Chapter 1: Born Laughing
The lab at Bowling Green State University was not a glamorous place to pursue a big idea. It was Ohio in the late 1990s, it smelled like rat, and the question Jaak Panksepp was about to ask out loud would get him laughed at by people who were, in his professional assessment, not nearly as thoughtful as the animals they declined to anthropomorphize.
His team had been running young rats together in play conditions — rough-and-tumble wrestling, the kind of thing juvenile rats do when left to their own devices. They'd been using a bat detector, one of those handheld instruments that converts ultrasonic frequencies into audible range, so a researcher can hear what's happening in a pitch the human ear otherwise misses entirely. And during play, the rats were chirping. At 50 kHz — above the threshold of human hearing, somewhere in the neighborhood of a dog whistle. During rough-and-tumble wrestling, the chirps increased dramatically. When the play stopped, they subsided.
Panksepp had the thought that would follow him for the rest of his career, and that his colleagues would later cite as the kind of thing a serious scientist was not supposed to think: what if that's laughter?
He began tickling the rats.
The chirps went up. He expected that. What he did not quite expect — what became the moment the entire hypothesis clicked into a different gear — was what the rats did next. They sought out the tickling hand. They circled back to it. They placed their small forepaws on it and presented themselves for more. Not for food. Not from any conditioning Panksepp had established. They came back for the thing that made them make that sound.
Anyone who has ever watched a golden retriever demand to have its belly scratched will find this scene immediately legible. Which is part of the point, and part of what made Panksepp's colleagues so uncomfortable. Science has a word for the mistake of inferring mental states in animals from behavior that looks human: anthropomorphism. Panksepp knew the word. He was a careful thinker and a thorough experimentalist, and he also thought his colleagues were letting a methodological worry talk them out of a finding that was sitting right there making 50 kHz chirps and asking to be tickled again.
He spent years trying to publish it. The paper finally appeared in Physiology & Behavior in 2003, under a title that included the phrase "laughing rats" in quotation marks — a gesture toward scientific caution that did nothing to prevent the prominent emotion researchers who reviewed it from being offended. Panksepp died in 2017. He did not live to see a 2023 study that used cross-species neuroimaging to confirm that the same structure in the midbrain — the periaqueductal gray, a cluster of neurons older than the cerebral cortex, older than language, older than anything you might think of as distinctly human — mediates both the rats' 50 kHz play chirps and human involuntary laughter. Same address. Same ancient building. Two tenants who arrived ten million years apart.
There is something in the shape of Panksepp's career that deserves a moment's attention before we move on. The emotional quality of that career — the decades of being professionally mocked for taking seriously a thing that turned out to be true — is not irrelevant to the chapter's argument. Here was a scientist who noticed something real, was told by his field that he was anthropomorphizing, and kept going anyway. His instrument was picking up laughter. His colleagues' objection was essentially aesthetic: this sounds too much like a story, too warm, too human-facing. But "it sounds like a story" is not actually an objection to data. The data were there. The rats were there, pressing their forepaws against the hand that had made them make that sound. The vindication was posthumous, as the best vindications often are.
But let's not get ahead of ourselves. There is a question lurking inside the rat story, and it is the question this chapter is going to answer: if the machinery for laughter lives in a part of the brain that predates culture, language, and the entire edifice of learned behavior — if it's down there in the basement with the autonomic nervous system and the pain-modulation circuitry — then what does that mean for what we think laughter is?
Here is the answer, and it will probably feel backwards: we did not develop laughter to express humor. We developed humor to fill the laughter system we already had.
The gun came first. Culture loads the ammunition.
You are thinking, perhaps, that this is a semantic distinction. That laughter and humor are the same thing, just described from different ends. Let me show you that they are not.
In 2003, the same year Panksepp's rat paper finally appeared, Birk Wild and colleagues at the University of Freiburg published a paper in the neurology journal Brain that contained one of the stranger experiments in the emotional-expression literature. They put subjects in scanning equipment and tickled their feet. Real, involuntary laughter erupted — the kind you cannot prevent, the kind that makes you squirm and beg the other person to stop while simultaneously being unable to do anything useful to prevent it. The scan lit up the periaqueductal gray, the hypothalamus, the amygdala. Ancient structures, every one of them. The brainstem. The places where the brain keeps its most fundamental, least culturally mediated operations.
