Navigating the social world requires predicting others’ behavior. The target article posits that abstract institutional representations increase predictability by delineating the formal rules and social norms that govern the behavior of different role holders within organizational systems. It claims further that individuals form expectations concerning others’ behavior by combining representations of others’ roles in the social structure (e.g., boss, teacher, and parent) and representations of others’ mental states (e.g., motives, beliefs, and feelings). However, the target article leaves open the questions of why and how individuals synthesize predictions derived from institutional and mental state representations.
Why individuals enrich institutional representations with mental state representations
The claim that institutional environments – with their roles, rules, and norms – enable calibrated expectations about human behavior implicitly assumes that institutional environments are strong situations that restrict variance in human behavior (Li et al., Reference Li, Balliet, Thielmann and de Vries2025). However, not all institutional environments are strong situations. Tight environments feature well-defined norms and enforce them strictly, whereas loose environments enable individuals to have many degrees of freedom when choosing their actions (Gelfand, Harrington, & Jackson, Reference Gelfand, Harrington and Jackson2017). Supervisors and subordinates may rely on their institutional representations to predict each other’s behavior in tight environments but will likely need additional information about others’ mental states to form accurate predictions in weak situations and loose environments (e.g., are they motivated to do a good job?).
Further, individuals may mentally represent the same institution differently. For example, employees may represent their workplace as a hierarchy or as a network, with distinct consequences for perceived norms (Moyal, Tan, & Halevy, Reference Moyal, Tan and Halevy2023). This variability in mental representations of institutions means that a theory of the institutional context (e.g., workplace) may not be enough to predict others’ behavior. Doing so requires understanding how others mentally represent the institution.
Within organizations, many employees experience role ambiguity and role conflict, which capture lacking clear information about their role demands and facing contradictory role-based expectations, respectively (Schmidt et al., Reference Schmidt, Roesler, Kusserow and Rau2014). Role ambiguity and role conflict may increase variability in in-role behavior. Even when expectations are clear, employees may engage in deviant behavior, break rules, and innovate. By falling short of expectations, exceeding expectations, or violating expectations in other ways, coworkers may surprise each other (Halevy, Miclau, & Lee, Reference Halevy, Miclau and Lee2026). The institutional stance is insufficient to reliably predict the behavior of role holders who do not behave as expected.
Consider the case of convenience store robberies. Research shows most people expect to find clerks behind the register when they enter a convenience store, even though clerks sometimes stock shelves, mop floors, and take bathroom breaks. Convenience store robberies that begin while clerks violate robbers’ expectations by being away from the cash register end up being significantly more violent (DeCelles et al., Reference DeCelles, Kouchaki and Halevy2022). Although the institution (convenience store) and roles (robber, clerk) remain the same irrespective of the clerk’s position in the store, this deviation from a role-based expectation shifts the interaction toward greater violence. As this example illustrates, knowing the institutional context and people’s roles in it may be insufficient for predicting how an interaction will unfold.
One institution mentioned frequently in the target article is hierarchy. There is considerable variation among hierarchies, not only across cultures and historical periods (Muthukrishna, Henrich, & Slingerland, Reference Muthukrishna, Henrich and Slingerland2021) but also within the same society and historical period. Hierarchies can be relatively steep or flat, well-functioning or dysfunctional, based on dominance or prestige, and so on (Cheng et al., Reference Cheng, Tracy, Foulsham, Kingstone and Henrich2013; Yu et al., Reference Yu, Greer, Halevy and Van Bunderen2019). Further, the same position in the hierarchy may entail different duties and responsibilities in different institutional contexts (e.g., a CEO in a private company, a public company, and a nonprofit organization). The fact that similar institutions often give rise to very different normative systems (i.e., organizational cultures) is evident also from the ubiquity of cultural clashes following organizational mergers and acquisitions (Weber & Camerer, Reference Weber and Camerer2003). Hence, knowledge of the institutional context and individuals’ roles may be insufficient to predict others’ actions.
How theories of mind help calibrate predictions derived from institutional theories
Our simple solution extends Figure 1c in the target article. We propose that people regularly integrate institutional and mental state representations in the same way they enrich nouns with adjectives. Because role representations are insufficient to form accurate predictions about others’ behavior, people supplement roles with descriptors that capture others’ mental states. Parents to multiple children, or instructors who teach multiple students, know that the roles “child” and “student” are insufficient to describe how different individuals will behave in the same role. This is why we have numerous adjectives that add meaningful nuance to social roles: a helpful child, a curious student, a messy roommate, a nasty neighbor, a kind boss. In all these cases, the mentalistic stance meaningfully clarifies what to expect from the role holder.
Importantly, the adjective does not have to represent an enduring personality trait. It can capture a transient state that helps calibrate expectations. Consider the role of a barista, mentioned repeatedly in the target article. Keeping the role holder (barista) and institution (café) constant, we may form different expectations depending on whether the barista seems overworked and tired or well-rested and energized on a given shift. Their expectations concerning our tipping behavior may likewise depend on their perceptions of our mood that day.
In sum, social prediction requires more than knowledge of institutional structure and people’s roles in it. It requires mental state inferences. Trust between two romantic partners or business partners cannot be based only on roles and norms; it must incorporate beliefs about others’ feelings, intentions, and capabilities. The same logic applies beyond dyads and small teams. The large-scale coordination and cooperation that are necessary to provide public goods and solve collective action problems need more than a theory of institutions; they require beliefs about others’ mental states. Hence, important as the institutional stance is for forming expectations, understanding how people enrich institutional models with mental state inferences is essential for a complete and coherent theory of social prediction.
