Chapter 1: The pharaoh and the pendant
The Westcar Papyrus, copied around 1600 BCE by an Egyptian scribe whose name we do not have, preserves an older tale about King Sneferu, second pharaoh of the Fourth Dynasty, builder of the first true pyramids, and on the day this story finds him, profoundly bored.
He has wandered the palace looking for "anything to amuse," which is the kind of activity that generally precedes either a treaty or a divorce. His chief lector-priest, a man named Djadjaemankh, prescribes a cure. He prescribes it with the clinical register of a court physician dispensing willow bark.
Take twenty of the most beautiful women in the palace, he advises. Give them ebony oars. Dress them in fishing nets in place of their clothes. Have them row the king around the palace lake.
The pharaoh perks up immediately.
I am going to camp here for a moment, because the moment is funnier than it looks. Whatever civilisation has accomplished in four thousand years, it has not stopped Sneferu from being a guy. The chief lector-priest's prescription, as preserved in Miriam Lichtheim's translation of the Berlin Papyrus, runs in the same flat ritualised cadence as everything else a chief lector-priest says: "Let there be brought to me twenty nets, and give these nets to these women in place of their clothes" (Lichtheim 1973). That sentence is the reason any of this survives. The scribe is treating the recommendation of fishnet-clad rowers as he would treat a state ritual, which is what makes it land. Set the same content in any other register and the scene is sleazy. In this register it is dry to the point of crystalline. The court is not embarrassed. The court is administrative.
Then the joke escalates.
One of the rowers loses a small turquoise fish-shaped pendant overboard. The king — being a king — offers her a replacement from the royal treasury, which has, after all, many turquoise fish. She refuses. She wants that pendant. The whole boating party comes to a halt: twenty women in fishnets stalled mid-stroke on a palace lake while a pharaoh and his magician negotiate with a single staff member about her jewellery. The negotiation goes on long enough that the lector-priest eventually has to recite a magical formula, lift one half of the lake into the air, stack it on the other half, locate the pendant on the dry bottom, hand it back, and let the water down. This is supposed to be the magic. The woman gets her pendant. The boating resumes. The king's mood is restored.
I have read this story a number of times now and I have not been able to stop noticing that the longest pause in it is over the jewellery.
There is something I want to say about the scribe's wink. The comic technique — deploying the most serious ritual register on the most ridiculous content — is recognisable, in its mechanics, to any modern reader who has watched a 60-Minutes-style interviewer ask a competitive eater how he prepares for a major competition, or read a New York Times wedding announcement about a couple who met in line at Buffalo Wild Wings. The technique has names in modern criticism (mock-formal, mock-heroic, register inversion) and a literature; it is also alive in 1600 BCE on the Westcar Papyrus, in hieratic Egyptian, in front of an audience asked to read it the same way I just did. The lector-priest's name, Djadjaemankh, is preserved. The pharaoh's name, Sneferu, is preserved. The scribe's name is not, which is the standard way of things in late Bronze Age Egyptian copying. I would like to know his name. I would like to thank him.
The Westcar Papyrus, in its received form, is a frame narrative: King Khufu, Sneferu's son and the builder of the Great Pyramid, sits in audience while his own sons take turns telling him stories about the wonders worked by magicians at the courts of his predecessors. The Sneferu story is the second one. It is told to entertain a king. It also happens to be the only one of the Westcar stories that opens with the line about the king being bored. The narrator, in other words, has constructed a story for an audience that includes the great-grandson of the king in the story, and the narrator's first move is to inform that audience that the king's mental state at the start was, more or less, Sunday afternoon, palace, no plans. There is a way of writing this scene that is reverent. The scribe is not writing that way. The scribe is winking, three and a half thousand years before any of us got here to read the wink.
The Westcar Papyrus itself is one papyrus roll, written in hieratic script and currently stored in the Egyptian collection of the Berlin Neues Museum as Papyrus Berlin 3033. It was acquired in the late 1830s by a British traveller named Henry Westcar, who left it to the Egyptologist Karl Richard Lepsius, who in turn placed it in Berlin in the late 1850s; Adolf Erman published the first scholarly edition in 1890. The hand is a single scribe's; the dating, by paleography, sits in the late Hyksos period — the Second Intermediate Period — which is to say somewhere on the far side of 1600 BCE. The story it tells, however, is older. The Sneferu episode is set in the Old Kingdom court four or five centuries earlier than that, and the linguistic features of the text suggest a Middle Kingdom original behind the Second-Intermediate copy (Lichtheim 1973, pp. 215–217; Lepper 2008). We have, in our hands, a copy of a copy of a story whose first telling we cannot recover, and which somebody, somewhere, decided was good enough to keep, and somebody after that decided was good enough to keep again, and some second somebody after that decided to recopy it because the first copy was on a roll falling apart. There are people in this chain. None of us knows their names. They each had to look at the line about the pendant and decide whether to copy it. They copied it.
I would like to put a question on the table, because the question is the engine of this book and there is no point being coy about it.
