Chapter 4: A Joe Miller

The book on my desk is a small duodecimo, the size of a deck of cards bound in calf, and the title page reads, in the long-vowelled typography of Hanoverian London, JOE MILLER's JESTS: or, the Wits Vade-Mecum. Being a Collection of the most Brilliant Jests; the politest Repartees; the most Elegant Bons-Mots, and most pleasant short Stories in the English Language. The compiler is identified as one Elijah Jenkins, Esq., which is a pseudonym. The man behind it was a jobbing playwright named John Mottley, working under a polite scrim because he did not want to be associated with a joke book. The book ran through three editions in 1739, the year of its first printing; through more than seventy by the end of the century. The actor on the cover, Joseph Miller of Drury Lane, had died the previous August. He had also, by every contemporary account, been a mostly silent comedian — a famous deadpan, dependable for a slow take and a lifted eyebrow but not famous, in his own right, for his jokes. The book was selling jokes the man named on the cover was famous for not telling.

I do not want to make too much of any of this just yet. What I want to point at is a smaller and stranger fact, which is that I can quote it. Specifically, I can quote it in the version that a barrister in 1739 read aloud to his clerks, and that an Albany apothecary read silently in 1796, and that a Liverpool housewife consulted in 1842 to settle a dinner-table argument about a lady's age, and that a bookseller in Calcutta pulled off an English-language reprint shelf in 1903 to recommend to a young subaltern who wanted to know what was funny in his grandfather's day. Five readers, separated by oceans and a century-and-a-half, all of them looking at the same sentence. The same commas. The same italics. The same dash on the same line. They were doing something the readers in the previous chapter could not have done. They were holding the joke in its exact original wording.

That is a thing print does and manuscript could not. Manuscript copies drift. Every scribe lays his own hand on the line, abbreviates this and expands that, drops a clause when his eye skips, glosses a hard word into the margin, and sometimes inadvertently improves the joke and sometimes ruins it. Print stops the drift. Two readers in two cities holding two copies of the third edition hold the same sentence. The punchline acquires a verbatim form. The punchline becomes, in some cases, a quotation. The quotation becomes, in still other cases, a brand. The brand — Joe Miller — becomes English slang for a stale joke before the seventy-third edition is printed, because once you can quote a joke verbatim, you can also catch the speaker recycling it. That's a Joe Miller enters mid-eighteenth-century English idiom; Smollett uses the phrase in his novels of the 1750s and 1760s; Boswell uses it in the Life of Johnson; by the end of the century the phrase is so universally legible in English drama and prose that writers can drop it without footnote. It is the first time, so far as I can find in English, that a joke book sold so well that its title became a noun.

I would like to camp on that for one more moment, because it changes what comedy is. In the manuscript era a joke was a type — a shape with a recognisable structure, told differently at every retelling, recovered by the reader from a thicket of scribal noise. In the print era a joke is an artefact. A printer in T. Read's shop on Dogwell Court, off Fleet Street, sets the punchline once. The compositor's metal lays the comma where the comma goes. After that it is not the joke's kind that travels but the joke itself — the wording, the punctuation, the line break. A 1620s reader in Suzhou holds a printed Xiaofu whose deadpan line about King Yama's ghost-minions is identical, character for character, to the line in his neighbour's copy across the courtyard. A 1739 reader in London quotes Jest #99 to a friend at a dinner in Bath, and the friend looks it up at home and finds the same words. A 1865 reader in Brooklyn opens that week's Saturday Press and finds a paragraph about a frog full of quail-shot, and the same paragraph is being read by a clerk in San Francisco at the same hour. The print era is the first era in which the joke survives intact across distance. This chapter is about what that means.

I want to try, before going further, to be specific about the object the chapter is named after, because the object matters. The duodecimo on my desk is roughly four-and-a-half inches tall by two-and-a-half wide, bound in calf, with raised bands across the spine, and the price — one shilling — printed on the title page. The text was set in a single Roman face, leaded modestly, justified to a narrow column. The paper was rag — pre-pulp, pre-bleach, faintly grey when fresh. The ink was a lampblack-and-linseed-oil mix the apprentice mixed by the pot. Each sheet was folded twelve to a gathering — that is what duodecimo means — sewn through the fold, trimmed to size, and bound. The compositor laid each line of type by hand, picking individual letters out of a wooden case organised by frequency, and the punctuation marks were laid the same way: comma, period, semicolon, dash, the italic switch when Mottley wanted emphasis, the small caps for the lady's MADAM, the occasional ligature. Each line had to be read backward in the stick. Then a forme of pages had to be locked together, inked, paper laid on, and the press pulled. A skilled team on a wooden press could clear a few hundred impressions in an afternoon. The book existed because, in 1739, human hands picked individual pieces of metal out of a case and laid them in a frame and pulled a press over them. After that, the book existed everywhere.

It means, mostly, that we can finally check our work. Comedy in the previous chapters was an inferential game: we had records of jokes told differently in a hundred manuscripts and we worked out what shape they had been before they fell into ink. Comedy from 1500 onward is testable. I can hold Joe Miller #99 in 1739, and I can hold Joe Miller #99 in a 1791 reprint, and I can hold the same jest in a Boston knockoff of 1820 and a New York knockoff of 1858, and I can verify, on the page, that the wording does not move. When the wording does not move, what surfaces is what was actually funny. The structure is exposed. The mechanics are visible. And — this is where the thesis sharpens, because the same comedic engine I have been chasing through clay tablet and papyrus and parchment is still running here — the same mechanics turn out to be at work in late Ming Suzhou, in Hanoverian London, in Edo, in Madrid, in Hartford. Different presses. Different languages. Different alphabets. Different paper. Same engine.

The four people I would like to spend the rest of this chapter with are a Suzhou collector named Feng Menglong, a London playwright-turned-hack named John Mottley, an Edo senryū editor whose pen name is now the name of the verse form he edited, and a Hartford-based Missouri humorist who in his lifetime would not stop denying the things he had said. They sat in different rooms in different centuries. They had no knowledge of one another. They would not have understood each other's languages without translators. What they share is a printed page and an instinct for what to put on it. The instinct is the same.

