Mature counterfactual reasoning in 4- and 5-year-olds.

Counterfactual reasoning is a hallmark of the human imagination. Recently, researchers have argued that children do not display genuine counterfactual reasoning until they can reason about events that are overdetermined and consider the removal of one of multiple causes that lead to the same outcome. This ability has been shown to emerge between 6 and 12 years of age. In 3 experiments, we used an overdetermined physical causation task to investigate preschoolers' ability to reason counterfactually. In Experiment 1a, preschoolers (N = 96) were presented with a "blicket-detector" machine. Children saw both overdetermined (2 causal blocks on a box) and single-cause trials (1 causal and 1 non-causal block) and were asked what would have happened if one of the two blocks had not been placed on the box. Four-year-olds' performance was above chance on both trial types, and 5-year-olds' performance was at ceiling, whereas 3-year-olds did not perform above chance on any trial types. These findings were replicated in Experiment 1b with 4- and 5-year-olds (N = 40) using more complex question wording. In Experiment 2 (N = 40, 4- and 5-year-olds), we introduced a temporal delay between the placement of the first and second block to test the robustness of children's counterfactual reasoning. Even on this more difficult version of the task, performance was significantly above chance. Given a clear and novel causal structure, preschoolers display adult-like counterfactual reasoning.