Three-year-olds show greater suspicion of circular arguments than adults

Children aren't as gullible as you might think. Early in life they display a discernment that psychologists call "epistemic vigilance". They are more likely to trust information from experts compared with novices, from kind people rather than meanies, and from those they are familiar with, as opposed to strangers. Now a study shows that even by age three, children are sceptical about circular arguments; in some cases even more than adults.

Hugo Mercier and his team presented 84 children aged 3 to 5 (and a control group of adults) with three illustrated vignettes in which a girl was looking for her dog. For each story, one character advised the girl of the dog's whereabouts with an argument based on what they'd seen: "The dog went this way because I've seen him go in this direction," (this is known as an "argument from perception" and it was spoken in a neutral voice played through speakers). A second character said the dog had gone in the other direction and gave a circular argument, "The dog went this way because he went in this direction" (also heard through speakers).

Children from age three and up, and the adults, more often chose to believe the character who based their testimony on what they'd seen rather than on a circular argument. This supports the idea that children from three and upwards have epistemic vigilance. "These results point to the existence of basic skills of argument evaluation that children would possess from at least three years of age onwards," the researchers said.

A developmental trend was for the older children to grow more consistent in their preferences. That is, as the children got older, they more often favoured either the argument from perception on every occasion, or (in a minority of cases) they favoured the circular argument on every occasion. Focusing on just those participants who always made the same choice, an intriguing pattern emerged. A minority of the four- and five-year-olds, and adults, always favoured the circular arguments, but none of the three-year-olds showed this pattern. In a sense then, some older children, and adults, were less sophisticated in their judgment of arguments than the three-year-olds.

How could this be? Mercier and his team think that as they get older, some children and adults become dependent on a rule of thumb that mistakes circular arguments for a sign of dominance or authority. When a person says that "the dog went this way because he went in this direction" this is interpreted as equivalent to an authoritative person saying, "this is the case because I say so."

To test this idea, the same children were tested on a similar task to before, but this time one character used a circular argument for a cat's location, while the other character provided no argument (i.e. they just said "The cat went this way"). Preference for circular arguments would be evidence that they are interpreted as having value beyond no argument at all. In this case, three-year-olds were equally likely to trust either form of advice, while a large number of four- and five-year-olds consistently chose to trust the circular arguments. That is, older children, but not the three-year-olds, saw more value in a circular argument than in no argument at all.

Many children display a distrust of circular arguments from a very early age. However, the findings also reveal an intriguing developmental trend, in which a minority of slightly older children begin to be seduced by circular arguments (a weakness that also persists in a minority of adults). This is likely due to them interpreting such arguments as a sign of authority. Such an inference requires a complexity of social thinking that is beyond three-year-olds. Ironically, this means that three-year-olds end up being more canny in their distrust of circular arguments than even some adults.

_________________________________

Mercier H, Bernard S, and Clément F (2014). Early sensitivity to arguments: How preschoolers weight circular arguments. Journal of experimental child psychology PMID: 24485755