The Uses of Laughter

I · The Man in the Mall

In the early nineteen-nineties, the neuroscientist Robert Provine went looking for laughter where it actually lives. Not in the laboratory, where the phenomenon proves shy, but in shopping malls, food courts, and the student union—anywhere people talked and could be observed without noticing it—where, notebook out, he logged twelve hundred episodes of spontaneous laughter and wrote down, each time, what had been said the instant before. The expectation, his and everyone’s, was that they were collecting jokes. They were not. By Provine’s count, only ten to fifteen per cent of the remarks that set off laughter were even remotely humorous. The typical triggers were sentences like “Where have you been?” and “It was nice meeting you, too”—lines no comedy club would book. Laughter, his transcripts insisted, is only occasionally applause for wit. Mostly it is something else.

The something else has structure. Laughter almost never interrupts: in over ninety-nine per cent of the twelve hundred episodes it arrived at the boundaries of phrases, punctuating speech the way a period punctuates a sentence—Provine called it the punctuation effect. Speakers laughed forty-six per cent more than their audiences, which is backwards if laughter is a review and exactly right if it is a signal. And in a companion study that had students keep laughter diaries for a week, laughter turned out to be thirty times more frequent in company than alone, once television and radio were set aside. Provine’s conclusion was the modest, radical one this essay takes as its premise: laughter is a social instrument first and a response to comedy a distant second. It is the sound of a relationship working—or being worked on.

That premise changes the assignment. If laughter is mostly relational, then the question “what is humor good for?” is really a question about relationships—and it can be audited, arena by arena. What does the joke buy at the office, where organizational research has priced it? At home, where marriage laboratories have filmed it? Among friends, where it appears to be doing chemistry? In courtship, where it functions as a signal read differently by the two people on the date? And in the body, where the claims are oldest and, it turns out, weakest? Behind all of it sit two harder questions. Whether men and women can play the same instrument on the same terms—the evidence says the instrument is shared and the terms are not. And what New York’s public stages reveal about who gets permission to use it.

A warning is owed before the dissection. E. B. White, introducing the anthology of American humor he and Katharine S. White edited in 1941, put it in the form of a caution: “Humor can be dissected, as a frog can, but the thing dies in the process and the innards are discouraging to any but the pure scientific mind.” The frog will not survive this essay either. The defense is that the innards have turned out to be genuinely interesting—and that knowing how the frog works has consequences, at the office and elsewhere, that the squeamish forgo at their own expense.

II · The Four Instruments

The science’s first useful gift is a distinction. In 2003 the psychologist Rod Martin, with Patricia Puhlik-Doris and colleagues, published the Humor Styles Questionnaire, the instrument that organized the modern field. It sorts habitual humor along two axes—whom the joke serves, and whether it is benign or barbed—yielding four styles. Affiliative humor is joking to warm a room: teasing gently, telling stories, easing tension. Self-enhancing humor is the private version—a comic lens kept trained on one’s own troubles, a way of not being crushed by Tuesday. Aggressive humor is wit with a target: sarcasm, ridicule, the put-down. And self-defeating humor is the habit of buying a room’s affection at one’s own expense—playing the fool, laughing along with one’s own denigration, amusing others as a form of appeasement.

The styles were built to predict well-being, and they do, in the directions the names suggest. A 2018 meta-analysis pooling forty-five samples and 12,734 people found the first two styles positively correlated with mental health—self-esteem, life satisfaction, optimism, less depression—at small-to-medium effect sizes; the self-defeating style ran the other way; and aggressive humor was, overall, simply unrelated to the joker’s own mental health, whatever it does to the target’s. The same analysis noted that the pattern shifts with geography and, for some associations, with sex—a small early warning that the rest of this essay will keep sounding: in humor, the material is universal and the terms are local.

One clarification matters enormously for what follows, because the loose phrase “a sense of humor” conceals the field’s central finding: kind is most of the story. And within kind, a distinction the questionnaire itself enforces: a self-deprecating joke is not a self-defeating style. The style is a habit—a standing offer to be the room’s punching bag—and it tracks poorer mental health wherever it is measured. The single self-mocking line, delivered from secure footing, is a different act altogether; Parts III and VII will price it, and the price turns out to depend on exactly how secure the footing is.

As for what a joke is, the leading one-line answer belongs to the behavioral scientists A. Peter McGraw and Caleb Warren: amusement arises from a violation—of logic, norm, or dignity—that is simultaneously appraised as benign. A joke is a small transgression wearing a safety certificate. The definition explains more than it seems to: nearly everything that follows, from the failed office joke to the Calvinist theater ban, is a dispute about who may issue the safety certificate, and for whom. One caveat before the tour begins: most of this literature is correlational, and cheerful people may simply have easier lives. Where an experiment or an intervention licenses a stronger verb, the essay will say so; elsewhere, “tracks” means tracks.

