Sunset here is around five now. Five. I watched it go down yesterday through the window and thought, well, that's the day done, and it was barely the afternoon.
And this is exactly the trap Daniel walked into with his prompt, because his instinct, my instinct, probably your instinct, is that living by the sun means lights out at five-ten and eleven hours of night.
His question is basically: who actually lives that way, and what happens to them? The traditional communities that still do it, the camping studies where modern people's body clocks snapped back to solar time within days, the individuals who've gone weeks or months without artificial light and lived to file a report. What happens to sleep, to mood, to daily life — and what did they find hardest.
And the answer to that last part is not what anyone guesses. Nobody guesses it. I've been sitting with these papers for a week and I still get it wrong when I try to predict it.
He said going to bed soon after dark sounds almost radical. It doesn't sound radical to me, it sounds like a Tuesday. But fine, let's look at the people who really do it.
So here's the surprise that anchors everything. The best data we have on humans living by the sun — and I mean actual wrist-device measurement, over a thousand nights — says they do not go to bed at sunset. They stay up two and a half to four hours past it, around fires, and then they sleep about six and a half hours.
Which is less than the average Tel Aviv commuter.
Less than the recommended minimum we've been lecturing people about for decades. And they report almost no insomnia. That's the paradox Daniel's prompt walks straight into: the romantic picture of the sun-aligned sleeper — in bed at dusk, ten hours of bliss — is wrong at both ends. They went to bed late and slept short.
The evidence splits three ways, doesn't it. There's the fieldwork — the Hadza in Tanzania, the San in Namibia, the Tsimane in Bolivia, plus the Amish as the low-tech society inside a modern country. There's Ken Wright's camping studies at Boulder, where they pulled electric light out of modern people's lives under controlled conditions. And then there are the self-experimenters, the people who just... turned everything off for a month and wrote about it.
And the through-line across all three is the same, which is what makes it worth an episode. What people lose when they live by the sun is not sleep quantity. It's the mismatch between their internal clock and the alarm clock. And the hardest part isn't the darkness. It's getting outside in winter.
That second one I don't buy yet. Let's see the receipts.
Receipts coming. Let's start where the fieldwork starts, because everyone assumes they know what pre-industrial sleep looks like and everyone is wrong.
The Yetish study. Twenty fifteen, Current Biology.
Yetish and a big team put activity monitors on ninety-four adults across three societies — Hadza hunter-gatherers, San in Namibia, Tsimane farmer-foragers in Bolivia — and got eleven hundred and sixty-five days of data. Actual measurement, not diaries, not grandpa's stories about the old country. Sleep time came out between five point seven and seven point one hours, averaging around six hours and twenty-five minutes. Near the low end of industrialized averages.
So the people with no electricity, no Netflix, no reason whatsoever to stay awake — slept about as much as a stressed accountant.
Less than we tell ourselves we need, certainly. And the timing: sleep onset ran two and a half to four point four hours after sunset. The paper's own phrasing is that sleep was confined to the latter portion of the dark period. They sat up. They talked. They kept small fires going.
Firelight — what were the actual light levels at night?
Below five lux. So dim orange, not black. That matters more than it sounds, because people imagine pre-industrial evening as this pitch-black void where sleep is the only alternative. It wasn't. It was a dim orange living room with better conversation than ours.
Five lux is a candle across the room. I read by less than that some nights and I'm told it's bad for me.
You're a sloth, your whole visual system is a compromise.
At least my compromises are ancestral.
Here's the finding I think is underreported, though. Temperature, not light, was the dominant sleep cue. The Hadza initiated sleep during a period of falling ambient temperature — always, that was the consistent signal. Not dusk. Cooling.
Hold on. That flips the whole conversation.
A bit, yes. And the Hadza data says the light story is incomplete. In a heated flat at a five p.m. sunset, the cue that's actually missing might not be darkness. It might be evening cooling. The sun goes down, the air cools, the body reads that as bedtime. We've thermostatted that signal out of existence.
