Theory of Humor Series, part 14, Inside Jokes by Hurley, Dennett and Adams, Review

This series of blog posts began as an analysis of the comic quality of the novel A Confederacy of Dunces (Confederacy), by John Kennedy Toole. After about ten posts, I had only just laid out the theory of humor I am using to frame the analysis. In early 2018, I was invited to write a book chapter on Toole and Confederacy, so now I will only use this series of blog posts to more fully articulate that theory of humor.

In my investigation, I have read the following book, and I will analyze it below.

Hurley, Matthew M., Daniel C. Dennett, and Reginald B. Adams. (2011). Inside Jokes: Using Humor to Reverse Engineer the Mind. Cambridge, MA: MIT Press.

Admittedly, Hurley, Dennett, and Adams announce in the subtitle of this book that they are aiming for a target other than merely a theory of humor. But theories of humor do not exist in a vacuum: they are usually part of a larger theory of psychology and human nature. So this effort is not unusual. That having been said, these authors minimize the social, and especially the disparagement, aspect of the humor, although it should be noted that they don't specifically disparage it per se. Their belief, naturally their active belief, is that the exploration of humor gives them entry into a process within the human mind that helps propel general intelligence. They argue that a truly intelligent computer would need a sense of humor.

(Now, they might counterargue that they have a definition for humor more narrow than mine, so within their narrow domain, their theory holds. They might then argue that they do not try to offer a theory for the broader domain that concerns me. I don't quite buy that argument, though it has some merits. See my discussion of Question #3 below.)

The book claims to be a new theory of humor; however, to me it is largely a very sophisticated and welcome update to Incongruity Theory. For example, the authors reject the theory of frames (Minsky) and the related theory of scripts (Attardo), but they put forward "mental spaces," which are dynamically created, light-weight frames in working memory. They then argue that the brain thinks using a lazy algorithm called "just in time spreading activation" or JITSA, using those mental spaces. If one uses the computer as a metaphor for the mind, one could say that they basically upgraded from the bulky design of Windows 98 to the light-weight, threaded architecture of Windows XP. To a cognitive scientist, it might be a big difference to go from scripts in long-term memory to activated mental spaces, but for the rest of us, it is, like, big whoop.

Now, to be fair, the theory is more sophisticated than the above computer analogy. For these authors, humor is generated by the discovery of a false committed active belief that was surreptitiously introduced into a mental space. So for the paraprosdokian "I've had the most wonderful evening, but this wasn't it," the first part of the sentence leads the listener to form a mental space in which the speaker is referring to the present evening. However, that committed belief was inferred by the listener as a customary flattery. The second half of the sentence causes the mind to reevaluate that covert committed belief and discover that the speaker was not referring to the present evening after all. Snap! Because the mind has to comprehend the world on the fly, it is regularly discovering that some beliefs it committed to are in fact false. But again, to me this amounts to a refinement of the general Incongruity Theory of Humor.

The major problem with their theory is that they start with an isolated mind; whereas, humans are essentially social animals. So, despite the fact that laughter (here a Duchenne belly laugh, not a polite laugh) is an involuntary mechanism and a social signal that is clearly contagious, they somehow argue that "Basic Humor" is a solitary activity of the isolated mind doing data-integrity maintenance on "committed beliefs." (For example, read pages 130-132.) They do modestly address the value of social capital (139), but that is about as far as it goes. In contrast, I fully agree with Gary Fine (1983): "... any adequate understanding of the dynamics of humor must include a social analysis" (Handbook of Humor Research, v. 1, 159).

During their review of earlier theories, Hurley and company acknowledge that Superiority Theory is the second strongest class of theories of humor behind Incongruity Theory, and they admit that it can offer insight into a large fraction of humor data, but then they proceed to minimize it. For example, when discussing Wyer and Collins, they wave off a need for the concept of diminishment (204). Why? Because it doesn't fit their goal of designing a new (non-social) artificial intelligence system that uses a sense of humor for data-integrity. To illustrate my point, here is an example from the book where they do discuss interpersonal humor:

Person A and B were wading across a river (151). (For the sake of pronouns, I will consider B male.) B slipped and fell in. If B had been genuinely and appropriately cautious, the authors argue that his fall would not have been funny to either A or B. However, A might have laughed at B because A may have attributed B's fall "to overzealousness or overconfidence, in which [B] hubristically assumed the task was easier than it proved to be."

