Why People Slept Twice a Night in the Middle Ages

Historian Roger Ekirch collected more than five hundred references to “first” and “second sleep” in diaries, court records, medical treatises, and poems. Their authors explain nothing: this is how you write about something everyone understands. A witness notes that she heard a noise “after first sleep” - and that was enough to know what time she meant.

The interval between sleeps wasn't empty. People prayed, added wood to the fire, mended clothes, talked in the dark, and visited neighbors across the yard. The French even had a separate word for this state - dorveille, “sleep-wakefulness.”

It wasn't about habit, but the length of darkness. In winter, the night lasted fourteen hours or more, and people can't sleep that long at a stretch. Sleep spread across the night and broke in the middle.

In the early nineties, Thomas Wehr at the National Institute of Mental Health gave people back a long night. For a month, participants spent fourteen hours in complete darkness - with no lamps, screens, or clocks.

After several weeks, their sleep split in two. A person would fall asleep, sleep for about four hours, wake up, spend an hour or two in calm semi-wakefulness - then fall asleep again. There was no anxiety in it: hormone levels during those hours resembled deep rest more than waking up.

The mechanism is simple. Melatonin is produced while it's dark; prolonged darkness extends that window, so sleep doesn't need to fit into one stretch. Electric light did the opposite: only about eight hours of darkness remain, and sleep gathers into one block.

It's easy to draw a neat conclusion: if two shifts are older, they must be better, and our eight hours straight are a corruption. But the same data offer an objection.