Chapter 3: The dirham, the lying horse, and the casket of charcoal

A miser in Basra

In the 860s of the common era, in the city of Basra at the head of the Persian Gulf, a writer whose contemporaries called him al-Jāḥiẓ — "the goggle-eyed," because his eyes bulged — finished a book about cheap people. The cheap of Basra. The cheap of Khurasan. The cheap of the Banū Tamīm and the Banū Sadūs, the cheap of the bazaar and the cheap at court, the cheap among the literate and the cheap among the illiterate. He had been listening for decades. He was now in his seventies, and the book he was finishing — Kitāb al-Bukhalāʾ, The Book of Misers — was less a treatise than a comic ethnography assembled by a man who had spent his entire life taking notes and refusing to be bored.

I want to start in Basra because al-Jāḥiẓ is the writer who first taught me that the medieval Arabic prose I had been told about and the medieval Arabic prose I encountered when I sat down with it were not the same body of work. I had been told, going in, that the prose of the early ʿAbbasid period was a lapidary thing, ornate, allusive, encrusted with quotation, the kind of writing one puts under glass. Some of it is. al-Jāḥiẓ is something else. He writes like a man who has just come back from the market and cannot wait to tell you what he has seen, in the order he saw it, with the prices.

The miser I want to introduce you to is Khālid ibn Yazīd, called al-Mukhāriq — the Cozener — a notorious tightwad in the Banū Tamīm quarter of southern Basra. The Banū Tamīm were one of the great Bedouin-descended Arab clans of the city, and the quarter that bore their name was a sprawling clan-neighbourhood along the canal-system that fed the southern part of town from the Shatt al-Arab. Khālid lived there. He was not a fictional miser. He was, on the evidence al-Jāḥiẓ gives us, a real man whom al-Jāḥiẓ either knew personally or knew at one short remove through informants whose accounts he treats as actionable. The book runs on actionable accounts. It runs on the kind of evidence that would be admissible in a small civil case.

Khālid is, by al-Jāḥiẓ's reckoning, a virtuoso of the small dishonesty. He short-weights at the mill. He pre-tastes at the wine merchant's. He brings guests through the long way to the dining mat so they will have to admire the courtyard and then leave before the food. The part al-Jāḥiẓ keeps coming back to, though, is his hands.

The scene is this. Khālid is sitting with members of his clan in the front courtyard or shaded entry of his house — al-Jāḥiẓ does not specify which, because al-Jāḥiẓ trusts his Basran reader to know where Banū Tamīm men sat in the afternoon in the 860s — when a beggar comes through. The protocol for the beggar is well-established. You give him a coin. You are not expected to give a generous coin. The currency of Basra in this period included two relevant denominations. The dirham was a struck silver coin of about three grams, a fingertip across. The fils, plural fulūs, was the local copper coin. A fils was the standard alms for an ordinary beggar; a dirham was — well, a dirham was real money. Forty fulūs to a dirham was a working rate among Basran moneychangers that decade.

The two coins were of strikingly different sizes. The dirham was small and dense — silver, a thumbnail wide, lighter than its volume suggested. The fils, struck under local Basran issue, were unusually large for copper: broad flans, thin strikes, sometimes broader across than the dirham itself. This is a small numismatic fact and it is the only reason the joke works.

Khālid reaches into his purse without looking. His fingers find a coin of the correct size — large, broad, palm-fillingly satisfying — and he passes it to the beggar. The beggar takes it. Khālid happens to glance down. Something small and silver-coloured has been left behind in his own palm. The dirham is in the beggar's hand. The fils is in his.

I want to camp here for a moment because al-Jāḥiẓ camps here. He does not give the moment to us as a setup-and-punchline. He gives it to us as a sequence of physical facts in the order they happened: the reach, the size, the handover, the look down. And then he gives us the sentence I have been waiting for. Khālid plucks the silver dirham back out of the beggar's hand and replaces it with the copper fils, and the beggar — the only person in the scene who does not, on the evidence we have, deserve any of this — is sent on his way with what he came in for.

The scene is over in three movements. The reach. The look. The substitution. al-Jāḥiẓ does not editorialise. He does not say can you imagine. He does not warn the reader to admire. He simply records, in a tone that is closer to forensic than to satirical, that this is what Khālid did and that Khālid did it before he had finished thinking about it, and that the only reason any of it had been possible in the first place was the unusual size of the Basran fils. The miser's reflex, in al-Jāḥiẓ's hands, is faster than his shame.

I have read this passage in two English translations — Jim Colville's and R. B. Serjeant's — and at the same place in both translations, I notice myself smiling at the same thing. It is not the substitution. The substitution is what the joke is about. The thing that does the comic work is the order of operations. Khālid doesn't see what he has given before he gives it. He sees what he has given when he looks down at his own hand and finds the wrong coin there. The mistake has already been made. He could let it go. He is among his clan. He could shrug, he could attribute it to charity, he could announce — for credit — that he had decided in the moment to be generous. He does none of these things. The hand goes back out. The coins switch places. al-Jāḥiẓ's prose, which is otherwise capable of long allusive arabesques, gets monosyllabic.

The thing I want to say about this is that I have read it now in three different places — Colville, Serjeant, and a French rendering by Charles Pellat done in the 1950s — and at no point in the encounter did I feel as though I was reading something historic. I felt as though I was reading something recent. The encounter has the shape of a New Yorker shorts page. It has the shape of a David Sedaris paragraph about a man at the airline counter. It has the shape, with one specific class of changes around the names of the coins, of a Twitter thread some friend of yours has just posted. The substitution scene is what happened. The fact that Basran fulūs were unusually large is the architectural detail that lets the substitution be feel-undetectable. The fact that the substitution is feel-undetectable is what gets Khālid into the wrong coin in the first place. Take any element out and the joke fails. al-Jāḥiẓ has not built it as a joke. He has recorded it, with the level of resolution that makes it a joke.

