00:00
Hey there, late-night wanderers. Tonight, one question. Why did our ancestors start believing in gods? And I want you to think about it the way an accountant would, because religion was the most expensive habit our species ever picked up. 34,000 years ago at Sungir in Russia, two children were buried wearing 13,000 beads carved from mammoth ivory by hand. Thousands of work hours put straight into the dirt and that is the cheap example. At Göbekli Tepe people with no metal and no wheel raised stone pillars weighing ten tons.
00:32
At Atapuerca someone dropped a brand new hand axe down a deep shaft without ever using it. Ancient humans basically invented the gift, then threw it away.
00:41
Evolution punishes waste like that, so belief had to be paying something back. So tonight we find out what. When this switched on, what in the brain produces it, and what it ended up building? And we start at the oldest place it could have begun, which is not a temple. It is a campfire. Hit that like button if you are into this kind of deep dive, and drop a comment telling me what city you're watching from and what time it is where you are. Now dim the lights and settle in. So let us go sit by that fire, because the fire is not just a cosy image to open with.
01:13
It is a real archaeological object, with a real address, a real date, and a real layer of ash you can hold in your hand. And it is the first moment in the human story where the daily schedule changed enough to leave room for something as gloriously impractical as a god.
01:29
Start with the obvious question, when did people actually get control of fire? Not stumble across a burning tree and back away politely, but keep flame, feed it, move it, own it.
01:41
The oldest serious candidates are two sites on different continents. At Wandavert Cave in South Africa, researchers found burnt bone fragments and ashed plant material roughly 30 metres inside the cave in layers close to a million years old. The depth matters. A lightning strike does not walk 30 metres into a hillside and set up shop in the back room. The other candidate is Gesher Benot Yaakov in northern Israel, dated to around three-quarters of a million years, where excavators mapped tiny burnt flint chips that cluster in the same spots across multiple occupation layers.
02:15
Nobody found a neat stone ring with a kettle hanging over it. What they found were statistical shadows of fire in the same corners, again and again, as if someone kept coming back to the same place to do the same thing.
02:27
Archaeologists sometimes call these phantom hearths, which is a wonderfully dramatic name for what is essentially a pattern of burnt crumbs. Here is where you need to be a little careful, and being careful is going to be the theme of this entire video. Evidence that fire happened is not the same as evidence that fire was managed. Grass fires happen. Bats leave guano that chemically mimics heat alteration. Sediments get cooked by processes that have nothing to do with anyone deliberately doing anything. For a long stretch of our past, fire use looks occasional, opportunistic and frankly a bit amateur, which is an entirely reasonable way to treat a substance that can kill everyone you love while you sleep.
03:06
The thing that changes the picture is not the earliest fire, it is the habitual one.
03:11
At Kizem Cave in Israel, about 300,000 years ago, there is a hearth that behaves like furniture. It is a defined patch of roughly four square metres near the middle of the cave, built up from layers of wood ash accumulated over generations, with charred bone all through it.
03:28
Around it the debris sorts itself by activity. In one zone people cut meat. In another they made small blades. The hearth is not a spot where a fire once happened. It is the centre of a room and the room has a floor plan. When archaeologists talk about the domestication of fire this is what they mean in practice. Not the first spark but the first permanent address. And once you have a permanent address with a permanent flame at the centre you have accidentally invented something no other animal on the planet had. You have invented the evening.
03:59
Before we get to the evening, it is worth asking why anyone would maintain a fire at all, because keeping one alive is genuinely annoying. Fuel has to be collected constantly. The fire has to be watched, fed, shielded from wind and rain, and somebody has to be responsible for it overnight. In exchange, you get warmth, you get a predator deterrent and you get the ability to cook, which is the benefit that most researchers think paid for the rest. Cooked food is softer, safer and easier to digest, and it releases more usable energy per bite than the raw version.
04:32
The biological anthropologist Richard Wrangham built an entire argument around this, proposing that cooking is what allowed our lineage to run a smaller digestive system and a much larger and more expensive brain at the same time, since you cannot afford both on a diet of raw fibrous plants and uncooked meat.
04:49
The chronology is argued over fiercely because the anatomical changes he points to are older than the confident fire evidence and a hypothesis that requires fire half a million years before we can prove fire is a hypothesis with a scheduling problem.
05:02
but whatever the exact timing, the general shape holds. Fire was not adopted as a mood enhancer. It was adopted because it made calories cheaper and everything else it did to us came along as a bonus, and the bonus turned out to be the interesting part. Think about what the night actually was before this. the sun sets, and your visual system, which is a daylight instrument with a rather poor night mode, drops out of usefulness. You cannot track game, you cannot spot roots, seeds, tubers, or that one specific shrub you walked 4km to find.
05:34
Meanwhile, a variety of large animals with far better night vision are just getting warmed up for their shift. The only rational strategy is to stop existing for 11 or 12 hours, find a tree, find a ledge, find a cave, become very quiet and wait for the universe to turn the lights back on. Fire deleted that. Not all of it, but a solid four to five hours at the front end of the night.
05:57
And the crucial detail, the detail everything in this chapter hangs on, is what kind of hours they were. They were hours you could not monetise. You could not hunt in them. You could not gather in them. Long distance travel was off the table. Even the domestic chores were limited because detailed handiwork by flickering orange light is a good way to ruin a tool and stab yourself.
06:18
What fire produced was not more productive time, it produced the exact opposite. It produced a large, recurring, unavoidable block of time in which a group of people were stuck together, awake, safe, warm, facing the same direction, with their hands free and nothing pressing to do.
06:35
If you want to know where culture comes from, that is the slot.
06:39
Idle time is not a luxury that shows up after civilization. Idle time is the raw material that civilization is made out of. Now, you cannot excavate a conversation. Ash preserves, stories do not. So how do we get from a heated patch of cave floor to any claim about what people were saying around it?
06:57
This is where one of the most quietly useful studies in the whole field comes in. The anthropologist Polly Wiesner spent decades working with the Yuhuansi of the Kalahari and she had something unusual in her files – recordings and transcriptions of ordinary conversation collected across years in a society where firelight evenings were still a normal part of life rather than a camping novelty. In 2014, she published an analysis comparing what people talked about during the day with what they talked about at night. The results are almost too clean.
07:28
During daylight hours, conversation is essentially an ongoing management meeting. Roughly a third of it is complaining, criticism and conflict. Who took what? Who is not pulling their weight? Who behaved badly at the waterhole? The rest is logistics, land, planning, gossip about people who are standing 40 metres away and jokes, mostly at the expense of the people standing 40 metres away. It is the sound of a small society doing the administrative work of not falling apart. Charming, necessary, and about as spiritually elevated as a group chat about a shared kitchen.
08:02
At night, around the fire, the content flips. About 80% of recorded conversation becomes story. Not complaints, not scheduling. Narrative. Long accounts of hunts that happened years ago. Travel to places most listeners will never visit. Genealogies that reach back through relatives nobody present has met, Songs, healing dances, accounts of spirits and of the dead, told by people who take them seriously. Critically, the criticism almost vanishes. Grievance is a daytime activity.
08:34
Nobody wants to relitigate a meat-sharing dispute while someone is three minutes into a good story.
08:39
There is a second pattern in that material that does not get quoted nearly enough, and it concerns who does the talking.
08:46
During the day, the loudest voices belong to people in their productive prime, because daytime conversation is about current business, and current business belongs to whoever is currently doing it.
08:57
At night, the floor shifts toward the old. The best storytellers in Riesner's recordings tend to be elders, and being a recognised storyteller is a genuine social position, not a consolation prize handed out to people who can no longer walk far, which quietly solves an economic problem that every foraging society has. A person in their 60s, in a world without agriculture, is a net cost in pure calorie terms. They bring in less than they consume. They cannot run down an animal or carry a heavy load for a day and a half, but they can carry the only copy of 40 years of accumulated information.
09:31
where the water was during the drought when everyone else was a child, which plant looks edible and is not, what the group did the last time a neighbouring band showed up in force. In a world with no writing, an old person is a library and the fire is the reading room. Nighttime storytelling is the mechanism that converts a biological liability into an indispensable asset, which is a rather elegant trick for a species to pull off without anybody designing it. It also means the authority in the circle belongs to the person with the longest memory, and authority based on memory is going to matter enormously later when the group starts needing somebody to speak for people who cannot speak for themselves anymore.
10:11
That flip is not mystical, it is mostly physics and a bit of social geometry. Firelight gives you a small bright circle inside an enormous dark nothing, which collapses your visual world down to a few metres. Your peripheral vision, which spends the whole day scanning for movement, has nothing to scan. Everyone is oriented inward, toward the same centre, in a rough ring, which means everyone is looking at roughly the same thing at the same time. Faces are lit softly from below and harsh expressions get gentler, which sounds like a small thing but genuinely changes how a group reads one another.
10:45
Bodies sit closer than they would at noon, because the warm zone is only so wide. And nobody can storm off mid-argument, because out there past the light is a temperature drop and a hyena with strong opinions. So you have a group with synchronized attention, reduced social distance, no immediate tasks and no escape route. That is not a campsite. That is a theatre, a church pew and a lecture hall with the same seating chart, several hundred thousand years before anyone builds any of them. Wiesner also made a point that goes beyond entertainment.
11:16
Those night stories were doing work. They carried geography, telling people what the land looks like eight days' walk away. They carried lore, in the form of characters who behave badly and characters who behave well. They carried kinship, which in a foraging society is not sentimental. It is your access network for water rights and refuge when your own territory fails. Daytime conversation manages the people present. Nighttime conversation manages the people absent, the people distant, and the people dead.
11:43
It extends the social world beyond the group you can see, and sit with that phrase for a second, because it is the hinge of this entire video.
11:51
Around a fire, humans developed the habit of talking, at length, with total seriousness, about agents who are not physically present. Ancestors, relatives across the desert, a hunter who died before you were born, beings nobody in the circle has ever met.
12:08
Every night for hours the group practices holding in mind a character who is not there, agreeing on what that character is like and caring about it emotionally. That is not religion yet, but it is the exact cognitive muscle religion runs on, getting trained in the dark, for free, as a side effect of having nothing else to do.
12:28
Now the honest caveat, because this video is going to keep handing you the reasons not to believe it. The Zhuhuanci are not a time machine, and they are not our ancestors preserved in amber. They are a modern people with their own long history, their own recent past of displacement and change, and their own contemporary lives.
12:45
Using their evenings as a window into 300,000 years ago is an analogy, not a measurement, and anthropologists have spent decades rightly beating up on the older habit of treating living foragers as museum exhibits.
12:58
What the study actually demonstrates is narrower and still very useful, that firelight reliably changes the content of human conversation in a measurable direction, away from grievance and toward narrative. It tells us what the medium does to the message. Whether people 300,000 years ago used that medium the way people did in the 20th century is an inference, and inferences are exactly what get overturned in the second half of this video. There is another reason to take the night seriously, and this one does survive in the ground.
13:29
Somewhere in that dark, people started making sound on purpose. In the caves of the Swabian Jura in southern Germany, in layers dated between roughly 42,000 and 43,000 years ago, excavators have recovered bone flutes, not scratched bone that might be a flute. Flutes, the clearest one from Holofells, is made from the radius of a griffon vulture, a bird whose wing bones are long, straight, thin-walled and naturally hollow, which is to say that somebody looked at the local wildlife and correctly identified the best available tubing supplier.
14:01
The surviving piece is about 22cm long, a bit under a centimetre across, and it carries five finger holes.
14:09
At the mouth end there are two carefully cut V-shaped notches where the player set their lips. The holes are not randomly punched, they are placed and the fine scraping marks around them show someone adjusting, testing and adjusting again.
14:23
That is already a professional object, but the nearby site of Geissen-Klöstele produced something that borders on absurd which is a flute made of mammoth ivory. ivory is not hollow. To build this, a craftsperson has to take a solid curved tusk section, split it lengthwise into two matched halves, hollow each half out along its length without breaking through the wall, carve the finger holes to match, then glue the two halves back together with a seam, airtight enough to hold a column of vibrating air.
14:52
If that seam leaks anywhere along its length, you have produced a very expensive stick. There is no practical reason to do this when the local sky is full of birds with perfectly good hollow bones. Somebody wanted an ivory one. Replicas of these instruments have been built and played and they work. They produce a clear, strong tone with a usable range of notes in intervals that a modern listener recognises as musical rather than as noise. Which brings up the point that actually matters here. A whistle is a signal. One pitch, one meaning, come here or watch out, and plenty of animals manage something like it.
15:27
An instrument with five tuned holes is not a signal. It is a device for producing a sequence that other people can learn, remember, reproduce and join.
15:38
It presumes an agreed system of pitches. It presumes repetition. It presumes an audience who will notice if you get it wrong. And music, unlike speech, does something specific to a group. It locks people onto a shared clock.
15:52
Voices align, bodies start moving in time, breathing falls into pattern. Try to imagine somebody playing a five-hole flute alone in the dark for their own private satisfaction. You can, but it is a strange picture, and it is not the picture the rest of the evidence supports. These caves were occupied. These were things played at people, among people, in the same season, probably in the same firelight we have been discussing. The flute is the first hard evidence of coordinated performance. which is a polite way of saying the first evidence of a group doing the same non-productive thing together on purpose.
16:27
Hold on to that, because coordinated non-productive activity is going to show up later in this story wearing much heavier costumes. It is also worth noticing what else was coming out of those same German caves at the same moment.
16:39
The Swabian Jura in this period was producing carved figurines in mammoth ivory, animals rendered with real anatomical attention and at least one object that does not correspond to anything you could walk outside and photograph. We will come back to that later because it belongs to a different part of the argument. The relevant point here is that the flutes are not an isolated eccentricity. They sit inside a cluster of activity in one small region across a few thousand years in which people poured skilled labour into objects with no nutritional return whatsoever.
17:11
A single strange artefact is an anecdote. A whole valley of them, layer after layer, is a culture with a budget line for the intangible. And the reason coordination matters is worth stating plainly because it sounds like a soft point and it is actually a hard one. A group that can move and sound in time with one another can do things a collection of individuals cannot. They can align labour. They can demonstrate to themselves and to anybody watching from the tree line that they are a single unit with shared rehearsal behind them.
17:42
Rehearsal is the operative word because you cannot fake having practice together. Whatever else a five-hole flute was for, it created situations where a group had to agree on a pattern and then execute it as one, and every one of those situations is a small proof of alignment. Keep that filed away.
17:59
By the end of this story, the question of how a group proves it is genuinely a group turns out to be the question the whole subject is built on. And now, the cautionary tale, which is the most important paragraph in this chapter.
18:13
In 1995 at Divya Babe in Slovenia, excavators pulled a fragment of a juvenile cave bear femur out of a layer associated with Neanderthals. The fragment had holes in it, roughly aligned holes. The announcement was irresistible. The oldest musical instrument in the world, 60,000 years old, played by Neanderthals, rewriting everything.
18:35
It appeared in books, documentaries, and an enormous number of confident internet posts. People built replicas and played them on television. Then the specialists went to work, and the specialists were not in a hurry to be charmed. Analysts examined the edges of the holes, compared them with bone assemblages chewed by carnivores, and found a much duller explanation. Hyenas and other scavengers crack open juvenile long bones for marrow, and their carnassial teeth punch through the thin walls in pairs, producing holes with a characteristic shape, spacing and internal damage.
19:08
There are also comparable cave bear bones from nearby sites, unambiguously chewed with unambiguously similar perforations.
19:17
The debate has never fully closed, and some researchers still argue for the flute, but the weight of opinion landed on the boring side.
19:24
The oldest musical instrument in the world is most probably someone's lunch. Take the lesson, because you will need it repeatedly. When we look at very old objects we are motivated observers. We want the past to be profound. We would rather find a Neanderthal artist than a hungry hyena, and that preference quietly shapes which explanation feels convincing. Bone does not care what we want. Neither does ash, ochre or stone. The discipline of this subject is the discipline of asking what else could have produced this, and then actually doing the work to rule it out, even when the boring answer is unwelcome.
19:59
So the fire chapter closes with a claim that is deliberately modest. Nobody invented gods around a hearth at Kesham Cave 300,000 years ago. What happened is more interesting than that and considerably better supported. Fire produced a recurring block of protected, useless shared time, that time filled with narrative because narrative is what the conditions favour. Narrative trained people to hold absent agents in mind and to agree about them, and somewhere along the line the group added sound that synchronised bodies as well as attention.
20:30
Myth is not a revelation that descended on our ancestors. Myth is a night shift technology, produced by the first species that ever had an evening to kill. But stories have a serious engineering flaw. They exist only while someone is speaking them. The moment the storyteller dies, the story goes with them, unless somebody happens to have been listening carefully and cares enough to repeat it.
20:52
If you want a shared idea to survive a bad winter, a migration or a generation gap, you eventually need to move it out of the air and into an object, which is how the first sacred substance in human history turned out to be dirt, red dirt specifically. Ochre is not a mineral in the strict sense so much as a family of earths loaded with iron oxides, mainly hematite and gothite, and it ranges from pale yellow through orange to a deep rusty red.
21:19
You find it in weathered soils and iron-rich rock all over the world. It is, physically speaking, extremely ordinary. It is also the first substance that humans consistently went out of their way to obtain, process, modify, transport and store, and it shows up in the archaeological record long before anyone was burying their dead with grave goods. That matters for the structure of the argument. If you want to know when humans started thinking symbolically, the tempting evidence is the dramatic stuff. the graves and the art, but the pigment gets there first by a wide margin and it gets there in a form that is hard to explain away.
21:57
The early end of the record sits at Pinnacle Point on the southern coast of South Africa, in a cave complex overlooking the Indian Ocean, in deposits dated to around 164,000 years ago.
22:09
Along with the earlier solid evidence for people systematically eating shellfish and making small bladelets, excavators recovered dozens of pieces of ochre showing clear signs of use, scraped surfaces, ground facets, and a preference, visible in the assemblage, for the reddest and most brilliant pieces rather than whatever happened to be nearest. Somebody was choosing. In a landscape full of coloured earth, they walked past the dull stuff. Pinnacle Point is the point where the evidence becomes hard to argue with, but it is not the beginning of the story.
22:40
There are older pigment traces scattered through the African record, including collected and apparently ground material from sites in Zambia and Kenya that push back considerably further, into a period when the anatomical distinction between us and our nearest relatives is still being negotiated. The trouble is that the deeper you go, the thinner the evidence gets and the louder the alternative explanations become.
23:03
Iron oxides occur naturally in sediment. Pieces get scuffed by geology. A fragment with a worn face in a very old layer is suggestive and nothing more. So the honest summary is that pigment interest may be very old indeed, but the point where it becomes undeniable, repeated and clearly selective is the Southern African Middle Stone Age. and since fairness demands it, our species was not alone in this. Neanderthal sites in Europe have produced collected pigment as well, including black manganese dioxide at a French site where the material was ground and shaped into what looked like crayon points.
23:38
It was immediately read as body decoration, which is the natural assumption.
23:42
Then researchers ran experiments and discovered something inconvenient. powdered manganese dioxide lowers the ignition temperature of wood shavings and dramatically improves the performance of a fire-starting kit, so the black crayon that everybody wanted to be makeup might be a component of a lighter. Both interpretations are still on the table and the episode is a useful reminder that our ancestors and their cousins were perfectly capable of having chemistry without having ceremony. Then there is Blombos Cave, a few hundred kilometres along the same coast, which has produced the single richest set of early symbolic material anywhere.
24:17
In layers around 77,000 years old, excavators found blocks of ochre whose surfaces had been ground flat and then deliberately engraved with cross-hatched geometric designs, a lattice of intersecting lines framed by a border.
24:32
One such block could be a bored person with a sharp flake. Blombos has produced more than a dozen engraved pieces, across multiple layers, with recurring structure. The design has a convention. Somebody learned it from somebody. And that is the difference between a doodle and a sign. A doodle happens once, a sign gets copied. The same site also gave up a silcrete flake bearing a set of deliberate lines drawn onto it with an ochre crayon, which is about as close as we get to catching a hand in the act. But the fine that should genuinely stop you is the workshop.
25:06
In 2011, a team led by Christopher Henshelwood published a description of two separate processing kits from Blombos, dated to around 100,000 years ago. Each kit is built around an abalone shell used as a container, inside a residue of ground ochre. Nearby, the quartzite hammerstones and grinders used to crush it, still carrying red powder in their pits.
25:28
Mixed into the residue, the chemical signatures have crushed bone, charcoal and fat, along with a small bone used as a stirrer and a vertebra that seems to have served as a stopper or a small mixing vessel. That is not a person rubbing a rock on their arm. That is a recipe. It has ingredients, a grinding stage, a binder, a mixing step and a storage vessel. Fat extends the mixture and helps it stick. Charcoal alters the shade and the working properties. The shell was chosen because it is a natural bowl that will not leak.
26:00
And the whole thing was prepared and then left, which is the detail that does the real work. Nobody prepares a batch of anything for storage unless they anticipate a future occasion for it. This is planning, it is chemistry by trial and error, and it is a division of labour between the people who fetch the ingredients, the people who grind, and the people who mix.
26:21
Around 100,000 years ago, somebody on that coast was in the paint business. The scale of supply makes the same point from the other direction.
26:29
At Lian Cavern at Nguenya in Eswatini, people did not simply pick up ochre where it was lying around. They followed a seam of iron-rich rock into a hillside and removed it systematically with stone tools over an enormous span of time. the excavated volume runs to a genuinely large tonnage of rock moved and the dates reach back tens of thousands of years, placing the operation deep in the Stone Age rather than in any later industrial context. This is usually described correctly as the oldest known mine on the planet, which produces one of my favourite facts in prehistory.
27:02
The first thing our species ever dug a mine for was not copper, not tin, not flint, not salt, it was colour.
27:09
We were mining pigment for tens of thousands of years before it occurred to anybody to mine metal. Priorities, unsurprisingly, were different back then. There is one more dimension to the supply picture, and it is the movement. Ochre travels. Pieces turn up in sights a long way from any source that could have produced them, sometimes tens of kilometres, occasionally much further, in an era when the only available transport was a person and their legs.
27:34
Carrying rock across a landscape is a real cost, measured in calories you do not get back, and nobody hauls a heavy dull-coloured stone for two days unless the stone has a value at the other end. That value implies either a strong personal need or an exchange. An exchange implies that a substance with no food value had become a thing worth having in its own right. Somewhere in the middle stone age a lump of coloured earth became an asset, which makes ochre a plausible candidate for the first commodity our species ever recognised.
28:02
The habit is also remarkably persistent, which is its own kind of evidence. Sites across Africa keep producing ochre assemblages over tens of thousands of years, some of them enormous, running to thousands of processed pieces accumulated in a single cave over long occupation.
28:19
This was not a fashion that flared up in one clever community and died out.
28:23
It is a continuous, geographically vast, multi-hundred-thousand-year commitment to a substance whose reliability as a food source is precisely zero.
28:33
And then the pigment shows up on bodies. At Kafza Cave in Israel, in layers dated to somewhere between 92,000 and 100,000 years ago, excavators recovered a large assemblage of ochre pieces, many of them heat-treated, which means people were deliberately roasting yellowish guthite to convert it into red hematite. That is not a trivial insight. Somebody figured out that a specific earth, heated in a specific way, changes colour permanently and then made a practice of it.
29:01
Ochre at Kafzeh appears both in the living area and in association with burials, and among those burials is an adolescent, around 13 years old, laid in a pit with the antlers of a fallow deer placed on the chest, the arms arranged around them.
29:15
The site also produced marine shells carried in from the coast, some of them pierced, some stained red. So far this all sounds like a fairly tidy case for ancient ritual. Here is where we slow down and do the unglamorous part, because the practical explanations for ochre are genuinely strong and archaeologists spent decades ignoring them. Ochre has an irritating number of legitimate uses. Smeared on skin, it reflects and absorbs ultraviolet light, functioning as a rough sunscreen, which is not a small benefit for a species running around equatorial Africa without clothing.
