Mardonios had urged the King not to abandon the enterprise even after the debacle at Salamis. According to Mardonios there was a way to invade the Peloponnese even without a ferry and he argued with the King that he could proceed to that invasion the following year if he had 300,000 men, that is, half of the army that had come to Greece in 380 B.C.
Mardonios marched with the King's army up to Thessaly and there he went into winter quarters. The following spring he was joined by 40,000 men under Artabazos who had followed the King in his withdrawal with an original force of 60,000. According to Herodotos the forces of Mardonios were 300,000 infantry plus cavalry; of the infantry, 50,000 had been provided by the Greek allies. This means that Mardonios had under his command the normal full strength of the Persian army, even though the cavalry did not by far come up to the table strength of 50,000 horsemen. But Herodotos states that, at the battle of Plataia that closed the campaign of Mardonios, the cavalry was the part of the Persian army that proved the greatest challenge to the Greeks.
Mardonios had a table of organization requiring 300,000 infantry men and tried to fill it up by all means. Herodotos reports that he put in line Egyptians who had never been in an army and originally had served as embarked marines in the fleet. Mardonios hoped to succeed by combining political manouvering with military action, since mere military impact had not succeeded. His plan was to force all the Greeks north of the Isthmus, mainly the Athenians, to desert the rest of the Greeks, with the result that even the Greeks of the Peloponnese who were defending the Isthmus would have collapsed. In the spring he made overtures to the Athenians, who wavered, but finally rejected Mardonios' enticing proposals. He tried to force them by invading Attika when the crops were about to be gathered, but the Athenians once again abandoned their city and withdrew to Salamis. By the end of the summer the Greeks had succeeded in producing an unexpected show of unity: they were able to gather an army of some 110,000 men. This army was so large that it was the Greeks' turn to have some difficulties with supply trains and with provisions of water.
The army of Mardonios, however, was still so strong that the Greeks kept avoiding battle until almost twelve months after Salamis, near the close of the military season, when the Persian army began to give signs of disintegration. The disintegration must have been unavoidable once it became clear that another year had passed without conclusive military or political results. Just before the Greek attack Artabazos, with 40,000 men under his command, deserted Mardonios and began to withdraw from Greece. When the Greeks finally attacked at Plataia the battle turned into butchery; Mardonios himself was not able to escape.
Herodotos' narrative of what happened after the battle of Salamis is clear and reasonable, but it is most obscure for the historians of the critical school, because they cannot account for the number of men engaged in the battle of Plataia, which was the only major land engagement in almost two years of campaigning and which sealed the fate of the war. According to Herodotos the Greek army consisted of 38,700 hoplites and of light armed men at the rate of one to each hoplite, plus 35,000 Spartan Helots. The hoplite force consisted of 10,000 Spartans, 8,000 Athenians, 5,000 Corinthians, and other lesser contingents from smaller cities which Herodotos enumerates in detail; the figures are perfectly consonant with what we know about the military power of each city.
Macan did not question the number of the hoplites, but reduced the rest of the army to one man to each hoplite. Eduard Meyer submitted a similar estimate, about 30,000 hoplites and as many lightarmed soldiers; the army of Mardonios would have consisted of 40,000 or 50,000 Asiatics and some thousands of Greeks. The lowest estimate, that of Obst, is 35,000 Greek soldiers; to the Persians Obst ascribed a force of 50,000 men. Beloch and De Sanctis settled for the figure of 50,000 Greeks, estimating the forces under the command of Mardonios as 50,000 Persians and 20,000 Greeks. Among the more recent writers, Berve estimates the Persians at 50,000 and the Greeks at 40,000, whereas Giannelli estimates the Persians at 100,000 and the Greeks at 70,000 of which half may have been hoplites.
There is no reason for doubting Herodotos' figure for the Greek forces, except for the fact that this affects the estimate of the Persian forces. As Glotz grants, one must assume that there were about three Persians to a Greek, since Persian soldiers were no match for Greek ones. Herodotos points out that even the very best of the Persian forces, the Persians themselves, although not inferior in courage and determination, were pitifully inferior to the Greeks in drill and in armor.
