by Oswyn Murray
Herodotus’ detailed figures for the Persian navy add up to 1207 triremes, which is usually held to be the total naval strength of Persia, rather than her actual strength in this war; for Herodotus is at pains to have an unbelievable 600 of them wrecked before Salamis, the first battle where it would have been possible for the Greeks to estimate precisely their opponents. Modern estimates therefore suggest perhaps 200,000 for the army and 600 for the navy. The exact figures are not important; what is clear is that the Greek strategy in 480 was determined by the fact that they were overwhelmingly outnumbered on land, and unable to meet the enemy in pitched battle despite their superiority in armour and training; while at sea they were still inferior in numbers as well as in skill, though on this element there was more hope of victory.
The defence of Greece therefore rested on the extent to which a common strategy could be achieved; Herodotus follows tradition in minimizing the difficulties in creating a unified command, and in particular presents the Greek negotiations which led to its establishment as a timeless jumble of episodes. Two Greek congresses established the principles of action. At the first in 481 ‘those Greeks who held the better opinion about Greece came together for deliberation and exchange of pledges’ (7.145). They agreed to end mutual feuds, in particular that between Athens and Aegina, and to send spies to Asia and ambassadors to the powerful uncommitted cities of Argos, Syracuse, Corcyra and Crete. They also agreed to give the command of the Greek forces to Sparta both on land and at sea; this reveals the extent to which the Peloponnesian League was the basis of the Greek resistance, and also the willingness of Athens to sacrifice her claims to the common interest. It was apparently now that an oath was taken to destroy those cities who joined the enemy without compulsion, and dedicate a tenth of the proceeds to Delphi (7.132).
The attempts to broaden the basis of the league against Persia failed. Argos, advised by Delphi, allegedly demanded impossible conditions, and in fact seems to have had an agreement with Persia (7.148ff); Crete was also warned off by Delphi (169); Corcyra promised help, but carefully failed to deliver her sixty ships. The case of Syracuse is more interesting.
Since about 505 a major tyranny had arisen at the city of Gela, under Kleandros and his brother Hippokrates; by 491 the Geloan empire extended over most of eastern Sicily, and Syracuse was the only major city still holding out. In that year Hippokrates was succeeded by his cavalry commander Gelon, who united himself by a double marriage tie to Theron, tyrant of Akragas. In 485 Gelon intervened in civil strife at Syracuse between the Gamoroi, the original settler aristocracy, and their native serfs the Killyrioi (7.155ff). He seized Syracuse and made it his capital, transferring the population of other cities to it, and handing Gela over to his brother Hieron. Thus by 481 virtually the whole of Greek Sicily was united under the control of three tyrants. Gelon allegedly offered a huge army to the mainland Greeks in return for the supreme command; in fact he was already preoccupied with the inevitable Carthaginian response to the unification of Greek Sicily. Hamilcar king of Carthage invaded in 480 with 300,000 troops (7.165). The fourth century historian Ephoros claimed that there was collusion between Carthage and Persia in the timing of the expedition (F.G.H. 70 Frag. 186), which is not impossible, for the Phoenicians must have been deeply worried by the influx of Greeks into the western Mediterranean which had resulted from the Persian pressure on Ionia; they may well have hoped to destroy Greek power simultaneously in east and west. At the same time of year, perhaps even on the same day as the battle of Salamis, Gelon and Theron decisively defeated Hamilcar at Himera, killing him and capturing the entire expeditionary force. The Carthaginians even feared an invasion of Africa; Gelon extracted 1000 talents as indemnity, and issued a famous series of victory coins, perhaps the most beautiful coins ever minted. The great temples of Akragas still stand as monuments to the suffering of the Carthaginian slaves, and individual citizens of Akragas are said to have possessed as many as 500 of them. In 474 Hieron defeated the Etruscans at the battle of Kyme, and the west was safe.
