Chapter 1 — Born Laughing
The device was meant for bats.
Jaak Panksepp's laboratory at Bowling Green State University in Ohio had acquired a bat detector — a gadget that does exactly what its name suggests, which is to say it converts ultrasonic frequencies into audible range and lets you eavesdrop on creatures who have been communicating in registers you can't hear. You can buy a bat detector in a birding catalog for about a hundred dollars. Panksepp had borrowed the idea of using one to listen to rats.
This was the late 1990s, and Panksepp was already a controversial figure in neuroscience for the sin of suggesting that animals might have emotional states — not just stimulus-response machinery, not just learned behavior shaped by reward and punishment, but actual affective experience. Emotions. The argument that a rat might want something, or enjoy something, or feel something recognizable to a human as positive or negative — this was not a popular position in a field that had spent decades scrubbing subjective language from animal research. His colleagues' eyebrows had a workout.
But he was watching young rats play, and something kept nagging at him. When the rats tumbled together in rough-and-tumble play — the biting and pinning and chasing that rat pups do with the focused intensity of small wrestlers — the bat detector picked up chirping. Bursts of it. At 50 kHz, far above anything a human ear could catch without help. The rats were making sounds he hadn't known they were making, and they were making them specifically when they played.
One morning, Panksepp had the thought.
What if that's laughter?
He was, he knew, about to become even more unpopular at seminars. But he tickled the rats.
The 50 kHz chirps increased dramatically.
Now here is the part that tends to stop people. Because the next finding was not that rats tolerate tickling, or that they chirped when physically stimulated, or anything else that a cautious scientist might describe in passive, affect-free prose. The next finding was that the rats sought out the tickling hand. After Panksepp tickled them and withdrew, the rats circled back. They placed their small forepaws on the hand. They presented themselves for more.
Not for food. There was no food involved. Not from conditioning — the experiment was designed to rule that out. When Panksepp offered a tickling hand and a neutral control hand, the rats chose the tickling hand. They came back, and came back again, pressing their tiny forepaws against a human hand, seeking the thing that had made them make that sound.
Panksepp spent years trying to get this published. Prominent emotion researchers told him he was anthropomorphizing. He was projecting human experience onto an animal. He was, in the academic phrasing, being insufficiently rigorous about the gulf between human and rat cognition — as if the gulf were so self-evidently vast that no evidence could cross it. The paper finally appeared in Physiology & Behavior in 2003, and it was met with roughly the reception you'd expect for a paper titled, with almost deliberately provocative scare quotes, "Laughing" rats and the evolutionary antecedents of human joy.
Jaak Panksepp died in 2017.
In 2023, six years later, a cross-species neuroimaging study confirmed that the same ancient brain structure — the periaqueductal gray, a midbrain region older than the cortex, older than language, older than anything you'd call culture — mediates both the rat's 50 kHz play-chirps and human involuntary laughter. The same address. The same machinery. Across two species separated by eighty million years of evolution.
He was right. The rat was laughing.
There is something I should tell you about the periaqueductal gray before we go further, because it will keep coming up and it is worth knowing what you're dealing with. The PAG — neuroscientists abbreviate it, which is what you do when a name is both technical and indispensable — sits in the midbrain, in the part of your brain that predates everything you'd associate with distinctly human cognition. Before the 2013 fMRI tickling study by Wattendorf and colleagues, the PAG was known primarily in connection with pain modulation and defensive responses — the ancient alarm system, the brainstem's emergency management. What Wattendorf found was that when you put a person in an fMRI scanner and tickle their foot until they laugh — involuntary, genuine, couldn't-stop-it-if-you-tried laughter — the PAG lights up more brightly than any other structure. More than the prefrontal cortex. More than the structures that plan, reason, have opinions, and attend faculty meetings. The old machinery. The basement of the brain.
The rat presses its paw against your hand. Something about that image. We'll be coming back to it.
Here is the thing about laughing that almost no one notices: there are two of them.
This seems like a strange claim. You've been doing it your whole life. You know what it feels like from the inside. But in 2003 — the same year Panksepp's rat paper finally appeared, which may or may not be a coincidence the universe intended — Birk Wild and colleagues at the University of Freiburg published a paper in the journal Brain that established something genuinely startling: involuntary laughter and voluntary laughter run on anatomically distinct neural pathways.
When you really, helplessly, can't-help-it laugh at something — the PAG, the brainstem, ancient subcortical structures, the evolutionarily old apartment building of the limbic system. When you're at a work party and your boss says something that doesn't quite qualify as funny but social norms require a response — that routes through the motor cortex. The planning brain. The "I'll have the chicken" part of your brain.