Hades makes inquiry

I want to start in Suzhou, around the 1610s, with a man named Feng Menglong. He is in his mid-forties. He has failed the imperial civil-service examinations — which in the late Ming was a normal thing for a literate scholar to be doing in his mid-forties, since the higher levels of the exam were almost designed to fail people — and his career, by official measures, is going nowhere. He is doing other things instead. He is editing collections of vernacular short fiction. He is editing folk songs. He is editing operas. And, in roughly the years between 1610 and the early 1620s, he is editing a joke book.

The book's title is Xiaofu: 笑府, "Treasury of Laughs." Feng is not the inventor of the late-Ming printed jest collection — that genre had been working in Chinese letters for a century already — but he is its most ambitious practitioner. Xiaofu runs to thirteen chapters, organised by topic: clergy, doctors, misers, in-laws, false scholars, hen-pecked husbands, the rich, the poor, the humorous and miscellaneous. The translator and editor who has done the most for the modern English-language reader is Pi-ching Hsu, whose Feng Menglong's Treasury of Laughs (Brill, 2015) publishes most of the Xiaofu on facing pages with the seventeenth-century Chinese on the left and her English on the right. I have spent a lot of time with Hsu's edition. Several of the jokes are very good. One of them is the joke I want to walk through now, because it is the joke I have not been able to stop thinking about.

The joke is short, and the title — Hsu's translation of Feng's chapter heading — is Hades Seeks a Good Physician. Feng's Chinese is 冥王訪名醫, which Hsu's English captures cleanly: King Yama, ruler of the underworld, makes inquiry after a famous doctor. The setup is procedural. King Yama, surveying his realm, observes that he has been receiving a steady stream of unjustly killed patients from the world above — quack doctors are sending him dead men in numbers he cannot keep up with — and decides he would like to engage a reliably good physician for himself, against his own eventual need. He dispatches his ghost-minions to the world of the living to find one. The ghosts ask, sensibly, how they will know which of the world's doctors is good. The instruction is direct. A good doctor is one whose dead patients are not loitering at his front gate. The wronged ghosts of the dead always congregate at the door of the man who killed them. Look for the gate where there are no ghosts.

The ghosts go up to the world of the living. They walk the streets. The streets, it turns out, are crowded — doctor after doctor with a small dispirited crowd of his own dead piled up at the threshold like patients waiting for the receptionist to admit them. The ghosts walk a long time. Eventually they come to a door with no ghosts at all in front of it. They are overjoyed. They burst in. They greet the physician. They explain their commission from King Yama. They congratulate him on being, by clear forensic evidence, the only competent doctor in the city. Whereupon the physician explains, modestly, that the office only opened yesterday.

I have stayed with this story now for some weeks, and I have not been able to stop noticing how clean it is. The setup is deadpan procedural. The instruction is given with the tone of a manager briefing a delivery driver. The ghosts' walk is described in an even, businesslike voice — the streets were crowded, they walked a long time. The punchline is one sentence, dropped without flourish. The architecture is the architecture of a New Yorker cartoon, a Saturday Night Live cold open, a contemporary doctor joke told over a beer. The setup primes the reader to expect a discovery; the punchline reveals that the search criterion was, all along, doing the wrong work. The doctor's gate is empty of ghosts because the doctor has not had time to kill anyone yet. The reader has been laughing, all along, alongside ghosts whose interpretive frame was about to fall over.

What I like most about the joke is the tenderness of the dead in it. Feng does not mock the ghosts. He does not even mock King Yama, who in the Chinese popular religion of the period is a real and terrifying figure: the judge of the underworld, the dispatcher of the guǐchāi, those otherworldly bailiffs who come for the soul at death. The ghost-minions in the joke are diligent. They are professional. They are doing their job carefully, with the dignified persistence of men serving a writ. The crowd of wronged dead at the wrong doctors' doors are not played for laughs at their own expense. They are loitering, aggrieved, the aggrieved word doing all the moral work — the joke takes the wronged dead seriously because the joke would not work otherwise. The comedy lives entirely in the punchline, where the ghosts' hopeful arrival at the empty gate is undone by a fact none of the systems involved had thought to check. The joke is generous to its dead.

Feng was writing inside a long Chinese tradition of doctor jokes. There are doctor jokes in the Shishuo Xinyu, the New Account of Tales of the World, compiled around the year 430 CE. There are doctor jokes in the Tang collections. There are doctor jokes in the Song. The complaint that the average physician is more dangerous than the disease is older than printing in any language — the line this man's medicines have killed more than the wars is a stock literary move from late antiquity onward, in Chinese, in Greek, in Arabic, eventually in Latin, eventually in everything. What is new in Xiaofu is not the target. What is new is the form: a printed joke book, sold for entertainment, organised by topic, set in a regular type-block, bound and trimmed at a Suzhou bookseller's shop, available on the open market in editions that did not require copying. The jokes were a consumer product. Feng curated them, edited them, sometimes rewrote them, signed his preface, and sent the manuscript to the printer. The printer set them once.

The set-once part matters because we have, in Xiaofu, a cleanly comparable experiment. Within a generation of Feng's death his joke book was being lifted into the Xiaolin Guangji, an early-Qing compilation that became the dominant Chinese jest book of the eighteenth century. The lifts are not always exact — Xiaolin Guangji compresses, condenses, occasionally rephrases — but the set-pieces are recognisable on the page. Feng's Hades Seeks a Good Physician lands in Xiaolin Guangji in basically the same wording. A reader in 1750 with a Xiaolin Guangji on his desk and a reader in 1620 with a Xiaofu on his desk are, on the relevant page, looking at the same printed Chinese. If a kid in Hangzhou in 1750 laughs, and a kid in Suzhou in 1620 laughs, the laugh is the same laugh in front of the same line — the difference is only that the second kid was reading a slightly later printing.