You can test that at home for free. Crack the window an hour before bed in January and see what your body does.
I'd say consult someone, but I'm the retired physician in the room, so: a crack, not an open pane, and Ezra's room stays shut.
Noted. What about napping? I feel like everyone assumes these societies siesta their way to a total.
Surprisingly rare. Under seven percent of winter days, under twenty-two percent of summer days. The separate Hadza study — Samson's team, thirty-three people, nearly four hundred days — found about six and a half hours at night plus naps averaging just under fifty minutes when they happened at all. And there's seasonality: Tsimane recorded in winter slept fifty-six minutes longer than in summer. Winter sleep is real, even at the equator-adjacent latitudes.
Fifty-six minutes is one long movie trailer less of sleep.
One More, actually, at winter Tsimane latitudes. Now the part I want your read on, because you live closer to the Hadza schedule than I do — the insomnia finding. Samson's quote is blunt: when they asked the Hadza whether their sleep was good, they almost universally said yes, it's totally fine — and that statistically doesn't match up with what we see in the West.
I sleep about six and a half hours and I'd tell you the same.
Would you, though? Honestly?
I'd tell you it's fine and then describe my night in enough detail to make it clear it was not fine. But that's me. The Hadza answer, I take at face value.
The interesting mechanism there — and I'm reasoning, not citing, this part is my read — is that insomnia complaints in the West are often about the mismatch. You're trying to sleep when your biology says not yet, and you're trying to wake when your biology says not yet, twice a day, every weekday, and we call the result a sleep disorder. If your sleep is anchored to your own physiology, being a bit short of hours might not register as suffering at all.
Which would mean half the sleep-aid aisle is treating a scheduling conflict.
That's more than half, I'd guess. Before we leave the traditional societies — the Amish, because they're the version of this living inside a rich industrialized country, which is the only version a listener could actually imitate. Lee and colleagues, twenty twenty, Lancaster County, Pennsylvania, thirty-three Old Order Amish adults with activity monitors. They woke before civil sunrise and slept after sunset regardless of season. Activity started before six-thirty in everyone but one participant.
And the comparison numbers against the rest of America?
Questionnaire data, thousands on each side — the Amish went to bed about eighty minutes earlier and got up about eighty-eight minutes earlier than the non-Amish comparison group. That's not a lifestyle tweak, that's a different time zone of living.
Eighty-eight minutes earlier. Ezra would defect to the Amish by Thursday.
He'd defect on the technology question alone. But here's the caveat I refuse to let get sanded off: the Amish are not light-free. The study's own conclusion is that waking before sunrise and sleeping after sunset irrespective of season implies artificial light exposure at all day lengths. They have lamps. Gas lamps, lanterns. What they don't have is the will to sit under them until one a.m. They're schedule-disciplined, not light-deprived.
That's the most deflating finding in the whole file, in a way. You can replicate the Amish sleep schedule tonight with a lamp and a spine.
There's even a heritability study — seven hundred Amish adults, and wake time showed meaningful heritability, around a fifth of the variation. Some of them are early risers by genes, not virtue. That should console every listener who's tried and failed to become a morning person.
It consoles me, and I was already consoled.
So that's the baseline: nobody goes to bed at sunset, they stay up around fires, they sleep less than we assume, and they feel fine about it. But that's observation of people who never had the alternative. The real test of Daniel's question is what happens when you take a modern, electrically-lit person and pull the plug. That's Wright's lab.
Ken Wright, Colorado Boulder. Walk me through the arc — he did this three times, didn't he?
Three times, and the sequence tells a story by itself. Summer twenty thirteen first: volunteers camped for a week, got roughly four times more light by day than in their normal lives, and — this is the design detail I love — no headlamps, no flashlights at night. Only firelight. On return, their melatonin onset had shifted nearly two hours earlier, landing close to sunset.
Two hours from one week.