Wild called them "partially independent systems." This is the polite description. What the lesion cases showed was somewhat less partial.
When patients suffer cortical damage — strokes, injuries, lesions of the kind that are always described as "significant" and "life-altering" in the case files — they sometimes lose something curious: the ability to perform a laugh on demand. Ask them to laugh for you, demonstrate the sound, fake it for a photograph, and they can't. The instruction gets processed, the will is there, but the motor production pathway has been disrupted. The cortical circuit is down.
These same patients, when they hear a good joke, or when someone tickles them, burst into genuine laughter.
Let me be clear about what this means. They cannot manufacture the laugh they use in social situations — the laugh that functions as punctuation, as affiliation, as signal of goodwill. They cannot produce the laugh that says I acknowledge your wit and wish to continue this pleasant interaction. That laugh is gone. But the ancient machinery — the subcortical circuit that has been running longer than the human species, longer than homo sapiens, longer than any cultural instruction manual for when to laugh and how — runs without it. Strip away the layer that culture operates on and the biological substrate keeps going, uninterrupted.
Two systems. Two completely different laughs. One is old enough to be in rats. One is new enough to attend your nephew's wedding and produce appropriate sounds at the toast.
This matters enormously for what follows. When we argue — and we are going to argue, at length, with genuine counterevidence, with cases of humor so culturally specific that it's been dead for two thousand years and counting — about whether laughter is universal, we need to be precise about which laugh we mean. The Wild et al. finding gives us a scalpel. The ancient laugh, the involuntary laugh, the laugh that routes through the part of your brain you share with a rat: that's what's universal. The laugh that says yes, perfectly appropriate, I acknowledge your wit: that's a performance, and performances are learned, and what is learned is not universal, and that's a different question — interesting but different.
The architecture is everywhere. The furniture is local.
In the summer of 1839, Charles Darwin was thirty years old, not yet famous, recently returned from the Beagle voyage, settling into married life, and had just become a father for the first time. His son William was a plump and evidently cooperative research subject.
Darwin kept a diary about William. This was not unusual for the period — observing infant development as a scientific project had a small but serious following among Victorian intellectuals who noticed that babies are essentially tiny humans who haven't had time to learn anything yet and are therefore potentially revealing about what's innate. What was unusual was what Darwin was looking for, and what he found it worth writing down.
On Day 113 of William's life, Darwin played peekaboo. He covered his face and uncovered it. His son laughed.
Darwin wrote: "surprise was the chief cause of the amusement, as is the case to a large extent with the wit of grown-up persons."
One sentence. His pre-linguistic infant — who had no language, who had not yet developed object permanence, who could not reliably understand that things continue to exist when hidden from view — had just demonstrated, in Darwin's eyes, the same mechanism as adult comedy. Surprise is the engine. The violation of expectation followed by its resolution. The momentary where did it go? and then the flood of there it is. The infant was laughing at the structure of the event. Not at a face, not at Darwin in particular, but at the pattern: tension, release, signal of safety.
He sat on this observation for thirty-eight years.
Darwin published his infant diary in 1877, in the journal Mind, when a French philosopher named Hippolyte Taine published his own child observation paper and gave Darwin the occasion to deploy William's data. By then Darwin had spent decades thinking about emotional expression, had written a whole book on the subject in 1872, and had been watching human faces and primate faces for the better part of four decades. The 1877 paper runs only ten pages. It is one of the most compressed pieces of developmental psychology ever written, anchored by a man playing peekaboo in his study on a summer afternoon and recognizing, in one sentence, what it was about.
What it was about was structure.
Jerome Bruner, the developmental psychologist, studied peekaboo systematically in the 1970s and confirmed what Darwin had intuited: infants don't respond to the specific face. They respond to the pattern. Vary the face — different person, different expression, swap in a stranger — the laugh still comes. Use an object instead of a face and run the same hide-reveal cycle: the laugh still comes. What the infant is responding to is the expectation cycle itself, the arousal and the resolution, the gone and the there. The form, not the content.
This is why peekaboo exists in essentially every documented culture. Some versions use sound rather than visual disappearance. Some use objects. Some have elaborate ritual variations that an anthropologist has dutifully recorded with a sense of discovery, before noticing that the underlying structure is identical to what his own grandmother did with a dish towel. The core is always identical: tension, release, signal of safety. An infant of four or five months — before language, before the cognitive development we associate with "getting" jokes — will laugh at this structure reliably, cross-culturally, because what the structure exploits is older than culture.
