When people take actions that cannot be understood as efficient means toward goals, observers often interpret the action as acts of communication. The target article focuses on one type of message that can be communicated in such situations: That the action is a ritual that defines a social group and that other groupmates should imitate such actions to maintain group cohesion and ingroup/outgroup lines. However, there are many cases where non-instrumental actions communicate another type of message: What kind of relationship the actor has with the target. For example, imagine you are observing two people who belong to the same group and are walking in one another's pathway. In one situation a person moves out of the way by stepping aside and continues walking. In another, the person bows deeply and then steps aside. While they both achieve the same goal of allowing the other person to pass, the second situation includes actions that are not necessary for achieving the goal of moving aside. How should the recipient of the action react to these two different actions? While we agree with the target article that the second action would be interpreted as social communication, we disagree it needs to be about marking ingroups from outgroups. The bower is going out of their way to acknowledge a specific type of relationship: a difference in rank exists. While this communicates something social, and indeed could communicate a desire for a relationship, it also communicates what kind of relationship those people have. In most cultures, the bower would be assumed to be lower ranked, while in some they may be assumed to be higher ranked (Kajanus, Afshordi, & Warneken, Reference Kajanus, Afshordi and Warneken2019). However, in either case the bowing does not need to be imitated for it to be meaningful for the bower, the person being bowed to, or the onlookers. In fact, most actions that communicate roles in hierarchical relationships are appropriately asymmetric.

There are many such actions including those that are cross-culturally salient, such as bowing, kneeling, gift-giving, and kissing, which are not efficient means toward instrumental goals, but also not meant to mark ingroups and outgroups by way of ritualized imitation. Instead, these actions acknowledge or establish certain types of relationships and relative roles within those relationships. Unlike actions that define group rituals, actions that communicate relationships can be asymmetric, and thus do not elicit copying. Indeed, many such actions may be an especially bad marker of ingroups because there are features of them that are shared across cultures. For example, in many cultures, lowering the body in an action communicates submission in a hierarchical relationship, and consensually transferring bodily fluids communicates an intimate or communal relationship (Fiske, Reference Fiske1992). Indeed, even infants recognize that a non-instrumental action that transfers saliva (putting one's finger in the other's mouth and then in one's own mouth) communicates a closer relationship than a similar action that does not transfer saliva (putting one's finger on the other's forehead, and then on one's own forehead; Thomas, Woo, Nettle, Spelke, & Saxe, Reference Thomas, Woo, Nettle, Spelke and Saxe2022): Infants predict that the recipient of the saliva-transferring action is more likely to respond to their social partner's distress, compared to the recipient of the forehead action. Neither of these actions has an obvious instrumental goal, and no imitation occurs in the study, yet infants use them to predict other actions associated with close affiliation or social intimacy. In summary, non-instrumental actions have social meanings, but not all such meanings concern rituals to mark the ingroup from the outgroup. This has implications for the cultural action framework proposed in the target article. In this framework, actions that have no obvious instrumental goal and that are causally opaque nudge observers into a “ritual stance.” However, if our proposal is correct, then these same attributes should sometimes nudge observers into something like a “relationship stance” where observers understand these actions as ways to communicate social relationships and roles within those relationships.

The critical question for future research is: How do people, including even infants, identify the relevant one, of the many possible, communicative goals of non-instrumental actions? When does a person interpret non-instrumental goals as ritual, when do they interpret them as a desire to communicate about a relationship, and when do they write off the action as meaningless? One possibility lies in proposals by Fiske (Reference Fiske1992); Kaufmann and Clément (Reference Kaufmann and Clément2014); and Thomsen and Carey (Reference Thomsen and Carey2013). These proposals suggest that infants are born with core concepts that allow them to recognize certain types of relationships, including the actions that go along with them. If this is true, then infants and children may be attuned to interpret other non-instrumental actions, which do not fit into these categories, as rituals, especially when they are imitated. While many rituals often “borrow” actions that would be a part of this innate repertoire, culturally specific rituals often include other arbitrary actions and symbols. For example, take the Catholic ritual of crossing oneself. Catholics often kneel while crossing themselves. Kneeling or making yourself small is a common marker of deference, indeed shared with other species (van Vugt & Tybur, Reference van Vugt, Tybur and Buss2014). Here the kneeling may serve to communicate deference to God. However, Catholics also cross themselves in a specific way. While the action carries meaning, it wouldn't be easily understood by someone who had never seen it before or didn't understand the context. It is an arbitrary symbol. Thus, new humans may solve the complex problem of understanding people's non-instrumental actions, first by asking whether they communicate about relationships and roles, then by computing whether they fit the structure of innate concepts of social relationships. If not, infants may then interpret them as ritual actions, especially when imitated. The bifocal stance theory could therefore be expanded to include non-instrumental actions that are meant to acknowledge or communicate desires for specific types of relationships or roles within them.