III · The Office

The cleanest demonstration of what a joke does at work involved a promotional testimonial. In the first two of eight experiments published in 2017, the management researchers T. Bradford Bitterly, Alison Wood Brooks, and Maurice Schweitzer had participants write promotional copy while a confederate delivered either a straight testimonial or one ending in a joke. In the first study the product was a pet-waste-removal service, and the joker was rated higher in status—5.03 against 4.43 on a seven-point scale. In the second, participants allocating “leadership points” for a group task gave the joker 6.66 to the straight speaker’s 4.85. A joke converted into command. The setting was experimental, not a New York office; the mechanism is what this essay brings back to the city.

The mechanism, the authors showed, is a two-part signal. Telling any joke broadcasts confidence—it announces that one has read the room and expects to win it—and confidence is credited toward status whether or not the joke lands. Competence, though, is credited only on success. A job candidate in one of their scenarios, asked where he saw himself in five years, answered, “Celebrating the fifth year anniversary of you asking me this question”; when the interviewer laughed, the candidate’s standing rose above the serious answer’s, and when the interviewer did not, it landed no higher than if he had answered straight. The joke that merely died, in other words, was free. The ruinous one was the inappropriate joke—“That’s what she said!”—which sank the teller’s status below the serious baseline even when it got a laugh: confidence up, competence down, net loss. The lesson is not that joking is risky but that the risk sits somewhere other than where nervous people put it. Falling flat is cheap; misjudging the room is not, and the laugh does not redeem it. The bet is also nearly universal. In the same team’s pilot survey of two hundred employed Americans, exactly one could not recall a coworker’s joke.

Around this sharp experimental core sits a wide correlational literature, and it is friendly to the joke. A 2012 meta-analysis by Jessica Mesmer-Magnus, David Glew, and Chockalingam Viswesvaran—forty-nine studies, 8,532 workers—found employee positive humor associated with less burnout and stress (corrected correlations of −0.23 and −0.25), and a supervisor’s humor associated with subordinates’ job satisfaction (0.39, from five studies) and with how well subordinates thought the boss was doing her job (0.45). Several of the cells in that table rest on three or four studies apiece, and the authors say plainly that such estimates should be read with caution; the essay passes the caution along. Still, the direction is consistent: rooms with benign humor in them are rooms people less often want to leave.

The complications are two. First, the leader’s joke deregulates. In three-wave field studies in China and the United States, Kai Chi Yam and colleagues found that a leader’s sense of humor raised engagement through warmer leader–member relationships—and simultaneously signaled that norms are negotiable, which predicted follower misbehavior. When the leader’s humor ran aggressive, the arrangement curdled: the misbehavior path strengthened and the engagement path disappeared. The joke, being a licensed violation, licenses violations. Second, the barbed instruments are specialists’ tools. Sarcasm, in experiments by Li Huang, Francesca Gino, and Adam Galinsky, roughly doubled creative problem-solving for both speaker and target—abstraction is apparently contagious—but it raised conflict except between parties who already trusted each other, and the finding remains a single team’s, unreplicated. Meanwhile the gentlest instrument compounds quietly: opening an e-mail negotiation with something funny, in a study of 122 executive MBA students by Terri Kurtzberg and colleagues, produced more trust and better joint outcomes, and in their follow-up, where the pie was fixed and only its division was at stake, it produced more even splits; and leaders who mock themselves are rated more considerate than leaders who mock others, and more effective, partly through the mediating variable of trust. Self-deprecation, note, is here a tool of the already elevated—a way of refunding some of the distance that status creates. Hold that thought; the dating literature prices it precisely.

IV · The License

Now the harder ledger. In 2019 the organizational psychologist Jonathan B. Evans and colleagues filmed four versions of the same presentation: a retail-store manager presenting her own store’s quarterly results to regional managers, played by a male or a female actor matched for age and appearance, delivering scripts identical but for five humorous lines. Two hundred and sixteen participants evaluated them. The men and women in the audience agreed with each other, which is the depressing part: the same jokes read as functional from the male manager—one viewer called him witty—and as disruptive from the female manager, of whom another viewer complained of “poor judgment in jokes.” Humor lifted the man’s status relative to his unfunny twin and lowered the woman’s relative to hers, and the difference flowed through to performance ratings and judgments of leadership capability. Same instrument, different license.

