The authors of this intriguing and thoughtful target article are keenly aware that there is room to worry about how much of our moral psychology can be comfortably assimilated into the resource-rational bargaining framework they propose (e.g. altruistic helping?). We explicitly lay all such concerns aside to argue that even for the sorts of examples and contexts to which their account seems to be applied most naturally and comfortably, the authors have neglected to consider or explain the most salient and puzzling features of human moral thought and experience and have thus failed to provide a theory of moral cognition at all.

As numerous authors have emphasized (e.g. Nichols, Reference Nichols2004, Joyce, Reference Joyce2006, Goodwin & Darley, Reference Goodwin and Darley2008, Stanford, Reference Stanford2018), humans famously treat moral norms, demands, and obligations not as fundamentally analogous to contracts or agreements between individuals but as standards of behavior applying to all individuals regardless of their own desires, preferences, or even explicit agreements. If I judge that slavery or stealing is wrong, I do not see this judgment as dependent on any sort of agreement or contract (not even an implicit or idealized one) nor do I think it’s wrongness could be suspended or overturned by any agreement: we think slavery would still be wrong even if a whole society or group agreed to permit it. Indeed we think it would remain wrong even if the whole world (including the slaves!) agreed that it was a permissible practice and that we ourselves would be making a profound moral mistake if we judged that it was acceptable no matter how many others agreed with us.

All this is to say that humans characteristically experience moral norms, demands, and obligations as (somehow!) externally imposed on us rather than agreed to by us or by the communities and social groups to which we belong. Such “externalization” is often regarded as the distinctive signature of genuinely moral norms (such as prohibitions against slavery or stealing) in contrast to our experience of merely conventional norms (such as rules concerning which spoon to use for soup, which side of the road to drive on, and whether gum-chewing is allowed in class). Generally (Nichols, Reference Nichols2004), moral norms but not conventional norms are regarded as “generalizable” (i.e. applying to people everywhere and everywhen, whether or not they know or care about them) and as beyond the power of an authority figure to suspend: children allow that a teacher can suspend a school rule about chewing gum for her classroom, for example, but not a rule prohibiting hitting or stealing from other students. We apply such externalized norms to ourselves and to others not only regardless of our and their desires or preferences but also regardless of any conventions, contracts, and even explicit agreements embraced by members of their society (or any other). Equally famously, we regard conventional norms as merely hypothetical imperatives, ultimately specifying what to do if we hope to achieve particular ends (like avoiding pain, preserving our social acceptance, or maximizing mutual gain), but we treat moral norms instead as categorical imperatives that tell us what we must and must not do no matter what we want or care about (Kant, 1785/Reference Kant1998). More recent research has challenged whether the traditionally recognized features of moral norms invariably cooccur, whether all and only moral norms exhibit these characteristic features, and even whether there is any distinctively moral domain of cognition to examine. But this new research does not threaten the idea that there are systematic differences between our experience and cognition of externalized and non-externalized norms or that prototypical moral norms are among those we externalize most strongly (indeed, nearly as much as empirical facts, see Goodwin & Darley, Reference Goodwin and Darley2008).

It is therefore telling that so many of the examples of norms and norm violations the authors consider throughout the paper concern conventional rather than moral norms (tipping, driving on the right, prohibiting nonresident street parking, restricting usage of a well, slurping soup, waiting in line, etc.). Such examples simply obscure the need to explain the most puzzling features of our experience and cognition of prototypical moral norms, demands, and obligations. The authors do, of course, suggest ways in which resource-rationality might leave us applying previously established or accepted norms or heuristics without any concern for present bargaining or negotiation. But these proposals do absolutely nothing to explain why such norms or heuristics should be experienced as having externalized features that so sharply distinguish them from the sorts of subjective preferences that motivate us, as well as the conventional norms and even explicit agreements that characteristically emerge, when agents with conflicting interests seek to maximize the mutual satisfaction of preferences through bargaining or negotiation.

We suggest a different (and more comprehensive) global picture. Human moral psychology evolved to protect our increasingly hypercooperative and altruistic dispositions from exploitation. To do so, it classifies a subset of our norms as special in ways that lead us to simultaneously (1) be motivated by those norms ourselves and (2) use adherence to the very same norms as a criterion in evaluating the desirability of candidate partners in social interaction. Externalization serves to ensure that precisely the same set of norms are used in both roles (avoiding both exploitation and exclusion), even under normative plasticity, that is, even as we introduce new norms and drop old ones, change existing norms, or apply existing norms in new ways (for details, see Stanford, Reference Stanford2018). On this view, what the authors have done is simply show in detail that the resolutions of some conflicts of interest between agents deploying norms (both externalized and nonexternalized) turn out to approximate ideal bargaining solutions. Given that human cooperation was and is adaptive because it allows us to maximize cooperative dividends while preserving incentives for all parties to keep cooperating, this is unsurprising. It is in any case not the key to understanding the emergence, adaptive advantages, phylogenetic trajectory, or fundamental operations of our moral psychology more generally.