Your voice versus your words
Fortune 500 recruiters preferred audio recordings of job pitches over written pitches specifically composed to be read.
Schroeder and Epley, 2015
Our tendency to dehumanize more or less according to the medium we’re using matters not just when we’re evaluating other people, but also when other people are evaluating us. All MBA students know how to give an elevator pitch, and most have rehearsed their personal pitch to recruiters—I suspect if you called one at 3 a.m. and asked for her pitch, she could recite it without lifting her head from the pillow—so Juliana Schroeder and I asked some Booth students to come in and give us theirs. Specifically, we asked for the pitch they’d give to their ideal employer, and then we again created video, audio, and written versions of each pitch. We also asked each student to write a pitch, because there could be meaningful differences between a pitch specifically composed to be read and one composed to be heard or watched.
We then asked evaluators to imagine that they were employers and report their impressions of how confident the person seemed, and how thoughtful, intelligent, and competent. They gave us their general impression, assessed how likable the person was, and indicated their interest in hiring the person.
It may not surprise you to learn that when the evaluators heard the students, as opposed to reading either their written pitch or a transcript of their spoken pitch, they described them as seeming more thoughtful, intelligent, and rational. They had a more favorable general impression of the participants, and they were more interested in hiring them. Again, adding video didn’t make much of a difference. What’s more, we got similar results when we repeated this experiment using actual Fortune 500 recruiters as evaluators instead of participants acting as employers.
Media evolve faster than humans do
Human history is long. Homo sapiens emerged on the planet somewhere around 300,000 years ago, but it took us about 295,000 years to start writing to each other. In all the intervening time, our brains evolved to communicate with each other in a particular way: face-to-face. We had physical interactions laced with voice or visual cues. As a species, we learned to communicate, to convey our states of mind, under those conditions.
And if writing is a recent innovation on the timeline of human history, electronic media are virtually brand new. It’s not surprising that we might find some gaps in our ability to use these tools especially effectively.
For instance, if I wanted to create a maximally dehumanizing medium for communicating with other people, I couldn’t do better than Twitter. Twitter is not only a largely text-based medium, but it is also a psychologically distancing medium. Other users are often identified by “handles” rather than by their own names, and the people you write to are “out there” on the internet somewhere rather than “right here” having a direct conversation with you. Add to that a character limit that makes it impossible even in text to communicate sophisticated or nuanced thought and you have the perfect platform for making other people seem like unthinking objects or unfeeling animals.
Facebook is a little more complicated. It was intended to connect people with each other. But it turns out that social connection is to Facebook what sugar is to Diet Coke: it seems like it’s there, but research again suggests not. My collaborators and I have found in experiments time and again that talking to others in person makes people feel better than they expect it will. It doesn’t even matter what they talk about—shallow stuff, deep stuff, whatever—they tend to love it more than they expect. But research on Facebook users finds that the more people use Facebook, the worse it seems to make them feel.
You might imagine that we’ll get better at using this technology over time. We’ll get better at using Twitter and Facebook and text-based media interaction in general over time. I don’t think so. Our data suggest that there’s something inherently dehumanizing about the cues that are present in text-only information. And you can’t artificially add cues, such as a person’s voice, into a text-based medium of communication that doesn’t include them to begin with. The only way you can do that is to use a voice-based medium instead.
Perhaps the most important question to ask is whether we are sufficiently media savvy to have a sense of these kinds of effects. Again, I think the answer is no. For one thing, avoiding the specific phenomenon I’ve been describing, the tendency to dehumanize in text-based communication, appears not to be intuitive for many people. We asked about a thousand participants in an online survey: If you were making an elevator pitch and wanted to be perceived as most intelligent, how would you choose to express your thoughts to someone? Would you choose to write or would you choose to speak? Seventy percent said they’d opt for writing.
Moreover, the medium through which we communicate with other people is often overlooked; the idea that we could be interacting with somebody through a different medium often doesn’t occur to us when we are in the midst of an interaction. I did a study some years ago with Justin Kruger, Jason Parker, and Zhi-Wen Ng, all then at the University of Illinois, in which we compared voice recordings to email, with participants sending messages via each that were either sincere or sarcastic. Not surprisingly, email recipients were not great at detecting sarcasm: their accuracy rate was 56 percent, not significantly better than random guessing. When they heard the message spoken aloud, of course, they were much more accurate. But just as important, both senders and receivers overestimated how easily sarcasm would be detected over email. They had no sensitivity at all to how the medium was affecting the message being conveyed.
This is particularly important because people’s preferred mode of communication is not necessarily the most effective one. University of Texas at Austin’s Amit Kumar and I have found in recent experiments that people tend to prefer email over a phone call, at least when it comes to reaching out to an old friend. Specifically, they expect a phone call to be more awkward, but when we put these two modes of communication to the test, we find that it’s not. Again, I think this suggests a misunderstanding of how media affect interactions.
The value of voice
The primacy of voice in communicating state of mind may not be completely intuitive when you communicate with all of your senses intact. But Hellen Keller, who lacked both hearing and sight, understood the importance of voice powerfully through her lived experience. Keller was once asked to speak at a conference advocating for kids who were also deaf and blind. In a letter to the organizer explaining why she was unable to attend, she wrote,
I’m just as deaf as I am blind. The problems of deafness are deeper and more complex, if not more important than those of blindness. Deafness is a much worse misfortune, for it means the loss of the most vital stimulus—the sound of the voice that brings language, sets thought astir, and keeps us in the intellectual company of man. . . . I’ve received letters from the parents of children who are either deaf or feebleminded. The parents could not say which. The doctor did not know or else he did not tell them the truth.
We don’t speak of other people being feebleminded anymore, but I think the tendency to infer that someone who lacks a voice, whom we don’t hear from directly, might be less mentally capable than other people still shows up in all of the data that we’ve been collecting. Technology can continue to develop new and more innovative modes of text-based communication, but it may never recreate the cues to another’s mind contained in the human voice.
Nicholas Epley is the John Templeton Keller Professor of Behavioral Science and the Neubauer Family Faculty Fellow at Chicago Booth. This essay is adapted from a presentation delivered at the Kilts Center’s Marketing Summit 2019.