Social structure learning in human anterior insula

Neural signatures underlying the effect of social structure on empathy and altruistic behaviors

Humans inhabit complex social networks, monitoring social structures that encompass both direct and indirect relationships. However, previous research primarily focused on direct relationships, leaving the neural basis of how social structure influences socioemotional processes understudied. This study addressed this gap by investigating the neural pathways underlying the influence of social structure on empathy and altruistic behaviors. During fMRI scanning, participants viewed painful or non-painful stimulation to innocent strangers who shared preferences with targets who had either treated participants fairly or unfairly. Afterwards, participants rated the pain experienced by these innocents and shared money with other innocents. Participants showed reduced empathic and altruistic responses toward innocents resembling unfair (vs. fair) targets, accompanied by heightened activation in regions crucial for emotion regulation and mentalizing, such as the lateral and medial prefrontal cortex. Furthermore, whole-brain and local neural patterns in the anterior insula and premotor cortex robustly discriminated painful (but not non-painful) stimulation of different innocents, suggesting that social structure altered emotional and sensorimotor aspects of empathy. These alterations might be driven by top-down regulation, as indicated by heightened functional connectivity between the lateral prefrontal cortex and sensorimotor areas, as well as between the anterior insula and subgenual anterior cingulate cortex when witnessing the pain of innocents resembling fair (vs. unfair) targets. Together, our work is the first to uncover the neural underpinnings through which human empathy and altruistic behaviors are shaped by social structure beyond direct self-other relationships.

Unifying Principles of Generalization: Past, Present, and Future

Generalization, defined as applying limited experiences to novel situations, represents a cornerstone of human intelligence. Our review traces the evolution and continuity of psychological theories of generalization, from its origins in concept learning (categorizing stimuli) and function learning (learning continuous input-output relationships) to domains such as reinforcement learning and latent structure learning. Historically, there have been fierce debates between approaches based on rule-based mechanisms, which rely on explicit hypotheses about environmental structure, and approaches based on similarity-based mechanisms, which leverage comparisons to prior instances. Each approach has unique advantages: Rules support rapid knowledge transfer, while similarity is computationally simple and flexible. Today, these debates have culminated in the development of hybrid models grounded in Bayesian principles, effectively marrying the precision of rules with the flexibility of similarity. The ongoing success of hybrid models not only bridges past dichotomies but also underscores the importance of integrating both rules and similarity for a comprehensive understanding of human generalization.

Replay shapes abstract cognitive maps for efficient social navigation

To make adaptive social decisions, people must anticipate how information flows through their social network. While this requires knowledge of how people are connected, networks are too large to have first-hand experience with every possible route between individuals. How, then, are people able to accurately track information flow through social networks? Here we find that people immediately cache abstract knowledge about social network structure as they learn who is friends with whom, which enables the identification of efficient routes between remotely connected individuals. These cognitive maps of social networks, which are built while learning, are then reshaped through overnight rest. During these extended periods of rest, a replay-like mechanism helps to make these maps increasingly abstract, which privileges improvements in social navigation accuracy for the longest communication paths that span distinct communities within the network. Together, these findings provide mechanistic insight into the sophisticated mental representations humans use for social navigation.

Processing of social closeness in the human brain

Healthy social life requires relationships in different levels of personal closeness. Based on ethological, sociological, and psychological evidence, social networks have been divided into five layers, gradually increasing in size and decreasing in personal closeness. Is this division also reflected in brain processing of social networks? During functional MRI, 21 participants compared their personal closeness to different individuals. We examined the brain volume showing differential activation for varying layers of closeness and found that a disproportionately large portion of this volume (80%) exhibited preference for individuals closest to participants, while separate brain regions showed preference for all other layers. Moreover, this bipartition reflected cortical preference for different sizes of physical spaces, as well as distinct subsystems of the default mode network. Our results support a division of the neurocognitive processing of social networks into two patterns depending on personal closeness, reflecting the unique role intimately close individuals play in our social lives.

