The developmentalist Jean Piaget described young children as "cognitive aliens". He considered their modes of thinking to be so different from those of adults that they were like members of a different species-- or even visitors from another planet! While many of us are familiar with children's cute misunderstandings or garbled information, we usually expect their basic thinking skills to be a lot like ours. This gives rise to common mistaken beliefs that children can (but refuse to) reason in the same ways adults do, and come to the same conclusions.
It's not too hard to figure out what someone else knows, but it can be hard to understand the logic of that person's thoughts. Most adults don't announce their reasoning, or walk us through it; they just tell us what their conclusions are. Children are the same, but unlike adults, even if we ask them why they think something, they usually can't explain it. That's why it takes clever experimental methods to figure out what kinds of reasoning children can or can't do, and at what ages they develop new abilities.
A study of counterfactual and hypothetical thinking revealed some of children's difficulties, and the times at which developmental change in these abilities occurs (Beck, R.S., Robinson, E.J., Carroll, D.J., & Apperly, I.A. (2006). Children's thinking about counterfactuals and future hypotheticals as possibilities. Child Development, Vol. 77(2), pp. 413-426). These two terms apply to thoughts about events that are possible but did not happen, or that could happen in the future. In counterfactual thinking, we try to deal both with things that actually did happen, and with alternatives that could have taken the place of the actual event. In hypothetical thinking, we consider the possibilities for events that have not yet occurred.
This study had young children play a game that involved a toy mouse sliding down a red slide. (This is a simplified description of the actual game.) The mouse could go through one of two gates and land either on a spotted card or on a striped card. The children could show which one they thought was going to happen by putting a ball of cotton on the right card, so the mouse wouldn't hurt himself. After repeating this game a few times, the children were asked future hypothetical questions ("If next time he goes the other way, where will he be?"), or counterfactual questions ("What if it had gone the other way, where would it be?"), or open counterfactual questions (" Could it have gone anywhere else?").
The children in this study were better at dealing with a single hypothetical event ("next time, the other way") than at dealing with counterfactuals, which demand that they think about more than one possibility and choose the alternatives. Three- and four-year-olds were quite limited in their abilities to do either of these thinking tasks. Five- and six-year-olds showed much better ability to think about several possible outcomes.
What practical implications come out of this and similar studies? One is that two- and three-year-olds do not have a very good understanding of things that MIGHT happen to them. If they didn't cut themselves on the sharp knife when they grabbed it this time, they don't grasp the possibility that they could be cut the next time they do this. If they ran into the street today and escaped being hit by a car, they don't see that things could be disastrously different tomorrow. Only at around age 5 do they begin to see that the events that happen are only a few among those that could happen. What happened yesterday gives us some guidance-- if cars didn't drive up on the sidewalk today, they probably won't tomorrow either--- but it doesn't provide us with information about all possible events. Children under 5 who seem to disregard adult warnings are cognitively immature, not deliberately disobedient and asking for trouble.
Can we do anything to speed up children's development of counterfactual and hypothetical thinking? This is the kind of question Piaget called "the American question": how to get them to do things faster, earlier, more easily-- and it does not have any very good answer. Our modeling and discussion of other cognitive abilities can be helpful to cognitive development, so that may be true for these abilities as well. A more important question may be, "how can adults use this information?". And the answer to that is that when we understand how young children think, we can protect them better from physical dangers, and save ourselves and them the annoyance of mutual cognitive alienness.