Little Intellects - New Thinking on the Birth of Cognition. By Anita Neal Harrison, Photos by Rob Hill, Illustrations by Blake Dinsdale

the first year of life, infants gain new skills at an astonishing pace. Parents watch for the motor milestones of rolling, sitting and first steps and proudly track their babies’ progress. But what about the progress being made in their minds? Cognitive development is not so easily witnessed, but MU researchers are finding ways to get inside babies’ heads, discovering amazing things about how and when intellect develops.

“We often come up with experiment ideas that require very sophisticated thinking on the part of the infants,” says Kristy vanMarle, an assistant professor in MU’s Department of Psychological Sciences and the director of the Development Cognition Lab. “Sometimes we wonder whether or not we’re giving them too much credit. But then you start seeing the data, one baby at a time, and often you see that they can do it. As a parent of three young children, it helps me remember that there’s probably so much more going on in their heads than we think or that may be showing on the surface.”

Cover Creation

Cover Creation  video
A photo illustration is born, step-by-step.

The question of what infants know and how they know it has been asked at least since ancient Greece, when philosophers debated whether humans are born with innate knowledge or whether all understanding starts with the senses. In more modern times, John Locke’s 17th-century assertion that infants have no innate knowledge led to the proliferation of the scientific method and its requirement that new knowledge be based on experiments and measurable evidence.

Infant boy interacting with a puppet.

Intellect Explorers
From left: Yuyan Luo, You-jung Choi and Kristy vanMarle.

Given Locke’s empiricist clout, it’s a little ironic that he did not come to his conclusions through experiments involving infants. He instead relied on pure reasoning. “If we will attentively consider new-born children, we shall have little reason to think that they bring many ideas into the world with them,” he wrote in An Essay Concerning Human Understanding.

To be fair, Locke did not have techniques available for testing infants’ cognitive abilities. Today’s researchers do.

One of the most revolutionary techniques is also incredibly simple. Pioneered by American developmental psychologist Robert Fantz in the late 1950s, “preferential looking” is based on the idea that infants’ gazes provide a window into their thoughts. Fantz showed that infants will look longer at things they find new or unexpected. This insight has helped subsequent researchers design preferential looking experiments that work like this. An infant is shown a stimulus. Researchers then measure the infants' gaze to determine whether they find that stimulus odd, or at least different from what they “know” should be happening.

Preferential looking was used in many of the studies that vanMarle and Sue Hespos, an associate professor in psychology at Northwestern University, included in their review exploring whether infants’ brains come equipped with “intuitive physics,” the ability, vanMarle explains, for infants to form “expectations about the objects around them, even though that knowledge is a skill that’s never been taught.”

Intuitive physics fits in with a “core knowledge” approach to cognitive development, a perspective that would make Locke and other no-innate-knowledge folks frown. “The idea is that over our evolutionary history, there have been certain skills or principles that have been selected for because they are adaptive,” vanMarle says. “Put another way, there is some knowledge about the world, like basic physics, that is so important for our survival that, rather than leaving it for us to learn through experience [and thus risk us not learning it], nature has built it right into the baby.”

VanMarle and Hespos looked at scientific literature spanning the past 30 years. Those studies showed, the researchers say, that even very young infants have detailed knowledge about how objects behave and interact. At two months — the earliest age at which testing can occur — infants show an understanding that a ball should stop when it hits a solid wall, and they show surprise when an object suddenly appears or disappears. By five months, infants have an expectation that a liquid will act differently than a solid. At six months they discriminate between cards showing eight dots and cards showing 16. In a previous publication, vanMarle determined that children as young as 10 months consistently choose larger amounts of food when presented with two different portion sizes.

Hespos and vanMarle organized these and other findings into what they call three knowledge domains — objects, substances and number concepts — and concluded that “certain core principles about these domains are present as early as we can test for them, and the nature of the underlying representation is best described as primitive initial concepts that are elaborated and refined through learning and experience.”

“So,” vanMarle adds, “not only is the knowledge in place very early in life, perhaps at birth, but it continues to guide our expectations and interactions with the world throughout the lifespan, highlighting a kind of continuity in the way the mind works.”

In addition to this confirmation of the core knowledge approach, the review, published in WIREs Cognitive Science, also resulted in new understanding about ontological development.

“An ‘ontological distinction’ can be described as a distinction made at the most fundamental level about the kinds of things there are in the world,” vanMarle explains. “One of the claims we make in our paper is that intuitive physics applies to objects, but that it is not clear whether non-object entities like non-solid substances are part of this domain. That is, although babies use their intuitive physics to predict how objects will behave in the world, they seem to apply different principles to entities that are not objects. In doing this, we were able to link up some relatively new research on infants’ reasoning about substances to make a judgment about the scope of intuitive physics and what kinds of entities it applies to. We conclude that substances may constitute a separate domain that is ruled by different principles.”

