Joe Toplyn writes a guest post for the Humor Code project:
Joe has written and produced television comedy for over twenty years. He won four Emmy Awards writing for Late Night with David Letterman, then went on to become co-head writer of The Tonight Show with Jay Leno, head writer of Late Show with David Letterman, head writer of The Caroline Rhea Show, and co-executive producer of Monk. He is also the author of the book Comedy Writing for Late-Night TV.
Why We Laugh at Late-Night TV
The benign violation theory of humor says that, as Peter McGraw puts it, “Laughter only occurs when something is both wrong and okay.” Does that theory crack the humor code of late-night comedy/talk shows? Yes, but only up to a point.
As I put together my book Comedy Writing for Late-Night TV I realized I needed a theory to explain why people laugh. A convincing theory of laughter would make all the comedy writing advice in my book seem less random.
So I did some research and decided that the benign violation theory in The Humor Code makes a lot of sense. But the more I thought about how jokes are actually written and about what makes one version of a joke funnier than another, the more I realized that I had to tweak the theory to make it fit the short-form comedy found on late-night TV. The amended theory that I use in my book I call the surprise theory of laughter.
The surprise theory of laughter: We laugh when we’re surprised that an incongruity turns out to be harmless.
My surprise theory differs from the benign violation theory in two key ways. The first key difference is this:
The surprise theory implies that a sequence of events, rather than simultaneous events, produces laughter. To be more specific, laughter results when something first seems incongruous (a violation) and then is revealed to be harmless (benign).
Let’s see how this sequence of events unfolds in a typical late-night monologue joke, one that I recently wrote and posted on Twitter:
Alaska Airlines frequent-flier lounges have pancake printers that make 180 pancakes an hour. So now passengers can join the Mile Wide Club.
When listeners hear “Mile Wide Club” their first thought is, “That’s wrong. It’s supposed to be ‘Mile High Club.’ That’s what airline passengers join.” But then a split-second later they realize that “Mile Wide Club” actually does make sense because eating all those pancakes would make passengers extremely fat. The initial violation of the listeners’ world view turns out to be benign. So the listeners laugh, like prehistoric cave people signaling to their cave families that some suspiciously rustling grass does not hide a deadly saber-toothed cat after all but instead a couple of mating monkeys.
The second key difference between the benign violation theory and my surprise theory of laughter is this:
The surprise theory acknowledges the importance of, yes, surprise in producing laughter. Building on Peter’s words, the surprise theory says that laughter only occurs when something is first wrong and then, surprisingly, okay.
To see how important surprise is, let’s remove some of it from the sample joke this way:
Alaska Airlines frequent-flier lounges have pancake printers that make 180 pancakes an hour. So now instead of joining the Mile High Club, passengers can join the Mile Wide Club.
That revised joke wouldn’t get as big a laugh, would it? The reason is because as soon as listeners hear “Mile High Club” they start expecting a play on those words as the punch line. So “Mile Wide Club” isn’t quite as surprising when it finally arrives. As Joke Maximizer #5 in my book advises: Don’t telegraph the punch line.
Here’s a different way I could rewrite the original joke to reduce the surprise:
Alaska Airlines frequent-flier lounges have pancake printers that make 180 pancakes an hour. So now joining the Mile Wide Club is something else passengers can do.
That revised joke puts what I call the laugh trigger—the most unexpected, surprising word or two in the punch line—in the middle of the joke. So when listeners hear “Mile Wide Club” they know the meat of the punch line before the joke is even finished. As a result, their laughter is more diffuse and not as strong. Writers can avoid that problem by applying my book’s Joke Maximizer #2 and ending their jokes on the laugh trigger.
Comedy Writing for Late-Night TV is packed with other practical tips, tricks, and techniques that demonstrate the surprise theory of laughter in action. Taken together they provide ample evidence that the benign violation theory only begins to explain what makes people laugh.
———————————————————————————–
Pete briefly responds.
-First, Joe is way funnier than me. So, you should think hard about what he has to say.
-Second, one challenge of creating a universal theory of humor is that a specific theory (such as Joe’s surprise theory) could outperform the more general theory when applied to a specific form of comedy (such as late night jokes). In that way, you should think hard about what Joe has to say.
-Third, I would caution Joe and others against over weighting the role that surprise plays in humor. It certainly matters in terms of creating a more arousing experience. However, surprise is not necessary for a humorous response. People often find a joke amusing even when hearing it over and over again. Joel and I recently enjoyed watching the famous “who’s on first” routine from Abbott and Costello. There are few surprises there (we have seen it many times), yet we were cracking up.
Thanks for the guest post Joe!
You make an excellent point, Pete. If comedy material has to be surprising to get a laugh, as I believe, why did you and Joel laugh at the “Who’s On First” routine even after you’d seen it many times?
I would argue that the routine still surprised you. It’s long–at eight minutes–very fast-paced, and dense with verbal gymnastics. I think it’s entirely possible that you laughed at harmless incongruities that hadn’t registered with you on previous viewings, or that you had forgotten.
The role of surprise in creating laughs is probably testable but one test that you mention in The Humor Code seems flawed to me. On page seven you describe how “two University of Tennessee professors had 44 undergraduates listen to a variety of Bill Cosby and Phyllis Diller routines. Before each punch line, the researchers stopped the tape and asked the students to predict what came next.” You go on to relate that “the predictable punch lines were rated considerably funnier [by a second group of students] than those that were unexpected.”
But I believe that stopping the taped performance of a joke and giving the students ample time to predict the punch line didn’t measure how surprising the punch line would be when it really counts, when the joke was delivered at full speed. What stopping the tape did was allow the students enough time to consult their vast internal database of past jokes and figure out which time-tested formula would probably be used this time to create the punch line.
Put another way, the very fact that the students could predict the punch line when they thought about the joke for a while strongly implies that the joke was written according to familiar formulas. That’s why the second group of students rated those jokes as funnier when they heard them at full speed the way they were designed to be heard—those familiar, proven joke-writing formulas worked. The formulas created surprising, harmless incongruities and resulted in laughs. I describe those formulas in Comedy Writing for Late-Night TV.
Another possible flaw in that University of Tennessee study, which I didn’t read, is how those two professors had their test subjects rate the funniness of the jokes. As someone whose employment has depended on making an audience laugh, I believe that the only accurate, objective way to measure the funniness of a joke is to measure the loudness and duration of the laughter it provokes.
So I ask, isn’t there a better way to test how important surprise is in creating laughter? For example, couldn’t a researcher take two groups of students, have a comic tell each group a slightly different version of the same joke—one version with the punch line telegraphed—and see which group laughs harder?
Or more simply, maybe a researcher could have a comic deliver a monologue to a group of students. Then, a week later, the comic would deliver the identical monologue to the same group of students. The researcher would measure whether the students laughed just as hard the second time.
Toplyn’s book is excellent.
Re benign violation, my first day writing for the Dean Martin Roast of Betty White, Dean’s head writer, the legendary Harry Crane, reviewed the material I turned in.
I still remember his comment to this day.
“Kid, your puns may garner a chuckle in Harvard yard, but when you write a gag for Dean, the punchline’s gotta be: ‘And then her ass fell off.’ Go back, lad, and give me ‘And then her ass fell off’.”
Puns alone: too benign. You gotta have a shocker.
That’s a great anecdote, Wayne. I laughed so hard my ass fell off.
Your story also makes the point that jokes should be consistent with the persona of whoever is delivering them. If someone tells a joke that doesn’t fit his or her established persona, the audience will be confused and distracted. Faced with too many violations at once, they won’t laugh as hard.