Tickling makes us laugh differently

Publicly released:
International
Photo by Gabe Pierce on Unsplash
Photo by Gabe Pierce on Unsplash

The way we laugh when tickled is unique, and both machines and humans can often tell the difference. Researchers took the sound components from almost 900 laughter clips related to tickling, watching something funny, seeing someone else's misfortune, or hearing a joke, and used these to train and test a machine learning classifier. The classifier was best at identifying tickling laughter - and so were human listeners. Over 200 participants listened to 30 of the clips and correctly identified tickling-induced laughter over 60% of the time, while a second group of listeners identified such laughter as sounding less controlled.

Media release

From: The Royal Society

Tickling Induces a Unique Type of Spontaneous Laughter

A new study reveals that laughter induced by tickling is uniquely different from other types of laughter, both acoustically and perceptually. Researchers collected 887 spontaneous laughter clips from various real-life situations, including comedic performances and playful pranks. Using machine learning analyses, they found that tickling-induced laughter has distinct acoustic properties compared to other types of laughter. Furthermore, over 400 listeners accurately differentiated tickling-induced laughter from others, and rated them as involving less vocal control and being more involuntary. These findings highlight tickling as an evolutionarily ancient play behaviour and showcase how machine learning can reveal patterns in human behaviours.

Multimedia

Female nontickling laughter
Male nontickling laughter
Female tickling laughter
Male tickling laughter

Attachments

Note: Not all attachments are visible to the general public. Research URLs will go live after the embargo ends.

Research The Royal Society, Web page URL will go live after the embargo lifts
Journal/
conference:
Biology Letters
Research:Paper
Organisation/s: University of Amsterdam, The Netherlands
Funder: R.G.K. and D.A.S. have received funding from an NWO Vici grant [VI.C.221.052] awarded to D.A.S. by the Dutch Research Council (NWO). Additionally, D.A.S., R.G.K., and R.S. are supported by the Netherlands eScience Center through the Small-Scale Initiatives in Machine Learning programme.
Media Contact/s
Contact details are only visible to registered journalists.