Normativity vs. uniqueness: effects of social relationship strength on neural representations of others

Understanding others involves inferring traits and intentions, a process complicated by our reliance on stereotypes and generalized information when we lack personal information. Yet, as relationships are formed, we shift toward nuanced and individualized perceptions of others. This study addresses how relationship strength influences the creation of unique or normative representations of others in key regions known to be involved in social cognition. Employing a round-robin interpersonal perception paradigm (N = 111, 20 groups of five to six people), we used functional magnetic resonance imaging to examine whether the strength of social relationships modulated the degree to which multivoxel patterns of activity that represented a specific other were similar to a normative average of all others in the study. Behaviorally, stronger social relationships were associated with more normative trait endorsements. Neural findings reveal that closer relationships lead to more unique representations in the medial prefrontal cortex and anterior insula, areas associated with mentalizing and person perception. Conversely, more generalized representations emerge in posterior regions like the posterior cingulate cortex, indicating a complex interplay between individuated and generalized processing of social information in the brain. These findings suggest that cortical regions typically associated with social cognition may compute different kinds of information when representing the distinctiveness of others.

Similarity Among Friends Serves as a Social Prior: The Assumption That "Birds of a Feather Flock Together" Shapes Social Decisions and Relationship Beliefs

Social interactions unfold within networks of relationships. How do beliefs about others' social ties shape-and how are they shaped by-expectations about how others will behave? Here, participants joined a fictive online game-playing community and interacted with its purported members, who varied in terms of their trustworthiness and apparent relationships with one another. Participants were less trusting of partners with untrustworthy friends, even after they consistently showed themselves to be trustworthy, and were less willing to engage with them in the future. To test whether people not only expect friends to behave similarly but also expect those who behave similarly to be friends, an incidental memory test was given. Participants were exceptionally likely to falsely remember similarly behaving partners as friends. Thus, people expect friendship to predict similar behavior and vice versa. These results suggest that knowledge of social networks and others' behavioral tendencies reciprocally interact to shape social thought and behavior.

Vulnerability or resilience? Examining trust asymmetry from the perspective of risk sources under descriptive versus experiential decision

Introduction

The investigation of trust vulnerability is important to the understanding of the potential mechanisms of trust formation and erosion. However, more effective exploration of trust vulnerability has been hindered due to the lack of consideration of risk sources and types of information.

Methods

This study extended the investigation of asymmetry to both social and natural risk under experiential and descriptive decisions. Using the trust game as the decision-making paradigm and money as the subject matter, the research employed experimental methods to examine how people perceive and make decisions after being positively and negatively affected by natural and social risks. A total of 286 college students were participated in our study. Study 1 (n = 138) and Study 2 (n = 148) explored asymmetry in experiential and descriptive decision separately.

Results

The findings indicated that when considering experiential information, negative information had a greater effect in reducing trust compared to the enhancing effect of positive information (t = -1.95, p = 0.050). Moreover, the study revealed that negative information had a stronger negative impact in the context of social risks rather than natural risks (t = -3.26, p = 0.002), suggesting that trust is vulnerable both internally and externally. Conversely, when considering descriptive information, the effect of both positive and negative information on trust was symmetrical, and the impact of negative information was less significant compared to that of natural risks, indicating that trust has a certain level of resilience (t = 2.25, p = 0.028).

Discussion

The study emphasizes the importance of refining risk sources and information characteristics in complex scenarios in order to improve understanding of trust enhancement and repair.

Structure learning principles of stereotype change

Why, when, and how do stereotypes change? This paper develops a computational account based on the principles of structure learning: stereotypes are governed by probabilistic beliefs about the assignment of individuals to groups. Two aspects of this account are particularly important. First, groups are flexibly constructed based on the distribution of traits across individuals; groups are not fixed, nor are they assumed to map on to categories we have to provide to the model. This allows the model to explain the phenomena of group discovery and subtyping, whereby deviant individuals are segregated from a group, thus protecting the group's stereotype. Second, groups are hierarchically structured, such that groups can be nested. This allows the model to explain the phenomenon of subgrouping, whereby a collection of deviant individuals is organized into a refinement of the superordinate group. The structure learning account also sheds light on several factors that determine stereotype change, including perceived group variability, individual typicality, cognitive load, and sample size.

