In May 2018 a one second audio clip divided the internet: in a Twitter poll of more than 500,000 people, 53 percent heard Laurel and 47 percent heard Yanny, and each side found the other impossible to understand. The recording began in 2007, when opera singer Jay Aubrey Jones recorded about 200,000 reference pronunciations for Vocabulary.com in a makeshift New York apartment studio. Eleven years later Katie Hetzel, a 15 year old freshman at Flowery Branch High School in Georgia, looked up laurel, heard Yanny, and shared the clip by replaying it through speakers and re-recording it in a noisy classroom.

From Instagram to Reddit to YouTuber Chloe Feldman on Twitter, the clip exploded, drawing in Ellen DeGeneres, Stephen King, Chrissy Teigen and the White House. Spectrograms showed both words living in the same degraded file: Laurel in the low frequencies and Yanny in high frequency noise. Age-related loss of high frequency hearing explains part of the split, but the episode also follows how the brain forces a single answer from an ambiguous signal and how researchers flipped listeners’ perception without touching the clip.

• Researchers recreated the Yanny pattern by splicing syllables from the same singer’s recordings of other words, including uncanny
• Audiology professor Benjamin Munson mapped Yanny to the high frequencies and Laurel to the low
• The New York Times pitch slider that turned the clip into clear Laurel or clear Yanny
• Hans Rutger Bosker’s study where a muffled lead-in sentence primed listeners to hear Yanny
• Look-alike illusions: the Brainstorm or green needle toy and Grover’s misheard line on Sesame Street