29:48
It has some insect-repellent properties. Mixed into hide processing, iron oxides act as a preservative and help prevent bacterial degradation of skins, which extends the life of leather substantially. And the best documented use is adhesive. When you are attaching a stone point to a wooden shaft, you make a mastic from plant resin, and loading that resin with ochre powder changes its mechanical behaviour, making the bond less brittle and more durable.
30:14
Experimental work has shown measurable improvement in hafted tool performance. There is even an argument that heat treatment was discovered as part of optimising glue rather than optimising colour, so a lump of red earth in a cave is not proof of anything spiritual. For a long stretch of the 20th century, a red stain in a paleolithic layer was treated as an automatic ticket to the word ritual, which is a classic case of finding what you brought with you. If your model of ancient humans is mystical, every stain is a ceremony. If your model is practical, every stain is glue.
30:46
Neither approach is doing any actual thinking. So let me hand you the tool that gets us out of this, because you will use it for the rest of this video. There are three tests that distinguish symbolic behaviour from ordinary utility and they work on almost any object. The first is excess. Does the effort exceed the function? Grinding a lump of ochre into powder is functional. Selecting for the brightest red when a duller piece would glue just as well is not. Carving an instrument out of solid ivory when hollow bird bone is free is not.
31:16
Excess is the fingerprint of meaning, because meaning is the only thing that makes extra effort feel necessary. The second is standardization. Does the pattern repeat?
31:27
One engraved block is an individual with time on their hands. A dozen engraved blocks sharing a design convention across thousands of years of deposits is a tradition, which means transmission, which means the mark carried information that more than one person could read. The third is context. Where does it appear and on what? The same substance means different things in different places. Ochre in a hide-working area is a tannery. Ochre on a grinding stone is production. Ochre on a corpse is something else entirely. And that last case is where the practical explanation quietly runs out of the road.
32:01
Consider what you're doing when you paint the dead. A body in a grave will not be going out in the sun. It will not be bitten by insects it needs protecting from. It will not scrape a hide, haft a spear or benefit in any physical way from anything you apply to it.
32:16
Every practical function of ochre is a function performed on the living. Apply it to someone who is permanently finished with practical functions, and you have spent real labour on a scarce processed material for a return that exists only in the minds of the people standing around the pit.
32:32
That is the whole argument in miniature, and it is the same argument the entire video is built on. The moment effort stops producing physical results and keeps happening anyway, you are no longer looking at economics. You are looking at meaning.
32:46
One more detail worth noticing before we move on, and it deserves a light touch rather than a heavy one. The colour choice is oddly consistent. People had access to a palette. There is white from clays and ash, black from charcoal and manganese, yellow from untreated guthite and various browns. Yet the substance that gets mined, processed, traded, hoarded, engraved, cooked and placed on the dead is overwhelmingly red, so consistently that people literally heated yellow earth to make it redder.
33:17
It is not much of a leap to connect that to the most vivid red any of these people ever saw, which came out of bodies at the two most significant moments in the life of a group, birth and death.
33:29
That association shows up in a great many cultures across a great many regions. I would just gently note that the fact that an idea is widespread and emotionally satisfying does not make it demonstrated, and we cannot interview anybody at Blombos about their colour preferences. File it as plausible, not proven.
33:47
What we can say, and say firmly, is this.
33:50
By around 100,000 years ago, humans were maintaining a technology whose only certain output was appearance. They knew where to find the raw material, they dug into hillsides to get it, they cooked it to change its properties, they compounded it with fat and bone and charcoal to a repeatable formula, they stored the results in containers, they carried it inland from the coast, they engraved it with patterns that other people could recognize, and they put it on bodies that were past all use. That is a vocabulary, not a language, not a religion, not a theology, but a stable set of marks and substances whose meaning lived in the heads of a community rather than in the object itself.
34:29
Stories from the fire now had somewhere to be stored. A design on an ochre block does not say anything on its own in exactly the same way that a wedding ring does not say anything on its own. It only works if there is a group of people who already share the story that explains it. So we have the setting, we have the storage medium, and we have the first real evidence that people were investing in things that did not feed them. The next step in the record is harder and considerably more contentious because it moves from what people put on the body to what people did with the body. It is also the place where this field has embarrassed itself more publicly and more often than anywhere else, with famous discoveries that turned out to be nothing and unglamorous discoveries that turned out to be everything.
35:11
A grave looks like the perfect piece of evidence. It is a body, deliberately placed, often in a pit somebody had to dig with tools that were not designed for digging, sometimes with objects arranged around it. Compared to arguing about crumbs of burnt flint, it feels almost like a signed confession. Here is a dead person. Here is somebody who cared enough to put them there. Surely that settles it. It does not settle it, and the reasons why are the most useful thing in this entire video. Because before we can ask when people started believing in something beyond death, we have to deal with the fact that this specific corner of archaeology has a long and colourful history of announcing profound discoveries that later turned out to be rodents.
35:52
Start with what a burial actually has to prove. Getting a corpse out of the living area is not a spiritual act, it is housekeeping.
36:01
A body left where people sleep attracts scavengers, and scavengers are not selective about which members of the household they take an interest in.
36:08
So any species with a fixed camp has a practical motive to move the dead somewhere else, and a few non-human animals do something in this neighbourhood. Elephants show prolonged interest in the remains of their own kind. Some primates carry dead infants for days. That is real and worth respecting, but it is not the same as excavating a hole, arranging a body in a specific posture, and putting things in with it that the living could have used.
36:33
What archaeologists look for, then, is a set of technical markers. Is the skeleton articulated, meaning the bones are still in anatomical order rather than scattered, which tells you the body was covered before it decomposed and nothing dragged pieces around? Is there a cut feature, a pit whose edges are visible as a change in sediment, proving somebody dug rather than nature dropping soil on top? Is the posture unnatural in a specific way, limbs flexed, hands placed, heads supported?
37:04
And are there objects present that cannot be explained as part of the surrounding rubbish? Each marker on its own can be produced by accident. A cave floor can settle. Sediment can slump into a depression and mimic a pit. Rodents are enthusiastic and tireless engineers who will happily move a small object several metres and put it somewhere extremely meaningful-looking.
37:26
Which brings us to the first of three famous cases and this one has the best name in prehistory attached to it.
37:32
In the hills of northern Spain, a limestone system called Atapuerca has been producing spectacular fossils for decades. Deep inside one branch of it, past a stretch of passages that take modern researchers a good while to negotiate on their stomachs, there is a vertical shaft dropping roughly 13 metres into a small chamber.
37:52
That chamber is called Cima de los Huesos and it contains the largest collection of ancient human fossils ever found in one place – more than 6,000 fragments, representing at minimum 28 individuals dated to around 430,000 years ago.
38:09
These are not our species. They belong to the lineage that leads toward Neanderthals, and DNA work on the site has helped establish that. Here is what makes it strange. The chamber has almost no large herbivore bones in it. If this were a simple natural trap, a pit that unlucky animals tumbled into, you would expect the standard menu. Deer, horses, the occasional bovid, all the species that wander into cave systems and misjudge a step.
38:36
Instead, the assemblage is overwhelmingly two things. Cave bears, which is entirely explicable because bears hibernate in caves and a certain number of them do not wake up, and people, dozens of people, in one hole, with the deer and horses conspicuously absent from the guest list.
38:54
The age profile is also odd. The dead skew heavily toward adolescents and young adults, which is not the demographic you would predict for accidental falls, though you can construct arguments in either direction and people have. And then there is the single most talked about object in the chamber, one hand axe, worked from red quartzite, a stone that is not the local everyday material and that stands out in colour. It shows no sign of having been used for anything. The excavators nicknamed it Excalibur, which is either the most evocative or the most prejudicial naming decision in modern archaeology depending on your mood, and they proposed it as a funerary offering.
39:32
one beautiful unused tool deliberately dropped into a shaft full of the dead. You will recognize the shape of that argument from the opening of this video. Effort that produces no physical return is the signature we are hunting. So is this the oldest funeral on earth, 400,000 years before anyone built a temple? Possibly.
39:52
and here are the ways it might not be. 1. The shaft may not always have looked like it does now. Cave systems rearrange themselves over hundreds of thousands of years. If there was once a more accessible opening, the chamber could have functioned as a genuine natural trap. with people falling in during exploration or while pursuing something and the demographic skew simply reflecting who does the risky exploring, which in most societies is teenagers and young adults. Teenagers, then as now, being the group most likely to look at a hole in the ground and decide to find out what is in it.
40:26
2. Water Caves flood, sediment moves, and material can be washed from one chamber into another over long spans. A concentration of bones in a low point is exactly what a hydrological system would produce. 3. Carnivores Bears and other animals move bones around and can accumulate them, and there is gnawing damage on some of the material. For, and this one bothers me the most, the hand axe might just be a hand axe. A single artefact in a deposit containing thousands of bone fragments is not a strong statistical statement.
40:59
If one unused tool went down a shaft along with everything else over a span of time, we are building a very large interpretation on a sample size of 1, and we already know that a sample size of 1 is how this field gets into trouble.
41:11
but there is a wrinkle that keeps the deliberate deposition argument alive and it comes from one particular skull.
41:18
Researchers examining the specimen known as Cranium 17 identified two separate impact fractures on the front of the skull and their analysis concluded that both were made by the same object arriving from different angles.
41:30
That pattern is difficult to produce by falling because falls do not usually hit the same small area twice from two directions.
41:37
The inference drawn by the team was that this individual died as a result of a confrontation rather than an accident, which matters for our question in a specific and slightly grim way. If the person did not fall, then the person did not walk into that shaft under their own power. Somebody carried them there. Whether that carrying was a ceremony, a clean-up, or something in between we cannot say, but it is transport, and transport requires a decision.
42:02
So Cima de los Huesos ends in an honest draw. It might be the deepest root of mortuary behaviour in the human line, or it might be geology plus bad luck plus a very good nickname. What it definitely is is a place where people, or something extremely close to people, were concentrated in one spot in a way that no ordinary process explains cleanly. Park it as unresolved, which is a perfectly respectable place for a 400,000-year-old question to live. The second case is the one that made it into every textbook, and it is the reason I keep insisting on scepticism.
42:36
In the 1950s, Ralph Salecki excavated Shanidar Cave in the mountains of northern Iraq and recovered the remains of several Neanderthals. One of them, designated Shanidar IV, became world-famous after soil samples taken from around the body were analysed and found to contain unusual concentrations of pollen, clustered rather than evenly spread, from a variety of flowering plants. The interpretation wrote itself. Neanderthals had laid their dead on a bed of flowers, chosen for colour, possibly for medicinal properties, in an act of tenderness roughly 60,000 years before anyone invented sentimentality.
43:13
It transformed public perception overnight. The knuckle-dragging brute of Victorian imagination became a grieving florist, and an enormous amount of popular writing was built on that single soil sample.
43:25
In 1999, Geoffrey Summer looked at the problem from a different angle and asked a question nobody had bothered with. What else lives in that cave? The answer is the Persian jerd, a burrowing rodent that is common throughout the region and that has one relevant habit. Jerds dig extensive tunnel systems and stock them with plant material, including flower heads and seeds in tidy caches. Their burrows run right through archaeological deposits and there were jerd remains at Shanidar. So the most economical explanation for concentrated clumps of pollen next to a Neanderthal is not a funeral.
43:58
It is a rodent pantry, installed at some point in the following 60,000 years by a small animal with excellent domestic instincts and no interest whatsoever in our research questions. There is also the perfectly ordinary possibility of pollen arriving on the boots and clothing of the excavation team, which is not a criticism of anybody. It is simply what happens when you walk through a spring landscape and then climb into a cave. I want to be precise about what this does and does not demolish. it does not prove Neanderthals were indifferent to their dead.
44:29
It proves that one specific piece of evidence, which carried an enormous amount of interpretive weight, does not carry that weight. That is a different and much more useful conclusion, and it is the sort of correction that makes the field trustworthy rather than embarrassing. And here is the twist that makes Shanidar worth the whole detour.
44:49
The site was reopened in the last decade by a new team, working with methods Seleki could only have dreamed about and what they found strengthens the case for intentional burial without any flowers at all. In the same area of the cave they recovered an articulated skeleton, now known as Shanidar Z, lying in a deliberate posture with the head resting in a way that suggests support.
45:10
More importantly the new work established that this was not an isolated event. Multiple individuals were placed in the same restricted part of the cave, close to a prominent upright rock that would have been a visible landmark in the gloom, across a span of time that means these were separate occasions rather than one disaster. That is a completely different quality of evidence and it is the moment to state the rule that will carry us through the rest of this subject. A single find is a hypothesis. A repeated pattern is a behaviour. The logic is simple, and it applies far beyond archaeology.
45:44
Any one-off event has an enormous space of possible causes, most of them boring, and you generally cannot eliminate them.
45:52
But coincidence does not organize itself. If bodies keep appearing in the same chamber, near the same landmark, in similar positions across generations, then whatever produced that pattern was a rule inside somebody's head. Chance produces scatter. Culture produces repetition. That is why the mundane Shanidar result several people put in the same corner over time is worth more than the glorious Shanidar result. One person on a bed of blossoms. The flowers were a story. The repetition is data. While we are in this cave, there is another individual here who deserves a mention, because he speaks to a related question.
46:28
Shanidar I was a man who lived with severe, long-term physical impairments. His skeleton records an injury to one side of the body that healed badly and left lasting damage, including to an arm and to the structures around one eye, and there is degeneration consistent with a long life afterward. The key word is afterward. He survived those injuries by years. In a mobile foraging society in mountainous terrain, an individual in that condition does not keep up on his own. Somebody shared food with him and somebody waited for him. That is not religion, and I am not going to pretend it is.
47:02
But it establishes that the emotional infrastructure was present.
47:06
A social world that carries its injured is a social world that has something to lose when a member dies. The third case is recent, extremely loud and still unresolved, which makes it a good illustration of how this argument actually happens rather than how it looks after 50 years of settling.
47:23
In a cave system in South Africa called Rising Star, teams working in exceptionally tight passages recovered the remains of a previously unknown member of the human family named Homo Naledi.
47:34
It is a strange creature by our standards, mixing features that look ancient with features that look modern, and crucially it has a brain roughly a third the size of ours. The material was dated to somewhere in the range of a few hundred thousand years ago, which places it in a period when our own ancestors were already producing considerably more elaborate technology. The remains sit deep inside the system, in chambers that are difficult to reach even with modern lighting and equipment.
48:01
The question of how they got there has been argued about since the first announcement. Then, in 2023, the research team led by Lee Berger went considerably further and published claims that these bodies had been deliberately buried in dug pits, that engraved marks on nearby cave walls were made by the same species, and that all of this happened long before comparable behaviour in our own lineage. If correct, it would break one of the load-bearing assumptions of this entire field, namely that you need a large brain and complex language to develop mortuary behaviour.
48:33
The reception was, to put it gently, not warm.
48:37
The papers were released as preprints alongside a major documentary and a book, which meant the public heard the conclusion before specialists could review the evidence.
48:45
When reviewers did get to it, several were blunt. They argued that the sedimentary evidence for deliberately dug pits was not demonstrated, that the features could be explained by natural processes moving bodies and sediment within the chamber, and that the dating and attribution of the wall markings were not established. Some reviewers recommended rejection in unusually direct language. Revised versions have followed. The team has continued to defend the interpretation and as things stand there is no consensus. I want to be careful here because it would be easy to turn this into a story about one researcher and it is not.
49:20
The interesting part is structural. The claim is exactly the kind that should require the strongest possible evidence because it overturns the most and the evidence is exactly the kind that is hardest to nail down because sediment in a deep chamber is a nightmare to interpret.
49:36
That combination guarantees a fight.
49:38
What you're watching, if you follow the Naledi debate, is not scientists being stubborn. It is the rule we just established being applied under maximum pressure by people who have every professional incentive to be right. So let us take that rule and point it at the cases where the answer is not in doubt.
49:55
At school, a rock shelter on Mount Carmel in Israel, in deposits somewhere around 100,000 years old, excavators found a burial that is difficult to explain as anything but deliberate.
50:06
The individual, usually referred to as School Five, was placed in a pit with the lower jaw of a wild boar held against the upper chest, the arms positioned around it.
50:16
Now, a boar jaw is not a useful item. You cannot cut with it, cook in it, or wear it comfortably. It is a piece of a dangerous animal, placed on a person who no longer needs protection from dangerous animals. Apply the test from earlier and it passes all three. There is effort beyond function, there is a context that excludes practical use, and it is not alone, because the same region produces comparable cases, including a double burial of a woman and a child, and the adolescent laid with antlers that we have already met.
50:47
One boar jaw is a curiosity. A regional habit of placing non-utilitarian animal parts on the dead is a practice. Then there is the site I opened this video with, and it deserves its details now rather than as a headline. At Sungir, in the cold plains of what is now Western Russia, around 34,000 years ago, someone buried an adult man and separately two children laid head to head in a single grave.
51:12
The beadwork is the famous part, but it is the rest of the inventory that ought to raise your eyebrows. The children were equipped with spears made from mammoth ivory, and ivory is not naturally straight, which means somebody worked out a method of softening and straightening a tusk into a shaft. The longest of these is well over two meters. There are carved disks, animal figures, and hundreds of pierced Arctic fox teeth. And if you have ever considered how many foxes and how much patience it takes to accumulate hundreds of specific teeth, you have grasped the point.
51:42
There is red pigment throughout, two children, in a mobile hunting society, in an ice age landscape, where every hour of skilled labour is an hour not spent on survival, and the community invested what has been estimated at thousands of hours of work into objects that were then covered with soil and never seen again.
52:01
This is not housekeeping. Nobody tidies a body away with a two-metre ivory spear. One more because it shows the same thing with a different flavour.
52:10
at dolny vestonets in moravia from a slightly later moment in the same ice age world there is a grave holding three young individuals arranged together in a deliberate composition with pigment applied around the heads and objects placed with them whatever story explains that arrangement is lost permanently but the arrangement itself is unmistakably a statement made by people who took the trouble to make it in frozen ground. So where does this leave us? The genuine argument among specialists is not whether ancient humans disposed of their dead deliberately.
52:42
That is not controversial. The argument is about when disposal became symbolic, which is to say when people started doing things to the dead that only make sense if the dead are still in some fashion participants.
52:53
A pit is logistics. A pit with a bore jaw, an unused blade, or 13,000 beads is a message, and a message implies a recipient.
53:03
The trouble is that the ground cannot tell us who they thought was receiving it. Dirt records actions, not intentions. To get at intentions, you have to stop digging and go look at the only machine that produces them, which is the one that has been sitting inside all these skulls the entire time. And the surprising thing about that machine is that it does not need to be taught any of this. The equipment that makes gods thinkable is standard issue. It installs itself in early childhood without instruction, and you can watch it running in a four-year-old who has never been to a religious service in their life.
53:34
Start with the oldest and dumbest part of the system, in the most affectionate possible sense of dumb. You're walking alone at dusk. Something rustles in the undergrowth to your left. Before any conscious thought occurs, your body has already made a decision. Heart rate up, head turning, muscles loaded.
53:52
Your brain did not calmly run a probability estimate on wind versus animal. It jumped straight to the assumption that something with intentions is over there, and it did so in a fraction of a second.
54:03
That reflex has a name in the literature. Researchers talk about a hyperactive agency detection device, an evolved tendency to attribute events to agents rather than to processes.
54:14
The anthropologist Stuart Guthrie laid out the broad case decades ago, arguing that we routinely interpret an ambiguous world in terms of faces, minds and purposes because that is the interpretation with the best payoff structure. Justin Barrett gave the mechanism its familiar name. The reason it exists is pure accounting and it is the cleanest piece of logic in this whole subject. There are two ways to be wrong. You can decide there is a predator when there is only wind, which costs you a jolt of adrenaline, a slightly undignified jump, and about four seconds.
54:45
Or you can decide there is only wind when there is a predator, which costs you everything, permanently, including all your future descendants.
54:53
Those errors are not remotely equal, so selection does not build a system calibrated for accuracy. It builds one calibrated to make the cheap mistake constantly in order to avoid the expensive one ever.
55:04
A smoke detector that occasionally goes off when you make toast is not badly designed. It is designed by someone who understands fire, so we are all walking around with an agent detector tuned to fire on the slightest provocation and the consequences are visible everywhere in modern life, which is the fun part. We see faces in wall sockets, in car fronts, in the arrangement of holes in a slice of bread. We apologize to furniture we bump into. We yell at software as though it can be shamed into behaving. Millions of otherwise sensible adults are convinced that their computer acts up specifically when they are in a hurry, which is not a belief about electronics.
55:40
It is an accusation of malice against a box.
55:44
Nobody taught us to do that. It is the default setting leaking through. Now scale that up to a world with no scientific explanation of anything. The wind stops the day before the hunt. A river that has run for generations dries up. Lightning hits a tree at the edge of camp and not the one beside it. Your detector, doing exactly what it was built to do, reports the same finding every time. Somebody did that. Not something. Somebody.
56:11
The second component makes the first much worse, and it is the one I find most entertaining because you can catch it in action at any dinner table with a small child present.
56:20
The psychologist Deborah Kellerman has spent years documenting what she calls promiscuous teleology, which has the tendency to explain objects and events by what they are for rather than by what caused them. Ask young children why rocks are pointy and they do not reach for erosion. They say the rocks are pointy so animals will not sit on them and get hurt, or so that animals can scratch themselves, ask about rain, and it exists so plants can grow, ask about mountains and they are there for climbing. Every feature of the natural world arrives with a job description attached.
56:53
The obvious response is that children are simply repeating what adults tell them, and that a decent education sorts it out.
57:00
Kellerman tested that, and the results are considerably less flattering. The preference shows up across different cultural and educational backgrounds, including in children raised by parents who explain the world in strictly mechanical terms.
57:13
More striking, adults do not lose it. They suppress it.
57:18
In experiments where adults are asked to judge explanatory statements under time pressure, the purpose-based explanations start slipping through as acceptable and this happens to people with serious scientific training.
57:30
In one well-known study, physical scientists at leading universities, given only a fraction of a second to respond, endorse statements of the form that a natural feature exists in order to accomplish something at significantly higher rates than when they had time to think. Sit with that for a second, because it is the most quietly devastating finding in this field. Purpose-based thinking is not a stage we grow out of. It is the factory default, permanently installed, and the scientific worldview is a piece of software running on top of it that requires continuous processing power to maintain.
58:03
Take away the time, the attention or the training and the old system comes straight back online. Every one of us is a teleologist wearing a lab coat and the lab coat is not load-bearing. Combine these two and the situation is already fairly advanced. You have a mind that assumes agents behind ambiguous events and assumes purposes behind natural features. That is most of a world view and nobody has taught it anything yet.
58:28
The third component is where death enters, and it is the one that does the heaviest lifting. Human beings are natural dualists. Not philosophically, but intuitively. We experience minds and bodies as different kinds of thing. Paul Bloom has argued at length that this is not a doctrine we absorb but a consequence of how the brain is built. We run two separate systems for understanding the world, one for physical objects and their mechanics and one for agents and their intentions. The object system handles trajectories, weight and collision.
59:00
The agent system handles goals, feelings and what somebody probably wants. They use different rules and they never fully merge. A rock is an object, a person is an agent. And when you meet a person, you are effectively running a model of their mind that is entirely separate from your model of their body, which creates a specific and consequential bug. Bodies are visible, so when one stops, you observe it directly. Minds were never visible in the first place. Your model of another person's mind was always an internal construction, assembled from their behaviour and maintained in your head.
59:34
The body's failure is data. The mind's failure is not because the mind was never in the room to begin with. Jesse Bearing ran the experiment that demonstrates this beautifully, and he ran it on children. The set-up involves a puppet show in which a mouse, going about its business, is caught and eaten by an alligator, at which point the researchers ask the children a series of questions about the mouse's current condition. Does the mouse need to eat now? Does it need to drink? Is its brain working?
