01
Lineage and distinctive method
Irvin Yalom's account of group psychotherapy emphasized cohesion, universality, interpersonal learning, hope and the corrective recapitulation of relational patterns. Other group traditions developed independently around education, behavior change, mutual aid and analytic process. The distinctive opportunity is multiplicity: feedback comes from more than one relationship, change can be witnessed, and the patient can both receive and offer something.
That opportunity is not automatic. Composition, screening, leadership, norms, culture and the carried protocol determine whether the social field becomes therapeutic, inert or harmful. A well-designed CBT group is evidence-based because it delivers CBT competently in a group, not because gathering people together produces a generic group effect.
02
What happens in the room
Some groups follow a clear curriculum: practice a skill, review homework, plan an experiment. Others use the relationships in the room as immediate material. A member may discover that their certainty that others are bored leads them to speak less, which leaves others unsure how to approach them. Feedback can make the loop visible without appointing the group as judge.
The leader protects conditions for learning. They regulate turn-taking, invite but do not compel participation, notice exclusion, address breaches and keep disagreement from becoming humiliation. Endings, missed sessions and silence are treated as meaningful without being moralized. Confidentiality is discussed honestly: the therapist can set expectations and explain limits, but cannot guarantee every member's behavior outside the room.
03
The Flow transition and choreography
The cross-cutting operator is relational; the protocol adds its own cognitive, behavioral, affective or attentional emphasis. Isolation and shame can form a closed attractor: “I am alone because I am defective; my aloneness proves the defect.” A group introduces multiple vectors at once—recognition, challenge, modeling, responsibility and care. That can decompress social state-space and make belonging possible without perfect performance.
Choreography begins with containment. Establish the frame, let cohesion develop, introduce enough difference for learning, repair when the field fractures, then help members carry new behavior outside. Too little challenge leaves the old system untouched; too much social intensity can amplify threat or reenact exclusion. The goal is not maximum openness. It is enough safety and friction for new relational trajectories to survive beyond the session.
04
Route-specific clinical uses
For depression, NICE includes several group-delivered options, including group CBT, behavioral activation and mindfulness-based programs, and meta-analytic evidence supports group psychotherapy. The exact intervention should always be named. In addiction, groups can provide skills, relapse planning, accountability and a recovery social world, but clinician-led psychotherapy, psychoeducation and mutual-help groups are different resources. For PTSD, groups can reduce isolation and teach regulation; NICE found the evidence for group trauma-focused CBT limited and did not treat it as equivalent to individual trauma-focused treatment. Group disclosure should never become informal exposure without consent and structure.
05
Alliance and integration
Alliance is distributed. Members form bonds with the leader, with one another and with the group as a whole. A strong leader–member relationship cannot compensate for scapegoating among members. Group treatment often integrates with individual therapy, medication, case management or medical addiction care. Coordination should prevent a participant from receiving contradictory risk plans or being asked to disclose private individual material to remain “authentic” in group.
06
Evidence and safety boundary
Outcome evidence is inseparable from the group protocol and population. Older meta-analyses combine heterogeneous approaches; average benefit cannot identify which format suits someone with acute suicidality, psychosis, intoxication, severe dissociation or interpersonal danger. Leaders must monitor contagion, coercion, harassment, romantic/financial boundary risks and destabilizing trauma detail. Peer support is valuable but not a replacement for emergency or medical care. Flow Hijacked's “corrective social attractor” is a useful image, not proof that every group creates belonging.