To learn about the distant past, about remote places, or about hidden causal processes, children must typically rely on others’ testimony. A considerable body of evidence has shown that children are ready to trust what they are told in these various domains. Indeed, they often do so even when what they are told contradicts their intuitions. Children typically cannot gather empirical evidence to confirm or disconfirm what an adult has told them. For example, they cannot easily gather evidence about long departed civilizations or extinct animals. In such cases, it is reasonable for children to accept what they have been told. However, there are situations where children can test an adult’s counter-intuitive claim, especially when they are learning about scientific topics. Presented with a counter-intuitive claim that is easy to test, do young children seize such opportunities or do they simply acquiesce to what they have been told?
We presented preschool and elementary school children with five different-sized Russian dolls and asked them to indicate the heaviest doll. Almost all children selected the biggest doll. Half of the children then heard a false, counter-intuitive claim (i.e., smallest = heaviest). The remaining children heard a claim confirming their initial intuition (i.e., biggest = heaviest). Children typically endorsed the experimenter’s claim even when it was counter-intuitive. During the experimenter’s subsequent absence, older children explored the dolls more if they had received counter-intuitive rather than confirming testimony, whereas younger children rarely explored regardless of the testimony they heard. Children who explored made judgments consistent with the dolls’ actual weight. Children who did not explore continued to endorse the experimenter’s testimony (Ronfard, Chen, Harris, under review, email for more information).
Thus, older children are prone to adopt an empirical stance in relation to counter-intuitive claims. Recognizing that such claims can be tested against the available evidence, they actively seek that evidence and revise their judgment of the unexpected claim accordingly. Younger children may not have explored following our counterintuitive claim because they may have placed greater trust in the testimony they received than older children. This interpretation is consistent with earlier findings showing that preschool children are surprisingly trusting of an informant’s testimony, even when the informant has a history of inaccuracy. One testable prediction, which we are currently investigating, is that younger children will increase their exploration when the informant appears to be less reliable than the informant in our current study.
One noteworthy implication of this finding is that it extends previous work on the impact of instruction on children’s exploration. Previous research has shown that instruction restricts exploration by reducing the number of hypotheses that children consider. Our results suggest, however, that for older children, whether instruction limits exploration depends on the exact nature of what they are told. When instruction does not conflict with children’s intuitions about what they observe, it may lead them to focus their exploration on a subset of the various possibilities that they would have investigated on their own (e.g., Bonawitz, Shafto, Gweon, Goodman, Spelke, & Schulz, 2011). This allows children to restrict their exploration in an efficient fashion. By contrast, when older children are presented with information that conflicts with their intuitions, such information helps them to learn by prompting them to consider possibilities they would not have considered otherwise, thereby increasing their exploration.