Something I want to say about al-Jāḥiẓ before I leave him to walk to China. He is writing this book in the same decade — the 860s — that Baghdad's House of Wisdom is finishing the second great wave of Greek-into-Arabic translation. The intellectual world he sits inside has Aristotle, Galen, and Ptolemy in pristine new Arabic editions; al-Jāḥiẓ himself moved between Basra and Baghdad and was, in Baghdad, on at least one occasion paid for a book by weight. His patrons were the chief judge of Basra and successive caliphal viziers. He was not a marginal figure writing in the corner. He was, by the metric of what his city paid for what its writers wrote, a successful author at a peak moment of his civilisation. And what he chose to write, at that peak moment, was a book in which he profiled a man called the Cozener pulling a silver coin back out of a beggar's hand.

I want to mark the city itself, briefly, because Basra in this period is the kind of place a contemporary reader is unlikely to have a picture of. It was built as a garrison town in the early caliphal expansion of the 630s, on a tidal estuary at the head of the Persian Gulf where the Tigris and Euphrates merge into the Shatt al-Arab. By the 860s it was perhaps the second or third largest city in the Islamic world after Baghdad, with a population that scholars have estimated at upwards of two hundred thousand — large by any pre-modern measure. It was a port, which meant it had Indian merchants and Sogdian traders and Ethiopian sailors moving through its bazaars; it was a centre of Arabic linguistic scholarship, which meant it had grammarians arguing across courtyards about case-endings; it was a centre of Muʿtazilī rational theology, which meant it had philosophers working out, in real time, the problem of how God could be just if the world contained children with toothache. al-Jāḥiẓ's Basra was, in other words, exactly the kind of city that produces a comic ethnographer. There were enough specific kinds of people to be specific about. Khālid pulling back the dirham is a Basran fact. The book is the city's mirror.

The book is a procession of portraits. al-Jāḥiẓ is not, structurally, doing anything different from what a contemporary observational comedian does in a club set: he is going person by person, building each one out of one or two physical specifics, and trusting the reader to do the laughing. The targets are not always individuals. The book also organises misers by region, with the Khurasanians serving as a recurring class of subject — al-Jāḥiẓ devotes substantial attention to them — and one of the standing jokes of the book is a kind of competitive ranking among regional varieties of stinginess, the same way contemporary stand-ups can get a long set out of regional varieties of bad behaviour at a wedding. He is mocking; he is also, because he is al-Jāḥiẓ, lifting the regional traditions onto the page so they can be read.

The disposition of the prose toward its subjects matters here, because it matters for the convergence claim. al-Jāḥiẓ is not, in this book, contemptuous of his misers. He is delighted by them. His misers are full people, with techniques and rationales and small dignities, and the book reads not as a tract against avarice but as a writer's catalogue of the variety of the human in respect of money. The position the prose takes toward Khālid is the same position Boccaccio's prose will take, five centuries later, toward Frate Cipolla — the writer is enjoying his subject too much to want him punished. That stance, too, is something the chapter wants to notice. Two writers, half a continent and half a millennium apart, taking the same affectionate-observational posture toward their comic subjects. The posture is not transmitted between them. The posture is independently invented.

His miser also tells me something I had not been told. al-Jāḥiẓ wrote Kitāb al-Bukhalāʾ roughly eight hundred years before Molière wrote L'Avare, roughly nine hundred years before the English-language miser tradition produced A Christmas Carol's Scrooge, and roughly eleven hundred years before Larry David started doing what he does. The character type — the man whose unwillingness to part with money produces the small daily comedy of his life — is, on the manuscript record, oldest in Arabic. The medieval Arab-prose tradition perfected the type before any of the European traditions had got to it. They didn't borrow it. The Arabic version was just there, finished, generations earlier, on the desk of a goggle-eyed man in Basra who was writing the kind of book he was paid by weight to deliver.

al-Jāḥiẓ wrote two hundred other books, by his own count, on subjects ranging from rhetoric to zoology — his Kitāb al-Ḥayawān, the Book of Animals, runs to seven volumes and contains arguments about cross-species mimicry that would not be out of place in a modern ethology seminar. Most of his books did not survive in full. The story I cannot stop thinking about is the one his medieval biographers tell about his death: he died, the story runs, in his Basran study, crushed by a collapsing pile of books he had been working through. Whether or not the story is accurate — and the medieval Arabic biographical tradition is, on the matter of writers' deaths, somewhat fond of poetic justice — it is the death his readers thought he ought to have had. A man buried under his own library. The Book of Misers survives the collapse. So does Khālid. So does the dirham going back into the hand.

A horse at court

I want to walk about five thousand miles east and roughly two and a half centuries earlier, to a court in northern China.

Hou Bai — 侯白 — was a Sui-dynasty literatus of low official rank, working in the imperial entourage of the Sui court sometime between 581 and 618 of the common era. The Sui ruled a unified China for less than forty years, sandwiched between the long fragmentation of the Northern and Southern Dynasties before them and the Tang ascendancy after them, and they ruled it with such administrative ambition — building the Grand Canal, rebuilding the Great Wall, attempting and failing to subjugate Goguryeo three times — that they exhausted the dynastic mandate in two generations. Hou Bai is one of the small handful of named comic voices we have from that thirty-eight-year window. The book associated with his name is called Qi Yan Lu, 啟顏錄 — Records to Crack a Smile, in the standard English rendering, or, more literally, Records of Inducing Laughter — and it is one of the earliest surviving Chinese jest-books.

I should be careful about the word "surviving." The original Qi Yan Lu as a self-contained book is lost. What we have are quotations from it embedded in two later compendia: the Taiping Guangji 太平廣記, the great Northern Song imperial encyclopaedia of strange and miscellaneous tales completed in 978 under the editorship of Li Fang, and the Shuofu 說郛, a Yuan-and-Ming-period anthology of literary excerpts. A modern critical compilation by the philologist Wang Liqi assembles the surviving fragments. What I am about to tell you about Hou Bai comes through that filtration: the quotations made it into the compendia because the editors of the compendia wanted them there. The selection is not Hou Bai's. The fact that the lying-horse anecdote and the three-bald-heads exit line both made the cut, in the same compendium, in the same century the Qi Yan Lu was being preserved for posterity, suggests that the Song editors had a fairly clear sense of which Hou Bai bits the tradition wanted to keep.

The lying-horse scene runs as follows.

