Listening for pleasure

In news sure to bring a song to the hearts of middle-of-the-road radio program managers, a new study has found that humans like songs that are familiar sounding and a little bit unpredictable.

It’s fun to consider what that might sound like.

A Drake take on “Let it Be”? Ed Sheeran snapping out a nice remix of “White Christmas”? A Taylor Swift update of “I Wanna Dance with Somebody”?

The study, published in the neuroscience journal JNeurosci, suggests that our musical preferences might be grounded in the way humans learn. 

Scientists have struggled to understand and explain why activities of apparently little evolutionary value, such as listening to music, bring so much pleasure. 

Previous studies have linked listening to music with activation in the reward centres of the brain, but they’ve produced conflicting results on how musical complexity relates to pleasure.

Lead author Benjamin Gold, of Montreal Neurological Institute, in Quebec, Canada, and his team used a computer model to determine the complexity of music fragments, and then asked listeners to rate how much they liked the songs on a scale from one to four.

Listeners preferred songs of “medium complexity”, a measure that involved features such as predictability and familiarity.

Songs that weren’t familiar to listeners were more liked if they were predictable. Familiar songs could be less predictable – they could contain more surprises – and listeners would still like them.

The study results correspond with insights into how the brain learns best: challenges and new situations of medium complexity provide the most enrichment with the least frustration.