Then they asked the same subjects to fake a laugh.
A completely different pathway fired. Voluntary, performed laughter — the social kind, the polite-at-someone's-joke kind, the nervous-in-an-interview kind — routes through the motor cortex, the relatively recent machinery of conscious action. It goes up, not down. It passes through the structures that plan and decide, not the ones that simply respond.
Two laughs. Two pathways. Only one of them is old enough to predate culture.
Wild's most vivid evidence came from the lesion cases. These were patients with damage to the cortical layer — damage that had stripped away their ability to control voluntary movement and expression. They couldn't produce laughter on demand. If you asked them to laugh, they could not do it. But if you told them a joke, or tickled them, they laughed anyway — genuinely, helplessly, the real thing. The cultural layer was gone. The ancient machinery ran without it.
The periaqueductal gray — I'll mention the name once and then use "the old address," because the name is not the interesting part — is a small cluster of neurons in the midbrain that scientists had previously associated primarily with pain modulation and defensive responses. It's one of the oldest structures in the vertebrate brain. The 2013 neuroimaging work by Wattendorf and colleagues tickled subjects in an fMRI scanner specifically to capture involuntary laughter in progress, and found the old address more activated than any frontal region. More active than anything in the prefrontal cortex, which is where your personality lives, where your language lives, where your culturally acquired judgments about what is and isn't appropriate live. When you really laugh — when something gets you, when it bypasses your composure and your dignity and makes you helpless — you are laughing from somewhere older than any of that. Older than language. Older than the species.
This should feel a little strange. We tend to think of humor as among the most sophisticated of human accomplishments — complex, learned, culturally variable, the province of wit and intelligence. And it is. But laughter itself is not a sophisticated accomplishment. Laughter is reflex. It is the brain responding the way it responds to sudden safe violations of expectation, and it does this in the brainstem before the cortex has had a chance to have an opinion about whether anything is funny.
The cortex has the opinion. The brainstem already laughed.
The distinction between those two operations can occasionally become visible in its failure modes, and when it does, the picture is vertiginous. Gelastic epilepsy — from the Greek gelos, laughter — is a seizure disorder in which spontaneous, uncontrollable laughter is produced by abnormal brain activity, most often from a benign tumor called a hypothalamic hamartoma pressing against the structures that manage emotional expression. The laughter is physiologically complete. It involves vocalization, the characteristic facial movement, the diaphragm, the whole mechanical performance. Observers in the room will instinctively smile in response. But patients consistently report that they feel no amusement during or after the episode. Many report feeling frightened. The laugh circuit has been hijacked by a lesion the size of a marble, and it is producing the output without any of the input that normally precedes it. The hardware is running without the software. This is not a metaphor: a person is laughing at a funeral, not because anything is funny, but because the old address is misfiring.
When you hear about a condition like gelastic epilepsy for the first time, the image is disturbing. But as evidence it is clarifying. It shows that the mechanism that produces laughter and the mechanism that generates the experience of finding something funny are separable systems — that "laughing" and "being amused" are not the same operation, merely expressed differently. The laugh can run without the humor. The brain has a laugh button, and the button can be pressed accidentally, by neurology rather than by comedy. What normally presses it — a joke, a tickle, the hide-reveal-delight cycle of peekaboo — is the cultural or social content that loads the machinery. But the machinery was there first.
On the 113th day of his son William's life, in the summer of 1839, Charles Darwin played peekaboo.
Darwin was thirty years old and not yet famous for anything except being the naturalist who had returned from the voyage of the Beagle with ideas that were still sorting themselves out in his notebooks. He was not yet the man the world would know. He was a new father, and he was doing what new fathers do, and when he covered and uncovered his face his infant son laughed. Darwin noted this in his developmental diary — the careful observational record he'd been keeping since William's birth, with the attention he'd have given a barnacle — and wrote a sentence that should be better known than it is:
"Surprise was the chief cause of the amusement, as is the case to a large extent with the wit of grown-up persons."
He connected his pre-linguistic infant's laugh to the structure of adult comedy in one sentence. Then he sat on the observation for thirty-eight years.
