discovered in 1970 after she had been kept locked in a bedroom and tied to the furniture for the first fourteen years of her life. These unfortunate children had almost no exposure to language, among other things, and provided an unusual opportunity for study. Neither ever achieved normal language skills. The fact that deaf babies do not automatically learn to talk also tells us that the skill is not purely innate. Yet if a deaf babyâs parents are fluent signers, he or she becomes a native user of sign language, which adheres to universal grammar, and the baby will follow the same path to fluency with visual language as hearing babies do with spoken language. So while speech may not be innate, certain patterns of language learning seem to be.
âLanguage is a super-interesting learning problem,â neuroscientist Elissa Newport told me as we sat in her new, somewhat bare office at Georgetown University Medical Center. Newport specializes in the acquisition of language. For twenty-three years, she was at the University of Rochester, the last twelve as chair of the Department of Brain and Cognitive Sciences. When I met her, she had just moved to Washington to head a new center for brain plasticity and recovery, where she is studying how young children who suffer a certain kind of stroke recover their language. Her straightforward, no-nonsense style is evident in everything from her short hair to her scientific approach. âWe know that languages of the world have a certain type of organization that you donât see in any other speciesâ communication system, and languages of the world have a lot of interesting profound similarities to one another. So thereâs a very interesting problem to explain: How did we get languages like that, and how do you learn them, and what kind of brain mechanisms are required to do that?â
When Alex was turning two, I hadnât yet realized how important the question of brain mechanisms would be. I was fixated on the second of Newportâs questions: How do you learn language? I sought out Newport because I wanted to understand not just what Alex couldnât do but also what other children could do and why. Following the lead of scientists, I knew I needed to understand what was typical in order to better make sense of what was atypical. What I found was worrisomeâfor what it was clear Alex didnât get as a babyâbut also a little bit reassuring, as I began to appreciate that he had managed to learn some important things about language with very little help from sound.
One fact scientists agree on isnât surprising anymore: To learn a language, itâs best to start young. âWe certainly learn languages as adults but not to the same degree of proficiency,â says Newport, âand thereâs much more variation among individuals as we get older.â To say that children are âbetterâ is too simplistic. âYoung children donât really learn faster or better,â says Newport. âThey learn more slowly. Itâs kind of tortoise and hare. If you look at people who move to a new country, adults are generally faster, they just donât get as far. They do it differently and they donât end up as good.â
Newport gave me an intriguing example. She has spent much of the past ten years making up what she calls âminiature languages.â In different studies, the same eight verbs and fifteen nouns carry different meanings. So âkleidumâ means âdragâ in one instance and âhead-buttâ in another (one can only imagine the story that language will tell). Words like âtombat,â ânagid,â and âmelnawgâ might mean âsingerâ or âbaby carriageâ or âshopping cart.â Newport teaches these limited strings of invented words to babies, children, and adults in her laboratory using pictures and videos. They are then tested on