Is there anything humans laugh at, across all the distance there is to cross between us — across continents, across languages, across centuries, across stones — that other humans, somewhere else, also laugh at? The Westcar scribe sat down at a desk in the Egyptian Second Intermediate Period sometime around 1600 BCE, and copied this scene from an earlier scroll, and the scroll he copied it from was already a copy. Three and a half thousand years of paper distance and an entire dead language sit between him and me. I read the paragraph about the pendant negotiation and I laugh. I would like to know whether that laugh is a coincidence — a smudge produced by my modern eyes accidentally finding shapes in old text — or whether the scribe, on the day he was copying it, also smiled at the same line, and copied it because of the smile.
I have begun to think it is the second one.
That is the claim I am going to try to land. Said plainly, it is that humor is one of the most stubbornly portable things humans make: more portable than recipes, far more portable than music, in some respects more portable than language itself, and certainly more portable than any of us were taught it was. Said carefully — because the stronger the claim is, the more careful — it is that there are comedic structures recognisable to any literate adult on any continent, in eras separated by millennia, in languages no living person speaks, on platforms that could not have been imagined when the joke was first told, and the recognition is not nostalgic, not academic, not gracious. It is the actual involuntary part of your face responding to the actual mechanics of the gag. The pharaoh perks up immediately. So do I. So, it appears, did the scribe.
A short defence of the project
I am not a folklorist or a historian or a comedian. I came to this question by accident, as a list. About six years ago I started keeping a note on my phone of jokes that should not have worked on me but did. The first entry on the list was a fart proverb from approximately 1900 BCE, which I will get to in a moment. The second was a sketch from a 1940 Mexican film I had stumbled into on a Saturday afternoon and could not stop laughing at, in Spanish, despite a Spanish vocabulary that begins with hola and ends shortly thereafter. The third was a two-panel webcomic of a dog in a hat in a burning kitchen sipping coffee. The fourth was a Greek punchline from approximately the year 360. The list grew.
By the time I was three or four pages in, I had a problem. I had been raised, as I think most people my age were, on the genteel assumption that humor is local — that a joke that lands in Brooklyn dies in Bishkek, that a punchline more than a generation old is a museum piece, that humor is among the most context-bound of arts and the test of whether something is funny is whether I laugh in my room with my friends. The list disagreed. The list had jokes from continents I had never been to and centuries I had not lived in and languages I could not read, and the line that runs through the list is that I had laughed at all of them, on first reading, in approximately the same beat the original audience appears to have laughed.
This book is the long version of the list.
A confession: I would have written a different book if I had thought it would land on a different thesis. I started reading with the working hypothesis that humor was beautifully, irreducibly local — every era and every place had its own unique grammar of laughter — and that the project would be to celebrate the fragmented loveliness of all the local versions. That book would have been easier to sell at a dinner party. It would have flattered the reader's worldview. It also turned out not to be true. The longer I sat with the material the more I had to give the original thesis up, and what replaced it is what you are holding. I will say this once and not again: I am as surprised by this finding as you may be when you reach the end. I have tried to write the kind of book that walks the find with the reader rather than presenting the conclusion already gift-wrapped at the door. Several of my favourite passages in what follows are passages in which I had to take something out — a clean local example I had been planning to use as a counterexample — because, on closer inspection, the clean local example wasn't local. I have left some of the seams of that process visible on purpose. I hope, by the time we land, the seams are part of the case.
It would not surprise me if some of the cases I keep are wrong on specifics. The slate has been pruned and re-pruned, and the entries that survive are the ones I have been able to source to a primary or scholarly text — Lichtheim for the Westcar Papyrus, Alster for the Sumerian proverbs, Berg and Baldwin for Philogelos, Pi-ching Hsu for the late Ming jest-books — but I am not a specialist in any of those traditions, and I have surely got things wrong. Where I have, I have tried to get them wrong in the direction of being too cautious about the claim rather than too eager. Where the data hedges, the prose hedges. Where the data is firm, the prose is firm. The places where the data is firm are the places that surprised me.
A pre-emptive answer to a reasonable objection: yes, of course there are jokes that don't travel. There are. The book has a chapter for them. It is the last one. The jokes that don't travel are the most interesting cases I know, and the chapter that examines them is the chapter that ought, if I have done my job, to make you trust the rest of the book. We will arrive there. For now — Sneferu and his pendant, and what the Sumerians thought funny enough to teach to children.
Things which have never happened
The earliest written joke we have, by current scholarly consensus and in the form that has the best claim to being a joke and not just a piece of philosophy with bad aim, is a fart joke. It is from a Sumerian proverb collection copied for scribal training around 1900 BCE in the cities of the Old Babylonian period, edited and translated by Bendt Alster in Proverbs of Ancient Sumer (Alster 1997), and catalogued by the Electronic Text Corpus of Sumerian Literature at Oxford as text c.6.1.01. The proverb runs:
Something which has never occurred since time immemorial: a young woman did not fart in her husband's embrace.
