When evolutionary anthropologist David Samson was living and working among the Hadza tribe in northern Tanzania, he noticed something puzzling: Their sleep was highly fragmented, short in duration, and low in “efficiency,” or actual time spent sleeping versus time in bed. This broken sleep was partly the consequence of activity and noise well into the night in the camps. People stayed up late telling stories, sharing food, and dancing. Yet, the Hadza uniformly reported high satisfaction with their sleep. It challenged the so-called Paleo sleep hypothesis, the notion that hunter-gatherer sleep must be optimally long and deep, but also ran counter to the medical orthodoxy that unbroken sleep is essential to a good night’s rest.
Featured Video
Samson traveled to Tanzania to hang out with the Hadza because he was trying to untangle what he calls the sleep paradox: Sleep is critical to human functioning, and yet, we sleep fewer total hours than any other ape, and are still arguably the most evolutionarily successful of the primates. This riddle sent him climbing into chimpanzee nests high in the trees and exploring the sleeping huts and practices of communities around Africa and Madagascar.
The stories he collected, people he met, and research findings he uncovered are vividly described in his new book The Sleepless Ape: The Story of Sleep in Human Evolution. I spoke with Samson about what we really need for a good night’s sleep, why we may be on the cusp of a “sleep enlightenment,” and the origins of what he calls the “lie-down-and-die” model of Western sleep.
What is the paradox of human sleep?
You can just say, “Oh, here’s where humans are relative to other primates on sleep.” But that misses the deeper evolutionary story. When I was a postdoc at Duke University about a decade ago, new data was emerging showing that even after controlling for brain size, body size, social order, and actual phylogenetic relatedness, humans are weird outliers—we’re the shortest sleeping primates, yet we pack in the most REM sleep relative to this short duration.
This emerged to me as one of these mysteries that I had to sink my teeth into, and it eventually became the thesis of the book. As humans, we all know what it’s like to have a bad night’s sleep. If you have only two or three hours under your belt, your cognition suffers, your capacity to plan ahead suffers, your social regulation suffers. REM sleep functions almost like a nightly behavioral therapist. Sleep deprivation also erodes your ability to resist disease and to resist injury. We know experientially on a day-to-day basis how desperately we need sleep, yet we’re the shortest sleeping primate ever recorded on the planet, by a very far measure. The owl monkey, by contrast, sleeps 17 hours in a 24-hour period. That’s the paradox in a nutshell.
Read more: “What We Can Learn from an Insomniac Fish”
Part of why we’re so different from other apes is that we descended from the trees around 1.8 million years ago and learned to sleep in tight formations on the ground, as you note in the book. What else was going on at that time that allowed this shift to happen?
I use the acronym SHELL to explain the shift.
S is for shelter. You can build a sleeping platform if you’re a great ape, or a bird-like nest if you’re a galago or a lemur. But what we were doing by 1.8 million years ago was building physical micro habitats out of, say, acacia, like the Hadza. That narrows the variations in the extremes of either cold or hot temperatures.
H is for heat. Mastery of fire, if you’re a proponent of Richard Wrangham’s cooking hypothesis, gave us a thermal regulatory buffer.
E is environmental preparation. We lived in camps and bands of adults that moved seasonally, much like foragers do, something you can see in the paleoanthropological record. Every time we moved, even generationally, we enhanced these environmental spaces, making it a little bit better for next season.
L number one is for lux, or light. Constant exposure to natural ambient light allowed our circadian rhythms to sync to real-time sensory signals around us.
L number two is for lookouts. Because we were living in that social group, we also had socially buffered security for our sleep sites.
Together, this created a novel sleep ecology, what I call a physical sleep exophenotype. An exophenotype is when a genetically driven instinct reshapes the environment, which in turn shapes a species behavior, feeding back into its evolution. A classic example is a beaver dam. Beavers create these fantastic feats of engineering, creating ponds artificially that then push them to specialize foraging skills suited to that artificial environment. That’s the kind of feedback loop that we got into one and a half million years ago that led to the drastic evolutionary changes in our sleep relative to other primates.
How did these shorter but more REM-saturated patterns of sleep influence human cognition? You mention one 2009 study in the book, “The Role of Sleep in Cognition and Emotion,” which was influential in your thinking about this.
Yes, that’s a brilliant paper by Matthew Walker, who makes the argument that REM sleep is a nighttime therapist. During REM sleep, you strip the emotional charge off a memory, so that when you recall it later while awake, you don’t have to re-experience the original emotion. It’s the science behind, “Sleep on it, honey, you’ll feel better in the morning.” You release the affective energy so that you can process the information without the emotional weight attached.
That’s why one of the leading edge therapies for PTSD now targets sleep quality directly. Trauma disrupts sleep, which prevents that affective release. Treat the sleep and the nightly passive therapy can resume, helping to resolve trauma symptoms.
If sleep is so important to so many functions, why is insomnia so common in humans?