The Sui court is hosting a foreign envoy. Foreign covers a lot of ground in this period — the Sui had diplomatic traffic from the Korean peninsula, from Turkic confederacies on the northwest frontier, from Sogdian merchant interests further west, from Indic Buddhist monastic networks. The envoy, whoever he is, is in the kind of mood foreign envoys have always been in everywhere when they have come to a richer court with a smaller treasury at home. He is overcompensating. He is talking down to staff. He is determined to be seen, this morning, finding fault with something.

The arena he picks for the fault-finding is the post-station horseflesh. The Sui court ran a vast horse-relay system; the stables at the capital were not just decorative, they were operational, and the inspection of a state-issue horse by a foreign visitor was a small piece of theatre with conventions about which a man like Hou Bai would have known a great deal. The envoy decides to ask a low-ranking attendant in plain dress for a price quote on a particular animal, which he will then dismiss as overpriced because that is the kind of thing he would, this morning, like to be seen doing.

What the envoy does not know is that the man in plain dress is Hou Bai. The fragments we have do not say why Hou Bai is in plain dress, but the inference scholars draw is the obvious one: he has done it on purpose, for sport, possibly with the connivance of his colleagues at court. He is in disguise the way a contemporary court correspondent might be in a Carhartt jacket on the loading dock. The masquerade is plausible because the Sui court had hundreds of men in plain dress around it at any given moment, and one more was unremarkable.

The envoy, in the middle of his pose, breaks wind. Audibly. The court is too well-trained to laugh out loud, but everyone present has heard it, and everyone present is waiting to see what the envoy will pretend just happened. The envoy recovers his pose, the colour rising in his face, and asks the man in plain dress what he would charge for a horse like the one in the stalls.

The servant looks at the horse. He looks at the envoy. He looks back at the horse. Then he says, in the flat tone of a man giving a real estimate to a real customer, that a horse that lies on its side and breaks wind is not worth a single coin.

I want to camp here too.

The retort is doing several things at once. On the surface it is an estimate of the horse — the kind of veterinary remark a real horse-trader might make about a real animal that has gone down in the stall, since horses do, occasionally, lie down and break wind, and it is a sign of colic that knocks the value of the animal toward zero. That meaning is the cover. Underneath the cover, the retort is also a price quote on the envoy: you are worth nothing. Underneath that, in the deepest layer, the retort is doing the thing the envoy had been waiting all morning for someone to do, which is to register the fart. He hadn't registered it. He had pretended it hadn't happened. He had recovered his pose. The man in plain dress, in apparent ignorance of the pose, has answered the question the envoy asked, and on his way through the answer he has — without lifting his head from the price-quote register — registered the fart with two specific words: lies on its side. The envoy is lying down. He is, at this moment, on his side. He has, just now, broken wind. The price for that horse is zero. The price for that horse is also the price for the envoy. The horse and the envoy are, in the geometry of the retort, the same animal.

The court, at this point, is permitted to laugh, because the apparent sense is decorous (a price quote on a horse) and the actual sense (the envoy is the horse) is contained in a phrase the speaker can deny meaning. The decorum is the cover. The envoy, who is not unintelligent, gets the joke at his own expense before the laughter has finished. He is reported, in the Qi Yan Lu fragments preserved through the Taiping Guangji, to have laughed the loudest. He had no choice. Either he laughed loudest or he insisted, in front of the assembled Sui court, that the comment be parsed for him line by line.

I want to mark, while we are here, what kind of author Hou Bai is, because the description matters for the convergence claim. The Sui court tradition kept a class of literate men whose role was the production of refined wit on demand — xiehui practitioners, in the Tang-period scholarly classification of comic prose, the writers of mocking and satirical pieces. Hou Bai was at or near the head of that class for his generation. He was, by any reasonable contemporary description, a comedy writer at court — a position the Tang would more clearly institutionalise after him, and which the Han court had had its earlier version of in the paiyou jesters whose biographies Sima Qian had collected eight centuries earlier in the Shiji. Comedy at the Chinese court was a continuous professional category by the time Hou Bai sat down to compile the Qi Yan Lu. The tradition al-Jāḥiẓ writes inside, the early ʿAbbasid adab prose tradition, is also a continuous professional category by the time of Kitāb al-Bukhalāʾ — comic writing as a recognised, paid, court-adjacent form of literary work. Two cultures separated by a continent and three hundred years had developed, independently, a class of professional comic writers attached to a literate court, producing books in their own languages on subjects ranging from miser-sketches to status-inversion retorts to court satire to literate-versus-clerical wit-debates. The convergence is not just at the level of individual jokes. The convergence is at the level of the kind of writer who would write the jokes down.

The second Hou Bai story I want to tell you involves three Buddhist monks.

It is a debate. The Sui court hosted public debates between the literati class and the Buddhist clerical class as a form of intellectual sport — Buddhism was at this point well-established in northern China, having spread across the Northern Wei and into the Sui court patronage system over the previous two centuries, and the question of whether the Buddhist tradition or the classical literary tradition held the higher ground was a live one with real stakes. The Buddhist establishment was a property holder, a tax-exempt land manager, and a parallel intellectual authority that could and did challenge the literati's monopoly on textual prestige. Public debates were one of the venues at which the relative status of the two traditions was negotiated. There were rules. There was an audience. There was a kind of running scoreboard. The audience would have included court literati, monks attached to court-patronised monasteries, courtiers without strong allegiance to either side, and — this is important — the emperor's representatives. The losing team in a high-profile debate had real consequences to absorb, in patronage and in face. Hou Bai, by virtue of being Hou Bai, is the literati team's anchor. The monks — there are three of them — make a credible run; they are doctrinally fluent, scriptural references at their fingertips, and the audience starts to tilt their way. The literati team begins to look as though it will lose.

Hou Bai stands up. The debate, before he speaks, has been running for some time. He has the audience's attention. He delivers what is essentially a closing line, and the closing line is a piece of gambling slang.

He says: three bald-heads can't beat one Lu.

I am going to take a moment to walk you through what is happening in those eight English syllables, because the line is doing more than one thing and the chapter is going to need to refer to the move it is making for the rest of the book.

