Friday, March 7, 2014

The battle of Plataea[1]

Antonis Mystriotis

The copyright of this website belongs to the author. Any use of this material bares the obligation of a proper citation.


The battle of Plataea, which was the greatest land battle of the classical era, is presented under a new perspective. The exact words of Herodotus are carefully analysed in order to better understand the battle tactics and reason the outcome. It is suggested that during the last day of the battle, the Greek commanders, while they came under serious pressure by the Persian cavalry, devised an ingenious stratagem for trapping the Persians in a region of the battlefield which was not suitable for cavalry attacks. The main concept of this complex strategic plan was that the Spartan regiment assisted by the Tegeans was separated from the bulk of the Greek army assuming the role of bait, while the Athenians guided by the Plataeans hid among the hills of Asopos Ridge. The Persians not seeing the Athenians took the risk to attack the isolated Spartans in a region of the battlefield which was unfavourable for the cavalry expecting that their numerical superiority was sufficient for the victory. When the hiding Athenians attacked, the Persian line collapsed, Mardonius was killed, and the Greeks won a brilliant victory.  

1.     Introduction

Despite the fact that the last book of Herodotus’ Histories is mainly dedicated to the battle of Plataea, its details remain obscure and uncertain. Although Herodotus presented the battle in length, his narration is vague and self-contradicting in many points. Recent scientific works (Shepherd, 2012; Cartledge, 2013) further clarified an older historical analysis (Green 1996), and revived the interest for this great battle in which the freedom of Greece was decided. However, many questions still remain open. In particular the events of the last day of the battle are difficult to explain, although they are documented in detail by Herodotus. In this article we attempt to interpret Herodotus’ story by carefully analyzing his exact words. A few interesting conclusions can be reached by following this approach.

2.     Preparations

The battle of Plataea decided the final defeat of the Persian army in Greece and the definite liberation of all Greece. It was the greatest land battle which took place in Greece during the 5th century BC. Approximately 150,000 men[2] participated in the battle from both sides and it lasted for about 15 days. Considering that the Greek heavy infantry included more than 35,000 hoplites, it was one of the greatest battles of all times involving heavily armoured infantry forming a phalanx. Moreover, the battle was decided on very delicate tactical moves which are worthwhile to study. 
It has to be noted that most probably the battle of Plataea was not foreseen in the original strategic plan of Themistocles for the second Persian war[3]. Indeed, the Athenians were expecting that Mardonius would withdraw from Greece as soon as was informed about the Greek naval expedition to Ionia. That campaign was setting the Persian possession of Minor Asia in danger and the communication and supply routes of Mardonius’ army in great risk. Therefore, the most reasonable scenario was that Mardonius would rush to Ionia to defend the Persian strongholds there. However, Mardonius against all expectations decided to stay in Greece and attempted to defeat the Greeks in a decisive battle. 
The Spartans after many hesitations and under pressure by their non-Peloponnesian allies, Athens, Megara, and Plataea, decided to dispatch a strong force of 5,000 Spartans assisted by a large number of helots and 5,000 perioeci[4] hoplites. This army, which roughly corresponded to half of the total manpower of Sparta, marched to Isthmus. There, they were joined by other Peloponnesian allies, who wished to join the expedition.
Once the Lacedaemonians had reached the Isthmus, they established their camp there. When the rest of the Peloponnesians (that is, the ones who were on the patriotic side) heard what the Lacedaemonians were doing, they decided that they should not be left behind— although some of them did so only when they actually saw the Spartiates taking to the field. (Herodotus, IX.19)
In the mean time, Mardonius, who had occupied Attica and burnt Athens for a second time, learnt about the Greek preparations and returned to Boeotia.
Next Mardonius heard that the Greeks had gathered in force on the Isthmus, so he turned back again and marched through Decelea, … to Tanagra, where he bivouacked for the night before making his way the next day to Scolus, which was within Theban territory. At Scolus, despite the fact that the Thebans were on the Persian side, he cut down all the trees on cultivated plots of land, not out of hostility, but because he really had no choice, since he wanted to have a defensive stockade built for his troops as a place of refuge in case the battle did not go well for him. His army was posted along the River Asopus, occupying an extensive stretch of land from Erythrae, past Hysiae, and all the way into Plataean territory. However, the actual stronghold which he built was not of course as large as that; it was square in shape, with each side being about ten stades[5] long. (Herodotus, IX.15)
The identification of the exact location of the Mardonius’ camp is a very important element for the correct analysis of the battle tactics. Despite the information provided by Herodotus, most of the landmarks he referred to, have either vanished or changed their names. Specifically, Erythrae and Hysiae were two towns which had been abandoned and ruined since the Roman times. Pausanias the geographer passed by their ruins in the 2nd century AD. Today, the exact location of these two towns is unknown. However, Pausanias reported that their ruins lay by the exit of the Eleutherae pass in the plain near Plataea. Therefore, the probable location of Hysiae is modern Kriekouki (Map 1) at the west side of the Eleutherae pass and Erythrae was probably located east of that road. Considering that Mardonius’ camp lied north of Asopos River, the information given by Herodotus indicates that the Persian army spanned the area between the roads Thebes – Erythrae and Thebes – Plataea (Map 1). The most probable location of the fortified Persian camp was just west of the road Thebes – Erythrae, north of Asopos.
The Peloponnesians marched from Isthmus to Eleusis, where they were joined by 8,000 Athenian troops, and after crossing Cithaeron Mountain entered Boeotia.
… the combined (Greek) forces made their way from the Isthmus to Eleusis, where they once again offered up sacrifices. The omens were favourable, and so they continued on their way, accompanied now by the Athenian troops, who had crossed over from Salamis and joined them at Eleusis. It was when they were at Erythrae in Boeotia that they found out about the Persian encampment on the Asopos, and then, in the light of this information, they took up a position opposite the Persians on the spurs of Mount Cithaeron. (Herodotus, IX.19)