And this is also why peekaboo works as a description of every joke you've ever found funny. The tension-resolution cycle that Sroufe identified in the 1970s as the engine of infant laughter — rising arousal, sudden release, signal that the arousal was safe — is the same cycle that underlies a punchline. Setup, beat, release. The joke works when the expectation is violated in a way that resolves safely, from a safe social position, with a recognized signal that the threat was not real. The baby at four months is doing the same thing the audience at a comedy club is doing. The machinery was installed at four months. The club is a very fancy peekaboo.
Sroufe and Wunsch tracked this development across the first year of life, documenting what elicited laughter from infants at four months versus eight months versus twelve months. At four months: physical stimulation, sudden movement, exaggerated faces. At eight months: social triggers begin to appear — the familiar voice, the familiar face, the social game. At twelve months: cognitive triggers, object play, the early forms of expectation-violation that require actual understanding of what the expectation was. The hardware develops from the body up. Touch first, then faces, then cognition. Laughter is there at the beginning, waiting for new inputs.
Darwin was thirty years old with a baby and a notebook. He found the engine. He wrote it down in one sentence. And then, being Darwin, he filed it and turned his attention to barnacles for the next several years.
You can argue that Darwin's son learned to laugh from watching Darwin laugh. This is the obvious objection, and it is not a foolish one. Infants are extraordinary imitators; the question of which emotional expressions are innate versus socially installed is genuinely contested in several areas of developmental psychology. Peekaboo at Day 113 is charming evidence, but it is not conclusive. William could have been copying his father's expression. He had had four months of watching Darwin's face.
For conclusive, you need the natural experiments — the cases where imitation is structurally impossible.
In 1964, a developmental psychologist named Daniel Freedman published a study in the Journal of Child Psychology and Psychiatry that has been quietly influential ever since. Freedman studied infants born blind — children who had never seen a face, never watched anyone smile or laugh, never had access to the visual model that imitation theory requires. The question was straightforward: would laughter and smiling still appear, on schedule, without any visual model to copy?
They did. Blind infants smiled at touch. They laughed during play. They produced the full range of appropriate positive vocalizations at the appropriate developmental moments, with roughly normal timing. No model existed. The behavior appeared anyway.
The hardware was installed. Nobody taught them the song. They were already singing it.
A 2008 acoustic study extended this finding in an unexpected direction. Congenitally deaf college students — people who had grown up without hearing, who had never heard what laughter sounds like — produced laughter that was acoustically indistinguishable from hearing people's. The same temporal pattern: the rhythmic bursts, the intervals, the way laughter accelerates and decelerates. The same pitch contour. The same structural signature that you would recognize immediately as laughter if you heard it on a tape, even with no other context. These people had never heard this sound. They produced it correctly anyway.
The argument from imitation requires a model. There were no models. The behavior appeared in full and correct form.
Irenäus Eibl-Eibesfeldt, the German ethologist, spent decades conducting fieldwork across dozens of cultures, filming emotional behavior using a mirror-prism lens that deflected the camera's apparent direction so that subjects thought they weren't being filmed. The goal was to capture unguarded behavior — the emotional expression that self-consciousness would modify or suppress — across the widest possible range of human populations. His 1989 synthesis, Human Ethology, is one of the most thorough comparative records of human emotional expression ever assembled.
Among his subjects were children born simultaneously deaf and blind. No visual models. No auditory models. Nothing to imitate at all. The question was as stark as it gets: does the behavior appear in the absence of any possible input from culture?
Eibl-Eibesfeldt's answer: "deaf and blind children smile when the mother caresses them; they laugh during play, cry when they hurt themselves, and emit all the appropriate sounds while doing so."
Nothing to imitate. All the right responses.
There is a thought experiment I keep returning to, not as argument but as image. Imagine you are trying to describe the color blue to someone who has been blind from birth. The concept requires a sensory experience they have never had. You can explain wavelengths. You can explain associations. But the experience of blue — that particular qualia — is, in some fundamental sense, inaccessible without the original input. Now imagine that the person spontaneously describes blue to you, accurately, with appropriate emotional resonance, having arrived at the concept from first principles without any external input at all.
That's not quite what's happening when a blind-deaf child laughs during play. But it is in the same neighborhood. The machinery is running without any external blueprint. The signal is correct without any signal to copy. That's the point.
