Word by Word
Written By Erin McKean
print
single page
Babies go from cooing to sentences in two years. How do they do that?
One night when I was eight months pregnant — absurdly pregnant, with a belly that had its own gravitational field — I suddenly sat up in bed and grabbed my husband's arm. "What if the baby doesn't like books?" I wailed.
My husband looked around our bedroom at the wall of bookshelves stacked two deep with books, and the piles of books on the floor, the bedside table, and the dresser, and said, "Fat chance." Then, realizing he had said the word "fat" to a hugely pregnant lady, he invented an errand elsewhere in the apartment.
In truth, I wasn't really worried about the baby's love for books, I was worried about his love for language. I'm a lexicographer — I edit dictionaries for a living — and the thought of a child who wouldn't appreciate the glory of a word like "serendipity" or the fun of "ginormous" just seemed too awful to consider.
Luckily for me, nature is on my side. There's no guarantee that the child of a football fan will know the difference between a wide receiver and a tight end, or that the foodie's offspring will ever eat anything more demanding than mac and cheese, but linguists have assured me that the language lover's kid — and any kid — will almost certainly become absolutely fluent in the parent's language, just as a matter of course. So, having never met a research topic I didn't like, I read what the books (and the articles, and darn near the whole Internet) had to say about the process of language acquisition. I found a few different theories about what's happening "in there" as children learn language. The first group (and the biggest and loudest) holds that children are primed to learn language, any language, just as birds are predisposed to sing. These folks, influenced by Noam Chomsky, often postulate something they call the LAD, or language acquisition device, a function of the brain that people use specifically to learn language. That human languages can be very different from each other doesn't seem to matter. It's as if every child comes with a drum, but whether he uses it to play heavy metal, salsa, or smooth jazz is up to him (or, perhaps, his parents).
An opposite view holds that learning a language is just like learning anything else, a problem solved by general intelligence, and it's simply because children spend so much time learning their language that they get so good, so fast. A group in the middle (I'm in this camp, if any) takes the position that children may not, in fact, be wired for language specifically, but that clearly language is a big part of the very human desire we have to connect with other people. "Kids have special skills for understanding other people," says Michael Tomasello, codirector of the Max Planck Institute for Evolutionary Anthropology, in Leipzig, Germany. "It's not just the same way they would learn arithmetic. We are specially prepared to communicate with others, and the structure of language is an outgrowth of that."
One thing most experts do agree on: The observable process by which most children reach certain language milestones happens in a fairly predictable order.
I could see this daily at home. Before language, my son Henry's desires seemed amorphous, requiring trained detective work; afterward they were disturbingly concrete. He changed from an obliging infant to a tiny martinet, all with the advent of his first word, "light." As soon as he could say that one little word, he would use it to demand to be held up to the light switch, which he proceeded to flick up and down furiously like a tiny Frankenstein animating his monster. Watching my son change from a soggy lump to a little rhetorician who could demand, cajole, beg, and, yes, whine was fascinating. Reading gave me lots of background, but nothing beat observing the labor of language acquisition in real time.
The hard work begins before birth. I remember feeling guilty about listening to expletive-laden rap music in my last trimester because I knew that my baby had hearing nearly as good as mine. Despite the gurgling of my stomach and the lub-dub of my heartbeat, sound came through clearly. Whether he knew what Ice Cube was rapping about is another story. Because of this eavesdropping, Henry liked his mother's voice best when he was born, even better than Grandma's (no, I didn't tell her that). And because he had also absorbed the rhythms of English, he preferred not just my own voice but my language. A 1986 study using rigged-up pacifiers with 4-day-old infants found that French babies sucked harder upon hearing tapes of someone speaking French than when they were listening to the same person speaking Russian — and for babies, sucking is a big thumbs-up.
But that didn't mean my infant son could carry on a conversation, in part because he was not yet physically able to speak. One of the reasons newborns don't emerge commenting on the weather is that their vocal tract isn't the right shape to make language sounds. (The very idea of a newborn speaking is so freaky that the eerie talking baby is a staple of horror movies.) Henry just wasn't up to vowels and consonants, much less the voice and mouth movement combination that results in the word "serendipity."
Although they can't yet make all the sounds they will need, babies under 6 months are capable of telling the difference between almost all the important sounds of the world's languages; they have, in effect, perfect pitch for speech. But keeping that perfect pitch is inefficient. Infants must figure out just which sounds they need for speaking the language they're growing up with. That's why Japanese adults have trouble telling the difference between "r" and "l" sounds: That difference doesn't matter in Japanese, so as babies they stopped paying attention to it. If they want to learn English later they have to retrain their brains to hear it and their mouths to produce it.
In fact, sorting out all the sounds of his language — and only his language — was one of Henry's main jobs, and by cooing a high-pitched "who's a good baby?" I was helping. Babies prefer "parentese" (it's almost impossible not to use it when talking to one), and linguists have found this kind of parentspeak to be nearly universal. Speaking parentese emphasizes those important vowels, making them stand out. But it's one thing to hear those sounds; it's another to be able to make them on demand. Henry, like most babies, started his language practice with cooing and gurgling noises at around 3 months, and in part because I was cooing back at him, he managed to pick up the typical sounds of English by about 6 months: The "ba-ba-ba" and "ma-ma-ma" sounds I shot at him he babbled right back.
For all babies, babbling is a kind of oral aerobics. Henry loved figuring out how his mouth worked and what sounds he could make — the louder the better, of course — and he relished his new toy. His babbling was not completely random. Some sounds and rhythms were more likely to show up than others because he was babbling in English. And by about 9 months, he could make speech sounds almost as well as I could (for what that's worth). The more common the sounds are in the language the child is practicing for, the more common those sounds are in his babbling. French babies babble in French, Chinese babies in Chinese — and so consistently that in a 1984 study, French-speaking adults listening to tapes of 6-month-old babies babbling in different languages were able to pick out the French-speaking babies nearly 75 percent of the time.
Next page: Deaf babies babble with their hands
