When is a joke too soon? A scientific inquiry.

Entry 2: When is a joke too soon? A scientific inquiry.

Photo illustration by Natalie Matthews-Ramo.

Photos by Anetlanda/Thinkstock and Creatas/Thinkstock. Photo illustration by Natalie Matthews-Ramo.

In the middle of 2001, the satirical newspaper The Onion moved its operation from its hometown of Madison, Wis., to New York City. This was a big break for the publication, and the staff was especially excited to put out their first issue from their new headquarters. It was scheduled to be published on Sept. 11, 2001.

That issue never hit newsstands. Instead, the Twin Towers disappeared from the Manhattan skyline, seemingly taking with them all potential for humor. Saturday Night Live, The Daily Show, all the late-night talk shows halted production. Time magazine declared, “The Age of Irony Comes to an End.”

“We wondered, ‘Is this the end?’ ” Todd Hanson, a former writer for The Onion, told us when we visited him in New York City. It was a fair question. Mark Twain famously said, “Humor is tragedy plus time.” But this was a tragedy of unprecedented proportions and might require an unprecedented amount of time before humor again felt appropriate. It wasn’t at all clear the newly transplanted Onion could wait it out.

When is it too soon to joke about something—and when is it too late? Dilemmas like the one The Onion faced led Peter McGraw, the academic half of our duo, to wonder if it might be possible to quantify when, exactly, jokes about a touchy subject start working—and when they become so worn out they evoke yawns instead of laughs.

Pete’s Humor Research Lab (HuRL) at the University of Colorado–Boulder launched an experiment when Hurricane Sandy began developing in the western Caribbean in late October 2012. As described in a recent Social Psychological and Personality Science journal article, in the days before the storm made landfall, the team collected three tweets from the recently launched @AHurricaneSandy, a real Twitter account that employed an abrasively humorous tone:

At various points of time as the storm progressed, different online survey participants rated the humorousness of the tweets on one scale and their offensiveness on another. The resulting ratings followed a curvilinear pattern: Participants thought the tweets were funniest when they were asked about them the day before the storm made landfall. Later, when millions were without power and the East Coast had sustained hundreds of casualties and billions of dollars in damage, the funniness of @AHurricaneSandy bottomed out. But as the trauma subsided, the humor ratings bounced back, peaking 36 days after the storm hit, only to decrease again as the emotional intensity of the catastrophe continued to fade into the past.

For the Sandy tweets, in other words, it wasn’t as simple as “too soon” or “too late.” Instead, the humor seemed to depend on the threat level—subjects found the tweets funniest when there was not too much danger, but also not too little.

This may seem like a straightforward observation, but most of the major humor theories that have been posited over the millennia don’t easily explain the phenomenon McGraw tracked during the storm. If people laugh to feel superior to others, as Plato and Aristotle conjectured, or for psychic relief, as Sigmund Freud believed, the pattern would have been the opposite: The funniest ratings should have occurred at the time of greatest human misfortune and trauma. And if humor arises when there’s an incongruity between what people expect to happen and what actually happens, a popular theory first put forward by 17th-century French philosopher Blaise Pascal, participants should have rated the tweets equally funny throughout the experiment, since the incongruity of the tweets (a hurricane acting like a wiseass) never changed.

But the @AHurricaneSandy experiment’s results appear to fit with the benign violation theory of humor, developed by McGraw and Caleb Warren. According to the theory, humor arises when a concept seems wrong or threatening but is simultaneously OK or safe. Before the storm hit, the jokes were funny because Sandy wasn’t too threatening. At the height of the storm, the jokes were too much of a violation; that’s when people rated the tweets the most offensive. With the passage of time, the offensiveness faded and humor returned. Months later still, when most people had moved on with their lives, the concept wasn’t funny because it was merely benign. (Interestingly, it doesn’t appear that the subject of Sandy had simply gotten stale with time; participants also rated the tweets’ irrelevance, and these ratings didn’t skyrocket as humor ratings dropped.)

The benign violation theory highlights the role of psychological distance in comedy: Not enough distance, and the joke offends; too much, and it bores. So how do you find that comedic sweet spot? Waiting for days, months, or even years before tackling a taboo subject is an obvious way to make a violation feel distant and therefore more benign. (South Park, for instance, declared in 2002 that AIDS was “finally funny.”) But as HuRL experiments have demonstrated, there are other ways to increase psychological distance besides simply waiting things out. In another study, participants read about a young woman who texted “Haiti” to donate $10 to a mobile charity program some 200 times—without realizing who would be footing the bill for her philanthropy:

People found this story more amusing when the woman was described as a stranger, rather than a close friend of the subject. In other words, a relatively extreme violation like accidentally spending $2,000 was less threatening and therefore funnier when researchers increased the psychological distance between the person experiencing the mishap and the person who’s supposed to laugh. On the other hand, study participants who read about a woman texting “Haiti” five times and accidentally charging herself $50 found the story more amusing when the woman was described as a friend rather than a stranger. So a relatively mild violation like this $50 mistake was only threatening enough to be funny when there was minimal distance between the subject of the joke and the person who’s supposed to get it.

So maybe we have it all wrong when we ask whether a joke is “too soon.” Perhaps a better question to ask is, “When is the punch line too close for comfort, and when is it too distant to matter?”

Hanson and his Onion colleagues looked at 9/11 this way—and less than two weeks after the towers came down, they tackled the tragedy head on, creating a whole issue devoted to the terrorist attacks. But they were exceedingly careful about their punch lines, keeping a safe comedic distance from the horrors that had transpired. They didn’t joke about the civilians who died that day or the new terror of flying in an airplane. Those subjects were too raw. Instead, they turned the terrorists into fools (“Hijackers Surprised to Find Selves in Hell,” read one article’s headline) and cracked wise about the strange aura of confusion and despair that had settled over the country (“Not Knowing What Else to Do, Woman Bakes American-Flag Cake,” read another.)

The day after the issue came out, people all over the country began faxing The Onion grateful comments, and fan mail flowed in by the thousands. “To me, it’s not about timing. It’s about legitimate versus illegitimate targets,” says Hanson. “If what you are saying is honest and legitimate and has a valid point, it’s going to be valid the day after, and it’s going to be valid 500 years later.”

Next up: Laughter, deconstructed. Researchers are discovering chuckles, guffaws, and cackles are far more complicated than anyone realized.

This series is adapted from the The Humor Code: A Global Search for What Makes Things Funny.