Baby bookies

Robert Adler

“Every child is born a genius,” my mother liked to say, usually with the rueful implication that it’s life’s hard knocks that dim that initial promise. A recent study reminded me of her firm belief in the innate brilliance of babies.

Here’s the core of the study:

A one-year-old baby sits on its caretaker’s lap in front of a computer screen. A movie runs for a few seconds showing four objects—three identical and one different--bouncing around inside a circle with an opening at the bottom. Most adults would recognize this as depicting a very simple lottery, but an infant, one assumes, knows nothing about such things.

The scene blanks out for a second. When the circle re-appears, one object tumbles out, leaving three behind.

The question is, can a one-year-old differentiate between outcomes that are more or less probable, that is between trials in which one of the three identical objects exits the circle versus less likely sequences when it’s the unique object that falls out?

If you predicted that one-year-olds can perform this remarkable probabilistic assessment, congratulations. Your high expectations for the cognitive abilities of infants have just been confirmed by an elegant series of experiments carried out by Ernõ Téglás and Luca Bonatti, cognitive scientists at the International School for Advanced Studies, in Trieste, Italy, and their colleagues, detailed in a recent paper (Proceedings of the National Academy of Sciences, November 27, 2007).

The researchers found that one-year-olds looked at improbable outcomes significantly longer than at probable ones—for 12.5 versus 9.3 seconds on average. The length of time infants pay attention to an event or scene is a well-established way of determining what gets and keeps their attention, for example things that are unexpected or surprising.

In other words, if one-year-olds could wager, they would be able to bet on the more likely outcome of this lottery based on an apparently innate sense of probability.

The team used carefully designed control sequences to see if the infants might be responding to non-probabilistic aspects of the scenes such as the objects’ shapes or colors. These tests confirmed that the babies looked longer at the unlikely outcomes simply because they violated their intuitive expectations.

“We found that infants can form these expectations, without any need to experience the outcomes of the lottery before,” says Bonatti. “Thus, they can reason about the future independently of their knowledge of the past.”

The researchers went on to show that similar intuitions about probability shape the responses of children at least through the age of five. At three, in fact, the children’s intuitive expectations—in this case anticipating whether a ball will escape from a box through the side with one hole or three holes--trumped what they actually experienced. Three-year-olds continued to favor the intuitively more probable outcomes even when the experimenters made them happen infrequently.

It was not until age five that children’s intuitive sense of probability could be modified over time by the actual frequency of the events they experienced.

So, not only are intuitions about probability in place by the age of one, independently of experience, but they turn out to be both strong and long-lasting.

These new findings add a surprisingly high-level skill to a long list of what Bonatti characterizes as the “stunning cognitive abilities” of babies and young children, revealed by research over the past few decades [New Scientist, 17 May 2003, p 42].

Bonatti notes that by four months of age, babies can differentiate between small numbers of objects, and even respond to addition or subtraction. Infants respond differently to “possible” and “impossible” events. They can categorize objects based on multiple characteristics simultaneously. They know that animate objects can do things that inanimate ones can’t, and that what they do is driven by goals. Most recently, researchers found that six-month old babies can infer character from behavior (New Scientist, 22 November, 2007).

Bonatti is convinced that we humans come into the world with vastly greater and more structured cognitive capabilities than suggested by John Locke’s blank slate or William James’ “buzzing, booming confusion.”

Still, many current developmental theories come down strongly on the “blank slate” side of the necessary interaction between innate capabilities and experience. For example, a widely accepted theory argues that the ability to assess probability necessarily requires previous experience with similar events. “This theory basically says that our ability to predict the future is entirely shaped by our knowledge of the past,” says Bonatti.”

Bonatti does not deny the importance of experience, but his findings emphasize what even very young babies bring to their encounter with the world. “Over and above our abilities to learn from experience, we also have logical and conceptual abilities that do not seem to derive from the world,” Bonatti says. “But, after all, why should this be surprising? We have ears, noses, eyes; why not also cognitive structures for reasoning about the future?”

Why not indeed?

If, as it seems, Bonatti’s findings show that there’s a lot more going on in the mind of a baby than meets the eye, or that modern-day “blank slate” theorists think possible, then it behooves us to look for even more high-level skills, and at even earlier ages. Like astronomers hunting for extra-solar planets, once our instruments allow us to find one, far more are sure to follow.

Ever since Copernicus dethroned the Earth from the center of the universe and Darwin showed that we are more closely related to apes than angels, science has dealt one blow after another to the human ego. In the face of this, I, for one, am thrilled to learn that human infants, at least, have much to crow about.

Not to mention that sometimes mothers get it right.


Robert Adler is a psychologist and science writer currently in Oaxaca, Mexico


Babies can spot the good, the bad, and the ugly
New Scientist, 22 November 2007
http://www.newscientist.com/channel/being-human/dn12948-babies-can-spot-the-good-the-bad-and-the-ugly.html

Bonatti and Téglás’ contact information
http://www.sissa.it/cns/lcd/members.htm

International School for Advanced Studies
http://www.sissa.it/cns/

Proceedings of the National Academy of Science, November 27 2007:
http://www.pnas.org/cgi/content/abstract/0700271104v1?maxtoshow=&HITS=10&hits=10&RESULTFORMAT=&fulltext=Bonatti&searchid=1&FIRSTINDEX=0&resourcetype=HWCIT

The movies seen by the one, three and five year old children:
http://www.pnas.org/cgi/content/full/0700271104/DC1#F5

What every baby knows
New Scientist, 17 May 2003, p v2
http://www.newscientist.com/article/mg17823955.400-what-every-baby-knows.html