Darwin, in 1872, wrote a book about emotional expression. The Expression of the Emotions in Man and Animals is, in the way of all Darwin's books, both more and less than its title suggests — it sprawls in directions the title doesn't quite announce, pursues side-questions with Victorian thoroughness, and contains, near chapter eight, a careful description of what happens when you tickle a chimpanzee or an orangutan.
They pant.
Not the voiced, musical "ha-ha-ha" of human laughter. Darwin was careful about this distinction. The ape's sound during play and tickling is a rapid, rhythmic exhalation, breathy rather than voiced, recognizable in structure though not in precise acoustic profile. But it is produced in response to the same triggers — social play, tickling, rough-and-tumble interaction with trusted companions — and Darwin concluded that it was "analogous to our laughter." The apes were doing something in the laughter family. The family resemblance was legible even without acoustic instrumentation. He had observation and the courage of his own taxonomy.
One hundred and thirty-seven years later, Marina Davila Ross and her colleagues at the University of Portsmouth went to Chimfunshi Wildlife Orphanage in Zambia and did the acoustic analysis that Darwin could not.
The study — published in Current Biology in 2009 — recorded play vocalizations from chimpanzees, bonobos, gorillas, orangutans, and humans during tickling. Then they did something elegant: they mapped the acoustic similarities between species onto the independently established evolutionary tree. If laughter evolved once in a common ancestor and was inherited by descendant species, acoustic similarities should track phylogenetic relatedness. If laughter evolved independently in different lineages through convergent evolution — the way eyes evolved independently in vertebrates and mollusks because seeing is useful — acoustic similarities might cluster differently, or randomly.
The result: acoustic similarities tracked phylogenetic relatedness almost exactly. Human laughter was most similar to chimpanzee laughter; chimpanzee laughter was more similar to bonobo laughter than to gorilla laughter; gorilla laughter clustered with the other African great apes away from orangutans; and so on down the tree. The laughter phylogeny matched the genetic phylogeny. Species by species, the acoustic distances between laughs correlated with the genetic distances between lineages.
What this means is that laughter did not evolve independently in humans and then independently in apes through convergent evolution, the way wings evolved separately in birds and bats because flight is useful. Laughter was there before the lineage split. It is an inherited vocalization, carried in the genome across branchings of a family tree that began diverging over ten to sixteen million years ago. The chimp's panting and your laugh at something that genuinely surprised you share a common ancestor that was doing the same thing before either of your species existed.
Darwin noticed the family resemblance in 1872 with nothing but his eyes and his willingness to say what he saw. Davila Ross gave him the acoustic proof in 2009.
Darwin's panting ape. Panksepp's chirping rats. The 2023 PAG study linking Ohio lab work to Zambia fieldwork across species lines. The line runs continuously from the midbrain structures of a rat to the convulsing laughter of a human at something genuinely, helplessly funny. It is not a metaphor. It is a phylogeny.
The gun was installed before anyone lived to pull the trigger.
Now I need to tell you about the jokes that died.
The comfortable version of the argument I've been making — laughter is universal, the machinery runs everywhere, biology over culture — is too comfortable. And comfortable arguments are the ones worth testing hardest, because they have a tendency to be true in the ways that are easy to verify and wrong in the ways that are easy to overlook. So here is the testing.
In 423 BCE, Aristophanes entered The Clouds at the City Dionysia in Athens. The City Dionysia was the Super Bowl of Athenian comic performance — public, competitive, civic, attended by thousands. Aristophanes was already a proven playwright with successful productions behind him, and The Clouds was his most ambitious and personally invested work. He was going after Socrates: a real person, recognizable, a street-corner presence, a man whose specific mannerisms and rhetorical habits his Athenian audience could perform back to each other at dinner parties.
It came last. Dead last, in a competition with three entrants.
Aristophanes was so offended and confused by this that he revised the play and complained publicly about the audience's obtuseness. The revised version is what survives. Scholars have been studying it for two and a half millennia. You can take an entire graduate seminar on it. You can read fifteen competing interpretations of why it failed at the Dionysia. And when you read the play itself, it still doesn't work.
Not in the sense that it's hard to translate, or requires explanation, or loses some texture across the centuries. It doesn't work in the more fundamental sense that you can read the setup, follow the logic, understand precisely what Aristophanes is doing, and produce no laugh. The joke mechanism is present. The mechanism does not fire.
