In her book The Rational Imagination, Byrne (Reference Byrne2005) describes some of the cognitive mechanisms underlying counterfactual thinking. Her guiding theoretical framework is informed primarily by the mental models tradition (Johnson-Laird Reference Johnson-Laird1983) but also by norm theory (Kahneman & Miller Reference Kahneman and Miller1986), which emphasize the lower-order building blocks of cognition. Byrne's main claim is that counterfactual thinking is inherently rational, which is to say disciplined, rooted to accurate inferences about reality rather than mere whimsy. We agree completely with this claim, but suggest that it perhaps does not go far enough. Counterfactual thinking is not only rational, but motivated (and motivating). Connected deeply to goal-oriented cognition, counterfactual thoughts contribute to the effective management of ongoing behavior. Bundled under a functional theory of counterfactual thinking (Epstude & Roese, submitted; Roese Reference Roese1994; Reference Roese1997; Roese & Olson Reference Roese, Olson and Zanna1997), we argue that taking into account the motivational and regulatory basis of the imagination helps to explain key findings about which the mental models tradition remains silent.

Consider the following empirical observations. Counterfactual thoughts are idealistic: they are more likely to focus on how the past might have been better than on how it might have been worse (Nasco & Marsh Reference Nasco and Marsh1999; Summerville & Roese, in press). Counterfactual thoughts are situationally reactive: they are more likely to appear after failure than after success (Roese & Hur Reference Roese and Hur1997). Counterfactual thoughts are problem-focused: they are more likely to focus on fixing a problem than on random life events (Roese et al. Reference Roese, Hur and Pennington1999). Counterfactual thoughts are egocentric: they are more likely to focus on the actions of oneself than on those of other people (White & Roese, submitted). Finally, counterfactual thoughts involve activation of the orbitofrontal region of the brain, a region previously linked to planning and problem solving (Coricelli et al. Reference Coricelli, Critchley, Joffily, O'Doherty, Sirigu and Dolan2005; Ursu & Carter Reference Ursu and Carter2005). What do these observations tell us about counterfactual thinking?

Taking these findings in hand, it becomes clear that the abundant research on goal pursuit (Carver & Scheier Reference Carver and Scheier1998; Fishbach & Ferguson Reference Fishbach, Ferguson, Kruglanski and Higgins2007; Higgins Reference Higgins2006; Lewin Reference Lewin1935) provides us with important insights into the form, function, and effect of counterfactual thinking. Goals may be defined as cognitive representations of desired ends and the means to achieve such ends (Fishbach & Ferguson Reference Fishbach, Ferguson, Kruglanski and Higgins2007). Counterfactual thoughts that occur in everyday life involve, for the most part, alternative means that “might have been” implemented so as to have obtained a desired end. We recently described how principles of motivation and goal cognition might explain counterfactual thinking (Epstude & Roese, submitted). Encountering a problem typically triggers an upward counterfactual (e.g., “If only I had studied harder, I would have passed”). Counterfactual thoughts themselves have as an inherent property such causal implications, and these directly fuel the activation of corresponding behavioral intentions (“I intend to study harder next time”), which in turn unleash corresponding corrective behavior (the student indeed studies harder the next time). To the extent that such behavior alleviates the original problem, this mechanism is effective in terms of regulating behavior in terms of goal pursuit. This regulatory mechanism is content-specific; that is, the information contained in the counterfactual directly translates into a related action.

In addition to a content-specific mechanism by which counterfactual thinking influences behavior, evidence also suggests a content-neutral mechanism. A content-neutral mechanism reflects how rather than what information is handled. For example, independent of their specific meaning, counterfactuals can exert an influence on attention and information processing, as in demonstrations of a counterfactual mind-set, which involves a heightened albeit generic tendency to consider alternatives (e.g., Galinsky et al. Reference Galinsky, Moskowitz and Skurnik2000). As another example, the negative affect that often springs from upward counterfactuals (which make the present look less desirable in contrast to a better alternative) may itself motivate behavior change (Markman et al. Reference Markman, McMullen, Elizaga and Mizoguchi2006). In addition, structural properties of counterfactual thoughts may evoke either approach or avoidance motivation (e.g., Roese et al. Reference Roese, Hur and Pennington1999).

The interplay between emotion and counterfactual thinking is pivotal. Regret is an unpleasant feeling state that depends on an upward counterfactual, an aching despair born of the realization that one might have made a better decision or achieved a better outcome (Roese Reference Roese2005). People are motivated to manage their regret experiences even as they draw insights from their regrets (Zeelenberg & Pieters Reference Zeelenberg and Pieters2007). Moreover, recent studies have linked mental health dysfunction to both an excess and a deficit in counterfactual thinking and regret. The principal consequences of upward counterfactual thinking (i.e., regret) are problem-solving insights and negative emotion; hence, excessive counterfactual thinking has been found to be associated with pathology rooted to excessive problem-focused cognitions (e.g., anxiety; Kocovski et al. Reference Kocovski, Endler, Rector and Flett2005) and excessive negative affect (e.g., depression; Markman & Miller Reference Markman and Miller2006). By contrast, a deficit of counterfactual thinking is associated with a deficit of problem-focused cognition (e.g., underachievement, work difficulty, social dysfunction) along with an absence of negative affect. Along these latter lines, schizophrenia has been shown to be associated with impaired counterfactual thinking (Roese et al., in press) and deficits in goal-related cognition (Brandstätter et al. Reference Brandstätter, Lengfelder and Gollwitzer2001). These studies suggest that there is an optimal level of counterfactual thinking and emotional reactivity to such inferences, and that both too much and too little may spell trouble for mental health.

An earlier generation of research on counterfactual thinking, dating from the 1980s and stimulated by the writings of Kahneman and Tversky (Reference Kahneman, Tversky, Kahneman, Slovic and Tversky1982), treated such thoughts as instances of bias, and hence, impediments to sound judgment and shrewd action. The work of Byrne and others has illuminated counterfactual thinking in a different light, as an instance of principled and rational imagination. Counterfactual thoughts do sometimes bring bias, yet balancing this cost is the larger benefit of the effective management of daily behavior. Counterfactual thinking, we argue, is best understood as an input to course correction, as an instance of goal cognition, and as an essential component of behavior regulation.