by Scott Weems
Consider the fact that your brain has between 10 billion and 100 billion neurons. That number is so large it’s meaningless, so let’s compare it to the number of people living on Earth, which is roughly 7 billion and growing. That’s pretty close to the lower estimate for the size of your brain, so let’s consider these two systems further. Imagine that at this very instant the entire population of the United States decided to scream. That would be comparable to the neural activity in your brain—at rest. A brain not at rest would be ten or a hundred times more active, so now it’s not just America screaming but all of Asia too. All it takes is for multiple parts of your brain to disagree, and pretty soon you’ve started World War III.
The brain manages this complexity the same way humans do—by building “governments.” As I’ve noted, its various regions are specialized for nearly everything we do, and although nobody knows exactly how many specialized modules the brain has, it’s probably close to the number of governments in the world. There has to be a way to manage all these voices, and for humans the solution is to create entities like the United Nations. The UN isn’t itself a government, as it has no land, economy, or political goals. It simply keeps an eye on everyone around it, hearing complaints and keeping troublemakers in check. The brain has a UN, too. It’s called the anterior cingulate.
Located near the center of our brain, just above the corpus callosum connecting the two cerebral hemispheres, the anterior cingulate is in a perfect position to oversee the rest of the brain (see Figure 2.1). In front is the frontal lobe, our primary reasoning center and the region responsible for initiating movement. Behind are the parietal and temporal lobes, which help with reasoning, as well as language and memory. And as part of the brain’s limbic region, the anterior cingulate is closely connected with the amygdala, the nucleus accumbens, and the ventral tegmental area—regions that, as noted earlier, are key to the dopamine reward circuit.
FIGURE 2.1. Selected regions of the human brain.
We know that the anterior cingulate is especially important for insight because we can observe activity in subjects’ brains prior to solving problems like our word triads. Most parts of the brain become less active as subjects prepare to solve these difficult problems, but the anterior cingulate is different. It becomes more active, because rather than coming up with solutions, it handles conflict. The Remote Semantic Associates task doesn’t appear at first to be one driven by conflict, but it is. As we already discussed, the solution is seldom the first one anybody thinks of. Coming up with a solution requires “holding back” more potent responses. The part of the brain that thinks it has the easy answer needs to be “shut up” so that softer voices can be heard. And telling others to shut up is exactly what the anterior cingulate is good at.
A good way to understand the anterior cingulate is to explore the Stroop task, named for John Ridley Stroop, who developed it in 1935. He found that when we are asked to identify the color of something, we are slower and less accurate when that thing is a color word. For example, it’s easy to identify the color of four asterisks printed in red, but much harder to provide a correct response when the items printed in red are the letters B-L-U-E. Why? Because now we have two competing responses. The human mind naturally wants to read, and preventing it from doing so is almost impossible. If you don’t believe me, try performing this simple experiment at home: watch an English-language movie tonight with the subtitles showing. I guarantee that you’ll be reading every word on the screen, even though you understand exactly what is being said.
What does this have to do with the anterior cingulate? Well, the Stroop task is exactly the kind of thing the anterior cingulate is specialized for, because it’s the only brain structure able to keep the reading regions silent so that the color-identifying regions can respond. And it’s especially effective at managing such control when we’re in a good mood, which is why the Stroop effect disappears when we’re happy. When subjects are asked to recall positive life events such as vacations or birthdays immediately before performing a Stroop task, they no longer have difficulty ignoring the conflicting color words. Just as insight is positively correlated with happiness, happy people are better at maintaining focus while identifying the color of fonts.
Mood and happiness are also important to constructing, as we’ll soon see. As the brain constantly debates over what to say or do, the anterior cingulate stays very busy, and so anything that provides help can have a big influence. Positive mood improves focus by helping the anterior cingulate hold back unwanted responses, such as ache in the insight problem and B-L-U-E in the Stroop task when the font is red. If the anterior cingulate is like the UN, then positive mood is its operating budget.
However, it’s important to realize that we are not passive actors in our environment. We don’t just take in information, we create it. We are constantly developing theories and expectations about our surroundings, then revising them when necessary. This phenomenon is observed not just in laboratory settings but on the scale of whole societies, past as well as present. Our ancestors interpreted lightning as anger from the gods, and eclipses as dragons eating the sun. Such beliefs invaded science too. Aristotle, who was born almost two thousand years before the invention of the microscope, thought that life spontaneously arose from slime and mud exposed to sunlight, because this was his only explanation for mold. And Isaac Newton, who lived during an age when chemistry offered no explanation for the mysterious existence of gold, wrote more than a million words about the subtleties of alchemy.
To show how ubiquitous our need is to construct such theories, and also to see its link with humor, let’s consider one last study before moving on to our second stage, reckoning. This study was conducted by the Swedish psychologist Göran Nerhardt, who wanted to know if he could induce laughter using materials that weren’t funny at all. He didn’t even tell his subjects that they would be participating in a humor study. Instead, he simply instructed them to pick up a series of objects and waited to see just how far their false expectations could take them.