3.     Phase 1 – Days 1-2

As soon as the two armies faced each other, Mardonius attempted to irritate the Greeks by a vehement cavalry attack in order to lure them to advance into the plain.
The Greeks persistently refused to come down from the hills to the plain, so Mardonius sent the whole of his cavalry against them. The cavalry was commanded by an eminent Persian called Masistius …, who rode a Nesaean horse, magnificently caparisoned, with equipment that included a golden bit. The horsemen advanced towards the Greeks and then attacked regiment by regiment. During these attacks, which inflicted severe losses on the Greeks, they taunted them by calling them women. (Herodotus, IX.20)
After a fierce battle between the Persian cavalry and the hoplites of Megara and Athens assisted by archers, Masistius fell off his horse and was killed. The Persians were left leaderless and decided to retreat to their camp. The victorious Greeks advanced cautiously in the lower hills towards Plataea.
Greek morale was considerably raised by the fact that they had not only withstood the assaults of the Persian cavalry, but had actually managed to push them back. … Next, they decided to come down off the hills and go to Plataea, because the land around there was clearly far more suitable than Erythrae as a place for them to establish themselves, for a number of reasons, including the fact that it had a better supply of fresh water. So they decided to move their camp down to Plataea, and in particular to the spring called Gargaphia which rises in that region, and to take up a position there in separate units. They picked up their gear and marched through the spurs of Cithaeron, past Hysiae, and into Plataean territory, where they formed up into various units based on their places of origin. The land they occupied by the Gargaphian Spring and the precinct of the hero Androcrates consisted of knolls and level ground. (Herodotus, IX.25)
The exact position of the Greek camp has to be derived from Herodotus’ description despite the difficulty in recognizing vanished or renamed landmarks, such as Gargaphia spring. It is widely believed that Gargaphia is the spring Alepotripa near Plataea. Nevertheless, P. Green in his work “The Greco-Persian wars” identifies it as another spring called Retsi near the Byzantine chapel of St. Dimitris, correlating this chapel with the location of monument of Androcrates or the shrine of Demeter mentioned by Herodotus. Both possible locations of Gargaphia indicate that the Greeks, after their victory against the Persian cavalry, camped in a hilly area called Asopos Ridge, south of the river and west of the road going from Erythrae to Thebes (Map 1).  Asopos Ridge is indicated on Map (1) by the line marking the 310m elevation. Considering that the altitude of the plain south of Thebes towards Plataea is roughly 270-290m above the sea level, the small hills forming the complex of Asopos Ridge rose 30-60m above the plain. Moreover, the riverbeds of a few tributaries of Asopos, which were dry during the summer, cut deeply the landscape making the terrain very uncomfortable for a cavalry charge. Since Mardonius’ camp was established north of Asopos, we may assume that the two camps were facing each other separated by the river.
Afterwards the Greeks— the late arrivals as well as the original members of the expedition— formed up at their posts. The disposition of the forces was as follows. A brigade of 10,000 Lacedaemonians held the right wing; 5,000 of these were Spartiates, and they were protected by 35,000 light-armed helots, seven for each man. The Spartiates reserved the place next to them for the Tegeans[6], in recognition of their prestige and their courage; the Tegean contingent consisted of 1,500 hoplites. The position next to the Tegeans was taken by the 5,000-strong contingent of Corinthians, who gained Pausanias’ permission to have the 300 from Potidaea in Pallene stand alongside them. Next came 600 Arcadians from Orchomenus, then a contingent of 3,000 from Sicyon. Next to the Sicyonians came 800 men from Epidaurus, then 1,000 from Troezen, 200 from Lepreum, 400 from Mycenae and Tiryns, and then 1,000 from Phleious. Alongside the Phleiasians stood 300 men from Hermione, then there was a contingent of 600 from Eretria and Styra, then 400 from Chalcis, and then 500 from Ambracia. Next to the Ambraciots stood 800 men from Leucas and Anactorium, and then 200 from Pale in Cephallenia. The next contingent consisted of 500 from Aegina, then there came 3,000 Megarians, 600 Plataeans, and finally, in the forward position on the left wing, there were 8,000 Athenians under the command of Aristides the son of Lysimachus. (Herodotus, IX.28)
This is a peculiar and confusing description of the Greek camp. Herodotus presents the Greek camp similarly to an army arranged in battle order ready to fight. Camping in a linear order as described above would expose the Greek army to cavalry attacks for the following ten days. Moreover, some of the Greek regiments would have to camp in the plain west of Asopos ridge, thus they would be constantly exposed to cavalry attacks. Therefore, it may be assumed that Herodotus’ story reflects the battle order at a later stage, not the actual camp arrangement. In reality the Greek camp should have been concentrated on the hills of Asopos ridge in a dense well protected formation. Gargaphia spring must have been inside that protected area allowing all Greek units to have access to the necessary water supplies. 
The Lacedaemonians were the only ones posted right by the spring; all the other Greeks had some way to go to reach it (exactly how far depending on where each contingent was deployed). They did not have far to go to the Asopos, but the Persian cavalry with their bows and arrows had made it impossible for them to fetch water from the river, and so, with the Asopos denied them, they had been going to the spring. (Herodotus, IX.36)

Map 1: The battlefield of Plataea. The green area represents land below the 310 m elevation above sea level. It represents flat terrain (altitude 270 – 310 m), which was most suitable for cavalry charges. The hilly region of Asopos Ridge, which is marked with dark texture, rises 20-60 m above the plain. The two main roads connecting Thebes with Plataea and Erythrae-Hysiae respectively are marked with yellow lines. The location of the “Island” is roughly marked on the map since the Oeroe riverbed cannot be recognized with accuracy in the present day.

Therefore, Herodotus’ comment, that only the Spartans had taken position near the spring, again concerned a later stage of the battle when the enemies were preparing for a clash. Indeed, the two enemies challenged each other attempting to lure their opponents to march into unfavorable territories. However, both commanders were prudent enough not to cross Asopos.

