Laughter, one of humanity's most universal expressions, could be a chuckle, a giggle, or a deep belly laugh. Published in the journal Biology Letters by the Royal Society, the study leverages machine learning and listener experiments to examine different kinds of laughs.

A Comprehensive Exploration of Laughter

The team compiled a dataset of 887 spontaneous laughs captured from real-life scenarios on YouTube. These laughs were categorized into four eliciting contexts:

• Tickling
• Humorous stimuli, such as comedy sketches or funny movies
• Witnessing someone’s misfortune
• Verbal jokes

Each video was  selected to ensure clear, unposed laughter produced by a single individual. This robust dataset formed the foundation for the study's analysis.

Distinct Acoustic Features and Listener Perceptions

Using machine learning, the researchers uncovered systematic acoustic differences between laughter types. The system could identify tickling-induced laughter with 62.5% accuracy—substantially better than chance—while distinguishing other types of laughter proved far more challenging.

Listener studies added another layer of insight. In one experiment, participants accurately identified whether laughs were tickling-induced or not, outperforming chance expectations. In another, participants rated laughter clips on dimensions such as arousal, positivity, and vocal control. Tickling-induced laughter consistently stood out, being perceived as more arousing and less controlled compared to other forms.

Evolutionary Insights

Tickling-induced laughter is deeply rooted in our evolutionary past. “This type of laughter likely originated over ten million years ago, rooted in the social play behaviors of our common ancestors with the primates. Just like chimpanzees chuckle during playful wrestling or dogs pant excitedly in a game of chase, this laughter reflects a spontaneous burst of joy, deeply ingrained in our biology.

The study highlights tickling-induced laughter's unique role in playful interactions and its enduring significance in fostering social bonds. By comparing laughter across contexts, the team provides a glimpse into the evolutionary origins of vocal expressions shared across species.