The occasion that finally pulled it into print was a paper by the French philosopher Hippolyte Taine, who had published his own infant developmental diary in the journal Mind in 1877 and mentioned that infants appear to respond to surprising stimuli. Darwin, then sixty-eight and by this point extremely famous, submitted his own diary excerpts to the same journal immediately. He'd been keeping this observation in a drawer since 1839. Something about someone else publishing a lesser version of it moved him to act. You can almost feel the sixty-eight-year-old Darwin reading Taine's paper and thinking: I have the better version of this. I have had it for almost four decades. Out it came.
There's something wonderful about the image of the greatest naturalist of the nineteenth century playing peekaboo — deliberately, carefully, scientifically — and writing it down in a field diary. He was, in that moment, doing exactly what the best popular science writing asks scientists to do: taking the ordinary seriously. Not dismissing the baby's laugh as simply what babies do, but asking why. And arriving, in one sentence, at what developmental psychologists would spend the next century confirming: the mechanism of humor is the mechanism of surprise.
In the 1970s, Jerome Bruner studied peekaboo systematically. Infant subjects, multiple ages, multiple variations. Vary the timing. Vary the face. Vary whether it's the parent or a stranger. The laugh still comes. What Bruner found, and what Darwin had intuited, was that infants were not laughing at the specific face or the specific game — they were responding to the structure. The hide-reveal-delight cycle: tension (where did the face go?), resolution (there it is!), signal of safety. The laugh is about the pattern. Bruner called this a "tension-resolution cycle," and the structure he was describing is identical to the structure of a joke. The setup loads an expectation. The punchline violates and resolves it. The laugh is the sound of the resolution.
What makes this significant for our purposes — for the argument about laughter being pre-cultural hardware — is the age at which this happens. Four months. Before language. Before the concept of object permanence is fully installed. Before the infant has anything like a cultural repertoire. A four-month-old child cannot have been taught to find the hide-reveal-delight cycle funny. There is no lesson available, no model to follow, no cultural knowledge to invoke. The pattern is funny before the child has any framework for understanding why. The pattern was always going to be funny. The brain was waiting for it.
Sroufe and Wunsch, studying infant laughter systematically in 1972, tracked what stimuli reliably elicited laughter across the first year of life and documented its developmental arc. At four months: physical stimulation. Bouncing, tickling, an exaggerated face looming close. At six months: social triggers. The familiar person doing the unexpected thing. At eight to twelve months: cognitive triggers. Incongruity, the object in the wrong place, the category violation. The sequence matters because it is a progression from body to relationship to idea — from the most ancient stimulus type to the most recently evolved. The infant arrives at cognitive humor last because the cognitive apparatus required for it is being built last. But the laugh is there from the beginning. The baby laughs before it understands object permanence. The response precedes the understanding, as it always does.
Darwin, watching William laugh in 1839, saw all of this and didn't quite have the language for it yet. The periaqueductal gray was decades from being identified. The fMRI was a century and a half away. But the father in the garden playing peekaboo was sitting directly on top of the chapter's argument: the laugh precedes the explanation. The response comes before the theory.
Here is where we remove the possibility of imitation as an explanation.
In 1964, developmental psychologist Daniel Freedman published a study of infants born blind — children who had never seen a face, never watched another person smile or laugh, never had a visual model to copy. At all the right developmental moments — at the appearance of the social smile, at the onset of laughter during play — these infants did what sighted infants did. They smiled at touch. They laughed when startled in the pleasurable way by a familiar caregiver. Freedman documented that they "emit all the appropriate sounds while doing so," in the words of the ethologist Irenäus Eibl-Eibesfeldt, whose fieldwork across dozens of cultures extended Freedman's finding to children born simultaneously deaf and blind: "deaf and blind children smile when the mother caresses them; they laugh during play, cry when they hurt themselves." No visual model. No auditory model. The behaviors appeared on schedule.
If laughter were a learned response, a culturally transmitted behavior, an imitation of adults — the blind child would be the test that blew it up. No model was present. The laugh arrived anyway.
Eibl-Eibesfeldt's research method is worth pausing on, because it was designed specifically to catch people being unguarded. He filmed subjects across dozens of cultures using a camera fitted with a mirror-prism lens — a device that pointed the lens ninety degrees away from where it appeared to be pointing. His subjects thought they were being filmed from one direction; the camera was actually capturing their faces from another angle. This is, admittedly, the kind of research methodology that would raise some eyebrows today, but what it captured was genuine spontaneous expression rather than the performed version people produce when they know a camera is on them. The expressions he documented — the laugh, the smile, the cry, the fear response — appeared consistently across cultures, across ages, across populations with vastly different social norms. The laughter and the smile were there before the cultures had a chance to train them into or out of existence.
