The casualties of Plataea

August 1st, 2022, marks the 2500th anniversary of the battle of Plataea in 479 BC. Plutarch (Aristides 19.7) provides us this precise date:

“This battle was fought on the fourth of the month Boedromion, as the Athenians reckon time; but according to the Boeotian calendar, on the twenty-seventh of the month Panemus.”

Even Plutarch notes that there were discrepancies in date in the ancient world due to the fact that “different peoples have different beginnings and endings for their months.” Nonetheless, now seems the perfect time to pen a brief blog about this most important of ancient battles.

We are given a remarkable breakdown of the forces at Plataea, especially in Herodotus (9.28.2-29.2). In a way, this is hardly surprising since Plataea is the culmination of Herodotus’ entire Histories. To some, indeed most, however, the numbers are unbelievable. According to Herodotus: “On the right wing were ten thousand Lacedaemonians; five thousand of these, who were Spartans, had a guard of thirty-five thousand light-armed helots, seven appointed for each man. (3) The Spartans chose the Tegeans for their neighbours in the battle, both to do them honour, and for their valour; there were of these fifteen hundred men-at-arms (hoplitai). Next to these in the line were five thousand Corinthians, at whose desire Pausanias permitted the three hundred Potidaeans from Pallene then present to stand by them. (4) Next to these were six hundred Arcadians from Orchomenus, and after them three thousand men of Sicyon. By these one thousand Troezenians were posted, and after them two hundred men of Lepreum, then four hundred from Mycenae and Tiryns, and next to them one thousand from Phlius. By these stood three hundred men of Hermione. (5) Next to the men of Hermione were six hundred Eretrians and Styreans; next to them, four hundred Chalcidians; next again, five hundred Ampraciots. After these stood eight hundred Leucadians and Anactorians, and next to them two hundred from Pale in Cephallenia; 6 after them in the array, five hundred Aeginetans; by them stood three thousand men of Megara, and next to these six hundred Plataeans. At the end, and first in the line, were the Athenians who held the left wing. They were eight thousand in number, and their general was Aristides son of Lysimachus. (29.1) All these, except the seven appointed to attend each Spartan, were men-at-arms (hoplitai), and the whole sum of them was thirty-eight thousand and seven hundred. This was the number of men-at-arms that mustered for war against the barbarian; as regards the number of the light-armed men, there were in the Spartan array seven for each man-at-arms, that is, thirty-five thousand, and every one of these was equipped for war. (2) The light-armed from the rest of Lacedaemon and Hellas were as one to every man-at-arms, and their number was thirty-four thousand and five hundred.”

Herodotus’ total number for the Greeks, therefore, is 38,700 hoplites from various city states, 35,000 Spartan helots operating as psiloi and 39,500 other light-armed psiloi from the other states. This gives us a total of 113,200. With the breakdown provided, as with the others in Herodotus’ work, this seems to be entirely reasonable. Plutarch (Aristides 11.1) also tells us there were 8,000 Athenians.

For the Persians, however, we are given the much larger number of 300,000 in Herodotus (9.32), where he estimates another 50,000 allied Greek troops. Likewise, Plutarch (Aristides 19.4), tells us there were 300,000 Persians. This is a much more unacceptable number and is usually reduced down drastically in modern works on the battle to ‘a more acceptable’ 80-120,000. Peculiarly, the Greek numbers are also reduced but you will often find them (or at least part of them) used with precision to give Spartan or Athenian numbers, not to mention the ratio of Helot to Spartan (7:1)). For instance, no one questions the 5,000 Spartan hoplites or the 8,000 Athenians (this was 1,000 men less than at Marathon in 490 BC - Cornelius Nepos Miltiades 5.1). Nonetheless, despite his precision, Herodotus’ numbers are considered untrustworthy (except when they are considered trustworthy).

Diodorus Siculus, however, sums up the numbers more succinctly and states (11.30.1) simply that “The total number of the Greeks approached one hundred thousand men, that of the barbarians some five hundred thousand.” Justin, in his Epitome of Pompeius Trogus’ Histories (2.14) also records that there were 100,000 Greeks. So, in terms of the Greeks at least, the numbers seem widely in agreement – between 100,000 and a little more than 110,000. For the Persians we find 300-500,000. On the other side of the coin, however, there is much less mistrust when Herodotus’ numbers in relation to Plataea are low. This is despite some scepticism even in the ancient world. In this regard I am not talking about the Persian casualties – Herodotus tells us that the Persians suffered 257,000 casualties and that only 40,000 men escaped with Artabazus (9.70.5): “Such a slaughter were the Greeks able to make, that of two hundred and sixty thousand who remained after Artabazus had fled with his forty thousand, scarcely three thousand were left alive.” That would represent 85% casualties of the army size he gives. We can note that, at this point, his 50,000 Greek allies are conveniently forgotten.