The Chinese word for bald in the Sui period was tū. Buddhist monks, then as now, shaved their heads; tū stood for the physical baldness, and by an entirely standard piece of metonymy, for the monk himself. Three bald-heads, in literal Chinese, is the three monks. Lu is where the line gets interesting. Lu is, simultaneously: a family name, written 盧, which positions the speaker as a bearer of a family-name on the other side of a clan-balance from Hou (侯) — Hou Bai is announcing his own house against the monks' headcount; and a piece of dice slang from the gambling game chu-bu (樗蒲), played with four wooden lots, in which the highest possible throw was called lu and lower throws had names that made up most of the vocabulary of low comedy. The proverb three small can't beat one Lu was already in circulation among gamblers as the equivalent of three of a kind can't beat a straight flush. The reference is to dice, and the literati would have known the reference, and the gamblers would have known the reference, and there is no evidence that anyone present had any difficulty with the double layer.

So what Hou Bai actually says, in one sentence, is this. In the gambling frame, our trio of low rolls cannot beat your single high roll. In the literati frame, your three bald-headed monks cannot beat me. The frame is hinged on the ambiguity of the syllable lu. The monks, who can hardly take offence at an analogy from gambling without confirming it, are stuck. Either they lose the debate gracefully or they lose it ungracefully by complaining about the comparison. They lose it gracefully. Hou Bai walks out. The literati team, by way of an exit line that nominally has nothing to do with the doctrine the debate had been about, has won.

I want to give what is happening here a name, because the book is going to need it.

Call it pun under literati cover. Or — the version I am going to settle on, since this is what the move actually does — call it pun-as-political-resistance. The structural shape of the move is this: in a setting where a class of speech is forbidden — disrespect for monks at court, an obscenity in a polite room, a contempt for an envoy from a power your own court is courting — the speaker says the forbidden thing through a pun whose surface meaning is decorous. The decorum is the cover. The pun is the obscenity. Both meanings are present. The audience that gets it gets both. The audience that gets only the surface gets only the surface and goes home satisfied that nothing inappropriate has been said. The forbidden meaning has been delivered, in public, to an audience that includes the people the meaning is forbidden in front of. And the speaker can, with a straight face, deny it. I was speaking of dice.

The Sui court contained gamblers and literati and clerics and people who were several of those things at once. The room can hear all the layers. The room can also, if pressed, confirm that nothing improper has been said.

I am going to come back to this move six chapters from now, in 2009, in another Chinese context, in which a children's-nature-song parody about a fictional grassland animal will turn out to be doing exactly what Hou Bai is doing here, against a censor and not a Buddhist debating team. The animal will not be a horse. The cover register will not be the price quote. The forbidden meaning will not be the envoy is the horse. But the move is the same move. Pun-as-political-resistance. Hou Bai is where it has its first preserved name. The move was older than him; he did not invent it, no one did; what he and his Song-period editors did was put the move on the manuscript record in a form a later civilisation could recognise. Fourteen centuries downstream, with a different language, a different empire, a different forbidden speech, the cover register kicks in again. By that point the line will not be three bald-heads can't beat one Lu. The line will be a song about an alpaca threatened by a river crab, sung by a children's chorus in the bland register of an animated educational video.

But that is six chapters away. For now, what I want you to hold is the shape: a forbidden statement, decorum-cover, pun, plausible deniability, audience splitting into the layer that hears it and the layer that doesn't. Three bald-heads can't beat one Lu. That is the shape.

Convergence, not transmission

Hou Bai is writing in northern China at the start of the seventh century. al-Jāḥiẓ is writing in southern Iraq at the end of the ninth. Neither has read the other. Neither has any plausible route to read the other.

I want this distinction to be clean before we get to Tuscany, because it is the distinction that does the chapter's work.

The Chinese jest-book tradition, in Hou Bai's period, is composed in literary Chinese; it is being transmitted on bamboo, silk, and paper inside an east Asian literary network whose western edge is Tang Dunhuang and whose far interlocutors are, at the most distant, Sogdian-speaking merchants on the Silk Road. The Arabic adab tradition al-Jāḥiẓ writes in is composed in classical Arabic on parchment and papyrus inside a network whose eastern edge is Khurasan and whose far interlocutors are, at the most distant, Indic-speaking traders in northwestern India. The two networks barely touch. There is no Hou Bai manuscript in al-Jāḥiẓ's Basra. There is no al-Jāḥiẓ manuscript in Hou Bai's Chang'an. The miser sketch is not a Chinese borrowing into Arabic literature, and the lying-horse retort is not an Arabic borrowing into Chinese literature, and there is no third tradition behind them both that either author has read.

When a comic structure shows up in two cultures with no contact between them, there are two things that can mean. The lazy meaning — the slogan-meaning, the dinner-party meaning — is that the cultures share a sense of humor. Everyone everywhere finds this funny. That is wedding-toast prose. The harder meaning is that the two cultures have independently arrived at the same comic engine. This is the meaning that does the work.

The lazy meaning is real enough — there are jokes that travel by transmission, and a wedding-toast sentence, properly used, has its place — and the easiest way to see what the lazy meaning looks like is to look at one. Juḥā.

Juḥā is the Arabic-Persian-Turkish folk wise-fool, and his stories really do travel. The character appears under different names — Juḥā in the Arab world, Nasreddin Hoca in the Turkish, Mulla Nasreddin in the Persian, Giufà in the Italian, Bahlul or Bohlul in some Persianate retellings — and the stories told about him from one tradition to the next are mostly the same stories. He is a scholar-fool whose foolishness is, on inspection, frequently wisdom in disguise; he locks his door against thieves and then carries the door with him; he loses a coin in the dark and looks for it under a streetlamp because the light is better there; he sits on the branch he is sawing off and is surprised to fall.

The Juḥā tradition is one of the cleanest cases of folkloric transmission in the medieval Mediterranean and Near East. The character is attested, by name, as early as Ibn al-Nadīm's Fihrist of around 987, which lists a Kitāb Nawādir Juḥā — a Juḥā-anecdote book — that has not survived. By the late medieval and early modern period the tradition has compiled at length in Arabic, Persian, and Turkish; in the modern period it has been printed, retold, translated, and adapted into stand-up routines from Cairo to Istanbul to Sarajevo. The stories travel by being retold. They are not converging from independent sources. They are being passed along.

