As Cushman notes, cognitive dissonance theory is “the best-known psychological theory of rationalization” (target article, sect. 2.1, para. 3). Cushman follows up this bold statement by claiming that cognitive dissonance theory “does not offer an ultimate explanation – an answer to the question, “Why did it evolve?’” (sect. 2.1, para. 4) However, cognitive theory is not a unitary construct upon which everyone agrees. Since Festinger's (Reference Festinger1957) original theory, a number of revisions and expansions have been proposed (for review, see Harmon-Jones Reference Harmon-Jones2019). One of these, the action-based model of cognitive dissonance, does offer an adaptive explanation for why dissonance evolved (Harmon-Jones & Harmon-Jones Reference Harmon-Jones and Harmon-Jones2002; Reference Harmon-Jones, Harmon-Jones and Harmon-Jones2019; Harmon-Jones et al. Reference Harmon-Jones, Amodio, Harmon-Jones and Zanna2009; Reference Harmon-Jones, Harmon-Jones and Levy2015a).

The adaptive function proposed by the action-based model differs from the one offered in the target article, and this difference is likely due to the target article's incomplete understanding of cognitive dissonance processes and how they differ from self-perception-like processes. That is, most scientists working on these theories have agreed that these two theories apply to different psychological situations (e.g., Fazio et al. Reference Fazio, Zanna and Cooper1977; Harmon-Jones et al. Reference Harmon-Jones, Armstrong, Olson, Albarracin and Johnson2019). Self-perception theory applies to situations in which organisms do not have strongly held perceptions, attitudes, or beliefs, and the self-perception process is concerned more with the formation or construction of perceptions, attitudes, or beliefs. In these situations, actions coldly cause rationalizations; the actions do not arouse negative affect, because the actions do not conflict with strongly held cognitions. In contrast, cognitive dissonance theory applies to situations in which organisms have strongly held perceptions, attitudes, or beliefs, and the dissonance process is one concerned with protecting one of those strongly held perceptions, attitudes, or beliefs. In these situations, actions motivate rationalizations because the actions aroused negative affect due to cognitive conflict (Fazio et al. Reference Fazio, Zanna and Cooper1977; Harmon-Jones Reference Harmon-Jones2000; Harmon-Jones et al. 1996).

The target article posits that the “function of rationalization, then, is to construct beliefs and desires that are consistent with the adaptive behaviors generated by non-rational processes, and then to adopt them” (sect. 3, para. 2). This psychological process is akin to a self-perception process, and we agree that this adaptive function of rationalization operates and applies to such self-perception-like situations. On the other hand, in dissonance-arousing situations, organisms are motivated to protect a strongly held “cognition” (e.g., perception, belief, emotion, behavior), and this cognition is the one to which the organism is most committed or regards as most true at the moment. In dissonance terms, this most strongly held, resistant-to-change cognition is known as the generative cognition, and cognitive changes will serve to protect or bolster this cognition. According to the action-based model, rationalization or protection of the generative cognition is posited to serve the adaptive function of enhancing effective action.

Why would rationalization be expected to enhance effective action? The action-based model posits that cognitions serve to guide action, that is, cognitions carry with them action tendencies. When cognitions conflict with one another, such that one cognition implies that another ought not to be true, their action tendencies are also likely to conflict. The organism will have a difficult time enacting a course of behavior when trying to follow the action tendencies evoked by conflicting cognitions. Thus, bringing cognitions more closely into alignment with the generative cognition makes the course of action clear, as the organism is not forever vacillating between alternative courses of action.

In line with the above, the action-based model posits that dissonance reduction is an approach-motivated process, and evidence supports this. For example, individuals who are higher in trait approach motivation engage in greater attitude change in the direction of a recent behavioral commitment (Harmon-Jones et al. Reference Harmon-Jones, Schmeichel, Inzlicht and Harmon-Jones2011). Similarly, manipulating approach motivation, via cognitive or embodied means, influences the amount of cognitive dissonance reduction, measured as attitude change (Harmon-Jones & Harmon-Jones Reference Harmon-Jones and Harmon-Jones2002; Harmon-Jones et al. Reference Harmon-Jones, Price and Harmon-Jones2015b). Furthermore, the degree of attitude change individuals engage in is positively related to their relative left frontal cortical activation, a pattern of neural responses that is associated with greater approach motivation (Harmon-Jones et al. Reference Harmon-Jones, Harmon-Jones, Fearn, Sigelman and Johnson2008). These and similar results support the idea that dissonance reduction is a motivated process, not merely a matter of cold information-processing as in self-perception.

Additional support for the idea that dissonance was selected for by evolution is provided by evidence that non-human animals engage in “attitude change” in the direction of behavioral commitments. For example, capuchin monkeys show reduced liking for treats they had previously rejected (Egan et al. Reference Egan, Bloom and Santos2010). Similarly, pigeons show a preference for conditioned stimuli for which they had to exert more effort over those for which they had to exert little effort (Zentall & Singer Reference Zentall and Singer2007). As with humans, these changes in preference are biased in the direction of protecting an important cognition (a previous behavior or behavioral choice). The results in non-human animals suggest that dissonance processes are evolutionarily old and may be adaptive in a variety of species.

The target article emphasizes that rationalization may lead the organism to adopt (through non-rational processes) beliefs that are true. In contrast, the action-based model proposes that rationalization may be adaptive whether the beliefs adopted are objectively true or not. Simply reducing motivational conflict may allow individuals to behave more effectively, regardless of the objective truth or falsity of their beliefs. Truth can be a slippery construct, and it is often difficult to say with certainty whether a particular course of action is objectively better than another. We propose that it is often adaptive to fully commit to one's decisions and pursue them without hesitation, even if the alternate course of action may have been objectively slightly “better.” Indeed, in cognitive dissonance research, rationalization often involves conflict between attitudes and motivations, rather than facts. Thus, we view the action-based model as a broader, more inclusive explanation for why dissonance processes evolved than that presented in the target article.