As a double insurance Gelon had sent a large sum of money to Delphi, to be presented to Xerxes if he won. Delphi in fact can be seen as consistently pro-Persian in this period, with her oracle constantly advising non-intervention or submission to Xerxes. When Athenian envoys arrived, even before they had put their question, the priestess counselled flight to the west, and only finally was prevailed upon to give a more ambiguous oracle, which ended:
… far-seeing Zeus grants to the Triton-born only a wooden wall to remain unsacked, which shall help you and your children. Yet do not await in peace the cavalry and the host of infantry that come from the mainland, but turn your back and withdraw; still you shall face him again: O divine Salamis, you shall destroy the children of women, when either Demeter is scattered or comes together.
(7–142)
Out of this defeatist prophecy with great difficulty Themistokles persuaded the Athenian people that the wooden wall was the navy, and that, since Salamis was called divine rather than hateful, the disaster prophesied would fall on the Persians:
The Athenians decided formally, taking counsel after the oracle, to meet the barbarian invading Greece with their ships and all their forces, in obedience to the god, together with those of the Greeks who were willing.
(7.144)
The lapidary phrases read like a contemporary decree.
The second congress of the Greek allies took place at the Isthmus in spring 480, and determined the strategy for the coming year. The importance of the Thessalian cavalry led them first to try to protect Thessaly by holding the northern line of Mount Olympus, and an expedition of 10,000 hoplites went to the vale of Tempe; but the pass could be too easily turned, the station was only suitable for checking the enemy by land, and an important section of the Thessalian ruling class had long been pro-Persian. They withdrew to the Isthmus again, and it was probably then that they decided to hold the Persians on both elements at once, at Thermopylae and Artemisium, and in second place at the Isthmus and Salamis. The choice of these two lines was inevitable once a double strategy had been decided on; but they also corresponded to an obvious division of interest between the Peloponnesian states behind the Isthmus, and the cities of central Greece, and especially Athens. This natural division finds constant expression in the narrative of Herodotus, and often threatened seriously the Greek resistance; indeed it may well be asked whether the Greek victory was due to their precarious unity, to their enemy’s mistakes, to chance, or to the heroism and intelligence of individual leaders. I shall not attempt to answer this question.
Certainly, to the modern strategist, Thermopylae and Artemisium seem to offer the best hope. The pass of Thermopylae was difficult to turn, and could easily be held against superior numbers; while at Artemisium the fleet could fight with their backs to the friendly island of Euboea, with safe harbours and an easy escape route down the channel between the island and the mainland, while their enemy had to moor or beach in an exposed bay. The only disadvantage was that the waters were too open to compensate fully for the superiority of the Persian fleet. Still this was the only place where army and fleet could deploy effectively side by side.
It is not easy to see whether the Greeks fully accepted such an analysis, or precisely what their strategy was on the two elements. Three factors make for uncertainty. The first is that since Herodotus tells the story of the two battles of Thermopylae and Artemisium separately, their detailed relationship is obscure. However it seems that the plan was to hold Thermopylae, while adopting a more offensive posture at sea: there were a number of raids on the Persian ships, and the Greeks finally accepted the Persian invitation to fight. Much must have depended on where the Persians chose to throw their emphasis. To them it must have been obvious that once Thermopylae had fallen, the Greek fleet would be forced to retire: their initial effort was therefore concentrated on land. On the other hand the exposed position of their fleet, reflected in the continuing stories of storm damage, made delay
particularly dangerous at sea.