I want to spend a paragraph on what is happening here, because the first time I read this line I did the involuntary thing where you laugh once and then immediately reread the page to make sure you read it correctly. The joke is built on what Alster catalogues as the paradoxical-formula — the speaker frames the joke as a litany of things the world has never seen — and the punch is that the slot is filled with an event that, of course, happens constantly. The setup primes you for an impossibility; the payoff is universality. The structure is the same one I would use to land "something the world has never seen: an opinion about pineapple on pizza." It is not just that the joke is recognisable across forty centuries. It is that the joke is built on a load-bearing comedic structure — defeated expectation, stated plainly, with an implied speaker who knows exactly what he is doing — and that the structure is teachable.
Because that is the part that gets me. Alster's edition makes it explicit, and the Oxford ETCSL catalogue makes it explicit: the proverb was scribal pedagogy. It sat on a tablet of model sentences that adults copied and recopied for the purpose of teaching children what good Sumerian looked like. The Old Babylonian equivalent of a primary-school grammar drill included a joke. Some twelve-year-old apprentice scribe, four thousand years ago, looked up at his teacher with a small grin he was not quite ready to commit to, because what was on his clay tablet, in the language he was learning to write, was not a virtue and not a fable. It was a fart joke about married couples. The teacher, being the teacher, knew what was coming and waited it out. The grin came anyway. The teacher had been twelve once. The whole institution of learning to write Sumerian was already running on humor in the year 1900 BCE; it would have been weird if it weren't.
What we know about Sumerian scribal training comes mostly from the schools of Nippur and Ur, which Assyriologists conventionally call by the term eduba, "tablet house," though the exact term is debated. The evidence is largely the tablets themselves: cracked, fragmentary, glossed in two languages, often signed at the bottom by named apprentices, sometimes corrected in the same hand by an instructor. We have surviving compositions in which a senior scribe describes a working day in the eduba — the recitations, the beatings (the eduba texts are, on the matter of beatings, candid), the apprentices snitching on each other, the bedtime complaints to a sympathetic mother — and we have a continuous Sumerian-language tradition of teaching by copying that produces tablet after tablet in the same handful of forms. The proverbs are one of those forms. They are not, on the surface, the rebellious or impious materials a modern reader might expect from material with comic punch. They are model sentences, kept in collections grouped by opening sign or by topic, and the topics include the proper care of livestock and the proper way to address one's elders and the proper structure of a witnessed contract. The model sentences sometimes also happen to be jokes, in the way that the model sentences of any reasonably sane educational tradition will sometimes also be jokes, because the way you remember a sentence is the way it makes you laugh.
That is a thing I had not been told.
I should also say that the proverb survives because somebody copied it. Then somebody copied the copy. Then somebody copied the copy of the copy, in the city of Nippur and a half-dozen other cities, on tablets that ended up in foundation deposits and trash heaps and student archives, until enough tablets sat in enough places that twentieth-century cuneiformists could collate them into an edition. None of the people in that chain of copying had any way of knowing that the joke would be readable in the year 2026. They copied it because it was good. The Westcar scribe also copied his scroll because it was good. The chain of survival is a chain of grins.
The objects themselves are intimate in scale. A typical Old Babylonian school tablet is hand-sized, brown or grey-brown clay pressed by a reed stylus and ruled into columns, with the Sumerian on one face and sometimes a smaller-handed Akkadian gloss riding above the harder lines for a student still learning the older language. The proverb tablets that Alster collates run in several physical formats: a teacher's exemplar at the top of the obverse with a junior imitation directly below it; a long extract on the obverse signed in the colophon by an apprentice — the hand of so-and-so, junior scribe — and dated by the regnal year of the king then on the throne; a single-line lentil tablet small enough to sit in a child's palm, the proverb scratched on the obverse and re-scratched on the reverse for practice. The corrector's hand, where it shows up, is sometimes the same hand that ruled the columns, returning to a misformed sign with a quick re-impression while the clay was still wet. None of those hands has a face I could draw. Each touched the line and chose to pass it along.
The proverb is one item in one collection. The collection is one of dozens of Sumerian proverb collections. The collections are part of a curriculum that lasted, in some form, for the better part of two thousand years. Sumerian was no longer anyone's mother tongue by the late second millennium BCE, but it remained the prestige language of scribal training in Mesopotamia for another fifteen hundred years, the way Latin remained the prestige language of European scholarship long after no one was speaking Latin at home. Generations of children in Babylonian and Assyrian cities, none of whom had a single Sumerian-speaking ancestor, sat at desks copying out the same proverbs that twelfth-generation native-Sumerian-speaking children had once copied. The continuity is the sort of thing that makes me put down my coffee. The fart joke — and dozens of similar items — was recopied by children who didn't even speak the language the joke was in, generation after generation, because their teachers told them to. The teachers told them to because the proverbs were the curriculum, and the curriculum was the curriculum. Somewhere in there a kid laughed.
What I had been told
I had been told humor was local. Specifically, I had been told that comedy is the most context-bound of the arts — a punchline depends on a thousand small contingencies of language, geography, era, in-group reference; you have to be there. I had been told that you cannot translate a joke; you can only retell it. I had been told that nothing dates faster than a joke book, that humor was disposable, that it was the first thing anthropologists drop when they are trying to convince themselves a culture is serious. I had been told that the shared laugh is a precious local commons, and that to laugh at something from another time or place was either a kind of theft or a kind of failure of imagination — you weren't, in fact, getting the same joke; you were getting a flattened version of it suitable for export.