Honesty requires the counter-evidence, which exists. In a 2001 survey of business-school alumni rating their own managers, Wayne Decker and Denise Rotondo found that women managers who used positive humor gained more in rated effectiveness than men who did—though negative humor cost them more, too. And a 2023 analysis of 2,407 TED talks by Ella Miron-Spektor and colleagues found humor increased a speaker’s influence especially for women; in a subset of ninety-two of the most influential talks, the benefit ran through perceived warmth and competence at once. The literatures do not agree, and the disagreement is informative: the vignette isolates the stereotype and finds it punitive; the field data measure the survivors—women whose humor has been polished against exactly that stereotype—and find it profitable. The honest summary for a working woman is not “don’t joke”; it is that her joke operates in a market with a prejudicial insurance premium, one that the best performers evidently recoup.

Behind the license question lurks the ability question, which the culture answers with a confidence the data do not support. A 2020 meta-analysis by Gil Greengross, Paul Silvia, and Emily Nusbaum—thirty-six effect sizes, 5,057 people—found that men’s attempts at humor were rated funnier on average, by about a third of a standard deviation: a real gap, and a small one, the kind that leaves the two distributions mostly sharing territory, and whose origin the authors themselves leave open between evolutionary and cultural accounts. What is not open is that the bookkeeping is rigged. In a caption-writing study by Laura Mickes and colleagues, raters found men’s captions modestly funnier—and then, on a memory test, misattributed the funny captions they had seen to men and the unfunny ones to women. The stereotype does not merely summarize the ledger; it edits it. Provine’s field notes add the behavioral asymmetry: in cross-gender conversations, women laughed 126 per cent more than the men they spoke with—women, he wrote, do most of the laughing, men most of the laugh-getting—which is a division of labor, not a difference in equipment.

So the answer at the office is this: nothing in the evidence says a woman needs a different kind of humor—the styles that help are the same four, aimed the same way. What differs is the tariff schedule. Her successful joke may be under-credited and her misfire over-billed; his collateral is cheaper. Whether that tariff is best answered with fewer jokes or with better ones is exactly the disagreement between the laboratory and the TED stage—and the field data, for what they are worth, side with the jokes.

V · Hearth

At home the accounting changes: nobody is allocating leadership points at the dinner table, or not officially. The widest view belongs to the communication scholar Jeffrey Hall, whose 2017 meta-analysis pooled forty-three samples—15,177 people, average age thirty-four, most of them actually partnered rather than undergraduates imagining it—and asked which humor tracked relationship satisfaction. Every one of the six positive humor types did; four of the five negative types tracked dissatisfaction; and the gradient is the finding: one’s own reported funniness mattered a little, a partner’s perceived funniness mattered more, and humor experienced together—inside jokes, shared comic memory, the couple’s private frequency—mattered most. “It’s good to have humor,” Hall told his university’s news service, in what may be the field’s most compressed abstract. “It’s better to see it in your partner. And it’s best to share it.”

The shared laugh can be caught on camera. The psychologists Laura Kurtz and Sara Algoe videotaped seventy-one couples telling the story of how they met and coded the laughter: not how much either partner laughed, but what share of the conversation they spent laughing simultaneously. That share—with all other laughter controlled—uniquely tracked relationship quality, closeness, and perceived support. Laughing at the same instant is a small behavioral proof that two nervous systems are running the same appraisal: the violation is benign, and it is benign for us. Their follow-up work suggests the mechanism is exactly that—shared laughter reads as evidence of similarity.

Conflict is where the instrument earns its keep. In the marriage laboratory of the psychologist John Gottman, 130 newlywed couples were filmed arguing, then followed for up to six years; the models built from those first conversations predicted divorce with eighty-three per cent accuracy. The variable that predicted both stability and happiness among the stable was not the absence of anger but the presence, inside the argument, of positive affect—humor, affection, interest—deployed contingently, at the moments it could de-escalate, physiologically soothing the partner (in these data, especially the husband; the effective soothers of husbands were often wives armed with a well-timed joke). Gottman’s own caution deserves quoting, because the self-help industry inverts it: “The admonition to be funny or to enjoy a partner’s attempts at humor are probably self-destructive mechanisms”—the humor must arrive naturally or not at all. Where does natural come from? A follow-up study by Janice Driver and Gottman filmed forty-nine couples at dinner and found that the silliness of ordinary evenings—playful bids, defined in the coding manual as good-natured teasing up to and including thumb wrestling—tracked the couple’s humor during conflict closely: the husband’s playful bids correlated 0.675 with it, the wife’s 0.626. The couple that clowns on Tuesday has jokes in the bank for Saturday’s fight. The evidence is observational, and the authors, who fitted the arrow in both directions and found it about equally strong either way, take the honest position that the traffic probably runs both ways.

