I heard “laurel.” But my coworker heard “yanny.” Welcome to “the dress” debate of 2018.
If you were on Twitter today, you also likely heard the viral four-second audio clip that, depending on the listener, sounded like one word or the other—or in some cases, both. So what weird trick of the human body is making people hear two different words from the same audio file?
Experts say it comes down to the frequencies we hear and, perhaps more importantly, the frequencies we expect to hear.
Brad Story from the University of Arizona's Speech Acoustics and Physiology Lab took a fine-toothed comb through the clip: “I'm pretty sure the original recording was 'laurel,'" he says. "The reason it can be confused is that there is a family of frequencies produced by the shape of our throat and mouth."
The three lowest frequencies are used to encode language as a sound wave. The third frequency distinguishes between l and r. This frequency is high for l, like at the beginning and end of "laurel," and low for r, as in the middle of "laurel."
To test this, Story recorded his own voice pronouncing both words and found similarities in the sound patterns for “yanny” and “laurel.” Because the original audio clip isn't exceptionally clear, it leaves room for interpretation—and that's where the mental controversy kicks in.
“The way you hear sound is influenced by your life in sound—what you know about sound,” says Nina Kraus from Northwestern University's Brain Volts lab. For example, she shared the two audio clips below. When listening to the first one, you hear white noise. When listening to the second, you hear a clear phrase.
Now listen to the first again, and you'll hear that phrase.
"Part of the answer is the difference between listening and hearing," says audiologist Douglas Beck, executive director of academic sciences at hearing aid manufacturer Oticon Inc. and senior editor of academic sciences at the audiology trade magazine Hearing Review. "Most people think of hearing as occurring in the ear, but hearing and listening actually occur in the brain."
"Hearing is simply perceiving sound," he adds. "That is, you can hear while you're asleep, and so in that regard, hearing is passive. Listening is attributing meaning to sound. A host of factors, such as working memory, expectations, language skills, cognitive ability, hearing, musical skills and training, attention and more, influence the many ways people listen to the same audio clip."
This background effect in the brain leads to a top-down approach to listening, Story says, in which your brain fills in any missing pieces with what you expect to hear. Because the viral version of the audio clip isn't crisp, it leaves ambiguity, and your brain fills in the rest.
“They were primed to hear 'laurel' or 'yanny,'” says Story. “They may have made their decision for what they're listening for” before the clip is even played. To truly test how someone perceives the clip, he notes, you would have to play it without any visual cues and simply ask, What do you hear?
As the clip spreads across social media, however, it may be getting harder to find truly unbiased listeners.