Acquiring sign language may be more innate than learned

Research on sign language supports theories that humans have a natural ability to structure language.

by Beth Azar
Monitor staff

For centuries deaf and hearing-impaired people have devised sign languages to communicate.And for some time, researchers debated whether such languages were distinct from oral language. Most now agree that they are. Indeed, the world?s sign languages tend to have more in common with each other than with the spoken languages of their country of origin and are as complex grammatically as some of the most complex spoken languages.

With that debate over, researchers have turned their attention to standard questions of language development, namely, how much of sign language acquisition is learned and how much is innate. They?re finding that at least some structural aspects of sign language are due less to parental input than a child?s own innate ability.

Remarkably resilient

For example, University of Rochester psychologist Elissa Newport, PhD, has used her research on sign language to ask whether children have an innate ability to form more complex languages than adults. In particular, she wanted to examine language development in children exposed to imperfect or abnormal language input.

In the hearing community, this may happen when two cultures with different languages begin to interact. Often, adults of the cultures develop a 'pidgin' language?a cobbled together combination of two languages that is usually simple and grammatically limited. If the two cultures begin to intermarry, the resulting children often begin speaking a 'creole' language based on the pidgin language they?re hearing from their parents, but far more complex.

To study how children move from the simple language of their parents to the more complex creole, Newport and her colleagues Jenny Singleton, PhD, and Danielle Ross, PhD, examined ASL acquisition in deaf children whose parents were late learners of the language. In earlier work, Newport established that late learners of ASL use a far less complex version of ASL than native speakers. Similarly, the parents in these later studies used few complex grammatical constructions, and those they did use, they used inconsistently. The children were exposed to ASL only at home so had to learn the language based on imperfect guidance from their parents.

One aspect of language the researchers examined was the use of a morpheme?parts of signs?that distinguish verbs of motion. A comparable structure in English is the addition of 'ed' and 'ing' to the end of verbs to mark time.

In one study of hearing parents with deaf children, Ross and Newport found that parents used these ASL morphemes correctly about half the time. The rest of the time they either omitted the morpheme or replaced them with an ungrammatical form. Despite this inconsistent input, children appear to learn to use the morpheme correctly within the same developmental window as native ASL speakers.

Native ASL speakers gradually learn to use this morpheme by about age 8. And between ages 8 and 10 children who learn ASL from their non-native speaking parents use it almost as well as native speakers between those ages. Even one child, Sarah, whose mother only used the correct morpheme 20 percent of the time was able to use the morpheme correctly most of the time.

These children may also be able to build certain aspects of grammatical structure into language with no input at all, says Newport. ASL allows speakers to move phrases to the beginning of a sentence and mark them as the topic of the sentence with a facial expression. In a study by Singleton and Newport, the non-native speaking parents of one child, Simon, used this rule in a small subset of the many ways it can be used. However, when the researchers tested Simon?s language comprehension of this grammatical marker, he scored perfectly. His parents, on the other hand, only understood the forms they use themselves.

'The learning process survives a lot of noise,' says Newport. 'Children certainly must learn many aspects of language. But they also come with innate biases about how languages should be organized.'

Grammar out of thin air

University of Chicago psychologist Susan Goldin-Meadow, PhD, couldn?t agree more. She?s spent her career studying deaf children who aren?t exposed to any formal sign language but who, nonetheless, spontaneously develop signs. For more than 20 years she?s studied deaf children of hearing parents in the United States, finding that the children and their parents often make up gestures with which to communicate. Even more intriguing is that the children?s gestures are organized into a rudimentary grammar while their parents? are not. This implies that the children are imposing a consistent structure on their signs, independent of what they?re learning from their parents, says Goldin-Meadow.

Recently, she and her colleague Carolyn Mylander extended these findings with a crosscultural study that examined spontaneous signing in Chinese children. They found that the Chinese children organize their signs in a consistent, grammar-like pattern that looks much like the pattern the American children use.

In particular, the children structure their signs in what?s called an ergative pattern, as opposed to the accusative pattern used in English. The difference between the two patterns is in how people mark the nouns in sentences that use transitive and intransitive verbs. Both ergative and accusative languages distinguish between the subject and the object of the verb in transitive sentences. For example, in the sentence 'the mouse eats cheese,' English distinguishes between the eater (subject) and the eaten (object) by placing the eater?mouse?before the verb?eats.

The difference between accusative and ergative languages comes with intransitive sentences, as in 'I ran to the corner.' English marks the runner the same way it marks the eater in the previous example, by putting it before the verb. But ergative languages?such as Chinook and Georgian?mark the runner like the eaten. If English were ergative, we?d say, 'ran me to the corner,' with the runner following the verb. The Chinese and American deaf children structured their signs according to an ergative pattern despite the fact that their native countries? languages are not ergative, Goldin-Meadow finds. In fact, the American children looked more similar to the Chinese children than they did to their own parents who showed no pattern to their signing, she says.

In contrast, Chinese mothers tended to follow the same pattern as their children. One interpretation is that Chinese children learned the pattern from their mothers, says Goldin-Meadow. But since Chinese and American children?s gestures follow the same patterns, it seems more likely the Chinese mothers learned from their children, she says.

'Given the salient differences between Chinese and American cultures, the structural similarities in the children?s gesture systems are striking,' write Goldin-Meadow and Mylander in their article, published in the journal Nature (Vol. 391, p. 279?281).

And although the debate over whether grammar as a whole is innate is far from over, both Newport and Goldin-Meadow believe their research in deaf children supports arguments that the structural aspects underlying language are something we are all born with.