Theory of Humor

First, two explanations I heard so far (in my own words, but I am trying to be correct):

The general line here is: joke = humor = laughter

Sigmund Freud:
One can't express aggression and sexual drive directly, as it is prohibited in the society, so these desires get sublimated in telling "jokes". If you look at jokes, they are either about somebody getting hurt, or they have sexual connotations.

Marvin Minsky:
There are not only general social prohibitions. There are also things your mother told you not to do - like stick your finger into your eye. So when you tell a story about something stupid, you attack the rules of common sense, in a safe and socially acceptable manner.
So Minsky adds jokes about stupidity. Freud's advocate could explain jokes about stupidity as manifestations of aggression against other people - stupid things *you* do are never funny. I wouldn't call it a strong argument though. Minsky also shifts attention from urges to social prohibitions a bit; this raises the level of intelligence in the reasons, and includes jokes about stupidity in an integrated manner.

There are some things though that both of these explanations do not account for:

  1. The biological origin of humor. The above explanations make humor appear as a function that is either programmed culturally, or if biological, then very recent - after the development of language, at least. These explanation suggest no ties with previous evolutionary development and pre-speech mental mechanisms.
  2. Social role of humor Seeing how much joy (emotional reward) people derive from humor, one could assume that there should be more serious social reason why such behavior would be rewarded, then letting out steam of urges.
  3. In many cases, people are ready to openly express more aggression, sexuality, and disagreement with authorities that they are suggesting in the jokes, so their jokes can hardly be viewed as a suppressed revolt. Peaceful people and innocent children find lots of things funny; children find things funny (such as peek-a-boo) that adults don't and wouldn't teach them; there is hardly any evidence that people with strongly suppressed anger or sexuality have more interest in jokes than people who do not have these interests, or feel free to express them.
  4. Most of references to sex, violence and stupidity are not funny. Let me try a few non-jokes:
  5. Are you laughing yet? Why not? These sentences contained all suggested ingredients of jokes. Maybe, these levels of references are socially acceptable? I think so, but we couldn't make these sentences funny by making "forbidden" references more explicit. Maybe they were too brief? Then imagine how much you'd laugh at a research paper on prostate cancer and corresponding mortality rates. Or maybe, some crucial ingredient of humor just wasn't there? Then what is that missing ingredient?

  6. There are lots of things we consider funny that do not have anything to do with sex, violence or stupidity. The above theories offer no explanations.
  7. Why do kids consider peek-a-boo funny? Why is it funny when I pull a pig out of my pocket during a conversation?

How about the following jokes: (bear with me, it's for scientific purpose)


  • "Why does the giraffe have such a long neck?" - "because its head is so far away from its body that a long neck is simply necessary to reach it"
  • "What is in common between a plum and an elephant?" - "they are both purple, except for the elephant."
  • "why didn't the skeleton cross the road?" - "Because it didn't have the guts".
  • "Two wrongs don't make a right, but three left turns do."
  • And a few quotations:

    I hope you found at least some of these funny. Now, how many references to sex, violence and stupidity did you see there?

    Enough of criticism. Now I want to suggest an explanation of humor that explains all of the above cases and has (or so it seems to me) apparent biological roots and social utility.

    The keyword here is "surprise". Or, "twist". Something unexpected. Something that breaks the rules - not the social rules! - but the rules of logic, of common sense, breaks your expectations. Like a pig pulled out of a pocket. Or the peek-a-boo where a child finds the transition [now you see me - now you don't] amusing, but adults, for whom it is not new, don't. Or a "punch line" that adds a twist that all the joke was built for.

    It is programmed into us biologically, to look for all kinds of exceptions in the outside world: changes in the level of signals, something suddenly appearing, blinking, bursting, jumping, etc. This attention to surprises is a result of millions of years of evolution, and can be traced from bacteria to humans. It is natural to expect that increasingly intelligent organisms would pay attention to increasingly complex surprises - including those that challenge their internal models of the world by suggesting unexpected connections between different ideas and interpretations. I heard a story about a dog that had its favorite joke, but my model example here would be closer to modern humans - e.g., a group of young Neanderthals. The first and most important common project of humans was joint construction of the mental models of the world. The cooperation went through the language, by sharing facts for building models, passing models that seemed right - AND sharing unexpected twists that either challenge the models, or help define limits of their applicability, or teach when [not] to use them, or just train your brain on amusing puzzles.

    Neanderthals were too stupid (just as modern humans) to consciously seek and share semantic surprises "for betterment of the common knowledge base". Built-in neural loop producing pleasure upon detection of semantic twists helped a lot.

    So, "funny" is an internal reaction to humor; it produces pleasure. Smile and laughter are social expressions of pleasure - but not just from humor. We smile or laugh when we see little kids play, or win lottery, or think of something nice, or are just happy. Humor is just one of the inputs for this reaction. When we tell people things, we try to make sure that all their elements are interesting, so we include twists, references to sex, things they may like to hear or see or things their enemies wouldn't like to see. Analogously, when we have a party, we provide space, drinks, light, heat, music, food, chairs, etc. This doesn't mean those are the same thing - they are different, have different reasons, and under other conditions may not be present together.

    In the course of history, people learned to abuse the natural pleasures they used to receive from simple natural urges and sensations. Pleasures we derived from sweets, fats, watching rapidly moving objects and noticing simple contradictions and little logical surprises that used to be important for finding better food, avoiding predators, and learning new things, have long been artificially satisfied with donuts, videogames, and jokes. If we want to understand their nature though, we should not play with current cultural artifacts, but look into the original reasons why these things were developed. And their roots were all developed for one purpose - to help the animals navigate the environment.

    Unfortunately, humans are still wired as if they lived in the jungle. They could understand the environment a lot better if they studied more science and read some good philosophy - but they are still trying to get pleasure exercising their reactions to [simulated/perverted] fast-moving game objects, or sharing silly little twists of logic. Many people realize quite well that these things are a waste of time, but they still can't help it: the atavistic urges are still running us. Sometimes people try to combine "learning and fun", but this requires a lot of tricks, as most natural implementations of "fun" have lost their intended utility quite awhile ago.

    This also explains why there are so many more freelance comedians than philosophers. And why inventors of good videogames are much richer than authors of great theories. It's because the comparisons of their values are performed by neural networks taken from animals.

    I personally find some bitter irony in this situation.


    other essays - home page