My commentary: The authors argue that A's mirth was generated merely by the discovery of B's false belief that the crossing would be easy. The purpose of the laughter was to signal that A should helpfully point out the flaws in B's reasoning. But the use of the word "hubristically" indicates something deeper. B didn't just commit to a false belief, he apparently deserved that dunking in A's estimation.

We can highlight this by elaborating on the fictional scenario. Maybe B was head of the expedition and was lording his position over the others. Maybe the rest of the group thought it was crazy to cross the river at that spot, but B was insisting that the group had to cross there. The dunk in the river proved B wrong to the delight of A and the rest. Ha, that B, what a pompous ass! Crow is on the menu for B.

As I have stated earlier in this series, the theory I support posits that a major function of humor within a social group is to adjust the social status hierarchy, specifically by diminishing the status of some member or members. It is a non-violent way of adjusting the dynamics of the group. In the typical group, high-status members enforce the group's rules for correct behavior, and their version of reality is the version acted upon by the group, so false committed beliefs are not usually limited to an isolated mind; instead, they are of deep concern to the whole group. For its part, self-deprecating humor allows the humorist to lower his or her own status, at times to improve the functioning of the group and group cohesion.

To further examine where Hurley and company have gone wrong, it is useful to know something about Dennett's philosophical work. Dennett calls the modern Theory of Mind "the Intentional Stance." The theory of mind is: we humans are able to perceive other beings in our universe, specifically other beings with minds, as having their own set of beliefs and intentions and mental states (143), and we can act accordingly. Children tend to develop this theory of mind or intentional stance around the age of four or five. You need a well-functioning intentional stance to be a good liar. Dennett had been promoting the intentional stance long before writers such as Alison Gopnik started promoting the theory of mind as a mainstream concept.

In this book, the authors state: "Using the intentional stance is how we manage our social lives, by modeling what other people believe" (144). That statement is at best very incomplete. We were a social species long before we evolved the ability to take an intentional stance. When I take my dog to the park to play with other dogs, it does not have an intentional stance, but it knows all about the dynamics of a pecking order, who is top dog and who is an underdog. We might manage some of our social behavior using an intentional stance, but we also manage it by establishing and negotiating status within a group, and the group status ranking system almost certainly predates the intentional stance by millions of years.

These authors claim that interpersonal humor is an "offspring" of their within-brain Basic Humor, but how we understand our world is very shared and social and always has been. To quote Penny from The Big Bang Theory, "Oh, so, you believe your friend, and your friend's wife and your own eyes over me?? Wow." I would say that emotional and psychological precursors to interpersonal humor very much predate the rich inner lives of homo sapiens with our intentional stances, complex mental spaces, and logical pre-frontal cortices. They call social play "kidding around" and "horsing around" because young goats and horses play, too.

To go back to our paraprosdokian, certainly the snap of surprise is triggered by the discovery that the "wonderful evening" was not the present evening of the utterance. But the zing that causes the laughter is generated by this new interpretation being a put-down. The speaker is disparaging the present evening, possibly breaking social decorum by challenging the evening's host, and doing so in a humorous mode that deflects a violent response. In 1974, Zilman and Bryant found that the intensity of humor was not related to the listeners sense of superiority, but was greatest when the target of the put-down was perceived to have deserved it (Journal of Experimental Social Psychology, vol. 10, pp. 480-488). So this joke would be mean-spirited if the present evening had actually been pleasant, but it would be very funny if everyone else at the party had been thinking the same thing, but they were too polite to say it out loud.

Twenty Questions

I apologize for making this review overly long, but I would like to discuss relevant details. The authors early in the book posed twenty questions that any theory would have to address, and at the end of the book, they summarize their theory by answering them. In the following section I will discuss some of those answers. What I find amusing is that they stick to their theory in some answers, but other answers address some of the shortcomings I see in their theory.