In 2008, researchers published an acoustic analysis of laughter in congenitally deaf college students — people who had been born deaf and had never heard laughter in their lives. They recorded the laugh vocalizations of deaf and hearing subjects and compared them. The temporal pattern was the same. The pitch contour was the same. The spectral features were the same. Deaf adults, without ever having heard a laugh, produced acoustically typical laughter. The sound of laughter was installed without instruction. It came preloaded.
It's hard to overstate how much work this finding does for the argument. One of the most common objections to laughter universality is that the specific acoustic form of laughter — the "ha-ha" pattern, the breathy panting — could be a cultural convention, transmitted through childhood imitation the way language is transmitted. The deaf adult data removes that possibility entirely. You cannot learn the sound of a thing you have never heard. The pattern is in the hardware.
What is in the hardware, precisely? A 2025 neuroimaging study found that at five months of age — before language, before much of what we recognize as social learning — infants already show enhanced functional connectivity in the brain networks associated with positive social expression, including early laughter. The infrastructure is organized before culture has had meaningful time to operate on it. The connections are being established in the first months of life, independently of whatever specific cultural environment the infant finds itself in.
The gun was there first. It was waiting for someone to load it.
At this point we have moved from a lab in Ohio to an fMRI scanner in Freiburg to a Victorian garden to studies of blind infants and deaf adults. The argument has been almost entirely anatomical. Let's lift up, geographically and evolutionarily, and look at the bigger picture.
In 2009, Marina Davila Ross went to Zambia with a microphone.
She worked at Chimfunshi Wildlife Orphanage, where infant and juvenile chimpanzees were housed, along with other ape species at facilities across Europe and Africa. Her project was to tickle them — systematically, with the appropriate permissions and approvals, across five different ape species — and to record what happened. Infant chimpanzees. Bonobos. Gorillas. Orangutans. And, as the human control, human infants.
What happened was that they all laughed. Not identically — there is no single acoustic signature that spans the great apes unchanged. The chimp laugh is a rapid breathy panting, pushed out on both the inhale and the exhale rather than only on the exhale the way human laughter works. The gorilla's version is deeper and less rapid. The orangutan's is most remote from our own. But the variations were not random. When Davila Ross and her colleagues analyzed the acoustic structure of the play vocalizations and mapped the similarities onto a chart, the result looked like a family tree. The species whose laughter sounded most similar to human laughter were the chimpanzees and bonobos — the species most closely related to us by genetics. Then gorillas. Then orangutans, whose laughter-equivalents were most distant from our own. The acoustic similarities tracked the evolutionary tree almost exactly. This was not convergent evolution — not five separate species independently arriving at the same vocalization because it happened to be useful. This was inheritance. An ancient sound, passed down through ten to sixteen million years of primate lineage, that each species had modified according to its own anatomical constraints but never fully abandoned.
Darwin noticed this in 1872, fifty-seven years before the acoustic analysis that proved it. In The Expression of the Emotions in Man and Animals — his other great book, the one overshadowed by The Origin of Species and The Descent of Man but arguably more interesting for our purposes — he described the sounds made by chimpanzees and orangutans being tickled: a rapid, breathy panting. Not the voiced ha-ha of human laughter, but produced in response to the same social triggers, in the same social situations, with the same apparent emotional valence. He concluded that "the young of the anthropomorphous apes express pleasure by uttering a reiterated sound, analogous to our laughter." Not identical. Analogous.
He was right. He was writing about inherited behavior in 1872. Davila Ross gave him the acoustic phonetics to prove it in 2009.
The line now runs: Bowling Green lab rats, Ohio, 1990s. Chimfunshi Wildlife Orphanage, Zambia, 2009. The Victorian garden, 1839. The fMRI scanner, Freiburg, 2003. The periaqueductal gray, midbrain, approximately 300 million years of vertebrate history. The laugh was always there. It was running before the jokes started.
Before we can wrap this argument up and present it as settled, there is a display issue to address, and it comes from a culture that has thought about this more carefully than most.
