Infant research has accumulated evidence showing that young humans can understand abstract interactional roles, guided by the individual goals of other agents. For example, infants can identify the role of a chaser as opposed to a chasee (Yin & Csibra, Reference Yin and Csibra2015) based on their intention to follow and catch another entity. Infants can also identify a giver as opposed to a taker (Tatone & Csibra, Reference Tatone and Csibra2024) or a communicator as opposed to another agent who does not or cannot engage in communicative information transfer (Tauzin & Gergely, Reference Tauzin and Gergely2018, Reference Tauzin and Gergely2019). The foundations of recognizing interactional roles might also be present in nonhuman animals, allowing them to identify predetermined interactional roles, such as “prey” or “predator,” in order to survive. In contrast to recognizing interactional roles, representing institutional roles should allow infants to understand that the behavior of an individual in a certain context is guided by the internalized aims and norms of the group to which this individual belongs. Crucially, however, there is no direct evidence that the comprehension of institutional roles – or, an institutional stance – is present early in development, nor that it has a specialized brain region or neural mechanisms dedicated to process such information. Although the absence of evidence is not evidence of absence, it is nevertheless necessary to examine whether an institutional stance constitutes a genuinely distinct ability, or whether it can be parsimoniously explained as an additional functional adaptation of an early-emerging cognitive skill with well-characterized neural underpinnings.

Accumulating empirical evidence suggests that human infants have precocious inferential skills to interpret the instrumental (Gergely & Csibra, Reference Gergely and Csibra2003) and communicative behavior of others (e.g., Tomasello et al., Reference Tomasello, Carpenter, Call, Behne and Moll2005; Vouloumanos, Onishi, & Pogue, Reference Vouloumanos, Onishi and Pogue2012; Vouloumanos, Martin, & Onishi, Reference Vouloumanos, Martin and Onishi2014) and perform adequate responses based on these inferences. When relying on these inferential skills, infants can take into account the situational context provided by the environment while assigning and maintaining representations of other agents’ goals and states of informedness. In communicative interactions, for instance, they can draw adequate inferences to interpret a request about retrieving a target object based on its location, the informedness of this person about it, and the intention of the communicator (Király, Oláh, & Kovács, Reference Király, Oláh and Kovács2023; Tauzin, Call, & Gergely, Reference Tauzin, Call and Gergely2024). Another line of research shows that, beyond their early inferential capacities, young infants can identify group membership on the basis of perceptual cues (Powell & Spelke, Reference Powell and Spelke2013, Reference Powell and Spelke2018). Infants expect members of a given group to behave similarly, interpreting such behavioral similarity as a signal of social affiliation (similar to other species, see e.g., Paukner et al., Reference Paukner, Suomi, Visalberghi and Ferrari2009).

Crucially, these abilities may form a sufficient basis to gradually understand and learn about institutional roles without possessing an institutional stance. By relying on their understanding of goal-directed behavior, infants may infer the instrumental goals of other agents in social interactions. For instance, they may understand that an exchange of objects (i.e., money and goods) occurs during shopping without yet understanding the institutional role of a shop clerk. In communicative interactions infants might also be able to draw inferences about the epistemic goals of others. For example, they may recognize that different educators possess the same relevant information about generic facts (Egyed, Király, & Gergely, Reference Egyed, Király and Gergely2013) even if they have no understanding of the institutional role of such individuals. Thus, by relying on the situational context as a cue and their ability to recognize group members, they may learn over the course of repeated interactions that, irrespective of the identity of the interacting individual, different people behave in a similar manner in given situations. Presumably, this is sufficient for infants to begin forming abstract representations of simple institutional roles (such as a shop clerk or a teacher) after repeated exposure to interactions with different individuals occupying these roles.

Importantly, this capacity is likely to be complemented by another early-emerging ability in human infants: their communicative and linguistic competence. Infants can infer that labels refer to the same group of entities (Pomiechowska et al., Reference Pomiechowska, Takács, Volein and Parise2024). Acquiring labels can, therefore, highlight commonalities across different individuals fulfilling the same role by referring to them with the same word, such as “clerk” or “educator,” through a categorization effect. Consequently, even with limited direct experience of institutional roles, children can learn about them through language.

I thus propose that identifying institutional roles is supported by infants’ sophisticated inferences about situational constraints, instrumental and communicative interactions, group behavior, and language at the beginning of development. These abilities have been shown to be present in young human infants and are supported by dedicated neural circuits that process domain-relevant information (such as the posterior superior temporal sulcus with respect to goals, see Begus et al. (Reference Begus, Curioni, Knoblich and Gergely2020); the right temporo-parietal junction with respect to informational asymmetries, see Hyde et al. (Reference Hyde, Boas, Blair and Carey2010); Liu et al. (Reference Liu, Moss, Ting and Hyde2025); the language circuit with respect to linguistic information). Possessing such skills may enable infants to recognize that certain interactional roles are stable across contexts and independent of the individual who occupies a given role in a specific interaction. This recognition requires active and presumably effortful inferential processes early in development in order to learn and generalize from experience about interactional roles, but it does not necessitate an institutional stance. Thus, understanding institutional roles may be a natural outcome of a prolonged learning process grounded in the aforementioned capacities, rather than an independent skill. This, in turn, may enable humans to develop fast access to, implicit recognition of, and sophisticated inferences about institutional roles as they grow older, through the automatization of these processes, similar to a music enthusiast who does not adopt a musical stance, but instead draws on a set of basic cognitive mechanisms to differentiate between genres. In addition to being more parsimonious, this account is consistent with current evidence, as the understanding of institutional roles appears to emerge slowly and gradually in children, perhaps without involving specialized brain circuits to understand institutions.