What Laughter Really Signals (Hint: It’s Rarely About the Joke)

Why do people laugh when nothing is funny?

Because laughter is mostly about the relationship, not the material. The psychologist Robert Provine and his students spent years quietly observing naturally occurring laughter among young adults in public places. In a sample of 1,200 laugh episodes, laughter “followed both statements and questions and material that did not seem humorous outside of the conversational context.” And the person talking, especially when that person was a woman, laughed more than the people listening.¹ Looking back on this work, Provine summed it up plainly: laughter “is not typically humor related.”³

Laughter is built for company

In an earlier study, Provine and Kenneth Fischer asked undergraduates to keep a log of every time they laughed, smiled or talked over a week. All three were “most frequent in social situations,” and smiles and laughs, like talking, were “performed primarily during social encounters.” The authors drew a practical conclusion: because laughing and smiling are social acts, they are “of limited value as indices of ongoing” emotional state.² In other words, a laugh is a message to someone, not a readout of how they feel inside.

Laughter also fits neatly around speech. In the 1,200 episodes Provine studied, laughter landed in pauses at the ends of phrases or sentences more than 99% of the time, rather than cutting into the middle of a phrase.¹ It works a little like punctuation: a small vocal mark that says “this is light,” “I’m with you,” or “we’re fine.”

Can you hear the difference between a real laugh and a polite one?

Somewhat. Nadine Lavan, Sophie Scott and Carolyn McGettigan recorded nine women laughing two ways: spontaneously, at videos they had chosen, and on purpose. The spontaneous laughs were higher in pitch, longer, and contained far more breathy, unvoiced sound. Listeners labelled about 69% of the spontaneous laughs as “real” and correctly identified about three-quarters of the posed laughs as posed.⁶ But the researchers point out that these laughs had been pre-selected because they were easy to categorize, so in everyday life accuracy is likely lower.⁶

A larger cross-cultural study played 36 laughs to 884 listeners in 21 societies around the world. People told spontaneous from volitional laughter apart with an accuracy of 56% to 69%, and the sound features linked to vocal arousal predicted their judgments fairly consistently across cultures.⁵ So there is something to hear. In a single moment, though, you will often be wrong.

It also helps to remember what a posed laugh is for. Researchers describe conversational laughter as a signal of things like agreement and liking.⁶ A polite laugh is not a lie. It is a courtesy: “I’m following you, and this is friendly.”

Shared laughter says something about the relationship

In one of the most striking studies, 966 listeners from 24 societies heard brief clips, about a second long, of two American college students laughing at the same moment in conversation. Some pairs were friends; others had just met. Listeners guessed which was which correctly 61% of the time, against 50% for guessing, with accuracy across sites ranging from 53% to 67%. Even listeners with essentially no familiarity with English could do it. Their “friends” judgments tracked features associated with spontaneous, aroused laughter, such as shorter bursts with less regular pitch and loudness.⁴

That is a real signal, but a noisy one. Listeners everywhere tended to call pairs of women friends whether they were or not, and the whole study rests on one set of 48 clips recorded in 2003.⁴ The authors themselves call it “strong preliminary evidence.”⁴

Something similar shows up in close relationships. Laura Kurtz and Sara Algoe coded the laughter of 71 couples as they talked about how they first met. Controlling for all the other laughter in the conversation, the share of time partners spent laughing simultaneously was uniquely linked to how they rated their relationship’s quality, closeness and support.⁷ That is a correlation, so it doesn’t show that laughing together causes closeness; it may simply reflect it. Either way, shared laughter looks like a better cue than laughter in general.

How to use this

Where the evidence runs out

A few honest caveats. Provine’s work is observational and draws on young adults in public places.¹ The friends-and-strangers study used a single set of short clips from American undergraduates.⁴ The real-versus-posed accuracy figures come from laughs recorded in controlled conditions, some of them pre-selected for being easy to tell apart.⁶ And the couples study is correlational.⁷ None of this research tells you what one particular nervous or awkward laugh means; the surrounding context does far more of that work than the sound itself.

What survives is a useful shift in perspective. Laughter is less a review of your material and more a running signal about the relationship: are we comfortable, are we on the same side, is this moment light? If you’d like to practice reading cues like these in context rather than in isolation, our guide to reading social cues and free practice scenarios are a low-stakes place to start.