One week of no electric light. Then the summer weekend version, twenty seventeen: fourteen volunteers, nine of them camping for a single summer weekend, five staying home as controls. Two days. That's it. The campers' melatonin rise shifted one point four hours earlier. Wright's line on it: a weekend of natural light got them sixty-nine percent of the shift they'd previously measured after a full week.
Sixty-nine percent from a weekend. That's the figure everyone's going to quote at dinner parties.
It's the most quotable and the most fragile number in the file, and we'll come back to why. But the winter study is the one that actually matches Daniel's premise — his prompt is a five p.m. sunset, that's a December question. Five volunteers, one week camping in Colorado near the solstice. By day they were exposed to thirteen times as much light as on their normal winter weekdays.
Thirteen times. In winter. From being outside in daylight instead of in an office.
Daylight in winter is weak per hour but the dose over a whole day outdoors is enormous compared to a life of windows. And the results: melatonin onset two point six hours earlier, sleep start two and a half hours earlier, and sleep duration two point three hours longer — nine point nine hours against seven point six at home.
Nearly ten hours. The winter camping trip is the best sleep of anyone's life.
The finding Wright highlighted, though, is subtler than the total. Their biological night lengthened to match the season — as it does in many animals. His phrase was that this had been assumed but never demonstrated. Humans living in electric light keep the same internal night length year-round; put them outside in December and their physiology stretches like a deer's.
We're the only animal that central-heats its own circadian system.
The control group is the quiet punchline of that study. The five who stayed home for the weekend did what modern people do — stayed up later, slept in later — which is exactly the pattern that produces Monday grogginess, what researchers call social jetlag. The campers, on a regular schedule, sidestepped it entirely. They fixed their Monday by leaving town on Friday.
The cure for the Monday is the Saturday.
And the fragility, because this is where I stop being impressed and start being honest. Wright says it plainly: people have to keep a regular wake and sleep cycle to retain the circadian reset they've gained. The weekend gets you sixty-nine percent of the way. Monday through Thursday of ordinary indoor living gives it back. This is not conversion, it's maintenance. It's less a cure than a subscription.
You pay in weekends, forever.
And the subscription model is what nobody selling sunrise lamps wants printed on the box. Now — Corn, you've done the overnight-awake thing for years, you're up while the whole flat sleeps. When you hear ten hours of winter-camp sleep, do you buy the mood claims that get attached to it? Because that's Daniel's other thread — what happens to mood and daily life.
I buy the direction, I'd want to check the size of it. Being awake at four a.m. is only pleasant when nothing is required of you afterward.
That's the right skepticism, and the honest answer is that the mood evidence is thinner than the timing evidence. The mechanism is solid — light does two separate things. It entrains the clock, morning light pushing it earlier, evening light pushing it later. And separately, it has an acute alerting effect. Dijk compares it to drinking a double espresso, and light also stimulates brain regions that regulate mood. Two jobs, one input.
Morning light is an espresso and evening light is a delay pedal.
That's better phrasing than the paper managed. Dijk's summary of the modern problem: in modern societies, at least on weekdays, we don't sleep in tune with our body clock. Evening light drags the clock later, and then the alarm fires while the biology still says night. The lamp and the alarm are playing tug-of-war with the same person.
And the person loses, twice a day, five days a week.
The measured mood data — the closest anyone's come is Linda Geddes, the BBC Future experiment from twenty eighteen. She ran three weeks: one maximizing daylight, one minimizing evening light after six p.m. — candlelight, dim red light — and one combining both, with baseline weeks between. Surrey's chronobiology team measured her melatonin and sleep properly, this wasn't a vibes diary.
What moved?
Average bedtime twenty-three hundred on the intervention weeks against twenty-three thirty-five at baseline. Melatonin released about an hour and a half earlier from the daylight week, two hours earlier when she avoided evening light. And her early-morning mood was significantly more positive. That last one is an n of one, and I want to keep it labeled as such — no study I know of has measured mood in people living solar-synced long-term. She's the datum.