The Juḥā story I have on my list is the donkey one. A neighbour comes to borrow Juḥā's donkey for a trip; Juḥā, not in the lending mood, tells the neighbour the donkey was sold yesterday. At that moment the donkey, very much still in the stable behind Juḥā, brays. The neighbour protests. Juḥā draws himself up and answers, with what one imagines as the dignity of an offended scholar, would you take the word of an ass against mine?

I should be honest about the dating of this one. The bray-and-answer punchline of the donkey story is securely attested in the late-medieval and early-modern Juḥā compilations; in the 9th-century fragments where the character first appears by name, the specific bray version is not confirmed. The story may have been current orally in the 9th century, and the character is — there is no question that Juḥā as a figure was in circulation when al-Jāḥiẓ was writing — but the bray punchline as we have it is a later attestation. The Marzolph scholarship on the Juḥā tradition is the place to read about this; what I want to flag, and not bury, is that the bray-line is not 9th century in the way the Khālid-and-the-dirham line is. It is something more like 13th century in its earliest secure form.

What the Juḥā tradition shows me, hedge included, is the lazy sense of universality. A joke that crosses Arabic, Persian, Turkish, Greek, and modern Albanian-Bosnian stand-up because it has been carried across all of those, by physical retelling and physical translation, from a single common ancestor inside the pre-modern Near Eastern story-stream. That is how transmission works. It is real, it produces real cross-cultural commonality, and it is not what this chapter is about.

What this chapter is about is the other case. Two cultures with no transmission, producing the same comic engine. al-Jāḥiẓ and Hou Bai. Khālid pulling back the dirham and the man in plain dress saying the horse is worth nothing. Neither a chain of retelling, neither a borrowed common source. Two writers, two thousand miles and three centuries apart, recording, on the manuscript record, the same comic shape. That is the structural sense of universality. That is the version of the claim the book is willing to defend. We need a third instance, and a fourth, and the rest of the chapter is going to give us them.

There is a small thing I want to flag about the manuscript medium itself, because the chapter's argument depends on it. The codex — the bound, written, copyable book — is what makes the convergence visible to us. Comedy from the Sui court and comedy from ʿAbbasid Basra would have had no encounter point at all without their writers' decision to commit the comedy to a recordable form, on a substrate that other writers could copy, in a script that would still be legible after the people in the room were dead. Every case in this chapter is a case we have because somebody, in each tradition, sat down and wrote it down. That is what the manuscript era is, structurally: the period in which comedy passes from being something said in a room to being something a stranger fifteen centuries later can sit in another room and read. Without the codex we would not know about Khālid; without the Taiping Guangji we would not know about the lying horse; without Boccaccio's eye for the local we would not know about Cipolla; without Poggio's Bugiale we would not know about Dante and the bones. The argument's evidence base is, exactly, what the codex caught. The cases that follow are the cases the codex caught.

A feather and a casket

The Decameron was begun in the year 1349 in Florence, while the city was still counting its dead from the second wave of the Black Death. The plague reached Florence in the spring of 1348; Boccaccio, then in his mid-thirties, was in the city or close to it through the worst of that summer. Estimates of the city's mortality run between fifty and seventy-five per cent in the months of the outbreak's peak. The frame story of the Decameron, in which ten young men and women retreat from plague-stricken Florence to a country villa for two weeks and entertain each other with a story a day apiece, is the writer's bid to set a workable comedy against an actually unworkable death-count, and the bid was successful in the way the strongest comic bids tend to be: by not flinching at what was making the comedy necessary in the first place.

The story I want to tell you is from Day 6, Tale 10, and Day 6 is the day on which the assembled company is telling stories about quick-wittedness. The novella is told by Dioneo, the most ribald of the ten storytellers, and it is set in the village of Certaldo, on a ridge in the Tuscan countryside about forty kilometres southwest of Florence. Boccaccio knew Certaldo well — his father's family was from there, and Boccaccio himself spent the last decade of his life there. The setting is not generic. The setting is local to the writer, and the writer tells the story like a man telling you about something that may or may not have happened in his actual home town.

I want to spend a sentence on Day 6 specifically, because the structure of the Decameron is doing comic work the way a stand-up special's set list does. Each of the ten days has a theme; each day's stories build to the day's last story; each day's last story is told by Dioneo, who is given license, by the brigata's rules, to ignore the day's theme if he wants to. Day 6's theme is, in Boccaccio's framing, the witty rejoinder — the words a person says in a tight spot to escape it — which makes Day 6 the Decameron's comedy day, the day on which the company has agreed to spend its hundred minutes of narrative entertainment specifically on the question of what good comeback delivery looks like. Dioneo, given the closer's slot on comedy day, does not deviate from the theme. He tells the longest, most virtuoso, and most structurally complete improvisation-recovery story in the entire collection. The Cipolla novella is, in the Decameron's engineered architecture, the punchline of the comedy day. Boccaccio knew where he was putting it.

The Franciscan relic-pedlar Frate Cipolla — Brother Onion — rolls into Certaldo on the saint's feast day. Cipolla is the sort of itinerant friar whose role in the late-medieval Tuscan religious economy was to circulate among the hill villages, preach, and collect alms in exchange for a glimpse of, or a touch from, a relic. The relics in circulation in the Tuscan friars' bag-trade in this period were a competitive market. There were always more pieces of the True Cross in pious hands than the True Cross had ever had pieces. There were, by some 14th-century church inventories, several distinct skulls of John the Baptist, in different reliquaries, in different cities. The relic economy was one of the load-bearing pieces of the late-medieval Italian church, and like any load-bearing piece, it had its grift.

Cipolla announces from the village square that he will, that very afternoon, in the church, display to the faithful a relic of unprecedented value: a feather from the wing of the Archangel Gabriel, dropped in the Virgin Mary's bedroom at the moment of the Annunciation. The crowd is excited. Two local gentlemen — Boccaccio names them, which is the part I want you to notice he does — Giovanni del Bragoniera and Biagio Pizzini — are excited too, but in a different register. They have known Cipolla for years. They have been on the receiving end of his pre-sermon hospitality and his post-sermon collection-plate and they think the feather is going to be funny. They are also bored. The two of them slip away from the company, walk over to the inn where Cipolla is taking his lunch, slip past his serving-boy Guccio Imbratta — Guccio is, in Boccaccio's portrait of him, a layabout currently flirting with a kitchen girl named Nuta — go up to Cipolla's room, and find the reliquary casket.