The role of the prefrontal cortex in social interactions of animal models and the implications for autism spectrum disorder

Social interaction is a complex behavior which requires the individual to integrate various internal processes, such as social motivation, social recognition, salience, reward, and emotional state, as well as external cues informing the individual of others' behavior, emotional state and social rank. This complex phenotype is susceptible to disruption in humans affected by neurodevelopmental and psychiatric disorders, including autism spectrum disorder (ASD). Multiple pieces of convergent evidence collected from studies of humans and rodents suggest that the prefrontal cortex (PFC) plays a pivotal role in social interactions, serving as a hub for motivation, affiliation, empathy, and social hierarchy. Indeed, disruption of the PFC circuitry results in social behavior deficits symptomatic of ASD. Here, we review this evidence and describe various ethologically relevant social behavior tasks which could be employed with rodent models to study the role of the PFC in social interactions. We also discuss the evidence linking the PFC to pathologies associated with ASD. Finally, we address specific questions regarding mechanisms employed by the PFC circuitry that may result in atypical social interactions in rodent models, which future studies should address.

Distrust as a form of inequality

Navigating social hierarchies is a ubiquitous aspect of human life. Social status shapes our thoughts, feelings, and actions toward others in various ways. However, it remains unclear how trust is conferred within hierarchies and how status-related cues are used when resources are on the line. This research fills this knowledge gap by examining how ascribed, consensus-based status appearance, and perceived status appearance impact investment decisions for high- and low-status partners during a Trust Game. In a series of pre-registered experiments, we examined the degree to which participants trusted unfamiliar others with financial investments when the only available information about that person was their socioeconomic status (SES). In Study 1, SES was ascribed. Studies 2 and 3 conveyed SES with visual antecedents (clothing). Across all three experiments, participants trusted high SES partners more than low SES partners. In addition, subjective perceptions of status based on visual cues were a stronger predictor of trust than consensus-based status judgments. This work highlights a high status-trust bias for decisions where an individual's money is on the line. In addition, high-status trust bias may occur simply because of an individual's subjective assumptions about another's rank.

Formalising social representation to explain psychiatric symptoms

Recent work in social cognition has moved beyond a focus on how people process social rewards to examine how healthy people represent other agents and how this is altered in psychiatric disorders. However, formal modelling of social representation has not kept pace with these changes, impeding our understanding of how core aspects of social cognition function, and fail, in psychopathology. Here, we suggest that belief-based computational models provide a basis for an integrated sociocognitive approach to psychiatry, with the potential to address important but unexamined pathologies of social representation, such as maladaptive schemas and illusory social agents.

Incorporating social knowledge structures into computational models

To navigate social interactions successfully, humans need to continuously learn about the personality traits of other people (e.g., how helpful or aggressive is the other person?). However, formal models that capture the complexities of social learning processes are currently lacking. In this study, we specify and test potential strategies that humans can employ for learning about others. Standard Rescorla-Wagner (RW) learning models only capture parts of the learning process because they neglect inherent knowledge structures and omit previously acquired knowledge. We therefore formalize two social knowledge structures and implement them in hybrid RW models to test their usefulness across multiple social learning tasks. We name these concepts granularity (knowledge structures about personality traits that can be utilized at different levels of detail during learning) and reference points (previous knowledge formalized into representations of average people within a social group). In five behavioural experiments, results from model comparisons and statistical analyses indicate that participants efficiently combine the concepts of granularity and reference points-with the specific combinations in models depending on the people and traits that participants learned about. Overall, our experiments demonstrate that variants of RW algorithms, which incorporate social knowledge structures, describe crucial aspects of the dynamics at play when people interact with each other.