Some of that is true. Some of that is true some of the time. Some of it is so reliably false that I have started to think the assertion was performing some other work — the work, perhaps, of warning us off comedy as a serious object of study. Comedy resists being studied. It dies under the lights. The jokes that get analysed stop being funny on contact, which is true, and which ought to be filed alongside the fact that the recipes that get analysed stop being meals on contact, and the songs that get analysed stop being songs on contact. Analysis stops things from being themselves. None of which means the comic mechanism dies under the lights. It means that particular piece of the show dies. The mechanism remains alive and reproducible. You just have to look at it sideways.
The ground I want to defend is narrower than the working slogan. I am not going to argue that everyone everywhere finds the same things funny. They don't. I am going to argue that there are recurring comedic structures — the deadpan-narrator move, the helpless body versus the accelerating mechanism, the self-deprecating reclamation of an insult, the literal-fool category error, the improvisation-under-pressure recovery, the deadpan surface concealing an obscenity — that recur across continents and millennia, that they recur convergently as well as by transmission, and that when they recur the recognition is not academic. It lands as a laugh. The laugh is the thing that makes the universality real, because the laugh is the thing the analytical apparatus cannot fake.
I understand that this is, by polite-dinner-party standards, a slightly impolite thing to argue. The polite version is to say that humor is "human" in some abstract uplifting sense, with an unstated retreat door that means whatever the speaker wants it to mean. The version I am actually arguing is the harder one. There are specific structural moves, with specific source cases, that show up in specific places at specific times, and I am going to lay them out. If the cases hold the argument holds. If they don't, no abstract uplift will save it.
The way I want to test the claim is to walk a case slate. Each chapter takes a body of evidence drawn from a specific era and a specific medium, and asks whether the comedic structures in it survive their own boundaries. The boundaries are five, and they are the places where the conventional wisdom would predict a joke to die. I have lost some of the bets and won others. The total ledger is the argument.
Five boundaries
The way I have come to think about the project is in terms of five boundaries that any candidate-universal has to cross before I am willing to call it one. Each boundary is a place where the conventional wisdom would predict the joke to die. Each, accordingly, is a place where finding the joke not dead means finding something worth a chapter.
The reason the boundaries matter as boundaries is that each one is the sort of thing that, by itself, ought to flatten any comedic structure that has to cross it. A joke that depends on a pun cannot survive translation — but the pun is not the only kind of comedic structure. A joke whose punchline names a person nobody remembers cannot survive a century — but the person is not the only thing that matters. The book's case is that comedy stores its load-bearing structures in places that the boundaries can't, in fact, reach. The pun goes; the paradox-formula stays. The local reference goes; the deadpan register stays. The local language goes; the helpless body stays. What is left is more than nothing. It is, in some cases, almost everything.
The first is cultural — across societies that have had no contact with each other, and that maintain mutually unintelligible structures of meaning, manners, and reference. If a comedic move shows up independently among a 9th-century ʿAbbasid prose stylist sketching a Basran miser, a 14th-century Tuscan friar caught with the wrong relic in the casket, and a 17th-century Chinese jest-book editor describing King Yama, ruler of the underworld, dispatching ghost-minions to find a doctor whose dead patients aren't loitering at the door, the move is doing something the cultures in question are not, in fact, sharing — which is the only sense in which "universal" is doing work. It is not the lazy sense (everyone has the same sense of humor). It is the structural sense (the same comedic engine is deployed in cultures that did not negotiate over its deployment). I have come to believe the structural sense is real. The lazy sense is, of course, not.
The second is temporal — across centuries, ideally millennia. Sneferu's lake is one of these. The Sumerian fart proverb is another. The bar I want to clear here is high. It is not enough to say that a 4th-century joke is "still relevant." Any joke is "still relevant" if you talk about it long enough; that is how seminars work. The bar is whether a contemporary reader laughs at it on first encounter, without scholarly briefing, in the same beat the original audience appears to have laughed at it. Several of the jokes in this book clear that bar. Some of them do it embarrassingly cleanly. There is a Greek punchline from somewhere between the fourth and fifth century of the common era, copied into a book that announces itself as a joke book, that is funny enough that a working stand-up named Jim Bowen retired several of its bits intact onto a 2008 Edinburgh Fringe stage and got laughs without telling the audience the bits were 1,600 years old. The audience gets to laugh first. They get to be told later. Some of them have written me to say that the second laugh, when they were told, was bigger than the first.
