Jara-Ettinger and Dunham argue that the institutional stance can be used to interpret and analyze behavior and that it is generative, allowing us to represent and create arbitrary structures. While most of their examples focus on roles for individuals like barista, assistant, or the most dominant member of a baboon troop, they also connect the institutional stance to social kinds like race, gender, and caste kinds, arguing that the institutional stance “offers the possibility of envisioning and enacting new and more just futures” (39). Here I draw out differences between these cognitive projects applied to, on the one hand, individuals playing roles and those roles themselves (e.g., a barista, the barista role), and, on the other, to individual kind members and social kinds themselves (e.g., a woman, the kind women). These suggest the institutional stance is not utilized in both cases, or, at least, that more work is needed to show that it is.

Imagine a basketball team. It has a structure with positions like point guard and center. Team members play these roles. Players’ behavior can be interpreted through the roles they play. Other basketball teams share the same structure, but on those teams, different people play each role. And each team might have different players at different times. We can also imagine roles themselves changing – for instance, the role of point guard coming to involve more shooting as part of its function.

Now think of social kinds like women and men or White, Black, and Asian people. Like all other kinds, these can vary in members across time. And like many other kinds, any member is usually thought to be a member for life. As Jara-Ettinger and Dunham note, many theorists have argued that these kinds should be theorized as structural. For example, gender might be analyzed in structural terms so that what it is to be a woman, for instance, is to be positioned as subordinate to men (Haslanger, Reference Haslanger2000). Here, theorists are not considering some particular group with a hierarchical structure (like a particular basketball team with its structure). Rather, it is the kinds themselves that are analyzed as structural (Ritchie, Reference Ritchie2020).

Does the institutional stance enable the interpretation of individual behavior through these social kinds? For instance, might people interpret the behavior of an individual woman or Black person not through their mental states, but through their being members of the kind of women or Black people, respectively? It seems that people do interpret behavior and aptitudes in terms of kind membership, but this is compatible with many ways of reasoning that do not involve the institutional stance. For instance, people might reason that these kinds have essences, and that this explains the behavior of kind members.

Further, there are several pressures that support the use of the institutional stance for roles like barista that do not apply well to gender, race, and caste. First, individuals can shift between roles across time. A creature might move up in a dominance hierarchy. A bee might shift from forager to receiver. Eric might be an assistant at one point, an acquaintance later. Second, there is fungibility in role-fillers. Steve might play the role of point guard, then Chris subs in. One employee at a car rental agency might step in to continue a conversation with a customer. By representing roles and structures themselves, we have the tools to apply different roles to the same person and a single role to different people. Being able to represent roles and structures themselves also provides resources to analyze and imagine new roles and structures.

However, neither of these features applies well to kinds like women, Brahmins, or Black people. Return to our example of gender social kinds. As noted above, people tend to think of gender as stable across a person’s lifetime. There are not many changes in roles played (in a game, after 5 pm, etc.) to suggest that these are roles that one might play only temporarily. And, it would be very strange (and sexist!) to imagine that Sarah could just “sub in” for Laurie as a woman. Roles like barista and kinds like women pattern differently on both features. So, it is not clear that interpreting via roles or imagining modification to roles (baristas, point guards) gives us resources to interpret via or imagine alternative ways a social kind might be.

Perhaps an appeal to structural thinking, which Jara-Ettinger and Dunham say “aligns” with their arguments about the institutional stance, will provide the needed support to bridge this gap (19). Research on structural thinking has found that children’s and adults’ explanations and judgments of mutability are sensitive to structural information (Vasilyeva & Lombrozo, Reference Vasilyeva and Lombrozo2020; Vasilyeva, Gopnik, & Lombrozo, Reference Vasilyeva, Gopnik and Lombrozo2018; Yang, Naas, & Dunham, Reference Yang, Naas and Dunham2021). The structural conditions tested rely on differences in access (e.g., to more education or to different toys) or constraints (e.g., fewer restrictions on hiring). Given different environmental structures, children and adults reason that a group (girls, boys, fictional Borunians) might hold higher-paying jobs or play with different toys. Does this work suggest that we represent social kinds themselves in structural ways (i.e., as positions or roles in structures)? Does it show that people take social kinds themselves to be changed in such scenarios? Not directly. Rather, what I take it to show is that judgments about the stability of a property, inductive potential, or adequacy of an explanation can be connected to external structural sources. But this is compatible with the kind itself, not figuring in the interpretation of behavior. It is compatible with non-structural or even essentialist views of kinds and kinds themselves remaining unchanged, even if in some environments, individuals in these kinds display different behaviors.

Future work on the institutional stance ought to further explore its applicability to social kinds themselves, especially if this is useful for analyzing, imagining, and manifesting a more just social world.