Duke University Alumni Magazine







WHEN BABIES TEACH
LABORATORY-THEATER
BY DENNIS MEREDITH

Setting the stage: Needham and the "props" for her abstract playlet;

Photo: Jim Wallace

Psychologist Amy Needham uses colorful cardboard boxes, key rings, and a scientist's ingenuity to discover the amazing aptitudes of the youngest infants.

lena Storelli enjoys absolutely the best seat in the house, nestled on her mom's lap at front-row center waiting for the show to begin. In fact, Elena's is the only seat in the house: She will be the sole audience member. Even mom, Anne Storelli, has to keep her eyes shut during the performance.

The white muslin curtain rises to reveal a brightly lit stage with the play's single character--a colored plastic key ring sitting on a yellow box. After a moment, another player appears: A purple-gloved hand emerges silently from stage right, grasps the ring, and, in a surprise ending, pulls it away from the keys as if it were a separate object. The curtain falls on this minimalist drama, and although it has held Elena's rapt attention, she does not applaud. After all, the dimpled, red-haired Elena is only five-and-a-half months old.

A delighted audience of one, whose amazement at objects' behavior is the one source of Needham's research on infant learning
Photo: Les Todd

Psychologist Amy Needham--who produced and directed this abstract playlet--needed no applause to gauge Elena's interest. Instead, Needham had two stealthy undergraduates peeking at Elena through holes in the tiny theater's side curtains, pressing joystick triggers to tell a computer precisely how long Elena's gaze rested on the stage action. Such a peculiar scenario of students-watching-babies-watching-key rings might seem slightly Kafkaesque. But Needham seeks not to create good art, only good science.

In her laboratory-theater, she has presented hundreds of infants with scenarios featuring key-rings, blocks, cylinders, octagons, and other objects. By precisely measuring the infants' reactions, Needham is exploring one of the most intriguing and important mysteries in psychology--how a baby's brain, born innocent of the world's ways, learns to perceive objects and predict how they will behave. To a baby, such an achievement is both remarkable and crucial for cognitive development. Unable to manipulate objects to learn about them, infants even a few months old nevertheless make great strides in figuring out how to process the avalanche of visual information bombarding them into an ordered object-world of cars, cans, clocks, and cats. Such early skills are the necessary foundation on which the baby builds, maturing from mere passive observer to active manipulator of objects.

"This is really a hot research topic right now, trying to figure out the differences between babies' visual abilities and their manual abilities," says Needham, whose work is sponsored by the National Institute of Child Health and Human Development. Discoveries such as hers have called into question long-held theories about infant cognition first developed by the famed Swiss psychologist Jean Piaget half a century ago. "Piaget didn't believe that children had a full-blown concept of objects until around age two," she explains. "He thought babies' own manipulations on objects were critically important for their learning about the world. He didn't think that babies engaged in much observational learning, so that watching their parents drop things would have been meaningless if the baby wasn't doing it himself or herself. But our new studies show that observational learning is important, that babies know a great deal before they're really acting on the world."

To begin to figure out how babies perceive objects, Needham and her fellow psychologists had to first overcome the daunting hurdle of babies' inability to tell experimenters what's going on inside their darling noggins. Very young babies couldn't even give researchers clues about their perceptions through such overt actions as reaching for objects. So, Needham and other scientists have come to depend on the simple fact that babies tend to look longer at things that surprise them. In the one experiment, baby Elena would have tended to look longer at the ring separating from the keys if the object she believed was one piece turned out, surprisingly, to be two.

The key-ring experiment is only the latest of a long line of studies that have brought Needham, an assistant professor in the department of psychology: experimental, ever greater insights into babies' cognitive development. In one of her seminal studies, she and her students showed babies a simple display of a curved, corrugated, yellow cylinder sitting smack against a tall, blue box covered with white squares. A mysterious gloved hand would appear and grasp the cylinder, in some cases dragging it away separately from the box, and in others, pulling the two objects as a single assemblage.

"In that experiment, we tried to address whether babies, seeing a stationary display, use the objects' physical features alone to figure out that these should be two separate things," she says. Other scientists' studies had indicated that young infants could only gain clues from the objects' motion and separation, not their features, to figure out that two objects were separate. "When I started my research as a graduate student, people commonly believed that if objects were stationary, and there weren't big spaces between them, young babies had no hope of figuring out whether they were two separate pieces or not. By this theory, for example, they would see the papers, pencils, staplers, and other objects on a desk as one big blob."

Her first experiments confirmed this conventional wisdom. She found that four-and-a-half-month-old babies didn't seem to grasp the concept that objects that looked different are likely separate entities. The infants didn't show particular suprise either when the yellow cylinder dragged the blue box across the stage --indicating they were one object--or when the cylinder moved as a separate object.