I do not, of course, know that they laughed. I know that the joke was set, set again, copied into a successor collection, kept in Qing-era circulation for two centuries, eventually reprinted in Republican-era editions, eventually translated into European languages in the twentieth century and into Hsu's modern English in the twenty-first. I know that I laughed when I read Hsu's version. I know that a friend whose Chinese is fluent laughed when I asked him to read me Feng's original aloud, in the version that lives in his copy of the modern reprint. I know that the friend's six-year-old, who heard the punchline in Mandarin in the register her father uses for serious things, also laughed and then asked whether the dead doctors got back into bed and went to sleep. The form of her question is the form a child asks of a true story.

I want to take a small detour, while we are here, because Feng was working in a print era that was global. Chinese movable-type printing had been around in various forms since the eleventh century, woodblock printing for several centuries before that — Suzhou was, in Feng's lifetime, one of the great book-publishing cities of the late Ming — and the European print revolution Gutenberg kicked off in the 1450s was, by the early seventeenth century, producing books at a pace and a price that no manuscript culture could match. While Feng Menglong was assembling Xiaofu in Suzhou, a part-time tax collector and discharged soldier in his late fifties, named Miguel de Cervantes Saavedra, was watching a Madrid printer named Juan de la Cuesta finish setting the first part of a novel called El ingenioso hidalgo Don Quixote de la Mancha. The novel was published in Madrid in January 1605. It would be, within a generation, the most reprinted book of secular fiction in Europe — pirate editions in Lisbon and Valencia within months, an authorised second printing later in 1605, an English translation by Thomas Shelton in 1612, a French in 1614, an Italian in 1622, and German versions before Cervantes was a quarter-century in the ground.

Don Quixote and Sancho Panza have, by the eighth chapter of Part One, been on the road for a few days. Don Quixote is a country gentleman who has read too many chivalric romances and has decided to be a knight errant. Sancho is his neighbour, a farmer of practical disposition, who has been promised an island and signed on as squire. They cross the plain of Montiel. Don Quixote sees, in the distance, thirty or forty windmills. He does not see windmills. He sees giants — enormous monstrous giants, the destruction of whom will inaugurate his glory and Sancho's island. He announces that fortune is guiding their affairs better than they could have hoped. He spurs his horse. He fixes his lance.

Sancho, in Edith Grossman's English translation, points out what is plainly there. Look, your grace, he says: those things over there are not giants. They are windmills. The arms his master takes for the giants' arms are the sails. Cervantes's Castilian, in the 1605 first edition Juan de la Cuesta set in Madrid, runs no son gigantes, sino molinos de viento, y lo que en ellos parecen brazos son las aspas — they are not giants but windmills, and what look like arms are the sails. Don Quixote, undeterred, has prepared his answer. The answer is that Sancho, being unfamiliar with the rules of knight-errantry, does not understand. If Sancho is afraid, Sancho should withdraw and pray.

Don Quixote charges. The wind is up. The sails are turning. The lance hits the first sail dead-on. The sail takes the lance, the lance takes the man, the man takes the horse — and the entire assembly is lifted: knight, lance, horse, all of it carried into the air on the shoulder of an indifferent agricultural machine, then dropped onto the plain hard enough that Don Quixote cannot stand. Sancho arrives at full pace on his donkey to find his master flat on his back next to a horse that is also flat on its back, both of them having been hammered by a sail that has continued, after the brief interruption, to do the job it was built to do. I told you so, sir, says Sancho. Don Quixote, dusting himself off, explains that the giants must have been transformed into windmills mid-charge by the malign sage Frestón, who is jealous of his prospective glory.

I am going to come back to this. I want only to lodge it here. The shape — knight, lance, sail, lift, fall — is one of the load-bearing comedic shapes of the print era. Cervantes wrote it in Castilian Spanish in 1605. It is the structural skeleton of a great deal of the slapstick that survives. It is a body in the way of a mechanism, and the body losing. The mechanism wins because the mechanism is indifferent. The book held it for me on the page; the page is why I have it; the page is where it survives. We will see this shape again, at least three more times, before this book is over.

The lady's age

The smaller book is back on the desk in front of me. The duodecimo, JOE MILLER's JESTS, T. Read's first edition of 1739, three months out from press. The compiler is John Mottley working under the pseudonym Elijah Jenkins, Esq.; the dedicatee is the public; the cover trades on the recently dead Drury Lane comedian.

A word, briefly, on the dead comedian, because the question of who Joe Miller actually was bears on what the book did. Joseph Miller (ca. 1684–1738) was a Drury Lane stock-company actor, a low-comedian by trade, a regular in Restoration-era and early-Hanoverian comedies who specialised in the kind of secondary character who anchors a scene with a flat, suspicious look. He was, by the testimony of contemporaries who knew him personally, near-illiterate; the standard story is that he learned his lines from his wife reading them to him at home. He was not a famous wit. He was a famous face — the face that a Drury Lane audience trusted to register the fact of a joke being told without ever delivering the joke himself. He died in mid-August 1738, of a sudden illness, in the middle of a long career. He was buried at the parish ground of St Clement Danes, the Drury Lane parish church. His widow had been left with little. Mottley, who had play-doctored at Drury Lane for years, knew the family's situation; the early reception of Joe Miller's Jests understands the book partly as a benefit project, with proceeds from the first edition reaching Miller's family before the publisher's ownership of the manuscript took over and the gravy stopped flowing. By that point the brand had detached from the man. The man was already in the ground. The book is two hundred and forty-seven numbered jests fitted into the small format. The jokes are not Joe Miller's jokes. Some of them are scraped from earlier English jest-books — Mottley clearly had at his elbow a copy of A Banquet of Jests (1630) and probably one of the Cabinet of Mirth reprints. Some are translated out of French collections. Some are imported from Italian. Some are stage anecdotes Mottley picked up at Drury Lane during his playwriting years. Some are, in fact, jokes that were doing the rounds at coffee-houses in 1738 and that Mottley wrote down before they could drift any further. The book is a sweep, in the modern editorial sense — Mottley going through the available stock and assembling a representative collection, weighted toward the kinds of jests that fit the coffee-house register.