Nerhardt’s task was quite straightforward. Subjects picked up objects of various weights (e.g., between 740 to 2,700 grams, roughly 2 to 6 pounds). Then they were asked to classify each on a 6-point scale, ranging from very light to very heavy. This sequence was repeated several times, after which subjects were given an object that was much lighter than the others—just a tenth of a pound. They weren’t told anything special about this last object. They were simply asked to make a series of judgments about items that weren’t funny at all.
Yet Nerhardt found that, when the subjects were asked to make a judgment about the final, wildly incongruous weight, most of them laughed despite being given no indication that it was intended as a joke. Not only that, but the more the weight differed from those lifted earlier, the more they laughed at this absurdly light one.
In the forty-plus years since this experiment was first conducted, the design has been varied several times, and each time the subjects’ reaction is the same—they find the last, incongruous weight funny. There’s nothing humorous about the weights themselves. The subjects simply have to construct an expectation. And when that expectation proves false, they have no other choice but to laugh.
Reckoning in a Confusing World
Now that we’ve explored the concept of constructing, let’s see what our brains do with all these wild expectations. Only by observing the consequences of our false starts can we understand why they so often lead to humor. This means familiarizing ourselves with reckoning, the jettisoning of our mistakes so that we can uncover new interpretations.
My guess is that if you asked a hundred experts what the key ingredient of humor is, most would say surprise. Surprise is special because it affects us in so many different ways. It’s what makes insight problems unique, because for these tasks we have no idea how close we are to a solution until we already have it. That’s what defines insight problems. Research by Janet Metcalfe at Indiana University showed that
confidence in being close to an answer for insight problems is inversely related to actual progress. In other words, the closer we think we are to a solution, the farther away we really are. Surprise isn’t a by-product of completing these tasks, it’s a requirement.
Surprise is important for humor the same way it’s important for insight—we take pleasure in being pulled from false assumptions. Punch lines catch us by surprise, and the more we set our expectations on one interpretation, the more we allow ourselves to be caught off-balance by the actual turn of a joke. A joke that you’ve heard before isn’t inherently less funny. It’s just old news, and so it no longer gives you surprise. An insight problem that you’ve seen before isn’t fun or challenging either, because you no longer need insight to solve it. You just need a bit of memory.
Reckoning is the process of reevaluating these misperceptions, usually leading to a pleasant surprise. We enjoy discovering our mistakes because surprise is one of our most valued emotions, as fundamental as happiness or pride. Scientists have even quantified the importance of surprise by asking people about recent emotional experiences. This is what Craig Smith of Stanford University did when he asked subjects literally thousands of questions regarding recent events in their lives, questions like “How pleasant or unpleasant was it to be in this situation?” and “When you were feeling happy, to what extent did you feel that you needed to exert yourself to deal with this situation?” Using advanced data analysis, he was able to locate the subjects’ emotions along certain dimensions, including pleasantness and the amount of effort they required from the person experiencing them. Figure 2.2 shows how surprise ranked, compared to other emotions.
FIGURE 2.2. Emotions as they vary by pleasantness and effort involved in their experience. Adapted from Craig Smith and Phoebe Ellsworth, “Patterns of Cognitive Appraisal in Emotion,” Journal of Personality and Social Psychology 48 (1985): 813–838. Published by the American Psychological Association. Adapted with permission.
As it turns out, surprise holds a special place, near the top of the diagram. Since the axes measure pleasantness and effort required for their experience, this means that surprise is one of the most positive and natural emotions we experience.
Surprise leads to pleasure in lots of contexts, not just humor. German psychologist and art theorist Rudolf Arnheim presented perhaps the most graceful example of pleasant surprises when he analyzed, of all things, a violin sonata by the Baroque composer Jean-Marie Leclair. Leclair, who wrote nearly a hundred major works in the mid-eighteenth century, was well known for creating sophisticated, cerebral violin concertos. In one of his last works, there’s a point near the middle where he suddenly includes a note that is harshly out of key. At first it sounds dissonant, and the listener wonders if perhaps there has been a mistake. But the same note occurs again, and then another surprising note, and soon we realize that the composer has switched keys in the middle of the performance. An examination of the music in written form reveals that the change is entirely intentional—a note written as B flat is identified as A sharp later in the same measure, conveying Leclair’s message that it serves different purposes for the old and new keys. In just a few notes the listener is compelled to discard previously held assumptions about the piece and to listen to it in an entirely new way. And the experience is richer for it.
Arnheim explains that such sudden shifts occur in architecture too. Take, for example, the Hôtel Matignon, the Paris mansion designed in 1725 by architect Jean Courtonne that now serves as the home of the French prime minister, Jean-Marc Ayrault. At the time it was built, tradition dictated that buildings be built symmetrically about an axis connecting the front and rear entrances. But this was impossible for the Hôtel Matignon given the surrounding streets, so the architect did the only thing he could—he shifted this axis inside the building itself. Visitors entering either entrance see everything laid out in the expected, symmetrical fashion. But further on, there’s a point where everything suddenly shifts and they’re off-center relative to the entrance they used, and are now centered about the opposite one. Some call it cheating, others call it brilliance, but everyone appreciates that this shift is what makes the building so pleasurable to live in—including its current resident.
