For three days, the Internet has been tearing itself apart over this sonic puzzle: Is the voice saying “Laurel” or “Yanny”?
Some people hear “Laurel” and will go to their graves insisting that there’s no other possibility. Others hear “Yanny” and nothing will budge their opinions. In-office polls put the perceptions right around 50-50.
The answer, as we now know, depends on the characteristics of your hearing and the speakers you’re listening to. In general, younger people and tinnier speakers hear “Yanny”; older people and deeper speakers tend toward “laurel.” (Here’s a more complete explanation of the illusion.)
We also now know that this astonishingly weird one-word recording began life on Vocabulary.com, where it still plays today. You can listen to its pronunciation right here. And we know that Katie Hetzel, a high-school freshman in Georgia, first discovered the sonic ambiguity of that recording; another student posted it on Instagram; and a third posted it to Reddit.
What nobody has known until today, though, is whose voice is speaking it.
Meet Jay Aubrey Jones
The speaker of the word “laurel” (or “Yanny,” if you insist) is actor/singer Jay Aubrey Jones. He had been entertaining audiences on stage and screen long before his one-word 2007 recording catapulted him to a strange kind of Internet fame.
He’s appeared in Broadway shows like the original “Cats” and the 1995 revival of “How to Succeed in Business Without Really Trying.” He sang “Porgy and Bess” at the Metropolitan Opera. He’s performed in musicals at the Yale Repertory Theater, at the Goodspeed Opera House, and Off-Broadway, and he’s had some TV roles.
And once, 11 years ago, he recorded 36,000 words, one at a time, over a period of two months, for Vocabulary.com. Including the word “Laurel.”
“It was early days, so we just gave them a laptop with the best microphone we could afford, and foamcore and egg-crate stuff, and a little booth they could set up in their apartment,” says Marc Tinkler, the president and chief technology officer of Vocabulary.com—and the man who hired Jones. “We designed some software that would check a word out from our server, and show it to them on the screen, and they would do the recording. It would send the recording back to us, and it would go to two other people to confirm that it was accurate.”