The Clouds required four simultaneous cultural dependencies to operate, and all four are gone. It needed an audience that knew Socrates personally, or knew someone who did — that had a live, specific, street-level relationship with this man as a recognizable annoyance, not as a historical figure studied in school. It needed an Athenian civic anxiety about sophistry, about the new rhetorical education corrupting the young, that wasn't abstract concern but felt civic panic — the specific alarm of people who believed they were watching their children be ruined in real time. It needed recognition of specific rhetorical modes that Aristophanes was parodying: techniques fashionable in 423 BCE and nowhere fashionable now, in the way a parody of a specific cable news pundit's verbal tics is legible only if you've been watching that pundit. And it needed acoustic wordplay — puns embedded in the rhythm of Greek meter that only work if you can hear the pulse of spoken Greek performance, the way a rhyme only rhymes in the original language.
Lose any one of these four dependencies and the joke collapses. Lose all four and you have a play that requires forty pages of footnotes explaining why it was once funny, and after the forty pages it still isn't funny, and you sit there with the flat feeling of having read a description of a meal rather than eaten one.
The biological substrate is intact. Pattern violation: present. Status reversal: present. Shared recognition: present in architecture if not in content. The machinery is there and in full working order. The cultural superstructure that loaded those mechanisms with live ammunition evaporated two and a half millennia ago. The gun is pointed at something that no longer exists, and the trigger is loaded with vapor.
The Philogelos is the oldest surviving joke book in the world — a Greek collection from roughly the fourth or fifth century CE, containing over two hundred and fifty jokes organized by character type, which means either that humanity has been writing down bad jokes since the Roman Empire or that the Roman Empire had better archival instincts than we credit. More than a hundred of the jokes concern the scholastikos — the absent-minded scholar, the man so over-educated as to be catastrophically incompetent at the basic requirements of life.
One scholastikos is asked how old he is and replies that he'll have to check his records. Another returns from a long trip abroad and, apparently having forgotten the state of his household when he left, asks his slave whether his father is still alive — he doesn't know. The structure of these jokes is transparent. Setup, beat, reveal of absurd incompetence. You can read them and see exactly how the mechanism is supposed to work.
Modern readers find them mildly amusing. Not funny in the way that catches you off guard and produces sound. Mildly amusing in the way that a diagram of a combustion engine is interesting — you can see how it would work; you just can't feel it working.
The gap is not linguistic. It is about the emotional valence of expertise.
The scholastikos jokes worked because over-education was threatening. The over-educated man had prioritized abstract knowledge at the expense of practical wisdom and was therefore, in the civic imagination of the ancient Mediterranean world, less of a functional human being — contemptible, a failed citizen, a man who had squandered his capacity for the real work of life on useless abstraction. The audience laughed down at the scholastikos. He deserved it. His incompetence was an indictment.
Modern readers laugh with the scholastikos — if they laugh at all. The absent-minded professor is now a charming archetype, possibly admirable, the genius too absorbed in important work to bother with trivial logistics. The butt of the joke has become a sympathetic figure, and the emotion that was contempt has quietly become mild affection. The humor mechanism — status incongruity, the gap between claimed expertise and demonstrated competence — is intact and firing. We see the joke. We see how it works. But the cultural signal that told the audience how to feel about the target has not merely faded; it has reversed polarity. Same punchline. Opposite planet.
Flyting is the third case, and it illuminates something slightly different.
From roughly the fifth through the sixteenth century in England and Scotland, flyting was a competitive public entertainment: two poets trading elaborately constructed insults in verse, before a royal court, with an audience trained to appreciate the craft. The Flyting of Dunbar and Kennedy — performed around 1503 before James IV of Scotland — is the most famous surviving example. If you read it, you will encounter something so aggressively alliterative, so luridly specific in its abuse, that contemporary scholars have reached for battle rap as the closest modern comparison. William Dunbar calls his opponent a "monstrous mouth" and a "rottin crok." Kennedy replies with equal technical precision and comparable anatomical inventory. The skill is extraordinary.
Modern readers can appreciate the craft. They cannot experience the event.
Flyting required a double register that no longer exists as a unified cultural form. The audience had to simultaneously experience the abuse as real enough to be humiliating — genuine threat, genuine attack on the target's honor and reputation, the kind of thing that, in another context, would require a response — and recognize it as performed enough to be safe, as a game with established rules, as theater rather than combat. This double experience — genuine threat and theatrical safety held in tension at the same moment — was the humor. Strip away the craft and you have sincere abuse. Strip away the threat and you have poetry recitation. The form required both at once, and the social context that enabled that double-reading — a court audience that shared norms about when performance began and ended, who trained in verbal dueling as a martial art of language — is irrecoverably gone.
