Is there such thing as a Caucasian, Asian, or African voice? And could you actually identify someone's race just by listening to him or her? Admittedly, the whole notion of race is nebulous to say the least — it's a concept that blends physiological and social differences indiscriminately. But there are still scientists trying to study the actual physiological differences between different racial groups. And one of those differences is how we sound to each other.
The moment you try to study the actual physical differences between races, you run into trouble. The data gets tangled pretty quickly because genetic makeup doesn't always match up with how you look. Sometimes the data are statistically tangled, since any physiological characteristic of a race, from height to skin color, will overlap with characteristics of other races. And when you start to look into the question of vocal differences, you run into sociological tangles as well. Humans are talented, and often unintentional, vocal mimics. We've always been unconscious mimics, which is how language spread around in the first place. So it's pretty much impossible to find such a thing as a completely uninfluenced voice.
In spite of the complications, there have been a few studies looking at race and voice, both from a physiological and a social standpoint.
Physiology of Voice
The studies on physiology of race, as it applies to voice, are few and far between. For one thing, there are the technical challenges of studying a living voice box, especially as it moves. Still, a few studies have delved into the vocal tract.
Speech therapists, for example, are helped by knowing the different physiology of their patients. So one fairly limited study looked at exactly what the dimensions of the vocal tract are in each person. Because people use everything from their throats to their noses when speaking, any people who have different facial features should have different voices.
The study used about one hundred and twenty individuals, of white American, African American, and Chinese descent, including both men and women. The researchers studied the dimensions of basically everything, from the lips to the bottom of the throat — working out exactly how much length and volume the different parts of the vocal tract gave different people. What they found were significant differences in some areas, including the inside of the mouth, the pharynx — the little pocket of space behind the flaps that come together when people swallow — and the vocal tract.
Unfortunately, the data was pretty inconclusive. Gender juggles the overall size and dimensions of the inside of the mouth. In males, researchers found an overall size gradient, with African American males having the smallest oral cavities, Chinese men having the largest, and white American males splitting the difference. Chinese women headed the list in terms of oral volume and just barely won out in total vocal tract volume, while white American women had by far the largest pharyngeal volume. African American women came in last in everything except oral volume, where they outdistanced white American women. The sound of a voice, then, doesn't just vary by race and gender, but by gender within race.
The researchers obtained these dimensions via acoustic pharyngometry. This procedure involves a computer that processes sound waves that bounce off the inside of a person's vocal tract. And different dimensions will produce different sounds. Few people, though, use the extreme edge of their vocal physiology in everyday speech. If anything, humans in a society try to sound like each other. But the different acoustics inside your mouth should have some impact, right?
The Sound of Speech
The data regarding the sound of different race's speech are even more limited than the last. There is one publicly available study, in which volunteers attempt to identify a person's race by the sound of their voice, but it involves only males, and only white American and African American subjects. A total of one hundred volunteers were asked to record vowel sounds. Those sounds were then put in pairs, and played to more volunteers.
The listeners were asked to identify which race each of the speakers on the tape were. Overall, they were successful sixty percent of the time. That's significant — but not too much more accurate than a coin flip. What's more interesting is that the sixty percent guess rate held despite the race, sex, or listening experience of the listener.
What was the difference in these vocal sounds? African American voices had greater frequency perturbation, greater amplitude perturbation, and a lower harmonics-to-noise ratio. This means that, on average, African American male voices varied in tone and loudness more than white American male voices. Is this a built-in measure of a racial difference in voice? Not necessarily. Studies have indicated that greater frequency perturbation is a good indicator of vocal cord health. In other words, it's yet another tangle.
So what have we learned? Possibly, if you cloned a number of people, raised them without any exposure to society, and forced air through their vocal tracts, they might produce different overall sounds, depending on race, and purely as a matter of physiology and acoustics. Right now, however, there's no real evidence of a characteristic physiological voice produced by people of different races. What differences there are, physiologically, are more than overwhelmed by age, health, and the deliberate use of the social voice.