They open it. There is a feather inside. It is a parrot's tail feather. They are not surprised by this; Cipolla's relic-bag is the sort of thing a paying audience does not get to inspect at close range, and a parrot's tail feather, in the dim light of a country church, can easily play an angel's wing-feather to a congregation that wants to see one. Giovanni and Biagio remove the feather. They look around the room for something the size and weight of a feather to put in its place. They find a piece of charcoal — possibly more than one piece; the Italian text uses the plural carboni. They drop the charcoal into the casket, close it, slip back out past Guccio (who has by now noticed nothing), rejoin the company at the square, and wait.

The afternoon arrives. The church fills. Cipolla mounts the pulpit and begins his sermon, which is — this is one of Boccaccio's pleasures — a tour de force of geographical fabrication. He has, he announces, travelled in pursuit of this feather to lands the congregation has never heard of, and his catalogue of those lands is a long string of invented and distorted place-names — Truffia, Buffia, the Bridge of Parione, the country of the Abruzzi where men go in clogs over the mountains and clothe pigs in their own intestines, the Parmesan mountains of grated cheese where the inhabitants do nothing but make ravioli and boil it in capon broth — delivered in the grave register of a homiletic geographer. The sermon, structurally, is a parody of relic-pilgrimage rhetoric, in which the friar establishes his authority by listing the holy sites he has personally visited. Cipolla's holy sites are jokes. He has not been to any of them. The congregation is rapt anyway.

He reaches the climax. He raises the casket above his head. He addresses the congregation. He invokes, with priestly gravity, the Archangel Gabriel and the Virgin Mary. He opens the casket.

He looks down into it.

There is no feather. There is charcoal.

I want to camp here, because what happens next is the load-bearing piece of the joke and it is not the substitution. The substitution is what makes the joke possible. What does the comic work is what Cipolla does in the next seven seconds of stage time.

He does not pause. He does not look around the church. He does not wonder aloud who has been in his casket. He does not so much as register surprise. He simply continues his sermon, in the same register, on the same tonal level, to the same congregation, and announces, with calm authority, that he has — in his preparatory eagerness — brought the wrong casket. The casket of the Archangel Gabriel's feather and the casket of the coals on which Saint Lawrence was roasted are very alike. Not seldom, Cipolla observes, I mistake one for the other. It is, providentially, the Saint Lawrence casket he has brought today; and the saint has, evidently, willed this. He invokes Saint Lawrence. He raises the casket. He explains, reverently, the relic the casket actually contains. He offers, by way of an unanticipated grace, to mark with these very coals a cross on the brow of any congregant who comes forward, and to grant by that mark — for the year — protection against fire.

Saint Lawrence, the historical saint, was martyred in Rome in 258 by being roasted to death on a gridiron. The story of his roasting is one of the most widely known martyrologies in the late-medieval Catholic calendar. The coals on which he was roasted are exactly the kind of relic that, in the inflationary relic economy of the period, could plausibly be in a Franciscan friar's casket. The pivot from feather to coals is, on the strict logic of the relic-grift, perfect. Cipolla has, in the seven seconds between opening the casket and resuming his sermon, found the precisely calibrated saint whose relic would naturally be a piece of charcoal.

The congregation comes forward. They line up. Cipolla draws crosses on their foreheads with the charcoal that Giovanni del Bragoniera and Biagio Pizzini placed in the casket two hours earlier. He is paid in alms. Boccaccio's narrator notes — in the kind of small final twist Boccaccio likes — that Cipolla collected more in alms that afternoon than he would have collected for the feather. The pranksters, Giovanni and Biagio, are sitting in the crowd watching this. They cannot stop laughing. Boccaccio's narrator notes that they laughed so hard they were afraid their jaws would crack. They keep their seats; Cipolla, who has by now spotted them and worked out what has happened, says nothing, because nothing he could say would improve the situation. The story ends with the friar's purse heavier and the village, on the whole, satisfied.

I have been reading and rereading this novella for years and the part I keep noticing is the pacing. Boccaccio has set a trap that Cipolla has been walking into for two pages. The substitution sets it. The geographical-tour sermon stretches it. The opening of the casket sprints it. And then, in the moment Boccaccio has been preparing, Cipolla does not fall. He pivots. Boccaccio could have written the scene as the trap closing on the friar. He has written it as the trap closing on the friar and the friar walking out the other side richer. The audience for the joke — Boccaccio's audience, the readers of the Decameron — is, structurally, the audience inside the church: they have been led to expect a punchline at the friar's expense, and the punchline is at the pranksters' expense instead, and the punchline is also, for the same reason, at the relic-economy's expense. The friar is not exposed. The relic-economy is exposed. The relic-economy is exposed in the safest possible way, which is by being parodied from inside its own audience by one of its own practitioners, with the practitioner emerging financially better off.

The audience that received the Decameron in the second half of the 14th century did so under several layers of complication. Boccaccio's book was not, on first reception, treated as a comic masterpiece. It was treated as something between a vernacular curiosity and a moral problem. It circulated, copied by hand, among Tuscan literate readers — the manuscript era is exactly that: literate Tuscany copies the Decameron for itself by the page — and was read with what one might call the standard medieval mixture of relish and unease. The Cipolla novella was singled out, even within the Decameron, as a piece that would have to be defended; relic-mockery was a delicate matter, even when the friar in the story was inventing his geography out of whole cloth. Boccaccio himself, in his old age, repented of the Decameron — or said he did, in letters that may themselves be performances of the kind of late-life rebranding aging writers go in for. The novella survived its writer's regret. It is what gets copied. It is what gets reprinted, and translated, and reprinted again, and put on lists of the works that defined what vernacular Italian could do. The Cipolla scene is, in fact, one of the things that taught the European prose tradition what comic prose could be. Without it, the next century is missing one of its templates. With it, the template is in place by the time the Tuscan humanists who knew the book by heart had the time to write their own.

