We welcome Tomasello's persuasive account of the evolutionary and developmental origins of moral obligation as a “bipolar attitude” towards moral decisions. Tomasello suggests that the binding quality of obligation generates a sense of a moral “we” as “a kind of cooperative identity” (sect. 2.1.1, para. 1). We argue that, while Tomasello's account of the origins of obligation to cultural group norms is correct in its outline, it overlooks significant aspects of the internal dynamics of how obligation and identity functions in cultural groups. We argue that the “we” is frequently divided so that the resulting moral obligations are multiple and sometimes contradictory. As such, moral obligation appears less bipolar and more multipolar, with its judgments looking less unidirectional than Tomasello assumes.

This moral “we” is an important claim that links Tomasello's account with work on social identity and morality (e.g., Aquino & Reed Reference Aquino and Reed2002). For us, Tomasello's picture (and the work on social identity of morality) offers only a partial account, since they overlook the possibility of a divided “we.” Namely, if a cultural group holds multiple moral values, those different values can drive different obligations for a single decision. For example, Kim and Jo are both Christians. Sam, a fellow Christian in their community, was caught stealing from both Kim and Jo. In response, Kim advocates punishment, while Jo advocates forgiveness – both supported by community norms. Sam breaks the wider community's obligations for selfish reasons, but each of Kim's and Jo's responses seems to the other to fail to meet significant obligations within the Christian community's norms. However, each is emphasising different but equally important moral motivations. This kind of phenomenon is echoed in moral ambivalence in other religious, political, and ideological groups. It can be explained straightforwardly by a social identity approach, in which the content and context of specific identities affects individuals’ thoughts and actions. Research by Van Tongeren et al. al. (Reference Van Tongeren, Welch, Davis, Green and Worthington2012; Reference Van Tongeren, Newbound and Johnson2016) supports this notion as they have shown how religious individuals’ judgments can be shaped by differing virtues valued within the same identity group. Our ongoing research indicates that differing values within the same social identity can lead to diverging and ostensibly contradictory moral decisions (Stewart Reference Stewart2016; Stewart et al. forthcoming Reference Stewart, Franks and Gleibs2019). Using variations on the trolley dilemma (Foot Reference Foot1967; Thomson Reference Thomson1985), we show how the content of a single identity can interact with different contexts to lead to contrasting moral obligations and equally contrasting moral decisions. This understanding adds nuances that are not accounted for in Tomasello's explanation of “excuses, justifications, or apologies,” since multiple obligations can create multiple reasonable responses within a social group.

This quality is not rare in morality. The ability to support apparent internal contradictions seems to be a cornerstone of successful moral, religious, and ideological systems (i.e., ones that have been transmitted across many generations of cumulative cultural evolution, and are implicated in cultural group selection: e.g., Richerson et al. Reference Richerson, Baldini, Bell, Demps, Frost, Hillis, Mathew, Newton, Naar, Newson, Ross, Smaldino, Waring and Zefferman2016). Curry et al. (Reference Curry, Mullins and Whitehouse2019) and others have demonstrated differing values of morality across and within cultural groups, which can lead to differing decisions over the same issue (for similar cases, see Graham et al. Reference Graham, Haidt, Koleva, Motyl, Iyer, Wojcik and Ditto2013; Greene Reference Greene2013; Legare & Gelman Reference Legare and Gelman2008; Hall et al. Reference Hall, Franks, Bauer and Bangerter2019). Extending Tomasello's concept of the “special social structure” for bipolar obligations, a complex moral system supports group cohesion. In complex cultural systems, failing to meet one obligation in fact may be following a competing obligation. Competing obligations may arise from different values within the group's moral system that apply to the decision in question but which are equally or more justifiable in the context. As such, the concept of obligation, when placed in concert with the social identity approach and its emphasis on context, offers a clearer understanding of how morality functions within and between cultural groups.

Our approach also suggests that the evolutionary picture of the origins of cultural-group obligation requires a role for a divided “we” with its multiple obligations. Tomasello's account does not provide this: He offers a view of separate moral hunter-gatherer tribes who do not interact to any significant degree, akin to tribes in Greene's Tragedy of Commonsense Morality (Reference Greene2013). Distinctive moral systems bind together groups and generate social selection pressures for good collaborators who can fulfil obligations. The prevalent moral norms for each group are “univalent” in directing behaviour in specific areas towards a single moral judgment and behaviour. This description is a plausible evolutionary starting point of cultural-group-based obligation. However, other accounts of hunter-gatherer sociality suggest a long evolutionary history of cross-tribe integration (e.g., Chapais Reference Chapais2008; Reference Chapais2011; Hill et al. Reference Hill, Walker, Božičević, Eder, Headland, Hewlett, Hurtado, Marlowe, Wiessner and Wood2011). In these accounts, groups combined via pair bonding and mating, providing descendants with both the genes of the integrated tribes as well as their cultural traditions. This inter-tribe integration and interaction lays the foundations for a divided “we” with multiple and possibly contradictory obligations for members of the combined group. We do not claim that such changes necessarily led to seamless and harmonious integration of those different moral values. Rather, our suggestion is that the recurrence of such situations helped create social selection pressures not only for good collaborators, but also for people who are skilled navigators through complex morality and its resulting multipolar obligations. In a world with a divided “we,” these are the collaborators who could most skilfully cooperate within a group holding divergent values from component tribes.

Moral systems are complex. Contradictory moral values are held between and within groups, where they can prompt different responses. The notion of obligation, when treated as monolithic, underplays the complexity and context-dependence of such responses. By treating obligation as a dynamic function, the vacillation between values provides a clearer understanding of how morality and group identity are related both today and in our evolutionary past.