In the answers to question one and two, the authors stick to their theory. Question one is "Is humor an adaptation?" and question two is "Where did humor come from?" They maintain that, as our ancestors evolved more sophisticated brains, we had a greater need for data integrity checking, so there was selective pressure for that feature of the mind. They admit that laughter probably comes from the play panting similar to that displayed by chimps, but that it was co-opted by humor. Because chimps do not have a full Theory of Mind, "the breadth of humor that is due to social circumstance and others' perspectives (which is the bulk of humor) is lost to them, and presumably to all other species" (291). Again, the authors cannot see precursors to interpersonal humor without an intentional stance.

In question three, "Why do we communicate humor?" the authors finally offer a origin story for humor with which I can agree. "The communication of humor may have begun as a way of causing our conspecifics to know that we were only half-serious with them during mock-aggression and play" (291). I would argue that that is the origin of all humor, as a signal of playful non-aggression, which has been co-opted as a tool to negotiate status within human groups. The authors agree: "Later, laughter was co-opted for usage in more complex social circumstances, especially the mate-attracting display of intellect and the trading of social capital in various manners" (291). Thus the widely observed social functions of humor sneak in as part of laughter's "trading of social capital." This scenario to my mind is the actual origin and one of the two major aspects of humor, the resolution of false committed beliefs being humor's other major aspect. The authors finish up this section on communication arguing that the sharing of jokes is an instance of the spreading of parasitic memes. (Dennett likes memetics, the poor dear!)

At the beginning of this essay, I said that these authors may be using a more narrow definition of humor, and I think this section shows it. What I consider to be the origin of humor they limit to the origin of laughter. If one limits humor to being strictly the snap of a correction to a false belief, then much of what I consider to be humor is merely the somewhat related phenomena of using humor toward social ends. I believe this argument is wrong, because of the deep social nature of humor, especially its status challenging edge, but it is an argument one could make to reconcile the present text with my own position.

Question seven finally attacks my issues head on: "Why does humor often get used for disparagement?" I will quote their answer at length. "Putting someone down by humorously demonstrating an infirmity in their cognitive capacities efficiently makes the humorist and the addressed audience look superior in comparison, enlisting the audience as like-minded allies and at the same time making the humorist appear good natured, not just angry or aggrieved. This is a common use of humor in modern society, but not its original or even secondary purpose, which is more plausibly the demonstration of intellectual prowess (with or without a target or butt of the joke) to potential mates and allies" (292).

Notice that the authors go to great lengths to insist that disparagement is not even be a secondary purpose of humor? I would agree that it is not secondary, but would suggest that employing humor toward social ends, such as adjusting the group status hierarchy, is, if not the primary driver, than at least one of the main drivers of the evolution of humor. And non-violently adjusting social status within the group preceded humor's function for data integrity or for displaying intelligence. And I'm not just kidding around.

Question thirteen is: "Why can humor be used as a social corrective?" In their answer, the authors limit the correction to errors in logic and inference, but a major type of correction that they ignore is the enforcement of social norms. Why not? Such behavior, in my humble opinion, would inch too close to group selection theory for Dennett. Boyd and Richerson, stay away!

On the positive side, Dennett and company are proponents of applying evolutionary theory to human psychology, so at least their theory has well-thought-out evolutionary justifications. But again, Dennett's agenda intrudes. I believe he is not a fan of David Sloan Wilson's Multilevel Selection Theory, which adds group selection to evolution and could help provide the needed explanatory framework for the evolution of such a social phenomenon as humor and laughter. The 2005 paper by Gervais and Wilson, "The evolution and functions of laughter and humor: A synthetic approach" (Quarterly Review of Biology, vol. 80, pp. 395-430) is briefly discussed in this book, but its main point about group selection is neglected. So we are left with humor as a data-integrity process largely in a social deprivation tank. It is good as far as it goes, but it misses one of the two core aspects of humor.