Learning whom to cooperate with: neurocomputational mechanisms for choosing cooperative partners

Cooperation is fundamental for survival and a functioning society. With substantial individual variability in cooperativeness, we must learn whom to cooperate with, and often make these decisions on behalf of others. Understanding how people learn about the cooperativeness of others, and the neurocomputational mechanisms supporting this learning, is therefore essential. During functional magnetic resonance imaging scanning, participants completed a novel cooperation-partner-choice task where they learned to choose between cooperative and uncooperative partners through trial-and-error both for themselves and vicariously for another person. Interestingly, when choosing for themselves, participants made faster and more exploitative choices than when choosing for another person. Activity in the ventral striatum preferentially responded to prediction errors (PEs) during self-learning, whereas activity in the perigenual anterior cingulate cortex (ACC) signaled both personal and vicarious PEs. Multivariate pattern analyses showed distinct coding of personal and vicarious choice-making and outcome processing in the temporoparietal junction (TPJ), dorsal ACC, and striatum. Moreover, in right TPJ the activity pattern that differentiated self and other outcomes was associated with individual differences in exploitation tendency. We reveal neurocomputational mechanisms supporting cooperative learning and show that this learning is reflected in trial-by-trial univariate signals and multivariate patterns that can distinguish personal and vicarious choices.

Latent structure learning as an alternative computation for group inference

In contrast to Pietraszewski's account, latent structure learning neither requires conflict nor relies on observation of explicit coalitional behavior to support group inference. This alternative addresses how even non-conflict-based groups may be defined and is supported by experimental evidence in human behavior.

Cytoarchitecture, probability maps and segregation of the human insula

The human insular cortex supports multifunctional integration including interoceptive, sensorimotor, cognitive and social-emotional processing. Different concepts of the underlying microstructure have been proposed over more than a century. However, a 3D map of the cytoarchitectonic segregation of the insula in standard reference space, that could be directly linked to neuroimaging experiments addressing different cognitive tasks, is not yet available. Here we analyzed the middle posterior and dorsal anterior insula with image analysis and a statistical mapping procedure to delineate cytoarchitectonic areas in ten human postmortem brains. 3D-probability maps of seven new areas with granular (Ig3, posterior), agranular (Ia1, posterior) and dysgranular (Id2-Id6, middle to dorsal anterior) cytoarchitecture have been calculated to represent the new areas in stereotaxic space. A hierarchical cluster analysis based on cytoarchitecture resulted in three distinct clusters in the superior posterior, inferior posterior and dorsal anterior insula, providing deeper insights into the structural organization of the insula. The maps are openly available to support future studies addressing relations between structure and function in the human insula.

Brain System for Social Categorization by Narrative Roles

The cognitive system applies categorical thinking to facilitate perception of the rich environment around us. In person cognition, research has focused on the roles of gender, race, age, or appearance in social categorical thinking. Here we investigated how narrative roles, as portrayed by different cinematic characters, are categorized in the neurocognitive system. Under functional MRI, 17 human participants (7 females) were asked to make different judgments regarding personality traits of 16 renowned cinematic characters representing four roles: hero, sidekick, mentor, and villain. Classification analysis showed a brain network, comprising the dorsal medial prefrontal cortex, the precuneus and the temporoparietal junction bilaterally, and the left occipital face area (OFA), to discriminate among the four roles. No such classification was found between other individual attributes including age or the associated film. Moreover, regions overlapping the default mode network (DMN) were found to better discriminate between roles, rather than the individual characters, while the OFA was found to better discriminate between individuals. These results demonstrate the inherent role of roles in person cognition, and suggest an intimate relation between roles categorization and self-referential activity.SIGNIFICANCE STATEMENT Social categorization, the assignment of different people in our social network to subgroups, is a powerful strategy in social cognition. How is this managed by the brain? We provide evidence that different characters from different stories, representing similar roles in their corresponding narrative, elicit similar brain activation patterns, as revealed by functional MRI. Unlike previous studies of social categorization, these brain activations were similar to those elicited by social cognition rather than face processing, and included regions at the prefrontal cortex, the precuneus, and the temporoparietal junction. The identified brain network significantly overlapped the default mode network. We suggest that social categorization by roles is fundamental to the cognitive system, relying on brain regions related to social cognition.

