Herodotus · Book VIII
Salamis
Σαλαμίς
5,497 words · 854 unique lemmas
Herodotus > How a Coalition Holds > Salamis
Salamis in September 480 BC decides the invasion in a single day. What makes it analytically interesting is that it should not have happened: the Greek coalition was on the verge of dissolving the night before the battle, with Peloponnesian commanders arguing for withdrawal to defend the Isthmus of Corinth and abandoning Athens entirely. Themistocles forced the engagement by sending a false message to Xerxes warning that the Greek fleet was about to scatter — guaranteeing a Persian night maneuver that made withdrawal impossible. The decisive battle of the Persian Wars was manufactured by deception.
The night before Salamis, Themistocles sends his slave Sicinnus across to the Persian fleet with a message: the Greek commanders are quarreling, the coalition is fracturing, and the fleet is about to retreat. If Xerxes moves now, he can trap and destroy it in detail. Every word is false. What Themistocles knows, and Xerxes does not, is that the Persian fleet is already exhausted from the Artemisium engagements, that the strait of Salamis is too narrow for Persian numbers to count, and that an overnight maneuver will fatigue the Persian oarsmen before the battle begins. The message is not a gamble. It is a carefully engineered trap, and Xerxes walks into it while watching from a golden throne on a hill above the strait.
Salamis is a masterclass in what Sunzi calls 'shaping the enemy' — creating conditions in which the opponent makes the move you need him to make, believing it is his own decision. Themistocles has understood three things that Xerxes has not.
First, Persian naval power is a function of numbers in open water. In the strait of Salamis — roughly a mile and a half wide at its narrowest — the 1,200 Persian triremes cannot maneuver in formation. They have to enter in column, engage a waiting enemy, and cannot easily reinforce or withdraw. Every advantage Persian numbers confer disappears. The Greeks chose Salamis as their fallback position not from desperation but from calculation.
Second, coalition cohesion is Themistocles' hardest problem — harder than the Persians. The Peloponnesian commanders, led by Eurybiades of Sparta, want to withdraw to the Isthmus. Their strategic logic is sound if you ignore what abandoning Athens means: it means Athens has no reason to keep its fleet in the coalition, and without the 200 Athenian triremes, the Greek fleet cannot fight at all. Themistocles has pointed this out repeatedly. No one moves. So he sends Sicinnus. Once the Persian fleet has sealed the strait overnight, there is no retreat. The coalition is forced to fight by geography — a geography Themistocles arranged.
Third, Xerxes' presence on the hill above the strait is itself a tactical liability. Xerxes is watching; his commanders know it. Every engagement will be conducted with an eye to appearing before the king, which is not the same as winning the engagement. Artemisia of Halicarnassus — commanding a Carian contingent in the Persian fleet, and one of the sharper minds in Herodotus's narrative — rams a Persian ally when she needs to escape, and Xerxes, watching from the hill, interprets this as a Greek ship being sunk. He remarks that his men have become women and his women men. The surveillance that is meant to incentivize valor instead incentivizes performance, and performance is not the same as effectiveness.
The battle itself lasts most of the day. The Persian fleet, entering the strait in column, encounters the Greek line waiting in the dawn light. The Greek triremes are heavier; the Persian oarsmen are tired from the overnight maneuver; the confined water means the Persians' rear ships push into their front as the front falters. The Persian fleet loses roughly 200 ships. Xerxes watches from the hill and sees the invasion's logic dissolve.
Cross-Civilizational Connection
Parallel: Themistocles' use of false intelligence to shape enemy action is precisely what Sunzi prescribes in Chapter 13 ('Use of Spies') and what the Zhanguoce documents in dozens of stratagems involving planted information, false defectors, and manufactured intelligence. The Qi chapters of the Zhanguoce (08–13) are especially rich in examples of commanders who understood that the enemy's decision-making apparatus was as much a target as his armies. Sicinnus is a Chinese-style stratagem in a Greek war. The Zhanguoce's term for this kind of operation — 反間計, the reversal of the spy — describes exactly what Themistocles does: he uses the Persian intelligence apparatus to deliver a message that serves his own purposes.
Difference: The Chinese strategic tradition is generally skeptical of battle as an instrument — 'attack by stratagem' ranks above 'fighting and winning in actual battle' throughout Sunzi, and the Zhanguoce's stratagems consistently aim at outcomes short of pitched engagement. Themistocles engineers Salamis not to avoid battle but to guarantee it, on his terms. This is a different use of deception: not to substitute for fighting but to determine where and how the fight happens. The Greek tradition, even when it deploys stratagem, ends in mass violence in a way the Chinese texts treat as regrettable necessity.
Limit: The naval dimension has no real Chinese Warring States parallel. The Warring States wars are land wars; the Zhanguoce contains almost nothing about naval strategy. The logistical and tactical specifics that made Salamis decisive — the geometry of trireme combat, the significance of oarsmen's fatigue, the confined water of the strait — have no analogs in the Chinese corpus. Cross-civilizational comparison breaks down precisely here: the medium of war is sufficiently different that principles drawn from one cannot be applied to the other without distortion.
Themistocles at Salamis demonstrates that coalition management and military strategy are the same problem at different scales. The Sicinnus deception solves two problems simultaneously: it forces the Persian fleet into disadvantageous terrain, and it removes the Greek coalition's exit option. Both effects are necessary; neither alone is sufficient. The strategic genius of Salamis is not the deception in isolation — it is the recognition that the Greek fleet's internal political crisis and the Persian fleet's tactical vulnerability could both be resolved by a single message sent in the night. That is a level of systems thinking that Xerxes, watching from his hill, never approaches.
See Also
- sunzi-bingfa/13-use-of-spies — Sunzi's taxonomy of intelligence operations — including the double agent and the use of enemy intelligence channels against the enemy — maps directly onto Themistocles' use of Sicinnus as a false defector carrying manufactured information.
- sunzi-bingfa/06-weak-points-and-strong — Sunzi's central principle — 'make the enemy come to you, do not go to the enemy' — is Themistocles' entire Salamis strategy. The Greeks choose the ground, force the Persian commitment, and wait. The chapter's analysis of how to shape the battlefield before the battle is textually precise to what Themistocles accomplishes.
- sunzi-bingfa/07-maneuvering — Sunzi's discussion of how to turn disadvantage into advantage through maneuver — 'the difficulty of tactical maneuvering consists in turning the devious into the direct' — speaks to Themistocles' indirect route to forcing the decisive engagement he needs.
- zhanguoce/08-qi-1 — The first Qi chapter documents the strategic tradition of Qi, one of the Warring States most sophisticated in intelligence operations and court intrigue — the closest Chinese analog to Athens' combination of commercial wealth, democratic politics, and strategic cunning under pressure.
Edition & Source
- Author
- Ἡρόδοτος Herodotus
- Greek Text
- Perseus Digital Library
- Translation
- G.C. Macaulay (1890)