In Japan, covering one's mouth while laughing is a practice that dates at least to the Heian period — roughly 800 to 1185 CE — and is connected to the concept of wa, the maintenance of group harmony and the avoidance of asserting individual presence at the expense of collective cohesion. In traditional contexts, an open laugh — full, unguarded, showing teeth — can be read as aggressive, self-assertive, a kind of social intrusion. The polite response is to cover: to contain the expression before it imposes itself on others. Contemporary surveys find the behavior more common among women, in formal settings, and among older age groups; it has been moderating with globalization. But its roots are a thousand years old.
What does this mean for the claim that laughter is universal? A skeptic might read the Japanese display norm as evidence against universality: if laughter were genuinely pre-cultural, why would culture be able to intercept it so reliably?
The answer is that intercept is precisely the right word. The capacity is not suppressed — the laugh still happens; it is covered. The biological substrate is present and firing. What the cultural norm has managed is the display of laughter to others, not the production of laughter in the body. This is an important distinction. Culture operates on the surface — on what you show other people, on what your face broadcasts, on the social signal your body sends. It does not, apparently, have reliable access to the ancient machinery running underneath. The laugh erupts; the hand comes up; the display is managed. The substrate is intact, running at its same old depth, producing the same old output. Culture intercepts the signal before others can receive it.
If anything, the Japanese mouth-covering norm is evidence for the biological argument, not against it. You do not develop an elaborate, culturally transmitted, thousand-year-old behavioral norm for managing an impulse that does not exist. The norm presupposes the impulse. The regulation presupposes the thing being regulated. The wa tradition confirms, in its thousand years of continuous practice, that the laugh impulse is present and powerful enough to require management.
Cross-cultural work on laughter display has documented similar dynamics in other contexts. In several West African traditions, laughter accompanies the delivery of news that would, in Western contexts, be received with solemnity. In several Southeast Asian contexts, laughter in response to bad news is a social lubricant, a signal that the speaker intends goodwill and the situation will not be allowed to become threatening. These uses of laughter — for grief management, for deflection, for social safety signaling — are real and culturally specific. But notice what they share: they are all uses of laughter, deployments of the mechanism for socially specific purposes. The mechanism is the constant. The use is the variable. If laughter were a culturally constructed behavior rather than a biological substrate, it could not be recruited for these additional social purposes any more than a tool can be repurposed if it doesn't exist.
And now we have to talk about the jokes that stopped.
This is the part of the argument that would be easy to skip, and the part the argument cannot survive without. Because here is the obvious objection: if laughter is pre-cultural hardware, why is almost nothing cross-culturally funny? The Yanomami obscene-names joke requires knowing Yanomami name taboos. The Aristophanes pun requires the specific lexical ambiguity in ancient Greek. A joke about a Sumerian measurement standard requires being Sumerian. The machinery may be universal. The triggers are not. And the triggers are — in some important sense — what humor is.
The honest answer is the one the chapter is built toward: the capacity is universal; the content is culturally specific. But to make that answer precise rather than evasive, we need to watch some jokes die.
Consider Aristophanes. In 423 BCE, he entered The Clouds at the City Dionysia in Athens — the major dramatic festival, the most prestigious competition — and it came last. Dead last. Aristophanes was mortified. He revised the play and complained, at some length in his other work, about the audience's obtuseness. The revised version is the one that has survived, and it still doesn't work for modern readers. It doesn't really work for classicists, who can explain every joke and still not quite feel them land.
The play mocks Socrates. Not the Socrates of the dialogues — the philosophical saint, the noble martyr — but the Socrates who was, in 423 BCE, a recognizable street-corner figure in Athens: a weird man who walked around barefoot asking unanswerable questions and corrupting young people with argumentation. The humor required several things to happen simultaneously. You needed to know Socrates personally, or know someone who did. You needed to share a specific Athenian anxiety about sophistry — about the corrosive effect of all this fancy new rhetoric on traditional values — that was entirely local to a specific decade in a specific city. You needed to be able to parse the parody of specific rhetorical techniques that were fashionable in 423 BCE and have not been fashionable since. And you needed to hear the wordplay that Aristophanes embedded in the Greek meter — puns that exist in the acoustic texture of the original performance and have no home in translation.
All four conditions are gone. You can read a forty-page scholarly introduction explaining each one, and the jokes still won't land. Not because the scholarly introduction was inadequate, but because the machinery requires the fuel and the fuel ran out two and a half millennia ago. The biological substrate — pattern violation, status reversal, shared recognition of a butt — is perfectly intact. The cultural superstructure that loaded those mechanisms has evaporated. The gun is still there. The ammunition is dust.