60:03
On the biological questions, children are clear-eyed and correct. The mouse is finished. No more eating, no more drinking, no more body functions. Then the questions shift. Is the mouse still angry at the alligator? Does it still miss its mother? Does it know that it is no longer alive? And here the same children, who just correctly reported that the mouse has ceased to function, say yes.
60:25
It misses its mother. It is still scared. It knows what happened. The mental attributes survive the alligator entirely intact. The detail that turns this from interesting into important is the age pattern. You would expect that if this is a religious idea learned from adults, older children would show it more, having had more exposure. The pattern runs the other way. The youngest children are the most likely to attribute continuing mental states to the dead, and the tendency decreases somewhat with age. Continuity of mind is not something we are taught.
60:56
It is something some of us are gradually talked out of. There is an even more basic constraint underneath all of it. Try right now to imagine your own non-existence, not your funeral, not your family afterward, not the world continuing without you. Imagine the actual experience of not existing. You cannot do it, and the reason is structural rather than emotional. Imagining anything requires an imaginer to be present in the imagined scene, observing it.
61:23
Every attempt to picture your own absence sneaks a spectator in through the back door. The mind simulation engine simply cannot render a scene with nobody in it, which means that at the intuitive level, personal non-existence is not a frightening idea. It is an unthinkable one. So the afterlife does not require an inventor, a prophet or a founding document. It falls directly out of ordinary equipment. We model other minds separately from bodies. We cannot observe a mind stopping. We cannot simulate our own absence. A four-year-old with no religious instruction will tell you a dead mouse is sad about being dead and that child is not being taught theology.
62:00
They are reporting the output of standard human cognition.
62:04
Now put the three components together and watch what happens to the landscape. everything that moves without an obvious cause acquires an agent everything that has a form acquires a purpose and agents persist after their bodies stop so the storm is not a pressure system it is somebody in a mood the river does not flow downhill because of gravity it goes where it wants and it can decide to stop the mountain is there for a reason and the reason belongs to somebody and the grandfather who died last winter did not end.
62:35
His body ended. He is presumably somewhere, retaining opinions, and he is quite possibly aware of how you have been behaving.
62:43
That is animism, and it is not a primitive philosophy that someone sat down and composed. It is what the world automatically looks like through unmodified human hardware. And here is the part that converts a perception into an institution. Once you have decided that a force has intentions, a whole strategy set unlocks. Because you already know how to deal with things that have intentions. You have been doing it since infancy. If a person is angry with you, you do not attempt to modify their internal chemistry. You apologize.
63:14
You offer something. You show respect. You perform the behaviours that signal you're not a threat and are worth keeping around. That toolkit is deep, automatic and extremely well practised. So when the reins fail, and the reins are somebody, the response is not mysterious. It is the same thing you would do with a neighbour you had offended. Find out what you did, offer something valuable, demonstrate respect publicly, ask politely, bring a gift and make sure it is a real gift because a cheap gift insults the recipient and everybody knows it.
63:46
That is the entire logic of offering, and once you see it, you cannot unsee it. Sacrifice is not a bizarre invention. It is interpersonal diplomacy applied to weather. The same instinct that makes you bring something to a dinner party makes a community give up a valuable animal to a river. It is negotiation with a counterparty who never answers, which is admittedly a difficult relationship to maintain. Though as anyone who has dealt with a large organisation can confirm, humans are remarkably good at continuing to submit requests into systems that never respond.
64:18
There is one more piece worth adding because it explains not just how these ideas arise but why some of them stick and most do not.
64:25
The anthropologist Pascal Boyer pointed out that the concepts that survive across generations tend to share a peculiar structure. They are almost entirely ordinary, with one or two violations of expectation. A being that is exactly like a person, except that it cannot be seen. A mountain that behaves like a mountain, except that it listens. A statue that is a normal statue, except that it hears requests. Concepts with that shape are memorable in a way that fully ordinary ideas are not, since a completely normal person is not worth mentioning, while a person who knows what you did last night is extremely worth mentioning.
65:01
But they are also transmissible in a way that wildly strange ideas are not, because a being that violates everything at once cannot be reasoned about. You need the ordinary framework intact to make any predictions at all. Keep almost all of the human template and break one rule, and you get something both striking and usable. which is exactly why the gods that populate human cultures are not abstract forces or impersonal geometries. They are people, with tempers and preferences and family problems, plus one impossible attribute.
65:32
Now the honesty section, because this chapter has been confident, and confidence is exactly when to slow down.
65:39
Everything just described is a family of models, not a proven mechanism. Agency detection is a well-supported concept, but calling it a device makes it sound like a discrete component with a location and a wiring diagram and it is nothing of the kind.
65:54
It is a description of a behavioral tendency that probably emerges from several interacting systems. More importantly, despite decades of enthusiastic headlines, there is no belief center in the brain. Claims of a dedicated spiritual module have not held up.
66:10
The most famous version of this involved an apparatus applying weak magnetic fields to the head, reported to induce a sense of a presence in the room. It generated a great deal of excitement and a memorable nickname. Then an independent team ran the procedure with proper controls, including participants who received no field at all while believing they had, and the sensed presence showed up just as reliably in the people receiving nothing. What the experiment had mostly measured was suggestibility. which is a genuine and interesting finding, just not the one anybody wanted.
66:43
Neuroimaging of prayer, meditation, and religious experience shows changes in activity across networks involved in attention, self-referential thought, language, and emotion.
66:54
In other words, religious cognition uses the ordinary furniture of the mind, arranged in a particular configuration. There is no dedicated organ, and anyone selling you a godspot is selling you a headline.
67:06
There is also a live disagreement about what all of this means. The account in this chapter treats religion as a by-product, an accidental consequence of systems that evolved for detecting predators and understanding other people. The competing account holds that religious belief was itself selected for because groups that had it out-competed groups that did not. That argument is not settled and we will come back to it properly at the end where it turns into a genuinely fierce scientific fight with real data on both sides.
67:35
But the by-product account has one enormous advantage that is worth stating plainly. It removes the need for an inventor. Nobody had to think of this. There is no moment where a clever individual around a fire looks up and proposes a new theory of the universe. You simply take a social primate that detects agents aggressively, explains objects by purpose, models minds as separable from bodies, cannot conceive of its own absence, and spends its evenings telling stories about people who are not present, and you let it live in a world it does not understand.
68:09
Religion is not an idea that had to be added. It is what you get by default. Which raises the obvious problem with everything I have just said. Defaults are cheap. A tendency to see agents and purposes explains why people might casually assume the river has moods. It does not explain why anyone would spend thousands of hours on beads for a grave or haul ten tons of limestone across a hillside. Between a passing intuition and an institution that consumes real resources, something has to supply the pressure, and there is one experience, universal to our species and, as far as we can tell, unique to it, that generates that pressure continuously, for free, whether anyone wants it or not.
68:50
Every animal on this planet avoids dying. That part is not special. A beetle runs from a shadow, a fish bolts from a movement, a gazelle reacts to a sound in the grass before it has identified what made the sound.
69:03
Avoidance of death is the oldest software in biology and it works beautifully without any understanding attached to it whatsoever. What our species has is something categorically different and considerably less pleasant. We do not merely avoid death. We know about it. We carry, permanently in the background of an ordinary afternoon, a correct model of the fact that we personally are going to stop, that everyone we love is going to stop, and that this is not a risk to be managed but an appointment already on the calendar with the time left blank. As far as anyone can tell, we are the only species carrying that particular file, and I want to be careful with that claim, because it gets thrown around far too casually.
69:41
Plenty of animals respond to death in ways that are moving to watch.
69:45
Elephants investigate the remains of their own kind, sometimes returning repeatedly. Chimpanzee mothers have been observed carrying dead infants for days. Certain birds gather noisily around a dead member of their species, though when researchers looked closely at that behaviour it appeared to be less a memorial service and more a rapid seminar on locations to avoid, which is arguably more useful and definitely more practical. All of this shows that other minds register death as significant.
70:13
none of it shows that any animal sits quietly on a warm afternoon with nothing threatening anywhere nearby and contemplates its own future non-existence the honest statement is that we have no evidence for it not that we have disproved it but the absence is conspicuous so where did we get this gift and can we return it The uncomfortable answer is that it arrived attached to the thing that makes us formidable. Human beings are unusually good at mental time travel, the ability to leave the present and run a detailed simulation of a moment that has not happened yet.
70:45
It is the foundation of nearly everything we do well. It lets you dry meat in autumn for a month you cannot see. It lets you place a trap where an animal will be tomorrow rather than where it is now. It lets you nap a blade tonight for a hunt next week, teach a child a skill they will need in a decade, and store water because you remember what the dry season did last time. No other animal plans on that scale. This capacity has been studied carefully and the researchers Thomas Suddendorf and Michael Korbalis gave it the name that stuck, arguing that the ability to project yourself into a future scenario and mentally inhabit it is one of the genuine dividing lines in cognition.
71:23
There is a long-running debate about how much of it other animals have. Certain birds cache food in ways that appear to anticipate future need, and some primates prepare tools in advance of using them which are real and impressive behaviours. What remains contested is whether any of it involves actually simulating a future moment as an experience, as opposed to executing a well-tuned rule that produces future-appropriate behaviour without anyone picturing anything. The distinction sounds academic until you notice that it is the entire difference between storing nuts and dreading old age.
71:55
But a simulation engine that can project forward has no natural stopping point. Once you can model next month, you can model next year. Once you can model next year, you can model the year you become too weak to walk and the year after that.
72:09
The exact same faculty that lets you prepare for winter eventually informs you, without being asked, that there is a final winter in the sequence and that you are in it at some point and that this is not an error in the model. That is the bill for foresight. We got planning and planning came bundled with dread and the subscription cannot be cancelled.
72:30
There is also a developmental angle here that makes the whole thing feel less abstract because every single one of us had to assemble this knowledge from scratch and researchers have mapped how it happens.
72:40
A mature understanding of death is not one idea. It is a bundle of separate components that children acquire at different rates. There is irreversibility, the recognition that this state does not end and the person does not come back.
72:54
There is non-functionality, the understanding that the body has stopped doing all the things bodies do. There is universality, the fact that it applies to every living thing without exception. There is causality, a grasp of what actually produces it rather than a magical account. And there is the last and hardest piece, personal inevitability, which is the moment a child understands that universality includes them. Studies generally find these assembling through early and middle childhood, with the full package typically in place somewhere around the ages of seven to ten, though the range is wide and heavily affected by experience.
73:30
A child who has lost a family member or lives with illness tends to get there earlier, which is not a happy finding but it is a consistent one. What interests me is the shape of that last step.
73:40
Most people, if they think about it, can locate a specific evening in childhood when the realisation arrived. Not the fact that people die, which children absorb fairly casually, but the personal version, usually delivered by a mind that was supposed to be going to sleep. It tends to be remembered vividly decades later, which is not how we usually remember things a seven-year-old thought about. So the knowledge is not cultural furniture that some societies install and others do not. It is a developmental milestone, arriving on roughly the same schedule as losing your baby teeth, and there is no version of childhood anywhere on earth that skips it.
74:16
Every human community in history has had to have an answer ready, because every generation of children has asked the same question at approximately the same age, usually at bedtime, usually at the worst possible moment for the adult being asked. Now consider what this does to a stress system that was never designed for it. The standard vertebrate arrangement is elegant. A threat appears. Your body floods with hormones that shift resources toward immediate survival, raising heart rate, sharpening attention, releasing stored energy, and shutting down anything that can wait, including digestion, tissue repair, growth, and immune activity.
74:54
You then run, fight, or hide. The threat resolves one way or another. The chemistry clears. Everything switches back on.
75:04
From start to finish, it is a few minutes of emergency followed by a return to normal service. The system works because it assumes the emergency is short, external and resolvable. All three assumptions are correct for a gazelle. All three are false for the specifically human problem because the threat is not out there in the grass, it is inside the skull.
75:25
It can be triggered by a thought, a memory, a stray association at three in the morning. It cannot be outrun.
75:32
It cannot be fought. And it never resolves, because unlike most things the anxiety system fires at, this one is completely accurate. You're not being irrational. That is the worst part of it. So our ancestors had an alarm system capable of running continuously in response to a correct prediction, about which nothing could be done, which is a design flaw of some magnitude.
75:54
The most influential attempt to build a science around this problem started with a book. In the early 1970s, the anthropologist Ernest Becker published an argument that human civilization is, at its foundation, a machine for managing the knowledge of mortality. His claim was that we cannot function while consciously holding that knowledge, so we construct systems of meaning that place us inside something larger and more durable than a body, and we spend our lives earning a position within those systems.
76:22
Becker did not live long after publishing it, which gives the whole thing an unfortunate poetry. A decade later, three social psychologists – Jeff Greenberg, Sheldon Solomon and Tom Piszczynski – decided that Becker's argument was too important to leave as literature and set out to test it in a laboratory.
76:39
The framework they built is called terror management theory and its core proposal is straightforward enough to state in a sentence. Awareness of death produces a potential for overwhelming anxiety and human beings manage it primarily through culture, by belonging to a shared system of meaning that will outlast them and by maintaining a sense of being a worthy participant in it.
77:00
The theory identifies two ways out of the problem. One is literal continuation, the proposition that some part of you does not stop when the body does. The other is symbolic continuation, and this one is available to everybody regardless of belief. Your children continue. Your group continues. Your name is repeated. The thing you built stands. The story you told gets retold by people who never met you. Symbolic immortality is why a person with no interest in an afterlife can still be desperate to leave something behind.
77:31
And it is why carving your initials into a tree feels obscurely satisfying, for reasons no one can articulate. And symbolic continuation is not a consolation prize, which becomes obvious the moment you look at how hard people work for it. A name recited in a genealogy is a form of existence. Being the person who taught a skill that is still being used two generations later is a form of existence. In societies without writing, the entire archive of who a person was consists of other people saying so, which means that securing a place in that spoken record is the only durable thing available.
78:06
It explains a behaviour that would otherwise be baffling, namely that people across history have consistently been willing to accept serious physical risk in exchange for being remembered well, and have consistently feared disgrace more than discomfort. A reputation is a piece of you stored outside your body, in hardware that does not die when you do.
78:25
The experimental program built around this became enormous, running to hundreds of studies over several decades. The basic method is consistent. You take two groups. One group is prompted to think about death in a personal way, usually by writing briefly about what will happen to them physically and what emotions the thought produces.
78:45
The control group writes about something unpleasant but unrelated, often dental work, which is a genuinely inspired choice because it controls for general discomfort without touching mortality. Then you measure something apparently unconnected. The reported results have a consistent shape. After a death prompt, people become more attached to their own group and more hostile to people who challenge its values. They judge rule breakers more harshly and reward those who defend shared beliefs more generously. In one widely discussed early study, working municipal judges given a mortality prompt set substantially higher bail amounts in a hypothetical case than judges in the control condition, which suggests that the effect reaches people whose entire professional training is built around not being influenced by their mood.
79:28
Other studies reported increased discomfort at the improper use of cultural symbols, greater attraction to confident charismatic leaders, stronger endorsement of religious belief, and increased conviction about the supernatural.
79:41
There is also a timing detail that is genuinely clever. The effects tend to be strongest not immediately after the prompt, but after a brief delay filled with an unrelated task.
79:52
When the thought of death is in focal awareness, people push it away with rational reassurance, telling themselves they are healthy and young and it is far off. It is once the thought has dropped out of conscious attention that the defensive behavior emerges quietly in judgments about apparently unrelated matters, which is a rather elegant account of why nobody experiences themselves as motivated by mortality while behaving in ways that look a great deal like it.
80:18
Now the part where I have to slow you down again, because this framework is popular enough that it often gets presented as established fact, and it is not.
80:27
Social psychology went through a difficult decade when researchers began systematically attempting to reproduce famous findings and discovering that a disturbing number of them did not hold up under stricter conditions. Terror management research was caught in that wave. A large coordinated replication effort, with multiple independent laboratories running the same protocol under agreed conditions, with far larger samples than the originals, failed to recover the basic mortality salience effect. The original researchers responded with methodological objections about how the replications were conducted, and that argument continues.
81:02
So here is the calibrated version.
81:05
The general observation that human beings manage the knowledge of death through shared systems of meaning is not seriously disputed by anyone. It is visible in every society that has ever been described. What is disputed is the specific laboratory claim that a short writing exercise reliably produces measurable shifts in unrelated judgments minutes later.
81:27
Treat terror management theory as an influential and productive way of thinking about the problem, one that generated real insight and real research rather than as a settled law of human behaviour. Frameworks can be extremely useful without being airtight, and this is one of the more useful ones in the business. What is not in dispute at all is the physiology, and this is where the argument gets teeth, because it converts an emotional problem into a survival statistic.
81:54
Chronic activation of the stress response is measurably bad for a body. Sustained elevation of stress hormone suppresses immune function, slows wound healing, interferes with sleep architecture, disrupts digestion, and over long periods contributes to cardiovascular damage.
82:11
In a modern context, this is a quality-of-life issue and a long-term health risk, and we talk about it in terms of burnout and blood pressure. Transplant the same physiology into a world without antibiotics, surgery, or any reliable medical intervention, and the stakes change completely. A suppressed immune system in the Upper Paleolithic is not a wellness concern. It is the difference between surviving an infected wound and not. It is the difference between recovering from a winter illness and being the reason the group has to dig a grave in frozen ground.
82:43
In an environment where minor injuries and ordinary infections were regularly fatal, anything that reduced chronic physiological stress across a population would have translated directly into more people surviving to reproduce. Which reframes the entire question of whether religious belief is true into something more interesting. Whatever else it was, it was a measurable intervention on a body. collective ritual, shared singing, repetitive prayer and communal ceremony reliably reduce self-reported anxiety, and they do so in ways that show up in physiological measures as well.
83:15
This is not a controversial or mystical claim. It is the same category of effect as the well-documented placebo response, in which an expectation of relief produces genuine, measurable changes in symptom experience, pain perception and stress physiology.
83:31
I should be exact about what placebo does and does not do, because this argument gets abused in both directions. Expectation does not shrink a tumour, mend a fracture or clear an infection. It does modulate the experience of pain, the level of physiological arousal and the subjective sense of control. And those three things are precisely the mechanisms through which chronic stress does its damage. So the claim here is narrow and defensible. Belief did not heal anybody. It reduced the background load of anxiety in a population, and lowering that load has real downstream effects on immune function and survival.
84:07
Religion functioned, among other things, as a population-scale intervention on the stress axis. There is a further detail that makes ritual an unusually strong version of this effect. Placebo responses are not uniform. They scale with the elaborateness of the procedure, the confidence of the person administering it, the seriousness of the setting, and the amount of attention the recipient receives.
84:29
An unremarkable pill in a plain room produces a modest response. An involved procedure, performed by a figure with recognised authority, in a setting marked as significant, in front of a community that is assembled specifically for the purpose, sits at the far end of that scale. Some researchers prefer to call the whole phenomenon the meaning response rather than the placebo effect, precisely because what does the work is not the inert substance but everything arranged around it, which means that ancient healing ritual, judged purely as a delivery system for that effect, was close to optimally designed, long, formal, sensory, communal, conducted by a specialist and aimed at a person lying at the centre of the group's undivided tension.
85:12
If you set out deliberately to maximize the psychological component of recovery using nothing but social materials, you would build something that looks remarkably like a ceremony.
85:22
Brain imaging supports the general picture without supporting anything dramatic. Studies of experienced practitioners during deep meditation and sustained prayer show shifts in the network of regions associated with self-referential thinking and mind-wandering, sometimes called the default mode network. which is the system that generates that constant internal monologue about yourself, your past and your worries.
85:47
Quieting that system corresponds with the reported experience of absorption and reduced self-focus and with a drop in rumination.
85:55
Early studies imaging contemplative nuns and experienced meditators found changes in areas involved in attention and in the sense of spatial self-location which fits the way people describe these states.
86:08
As mentioned before, none of this locates belief anywhere. It shows that deep ritual practice recruits ordinary attentional and self-monitoring systems and turns down the volume on the part of the mind that worries, which is, in fairness, the whole product. All of this becomes far more urgent once you look at how often our ancestors actually needed it.
86:28
The statistics on Paleolithic mortality require a bit of care because the headline number badly misleads people. You will often see life expectancy in the range of 30 to 35 years and the natural reading is that the world was full of people collapsing from old age at 32.
86:44
That is not what the figure means. Life expectancy at birth is an average across everyone and it gets dragged down enormously by deaths in the first years of life.
86:54
Once an individual survived childhood, the picture improves sharply. Skeletal evidence and comparative work with recent foraging populations suggest that a person who reached 15 had a reasonable chance of reaching their 50s or beyond, and older individuals are well documented in the archaeological record.
87:11
The brutal number is the other one. Estimates for pre-modern and foraging populations commonly put the proportion of children who died before reaching adulthood somewhere between a third and a half.
87:22
Roughly a quarter died in the first year. Turn that from a percentage into a life. A woman who survives to 40 in this world has likely given birth several times and has buried some of those children. She has almost certainly lost siblings. She has watched people she grew up with die from injuries that a modern clinic would handle in an afternoon. Grief in that world is not an occasional catastrophe that interrupts an otherwise stable life. It is a recurring condition arriving every few years for everyone, forever. It is weather.
87:54
And grief is a genuinely strange cognitive problem, which is worth spelling out because it explains why ritual works.
88:01
Remember that our model of another person lives in our own head. When you're close to someone, that model is extremely detailed and it runs constantly generating predictions. It tells you what they will say about the news before you tell them. It tells you which noise at the entrance is them arriving. It expects them at the fire in the evening because they have been at the fire every evening for 20 years. When the person dies, the body stops but the model does not. It keeps producing predictions on schedule and every one of them now fails. You turn to say something and there is nobody to say it to.
88:33
You hear a sound and your mind delivers their name before catching itself. Each of those moments is a small error signal and the mind treats error signals as something to be resolved by searching.
88:44
That is why grief feels like looking for someone. The searching cannot succeed, so it repeats, sometimes for years. It is also worth stating the group-level version of this problem, because we have been discussing grief as a private experience, and in a small band it was never private. A person deep in bereavement is not a fully functioning member of a hunting party. Attention is impaired, sleep is disrupted, reaction times slow, and risk assessment degrades, which in a landscape full of large animals and sharp edges is an active danger to everyone present.
89:16
A household that stops functioning stops producing food. If several people are affected at once, which is what happens when a family loses someone, a small group can lose a meaningful share of its working capacity at precisely the moment it can least afford to.
89:30
Restoring people to function was therefore not a kindness the group extended when it could spare the time. It was maintenance on the only machinery the group had.
89:39
Now consider what a funeral rite is, viewed as a piece of engineering rather than as a set of beliefs. It does not delete the model. Nothing can do that, and cultures that have tried have not enjoyed the results. What it does is relocate it. The person is not gone, they have moved, and here is where they went, and here is how you now interact with them. Bring food to this place. Speak to them at this time of year. Keep their name in the account of the family. The prediction engine is not shut down, which is impossible. It is given a new and sustainable set of expectations that can actually be met.
90:12
Which leads to the most practical point in this chapter, and the reason I would argue religion earned its price here before it earned it anywhere else. Humans cope badly with situations where nothing can be done. Not just emotionally, but physiologically.
90:26
Research on helplessness has consistently found that the harmful effects of a stressor increase sharply when the individual has no available response, and decrease when they have any response at all, even a fairly ineffective one. Control, or the experience of control, is itself protective. A situation that offers no action is corrosive in a way that a difficult situation with a task list simply is not.
90:49
There is a classic field observation that makes this concrete and it comes from the Trobriand Islands rather than from any laboratory.
90:57
The anthropologist Bronislaw Malinowski noticed that the islanders he worked with fished in two quite different settings. Inside the lagoon, the water is sheltered, the conditions are predictable, the techniques are reliable, and a fishing trip is essentially a job.
91:12
Beyond the reef, in open water, everything changes. The catch is larger, but the weather is dangerous. The outcome is unpredictable, and people occasionally do not come back.