Tubular sensation: In this experiment, above, a mysterious, gloved hand appears to grasp the cylinder, in some cases dragging it away separately from the box: "We tried to address whether babies, seeing a stationary display, use the objects' features alone to figure out that these should be separate things," says Needham; left, the baby's mother, a non-participant, keeps her eyes closed during the experiments to keep from evoking a reaction
Photos by Les Todd

Needham didn't give up on the possibility that young babies could use "featural" information to sort out objects. (Featural information refers to the visual appearance, or visual features, of an object--color, size, shape, and so forth. Physical information encompasses how the object behaves according to physical laws--whether it falls, for example, or moves through other objects.)

For one thing, she reasoned that young babies might be overwhelmed by the demands of visually processing a complex scene. So, in follow-up experiments, she simplified the objects and how they were viewed, straightening out the yellow cylinder and orienting the blue box face-on to the babies' view, instead of corner-on.

Sure enough, these experiments revealed for the first time that four-and-a-half-month-old babies did look significantly longer at the struck-together cylinder-box than at the one that moved apart. "We've found that, as long as the objects are simple and the information clear, babies can start using the color and shape of objects as clues."

Photo: Les Todd

The featured players in the trickiest production of her "Theater of the Inferred" were two yellow octagons and a retractable metal blade, which she enlisted to help her discover how infants cope with multiple conflicting cues about objects. In these experiments, the curtain opened to show the baby the two octagons sitting together. Next, the gloved hand would appear holding a metal blade encased in a bright red wooden frame. In some trials, the hand would merely set the blade down beside the octagons; in others the hand would push the blade down between the two octagons. Although the baby perceived the blade to slice between the octagons, it actually retracted, maintaining their connection. Next, the ever-busy hand appeared and grasped one octagon to drag it across the stage, either drawing the other octagon along or not.

Needham discovered that the babies who saw the blade simply sit beside the octagons showed more surprise by looking longer when the shapes moved apart and those who saw the blade apparently slice the octagons apart seemed more surprised when the octagons slid along as one piece. "This experiment showed that babies used both featural and physical cues to decide about the connectedness of objects. And when there's a conflict between the two, they choose the interpretation based on the more reliable physical information."

In a related bit of experimental sleight-of-hand, Needham explored whether eight-month-old babies understood the gravity of situations, so to speak. In this experiment, the curtain rose to reveal the bent yellow cylinder either resting on the stage against the box or up off the stage against the box. In both scenarios, the hand appeared, grasped the cylinder, and either pulled it apart from the box or dragged the two as a unit. Theorized Needham, "If the infants used the features of the objects to segregate the display, they would see the display as composed of two separate units. However, if they used the physical information like gravity to segregate the display, they would see the cylinder-up display as composed of a single unit."

Her baby-data showed that the infants did expect the display with the cylinder resting on the stage to be two separate objects and the display with the cylinder off the stage to be a single object. "This suggests that by eight months, babies can use object features, but they are also sensitive to the support of objects," says Needham. "And they seem to have a ranking system in which they go with the physical information first, because it's more likely to be accurate than the colors and patterns and shapes."

In another gravity-defying experiment, Needham and her students showed four-and-a-half-month-old infants the gloved hand either stacking one box safely on another or performing the seemingly impossible act of setting the second box in mid-air. As expected, the infants in this experiment gazed longer at the "impossible" event of the box hovering with no support. This particular experiment also showed how complicated and subtle baby cognition studies can prove. "The babies might have looked longer at the impossible event, not because they were surprised that the box didn't fall, but simply because they had never seen the deliberate release of an object in mid-air." To test this confounding possibility, Needham conducted yet another experiment in which the box actually fell after being released in mid-air. "Our reasoning was that, if the infants in the first experiment looked longer just because the box was released, they should look longer when it is released and falls than when it is set on the other box."

The results showed that infants were no more interested in the box that fell than the one that simply rested on the other box. Quandary solved: It was the physical "impossibility" of the event that fascinated the babies.

Sense of gravity: For the first stage of this experiment, one box is safely stacked on the other, and then the gloved hand seems to set the second box in mid-air. As expected, the four-month-old looked at the event longer, fascinated by its seeming impossibility
Photos by Les Todd

As clever as young babies might seem, says Needham, her tests reveal that they still have only a na夫e mental model of gravity and support. "While young babies find it surprising when an object just hovers in space, if they see even a finger touching it from the side, they think that's adequate support." Needham is further exploring the limits of babies' models of support by showing them scenes in which one box is perched atop another too precariously to support it. She is testing just how far the top box can be pushed out before infants find support impossible.

Believing in the old adage "Experience is the best teacher," Needham is also testing the limits of babies' ability to learn from experience with objects. In one set of experiments, she found that, when babies were not allowed the five-second exposure that gave them a chance to familiarize themselves with a display, they couldn't seem to decide whether two different-looking objects were separate or not. Conversely, she found that giving babies a preview of the experiment, by showing them either

object separately, enhanced their ability to perceive a separate cylinder and box. Even bringing the objects to the babies' homes as long as twenty-four hours before the experiment was enough to help them in the experiment.