And the sexes? Here the asymmetry all but disappears, and what remains of it points the wrong way for the evolutionary story. Hall tested the question directly in 2019: his survey of couples found no sex difference in how much a partner’s humor mattered, while his ten-study meta-analysis found one—a partner’s humor production was associated more strongly with men’s satisfaction than with women’s, the reverse of what a courtship-display account predicts. Long-term love, it appears, is humor’s most nearly unisex market. The asterisk is teasing. Dacher Keltner and colleagues, studying couples, found the satisfied ones teased with less face-threat and more repair—and that women reported being teased as more distressing than men did. The joke at the hearth is symmetrical; the pinch is not.

VI · The Company of Friends

Friendship is where laughter does its plainest chemical work, and the evidence here includes actual chemistry. Start with manufacture: in a 2004 experiment from Arthur Aron’s closeness laboratory, ninety-six strangers were paired off and run through structured tasks engineered either to strike shared humor or not—and the pairs who had laughed together reported feeling significantly closer, an effect carried partly by the sense that the encounter had enlarged them and partly by simple distraction from first-meeting awkwardness. Manufactured mirth, between strangers, manufactured intimacy. The linguist Jennifer Coates, transcribing real friends, described what the manufactured version imitates: humorous talk among intimates runs in a “play frame” where speakers overlap, repeat, and build each other’s sentences—less like alternating monologue than like jazz.

Now the chemistry. The evolutionary psychologist Robin Dunbar and colleagues ran six experiments—laboratory comedy screenings and live shows, including at the Edinburgh Fringe—measuring pain thresholds before and after, on the logic that pain tolerance indexes endorphin release. Thresholds rose after shared laughter and not after equally pleasant but unfunny material: the effect followed the laughing, not the good mood. A Finnish–British team later closed the loop with positron-emission tomography: social laughter triggered opioid release in the thalamus, caudate nucleus, and anterior insula, and baseline receptor availability predicted how readily people laughed. In Dunbar’s interpretation, laughter is a grooming technology—the primate bonding ritual scaled from two nitpicking partners to a whole table at once. It is the cheapest social glue the species manufactures, which is presumably why Provine’s mall transcripts found it stuck to sentences as banal as “Where have you been?”

Humor also guards the self. George Vaillant, custodian of the Grant Study’s 268 Harvard men, ranked humor among the “mature” psychological defenses—the adaptations that metabolize suffering without denying it. Viktor Frankl, remembering the camps, called humor “another of the soul’s weapons in the fight for self-preservation.” The laboratory agrees at its smaller scale: reappraising distressing images with good-natured humor beats mean-spirited humor at lifting mood, and in the classic moderation studies, students with a live sense of humor showed weaker links between bad life events and bad months. The style distinction returns here with force: it is the benign instruments that buffer. The self-defeating habit, the meta-analyses say, keeps bad company—more depression, less esteem—a coping strategy that bills the account it claims to protect.

What humor will not reliably do is medicine, and the honest essay says so. Rod Martin’s stern 2001 review found the immunity studies inconsistent, the analgesia unspecific—negative emotions numbed as well—and “no evidence of increased longevity with greater humor.” The intervention literature since is warmer but wobbly: a 2019 systematic review of laughter-inducing therapies found real improvements in depression, shrinking once the randomized trials were separated out, with the review’s own verdict that study quality was low throughout—and one deadpan twist: simulated laughter exercises, no jokes involved, outperformed the humorous kind, a gap that itself thins to nothing among the randomized trials. The medicine, such as it is, appears to be the laughing and the company, not the comedy. The grandest data are Norwegian: in the HUNT cohort, 53,556 adults followed for fifteen years, women with high scores on the cognitive component of humor had roughly half the all-cause mortality of low scorers (a hazard ratio of 0.52 in the adjusted models), with the cardiovascular and infection numbers more striking still—while for men the association held only for death by infection, and the earlier seven-year analysis found the whole effect fading after sixty-five. Observational, unexplained, and unreplicated at that scale; but it is a curious fact worth carrying forward that even the mortality tables decline to treat the sexes symmetrically.

