We welcome Julian Jara-Ettinger and Yarrow Dunham’s argument that the understanding and prediction of others’ behavior cannot be reduced to mental-state reasoning. Their article contributes to our long-standing efforts to demonstrate that social cognition is not reducible to theory of mind (Clément et al., Reference Clément, Bernard and Kaufmann2011; Kaufmann & Clément, Reference Kaufmann and Clément2007, Reference Kaufmann and Clément2014). Yet, although we are very sympathetic to their position, we have some reservations about their “institutional stance.”
The foundation(s) of institutions
In our view, Jara-Ettinger and Dunham’s main hypothesis is not controversial: in addition to navigating the social world by representing other minds, humans and, to some extent, other animals are able to “interpret social interactions in terms of role-based structures that constrain and regulate behavior via rule-like behavioral expectations” (p. 1). What is more problematic, however, is the claim that this ability is inherently institutional. The word institution appeared in Middle English before c. 1400, where it referred to “the action of establishing or founding (a system of government, a religious order, etc.)” (Etymonline). Put differently, institutions require that a social group collectively endows a physical substrate (an object, gesture, or action) with a specific meaning or function. To account for the constitutive backbones of institutional facts, and the collective recognition of “status function” they involve, the philosopher John Searle uses the formula “X counts as Y in context C” (Searle, Reference Searle1995): a piece of paper X (such as a ten-dollar bill), for instance, counts as a bearer of a certain value Y in the context of economic exchange C. Institutions are therefore complex phenomena. From a psychological perspective, they require the capacity to recognize that objects, persons, or events can possess properties that are not intrinsic to them but are instead added through a process of collective attribution. From a sociological perspective, institutions are sustained by intricate cultural devices (juridical rules, administrative organizations, police forces, etc.) that “objectively” constrain and limit agents’ own wills. Given the impersonal and external dimension of these constraints, it seems odd to restrict institutions to their “mentalistic” counterpart by defining them as a “type of cognitive representation consisting of a set of interconnected roles with a level of persistence and continuity” (sect. 4.1, para. 1).
Once they have defined their institutional stance, the authors attempt to imagine an evolutionary scenario that could explain how non-human animals might also use it to make sense of their conspecifics’ behavior. To do so, they must assume that social animals live within “proto-institutional” systems. One of their examples is ants, which live in complex social structures with clearly defined roles; such a system could be considered proto-institutional “because members of the social group do not, or at least need not, have any representation of roles or of the social structure itself” (sect 3.1, para. 3). Yet comparing insects’ social systems to institutions falls into what we could call retrospective determinism: it reads the traits of our institutional systems back into their supposed evolutionary precursors, as though these systems had been moving toward our own all along.
From institutions to social forms
Avoiding this teleological trap while preserving the idea that social cognition cannot be reduced to mentalizing requires a cognitively and evolutionarily parsimonious understanding of what social means. If we follow the sociologist Georg Simmel (Reference Simmel1910), the ontological building blocks of social life are elementary forms of reciprocal relationships: exchange, hierarchy, alliance, competition, nurturance, and so on. Those relationships are primarily recognizable through the observable behavior of the individuals embedded in them. For instance, the low posture and shifty gaze of a chimpanzee approaching a dominant member indicate – almost in a perceptual manner – its social position within the group. This position comes with a set of expectations about what is likely to happen next. Observers will, for example, be highly surprised when witnessing a sudden attack by a subordinate on a dominant individual (Slocombe et al., Reference Slocombe, Kaller, Call and Zuberbühler2010). Children likewise develop early expectations about how individuals behave within hierarchical relationships (Charafeddine et al., Reference Charafeddine, Mercier, Clément, Kaufmann, Berchtold, Reboul and Van der Henst2015; Thomsen et al., Reference Thomsen, Frankenhuis, Ingold-Smith and Carey2011). The fact that such anticipations appear both in non-human primates and in young children suggests that the underlying cognitive mechanisms do not require a high degree of complexity. Interpreting these mechanisms as a rudimentary layer of a fully fledged “institutional stance” – one that presupposes representations of social structures built on internal models of “interconnected roles that regulate and guide how behavior unfolds over space and time” (sect. 4, para. 1) – does not seem evolutionarily plausible.
From the institutional stance to the social stance
To avoid moving backwards along the evolutionary path, it is better to follow the chronological order in which social structures – and the cognitive processes that support them – emerged (Kaufmann & Clément, Reference Kaufmann and Clément2014, Reference Kaufmann and Clément2024). This also preserves the arguments made by Jara-Ettinger and Dunham, provided that elementary forms of social life are not regarded as proto-institutional. Rather, they consist of reciprocal relations that social agents learn to process, progressively acquiring the ability to identify their essential properties – what we have called “naïve sociology.”
We are therefore more optimistic than they are about the role of scripts in how we use our “social stance” to make sense of others’ behaviors (Taylor et al., Reference Taylor, Gönül, Alexander, Züberbühler, Clément and Glock2023). Before becoming institutional experts, social agents observe the interactions around them and extract regular patterns that structure their social world. This external reality can be considered as “objective” as physical reality and cannot be reduced to internal representations alone.
This view has at least two advantages. From an epistemological standpoint, it avoids “retrospective determinism” while capturing both evolutionary continuity and the novelty of human sociality. From a cognitive standpoint, it grounds the institutional stance in a more elementary social stance, such as naïve sociology, that enables social agents to navigate a shared realm of social interactions well before the emergence of the roles, statuses, and normative commitments characteristic of institutions.