The second factor in the analysis of the Greek strategy is the problem of Greek numbers. At sea Artemisium was an all-out effort: the fleet consisted of 271 triremes with 53 apparently covering the line of retreat (8.1); this comprised the entire Athenian navy of 200 and roughly three quarters of the Peloponnesian fleet (including full contingents from Corinth and Megara, but not Aegina). The land numbers are a different matter. The figures given by Herodotus show only 300 Spartans, 2120 Arcadians and 400 Corinthians (the total of 4000 Peloponnesian troops in the inscription probably includes helots); in addition there were 1100 Boeotians and 1000 Phocians. This figure should be compared with the forces united for the battle of Plataea in 479: 38,700 men, mostly Peloponnesians. The discrepancy cannot be explained by the absence of 8000 Athenians, and perhaps others, fighting on board ship. There was too a clear reluctance on the part of Peloponnesian states to commit their troops so far north; many stayed behind to fortify the Isthmus, and the Spartans themselves claimed that their contingent was an advance force, while their main army was again delayed for the festival of the Karneia. One reason for the fall of Thermopylae was a lack of manpower; for the mountain track which turned the position was guarded only by the Phocians. Either then the plan to fight at the Isthmus was a genuine alternative, and Thermopylae only a holding position, or Greek unity split on geographical lines, and the Peloponnesian cities in particular failed to support the advanced position agreed on. The discrepancy between land and sea forces suggests the latter. If this is the correct alternative, then only the heroism of Leonidas saved the cause of unity, when there might so easily have been recrimination; after direct assault had failed, a Greek traitor told the Persians of the mountain path. When Leonidas heard that the position was turned, he dismissed the allies, but remained with his three hundred Spartans (and the Boeotians who refused to leave him), to die a hero’s death:
Tell them in Lacedaemon, passer-by,
obedient to their orders, here we lie.
(7.228)
His gesture saved Greece; the Spartans had kept their word.
The third factor is the ‘decree of Themistokles’, an inscription discovered in 1959 in Troezen in the east Peloponnese opposite Attica. The stone is almost complete and relatively well preserved; it reads (with a few minor uncertainties) as follows:
Gods
It was resolved by the council and the assembly;
Themistokles son of Neokles of Phrearrhioi proposed:
To entrust the city to Athene protectress of Athens and all the other gods, to give protection and defence against the barbarian on behalf of the land. All the Athenians and the foreigners resident in Athens are to place their children and women in Troezen [21 letters missing] the founder of the land; they are to place the old men and the movable property on Salamis. The treasurers and the priestesses are to remain on the Acropolis guarding the possessions of the gods. All the rest of the Athenians and the foreigners who have reached military age are to embark on the 200 ships prepared, and fight against the barbarian for the freedom of themselves and the rest of the Greeks, alongside the Lacedaemonians and Corinthians and Aeginetans and the others who are willing to share in the danger. The generals shall appoint 200 trierarchs (captains), one in charge of each ship, beginning tomorrow, from those who possess land and a home in Athens and who have legitimate children and are not more than fifty years old, and the lot shall determine their ships. They shall choose marines, ten for each ship, from those aged between twenty and thirty, and four archers, and they shall appoint by lot the skilled officers for the ships when they draw lots for the trierarchs. The generals shall write up the rest by ship on white boards, the Athenians from the deme registers, the foreigners from the list of names registered with the polemarchos. They shall write them up assigned to 200 divisions, up to the number of 100 (men per division), and shall add to each division the name of the trireme and of the trierarch and of the officers, so that they may know On which trireme each division shall embark. When all the divisions are listed and divided by lot among the triremes, the council and the generals are to fill all the 200 ships, sacrificing an offering to Zeus the all powerful, and Athene and (or ‘of’) Victory, and Poseidon the Saviour.
When the ships are filled they are to bring aid at Artemisium in Euboea with a hundred of them, and with a hundred of them they are to lie at anchor around Salamis and the rest of Attica and defend the land. In order that all the Athenians may defend themselves against the barbarian with a single mind, those who have been exiled for ten years are to go to Salamis and wait there until the people comes to a decision about them; but those [who have lost their rights?]…
(Greek Historical Inscriptions no. 23 = 55F)
The date and purpose of the inscription are not seriously in dispute. It was carved in the late fourth or early third century BC, and set up at Troezen, one of the Athenian places of refuge (note the absence of Aegina at this point), as a reminder of the unity and courage that the Greeks had once shown in the face of a foreign invader. Its immediate aim was probably political rather than purely commemorative: its language recalls an occasion like that in 323 BC, when the Athenians decided to revolt from Macedonian rule on the news of the death of Alexander the Great:
Immediately the orators, embodying the wishes of the mob, wrote a decree that the people should take counsel for the common safety of the Greeks and free the cities subject to garrisons; they should prepare 40 quadriremes and 200 triremes, that all Athenians up to the age of 40 should be mobilized, and three tribes should guard Attica, while seven should be ready for campaigns beyond the frontiers. They should send out envoys to visit the Greek cities and inform them that formerly the people, believing that all Greece was the common fatherland of the Greeks, had fought against the barbarians who came to enslave them, and that now too the people think it right to risk lives and money and ships on behalf of the common salvation of the Greeks.
