Why Do Humans Laugh? The Evolutionary Purpose of Humor

Laughter is one of those behaviors that is so familiar it barely registers as strange. But step back for a second and it is genuinely odd: a loud, rhythmic, involuntary burst of sound that takes over your breathing, sometimes your whole body, and that you cannot easily produce on command. Infants laugh before they can speak. People born deaf and blind laugh the same way sighted, hearing people do, despite never having seen or heard anyone else do it [1]. That combination, universal, early-developing, and largely involuntary, is exactly the kind of pattern that points to something with a real evolutionary history rather than a purely learned social habit.

For most of the last two thousand years, philosophers treated laughter as basically the same question as “what is funny.” That turns out to be a mistake. Laughter and humor are related but separate things, and treating them as identical has held back a proper evolutionary account for a long time [5]. Once researchers started separating the two, and started taking seriously the idea that laughter shows up in other primates and even rats, a much clearer picture of its origins started to form.

Two kinds of laughter, not one

A key distinction that gets missed in a lot of casual writing on this topic is that there are two functionally different types of laughter, and they come from different neural pathways [1]. Duchenne laughter is the spontaneous, involuntary kind, tied directly to genuine positive emotion. It originates in subcortical, limbic, and brainstem circuitry, the same general systems involved in emotional expression across mammals. Non-Duchenne laughter is closer to a learned motor skill. It originates in frontal and premotor regions, the parts of the brain associated with voluntary control, and shows up as nervous laughter, conversational laughter, or the kind of laughter used strategically to smooth over an awkward moment.

The evidence for this split is fairly striking. There are documented cases of people who cannot voluntarily produce a laugh on command but laugh normally at something genuinely funny, and the reverse: people who can mimic laughter voluntarily but do not laugh spontaneously in response to humor [1]. That double dissociation is hard to explain unless these really are two separate systems that happen to produce a similar sound.

This matters for the evolutionary story because most of everyday laughter is actually the non-Duchenne, conversational kind. Robert Provine’s observational studies of people talking in public found that laughter occurs about thirty times more often in social situations than when someone is alone, and that only a small fraction of it follows anything resembling a joke [5]. The person doing the talking laughs more than the person listening. Laughter, in other words, is mostly a social tool used during ordinary conversation, not a reaction to comedy.

Where it actually came from: play, not punchlines

If laughter did not start as a reaction to jokes, where did it start? The strongest evidence points to physical play, specifically tickling and rough-and-tumble play among young primates.

Chimpanzees, gorillas, and orangutans all produce a “play pant,” a rhythmic panting vocalization that occurs during tickling and chasing games, accompanied by a relaxed, open-mouth facial expression [1,2]. This is considered a genuine homologue of human laughter, meaning it shares a common evolutionary origin rather than just looking similar by coincidence. Since this facial and vocal pattern shows up across the great apes, its roots likely go back at least 6.5 million years, to the last common ancestor of humans and chimpanzees, and possibly further, since orangutans share it too, which pushes the estimate closer to 14 million years [3].

Tickling itself is not a random target. The parts of the body most sensitive to tickling, the neck, ribs, underarms, soles of the feet, are also the areas most exposed during a predator attack or a physical fight [3]. Tickling has been described as a kind of “mock attack,” and the laughter it produces may have originally functioned to signal that an aggressive-looking approach was actually safe and playful.

Over time, this laughter response seems to have become detached from the need for actual physical contact. Anticipating a tickle, even the mere approach of a hand, can trigger laughter before contact happens at all [2]. From there, the same response could extend to other situations sharing the same underlying structure: an unexpected, potentially threatening event that turns out to be harmless. This broader category has been termed “nonserious social incongruity” [1], and it explains a huge range of laughter triggers well beyond tickling itself, including slapstick, near-misses, and eventually verbal jokes.

The false alarm theory

One of the more elegant explanations for why laughter takes the specific form it does, loud, sudden, and explosive, comes from neuroscientist V.S. Ramachandran. His false alarm theory proposes that laughter functions to tell the people around you that a potential threat turned out to be nothing [4].

Ramachandran’s own example is useful here. Imagine hearing repeated loud thumps from downstairs at night. You tense up, your limbic system activates, you grab a flashlight and go check it out, fully alert. Then you discover the source was just a cat that knocked over a vase. That sudden shift, from a genuinely threatening interpretation to a trivial one, is what triggers laughter, and the loud, unmistakable sound of it is thought to have evolved specifically to broadcast that information to everyone nearby: stand down, there is nothing to worry about [4].