Neural Mechanisms of Intergroup Exclusion and Retaliatory Aggression

Aggression occurs frequently and severely between rival groups. Although there has been much study into the psychological and socio-ecological determinants of intergroup aggression, the neuroscience of this phenomenon remains incomplete. To examine the neural correlates of aggression directed at outgroup (versus ingroup) targets, we recruited 35 healthy young male participants who were current or former students of the same university. While undergoing functional MRI, participants completed an aggression task against both an ingroup and an outgroup opponent in which their opponents repeatedly provoked them at varying levels and then participants could retaliate. Participants were then socially included and then excluded by two outgroup members and then completed the same aggression task against the same two opponents. Both before and after outgroup exclusion, aggression towards outgroup members was positively associated with activity in the ventral striatum during decisions about how aggressive to be towards their outgroup opponent. Aggression towards outgroup members was also linked to greater post-exclusion activity in the rostral and dorsal medial prefrontal cortex during provocation from their outgroup opponent. These altered patterns of brain activity suggest that frontostriatal mechanisms may play a significant role in motivating aggression towards outgroup members.

Transforming Social Perspectives with Cognitive Maps

Growing evidence suggests that cognitive maps represent relations between social knowledge similar to how spatial locations are represented in an environment. Notably, the extant human medial temporal lobe literature assumes associations between social stimuli follow a linear associative mapping from an egocentric viewpoint to a cognitive map. Yet, this form of associative social memory doesn't account for a core phenomenon of social interactions in which social knowledge learned via comparisons to the self, other individuals, or social networks are assimilated within a single frame of reference. We argue that hippocampal-entorhinal coordinate transformations, known to integrate egocentric and allocentric spatial cues, inform social perspective switching between the self and others. We present evidence that the hippocampal formation helps inform social interactions by relating self versus other social attribute comparisons to society in general, which can afford rapid and flexible assimilation of knowledge about the relationship between the self and social networks of varying proximities. We conclude by discussing the ramifications of cognitive maps in aiding this social perspective transformation process in states of health and disease.

Neural substrates of shared visual experiences: a hyperscanning fMRI study

Sharing experience is a fundamental human social cognition. Since visual experience is a mental state directed toward the world, we hypothesized that sharing visual experience is mediated by joint attention (JA) for sharing directedness and mentalizing for mental state inferences. We conducted a hyperscanning functional magnetic resonance imaging with 44 healthy adult volunteers to test this hypothesis. We employed spoken-language-cued spatial and feature-based JA tasks. The initiator attracts the partner's attention by a verbal command to a spatial location or an object feature to which the responder directs their attention. Pair-specific inter-individual neural synchronization of task-specific activities was found in the right anterior insular cortex (AIC)-inferior frontal gyrus (IFG) complex, the core node of JA and salience network, and the right posterior superior temporal sulcus, which represents the shared categories of the target. The right AIC-IFG also showed inter-individual synchronization of the residual time-series data, along with the right temporoparietal junction and dorsomedial prefrontal cortex-the core components for mentalization and the default mode network (DMN). This background synchronization represents sharing the belief of sharing the situation. Thus, shared visual experiences are represented by coherent coordination between the DMN and salience network linked through the right AIC-IFG.

Social Relationship Strength Modulates the Similarity of Brain-to-Brain Representations of Group Members

Within our societies, humans form co-operative groups with diverse levels of relationship quality among individual group members. In establishing relationships with others, we use attitudes and beliefs about group members and the group as a whole to establish relationships with particular members of our social networks. However, we have yet to understand how brain responses to group members facilitate relationship quality between pairs of individuals. We address this here using a round-robin interpersonal perception paradigm in which each participant was both a perceiver and target for every other member of their group in a set of 20 unique groups of between 5 and 6 members in each (total N = 111). Using functional magnetic resonance imaging, we show that measures of social relationship strength modulate the brain-to-brain multivoxel similarity patterns between pairs of participants' responses when perceiving other members of their group in regions of the brain implicated in social cognition. These results provide evidence for a brain mechanism of social cognitive processes serving interpersonal relationship strength among group members.

