Are children able to draw normatively correct probabilistic inferences? The readers may find this question rather odd given the well-documented difficulties of intelligent adults to solve various sorts of probability problems. The aim of this chapter is to show that children do possess an elementary competence to reason about uncertain events, and that the existence of such a competence sheds light on the properties of adults' reasoning and its relation to the normative models of probability.

There is an ongoing controversy in the adult literature about the factors that hinder correct probabilistic reasoning in naive individuals (i.e. those who have not mastered the probability calculus or other formal calculi, like standard logic). According to an influential, evolutionary view, naive individuals fail simply because the human mind is inherently unable to reason about the probability of single events, and only individuals trained in probability calculus can succeed. Besides empirical evidence, a normative argument appears to support such a view: since a norm for measuring chance (i.e. the probability calculus) emerged only in the seventeenth century, it follows that probabilistic reasoning is not a natural, mental capacity and that before the seventeenth century individuals could not even consider the possibility of such a yardstick. An alternative is the extensional competence view. Under this view, untutored individuals are able to draw correct probabilistic inferences, even if they have not learnt the probability calculus.