Drop the FryJudging Elizabeth Holmes’s Voice Comes down to Bias

Elizabeth Holmes, a young white woman with her blond hair in a ponytail, stands in a black turtleneck in a medical lab

Elizabeth Holmes (Photo credit: HBO)

One of my first jobs was at a men’s magazine. I was the sole woman in the office, and one of my editors told me the only workspace available was at the front desk in the lobby, away from the rest of the staff. It’s the only desk left, sorry!” (Spoiler: He was lying.) “And you wouldn’t mind answering calls and grabbing packages that come in while you happen to be sitting there, would you?” “Of course not,” I remember saying. I was young. It was one of my first paid writing jobs, and maybe this was just how things worked. As much as I enjoyed my work and my coworkers, I felt like a glorified receptionist for the remainder of the summer. If I wanted to eat lunch away from my desk, editors would imply that one of my male coworkers couldn’t fill in. It had to be me (“You’re just so good at it!”).

One delivery man would, whenever he came to the office, critique everything from my facial features to my chipped nail polish to the attractiveness of my outfit. Regular callers would inform me that my voice wasn’t “perky” enough, and I eventually stopped correcting them for referring to me as a secretary. And though being a secretary isn’t a shameful profession, of course, I felt ashamed to go from being in editorial meetings where I was treated like “one of the guys” to signing for packages and being called “honey.” According to a 2016 U.S. Department of Labor survey, more than 94 percent of secretaries and administrative assistants are female. In fact, secretarial work has remained the top job for women since the 1950s, when companies realized they could pay women—who were generally less educated—less to do administrative work.

Women are still playing that role in ways we might not even realize. Amazon’s Alexa, Apple’s Siri, Microsoft’s Cortana, and Google Assistant have one thing in common: They all have female-identified voices and names. You can command Alexa to play a song and then stop and engage in dialogue with Siri, whose rudimentary responses are catered entirely to you. Subway announcements, GPS navigation systems, and phone operators are among the technologies that typically use a female persona. In a recent interview with Business Insider, an Amazon executive said that people “opt for a woman to assist us” because her voice is more “pleasing,” “sympathetic,” and “better received.”

Though Alexa’s name is meant to reference the Library of Alexandria, there’s likely a reason the company chose Alexa over other names, like Alex, Lex, or a simple Al. Siri is named after the Norwegian word for “beautiful woman who leads you to victory,” and Microsoft’s Cortana for an artificial intelligence character in the video game Halo who has the outlandish proportions of a Barbie doll and sports a rather lewd blue costume. We already know that men dominate the tech industry, and the AI sector in particular. Alexa, Siri and Cortana, are not neutral decisions. It has always made sense that women who assume positions of power do so knowing that their voices will be uniquely scrutinized. Margaret Thatcher was famously advised by Laurence Olivier to hire a voice coach in order to dramatically drop the pitch of her voice. Hillary Clinton’s voice was called everything from “shrill” to “excruciating” and regularly compared to a nagging wife.

Most recently, there was Elizabeth Holmes, the disgraced founder and former CEO of the now-defunct blood-testing company Theranos. In addition to the Steve Jobs-ian black turtlenecks that became her trademark uniform, Holmes sported an unusually deep voice that former friends and classmates have said didn’t exist pre-Theranos, alleging that it was an effort to win the respect of her colleagues. Holmes has done a whole lot we should fault her for—like the actual criminal fraud she’ll be tried for in 2020—but cultivating a deeper voice seems like her most understandable move. According to Tim Bressmann, an associate professor of speech-language pathology at the University of Toronto, “Not using your fundamental frequency can be a way to express social power or domination. We all do it: To speak to our supervisor, we’ll raise our voice to make it seem smaller, but if we’re speaking to an underling, we lower it. Because a woman’s voice is naturally smaller, to reach a lower register in order to compete with a male frequency, at some point some may have to drop into the fry register.”

Bressmann is referring to vocal fry, a phenomenon most often associated with the Kardashians and commonly written off as shallow and ditzy. But to Bressman and any speech pathologist you’re likely to ask, vocal fry is simply a way of reaching a lower pitch and, for many people, appears at the end of a sentence when we’ve used up a breath. That means, of course, it’s not just a female thing: In fact, among the many men who have been known to use it are esteemed talkers Noam Chomsky and Ira Glass. They just haven’t received the same amount of criticism as their female counterparts because, as Bressmann puts it, “men just get away with more.”

Voice assistants, after all, are a proliferation of a role women have been assigned for decades, which is shocking considering the evolution of the form, but not so considering those at work behind the scenes.

Tweet this

Indeed, whether low or high, modulated or not, a woman’s voice can’t avoid policing. Slate podcaster Jessica Grose took up the subject with Stanford linguistics professor Penny Eckert on NPR’s Fresh Air in 2015 after Grose found her own voice criticized. Eckert admitted that she was “shocked the first time I heard [vocal fry] on NPR. I thought, ‘Oh my god, how can this person be talking like this on the radio?’ Then I played it for my students, and I said, ‘How does she sound?’ and they said, ‘Good, authoritative.’ And that was when I knew that I had a problem. … That I was not a part of the generation that understood what that style means. … There’s been a change and those of us who are bothered by some of these features are probably just getting old.”

Meaning, yet again, judging women’s voices comes down to bias. For a certain demographic, something like vocal fry can be distracting, but for another, it’s just the way people speak now. It seems time for technology to adapt, too. A promising development arrived in March when a team of researchers unveiled the world’s first genderless voice assistant. In a video posted to YouTube, Q introduced itself: “Think of me like Siri or Alexa, but neither male nor female. I’m created for a future where we are no longer defined by gender but rather how we define ourselves.” The voice was developed by a team of researchers, sound designers, and linguists, and based on male, female, transgender, and nonbinary voices. Still in early stages, the developers are hoping to see companies use it expressly to challenge gender stereotypes. And in an effort to combat the way we decide who belongs in positions of power, Google now randomly assigns its assistant’s gender for users. Similarly, in response to a ChildWise report in 2018 that found kids who grow up barking orders at virtual assistants are more likely to use that communication as adults, Google also recently added a new feature called “Pretty Please” that encourages kids to be polite, as has Amazon. If they say “please” after asking for the weather, for example, the respective assistant will respond with “thanks for asking so nicely.”

These are certainly baby steps, but as the UNESCO study notes, Alexa, Siri, Google Assistant, and Cortana are installed on more than two billion devices around the world, while over two-thirds of voice assistants offer female-only voices. Artificial intelligence is only growing more common and more humane. Voice assistants, after all, are a proliferation of a role women have been assigned for decades, which is shocking considering the evolution of the form, but not so considering those at work behind the scenes.


by Sadaf Ahsan
View profile »

Sadaf Ahsan is a freelance arts and culture writer and editor based in Toronto. You can find her work in MEL Magazine, Refinery29 and Flare. She is 50% hair.