I want to spend most of this section on Jest #99, which is forty-seven words long including the heading.

The full text, from the 1739 first edition, runs:

A Lady's Age happening to be questioned, she affirmed, she was but Forty, and call'd upon a Gentleman that was in Company for his Opinion; Cousin, said she, do you believe I am in the Right, when I say I am but Forty? — I ought not to dispute it, Madam, reply'd he, for I have heard you say so these ten Years.

There are two ways into a joke this short. The first is to go at the mechanics. The second is to go at the company.

The mechanics are tidy. There is a speaker who has just made a statement. The speaker turns to a witness. The witness is — by the way the speaker frames the question — being asked to ratify the statement. The witness ratifies, in form, by promising not to dispute. The witness then pivots, in his very next clause, on the word for — which the witness has, in that single syllable, made into the most violent connective in the English language. I ought not to dispute it, Madam, for I have heard you say so these ten Years. The ratification has been a lever. The lever has been pulled. The fact ostensibly underwriting the statement undoes it. The lady is not forty. The lady has been forty for ten years.

I want to be precise about what is happening, because the move is older than Joe Miller and is going to come back, in this chapter and in the chapters after it, in a couple of different costumes. The move is the witness whose support undoes the speaker. It depends on a witness's deadpan compliance with the speaker's frame at the level of grammar, while the witness's actual content explodes the frame. The witness is not contradicting the speaker. The witness is agreeing. The witness is so agreeable that he is, in his agreeableness, supplying the long historical record that makes the speaker's statement structurally impossible. It is a compliance that is also a knife.

I came at this jest, the first time I read it, the way most people probably do — I laughed once, then re-read it to see how Mottley had set the trap, then laughed again at the way Mottley uses for. The little three-letter conjunction is doing the entire violence. The whole structure runs on that one syllable. Mottley sets it at the comma where the comma must go. The compositor at T. Read's shop in Dogwell Court, near White-Friars off Fleet Street, laid the comma in metal, the comma went into the forme, the forme went on the press, and the comma is in every reprint of Joe Miller I can find. The 1791 reprint has it. The 1799 American reprint has it. The 1830s Liverpool printing has it. I ought not to dispute it, Madam, for I have heard you say so these ten Years. You can read the joke aloud in 2026 and the comma is still where it was. This is the verbatim survival the chapter is named after.

But the jest is not an isolated specimen, and this is the place where I want to pull a thread. There is a joke in Philogelos, the fourth-century Greek joke book I spent the back half of Chapter 2 with — the dedication-page-of-Western-comedy joke book that Hierocles and Philagrius compiled around the year 400 CE. The joke is number eighteen in the manuscript order. A man comes back to the slave-seller and complains that the slave he has just bought is dead. The slave-seller — and this is Berg's English, but Baldwin's reading is structurally the same — draws himself up and replies: By the gods! When he was with me, he never did anything like that.

I want to point at this for a second, because what is happening at the slave-seller's stall in fourth-century Greek and what is happening in a Hanoverian London drawing-room in 1739 is the same comedic manoeuvre. The slave-seller, accused of having sold a defective product, defends himself with a clause-level grammatical compliance — yes, of course, when he was with me — that supplies the very evidence that makes his defence absurd. The slave never died on his watch. Death is a behaviour the slave personally took up after the sale. The seller's defence, like the witness's reply at the lady's drawing-room, is a ratification that is also a knife. The mechanism is identical.

I want to be specific about how identical. In each case the speaker is in trouble. In each case the speaker turns to an authority — the lady to her cousin, the slave-seller to the gods — and asks, with grammatical innocence, for support. In each case the support arrives, in the form of a clause that is ostensibly compliant. In each case the support contains, in a tiny piece of historical or referential information, the entire load that destroys the original statement. I have heard you say so these ten Years. When he was with me, he never did anything like that. The lever is the same. The pivot word is for in 1739 English, gar (for, because) in the Greek of Philogelos. The two-letter conjunction is doing the same work in two languages thirteen-and-a-half centuries apart. I do not think the Philogelos slave-seller travelled to Hanoverian London by direct transmission. I suspect Mottley had not read Philogelos; in 1739 Hierocles and Philagrius were still mostly in manuscript and would not be properly edited and printed until well into the next century. What I think is that the manoeuvre — the deadpan compliance that delivers a knife — is one of the half-dozen recurring shapes the print era inherited and then refined. Philogelos found it. Mottley found it. They found it because it was findable. It is sitting in the structure of the move, waiting.

Let me say more about why it lands. The reason the slave-seller is funny, and the reason the witness in Joe Miller #99 is funny, is the register. Both speakers are calm. Both are formally unimpeachable. Both are giving you, on the surface, a reasonable line of agreement. Neither breaks the deadpan. The slave-seller is indignant — this is critical — but his indignation is an indignation aimed at the wrong target. He is offended that he is being held responsible for a slave who personally chose to die. The witness in 1739 is courteous — also critical — but his courtesy is the courtesy of a man so polite he cannot bear to dispute a lady, which is why he is the most dangerous person in the room. The deadpan voice does not understand what is funny about what it is saying. The reader does. The recognition is the laugh. Both jokes refuse, structurally, to register what is plainly happening, and the refusal is the joke.

I will keep coming back to this. The deadpan-narrator move is one of the load-bearing inventions of the comic engine, and it shows up in every era this book covers, in every medium this book covers, with a regularity that has stopped surprising me and started arguing for the thesis on its own. We will see it again before the end of this chapter, in 1865, in California, in a frog. We will see it later in a British pet shop with a customer and an evasive parrot. We will see it on the internet, in 2013, in a webcomic of a dog drinking coffee in a kitchen on fire. The slave-seller's voice is in all of them. So is the lady's cousin's. The voice is older than print, survives print, survives broadcast, survives the algorithm. The voice is, so far as I can tell, what the comedic mind sounds like to itself when it knows the joke is happening to someone other than the speaker.

Now — about Mottley.

