Our history and values are embedded in our language, says LSA Ph.D. candidate Kelly Wright. When we speak, we communicate a lot more than the meaning of our words. Our voice can indicate our age, our ethnicity, or our education—information that can make us vulnerable to discrimination or help us connect with others. There’s a lot we can understand about other people—and ourselves—by learning to listen more closely to someone’s voice.
LSA: How did your training as a linguist help you investigate housing discrimination?
Kelly Wright: One area of my research seeks to trace linguistic profiling, a phenomenon that was developed by a linguist named John Baugh. It’s parallel to the concept of racial profiling—that when we see a person, we use information about their appearance to decide how we want to interact with them. Linguistic profiling operates on the same principle—it's just that you’re hearing instead of seeing. My project looked at how linguistic discrimination works in the rental market.
After designing a field experiment that tests for discriminatory behavior called an audit study, I did a search on Craigslist and found 90 rental ads that included phone numbers and called them from the lab. My training in linguistics helped me assess the landlords’ responses to three different voices. The landlords—who were in three different, class-separated neighborhoods—thought they were hearing a different person in each call, but it was always me—a mixed-race Black woman from Knoxville, Tennessee, one of the largest cities in Appalachia—using her three dialects.
A protected class is a category like race or gender that is legally protected from discrimination. I found listeners are able to accurately identify when people belong to a protected class just from hearing their voice. This is something we’ve known about for 40 years, and there’s a mountain of scientific evidence that confirms this. But I found that listeners make rapid judgments about a person's character based on those identities too. It’s not only that I hear a Black woman speaking, it’s that I hear a Black woman speaking and I don't trust her as much as a white woman, or I don't think she's very educated or competent. And those decisions about trustworthiness and competence affect a renter’s chance of finding a home in whatever neighborhood they can afford.
LSA: Is linguistic profiling also legally recognized as discrimination?
KW: When it comes to voice, discrimination law is very asymmetrical. It can be used to help convict you, but it cannot protect you. If you said, “I heard a Black person robbing that store,” or, “There was a large man outside my door threatening me,” those identities can be used as prosecutorial evidence to send somebody to jail.
But at the same time, you can't claim someone discriminated against you because of your voice. You can't say, “I called that person and I know that they knew I was Black because I sound Black and that’s why they wouldn’t show me the apartment.” It’s a double standard. Legal discrimination has to happen through physical proximity. The law says I have to be standing in front of you for you to identify me and discriminate against me as part of a protected class.
Using “desirable” attributes identified by rental professionals in a national survey, Wright mapped landlords’ responses to each of her dialects: African American language (AAL), Mainstream US English (MUSE) and Southern American (SA). Overall, MUSE received the most positive ratings, though SA was similar or equal in several categories, while AAL was rated as both pleasant and least trustworthy, least attractive, and least confident.
LSA: What is it that we hear in a person’s voice that conveys information about them, such as their race?
KW: We really don't know! Black speakers aren’t a monolith, so it has been challenging to objectively define what’s common about African American languages. It might be speech rate, it might be tone, or some vowels might be held just a little longer—and by longer, I'm talking about fractions of a millisecond. We’ve measured all of these things on spectrograms, but there’s not a clearly measurable explanation. More colloquially, some people say if you “put bass in your voice” that conveys Blackness, which is not something we can measure on a spectrogram.
Still, when most people hear a Black person speaking, they know they’re Black. It's something that goes beyond phonetics, phonology, lexical items, morphosyntactic variation—all of those things that sociolinguists, or in linguistics generally, typically identify as the places where languages vary. There is something very intimately tied to identity in ways that we have yet to establish in all of our ways of measuring language. I’m very excited to be doing work on the cutting edge by working with these exact questions.
LSA: There are a lot of ways people can use speech as a tool to discriminate—from forbidding someone to speak a certain language, for example, or assuming that someone who speaks with a regional dialect or non-standard English isn’t smart. As a linguist, how do you think about those kinds of discrimination?
KW: In those examples, language reveals a lot about our culture and our history. What we think of as “grammar” is a raced, classed, gendered, and colonized construction. My research (and that of many others) shows that discrimination can occur through voice alone—and that it has tangible consequences. In order to combat linguistic discrimination, we need to advocate for equity for people who don’t communicate in White Standard Spoken English. In fact, we have an organization of students, faculty, and staff on campus called Language Matters that deals directly with these issues.
The American standard English we speak is a legacy of the anti-immigrant sentiments that swept the United States in the early twentieth century. In other countries, the national standard languages are associated with the largest population centers. Parisian French is the standard variety of France. In England, it’s London—the Queen’s English. So why isn't New York the standard in the United States? It's because of xenophobia.
At the turn of the twentieth century, immigrants from Eastern Europe and other places poured into New York and Philadelphia and other coastal cities. People didn't want to sound foreign or accented so they gravitated towards pronunciation associated with what they considered a pure region of the country: the Midwest. The Midwest speech pattern was explicitly race neutral and ethnically unaccented. Over time, it spread through prominent newscasters, such as Walter Cronkite and Dan Rather, and became associated with people who we considered to have better values.
LSA: If Midwestern speech is the standard, does that mean Midwesterners don’t have accents?
KW: Everybody has an accent, and what's interesting about your accent, about any individual voice, is that it carries a lot of your history. My voice is Southern and it's Black because that's where I grew up and that’s my heritage. My mother is a German immigrant from upstate New York. Her speech is very standard because she learned English in school. Her parents didn't speak it, so that's the only English she knew when she was a child.
All this interesting stuff comes into your voice and then you get to a point where you think, “Okay—this is who I am and this is how I want to sound.” I don’t think people observe their own voices enough to realize what's really cool about it.
LSA: Why is our language such an important signifier?
KW: I love this question because language is integral to our experience of the world. The same thing that structures our thoughts and memories, this cognitive engine that we have for making sense of the world and seating ourselves within it, is the exact same engine that helps us communicate. It's how we orient ourselves to other people, and it's how we create and express our personal identities, our collective histories, our knowledge, and our ancestries. They are all baked into the linguistic material that surrounds us.
When we encounter another person, we use all the information we can to determine who they are and to predict how they'll react to us. In a perfect world, prejudice wouldn't affect our next step, our next action. But in our world our ideological predisposition and stances mold much of what we attend to in someone’s language.
Because our brains work the way they do, there are things that you produce in your voice that you yourself don't hear. We do the same thing when we’re listening to other people. They might produce features that you don't hear, or they might not produce features that you think you hear. Your voice, your actual voice, says a lot about you. It has all these wonderful features about it. A lot of people think, “I don't have an accent.” But you do and your voice is amazing.