First, I’m not a clinician. Always consult your physician before you listen to me. I’m just an evolutionary anthropologist. But my intuition is that we over-problematize our sleep. One of the framings from evolutionary medicine involves hypervigilance. Let’s say somebody comes in with insomnia, you can reframe it as, “Hypervigilance was probably quite adaptive in an ancestral condition.” What you have is a perfectly normal outcropping of a psychological adaptation, not something broken. You’re hypervigilant at night, which probably served many of your ancestors very well. You wouldn’t be here if not for some level of hypervigilance. Reframing it that way alone tends to reduce symptoms.
The problem arises when your body can’t distinguish, on a mechanistic level, between your 9 a.m. PowerPoint to the board and a tiger stalking you in the bush. The HPA axis—the hypothalamic-pituitary-adrenal axis—cannot distinguish between the two. It burns hot either way, and that can produce insomnia.
The other angle is an evolutionary mismatch in our circadian physiology. We evolved for how things were, not how things are. Today, we spend way too much time indoors. We’re sedentary. Right now, I’m in a temperature-controlled room, and it’s hot outside. My body’s being denied crucial information about temperature oscillation. That information normally helps a dozen or so independent clocks in my body coordinate physiological processes. Light and the temperature cues are crucial, and we’ve been disrupting them ever since Edison’s light bulb. Maybe even earlier, since the invention of fire, although fire’s much less a culprit because it doesn’t emit blue-wave light. Blue-wave light is the real problem because it inhibits melatonin production.
You say we evolved for how things were****, not how things are, so to get a picture of how things were, you studied the sleep customs of a number of hunter-gather tribes, like the Hadza and Himba, and Malagasy communities in Madagascar. What does a good night’s sleep look like for them?
Well, there’s a lot of variation by ecology and socio-technological adaptation. Not all small-scale societies are alike, but there are patterns. One paper we just published when I was writing the book found that people in large-scale societies were sleeping longer, and with higher sleep quality than people in small-scale off-grid societies. Sleep efficiency—time spent asleep versus time spent in bed—averages around 88 percent in the global north versus 74 percent in small scale off-grid societies. Sleep scientists are usually thrilled to improve someone’s sleep by 10 or 15 minutes a night. These are categorically larger differences.
It’s kind of counterintuitive, right? It flies in the face of the sleep epidemic hypothesis. For the past 15 years, the CDC has painted modern global northern sleepers as the worst sleeping cohort of all time, like we’re in a sleep dystopia. I never bought it, because I’d just started doing field work in small-scale foraging societies. I’m like, “No way. They’re partying all night.” But they can, because there’s no top-down pressure to go into the office at 8 a.m. They’re free to catch supplemental sleep as needed throughout the 24-hour period.
Most importantly, when we measured their circadian function directly, small-scale communities had much higher amplitude circadian rhythms than global north populations, which I think is why they report being happy with their sleep even though it’s more fragmented. That satisfaction with their sleep seems to be underpinned by healthy, strong circadian function.
How do you square the fact that these small-scale societies are so happy with their fragmented sleep with the clinical evidence that links short, broken sleep to so many different disorders and health problems, such as diabetes, obesity, Alzheimer’s, heart disease, cancer, and infertility?
The science is catching up. Sleep and chronobiology, the study of circadian rhythms, have historically been completely separate fields. Now, they sit side-by-side at conferences, because they’re deeply interlinked. Looking at sleep duration in isolation doesn’t tell the whole story. Six and a half hours of sleep with poor circadian function feels fundamentally different from six and a half hours with strong circadian function, even for the same person.
Across cultures, the human sleep average clusters just under seven hours, which looks like a robust signal of healthy sleep. But if you go an hour or two under, or an hour or two over, it tends to be bad. Oversleeping is even a marker of underlying comorbidities like depression. Getting more than eight and a half hours of sleep consistently is associated with increased mortality. There’s a sweet spot.
How can modern humans actually arrange their lives to sleep in a way that’s more aligned with circadian rhythms, that isn’t against the clock? I don’t see 9 a.m. meetings, 9-to-5 work schedules, or air conditioning going away any time soon.
The good news is that we’re actually sleeping okay. 7.1 hours a night sits in the sweet spot. There are gains that we can be very happy about for many of us in the West, though they’re not evenly distributed. We sleep in relatively safe, temperature-controlled environments (18 degrees Celsius is considered optimal for sleep depth), on beautiful plush beds. The Hadza sleep on a centimeter of textile or hide and half of them don’t use a pillow. It’s very spartan. But 98 percent of them report being happy with their sleep.
I’m not recommending a paleo sleep diet where we all go out in the bush and sleep. I don’t think that’s the answer. I think we hold our gains, but we become way more aware of circadian function, cuing our bodies with the environmental signals it evolved to expect. That means a minimum 15 percent of your waking hours are spent outside. Most people living in the global north fall well below that threshold. We’re on the cusp of a sleep enlightenment.