4.     Phase 2 – Days 3-9

The day after they had all taken up their positions, people by people and regiment by regiment, both sides offered up sacrifices. (Herodotus, IX.33) but …The entrails gave favourable omens for the Greeks if they remained on the defensive, but not if they crossed the Asopos and took the fight to the enemy. (Herodotus, IX.36)
Mardonius was under pressure to take the offensive since his position was strategically weak. He knew about the Greek naval campaign to Ionia. If that expedition succeeded to set Hellespont under Greek control, his communication and supply lines would have been terminally cut.  
Although Mardonius wanted to be the one to attack, the entrails also gave him omens that were favourable for defence but unsuited to attack. He too used the Greek method of divination by examination of the entrails, (Herodotus, IX.37)
However, Mardonius was not succeeding in luring the Greeks to make a wrong move. Therefore, he had no other option but to delay his attack.
Since the omens received not just by the Persians themselves but also by the Greeks in the Persian army … warned against engaging the enemy, and because there was a constant influx of men into the Greek army, which was consequently increasing in size, a Theban called Timagenidas the son of Herpys advised Mardonius to patrol the passes over Cithaeron, on the grounds that a great many of the Greeks who were constantly flooding in every day could be caught there. (Herodotus, IX.38)

5.     Phase 3 – Days 10-14

Additionally, the Thebans advised Mardonius not to attempt an attack against the hills of Asopos Ridge where the superiority of the Persian cavalry would not have been effective. As an alternative, they suggested him to try to cut the supply routes of the Greek army at Cithaeron hoping that such an action would irritate the Greek commanders and provoke a counterattack in the plains along the two roads Plataea-Thebes or Erythrae-Thebes, where the Greek phalanx would be vulnerable to cavalry attacks.
The two sides had been facing each other for eight days when Timagenidas put this suggestion to Mardonius. Realizing that it was a good idea, Mardonius sent his cavalry that night to the pass on Cithaeron which leads towards Plataea— the pass known to the Boeotians as Three Heads, and to the Athenians as Oak Heads. This was an effective mission for the Persian horsemen, because they captured fifty yoke-animals (along with their carters) as they were coming down on to the plain with food from the Peloponnese for the Greek army. But once they had taken this prey the Persians turned to indiscriminate murder, slaughtering humans and animals alike. When they had had their fill of killing, they rounded up the remnants of the baggage train and drove them back to Mardonius and the Persian encampment. (Herodotus, IX.39)
The success of the Persian attack certainly made the Greek commanders nervous. The next day they attempted to block the road from Thebes to Plataea by a solid line of hoplites spanning all the plain between Asopos Ridge and Pyrgos Hill. Probably, the battle order described by Herodotus above refers to that phase of the battle.
Two more days passed after this incident, with neither side being prepared to start the battle. The Persians advanced right up to the Asopus to test the Greeks, but neither side actually crossed the river. Mardonius’ cavalry, however, was constantly attacking and harassing the Greeks, because the Thebans (who were staunch and belligerent supporters of the Persian cause) kept guiding the cavalry to within striking distance of the enemy, at which point the Persians and Medes would take over and perform deeds of valour. (Herodotus, IX.40)
In spite of the strong pressure of the Persian cavalry at the phalanx, the Greek hoplites held their position, although with many casualties.
Then Mardonius the son of Gobryas and Artabazus the son of Pharnaces, who was one of Xerxes’ particular favourites, held a meeting and discussed the situation. (Herodotus, IX.41)
Then in the eleventh day after the Greeks camped on Asopos Ridge (probably the twelfth or the thirteenth day of the battle), the two top commanders of the Persian army, namely Mardonius and Artabazus had a long discussion on the strategy to be followed. The reason for this discussion might be new information coming from East concerning the Persian naval defeat in the battle of Mycale. Mardonius became now desperate for a quick victory, otherwise his communication route to Asia was in severe danger. The strategic disadvantage of the Persian position was heavily weighing in his mind and could easily lead to tactical errors.
During that discussion Artabazus expressed the opinion that their position was very delicate and risky, and suggested to withdraw to Asia or at least to Thebes or Thessaly as soon as possible. However, Mardonius had invested too much of his personal reputation in the conquest of Greece and was not interested in a safe but inglorious return home. Hence he decided for an attack despite the difficulties. His plan for the battle is not precisely known and has to be guessed from Herodotus’ narration. 
After Mardonius’ question about the oracles and his words of encouragement night fell and guards were posted. Late at night, when both camps had apparently fallen quiet and almost everyone was asleep, Alexander the son of Amyntas, the commander and king of the Macedonians, rode up to the Athenian sentries and asked to meet with their commanders. (Herodotus, IX.44)
On hearing this report the Athenian commanders lost no time in following the sentries back to their posts, where they met up with Alexander. ‘Men of Athens,’ Alexander said, ‘please take what I have to say to you as a token of my good faith. You must keep it to yourselves and tell no one except Pausanias, because otherwise you might destroy me. … So I’m telling you that Mardonius and the Persian army have found it impossible to receive favourable omens from their sacrifices; if they had, battle would have been joined long ago. But now he has decided to ignore the omens and to attack at dawn— I imagine because he is afraid of your army increasing any further in size. Get ready to face an attack, then. In fact, even if he puts it off and does not join battle, you should just maintain your position and be patient, because he only has enough supplies left for a few days. (Herodotus, IX.45)