Here, however, there seems to be as much disagreement as there was with the numbers who fought. Plutarch (Aristides 19.4) agrees with Herodotus: “Out of three hundred thousand, only forty thousand, it is said, made their escape with Artabazus.” Plutarch doesn’t include the 3,000 left alive mentioned in Herodotus (and so his battle is even bloodier). Diodorus, however, even though he had 500,000 Persians at the battle, records that only 100,000 of them died (11.32.5): “in the end, when the Greeks had slaughtered more than one hundred thousand of the barbarians, they reluctantly ceased slaying the enemy”.

Following Ephorus, Diodorus tells us (11.33.1) that Artabazus fled with as many as four hundred thousand Persians (not the forty thousand of Plutarch and Herodotus). Despite our unwillingness to accept Diodorus’ larger number of Persians, his number of casualties seems more reasonable, at least as a proportion. His number would still represent 20% casualties, however.

For the Greek casualties at Plataea, we are dealing with much greater discrepancies in the ancient sources as well as a much greater degree of unreasonableness, if only because in most of our sources they are so low. Herodotus tells us that there were only 160 casualties (9.70.5): “Of the Lacedaemonians from Sparta ninety-one all together were killed in battle; of the Tegeans, seventeen and of the Athenians, fifty-two.” Using Herodotus’ numbers (160÷113,200), this represents losses of only 0.14%. Athens, with 8,000 men suffered 0.65% casualties (52÷8,000), Tegea, which only brought 1,500 hoplites, suffered 1.13% (17÷1,500), and Sparta suffered the most with 1.82% (91÷5,000).

There are many ancient battles with disproportionate casualty numbers – at the battle of Marathon 192 Athenians, and 11 Plataeans fell as opposed to 6,400 Persians (Herodotus 6.117). At the battle of Amphipolis in 422 BC, the Spartans only suffered 7 casualties whilst inflicting 600 (50%) on the Athenian force of 1,200 (Thucydides 5.11.2). Later, too, such as at the battle of Argentoratum in AD 357, the forces of Julian also only suffered 243 casualties whilst inflicting 8,000 (Ammianus 16.12.63). At the battle of Cunaxa in 401 BC, despite being on the losing side, the Greek hoplites did not suffer a single casualty according to Xenophon (who was a participant) (Anabasis 1.8.20 - Diodorus corroborates this (14.24.5-6); Artaxerxes lost 15,000 men, and Cyrus’ forces lost only 3,000 men. Of the mercenaries, not a single man fell although several were wounded). The small numbers of casualties don’t seem to attract anywhere near the criticism (ancient and modern) that vast overestimates of army sizes do. These low numbers for Plataea are fascinating in themselves since even where we had agreement on the large numbers of troops involved, we now find dissent in the low number of casualties.

Plutarch adds more detail to Herodotus’ low numbers (Aristides 19.4-5) “Of those who contended on behalf of Hellas, there fell in all one thousand three hundred and sixty. Of these, fifty-two were Athenians, all of the Aeantid tribe, according to Cleidemus, which made the bravest contest (for which reason the Aeantids used to sacrifice regularly to the Sphragitic nymphs the sacrifice ordained by the Pythian oracle for the victory, receiving the expenses therefor from the public funds); ninety-one were Lacedaemonians, and sixteen were men of Tegea.” Plutarch agrees with Herodotus’ ninety-one from Sparta and fifty-two from Athens (it is remarkable that all fell from a single tribe, Aiantis – more of this in a moment). For Tegea there is the discrepancy of a single casualty (you will find many modern works stating Herodotus has 159 casualties – bringing the two into agreement). Plutarch, however, has an additional 1,201 casualties amongst the Greeks, so representing the remaining city states. So, using Herodotus’ overall numbers of 38,700 hoplites, this would seem to represent 3.51% casualties (I have only used the hoplite figures since they seem to be the only casualties recorded). Plutarch also admonishes Herodotus for his low number (and the fact he claims only the men of Sparta, Tegea and Athens fought the Persians) – (Aristides 19.7): “Astonishing, therefore, is the statement of Herodotus, where he says that these one hundred and fifty nine represented the only Hellenes who engaged the enemy, and that not one of the rest did so. Surely the total number of those who fell, as well as the monuments erected over them, testifies that the success was a common one.”

Diodorus (again probably using Ephorus) offers a much different picture. He states (11.30.1) that there were more than 10,000 Greek dead. In terms of casualties, he therefore seems to paint the most reasonable numbers – 10% Greek losses and 20% Persian losses in the Greek victory.