91:23
The ritual behaviour mapped onto that difference exactly. Lagoon fishing involved almost no ceremony. Open sea expeditions involved a considerable amount of it, performed carefully, with attention to getting the details right. Same people, same beliefs, same week. The variable was not piety. It was uncertainty. That observation has been repeated in various forms since, and it points at something important about what ritual is actually for. It concentrates where outcomes matter and control is absent, which is why you find it clustered around childbirth, illness, weather, travel, warfare and death, and largely absent from tasks where competence reliably produces the desired result.
92:06
Nobody performs a ceremony before tying a knot they have tied 10,000 times, and it is why the behaviour survives in exactly the same distribution today, which we will get to at the very end of this story, once we have earned it. Death offers no action. Nothing in the standard human toolkit applies. You cannot fight it, fix it, negotiate it, or wait it out.
92:29
A person facing that with no script is left standing in a camp with a body and no idea what to do with themselves or their hands.
92:37
Ritual supplies exactly what is missing. It provides a sequence. Wash the body. Wrap it. Apply the pigment. Carry it along a particular route. Place it in a particular orientation. Put certain objects with it. Sing the specific songs in the specific order. Return at intervals to bring something.
92:56
Suddenly there is a procedure, and a procedure has steps, and steps can be executed correctly, and executing them correctly produces a genuine feeling of having done something right at a moment when nothing right is available. Notice what that means for our central question. The procedure does not need to be true to work. The benefit does not depend on anything actually receiving the offering. The benefit is that a shattered person has somewhere to put their hands for three days and a group has a coordinated way of behaving during the period when it is most likely to fall apart.
93:28
There are two more functions worth naming because they are easy to miss. The first is scheduling.
93:34
Rites of mourning do not only tell people how to grieve, they tell people when to start and when to stop. There is a period, it has a defined length, and at its end there is a marked transition back to ordinary life. That sounds bureaucratic, and it is, and that is precisely the value. A society that gives grief a shape also gives it an exit and gives everyone around the grieving person a clear signal about how to treat them at each stage.
94:01
Left unstructured, mourning has no natural endpoint and no agreed etiquette, which is a problem modern societies have and are not handling especially well. The second is mobilisation. A funeral rite is a summons. It requires that people show up, and once they have shown up, they are physically present with a household in crisis. in practice that means food arriving work being covered children being watched and a bereaved person being surrounded rather than isolated for the exact period when isolation would be most damaging the theology whatever it was is the reason people came the support is what they brought with them and when they came they did something together that has a measurable physical effect which brings us to the last mechanism in this chapter Human groups engaged in synchronized activity, moving in time, singing together, chanting in unison, show a specific and repeatable response.
94:56
The most common way of measuring it is indirect but effective. Researchers test pain threshold before and after an activity, usually with a pressure device, because endorphin activity raises the level at which discomfort becomes intolerable. Endorphins are part of the same system involved in social bonding across a range of species, so the pain threshold works as a usable proxy.
95:19
A well-known study of this compared rowers training together in a boat, with the same rowers exercising alone on machines, with the physical workload matched.
95:28
The synchronized condition produced a significantly higher pain threshold than the solo condition. Same effort, same muscles, same duration, same… The only variable was whether the movement was shared and coordinated, and coordination alone changed the chemistry. Singing produces a comparable result. Studies of choirs have found increased pain thresholds and increased reported closeness among participants after singing together, and one of the more useful findings is that the effect appears even in very large groups of singers who do not know each other.
95:58
Which is notable, because most forms of social bonding do not scale at all. Sitting with someone, grooming, one-to-one attention – all of these are strictly limited by time and by the number of hours in a day. Singing is not. A hundred people can bond simultaneously through a single shared activity, in the time it takes to perform one song. There is a second, quieter mechanism at work in group singing that has nothing to do with endorphins, and it is my favourite finding in this area because it is so concretely physical. Singing forces a particular breathing pattern.
96:31
You take a quick breath at the start of a phrase, then exhale slowly and under control for the length of that phrase. Slow controlled exhalation activates the branch of the nervous system responsible for calming the body down, which is why deliberate slow breathing is the first instruction in essentially every anxiety management technique ever devised. Now add the group. When people sing the same melody together, they breathe at the same moments because the phrases dictate it. Researchers monitoring singers have found that heart rate variability begins to align across the group, driven by that shared respiratory rhythm, with the strongest synchronization occurring during structured unison singing such as slow chanting.
97:10
A group singing in unison is, quite literally, operating on one shared breathing pattern and drifting toward a shared cardiac rhythm.
97:19
so a community gathered around a death and singing together is not merely expressing something. Every individual in that circle is running a physiological procedure known to reduce arousal and running it in lockstep with everyone else. The comfort is not metaphorical, it is respiratory. I will flag the usual caution. Pain threshold is an inference, not a direct measurement of endorphin release. Sample sizes in this area are often modest and the effects are real but not enormous.
97:48
What survives the caution is the direction of the finding, which has held up across different activities and different populations. Doing the same thing at the same time with other people changes how you feel about them and how your body processes discomfort.
98:03
Now assemble the whole apparatus and look at it as a solution rather than as a set of curious findings. A species discovers, as a side effect of being able to plan, that it is going to end. This produces a chronic stress load with no natural resolution, in an environment where chronic stress carries real mortality costs. That same species loses roughly half its children and buries its siblings and parents at regular intervals, meaning the abstract knowledge is constantly refreshed by concrete loss.
98:33
and the situation offers no effective action, which is the specific condition under which human beings deteriorate fastest.
98:40
Into that gap comes a practice that provides a shared story explaining where the person went, a sequence of actions to perform when no action helps, a schedule that bounds the mourning period, a summons that brings material support to a household in crisis, and a synchronized collective activity that measurably reduces anxiety and physical distress while strengthening the bonds of the group performing it. You do not have to believe a single metaphysical claim to recognize that as an extraordinarily good piece of technology. Nothing else available at the time came close.
99:13
There was no medicine for it, no philosophy yet, no institution, no alternative comfort of any kind. Whatever else religion later became, and it became a great many things, grief is the first problem it's solved better than anything else on offer.
99:28
But a script is only half of a working system. A sequence of actions still needs content. The people performing it need to know who they are addressing, what that being looks like, where it lives, what it wants, and what happened in the story that explains why any of this is done this way. The dead have to go somewhere specific, and somewhere specific has to be imagined by someone and then agreed on by everyone else. which means that at some point the invisible had to be given a face. And remarkably, we can find the moment when people started drawing it.
100:00
In December of 1994, three cavers exploring a cliff face above a river gorge in southern France felt a draft coming out of a pile of rubble. A draft means space behind the rock, so they cleared the debris, squeezed through, and dropped into a chamber that no human being had entered in more than 20,000 years.
100:19
The floor was littered with the bones of cave bears. The walls were covered in animals. Chauvet, as the cave came to be called after one of the three, contains over 400 images, and the earliest of them date to roughly 36,000 years ago.
100:33
Which caused an immediate and slightly humiliating problem for everyone who studied this material, because the images are excellent. Not promising. Not charmingly primitive. Excellent. Whoever worked in that cave prepared their surfaces, scraping the wall to expose a lighter layer so the figures would stand out against a brighter background, which is essentially priming a canvas. They used charcoal and applied it with control, smudging and blending to produce shading along an animal's flank so the body reads as three-dimensional. They employed perspective, drawing the far horn of a bison shorter than the near one, a convention that formal European painting somehow misplaced for a considerable stretch of later history.
101:14
They use the natural swellings and hollows of the rock, so that a bulge in the limestone becomes the shoulder of a bison.
101:21
And in one of the most striking panels, a group of lions is drawn with overlapping heads, receding into the distance and multiple leg lines that read unmistakably as motion, which means somebody in the last ice age independently invented a technique that would later be used for animated film.
101:37
The reason any of this survived is a piece of luck so convenient it sounds invented.
101:42
At some point roughly 20,000 years ago, the cliff above the original entrance gave way and sealed the cave completely. Everything inside was placed in storage. The floor surfaces, the charcoal, the bones, the footprints and the pigment, all held at stable humidity in the dark with no visitors, no vandalism and no Victorian gentlemen scrubbing at the walls to see what was underneath.
102:04
When the modern explorers came through, they found claw marks where bears had raked the soft walls, some of them underneath the paintings and some drawn over the top, which lets researchers work out the order of events across thousands of years. It also means the cave was shared. People did not have exclusive use of this space. They went into a chamber that large predators used as a den, carrying fire to draw on the walls.
102:27
The subject matter is peculiar, too. If you assume cave painting is about hunting magic, you would expect the walls to be full of dinner. Instead, the dominant species at Chauvet are the ones nobody was eating. Lions, rhinoceroses, mammoths, bears. These are the animals that could kill you, drawn in the greatest numbers and with the most attention, by people whose actual diet leaned heavily on reindeer and horse. Whatever these images were for, a shopping list is not it. And then there is the chamber with the skull.
102:57
In one of the deeper halls a block of limestone had fallen from the ceiling at some point, forming a flat-topped slab roughly at waist height near the centre of the space. On top of it, positioned deliberately, is the skull of a cave bear, facing outward.
103:12
On the floor around it are dozens more bear skulls. Cave bears died in this cave in large numbers, so the raw material was abundant. but bones do not climb onto blocks. Somebody picked one up, carried it to the middle of the room and set it on the only piece of furniture the room had, in a chamber decorated with images deep underground by torchlight. We cannot say what it meant. We can say with complete confidence that it was arranged, that arranging it served no practical purpose whatsoever, and that if you found the same configuration in a building from any later period, no one would hesitate to call it an altar.
103:46
There is one more thing at Chauvet worth pausing on. Hanging from the ceiling in a rear chamber is a pendant of rock, a stalactite formation that comes down like a hanging pillar. Drawn on it is a composite figure. The lower portion is the lower body of a woman, rendered in the schematic way these things usually are. Emerging from and above that is the head and shoulders of a bison. The two are not adjacent images that happen to overlap. They are integrated into a single figure. Hold that thought because it is going to turn out to be the most important kind of image in this entire chapter.
104:19
Roughly 20,000 years after Chauvet and a few hundred kilometres away, we get the sight everyone has heard of. Lascaux was founded in 1940 by a group of teenagers, one of whom was following a dog that had gone down a hole, which remains the single most productive act of canine misbehaviour in the history of archaeology.
104:37
The paintings there date to around 17,000 years ago and they are, in terms of scale and ambition, on another level. Enormous bulls, some over five metres long, wrap around a chamber that has been called a cathedral so often that the word has stopped meaning anything. There are horses, deer swimming a river rendered as a row of heads, and abstract sign systems that nobody has convincingly decoded. But the part of Lascaux that matters for our argument is not the Grand Hall. It is a narrow shaft, accessible only by descending several metres, which contains the single strangest composition in Paleolithic art.
105:13
At the bottom of that shaft there is a scene. A bison stands with its head lowered, its body wounded, entrails visible clearly in extremity. A spear or spear-thrower lies nearby. To the left a woolly rhinoceros appears to be walking away, seemingly uninterested in the entire business. And between them, falling backwards at an angle, is a human figure. He is drawn with four fingers on each hand rather than five. He is unmistakably male, and his head is not a human head. It is a bird head with a beak on a long neck.
105:46
Below him is a staff or pole topped with the figure of a bird. Every generation of researchers has produced a reading of this scene, a hunting accident memorialised on the wall, a specific event from a story the community shared. A soul departing the body in the form of a bird, a ritual specialist in a trance state, which we will come to properly in a moment.
106:07
What is inarguable is that it is a scene with multiple elements arranged in relation to one another, which is rare in this period, and that it includes a person who is not entirely a person. While we are here, notice a pattern that runs through nearly all Ice Age cave art and is almost never commented on. The animals are magnificent. The humans are terrible.
106:29
In the same caves where a horse is rendered with an accuracy that lets modern specialists identify the species, subspecies and seasonal coat, the human figures are stick-like, distorted, faceless or reduced to a few lines. These are the same artists. They had not forgotten how to draw.
106:46
A painter capable of shading a bison's shoulder is entirely capable of drawing a recognisable face, and simply did not. So this is a rule, not a limitation. Something about depicting a human being straight on was either uninteresting, unnecessary, or not allowed. And the exceptions to the rule, the human figures that do get attention and detail, are almost always the ones that are partly animal. Now, for most of the 20th century, this entire discussion was a European one, which produced a very convenient story.
107:15
symbolic imagery blossoms in ice age europe therefore something special happened in europe therefore the human revolution has a postcode that story survived on the strength of the evidence available and the fact that european caves had been studied for a century and other regions had not it has now collapsed and the site of the collapse is indonesia on the island of sulawesi there are limestone cast towers riddled with caves and shelters and in them are paintings that local people have known about for generations and that researchers long assumed were relatively recent because the tropics are hard on pigment and nothing there looked as old as europe Then teams began applying uranium series dating to the thin layers of calcite, sometimes called cave popcorn, that slowly grow over painted surfaces.
108:01
If a mineral crust has formed on top of a painting, the painting is older than the crust and the crust can be dated. The results rearranged the field. A hand stencil turned out to be at least around 40,000 years old. A painting of a wild pig came in at around 45,000. And in 2019, a team including Maxime Aubert and Adam Brum published a panel from a site called Liangbulu Supong 4, dated to a minimum of roughly 43,900 years ago that is not a single figure at all. It is a scene and a busy one.
108:33
Several wild animals, identifiable as Sulawesi warty pigs and small buffalo called a Noah, are depicted along with a group of considerably smaller figures approaching them, apparently carrying long objects that look like spears or ropes. The small figures are roughly human in shape, but several of them have additions. One appears to have a tail, another has an elongated snout or beak-like face. These are not people wearing hats. The published interpretation is that they are therianthropes, part human and part animal.
109:03
If that reading holds, then the oldest known narrative scene in the world, older than anything comparable in Europe, is a story involving beings that do not exist. And in 2024 the same research programme went further, publishing a scene from another Sulawesi cave using a refined dating method applied at much finer resolution, with an age estimate exceeding 51,000 years, again showing multiple figures interacting with an animal.
109:29
There is a detail in that comparison that deserves its own moment because it is easy to skate past.
109:35
Narrative scenes are genuinely rare in Ice Age Europe. The overwhelming majority of European cave imagery consists of individual animals, magnificent but isolated, arranged on a wall without any obvious relationship to one another. Figures interacting, doing something to each other in a way that implies a sequence of events, appear in only a handful of places, and the shaft at Lascaux is famous precisely because it is such an exception.
110:01
Sulawesi hands us the opposite. The oldest images there are scenes, multiple actors, oriented toward one another, engaged in a shared action. Whatever the dating debate eventually settles on, the composition itself is not in question, and a composition is a sentence rather than a word. Somebody was not recording what an animal looks like, they were recording what happened, or what happens, or what should happen.
110:26
That is the difference between an illustration and a story, and a story is the form every religious tradition in history has been transmitted in. I want to be honest about the limits of these dates, because they are doing a lot of work. Uranium series dating of calcite gives a minimum age for what lies beneath, and it depends on assumptions about whether the mineral system stayed closed over that span, meaning nothing leached in or out.
110:50
Critics have raised exactly that objection, and the debate is live.
110:54
The newer laser-based method addresses part of the concern by sampling extremely thin layers rather than averaging across a thick crust, which is a genuine improvement. Still, treat these figures as strong minimum estimates from a young and rapidly improving technique rather than as fixed points. But even at the conservative end, the implication is enormous. Sophisticated narrative imagery, including composite beings, appears at both ends of Eurasia at roughly the same time, in populations separated by the entire landmass of Asia and by thousands of years of separate history.
111:28
Nobody carried the idea from France to Sulawesi. There was no correspondence, which leaves two possible explanations and both are interesting. Either the capacity is old enough that it travelled with our species as it dispersed out of Africa, packed in the luggage alongside language and fire, or it emerges reliably wherever you put modern human minds in social groups with time and pigment, which would make it less an invention than an inevitability. Either way, the notion that any one region invented the symbolic imagination is finished, and it deserved to be.
112:01
There is a further frontier in this discussion that is currently generating a great deal of heat, and it concerns dates even older than the Sulawesi figures. In 2018, a team applying the same calcite dating approach to marks in three Spanish caves reported ages in excess of 64,000 years. The marks in question are not animals. They're red lines, a ladder-shaped design, a hand stencil, and areas of applied pigment.
112:28
The significance of the number is that at that date our species was not established in Western Europe. If the ages are right, the makers were Neanderthals. The response was immediate and vigorous. Specialists challenged the sampling, arguing that the mineral crusts in question are prone to exactly the open-system behaviour that can make calcite look far older than it is, that contamination from underlying rock can distort results, and that the sample material might not have any reliable relationship with the pigment beneath it.
112:58
The original team defended the analysis and published further work.
113:02
As of now, the claim is neither established nor dead. It is a useful case because it demonstrates that the rule we set earlier is not a museum piece. A single dating study making an extraordinary claim gets treated exactly as it should be, with intense scrutiny, by people who would be delighted if it were true and are not willing to accept it until the method is bulletproof.
113:24
If the result survives, the field will absorb it. If it does not, nothing else in this chapter changes. So let us deal properly with these hybrid beings, because they are the most philosophically loaded objects in the Paleolithic.
113:37
The best preserved example is not painted at all. It is carved, it comes from a cave in the Swabian Jura in Germany, and it is the object I mentioned earlier as not corresponding to anything you could walk outside and photograph.
113:50
Found in fragments in 1939 and painstakingly reassembled over decades from hundreds of ivory pieces, it stands about 31 cm tall and is around 40,000 years old. The body is human, upright, with human legs and arms held against the sides. The head is unmistakably that of a lion, with the muzzle, ears and jaw carefully modelled.
114:12
It has been estimated through experimental replication with period-appropriate flint tools that carving it took something in the region of hundreds of hours, in an ice age environment, by a person who could have been doing almost anything else, including acquiring food or sleeping. Now think about what has to be true inside a mind for that object to exist. To draw a horse, you need memory and skill. You have seen horses. The image is a compressed record of an observation, and while doing it well is difficult, the content of the thought is simply a horse.
114:44
To carve a lion-headed person, you need something categorically different. You need to hold two incompatible categories in mind at once, animal and human, deliberately violate the boundary between them, construct a coherent form that has never existed anywhere in the observable world, and then commit hundreds of hours of labor to giving it a physical body.
115:05
That is the ability to model a reality that is not merely absent but impossible.
115:10
And it is exactly the ability required for every subsequent development in this story. A being that cannot be seen, that lives beyond the world, that has a name and preferences and a history, is precisely this operation applied to something even less tangible than a lion. There is a further requirement that is easy to overlook. Making one is a private act of imagination. Making one that other people recognize is a social fact. Composite beings recur across sites, across regions, and across tens of thousands of years, which means they were not one person's odd dream.
115:44
They were shared. There was a convention about what these beings looked like, and presumably about who they were and what they did. The image is a physical index to a story that everybody in the group already knew, in the same way a modern religious symbol carries an enormous amount of narrative that is nowhere present in the shape itself.
116:03
And notice the structure of these figures. They are almost entirely ordinary, with one rule broken. a human body with an animal head, a person with a tail. This is exactly the shape of concept we discussed earlier as being unusually memorable and unusually easy to transmit, now rendered in ivory and charcoal. Whether or not anyone was consciously optimising for that, the surviving imagery has the profile you would predict. There is one famous therianthrope that comes with a warning attached, and this video would be incomplete without it.
116:36
In a cave system in the French Pyrenees known as Trois Frères, high on a wall above a chamber packed with engraved animals, there is a figure usually called the Sorcerer. In the version everyone has seen, it has antlers, an owl-like face with large staring eyes, the ears of a stag, the forelimbs of an animal, a tail and human legs in a posture suggesting movement.
116:58
It is one of the most reproduced images in the study of prehistory and it has anchored a century of argument about masked ritual specialists. The complication is that the famous version is a drawing. It was produced by the priest and prehistorian Henri Bray, who dominated this field for decades and who published his careful renderings of engravings that are, in the actual cave, faint, superimposed on other lines and located in a spot that is genuinely difficult to see properly.
117:25
Later researchers who examined the panel and the available photographs raised an awkward point.
117:31
Some of the features that make the image so compelling, particularly around the eyes and parts of the antlers, are hard to verify on the wall itself.
117:38
Defenders reply that Bruy worked in the cave for extended periods with lamps at multiple angles and could see relationships in the engraving that a photograph flattens, which is a fair argument since engraved lines genuinely do read differently under raking light.
117:53
The result is an image that is probably real in outline and possibly enhanced in detail, sitting at the foundation of an entire interpretive tradition. I am not telling you this to discredit anything. The Sulawesian-Swabian evidence carries the argument for composite beings perfectly well on its own. I am telling you because it is the same lesson as the rodent pantry and the chewed bear bone arriving from a third direction.
118:18
Our evidence about prehistoric belief passes through modern hands, and modern hands have preferences. A researcher who expects to find a masked ritual specialist will see one more readily than a researcher who does not, and nobody involved has to be dishonest for that to happen.
118:34
Now we come to the layer of this evidence that I find most persuasive, and it has nothing to do with what is on the walls. It is about where the walls are. If you were making images for an audience you would put them where the audience is. Paleolithic people spent most of their lives in the open, in rock shelters and at cave entrances where there is light and air flow and room to work. That is where the domestic debris is, hearths, butchered bone, tool production, all the accumulated garbage of daily living. It is the obvious place for decoration. The images are overwhelmingly not there.
119:07
They are deep, sometimes hundreds of metres in, past sections where the passage narrows enough that an adult has to crawl.
119:15
At Nyo, the principal painted chamber lies the better part of a kilometre from the entrance. Some decorated spaces can only be reached through squeezes that require removing equipment and going through flat.
119:26
Several caves contain art in places where the artist could not have stood upright or stepped back far enough to see their own work as a whole. To get there at all, you need portable fire and here the archaeology is beautifully specific.
119:40
excavations of recovered stone lamps, shallow bowls hollowed into sandstone or limestone, which burned rendered animal fat with a wick of vegetable fibre or moss. A famous example from Lascaux is carefully shaped with a carved handle. An analysis of the residue in similar lamps has identified both the fat and traces of the plant material used as wicking.
120:00
Experiments with replicas show they produce a small, steady, low flame that does not produce much smoke, which matters enormously in a confined space where you would prefer to keep breathing.
120:11
Torches were used as well, and they left signatures. Walls in several caves carry dark streaks where a burning torch was deliberately raked against the stone to knock off the charred end and revive the flame, the Ice Age equivalent of tapping ash off a cigarette. Charcoal fragments from those torches have been recovered from cave floors, some of them dateable, which occasionally lets researchers reconstruct a route and a sequence of visits. The pigments themselves add another layer of planning. Analysis of paint from different panels has shown that the mixtures were not uniform.
120:42
Different recipes appear in different parts of the same cave, combining colouring minerals with various extenders and ground materials, which means batches were prepared separately, probably on separate occasions, possibly by different hands. Some of the material was brought in from outside the cave entirely, so the sequence runs, locate and collect the raw material, prepare the mixture to a specific formula, transport it into the dark along with the means of making light and apply it in a place where nobody will encounter it by accident. Add up the logistics.
121:14
Someone had to render fat, prepare wicks, carry lamps or torches with a limited burn time, navigate passages in darkness so complete that your eyes never adjust, remember the way out, and then perform detailed technical work by a flickering flame while lying on their side in a cold space, with the nearest daylight a considerable distance away. It is worth trying to feel the conditions rather than just list them, because the list undersells it. A deep cave is not merely dark, it is a kind of dark that has no equivalent on the surface.
121:46
There is no ambient glow to adapt to, no faint outline of your own hand, no gradual return of vision after a few minutes.
121:53
If the flame goes out, the visual system simply has nothing to work with, and people who have experienced it describe a rapid loss of any sense of direction or scale. The temperature is constant and cool, the air is still and the only sounds are the ones you make and the occasional movement of water.