Mottley was not a great writer. He wrote two unsuccessful tragedies — The Imperial Captives in 1720 and Antiochus in 1721 — neither of which had the legs to outlast their initial run. He continued, after Whincop, the early reference work A List of all the Dramatic Authors, which appeared in 1747. He worked, mostly, as a hack: paid by the sheet, paid for translations, paid for pamphlets, paid by Drury Lane managers for play-doctoring, paid for whatever brought him through the next month. Joe Miller's Jests was, on Mottley's part, an editorial sweep done quickly. The 1739 title page promises the most Brilliant Jests; the politest Repartees; the most Elegant Bons-Mots, and most pleasant short Stories in the English Language — the marketer's voice doing the work the marketer is paid to do. The implicit claim is that what is in the book is what Joe Miller would have laughed at, which is a different claim from the claim that Joe Miller said any of it. Mottley does not, anywhere in the book, attribute a specific jest to Miller. The cover does the attribution; the jests, inside, are a coffee-house anthology with a Drury Lane portrait pasted on the front. He did not list himself on the title page — Elijah Jenkins, Esq. is the supplied authorial fiction — and he was not, in his lifetime, paid royalties on the book that made him bibliographically immortal, since he sold the manuscript to T. Read for a flat fee. By the time it was clear how much money the book was making, somebody else was making it. The book, in the form that mattered, became cultural property within a year of its publication. That's a Joe Miller enters the language. Smollett uses the phrase. Boswell uses the phrase. By the end of the century an English novelist can drop a Joe Miller into a paragraph the way a modern essayist drops in a cliché — readers will catch it. The book has done a thing the manuscript era could not do: it has converted a joke book's title into a unit of currency.

Now I want to come back to the windmill, because the windmill is what the print era's verbatim survival makes structurally available. In 1605, in Madrid, Cervantes set in metal the scene I described above. The scene has a knight, a lance, a sail, a lift, and a fall. The knight is undone by a piece of agricultural infrastructure that does not know him and would not respect him if it did. The shape — body in the way of a mechanism — is verbatim available to a reader in 1605 in Madrid, a reader in 1612 in London reading Shelton's English, a reader in 1622 in Florence reading Franciosini's Italian, a reader in any later century in any later language. The sail's arc is the same arc. The fall is the same fall. The body is the same body. The lance is the same lance. Cervantes set it, and it stayed set, and the eye of the reader, whenever the page is opened, watches the same lift. There are not many things in human cultural production that work this way. The print era found a thing that does.

The Joe Miller witness is keeping company with the windmill scene at a level the reader is meant to feel. Both turn on a deadpan figure landing a small weight of fact in the right place at the right comma. Both are printed in the seventeenth or early eighteenth century at the head of European publishing. Both have stayed where they were set. The witness is in a drawing-room, not on a plain in La Mancha, and the comeuppance is social rather than physical. But the page is the same kind of page. The metal type is the same kind of metal type. And the joke, on the page, has not moved.

Seventeen syllables

I want to step out of English now and into seventeen-syllable Japanese, because the print era was global and the same year that Mottley was sweeping coffee-house jests for T. Read, a parallel print apparatus on the other side of the world was selling small books of comic verse to townspeople in Edo. The form is senryū. It is, by syllable count, a haiku. It uses the same 5-7-5 structure. It is generally written in two-clause logic, with the kireji — the cutting word — between the second and third lines. What it is not is the natural-world contemplation of the haiku Bashō and his school had established in the previous century. Senryū drops the cherry-blossom and turns the camera on people. Specifically: people doing the things people do.

The form takes its name from the man who first edited it for print. Karai Hachiemon, of Asakusa Shinmonchō in Edo, was a nanushi — a ward headman, a kind of municipal middle manager — who in his off-hours ran a literary parlour game called maekuzuke, cap-verse capping. The game worked like this: a poet supplies a 7-7 second couplet, and players supply a 5-7-5 first verse that links to the cap. The capping was popular. Karai held judging sessions, the verses were read aloud, the best were retained. Karai's pen name was Senryū, "river willow." Beginning in 1765, in collaboration with the printer Goryōken Arubeshi, Karai published the first volume of an anthology drawn from his judgments: Haifū Yanagidaru — Yanagidaru the willow-cask, Haifū the poetic style. The book sold. The series ran, edited by Karai during his life and continued by his successors after his death in 1790, into the 1840s — twenty-four volumes, each carrying hundreds of compressed comic vignettes about the floating world, the back streets, the in-laws, the merchant houses, the priests, the unfaithful husbands, the unwise sons, the brothel-quarter, the long-suffering wives. By the time the series stopped, senryū had become a noun. The form had taken its editor's pen name.

The verse I want to walk through is the textbook example of the form, included in the first volume in 1765, and one of the verses Makoto Ueda places near the front of his English-language Light Verse from the Floating World. The text in romaji:

dorobō o toraete mireba waga ko nari

— where dorobō is thief, toraete mireba is having caught (him), upon looking, and waga ko nari is (he) is my own son. Ueda's English on the facing page: the thief / I caught and looked at — / my own son. R. H. Blyth, in the 1949 edition that introduced the form to most English-language readers, glosses it as the thief I caught was my own son.

The verse is doing a quantity of work that I want to slow down to examine. There is a chase. There is a capture. There is a recognition. There is the deflation. The whole comic arc — pursuit, struggle, hand on the collar, drag into the lamplight, look at the face — fits into seventeen syllables. Senryū's editorial discipline supplied the cut, the kireji, exactly between toraete mireba and waga ko nari. The cut is the punchline. The reader's eye crosses the white space and the kireji and emerges on the other side with the pivot already accomplished.

I have read this verse alongside readers whose Japanese is fluent and readers whose Japanese is none. The verse lands either way. The reader who has the romaji in front of him laughs at waga ko nari — at the way the closing copula completes the trap. The reader who has only the English laughs at my own son. The structure is the carrier. The structure is so compact that it survives translation in the way a folded paper crane survives travel — the paper is different, the figure on the table is the same.