Herodotos' figures about the forces engaged at Plataia do not contain any element that can be questioned on specific grounds. The list of the cities that contributed to the Greek forces, as given by Herodotos, is confirmed by the inscription on the serpent-column that was dedicated at Delphoi shortly after the battle and which is still extant today. A similar list is quoted by Pausanias (V 23) as having been seen by him inscribed on the pedestal of a statue of Zeus erected at Olympia. As I have said, the size of each Greek contingent as given by Herodotos corresponds to what we know of the military capabilities of each city.
Hammond, retreating from the extreme positions of the critical school, estimated the Greeks at 100,000 and the Persians at 300,000, but did not explain how Mardonios could marshall an army several times larger than that which had invaded Greece about two years earlier.
Critical historians are bound by their low estimate of the size of Xerxes' army at the beginning of the campaign. In order to justify in some way the size of Mardonios' army, all these scholars must follow Niebuhr in claiming that the entire Persian army was left with Mardonios and that the disastrous retreat of King Xerxes, described by Aischylos and Herodotos, is sheer fabrication. For Meyer only the contingent commanded by Artabazos escorted the King, returning later to join Mardonios. According to De Sanctis "the bulk" of the army remained with Mardonios. Glotz agrees with Meyer, but he tries to soften the wording, though not the substance, of the contention that the disastrous retreat of the King's army is the product of mythological imagination.
If these scholars were right, Aischylos' Persians should be read as a comedy rather than as a tragedy. Perikles, who put this play on stage as a political gesture in favor of a rapprochement between Athens and Persia, must have been in truth putting forth his candidacy for the position of jester at the court of Susa.
But even granting that all the Persians forces were left with Mardonios, the accounts of the battle of Plataia indicate that the Persians had a most substantial army. Beloch and De Sanctis try to evade the issue by a further revision of the historical tradition: there was not much of a battle of Plataia. For once, they accept one of the figures cited by Herodotos: the 40,000 men who, according to Herodotos, deserted Mardonios' army on the eve of the battle, in reality were the number of Persians who were able to withdraw after the battle. Since the Persian army would not have been much larger than 40,000 men, it follows that it went away unscathed. According to Beloch the Greek losses were small, not because the Greeks were able to cut to pieces the disorganized Persians, but because the Persians slipped away.
All told, the history of the events of 480 and 479 B.C. should be rewritten as follows: a rather small Persian expeditionary force was able to invade Greece, to ravage, undisturbed, the country for two years, to destroy several cities among which was Athens, and to withdraw with limited losses.
The conclusion that the most important military campaign of ancient history was a rather modest affair was used to question all ancient information about great military actions. The current approach is well summarized by a scholar of Carthaginian history, Pierre Hubac. In analyzing the wars between Carthaginians and Greeks and between Carthaginians and Romans, he declares that sources that mention Carthaginian armies of 50,000 men can be completely disregarded, since Europe did not see armies of this size until recent times. According to him the figures provided by ancient historians would be valuable only as a subject for a psychoanalytic investigation. His remarks can be abstracted as follows:
Should we say that every number used by ancient historians must be presumed to be false? . . . We should rather be aware of the use that the ancients made of numbers. The real value of these numbers does not correspond at all to its true value; it does not correspond at all to the value and the use that we give to numbers in our age. For us, to quote a number means to exhibit precision, to give the result of a mathematical operation . . . Scientific method compels us to submit precise figures only with prudence and with respect for exactness . . . the mania for numerical data is a recent one in which neither the Orient nor Antiquity would have liked to compete with us . . . Antiquity plays with numbers as with sparkling stones, making them shine . . . The Orient and Antiquity never knew how to distinguish dream-states from wakefulness. They did not find the real attractive; the dazzlement of dreams was considered much preferable. Could we ask a storyteller to act like a scholar: to have method, to respect documents, to be obsessed with numerical precision, to be concerned with statistics? (93)
Pierre Hubac, Carthage second ed. (Paris, 1952), pp. 122f.