The neural underpinnings of intergroup social cognition: an fMRI meta-analysis

Roughly 20 years of functional magnetic resonance imaging (fMRI) studies have investigated the neural correlates underlying engagement in social cognition (e.g. empathy and emotion perception) about targets spanning various social categories (e.g. race and gender). Yet, findings from individual studies remain mixed. In the present quantitative functional neuroimaging meta-analysis, we summarized across 50 fMRI studies of social cognition to identify consistent differences in neural activation as a function of whether the target of social cognition was an in-group or out-group member. We investigated if such differences varied according to a specific social category (i.e. race) and specific social cognitive processes (i.e. empathy and emotion perception). We found that social cognition about in-group members was more reliably related to activity in brain regions associated with mentalizing (e.g. dorsomedial prefrontal cortex), whereas social cognition about out-group members was more reliably related to activity in regions associated with exogenous attention and salience (e.g. anterior insula). These findings replicated for studies specifically focused on the social category of race, and we further found intergroup differences in neural activation during empathy and emotion perception tasks. These results help shed light on the neural mechanisms underlying social cognition across group lines.

Causal Inference Gates Corticostriatal Learning

Attributing outcomes to your own actions or to external causes is essential for appropriately learning which actions lead to reward and which actions do not. Our previous work showed that this type of credit assignment is best explained by a Bayesian reinforcement learning model which posits that beliefs about the causal structure of the environment modulate reward prediction errors (RPEs) during action value updating. In this study, we investigated the brain networks underlying reinforcement learning that are influenced by causal beliefs using functional magnetic resonance imaging while human participants (n = 31; 13 males, 18 females) completed a behavioral task that manipulated beliefs about causal structure. We found evidence that RPEs modulated by causal beliefs are represented in dorsal striatum, while standard (unmodulated) RPEs are represented in ventral striatum. Further analyses revealed that beliefs about causal structure are represented in anterior insula and inferior frontal gyrus. Finally, structural equation modeling revealed effective connectivity from anterior insula to dorsal striatum. Together, these results are consistent with a possible neural architecture in which causal beliefs in anterior insula are integrated with prediction error signals in dorsal striatum to update action values.SIGNIFICANCE STATEMENT Learning which actions lead to reward-a process known as reinforcement learning-is essential for survival. Inferring the causes of observed outcomes-a process known as causal inference-is crucial for appropriately assigning credit to one's own actions and restricting learning to effective action-outcome contingencies. Previous studies have linked reinforcement learning to the striatum, and causal inference to prefrontal regions, yet how these neural processes interact to guide adaptive behavior remains poorly understood. Here, we found evidence that causal beliefs represented in the prefrontal cortex modulate action value updating in posterior striatum, separately from the unmodulated action value update in ventral striatum posited by standard reinforcement learning models.

Towards a computational theory of social groups: A finite set of cognitive primitives for representing any and all social groups in the context of conflict

We don't yet have adequate theories of what the human mind is representing when it represents a social group. Worse still, many people think we do. This mistaken belief is a consequence of the state of play: Until now, researchers have relied on their own intuitions to link up the concept social group on the one hand, and the results of particular studies or models on the other. While necessary, this reliance on intuition has been purchased at considerable cost. When looked at soberly, existing theories of social groups are either (i) literal, but not remotely adequate (such as models built atop economic games), or (ii) simply metaphorical (typically a subsumption or containment metaphor). Intuition is filling in the gaps of an explicit theory. This paper presents a computational theory of what, literally, a group representation is in the context of conflict: it is the assignment of agents to specific roles within a small number of triadic interaction types. This "mental definition" of a group paves the way for a computational theory of social groups-in that it provides a theory of what exactly the information-processing problem of representing and reasoning about a group is. For psychologists, this paper offers a different way to conceptualize and study groups, and suggests that a non-tautological definition of a social group is possible. For cognitive scientists, this paper provides a computational benchmark against which natural and artificial intelligences can be held.

