Long before they reach school age, young children understand that people possess beliefs, wishes and intentions that operate, unseen, inside their heads. This "theory of mind" understanding plays a critical role in children's social interactions, enabling them to understand, predict and manipulate the actions of other people.

A recent series of studies led by Yale University developmental psychologist Paul Bloom, PhD, indicates that children also use their theory of mind understanding to label objects that other people create, as well as their own creations.

How children organize and categorize the world around them has been of long-standing interest to developmental psychologists. The new findings contradict one commonly held notion--that children depend most heavily on perceptual cues such as shape when categorizing and naming objects.

When it comes to labeling objects in the world around them, "Children are not perception-bound," concludes Bloom. "This work shows that theory of mind understanding plays out in terms of how even very young children think about the physical world. Beginning in early childhood, people intuitively recognize that names are symbols for things that share deep, essential properties. And in the case of artifacts--things that people create--these essential properties seem to correspond to the intention that underlies the artifact's creation."

Seeking deeper meaning

Bloom first became interested in how children's understanding of intentions affects how they think about the objects they encounter after observing his young son, Max, drawing. Max insisted that his drawings represented familiar things--Mommy, for example--even though they weren't very accurate renditions.

In an initial effort to probe how children come to understand the naming of objects, he and colleague Lori Markson, PhD, a postdoctoral fellow at the Massachusetts Institute of Technology, tested whether young children named drawings of objects based on what the objects looked like or on what they thought the creator was trying to draw. Bloom and Markson asked 3- and 4-year-olds to draw pictures of four different objects: a balloon, a lollipop, themselves and the experimenter. After a short break, the experimenter asked the children to describe what each picture represented.

Assuming that preschoolers' drawings of a balloon and a lollipop, or of one person and another, are unlikely to systematically differ in appearance, the researchers argued that if children correctly label their drawings, they must be inferring what the picture is based on what they had intended it to be.

Bloom and Markson's results, published in the journal Psychological Science (Vol. 9) in 1998, showed that 3-year-olds labeled their pictures correctly 73 percent of the time, and 4-year-olds did so 87 percent of the time.

A second phase of the study corroborated that finding. Bloom and Markson showed the children several pictures--each depicting one large and one small oval--that had ostensibly been drawn by another child. The experimenter explained that the drawings were of objects such as an elephant and a mouse or a tree and a spider, but that because the other child had a broken arm, the drawings didn't always look much like what he or she had intended.

Consistent with the earlier results, the researchers found that when the children were asked to describe the objects in the drawings, they used the size of the ovals and what they knew about the objects to decipher what the artist had intended to draw. For example, the children tended to describe a large oval as an elephant and a small oval as a mouse, rather than to do the opposite.

In a third part of Bloom and Markson's study, children were shown pictures of three vertical ovals and one horizontal oval, representing items such as three pigs and one chicken. Children were asked to point out which was the odd object--for example, the chicken. Although 3-year-olds did not perform any better than chance, the results again indicated that 4-year-olds relied on their understanding of the artist's intentions, more often than not calling the horizontal oval the chicken.

Bloom believes the findings indicate that children are "naïve essentialists"--that far from perceiving the world only superficially, they attend to objects' deeper, more essential properties, including the intentions of their creators.

"Common sense tells you that kids would think that if it's a picture that looks like a rabbit, then it's a rabbit," he observes. "It's really striking that that's not the case. Kids are really sensitive to what the person intended to do."

A matter of artifacts

Working with University of Michigan colleague Susan Gelman, PhD, who has found results similar to Bloom and Markson's, Bloom has recently found that children's naïve essentialism extends beyond the world of representational art to physical objects, or artifacts, created by people.

Bloom and Gelman showed children objects that resembled items such as a paper hat or a Plexiglas knife. They told the children either that an object had been created intentionally by someone or that it had been made by accident.

The results, published in the journal Cognition (Vol. 76) in 2000, revealed that children who believed the objects had been created intentionally tended to name them according to what they resembled--a hat or a knife, for example. In contrast, children who were told the items were made by accident tended not to use the familiar names when describing the objects. Instead, they often referred to what the item was made of--for example, describing the paper hat as "paper" or the Plexiglas knife as "plastic."

The extension of the earlier findings with artwork to familiar, everyday objects "suggests that there's a very basic way in which reasoning about mental states is implicated in word use," says Gelman. "The fact that we find that understanding of intentions is guiding the way children name all kinds of objects suggests that it's very deeply ingrained in the language system."

University of Oregon psychologist Dare Baldwin, PhD, who studies the emergence of theory of mind understanding in babies, comments that the research demonstrates that "the ability to appreciate people's intentions plays an amazingly central role in children's understanding of a variety of things they encounter in the world."

Autism: a special case

To further test whether children name objects based on their understanding of the creator's intentions, Bloom teamed with psychologists Frances Abell, PhD, and Uta Frith, PhD, of University College London, and Francesca Happé, PhD, of London University.

The group reasoned that if children's theory of mind guides how they categorize and name objects, then children with autism--who some psychologists believe lack theory of mind understanding--should show impaired ability to name objects when only the creator's intention is clear.

In their study, which compared children with and without autism, the researchers placed four toy cars or airplanes on a table between a child and an experimenter. In one task, the experimenter turned and looked intently at one of the four cars or airplanes and drew a picture of it with a pencil. In another task, the experimenter asked the child to draw one of the four objects. In each task, after the drawing was complete, the experimenter asked the child what it was a picture of.

As they had expected, Bloom and his colleagues found that the children with autism did more poorly than did the children without autism--even though on average, the quality of autistic children's drawings was much better than those of children who did not have autism. What's more, regardless of whether or not they had autism, children's performance on a standard test of theory of mind understanding was related to how well they did at naming the drawn cars or airplanes.

This study corroborates the earlier findings and suggests an intriguing possibility: that the inability to appreciate one's own or others' mental states might be partly responsible for the emergence of "savant" artistic skills in autistic children.

"Maybe autistic children don't understand that a picture can represent something just because of the artist's intention, so the only way they can draw a picture that satisfies them is to draw one that strongly resembles reality," suggests Happé. "Perhaps this is an interesting example where extreme talent might come about because of a difficulty in an area of social understanding."