She also discovered limits to babies' ability to learn from experience. For one thing, the box they saw beforehand had to be almost exactly the same one used in the experiment or the babies could not connect it with the experiment box. Says Needham, "If we showed four-month-olds, say, a purple box with yellow dots, then later in the lab used a same-sized blue box with white squares, they couldn't make the connection. The only time they could extrapolate from their experience was when we showed them a blue box with white circles, and in the lab used a blue box with white squares."

Needham extended her study to showing babies bunches of boxes. "We wondered whether, if we showed babies a lot of different boxes together before the test, that they would create a mental category of boxes that would help them later," she says. "So, we did the experiment, and it worked. Those babies apparently formed some kind of category of colored, decorated boxes that exist as separate objects. And, they used this category in the lab tests to help them see the cylinder as separate from the box, even though they had not seen that particular box before." Again, she found limits to baby thought. Her experiments showed babies could not use a prior encounter with three identical boxes--different from the test box--to create a category of boxes. Nor could they make a category when shown only two different boxes before the experiment. "Thus, infants may need experience with different exemplars of a category, and may need more than one set of differences, to generalize across these differences to form a category to which these objects could belong."

Understanding such details of early learning by experience is important, she says, because babies are not likely born with their brains pre-wired with basic ideas of object support or other rules of physics. "Our position is that infants are born, not with substantive beliefs about objects, but rather with highly constrained learning mechanisms that enable them to quickly arrive at important generalizations about objects."

Now, Needham is showing babies such as Elena other objects to explore further their concepts of categories. "We're showing babies key rings and paint brushes to see whether their experience of watching parents or siblings use such things has affected their perceptions," she explains. "Because the parts of a key ring or a paint brush look very different, those babies who analyze only features might normally expect the parts to be separate objects. In contrast, if babies are developing some kind of category knowledge about these objects, as they get older they might override whatever featural analysis they do and assume the ring to be attached to the keys, or the handle to the paint brush." Needham's results so far confirm this developmental learning theory. She has found, for example, that twelve-month-olds are surprised when a paint brush handle comes off, while eight-month-olds don't seem to be surprised.


Needham is also exploring whether temperamental differences among babies affect their ability to perceive objects. By placing objects in young babies' hands and observing their reactions, she can determine whether individual babies are "high explorers"--who typically handle objects and stick them in their mouths to learn about them--or "low explorers" who do less manipulation. "We've found that the high explorers are more likely to see the cylinder and box display as two separate pieces, and the low explorers are less likely," she says. "We're still not sure why, but we think that the high explorers gather more data about objects, such as color and texture and feel, that help them form rules about objects." She emphasizes that the low-explorer babies are not necessarily more passive than high-explorers, but just slower to develop their strategies for exploring objects.

Besides a steady stream of some eighty babies a month coming through Needham's lab, there are many undergraduates; she encourages student involvement in her experiments. "We get a lot of students who love babies, and my work integrates that interest with their academic pursuits. For students who are interested in graduate school, this research gives them a much better sense of whether they're interested in pursuing their studies in psychology." Needham, who teaches undergraduate classes on developmental psychology, sees laboratory experience as giving a very different view of the field from that gained in the classroom. "I tell students that, even if they didn't like my classes, they shouldn't be dissuaded from continuing in psychology, because research can be much different and more interesting than a classroom experience."

Her work has attracted some undergraduates to stay a year after graduation to work as laboratory managers and co-author research papers--giving them a valuable start on their own research careers. Former lab manager Erika Holz '95 is now completing a master's degree in speech pathology; former manager Cynthia Ramirez '97 has entered a doctoral program in clinical psychology; and this year's manager, Avani Modi '98, plans to do graduate work in clinical psychology.

Needham's work has yielded practical insights into baby-raising, as well. "If you know what perceptual skill your child is working on or capable of at a given time, you can provide them with experiences more tailored to what they're ready for," she says. "Years ago, people didn't know much about young babies' visual abilities, so they weren't sure about the best toys to give them. But now we know that their contrast sensitivity is not very good, so if you give them pastel toys, they're not going to distinguish them from the environment very easily. Toys in vivid colors of black, white, or red will likely attract infants' attention better than pastels."

Needham theorizes that babies benefit most from experiences in the real world, and not just from designer toys: "In the normal visual environment, there's a very healthy diet of angles and colors and shadings and movement that can help babies learn about objects. Just letting your baby watch you vacuum or wash the dishes or chop wood provides a lot of interesting sights to help both their visual and cognitive development." But her intricate experiments also reveal that studying the mysteries of infant development is no child's play; that each finding is only a baby step toward understanding a stunningly complex process.




Share your comments

Have an account?

Sign in to comment

No Account?

Email the editor