Adoptees offer clues on skills of language

Harvard research uncovers pattern

Julie Ross helped her son Andy with spelling and language while home schooling in their Durham, N.C., residence. Julie Ross helped her son Andy with spelling and language while home schooling in their Durham, N.C., residence. (Davis Turner for The Boston Globe)
By Carolyn Y. Johnson
Globe Staff / February 15, 2010

E-mail this article

Invalid E-mail address
Invalid E-mail address

Sending your article

Your article has been sent.

  • E-mail|
  • Print|
  • Reprints|
  • |
Text size +

As Andy Ross learned to speak English, he progressed from simple word combinations like “Andy shoe’’ to the more complex “my red shoe,’’ just like any toddler.

But Andy was older when he began to learn English, after being adopted from Russia, and his chatter - taped in weekly sessions - has provided scientists important clues about how language develops.

Harvard psychologists are finding that preschool-age children adopted from foreign countries learn English in the same sequence as babies: starting with single words and progressing to word combinations and complex grammar.

That means it is not the maturity of the brain but the nature of language itself that dictates how it is learned, the Harvard scientists say.

“Because babies are immature in so many ways, it’s easy to assume their language is simple because their minds are simple,’’ said Jesse Snedeker, the Harvard psychologist leading the research. But this appears not to be the case, she said.

The research is helping scientists and families understand what to expect when a child who may already know one language is plunged into a new one. And it could eventually lead to better ways to teach language.

“Knowing which kinds of things have to come before, because it’s inherent in the language system being used, strikes me as something important for an intervention’’ in children with language problems, said Susan Goldin-Meadow, a psychology professor at the University of Chicago who was not involved in the study.

She said the Harvard work is exciting because it begins to get at an underlying question that has been difficult to study. Unraveling how and why we learn to speak can be far from straightforward, with much happening simultaneously in the brains and bodies of growing babies.

Scientists such as Goldin-Meadow, who has studied gesture communication systems developed by deaf children, have long depended on events that amount to natural experiments to probe language development, because no experimenter could wrest children away at birth and delay their exposure to language just to see what happens.

An apocryphal early example of the longstanding interest in the origins of language comes from the writings of Herodotus, in which a king, intent on finding out whether the Egyptians or the Phrygians were more ancient, ordered two children raised in complete isolation to find out what their first word would be. Whatever language they spoke first would indicate which group was the oldest.

Today, psychologists depend on groups of people who have experienced aberrations during development to study language. Deaf children who do not learn a language until later in life and children not exposed to language in the early years, such as “Genie,’’ a California girl who was locked in a room the first 13 years of her life, have been studied to better understand natural languages people can develop on their own and to better understand a “critical period’’ during which children must acquire language.

Julie Ross of Durham, N.C., Andy’s mother, said she decided to participate in the Harvard research “to become part of something that would help ourselves at that time, but also for families in the future.’’ Ross, who adopted two boys from Russia - Andy at 2 1/2 and Max at nearly 3 - said adoptive parents often don’t have the same language benchmarks to understand whether their children are doing well or not, and this work helped her understand Andy’s progress.

Snedeker is motivated less by such practical issues than by trying to answer the basic question of how language develops in children.

Of course, babies must develop certain cognitive abilities before they are capable of speech, but it appears that cognitive immaturity isn’t what dictates how children learn a language, said Snedeker, who has studied 141 older international adoptees, some for as long as 15 months. If that were the case, she would expect to see older children who were learning a language to leapfrog some of the usual stages. Perhaps they would learn many verbs and prepositions that are usually learned later, interspersed with nouns like “cookie.’’ But they don’t.

Children, Snedeker said, “might need to learn words like ‘ball’ and ‘shelf’ before they can learn a word like ‘on’ ’’ and produce a sentence like “the ball is on the shelf.’’

Her research has revealed that older children are quicker at picking up words than babies, but they go through the same stages and make the same sorts of grammatical mistakes as their younger counterparts.

They reach milestones based on the time since their arrival in the United States, not on their chronological age. After being in this country for three months, a study participant named Jason, who was 3 years, 9 months old, talked in short, telegraphic utterances typical of a 24-month-old: “Mama cookie’’ and “Mama look helicopter,’’ he said in one video. On the other hand, a 5 1/2-year-old named Regina, who was adopted from Russia and had been in the country for less than six weeks, sometimes uttered words from her native language and said single English words one at a time.

The one difference Snedeker has observed is that older children are much quicker to use words that refer to time, the future and the past. That, she said, suggests that there are at least some areas in which younger children may lack the cognitive maturity to grasp a concept.

Charles A. Nelson, a pediatrics professor at Harvard Medical School and head of a cognitive neuroscience laboratory at Children’s Hospital Boston, has studied the effects of institutionalization on children. He cautioned that the children’s experiences before they were adopted may have affected their cognitive development and could have altered their abilities to learn language. But he said Snedeker’s approach is clever and follows a long tradition of psychologists who have studied language using natural experiments.

“With animals you can control the environment,’’ Nelson said. “There, you have tremendous control. You can’t do that with humans - you depend on these accidents of nature.’’

Carolyn Y. Johnson can be reached at