Here is the most precise illustration of this problem available: there is a joke in Aristophanes that requires knowing a specific Greek word that means both "lantern" and a type of fish. Without that knowledge, the punchline is literally not there. The word is gone from the translation, because the word cannot be translated. What appears in the footnotes as "[untranslatable wordplay]" was, in the original performance, the entire payload. The joke was the pun. The pun is now an absence. Translators of Aristophanes have faced this problem consistently for two centuries — the choice between translating the words and losing the joke, or inventing a new joke and losing the words. There is no third option. The humor was the specific sound of a specific language at a specific moment, and that moment is gone.
Now consider the Philogelos — the collection of 264 jokes that survives from the Greek-speaking Roman world, probably compiled around the fourth or fifth century CE. The Philogelos contains a long series of jokes about the scholastikos — the absent-minded scholar, the man so thoroughly educated in abstract things that he has become catastrophically incompetent at basic life tasks. In one joke, a scholastikos is asked how old he is and replies that he'll have to check his records. In another, he returns from abroad and asks his slave whether his father is still alive — because he doesn't know. In the most famous one, he stands in front of a mirror with his eyes closed so he can see how he looks when asleep.
These jokes still land. You can feel the mechanism fire. You see the setup, you see the error accumulate, you get the punchline. The absent-minded professor is a living archetype; you have met several.
But something is slightly off about the register. Modern readers laugh with the scholastikos, finding him charming, perhaps admirable in his dedication to unworldly things. The original audience laughed down. Over-education was, in the Greco-Roman world, a form of contemptibility — a man who had prioritized abstract knowledge over practical wisdom was less of a man, less of a citizen. The scholastikos was not an endearing eccentric. He was a failure. The audience was invited to feel superior to him, and they did.
The humor mechanism — status incongruity, the comedy of someone behaving in a way that violates their presumed competence — is intact and firing. But the cultural content that told the audience what to feel about the target has inverted. The butt of the joke has become, in modern eyes, someone most readers would rather be than most people they know. Same punchline. Different emotional register. The machinery is working; the direction of the joke has reversed.
And then there is flyting, which is the case that most precisely illustrates what humor requires beyond the biological substrate.
From roughly the fifth through the sixteenth centuries in England and Scotland, flyting was a recognized performing art: two poets trading elaborately structured, devastatingly personal insults in verse, before a royal court, with the formality of a ritual duel. The Flyting of Dunbar and Kennedy, performed before James IV of Scotland around 1503, is the most famous example — so aggressively alliterative, so luridly specific in its abuse, that scholars have compared it to battle rap, which is a comparison that holds up better than it should across five hundred years and a social context so different as to be almost unrecognizable. Kennedy calls Dunbar a "monstrous mouth" and a "rottin crok." Dunbar responds with equal technical precision and equal savagery.
Modern readers can appreciate the craft. They cannot experience the event.
What made flyting funny — not just impressive, but genuinely funny, in the way that the assembled court apparently found it genuinely funny — was a double register that is now simply unavailable. The audience had to simultaneously experience the insults as real enough to be humiliating and performed enough to be safe. The humor lived in the gap between those two experiences, in the strange pleasure of watching abuse that was genuinely wounding being received by someone who had consented to receive it. Strip the craft — remove the metrical precision and the studied alliteration — and you have sincere abuse. Strip the threat — remove the sense that the insults were actually drawing blood — and you have a poetry recitation. The form required holding both at once, and the social apparatus that allowed a court audience to do that — the shared norms about when performance began and ended, the understood rule about what was game and what was earnest — is gone. Not inaccessible. Gone.
The insult-contest impulse is not gone. It survives in rap battles, in roasts, in the competitive ribbing that occurs in virtually every social group that has been studied anywhere in the world. The impulse is, as it turns out, universal — which is exactly what you'd predict if the underlying mechanism were pre-cultural. But the form flyting took was mortal. It required a specific social apparatus to hold it together, and that apparatus dissolved. The biology persisted. The content expired.
These three cases — Aristophanes, the scholastikos, flyting — form a graduated argument. The Clouds shows humor dying when all four simultaneous cultural dependencies vanish at once. The scholastikos shows humor surviving mechanically while its emotional valence inverts. Flyting shows the humor impulse surviving while the specific form that housed it collapses. Together they say: the machinery is constant. The fuel is not. And the fuel can expire, invert, or fragment, leaving the machinery running on empty.
