News & Events

by Danny Lachance

Yang Zhang’s mother was probably committing a crime when she taught him, at the age of five, some basic English phrases—“My name is Yang" and “The pot is under the table." China, where he grew up, was closed then, and instructing children in a language as emblematic of the West as English was forbidden.

But Zhang doesn’t remember a sense of trepidation in learning English so furtively. He remembers, instead, feeling immense curiosity.

“I grew up in the countryside," he says. “We didn’t have television. English was like a different world.” He felt a sense of awe, he says, when he realized that these foreign sounds were, for others, as natural as breathing. He marveled at the way that English speakers could add phrases to the ends of their sentences ad infinitum and still remain grammatically correct—a linguistic feature that was very different in his native language.

That sense of wonderment has endured. Almost 30 years later, Zhang continues to ponder the mysteries of human language acquisition and production. These days, however, he’s doing so with state-of-the-art lab equipment, research assistants, and sophisticated experiments designed to reveal the way language functions in the brains of everyone from Japanese infants to Finnish adults.

Warped sensibilities

No theory has been more compelling and controversial in explaining how language develops in the brain than Noam Chomsky’s.

Language, Chomsky theorized, is seeded in our genes. Thanks to our DNA, the structures of our brains already contain a universal grammar and phonetics— certain rules about what sounds can combine to form words or how words can combine to form sentences—to which all languages, despite their immense variation, must conform.

When we are infants, Chomsky has argued, the sounds of our native language, swirling around in the atmosphere, trigger a particular pattern from the range of possible patterns allowed for by the universal grammar and phonetics. As our language faculties grow, a kind of specialization takes place. The pattern, or language, that has been selected by our environment grows and flourishes.

But for Zhang and his former adviser Patricia Kuhl (Ph.D., ’73), a luminary in the field, Chomsky’s theory seemed to be missing something important. They sensed that Chomsky’s theory was underestimating the effect of the environment on language acquisition. “We don’t dispute that there’s a genetic predisposition for language. But what exactly is being picked up by the infants from the environment is the question," Zhang explains.

Take, for instance, our ability to distinguish and produce sounds. Spoken Japanese, for instance, doesn’t contain the sounds /r/ and /l/ as spoken English does. As a result, by the time they are adults, untrained Japanese speakers cannot distinguish between the two sounds when they hear them. That’s why nonnative English speakers whose first language is Japanese often mispronounce and misspell a word like lollypop.

In a now-famous study, Kuhl demonstrated that the ability to distinguish between native language sounds is gained as early as the first year of life. Japanese and American infants were exposed to sounds both indigenous to and absent from their native language. At seven months, American and Japanese babies were equally good at recognizing the distinction between the English sounds /r/ and /l/. But at 11 months, American infants were significantly more likely to recognize the distinction between these sounds than Japanese infants were; American infants’ ability to tell /r/ from /l/ increased over time, while Japanese infants’ ability to distinguish those sounds decreased.

Ambient language, Kuhl and Zhang agree, plays a much larger role than simply selecting one pattern among many that are latent in our brains. Sounds unique to a language appear to be mapped onto the brain over time, and mapping, in turn, makes the sounds easier to recognize and, eventually, produce. In a way, language warps our brains, reshaping them in ways that go beyond simply calling on what is already there.

Babbling may be part of this warping process. “Babies can’t avoid babbling," Zhang notes. “They can’t avoid imitating whatever is given to them in the environment if it sounds like their native language or whatever their mother or father is trying to speak to them. They’re trying to imitate."

And by imitating, babies are rehearsing, listening to, and then encoding the distinction between sounds, a distinction they will eventually need in order to distinguish words from one another. “More or less, children are teaching themselves in this way. Their mind is actively involved," Zhang says.

Muddled Multitasking?

Zhang’s current work is dedicated to understanding more precisely how the brain processes ambient language. With this information, he hopes, researchers will be able to identify exactly why the language-related symptoms of autism and dyslexia occur. That knowledge, in turn, may become the basis for successful treatments or cures.

Zhang hopes to contribute to our understanding of the relationship between the brain’s processing of nonlinguistic (or paralinguistic) qualities of language sounds, like the timbre and pitch of a voice, and its processing of those same sounds’ linguistic qualities, like syllables and sentences. In a recent study, he tested the relation between our ability to identify the (paralinguistic) gender of a speaker and our ability to distinguish between different syllables. He connected subjects to noninvasive electrophysiological equipment that could measure the frequency and location of their brainwaves, and then exposed them to linguistic sounds.

He found that when he asked the subjects to identify the gender of the speaker, gamma waves appeared in a specific region of the brain in ways particular to the gender of the voice the subjects were hearing. When he asked subjects to focus on the kinds of native language syllables being spoken to them, though, their brain activity shifted: waves of a different frequency appeared in a different region of the brain. Zhang also found that we are much more accurate at recognizing gender than syllables, suggesting that gender recognition is a more primitive, primary process.

By mapping these differences, Zhang hopes to bring us one step closer to understanding how and where the brain processes the large amount of information that is thrown at it all at once. Since these processes occur separately, each requires the allocation of resources in the brain. And they have the potential to interact with one another.

“The brain has a very sophisticated way of dealing with different types of linguistic and paralinguistic information," Zhang says. That sophistication and complexity enables us to do what other animals cannot do. But it may also become a liability; greater complexity taxes the brain more and introduces more points in the process where things can potentially go awry. For those with language disabilities, Zhang hypothesizes, the more primitive ability to recognize gender differences may siphon resources from or interfere with the more advanced ability to recognize syllables.

“For children with autism, voice and gender recognition may interfere with their ability to recognize words," he says.

It’s only a hypothesis at this point—one that needs to be tested by extensive comparative studies of normal and language- disordered populations. But Zhang is convinced that researchers aren’t paying enough attention to the way that the processing of paralinguistic sounds may affect children’s acquisition and production of language.

“It’s not just about the message itself," he says. “It’s about the speaker, too."

A History of Innovation

Dr. Kuhl presented the keynote address to honor Shevlin Hall’s 101st year in April 2007. During her visit Dr. Kuhl received the Outstanding Achievement award in recognition of her exceptional career and profound contributions. This is the most distinguished award given to alumni of the University of Minnesota.