Abstract
We praise Heyes for fleshing out a long needed integrative theoretical framework of norm psychology that offers a convincing and comprehensive alternative to the current nativist cognitive-evolutionary accounts. In our view, the particular strength of Heyes’s framework compared with the nativist account is that it avoids falling into the pitfalls of narrow modularity accounts (Pietraszewski & Wertz, 2022) by drawing on a domain-general mechanism as an important basis of (human) norm psychology. In this way, and as illustrated by recent developments in evolutionary psychology (Narvaez et al., 2022; Pietraszewski & Wertz, 2022), social neuroscience (Ramsey & Ward, 2020), and our own work (Duderstadt et al., 2022; Germar et al., 2016; Germar & Mojzisch, 2019), norm psychology can capitalize on well-established psychological laws and general principles (i.e., use strong theorizing; Fiedler, 2017), stay connected to other fields of psychology (i.e., be integrative rather than isolated), and generate and test new hypotheses.
At the same time, however, we think that some important conceptual groundwork is still necessary to refine Heyes’s conceptualization of norm psychology and Heyes’s cultural-evolutionary account to properly set the stage for diagnostic tests of competing hypotheses and, more specifically, to test Heyes’s cultural-evolutionary account against current nativist cognitive-evolutionary accounts.
First, we think that it is crucial to further refine the explanatory target of norm psychology by recognizing the full life cycle of norms (Legros & Cislaghi, 2020). In particular, norms should be defined as a group-level phenomenon. Thus, norms emerge when the majority of a group repeatedly shows the same behavior (i.e., a descriptive norm) or the same evaluation of a behavior (i.e., an injunctive norm or a prohibition) in a specific (set of) context(s) for a certain period of time. This behavioral regularity is “social” in the sense that it emerges from the interaction of group members. Simply stated, the life cycle includes the emergence, the stabilization and persistence, the change or mutation, and the removal of a norm. Consequently, the task of norm psychology should be to elucidate the mechanisms (i.e., [neuro]cognitive, affective, and motivational processes) driving the behaviors contributing to this life cycle. In other words, we argue that Heyes’s behavioral categories (compliance, enforcement, and commentary) are somewhat too narrow. These categories focus only on the stabilization, persistence, and spread of norms. However, the change of norms driven by the independence and deviance of individuals (e.g., minority influence; Germar & Mojzisch, 2021) is equally important for the life cycle of norms (only the commentary category seems to represent these processes). Hence, Heyes should be more explicit about how the full life cycle of norms is represented by the proposed behavioral categories. Likewise, Heyes should be more explicit about how the proposed domain-general and domain-specific mechanisms contribute to all stages of norms’ life cycle.
More generally, addressing the complete life cycle of norms and, hence, broadening the scope of what norm psychology should explain is vital for evaluating competing psychological accounts concerning their parsimony, their explanatory breadth, and their predictive power (Murayama et al., 2014; Trafimow & Uhalt, 2015).
Second, we caution against organizing the psychological mechanisms of norm psychology within a dual-system framework. As demonstrated by Keren and Schul (2009), dual-system models have at least two fundamental problems. The first problem is the use of dichotomies as qualitative labels to characterize the two classes of processes (e.g., in Heyes’s account: fast vs. slow, automatic vs. controlled, associative vs. rule-based). This often leads to the “not-the-liver” fallacy (i.e., attributing everything that is not a particular attribute to only one other contrasting attribute, e.g., everything that is not “automatic” is necessarily “controlled”). Furthermore, intrinsically continuous dimensions (e.g., ranging from “slow” to “fast”) and developmental trajectories (e.g., initially “controlled” then becoming “automatic”) are erroneously characterized as dichotomies (e.g., either “slow” or “fast” or either “controlled” or “automatic”). For example, analogous to learning to drive (the example used by Keren & Schul, 2009), learning to navigate in a new social context with unknown norms might initially require attentional resources and cognitive control, but with practice and experience, socially navigating in this context becomes “automatic.” Even when Heyes’s model allows for such developmental trajectories, a new fundamental problem arises: How does the transition between the two qualitatively distinct systems occur?
The second problem is that dual-system accounts necessarily imply that the two systems do not only differ in one (fast vs. slow) but in all attributes across all contexts (fast and automatic and associative and . . . vs. slow and controlled and rule-based and . . .), which is hard or perhaps even impossible to prove (e.g., with only three binary features, the existence of five out of six possible combinations must be disproved). Similar to other research areas (Erb et al., 2003; Keren & Schul, 2009), Heyes’s review of the norm literature does not reveal this level of evidence.
As discussed in more detail elsewhere (Garcia-Marques & Ferreira, 2011; Keren & Schul, 2009; Mugg, 2016), these problems lead to rather vague theoretical models that are difficult to falsify. Thus, we think that Heyes’s cultural-evolutionary account could profit from dropping the dead weight of the dual-system assumptions. Instead, Heyes’s account should focus more on developing the central claims of her account: Genetically inherited domain-general processes (that are not necessarily fast, automatic, etc.) and socially inherited domain-specific processes (that are not necessarily slow, controlled, etc.) underlie human normative behavior (which contribute to all stages of norms’ life cycle).
Then, research can test Heyes’s account against the nativist in two successive steps. First, research should ask whether a particular normative behavior is driven by domain-general (Heyes’s account) or domain-specific processes (nativist accounts). And only when there is evidence for a domain-specific mechanism should research then ask whether this mechanism is predominantly genetically (nativist accounts) versus socially inherited (Heyes’s account). Fortunately, Heyes already developed the excellent poverty-wealth scheme that studies can apply to answer these two questions.
