Why do human voices sound the way they do?

By Professor Sophie Scott
Institute of Cognitive Neuroscience, University College London

Sophie Scott (r) has been testing BBC presenter Clare Balding's voice

New research into how the human voice works and interacts with the brain is being undertaken by the University College London's (UCL) Institute of Cognitive Neuroscience for BBC Radio 4's Vox Project.

The human voice is possibly the most complex sound in nature. We use it to express ourselves in speech and song, and it conveys a multitude of clues about us.

Someone talking to us on the phone can usually tell if we are a man or a woman, have a guess at our age, know if we are unwell or in a bad mood, and identify our geographical origins.

Vibrating folds

As a sound, we start to produce speech by producing a very finely controlled stream of breath from our lungs, and this vibrates our vocal folds - membranes in our windpipe that we hold together for speech.

Just as when a saxophone player uses their breath to vibrate its reed, the vibrations of our vocal folds create a buzzing sound.

Men and women have different sized vocal folds - men's being longer and thicker. Men also have longer vocal tracts - the distance between the vocal folds and the lips (this is what happens when boys' voices break in adolescence).

The differences in how our larynxs are shaped affect how we sound Slideshow: Clare Balding's test

These differences mean that, just as a cello sounds lower than a violin, that men's voices are lower in pitch and with a wider range than women's voices.

Anatomy is not necessarily destiny in voices however, and different ways of holding the vocal folds together when we speak or sing means that people can sound breathy when they speak, or sing in a falsetto voice.

There are important cultural factors at play as well - many American speakers use a mode of speaking called "creaky voice", which is less common among British speakers.

Illness can also affect the types of sound made at the larynx, due to problems with breathing or inflammation of the colds themselves (remember that episode of Friends when Phoebe had a cold and was delighted with her "sexy voice").

The buzzy sound produced by the vocal folds has a pitch, and we can vary the rate of vibration, and raise or lower our vocal folds to vary this pitch.

Within any one person, their voice will change a lot, not just to do with mood or health, but also based on their social setting

In all the spoken languages of the world, pitch is used to give our speech melody, inflections and emphasis, and it also plays a large part in how we express emotion in our voices. The control and variation of pitch is also of course central to song.

We shape the buzzing sound made by the vocal folds by how we move our lips, tongues, jaw and soft palate to make the sounds of speech.

If the vocal folds produce the music of speech, these articulators impose a linguistic signal upon this music.

Try saying "ah, or, ooo, ur, eee" aloud while not changing the pitch of your voice, and see how the ways that you move your mouth lips and tongue change the timbre of your voice while the pitch stays the same.

It is this shaping of the sounds that gives human speech much of its complexity.

Prof Scott says the human voice could be the most complex sound in nature

Difference in the anatomical structure of our larynx, vocal tract and articulators lead to different acoustic characteristics in people's voices.

In the same way that two violins might sound different due to minute differences in their shape, or the wood they are made from, two different people can sound very different due to the way their articulators are shaped.

There are also large differences between people in how we move our articulators to speak: some people speak with quite clenched teeth, or with lots of nasality.

'Regional accents'

Furthermore, when we talk about differences between regional accents of English, we are talking about systematic differences in how people produce certain speech sounds (for instance, the sound in the middle of the word butter is pronounced as a "d" sound by many North American English speakers, and as a glottal stop by many London speakers).

There are also big differences in how people pattern the music of speech that can also vary with regional accent.

It is also important to note that within any one person, their voice will change a lot, not just to do with mood or health, but also based on their social setting and how they perceive themselves as within that.

This change can be something which we have more or less control over.

For example, I am aware that I sound very different when I am buying fruit in a market than when I am giving a lecture, but I also apparently sound very different when I am talking to my mother, and I am not aware of that at all.

This variability means that, for example, it has proved hard to automate computer systems to recognize voices for forensic analysis.

So the complexity of the sounds of our voices is matched by the complexity of their use. When we speak, our voices reveal a great deal about ourselves, over and above the messages that our words conveys.

Clare Balding and Sophie Scott present The Vox Project series on BBC Radio 4 in January.

The UCL team is collecting audio from the public for the research here.