Hitobashira

A dam keeps failing. A bridge washes out every time the river rises. A castle keep cracks no matter how it's built.

And the tradition says there was a remedy: bury a person in the foundation.

The word is hitobashira — "human pillar." A life placed under the structure, as a prayer made permanent, so that the thing would finally stand.

The idea is old enough to appear in the Nihon Shoki, the eighth-century chronicle, where an oracle at an embankment calls for it.

The cases the tradition names

Maruoka Castle. The keep would not hold. A blind woman named O-Shizu is said to have volunteered for the central pillar, on one condition: that her son be raised to samurai status.

The keep stood. The bargain was not honoured.

And the tradition says her resentment surfaced afterwards — that the moat's seasonal rise, every spring, is her weeping. The castle still stands; the moat still rises.

That is the standard Japanese ghost mechanism, applied to a contract. Not grief, not madness — a breach of terms, enforced from underneath the building by the party who performed first.

Matsue Ōhashi bridge. The pilings would not hold, and a man named Gensuke was taken.

And here is the detail that is harder to sit with than any of the rest: the bridge remained immovable for three hundred years afterwards.

The tradition remembers the sacrifice as having worked. That's the part that lingers. A story about a murder that failed would be one thing. A story in which the town got exactly what it paid for is another.

Matsue Castle carries a version too — a dancing girl said to be in the stone walls, connected locally to a rule forbidding girls to dance in the streets below. Lafcadio Hearn recorded that one.

Did it actually happen

The honest answer is that scholars are not agreed, and the reasons are worth stating.

The archaeological evidence is thin and contested, and the construction failures these legends explain have perfectly good architectural causes — bad ground, inadequate footings, a river that no premodern engineering was going to master. Buildings failed because building was hard. Many of these stories may be explanation rather than memory.

But the stories exist, in quantity, attached to real named structures, and they were told by the people who lived beside them.

Which means the useful way to read them is not as a verdict on whether it happened, but as something a culture wrote about itself: a record of what people believed their safety had cost.

Both readings, in the end, explain the hauntings equally well. A building bought with a life ought to be haunted. And a culture that keeps telling itself these stories is doing something like auditing its own foundations — checking, over and over, in public, what was under there.

Maruoka and Matsue are both still standing. You can go and look at them.