One detail in Plutarch’s account, however, makes the case that his and Herodotus' numbers may be reliable. Plutarch tells us that all the fifty-two Athenian losses were all suffered by a single tribe, Aiantis. We do not know where Aiantis were stationed that day in the Athenian line but the only action of note undertaken by the Athenians was that they fought (and killed to a man) the three hundred Thebans who were fighting on the Persian side (Herodotus 9.67, Plutarch Aristides 19.2). Leaving aside who these 300 may have been (I have written on their possible identity in AW 14.2 - suffice to say this unit may have been a precursor to the later Theban Sacred Band which also numbered 300 members), we can explore how they were overcome by the Athenians. To do this we first need to consider the make up of Athenian hoplite armies and the depth of their line.

Both Herodotus (9.28.6) and Plutarch (Aristides 11.1) tell us that there were 8,000 men brought to the battle of Plataea by Athens in 479 BC. 8,000 men would roughly divide into ten, 800-man taxeis with each one (a taxis) representing one of the ten tribes (phylai) of Athens. Each taxis was named after the tribe (so Erechtheis, Aigeis, Pandionis, Leontis, Acamantis, Oineis, Cecropis, Hippothontis, Aiantis, and Antiochis) and commanded by a taxiarch elected from that tribe. Below the taxis, there were a certain number of lochoi, each commanded by a lochagos. Unfortunately, we have very little evidence on the size of the lochos at Athens – for reasons way too complex to include here, I think it was usually 30 men. This was how every Athenian army was divided – the evidence is scattered but the best evidence is Herodotus 6.111.1, Thucydides 6.98.4, 6.101.6, Xenophon Hellenica 4.2.19, Lysias For Mantitheus/Oration 16 15, Aristotle Athenian Constitution 61.3, Demosthenes Funeral Oration 27-31). Wherever Aiantis was stationed on the day of Plataea, it bore the brunt of the fighting; and the only fighting of note was against the Theban 300. This brings us to consider the depth of the line in Hoplite armies.

We hear of other a variety of phalanx depths (8: (Thucydides 4.94.1, 5.68.3, 6.67.1,Xenophon Hellenica 2.4.34), 10: (Hellenica 2.4.11-12), 12: (Hellenica 6.4.12), 16: at Syracuse (Thucydides 6.67.2), and later at Nemea (Hellenica 4.2.18), even 25: for the Thebans at Delium (Thucydides 4.93.4) all the way to 50: (Critias in the Piraeus in 404, Hellenica 2.4.11-12, and, most famously, the Thebans at Leuctra (Hellenica 6.4.12). For all the battles of the classical phalanx, however, this is actually very little evidence, and the truth is we are most often not told the depth of hoplite phalanx. The best evidence is Thucydides’ account of the battle of Mantinea in 418 BC where we are told that the individual Spartan commanders designated a different depth per lochos (the Spartan lochos was a larger division of the army than the Athenian lochos) but which averaged out to 8 men. At Plataea, a depth of eight men would mean each Athenian taxis had a frontage of 100 men. Although we do not know how the line was arrayed, it makes most sense then, that Aiantis bore the brunt of that fighting, probably in the centre of the Athenian line. Opposite them, the 300 Thebans, if drawn up similarly eight deep would only have had a frontage of roughly 37 men. If four, only 74 men – few enough to only face a single Athenian taxis. We do not know the depth of the Theban unit, but it would, in all probability, have had less of a frontage than a single Athenian taxis. We can imagine that the remaining Athenian taxeis swung around the flanks of the Thebans and accelerated their destruction (without the loss of a man). These troops would have suffered much less risk than the taxis engaged in a front-on fight with the Thebans.

Thus, Aiantis suffering all of Athens’ casualties on the day of Plataea, is certainly possible. This observation, then, causes us to reconsider the numbers. If fifty-two casualties from 8,000 Athenians is reasonable, 8,000 Athenians then also becomes reasonable. If so, then the breakdown of troops for the Greeks – between 100,000 and 113,000 – also becomes more reasonable. If this is the case, and we do not need to emend that number down, as is done so often, what does it mean for the larger numbers of the Persians? I do not suggest that we should accept uncritically the vast numbers (even the small ones) our sources give us but, perhaps, the readiness some have shown to reject them and revise them down is just a little too quick off the mark.

1 comment

I have always wondered about that business where there were supposedly 7 helots for every Spartan. Aside from the fact that this would make for an awful lot of light troops for a Classical Greek hoplite army (not even considering the other Greek light troops), it seems unlikely that the Spartiates would ever have risked arming that many helots at one moment. They were famously concerned about revolt, but this would seem to be handing the helots a chance to simply mob the majority of the Spartiates out of existence.

Jamie Gagnon

Leave a comment