Human Representation Learning

The central theme of this review is the dynamic interaction between information selection and learning. We pose a fundamental question about this interaction: How do we learn what features of our experiences are worth learning about? In humans, this process depends on attention and memory, two cognitive functions that together constrain representations of the world to features that are relevant for goal attainment. Recent evidence suggests that the representations shaped by attention and memory are themselves inferred from experience with each task. We review this evidence and place it in the context of work that has explicitly characterized representation learning as statistical inference. We discuss how inference can be scaled to real-world decisions by approximating beliefs based on a small number of experiences. Finally, we highlight some implications of this inference process for human decision-making in social environments.

The neural basis of ideological differences in race categorization

Multiracial individuals are often categorized as members of their 'socially subordinate' racial group-a form of social discrimination termed hypodescent-with political conservatives more likely than liberals to show this bias. Although hypodescent has been linked to racial hierarchy preservation motives, it remains unclear how political ideology influences categorization: Do conservatives and liberals see, feel or think about mixed-race faces differently? Do they differ in sensitivity to Black prototypicality (i.e. skin tone darkness and Afrocentric features) or racial ambiguity (i.e. categorization difficulty) of Black/White mixed-race faces? To help answer these questions, we collected a politically diverse sample of White participants and had them categorize mixed-race faces as Black or White during functional neuroimaging. We found that conservatism was related to greater anterior insula activity to racially ambiguous faces, and this pattern of brain activation mediated conservatives' use of hypodescent. This demonstrates that conservatives' greater sensitivity to racial ambiguity (rather than Black prototypicality) gives rise to greater categorization of mixed-race individuals into the socially subordinate group and tentatively suggests that conservatives may differ from liberals in their affective reactions to mixed-race faces. Implications for the study of race categorization and political psychology are discussed. This article is part of the theme issue 'The political brain: neurocognitive and computational mechanisms'.

Reframing social categorization as latent structure learning for understanding political behaviour

Affiliating with political parties, voting and building coalitions all contribute to the functioning of our political systems. One core component of this is social categorization-being able to recognize others as fellow in-group members or members of the out-group. Without this capacity, we would be unable to coordinate with in-group members or avoid out-group members. Past research in social psychology and cognitive neuroscience examining social categorization has suggested that one way to identify in-group members may be to directly compute the similarity between oneself and the target (dyadic similarity). This model, however, does not account for the fact that the group membership brought to bear is context-dependent. This review argues that a more comprehensive understanding of how we build representations of social categories (and the subsequent impact on our behaviours) must first expand our conceptualization of social categorization beyond simple dyadic similarity. Furthermore, a generalizable account of social categorization must also provide domain-general, quantitative predictions for us to test hypotheses about social categorization. Here, we introduce an alternative model-one in which we infer latent groups of people through latent structure learning. We examine experimental evidence for this account and discuss potential implications for understanding the political mind. This article is part of the theme issue 'The political brain: neurocognitive and computational mechanisms'.

The Social Neuroscience of Prejudice

The social neuroscience approach to prejudice investigates the psychology of intergroup bias by integrating models and methods of neuroscience with the social psychology of prejudice, stereotyping, and discrimination. Here, we review major contemporary lines of inquiry, including current accounts of group-based categorization; formation and updating of prejudice and stereotypes; effects of prejudice on perception, emotion, and decision making; and the self-regulation of prejudice. In each section, we discuss key social neuroscience findings, consider interpretational challenges and connections with the behavioral literature, and highlight how they advance psychological theories of prejudice. We conclude by discussing the next-generation questions that will continue to guide the social neuroscience approach toward addressing major societal issues of prejudice and discrimination.