Does that mean being specifically outdoors, or is being near a window enough? Is it just getting natural light?
It’s light and temperature. Some emerging research suggests Earth’s magnetism matters, as well. Being barefoot on the ground versus 30 stories high in a downtown skyscraper puts you in a completely different magnetic environment. There might even be cues that science hasn’t measured yet. There always are. We’re just figuring this stuff out.
Window type matters, too. Energy-efficiency regulations, at least in Canada, have mandated glass that blocks a lot of full-spectrum light. Older pre-1980s windows let a lot more of that light through. You can literally feel the heat of full-spectrum light coming through them, which is thought to help “prime mitochondria function.” In the name of efficiency, we may be quietly disrupting our circadian physiology.
Read more: “Why Vivid Dreams Make for Better Sleep”
I was fascinated to read that sleeping with a partner or in groups is supposed to support healthy sleep. What is going on here?
It depends on the relationship. If you’re in a good relationship, sleeping together is great for your sleep. In a bad one, it hurts. It’s fairly intuitive. The underlying idea is that absent a state that protects you, a social group is one of the best ways to secure safe sleep. For nearly all of human history, it was up to the group to shore up the sleep environment.
There’s also the first-night effect. Sleeping alone in a novel place produces a worse first night’s sleep. A new hotel room is the classic example. It fades the second night if you stay put, but resets if you keep moving. There is also a kind of “social tax.” People who report feeling isolated or lonely tend to run metabolically hotter, burning more calories per unit time, plausibly because they’re unconsciously aware that the only eyes protecting them from uncertainty in the environment are their own. This aligns with Jim Coan’s social baseline theory.
Is this an argument for siblings sharing bedrooms when they’re growing up?
I’m all for it. I’ll encourage it with my own kids. We’re a pretty pro-co-sleeping family. I think two or three siblings in the same room is pretty manageable versus, say, the 180-square-foot Hadza huts I saw housing a mom, a dad, and nine other kids.
You use this term “lie down and die model,” to describe sleep in the modern global north. Where does this model come from?
I grew up with the folk wisdom of “Get your eight hours.” In Wild Nights, a cultural history of sleep, author Benjamin Reiss traces it to post-industrial labor negotiations. Unions argued that a third of the day should be set aside for rest—eight hours. But that wasn’t a reflection of biological need. It was a floor carved out in negotiations over time away from work.
That history shaped what we now think of as normal sleep, and points directly to the lie down and die model—the assumption that you owe peak productivity the other two-thirds of the day. A more dynamic and flexible, polyphasic sleep pattern, would threaten the rigid 9-to-5 model, so the model itself discourages it. It’s a product of particular sociocultural and historical factors, not biology.
Has human sleep finished evolving, or will we continue to adapt?
I’d be very cautious about trying to engineer humans to sleep less. We know short-sleep genes exist: DEC2, for instance, where some individuals can function well on only four to five hours of sleep. In theory, you could edit for it with CRISPR. There might be legitimate use cases: long-duration space flight, high-pressure emergency medicine. Artificial selection already happens in a sense. U.S. Army Ranger School is notoriously built around sleep deprivation. They’ll go two weeks on three hours sleep a night, and they’re performing at the most extreme levels of human performance day in and day out—effectively selecting for people with an unusually DEC2-like capacity.
I’d still be cautious about trying to engineer reduced sleep. Immune strengthening and cellular restoration depend on slow-wave sleep, and for a complex social animal, emotional regulation depends on REM.
Beyond genetics, body mass is one of the most powerful predictors of total sleep duration across all animal life. Sleep duration scales with mass and energy consumption as a power law. Whatever the future of human sleep is, we need to understand these constraints before intervening genetically. And if humans were ever to migrate to Mars or another planet, we’d need to replicate Earth’s 24-hour light and temperature cues with real scientific precision. Getting that wrong would be a mission critical failure.
What is your sleep like? Do you have any sleep rituals, and has this research changed the way you sleep?
I sleep well, but not because I obsess over it. It comes from paying close attention to circadian function. I anchor my mornings with full-spectrum outdoor light. My son and I go outside, say hi to the sun, and I drink my cup of coffee. That’s one of the most effective ways to start your circadian “timer.” At noon, if I have access to green space and sunlight, I try to get outdoor temperature exposure, too, since that gives the body both time and thermal cues. Green plants also reflect infrared light, which is thought to support mitochondrial function. After sundown, I’m careful to avoid blue light. It’s abundant, but nutritionally poor light, especially at night. I also don’t eat three to four hours before bed.
Over the past four or five years, as I’ve come to understand and respect my circadian function, I’ve had the best sleep of my life. 
Enjoying Nautilus? Subscribe to our free newsletter.
Lead image: Darumo / Adobe Stock
Instagram Engineering
Shopify Engineering