A computer-like brain

ANNE McILROY

Hamilton From Saturday's Globe and Mail

Joseph Visser doesn't like hats, so he looks skeptical when a researcher slips a dangly wire wig of 128 electrodes over his head. He is six months old, and this is just another outing with his mom.

For Laurel Trainor, a researcher at McMaster University, Joseph's visit to her lab is a chance to figure out how babies learn language. The white plastic net is hooked up to an electroencephalograph, or EEG, that monitors electrical signals from Joseph's brain as he hears a series of clicks.

While babies use many non-verbal ways to communicate with their parents, including crying, smiles and other facial expressions, none are that precise. From the moment infants are born, they are learning the spoken language they hear around them. Dr. Trainor is one of many researchers studying how.

Newborns prefer hearing speech to non-speech sounds, and can discriminate words that convey meaning on their own — nouns and verbs — from those that don't, such as articles and conjunctions, says Janet Werker, a researcher at the University of British Columbia.

Dr. Werker's research has added to the mounting evidence that infants, like little computers, learn by statistically analyzing the impressions they receive. In the case of language, this analysis allows them to distinguish what is variable and focus on what is constant.

For example, if you play a baby a tape of nonsense syllables in which three sounds bo-gi-da are often found together, babies will learn to recognize bo-gi-da. They do the same with the real speech they hear from their parents and others.

At six months of age, babies recognize the sounds from many languages — not just the ones spoken by their parents. There is no point in playing Mandarin tapes for junior, however, unless you plan on speaking it to him. Babies lose this ability before they turn one unless they hear a language consistently.

For her part, Dr. Trainor wants to know if Joseph can tell when sounds start and stop — an important skill for recognizing words. She has found that even young babies can tell when there is a gap of only 16 milliseconds between the sounds. Now, she is hoping to find a way to predict if infants who have difficulty distinguishing spaces between sounds will have trouble with language and speech as they grow older.

In English, for example, the only difference between pa and ba is the delay of a few milliseconds. (If you put your fingers on your larynx, or voice box, you will notice that although your mouth and tongue are in exactly the same position for both, your vocal chords vibrate later for pa.)

She also does experiments with music, and has found that before they understand a single word, young babies can discriminate changes of one note in a melody, and show an emotional response to music. By six months, they can remember melodies for weeks.

"Before they are linguistic beings, they are communicating and interacting through music," Dr. Trainor says.

Just as babies seem built to learn language, mothers and fathers also seem to naturally change the way they speak to infants in a way that makes it easier for their babies to learn a language.

Listen to Patricia Monzon, who is babbling in motherese — that is, baby talk — to her son Dez: "Deeeeeeeezby, ahhhhhh. Where is the baaaby boooy? Where is the goooood booooy. Where is the goood boooy. Dunk dunk went Kermit the frog, one day. Dunk Dunk went Kermit the frog. Nose, nose, nose where is your nooooose?"

Motherese is high-pitched, slow-motion and repetitive. Taken out of context, it sounds ridiculous, but researchers believe that it is what babies need to hear.

Ms. Monzon is taking part in an experiment on motherese — also known as infant-directed speech — at Dr. Trainor's lab. She's speaking into a microphone. Her five-month-old son is in a soundproof booth behind her. She thinks Dez can hear her, but in fact, he's not hearing a single nonsensical word.

Dr. Trainor wants to know if motherese is innate, or something parents learn, encouraged by the enthusiastic reaction of their infants when they speak in a high-pitched voice. She is watching to see if moms like Ms. Monzon change the way they talk when they don't get the same positive response.

So far, they have found that mothers simply try harder when their babies don't respond as usual. They don't switch to more adult-like speech patterns. "If they have learned it, they have learned it very, very well," she says.

Adding to the evidence that motherese is instinctual is the fact that it has been found in every culture researchers have studied. It is most pronounced in American English speakers, followed by Canadian English speakers. It is least noticeable in Japanese, a culture in which language is less emotionally expressive.

Dr. Trainor believes that it is the emotional content of infant-directed speech that makes it so attractive to babies, getting them to pay attention and helping them to learn. Babies don't understand the meaning of words, and the only way adults can communicate with them verbally is by maximizing the display of emotion.

"Why don't we talk in infant-directed speech to adults most of the time? Probably because we've learned socially that you shouldn't let all your emotions out when you are talking to other adults. . . . Unless you are two lovers alone in a room. Then you would see something much closer to it."

If there is a message for parents in her research, it is that they should trust their instincts more.

"You sometimes wonder if parents wouldn't be better off being a little more intuitive. When you look at something like infant-directed speech, they just do it. "

Join the Discussion:

Sorted by: Oldest first
  • Newest to Oldest
  • Oldest to Newest
  • Most thumbs-up

Latest Comments

Sponsored Links

Most Popular in The Globe and Mail