People can tell a lot by the sound of your voice—your mood, your hometown, and even whether you’re a friend or an enemy. Now, a group of French researchers has figured out which vocal intonations make a person sound more trustworthy or competent, using a new computer program that can transform the pitch patterns of our voices.
First, the researchers built their own voice processing software, which they used to create hundreds of random intonations of a recording of the word “bonjour”—“hello” in French—by both male and female speakers. Then, they asked two groups of about 20 volunteers each to listen to about 700 pairs of recordings; they used their responses to reconstruct optimal pitch patterns for both trustworthiness and competence.
The team found that listeners clearly associated specific intonations with each social trait, regardless of their own gender or that of the speaker, they reported this week in the Proceedings of the National Academy of Sciences.
The differences between recordings are subtle; in the lab, it almost seems that the listeners are just picking randomly, says study co-author Jean-Julien Aucouturier of CNRS, France’s national research agency in Paris. “They can’t explain [their choices], but they all have the same mental representation in their head,” he adds. One consistent feature of competent voices was that they would descend in pitch; trustworthy voices would rise quickly at the end of the word.
Because the different intonations are generated by a software (freely available to other researchers, and to anyone who wants to play with it) rather than an actor, the method could be applied to other languages or cultures. It could also be used one day to assess brain stroke patients, who typically have difficulty making social judgements based on a person’s intonation.