VII · The Market for Wit

Courtship is where humor’s reputation was made, and the headline is real. In the psychologist Richard Lippa’s BBC internet survey—119,733 men and 98,462 women across fifty-three nations, each picking the three traits they most wanted in a partner—humor ranked second overall, behind only intelligence. The sexes differ at the margins in the expected directions, and unequally: women ranked humor a little higher than men did, by about a fifth of a standard deviation, while men ranked good looks higher by more than twice that. It is beneath the headline that the market turns strange. In an experiment by Eric Bressler and Sigal Balshine, 210 students evaluated photographs paired with funny or bland statements attributed to the person shown: women preferred the funny man, and men were indifferent to the funny woman. Stranger still, the funny targets of both sexes were rated less intelligent and less trustworthy—and chosen anyway. Whatever the joke was selling, it was not certified brainpower.

The field’s most replicated asymmetry is about roles, not ability. Asked to choose between a partner who makes them laugh and a partner who laughs at their jokes, women in Bressler, Rod Martin, and Balshine’s studies took the producer, for dates, for the long term, even for friendship; men took the appreciator, most clearly for dates. A decade later Liana Hone, William Hurwitz, and Debra Lieberman replicated the pattern and priced it with a budget task: allotting scarce “mate dollars,” men spent three-fifths of their budget on a partner’s receptivity to their humor—treating being laughed at as a necessity—while women bought production first. “We have no general ‘sense of humor,’” the authors concluded; there are two markets wearing one name. The advertising archive concurs: in 3,745 newspaper personal ads counted by Provine on a single April day in 1996, women were sixty-two per cent more likely to mention laughter, mostly seeking it, while men mostly offered it; the online-dating literature has replayed the pattern ever since—his profile advertises wit, hers requests someone who makes her laugh.

Why should she audit and he perform? The famous account is the evolutionary psychologist Geoffrey Miller’s: humor as a fitness display, a hard-to-fake flash of intelligence, and there is evidence for it—in a study of four hundred students, caption-writing ability correlated with vocabulary and abstract reasoning, predicted lifetime partner counts, and showed no sex difference in how well it tracked intelligence—which the authors read as humor being an equally reliable cue in each sex. But the display theory has been mauled by fieldwork. Jeffrey Hall, across three studies, found humor production unrelated to intelligence as he measured it—by grades and college-entrance scores—and tied instead to extroversion, with warmth turning out to be the correlate of appreciating humor rather than producing it; Bressler’s subjects, remember, rated the funny less intelligent; and in transcripts of 947 four-minute speed dates analyzed by Daniel McFarland, Dan Jurafsky, and Craig Rawlings, the laughter that tracked a mutual “click” was strikingly often the man’s—men laughed, matched the woman’s laughter, offered appreciations; “a man signals his intent to select a woman when he laughs,” the authors wrote. In Hall’s own recordings of first encounters between strangers, what tracked interest in both parties was laughing at the same moment. The audit reading: courtship humor is less an IQ test than an attention test—proof that two people are appraising the world in the same key, with her laughter, as the ethologists found long ago, the single best telltale of where the encounter is going.

And self-deprecation, the instrument everyone is told to reach for? It is a luxury good. Gil Greengross and Geoffrey Miller had students hear presenters of high or low status mock themselves or mock rivals: self-deprecating humor from the high-status presenter raised long-term attractiveness in both sexes’ ratings; from the low-status presenter it did nothing at all. An earlier study found self-ridicule helped the physically attractive, men especially, and only in ratings for a serious relationship. The mechanism is the one the office already taught: mocking yourself spends status, which requires having it, and audibly being able to afford the loss. From a professor, “I’m useless before coffee” is charm; from the intern it is information. The self-deprecating joke, in short, is interest paid on capital—and the couple that survives the courtship graduates into Part V’s better economy, where the shared laugh replaces the performed one and the sexes finally trade at par.

VIII · New York’s Stage

New York made the joke into a public performance. The Friars Club developed the roast as a ritual of affectionate demolition; the Improv, Catch a Rising Star, the Comic Strip, and the Comedy Cellar gave performers rooms in which a failed line could be tried again. These histories do not establish a single New York temperament. They show institutions built to let an audience judge a joke in public.

The city also has private rooms, where the rules are different. Deborah Tannen’s account of New York Jewish conversational style describes cooperative overlap: talk that might look like interruption to an outsider can signal attention among people sharing the register. Her recorded dinner took place outside the city, among speakers raised here, and it cannot stand for every New Yorker. It does make the mechanism visible. A joke, like an overlap, needs a receiver who knows what game is being played.