The practical benefits of understanding our initial knowledge and how it’s structured range from developing pedagogical techniques to predicting which concepts will be harder for students to learn, vanMarle says.“It might also give insight into the source of developmental disorders. It is hard to know how a system has gone awry unless you know what it looks like when it’s intact and functioning properly. ... There are lots and lots of reasons why doing basic research like this can help us improve education, develop interventions and, more generally, give parents information that will help them understand their children better.”

At the same time that babies are figuring out the physical world, they are also sorting out what to expect in the social world. That’s the area of cognitive development that Yuyan Luo, an associate professor of psychological sciences, and You-jung Choi, a doctoral candidate at MU, explored in a study testing how much infants understand about appropriate social behavior.

“From the moment we wake up to the moment we go to bed, we constantly interact with others,” Choi says, explaining what drew her to this research. “At any given time, we can interact with a stranger on the street, someone we know, or perhaps someone we love. Based on the interactions, we can initiate a new social relationship, maintain the relationship or terminate the relationship. Most adults are comfortable with this routine. I wondered how infants deal with these social situations.”

For the research, Choi and Luo created puppet shows that presented infants with complex social situations. The shows began simply enough, with an introduction to two puppet friends, A and B, who bounced together, clapped together and silently laughed together. Then A exited, and a third puppet, C, entered. And with her entrance, the Sesame Street plot took a Punch and Judy turn. Instead of welcoming C, B deliberately knocked the new puppet down. Meanwhile, A watched the attack through a side window.

How, Choi wondered, would infants watching this scene expect A to behave toward B after witnessing B’s bad behavior?

The question gets at whether babies possess a “theory of mind,” or an understanding of other people’s intentions and beliefs. It also explores whether babies have social evaluation skills, or can assess others’ personality traits through observing their behavior. Previous research had shown that these cognitive capacities emerge in infancy. Choi and Luo designed the study to test whether babies at 13 months of age had the mind and social evaluation skills necessary to make sense of complicated social situations.

The test came in the puppet show’s final scene. After B knocked down C, C left the scene and A returned. At this point, babies watched either: one, a "together" event, in which A and B engaged in a new interaction, swaying and laughing; or two, an "alone event," in which only B swayed and laughed while A stood still.

As with vanMarle’s study, Choi and Luo used preferential looking to determine which scenario best met the babies’ expectations. If babies looked longer at the together event, that would indicate they were socially savvy enough to be surprised that A continued to be friends with a misbehaving meanie. If they looked longer at the alone event, that would indicate the babies didn’t get what was appropriate social behavior.

The result? Babies looked at the together event significantly longer, 40.2 seconds versus 32.2, revealing they did know enough to expect A to snub B.

“This to us indicates infants have strong feelings about how people should deal with a character who hits others. Even his or her acquaintance, or ‘friend,’ should do something about it,” Choi says.

Researchers Yuyan Luo, You-jung Choi and Kristy vanMarle.

Playing Nice
Controlled scenarios with puppets have shown that even very young children have strong feelings about appropriate interactions with others.

But what if A had not seen B’s misbehavior? Would babies still expect A to ignore B because they, the babies, know B is not to be trusted? Or would they have enough social-cognitive prowess to understand that, while they know B is a meanie, A is not privy to this information and, consequently, holds a false belief that B is a worthy friend?

To find out, Choi and Luo ran the same show, except this time A did not witness B’s naughty behavior. Babies watching this “false-belief” scenario looked at the together event for an average of 27.4 seconds and the alone event for 33.7 seconds, suggesting they expected the clueless A to continue to be friends with B.

“The true- and false-belief conditions differed only in whether or not A was present when B harmed C, but yielded opposite results,” Choi and Luo wrote in their study. “This suggests that infants considered A’s perspective to predict A’s behavior.”

Choi say she was “really amazed” by this result. “This shows that infants can take others’ perspectives, which is difficult at times for some of us adults,” she says.

Choi and Luo also wanted to know if babies would evaluate the situation differently if B’s misbehavior was an accident. They ran another show, putting A back at the window during B’s transgression. But this time, B knocked C down while looking in the opposite direction.

Babies watching this accidental wrong were nearly equally likely to expect a snub as an acceptance in the final scene, watching the together event for 32.2 seconds and the alone event for 34.3.

