Distinguished Lecture: Professor Eve V. Clark: Language Acquisition and Expertise

 

The 2017 Bing Distinguished Lecture, held in East Room on May 11 and titled “Language Acquisition and Expertise,” certainly lived up to its billing. Stanford linguistics professor Eve Clark is one of the world’s leading authorities on how children acquire language. Since coming to Stanford in the early 1970s, Clark acquired much of her own expertise at Bing, and she began her talk by thanking the school’s faculty and families for supporting her research over the years. Bing afterwards honored Clark on the occasion of her retirement with a reception at the Tower House. 

 

In her valedictory Bing lecture, Clark showed how a young child learning to speak, an Olympic athlete and a chess grandmaster depend on similar learning processes to acquire expertise in their respective fields. Clark then followed up her lecture with a Q&A about baby talk, bilingualism and how parents and caregivers can best support their little language learners.

 

 

Nobody is born speaking, any more than Serena Williams was born playing Grand Slam tennis. Modern studies show that no matter how much innate talent anyone has, this alone is not enough to develop true mastery. It takes hard work and the right conditions. Whether it is a child learning to speak or a future world champion learning the game of tennis, gaining expertise in any field requires extensive early exposure, immediate feedback from an expert and deliberate practice.

 

She compared the process by which children learn to speak to the path chess players take to become masters, drawing on the findings of Herbert Simon and William Chase in the 1970s at Carnegie Mellon University. In a 1973 article in American Scientist titled “Skill in Chess,” they stated:

 

 

Simon and Chase studied chess because it required no motor skills, but in the 1990s Simon’s student Anders Ericsson began studying what it takes to become an expert in other areas. Ericsson studied skills ranging from sports and music to typing, mathematics and X-ray diagnosis. In all these fields, he found that the learning process was essentially the same.

 

What is expertise? Clark showed a number of examples of child experts in various fields: 8-year-old Samuel Reshevsky playing a roomful of bearded chess masters in Paris in 1920; 14-year-old Nadia Comaneci scoring a perfect 10 in the 1976 Montreal Olympics; 6-year-old violin virtuoso Jascha Heifetz in 1907; and a 4-year-old dinosaur fanatic. A baby learns to talk, she says, in the same way that these other child experts acquire their skills. Expertise, however, is relative. Even the most gifted child prodigies differ from each other and will not attain their full powers until maturity. 

 

And it isn’t just any 10,000 hours of practice: Ericsson concluded that to master any domain, intellectual or physical, the type of practice itself mattered as much as the amount of time spent practicing.

 

 

So how should those 10,000 hours be spent?

 

Clark showed that, just as in chess, expertise in language requires:

• Guided exposure: Listening to adult speech from birth is the Suzuki method of early language acquisition. 

• Immediate feedback from an expert: When a child makes an error, the adult immediately gives feedback, and the child corrects the error.

• Deliberate practice on one’s own, consciously and frequently: That cute baby babble is actually serious work.

• Monitoring of performance: Children have to become conscious of their own errors so they can correct them on their own.

• Extended immersion: Passion for a subject makes for greater expertise.

 

 

The type and timing of linguistic exposure is essential for success. The more engaged speech directed to a child in the first years of life, the better the outcome. Overheard conversation, watching TV and other passive auditory stimulation simply don’t cut it. What counts is an adult speaking directly with the child before the age of 4. Children who hear more child-directed speech by 19 months can recognize familiar words faster than children who have heard less.

 

To illustrate her point about the necessity for engaged, child-directed speech, Clark cited a fascinating case of a hearing child of deaf parents. The parents had had their child watch TV in an attempt to teach him spoken language, but at age 3, the child could produce only ad jingles. But after a year of intensive one-on-one language interaction with an adult at preschool, the child began to catch up with his peers. 

 

Socioeconomic factors play a big role in the quality of this critical early exposure. For example, a 1995 study by Betty Hart and Todd Risley showed that the number of words spoken per hour to a child varies enormously with the parents’ socioeconomic status, with higher status parents talking more to their children. A follow-up study showed that greater engaged exposure at the age of 3 predicted greater verbal ability at age 10.

