Are babies born knowing a little physics?

Perhaps a little, and the argument is over exactly what that means. Piaget's theory was domain general: it predicted that cognitive maturation happens at the same time across domains such as maths, logic, physics and language. Using looking-time methods, Elizabeth Spelke and her colleagues have argued instead that infants have structured expectations about the physical and social world long before they acquire language.

The method is to watch what babies look at. Researchers habituate a baby to an image, show a new one with more or fewer objects, and infer from a longer look that the baby can tell the quantities apart. The expectations Spelke describes include that objects move along continuous paths, cannot pass through solid barriers, and persist when hidden. Renée Baillargeon is best known for research showing that infants have an intuitive awareness of solidity, containment and occlusion at a young age.

Spelke's proposal is called Core Knowledge Theory. Infants, she proposes, rely on a small set of domain-specific systems, for objects, approximate number, space, goal-directed action and social categories, which are seen as products of evolution and building blocks for later thought. It is an alternative to William James's hypothesis that babies are born with no distinctive cognitive abilities.

But the researchers disagree with each other, and others disagree with them. Baillargeon claims infants come with innate learning mechanisms, while Spelke argues that they are born with substantive knowledge about objects. Schoner and Thelen argued that longer looks may only show that infants notice a difference, not surprise at an impossible event, and a dynamic systems approach has strongly challenged some of the core knowledge school's basic presumptions.