122:10
Now put a person in that environment for an extended period by a small unsteady flame surrounded by images of enormous animals that appear and disappear as the light shifts across the rock. The technique of using natural relief mentioned earlier does something specific under those conditions. A moving flame makes the modelled shapes swell and recede, so the figures do not hold still.
122:33
Researchers who have worked in these caves with reconstructed lamps consistently report the same impression that the animals appear to move, Whether that effect was deliberate is unprovable. That it occurred is simply physics. Nobody does that for decoration. Nobody does it for an audience either, because there is no audience down there and there cannot be more than a handful of people in most of these spaces at once, which reframes the entire question. A place that requires effort, equipment, knowledge and courage to enter, and that cannot accommodate a crowd, is not a gallery.
123:05
It is a restricted venue. The difficulty is not an inconvenience to be tolerated, it is a feature that determines who gets in. Everything about the arrangement says that access was limited, and the limitation was the point. And before anyone objects that these people simply had nowhere else to put an image, they did.
123:23
The same period produced a large body of portable art, small carved figures in ivory, bone and stone, animals rendered on spear throwers and tool handles, and flat stone plaques engraved with images that could be carried, handled, and passed around.
123:39
One cave in France produced hundreds of engraved plaquettes. These objects circulated in daylight, in the living areas, among everybody. So image-making was not confined to the deep zone, which makes the deep zone a choice rather than a necessity. Certain images, in certain caves, at certain distances from daylight, were placed where almost no one would see them. The portable pieces were the public version. What happened underground was something else, and the two coexisted, made by the same communities in the same centuries.
124:11
There is evidence about who did go in, and some of it is quietly astonishing. Several caves preserve human footprints in soft clay floors, undisturbed for tens of thousands of years, and a number of them belong to children and adolescents.
124:25
at one site there are the heel impressions of young people in a chamber containing modelled clay-bison at chauvet a trail of footprints made by a child runs alongside the prints of a large canid though whether the two were made at the same moment is disputed and i am obliged to tell you that even though the version where they walk together is better Children deep in decorated caves is not what you would expect from an artist's workshop. It is exactly what you would expect from a ceremony that involves bringing the young to a place they have never been, by lamplight, into the dark, to be shown something.
124:58
The other detail that changes how you read these spaces is time. Many panels are not single compositions but accumulations. Images are drawn over earlier images, sometimes centuries or millennia apart, sometimes carefully positioned to incorporate what was already there, sometimes simply crossing it.
125:16
Certain caves were visited, decorated, abandoned and returned to across spans that dwarf the entire history of any modern institution.
125:24
A wall might hold work separated by more time than separates us from the invention of writing.
125:29
That pattern tells you the place itself held status, independent of any individual image on it. People were not coming to admire what an ancestor made, they were coming to do the thing that is done here, and adding to it was part of the doing.
125:43
That is the behaviour of a site, not of a collection.
125:47
I should note that not everyone reads this material the same way and the alternatives are serious. There is a long tradition that approached cave art structurally, cataloguing which species appear in which parts of caves and arguing that the images form an organized system with consistent spatial rules, essentially a grammar rather than a set of illustrations.
126:10
That approach produced real insight and does not require any assumption about ritual specialists.
126:16
Others have emphasized initiation, group identity, or information storage about animal behavior and seasonality. And there remains a perfectly respectable minimal position, which is that we can describe the patterns and should be extremely modest about naming their purpose. What nobody seriously argues anymore is that this was idle decoration made by people with spare time. The costs are too high, the locations too deliberate, the conventions too consistent, and the beings too impossible. There is one more category of image that deserves a mention because it is the most human thing in any of these caves.
126:51
Across the world, from France to Spain to Argentina to Sulawesi, people put a hand against the rock and blue pigment around it, leaving a stencil. The technique is identical everywhere, which is unsurprising given that there are not many ways to do it. The result is a negative image, the one part of the wall that stayed clean. Studies of hand size and finger proportions have suggested that a substantial share of these stencils were made by women, which contradicted a century of casual assumption about who was doing this, though the method has been challenged and the question is not closed.
127:23
In some caves, many stencils appear to show missing finger segments, which has generated a range of explanations including deliberate bending of fingers to make signs, cultural practices of alteration and cold injury, with no agreement in sight. And the stencils include small hands, children's hands pressed against a wall deep underground with an adult presumably holding the pigment.
127:46
Whatever these caves were for, children were brought into them and someone helped them leave a mark that is still there fifty centuries after every other trace of their existence disappeared. That is not a hunting technique, that is membership. So the picture at the end of this chapter is considerably sharper than at the start. by at least 40,000 years ago and probably earlier, people in widely separated parts of the world were producing images of beings that do not exist, according to shared conventions that other members of their community could read.
128:18
They were taking these activities into spaces that required fire, planning and physical difficulty to reach, spaces that could hold only a few people at a time and that had no domestic use whatsoever. They were returning to the same spaces across enormous stretches of time. They were bringing children. The invisible had been given a form, the form had been agreed on, and there was now a place where the agreed thing was done, and there is a property of those places that the people using them could not have failed to notice, because it announces itself the moment anyone speaks.
128:49
A deep limestone chamber does something to sound. It answers you. It makes a single voice into several, holds a note after you have stopped, and in certain specific spots it makes the stone itself seem to reply. which raises a question somebody eventually thought to test. If the images are in the deepest parts of the caves, are they in the loudest ones?
129:10
The person who thought to test it was a French musicologist named Igor Resnikov, and his method was refreshingly low-tech. He went into decorated caves and he sang. Working with the archaeologist Michel Dauvoir in sites including Arcy Secure and the deep chambers of Nioh, Reznikov moved through the passages in stages, stopping every metre or two to produce a sustained tone with his voice and noting how the cave responded. Some spots swallow sound almost completely, others return it immediately and flatly, like speaking into a cupboard, and in certain specific locations the chamber comes alive, holding the note after you stop, multiplying a single voice into what sounds like several, or throwing the sound back from a direction you did not send it.
129:54
Then he mapped those acoustic hotspots against the locations of the images. The correlation he reported was strong. Painted and engraved figures cluster disproportionately at or near points of maximum resonance. Deadened areas, where sound goes to die, are comparatively bare. And most intriguingly, in several places where the acoustics are remarkable but the rock offers no decent surface to work on, there are still marks. not animals, just small red dots or short strokes, as if somebody needed to record that this particular spot was significant even though a picture was not practical here.
130:28
The suggestion is that people were locating something by ear and then marking it by hand. If that is right, it changes the sequence of events in a rather satisfying way. The conventional picture has someone entering the cave, finding a good flat surface and painting on it. The acoustic picture has someone entering the cave, listening their way through the dark, finding the place where the stone answers back and marking that place. The wall is not the reason. The wall is the address. It is worth spending a moment on what these effects would have meant to someone with no theory of acoustics, because we have the enormous advantage of knowing why a cave echoes and they did not.
131:04
A reflected sound is, physically, the same sound arriving late, but it does not present itself that way to the ear.
131:12
A voice returned from a curved stone surface is altered, usually lower in the higher frequencies, often slightly blurred, and it arrives from a direction that has nothing to do with where you are facing.
131:23
In several caves, the reflection off a distant wall comes back as a low rumble that people have compared to the sound of large animals moving, which in a chamber decorated with bison and mammoths is not a neutral coincidence. So a person stands in the dark, makes a sound, and something in the rock replies in a voice that is not quite theirs, from a direction where nobody is standing. Now apply the machinery we discussed earlier, the mind that assigns an agent to any ambiguous event. There is only one available conclusion and it is not a lecture on reflection coefficients.
131:55
Something down here heard you and responded. That is not superstition. That is the correct output of a normal brain given the available data. There is one further piece of evidence in these caves that requires no interpretation of resonance at all, because it is physical damage.
132:10
In a number of decorated caves, the natural mineral formations, the columns and curtains of calcite that grow down from ceilings and up from floors carry marks of repeated striking. These formations ring when hit. A thin sheet of calcite produces a clear sustained tone and different thicknesses produce different pitches, so a single formation can function as a set of tuned bars. At several sites, researchers have identified percussion marks concentrated on exactly the parts of a formation that sound best, sometimes with pigment traces around them and sometimes with painted marks adjacent.
132:44
Formations that produce a dull thud tend not to be marked. These are usually called lithophones, and they are the oldest surviving fixed musical instruments in the world by a considerable margin, in the sense that they were not made at all, only selected and used. So people went into the deep zone, located the parts of the rock that produced the best tone and struck them repeatedly enough to leave visible wear on stone.
133:08
Whatever else was happening down there, it was not quiet.
133:12
Now, let us do the obligatory damage to this idea, because it is popular enough to need it. The first problem is that caves are not static. Floor levels in many decorated caves have changed substantially, sometimes by metres, through sediment accumulation or through excavation by early researchers who were not thinking about acoustics while they removed deposits. Blocks have fallen. Passages have been widened for modern access. Measuring resonance today in a space that has been physically altered tells you about the space as it now exists, and the relationship between that and the acoustics of 40,000 years ago is an estimate at best.
133:50
The second problem is confounding.
133:52
Resonant spots tend to be in wider chambers, at narrowings between spaces and at prominent features, because those are the shapes that produce interesting sound. Those same locations are also where a person would naturally stop, where the passage opens up enough to work, and where a wall is visible and prominent. So a correlation between images and resonance might reflect nothing more than the fact that both are attracted to the same architectural features for entirely separate reasons.
134:19
The third problem is method. Much of the early work relied on a trained human voice and a trained human ear rather than on instruments, which is not nothing, since Reznikov is a specialist in vocal acoustics, but it is not a measurement anyone else can precisely reproduce.
134:35
Later studies have used proper acoustic equipment and impulse response analysis, and the results have been mixed, with some supporting a relationship and others finding it weaker than claimed. So where does that leave us? With something genuinely interesting and not proven. The honest position is that the resonance hypothesis is plausible, partially supported and impossible to establish with certainty. What is certain, and requires no statistics at all, is the underlying physical fact. These are enclosed stone spaces with hard, uneven surfaces, and they do dramatic things to sound.
135:09
Anyone who went into one and made a noise found that out within the first ten seconds. you do not need a research programme to notice that the cave is answering you. The same caution, only more so, applies to the most famous claim in this entire area, which concerns an underground complex on Malta.
135:26
The Hippogym of Halsaflieni is a multi-level structure cut directly down into soft limestone, built and expanded over centuries in the 4th and 3rd millennia before the Common Era. It was not a natural cave. People carved out chambers, passages and staircases with stone tools and in several places they carved imitations of above-ground architecture into the rock, reproducing lintels and corbelled ceilings in a place where no structural support was needed at all.
135:52
Somebody went to considerable trouble to make an underground void look like a building. One small chamber has acquired a reputation. A niche cut into the wall of what is usually called the oracle chamber produces a striking effect. A low male voice spoken into it resonates powerfully and seems to fill the whole complex, and the sound quality is noticeably different from anywhere else in the structure.
136:15
Measurements published by acoustic researchers identified a strong resonance in the neighbourhood of 110 Hz, which sits comfortably in the range of an adult male voice or a chanted tone. Then the claim escaped into the wild. You will find a great deal of material asserting that this specific frequency was deliberately engineered. that it produces distinctive changes in brain activity, that it shifts language processing between hemispheres, and that the builders therefore possessed sophisticated knowledge of neuroacoustics. Here is what can actually be supported.
136:48
The resonance is real and measurable. Small chambers of that size in that material will resonate somewhere in that region as a straightforward consequence of their dimensions, which means the frequency may be a by-product of the room rather than a target of the design.
137:02
The studies claiming distinctive neurological effects at that exact frequency are limited, have small samples, and have not been convincingly replicated. And the leap from a measurement in a room to a theory about the intentions of people 5,000 years ago is exactly the kind of leap this video keeps warning you about. What survives is still worth having. People dug a large complex into rock by hand, made it resemble a building they did not need to imitate, and used it for activities in a space that does unusual things to the human voice. Whether they tuned it on purpose or simply noticed what they had built and liked it, they were working underground with sound.
137:39
For a case where the engineering is not in doubt, we have to travel to the Andes and to a site that has the distinction of being the closest thing prehistory offers to a purpose-built experience. Chavin de Huantar sits high in the Peruvian highlands at the meeting of two rivers, in a valley that connects the coast, the mountains, and the forested lowlands. It flourished roughly between 1200 and 500 years before the Common Era, well before the Inca, and before any state in the region, and it appears to have been a destination.
138:08
people came to it from considerable distances bringing materials and styles from multiple ecological zones which is the archaeological signature of a pilgrimage centre rather than a town from the outside it is a set of massive stone platforms with sunken plazas What makes it extraordinary is what is inside them. The platforms are not solid. They are honeycombed with narrow stone-lined galleries, corridors barely wide enough for one person, running on multiple levels, connected by staircases and by shafts.
138:39
There is no natural light in any of them. Some corridors open into small chambers, some connect to other corridors, some do not obviously go anywhere useful at all.
138:49
The architects built in ventilation ducts, narrow rectangular channels cut through the masonry, which served the practical purpose of keeping air moving in a sealed stone labyrinth.
138:59
They also built an extensive system of water channels running beneath and through the structure, fed from the rivers, engineered with drops and turns.
139:08
Both systems carry sound. A duct connects galleries acoustically as well as pneumatically, which means a noise produced in one corridor emerges somewhere else entirely, arriving from a wall or from a hole in the floor rather than from anything a person can see. At the heart of the oldest section, at the intersection of a cross-shaped gallery, stands a carved granite monolith over four metres tall, a tapering shaft with a relief figure that combines human and feline features with serpents and fangs. It is wedged into a space that seems designed around it, in near-total darkness, reachable only by walking a considerable distance through the corridors.
139:46
There is a structural detail around that monolith that is hard to read innocently. The chamber holding it is not a simple dead end. There is a shaft above it connecting to a gallery on the level overhead, which means the space directly above the carved figure is occupied, accessible and acoustically linked to the chamber below.
140:04
Whether that arrangement was used to deliver sound from above the figure is not something anybody can prove, but the plumbing is there and it did not install itself by accident. It is also worth noting how few people the galleries can hold. These are single-file corridors. Whatever happened inside, it happened to individuals or to very small groups, in sequence, while everyone else waited outside in the plazas. That is not a mass service. That is an appointment system, and one that inherently produces two categories of person, those who have been inside and those who have not.
140:39
In 2001, excavation of one gallery produced 20 conch-shell trumpets. The shells are a large marine species that does not come from anywhere near Chavin, which means they were carried up from the coast. They had been modified into working instruments with a cut mouthpiece, several were carved with the site's characteristic imagery, and they are still playable. Instruments of this type are known in the Andes as pututus, and they produce a loud, penetrating, slightly rough tone. A research program led by Miriam Kohler has studied the acoustics of these galleries systematically, using the actual shells alongside modern measurement techniques, and the results are specific.
141:17
The corridors do not simply echo, they alter the directional cues a listener uses to locate a sound source, so a person standing in a gallery has serious difficulty telling where a tone is coming from.
141:28
In some configurations the sound seems to arrive from everywhere or from the wrong side. The architecture also filters certain frequencies while reinforcing others, changing the character of the instrument depending on where it is played and where it is heard.
141:42
Now, assemble the visitor experience. A person has travelled for days or weeks to reach a site they have heard about their entire life. They are brought into a stone corridor with no light, underground, in a building whose interior they cannot see and whose layout they do not know. Water is moving somewhere beneath their feet, producing a continuous low roar with no visible source. A trumpet sounds, and it is not possible to say from which direction. They are led through turns they cannot map. and at some point in the dark they are brought in front of a carved figure that is part person and part predator.
142:18
Archaeologists who work on this site do not describe it as a temple with good acoustics. They describe it as an instrument for managing perception, and I think that is exactly right.
142:29
Every element that would help a person stay oriented has been removed, and every element that undermines orientation has been installed. This is deliberate, it is expensive, and it was built by people who clearly understood what they were doing to the human nervous system, even if they would have described it in entirely different language. And they had chemical assistance, which brings us to the next layer. The stone art at Chavine includes a relief of a figure holding an upright columnar cactus, and the species depicted is recognisable. It is the plant commonly known as San Pedro, which is native to the Andes and contains psychoactive alkaloids, and it has a documented history of ceremonial use in the region that continues into recent times.
143:10
The same iconographic tradition at the site is full of imagery that reads as transformation, with human features merging into those of jaguars, birds of prey and serpents, and with figures whose faces show the sort of physiological effects associated with certain plant preparations.
143:27
Excavations have also recovered the material apparatus. Carved bone tubes and small trays of the kind used across the Andes for administering powdered plant preparations nasally appear in the archaeological record at Chavine and at related sites. Residue analysis on similar equipment elsewhere in the Andes has identified compounds from specific plants, so the identification of the practice is not merely inferred from shape.
143:52
the same pattern of ritual pharmacology appears on other continents and in other periods. In Southern California, the Chumash and neighboring peoples used preparations of a plant in the nightshade family, known there as toloch, in initiation ceremonies for young people under the supervision of experienced practitioners, and in 2020, researchers working at a painted rock shelter in that region published a striking find.
144:16
wads of chewed plant fibre, tucked into crevices in the ceiling, which analysis identified as coming from that same plant. Directly beside them, painted on the rock, is a design that unfurls in a spiral, which the researchers noted closely resembles the way the flower of that plant opens. That is an unusually direct association between a specific substance, a specific image and a specific place, and it is worth stating what it does and does not show.
144:45
It does not show that all rock art everywhere was produced under such influence. It shows that at one site, in one culture, the substance and the imagery and the location are found together, exactly as the ethnographic record for that region describes.
144:59
I should be clear and unambiguous here. The plants involved in these traditions are seriously toxic, several are potentially lethal, and the ceremonies described always involved extensive cultural knowledge, specific preparation and experience supervision, which is precisely why they were restricted practices rather than casual ones. This is a historical account of what people did, not a suggestion about what anyone should do. Alcohol belongs in this section too, and it shows up early. Chemical analysis of residues absorbed into the pottery of a Neolithic village in China identified a fermented beverage made from a combination of ingredients including fruit, honey and grain, dating to around 7,000 years before the Common Era.
145:42
In Western Asia, early Neolithic sites include large carved stone vessels that some researchers have argued held fermented liquid, based on chemical traces, though the identification is contested and we will come back to that region later. Fermentation is not a difficult technology to discover. It discovers itself repeatedly in any container of sweet liquid left alone for a few days, and humans appear to have noticed extremely quickly. What all of these have in common is not intoxication for its own sake. In the context we are discussing, the substances are restricted to particular occasions, particular places, and often particular people.
146:18
They are part of an assembled experience rather than the whole of it. Which brings us to the most ambitious attempt to connect all of this to the images themselves and to a theory that has generated 40 years of argument.
146:30
The South African archaeologist David Lewis Williams built his model on two foundations. The first was laboratory and clinical work on the visual effects reported by people in various altered states, which includes a consistent catalogue of geometric forms, grids and lattices, sets of parallel lines, dots and flecks, zigzags, nested curves, and spirals or tunnel shapes. These are called entoptic phenomena, meaning they originate within the visual system rather than in the outside world, which is why they are reported by people from any culture regardless of what they expect to see.
147:06
The second foundation was ethnography, particularly the detailed accounts of sand rock art and its relationship to healing dance traditions in southern Africa. where the connection between the imagery and the practice is documented rather than guessed at. From these he proposed a three-stage model. In the first stage, the geometric forms appear. In the second, the mind begins interpreting those forms as objects, so a set of nested curves becomes a hill, a zigzag becomes a snake, and the content is supplied by the individual's own cultural repertoire.
147:38
in the third and deepest stage the person experiences a sense of moving through a tunnel or vortex and emerging into a fully realized environment populated by beings frequently including a sensation of merging with an animal then he pointed at the walls Paleolithic cave art is full of geometric signs and they have never been satisfactorily explained. There are grids, rows of dots, zigzags, nested curves and spirals across European caves and elsewhere, sitting alongside the animals and usually ignored by everyone who prefers looking at the horses.
148:11
Lewis Williams argued that the same catalogue appears because the same visual system generated it, and that the therianthropic figures correspond to the deeper stage, the point at which the boundary between person and animal dissolves. It is an elegant theory and it explains a great deal in one move, which is precisely why you should hold it at arm's length.
148:30
The central criticism is that it is very difficult to falsify. The catalogue of geometric forms is broad enough that almost any abstract mark can be assigned to it. A grid is entoptic. A zigzag is entoptic. A row of dots is entoptic.
148:45
but grids, zigzags and rows of dots are also the most basic marks a human hand can make and they appear on basketry, textiles, body decoration, tally systems and the doodles of bored office workers, none of which require an altered state.
149:01
If every abstract sign confirms the theory and no abstract sign could contradict it, the theory is not doing the work it appears to be doing. There are further objections. The three-stage sequence derives largely from studies conducted within particular research traditions and may be less universal than presented.
149:19
The extension of a specific Southern African ethnographic situation, where the link is genuinely documented to Ice Age Europe, where nothing is documented, is an enormous analogical stretch. and there is a tendency in the wider literature to treat any ambiguous prehistoric behaviour as evidence of trance, which flattens an enormous range of possible human activity into one explanation.
149:43
Lewis Williams and his supporters have responded to much of this, and the debate has produced better work on all sides. The reasonable summary is that altered states were almost certainly part of the ritual repertoire of many ancient societies, since they are part of the ritual repertoire of nearly every society we can observe directly, but that the specific claim of decoding cave imagery through a fixed neuropsychological sequence remains unproven and probably unprovable. The final component in the toolkit is the one that requires no substance at all, only a stick and patience.
150:15
Rhythmic drumming is a near-universal feature of ceremonial practice and there is a long-standing idea that it works by driving brain activity toward a particular frequency band.
150:25
The proposal, going back to experimental work in the early 1960s, is that a steady repetitive beat in roughly the four to eight cycles per second range encourages activity in the theta range, which is associated with drowsy, inwardly focused states on the border of sleep, and that sustained exposure can shift a person toward trance. The idea is appealing, and the evidence is softer than the popular version suggests. Studies using modern recording techniques have found effects that are real but generally modest, and the results vary considerably between individuals.
150:59
More importantly, several studies suggest that what matters most is not the frequency of the beat but the context surrounding it – expectation, prior training, cultural familiarity with what is supposed to happen, the presence of a group, physical movement, and the state of the body. A person who has been dancing for four hours in a hot enclosed space without food will reach an unusual state regardless of the precise tempo. That last point is the one I want to leave you with because it corrects a mistake that runs through most popular treatments of this subject.
151:31
There is no single lever. Nobody in the ancient world was pulling one switch labelled trance. What these practices did was stack conditions and the stack is what does the work. Consider what is actually being combined. Darkness, which removes visual reference and induces the mind to generate imagery on its own. A phenomenon documented in modern sensory reduction experiments where people in featureless environments begin to see shapes, movement and eventually complex scenes within a remarkably short time. enclosure, which removes spatial orientation and the ability to leave, sound that cannot be localised, which defeats one of the most fundamental systems the brain uses to model its surroundings, repetitive rhythm, sustained well past the point where attention would normally wander, physical exertion, often for hours, fasting or restricted food beforehand which is a standard feature of ceremonial preparation across the world, sleep deprivation which on its own reliably produces perceptual disturbances, heat, smoke, group synchrony which as we have seen alters the body's chemistry, and in some traditions a plant preparation on top of all of it.
152:41
Every one of those conditions produces a measurable effect on perception in isolation. Applied together, deliberately, to a person who has been told what to expect, they produce something reliable enough that entire institutions were built on it, and the ingredient that binds the whole stack together costs nothing, which is expectation. A participant does not arrive as a blank instrument. They arrive having heard about this place, and this night since childhood, having been told what kind of being lives here, what it looks like, what it sounds like, and what it tends to do to people who meet it.
