Johns Hopkins Gazette: September 29, 1997

Babies May Recall
Complex Words

Psychologist continues
innovative work into
how children learn

Emil Venere
News and Information

Be careful what you say; little ears might be listening.

And remembering.

Recent findings in psychology research suggest that infants begin remembering relatively complex words long before they learn their meanings. The babies seem to pick up on the vocabulary even when they hear new words in tape-recorded stories without the benefit of any other stimuli.

Audio-taped children's stories containing words like peccaries and python were played to 8-month-old infants once a day for 10 days; two weeks later, 36 words that occurred most frequently in the stories were played back to the babies in list form.

Perhaps the most remarkable finding was that the babies recognized the words even though they sounded different in list form than they did in the stories.

"When we just read a list, we actually pronounce those words a little bit differently; those words have a very different acoustic form than they had in the stories," said Peter Jusczyk, a Hopkins professor in the Department of Psychology.

But the experiments indicated that the infants remembered the words they had heard in the stories, suggesting that babies memorize words that occur frequently in speech, an important prerequisite for learning language.

The findings were detailed in a paper that was published on Sept. 26 in the journal Science.

Although much work has been conducted to investigate how children learn the meanings of words, there has been little research aimed at learning how infants focus on the sounds of words, said Jusczyk, who co-authored the paper with Elizabeth A. Hohne, a psychologist at AT&T Labs in Holmdel, N.J.

The scientists tape-recorded women narrating three different children's stories, each lasting about 10 minutes. Then researchers visited the homes of 15 infants, playing the stories to them every day for 10 days. In the end, the babies had heard each story 10 times.

The psychologists identified the 36 content words--usually nouns--that occurred most frequently in the stories. Then they arranged those words in lists of 12 words each. Two weeks after the final visit to the infants' homes, the babies were brought to Jusczyk's lab at Johns Hopkins. One at a time they were placed inside a special testing booth, where they listened to the lists containing the words that occurred most frequently in the stories. Then they listened to lists of other, similar-sounding words that did not occur in the stories.

A light flashed above the speaker through which the tape recording was played. When the infants looked at the light, the word lists began and continued to play as long as the infants looked toward the light. Babies who stopped listening to the words looked away from the light, telling the researchers how long the infants had listened to specific lists of words.

"What we found was that the babies listened longer to the lists of words from the stories, significantly longer," Jusczyk said.

Previous research using the technique has shown that infants tend to listen longer to words that are more familiar to them. The researchers, however, wanted to make sure that the infants were not listening longer to the story words simply because they found them more interesting, so they brought a new group of infants to the lab who had never heard the stories on tape. When those infants heard the lists of the story words and the non-story words, they showed no preference and actually listened slightly longer to the non-story words.

"That showed us that the experience the babies had had at home listening to the stories had an impact on what they really remembered," Jusczyk said. He noted that the infants learned the words even though they never had any personal contact with the women who narrated the stories. "So, imagine what happens when you actually have the baby in your arms, and you are reading the story and you are turning the pages of the book," he said. "You'd expect that they would be even more inclined to store some of that information."

The babies who had never heard the stories listened an average of about 6 seconds to the story words and slightly more than that for the non-story words. The infants who had heard the stories listened an average of less than six seconds to the non-story words but nearly seven seconds to the story words.

"A second doesn't sound like a lot of time, but it's consistent," Jusczyk said. "The whole object of this was to see whether, when infants are listening to people talk or listening to people read stories to them, they are storing any information away about sound patterns that occur frequently."

Ultimately, scientists are trying to learn how young children are able to learn and master the complexities of language, a difficult task for the adult brain and the most powerful computers.

At about 18 months of age, a child's vocabulary and grasp of language suddenly expand, and scientists don't know why. One possible explanation is that children may begin storing the sounds and meanings of words while they are infants, and suddenly they are able to connect the words with meanings.

"It's sort of like working on a puzzle. You get a few pieces and then everything falls into place," Jusczyk said.

Learning words requires storing both sounds and meanings. This study shows that infants sometimes store the sounds of words, even when they have not yet learned the meanings, he said.

Jusczyk has written a book about the history of research into how infants learn language, focusing on the first year of life. His book, The Discovery of Spoken Language, was published earlier this year.

Go back to Previous Page

Go to Gazette Homepage