Key Takeaways
- Before electric light, segmented sleep was the norm — people slept in two shifts with a 1-2 hour waking period around midnight
- Historian Roger Ekirch found hundreds of references to "first sleep" and "second sleep" in pre-industrial European texts
- A 1992 lab study recreated 14-hour darkness conditions and participants spontaneously returned to biphasic patterns within weeks
- Artificial light collapsed the two-phase pattern into a single consolidated block — this happened relatively recently, within the last 200 years
- If you wake up around midnight feeling calm and alert, you may not have insomnia — you might be experiencing an ancestral sleep pattern
If you wake up at 1am feeling completely alert and can't figure out what's wrong with you, this article is for you. I used to lie there stressing about being awake, which made it worse, which made me more awake, until eventually I'd fall back asleep around 4am and feel terrible at 7. The cycle ran for months.
Then I read about Roger Ekirch's work, and something genuinely shifted for me. Not because it fixed my sleep — it didn't, immediately — but because it reframed what was happening. I wasn't broken. I was possibly just doing something my ancestors did on purpose.
01 Roger Ekirch and the Discovery of Segmented Sleep
Roger Ekirch is a historian at Virginia Tech who spent years researching sleep in preindustrial Europe. Over a decade of archival work, he combed through diaries, court records, medical texts, literature, and letters from the medieval period through the early 19th century. He published his findings in 2001, and then in his 2005 book "At Day's Close: Night in Times Past."
What he found was startling: across hundreds of sources, people casually referred to "first sleep" and "second sleep" as distinct periods, separated by a waking interval of roughly one to two hours. This wasn't described as insomnia. It wasn't described as a problem. It was just how the night worked[1].
"He would wake after his first sleep and then lie awake for the better part of an hour..."
— From a 16th century English account"We shall also speak of 'broken' sleep between first and second sleep, as being a natural custom of the body..."
— 15th century medical text (paraphrased)References to "first sleep" appear in sources ranging from Homer's Odyssey to 17th century French medical journals to colonial American diaries.
— Ekirch, At Day's Close (2005)The waking period wasn't passive. People did things. They prayed. They had sex. They talked with their spouses. They got up to check on animals or the fire. Some wrote in journals or did creative work, finding the midnight quiet conducive to thinking. A 15th century French physician actually recommended the period between first and second sleep as the best time for sex because "at that time they will enjoy it more and do it better."
This was not a small sample or a regional pattern. Ekirch found references across England, France, the Netherlands, Italy, and colonial America. The medieval phrase was "first sleep" (premier somme in French, primo somno in Latin). By the 17th century it appears in Shakespeare, Cervantes, and countless lesser-known texts. It was just... what sleep was.
02 The 1992 Lab Study That Recreated It
Thomas Wehr was a sleep researcher at the National Institute of Mental Health who, almost independently of Ekirch's historical work, ran an experiment in 1992 that accidentally confirmed the same thing[2]. He took a group of volunteers and put them in conditions of 14 hours of darkness per night — roughly what pre-industrial humans would have experienced in winter at mid-latitudes.
Within a few weeks, every participant had shifted to a biphasic pattern. They'd fall asleep, sleep for about four hours, wake up for one to two hours, and then sleep for another four hours. The waking period wasn't anxious. Participants reported feeling calm, reflective, and unusually clear-headed during it. Their prolactin levels (a hormone associated with quiescence and calm) were significantly elevated during the inter-sleep period.
Wehr interpreted this as evidence that consolidated eight-hour sleep is an artifact of artificial light, not a biological imperative. When the brain is given enough darkness — more than modern electric light allows — it naturally divides the night into two sleep phases.
"The evidence strongly suggests that segmented sleep is the ancestral human sleep pattern, and consolidated sleep the modern artifact."
— Roger Ekirch, "Sleep We Have Lost" (2001)
03 Pre-Industrial Sleep Wasn't Actually Better
Here's where I want to push back a little on the romanticization that sometimes surrounds the "ancestral sleep" narrative. Biphasic sleep was natural, yes, but the conditions around it were genuinely terrible by modern standards. This is worth being honest about.
The Cold Problem
Before central heating, winter nights in Northern Europe were genuinely dangerous. People often shared beds with multiple family members, and sometimes animals, for warmth. Cold-related awakenings were frequent, and going to bed was a much more elaborate ritual of preparation against freezing.
Parasites and Vermin
Bed bugs, fleas, and lice were endemic. The phrase "sleep tight, don't let the bedbugs bite" is not metaphorical — it was practical advice, and the bugs frequently won. Regular awakening from bites was a normal part of pre-industrial sleep.
Shared Beds, Shared Problems
Privacy in sleep is a modern concept. Most pre-industrial people slept alongside spouses, children, servants, or strangers at inns. One person's nightmare, cough, or nocturnal prayer woke the others. The social intimacy was sometimes valued; the sleep disruption less so.
