What do we mean when we call an association 'my group'? One person values the friends made there, another what it represents, and a third feels that its fate forms part of their own story. The same words can describe different bonds.
Belonging is not a single feeling
People often say 'we are a group' as if the bond were obvious. Yet the same sentence can refer to affection for particular companions, identification with a category or a sense that the group's fate is woven into one's own life. Jack W. Klein, William J. Bingley, Matthew J. Hornsey and William B. Swann Jr. review three research traditions that have each explained part of this terrain. Their aim is to connect them without pretending that they measured the same thing all along.
This matters because the bond can predict different behavior. A person may keep volunteering because friends would miss them. Another may defend the same association because it stands for a valued identity. A third may experience a threat to it as personal. Describing all three as 'high attachment' loses the path from connection to action. It also obscures why a group of strangers can act collectively while people with warm personal ties may not share a political identity.
The review is theoretical and integrative. It draws on evidence from group experiments, social networks, identity research and studies of unusually strong group commitment, then proposes a framework that separates kinds of ties and their relationship to the self. These distinctions allow researchers to ask about a person's bond with a group. Testing them requires studying relationships and behavior; the theoretical categories alone cannot classify an actual member.
Cohesion begins with relationships among people
The oldest tradition here, social cohesion, begins with the feelings and interactions among members. Repeated contact, attraction and cooperation can hold a group together. A small work team may be connected because members know each other's habits, trust one another and expect mutual help. From this view the group is not a mysterious entity above its members; it is a pattern of ties that can be observed as a network.
That insight remains valuable. One can ask whether ties are dense or depend on a few central people, whether all members are recognized as individuals, or whether a move to another building weakens everyday contact. The review also notes that relationships have different forms: a family, task group and loose association are not organized by identical expectations. Treating every group as a network of interchangeable friendships would flatten this variety.
Cohesion alone, however, cannot explain every collective bond. People may favor members of a category they have never met. A stranger at a demonstration may help another because they recognize a shared cause, not because they have formed a personal friendship. That puzzle helped launch the second tradition. The authors retain cohesion's account of interpersonal ties while asking where it stops.
A shared category connects strangers
The social identity approach moved attention from affection between individuals to the recognition of common group membership. Classic minimal-group experiments assigned participants to arbitrary categories, sometimes by chance, without any face-to-face history. Even then, participants could favor their assigned group. The result was striking because familiar explanations based on long friendship or mutual benefit were insufficient. The category itself could become psychologically meaningful in the situation.
This does not say that any label creates durable solidarity. The context must make a category relevant, and norms about what 'we' do shape action. The social identity tradition is particularly powerful where behavior depends on relations between groups rather than the unique qualities of each individual. It can explain why people who do not know one another coordinate around a shared identity or why a boundary between 'us' and 'them' becomes salient.
The price of this emphasis, the review argues, can be insufficient attention to enduring personal commitments. If an identity matters only when the immediate setting activates it, how do we explain a person who protects a group across years and situations? Some later social identity work recognizes stable elements of identification. The integrative framework makes room for both the situational cue and the more persistent self.
Fusion brings the enduring self into the group
Identity fusion theory was developed partly to understand exceptionally strong commitment, including willingness to make costly sacrifices for a group. It emphasizes a firmly held sense of self that is aligned with the collective without simply disappearing into it. The review describes studies in which people strongly fused with their country responded to challenges to their personal self-concept by increasing expressed willingness to fight or die for the country. Such results distinguish fusion from a merely situational label.
The authors also discuss a sense of reciprocal strength: people feel supported by the group and motivated to strengthen it in return. Close ties to fellow members can develop alongside this alignment. Fusion is therefore not simply 'more' of the same identification, just as friendship is not a lower score on a fusion scale. The source and organization of the bond differ.
Care is needed when interpreting the dramatic outcomes used in some research. A questionnaire about willingness to sacrifice is not an observed violent act, and strong bonds can support prosocial as well as harmful behavior. The review asks which phenomena each theory is best equipped to explain; it does not brand highly committed members as dangerous.
The framework separates ties and time
To bring the traditions together, Klein and colleagues distinguish interpersonal ties from categorical ties. The first connect a person with specific others; the second connect them through membership in a recognized group. They also ask whether the relevant part of the self is mainly activated by the immediate situation or remains important across situations. These dimensions create a more useful map than a single question about attachment.
A friendship network sits naturally in one region of the map; a minimal group assigned by chance sits elsewhere. A person whose group identity is central to their life occupies another space. These are illustrations of theoretical emphases, not sealed boxes. In a neighborhood association, personal friendships, a shared civic category and deeply held identity may coexist and reinforce each other. The framework helps ask which is doing explanatory work in a particular event.
It also reconciles a debate about causality. Some cohesion accounts begin with liking individuals and forming a group. Social identity accounts can run the arrow in the other direction: recognizing a common category may generate attraction to fellow members. Neither sequence has to cover every case. By specifying the kind of tie and setting, the review turns an apparent contradiction into a question that can be studied.
A network of identities, not one label
The authors' proposed network model describes the self-concept as a collection of representations and ties rather than a single group label. A person can be a parent, neighbor, professional and member of a movement. Some connections among these parts are strong and enduring; others become active in a particular encounter. A threat or opportunity may spread across the network differently depending on how a group connects with other valued parts of the self.
This model is an attempt to state how the three traditions might work together. It can accommodate personal attraction, category-based identification and the alignment of personal and group selves. It also suggests why one person may react intensely to a group event while another, who reports the same nominal membership, barely notices it. The number of memberships alone cannot tell us the strength or structure of their connections.
The model remains a proposal with empirical work ahead. A persuasive test would have to measure different ties and self-concept connections over time, then show which patterns predict behavior beyond a broad identification score. It should also examine settings where a group changes, splits or comes into conflict with another identity. The review offers a research agenda, not a complete map of every human affiliation.
The map must travel beyond familiar settings
The authors acknowledge questions that their synthesis leaves open. Culture can change what counts as a group and which bonds people notice first. In some settings, a group may be understood mainly through relational networks rather than a sharp boundary between categories. Norms also matter: feeling connected does not dictate one fixed action if a group's expectations change. The three approaches have not been tested equally across societies and kinds of group.
The strength of belonging does not by itself explain how it works. We also need to ask who connects them, what the group represents, how stable that connection is and what the present situation activates. That richer account can explain why joining, staying, helping, protesting or leaving may follow different paths, even when people use the same word 'we'.
Understanding how a group forms part of someone's self-concept requires multiple measures and observation of behavior. The reported intensity of belonging cannot reveal all the ties sustaining it. The framework distinguishes those relationships to examine how they combine: friendship with other members, identification with the group and a deeply personal commitment may coexist, but should not be treated as interchangeable.
What to keep
- Personal ties and category-based ties are distinct.
- Enduring identity and current situation both matter.
- A network model links, rather than erases, the three research traditions.
The nuance
What this story does not let us claim
The proposed network model integrates theories; it does not reduce every group bond to one intensity scale.
Go deeper: concepts and optional reading practice
Reversible cards
Test the concept
Topic assessment
Check whether you can explain the idea, not merely recognize it.
Every answer is grounded in the explanation above. Some questions introduce new cases to test whether you can apply and analyze the ideas, not merely recall them.
Source & scope
Toward a Unified Theory of Group Connection. Jack W. Klein, William J. Bingley, Matthew J. Hornsey & William B. Swann, Jr.. Annual Review of Psychology, vol. 78 (2027). First published as a Review in Advance on September 29, 2026.
This editorial guide is an independent explanation. Consult the original publication for the studies, references and full context.
Open the original review ↗