The target article proposes that domain-general psychological mechanisms are essential to creating cohesive, effective human groups. I agree, but elaborate on what domain-general mechanisms are, how they can evolve, and how they may control domain-specific mechanisms.

The target article avoids dealing with the difficulties involved in proposing domain-general psychological mechanisms that have been raised by evolutionary psychologists, relying on an analogy with the immune system that does not shed light on the evolution of domain-generality.

Domain-generality is evolutionarily ancient. Although there are a variety of evolved, special-purpose learning mechanisms, learning is also characterized by domain-general mechanisms that are able to achieve evolutionary goals by making novel and serendipitous associations with environmental cues. Such mechanisms are domain-general because they are able to respond adaptively to ephemeral, nonrecurrent environmental regularities that are detectable by the organism's sense organs (i.e., they are not restricted to statistical regularities over evolutionary time, as required by evolutionary psychologists; e.g., Barrett & Kurzban Reference Barrett and Kurzban2012; Tooby & Cosmides Reference Tooby, Cosmides, Barkow, Cosmides and Tooby1992; see MacDonald Reference MacDonald and Kush2013). Such mechanisms can evolve because of the evolution of affective cues (prototypically pain and pleasure) that signal the attainment or nonattainment of evolutionary goals (e.g., satiation of hunger, the pleasure of sexual intercourse) (MacDonald Reference MacDonald1991; Reference MacDonald and Kush2013; MacDonald & Hershberger Reference MacDonald, Hershberger, Burgess and MacDonald2005). This then allows humans and animals to alter their behavior in response to ephemeral, nonrecurrent environmental contingencies.

Affective motivational mechanisms imply a set of adaptive problems to be solved but whose solution is underspecified. Such systems enable the evolution of any cognitive mechanism, no matter how opportunistic, flexible, or domain-general, that is able to solve the problem. Humans evolved the domain-general symbolic systems and reframing processes emphasized in the target article, as well as mechanisms underlying general intelligence, particularly the executive functions of working memory and a central executive able to direct attention and manipulate information that it receives from inputs from specialized, domain-specific mechanisms (e.g., spatial and symbolically coded information) (Chiappe & MacDonald Reference Chiappe and MacDonald2005; Geary Reference Geary2004).

The target article reviews several validated programs for change, but no attention is given to exactly what evolved mechanisms are involved and how domain-specific and domain-general they are. I argue that an evolutionary science of change must carefully tease out domain-specific, modular mechanisms from domain-general mechanisms, and in the case of the latter, must be clear on exactly how they are domain-general and how this promotes change.

An important contrast is between explicit and implicit processing. Implicit processing is characteristic of modules emphasized – often to the exclusion of domain-general mechanisms (e.g., Tooby & Cosmides Reference Tooby, Cosmides, Barkow, Cosmides and Tooby1992) – by evolutionary psychologists. Modules are evolutionarily ancient and their operation is fast, unconscious, automatic, and domain-specific (designed to solve specific problems). On the other hand, explicit processing is relatively recent and processing is conscious, relatively slow, effortful, and domain-general (Stanovich Reference Stanovich2004). Explicit and implicit processing are intimately related. Particularly important is the effortful control of implicit processing related to social and emotional behavior, including control over evolved modules designed to solve problems of survival and reproduction that were recurrent over evolutionary time (MacDonald Reference MacDonald2008; Reference MacDonald2009; Reference MacDonald2010). The inputs to effortful control mechanisms include a wide range of nonrecurrent information – information resulting not from evolutionary regularities but from explicit appraisals of costs and benefits related to the contemporary world of DNA testing and video recording.

The control of evolved modules by domain-general explicit processing is a critical aspect of behavioral change. MacDonald (Reference MacDonald2008) describes the psychology and neurobiology of how effortful control can result in behavioral change in the case of aggression, ethnocentrism, emotional behavior, and drug use.

Domain-general mechanisms can indeed produce a wide variety of phenotypes for selection to act upon (the analogy of the immune system used in the target article), but it is not just a matter of randomly producing thousands of hopeful monsters. Explicit processing involves taking in information from a variety of modular systems (e.g., perceptual information, affective desires such as sexual desire), coordinating and integrating it, and making plans of action that may involve effortful control of modular desires (Geary 2005; MacDonald Reference MacDonald2008). But it may also include envisioning possible future states (e.g., utopian visions of the future central to Marxist ideology made possible by symbolic processing) and rationalizing group aims (MacDonald Reference MacDonald2009; Reference MacDonald2010), and finally deciding to act.

Explicit processing is likely unique to humans, or at least is so highly elaborated among humans that there is a qualitative difference between humans and animals. Explicit processing is required to enable at least some of the qualities essential for successful groups noted in the target article, including the following:

• Group identity. Some animals have a strong sense of group identity and group boundaries (e.g., van der Dennen Reference van der Dennen, Thienpont and Cliquet1999) that is presumably modular. However, the rights and obligations of human group membership are typically explicitly articulated – they are often formally written and may be subject to judicial oversight (e.g., being in the military, a union, or a religious sect). The vast differences among human groups in rights and obligations and the fact that rights and obligations may change rapidly in the contemporary world strongly suggest that they are processed explicitly rather than exclusively via evolved mechanisms designed to track regularities of group living over evolutionary time.

• Proportional costs and benefits; monitoring. Explicit processing allows people to build explicit representations of others' reputations (e.g., for shirking communal work) and the costs and benefits of actions, thus enabling human cooperation (MacDonald Reference MacDonald2008). Whereas modular mechanisms have built-in assessments of costs and benefits (e.g., Buss & Shackelford Reference Buss and Shackelford1997), explicit representations of others' reputations are able to track rapidly changing, novel environmental contexts (e.g., developing new technology to assess others' reputations) and are able to form explicit memories and written records of past interactions with others, thus enabling indirect reciprocity (e.g., Semmann et al. Reference Semmann, Krambeck and Milinski2005; Smith Reference Smith2005).

• Punishment; fast and fair conflict resolution. Assessing the fairness of conflict resolution is often complex, requiring explicit knowledge of contemporary contexts incomprehensible by mechanisms attuned to regularities of the evolutionary past (e.g., assessing whether a group member is embezzling funds via computer fraud). On the basis of such explicit assessments, punishment can be finely graded, from gentle reprimands to expulsion, to ensure the viability of the group.