The third is linguistic — humor that travels through translation, or sidesteps language entirely. This is a harder boundary than the others, because language is where comedy stores most of its precision instruments — puns, syllabic compression, register shift, dialect — and translation is where those instruments get bent. The interesting cases are the ones where something other than the local linguistic instrument is doing the comedic work, and the something-other-than survives the trip. Take a Mandarin tonal substitution that exists only in the four-tone phonology of Putonghua: in 2009, in response to a tightening of online-content rules, a children's-nature-song parody began circulating on Chinese forums about a fictional grassland animal called the cǎonímǎ — written 草泥马, "grass mud horse," but one tone away from a maternal obscenity directed at the listener — whose habitat was being colonised by the héxiè, the river crab, one tone away from the official term for a "harmonious society" (Wines 2009). The pun lives only in Mandarin. The shape — a children's chorus serenely cataloguing the threats to a happy ecosystem in the bland register of an animated educational video — is portable on the strength of the shape alone, and a non-Mandarin reader who has the substitution explained laughs in the same beat the original chorus did. The cleanest examples are the wordless ones. In February 1991 a Japanese animated children's series called The Brave Fighter of Sun Fighbird aired its third episode, in which a newly embodied android named Yutaro Katori was walked through a meadow as part of his terrestrial-life education. He pointed at a passing butterfly. With his finger up, his eyebrows up, and his mouth half-open in the universal cel-animation register of sincere confusion, he asked, "Is this a pigeon?" The frame was held for two beats. Twenty-seven years later, in the spring of 2018, that single frame became a global wordless template for confident misidentification, used in English, Spanish, French, Portuguese, German, Italian, Korean, Japanese, and Chinese tweets in a six-week window, including one Netflix corporate post about casting twenty-eight-year-olds as teenagers (Feldman 2018). The comedy is the gap between confidence and competence. The speaker is calmly, publicly, and on the record wrong about a trivially identifiable fact, with their face still visible. No language at all, in the end. A face.
The fourth is belief — across worldview lines that the parties to the joke take to be so important they would die for them. This is the boundary where humor can stop being collegial and start being warfare. It is also the boundary where the cases get most interesting, because some of the best comedy — the relic-pedlar busted with charcoal in the casket and improvising it into the bones of Saint Lawrence in front of a paying congregation, the Buddhist monks debating the literati wit, the censored Mandarin homophone repurposed by a children's chorus into anti-state singalong — runs precisely on belief friction without crossing into hatred. The book engages the cases where humor does cross into hatred. They are the gravest material in it. They are also material where the engine is, unmistakably, the same engine that made you laugh on page two. We will have to deal with that.
The fifth is technological — humor that travels through and across new media. The joke that starts as an oral fart proverb and ends as a clay tablet. The clay tablet copied onto papyrus, copied into manuscript, copied into print, copied onto film, copied into a sitcom, copied onto a smartphone, copied onto a meme template. The interesting thing about this boundary is that it is the one we tend to assume disrupts comedy most — of course humor is "different now," of course the meme has nothing to do with the village comedian, of course the algorithm produces something other than what your grandmother laughed at. And the interesting thing about looking at the cases is that this assumption is the most consistently wrong of the five. Sneferu's bored-rich-guy energy is on TikTok every Sunday. The Sumerian "things which have never happened" deadpan formula is the structural skeleton of every list-format meme of the last twenty years. The chief lector-priest's clinical register prescribing fishnet rowers is recognisably the voice of the subreddit unethical life pro tips. We will get there.
The five boundaries do not all show up in every case, and they do not have to. The book's claim is that across the case slate as a whole the boundaries are crossed; that some cases cross several at once; and that no single boundary, on its own, is the thing keeping a joke from travelling. A joke that fails to cross one boundary may cross four. A joke that fails to cross any of them — and there are some — is a counterexample, and there is a chapter for those.
An era spine
The body of the book runs in chronological order, by medium, because medium is what determines what comedy can survive of itself in transit. There are five chapters under that spine.
The oral and pre-text antiquity — through roughly 500 CE — is where the project begins. The opening chapter past this one looks at the era in which comedy survived because it was memorable enough to be repeated and durable enough, when writing arrived, to be among the first things written down. The Sumerian fart proverb sits in that chapter; so does Sneferu's lake; so does a 4th-century Greek joke book that announces itself as a joke book and contains punchlines a contemporary stand-up has been known to retire intact. The Greek collection is named Philogelos, "The Laughter-Lover," and contains 264 jokes attributed to two compilers named Hierocles and Philagrius (Berg 2008; Baldwin 1983). Read it on a Saturday afternoon and you will find your own act in pieces. The British comedian Jim Bowen has been performing several of its bits, untouched, since around 2008; the audiences laugh on the first beat, and only some of them, when they are told afterwards that the joke is sixteen hundred years old, laugh harder on the second. The case I want to make in the next chapter is that humor was already polished — already engineered, already taught, already in stock — before literacy could reach it. Writing did not invent the joke. Writing caught the joke, in the way a butterfly net catches what was already in the air, and the air had been thick with the stuff for as long as anybody had been making language out of breath.