Social-Structure Learning

Social-structure learning is the process by which social groups are identified on the basis of experience. Building on models of structure learning in other domains, we formalize this problem within a Bayesian framework. According to this framework, the probabilistic assignment of individuals to groups is computed by combining information about individuals with prior beliefs about group structure. Experiments with adults and children provide support for this framework, ruling out alternative accounts based on dyadic similarity. More broadly, we highlight the implications of social-structure learning for intergroup cognition, stereotype updating, and coalition formation.

Group Cooperation, Carrying-Capacity Stress, and Intergroup Conflict

Peaceful intergroup relations deteriorate when individuals engage in parochial cooperation and parochial competition. To understand when and why intergroup relations change from peaceful to violent, we present a theoretical framework mapping out the different interdependence structures between groups. According to this framework, cooperation can lead to group expansion and ultimately to carrying-capacity stress. In such cases of endogenously created carrying-capacity stress, intergroup relations are more likely to become negatively interdependent, and parochial competition can emerge as a response. We discuss the cognitive, neural, and hormonal building blocks of parochial cooperation, and conclude that conflict between groups can be the inadvertent consequence of human preparedness - biological and cultural - to solve cooperation problems within groups.

How Does the Brain Infer Hidden Social Structures?

Many everyday thoughts and actions are shaped not only by our direct relationships with others, but also by our knowledge of relations between third-parties. Lau et al. recently demonstrated how knowledge of one type of social relation - interpersonal similarity - shapes cognition and behavior, and shed light on the neural basis of such phenomena.

Social structure learning in human anterior insula

Humans form social coalitions in every society, yet we know little about how we learn and represent social group boundaries. Here we derive predictions from a computational model of latent structure learning to move beyond explicit category labels and interpersonal, or dyadic, similarity as the sole inputs to social group representations. Using a model-based analysis of functional neuroimaging data, we find that separate areas correlate with dyadic similarity and latent structure learning. Trial-by-trial estimates of ‘allyship’ based on dyadic similarity between participants and each agent recruited medial prefrontal cortex/pregenual anterior cingulate (pgACC). Latent social group structure-based allyship estimates, in contrast, recruited right anterior insula (rAI). Variability in the brain signal from rAI improved prediction of variability in ally-choice behavior, whereas variability from the pgACC did not. These results provide novel insights into the psychological and neural mechanisms by which people learn to distinguish ‘us’ from ‘them.’

In every society, people form social coalitions — we draw boundaries between 'us' and 'them'. But how do we decide who is one of 'us' and who is one of 'them'? One way is to use arbitrary categories. For example, we say that those living 49 degrees north of the Earth’s equator are Canadian, whereas those living south of it are American. Another possibility is to use physical characteristics. But what about when neither of these options are available?

By monitoring brain activity in healthy volunteers learning about other people’s political values, Lau et al. obtained insights into how people make these decisions. Participants lying in a brain scanner were asked to report their position on a political issue. They then learned the positions of three other hypothetical participants – A, B and C – on the same issue. After repeating this procedure for eight different issues, the volunteers had to decide whether they would align with A or with B on a 'mystery' political issue.

So how do participants choose between A and B? One possibility is that they simply choose whichever one has views most similar to their own. If this is the case, the views of hypothetical person C should not affect their decision. But in practice, C's views – specifically how much they resemble the volunteer's own – do influence whether the volunteer chooses A or B. This suggests that we choose our allies based on more than just their similarity to ourselves.

Using a mathematical model, Lau et al. show that volunteers also take into account how similar the views of the other ‘participants’ are to each other. In other words, they consider the structure of the social group as a whole. Moreover, the results from brain imaging show that different regions of the brain are active when volunteers track the structure of the entire group, as opposed to their own similarity with each individual.

Notably though, the activity of the group-tracking region explains people's alignment choices better than the activity of the similarity-tracking region. This suggests that we base our judgments of 'us' versus 'them' more on the structure of the group as a whole than on our own similarity with individual group members. Understanding how we determine whether others are on the same ‘team’ as ourselves could ultimately help us find ways to reduce bias and discrimination between groups.