Like many aspects of human behavior, laughter is complicated.

In a recent article for Time, Dr. Greg Bryant, an associate professor at UCLA outlines a study he conducted at his Vocal Communication Lab. There, he and his research team played recorded laughs to participants and asked them to distinguish whether the laugh was “real” or “fake”. The real laughter was from live conversations between friends in a laboratory setting and the fake laughter was produced on command.

Interestingly, listeners were able to tell the “real” laughs from the “fake” laughs about 70 percent of the time. Which means 30 percent of the time, they couldn’t tell the difference. Bryant was interested in why people fell for the fake laughs.

Bryant says, “Laughter triggers the release of brain endorphins that make us feel good, and it reduces stress. There is even evidence that we experience a temporary slight muscle weakness called cataplexy when we laugh, so we could be communicating that we are unlikely (or relatively unable) to attack. But laughter is not always made in fun, and can be quite hurtful (e.g., teasing). Laughter is a powerful signal with huge communicative flexibility.

A fake laugh is produced with a slightly different set of vocal muscles controlled by a different part of our brain. The result is that there are subtle features of the laughs that sound like speech, and recent evidence suggests people are unconsciously quite sensitive to them…The ability to be a good faker has its advantages, so there has likely been evolutionary pressure to fake it well, with subsequent pressure on listeners to be good “faker detectors.” This “arms race” dynamic, as it’s called in evolutionary biology, results in good fakers, and good fake detectors, as evidenced by many recent studies, including my own.”

Dr. Bryant suggests that the reasons we laugh are as complicated as our social lives, and relate closely to our personal relationships and communicative strategies. He states that one focus of researchers now is trying to decipher the relationship between specific sound features of our laughs—from loud belly laughs to quiet snickering—and what listeners perceive those features to mean.