The impulse itself, however, has not gone anywhere. Rap battles work because they operate within a recognized performance frame. Everyone in the room knows the rules. Everyone knows that "going in" is theater, that the performer and the audience share a contract about the boundary between performance and genuine attack. The insult contest as competitive form appears in cultures across the world; scholars have documented some version of it on every inhabited continent. Flyting didn't die because the impulse died. It died because the specific social apparatus that held the form together — the court, the shared norms, the trained audience — dissolved. The impulse migrated to new containers and found new audiences.
These three cases are not exceptions to the universality argument. They are the universality argument, made precise.
The Clouds shows what happens when all the cultural dependencies expire simultaneously: the machinery runs on empty, producing nothing. The scholastikos shows the machinery intact and operational, the mechanism firing, but the cultural content that powered it inverted — same mechanism, opposite emotional register. Flyting shows the impulse surviving the dissolution of the form, decanting into new containers across centuries. The biology is the constant. The content is the variable. The variable can expire, invert, or fragment, and when it does the machine keeps running, waiting for new material to load.
There is a more serious challenge to the universality claim that I've been allowing to wait in the anteroom, and it deserves genuine engagement rather than a polite paragraph before moving on.
Lisa Feldman Barrett is a neuroscientist at Northeastern University, and her constructed emotion theory is the most sophisticated scholarly challenge to claims of emotional universality currently in the field. Barrett's argument is not that people in different cultures are biologically different, or that emotional expression is purely arbitrary. It is more precise than that: emotions, on her account, are not universal biological programs that fire reliably in specific circumstances and produce consistent expressions. They are constructed experiences — assembled from interoceptive sensations (the raw physical data of what's happening in your body) and cultural concepts (the learned categories your brain uses to interpret those sensations). On this view, what we call "laughter" is not a natural kind. It is a concept, and concepts vary across cultures.
This is not a fringe position. Barrett's 2017 book, How Emotions Are Made, has reshaped how many researchers think about the universality of emotion, and the critiques it leveled at earlier cross-cultural work have not been fully answered.
Barrett's Namibia data are directly relevant to this chapter. In a 2014 study, Marie Gendron, Barrett, and colleagues played emotional vocalizations to Himba participants in rural Namibia — people with minimal exposure to Western media, chosen specifically to test cross-cultural recognition without the confound of shared cultural context — and asked them to respond freely rather than choose from a preset Western list of emotion categories. Most emotional sounds were not reliably matched to the categories Western psychology expected. Fear sounds were not consistently categorized as fear. Anger sounds were not consistently categorized as anger.
Laughter was the exception. It was consistently identified as positive and amusing across cultures.
Laughter was the most cross-culturally robust vocalization tested.
But "most robust" does not mean "perfectly universal," and Barrett's point is precisely about the residual variation. Laughter passed the cross-cultural test that most emotions failed, but the conceptual layer around laughter — what it means in a specific context, what social category it belongs to, what emotion it indexes — varied significantly. The finding cuts in two directions: laughter cleared the cross-cultural bar, but the bar itself revealed how much of emotional expression we wrongly assumed was universal.
The answer to Barrett, though, lives in the two-pathway anatomy. Barrett's critique applies most forcefully to the performed laugh — the social, motor-cortex laugh, the one that runs through the cultural layer, the one that cortical damage can remove. That laugh is learned, culturally inflected, subject to display rules and social norms and all the construction that Barrett's theory describes. But the pre-cultural claim in this chapter does not rest on the performed laugh. It rests on the involuntary laugh: subcortical, ancient, running through structures that predate the cultural apparatus by millions of years.
Wild et al.'s lesion patients are the direct evidence. The cortical layer — the layer where Barrett's construction would operate — was removed. The ancient machinery kept running without it. The patients who could no longer perform a laugh on demand still burst into genuine, involuntary laughter at jokes and tickling. Whatever cultural construction is happening to laughter, it happens in the prefrontal cortex, not in the periaqueductal gray. Strip the cortex and you have the raw, pre-cultural mechanism.
The raw mechanism is what the rats have.
Barrett is right that the performed, social, culturally-inflected laugh is a construction. She may well be right about the precise scope of that construction and its cross-cultural variability. The argument here doesn't require that laughter mean the same thing everywhere, or that the conceptual categories surrounding laughter be identical across cultures. It requires only that the capacity to produce involuntary laughter — the subcortical, ancient-machinery response to tickling, to surprise-and-release, to social play — was installed before culture arrived to shape what to do with it.
Culture shapes the trigger and the meaning. Biology installed the capacity. Barrett is describing the second floor of a building whose basement is not under construction.