The manuscript era — roughly 500 to 1500 — is where comedy starts circulating. The codex lets a punchline travel further than its teller; the same comedic forms (the miser sketch, the improvisation-under-pressure recovery, the named-patron pun, the wise-fool retort) start showing up in 9th-century ʿAbbasid Basra and 14th-century Tuscany and the courts of imperial China and Persian-speaking Anatolia, and they are showing up convergently, not just by transmission. The al-Jāḥiẓ vignette of a miser pulling back a silver coin from a beggar's hand the moment he realises he has overpaid him by accident — a man whose reflex to keep his money is faster than his shame at having been seen reaching for it — could be sliced into a New Yorker shorts page tomorrow and miss nothing in the move from parchment to print (al-Jāḥiẓ, ca. 860; Colville 1999). The vignette sits in Kitāb al-Bukhalāʾ, the Book of Misers, written from a Basran shop-and-court vantage at the head of the Persian Gulf in the same decade Baghdad's House of Wisdom was finishing the second wave of Greek-into-Arabic translation; the silver coin is a dirham, weighed and counted by hand, and the speed of the reflex is a small fact of the body. The Boccaccio relic-pedlar saved by improvisation in front of his own congregation, opening the casket to find it filled with charcoal where the angel-feather ought to be and announcing, without missing a beat, that these are the very coals on which Saint Lawrence was roasted, is the structural skeleton of every confidence-restoration scene since (Decameron VI.10, ca. 1349–53). Boccaccio sets the scene in his own native Tuscan village of Certaldo on the saint's feast day; the relic the friar Cipolla has pre-promised the congregation is a feather from the wing of the Archangel Gabriel; the swap to charcoal is performed earlier on the road by two pranking gentlemen who have rifled the friar's bag while he is at lunch; the recovery, when Cipolla opens the box at the pulpit and sees what is in it, is delivered with the timing of a stage routine. The Arabic-Persian-Turkish wise-fool tradition — Juḥā in the Arab world, Nasreddin Hoca in the Turkish, Mulla Nasreddin in the Persian — runs from at least the 9th century into the present, telling the same dozen jokes at the same dozen punchlines, in mutually unintelligible languages, to audiences that would not recognise each other's faces. Three civilisations sit down with reed pen and parchment and write the same kinds of jokes. They are not borrowing from each other. They are, in some structural sense, drawing from the same well.
The print era — roughly 1500 to 1900 — is where comedy gets fixed in its exact wording. Print does to a joke what a recording does to a song. The punchline acquires a verbatim form a reader can quote, which is how Joe Miller's Jests of 1739 became a brand and "a Joe Miller" English slang for a stale joke (Mottley 1739). The era spine here is verbatim survival: a punchline that lands in Philogelos in the 4th century shows up cleanly in Joe Miller in 1739, and the Joe Miller witness-undoing-the-speaker structure shows up cleanly in Mark Twain's frog with a bellyful of quail-shot in 1865 (Twain 1865). The joke is no longer just a recurrent shape. It is a transmissible artefact, copyable across centuries the way an instrument tuning is copyable across orchestras. In Edo Japan the same beats compress to seventeen syllables. The thief I caught, when I looked at him, was my own son: the entire arc of the chase, the capture, the recognition, and the deflation, in three lines (Karai Senryū 1765). The form is senryū, named for its 18th-century editor, and it sells volume after volume — the joke as published commodity, in a form so structurally pure that the modern English reader, on first encounter, hits the recognition shock in the same beat the original Edo reader did. By the late nineteenth century an American humorist can sit at a desk in Hartford and freeze the cadence of a Western miner at the speed it works on a Western miner, and the cadence will play, identically, to a London reader of the Saturday Press who has never heard a Western miner talk in his life.
The broadcast era — roughly 1900 to 2000 — is where the comic body travels across continents. Charlie Chaplin, snowbound in a Klondike cabin in 1925, boils his boot, carves it like a roast bird, twirls the laces around his fork like spaghetti, and eats the bent nails one at a time with the delicacy of a diner working through fish bones (the boots and laces, in fact, were licorice; Chaplin shot the scene in roughly sixty-three takes and was eventually hospitalised for an insulin reaction; Robinson 1985). The same physical scene plays the same way in Berlin, Tokyo, Buenos Aires, and Los Angeles in the same week. Lucille Ball and Vivian Vance, twenty-seven years later, stationed at a chocolate conveyor belt that accelerates past the speed of human hands, stuff the unwrapped chocolates into their hats and the fronts of their uniforms and, when the hats and the uniforms run out of room, into their own cheeks like squirrels going to ground for the winter, in front of forty million Monday-night American living rooms. The forewoman returns, surveys the suspiciously empty conveyor, and shouts back through the door: speed it up (Asher, I Love Lucy, 15 September 1952). A Mexican comic in 1940 — Mario Moreno, working under the stage name Cantinflas — could improvise a courtroom sequence in Ahí está el detalle, in which his character is on the stand for a killing he did not commit, so stuffed with subordinate clauses, dangling pronouns, deferred subjects, and adjectives in search of nouns that the prosecutor and the judge had to ask each other to translate the previous sentence; the Spanish language formally added the verb cantinflear to the Diccionario de la lengua española, "to talk a great deal without saying anything" (Real Academia Española 1992). The bit is funny in 2026 in any language with subordinate clauses, which is to say all of them. The body is part of the joke. The body, it turns out, is portable in a way the language was not.
