The goal of historians consists in constructing counterfactuals that deviate as little as possible from what really happened. Conjectures of the type “if only …” must make the least amount of changes possible to stories, in order to isolate the essential causes of what happened. From the kind of “if only” conjecture a historian uses, we understand the structure of his or her explanations and the nature of the causes that he or she identifies for a given episode in the past (Tetlock et al. Reference Tetlock, Lebow and Parker2005). When, on the contrary, we construct alternatives to reality in order to invent new technologies or scientific theories, we are constrained by the criteria which the community we belong to accepts. As is the case with historians, we cannot work from fantasy.

Fantasy is nothing other than the imagination restrained by fewer constraints. When we create a story of fantasy, science fiction, or magic, we can violate certain principles of reality, but do not arrive at the absurd because it would not serve our purposes. For example, according to the principles of naïve physics, we are able to move at very high speeds, fly, travel backwards and forwards in time, disappear, make difficult calculations, preserve excellent memory, and transform objects. Nevertheless, we always begin from characteristics of the world with which we are familiar. What we do is imagine powers that we do not have, but that are conceivable. So much so that these powers usually belong to other nonhuman entities: speed (missiles), going backwards in time (history books), transforming objects (technologies), calculations and memory (computers), disappearing (complete mimicry of certain animals), and so forth.

When the purpose is to console ourselves or others, in order to encourage or deceive ourselves, or to make ourselves feel guilty, we resort to the notion of control. We construct counterfactuals to demonstrate that avoiding a certain situation was beyond our power, or that it was in our power, but we are to blame for not having avoided it. To exalt in ourselves, instill hope, or deceive ourselves, we imagine alternative worlds that are “worse off” without our intervention. To make ourselves feel guilty, we imagine that if we had not existed, “better” alternative worlds would have.

In childhood, we are unable to create counterfactuals with respect to the contents of our mind. Children believe that what is inside their minds and what is inside the minds of others is the same. For this reason a child cannot believe that others have false beliefs (they have the same beliefs as the child and these beliefs are true), until he or she reaches full development between the ages of one and five (Surian et al. Reference Surian, Caldi and Sperber2007). Adults become so sophisticated that they are able to comprehend statements such as the one made by the former Federal Reserve chairman, Alan Greenspan: “I know you believe you understand what you think I said, but I am not sure you realize that what you heard is not what I meant” (Resche Reference Resche2004, p. 731).

Ruth Byrne says that “people do not tend to imagine ‘miracle-world’ alternatives” (Byrne Reference Byrne2005, p. 191). However, if the goal is to construct a religion, it is better to do just that. “Religions are costly, hard-to-fake commitments to a counterintuitive world of super-natural causes and being” (Atran Reference Atran2002, p. 264). The experiments of Scott Atran on the impact and memorizability of intuitive and minimally counterintuitive beliefs show that the delayed one-week recall presents the following sequence of remembering: intuitive and ordinary, intuitive but bizarre, minimally counterintuitive, and maximally counterintuitive. The best type of counterfactual to use for religious believers is the intuitive but bizarre domain of beliefs: for example, floating pencil, dangling cat, blinking newspaper. These are the couplings typical of miracles: walking on water, restoring sight to the blind, multiplying food, curing the ill. From the viewpoint of naïve physics, such things are impossible, but believable. In earthly matters, which are matters of explanations and calculation, one prefers a probable counterfactual to an improbable one for explaining things. In unearthly matters, an improbable belief is worse than an impossible belief for building faith in a religion. It is on such grounds that Oscar Wilde (Reference Wilde1889/1989, p. 990), another great Dubliner like Ruth Byrne, criticized the Church of England and anticipated the experiments of Atran: “The growth of common sense in the English Church is a thing very much to be regretted. It is really a degrading concession to a low form of realism. It is silly, too. It springs from an entire ignorance of psychology. Man can believe the impossible, but man can never believe the improbable.”