You have
Access
We welcome Julian Jara-Ettinger and Yarrow Dunham’s argument that the understanding and prediction of others’ behavior cannot be reduced to mental-state reasoning. Their article contributes to our long-standing efforts to demonstrate that social cognition is not reducible to theory of mind (Clément et al., Reference Clément, Bernard and Kaufmann2011; Kaufmann & Clément, Reference Kaufmann and Clément2007, Reference Kaufmann and Clément2014). Yet, although we are very sympathetic to their position, we have some reservations about their “institutional stance.”
The foundation(s) of institutions
In our view, Jara-Ettinger and Dunham’s main hypothesis is not controversial: in addition to navigating the social world by representing other minds, humans and, to some extent, other animals are able to “interpret social interactions in terms of role-based structures that constrain and regulate behavior via rule-like behavioral expectations” (p. 1). What is more problematic, however, is the claim that this ability is inherently institutional. The word institution appeared in Middle English before c. 1400, where it referred to “the action of establishing or founding (a system of government, a religious order, etc.)” (Etymonline). Put differently, institutions require that a social group collectively endows a physical substrate (an object, gesture, or action) with a specific meaning or function. To account for the constitutive backbones of institutional facts, and the collective recognition of “status function” they involve, the philosopher John Searle uses the formula “X counts as Y in context C” (Searle, Reference Searle1995): a piece of paper X (such as a ten-dollar bill), for instance, counts as a bearer of a certain value Y in the context of economic exchange C. Institutions are therefore complex phenomena. From a psychological perspective, they require the capacity to recognize that objects, persons, or events can possess properties that are not intrinsic to them but are instead added through a process of collective attribution. From a sociological perspective, institutions are sustained by intricate cultural devices (juridical rules, administrative organizations, police forces, etc.) that “objectively” constrain and limit agents’ own wills. Given the impersonal and external dimension of these constraints, it seems odd to restrict institutions to their “mentalistic” counterpart by defining them as a “type of cognitive representation consisting of a set of interconnected roles with a level of persistence and continuity” (sect. 4.1, para. 1).
Once they have defined their institutional stance, the authors attempt to imagine an evolutionary scenario that could explain how non-human animals might also use it to make sense of their conspecifics’ behavior. To do so, they must assume that social animals live within “proto-institutional” systems. One of their examples is ants, which live in complex social structures with clearly defined roles; such a system could be considered proto-institutional “because members of the social group do not, or at least need not, have any representation of roles or of the social structure itself” (sect 3.1, para. 3). Yet comparing insects’ social systems to institutions falls into what we could call retrospective determinism: it reads the traits of our institutional systems back into their supposed evolutionary precursors, as though these systems had been moving toward our own all along.
From institutions to social forms
Avoiding this teleological trap while preserving the idea that social cognition cannot be reduced to mentalizing requires a cognitively and evolutionarily parsimonious understanding of what social means. If we follow the sociologist Georg Simmel (Reference Simmel1910), the ontological building blocks of social life are elementary forms of reciprocal relationships: exchange, hierarchy, alliance, competition, nurturance, and so on. Those relationships are primarily recognizable through the observable behavior of the individuals embedded in them. For instance, the low posture and shifty gaze of a chimpanzee approaching a dominant member indicate – almost in a perceptual manner – its social position within the group. This position comes with a set of expectations about what is likely to happen next. Observers will, for example, be highly surprised when witnessing a sudden attack by a subordinate on a dominant individual (Slocombe et al., Reference Slocombe, Kaller, Call and Zuberbühler2010). Children likewise develop early expectations about how individuals behave within hierarchical relationships (Charafeddine et al., Reference Charafeddine, Mercier, Clément, Kaufmann, Berchtold, Reboul and Van der Henst2015; Thomsen et al., Reference Thomsen, Frankenhuis, Ingold-Smith and Carey2011). The fact that such anticipations appear both in non-human primates and in young children suggests that the underlying cognitive mechanisms do not require a high degree of complexity. Interpreting these mechanisms as a rudimentary layer of a fully fledged “institutional stance” – one that presupposes representations of social structures built on internal models of “interconnected roles that regulate and guide how behavior unfolds over space and time” (sect. 4, para. 1) – does not seem evolutionarily plausible.
From the institutional stance to the social stance
To avoid moving backwards along the evolutionary path, it is better to follow the chronological order in which social structures – and the cognitive processes that support them – emerged (Kaufmann & Clément, Reference Kaufmann and Clément2014, Reference Kaufmann and Clément2024). This also preserves the arguments made by Jara-Ettinger and Dunham, provided that elementary forms of social life are not regarded as proto-institutional. Rather, they consist of reciprocal relations that social agents learn to process, progressively acquiring the ability to identify their essential properties – what we have called “naïve sociology.”
We are therefore more optimistic than they are about the role of scripts in how we use our “social stance” to make sense of others’ behaviors (Taylor et al., Reference Taylor, Gönül, Alexander, Züberbühler, Clément and Glock2023). Before becoming institutional experts, social agents observe the interactions around them and extract regular patterns that structure their social world. This external reality can be considered as “objective” as physical reality and cannot be reduced to internal representations alone.
This view has at least two advantages. From an epistemological standpoint, it avoids “retrospective determinism” while capturing both evolutionary continuity and the novelty of human sociality. From a cognitive standpoint, it grounds the institutional stance in a more elementary social stance, such as naïve sociology, that enables social agents to navigate a shared realm of social interactions well before the emergence of the roles, statuses, and normative commitments characteristic of institutions.
Financial support
This research received no specific grant from any funding agency, commercial or not-for-profit sectors
Competing interests
None