153:15
The mind under sensory reduction does not generate random content. It generates content from what is already loaded, which is why the visions reported in any tradition tend to feature that tradition's cast of characters rather than someone else's.
153:29
This is the part that makes the whole system self-confirming, and it is elegant enough to be slightly unnerving. The stories create the expectation. The expectation shapes the experience. The experience is remembered as an event rather than as a construction, and the report of that event then goes back into the pool of stories as first-hand testimony from someone the community knows and trusts, which strengthens the expectation for the next person. Nobody in this loop is lying, that is the important part. Every participant is accurately reporting what they perceived.
154:01
There is one more mechanism worth mentioning because it is delightfully specific.
154:06
Sound below the threshold of hearing, generally called infrasound, has been associated with reports of unease and of a sense of presence.
154:14
The best-known anecdote involves an engineer who investigated reports of a haunted laboratory and traced the effect to a fan producing a standing wave at around 19 cycles per second, a subsequent concert experiment in which infrasound was covertly added to some pieces of music found that audiences reported more unusual feelings during the affected pieces.
154:34
As with everything else in this chapter, the effects are modest, the studies are limited, and the popular version overstates it considerably.
154:42
But large enclosed stone spaces, moving water, and low resonant tones are exactly the conditions that generate low frequency energy, and it would be strange if it had no effect at all.
154:53
There is a final, practical reason these experiences were staged rather than left to chance, and it has to do with what happens afterwards.
155:01
An unstructured, overwhelming experience is not automatically useful. It can be frightening, disorganising and hard to integrate, and a person left alone with one may simply be damaged by it. What ceremonial contexts supply in every tradition that has been described in detail is the framing around the event – preparation beforehand, supervision during, and interpretation afterwards – in which what happened is explained, named, and fitted into the shared account of the world. That final step is the one that converts a private episode into communal knowledge.
155:34
Without it, you have a person who had a strange night. With it, you have a testimony that belongs to everybody. So what is the conclusion of all this? Ritual was never only about transmitting ideas. If the goal had been to teach a story, you could do that by the fire, in comfort, in daylight, for free. The goal was to produce an experience, and the architecture, the darkness, the sound, the rhythm, the movement, the fasting and in some cases the pharmacology were the means of producing it reliably, on schedule, in a controlled setting, for people who arrived expecting it.
156:09
And the crucial thing about experiences produced this way is how the mind files them afterwards. A thought is something you recognize as your own. You know you constructed it, you can trace how you got there, and it carries the unmistakable signature of internal origin. A perception is different. It arrives unbidden, it has sensory texture, it comes with emotional weight, and it does not feel authored.
156:33
When somebody in a dark chamber hears a voice arriving from the wrong direction, sees the figures on the rock shift as the flame moves, and experiences the presence of something that is not physically there, their mind does not categorize that as an idea they had. It categorizes it as something that happened to them. That distinction is everything. An idea about a spirit is a proposition, and propositions can be doubted, argued with, and forgotten.
156:59
An encounter with a spirit is a memory, and memories are evidence. You do not believe your own experiences, you simply have them, and afterwards you report what occurred. This is how a story told around a fire becomes a fact known from personal knowledge. Not by persuasion, but by staging. The community did not need to convince anyone that the beings on the wall were real. It needed only to build a situation in which people met them and then let ordinary human memory do the rest.
157:27
but a situation like that does not assemble itself. Somebody has to know which passage leads where, which sequence comes in what order, how long the rhythm continues, what quantity of which plant is safe, and what to do when a participant becomes frightened or unwell in a confined space underground.
157:44
somebody has to have done it before more than once successfully which means that at some point in this story the experience stops being something a community produces collectively and starts being something a particular person provides and when a society begins investing in a specialist the specialist starts showing up in the ground In the hills of the Western Galilee there is a small cave set high in a cliff above a stream gorge, difficult to reach, with no particular domestic advantages. It is called Hillazontachtit, and around 12,000 years ago a community carried their dead up to it.
158:19
excavations led by Leo Grossman of the Hebrew University recovered the remains of at least 28 individuals, most of them treated in the ordinary manner of the period. One of them was not.
158:31
The publication describing that burial appeared in 2008, and it remains the single most persuasive piece of evidence that ritual specialists existed before agriculture, before writing, before anything we would call a society with institutions. Start with the grave itself, because the container tells you as much as the contents. This was not a scrape in loose sediment. Somebody cut an oval pit into the bedrock floor of the cave, then lined the walls and the floor of that pit with limestone slabs, essentially building a small stone chamber.
159:02
The body was placed inside and then wedged in position with additional stones packed against the head, the pelvis and along the sides, holding the posture in place. A large flat slab was set over the top to seal it. Constructing this in limestone with the tools available was a serious piece of work and every stage of it was done for a person who was already dead.
159:24
It is worth knowing who these people were because the timing is not incidental. The culture this burial belongs to is called the Natufian, and it occupied the eastern Mediterranean coast and its hinterland in the closing millennia of the Ice Age. The Natufians were foragers. They did not farm, they had no domesticated animals apart from a relationship with dogs, and their diet came from wild plants and wild game. but they were doing something no forager population had done before at this scale – staying put. Natufian sites include structures with stone foundations, built in clusters, occupied across seasons and in some cases across generations.
160:02
They include heavy equipment, stone mortars and grinding installations that are far too large to carry and therefore announce that nobody intended to carry them. They include blades with the characteristic sheen that comes from cutting quantities of wild cereal stems along with storage features.
160:18
The landscape they lived in was, for a period, generous enough to support permanent residents without agriculture. And the most relevant innovation of all, they buried their dead in the same places repeatedly in what we can reasonably call cemeteries.
160:31
That combination is the hinge of this entire section. A mobile group leaves its dead scattered across a territory it passes through. A settled group accumulates them in a fixed location, which means the dead become a population with a location and the location is one the living return to deliberately. It also explains why the resources for an event like this were available at all. A feast requires a concentration of food and a concentration of people in one place at one time, and both of those are difficult when everybody is moving in small bands across a wide territory.
161:04
Sedentism makes ceremony affordable. I should add that the Natufian period was not a comfortable one throughout. It spans a sharp climatic downturn, a return to cold and dry conditions lasting over a thousand years, which put real pressure on the wild resources these settlements depended on. Populations in this region were, during exactly this window, working out how to survive staying still in a landscape that had stopped cooperating. That is the context in which somebody decided to spend 50 tortoises on a funeral. The occupant was a woman, roughly 45 years old, notably small in stature, and her skeleton records a lifelong physical condition.
161:43
There are deformations affecting the spine and the pelvis, of a kind that would have produced a distinctly asymmetrical way of moving. In practical terms, she walked with a pronounced limp, and her body was visibly different from everybody else's, from early life until the end of it. That is worth holding on to, because it is going to come back.
162:02
Now for what went in with her, and I want you to keep a running tally of how useful each item is. Fifty complete tortoise shells. Not fragments, not a handful. Fifty.
162:12
the wing bone of a golden eagle the forearm of a wild boar the pelvis of a leopard the tail bones of an aurochs which is the large wild cattle species of the period the skulls of two martins and resting against the body a complete human foot belonging to an individual other than the woman in the grave nothing on that list is a tool nothing on it is a container a weapon a garment or a supply there is not a single item that a person could use for anything except being that item the human foot deserves its own moment because it is the element that most people skip past and it may be the most informative item on the list a complete foot from a person other than the woman in the grave placed with the body Bones of a hand or foot are among the parts most likely to come apart and go missing when a body decomposes which means a complete articulated foot had to be either taken while the anatomy was still held together or carefully collected and kept.
163:09
Either way, somebody was in possession of part of another human being and considered it appropriate to include. That tells us that human remains in this community were not simply disposed of and forgotten. They could be retained, transported and deployed for a purpose as material with a use. Keep that in mind because it is the small beginning of a practice that within a few thousand years becomes an organised institution and one of the defining features of the world that followed. Consider what acquiring this collection actually required.
163:40
Tortoises do not congregate and are not caught in numbers by chance. Each one is an individual encounter found while walking, picked up, carried back.
163:50
Fifty of them means either sustained collecting over a considerable period or a large number of people sent out to look. The leopard is a different problem entirely. Obtaining part of a large predator in this landscape with the technology of 12,000 years ago is a serious and dangerous undertaking and nobody does it by accident. A golden eagle is not easily obtained either. Martins are small, quick and of no dietary consequence whatsoever, so the set is not a selection of available animals. It is a curated collection of the rare, the dangerous and the difficult, assembled deliberately and then removed from circulation permanently.
164:27
There is one more thing about that inventory that repays attention, and it concerns which parts were chosen. Look at the list again and notice that in almost every case, the item is the part of the animal that carries its defining capability. The wing of a bird that flies higher than anything else in that sky. The pelvis of the fastest and most elusive predator in the region. the forearm of an animal known for its strength and its refusal to yield, the tail of the largest land animal anybody there had ever seen, the shell of a creature that carries its own house and outlives almost everything.
165:02
This is not a random assortment of exotic bits. It is a selection organised around a principle, and the principle appears to be capability rather than species.
165:12
somebody chose the flying part the running part the striking part if you were designing a set of objects to represent a range of powers that a human being does not have you would arrive at something remarkably close to this i am not going to claim that is what it means since we established that intentions do not fossilize but the internal consistency of the set is a fact about the set and it is the kind of fact that makes a random accumulation extremely unlikely The tortoises carry a second piece of information.
165:44
Their remains show the marks of butchery and heating, which means the animals were processed and consumed rather than simply killed for their shells. The same layer contains the remains of an aurochs. Somebody calculated, on the basis of the meat available, that this represents food for a gathering of perhaps 35 or so people, in a period when a local group might not have been much larger than that. which means the funeral was not a disposal. It was an event with attendance, catering and a structure, a feast, a highly elaborate burial, and then the deliberate sealing of the grave, with evidence that the chamber itself was closed off afterwards.
166:19
Try to price this in the currency of the time, a specially constructed stone-lined pit, weeks of collecting, a hunt for a large predator undertaken for a body part rather than for meat or hide, a feast that consumes an enormous quantity of protein in a single occasion, in a world with no refrigeration and no reliable surplus, and a set of objects taken out of the economy forever. You do not spend that on a grandmother because you loved her, you spend it because of what she was to the group, and here is the argument that makes this burial decisive rather than merely spectacular.
166:51
She is in a cemetery. There are more than two dozen other people from the same community, from the same period, in the same cave, and their burials are ordinary. We are not comparing this grave against some imagined baseline. We are comparing it against her own neighbours, buried by the same hands under the same customs. The rite performed for her was unique, and uniqueness within a known population is the strongest signal archaeology can give you about social role. The choice of the cave itself is worth a line, since it is not the sort of place anybody would settle in.
167:23
It sits high in a cliff face above a gorge, awkward to reach, cramped and offering nothing in the way of comfort or convenience. Carrying bodies up there, along with the stone, the animals and everything required for a gathering, was a decision that made the process considerably harder than it needed to be, which is precisely the signature we have been tracking all along. When a location is chosen despite being inconvenient, the inconvenience is usually part of the meaning. Now, before this becomes too tidy, let us give the objections a fair hearing because there are some.
167:56
The first is the wedging. The body was packed in with stones and the grave was sealed with a slab, and one reading of that arrangement is not reverence but containment. Cross-cultural accounts include practices in which particular individuals are buried with weights, coverings or restraints, the underlying concern being that this person should stay where they were put.
168:15
On that reading, the elaborate treatment marks fear rather than honour, and the woman was not a revered figure but a dangerous one. I would point out that this objection does not damage the main conclusion nearly as much as it appears to. Fear and reverence are not opposites in this context, and a person believed capable of causing harm after death is a person believed to have unusual power, which is precisely the category we are trying to identify.
168:41
A community that stages a feast for 35 people and then carefully seals a grave is not indifferent to the occupant either way. The second objection is simpler. Perhaps she was high status for reasons having nothing to do with ritual. The trouble is that the usual routes to status in a forager society, being a productive hunter, a skilled maker or the head of a large family, do not predict this particular inventory. A celebrated hunter gets tools, weapons and parts of the animals that matter to hunting. An important elder might get ornaments and quantity.
169:14
Neither explanation generates 50 tortoises, a predator's pelvis and parts of a bird of prey. The kit does not look like an economy. It looks like a set of references. The third objection is the honest one and it applies to everything in this chapter. We are reading intentions from objects, and objects do not speak. What we can establish is the shape of the behaviour, extraordinary investment, singular treatment, non-utilitarian materials from species chosen for rarity and danger, and a communal gathering.
169:45
What that meant to the people who did it is inference. Now take the same pattern and move it 4,000 kilometres northwest and forward about 3,000 years. In 1934, workers digging for construction near the town of Bad Durenberg in central Germany encountered a burial that turned out to be one of the richest of its period in Europe. The date places it in the Mesolithic, roughly 9,000 years ago, among foragers living after the Ice Age but well before farming reached the region. The grave contained a woman, between 25 and 40 years old, and an infant.
170:18
She was not laid out flat. She was placed in a seated or upright position, which required the grave to be constructed to hold her that way. the burial was saturated with red pigment, to the extent that the surrounding soil is stained. The inventory runs to dozens of items, and it is a familiar kind of list. Sections of red deer antler, including pieces from the skull with the antlers still attached, which is the standard raw material for a headdress, and is found in that role at other sites of the period. bore tusks a large number of animal teeth perforated for attachment bone points worked flint shell and the remains of a turtle bones of cranes and other birds and then there is the detail that made this burial famous a second time nearly 90 years after it was found detailed modern examination of her skeleton identified an anatomical variation in the first vertebra of the neck the ring-shaped bone that supports the skull.
171:16
In her case, the bone shows an unusual form affecting the channel through which a major artery passes on its way to the brain.
171:24
Variations of this general kind exist in modern populations and are usually of no consequence, and I want to be careful not to overstate anything here. The research team raised the possibility explicitly as a hypothesis. that this particular configuration might have made her more susceptible to unusual sensations when the head was tilted far back for an extended period, potentially including visual disturbance or dizziness. In a cultural context where such sensations are interpreted as contact with another order of reality, a person who could produce them reliably, by adopting a posture, would occupy an unusual position indeed.
172:01
I will flag the obvious. This is speculation built on an anatomical observation. It cannot be tested and the researchers themselves presented it cautiously. It has been repeated in popular accounts with far more confidence than the original claim carried. Treat it as an intriguing possibility rather than as an established fact and notice that even if it is entirely wrong, the grave remains extraordinary on the strength of everything else in it. The burial received a comprehensive modern reinvestigation in the years around 2020, which is a reminder that excavated material does not stop producing information.
172:37
The original 1934 recovery was a rescue operation by the standards of the time, so researchers went back to the site itself, relocated the exact spot and excavated the surrounding area with current methods, recovering additional material that the first dig had missed. Genetic analysis produced a portrait consistent with what we now know about European foragers of that period, a population whose appearance differed from later farming populations in ways that surprised a great many people when these results first started appearing.
173:08
The relationship between the woman and the infant placed with her was also examined, and the picture turned out to be more complicated than a simple mother and child, which is a useful reminder that the obvious reading of a double burial is not always the correct one. The team also reconstructed what the antler headdress would have looked like in use, based on the surviving worked pieces and on comparable finds from other sites of the period, where portions of deer skull with antlers attached were deliberately thinned and perforated in ways that only make sense for wearing. so we are not guessing that these were worn.
173:39
We know that somebody hollowed and pierced them for exactly that purpose. What these two burials share, across an enormous distance in space and time, is a combination that starts to look like a signature, a body marked by physical difference, a grave built rather than dug, a large assortment of objects from species selected for rarity rather than for food, heavy use of pigment, and a rite that does not match anything else in the same community.
174:07
It is also worth being concrete about what such a role consists of, because the popular image reduces it to dramatic performance and the reality is closer to a small general practice. In societies where these specialists have been described directly, the job is a bundle of services rather than a single function.
174:24
They are consulted when someone is ill, which involves both practical knowledge of plants and the ceremonial framing we discussed earlier. They are asked to locate things, including game, lost people and lost objects, generally through some form of divination. They are called on before risky undertakings. They handle the passages in a life, including naming, coming of age, marriage and death. They frequently act as mediators in disputes because a person who speaks for something outside the community is well positioned to propose a settlement nobody can easily refuse.
174:56
And they carry and recite the group's account of itself. That is a considerable amount of work and it is generally compensated. Payment in the ethnographic record takes the form of food, goods, labour or obligation, and the expectation of payment is usually explicit. Nobody in these traditions imagines the specialist works for free, and a number of recorded traditions hold that a ceremony performed without proper compensation will not be effective, which is either a profound spiritual principle or the oldest invoice in human history, depending on how cynical you are feeling.
175:29
There is also a risk built into the role that rarely gets mentioned. A person who claims the ability to influence outcomes is a person who can be blamed when outcomes go badly.
175:39
The ethnographic record includes plenty of accounts of practitioners being held responsible for illness, death or failed hunts, sometimes with serious consequences for them personally.
175:49
This is not a comfortable position with guaranteed deference. It is a high-visibility job with an unpredictable success rate in a community that remembers everything and has nowhere else to go.
176:01
And that first element, the physical difference, recurs often enough to require comment. Across the ethnographic record, in societies that have been described directly by observers, The role of ritual specialist is disproportionately associated with individuals marked out by some form of bodily or experiential difference. This appears in accounts from Siberia, from the Amazon basin, from Korea, and from many other regions, in traditions that have no historical connection to one another. The pattern is not that such people were merely tolerated, it is that the difference was frequently understood as the qualification itself.
176:37
in many siberian traditions the vocation is described as arriving unbidden often through a period of serious illness or crisis that the community interprets as a summons after which the person either accepts the role and recovers or continues to suffer The Korean tradition has a specific name for that initiatory illness, and the resolution involves formal acceptance of the role under the guidance of an established practitioner. In parts of Amazonia, similar accounts describe the calling as something that selects the person rather than something the person chooses.
177:09
Now, there are two ways to read this, and I think both are partly right. The cultural reading is that these societies had a coherent theory about who can cross boundaries. A specialist's job is to move between the ordinary world and another one, and a person who already stands at an angle to ordinary human existence, whose body works differently, or whose experience includes things others do not experience, is a natural candidate.
177:34
The logic is symbolic, and it is perfectly consistent. You send the person who is already partway across.
177:41
The practical reading is less romantic and I suspect equally operative. Consider a foraging community with a member who cannot move fast, cannot travel far, cannot carry loads or cannot participate in the hunt. In pure economic terms that person is a cost. But the group has another set of needs that require no mobility at all, remembering the songs, holding the sequence of the ceremony, knowing which plant is used for what, maintaining the account of the ancestors and being present at camp when everyone else is out.
178:11
We saw this logic already with the value of elders and it applies with even more force here. A society that finds a productive role for people who cannot forage is a society that keeps knowledge it would otherwise lose and it gets a specialist in exchange for food it was going to share anyway.
178:29
The two readings reinforce each other. The role is symbolically appropriate and materially sensible at the same time, which is generally how durable social arrangements work. Here comes the necessary correction, and it is a significant one, because the word everyone reaches for at this point in the conversation is doing a great deal of damage. The term shaman is not a general anthropological category that happens to fit societies worldwide. It comes from a specific language family in Siberia and originally described a specific set of practitioners in a specific cultural context.
179:03
In the 18th and 19th centuries, European observers took that word and applied it to every ritual specialist they encountered anywhere on the planet, which had the effect of collapsing an enormous variety of distinct roles, obligations, techniques and social positions into one imported label.
179:19
The practical consequence for archaeology is serious. When we find an unusual burial and call the occupant a shaman, we import a whole package of assumptions – drumming, trance, spirit flight, healing, a costume, and a particular relationship to the community. none of that is present in the evidence what is present is an unusual grave so the responsible approach is to describe the behavior rather than to name the person what we can say about these burials is that the community recognized a category of individual who received treatment given to no one else that the treatment involved materials with no practical function drawn from species that are difficult to obtain and that acquiring those materials required organized collective effort That is a role.
180:03
Whether its occupant drummed, sang, healed, negotiated with the dead, dreamed, or did something we have not thought of is unknown. And this is where the criteria matter because plenty of graves contain objects and most of them tell us nothing of the kind. Grave goods alone prove very little. A person buried with a fine blade might have owned a fine blade. Ornaments indicate wealth or affection or fashion. Quantity by itself indicates a family with resources. None of that gets you to a specialist.
180:33
What does the work is the combination of four things. First, a constructed grave, meaning labour invested in the container rather than just the act of covering. Second, an inventory that is systematically non-utilitarian, made up of parts of animals rather than usable products, and weighted towards species that are rare, dangerous or hard to catch. Third, evidence of a collective event, which tells you the group participated rather than a family. And fourth, and most decisive, singularity within a known population.
181:05
The right happened once, and everyone else from the same community was treated differently.
181:10
Apply those four tests to the two burials we have discussed and both pass on every count. Apply them to the vast majority of prehistoric graves and they fail immediately, which is exactly what you want from a test. There are other candidates scattered through the record and they are worth a brief mention because they show the pattern is not confined to two accidents of preservation. From the Ice Age hunting cultures of Central Europe there is a burial of a man accompanied by an unusual assemblage, including a carved ivory figure with separately made limbs, apparently designed to be jointed, which is not a category of object that appears anywhere else in domestic life.
181:46
Elsewhere in the same region there are individual burials with pigment, perforated teeth in quantity, and animal parts selected with the same logic we have been describing. Each on its own is a hypothesis. Together they describe a recurring type. So what does the existence of this role actually mean for our larger question? It means that by the end of the Paleolithic, at the latest, religion had crossed a threshold that changes its character completely. Up to this point everything in this story has been paid for in objects and effort. Pigment prepared and applied, beads carved, images placed in caves that cost fire and fear to enter, instruments made.
182:24
All expensive, all real, but all of it something a person does in the time left over from staying alive. A specialist is a different category of expense. A specialist is a human being whose hours are allocated to this rather than to subsistence. even in a part-time arrangement and in forager societies these roles are usually part-time rather than full occupations, the group is now underwriting a person's expertise. Somebody has to learn the sequences, and learning takes years, and the years spent learning are years not spent becoming an excellent hunter or a skilled maker of tools.
182:59
somebody has to be available when needed rather than wherever the food is and somebody has to have taught them which means the knowledge has a lineage with an apprenticeship at the bottom and an accumulated body of practice at the top There is a consequence of apprenticeship that deserves more attention than it usually gets because it changes the nature of the thing being transmitted. The moment a practice has to be taught it acquires a correct version. A person performing a private ceremony can improvise and nobody is in a position to object. A person teaching a ceremony to a student has to specify what comes first, what words are used, how long each part lasts and what counts as having done it properly.
183:38
The teacher will correct the student, and correction requires a standard. That is how a fluid, improvised practice hardens into a procedure, and once a procedure exists three things follow almost automatically. There is now such a thing as performing it wrongly, which means there is now such a thing as a failure caused by human error rather than by the indifference of whatever was being addressed. There is now a reason to preserve the details exactly, since nobody knows which part is the part that works. And there is now a body of knowledge that exists independently of any individual, held across a chain of teachers and students, capable of outliving all of them.
184:17
Notice that this is also the beginning of authority in its most durable form. The teacher is not merely older, they are the one who knows the right way and the right way is not a matter of opinion. Disagreements about practice now have an answer and the answer belongs to whoever stands at the older end of the chain.
184:34
That is the moment when religion stops being a set of things people do and starts being something people are. It also creates a structural feature with enormous consequences later. As soon as there is a person who performs the ceremony for others, there is a distinction between those who can and those who cannot.
184:51
Access to the invisible world is no longer symmetrical. It runs through somebody.
184:57
and a channel that runs through a person is a channel that person can, in principle, regulate. I want to be measured about this because it is easy to turn into a cynical story about manipulation, and there is no evidence for that here.