What surprised me on first encounter is what print did to senryū. The form had existed orally — in maekuzuke parlour games — for a least a half-century before Karai started anthologising. It was, before 1765, a parlour activity. It became, after 1765, a publication. People wrote senryū at home and submitted them to the editors. The form acquired editorial conventions. Editors like Karai — and after him, his son and the syndicates of senryūshi who took over the press — developed an aesthetic. The published senryū became more compressed, more economical, more pointed, than the parlour-game senryū that fed them. The form tightened under the pressure of editorial selection. Haifū Yanagidaru's editors kept some, dropped most. The keepers became models. The models became conventions. The conventions became, by the early nineteenth century, a genre. Print was the genre's habitat.

I want to point at one more thing about the thief verse, which is what kind of joke it is. The verse is not a verbal joke. It does not depend on a pun, a homophone, or a culturally specific reference. The recognition shock is structural. A homeowner hears something. The homeowner gives chase. The homeowner is in a particular kind of mental state — alert, adrenalised, righteously aggrieved, the mind running ahead of the body and the body running ahead of the mind. The homeowner gets a hand on the intruder. The homeowner pulls the intruder into the lamplight. The homeowner looks at the intruder's face. The face is his son's. Everything that happens after that — the suspended fist, the recalibration, the older life-question of what has this kid been doing and the smaller life-question of do I tell his mother — is not in the verse. The verse stops at the look.

The reason I think the verse travels is that the experience of the look is universal, and the verse is short enough that there is no other content to get in the way of the reader's recognition. I caught the thief. I looked at him. He is my child. The compression is doing the work the deadpan-narrator move does in Philogelos and in Joe Miller — a refusal to expand into the emotional consequence of what has just happened. The narrator does not laugh first. The narrator does not say what a turn. The narrator does not psychologise. The narrator just lays the three lines down and stops at the copula. The reader is left to do everything else.

The other thing the verse does, which I want to flag, is that it preserves the print era's working ratio. Karai, like Mottley, was an editor working with submitted material. He was running a contest — a parlour-game competition, with prizes — and his judgement determined what got into print. He was, in effect, a comedic curator. By 1765 the senryū form was a printed market commodity, sold by the volume, with named editors taking credit for selection rather than for original composition. Haifū Yanagidaru's title page lists Karai Senryū as the senja — the one-who-selects, the chooser — not the author. The verses inside are by hundreds of submitting hands. Karai's job is to say which. This is the job Mottley does for Joe Miller's Jests. This is the job Feng Menglong does for Xiaofu. The print era invented a profession that could only exist in print: the editorial curator of jokes. The thief verse is not Karai's. The thief verse is in a Karai book because he chose it.

There is something I want to flag about the rest of Haifū Yanagidaru before I leave the form, because the thief verse is not an outlier. The series runs to twenty-four volumes, and the territory the verses cover is the small, embarrassing, recognisable rooms of Edo life: the unfaithful husband, the long-suffering wife, the disappointed in-laws, the unwise son, the merchant who pretends not to know what his apprentice is up to, the priest with too much wine in him, the brothel-quarter regular caught by the new rules. The recurring move is a small recognisable embarrassment, set down in the smallest formal container the language has, with the punchline at the kireji and not a syllable wasted on it. None of those subjects depend on a pun. None depend on a culturally specific reference that does not also have an analogue in any other settled human society. They are jokes about people, of the kinds people are. They were sold by the volume. People bought them by the volume. The form's commercial life — the volumes piling up on the shelves of Edo booksellers from 1765 into the 1840s — is itself an argument: the same shapes of human embarrassment that work in any room work in print, in seventeen syllables, sold for money.

What we can say with confidence is that the thief verse was being read in 1765 in Edo, and is being read in 2026 in places that did not exist in 1765, and that on first encounter — without scholarly briefing, without paratext, without historical context — the verse lands. The recognition shock is the same shock. The look is the same look. The cousin's verdict in Joe Miller and the homeowner's recognition in Yanagidaru are not the same joke, but they were keepable on the same kind of page in the same century, on opposite sides of the world, by editors who did not know that the other was working.

A double handful of shot

I want to land this chapter a little further west, and a hundred years on, with a paragraph from a New York newspaper that is a piece of evidence a working author cannot anticipate.

In November of 1865 a struggling humorist named Samuel Clemens — broke, in his second year of trying to make a living from print, late of Nevada and Hawaii, currently in San Francisco, writing under the pen name Mark Twain — sent a story to the editor of the New York Saturday Press. The story was already late. Artemus Ward, the established humorist and Clemens's friend, had asked for a contribution to a forthcoming book Ward was editing, and Clemens had been working on a piece about a jumping frog all autumn, drafting and redrafting, frustrated with the cadence. The story missed the book deadline. Ward forwarded it to the Saturday Press, where the editor, Henry Clapp Jr., ran it on November 18, 1865, in the Press's Saturday issue, under the title Jim Smiley and His Jumping Frog. It made Clemens, in the New York newspaper world, a sensation by the following Saturday. Newspapers from Boston to Sacramento were copying it. By the time Clemens's first book came out, in 1867 — The Celebrated Jumping Frog of Calaveras County, and Other Sketches — the title piece was so famous that the book was sold under its name. Clemens spent the rest of his life mildly resenting the frog and being unable to escape it.

The story is not the frog. The story is the narrator. There are two narrators, in fact. The outer narrator is a polite Eastern stranger — Twain himself, lightly fictionalised — who has been asked, by an unnamed friend, to call on a man named Simon Wheeler at a tavern in Angel's Camp, California, and ask Wheeler about a man named Leonidas W. Smiley. The friend has confided that asking will produce useful information. The outer narrator does as he is asked. Wheeler corners him at the tavern stove, blockades his exit with a chair, and begins to tell him, in a level voice and at deadpan length, the story not of Leonidas W. Smiley but of Jim Smiley — a compulsive gambler — and Jim Smiley's prize jumping frog, Daniel Webster. By the second paragraph of Wheeler's monologue the outer narrator has realised that his friend has played a trick on him: there is no Leonidas W. Smiley. The friend wanted the outer narrator pinned to a chair while Wheeler talked about Jim. By that point Wheeler is past the introduction and into the frog. The outer narrator cannot extricate himself.

