Voices carry so much information. Joy and anger, desires, comfort, vocabulary lessons. As babies learn about their world, the voice of their mother is a particularly powerful tool. One way mothers wield that tool is by speaking in the often ridiculous, occasionally condescending baby talk.
Also called “motherese,” this is a high-pitched, exaggerated language full of short, slow phrases and big vocal swoops. And when confronted with a tiny human, pretty much everybody — not just mothers, fathers and grandparents — instinctively does it.
Now, a study has turned up another way mothers modulate their voice during baby talk. Instead of focusing on changes such as pitch and rhythm, the researchers focused on timbre, the “color” or quality of a sound.
Timbre is a little bit nebulous, kind of a “know it when you hear it” sort of thing. For instance, the timbre of a reedy clarinet differs from a bombastic trumpet, even when both instruments are hitting the same note. The same is true for voices: When you hear the song “Hurt,” you don’t need to check whether it’s Nine Inch Nails’ Trent Reznor or Johnny Cash singing it. The vocal fingerprints make it obvious.
It turns out that timbre isn’t set in stone. People — mothers, in particular — change their timbre, depending on whether they’re talking to their baby or to an adult, scientists report online October 12 in Current Biology.
For the study, 12 English-speaking moms brought their babies into a Princeton lab. Researchers recorded the women talking to or reading to their 7- to 12-month old babies, and talking with an adult.
The unique “color” or quality of a voice is called timbre. You can hear timbre differences here, when two women say the same sound.
An algorithm sorted through timbre data taken from both baby- and adult-directed speech, and used this input to make a mathematical classifier. Based on snippets of speech, the classifier then could tell whether a mother was talking with an adult or with her baby. The timbre differences between baby- and adult-directed speech were obvious enough that a computer program could tell them apart.
Similar timbre shifts were obvious in other languages, too, the researchers found. These baby-directed shifts happened in 12 different women who spoke Cantonese, French, German, Hebrew, Hungarian, Polish, Russian, Mandarin or Spanish — a consistency that suggests this aspect of baby talk is universal.
Defined mathematically, these timbre shifts were consistent across women and across languages, but it’s still not clear what vocal qualities drove the change. “It likely combines several features, such as brightness, breathiness, purity or nasality,” says study coauthor Elise Piazza, a cognitive neuroscientist at Princeton University. She and her colleagues plan on studying these attributes to see whether babies pay more attention to some of them.
It’s not yet known whether babies perceive and use the timbre information from their mother. Babies recognize their mother’s voice; it’s possible they recognize their mother’s baby-directed timbre, too. Babies can tell timbre differences between musical instruments, so they can probably detect timbre differences in spoken language, Piazza says.
The work “highlights a new cue that mothers implicitly use,” Piazza says. The purpose of this cue isn’t clear yet, but the researchers suspect that the timbre change may emotionally engage babies and help them learn language.
People may not reserve timbre shifts just for babies, Piazza points out. Politicians talking to voters, middle school teachers talking to a classroom, and lovers whispering to each other may all tweak their timbre to convey … something.