185:11
The specialist in a small forager community is not a priest with power over a population. They are a neighbour with an unusual job subject to the same social pressures as everyone else, and in the ethnographic record such figures are often regarded with ambivalence rather than deference. Because a person who can help in extraordinary ways is a person who might do other things too. The concentration of authority comes much later and requires much larger societies.
185:37
but the seed is here, in a stone-lined pit in a Galilean cave, sealed under a slab with the remains of a feast around it.
185:45
A community decided that one of its members was categorically different from the others and marked that decision with more resources than they spent on anything else we can detect.
185:55
And notice what has happened to the geography of all this. The specialist is buried inside a cave used repeatedly as a cemetery by a community that was, in this period, beginning to stay in one place rather than moving continuously.
186:08
That is a new development in itself. As long as people are mobile, the dead are left behind in places you may or may not return to. Once people stop moving, the dead stay where the living are. Which raises a question that had never needed an answer before. If you're not going anywhere, and your grandparents are buried under the ground you live on, what exactly is your relationship with them now? The answer that emerged over the following few thousand years is one of the strangest and most consequential things our species has ever done. People stopped leaving the dead in the ground entirely.
186:40
They started taking parts of them back into the house. Sometime around 9,000 years before the Common Era, in the villages spreading across the Levant, People began doing something with their dead that has no precedent anywhere in the record and that is, depending on your temperament, either deeply moving or deeply unsettling. I would suggest both.
187:00
The sequence went like this. A person died and was buried in the ordinary way, usually in a pit beneath the floor of the house they had lived in. Time passed, generally a year or more, long enough for the soft tissue to decompose completely. Then the floor was opened again and the skull was removed and taken out. What happened next is the part that changes everything. The skull was cleaned, the lower jaw usually removed, and then a specialist took lime plaster and modelled a face onto the bone. Cheeks, brow, the bridge and wings of a nose, lips, a chin.
187:34
The eye sockets were filled and set with shells, sometimes a pair of cowries laid to suggest lids over an open eye, sometimes flat bivalve shells with a line of pigment across them. Some were painted. Some carry traces of applied colour in bands across the crown, which may represent hair or a headdress. And then the face went back into the house, where the living could see it. These objects are, by any reasonable definition, the oldest portraits in the world. They predate the first writing by about 4,000 years. They predate the pyramids by longer than the pyramids predate us.
188:06
And unlike a painting or a statue, each one is a portrait with the original sitter physically inside it.
188:13
The period this belongs to is called the pre-pottery Neolithic B, which is an unglamorous name for one of the most consequential stretches of human history. Its people farmed, kept early domestic animals and lived in permanent villages of rectangular mud-brick houses, some of them large enough to count as towns by any standard of the time. They had not yet invented pottery, which is the detail the name is built around, and which is genuinely odd when you consider what they had invented instead. because they had lime plaster, and lime plaster is not a simple material.
188:45
To make it, you take limestone and heat it to somewhere in the region of 800 degrees sustained, which requires a substantial fuel load and considerable technical control. The resulting quicklime is then slaked with water, a process that is energetic and requires care, and the product is a workable pace that sets hard into an artificial stone.
189:05
These communities were using it for floors, for wall surfaces, and for objects, and the fuel cost was serious enough that some researchers have suggested plaster production contributed to local deforestation. So a society with no pots in a world with no metal was burning a meaningful share of its firewood to produce a material that it used, among other things, to reconstruct the faces of the dead.
189:28
That is an allocation of resources that tells you where the priorities were. The first of these to be recognized came out of Jericho, the long-occupied mound beside a spring in the Jordan Valley, during Kathleen Kenyon's excavations in the 1950s.
189:42
Kenyon was digging through the deep sequence of the site when she encountered a group of plastered skulls deposited together beneath a floor. Seven of them, in one cache, faces modelled, shells in the eyes, arranged as a set. Since then the practice has been identified at site after site across a wide region. Ain Ghazal in Jordan, on the outskirts of what is now Amman, produced examples along with something else remarkable, which we will get to. Tel Aswad in Syria produced a whole group of them deposited together. Kefar Ha Horesh, Yiftahel, Besamun and others in the southern Levant have contributed further examples and there are related finds from Anatolia.
190:22
The total now runs to several dozen individuals across an area stretching from central Turkey to Jordan.
190:28
One site in that list deserves separate attention because it suggests the practice had dedicated facilities.
190:34
Kefar Ha Horesh, in the hills of Lower Galilee, does not look like an ordinary village. It has plastered surfaces, enclosures and installations, but relatively little of the domestic debris you would expect from a place where people spent their lives.
190:49
What it has in abundance is burials, along with deposits combining human and animal remains, in arrangements that are clearly deliberate. In one case, bones were laid out to form a shape. In others, human remains were interred together with parts of wild cattle, and there are deposits that appear to represent a gathering with a considerable amount of meat consumed on site. The interpretation offered by the excavators is that this was a regional mortuary and ceremonial centre, serving multiple settlements in the surrounding landscape rather than one community.
191:20
If that is right, then the treatment of the dead in this period was not only a household matter, some of it happened at a shared facility that people travelled to, which implies coordination between villages and a degree of standardisation in how the work was done. That number matters more than any individual object, and the reason should be familiar by now. A single modelled skull is an eccentric episode. Several dozen, spread across hundreds of kilometres in several centuries, sharing the same technique and the same basic conventions is an institution. People in different villages, some of whom never met, were doing the same thing to their dead in the same way.
191:55
And while we are at Ein Ghazal, the other thing found there deserves its moment, because it demonstrates that this society had a broader project in mind. Buried in pits, carefully deposited rather than discarded, excavators found caches of large human figures made of lime plaster built up over a framework of bundled reeds, the largest stand close to a metre tall. They have flattened bodies, simplified limbs and heads with wide staring eyes, outlined in bitumen with painted details. Some have two heads on a single body.
192:27
They were among the oldest large-scale representations of the human form anywhere, and they were made by the same communities producing the modelled skulls, using the same material and the same technical knowledge. Then, at some point, they were deliberately buried in groups, apparently after a period of use, without damage, in a manner that reads as retirement rather than disposal.
192:48
Whatever these figures were, they were not thrown away when they stopped being needed. They were decommissioned. Now let us slow down and look at the process behind the modelled skulls, because the sequence of steps contains the whole argument.
193:02
Stage 1. The person is buried beneath the floor of an occupied house, not in a cemetery outside the settlement, not in a cave up a cliff, under the room where the family sleeps, cooks and works. Stage 2. Time passes. A year, probably more. The household continues to live directly above the grave, knowing exactly who is there and precisely where.
193:26
Stage 3. The floor is opened and the skull is retrieved. This is not a casual act. It requires breaking a plaster floor that somebody made at considerable expense and it requires handling the remains of a person the household knew personally. Stage 4. The skull is prepared and the face is modelled, which requires a skilled person, the right materials and probably several days of work. Stage 5. The object exists in the world of the living. Wear exactly as debated since most were found in their final deposit rather than in use, but wear patterns on some examples and the care taken in their manufacture suggest they were handled, displayed or kept somewhere accessible rather than immediately buried.
194:07
stage six eventually sometimes after a considerable period the object is buried again frequently in a group with others beneath a floor so a single dead person could remain an active participant in household life for years in three distinct phases each requiring labor and decision the funeral which earlier in this story marked the end of the process has become the first item on an agenda There is a detail in the modelling that has generated a long argument, and it is worth walking through because it changes the interpretation substantially.
194:41
Are these likenesses? Was somebody trying to reproduce the actual appearance of a specific individual, working from memory and from the underlying bone structure? Or is the modelled face a generic idealised form applied to whoever happened to be under the plaster?
194:55
The honest answer is that we cannot tell. The faces differ from one another in proportion and detail, which is consistent with individual portraiture. But skull shapes differ too, and plaster applied over different bone will produce different results without anyone intending it.
195:11
The technique also imposes conventions that recur across sites, which points toward a shared formula rather than free artistic observation.
195:19
What complicates the simple reading further is who these skulls belong to. The popular version of this practice is ancestor veneration in the narrow sense, meaning respected male elders preserved as founders of a lineage. The actual sample includes women, and it includes children.
195:36
There are modelled skulls of individuals who died young enough that they can have founded nothing and led no one. That is a genuinely informative complication. It suggests that whatever selected a person for this treatment was not simply seniority or achievement. It may have been lineage position, cause of death, circumstance, or something we have no access to at all.
195:57
The practice is not a straightforward hall of fame. Now let us move north and forward a few centuries to the site that shows us what living with the dead actually looked like on a daily basis. Çatalhöyük occupies a mound on the Konya plain in central Turkey and it was occupied roughly between 7100 and 5900 years before the Common Era.
196:16
At its height it held a population that ran into the thousands, which makes it one of the largest settlements on earth at the time, and it has no streets. The houses are packed directly against one another, sharing walls with no lanes, no alleys, and no open plazas in the excavated areas. Movement through the settlement happened across the rooftops, which formed a continuous surface at varying levels, and people entered their homes by ladder through an opening in the ceiling that also served as the smoke hole for the hearth.
196:44
Climbing down a ladder into your own house through the chimney every single day is a design choice that tells you the community valued something about this arrangement more than it valued convenience.
196:55
Inside, the houses follow a consistent plan with remarkable persistence. There is a main room with plastered walls and floors, raised platforms along certain sides, a hearth and oven near the entrance, and storage areas.
197:08
The platforms are the significant part because they are both the furniture of daily life and the location of the graves. The dead at Çatalhöyük were buried inside the houses, in pits cut beneath the platforms in the rooms where people slept and ate. Some houses contain a handful, others contain dozens. Individuals were interred over long periods, with later burials sometimes disturbing earlier ones, and bones were occasionally removed and moved elsewhere.
197:34
There is a finding from Çatalhöyük that cuts directly against the tidy version of this chapter, and intellectual honesty requires putting it right here rather than burying it later. If burial beneath a house is about lineage, then the people under one floor ought to be a family.
197:50
Researchers have tested that, using inherited features of teeth and skulls and more recently ancient genetic material, and the results have been consistently awkward.
198:00
The individuals interred within a single building do not show the degree of biological relatedness the simple model predicts. Some are related, many appear not to be. Nobody's quite sure what to make of this, and several explanations are on the table. Household membership in this society may have been defined by residence, adoption, or some social arrangement rather than by descent, which would make these buildings the graves of a household in the sense of a group who lived together rather than a group who shared ancestry. Bodies may have been moved between structures, or the category that determined where a person was buried may have been something we have not identified at all.
198:35
What it demonstrates is that our instinctive reading of a house full of graves as a family plot is an assumption, and that the people of 9,000 years ago were not obliged to organize themselves along the lines we find intuitive. Keep that in mind through the next section, because I'm about to build an argument that leans on inheritance, and this is the strongest objection to it. And the houses were rebuilt. When a building reached the end of its life, the walls were dismantled to a certain height, the interior was filled, and a new house was constructed directly on top, following the same plan, in the same position, with the platforms in the same places.
199:11
This happened repeatedly, building on building, sometimes through half a dozen or more cycles across generations. A single spot on that mound might hold a vertical stack of successive homes, each one a near copy of the last, with the accumulated dead of the family beneath them all. Ian Hodder, who directed work at the site for decades, noted that the burials are not distributed evenly.
199:33
Some structures contain many more than others and show a longer sequence of rebuilding and these have been described as houses that held the memory of the group, functioning as a focus for a wider set of households rather than simply as a family dwelling.
199:48
The interiors were also elaborate in ways that have nothing to do with comfort. Walls carry painted designs, including geometric panels and in some cases scenes. There are reliefs modelled in plaster. Most striking, the horn cores of wild cattle were set into benches and pillars and mounted on walls, physically installed into the architecture, along with other animal parts embedded in the plaster.
200:11
In some rooms the effect must have been overwhelming. A low, dim, smoke-stained space with painted walls, horns projecting from the furniture, and the family's dead directly underfoot. And in one of these houses excavators found a burial that sums up the whole chapter. A woman was interred beneath a platform, and cradled in her arms was a plastered human skull, its face modelled in the manner we have been discussing. We cannot know the relationship. We cannot know whether she chose it or whether the community decided. What we can say is that at the moment of her burial, somebody considered it appropriate or necessary that she go into the ground holding the face of a specific other person who had died before her.
200:53
That is not a symbolic gesture toward the dead in general. It is a link between two named individuals made physical and deposited under the floor of an occupied house. There is a third piece of evidence in this regional picture and it comes from a site we are going to spend the next chapter on, so I will keep it brief and specific.
201:11
In 2017, a research team led by Julia Greski published an analysis in science advances of human skull fragments recovered at Göbekli Tepe in southeastern Turkey, a site occupied in the centuries before the villages we have just been discussing. Among thousands of fragments of human bone from the deposits, three skulls showed a distinctive set of modifications. The alterations consist of deep grooves cut into the bone along the vertical axis of the skull, executed with stone tools, deliberate and controlled rather than incidental.
201:43
The cuts are not the shallow marks associated with removing soft tissue, they're carved channels. On one specimen there is also a perforation drilled through the bone, positioned in a way that would allow a cord to pass through. The researchers propose that the most plausible explanation is suspension. A drilled hole and a carved groove taken together are the standard signature of an object intended to be hung and the cuts would help a cord sit without slipping. Two caveats belong with this immediately. First, these three are a tiny fraction of the human bone from the site, so whatever this practice was it applied to a minority.
202:18
Second, and more important, we do not know whether the individuals concerned were honoured, feared, or something else entirely, and the published work is careful on exactly this point. Modified human remains appear in the ethnographic record in the context of revered ancestors and in the context of defeated opponents, and the bone itself does not distinguish between them.
202:39
What the finding does establish is that in this region, in this period, human skulls were being worked, modified and prepared for display or suspension.
202:49
Combined with everything else in this chapter, we are looking at a broad regional concern with the head as the part of a person worth keeping. It is worth pausing on how ordinary all of this would have felt from the inside, because the modern reaction to it is not the relevant one. We separate the dead from the living as a matter of course. There are dedicated places, dedicated professionals, and a strong cultural expectation that a body will be removed promptly and handled by people trained for it.
203:17
Most people in wealthy countries today reach middle age without ever having been in a room with a dead person for more than a few minutes. That arrangement is recent and it is unusual. For the people in these villages none of it applied. The dead were not elsewhere. They were three feet below the sleeping platform and their removal, preparation and return were tasks the household carried out, much as it carried out repairs to the roof. A child growing up in that house would have seen a floor opened, seen what was underneath, and understood the process the way a child today understands that recycling goes out on a particular day, which means the plastered face on the shelf was not a macabre object in that setting, it was a relative, in the form relatives take after a while.
204:00
Now, for the question that has been hanging over all of this, why? What does a society actually get out of maintaining relationships with the dead that require this much plaster, labor, floor demolition, and careful handling? The emotional answer is obvious, and it is not wrong. These are people who have lost family and want to keep them close. Grief does not require a functional explanation and I am not going to pretend that every act in the ancient world was a piece of social engineering.
204:27
But the emotional answer does not explain the specific form. It does not explain why the skull, rather than another part, It does not explain why the face is reconstructed so that the individual is recognizable rather than anonymous. And it does not explain why this practice appears in exactly the period it does, spreading across a region at the same time as something else. Here is what else was happening. These were the first societies in human history with fixed, heritable assets. Think about what that means as a genuinely new problem. A foraging band moves across a territory and holds rights that are flexible, negotiated and shared because nobody can own a place they visit twice a year.
205:05
A farming village cannot work that way. A field that was cleared of stones and worked for a decade represents accumulated labour, and its productivity belongs to whoever cleared it. A house built of mud brick with plastered floors represents months of work. A well or a spring-fed channel, a stand of fruit or nut trees planted by somebody now dead, a storage pit, a flock. All of these are things that persist, that took effort to create and that will still exist when their maker does not.
205:35
So a question arises that had never needed an answer in 300,000 years of human existence. When the person who made this dies, who gets it? and the answer every agricultural society in history has arrived at is their descendants, which immediately generates a second question, and this is the one that matters. How do you prove you are a descendant? You have no writing. You have no documents, no registry, no court, no sealed deed. You have no institution of any kind that keeps records, because those institutions do not exist yet and will not for thousands of years.
206:11
What you have is memory and testimony, which is to say you have claims, and claims can be disputed by anyone with the nerve. Now look again at what these people built. The ancestor is under the floor of the house. The house is the asset. The skull of a specific named individual is preserved, given a recognisable face, and kept by the household.
206:33
The genealogy connecting that individual to the living occupants is recited and there is a physical object present during the recitation, sitting in the room, looking at everyone with shell eyes. That is a title deed. It is a title deed with a face on it. The house is not merely where the family lives, it is the physical proof of the family's continuity, and the proof is embedded in the structure itself. Rebuilding on exactly the same footprint, generation after generation in the same plan, is not architectural conservatism or a failure of imagination, it is the maintenance of a claim.
207:08
The plot has not changed hands, the line is unbroken, and the evidence for both of those statements is stacked vertically beneath the floor in the form of every previous occupant. In a world without documents, the dead are the archive. They are the only records technology available, and the face modelled in plaster is the entry in the register, made specific so that it refers to this ancestor and not to some other families.
207:33
Once you see the practice this way, several details stop being mysterious. The individuality of the faces matters because the claim is specific, not descent from ancestors in general, but from this one. The retrieval after decomposition makes sense because the skull is the part that distinguishes one person from another and the final burial in groups beneath a floor is the closing of an account, the archive consolidated and secured in the property it authenticates. There is a second function and it is the one that carries the most weight for the rest of this story.
208:04
An ancestor who is present is an ancestor who is watching. Up to this point in our account the invisible participants in human life have been mostly impersonal or distant, forces with intentions, beings encountered in altered states, agents in stories.
208:20
What the ancestor cult introduces is something new, a supervising presence that is specific, local, named, and deeply interested in the conduct of particular people. Your ancestors know you. They know what your obligations are because they set them. They are in the room, and they can be displeased. That is the smallest possible unit of moral surveillance, and it is enormously effective because it operates on exactly the scale where it can be enforced. Nobody needs a doctrine about a universal moral order. What they need is the conviction that the grandmother under the floor is aware of how her grandson is treating his obligations and that if things go badly for the household, it may not be bad luck.
209:01
And unlike a living authority, an ancestor cannot be avoided, argued with, or outlasted. They do not sleep, they do not travel, and they are not going to die.
209:12
Now, the caution section, because this chapter has been building a rather elegant machine and elegant machines deserve scrutiny. First, and most straightforwardly, secondary treatment of the dead can have entirely practical drivers. In a densely packed settlement with no space to expand, the question of where to put people is a real logistical problem, and clearing older burials to make room for newer ones is a solution that requires no belief whatsoever. Bones get moved because a family needs the space under the platform and that is the whole of it.
209:43
Second, there are hygiene and management considerations. Living directly above burials creates practical issues and the sequence of decomposition followed by retrieval might reflect a sensible waiting period rather than a ritual calendar. Communities that bury inside their houses have a strong incentive to develop a procedure, and procedures look like rituals from a distance. Third, mobility. Households move, structures are abandoned, and the portable dead are portable precisely because they have been reduced to a manageable form.
210:15
A skull can go with you to a new house. A complete burial cannot.
210:19
Fourth, the interpretive problem raised by the Gobekli Tepe material applies more broadly. We have decided to read this entire practice as veneration, because that is the emotionally coherent reading, and because it matches ethnographic ancestor cults in later periods. But the same physical evidence would be produced by practices with very different meanings, including the display of remains taken from outside the community. The bone does not come with a label. Fifth, and most fundamentally, the property argument I have just spent several minutes constructing is a functional interpretation, and functional interpretations have a specific hazard.
210:57
They are satisfying, they explain a great deal, and they cannot be tested against the object.
211:02
Nobody at Jericho left us a statement about inheritance. What we have is a correlation in time between the appearance of fixed assets and the appearance of curated ancestors, and correlation in this field is the beginning of an argument rather than the end of one. There is one further observation worth making which is that this practice did not last forever. The large villages of this period ran into serious trouble toward the end of the eighth millennium before the Common Era. Several of the bigger settlements contracted sharply or were abandoned, and the reasons debated include soil exhaustion, deforestation driven partly by the fuel demands we have already discussed, climatic stress, and the simple difficulty of holding a large community together without any of the institutions that later societies invented for the purpose.
211:46
The curated skulls fade out with them. Later periods in the region handle the dead differently, generally moving them out of the house and out of the settlement, into cemeteries and eventually into monumental tombs that are somebody else's problem entirely.
212:01
So this was a solution suited to a specific moment. Permanent villages large enough to have property but small enough that every household could hold its own records under its own floor.
212:11
Once settlements outgrew that, the arrangement broke and the sacred had to be relocated somewhere it could serve more than one family. Which is the direction this story is about to go. What I would defend without hesitation is the weaker claim, which is still substantial.
212:27
In the early farming villages of this region, the dead were not removed from society. They were retained, physically, inside the space of the living, in a form that identified them as specific individuals through a process requiring significant skill and expense sustained over centuries and across a wide area.
212:46
That is a different relationship with death from anything earlier in this story.
212:51
The graves of the Ice Age, however elaborate, were endings. These are not endings. The person continues to have a location, a face and a role, and the household continues to have obligations toward them.
213:04
And that shift produces something that had never existed before – a community that includes both the living and the dead as ongoing members, with the dead holding the senior positions. Consider how that changes the experience of being alive in such a village. You inhabit a building your great-grandparents built, standing on the remains of the one before it. Beneath your sleeping platform are people whose names you know and whose faces you may have seen reconstructed in plaster. Your right to be there rests on your connection to them. Your conduct is visible to them, and when you die, you will be added to the sequence, and your descendants will stand on you.
213:41
That is an extraordinarily stable arrangement, and stability is exactly what a farming society needs, because farming punishes disruption in a way that foraging does not.
213:51
miss a planting season and you do not eat next year so by around 8 000 years before the common era we have arrived somewhere specific there is a class of ritual specialist as we saw there is a practice of maintaining the dead as identified individuals with continuing presence there is a physical location that holds all of it namely the house which is simultaneously a dwelling a tomb an archive and a proof of ownership But notice the limitation built into that arrangement. Everything is happening at the scale of a single household.
214:24
The ancestors under one floor belong to one family and authenticate one family's claims. They are watching one household's conduct. A neighbor's ancestors have no authority over you, and yours have none over them.
214:36
which is fine while a settlement is small. It becomes a problem the moment a community grows large enough that households need to cooperate with people they are not related to, on projects that no single family can undertake. And at some point in this region, in the centuries before any of the villages we have just discussed, somebody solved that problem in the most ambitious way imaginable. They stopped putting the sacred inside the house. They built it a place of its own, out of stone, on a hilltop, at a scale that required hundreds of people who did not share a floor to work together.
215:10
In 1963, a survey team working through southeastern Turkey walked over a low hill on a limestone ridge north of the modern city of Şanlıufer. They noted broken limestone slabs sticking out of the ground, concluded they were probably gravestones from a medieval cemetery, wrote it up, and moved on.
215:29
It is difficult to overstate how wrong that was, and equally difficult to blame them, since nobody had any reason to expect what was underneath.
215:37
Three decades later, a German archaeologist named Klaus Schmidt was looking for a site to work on in the region, read the old survey notes and went to see the hill himself. He recognized within a short time that the slabs were not gravestones, they were worked limestone and there were a great many of them. He began excavating in 1995 and he spent the rest of his life there. The hill is called Gobekli Tepe, which translates roughly as Potbelly Hill and it is not a natural formation. It is a mound about 15 metres high and 300 metres across, built up over centuries from the accumulated debris of human activity.
216:14
and the material at the bottom of it dates to around 9,500 years before the Common Era. To place that properly, it is roughly 6,000 years before the first writing, 7,000 before the first pyramid, and about 5,500 years before anybody in Britain thought about putting up a stone circle. It belongs to the pre-pottery Neolithic A, the phase before the villages of the previous chapter, and it was built by people who had not yet domesticated a single plant or animal.