One well-instrumented woman in Bristol is carrying the entire mood claim.
Carrying it cheerfully, apparently, which is at least on-theme. What I find more telling is her account of what was hard. Because Daniel asked exactly that — what did people find hardest — and Geddes is unambiguous. Not the darkness. Getting outside in winter. Her words: it's hard to overcome the belief that because it's winter, it will be cold and miserable outside.
The hard part was the daylight side.
She borrowed a Swedish friend's line — there's no such thing as bad weather, only inappropriate clothing — and by the end reported that the more she did it, the more getting outdoors in winter felt like a treat rather than a chore. But she had to be dragged to that conclusion. The darkness was easy. The lunchtime walk was the battle.
Which makes the whole thing a plot twist. Everyone buys the candles and skips the coat.
The month-long self-experimenters all hit the same wall, from different directions. One writer had, in her phrase, some of the best nights of sleep — and also fitful ones; it wasn't monotonic — and found shutting off her phone at eight helped. Another had to reorganize her days so cooking happened in daylight, because preparing dinner by candlelight turned out to be impractical. The friction was logistics, not sleep. A third found herself falling asleep at nine forty-five by week two, waking without an alarm — earlier, rested, and slower in the best way. And the family who did it together concluded that going without light causes you to slow down, and that they'd assumed they'd be able to do everything normally.
They could not do everything normally, news at eleven. Or rather, at five.
The pattern across all of them: sleep sorts itself out fast. Mood seems to follow. What doesn't sort itself out is the shape of the day — cooking, reading, working, walking the dog, all of it has to be re-liturgied around where the light actually is. The body adapts in a weekend. The schedule takes a month, and most people don't have a schedule to spare.
Daniel's prompt also asked about adventurers, expeditions, polar parties — weeks or months out. What did you find there?
Honestly, not what the question implies. I went looking for a measured account of a polar expedition or a long crossing without artificial light — sleep diaries, mood data, anything systematic — and it isn't there. The expedition literature is travel writing. The closest rigorous things we have are Wright's winter week and Geddes' weeks at home. Multi-week data comes from self-experimenters, and multi-month data doesn't exist in measured form at all.
So the romantic version — the explorer at one with the polar night — is unmeasured.
Unmeasured and possibly unexperienced, because real expeditions burn fuel for light on purpose. The people with data are a woman in Bristol and five Coloradans in December. I find that gap almost more interesting than a study would be. The best-documented method for going dark is also the most domestic one.
The frontier is a kitchen with the lights off.
Which is what the whole file keeps saying, if you stack it up. The exotic version of this life doesn't exist in the data. The ordinary version does, and it works.
Hilbert's Mid-Show Haiku. Sleeping by the sun. The fire dies down to coals. Someone turns, and turns. Morning arrives without permission. The kettle is cold.
Five lines. He'll argue it's a regulation haiku if asked.
He'll have a citation too, from some manual nobody's heard of. So — the contradiction I promised, because it's a real one and it sits right in this episode's machinery. Ekirch's segmented sleep thesis. Roger Ekirch, historian, two thousand five, At Day's Close — the argument that pre-industrial Europeans slept in two chunks, a first sleep and a second sleep with a wakeful hour or two between, and that our modern single block is the aberration.
The midnight watch.
We did, and I want to say the follow-up out loud because it's uncomfortable for the thesis: the actigraphy fieldwork does not confirm it. Yetish's data from three living pre-industrial societies found consolidated sleep, confined to the latter portion of the dark period. One block, not two. These are people with no electric light, no clocks, no reason to sleep in the shape we do — and they sleep in the shape we do, just timed differently.
So the diaries and court records Ekirch worked from say two sleeps, and the wrist monitors on actual sun-living people say one.