No Mattresses as We Know Them
Straw pallets or straw-stuffed sacks were the norm for most of history, not the exception. Even the wealthy slept on feather beds that were enormously expensive and often damp. Chronic back pain from poor sleeping surfaces was universal.
The biphasic pattern was natural, but "natural" in this case coexisted with cold, vermin, noise, shared beds, and the general physical discomfort of pre-industrial life. The waking period between sleeps was probably also sometimes driven by sheer discomfort rather than tranquil reflective time, despite how some sources describe it.
04 How Artificial Light Changed Everything
The transition from segmented to consolidated sleep happened faster than you might expect. Gas lighting became widespread in European and American cities in the 1820s and 1830s. Electric light followed in the 1880s. Within about 50 years, references to "first sleep" and "second sleep" in written materials essentially disappear. The terms fall out of use so completely that by the mid-20th century, nobody remembered they had existed.
What artificial light did was extend the period of wakefulness into the evening, compressing the night. Instead of going to bed at 8pm and rising with first light, people stayed up until 10 or 11pm and needed to wake up at the same time for work. The 8-hour sleep window shrank. With less time in bed, the biology adapted to maximize sleep efficiency — consolidating two phases into one, eliminating the waking interval that used to separate them.
Pre-Industrial (pre-1820)
Modern (post-electricity)
05 Sleeping Pills Through History, and What They Tell Us
People have been trying to chemically modify sleep for as long as we have recorded history. The history of sleep medication is mostly a history of discovering substances that work tolerably well in the short term and cause problems over time.
Opium and herbs
Poppy-derived preparations, mandrake, and valerian appear in ancient Greek and Roman medical texts as sleep aids. Effective at sedation; addictive, with known withdrawal effects described in ancient sources.
Laudanum (opium in alcohol)
Became widely available as a patent medicine. Broadly used for insomnia, pain, and anxiety across all social classes. The Victorians had a significant laudanum dependency problem that's rarely acknowledged in popular history.
Chloral hydrate and bromides
First synthetic sleep medications. Chloral hydrate (the original "knockout drops") is still occasionally used today. Barbiturates followed shortly after — highly effective sedatives with narrow safety margins and severe dependence potential.
Benzodiazepines, then Z-drugs
Valium and its relatives were marketed as safe alternatives to barbiturates, until their dependence potential became clear. Z-drugs (zolpidem/Ambien, zopiclone) followed with similar promises. More recent options include orexin antagonists (Belsomra/suvorexant), which work on a different mechanism with less dependence risk.
The pattern across all of these is that we keep reinventing the same wheel: a new drug that works short-term, followed by recognition that it creates problems with chronic use, followed by the next drug. The honest assessment is that no currently available sleeping medication fully preserves normal sleep architecture, and most are best used for brief periods rather than as chronic management.
What This Means for Middle-of-the-Night Waking
If you regularly wake up around midnight to 2am feeling calm and alert, and this pattern has been distressing you, it's worth considering whether you might be experiencing a historical norm rather than a disorder. The clinical criteria for insomnia include distress and functional impairment — if the waking is peaceful and you fall back asleep without too much trouble, it may not require treatment.
That said: if the waking is anxious, prolonged, accompanied by racing thoughts, and leaves you impaired the next day — that's worth addressing. The historical context doesn't make all middle-of-the-night waking benign. It just removes the assumption that any waking is automatically pathological.
What to actually do with this information
The first thing is to stop catastrophizing about waking up at night. Before electricity, everyone did this. The waking period between first and second sleep was used for prayer, reflection, sex, and conversation. Nobody wrote in their diary "I woke up at midnight again, something is wrong with me." They just got up and did things.
The second thing: if you do wake up in the night, try not to try too hard to go back to sleep. Straining at sleep is the most reliable way to prevent it. Reading, quiet reflection, getting up briefly — all of these are things your ancestors did and considered normal. The 8-hour unbroken block is the historical anomaly, not the waking.
The bigger lesson from sleep history isn't that we should go back to biphasic sleep — it's that there's more than one way to be a normally sleeping human. The modern consolidated block works well for most people most of the time. But if yours doesn't consolidate perfectly every night, you're in very long company.
Sources & Further Reading
- "Sleep We Have Lost: Pre-Industrial Slumber in the British Isles." The American Historical Review, 106(2), 343–386. (2001) JSTOR →
- "In short photoperiods, human sleep is biphasic." Journal of Sleep Research, 1(2), 103–107. (1992) PubMed →
- "At Day's Close: Night in Times Past." W. W. Norton & Company. (2005) Amazon →
- "Overcoming Insomnia and Sleep Problems: A self-help guide using Cognitive Behavioral Techniques." Constable and Robinson. (2006) Sleep Foundation CBT-I →