216:42
What they built is a series of enclosures, oval or roughly circular, cut down into the ground.
216:48
Around the inner edge of each runs a wall with a stone bench and set into that wall at intervals are pillars standing upright. Each pillar is a single piece of limestone carved into a T-shape with a long vertical shaft and a broad horizontal slab across the top and in the middle of each enclosure stand two more pillars, larger than all the rest facing each other.
217:08
In the enclosure known as D, the two central pillars stand about 5.5 metres tall and weigh somewhere in the region of 8 to 10 tonnes each. Now hold that number against the available technology. There was no metal of any kind. There was no wheel. There were no draft animals because cattle were still wild and would remain so for another 2,000 years. There was no writing to record a plan and no state to compel anybody to work.
217:33
what existed was stone tools rope timber and people the quarry has been located and it is on the same ridge still lying in it is an unfinished pillar around seven meters long estimated at something like 50 tons abandoned before it was fully separated from the bedrock somebody was attempting that whether they intended to move it or gave up part way through is unknown but the ambition is recorded in the rock Moving a ten-ton block without wheels means levers, sledges, rollers, ropes and a substantial number of people pulling in coordination over a distance of a few hundred metres uphill.
218:10
estimates of the workforce required vary, but every version of the calculation arrives at the same conclusion. This is not a project a family or a band can undertake. It requires a crowd, organised, working to a plan for an extended period. And then there is what the pillars have on them. Many are carved in relief with animals, and the selection is pointed. foxes with teeth bared, wild boars, snakes sometimes in dense arrays flowing down the shaft, scorpions, cranes and vultures, aurochs, spiders, leopards.
218:42
What is almost entirely missing is the category you would expect if these were about the food supply, because the actual diet at the site was dominated by gazelle and gazelle barely appear.
218:52
As with the painted caves of the Ice Age, the creatures on the walls are not the creatures on the menu, but the most important carvings on these pillars are not the animals. They are the details that are easy to miss until somebody points them out and then impossible to unsee. On the central pillars of several enclosures, running down the narrow sides of the vertical shaft, there are arms, carved in low relief, bent at the elbow, with hands that come around to the front and meet above the waist, fingers extended toward each other. Around the base, at the level where a waist would be, there is a belt, rendered in detail with what appears to be a buckle and decorative elements, and hanging below it the suggestion of a garment, in at least one case interpreted as an animal pelt.
219:35
So the pillar is not a pillar, it is a body, a schematic, colossal stylised body, standing upright, with the broad horizontal slab at the top representing the head, viewed from an angle that makes it entirely flat and entirely faceless.
219:50
That last part is a choice. These people could carve a fox so well you can identify the species. The blank head is not a limitation of skill. Somebody decided that these beings should have a form, arms, hands, clothing, and a stance, and no face at all. And the arrangement is consistent.
220:09
Two large figures stand in the centre, facing each other, and around them in a ring stand smaller figures, all of them facing inward toward the centre pair. Whatever is happening in that enclosure, the smaller beings are attending it.
220:22
For years this was one hill with a spectacular set of monuments, which raised an obvious and uncomfortable question. Was it a one-off? A single extraordinary sight tells you that something remarkable happened once, and as we established a long way back, one instance is a hypothesis. That question has now been answered emphatically, and the answer is the most exciting development in this field in a generation.
220:46
Since 2021, Turkey has been running a coordinated research programme across the wider region under the name Tash Tepele, meaning stone hills, covering roughly a dozen sites of the same period spread across the Şanlöfer area. These are not variations on one theme discovered by luck. They are a landscape of related settlements, contemporary with one another, sharing architectural conventions and iconography, and several of them have produced material that Göbekli Tepe did not.
221:13
Karahan Tepe, excavated under the direction of Nekmikarel, is the standout. Parts of this site were not built up from blocks at all. They were carved directly down into the living bedrock, which is a fundamentally different construction method and arguably a more demanding one since you cannot correct a mistake by replacing a stone.
221:33
One chamber there has become the defining image of the whole programme. Cut into the rock floor of a room are eleven upright pillars, shaped and left standing as the surrounding rock was removed around them, arranged in a group.
221:45
They are phallic in form which the excavators describe plainly and which I will note and move past. and at one end of the chamber, emerging from the rock wall at head height, is a carved human head with a face looking out over the room. The neck extends back into the wall in a way that has been compared to a serpent. So somebody entered a space hollowed out of solid rock, and at the end of it a face was watching them, which, after a chapter spent establishing that the pillars elsewhere were deliberately given no faces, tells you these communities had a repertoire of choices and were using them for different purposes.
222:19
Another site in the group, Sabus, produced something even rarer, a narrative panel carved in relief on a bench inside a communal building. It shows a human figure flanked on both sides by leopards, and beside that, a second figure holding something in his hand, facing a large bull with prominent horns. Two scenes, on one wall, with figures in relation to one another rather than isolated animals. You will recall how rare narrative composition is in the deep past. Here it is, in stone, on a bench where people sat, in a building meant for gathering, and from the same region and period comes a life-size human statue, recovered near a spring in the middle of Shanlufa itself, carved in limestone, standing about 1.9 metres tall, with eyes made of inset obsidian and hands held over the front of the body in a posture that echoes the pillars, It is the oldest known life-size sculpture of a human being anywhere in the world, and it has a face, and the face is staring directly at you.
223:20
Now for the part of this chapter that most accounts skip, and that I think is considerably more interesting than the version everyone knows, When Gobekli Tepe entered public awareness, it arrived with a headline attached, the world's first temple, built by hunter-gatherers, with no settlement around it. Nobody lived there. It was a sacred precinct in an empty landscape, visited by dispersed foraging bands who gathered periodically to build and to feast and then dispersed again. Klaus Schmidt argued this, and he argued something bolder on top of it.
223:52
If people had to gather in large numbers to build and to feed the builders, then the demand for a reliable food supply at one location could have driven the intensification of wild cereal use and eventually the domestication of grain. In short, first came the temple and agriculture followed to support it. It is a wonderful thesis. It inverts the entire standard sequence in which farming produces surplus, surplus produces cities, and cities produce religion. Here, religion comes first and drags the rest of civilization along behind it.
224:24
Since the late 2010s, the research team at the site, led by Lee Clare, has been steadily publishing findings that complicate this picture, not demolish it, complicate it, which is usually how real revisions work. The first problem is the domestic evidence. Excavation and analysis have produced thousands of grinding stones and handstones from the site, which is an enormous quantity of food processing equipment. There are traces of wild cereal being processed at scale. There are substantial deposits of animal bone consistent with sustained occupation rather than occasional feasting alone.
224:57
The second is water. The ridge has no spring and no stream, which was one of the original arguments that nobody could have lived there permanently.
225:05
Then the team identified cisterns cut into the bedrock, designed to collect rainwater, with a combined capacity in the region of 150 cubic metres along with channels that appear to have been cut to direct runoff into them. That is a water supply system, and water supply systems are built by people who intend to stay.
225:24
The third is the architecture itself. Alongside the famous large enclosures, excavation has revealed smaller rectangular buildings, some with their own smaller T-shaped pillars, with terrazzo floors of polished lime and with features suggesting they were roofed.
225:39
Some of the large enclosures may themselves have been covered rather than open to the sky.
225:44
The current formulation, as the excavators now put it, is milder and better. This was a settlement with exceptional communal buildings. People lived on the hill. They also built things there that no household could build alone.
225:58
I want to be clear about what this does and does not change, because a lot of popular coverage has treated it as a debunking, and it is not. It does not make the monuments less remarkable. The pillars are still there, still ten tons, still carved, still erected without metal or wheels or animals by people who had not domesticated anything. It does not make the site later, or smaller, or more ordinary. What it removes is the sharp and slightly mystical image of an empty sanctuary in a wilderness, tended by nomads who appeared from nowhere and vanished again.
226:32
What it puts in its place is a community that lived in one location, solved the problem of water, processed wild grain in industrial quantities, and devoted a vast share of its collective labour to building enormous communal structures full of beings with no faces.
226:48
Honestly, I prefer the revised version. The first one was a mystery. The second is a society.
226:55
There is a further element of the original interpretation that has come in for revision and it is worth knowing because it appears in almost every popular account. Schmidt observed that the enclosures were filled with rubble, bone and sediment and concluded that they had been deliberately buried when their period of use ended, intentionally decommissioned and covered over, which would be a striking ritual act in its own right and would also explain the extraordinary preservation.
227:20
more recent analysis has questioned that the fill-in at least some enclosures looks consistent with material sliding in from the surrounding slope along with collapse and the ordinary accumulation of debris from a settlement built on a mound of its own making deliberate closure may still have happened in particular cases but the picture of a whole complex ceremonially entombed by its builders has become harder to defend It is a small thing, and it changes nothing about the pillars, but it is a good example of how an early interpretation, once it enters circulation, gets repeated long after the people actually digging have moved on from it.
227:56
Now, the wheat. This is the part where causation gets slippery, and it is worth being careful.
228:02
Genetic studies of einkorn wheat, one of the founder crops of Near Eastern agriculture, compared cultivated varieties with wild populations across the region and found that the wild stands most closely related to the domesticated form grow on the slopes of the Karakadar range in southeastern Anatolia.
228:19
which is to say, in the same corner of the world we have been discussing, within a manageable distance of these sites. So the picture is genuinely striking. In one region, in a window of a few centuries, you find the earliest monumental communal architecture anywhere and the ancestral population of one of the first domesticated grains. And yet, later work has complicated the neat single-origin story considerably. domestication now looks like a protracted process rather than an event taking well over a thousand years involving multiple populations and probably multiple regions across the fertile crescent with cultivation of wild plants long preceding the appearance of genuinely domesticated forms It is less a moment of invention and more a long, gradual, mostly unintentional drift in which humans kept selecting for traits like seeds that stay on the stalk instead of scattering, without any idea that they were doing so.
229:14
Which means we have a correlation in space and time between monumental building and early cultivation. And correlation, as this video has said more than once, is where an argument starts. There are at least three ways to read the relationship. Religion drove agriculture, which is Schmitt's version. The gatherings demanded food, the demand drove intensification. Agriculture enabled religion, which is the traditional version. Enough wild grain in the region made large gatherings feasible, and the gatherings produced the monuments. Or both were responses to the same underlying situation, namely a rich region at the end of the Ice Age where population was rising, mobility was declining, and people needed new ways to organise themselves.
229:55
I would replace the dramatic claim with a precise one and I think the precise one is more interesting anyway.
230:01
Monumental communal ritual and the transition to food production emerged in the same region, within the same few centuries, among the same populations, and each made the other more viable. Building on that scale requires feeding a crowd. Feeding a crowd rewards more intensive use of grain. More intensive use of grains supports a larger, more settled population. A larger settled population can build more. That is not a cause and an effect. It is a loop, and loops are how most genuinely transformative changes in history actually work.
230:34
but there is a problem hiding inside that loop, and it is the problem that the rest of human history has been dealing with ever since. Large-scale cooperation is fragile. The moment you have a project that a hundred people contribute to, you have a project that one person can benefit from without contributing to. Somebody can show up for the feast without hauling the stone. Somebody can take from the shared store without adding to it.
230:57
In a small band, this is manageable, because everybody knows everybody, and a reputation follows you permanently.
231:04
In a community of hundreds, working alongside people from other households, it becomes a genuine structural weakness. So a society that wants to operate at that scale needs some way to tell the committed from the opportunistic. It needs a test, and it turns out that the peculiar, expensive, apparently pointless demands that religions make of their members are extremely good at exactly that.
231:28
The framework for understanding this comes from signaling theory, which biologists developed to explain why animals produce displays that appear to be actively harmful to their own interests.
231:38
The core insight is that a signal is only worth trusting if it is expensive to produce, because a cheap signal can be faked by anyone. Saying you're strong costs nothing. Doing something that only a strong individual could do is information.
231:53
Applied to human groups, the logic is uncomfortable and rather elegant. If you need to identify who is genuinely committed to a community, asking them is useless because a person planning to take advantage will say yes with enthusiasm.
232:06
What works is requiring behaviour that someone with one foot out the door would not be willing to perform. Sit through a ceremony lasting several hours, repeatedly, when you could be working.
232:18
Abstain from certain foods permanently, including ones that are convenient and nutritious. Fast for extended periods. Dress in a distinctive manner that makes you identifiable and therefore accountable everywhere you go. Undergo an initiation that is difficult or painful. Hand over a fixed share of what you produce. Accept restrictions on marriage, on trade, on your daily schedule. Each of these looks irrational in isolation and in fact none of them is. The cost is the function, someone willing to pay it, year after year, in public, is demonstrating that they have invested too much in this group to casually defect from it and everyone else can see them doing it.
232:58
The empirical work on this is better than you might expect because 19th century America accidentally ran the experiment. During that period, a remarkable number of utopian communes were founded across the United States. Some were religious, organized around a shared faith and its requirements. Others were secular, organized around political or economic principles, socialist, anarchist, or simply idealistic.
233:23
All of them faced the same fundamental challenge, getting a group of unrelated adults to pool resources and work for collective benefit without falling apart.
233:33
Richard Sosis and Eric Bressler assembled data on roughly 200 of these communities and tracked how long they survived. The difference was not subtle. In any given year, religious communes were around four times more likely to still be in existence than secular ones.
233:49
Many of the secular experiments dissolved within a few years. Some of the religious ones persisted for generations. Then they looked at why, and the finding is the one that matters. They counted the costly requirements each commune imposed on its members, things like dietary restrictions, dress codes, limits on contact with outsiders, and mandatory communal activity. Among the religious communes, the number of such requirements predicted longevity. More demands, longer survival.
234:17
Among the secular communes the same relationship did not hold. Imposing heavy requirements did not help them. That asymmetry is the interesting part, and it suggests that the sacrifice by itself is not sufficient. The costs appear to work when they are justified by something not open to negotiation.
234:34
A rule grounded in a shared belief about the order of the world is a different kind of rule from one agreed at a meeting because the second one can be revisited at the next meeting by anybody who finds it inconvenient. You cannot put a divine command on the agenda.
234:48
A second study takes the argument out of history and into a controlled setting. Sosis, working with the economist Bradley Ruffle, ran cooperation experiments on Israeli kibbutzim using a standard game in which participants can either take from a common pool for personal gain or restrain themselves for the collective benefit.
235:06
The design lets you measure cooperative behaviour with actual money at stake rather than relying on what people say about themselves. They compared members of religious kibbutzim with members of secular ones. The religious participants were, on average, more cooperative. And within the religious group, the strongest predictor was participation in collective public prayer. The men who attended synagogue regularly behaved most cooperatively of all, and the effect tracked communal attendance rather than private devotion.
235:34
That detail is the crux of the whole theory. What predicted cooperative behavior was not belief measured privately. It was showing up publicly, repeatedly, where other members could see you. Now we have to follow this logic to the place it becomes disturbing, because a video that only shows the flattering version of an argument is not being straight with you.
235:54
If costly signals bind a community, and if cost is the mechanism, then there is no natural upper limit built into the theory.
236:02
And the archaeological record from the ancient Mediterranean contains a case that sits at the extreme end, which historians have argued about for a century.
236:10
in carthage and at related sites across the western mediterranean there are enclosures known as toppets containing large numbers of small urns holding cremated remains of infants and young animals with carved stone markers above them many bearing dedicatory inscriptions to deities The sites were in use for centuries. Classical authors, mostly writing from the perspective of Carthage's enemies, described these as sites of child sacrifice. For a long stretch of the 20th century, many scholars regarded those accounts as hostile propaganda from a civilization that had every motive to slander a defeated rival, which is a sound instinct in general.
236:49
A group of researchers, including Paolo Jella and Josephine Quinn, published a reassessment in 2013, arguing that the traditional interpretation deserves to be taken seriously, on the grounds that the inscription's record vows to deities, that the urns were treated as offerings rather than as ordinary burials. that the age distribution does not match a normal cemetery population and that the classical sources for all their bias are numerous and consistent others have pushed back firmly arguing that infant mortality in the ancient world was extremely high that a dedicated burial ground for those who died in the first weeks of life would produce precisely this kind of assemblage and that skeletal analysis of the remains is consistent with deaths around or shortly after birth rather than with selection at later age The debate is not settled and I am not going to pretend to settle it.
237:40
What I will say is why it belongs in this chapter.
237:43
If costly signalling explains the stability of religious communities then the theory has no built-in stopping point and the question of where societies drew that line is a real historical question rather than a rhetorical one. It also illustrates the limits of what archaeology can tell you. Two teams, examining the same material with professional care, reach opposite conclusions about what people intended because intention does not survive. Now for the largest question in the field, and the one with the best ongoing argument. Everything we have discussed so far operates at a scale where people know each other – a band, a village, a commune, a kibbutz.
238:20
Within that range, reputation does most of the work, because you will see everyone again and they will remember.
238:28
The upper limit on that kind of society is usually given as somewhere around 150 individuals, the figure associated with Robin Dunbar's work relating group size to brain structure in primates, and used as a rough estimate of how many relationships a person can maintain with any real knowledge of who each one is.
238:45
The number is itself disputed and should be treated as an order of magnitude rather than a threshold, but the underlying point survives. There is a limit to how many people you can personally keep track of. Above that limit, cooperation runs into a wall. You are now dealing with people you will never see again, whose reputations you cannot check, who have no relationship with anyone you know.
239:08
Every classic problem of large-scale society appears at once. Yet human beings built cities, states and empires, involving cooperation among total strangers on an enormous scale. Something solved the problem. Aranor and Zion proposed an answer in his book Big Gods, and it is the most influential recent argument in this field. His claim is that as societies grew, they came to be dominated by a specific kind of deity. one that is concerned with morality, that monitors human behaviour, that knows what people do when nobody's watching, and that responds accordingly.
239:44
The key word is monitoring. A local ancestor watches one household, as we saw. A powerful god who is not tied to one house or one family watches everyone who accepts it, everywhere, all the time. Two strangers who both believe themselves observed by the same vigilant power have a basis for trusting each other that requires no shared history at all, and societies that had that arrangement, Norenzayan argues, could grow larger and hold together better than societies that did not, which over time meant they spread.
240:16
It is a compelling case, supported by experiments showing that people behave more cooperatively when reminded of supervision. including work suggesting that even the suggestion of being observed shifts behaviour in economic games, though I should note that this specific experimental literature has itself run into replication difficulties, which by now should surprise nobody.
240:37
and in 2019 it ran into a much bigger problem. A team led by Harvey Whitehouse, working with a large historical database called Sashat, which codes hundreds of variables for hundreds of societies across thousands of years set out to test the sequence directly. If moralising gods enabled social complexity, they should show up before societies became complex. If they appear afterward, the causal story runs backwards. Their analysis, published in Nature, reported the opposite of the prediction.
241:07
Across their sample, moralising gods generally appeared after societies crossed a threshold of complexity, which they placed in the region of a million people. On that reading, big gods are a consequence of large-scale society rather than its cause, arriving to help manage a system that already existed.
241:25
The response was swift and technical. The central criticism concerned how absence of evidence was handled.
241:32
In many of the coded societies, the relevant information comes from written sources and writing appears late.
241:38
If a society has no surviving record of its beliefs, coding that as an absence of moralising gods rather than as missing data will systematically make such gods appear later than they were. Critics argued that the apparent sequence could be an artefact of when societies started writing things down rather than when they started believing them. The paper was retracted in 2021.
242:00
I want to be careful about what that means, because retraction sounds like a verdict and it is not. The retraction reflected unresolved concerns about the data treatment, not a demonstration that the opposite conclusion is true. The underlying question, whether moralising gods enabled large societies or emerged from them, remains genuinely open and both camps continue to publish.
242:22
And frankly, this episode is my favourite thing in the entire chapter because it shows the process working.
242:29
A bold theory was proposed. A large empirical test was run. It produced a result contradicting the theory. Specialists examined the method and found a serious flaw. The paper was withdrawn. Nobody's career ended, no conspiracy was involved, and the question is still being worked on with better methods. That is not science failing. That is the only system humans have ever built that reliably catches its own mistakes, doing the thing it exists to do, in public, on a question about which everyone involved has strong opinions, which leaves one last question, and it is the one that brings all of this home.
243:06
If this entire apparatus is as deeply rooted as I have been arguing, then it should not have vanished simply because we acquired better explanations for lightning. It should still be running, quietly, underneath everything else. It is, and you can catch yourself doing it before the end of today. You knock on wood after saying something optimistic and if there is no wood available you will feel a small irrational pull to find some. You avoid walking under a ladder and the sensible explanation about falling objects is a justification you constructed afterward.
243:37
You have an item of clothing you consider fortunate and you know perfectly well that fabric has no influence on outcomes and you will still wear it to the interview.
243:45
Professional athletes are the clearest case because their behaviour is recorded in high definition and repeated thousands of times.
243:53
Watch the fixed sequences performed before a serve, a free throw or a penalty kick. The same number of bounces, the same adjustments, the same breath, in the same order every single time by adults whose careers depend on performing under pressure. And here is the thing that should stop you from feeling superior about it. It works. Experimental studies have found that performing a personal ritual before a stressful task reduces self-reported anxiety and can improve performance, including in tasks requiring fine motor precision, and that the effect appears even for arbitrary sequences invented by the experimenters.
244:31
Some of this literature, particularly the work on luck-related priming, has had its own replication troubles, so treat individual studies with the caution we have been applying all along.
244:41
but the broad finding that a fixed preparatory routine reduces anxiety and steadies performance is consistent with a large body of work on control and stress.
244:51
Notice that it also fits the distribution we identified much earlier. Where competence produces reliable results, nobody bothers with ritual. It clusters exactly where skill runs out and chance takes over.
245:04
A batter facing an unpredictable pitch has a routine. A batter walking to the dugout does not. So none of this is a residue of ignorance. It is the same mechanism, running the same program, in people who can explain the physics of lightning and the epidemiology of disease. The knowledge did not remove the machinery, because the machinery was never about knowledge in the first place. It was about the gap between what we can control and what we care about, and that gap has not narrowed nearly as much as we like to think. So, after all of this, let us return to the question we started with and answer it properly.
245:38
Why did ancient humans start believing in gods? Not because they were stupid, not because they were credulous, or frightened of thunder, or unable to think clearly. They started because they had minds that assign intentions to ambiguous events and purposes to natural forms, because those minds could not conceive of their own ending, and therefore could not conceive of anyone else's, and because they lived in a world that regularly took their children and offered nothing to do about it. And it continued and grew and got expensive because it worked.
246:11
It gave grief a procedure and a schedule. It reduced the chronic anxiety of a species that knew too much about its own future in an era when chronic anxiety could kill you. It bound groups together with synchronized sound and shared sacrifice. It preserved knowledge across generations by attaching it to people whose job was to remember. It recorded who belonged to which family and which land before anyone could write.
246:35
and it made it possible for strangers to trust each other enough to lift a ten-ton pillar together, those 13,000 beads in a grave, which we opened with, were not wasted. They bought a community that could survive the loss of two children without dissolving. The pillars on that Turkish hillside were not a misallocation of labour. They were the first proof that hundreds of unrelated people could be organised around something none of them could see.
246:59
Belief in gods was not a bug in human reasoning. It was the first working social infrastructure our species ever built, and every institution that came afterward, including the ones that now explain the lightning, was constructed on top of it and inherited its foundations. We are still living in that building. We have renovated extensively and some of us have changed the name on the door but nobody has replaced the footings. A couple of loose threads I deliberately left hanging, which deserve their own nights. The question of what Neanderthals actually believed, which we touched on and then walked carefully around, is a whole story in itself and the evidence has changed considerably in the last decade.
247:36
and the stone hills of southeastern Anatolia are being excavated right now, with new material coming out of the ground every season, which means the chapter I just told you is going to need rewriting sooner than most. Both of those are coming. For now, that is the full length of it, from a hearth in a cave 300,000 years ago to the wooden table you may have just tapped without noticing. Thank you for staying with me through all of it. Sleep well and sweet dreams.