The studies in this essay distinguish a leader’s joke, a couple’s joke, a friend’s joke, and a stranger’s. New York gives all four their stages, but the stage does not determine the outcome. A roast among friends can bond; the same barb from a stranger can wound. The city’s comedy history offers an example of permission being negotiated in public. The science supplies the more durable rule: who laughs, who may answer, and who bears the cost of a failed line matter more than the fact that a line was delivered.

IX · The Uses

What, then, does the evidence license a reader to conclude? Five uses, in descending order of reliability. Laughter binds: shared laughter builds closeness between strangers, tracks the health of marriages, and releases the bonding chemistry directly—this is the firmest finding in the file. Humor pays at work: the successful joke buys status and leadership at measurable rates, and the flat one costs little—what costs is the joke that violates the room, which is charged whether or not it raises a laugh. At home, the asset is not being funny but laughing together—cultivated, if the observational data are to be believed, in the low-stakes silliness of ordinary evenings, and never on command. In courtship the instrument is a signal read asymmetrically—he tends to perform, she tends to audit, and the self-mocking joke yields only to those who can visibly afford it. And for the body: modestly, socially, and—if Norway generalizes—mostly for women, a fact nobody yet explains.

Which leaves the question this essay set itself: whether the sexes must reach for different humor to collect the same benefit. The file supports a two-part answer. The kinds that help are the same kinds: every ledger in this essay—mental health, marriage, office, friendship—favors the benign styles, aimed with rather than at, and bills the aggressive and self-abasing habits, for men and women alike. No study here finds women needing a separate comic repertoire. What differs is the tariff: the office vignettes bill her joke at a premium the field data say good performers recoup; courtship assigns the sexes different default roles in the same transaction; and self-deprecation, the most democratic-looking instrument of all, is priced by status, which the sexes do not hold equally. The asymmetry, in short, sits in the audience, not the instrument—and audiences, unlike instruments, are renegotiable, a renegotiation visibly underway between the laboratory’s verdict and the TED stage’s.

Then the humility, which is house doctrine. Most of these findings were bought from undergraduates in two rich countries; most are correlations; several of the sharpest are single experiments awaiting their replications; and the New York setting is this essay’s argument, not anyone’s dataset. White was right about the frog: dissected, it dies on the table, and no one who has read a variance decomposition has ever laughed at one. But the dissection did find something the joke books miss. Go back to the mall, to the man with the notebook, and read again what people were actually laughing at. Where have you been? It was nice meeting you, too. Not wit—arrival, recognition, the small ceremonies of being expected somewhere. The instrument measured in these pages turns out to run less on cleverness than on company, which is presumably why it works at the office, at the hearth, among friends, and across a speed-dating table—and why it survives translation between a city that performs it nightly and a city that saves it, dry and cast in chocolate, for December. The laughter was never about the joke. It was about the meeting.

Notes on Sources

Provine, R. R., & Fischer, K. R. (1989), diary study of social laughter; Provine, R. R. (1993), laughter punctuates speech. Ethology. The thirty-times, ten-to-fifteen-per-cent, forty-six-per-cent and 126-per-cent figures are as restated by Provine (2004), Current Directions in Psychological Science; the ninety-nine per cent is from the 1993 paper itself; popular treatment in Laughter: A Scientific Investigation (Viking, 2000), source of the 3,745 personal ads and the sixty-two per cent.

White, E. B. (1941). Preface to A Subtreasury of American Humor, edited by E. B. White and Katharine S. White; the preface is signed by him alone, and the magazine version of the same year reads “purely scientific mind.”

Martin, R. A., Puhlik-Doris, P., Larsen, G., Gray, J., & Weir, K. (2003). The Humor Styles Questionnaire. Journal of Research in Personality. N = 1,195 across development samples.

Schneider, M., Voracek, M., & Tran, U. S. (2018). “A joke a day keeps the doctor away?” Meta-analysis, 45 samples, N = 12,734. Scandinavian Journal of Psychology. Per-style pooled coefficients sit behind a paywall and are not quoted here.

McGraw, A. P., & Warren, C. (2010). Benign violations. Psychological Science.

Bitterly, T. B., Brooks, A. W., & Schweitzer, M. E. (2017). Risky business: when humor increases and decreases status. Journal of Personality and Social Psychology. Source of the promotional-testimonial studies, the interview studies, and the status means quoted. American samples.

Mesmer-Magnus, J., Glew, D. J., & Viswesvaran, C. (2012). Meta-analysis of positive humor in the workplace. Journal of Managerial Psychology. k = 49, n = 8,532; corrected correlations; several cells rest on three to five studies, as the authors warn.