The story is Wheeler's voice. Wheeler tells, in a slow, even, ungrammatical, unembellished register — his sentences run long, his transitions are anyway and well, thishyer and but he was kind of fond of him — the comprehensive history of Jim Smiley's gambling. Smiley would bet on anything. Horses. Dogfights. Catfights. The parson's wife's recovery from a sickness, on which Smiley, when the parson reported that with the blessing of Prov'dence she'd get well yet, said before he thought, Well, I'll resk two-and-a-half she don't anyway. Wheeler's voice is the central comic instrument. Wheeler's voice does not break. It is a very long voice, and inside it Twain has fitted a story like a ship in a bottle.

Eventually Wheeler arrives at the frog. Smiley caught a frog in the swamp. He named the frog Daniel Webster. He educated the frog in jumping for three months — turned him out in the back yard with a piece of cloth on the ground for a launch pad, and worked with him patiently, until the frog could outjump anything of his breed you ever see. Smiley used to keep the frog in a small lattice box, and would carry it down to the tavern, and would set the box on the bar, and would offer to bet anyone in the place, on the spot, that Daniel Webster could outjump any other frog in Calaveras County. Smiley was not bluffing. Smiley had spent three months on the frog. The frog was a champion.

One day a stranger came into camp. The stranger looked at Smiley's box. What might it be that you've got in the box? Smiley says, sort of indifferent, that it might be a parrot, or it might be a canary, but it ain't — it's only just a frog. The stranger asks to see the frog. Smiley hands him the box. The stranger looks at the frog a long time. The stranger gives, deadpan, his assessment: I don't see no p'ints about that frog that's any better'n any other frog.

I am going to slow down here. The line is forty syllables that have done as much work in American letters as any forty syllables in the literature. I don't see no p'ints about that frog that's any better'n any other frog. The stranger, on first inspection, supplies the comedic key. He is not impressed. He is not, on the evidence of his own eyes, going to be impressed. He is going to maintain — with the same deadpan voice the slave-seller used at his stall in 4th-century Greek, with the same compliance the cousin used in the lady's drawing-room in 1739 — that nothing remarkable is happening to him. The stranger has never seen a frog this special. The stranger is going to keep saying so.

Smiley takes the bait. Smiley has a champion frog. Smiley likes betting. Smiley says he will wager forty dollars that the frog is better than any other frog. The stranger, sad and reflective, says he is a stranger to the camp and ain't got no frog. But, says the stranger, if I had a frog I'd bet you. Smiley, immediate, replies: That's all right — that's all right — if you'll hold my box a minute, I'll go and get you a frog. Smiley is gone. The stranger is sitting at the bar with Smiley's box and forty dollars at stake. The frog Smiley has trained for three months is in the box. Smiley is at the swamp.

I want to handle this part with care, because what happens next is the Philogelos slave-seller in 1865 American clothes.

The stranger sits with the box. The stranger thinks for a moment. The stranger gets out a teaspoon. The stranger pries open the frog's mouth. The stranger fills the frog with quail-shot — filled him pretty near up to his chin, in Wheeler's voice — and sets him carefully back down on the floor of the box. Twain's prose is almost dispassionate about the act. The teaspoon. The quail-shot. The careful dosing. The narrator does not editorialise. The narrator stays in Wheeler's voice. The reader is being shown, with the dignity of a procedural manual, the loading of small bird's-bore lead into a stationary live amphibian.

Smiley returns from the swamp with a competing frog. Smiley sets up the contest. The two frogs are placed even on the floor. Smiley counts down. One — two — three — git! and he and the stranger touched their frogs from behind. The stranger's new frog hops off briskly. Daniel Webster — the three-month-trained champion — couldn't budge; he was planted as solid as a church, Wheeler says, and he couldn't no more stir than if he was anchored out. Smiley is dumbfounded. Smiley does not understand. Smiley has never seen Daniel Webster fail. Smiley's frog has done his three-month performance for three months without missing a beat. The frog now sits, immobile, while the stranger's swamp-caught frog clears the room.

The stranger collects the forty dollars. The stranger walks toward the door. The stranger gives, as he leaves, his deadpan exit verdict: Well, I don't see no p'ints about that frog that's any better'n any other frog. He has repeated himself. The line is the same line he gave on inspection. The stranger's voice has not moved. The stranger's voice will not move. The stranger will not register, on his face or in his speech, the existence of the operation he has just performed with the teaspoon and the lead. He shuts the door.

Smiley is left with his frog. Smiley is bewildered. Smiley scratches his head. Smiley stoops down and lifts the frog: Why blame my cats if he don't weigh five pound! Smiley has the frog by the loose skin around the middle. The frog does not feel right. The frog feels, by Smiley's hand, much heavier than the frog he trained for three months. The frog's loose skin is full of something. Smiley turns the frog over.

The frog belched out a double handful of shot.

I am going to stop and let that sentence sit, because it is a sentence, and because I have read it many times now and on each reading the sentence does the same work. The frog belched out a double handful of shot. It is two clauses. Subject-verb-object. The pivot is the verb belched. The verb is doing in 1865 what for did in 1739 and what gar did in 400 CE. The verb is the lever. The reader has been waiting, structurally, for a release of the loaded frog's content; the sentence releases it. The release is not a was found to contain. The release is not a was discovered to be filled with. The release is belched, which is the verb the loaded frog has earned. A double handful of shot is the quantity the teaspoon has been measuring. Twain has accumulated all of the necessary stock — the frog, the box, the stranger's teaspoon, the quail-shot, the contest, the immobility, Smiley's bewilderment, the heft, the loose skin — and the sentence collects the principal at the end with the precision of a banker closing books at six o'clock. The sentence is the close of the joke. It is also, by my count, the longest-fuse joke in American English up to that decade.

