Most of us take our voices for granted, but New Yorker writer John Colapinto got a scare several years ago when his failed him.
After working every day, mostly in silence, he damaged his vocal cords while singing with a rock band in the evenings after work.
"I would jump up at the end of the day, take the subway to our rehearsal space and then just start wailing over there, cranked up guitars and drums. I mean, zero to 60 with my voice, just crazy," he says. "Anyone that knows anything about proper singing knows you don't do that."
Colapinto's voice became raspy as a result of a polyp — a bump on the edge of his vocal cord. In his new book, This Is the Voice, he describes his own injury and journey exploring the complexity of human speech.
The term "vocal cord," he learned, can be misleading.
"Our vocal cords don't produce sound like a violin string or a guitar string," he says. "In fact, it's a valve at the top of our throat that, under pressure from the lungs, vibrates in this way that chops the air into these pulses."
Colapinto says even the simplest of sentences — like "give me the salt" — also requires a complex set of the vocal "gymnastics" of the tongue and throat that relies on timing and gives rise to a sort of music that is human speech.
"We tend not to think of how remarkable that is," he says.
On meeting a woman in an elevator who alerted him to his voice injury
I said to her, "What floor?," as one does in New York — you're going to push the button for someone. And [with] those two words, she said, "Oh, you've got a serious voice injury." And I said, "Oh, it's nothing. It'll clear up." And she said, "No, no, no. I work with Broadway singers." ...
She read me like a book. She said, "I bet you get kind of tired at the end of the day, because you're using all your muscles. You're having to work harder with your back and your abs and your hip flexors, all of these muscles we use in order to actually push the air out when we speak." She even said, "I bet your neck gets pretty sore." ... It was amazing.
But the last thing she said to me was, "You should at least get a laryngologist to look at that, because it could be something else."
I grew up in a medical family. "Something else" is the approved euphemism for cancer or some kind of dangerous growth. So I immediately made an appointment with one of the top vocal surgeons in the world. ... He looked in my throat ... and said, "You've got a pretty major polyp."
On how vocal polyps develop
This is one of the mysteries of the voice. We still know so little about it, but as far as doctors understand, these polyps really begin with bleeding within the vocal cord itself, which is effectively a bruise. And if you bleed without any stanching of that bleeding, you can develop this scar tissue in this bump. ...
If you have a pure, clear singing tone like Barbra Streisand, let's say, you indisputably have very clean edges to your vocal cords. They meet flush; they chop the air really nicely. If you've got a lump on the vocal cord like me, you get all this extra turbulence. You get air passing through in funny ways, and you get rattles and all of these other things. The vocal cord is also burdened with this extra mass on it. So I cannot as effectively and smoothly change the pitch of my voice — because we really do that by tightening or slacking the tension on the vocal cord, and that's been kind of messed up for me, too.
On the delicate surgery to repair vocal cords and remove polyps
What [surgeon Steven Zeitels] does is he gets the person lying on their back, puts them unconscious, and also introduces a paralyzing agent to the entire body, because the slightest move while he's operating on a vocal cord will only make matters worse — because he'll nick healthy tissue or something. But the person's lying on their back and their head is tilted back. And then this instrument is introduced down the throat that pries the vocal tract open. The weight of the body is used to kind of pry the whole vocal tract open, and he then reaches down the vocal tract, the throat, with these tiny scalpels and little devices for plucking at the flesh, and he's got them in both hands, and they're on, like, long knitting needle extensions that go down the throat. And he's looking through a stereomicroscope that's aimed down there with a little light on it. And he's making these meticulously tiny movements with his fingers to do this astonishingly delicate surgery on these vocal cords. ... The patient has to remain silent for six weeks afterwards so that everything can heal on its own. ... It really is this exceedingly delicate surgery.
On his decision not to get the surgery, and how it affects his life
I really can't sing properly anymore with anything like the singing voice I used to have. I go off pitch and so on. But it was really in my speaking voice that [Zeitels] said, "You've got a lot of trouble there, and you don't know it, but you are doing all sorts of stuff to speak around your injury."
And what he meant by that was I was dropping my voice into a pitch where the part of my vocal cord that would still chop the air cleanly would do that. But it meant that I was, to a degree, removing what linguists call the prosody from my speech.
And what that is, is the musical up and down — the melody that we actually fit all the lyrics of our speech to. We don't really realize we're doing it until we hear someone talk like that. And you cannot listen to someone talk like that. It's boring. It's irritating and you lose focus on it.
We're actually performing a beautiful singing performance in everything we say, from answering the phone when we say "hello" and to longer sentences. ...
There's another type of prosody called emotional prosody — which is almost unconscious or is unconscious. And that's how we ride the music to kind of suggest happiness, sadness — even sarcasm, irony. Well, Dr. Zeitels said to me, "Because you're dropping your voice into this more monotonic range, you're really removing the emotion from your voice, or some of it." That hit me hard.
On AI engineers working to make computers sound more human
Maybe 10 years ago, the big tech companies realized, OK, we can kind of make our computers speak the language part, but they don't sound like humans because there's no emotion, there's no prosody, there's no melody. Well, the way to get that in, they've discovered, is also machine learning.
So they take samples of human speech, which they have to label "anger," "fear," "joy," and they play it into the computer's massive amounts of data. The computer listens to this stuff and sorts it out and begins to assign particular emotions to those melodies in the speech. Computers are now almost as good as humans at hearing the broad emotions in our speech and to be able to produce it in synthetic speech, computer speech. So we are heading into, really, a brave new world where we are really not going to be able to tell that it's a computer that's speaking to us.
Sam Briger and Seth Kelley produced and edited this interview for broadcast. Bridget Bentz, Molly Seavy-Nesper and Deborah Franklin adapted it for the Web.