Yam, K. C., Christian, M. S., Wei, W., Liao, Z., & Nai, J. (2018). The mixed blessing of leader sense of humor. Academy of Management Journal. Three-wave field studies in China (N = 215) and the United States (N = 200 completing); the deviance path was tested in both samples, the engagement path and the aggressive-humor moderations in the American one.

Huang, L., Gino, F., & Galinsky, A. D. (2015). Sarcasm and creativity. Organizational Behavior and Human Decision Processes.† (Single team, small cells, no direct replication; unrelated misconduct findings against the second author do not concern this paper, but caution is doubled.)

Kurtzberg, T. R., Naquin, C. E., & Belkin, L. Y. (2009). Humor in online negotiations. International Journal of Conflict Management. The trust and joint-gain results are from their first study (122 executive MBA students); the second (216 MBA students) tested a purely distributive negotiation. Hoption, C., Barling, J., & Turner, N. (2013). Leadership & Organization Development Journal (self-deprecating vs. aggressive leader humor). Gkorezis, P., & Bellou, V. (2016). LODJ (effectiveness via trust; cross-sectional).

Evans, J. B., Slaughter, J. E., Ellis, A. P. J., & Rivin, J. M. (2019). Gender and the evaluation of humor at work. Journal of Applied Psychology. N = 216; the two participant remarks quoted here appear in the authors’ own account of the study in the Harvard Business Review, March 2019, not in the journal article.† (Single vignette experiment, no published replication; the field studies below point the other way, and the essay reports the dispute.)

Decker, W. H., & Rotondo, D. M. (2001). Gender, humor type, and perceived leader effectiveness. Journal of Managerial Issues.† (Correlational survey, single-source ratings.) Miron-Spektor, E., Bear, J. B., & Eliav, E. (2023). Think funny, think female. Academy of Management Discoveries (online first, 2022). 2,407 TED talks, observational; the warmth-and-competence mechanism comes from a ninety-two-talk subset.

Greengross, G., Silvia, P. J., & Nusbaum, E. C. (2020). Sex differences in humor production ability: a meta-analysis. Journal of Research in Personality. d = 0.321 across 36 effect sizes, N = 5,057; the evolutionary-versus-cultural interpretation is left open by the authors. Mickes, L., Walker, D. E., Parris, J. L., Mankoff, R., & Christenfeld, N. J. S. (2012). Who’s funny. Psychonomic Bulletin & Review.

Hall, J. A. (2017). Humor in romantic relationships: a meta-analysis. Personal Relationships. 43 samples, N = 15,177; quotes from the University of Kansas news service. Hall, J. A. (2019). Humor — its couples survey (N = 246) found no moderation by sex; its Study 1 meta-analysis (k = 10, N = 2,167) found partner humor production more strongly associated with men’s satisfaction, which the author notes runs contrary to prediction.

Kurtz, L. E., & Algoe, S. B. (2015). Shared laughter as behavioral indicator of relationship well-being. Personal Relationships; and (2017) Journal of Nonverbal Behavior.

Gottman, J. M., Coan, J., Carrère, S., & Swanson, C. (1998). Predicting marital happiness and stability from newlywed interactions. Journal of Marriage and the Family. Humor is a component of the coded positive-affect construct. Driver, J. L., & Gottman, J. M. (2004). Family Process; the two playful-bid correlations and the authors’ bidirectional inference are from its Table 2 and discussion. Quotations from both papers are verbatim.

Keltner, D., Young, R. C., Heerey, E. A., Oemig, C., & Monarch, N. D. (1998). Teasing in hierarchical and intimate relations. JPSP; Keltner, D., et al. (2001). Just teasing. Psychological Bulletin.

Fraley, B., & Aron, A. (2004). Shared humorous experience and closeness. Personal Relationships. N = 96. Coates, J. (2007). Talk in a play frame. Journal of Pragmatics.

Dunbar, R. I. M., et al. (2011, print 2012). Social laughter is correlated with an elevated pain threshold. Proceedings of the Royal Society B. Manninen, S., et al. (2017). Social laughter triggers endogenous opioid release. Journal of Neuroscience.

Vaillant, G. E. (1977). Adaptation to Life (Grant Study, 268 men). Frankl, V. E. (1946). Man’s Search for Meaning. Samson, A. C., & Gross, J. J. (2012). Cognition & Emotion. Martin, R. A., & Lefcourt, H. M. (1983). JPSP.

Martin, R. A. (2001). Humor, laughter, and physical health. Psychological Bulletin. van der Wal, C. N., & Kok, R. N. (2019). Laughter-inducing therapies. Social Science & Medicine.† (The review’s own verdict: low study quality throughout. Its simulated-beats-humorous contrast is a descriptive average across all designs; restricted to the randomized trials, the difference is not significant, and the essay says so.)