The nightly visit of Alexander to the Athenian camp has raised many discussions among the modern historians. If the exact content of his message is taken literally, his determination to the Greek cause becomes questionable. Indeed, he advised the Greeks to keep their positions waiting for a Persian attack. He was basing his advice on the lack of supplies in the Persian camp. As Herodotus mentions just a few paragraphs above, the Persians did not have such a logistical problem: “It was Artabazus’ opinion that they should strike camp as soon as possible and withdraw the entire army to the shelter of the walls of Thebes, where there was a good stock of supplies for them and plenty of fodder for the yoke-animals. (Herodotus, IX.41)”. If Alexander was communicating wrong information, was he actually a traitor and not a hero for the Greeks? Was he sent by Mardonius to confuse the Greek commanders by advising them to retain their defensive position across the south bank of Asopos?
The Athenian commanders went straight to the right wing and passed on to Pausanias the information Alexander had given them. His response to their report, because he was afraid of the Persians, was to say, ‘So battle is to be joined at dawn. You Athenians had better take on the Persians, while we meet the Boeotians and the other Greeks who are currently ranged against you. You’re familiar with Persian tactics, because you’ve already fought them at Marathon, while we are untried and ignorant of them. On the other hand, while we may have no experience of the Persians, we are experts in Boeotians and Thessalians. So go and collect your gear, and then move over to this wing, while we swap with you and take the left wing.’ (Herodotus, IX.46)
This information raises another question: If Alexander advised the Greeks to stay at their current positions protecting the road Thebes – Plataea, why did they move immediately after the message had reached them? A possible explanation is that Herodotus mixed the exact information he obtained from eyewitnesses with his own thoughts or perceptions. The manoeuvres of the Greek army indicate that they received a different message to the one reported by Herodotus. A reasonable explanation of the tactical moves of the Greeks is that Alexander informed them that the Persians were planning to attack along the Thebes – Erythrae road in the next morning. Therefore, if they remained in their original position across the Thebes – Plataea road, the Persians could have captured the exit of the Eleutherae pass and outflank the Greeks threatening seriously their supply routes. 
The reallocation of the Athenian regiment at the right of the Spartans probably means that the Athenians were sent to protect the road Thebes – Erythrae. Hence they moved from Pyrgos hill to the east slopes of Asopos Ridge. Early the next morning more moves of various units took place, which are difficult to interpret.
Since both parties liked the arrangement, they set about exchanging positions as day began to break. The Boeotians, however, realized what was going on and told Mardonius, who immediately tried to make changes within his forces too, by bringing the Persians across to face the Lacedaemonians. When this came to Pausanias’ attention, he realized that his manoeuvre had been detected, so he led the Spartiates back to the right wing, and Mardonius followed his example with regard to his left wing. (Herodotus, IX.47)
The analysis of these manoeuvres is intriguing. Most probably, Mardonius discovered that his plan did not take the Greeks by surprise, as he was expecting. Then realizing that the Greek defence was now weaker at the Plataea road, he ordered his cavalry to attack again along that road. The Athenians hastily returned back to the old position, but the whole Greek line was now disordered leaving several weak points. In this way, the cavalry found their way through to Gargaphia spring and destroyed it.
.. he (Mardonius) ordered his cavalry to charge the Greek lines. Every unit of the Greek army took casualties from the javelins and arrows of the Persian cavalry as they bore down on them, since they were faced with expert mounted archers to whom they could not get close. The Persian cavalry also churned up and blocked the Gargaphian Spring, which had been supplying the whole Greek army with fresh water. (Herodotus, IX.49)
The Greek position has become difficult. They had suffered many casualties the day before, and they had lost their water supply. Moreover, they had to defend a very long front of about eight kilometres long, which was exceeding the strength of their forces. A reasonable reaction would be to retreat towards Cithaeron.
Under these circumstances the Greek commanders met with Pausanias on the right wing to discuss various matters, including the loss of the army’s water supply and their harassment by the Persian cavalry. There were other items on the agenda because these events were not the only or even the main problems facing them: they had also run out of provisions, and the retainers of theirs who had been dispatched to the Peloponnese to bring them fresh supplies had been cut off by the cavalry and could no longer get through to the Greek camp. (Herodotus, IX.50)
The upshot of the commanders’ conference was that they decided to move their forces to the island, if the Persians refrained from joining battle that day. This island is located in front of the town of Plataea, ten stades away from the Asopos and the Gargaphian Spring, where they were based at the time. It is a kind of inland island: a river— the River Oëroë, which the locals hold to be the daughter of Asopos— divides further upstream on its way down from Cithaeron to the plain and the two branches of the river remain separate from each other for about three stades before merging again. So they decided to move here, because then they would not only have plenty of water, but also the cavalry could not inflict the casualties on them that they could when they were able to come straight at them. They decided to make the move during the night, at the time of the second watch, so that the Persians would not notice them setting out and also so as to avoid having the cavalry on their heels harassing them. They decided as well that once they had reached this new site … they would dispatch half their troops to Cithaeron under cover of darkness to meet up with the retainers of theirs who had gone to fetch supplies and were trapped on Cithaeron. (Herodotus, IX.51)

