Regular speech has a sing-song quality that's easy to miss. When we talk, our voices rise and fall. Sometimes the tune stands out, like the infamous “rising high terminal” of uptalk–the upward lilt that makes, like, everything sound like a question? But most often the pitch falls away, and our attention focuses on the words' meanings. Yet as Elizabeth Hellmuth Margulis writes for Aeon, the song of speech can be pulled back out with one simple trick: repetition.
By repeating a word of phrase over and over and over again, your brain starts to see the song that lies beneath. Even when that same phrase is played back again as part of a longer sentence, says Margulis, the sound seems to transform.
By way of example, Margulis presents two audio samples. One has a short phrase repeated out of context:
And the second has the original phrase. It seems, says Margulis, “as if the speaker has broken into song, Disney-style.”
This illusion is known as the speech-to-sound illusion, and it's a testament to the blurry boundary between speech and song.
“No matter the constituent material,” says Margulis, “whether it’s strings of syllables or strings of pitches, it seems that the brute force of repetition can work to musicalise sequences of sounds, triggering a profound shift in the way we hear them.”
You can push this illusion pretty far, too. Take, for instance, the music of the experimental Canadian band TAS 1000, who used recordings lifted from a found answering machine as lead vocals for an entire, surprisingly catch, album.