(Diodorus 18.10)
The language of this decree echoes that of the Themistokles decree; and it was most probably on this occasion (or another similar to it) that the Troezen inscription was carved by the people of Troezen, as a sign that they accepted the new alliance. The Themistokles decree had in fact been serving such propaganda purposes since at least 348, when the orator Aischines used it (Demosthenes 19.303: this is the earliest surviving reference to it). The obvious question arises, was it a forgery made in the interests of the call to political unity in the fourth century, when a number of other Persian War documents are known or thought to have been forged – for instance the obviously false ‘decree of Miltiades’, also used by Aischines?
On the issue of authenticity no agreement has been or ever will be reached. The decree is clearly intended to be that mentioned by Herodotus at 7.144 (p. 293). But form and language owe a great deal to the fourth century, and this debt is not just the consequence of a superficial reworking of an earlier document. The whole structure of the decree is literary; its coherence and organization are unparalleled in any genuine inscription of the period; it is a synthetic attempt to cover all aspects of the great event. Such arguments suggest at the least a fundamental reworking of whatever material was available. On the other side certain details seem implausible as inventions of a later forger; two are particularly important. The ships are to be filled ‘up to the number of a hundred’; the standard number of rowers in a trireme was as far as is known always 200, and there is no reason to suppose that any forger would have thought otherwise. If the Athenian ships were to be only half full, this might suggest a serous shortage of manpower: in fact Herodotus says that 20 ships were lent to colonists from Chalcis (8.1); and on other grounds it might be reasonable to put the total adult male population in 480 at nearer 20,000 than 40,000, given the known total of hoplites as 8000. It is tempting to believe that this detail at least is genuine; if so, such a fact could only have been preserved in a document, which must have contained other details.
The s
econd point is precisely the strategy implied in the decree, which splits the Athenian navy into two equal parts, and implies that Salamis and Artemisium were seen as equally important from the start; such a view seems to run counter to Herodotus. Paradoxically the conflict of evidence is easier to reconcile if the information in the decree is genuine: in that case the decree represents the plan, and Herodotus’ narrative the actual course of the action; when it was seen that the Persians were not going to mount another Marathon landing, it became possible to send all ships to Artemisium. A vestige of this original division of forces might be detected in Herodotus, who mentions a squadron of 53 Athenian ships arriving on the last day of the battle (8.14); they had apparently been guarding the narrows of Euboea against a Persian encirclement. On the other hand if the decree is in this respect a forgery, its divergence from Herodotus is more serious, for a forger should compile his document in conformity with what he believed actually to have happened; the decree then becomes evidence for an alternative version of the events at the battle of Artemisium, and so a different account of the Greek strategy; yet no surviving literary source offers such an alternative. This uncomfortable conclusion might be avoided if it could be shown that the forger had special reasons for giving an unorthodox account of the strategy of the Athenians. For instance he may have wanted to include in his document references to both the great naval battles of the war: an unnatural striving after completeness is a common fault of forgers, and one which has already been noted in this document. Or the forger may have intended his document to endorse a particular strategy appropriate to his own day; here we might compare the strategy of split forces put forward in 323, a stategy which is there associated with the Persian War, and which may well go back to the 340s and the first production of this document. Such problems will ensure that the Greek strategy in the great Persian War remains a matter for debate.