6.     The final confrontation

On the following day the plains and the low hills of the land between Asopos River and Cithaeron Mountain witnessed the greatest battle of the classical era. Approximately one hundred and fifty thousand men participated in a battle, which took place in a much extended battlefield (approximately 10 x 5 km). Mardonius had the tactical advantage of a larger army which consisted of different units with complementary capabilities. He had in his disposition heavy and light cavalry, a large body of Asian archers and a phalanx with heavily armoured Boeotian hoplites. This last addition was an innovation for the Persian army and gave important flexibility to his plans.
On the opposite side, the Greeks solely relied on their heavy infantry, which was 38,500 men strong according to Herodotus. Herodotus also refers to a numerous body of 63,000 lightly armed men on the Greek side. However the events of the battle indicate that the Greeks were lacking the support of a light infantry. For example, in the last day of the battle, the Spartans asked the Athenians to send them their archers (“If for some reason you can’t come yourselves, please send us your archers…” Herodotus, IX.60). This fact indicates that the Spartans had a limited light infantry support. Therefore, we may estimate that the total light infantry of the Greek army did not exceed 10,000 soldiers. The lack of versatility of the Greek army was limiting its tactical flexibility.
The battle was complex and difficult to analyze. Different units were moving independently to each other over a wide landscape and fighting was intense at several locations of the battlefield simultaneously. The commanders had great difficulties in coordinating the movements of their troops and had to use mounted messengers to communicate their orders.  The eyewitnesses who provided information about the battle to Herodotus could not have a complete view of the actions, so the synchronization of the events as presented by Herodotus is doubtful. The analysis of these complex battle tactics requires the careful reading of Herodotus’ story as a first step.
After they reached these decisions, the whole of the rest of the day was taken up with the constant burden of cavalry attacks, until the horsemen disengaged late in the afternoon. Night fell and the time agreed for departure arrived. The bulk of the army broke camp and left, but they had no intention of going to the appointed place: as soon as they started out, all they wanted to do was get away from the Persian cavalry, so they headed for the town of Plataea. On the way, however, they came to the temple of Hera, which stands in front of the town, twenty stades away from the Gargaphian Spring, and took up a position in front of it. (Herodotus, IX.52)
So they established themselves by the temple of Hera. Now, when Pausanias saw them leaving the camp he assumed that they were going to the appointed place, so he instructed the Lacedaemonians to collect their gear as well and follow the others’ lead. Most of his officers were prepared to obey Pausanias, but Amompharetus the son of Poliadas, who was the commander in charge of the company from Pitana, declared that as long as he had any say in the matter he would never bring shame to Sparta by retreating from the ‘strangers’. (Herodotus, IX.53)
So Pausanias and Euryanax[7] were trying to win over Amompharetus, since the Lacedaemonians and Tegeans had been left behind on their own. Meanwhile, the Athenians had also not moved from their post, … Once most of the army had decamped, they sent one of their men on horseback …  to ask Pausanias for instructions. (Herodotus, IX.54)
Pausanias then … turned to the Athenian, who had asked the question he had been sent with, and told him to let the Athenians know the difficulty of his situation; he requested that they link up with the Lacedaemonians and, as far as the withdrawal was concerned, that they follow the Lacedaemonian lead. (Herodotus, IX.55)
The messenger returned to the Athenian lines. As day began to break, the Lacedaemonians had still not resolved their differences. All this time Pausanias had stayed put, but now he judged— rightly, as it turned out— that Amompharetus would not let himself be left behind if the rest of the Lacedaemonians marched away, so he gave the order and began to lead all the Lacedaemonians except for Amompharetus and his men away through the hills, and the Tegeans fell in behind them. As instructed, the Athenians took the alternate route from that taken by the Lacedaemonians: whereas the Lacedaemonians kept to the hillocks and the spurs of Cithaeron because they were afraid of the Persian cavalry, the Athenians made their way down to the plain. (Herodotus, IX.56)
… when Pausanias’ troops were some way off, he (Amompharetus) saw that they really were abandoning him, so when his company had collected their gear he (Amompharetus) led them at a slow pace towards the other column, which had opened up a gap of about four stades and had halted on the River Moloeis, at a place called Argiopius (where there is also a sanctuary of Demeter of Eleusis), to wait for Amompharetus’ company. … Just as Amompharetus and his men met up with the rest of the Lacedaemonians, the Persian cavalry attacked in full force. (Herodotus, IX.57)
(Mardonius) he led the Persians at the double across the Asopus and after the Greeks, who he believed were trying to run away. In actual fact it was only the Lacedaemonians and Tegeans that he went for, because the knolls blocked his view of the Athenians, who had headed for the plain. As soon as the officers in charge of the remaining units of the invading army saw the Persians setting out in pursuit of the Greeks, they gave the signal for their men to join in the chase, and before long an undisciplined and chaotic mob of shouting soldiers was running as fast as they could after the Greeks, convinced that they would make short work of them. (Herodotus, IX.59)
As soon as the cavalry began to attack his men, Pausanias sent a messenger on horseback to the Athenians with the following message: ‘Men of Athens, the main battle is about to begin, and the outcome will decide whether Greece is to be free or enslaved…In fact, though, they have come in full force against us, so you should come and support us, since we are the ones who are particularly hard pressed. If for some reason you can’t come yourselves, please send us your archers…’ (Herodotus, IX.60)
On receiving this message, the Athenians wanted to go and provide all the help they could, but when they were on their way they were set upon by the pro-Persian Greeks who had been deployed against them. This attack put them under so much pressure that they found it impossible to go and help the Lacedaemonians. … they (Lacedaemonians) were about to join battle with Mardonius and as much of his army as was there, but the omens were unfavourable, and many of their men fell, with many more wounded, while the sacrifices were taking place, because the Persians formed their wickerwork shields into a barricade and continuously rained arrows down on the Greeks. In this situation, with the Spartiates under heavy pressure and the omens unfavourable, Pausanias looked towards the Plataeans’ temple of Hera, invoked the goddess, and asked her not to let their hopes prove to be false. (Herodotus, IX.61)
Pausanias was in the middle of his prayers when the Tegeans precipitately started forward to attack the Persians, and then, just as Pausanias finished praying, the Lacedaemonians received good omens. Now that the situation was at last favourable, the Lacedaemonians proceeded to attack the Persians as well, and the Persians laid aside their bows and prepared to meet them head on. The first phase of the battle took place at the wickerwork barricade, until that was knocked down, and then a fierce battle raged for a long time around the temple of Demeter. Eventually the two sides ended up grappling with each other, as the Persians caught hold of the Greeks’ spears and broke them off short. In courage and strength the Persians and the Greeks were evenly matched, but the Persians wore no armour; besides, they did not have the skill and expertise of their opponents. They would rush forward ahead of the main body of troops, one by one, or in groups of ten or so, and attack the Spartiates, only to be cut down. (Herodotus, IX.62)