 

One of the most surprising findings was about day care vs. being cared for at home. Parents are often concerned about having young children in day care, but studies show that, at least when it comes to acquiring language, children cared for outside the home may have an advantage. After one year, children in day care had greater vocabulary and sentence complexity than children at home, possibly because they had more practice as a result of explaining themselves to caregivers who did not know them well, and then talking about their day to their parents once they got home. However, noted Clark, that this study was carried out in France, where day care must meet strict national standards for adult-child ratios and quality of care.

 

 

Immediate feedback on performance is a critical requirement for developing expertise, and the best feedback is immediate, clear and usable by the child. Like any good coach, adults combine positive and negative feedback in a way that scaffolds the child’s meaning and keeps the conversation going. 

 

Positive feedback expands on what the child is trying to say. For example, when a child says, “Nose,” the parent replies, “He has a pointed nose, doesn’t he?”

When a child makes a mistake, adults rarely stop the conversation to say, “No, that’s wrong.” Most adults reformulate the child’s speech to check that they have understood the child’s meaning (“Do you want me to put this back? There you go.”) or to provide embedded corrections that likewise preserve the flow of conversation. For example, when a child says, “Don’t fall me downstairs!” the parent reassures, “No, I wouldn’t drop you downstairs.” 

 

Clark discusses a scenario where a 1 1/2-year-old, looking at a picture of an owl, says, “Duck, duck.” The mother reformulates by saying that it’s a bird called an owl that goes “hoo, hoo.” The child learns a new category: bird; a new word: owl; and a distinguishing property of owls: hoo, hoo. 

 

Most children’s mistakes involve pronunciation, choosing the right words or using the correct syntax or grammar. When adults reformulate these errors, children will either repeat the corrected form offered by the adult or acknowledge it with an “uh-huh,” though sometimes the child will reject the adult’s interpretation (it’s a misunderstanding) or opt out of the conversation. Clark’s research shows that a high percentage of children’s language errors are reformulated (up to 60 percent at first, decreasing with age as children make fewer and fewer errors).

 

 

Does your toddler babble nonstop? Mastering anything takes practice, practice, practice on one’s own, and those bedtime monologues and pretend-play dialogues have a key role in learning to speak. Clark cited studies by Ruth Weir that showed how children intensely rehearse words and role-play speech to themselves and with each other. Children will repeat conversations and combinations over and over, almost like a language lab exercise.

 

In pretend-play dialogue, children are very sophisticated. In studies by Elaine Andersen, carried out at Bing, children can differentiate roles by making deep voices for males, higher voices for females, and high-pitched baby talk for children, assigning different syntax, accent, status and vocabulary to the different roles. They also differentiate stage directions to their playmates from the main dialogue by putting the directions in the past tense (“I was the mother, you were going to the ball”). 

 

Practice makes perfect.

 

 

 

External feedback is essential for developing expertise, but so is the ability to monitor one’s own performance, recognize a mistake and correct it. When children practice speaking, they mentally compare their own utterances with what they have represented in memory for adult versions, and then they try to make any repairs needed in order to match the adult versions as closely as they can. They can tell the difference between how they say things and how an adult does, even if developmentally they can’t yet manage to produce the adult version themselves. 

 

When asked why children who acquire a second language early often have no accent, while adult learners always seem to have one, Clark suggested that something changes in a child’s motor development that affects their ability to make fine adjustments to pronunciation. Although there is no time limit for acquiring another language, there does seem to be a time limit—about the age of 13—for acquiring one without an accent. 

 

 

By the time they are 2, children are initiating two-thirds of their conversations with adults. By the time they are 4, they are playing word games, clamoring for stories, telling bad jokes and making bad puns. As in any field of skilled performance, the more time spent immersed in the topic with extensive feedback, practice and monitoring, the greater the eventual skill. By the time children are 4, they usually have spent over 10,000 hours practicing language, but they won’t reach adult-like linguistic mastery for a long time: In 10 more years, they will be fairly competent, but even among adults, there is extensive variation in individual linguistic ability. 

 

To get an idea of how immersed a child is in language acquisition, compare the average 3,640 hours a year a preschooler spends practicing a first language with the 180 hours per year averaged by a college student taking a second language in the classroom. A college-level course of one hour a day plus one hour a week in language lab still amounts to only 5 percent of the language immersion and practice a child of 4 has achieved. 

