Kids Are the Champs at Learning Language

It’s a given that some kids are better than others at throwing a baseball or memorizing a Shakespearean soliloquy.  Whether it’s due to inherent talent or education, or some combination of both, might be open to question.  But there is one thing that all children are amazingly good at, no matter how good or bad their teachers are: learning language.

On the face of it, it may not seem all that remarkable that children learn language – they all do it, and more or less around the same time when they’re young.  But when you think about how difficult it is to learn a second language as an adult (many of us struggle with learning Spanish verb tenses, walking around in earphones endlessly listening to Berlitz tapes), the ease with which young children learn the incredibly complex structure of language is extraordinary.  Elissa Newport, the George Eastman Professor of Brain and Cognitive Sciences and Linguistics at the University of Rochester, has spent her career trying to understand how children are such expert language learners and even creators.

“Children can be a force in changing and forming languages,” she said.

In a talk entitled “How Children Shape Language: Language Acquisition and the Emergence of Signed and Spoken Language” Newport discussed her research on language formation and learning, highlighting the unique avenue she has been fortunate to use – sign languages.  Spoken languages have been around for a long time, but sign languages are relatively new, and Newport and her colleagues have been lucky to witness the birth of one sign language in particular, in Nicaragua in the 1980s.  Newport’s talk is part of the National Institutes of Health Behavioral and Social Sciences Lecture Series.

It’s important to note that sign languages, of which American Sign Language is only one,  are fully equivalent to spoken languages in terms of grammatical structure, complexity and geographical variation (in fact, American Sign Language and British Sign Language have very little in common!), and utilize the same brain structures that spoken languages do. “Human language acquisition looks the same, regardless of whether language is spoken or signed,” Newport said.

Sign languages are of particular interest to developmental linguists because they provide the rare opportunity to observe a language starting from scratch and how children learn a new language.  Nicaraguan Sign Language has proven to be the perfect case study, as Newport’s colleagues Ann Senghas at Barnard College and Marie Coppola at the University of Chicago discovered.

Deaf people in Nicaragua were historically isolated from one another and received no education.  After the Sandanista Revolution in the early 1980s, the first deaf school was formed in Managua and deaf people began to socialize for the first time.  Their classes were in spoken Spanish which was completely useless to them, but what happened was magical: they began to converse with one another in their “homemade” signs, and eventually formed a pidgin sign system.  Pidgins aren’t full languages, but they give rise to creoles, which are grammatically more complex and provide a mechanism by which full languages form and expand.  A second generation came along, and the difference was remarkable: these children’s signing was faster, more spatially modulated, and closer to a full sign language.  What’s key here, though, is that despite the poor model they had (the pidgin signing), the kids added the grammatical structure and regularized the inconsistencies to form a much more complex language than what they were given.

This ability to ignore poor input has also been demonstrated by Newport in the lab.

In a series of experiments with deaf children who have hearing parents, she found that despite parents who are poor signers (most of these parents learned ASL later in life, and thus never become fully fluent), the children ignored their parents’ errors and signed correctly most of the time.  An especially convincing case was a mother who signed correctly only 15 percent of the time but whose daughter reached 80 percent, with her mother as the primary language model.  This is astounding, given that adults who are exposed to poor language models do much, much worse.

“Children don’t always learn what they’re exposed to,” Newport said.  In this case, ignoring your teacher might be a good thing.

The other phenomenon at play here is the well-known critical period.  The exact boundaries of this developmental stage are still debatable, but generally it spans from birth to about age six or seven.  After that, even with maximum exposure, children don’t learn language nearly as well.  This applies to secondary as well as primary language acquisition.  Newport conducted studies looking at second language acquisition in hearing children and adults, and the children consistently outperformed the adults.  This is not to say that adults can’t ever learn a second language, only that they are more variable whereas children are uniformly excellent at it, as Newport argues.  It turns out that this linear relationship between age and competency also exists in sign language users: young, native users of ASL perform better on grammatical judgment tasks than do adults, even if both groups have the same amount of exposure to the language.

Newport and her colleagues rightly claim that all children are talented when it comes to learning and, in extreme circumstances, inventing language.  They do so when they have only a pseudo-language to learn from, and when their parents aren’t fluent.  Researchers have made significant strides in establishing the legitimacy of sign languages, and these are only enhanced by the finding of a critical period of acquisition in signed as well as spoken languages.  Future research that Newport plans to conduct involves looking more closely at the statistical regularities of naturally-acquired language, as well as understanding the constraints and biases in these computational processes which may give rise to phenomena like creolization.

Adults may be able to figure out complex math problems or expound on the theory of relativity, but when it comes to learning language, children are the champs.