The algorithmic age — Ch6 — is where the joke decouples from speaker, language, and country. A 1991 Japanese anime freeze-frame becomes a wordless template for confident misidentification. A Mandarin tonal substitution turns the children's-nature-documentary register into uncensorable obscenity sung by a children's chorus, with the obscenity hidden inside an alpaca threatened by a river crab (Wines 2009). A photograph of a Japanese kindergarten teacher's rescue Shiba Inu side-eyeing the camera from a couch becomes a global Comic-Sans-rendered grammar of a dog's broken interior monologue, and within four years there is a cryptocurrency built as the punchline (Sato 2010; Chayka 2013). A two-panel American webcomic, unmodified, becomes the planet's vocabulary for presiding calmly over a room on fire (Green 2013). What survives, in the cases I have come to like best, is some piece of comedic structure stripped down so far that it does not need a sentence to deliver it. A face. A pose. A grammar. A two-panel webcomic in which nothing is said and everything is conceded. The reader — across continents the joke's author has never visited, in languages the joke's author cannot read — recognises the move and laughs. The recognition is the point. By the end of Ch6, what is left is what survives when speaker, language, and country are stripped away. There is, it turns out, a lot.
Each chapter answers the previous chapter's question with new material. The chapter on antiquity establishes age. The chapter on manuscripts establishes convergence. The chapter on print establishes verbatim survival. The chapter on broadcast establishes embodied portability. The chapter on the algorithmic age establishes structural decoupling. By the end of the body, the case is in five different mediums and seven different traditions, and the same handful of comedic engines has run through all of them.
The seventh and last chapter is the one where I owe the rest of the book its honesty.
A short forecast of the counterexamples
I said earlier that the longer I read in this material the harder I found it to keep my original thesis. I should also say that the cases that gave me trouble — the cases that should, in principle, falsify the universality claim — did not all fall over when I pushed them. Some held. Some qualified. One broke locally, in a way that I have to concede in the prose because I cannot make the data say otherwise. They are the most interesting eight cases I know in the literature. The chapter that walks them is the chapter I rewrote the most times.
I want to set the cases out briefly here, because honest argument involves naming the ones you cannot make go away, and because some readers will reasonably want to know where the book lands before they sign on for the trip. The chapter that does the actual work walks them at length, in roughly the order Ch7 will take them. At the end of it I will have tested the universality claim against material that should, on every prior reading I had access to, have broken it. Some of the material did. The claim that survives is narrower than the working slogan and stronger than I had expected to be able to defend. What follows is the shape of the slate.
There is a Soviet-era anekdot that runs: Lenin, Stalin, Khrushchev, and Brezhnev are riding a train. The train stops because the rails have run out. Lenin proposes a subbotnik — a "voluntary" Saturday work-shift to lay more track. Stalin orders the engineer shot. Khrushchev says, take up the rails behind us and lay them in front. Brezhnev draws the curtains and says: Comrades — let's pretend we're moving (Lewis 2008). The form of the joke (a procession of leaders, each more degraded than the last, ending in the punchline that names the ambient corruption) is universal — every authoritarian state has produced this shape; one of them produced thousands, in samizdat. The vocabulary, though, is locked. Without the Soviet political-economic vocabulary the joke is shaped like a joke and there is no laugh. Lock at the vehicle, universal at the engine. That is one kind of qualifying.
There is a Buster Keaton sequence from 1928 in which a two-tonne false front of a frame house topples onto a dazed young man, and an open second-story attic window passes cleanly over his head, clearing his shoulders by two inches on either side. The whole gag's payload is the precise spatial coincidence of two measured objects, and there is no way to tell it without showing it. Read the description; nothing happens. Watch the shot; involuntary intake of breath (Steamboat Bill, Jr., 1928; Curtis 2022). That is medium-locked at the level of the gag. The lock is narrow — every visual medium since 1928 has reproduced or quoted the move — but the lock is real. And there is Guillaume Apollinaire's Il Pleut, from his 1918 Calligrammes, in which five oblique columns of letters fall diagonally down the page in slanting streaks — phrases about voices and memories raining, with the typeset letters being the rain (Apollinaire 1918; Bohn 1986). Read aloud, Il Pleut is five sentences about rain. On the page, the typography performs its own metaphor and the recognition is the small oh, there it is of typography clowning through its own subject. Strip the page and the gag is gone before the first comma. The lock there is also narrow — any culture with a writing surface to play on has produced its own version of the move, from the Hellenistic technopaegnia of Simias of Rhodes forward — but it is also real. Both, in their different ways, are kinds of qualifying.
Then there are the harder ones. There is a 1979 Monty Python film whose crucifixion finale — "Always look on the bright side of life," whistled by men nailed to crosses — was banned for blasphemy in country after country in the year of its release, and which is now sung at football stadiums and at funerals across precisely those countries. There is a 2005 Danish editorial cartoon by Kurt Westergaard, whose intent the cartoonist has explained, whose framing is satirical, and which a substantial number of perfectly decent humans encountered as blasphemy and could not get to the laugh. The first case holds with a time delay; the bans dissolved and the song stayed. The second case is the genuine concession of the book. There are sacred lines that humor, in any framing acceptable to both sides, cannot cross. I do not pretend otherwise. The chapter does not flinch and does not soften. I will not be telling you that the laugh wins.