 

 

Among the diverse audience in East Room listening to Clark’s talk, the topic of bilingualism was of intense interest. 

 

Many parents wonder whether learning a second language from birth will hinder learning the first language. Teachers in progressive, diverse Bay Area schools sometimes still advise parents to avoid introducing a second language too early, but Clark has some counter-advice: Ignore this. 

 

Studies show that up until 30 months, children who are introduced to two languages from birth are just like monolinguals, only doubled. They have vocabularies and speaking abilities comparable to monolingual children of the same age in both languages. They often understand more words in both languages than their monolingual counterparts.

 

After the age of 30 months, when children start going to outside day care or are exposed to wider social situations, it’s hard to maintain a perfect balance of engaged exposure in both languages. Clark’s message to bilingual parents is that they should continue to speak and read with their children in the home language as much as possible. Children benefit from early engaged exposure to language in any form. Finally, she pointed out that the majority of the world’s population is bilingual or multilingual.

 

 

 

Professor Clark answered questions from the audience and later spoke with the author, who has condensed and edited the following Q & A.

 

 

“Baby talk” is often parents pretending to talk with their babies until the children are ready to respond in real conversations. At first, parents simply talk at their babies in a one-way conversation: “Do you need a diaper change? Let’s get you clean. Are you hungry? It’s nearly time for dinner.” When talking with their children, from about age 1 on, parents may use simpler language or substitute child-directed words like “kitty” or “meow-meow.” However, if the parent says “cat,” the child will often point to a picture of a cat even if the child is still saying “meow-meow.” 

 

One of my students did a study that showed that even if parents deny ever using baby talk, they will often alter their speech when talking with their young children to give them instructions, for instance, or to explain something new. The higher pitch used by parents in many cultures when speaking to children is an example of how such talk is really a way to alert the child that the speech is directed to the child, not to someone else.

 

 

It has become popular because some parents feel it helps children communicate their meaning earlier, and so avoids frustration. I see it as just another form of language, and it should not affect language acquisition one way or another as long as the parent is speaking while giving the sign. As babies acquire spoken words, they generally drop the baby-signs.

 

 

Not really, since most reformulations happen when adults are trying to clarify what a child means or to expand on something they said. Often, adults are not even aware they are offering corrections to the child.

 

 

There is a great deal of resistance to bilingualism in this country. An elementary school teacher once told the mother of one of my son’s friends that she should not speak Italian to her son to avoid confusing him or hinder his progress in English. I told her to ignore this advice and to keep speaking and reading to him in Italian as much as possible. I nearly went to see the principal to talk about changing this kind of thinking.

 

Young children, even if they do not speak English well at first, will catch up very quickly without any special preparation once they are immersed in an English-speaking school or preschool where most of the other children speak English. They will be just fine, unless there are unusual circumstances.  

 

Children almost never confuse languages. They know exactly what words and constructions belong to which language. Acquiring two languages at once is not usually a problem, and knowing a second language is an enormous benefit.

 

 

Young children learn through normal conversation, talking about everyday things with their caregivers and playing with peers, not in language classes. To acquire a second language simultaneously with the first, children often associate a language with a particular person or place. It is important to have a “constant speaker” like a parent or a caregiver who speaks to the child only in the other language, or to have a defined space where that language is spoken. For example, the child knows to speak English at school and only Mandarin at home. It is also important to read in the home language as much as possible. Children are also motivated to learn something if they see that it is relevant in their lives—for instance, when traveling or having visitors from abroad.

 

True bilingualism, when a person feels equally comfortable in either language, is certainly possible, although a person may not be fully fluent when speaking about a specialized domain (like linguistics) in both languages. Nobody is ever perfectly bilingual on every topic. 

 

Anyone can learn a second language at any time, although adults and children learn language differently. There is no age limit to becoming bilingual, although there is a time limit to sounding like a native speaker. After the age of 10–12, it is highly unlikely that you will avoid having some sort of accent. There is something about the way children monitor adult sounds and fine-tune their own pronunciation that changes or is lost after that age.

 

 

Talk with your children, a lot. Listen to your children, a lot. Read with your children, a lot.