“I was surprised by the accidental scenario, too,” Choi says. “These results indicate that 13-month-old infants are sensitive to others’ intentions.”

These findings, published in the journal Psychological Science, established Choi and Luo’s study as among the first to demonstrate that preverbal infants can use their theory of mind and social-evaluation skills to make sense of complex social situations.

“Based on this research, we are looking at similar situations — how a pro-social behavior, for example, helping, will change the dynamics of social interactions,” Choi says.

The most compelling reason to study how infants learn to make sense of the world, says Choi, involves better supporting their development.

“For example, what babies understand about positive social interactions or even positive relationships and what contributes to the formation of such relationships,” she says. “We are trying to understand what babies learn.”

For her part, vanMarle agrees that supporting infants’ cognitive development is an important reason to do cognitive development research. But it’s also about larger issues.

“Studying the infant mind is fascinating not just because it can shed light on the mental lives of babies, but also because it can give us answers to questions that lie at the heart of philosophical — even metaphysical — inquiry,” vanMarle says.

“Young infants provide us with a look at the mind before it has been much affected by learning or culture, before it has been modified and shaped by language. Questions [about the origins of cognition] cannot be asked by studying adults or children who already have so much cultural ‘baggage.’ Babies, as cute and cuddly as they are, are also the perfect study subject if you’re interested in how minds, in general, might work.”

‘Amazing Ability’

Still more insights into early brain powers are coming from the MU School of Health Professions, where Kathryn Brady, now an assistant professor at Southern Illinois University Edwardsville, completed her dissertation on early language acquisition. Judith Goodman, an associate professor in the MU School of Health Professions and chair of the Department of Communication Science and Disorders, served as Brady’s advisor.

Brady’s study, published in the American Journal of Speech-Language Pathology, focused on how toddlers from 18 months to 36 months use various cues to infer and retain the meanings of words.

“[American philosopher Willard Van Orman] Quine raised a now classic philosophical question of how, when a child hears a word, is he able to figure out what it refers to?” Goodman says. “If you were in a remote country that speaks a different language and someone shouted, ‘Gavagai!’ just as a rabbit hopped by, it seems logical that you might think that Gavagai refers to the rabbit. But why? It could refer to hopping or fur or long ears or the powder-puff tail or even a concept like ‘the ears and head’ of the rabbit."

“This is the situation young children will find themselves in hundreds or thousands of times a day,” Goodman continues. “How can they solve this problem of lots of possible hypotheses for all the new words they are hearing? And yet, between 12 months of age and when they start school, they will learn between 9,000 and 14,000 words. However they are solving this puzzle, they are very, very good at it.”

Brady and Goodman set up a study to explore which kinds of cues toddlers find most helpful when “fast mapping,” that is, making their first stab at inferring new-word referents. Toddlers were each taught six words, using three types of cues that were presented alone or in pairs. The researchers then recorded the children’s ability to accurately guess what the words meant.

“When children were presented with a new word and asked to choose between an item for which they already had a name and an unfamiliar object, they appropriately assigned the new word to the unfamiliar object. This ability improved as children aged,” Goodman says. “The toddlers’ ability to infer a word’s meaning from linguistic context, such as figuring out that a ‘kiwi’ must be a food item when they hear, ‘Sammy eats the kiwi,’ also improved as the children aged. However, using social cues, such as eye gaze, became less effective as the children matured. By 36 months of age, children were less likely to assume a word referred to the particular object a speaker was looking at — looking at a kiwi when teaching the child the word ‘kiwi’ — than younger children were.”

In another unexpected twist, the study also revealed that the toddlers’ performances were, on average, no better with two cues than with one.

Along with exploring the roles of cues in fast mapping, Brady and Goodman also examined word retention. Here their tests showed that one day after the children learned the six new words, they were most likely to remember the first three. This finding of a "retention limit" corroborated earlier research. But Brady and Goodman did these previous studies one better by discovering that different kinds of cues result in different levels of retention success — an important point for clinicians, teachers and parents wanting to help children build vocabularies.

“The study strengthens the idea that children are open to — and able to benefit from — a variety of information in their environment and that they potentially ‘learn’ which types of information are most useful to them in different situations,” Brady says. “This is a pretty amazing, especially when we think about how young these children are. The finding that the older children retain the meanings of words they have inferred — even when no one confirms whether those meanings were correct or not — also suggests an amazing cognitive ability.”

Post a Comment

Reader comments are reviewed by Illumination staff before they are posted, so please keep your message civil and appropriate. All fields are required.

– Will not be published

Back to Top

University of Missouri

Published by the Office of Research

© 2021 The Curators of the University of Missouri