Mardonius rode into battle on his white horse, surrounded by his elite battalion of a thousand first-rate soldiers, and wherever he put in a personal appearance the Persians made things particularly difficult for their opponents. As long as Mardonius was alive, the Persians held their ground and fought back, inflicting heavy casualties on the Lacedaemonians. But after he had been killed and the men of his battalion, the most effective troops on the Persian side, had been cut down, all the others turned and fled before the Lacedaemonians. Their destruction was due more than anything else to the fact that they wore no armour: it was a case of light-armed soldiers taking on hoplites. (Herodotus, IX.63)
When the Persians were routed by the Lacedaemonians at Plataea, they fled in disorder back to their encampment and to the wooden stronghold they had built on Theban land. (Herodotus, IX.65)
Although the rest of the Greeks on the Persian side deliberately fought below their best, the battle between the Boeotians and the Athenians lasted a long time, because those of the Thebans who had collaborated with the enemy were fully committed to the battle and refused to fight below their full capabilities. The outcome was that their three hundred best and bravest men were killed at Plataea by the Athenians. When the Boeotian forces too were beaten back, they retreated to Thebes, but by a different route to the one taken by the Persians and the rest of the allied troops on the Persian side, all of whom— the full complement— fled without having struck a single blow or displayed any courage at all. (Herodotus, IX.67)
In the end, then, the whole Persian army was beaten back, except for the cavalry (and especially the Boeotian cavalry), which proved invaluable to those who were fleeing, because they stayed on the side nearest the enemy and so shielded their allies, as they fled, from the victorious Greeks, who gave chase, harrying and slaughtering men from Xerxes’ army. (Herodotus, IX.68)
During this rout, word reached the rest of the Greeks (the ones who had taken up a position by the temple of Hera and had not taken part in the fighting) that a battle had taken place which Pausanias and his men had won. As soon as they heard the news, they set out in complete disarray. The Corinthians and the units deployed near them took the high road and headed straight for the sanctuary of Demeter following the foothills and knolls, while the Megarians, Phleiasians, and so on made their way through the level ground of the plain. As the Megarians and Phleiasians drew near the enemy, they were spotted by the Theban cavalry under Asopodorus the son of Timander. The Thebans could see that there was no discipline to their hasty approach, so they charged into the attack, cut down six hundred of them, and chased the scattered survivors back to Cithaeron. (Herodotus, IX.69)
The battle began during the night when the Greek regiments decided to retreat towards Cithaeron. Here is a summary of the events that followed:
1)    All the Greek units except the Spartans and the Athenians were ordered to move towards Plataea and hold position at a place called Oeroe island which lies between two branches of the river Oeroe. That position was along the Plataea – Thebes road as Pausanias the geographer reports. However, this army moved further to the south and took position in front of the temple of Hera in order to “get away from the Persian cavalry”.
2)    The Spartans also began to move but with a delay. One of their units (a company) stubbornly delayed its retreat.
3)    The Athenians communicated with the Spartans by a mounted messenger. They received the instruction to “link up with the Lacedaemonians and, as far as the withdrawal was concerned, that they follow the Lacedaemonian lead”.
4)    At dawn the Lacedaemonians finally moved. However, the company that was refusing to retreat held its position.
5)    After the main body of the Spartan army moved about 800 m, the “rebellious” company also started moving but at a slow pace. They finally caught up with the rest of the Spartans.
6)  When Mardonius found out that the Greeks had retreated, he ordered an immediate attack against the Lacedaemonians. From his position he could see only the Spartans and the Tegeans moving south along the higher ground of Asopos Ridge.
7)  The Athenians were moving on a lower ground and could not be seen by the Persians. Considering that they were positioned at the left wing of the Greek battle line, they were probably marching along the west slopes of Asopos Ridge, while the Persians were attacking the Spartans at its east slopes. This scenario raises another question: If the Athenians had been marching along a path which was visible from the Thebes-Plataea road, they should have been spotted by the pro-Persian cavalry which was attacking the rest of the Greek forces near Plataea. Herodotus, however, does not mention any engagement between the Athenian and the pro-Persian cavalry. Therefore, the Athenians probably took a road which was crossing the hills of Asopos Ridge.
8)    The Persian cavalry was the first to attack the Lacedaemonians. As a result the Spartans could not retreat anymore and took defensive position near a place protected by the gully of a tributary of Asopos called Moloeis.
9)    In the meantime the Mardonius’ light infantry which consisted of the Persian and other Asian units advanced and after making contact with the Lacedaemonians started pouring arrows against them.
10) The Spartans kept their line but did not counterattack despite their casualties due to the arrows.
11) Pausanias sent a mounted messenger to the Athenians asking for help.
12) The Athenians rushed to help the Spartans but at that moment the Boeotian heavy infantry attacked them. The two enemy Greek armies got engaged in a fierce battle, where the Athenians won.
13) At the time Pausanias asked for help from the Athenian regiment, he also ordered his hoplites to attack the Asian archers. After the Spartan phalanx came in close contact with the light Asian infantry heavy fighting took place.
14) The Asian infantry did not have the weaponry to cope against the Spartan phalanx. Mardonius with his elite troops fought bravely refusing to retreat. When he was killed, the Persian battle line collapsed.
15) The Persians fled to their fortified camp pursued by the Spartans.
16) The army of the other Greek cities joined the battle but suffered many casualties when the Boeotian and probably the Thessalian cavalries attacked them.
17) The Boeotians found refuge back to Thebes retreating along a different road to the one taken by the Persians. They had lost 300 hoplites in the battle with the Athenians.
Although the description of the battle by Herodotus is exciting and detailed, the information includes discrepancies and several of the actions cannot be easily justified. Many questions are raised:
1)    Why did the Spartans wait in their original position until dawn?
2)    Why did the Spartans sustain the shower of the Persian arrows without reacting? What were they waiting for?
3)    What was the path followed by the Athenians?  Why they could not be seen by Mardonius? Was this accidental or intentional?  What was their actual role in the battle?
4)    Why was Mardonius taken by surprise and failed to retreat to a safer place when the Spartans charged? His cavalry could have intervened to allow the safe retreat of the Persian infantry when their assault failed. Why did they fail to do so?