I want to say one more thing about what Boccaccio is doing here, because it has to do with the convergence claim of this chapter. Cipolla is improvising. The improvisation under pressure is the load-bearing comic engine: a man who has been caught is required to keep talking, in the same register, without breaking, until the situation has resolved itself in his favour. The shape of the engine is exactly the shape of the engine that drives, six and a half centuries later, every working sitcom episode in the Curb Your Enthusiasm tradition, in which a Larry David or a John Cleese has been caught in a small crime and must improvise, in real time, a face-saving extension of the lie. The 14th century has the friar in front of a duped congregation. The 21st century has Larry in front of a deli counter. The cover stories are different. The comic structure is the same. Boccaccio did not invent it. al-Jāḥiẓ did not invent it. Hou Bai did not invent it. They each, in their own century, in their own language, in their own cultural frame, recorded a working instance of the engine on parchment, papyrus, or paper, and the engines they recorded were the same engine.

A heap of bones

I want to leave Tuscany briefly and return to it, because the next case is also Tuscan but a hundred years later, and because the way it travels through the manuscript record demonstrates something the Cipolla story doesn't: the way a piece of comic dialogue can survive four-step transmission and arrive in print intact.

Poggio Bracciolini was a papal secretary in the curial bureaucracy of Rome through most of the first half of the 15th century. He served under five popes. He was a humanist of the second generation — the generation after Petrarch and Boccaccio, the one that did the systematic recovery of classical Latin manuscripts from German monastery libraries. Poggio personally rediscovered Quintilian's Institutio Oratoria, Lucretius's De Rerum Natura, and several of Cicero's lost orations. He wrote in Latin. He wrote a great deal — letters, dialogues, treatises on avarice and on the unhappiness of princes — but the book that carried his name through the next four centuries was a small one, written in spare hours on the curial circuit, called the Liber Facetiarum, the Book of Witticisms, or simply the Facetiae.

The Facetiae is a collection of 273 short anecdotes in classical Latin: tavern jokes, court gossip, sermons gone wrong, the funny things people had said in the curial circles Poggio worked in. He composed it from roughly 1438 to 1452. After his death it became one of the first jest-books printed in Europe — first printed in 1470, then reprinted across the next century in dozens of editions in Latin, Italian, French, German, and English. The Facetiae is the book that fixes the European jest-book in its mature print form a generation before Joe Miller will do the same in English. It is a critical hinge in the history of how a joke gets from oral wit to printed quotation.

Poggio in his preface — the Facetiae opens with one — is candid about what the book is. It is, he writes, a record of the things he and his colleagues said over wine in the Bugiale, a back-room at the papal curia where the secretaries of the apostolic chancery gathered between official commissions. The Bugiale — the word can be rendered as something close to the lying-room — was a workplace social space, in the sense that any office's corner conversation is a workplace social space, and what Poggio claims to be doing is preserving the quality of the office gossip in a Latin sufficiently elegant that the office gossip will be readable in the next century. He is not wrong. The next century reads it. So does the one after that.

The anecdote I want to tell you is Facetia LVIII. It is set, by Poggio, at the dinner table of Cangrande della Scala — Can Grande, "Big Dog," lord of Verona, the most powerful patron of Dante's later years. The Della Scala family ruled Verona through most of the 13th and early 14th centuries, and the family had a habit of naming its male heirs after dogs: Cangrande's first name was, literally, Big Dog; his nephew Mastino II — Mastiff the Second — succeeded him; an earlier Mastino had founded the dynasty's lordship in the previous century. The whole house is, in Italian, named for dogs.

Dante is at dinner. He has been a long-term guest at the Della Scala court — he wrote much of the Paradiso there, between roughly 1312 and 1318 — and the court entertainments include, by 14th-century courtly standard, practical jokes. The attendants at the dinner have been instructed by some unnamed party — possibly Cangrande himself, possibly one of the courtiers — to pile gnawed bones at Dante's feet under the table during the meal. The bones come from the diners' plates. The diners are eating, the conversation is going, and the heap is growing, and at some appointed moment in the meal the company looks down and notices, with a great communal show of surprise, that Dante's place is surrounded by bones. The heap is laughed at. The laughter is the practical joke. The setup, on the grammatical surface, is that Dante has eaten more than anyone else and has dropped his own bones at his feet like a starving dog.

Dante stands up. He delivers, in flat Tuscan, the comeback Poggio preserves a hundred years later in his Latin: non è da stupire se i cani hanno mangiato i loro ossi — io non son cane. It is no wonder if the dogs have eaten their own bones — I am no dog.

The line has to be parsed in Italian to land. The pun is on cane, which in 14th-century Tuscan is at once the common noun for dog and the family name of his hosts. Cangrande's family name is Cane. The dogs in the room are not Dante. The dogs in the room are the diners around him, which is to say his hosts. The implication, the part of the line that does the comeback, is that Dante is the only person at the table who has not eaten the bones, because Dante is the only person at the table who is not, by family name and by a certain implication of behaviour, a dog. The line is short. It is one sentence. It is delivered standing. Dante does not laugh. Cangrande, on Poggio's account, is the first to laugh, because Cangrande was a man with the kind of dignity that could afford to laugh at a comeback at his own expense.

I want you to notice how many steps removed from the event this anecdote is by the time it lands on Poggio's desk. Dante dies in 1321. The dinner, if it happened, happened sometime between roughly 1312 and 1318. By Poggio's time — he is writing the Facetiae a hundred and twenty years after the dinner — the anecdote is a piece of received culture in Italian humanist circles. Poggio has not heard it from Dante. He has not heard it from anyone who heard it from Dante. He has heard it from someone who heard it from someone, and the chain is at minimum three or four bodies long. And what Poggio writes down, in mid-15th-century classical Latin in Rome, is a comeback that pivots on a Tuscan pun nobody in his Roman audience needs to be told to admire — the cane of the Della Scala family is by then proverbial in literate Italian. The line has survived four-step oral transmission across the entire Italian peninsula and into the next dynastic period of Italian high culture. It has survived because it was good. The Della Scala family who were the joke's victims were politically defunct by the time Poggio published; the politics of the dinner were dead with them. What was alive, and worth a facetia, was the punchline. The punchline travelled even when the politics didn't.