Back to the rats for a moment, because Panksepp observed something I've been holding in reserve.
During play, rats produce not only the 50 kHz chirp but also, in situations of distress — isolation, threat, pain — a 22 kHz cry. Different frequency, different context, different behavioral prediction. The 50 kHz chirp predicts approach: the rat seeks out the tickling hand, seeks play, seeks the interaction that generated the sound. The 22 kHz vocalization predicts withdrawal, freezing, the behavioral signature of something bad. Two vocal signals. Two opposite motivational states. Two systems.
Sound familiar?
What Panksepp had found, in the rat play-chirp, was an early instance of what he called the PLAY system — one of several primary emotional systems he identified across mammals. Not a learned behavior, not a performance in response to social expectation, but an affective-motivational circuit that generates specific vocalizations and specific approach behaviors in the absence of any training or cultural template. The rat is not pressing its paw against the tickling hand because it learned that paw-pressing produces rewards. The rat is pressing its paw against the tickling hand because something in its ancient brain registered the tickling as good and generated a motivation to have more of it. The 50 kHz chirp is the sound that motivation makes.
This is what makes the image strange and arresting. The rat is not performing. It is not executing a learned script. It is doing what the biology made it do, which is to seek, again, the thing that made it make that sound.
There is no tidy scientific word for what that looks like from the outside. From the outside, it looks like delight.
A note on kuru, which is a terrible disease and belongs in this chapter anyway.
In the 1950s, D. Carleton Gajdusek — who would later receive the Nobel Prize for the work — documented a fatal prion disease spreading through the Fore people of Papua New Guinea. The disease was transmitted through funeral practices involving ritual consumption of the deceased. In its early ambulatory stage, before motor symptoms became severe, kuru produced spontaneous uncontrolled laughter: episodes of involuntary laughing, unrelated to any external stimulus or apparent emotional state, generated by the neural degeneration progressing through the patient's brain.
The Fore called it the laughing disease. It was not amusement. It was the laughter machinery firing because the trigger mechanism was being destroyed — the gun going off not because anyone loaded it but because the firing mechanism was coming apart. Biology producing the signal with culture contributing nothing.
I raise kuru not to make laughter seem sinister but because it is the cleanest possible demonstration of the distinction this chapter has been building. The laughter was real, in the sense that it was neurologically identical to laughter produced by something actually funny. An outside observer who didn't know what was happening would see someone laughing. The neural substrate — the mechanism — was firing. The content, the emotion, the experience, the thing being laughed at: absent. The substrate ran without the software.
There is a related case that Gajdusek didn't have to travel to Papua New Guinea to find. A 2018 case report documented a forty-year-old man who had experienced spontaneous involuntary laughter — unrelated to external stimuli, occurring two to three times per week — since he was eight years old. At forty, an MRI revealed the cause: a hypothalamic hamartoma, a congenital benign tumor on a brain region governing emotional expression. For thirty-two years his laughter had been a seizure. Neurologically identical to real laughter from the outside. He had been generating the signal, correctly, twice a week, for three decades, with nothing funny happening.
The gun fires when the brain breaks down. That is how you know where the gun is.
Venezuela, 1964. Napoleon Chagnon, a young American anthropologist, has been living among the Yanomami for five months, doing what field anthropologists do: building relationships over enormous investments of time, learning the language, collecting genealogical data. The Yanomami have a name taboo — using someone's real name in their presence, or in the presence of their close kin, is a significant transgression, a violation of basic social respect. You work around this with kinship terms and nicknames. Chagnon needed real names for his genealogical records.
His informants had been providing names.
The headman's name, as provided to Chagnon: "long dong." The headman's wife: "fart breath." Various other community members had been assigned "eagle shit," "asshole," and variations on the theme. Chagnon had been carefully and seriously recording these names in his notebook, behaving with the professional dignity of a scholar doing important work. Word was spreading between villages. There was an anthropologist walking around with a notebook full of obscene names that he believed were real genealogical data. He had been doing this for months. He kept taking notes.
The exposure happened when Chagnon mentioned the headman's wife's name in front of the headman, in a neighboring village, with an audience that knew exactly what the name meant.
"A stunned silence followed," Chagnon wrote, "and then a villagewide roar of uncontrollable laughter, choking, gasping, and howling."
