Teaching babies to err

Infants’ flubs on a hiding game may arise from a natural learning strategy

Babies may look like they’re stuck in the here and now, but new research suggests they use what adults do and say as opportunities to generalize about the world.

Oddly enough, the findings link infants’ readiness for social learning to their tendency to make what looks like an obvious error, at least to adults. Babies will keep looking in the same hiding place for a toy after it’s been moved elsewhere.

This puzzling behavior, studied by scientists for more than 50 years, hinges on babies’ natural tendency to generalize about the meaning of social signals they receive from adults, say cognitive ethologist József Topál of the Hungarian Academy of Sciences in Budapest and his colleagues.

Topál’s group proposes that social signals from an experimenter repeatedly placing a toy under a cup allow 10-month-olds watching to make a practical inference, akin to “this kind of object is found under that cup.” Upon seeing the toy get hidden under another cup, a few seconds later the infant acts on his or her previous knowledge and reaches for the first cup, the researchers conclude in the Sept. 26 Science. This effect is called the “A-not-B error,” since babies choose the A cup after the toy has been moved to the B cup.

But if the experimenter avoids looking at or otherwise communicating with an infant during the hiding game, the child frequently chooses the B cup after a switch, the scientists find. Absent any social signals from the adult, infants interpret the task as a here-and-now event and search for the toy where it was last seen, the team proposes.

“Although a disposition to interpret others’ communication signals prepares infants to learn, in certain situations it can misguide their performance,” Topál says.

Researchers had earlier ascribed infants’ errors on the hiding game to memory deficits, to an inability to stop reaching toward the first cup and to a preference for reaching toward the original hiding spot.

Topál’s results partly explain the A-not-B error, remarks cognitive neuroscientist Adele Diamond of the University of British Columbia in Vancouver. Infants’ inability to stop behaving in ways that have worked in the past, even when the scenario is different and demands new behavior, also contributes to the effect, Diamond says.

“The A-not-B error actually marks a major achievement in cognitive development,” says psychologist Alfredo Pereira of Indiana University in Bloomington. At around 10 months of age, infants start to build stable memories of past events, although they’re not yet able to call up these memories flexibly, adjusting for changes, Pereira asserts.

Topál’s team divided 42 infants, all age 10 months, into three groups.

In one group, an experimenter established eye contact with each child while smiling and saying, “Hello baby, look here!” Then the experimenter hid a toy under one of two cups while looking back and forth between the infant and the cup. After four seconds, the experimenter pushed both cups forward so that the infant, sitting on the mother’s lap, could reach for either one. This process was repeated four times. The experimenter then hid the toy under the other cup and waited four seconds before pushing both cups forward.

In a second group, the experimenter played the same hiding game with infants while looking down, maintaining a neutral expression and saying nothing.

In a third group, the experimenter manipulated the cups and the toy from behind a curtain. At various points during the game, infants could see only the experimenter’s hands.

Among infants the experimenter communicated with, 12 of 14 persisted in reaching for the first cup after the toy had been moved. Only 6 of 14 infants in the second group, who received no communication, made the same error and 5 of 14 in the third group, who didn’t even see an experimenter, chose the wrong cup.

Infants denied social cues could still make some errors, and these mistakes may have also reflected limitations on memory and other mental faculties, the researchers suggest.

When infants repeatedly see a toy put under one cup and reach for that cup after short delays, it creates a stable memory. When the toy gets put under another cup, infants reach for the first cup based on their stable memory. Interpretation of others’ social signals need not play a role in this process, Pereira argues.

Bruce Bower has written about the behavioral sciences for Science News since 1984. He writes about psychology, anthropology, archaeology and mental health issues.

More Stories from Science News on Humans