Regular concert-goers are used to seeing singers use expressive and even very dramatic facial expressions - that's because it works.

Music and speech are alike in that they use both facial and acoustic cues to engage listeners in an emotional experience and so  a team of researchers at McGill University wondered what roles these different cues played in conveying emotions.

To find out, they did an experiment where participants were offered recordings of short phrases (neutral statements, all of which were seven syllables long, such as "children tapping to the beat" or "people talking by the door"), which were then spoken or sung with a variety of emotions.

Participants were then offered these recordings in three different formats: either audio alone, video alone (with no sound), or full audio-video recordings, and were asked to identify the emotions that the performers' intended to convey. 

The researchers discovered that when it came to song, although the participants had a hard time recognizing the emotion based on the audio recording alone, once visual cues were added, the observers' understanding of the emotions the music sought to convey improved dramatically.

In contrast, participants were much better able to recognize emotion in speech, whether they were listening to audio alone, watching a video without sound, or seeing both at the same time.

As a result, the researchers believe that visual cues play a much more important role in the understanding of the emotions being conveyed by music than they do in the understanding of speech.

The research was funded by The Natural Sciences and Engineering Research Council of Canada (NSERC) and by the Canada Research Chairs Program.

Citation: The Quarterly Journal of Experimental Psychology, DOI:10.1080/​17470218.2014.971034