and both infantry and cavalry from Persia, Media, Bactria, and the Saka and Indian frontiers," with the cavalry perhaps 4,000 strong, of whom the thousand best were picked out by Mardonius himself.^[10] Kuhrt, following Briant, reaches the same order from the Greek side of the field: "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."^[11]

The composition matters as much as the count, and it is the fact the Greek framing most obscures. This was not "Persians" against "Greeks." Between a quarter and a third of Mardonius' army was Greek: hoplites and cavalry called up from every medizing community from Macedonia through Thessaly to Boeotia, and above all from Thebes, whose oligarchs were the mainstay of the Persian cause in central Greece.^[4] At Plataea these Greek subjects held a third of the line, drawn up opposite the Athenians, Plataeans and Megarians, while the Iranian contingents faced the Spartans and the rest. Their participation, as Hyland puts it, was "orchestrated as a climactic proof of the loyalty that had already fed the Persian army for so many months, bringing together Greek friends and former opponents" to witness the empire's power to punish resistance and reward cooperation.^[4] The Companion draws the general point: "from the Thessalians in the north to the Thebans in Boeotia, Greeks fought on the side of the empire."^[1] The word "Persian" for the army beaten at Plataea is, once again, a category the Greek narrative supplies for its own purposes.

The cost to those Greek clients is caught in one of the most affecting passages in all of Herodotus, and one of the few that lets a Persian voice be heard. At Attaginus' banquet at Thebes, on the eve of the campaign, a Persian shared a couch with a Greek of Orchomenos and, in tears, told him what he foresaw:

"Dost thou see these Persians who are feasting here, and the army which we left behind encamped upon the river? Of all these, when a little time has gone by, thou shalt see but very few surviving... that which is destined to come from God, it is impossible for a man to avert... And these things which I say many of us Persians know well; yet we go with the rest being bound in the bonds of necessity: and the most hateful grief of all human griefs is this, to have knowledge of the truth but no power over the event." (Herodotus 9.16, trans. Macaulay)^[12]

The speech is Herodotus' composition, and it serves his theme of a doom foreseen; but it also grants the men of the King's army an interior life, a foreboding and a fatalism, that the rest of his narrative denies them.

The battle, and the death of Mardonius

When the Persians came on, the fighting fell into two separate actions. On one wing the medizing Greeks, the Thebans foremost, closed with the Athenians and fought hard, losing three hundred of their best before they broke. On the other, the crux of the day, the Persian infantry under Mardonius met the Spartans and Tegeans. Herodotus' account of that clash is detailed and, for once, largely free of Homeric ornament: a long archery barrage from behind a barricade of wicker shields, then the Spartan advance, hand-to-hand fighting along the barricade until it fell, and a ferocious struggle by the temple of Demeter. What Herodotus emphasises is the courage of the Persians and, at the same time, their supposed helplessness:

"Now in courage and in strength the Persians were not inferior to the others, but they were without defensive armour, and moreover they were unversed in war and unequal to their opponents in skill; and they would dart out one at a time or in groups of about ten together, some more and some less, and fall upon the Spartans and perish." (Herodotus 9.62, trans. Macaulay)^[13]

The battle turned on Mardonius himself:

"In the place where Mardonios himself was, riding on a white horse and having about him the thousand best men of the Persians chosen out from the rest, here, I say, they pressed upon their opponents most of all: and so long as Mardonios survived, they held out against them, and defending themselves they cast down many of the Lacedemonians; but when Mardonios was slain and the men who were ranged about his person, which was the strongest portion of the whole army, had fallen, then the others too turned and gave way." (Herodotus 9.63, trans. Macaulay)^[13]

Mardonius' death is the hinge. A Persian noble, unlike a king who could watch from a hillside, was expected to prove his courage in person, and riding a white horse at the head of his guard made him conspicuous and made him a target.^[14] When he fell, the elite corps around him was destroyed and the army's spirit broke; and the medizing Greeks, seeing the Persians in flight, ran before they had fully engaged. Herodotus makes this the proof of his thesis that the whole enterprise rested on the Persians alone:

"And this is an additional proof to me that all the fortunes of the Barbarians depended upon the Persians, namely that at that time these men fled before they had even engaged with the enemy, because they saw the Persians doing so." (Herodotus 9.68, trans. Macaulay)^[15]

One large body of the army never fought at all. Artabazus, who had counselled against the battle, held his division back, and when he saw the day lost led it away northward at speed toward the Hellespont; he brought the greater part of his men home across Thrace, and was afterward rewarded by Xerxes with the satrapy of Daskyleion, which suggests the King was satisfied with his management of the retreat whatever Herodotus says of his motives.^[14] The Persians who did not escape fled to the fortified camp, where the Athenians, expert at storming walls since the Ionian wars, broke in and killed them; three thousand, Herodotus says, were taken alive. Pausanias took Mardonius' camp and the tent Xerxes had left him, and marvelled at its gold and silver.