This also explains why laughter is contagious. If the entire value of the signal comes from spreading quickly through a group, then a laughing response to laughter itself would have been directly favored by selection [4]. It is also why the theory extends naturally to slapstick comedy. Watching someone slip on a banana peel and get up laughing is funny. Watching them hit their head and not get up is not. The difference is entirely about whether the anomaly resolved as trivial or serious.

A signal built for groups, not individuals

Zooming out further, several researchers have argued that laughter’s real evolutionary payoff was never really about the individual laugher at all, it was about the group. Duchenne laughter appears to function as a form of emotional contagion, mediated by a mirror-neuron-like system in the brain that activates similar motor and emotional circuits in someone hearing laughter as in someone producing it [1,5]. Hearing genuine laughter tends to induce positive affect and often more laughter in the listener, which couples the emotional states of everyone in the group and can promote cooperative, resource-building social play.

This idea connects well with multilevel selection theory, the framework recognizing that natural selection can act at the level of groups as well as individuals. Laughter imposes very little cost on the individual who produces it, but a group whose members laugh together and coordinate their emotional states through contagious laughter may function better as a unit than a group that does not [1]. Since laughter spreads socially rather than only genetically, it can create differences between groups faster than genetic transmission alone would allow, which is exactly the kind of dynamic that favors group-level selection.

There is also a comparative angle worth mentioning here. Rats, of all animals, produce ultrasonic 50 kHz chirps during play and when tickled by researchers, and these chirps share several properties with primate laughter: they show up during positive social interaction, appear more in younger animals, and decrease under stress [5]. Whether this counts as a true evolutionary homologue or simply a case of convergent evolution toward a similar solution is still debated, but it strongly supports the idea that the underlying reward circuitry involved in laughter, particularly regions like the nucleus accumbens, has deep and conserved roots across mammals.

When did this actually happen?

Piecing together a timeline requires combining several types of evidence: the phylogeny of great ape laughter, fossil evidence of bipedalism, and what is known about brain evolution and the timing of language.

Since laughter-like vocalizations exist across all the great apes, the underlying trait likely predates the split between the hominid line and other apes, somewhere between 5 and 7 million years ago [1]. But laughter as we know it today, ritualized, vocalized, and freed from the one-breath-per-pant constraint seen in chimpanzees, likely required bipedalism, since walking upright loosened the mechanical coupling between breathing and movement that exists in quadrupedal locomotion [1]. That places a major transition in laughter’s evolution somewhere between roughly 4 and 2 million years ago, well before the appreciable brain expansion associated with language, which did not begin in earnest until around 2 million years ago with Homo habilis [1,3].

That ordering matters. It implies Duchenne laughter and its associated play behavior were already fully established in the hominid behavioral repertoire before language, theory of mind, and complex culture evolved. Rather than laughter being invented alongside jokes, it looks much more like the reverse: an ancient, emotionally grounded signal that later got co-opted by newer cognitive machinery, like language and the ability to model other people’s mental states, to produce the enormous range of humor we see in humans today, from puns to sarcasm to stand-up comedy [1,3].

The takeaway

Laughter did not start with jokes. It started with tickling, play, and the practical problem of coordinating safety and positive emotion within a social group of increasingly bipedal, increasingly vulnerable hominids on the African savanna. Somewhere between 4 and 2 million years ago, an ancient vocal signal shared with the other great apes became ritualized into something distinctly human, and only much later did language, symbolic thought, and theory of mind arrive to build formal humor on top of that older foundation. The joke, in the evolutionary sense, really is the last thing to show up.

Common Questions

Is laughter always connected to humor?

No. Studies of everyday conversation have found that most laughter happens in response to ordinary comments, not jokes, and that the person speaking laughs more than the person listening. Laughter seems to have started as a social signal long before jokes existed.

Do animals laugh?

Chimpanzees, gorillas, and orangutans all produce a panting vocalization during play and tickling that is considered homologous to human laughter. Rats produce ultrasonic chirps during play and tickling that researchers consider a related, if not identical, signal.

Why do we laugh when someone slips and falls but not when they get seriously hurt?

This is central to the false alarm theory. Laughter is thought to signal that a perceived threat turned out to be trivial. If the fall is genuinely harmful, the incongruity resolves as serious rather than trivial, and laughter is replaced by concern.

Are there different types of laughter?

Yes. Researchers distinguish Duchenne laughter, which is spontaneous and tied to genuine positive emotion, from non-Duchenne laughter, which is more voluntary and used strategically in conversation, such as nervous laughter or laughter used to smooth an interaction.

References