The false-belief task—where children determine what another person will do when that person holds a false belief about a situation—taps representational theory of mind. Success around age 4-5 indicates children understand that others have beliefs independent of reality and that actions follow from others' beliefs rather than objective facts, a cornerstone of social cognition.
To understand what makes the false-belief task significant, start with what you know about theory of mind development: children gradually develop the ability to attribute mental states — beliefs, desires, intentions — to others as distinct from their own. The false-belief task is the benchmark test for a specific, pivotal milestone in this development: the ability to represent *another person's belief* as separate from how the world actually is.
The classic Sally-Anne task goes like this: Sally puts her marble in a basket and leaves the room. Anne moves the marble to a box. Sally returns. Where will Sally look for her marble? Three-year-olds typically answer "the box" — where the marble actually is. Four- to five-year-olds answer "the basket" — where Sally *believes* it to be. The younger child fails because they cannot represent Sally's belief as distinct from reality. They know the truth and attribute that knowledge to Sally, a phenomenon that reflects the absence of representational understanding of mind. The child treats minds as transparent windows onto the world rather than as systems that construct potentially incorrect models of it.
Passing the false-belief task requires a specific cognitive achievement. The child must hold two simultaneous representations: what the world is like (marble in box) and what Sally *thinks* the world is like (marble in basket). This is a second-order representation — a representation of a representation. Before age 4, this recursive embedding is unavailable; children's social predictions are anchored to reality, not to others' mental states. The transition around age 4-5 is one of the most studied and replicable findings in developmental psychology, having been replicated across dozens of cultures and dozens of methodological variations.
Understanding this task matters because it identifies the cognitive threshold that unlocks full-complexity social cognition. You can only deceive someone if you understand they have beliefs you can manipulate. You can only make a promise if you understand the other person will hold a mental representation of your commitment. You can only understand that two witnesses to the same event can draw different conclusions if you grasp that each is working from their own belief, not from reality directly. Children who pass the false-belief task reliably begin engaging in all these activities — and their social world becomes populated not just by agents who *act*, but by agents who *believe, intend, and know things you may not*.
One important nuance: later research using violation-of-expectation paradigms (measuring looking time rather than verbal responses) found that infants as young as 15 months behave as if they track others' false beliefs implicitly. Whether this reflects the same underlying competence or a distinct precursor system is actively debated. What the false-belief task specifically measures is the *explicit, deliberate, behavioral* deployment of ToM — and this emerges, reliably and robustly, at around age 4 to 5.