Drawing on developmental, comparative, and philosophical perspectives, Tomasello gives an account of moral obligation, which he argues has received “almost total neglect” from psychologists. We agree with Tomasello that a sense of obligation arises from cooperative agreements between humans, often functioning as a powerful “stick” to prevent behavior that would lead to one's own guilt or partner resentment. In fact, psychologists focusing on dyadic and group behavior have long discussed such means of prioritizing “we” over “me” in human societies: Kelley's (Reference Kelley1982) discussion of transformations-of-motivation from individual to shared concerns is a key example. The importance of cooperative norms and the risk of resentment and guilt implied by breaking them also have been emphasized by researchers such as Walster et al. (Reference Walster, Walster and Berscheid1978) who traced shifts from “me” to “we” and identified resultant obligations as part of their equity theory. Others have noted that moral judgments of the self and other are caused by meeting or failing to meet cooperative expectations, the nature of which often vary by relational context (e.g., Bloom Reference Bloom2011; Clark & Boothby Reference Clark and Boothby2013; Rai & Fiske Reference Rai and Fiske2011).

Different relationship types have been distinguished in terms of the adaptive functions they serve (Bugental Reference Bugental2000; Clark & Mills Reference Clark and Mills1979, Reference Clark, Mills, Van Lange, Kruglanski and Higgins2012; Fiske Reference Fiske1992); the demands people place on cooperative partners depend on the origins and nature of their interdependence (Bugental Reference Bugental2000; Clark & Mills Reference Clark and Mills1979). We suggest that taking into account who “we” are (in Tomasello's sense), as well as “why” and “how” we are cooperating based on the functions normatively served within our relationship(s), will be crucial for making progress in the social psychology of moral obligation.

Consider Wendy, who could easily provide a free hot meal to a hungry young child named Peter, but neglects to do so. Has she breached a moral obligation? It depends on the nature of the relationship between them. If Wendy is Peter's mother, the answer is probably yes (barring unusual circumstances). If Wendy is the unrelated owner of a local restaurant, the answer is probably no. Or consider John, who fails to pay his driver, Susan, for a ride to the airport. If Susan is his taxi driver, he likely has breached an obligation. If Susan is his sister or spouse, however, he likely has not.

These examples highlight a difference between two relationship types described in the literature: communal and exchange (Clark & Mills Reference Clark and Mills1979, Reference Clark, Mills, Van Lange, Kruglanski and Higgins2012; see also Fiske Reference Fiske1992). Communal relationships are often exemplified by friends, family, and romantic partners. In these relationships, people (normatively) assume a special responsibility for one another's welfare. They track each other's needs and desires (Clark et al. Reference Clark, Mills and Corcoran1989), note the responsiveness of the other to their own needs (Clark et al. Reference Clark, Dubash and Mills1998), and offer non-contingent support as necessary to promote partner welfare (Clark et al. Reference Clark, Ouellette, Powell and Milberg1987). Failure to offer such support reduces liking, elicits hurt feelings (Clark & Mills Reference Clark and Mills1979; Lemay et al. Reference Lemay, Overall and Clark2010), and, we expect, triggers negative moral judgments (e.g., resentment). By contrast, failure to directly compensate the other for needs-responsive support in a communal relationship typically does not cause such negative responses (as with John/Susan above).

Compare this to exchange relationships, often exemplified by casual acquaintances or customers/sellers. In these relationships, needs typically are not tracked (unless for purposes of selling or exchanging), but the other's contributions to joint tasks are tracked (Clark et al. Reference Clark, Mills and Corcoran1989; Clark Reference Clark1984). Goods and services are (normatively) provided on a contingent basis, and costly help is generally not offered in non-emergency situations (Clark et al. Reference Clark, Ouellette, Powell and Milberg1987). Here, repaying debts and willingness to accept payments leads to enhanced liking (Clark & Mills Reference Clark and Mills1979) and does not elicit negative judgments (McGraw & Tetlock Reference McGraw and Tetlock2005), whereas failure to pay likely would be judged as morally objectionable (and refusal to accept payment would seem strange and uncomfortable, if not immoral). The degree of perceived wrongfulness of an action likewise depends on relational context (Simpson & Laham Reference Simpson and Laham2015; Simpson et al. Reference Simpson, Laham and Fiske2016; Tepe & Adymi-Karakulak Reference Tepe and Adymi-Karakulak2018).

Communal and exchange relationships serve different functions. Bugental (Table 1. She ties each one to a specific adaptive goal or recurrent coordination problem faced by our species.

Table 1. Relationship functions, adapted from Bugental (Reference Bugental2000).

Everyday relationships serve one or more of these functions to varying degrees in different contexts. For example, a parent-infant relationship normatively serves the attachment and hierarchy functions across almost every context, but not the mating, reciprocity, or coalition functions. Teammates are expected to serve the coalition function, with a captain also serving the hierarchy function, but in most societies the captain normatively does not serve the mating, attachment, or reciprocity functions with other team members.

Feelings of moral obligation (and judgments of blame for failing to uphold such obligations) often will be specific to the functions that are central to the relationship at hand. So, a parent would be heavily blamed (and would likely feel guilt) for failing to serve the attachment function with their infant, but a captain would not be so blamed (nor would likely feel guilt) for failing to serve this function with a teammate, and so forth.

Two final points regarding the importance of asking who “we” are. Tomasello focuses on “we” relations that are voluntarily entered into by individuals who regard each other as relevantly equal in terms of obligations, rights, and power; but the parent-infant example highlights that at least some relationships are unequal in these respects and may be non-voluntarily entered into. Indeed, some “we” relations are imposed on people by the situations in which they find themselves, some of which may be functional for one party to the relationship but not the other (see Kelley et al. Reference Kelley, Holmes, Kerr, Reis, Rusbult and Van Lange2003).

Second, Tomasello focuses on negative judgments resulting from obligation-failures (the sticks). It will be fruitful to consider the personal and interpersonal rewards to be gained by obligation-fulfillment (the carrots) as well. Previous work suggests that serving relationship functions in ways that are desired and which exceed normative expectations will disrupt smooth, habitual interdependent routines and will likely elicit positive emotions (Berscheid & Ammazalorso Reference Berscheid, Ammazalorso, Fletcher and Clark2001). Might moral praiseworthiness judgments be similarly elicited?

To summarize, the varied nature of relational contexts and functions shape moral judgments (Bloom Reference Bloom2011; Clark & Boothby Reference Clark and Boothby2013; Haidt & Baron Reference Haidt and Baron1996; Rai & Fiske Reference Rai and Fiske2011), as well as many other psychological phenomena (Reis Reference Reis2008; Clark et al. Reference Clark, Lemay, Reis, Funder and Sherman2017). Future work should consider who “we” are, taking into account (1) the different functional and normative bases of different relationship types, (2) both voluntary and non-voluntary forms of interdependence, and (3) both positive and negative moral judgments.