The problem of reading a prophecy
Many famous replies from Delphi share one feature: they were vague enough that more than one reading seemed possible. The ancient sources that preserve them, mostly Herodotus, wrote after the events, so the stories often serve a moral as well as a record. This article looks at two cases and then at what they suggest about how oracles were understood.
The stories are not all equally secure. Herodotus lived in the fifth century BCE and collected traditions about earlier events, so some of what he reports may have been shaped by what people later believed the oracle should have said.
Croesus and the great empire
Croesus was the king of Lydia in the mid-sixth century BCE, ruler of a wealthy kingdom in western Anatolia. As Herodotus tells it, he tested several oracles by sending messengers to ask on a set day what he was doing at that moment. Delphi, and the shrine of Amphiaraus, passed the test, according to the story. Croesus was impressed, sent gifts of great value, and then asked whether he should march against Persia.
The answer, as reported, was that if Croesus crossed the river Halys he would destroy a great empire. He took it as encouragement and attacked, around 546 BCE. He was defeated by Cyrus the Great, and the empire destroyed was his own. In Herodotus’s account, Croesus complained to Delphi and received an answer that he should have asked which empire. The reply is a textbook case of a statement that was true whatever happened.
The same king, according to Herodotus, had also been told that his rule would last until a mule became king of the Medes, a reference interpreted after the fact as Cyrus, born of a Persian father and a Median mother. Modern historians generally accept that Croesus was defeated by Cyrus, but they are less sure about the oracle’s wording. It is possible that the story was made up or reshaped to explain a surprising fall, and perhaps to protect Delphi’s reputation after the Persians won. The Lydian kingdom is associated with early Lydian gold coinage, which makes the wealth in the story plausible, though the story of the gifts also serves a literary purpose.
The wooden wall
The second famous case concerns Athens in 480 BCE, as Xerxes’ army approached. Herodotus says Athenian envoys went to Delphi and received a terrible reply telling them to flee to the ends of the earth. They asked again, and this time were told that a wooden wall would not fail and that “divine Salamis” would destroy the sons of women.
Interpretation then became a political matter. Some Athenians thought the wooden wall meant the old thorn fence on the Acropolis. Themistocles argued instead that it meant ships, and that “divine Salamis” must refer to a victory, not a disaster, because the god would not call a place divine if it were to bring ruin to the Greeks. His reading prevailed. The fleet of triremes, supported by the Greek alliance, fought at Salamis, and the result is often discussed as an example of Themistocles naval strategy. The oracle’s language fits the outcome neatly, and this is exactly what makes some historians suspicious. The story may reflect events remembered after the victory, or an exchange adjusted to fit them.
The Greek trireme was the main warship in the battle, which explains why a wooden wall could so easily be read as ships. Yet ancient readers did not see ambiguity as a flaw. Instead, they treated the work of interpretation as part of the process. The oracle supplied the words, and people had to decide what they meant.
What the cases show
Several patterns appear in both stories.
- Hindsight shapes the record. We usually hear about a reply once its meaning is clear, which creates the impression of accuracy. Replies that proved wrong or meaningless were probably forgotten.
- Interpretation is political. The wooden wall was read differently by different groups, and the winning reading was the one with a strong advocate.
- Responsibility shifts to the inquirer. In Herodotus, Croesus is blamed for failing to ask a clarifying question, which protects the god.
- Ambiguity is not unique to Delphi. Many divination systems around the world, from omens to dreams, are open to interpretation. This is part of how they survive disappointments.
A cautious modern view is that the form of these replies, and their wording, may not be reliable, and that the stories tell us more about how Greeks thought about divine communication than about exact words spoken at Delphi. It is also worth noting that in practice the oracle often dealt with simpler questions, such as which god to worship, to which a yes or no reply was possible. The famous riddles are the exception, remembered because they made good stories.