Like a footprint in wet concrete, the first language a baby hears makes an impression that lasts for years, regardless of what follows. Later, children even as old as ten who are adopted or immigrate can completely forget their first language. But even if they do not consciously remember their mother tongue, their brains retain its traces, according to a study published this week in PNAS, led by Lara Pierce of McGill University.
The team studied girls who were adopted from China into French-speaking families. None of the adoptees could speak Chinese, but their brain scans showed a subtler story. Their brains processed tones used in the Chinese language like bilingual Chinese-French speakers, not like children who spoke only French. “People’s behavior is not necessarily an indication of how their brain is functioning,” said Fred Genesee, professor emeritus at McGill and senior author of the study.
The study looked at 44 girls. Twenty-one were adopted from China between 6 and 25 months of age (13 months on average). They compared the adoptees to two control groups: 11 girls who spoke only French, and 12 girls who spoke both Chinese and French.
At the time of the study, the girls were 9-17 years old. While inside an fMRI brain scanner, they heard pairs of pseudowords that used Chinese tones. Some of the pairs were identical, and in some, the tones of the last syllables were different. The girls had to push a button when the last syllables were different in tone. The Chinese language uses tones, while French does not.
All three groups of girls were equally accurate and quick at detecting tonal differences. The French-speaking girls’ right brains were activated, but not their left brains, where language is processed. “They were treating these sounds just as sounds,” not as language, said Genesee.
By contrast, the adoptees’ left brains were activated, a pattern that matched that of bilingual Chinese-French speakers. Activation in the left hemisphere was correlated with age of adoption, but not with how long it had been since a child was adopted.
The results may have implications for how brains function when they learn a second language. “If the brain has been fine-tuned” to be sensitive to the first language it is exposed to as a baby, said Genesee, “then learning a second language is always going to be different than learning a first language.” This doesn’t mean they can’t learn a new language well—Genesee said the adoptees’ test scores are within the typical range of monolingual French speakers. But they are probably “using brain systems that are different from what monolingual speakers use.” The team is looking more closely at how the adoptees’ brains process French, and plans to publish the new data soon.
Early language studies often look at adopted children. For example, Christophe Pallier, a senior research scientist in the neuroimaging of language in Paris, in 2003 published research that found that adoptees from Korea did not consciously remember Korean that they were exposed to in their first few years. But the adopted girls from China in the McGill study were particularly useful research subjects. Families often give up children for adoption because of medical problems, developmental issues, or extreme hardship, which could confound the results of a language acquisition study. But China’s one-child policy and cultural preference for boys means that many otherwise healthy girls are given up for adoption.
Pallier said recent behavioral experiments have shown people have an advantage at relearning a language they were exposed to as children. A study in 2009 led by Jeffrey Bowers found that native English speakers under forty years old could relearn Hindu and Zulu sounds more quickly if they had been exposed to them for 4-10 years during childhood, mostly through their nannies. “The only explanation” for these findings, he said, “is that you have kept something” in the brain. “The cool thing” about the new study is that the subjects went in cold, without learning any of their lost first language beforehand, unlike in earlier studies. “I think it’s a step forward,” said Pallier.