There is one more objection to handle before we can close, and it comes from the most rigorous quarter available.
Lisa Feldman Barrett, the neuroscientist and emotion researcher, argues in her 2017 book How Emotions Are Made that emotions are not universal biological programs expressed identically across cultures. They are constructed experiences, assembled from interoceptive sensations and cultural concepts. On Barrett's view, there is no universal "laughter emotion" that fires the same way everywhere — even laughter is a concept that varies cross-culturally, shaped by the culture that surrounds it.
She is not wrong about the cultural variability of emotional expression. Her Namibia data are directly relevant here. In 2014, Gendron, Roberson, van der Vyver, and Barrett played emotional vocalizations to participants from the Himba, a semi-nomadic pastoralist people in rural Namibia with minimal Western media exposure. Rather than using the forced-choice method — picking from a predetermined list of emotion categories — they allowed free labeling: what is this person experiencing? Most emotional vocalizations did not map reliably onto Western categories. The Himba heard something in an angry vocalization that the Western model didn't predict. The fear sound mapped differently. The category structure of emotion turned out not to be universal in the way earlier researchers had assumed.
Laughter was the exception.
Laughter was the most cross-culturally robust vocalization in the study — the one that participants across cultural backgrounds consistently identified as positive and amusing. It wasn't perfect; the conceptual layer around laughter (what it means in a specific social context, what emotion category it belongs to) varied. But the basic identification — this sound signals something pleasurable — held up across the widest range of cultural backgrounds tested. Barrett's study, designed to challenge universalist claims about emotional expression, found that laughter was the one vocalization that held up across all the cultural groups tested. The result cuts both ways: laughter passed the test that most emotions failed, but "passing the test" still means passing a test that was designed to find failures, and the test found some.
Here's the key move, though. Barrett's critique applies most forcefully to voluntary, performed laughter — the social kind, the polite kind, the kind that routes through the motor cortex. When we perform laughter, when we laugh because the situation calls for it, when we laugh at our boss's joke or out of social solidarity — that laughter is culturally constructed. It is learned behavior, shaped by social expectations, running through the cortex where culture lives.
The pre-cultural argument rests on involuntary laughter. The kind that erupts before you can stop it, the kind the rats were making, the kind that fires in Wild's lesion patients who can no longer fake a laugh but still burst out laughing at jokes and tickling. That laughter is subcortical. It is, in the most literal anatomical sense, below culture. Barrett's constructivist machinery is powerful, but it requires the cortex to operate on, and the laugh we're talking about is happening somewhere the cortex doesn't get a vote.
This is not a refutation of Barrett. She is right about the cultural construction of emotional expression, right about the variability of emotional categories, right that the old "universal emotions" literature overclaimed. But there are two kinds of laughter running on two different pathways, and only one of them is in her territory. The ancient one — the brainstem one, the one the rats are making at 50 kHz, the one the blind infants produce on schedule, the one the deaf adults produce with acoustically perfect fidelity to a sound they have never heard — that one is running below the layer she is analyzing. The Japanese display norm is cultural management of the surface. The gelastic seizure is the substrate running without management at all. Both confirm the same thing: there is something underneath, running independently, that culture works with rather than on.
Venezuela, 1964. Napoleon Chagnon, anthropologist, five months into his first extended fieldwork among the Yanomami of the upper Orinoco.
He had been collecting genealogical data, the foundational work for understanding kinship and social structure. It had been slow going. The Yanomami have a strong name taboo — using the name of a living person in their presence is deeply rude, an invasion of their personal identity — and so his informants had been, understandably, reluctant to provide real names. He had worked around this by asking about other villages' inhabitants, not the people immediately present.
What he had not understood — what he would not understand for five months — was that he was being fed invented names. Not random nonsense names. Names carefully constructed to be maximally obscene. The headman's name, which he had solemnly written in his notebook: "long dong." The headman's daughter: "fart breath." Other villagers: "eagle shit," "asshole." Chagnon had been dutifully recording these, practicing them, preparing to use them in his research.
The Yanomami, meanwhile, had been watching him do this with building delight. The game had spread to neighboring villages. People who had never met Chagnon knew about the anthropologist writing down the obscene names.