5)    What was the role of the army of the small Greek cities? Why were they sent to a position far away from the Lacedaemonians? Keeping all the Greek army united would have been safer. If they had actually taken position near the walls of Plataea, how is it possible to chase the Persians who were east of Asopos Ridge?
6)    What was the role of the medizing Greek cavalry? Why they were not able to see the Athenians who were marching at the west slopes of Asopos Ridge, while they were attacking the Greeks who had taken position near Plataea. If the Athenians were marching at the plain along the Thebes - Plataea road, the cavalry should have spotted them and attacked them.
Despite its discrepancies, a second careful reading of Herodotus’ narration may give clues of what actually happened during that dramatic morning. Here is a plausible interpretation of those events:
When the Greek commanders met in the night before the final battle, they realized their difficult position.  If the Persians managed to cut their supply lines, they had to step down to the plain and fight in an unfavourable terrain. At this difficult moment the Greek commanders proved their superior tactical skills. They worked out a battle plan which was by far superior to what an average general would expect. Who was the mind behind this ingenious stratagem is not known with certainty. Pausanias, who was the commander-in-chief, certainly had a strong contribution to the development of the plan. However, Aristides with his great experience of previous battles with the Persians (Marathon and Salamis) most probably helped as well. After all, this battle plan had a strong flavour of the Athenian way of thinking. In a difficult situation the Athenians were recognizing only opportunities not threats.
The Greeks based their plan on the current psychology of Mardonius. After his successes of the previous days, his increased self confidence made him susceptible to tactical errors. Therefore, it would be easier to lure him into a trap. However, they knew that Mardonius and his commanding officers were experienced and it would not be easy to deceive them. The Greek generals had to provide a bait, large enough to make the trap irresistible. Such a piece of bait would be a large division of the Greek army. In particular the mighty Lacedaemonian regiment would be a really tempting and hard to refuse offer. The Greek generals knew that the Lacedaemonians despite their high military skill, if separated from the rest of the Greek army and had to face the bulk of the Persian army their chances to get through would be very limited, so they were a convincing bait. Therefore, it was decided that the Spartans would take the role of the bait. After all the Spartans were the only Greek unit which had the training and the discipline to play this difficult role and escape.
After selecting the bait, the Greek generals had to set the trap. A second Greek division would hide and surprise the enemy who would run for the bait. This second Greek army had to be very disciplined and determined. This role was given to the Athenians assisted by the Plataeans, who had excellent knowledge of the landscape. The concept of a vanishing army of 8,500 hoplites, who managed to become invisible in a treeless landscape by exploiting just the screening effect of low hills of only 30-50 m height, is really incredible. Only the most imaginative strategist commanding a well disciplined army could conceive and execute such a trick.
The plan required a third Greek division for covering up the trap. That regiment had to be numerous but it would have limited participation in the fighting. The role of this third Greek army would be to confuse Mardonius so that the hiding Athenians would skip the attention of the Persians. Indeed, when in that early morning Mardonius saw a numerous Greek army taking battle position in front of Plataea protecting the Dryos Kefalae pass, and the Lacedaemonians rushing towards Erythrae, he could not suspect that a third Greek army was hiding among the hills of Asopos Ridge.
With the first light of that fatal day, Mardonius saw the battle line of his enemies in a very favourable arrangement for his plans. He immediately worked out a plan to exploit the division of the Greek forces. He first decided to send the Boeotian cavalry probably assisted by a few light Asian units against the bulk of the Greek forces which were protecting the road to Plataea. He was expecting that the cavalry attacks would immobilize these hoplites, so they could not offer any help to the Lacedaemonians. Then he directed all of his best units against the Spartans. These included the Persian cavalry, the Persian light infantry and the Boeotian heavy infantry. His plan was to attack the Spartans through the Erythrae road first by the cavalry, and then by all his infantry units. He was planning to outflank the Lacedaemonian phalanx by using his numeric superiority and surround the Spartans. For this reason he ordered an attack along two lines. The Persian cavalry and infantry attacked along the Erythrae road at the right wing of the Spartans. Then the medizing Greeks, mostly Theban hoplites marched through the hills of Asopos Ridge to attack the left wing of the Lacedaemonians. This last division of the Persian army was roughly equal in number to the Lacedaemonian army and similarly equipped. Therefore, the Spartans would have to fight against an almost equivalent phalanx supported by a much more numerous Asian light infantry of approximately 40,000 men and the Persian cavalry. The defence of the Spartans would be even more difficult considering that the attack would take place from two different directions. Now Mardonius was certain for his victory.
The cavalry attacking the army of the small Greek cities, which was retreating towards Plataea, had an easy job. The units of that Greek phalanx which were exposed in a low ground suffered heavy blows and had to retreat uphill to the temple of Hera near Plataea (see above Herodotus, IX.52). It was probably in this phase of the battle that the Megarians and the Phleiasians suffered heavy casualties since they were holding the right wing of that Greek phalanx and were positioned at a lower ground where they were more exposed to cavalry attacks. Herodotus wrongly reports this incident at the last stage of the battle, probably because his Athenian sources (eyewitnesses) wished to underestimate the contribution of the Megarians and the Corinthians in the victory. It is more reasonable to assume that at the end of the battle the Boeotian cavalry was more interested in securing the retreat of the defeated Theban infantry back to their city rather than attacking Plataea.
In the mean time, the Persians crossed Asopos and ran after the Spartans. First the cavalry attacked and forced the Spartans to stop their retreat towards Hysiae and resume battle position. Then the Persian infantry approached marching along the Erythrae road and started shooting arrows against the Lacedaemonians. The phalanx of the Thebans was advancing probably through the hills of Asopos Ridge aiming to attack the left wing of the Spartans. Both the Persians and the Thebans did not suspect the presence of the hiding Athenians and Plataeans. Moreover, the Boeotian cavalry, which was attacking the Greeks along the Plataea road, also failed to see the Athenians. The Athenians led by Plataean guides had crossed the plain from Pyrgos Hill to Asopos Ridge in the darkness of the night before, and took position at a lower ground among the hills. Thus they made themselves invisible from both the Plataea and Erythrae roads exploiting what is known as the “illusion of the low hills”. Low hills separated by gullies can hide large numbers of soldiers, while if there are mountains in the background, the viewer gets the impression of seeing an empty landscape. Other famous generals, such as Hannibal[8] or Frederick the Great[9], exploited this deception for setting traps.

