News

Smile -and the world can hear you

Wed, Jan 9, 2008

Smiling affects how we speak, to the point that listeners can identify the type of smile based on sound alone, according to a study by scientists at the University of Portsmouth.

The research, which also suggested that some people have "smilier" voices than others, adds to the growing body of evidence that smiling and other expressions pack a strong informational punch and may even impact us on a subliminal level.

"When we listen to people speaking we may be picking up on all sorts of cues, even unconsciously, which help us to interpret the speaker," said lead author of the report, Amy Drahota.

Drahota, a research fellow in the School of Health Sciences and Social Work at the University of Portsmouth, recorded interviews with volunteers that required them to respond, "I do in the summer," no matter the question. Examples of questions included, "Do you ever sunbathe?" and, "Do you go skinny dipping?"

"The interview was deliberately built up to begin serious and then become gradually more amusing and strange, and potentially slightly embarrassing," Drahota explained.

"All the while the speakers were 'admitting' what they do in the summer, even if it wasn't true, which made the interview very bizarre to the speakers which would have made them smile."

Researchers videotaped the volunteers and then categorized their smile types. It's believed that some 50 different types of smiles exist, ranging from triumphant ones to those that convey bitterness. For the purposes of this study, however, the scientists focused on four types.

Drahota described the first as an open smile "in which the lips are drawn back, the cheeks are raised and crows-feet wrinkles appear around the eyes." Technically this is called a Duchenne smile, which may be the truest and most intense of all.

The second smile type is like the Duchenne, only minus the "smiley eyes." The third is a suppressed smile, "where the speaker is trying to hide their smile by pulling their lips in or down as they speak." Finally, they denoted times when the speakers weren't smiling at all.

The audio for the interviews was then played back to another group of test subjects. Even without seeing the speakers, the listeners were able to hear the different types of smile the speaker made as he or she went through the wacky interview.

“A voice contains a variety of acoustical characteristics” said Drahota. “It’s possible that we interpret these ‘flavours’ in someone’s voice almost without noticing.”

The research may be useful for further work on computer generated speech programmes, computer games, automated telephone systems, and embodied conversational agents which are used commercially- for example as information providers in information kiosks, Drahota suggests.

Drahota is the co-author or a report with colleagues Alan Costall and Vasudevi Reddy which will be published in the journal, Speech Communication.