Chagnon was visiting a neighboring village when he mentioned the headman's wife's name — the woman he had been told was called something that, in Yanomami, was deeply inappropriate — in ordinary conversation.
He described what happened: "A stunned silence followed, and then a villagewide roar of uncontrollable laughter, choking, gasping, and howling followed."
Five months. They had built a five-month prank requiring transgression, careful escalation, multiple participants, and a final public reveal — a prank whose structure maps directly onto every comedic form we have discussed in this chapter. A setup that loads an expectation. A punchline that violates it. Transgression made safe by the performance frame of the joke. And that final, irreducible, cross-cultural response: the roar. The choking. The howling. The sound the old address makes when the gun fires.
Chagnon had to discard five months of field data and start over.
The Yanomami joke required knowing Yanomami name taboos. It required five months of careful setup. It required an anthropologist diligent enough to keep writing down what he was told without checking. The cultural content was entirely, irreducibly local — nothing about the specific material would have meant anything to someone outside that community, that tradition, that moment. If you tried to tell this joke in a different language to a different audience, you would be recounting a story about someone being told embarrassing things in a language they didn't know, which is not funny in the same way.
But the laughter was the same laughter.
The mechanism that fired in that Yanomami village — the ancient brainstem machinery running on culturally specific fuel, the tension-resolution cycle completing in a roar of choking and howling — is the same mechanism that was running in the Bowling Green lab, the same mechanism Darwin documented in Victorian England, the same mechanism the Freiburg scanner lit up in the brainstem. The biological constant. The cultural variable.
Consider what the Yanomami constructed. They identified a target — the foreign anthropologist, conspicuously earnest, conspicuously writing everything down. They identified a mechanism — the name taboo, the deepest available cultural violation. They built a setup lasting five months, during which the victim obligingly kept performing the setup by continuing to record. They managed the timing, containing the payoff until the moment of maximum effect: a public context, a neighboring village, the moment when the name entered actual use in an actual conversation. The construction of the joke was meticulous, patient, formally sophisticated. The structure it built was universal: victim, transgression, delayed reveal, public explosion.
The laughter at the end was not Western laughter or Yanomami laughter. It was human laughter, the kind that runs on the oldest hardware in the building, the kind that the rats made in Ohio and the chimps made in Zambia and the blind infants produced in the absence of any face to copy. The fuel was entirely local. The mechanism was as old as the vertebrate nervous system.
The machinery is universal. It was running before culture gave it anything to work with. It will run after every joke we know has turned to dust. The biology is the constant. The content is the variable. And the variable will expire, and invert, and fragment, leaving the mechanism running on empty until the next culture loads it with new ammunition.
But the gun will still be there. In the brainstem. At the same old address. Running at 50 kHz, pressing its small forepaws against the tickling hand and presenting itself for more.
Author's note on sources: The Wild et al. two-pathway model and lesion cases appear in Brain 126(10):2121 (2003). The Wattendorf et al. PAG imaging study is in Cerebral Cortex 23(6):1280 (2013). Cross-species PAG confirmation: PubMed 37516112 (2023). Panksepp & Burgdorf, "Laughing rats": Physiology & Behavior 79(3):533 (2003), PubMed 12954448. Davila Ross et al. phylogenetics: Current Biology 19(13):1106 (2009). Darwin's Expression of the Emotions (1872), Darwin's infant diary: Mind os-2(7):285 (1877). Sroufe & Wunsch, infant laughter onset: Child Development 43(4):1326 (1972). Freedman blind infant study: J. Child Psychology and Psychiatry 5:171 (1964), PubMed 14262770. Eibl-Eibesfeldt: Human Ethology, Aldine de Gruyter (1989). Deaf adult laughter acoustics: PMC 2809699 (2008). Infant DMN at 5 months: PMC 12464472 (2025). Gelastic epilepsy and hypothalamic hamartoma: PMC6277842 (2018). Gendron et al. Himba vocalizations: PMC 3989551 (2014). Chagnon: The Yanomamö (1968 and subsequent editions). Japanese mouth-covering cultural norm: Sauter et al. (2024), J. Nonverbal Behavior. The flyting comparison to hip hop: Flynn (2014), Oral Tradition 29(1). Mary Beard on the Philogelos: Laughter in Ancient Rome, UC Press (2014). Barrett: How Emotions Are Made, Houghton Mifflin Harcourt (2017).