Smiley, for his part, lands the verbal closer with the lift of a slow take. He see how it was, and he was the maddest man — he set the frog down and took out after that feller, but he never ketched him. And then, just before Wheeler's voice releases the outer narrator from the chair and lets him stand: If he ain't full of shot, I'll be cuss'd. The conditional is the Philogelos slave-seller's grammar. If he ain't full of shot, says Smiley — like the slave-seller in fourth-century Greek saying when he was with me, he never did anything like that. Smiley's voice has caught up to what the reader has been seeing for the past two paragraphs. Smiley is closing his hand around the obvious. The narrator does not break. Wheeler tells this in the same level cadence he has been using all afternoon. The outer narrator — Twain — has not interrupted Wheeler in any of this. The outer narrator's complaint that he has been trapped at the tavern stove against his will is, by this point, the reader's complaint, except the reader is loving it.

I said I would land the chapter here, and I want to say what I am landing. The stranger is the slave-seller. He is the slave-seller from Chapter 2 of this book, in Greek, around the year 400, holding the same line in a different costume. The teaspoon is the operation the slave-seller will not register having performed. I don't see no p'ints about that frog that's any better'n any other frog is when he was with me, he never did anything like that. The stranger refuses to acknowledge what he himself has just done. The voice does not break. The voice does not even concede. Twain wrote it in 1865. Mottley printed it, in a very different costume, in 1739. Hierocles or Philagrius copied it, in another costume again, around the year 400. The stranger's deadpan — bored, courteous, untouched — is a voice the reader has been hearing across this book in three languages and twenty centuries, and on each appearance the voice is doing the same job: refusing to register, in its own grammar, the obvious thing happening in the room. The reader does the registering. The reader's registering is the laugh.

There is something I want to add about Twain's instinct here, because Twain himself was the most articulate analyst of the move he had just executed. In October 1895, thirty years after the Saturday Press paragraph went out, Twain sat down at his desk and wrote, for The Youth's Companion, an essay called How to Tell a Story. The essay is the working manual. There are, Twain says, three kinds of stories: the comic story, the witty story, and the humorous story. The comic story is told with a flourish; the teller laughs first and points at the joke. The humorous story is American, the comic story is English, the witty story is French. The humorous story is the deadpan. The teller of the humorous story does not betray, in voice or face, that there is anything funny in what he is saying. The teller maintains the level register. The teller will, if anything, throw in a small confused pause, an unnecessary detail, a sigh — an oblique gesture in the direction of bewilderment about why his audience is laughing. The teller will, if at all possible, fail to land the punchline. The teller will bury the punchline somewhere midway through the next paragraph, where the listener has to dig it out. The humorous story may be spun out to great length, and may wander around as much as it pleases, and arrive nowhere in particular. Wheeler is the doctrine. The frog is the doctrine. I don't see no p'ints about that frog that's any better'n any other frog is the teller refusing to land it.

There is one further thing I want to mention about the publication, because it bears on the chapter's argument. Jim Smiley and His Jumping Frog went out in the Saturday Press on 18 November 1865. Within weeks the story was being lifted, in whole or in summary, by other newspapers — the San Francisco Californian reprinted it before the year was out, New England and mid-Atlantic papers picked it up in late 1865 and through the winter and spring of 1866, and by the time C. H. Webb collected it as the title piece of Clemens's first book in April 1867, the story was already broadly known. The spread is the print-era machinery in action. A piece set in metal in one New York shop on a Wednesday afternoon was, within weeks, available in the same wording to readers in Boston, in Hartford, in Chicago, in St. Louis, in San Francisco, in towns whose newspapers had cooperative reprinting arrangements with the Saturday Press and in towns whose editors simply read the Press and lifted what they liked. The verb belched was in the same place every time. Twain did not have to be in any of those rooms. Twain did not have to know any of those readers. The verb did the travelling. The page held it.

Twain knew, in 1895, exactly what he had built in 1865. He had built the deadpan, in print, in American English, frozen on the page of a New York Saturday newspaper, in a register that survives. The frog belched out a double handful of shot is one of the sentences I would point at if a reader asked me what print does. Print does this. Print holds the line at the speed it works. A reader in San Francisco in November 1865 reads belched and laughs. A reader in Cincinnati in March 1866 reads belched and laughs. A reader in Hartford in 1903 reads belched and laughs. A reader on a phone in a kitchen in 2026 reads belched and laughs. The verb is in the same place each time. The cadence is in the same place each time. The pause before if he ain't full of shot, I'll be cuss'd is on the same printed comma that Henry Clapp Jr.'s compositor laid down in November 1865. The reader, every time, gets the joke at the speed Twain set it.

I want to add one observation about the stranger before I close. The stranger has no name. He is in the story for two pages. He performs the operation on the frog with a teaspoon. He pockets the forty dollars. He walks out. He never returns. Wheeler — Twain's narrator-of-the-narrator — never tries to identify him. The stranger is a craftsman. The stranger is the travelling instance of the deadpan. He shows up in town, he sees a champion frog, he loads the champion frog with quail-shot, he collects, and he leaves. His exit line is the same as his arrival line. His face does not change. His voice does not change. He is, for the duration of his appearance, a perfectly anonymous specimen of a comedic stance the print era was learning, around 1865, to recognise. He is the slave-seller. He is the cousin in the drawing-room. He is, when this book reaches its broadcast era, going to be every customer at every counter who calmly insists his parrot is just resting. The stranger has been in the room, in different costumes, for at least 1,500 years. Twain put him in 1865 California in a hat and walked him out the door of a tavern in Angel's Camp.

I would like to leave you with him. Smiley is taking out after him; Smiley will never catch him; Smiley will return to the tavern eventually, kick the dust off his boot, wonder how on God's earth he was beaten by a frog he trained for three months. The frog will eventually be retired. The teaspoon's owner will be on a stage-coach to Stockton. The forty dollars will be in his coat pocket. The line of his coat will be slightly heavier on the right side from the small leftover quantity of birdshot the stranger has not, in the end, used. He will, when he reaches Stockton, find another stranger with another box, and he will ask, deadpan, what the man has in there.

He has been working, in different rooms, for fifteen centuries. He will work for as long as anyone is putting words on a page that another reader, in another century, can turn over in his hand and read.