What’s In a Voice?

The spectrum of sounds humans emit are not only the lifeblood of conversation—they are central to our sense of who we are and where we come from. In his book Now You’re Talking, acoustic engineer and BBC radio presenter Trevor Cox zooms in on the basic mechanics of our vocal abilities, broadly surveys the ways our speaking voices define and reflect identity, and asks what it means when this most human quality can be replicated by computers.

Voice-based judgments can be well off the mark. Does it matter?

Our evaluations can be quite benign, but in some cases they’re not. They often reflect typical prejudices. There is evidence, for instance, that women do better in elections if they drop the pitch of their voice a bit. And if you look at people complaining about a feature of voices, such as the creaking sound of vocal fry, it’s often women’s voices they complain about, even if men’s have the same feature.

What do our responses to accents show about how we deal with others?

I had a student from America who would complain that everyone assumed she was a kind of generic American. Of course, America is hugely diverse, but there is a tendency to compartmentalize people. Do we need to know the finer details? Since I often work in America, I do. But for most people the finer details don’t matter. So they paint with a broad brush, which actually is what a lot of cognition is trying to do: put things into categories so we know how to react to them.

You once investigated how we perceive screams. How do they fit into the bigger picture?

A scream is a rather important signal. Without it, babies wouldn’t be able to thrive because they couldn’t elicit the help of their parents. Screams, and presumably laughter, must predate speech; some of these basic signals we see in chimpanzees, which we diverged from 7 million years ago or so. When some people are really frightened, they let control of their vocal folds go, so they no longer open and close nicely. You get this roughness to the sound, a sign that you’re really screaming as opposed to just playacting.

As artificial voices grow more sophisticated, what’s one use that stands out to you?

Work on individualized speech synthesis is being done to help people who lose their voice for medical reasons. They end up selecting words off of a screen, and to be able to produce their own voice helps them maintain a little bit more of their identity. But making them individualized is hard work.

Is there a major difference between our voices and other indicators of identity?
We all make a conscious effort to cut our hair in a particular way or wear a particular shirt. You could choose to start talking in a different voice, too—we could retrain ourselves if we were really determined to—but it seems to be a bit of identity that most of us accept for what it is.

Facebook image: Pressmaster/Shutterstock