Consider what this joke required. A shared understanding of the name taboo — the social rule that made the names transgressive. A victim who didn't know the rule, moving through the village in good faith with his notebook, building the setup with every new interview. A coordination problem: multiple informants had to consistently provide obscene names without anyone breaking and warning the anthropologist. A delay — five months of careful maintenance, the joke getting richer as it spread to neighboring villages, the payoff building with every new person who learned what was in Chagnon's notebook. And a reveal that required the victim to unknowingly detonate his own punchline, in public, in front of witnesses.
This is a five-beat joke structure: transgression, victim, setup, escalation, reveal. Scholars of comedy would recognize it immediately. The Yanomami built it without access to Western joke-writing conventions, in a society without professional comedians, because the machinery was there and the opportunity was good. The mechanism recognized what it had to work with and knew exactly what to do.
The specific content — the name taboo, the particular transgressive vocabulary of a specific Amazonian linguistic community, the social weight that names carry in Yanomami culture — is mortal. It belongs to a specific place and time. Another culture, another anthropologist, another set of social rules: the joke doesn't work. The trigger is local.
But the five-beat structure, the delight in transgression and delayed reveal, the explosion of communal laughter at a perfect punchline: those are not local. Those are the biology finding something to work with.
Chagnon had to discard five months of genealogical data and start over. The Yanomami, one imagines, regarded this as a reasonable price for the best joke of 1964.
Something that does not get said often enough about laughter: it is almost always social. Not in the sense that it requires a performer and an audience — though it usually has them — but in the sense that the machinery was shaped by millions of years of social evolution. You are several times more likely to laugh at something if someone is with you when it happens. You are dramatically less likely to laugh alone at the same thing. Robert Provine, a neuroscientist who spent years studying natural laughter in public spaces rather than labs, found that people laugh at the mundanest social remarks far more often than at anything designed to be funny — at things like "I'll talk to you later" or "it was nice meeting you" — statements whose humor content is essentially zero but whose social function is to signal warmth and connection.
This is not the laugh we've been building the case for. This is the performed laugh, the motor-cortex laugh, the laugh as social bonding behavior. It is real and important and most of what humans do with laughter most of the time. But it is also the laugh that would not exist if the ancient machinery hadn't been there first — the substrate that culture then recruited, and shaped, and deployed for a hundred social purposes that have nothing to do with anything being funny.
The ancient laugh and the social laugh are not in conflict. They are in sequence. First came the mechanism, older than language, carried in the same brain structures as rats and apes. Then came culture, which found the mechanism enormously useful and built an entire second system on top of it. The building is old. The renovation is recent. You live in both at once.
The reason Panksepp's 2023 vindication lands the way it does — the way, if you read about it and know the history, there is something that feels like justice in the finding — is that he spent years being told he was committing the sin of anthropomorphism. Of projecting human experience onto animals. Of seeing what he wanted to see.
What he actually saw was a rat pressing its forepaws against a human hand. Circling back, presenting itself, seeking more of the thing that made it make that sound. He saw this, thought this looks like delight, wrote it down, and spent years defending the observation against people who had decided in advance that such observations were methodologically impermissible.
The 2023 paper didn't vindicate him because it proved that he was right about everything, or that rat and human subjective experiences are identical, or that the gulf between rat cognition and human cognition doesn't exist. It vindicated him because it confirmed that the machinery is shared. The periaqueductal gray that lights up in the fMRI scanner when a human laughs involuntarily is the same structure that mediates the 50 kHz chirp in rats during play. The same ancient address. The same circuit.
This is not a metaphor. It is not that rat play and human laughter are analogous in some loose poetic sense. The same brain structure does the same job in both species because it was there before the species diverged. Panksepp wasn't anthropomorphizing. He was recognizing a homology.
There is a version of this book that is only about human humor — the jokes, the cultural contexts, the infinitely varied ways that human beings have found to make each other laugh across four thousand years of recorded history. That would be a fine book. But it would be missing the deep structure, the part that doesn't change, the part that exists not because any culture invented it but because it was given to us — and to the rats, and to the apes, and to everything that ever played — before culture existed to give us anything.
The Yanomami aren't laughing because they learned to laugh from watching other people laugh. Chagnon's blind-deaf infant counterparts in Eibl-Eibesfeldt's fieldwork weren't laughing because they copied anyone. The rats aren't laughing because anyone told them to. The laughing is the starting condition, not the result. Everything else in this book — the jokes, the stand-up traditions, the court jesters and the satirists and the absurdists, the things that are hilarious in one culture and incomprehensible in another, all of the wildly specific and local and mortal content that human beings have been loading into this machine for as long as there have been human beings — is the software. Rich, various, endlessly inventive software.
But you were born with the hardware already running.