Psycho Babble

How We Talk to Babies

By Jessica Love | August 9, 2012


Not long ago, a fellow psycholinguist confessed to me that she would never pursue research on infants. Not because she hates them, but because she loves them. “I just don’t know how I would ever get any work done,” she gushed. “It’s hard enough knowing they’re in the same building.”

Babies have a way with us. We stare unabashedly at them and smile and make silly faces. Our speech changes, too. Baby talk, known among researchers as infant-directed speech, motherese, and parentese (as fathers are not immune), is well—even exhaustively—studied.

When we talk to babies, we tend to use shorter, simpler sentences, which we utter more slowly. Our pitch becomes higher and more variable, our affect warmer, and our vowels more exaggerated. “Oh suuuch a niiice baaaby!” we coo. And by “we,” I mean just about everyone everywhere; baby talk exists, in a (somewhat) similar form, in every language in which it has been studied. (The best known counter-example may be the Kaluli, an indigenous people from Papua New Guinea whose parenting practices were documented by anthropologists Elinor Ochs and Bambi Schieffelin. And yet, Stanford researcher Anne Fernald and her colleagues have argued that even though Kaluli “mothers do not address their infants directly, they frequently speak ‘for’ the infant in high-pitched voice,” and thus may be “functionally … speaking motherese.”)

So why do we do this? For one, infants like baby talk. Pretty much from the get-go they’ll listen to it longer than they’ll listen to speech directed toward adults. Baby talk is also a particularly efficient conduit of speakers’ emotions and intentions. All speech transmits this information to some degree; a non-English speaker can nonetheless listen to an English speaker’s words and do a decent job of deciding whether they’re listening to chastisement, approval, comfort-giving, or attention-seeking. But present that non-English speaker with English baby talk and she’ll do an even better job. Baby talk, with its theatrical sing-songedness, wears its heart on its sleeve.

Finally, and slightly more controversially, baby talk may actually help infants learn their native language. In 2009, for instance, researcher Leher Singh, now at the National University of Singapore, collaborated with colleagues at Boston University and the University of Vermont on a study that exposed seven-and-a-half-month-old infants to two simple but as-yet-unlearned words (bike and hat). The words were either presented in infant-directed or adult-directed speech. Then, 24 hours later, the researchers tested the infants’ memory for the words they’d heard the day before (compared to similar control words like tree and pear). Infants listened longer to passages containing the familiar words than the control words, but only if the words had originally been presented in infant-directed speech. That is, they only showed learning when bike and hat had been cooed to them in baby talk.

So is baby talk a happy accident, or a smart parenting strategy developed over the ages? We may never know exactly how it came to be, but we do know that we speak differently to babies, to whom we hope to teach language, than to our pets, whom we just hope to house-train. According to University of Western Sydney researcher Denis Burnham and his colleagues, pet talk is in many ways similar to baby talk, sharing the same high-pitched tone and happy affect. (Surely you are as guilty of this as I am.) But when talking to pets, we do not hyper-articulate our vowels. No matter how clearly we sound our vowels, we know in our heart of hearts that our cat will remain stubbornly ignorant of the intricacies of the English language.

When talking to adult foreigners, interestingly enough, we do hyper-articulate our vowels. We drop the happy affect and high-pitched tones of baby talk—they’re adults after all—but we adopt the exaggerated vowels. That lost tourist who asks for our help will care very much whether he’s headed toward the shop or the sheep.

Permission required for reprinting, reproducing, or other uses.

Comments powered by Disqus