Baby's first robot

If you could only learn a language with the innocent receptivity of a young child.

That adage, repeated ad nauseam, once an adult has decided to learn French or Tagalog engenders endless debate. Is it possible to create a teaching method or mental state that rewires the brain in a way that allows the seamless production of words, syntax and intonations typical of a native speaker? Should adults, equipped with executive faculties that a baby won’t fully access for decades, even try to replicate the pliability of the sounds and semantics of the child's new world?

The bigger question, though, is how does a baby or toddler do it at all? There’s the endless uttering of the Ba Bas and the Ga Gas, as if an infant is turning the dials on a staticky analog shortwave radio to look for infant hour on China Radio International or Deutsche Welle.

How does ba-ba-buh-baby tune into the ultimate social network, the global nation of the spoken word? Neuroscientists are still working on that one.

One new entry trying to address a piece of that puzzle is a recent Journal of Neuroscience article that looks at what happens when baby is turning those mental knobs—and what happens when researchers intervene to help out with the tuning process.

Rutgers neuroscientist April Benasich and her research team had a few dozen infants who had just reached their four month birthday listen to short, rapidly changing patterns of sounds—rhythmic tones, chirps and the like—intended to help their brains process the types of acoustic timing cues that distinguish one syllable or word from another.

Babies who took an active role in focusing on these tiny sound differences had quicker and more accurate brain responses at seven months of age when they heard new and increasingly complex sounds than did a group who only listened passively or others who had never heard the sounds.

Babies in the active group were trained to pay attention to these sound cues. If they listened carefully, they were rewarded with a short snippet of an amusing video. The babies in this group would then undertake this routine on their own, tracking the ever-faster and more complex sound barrage in exchange for a bit more baby flick. All the while, the investigators tracked electrical activity in the babies’ brains, which showed that the active group could better focus on the critical acoustic important for language learning.

Benasich is now developing an interactive game, replete with video snippets as rewards, that would let infants listen to the retinue of changing sounds and then signal they can hear the differences by moving their eyes within a given interval to a designated spot on a small robot figure, The prototype looks vaguely like a cybernetic Mickey Mouse (photo).

At Rutgers, Benasich was a post-doctoral fellow with Paula Tallal. Along with Michael Merzenich from the University of California San Francisco, Tallal developed exercises using speech sounds to help kids learn better. The intent of a baby brain game would perform a similar role, particularly for kids at risk for language difficulties.

Brain games' present karma is not the greatest. Studies have cast doubt on their effectiveness for both children and for older adults at risk for cognitive decline. A group of 69 scholars—put together by the Stanford Center for Longevity and the Max Planck Institute for Human Development— issued a statement on Oct. 20 that noted that brain games for older adults have “no compelling scientific evidence” for their effectiveness. Some review studies have come to similar conclusions for the children's games.

For your average baby, it’s possible that an involved parent may suffice. The exaggerated way most parents talk to an infant—"Whooose the preteee baaybeee?"—may help ready the child for learning first words. Patricia Kuhl from the University of Washington and colleagues published a small study earlier this year that showed that the higher pitch, slower tempo and the overemphasized tones of “parentese” are correlated with future language development

In principle the robot brain game could have advantages over just talking to baby. It could focus the child on the elemental building blocks of language without a bias toward Spanish, French or Tagalog. That would allow the natural plasticity of the young child’s brain to adapt to more than one language. And Benasich thinks that the game will be better than just a "preteeee bayaybee" in cultivating development of the neural circuitry needed for processing speech.

A consensus on how the first year of life serves as boot camp for the gabfest to come has yet to emerge. Going from incoherent babble to “Daddy, why is the sky blue?” will persist as a challenge for the neuroscience intelligentsia.


Image Source: April Benasich