How Did Humans Sleep Before Electric Light?

You've probably come across the claim that before electric light, humans didn't sleep through the night at all, they slept in two distinct chunks with an hour or so of quiet wakefulness in between. It's a great story, and it's based on real historical evidence. It's also more contested among historians than the version that usually gets repeated, and the nuance is the more interesting part.

The historian who found it

This whole idea traces back to historian Roger Ekirch, who spent years combing through diaries, court records, medical texts, and literature from the 15th through 18th centuries. He found hundreds of casual, matter-of-fact references to a "first sleep" and "second sleep," with an interval of wakefulness in between that people sometimes called simply "the watch" or "watching."¹

The references weren't dramatic or medical, they were completely offhand, the way you'd mention checking your phone. People used the between-sleep period for all sorts of ordinary things: praying, having sex, chatting with a bedmate, smoking, tending to a fire, even popping over to a neighbor's house at 1am for a chat, which is a level of casual midnight socializing that would get you a very concerned group text today. Chaucer and Dickens both mention it in passing, as something readers would recognize without explanation. Then, starting in the late 1600s, these references quietly disappear from the written record, tracking closely with the spread of artificial lighting in wealthier households, and later, the wider population.

The famous experiment that seemed to prove it in a lab

In the early 1990s, psychiatrist Thomas Wehr ran an experiment where volunteers spent a month in an environment with only 10 hours of light per day, mimicking a natural winter photoperiod with no artificial evening lighting. In the first couple of weeks, people mostly slept long, consolidated stretches, likely paying down existing sleep debt. But by around the fourth week, their sleep had reorganized into two distinct bouts of three to five hours each, separated by one to three hours of quiet wakefulness.² It looked like a real-time recreation of exactly what Ekirch had found in the historical record. Modern sleep science describes a different, single-block architecture as the norm, see why the first 90 minutes of sleep matter most for how that actually works.

Where it gets more complicated

A more recent historical reconsideration, published in the journal Medical History, pushed back on how universal and standard segmented sleep actually was.³ The critique isn't that Ekirch's sources are fake, they're real, it's that the practice may have been less uniform across time, place, and social class than the popular narrative suggests, and that some of the language could plausibly describe looser, more varied sleep habits rather than one single, near-universal biological pattern everyone followed every single night.

Wehr's lab result is real too, but it's one small study, under a specific artificial condition (extended winter-length darkness with nothing else to do), a fascinating data point rather than proof of humanity's one true default sleep pattern for all of pre-industrial history.

Putting the two together: segmented sleep was clearly a real and, in some times and places, common practice, and it's biologically plausible under specific conditions of extended darkness. Whether it was the single universal human pattern before electric light, or one variant among several that coexisted with more consolidated sleep, is still a properly open historical question.

Should you try it?

Probably not on purpose, and please do not set an alarm for 2am to "reclaim your ancestral sleep watch." Wehr's segmented pattern only emerged after weeks in a specific, extended-darkness environment nothing like a normal modern night, and there's no solid evidence that artificially splitting your own sleep in a normally-lit modern life recreates any benefit. If you do wake up in the middle of the night and find yourself lying there for a while, it's worth knowing this isn't necessarily "broken" sleep by some ancient standard, historically, plenty of people spent a chunk of the night quietly awake as a completely normal, unremarkable thing, not a crisis.

Historically, plenty of people spent a chunk of the night quietly awake as a completely normal, unremarkable thing, not a crisis.

History is messier than the theory

The two-sleeps story is real evidence and legitimately fascinating, with more open questions attached than the popular version usually lets on. Sleep has probably always looked more varied, across history and across individuals, than any single tidy narrative, including this one, wants it to be.

Sources