The anatomy of a defeat

Why, then, was the battle lost, if the campaign had gone Mardonius' way? Here the Persian-side reading and the older Greek-shaped one diverge, and the divergence is a live scholarly question. The traditional explanation, which Herodotus supplies and which Briant to a considerable degree accepts, is material and tactical: once the Greek hoplites carried the wicker barricade, the Persian infantry, "deficient in armour, untrained, and greatly inferior in skill" (Herodotus 9.62), could not stand against heavily armed spearmen.^[16] On this account the Persians had not grasped the implications of the Greek "hoplite revolution," and lost, at bottom, because a light-armed archer-army was outmatched by the phalanx in the close fight. Briant qualifies this in one crucial way that is itself a Persian-side correction: Herodotus "cannot be referring to the Persians proper," whose upbringing made them formidable fighters; and, more important, the defeat was not Mardonius' personal blunder but structural. "It is likely that any Persian general in the same situation would have made the same tactical decision, quite simply because this decision corresponded to requirements relating to the armaments and the combat style of the royal army."^[16] The army fought the way it was built to fight; the failure, if failure it was, lay in a system, not a man.

Recent scholarship has gone further and questioned the material premise itself. Hyland's reconstruction, drawing on the arms and armour recovered from Persian sites, on the seal-images of combat, and on reconstructive testing, argues that Persian infantry were not the naked victims of Herodotus' account. Body armour of layered linen or scale, tall wicker gerra shields and smaller oval ones, spears no shorter than the Greek, and swords for the close fight were all standard; the picture of helpless men "like naked men against hoplites" is a Greek trope, sharpened by Herodotus perhaps in deliberate inversion of the heroic-nudity motif in Athenian vase-painting.^[14] Manning reaches the same verdict on the specific charge that has done most work in the tradition, that the Persian spears were too short: "Herodotus' claim that the Persian spears at Plataea were too short is not supported by contemporary art," and the assertion is "embedded in a rhetoric of abuse" that modern readers should treat sceptically.^[17] Herodotus himself, Manning notes, says the Persians lacked "equipment and training," not manhood; whether anoploi, his word, means "without armour" or merely "without shields" is disputed, and the shields may have gone into the barricade.^[17]

If the equipment was adequate, the defeat needs a different explanation, and Hyland locates it in the conduct of the final advance rather than in any inferiority of the Persian soldier. Pursuing the retreating Greeks, Mardonius pushed his infantry down a single road, which caused the column to congest and the rear to straggle; by the time he engaged the Spartans he had brought only a third to a half of his infantry into action, no more than matched Pausanias' force, and could not use his numbers to outflank. Artabazus' whole division, caught in the traffic, never arrived. "Defeat arose not from inferiority in armament, but from the failure to concentrate sufficient soldiers at the decisive point, and as in many battles, from a combination of tactical errors and simple bad luck."^[14] The last of these was Mardonius' own death, an accident of the close fighting that, in Persian terms, would have read as the withdrawal of divine favour, and that broke the army when it might otherwise have held. "It was not foolish for Mardonios to react to Pausanias's retreat by giving battle," Hyland concludes; "it is tempting to wonder whether the Persians could have rallied to secure a different outcome if Mardonios had only ducked."^[14]

Mycale, and the withdrawal from Europe

Herodotus reports that on the same day as Plataea the Greek fleet destroyed a second Persian army at Mycale, on the Ionian shore below Samos, where the ships beached and the troops Xerxes had left to guard Ionia under Tigranes were overwhelmed. The synchronism is a literary flourish and probably not literally true, but the two events were close in time and their combined effect was decisive: the eastern Aegean fell open, and Herodotus dates from Mycale what he calls the "second Ionian revolt," as Samian, Chian and other contingents turned on the King.^[18] The two defeats together, not Salamis alone, ended the invasion. Artabazus' surviving column made its long way back across Thrace to the Bosporus; Xerxes, who had remained at Sardis through the summer, now turned inland, delayed on the way by a fresh revolt in Babylonia that the defeats may have helped to trigger.^[16]

It is important, in weighing what Plataea decided, not to overstate it in the Greek manner and not to understate it. Briant's account of the losses is deflationary: the impression of catastrophic Persian casualties "basically comes from Aeschylus," and "these losses were not in any way decisive: the Persian people continued to thrive."^[19] The territorial reckoning is where the real loss lay, and even that was bounded: the Aegean islands and the European foothold, Thrace and Macedon, were gone, but "when Xerxes left Sardis, and even at the end of 479, the losses in Asia Minor proper were minimal."^[19] The empire's heartland was untouched, and the administrative record shows the state functioning in good order for two decades after. What Plataea destroyed was the particular project of adding Greece, and, less tangibly, it compounded the wound Salamis had opened in the royal image: the campaign meant to redeem the King's honour by a fresh proof of divine favour had instead produced a second, worse failure of that proof, this time on land and in the person of the King's own general and kinsman.

Why the reading matters

Plataea concentrates the compendium's method under the hardest conditions. It is the empire's one great land battle, the place where the army the inscriptions boast of met the phalanx in the open and was beaten; and it is known from a single hostile source that made the beaten general a villain, the beaten soldiers helpless cowards, and the whole a parable of pride punished by the gods. The battle was real and the defeat was heavy. But the Persian-side account, built from the same Herodotus read against his grain and set beside the material and comparative evidence, tells a different story from the legend: an army of some tens of thousands, a large part of it Greek, under a general who fought a shrewd cavalry campaign and had all but won it by cutting his enemy off from water; a Persian infantry no worse armed than its opponents; and a defeat that turned less on any inferiority of the Persian than on a congested advance, a fraction of the force engaged, and the chance death of the commander on his white horse. That the empire could lose such a battle, and remain the empire, is the point the victors' sources are least able to see, and the one the Persian reading most needs to keep in view.