You have
Access
Navigating the social world requires predicting others’ behavior. The target article posits that abstract institutional representations increase predictability by delineating the formal rules and social norms that govern the behavior of different role holders within organizational systems. It claims further that individuals form expectations concerning others’ behavior by combining representations of others’ roles in the social structure (e.g., boss, teacher, and parent) and representations of others’ mental states (e.g., motives, beliefs, and feelings). However, the target article leaves open the questions of why and how individuals synthesize predictions derived from institutional and mental state representations.
Why individuals enrich institutional representations with mental state representations
The claim that institutional environments – with their roles, rules, and norms – enable calibrated expectations about human behavior implicitly assumes that institutional environments are strong situations that restrict variance in human behavior (Li et al., Reference Li, Balliet, Thielmann and de Vries2025). However, not all institutional environments are strong situations. Tight environments feature well-defined norms and enforce them strictly, whereas loose environments enable individuals to have many degrees of freedom when choosing their actions (Gelfand, Harrington, & Jackson, Reference Gelfand, Harrington and Jackson2017). Supervisors and subordinates may rely on their institutional representations to predict each other’s behavior in tight environments but will likely need additional information about others’ mental states to form accurate predictions in weak situations and loose environments (e.g., are they motivated to do a good job?).
Further, individuals may mentally represent the same institution differently. For example, employees may represent their workplace as a hierarchy or as a network, with distinct consequences for perceived norms (Moyal, Tan, & Halevy, Reference Moyal, Tan and Halevy2023). This variability in mental representations of institutions means that a theory of the institutional context (e.g., workplace) may not be enough to predict others’ behavior. Doing so requires understanding how others mentally represent the institution.
Within organizations, many employees experience role ambiguity and role conflict, which capture lacking clear information about their role demands and facing contradictory role-based expectations, respectively (Schmidt et al., Reference Schmidt, Roesler, Kusserow and Rau2014). Role ambiguity and role conflict may increase variability in in-role behavior. Even when expectations are clear, employees may engage in deviant behavior, break rules, and innovate. By falling short of expectations, exceeding expectations, or violating expectations in other ways, coworkers may surprise each other (Halevy, Miclau, & Lee, Reference Halevy, Miclau and Lee2026). The institutional stance is insufficient to reliably predict the behavior of role holders who do not behave as expected.
Consider the case of convenience store robberies. Research shows most people expect to find clerks behind the register when they enter a convenience store, even though clerks sometimes stock shelves, mop floors, and take bathroom breaks. Convenience store robberies that begin while clerks violate robbers’ expectations by being away from the cash register end up being significantly more violent (DeCelles et al., Reference DeCelles, Kouchaki and Halevy2022). Although the institution (convenience store) and roles (robber, clerk) remain the same irrespective of the clerk’s position in the store, this deviation from a role-based expectation shifts the interaction toward greater violence. As this example illustrates, knowing the institutional context and people’s roles in it may be insufficient for predicting how an interaction will unfold.
One institution mentioned frequently in the target article is hierarchy. There is considerable variation among hierarchies, not only across cultures and historical periods (Muthukrishna, Henrich, & Slingerland, Reference Muthukrishna, Henrich and Slingerland2021) but also within the same society and historical period. Hierarchies can be relatively steep or flat, well-functioning or dysfunctional, based on dominance or prestige, and so on (Cheng et al., Reference Cheng, Tracy, Foulsham, Kingstone and Henrich2013; Yu et al., Reference Yu, Greer, Halevy and Van Bunderen2019). Further, the same position in the hierarchy may entail different duties and responsibilities in different institutional contexts (e.g., a CEO in a private company, a public company, and a nonprofit organization). The fact that similar institutions often give rise to very different normative systems (i.e., organizational cultures) is evident also from the ubiquity of cultural clashes following organizational mergers and acquisitions (Weber & Camerer, Reference Weber and Camerer2003). Hence, knowledge of the institutional context and individuals’ roles may be insufficient to predict others’ actions.
How theories of mind help calibrate predictions derived from institutional theories
Our simple solution extends Figure 1c in the target article. We propose that people regularly integrate institutional and mental state representations in the same way they enrich nouns with adjectives. Because role representations are insufficient to form accurate predictions about others’ behavior, people supplement roles with descriptors that capture others’ mental states. Parents to multiple children, or instructors who teach multiple students, know that the roles “child” and “student” are insufficient to describe how different individuals will behave in the same role. This is why we have numerous adjectives that add meaningful nuance to social roles: a helpful child, a curious student, a messy roommate, a nasty neighbor, a kind boss. In all these cases, the mentalistic stance meaningfully clarifies what to expect from the role holder.
Importantly, the adjective does not have to represent an enduring personality trait. It can capture a transient state that helps calibrate expectations. Consider the role of a barista, mentioned repeatedly in the target article. Keeping the role holder (barista) and institution (café) constant, we may form different expectations depending on whether the barista seems overworked and tired or well-rested and energized on a given shift. Their expectations concerning our tipping behavior may likewise depend on their perceptions of our mood that day.
In sum, social prediction requires more than knowledge of institutional structure and people’s roles in it. It requires mental state inferences. Trust between two romantic partners or business partners cannot be based only on roles and norms; it must incorporate beliefs about others’ feelings, intentions, and capabilities. The same logic applies beyond dyads and small teams. The large-scale coordination and cooperation that are necessary to provide public goods and solve collective action problems need more than a theory of institutions; they require beliefs about others’ mental states. Hence, important as the institutional stance is for forming expectations, understanding how people enrich institutional models with mental state inferences is essential for a complete and coherent theory of social prediction.
Financial support
This research received no specific grant from any funding agency, commercial, or not-for-profit sectors.
Competing interests
None.