The reason I include the Dante anecdote in this chapter, alongside the Cipolla one and the al-Jāḥiẓ one and the Hou Bai one, is that it is a fourth instance of the chapter's main move — comedic structures that emerge in cultures whose contact with each other is partial, sporadic, or outright nil — but it is also, importantly, an instance of the transmissive sense rather than the convergent. The Cipolla story and the Dante story are both Tuscan. They are both being recorded by 14th- and 15th-century Italian humanists who are at one or two removes from the same intellectual community. The fact that both the improvisation-under-pressure recovery and the named-patron-pun comeback are being preserved by the same community, with the same prose discipline, in the same century, is a fact about Tuscan letters. It is not an instance of convergence. The convergence claim of this chapter is between the Tuscan-Italian strand and the Arabic-Persian-Chinese strand. Inside the Tuscan strand the cases are linked by transmission; that is what makes them a strand. The Dante anecdote is here to demonstrate that even when the politics around a punchline are dead — the Della Scala family is gone, Verona is no longer ruled by men named after dogs, the courtly conventions of attendants piling bones under tables are gone — the punchline travels. It survives in print because it was good enough to be repeated in oral form long enough to reach a man who would write it down.

By contrast, Hou Bai is not in the same network as al-Jāḥiẓ, and al-Jāḥiẓ is not in the same network as Boccaccio, and Boccaccio is not in the same network as Hou Bai. The triangulation is across strands. That is where the convergent claim lives.

What three writers were doing

I want to close the chapter on a small observation that I cannot get out of my head.

al-Jāḥiẓ in Basra in the 860s, Hou Bai in northern China sometime around 600, and Boccaccio in Florence in 1349 are doing the same job. Each of them is sitting down at a desk with a piece of parchment, papyrus, or paper, and recording, with deliberate prose discipline, the comic things that happened around him. None of them invents the comedy. The comedy is in the room — in the courtyard with the beggar and the silver coin; in the courtroom with the envoy and the lying horse; in the church with the casket and the charcoal. What the writers do is preserve. They preserve at the level of granularity that makes the joke survive: the specific weight of the dirham, the specific phrase lies on its side, the specific casket-mix-up line, the specific feather and the specific charcoal. The granularity is the medium-level of the chapter. The codex, the manuscript, the bound volume of jokes — these are the technologies that let an observation at the granularity of the size of a Basran fils travel into the next century and the next continent.

The convergence is not in what they are writing about. The Banū Tamīm beggar, the Sui-court envoy, the Certaldo congregation are not, on any plausible reading, the same situation. The convergence is in what the writers think is worth recording. al-Jāḥiẓ and Hou Bai and Boccaccio look at three completely different scenes in three completely different centuries, and in each of those scenes they identify, for posterity, a particular kind of moment: the moment at which a person's reflex outruns their dignity, the moment at which decorum becomes the cover for an obscenity, the moment at which a man caught in a small crime improvises his way back into authority. The moments are the same kind of moment. The granularity at which they are preserved is the same granularity. The comic engine is the same engine.

Nothing about the cultures producing these scenes is the same. The pre-modern Sui imperial court, the ʿAbbasid Basran clan-quarter, and the post-plague Tuscan parish church are radically different societies. They run on different foods. They worship in different idioms. They write in different scripts. They organise their gender relations and their property relations and their political relations in different ways. They would not, on first encounter, recognise each other as comprehensible. What they recognise — and this is what the chapter argues — is the engine. Khālid pulls back the dirham. The man in plain dress says the horse is worth nothing. Cipolla calls the charcoal Saint Lawrence. The audiences laugh in three different languages, in three different centuries, on three different continents. The audience of this book — me, you — recognises all three.

There is a temptation, on closing a chapter like this one, to land on a pious sentence about how this proves we are all the same under the skin. We are not. The Basran clan-quarter and the Sui court and the Tuscan parish were not the same, and we are not the same as any of them. What is the same is narrower. Something specific about a comic shape — the precise small mechanism that makes a man's reflex faster than his shame, or that makes a price-quote on a horse become an obscenity at an envoy, or that makes a casket of charcoal become the relic of a saint roasted on a gridiron — has been independently invented at least three times on the manuscript record, and the inventions are recognisable to each other.

I should also say what the chapter does not prove. It does not prove that the same comic engines are present in every literate culture between 500 and 1500. The chapter has only walked four cases and contrasted them with one. The argument is that within those four cases there are at least three independent inventions of the same engine across cultures with no transmission between them, and that this is not the kind of coincidence a thesis can dismiss as accidental. The fact that the case slate could be longer — that there are Sanskrit jest-traditions and Byzantine wit-collections and Sogdian fragments I have not opened — is the chapter's standing invitation: the longer the slate, the stronger the inference. What I have written is enough to put the inference on the table. It is not enough to mistake for a complete proof. The completion is, in the strict sense, ongoing scholarship; the thesis the book defends rides on the inference being directionally correct, not on the slate being closed.

What I do feel sure of, after several years with this material, is that the structural sense of universality — the harder sense, the sense in which independent inventions of the same comic engine are turning up in cultures that did not share a text base — is real, and the cases are not isolated. The miser sketch, the status-inversion retort, and the improvisation-under-pressure recovery are three engines among more. The chapters that follow add others: the Edo-period domestic-comedy beat in seventeen syllables, the late-Ming printed jest in an inversion-punchline rhythm, the 18th-century English witness-undoing-the-speaker structure, the 19th-century American tall-tale deadpan. The list grows. So far, on every engine the slate has named, the same finding holds: the engine appears in places that did not negotiate over its arrival. The convergence is not in any one case. The convergence is in the frequency of the convergence.

Within Poggio's own lifetime, the printing press is already at work in northern Europe; the Facetiae itself will be printed in 1470, eleven years after his death, and from there a punchline that has had to travel for a thousand years by oral retelling and patient scribal copying will start to travel verbatim, in tens of thousands of copies, across centuries and across languages. The same comic engines will start to be reproduced not in independent inventions but in literal word-for-word reuse — and that, the case for verbatim survival rather than convergent re-emergence, is the argument the next chapter makes. For now, what I want you to hold is that before any of those engines could survive the print run, they had been reinvented, independently, by men in three different cultures who had never read each other.

Khālid pulls back the dirham.

The horse is worth nothing.

I mistake one for the other.

Three sentences, three centuries, three continents. The writers who recorded them did not know each other's names. We know all three of theirs.