Romundstad, S., Svebak, S., Holen, A., & Holmen, J. (2016). Sense of humor and causes of mortality, the HUNT study. Psychosomatic Medicine. N = 53,556; hazard ratios as quoted.† (Observational cohort; the sex split is unexplained and unreplicated at this scale; the earlier 7-year analysis, Svebak et al. 2010, found the association fading after about sixty-five.)

Lippa, R. A. (2007). Preferred traits of mates in a cross-national study. Archives of Sexual Behavior. Bressler, E. R., & Balshine, S. (2006); Bressler, E. R., Martin, R. A., & Balshine, S. (2006). Evolution and Human Behavior. Hone, L. S. E., Hurwitz, W., & Lieberman, D. (2015). Evolutionary Psychology. Wilbur, C. J., & Campbell, L. (2011). Personality and Social Psychology Bulletin. Whitty, M. T. (2008). Computers in Human Behavior.

Miller, G. (2000). The Mating Mind (Doubleday).† (The fitness-indicator reading of humor is contested; Hall 2015, Evolutionary Psychology, found production unrelated to measured intelligence, and Bressler’s funny targets were rated less intelligent. Supporting it: Greengross, G., & Miller, G. 2011, Intelligence, N = 400.) McFarland, D. A., Jurafsky, D., & Rawlings, C. M. (2013). Making the connection. American Journal of Sociology. Grammer, K., & Eibl-Eibesfeldt, I. (1990), the woman’s-laughter finding, as described in the Bressler papers.

Greengross, G., & Miller, G. F. (2008). Dissing oneself versus dissing rivals. Evolutionary Psychology. (For low-status presenters the result is a null, not a demonstrated backfire.) Lundy, D. E., Tan, J., & Cunningham, M. R. (1998). Personal Relationships.† (The benefit was confined to ratings for a serious relationship, and the study did not statistically compare women’s and men’s preferences.)

On New York: the New York Friars Club (founded 1904; member roasts a formal tradition from about 1949–50; defunct by 2024); club histories for the Improv (opened April 1963 as a coffee house; its first comedian appeared in 1964), Catch a Rising Star (1972), the Comic Strip (1975–76, accounts differ), and the Comedy Cellar (1981 by its own account); the New Yorker prospectus (1925) via its published reproductions; Freud, S. (1905), Jokes and Their Relation to the Unconscious (Strachey translation).† (The self-mockery thesis is contested: Ben-Amos, D. 1973, Western Folklore.) Tannen, D. (1981). New York Jewish conversational style. International Journal of the Sociology of Language; her recorded dinner took place in California among speakers raised in New York, and her term for the overlap is cooperative overlap. Marx, G., Groucho and Me (1959), the telegram as printed there; the variant “people like me” belongs to an earlier retelling, per Quote Investigator’s sourcing. Rogerson-Revell, P. (2007). Humour in business: a double-edged sword. Journal of Pragmatics. Meyer, E. (2014). The Culture Map (PublicAffairs).† (Practitioner observation, not peer-reviewed.)

Jiang, T., Li, H., & Hou, Y. (2019). Cultural differences in humor perception, usage, and implications. Frontiers in Psychology.

† marks findings that are contested, methodologically fragile, or debated at the level of interpretation. Figures are stylized schematics after the cited studies, never replotted data, except Fig. 4, whose proportions follow the cited means. Meta-analytic correlations quoted are the authors’ corrected estimates. Where a number appears in the text—the 6.66 leadership points, the third of a standard deviation, the hazard ratio of 0.52—it is the cited authors’ estimate. Fraley and Aron’s abstract gives ninety-six without saying whether it counts individuals or pairs. The sex differences in Part VII are Lippa’s d = 0.22 for humor and 0.55 for good looks. Most samples are Western, educated, and young; the New York stage in Part VIII is an editorial reading of institutions, not a citywide measure of humor norms.

Drafted with AI (Claude, Anthropic) as a private research synthesis, August 17, 2026, from a claim inventory built study by study, then put through an adversarial pass whose instructions were to refute rather than confirm. That pass found errors—a reversed finding about failed jokes, a misattributed quotation, a preface credited to two hands instead of one—and they are corrected above rather than quietly kept; contested science is flagged rather than smoothed. Set, with affection and without affiliation, in the manner of a certain Manhattan weekly—which is not edited for the old lady in Dubuque either.New York revision drafted with AI on October 12, 2026; editorial review pending. ◆ ◆ ◆