The Roman soldiers and the missing date.

A young man working in a York cellar sees Roman soldiers pass through, their legs apparently hidden below the floor. Later, a Roman road is found at the right depth. It sounds like the detail nobody could have known. Before accepting that, I want three dates: the experience, the excavation and the first recorded account.

Take the witness seriously

Harry Martindale’s reported experience at Treasurer’s House is unusually memorable. The claimed sighting dates to 1953; the story familiar to readers was recorded later. His sincerity and the cause of the experience are two different questions.

We can treat a frightened witness with respect without deciding that his interpretation has been independently verified. Ridicule would tell us nothing about the cellar.

A match needs a chronology

The famous detail is the relationship between the apparent height of the soldiers and the Roman road. To count as a prediction, it would need to be documented before the relevant archaeological information was available.

Comparing the excavation history with the later publication of the account exposes the problem. It is a much more demanding test than simply observing that two versions of the story now agree. Information can travel between an event and the eventual telling of it.

Consistency is not the same as confirmation

Uniform details and additional witnesses are often offered as support. They too need identifiable sources. Who recognised the equipment? When did another witness describe a similar event? Was that account recorded independently?

The result may be a gap we cannot close. That is a legitimate conclusion. A story does not become a Roman legion merely because the archive cannot tell us exactly what a young man experienced in a cellar decades ago.