Figure 1: Elevation profile along an E-W cross section of Asopus Ridge at about halfway between Cithaeron’s feet and Asopos River. Points A and D correspond to the Thebes-Plataea and Thebes-Erythrae roads respectively. Positions B and C are invisible from both A and D view points. In particular, position B is the most probable hiding place of the Athenians.

As in Salamis, a fatal trap was set for the Persians proving the tactical superiority of the Greek high command. As the Thebans were approaching the Spartans, Pausanias, who was able to observe the whole battlefield, signalled the Athenians by a mounted messenger that this was the time to get out of their hideout. At the same time he ordered the Spartan phalanx forward. Mardonius was waiting in vain for the Thebans to appear. Using his elite squadron he was intervening at all the critical points of the Persian battle line attempting to keep his position until the Thebans arrived. But the Thebans had been engaged in battle with the Athenians and never came to his support. Mardonius was killed fighting bravely. Then the Persian battle order collapsed.
Artabazus was commanding various Asian regiments and he was assigned to protect the central area of the battlefield which lies between the two roads to Plataea and Erythrae. As soon as he saw the defeat of the Persians, he gathered the survivors and using the protection of his cavalry, withdrew towards Thessaly taking with him the bulk of the Persian forces that escaped the battle. Artabazus made no effort to save the supplies gathered in the fortified Persian camp since he knew that he did not have enough forces to defend it. After all, his main intention was to move as fast as possible to Minor Asia in order to save the Persian presence there.
Now, Artabazus the son of Pharnaces had disapproved of the campaign right from the very start, when Mardonius had been left in Greece by Xerxes, and had often tried, without success, to dissuade Mardonius from joining battle. Since he was unhappy with Mardonius’ tactics anyway, then, this is what he did. He was responsible for a sizeable force of about forty thousand men, and because he had no doubts about the final outcome of the battle that was under way, he had them adopt a tight formation and told them to follow his lead wherever he went and at whatever pace he set. Having issued these instructions he marched them out as if they were going to join the battle. When they were some way down the road, however, he saw that the Persians were already in flight. At that point he changed formation and began to run as fast as possible away from the battlefield, but not towards the stronghold or Thebes with its defensive walls. Instead, he made for Phocis, because he wanted to get to the Hellespont without delay. (Herodotus, IX.66)
Soon after the battle the Greeks attacked the fortified Persian camp which was probably defended by a small garrison that did not manage to escape with Artabazus. After a short siege, they managed to break in. The spoils were plenty and rich beyond any imagination. 
The main body of the invading army, including the Persians, had taken refuge inside their wooden stronghold. They managed to climb up the towers before the Lacedaemonians arrived, and then they reinforced its walls as best they could. The Lacedaemonian attack initiated a fairly tough battle for the wall, because until the Athenians arrived the defenders were getting the better of the Lacedaemonians, who did not know how to go about attacking fortified structures. But once the Athenians attacked the stronghold, a fierce and protracted battle took place. Eventually, thanks to their courage and persistence, the Athenians succeeded in scaling the wall and making enough of a breach in it for the Greeks to pour in. The first to enter the stronghold were the Tegeans, and it was they who plundered Mardonius’ pavilion, from which they took various objects including the manger Mardonius had used for his horses. (Herodotus, IX.70)
Herodotus reports only a small number of casualties among the Spartans and the Athenians. It is not clear whether these numbers refer to the battle of the last day or for the full duration of the hostilities. Anyway, if these numbers are correct, they prove that the trap worked very efficiently. The Persian casualties are usually exaggerated. As it was explained above, most Persian units are expected to escape under the protection of their cavalry.
… a total of 91 Lacedaemonians from Sparta lost their lives in the battle, along with 16 Tegeans and 52 Athenians. (Herodotus, IX.70)
The Persians were the best of the infantry from the invading army, while the best cavalry unit was that of the Sacae, and the individual prize for valour was held to belong to Mardonius. On the Greek side, although both the Tegeans and the Athenians proved their worth, the Lacedaemonians outshone everyone else. (Herodotus, IX.71)
Herodotus praises the Spartans as deserving the prize of valour more than any other Greek regiment. However, he is not explicit whether the prize was actually given to them. On the contrary, Plutarch is his “Life of Aristides” reports a different story:
After this, the Athenians not yielding the honour of the day to the Lacedaemonians, nor consenting they should erect a trophy, things were not far from being ruined by dissension among the armed Greeks; had not Aristides, by much soothing and counselling the commanders, especially Leocrates and Myronides, pacified and persuaded them to leave the thing to the decision of the Greeks. And on their proceeding to discuss the matter, Theogiton, the Megarian, declared the honour of the victory was to be given some other city, if they would prevent a civil war; after him Cleocritus of Corinth rising up, made people think he would ask the palm for the Corinthians (for next to Sparta and Athens, Corinth was in greatest estimation); but he delivered his opinion, to the general admiration, in favour of the Plataeans; (Plutarch –Aristides 20)
Despite the anecdotic character of the above incident, which is typical in Plutarch’s narrations, this event, if actual, strengthens the theory of the trap as presented above. Although some details may be fabricated to make Plutarch’s story pleasant and exciting to the reader, the final decision of the Greeks to award the prize of valour to the Plataeans demonstrates their critical role in setting up the trap by guiding the Athenians to the exact place where they were invisible to all enemies.
After the battle of Plataea, the Persians withdrew from Southern Greece. They tried to save their strongholds in Macedonia and Thrace, but this was possible only for a short time.
Cartledge P. (2013), “After Thermopylae: The Oath of Plataea and the End of the Graeco-Persian Wars”, Oxford University Press, NY 10016, USA.
Green P. (1996), “The Greco-Persian wars”, University of California Press, Ltd.
Macaulay G. C. (1890), “The History of Herodotus”, Macmillan, London and NY.
Mystriotis A. (2013), “507-450 BC - The 57 years which gave birth to Democracy”, Amazon – Kindle edition 2013.
Rawlinson G. (1996), “The History of Herodotus”, Wordsworth Editions Limited. Provided by The Internet Classics Archive at
Shepherd W. and Dennis P. (2012), “Plataea 479 BC: Greece's greatest victory”, Osprey Publishing, Oxford OX2 0PH, UK.
Waterfield R., and Dewald C. (1998), “Herodotus - The Histories”, (Oxford World's Classics), Oxford University Press.

[1] This article is also included as an Appendix in the book Democracy in Ancient Athens: The Rise, 507- 450 B.C.  (Amazon Kindle edition 2013).
[2] About three times the number of soldiers who participated in the battle of Marathon.
[3] For more details about Themistocles’ strategic plan see Mystriotis (2013).
[4] The perioeci were also Lacedaemonians but of an inferior military training than the elite Spartan hoplites. Their name indicates that they were living outside the city of Sparta.
[5] A stade approximately equals 180 m.
[6] Tegea was a city in Arcadia, north of Sparta. It traditionally was a close ally of Sparta. In the battle of Plataea, the Tegeans fought together with the Spartans during the decisive last day of the battle.
[7] Euryanax was Pausanias’ cousin and his co-commander of the Spartan army in the battle of Plataea
[8] In the battle of Trevia River, Hannibal hid one thousand horsemen and one thousand hoplites in a gully of a tributary of Trevia. As the Roman army advanced the hiding troops found themselves at the back of the enemy and attacked from behind.
[9] Battle of Rossbach