Byrne (Reference Byrne2005) argues that imagination is crucial to adult cognition. To support this claim, she demonstrates that the ability to envision alternatives to reality is regularly employed by adults when they reason about how the world may be or may have been. Because imagination is especially salient in childhood, one might also ask whether the imagination plays a similar role in children's reasoning. Does imagination allow children to reason counterfactually? The standard view is that, in infancy, we are tied to reality, unable to consider alternatives (Perner Reference Perner1991). Considering this, how do we develop into rational imaginers? Although Byrne argues that rational principles underlie imaginative thought, we suggest that the opposite is true: that imagination underlies rationality. Here, we consider how developmental research on pretense, imagination, and inhibitory control demonstrates that early imaginative thought may provide the foundation for counterfactual reasoning.
By regarding pretense as a type of counterfactual thought (p. 2), Byrne seems to suggest that imaginative and rational thought are governed by the same processes. It is certainly true that pretense and counterfactual reasoning share a common skill: namely, the ability to mentally create an alternative to reality. In pretense, when a child pretends that a block is a cookie, he or she must mentally represent an alternative world in which he or she is acting with a cookie, not a block. However, reality cannot be completely ignored. A real representation of the world must be maintained simultaneously with the pretend representation because, if reality did not limit pretense behavior, the child may act irrationally, such as attempting to actually eat the cookie (Leslie Reference Leslie1987). Thus, to pretend successfully, an individual must concurrently create an alternative reality while maintaining an accurate representation of reality.
Counterfactual reasoning also requires the ability to mentally represent alternative realities, because the initial step in addressing a counterfactual situation is mentally changing an aspect of reality. To reason about the question “Would the paper have blown away if I had closed the window?” one's mental representation of reality (i.e., an open window) must first be changed to the alternative state (i.e., a closed window). As Byrne discusses, if one were unable to complete this first step and imagine the alternative reality, she or he would also be unable to rationally answer the question about how the present might have been different.
Although the initial step of creating an alternative representation is the same in both pretense and counterfactual reasoning, these two processes are actually quite different. Consider the steps following the creation of the alternative in each case. In pretense, an individual enters the alternative world and acts online from within the pretend framework. Little thought about the real world is required because the pretend world is not entirely constrained by reality. Although a representation of reality must be maintained and kept separate from the alternative to avoid confusion, anything can happen in the pretend world. Pretending a block is a cookie is just as permissible as pretending to have tea with the Queen of England. Furthermore, there is no overt goal to pretend play. It exists solely in the present and does not require any explicit comparison between the real and alternative state. The two states coexist without conflict; it is not problematic to mentally represent the block as both a block and a cookie.
However, counterfactual reasoning does have an explicit goal, namely, to determine how the present reality might be different given an alternative premise. This situation necessitates a much closer relationship between real and alternative representations than in pretense. According to Byrne, following the creation of an alternative state in counterfactual reasoning, the individual manipulates that alternative based upon what actually exists or actually has happened in the real world. When the alternative temporally catches up to the present reality, one must explicitly compare the two to determine whether the change made in the alternative would have affected the current reality. This clearly differs from pretense because counterfactual reasoning requires not only representing the real and the alternative, but also comparing the two.
How do children develop this latter, more complex ability? We hypothesize that the development of inhibitory control facilitates counterfactual reasoning because inhibitory control is necessary to manage two conflicting representations successfully. Consider a simple counterfactual problem, such as reasoning from a false premise. In the example, “All fish live in trees. Where does Bob the fish live?” the alternative (i.e., all fish live in trees) directly conflicts with reality (i.e., all fish live in water). To answer this question correctly, one would have to inhibit the real world in order to reason based on the alternative. Preschool-age children reliably make “realist” errors when asked this question, answering in a manner that conforms to what they know about the real world. Given that children of this age have no trouble creating alternative states in pretense or making deductions that do not conflict with reality (Hawkins et al. Reference Hawkins, Pea, Glick and Scribner1984), we argue that children make these realist errors because they do not yet have the inhibitory control required to disregard the real state of the world. In other words, they cannot inhibit their knowledge that fish live in water in order to answer based upon the alternative premise (Lillard Reference Lillard, Brownell and Kopp2007; Ma Reference Ma2007).
If the inhibitory demands are removed from counterfactual reasoning tasks, children's performance should improve. This is exactly what happens. When children are explicitly instructed to use their imaginations in counterfactual deductive reasoning tasks such as the aforementioned Bob-the-fish example, they are more successful than children who are not given an imagination warm-up (Dias & Harris Reference Dias and Harris1988; Reference Dias and Harris1990; Richards & Sanderson Reference Richards and Sanderson1999). By instructing children to imagine, the inhibitory load is reduced by turning a counterfactual problem into a pretense situation. When children use their imaginations, they set up an alternative that does not require comparison to reality. Thus, the inhibitory demand is markedly reduced because children do not have to shift repeatedly between the alternative and real representations.
Overall, developmental research concurs with Byrne's thesis that imaginative thought is required for counterfactual thought. However, we argue that such research also clarifies the relationship between the two by showing how children's early imaginative thought, in conjunction with inhibitory control, may facilitate the ability to reason counterfactually.