These phenomena have an equivalent in the realm of humor, and it’s called paraprosdokia. Paraprosdokia is speech that involves a sudden and surprising shift in reference, usually for comedic effect. Take, for example, the following quote by Stephen Colbert: “If I am reading this graph correctly, I’d be very surprised.” Colbert was looking at polling data for the 2008 presidential elections—data that under even the best of circumstances would be difficult to interpret. At first it sounds like he’s preparing an insightful and cutting remark. Instead, we realize he’s basking in the ignorance we all feel when trying to interpret such numbers. The joke required no setup or punch line. All it needed was for the listener to “jump the gun” regarding what Colbert was actually saying.
Not surprisingly, the brain region responsible for catching these false starts is the anterior cingulate. We know this from studies like the one conducted by biologist Karli Watson of the California Institute of Technology, who wanted to see if any particular brain region was especially important for surprise. To do that, she showed subjects cartoons while they were monitored using an MRI scanner, and (as in previous studies) she made sure that some cartoons were funny whereas others were not. As an additional manipulation, she varied the nature of the cartoons so that some relied on sight gags whereas others depended on captions and language. Variations like this can have big impacts on how the brain responds, since visual centers are very different from language ones—so she expected the jokes to enlist entirely different regions. But were any regions activated in common?
The answer, of course, was yes. Both the dopamine centers and the anterior cingulate were active for each kind of joke. Not only that, but the funnier the jokes, the more engaged was each subject’s anterior cingulate.
Studies like this provide a great example of reckoning because they show that what elicits laughter isn’t the content of the joke but the way our brain works through the conflict the joke elicits. This can be seen in Colbert’s quip as well as in Leclair’s violin sonata and Courtonne’s Hôtel Matignon. We take joy in recognizing our mistakes. Though we often think of punch lines as involving misdirection, it’s actually our anxious brains that supply the false interpretations. There were no dissonant notes in Leclair’s sonata, just as there was no actual contradiction in Colbert’s one-liner. The enjoyment of both comes solely from overriding a false expectation created within ourselves. In this way, reckoning builds on constructing by forcing us to reexamine false expectations.
To see how all this eventually turns into a joke, let’s finally explore the concept of resolving.
Resolving with Scripts
A large woman sits down at a lunch counter and orders a whole fruitcake. “Shall I cut it into four or eight pieces?” asks the waitress.
“Don’t cut it,” replies the woman. “I’m on a diet.”
Is this joke funny? Unless you have a special affinity for fruitcake humor, your answer is probably no. But at first glance it seems like it should be, because the woman’s response is definitely surprising. It’s so surprising that it makes no sense at all. Consider, then, this alternate ending:
A large woman sits down at a lunch counter and orders a whole fruitcake. “Shall I cut it into four or eight pieces?” asks the waitress.
“Four,” replies the woman. “I’m on a diet.”
Now is it funny? Again, you probably didn’t laugh out loud, but I bet you at least found it funnier than the first version. The reason is that this second version provides an explanation for the sudden shift in perspective. It isn’t enough just to introduce surprise in a joke; we must also provide a shift in perspective. I call this third stage of the humor process resolving.
When studying humor, we need a way to characterize the expected and
actual outcomes of a joke. For our fruitcake story, we see there are several words signaling an expectation of gluttony. There’s the fact that the woman orders a whole fruitcake, not just a slice. She’s also described as large. All this background suggests that she’s really looking forward to the cake. When she asks for four slices instead of eight, one interpretation—the one influenced by her weight—is that she thinks four slices means fewer calories. The other interpretation, the right one, is that strokes of the knife have nothing to do with calories or the amount of cake.
Pretty tedious, huh? After such an analysis, it’s clear why dissecting humor is often likened to analyzing a spider’s web in terms of geometry. It loses its grace.
I apologize for breaking down such a bland joke, and I promise not to do it again. But it’s important to recognize that joke construction is complicated. To compare contrasting meanings, we need a scientific way to characterize all the false assumptions involved in the joke. We need a way to measure distances between intended and unintended meanings to get an idea how funny a joke can be. And, perhaps most importantly, we need to understand why people laugh at some incongruities—such as a woman thinking four large slices of fruitcake are healthier than eight small ones—when much bigger incongruities—such as a woman walking into a diner and ordering an entire fruitcake—are seemingly ignored. To do that, we need to understand scripts.
After graduating from the University of California with a PhD in psychology, I initially worked as a postdoctoral researcher with computer scientist and neurologist James Reggia. I was excited to work with Reggia because he was interested in nearly everything. He studied not only hemispheric laterality (my own specialty) but also language and memory. He specialized in artificial intelligence and chaotic swarming, an emerging field that uses artificial life to examine large-scale problem spaces. He even taught classes on machine evolution and expert systems. In short, he was the kind of person who knew something about nearly everything. So, when we first met in a restaurant in Columbia, Maryland, his first words to me were a surprise.