That's the contradiction, stated flat. Possible resolutions — and I'm speculating now — the historical populations Ekirch studied were Europeans in cold, dark winters with long nights; the Hadza and Tsimane latitudes are different, the nights are shorter. A twelve-hour subtropical night may not leave room for a middle waking the way a sixteen-hour northern one does. Or the segmented pattern was cultural, northern-European, particular, and not the human default. But the honest position is that it's unresolved, and anyone telling you first sleep, second sleep is how humans naturally do it is going past the direct measurement.
The direct measurement, it should be said, of societies that aren't timekeeping ones. People wake at night and don't file a report about it.
True, and actigraphy scoring a wakeful hour in the middle as one thing versus two is its own argument. I hold it more loosely than I did before reading the fieldwork, put it that way.
Where does that leave Daniel's actual question, then? He wanted to know what happens when you live by the sun. Stack it up.
You sleep about six and a half hours, later than the sunset you're romancing, and you don't miss anything. Your clock stops fighting your alarm. If you're winter camping, you sleep nearly ten hours and your internal night stretches to fit the season like every other mammal's. Your mornings, on the one measured account, feel better. And the price is paid in daylight logistics, not darkness — you have to be outside at midday, in December, in a coat, whether you feel like it or not.
I did that. Sort of. One winter, the flat's power went out and I didn't get it fixed. Corn, I heard your bit about the five a.m. kettle being cold, and it wasn't — I heated it on the gas ring. Anyway. Six weeks, no electric light, not as an experiment. I just wasn't paying for a callout for a fuse box I could see perfectly well from the floor.
You stayed in the dark six weeks over a fuse.
I had candles. A lot of candles. And it went how your studies say. I didn't go to bed at sunset, that's nonsense, I sat up reading till half ten, eleven, by candlelight. Slept six hours, six and a half, felt fine. The hard part wasn't the dark either. It was the football. My brother used to tape the second half for me at his place, and I'd go round Sunday and watch it, then walk home in the dark on purpose, which I maintained then and maintain now was the correct order to do it in. I kept a log of the candles. Forty-seven, for the month I read a three-hundred-page novel. That's a real figure, I still know it. People say candles are cheap. Per novel, they're not.
Forty-seven candles is a book with a cover price.
The box said forty per pack, so I was one into the next box, which is the kind of thing that stays with a man. I've still got the stub of the last one, in the drawer by the mixing desk. It proves something about that winter, and when I remember what, I'll say.
The drawer by the desk. That's where the good fuses live.
Among other things. Anyway, Herman, you said nobody's measured the mood part. I'll tell you the mood part for free: I was furious about the fuse box, and after week two I wasn't furious about anything. Read into that what you like.
Six weeks, six and a half hours, no complaints. He's the Hadza of the second floor.
With a candle ledger. The one thing I want to keep from what he said is the walking home in the dark on purpose — that's a man who found the shape of the day and kept it.
And the forty-seven. That's a real unit of reading nobody has ever measured. Right — the "one thing," as we do. If you take one thing from this episode: it's the sixty-nine percent figure with its catch. A weekend outdoors moves your clock most of the way back to solar time — measured, replicated, real. And you hand it back within days of ordinary indoor living unless the schedule stays regular. Conversion is cheap. Staying converted is the whole job.
And the radical act was never the five p.m. bedtime. It's getting up before dawn and getting outside at noon in December — the daylight side, not the candle side. That's the part every account says was hardest, and the part every listener will skip.
The morning walk is the cheapest experiment in the whole file, and the most avoided. Even Hilbert walked home in the dark rather than out in the light — habit beats knowledge every time.
He'd say the dark walk was the correct order.
He would, and I've stopped betting against him. Thanks as always to Hilbert Flumingtop at the desk, producing this thing we do. If you want more of this, try episode thirty-two, When Night Vanished; episode six sixteen, The Midnight Myth; and episode fifteen fourteen, The Midnight Watch. This has been My Weird Prompts, with me, Herman Poppleberry, and Corn, who was awake for all of it, as usual. If the show earns it, leave us a review — it's how other curious people find the rabbit hole. We'll be back soon.
See you at sunset. Or, per the data, four hours after.