And then there are the hardest two. There is the 1828 Louisville debut of a song-and-dance routine that depended on a Black target and a white audience, and that within a generation gave the body of American segregation laws its name (Lott 1993). There is a Rwandan radio station in 1993 and 1994 whose most popular animateur, named Kantano Habimana, used the recurring bits, the running gags, the host-with-audience pop-music-and-banter register, and the mocking caricature any culture would recognise as comedy, to organise the murder of approximately 800,000 people in roughly a hundred days (Thompson 2007). The mechanism — running bits, callbacks, mocking caricature, host complicity-with-audience — is the same mechanism that runs a Sumerian fart proverb and a Boccaccio sermon and a Mexican courtroom scene and a Doge meme. The thesis as I will state it does not survive without taking that on. The chapter does not celebrate; the chapter does not flinch. The same engine that unites can be aimed.
The book argues the thesis honestly, which means the last chapter walks each of those eight cases and lands the argument anyway — sharpened, not softened, by the stress test. The reader closes the book aligned, not lectured into agreement, with the claim that humor is one of the most stubbornly portable things humans make and that this matters because of, not despite, the fact that the same engine can be aimed.
The boating party
I will close with the lake, because the lake is where I started.
I have, by now, several years' worth of this list on my phone. Some of the entries have moved into the body of this book; many have been cut for being too thinly sourced; several have been replaced by better cases I found later. The list is in better shape than it was. The thing that has not changed is the texture of the experience that produced the list in the first place. I read Lichtheim's translation of the Westcar Papyrus and I laugh at the bored pharaoh, and I notice myself laughing, and the laugh runs along the same shape that the Sumerian schoolboy's grin ran along thirty-six centuries earlier when he had to copy the line about the impossibility of farting. I do not know what the schoolboy's name was. I do not know what the Westcar scribe's name was. I have never heard a Sumerian fart proverb told aloud in the language it was written in. I am, in the technical sense, a stranger to all of these people. There is, also, this thing in common with them: when one of us encountered the line about the pendant negotiation, or the line about the impossible-but-actually-constant fart, or the line about King Yama dispatching ghost-minions to find a doctor, the involuntary part of the face responded. The involuntary part of the face is the smallest piece of evidence for this thesis I have. It is also the only piece I do not know how to argue away.
I want to say what changes for me when I sit with this thesis seriously. It is not, as I might have expected, a kind of triumphalism — the comforting old story that everyone everywhere is fundamentally the same and isn't that lovely. The Sumerians are not lovely, in any uncomplicated sense, and neither are the Egyptians, and neither are we. They had a society. We have a society. The societies were not the same and the people inside them were not the same. What I gain from the find is smaller and stranger. It is the recognition that a specific twelve-year-old, on a specific morning, in a specific Mesopotamian city, copying a specific clay tablet, made the specific face that you and I make when something dumb and true lands. The face is a real piece of evidence. The face is in fact one of the only pieces of evidence I have for the inner life of a person separated from me by every available form of distance. The face is something I can read. That is not nothing. It is, on balance, more than I had any right to expect.
I would like to take you on a trip with me through some of the rest of this list. The trip moves, as I said, by medium. You will spend the next chapter in pre-literate antiquity with the fart proverb and Sneferu and a 4th-century Greek joke book that catches the deadpan voice in mid-stride. After that you will sit at a desk with a 9th-century Basran sketching a man too cheap to overpay a beggar by accident, and a fourteenth-century Florentine writing about a relic-pedlar who turns chunks of charcoal into the bones of Saint Lawrence in front of a duped congregation that buys it twice. After that you will hold an English jest-book of 1739 that tells you, among other things, exactly how a joke first became a brand — and then you will turn the page to a Mark Twain frog with a bellyful of quail-shot, in 1865, executing a structural manoeuvre indistinguishable from one in Philogelos in 360 CE. After that you will watch Charlie Chaplin eat his shoe and Lucille Ball stuff chocolates into her hat. After that you will encounter a Mandarin pun threatened by a river crab, an anime android pointing at a butterfly, a Japanese teacher's rescue Shiba on a couch, and a calmly burning dog in a hat sipping coffee. And then, finally, you will sit with me as I take seriously the eight cases that should break this thesis — a Soviet train, a Buster Keaton wall, a 1979 hymn whistled from a cross, a Danish cartoon, an 1828 Louisville stage routine, a Rwandan radio host's running bit — and as I tell you which of them hold and which of them qualify and which of them broke locally, and which of them I had to rewrite the chapter for.
There is a direction this trip is going. It is not a sentimental direction. It is a direction in which the case for what humans have in common gets sharper, not softer, and the cases where the case fails get drawn cleaner, not blurrier. The thesis as I will state it has costs. Those costs are part of the package. The reader who would prefer a sentimental version of the argument will not find one in what follows. The reader who has, on prior reading or prior life experience, become quietly cynical about the slogans of universal human nature will, I hope, find the actual cases more interesting than the slogans.
I am the friend pulling you by the sleeve. I am also, like you, the audience. I have been turning this over for several years now, and the more I turn it over the more I have to keep going.
The pharaoh is bored. The lake is calm. The lector-priest is about to make his prescription. You can hear him clear his throat. The boating party is about to begin.