Tavistock group relations

1. Introduction

Version followed: the Tavistock Institute's Leicester Conference model as currently published (Leicester Conference 2019/2025 materials: Small Study Group, Large Study Group, Inter-Group Event, Whole System Event, Embodying Role Event, Review and Application Groups, plenaries), with the A.K. Rice Institute (AKRI, USA) variants where they add procedural detail (GREX 2012 design, De Loach's Institutional System Event, the AKRI/GRI consultant competencies). Bion's 1940s–50s group work is the theoretical origin, not a procedure followed here.

What it is. Tavistock group relations is not a psychotherapy. It is an experiential educational method, delivered as residential or non-residential "conferences" (a weekend to two weeks), in which members study authority, leadership, role and organisation by examining their own behaviour in a temporary institution as it happens. Its sources describe the events as "conferences, workshops, or educational events" rather than therapy, and current AKRI material states plainly that the work is not a substitute for mental health care. [source: Tavistock Institute, What is a group relations conference; Gross 2015 SAGE entry; AKRI LDRS 2024]

History in brief. Wilfred Bion's post-war groups at the Tavistock Clinic, collected in Experiences in Groups (1961), supplied the core theory (work group vs. basic assumptions). Before that, Bion and John Rickman ran the first Northfield experiment, a short-lived experiment in patient self-organisation at a military psychiatric hospital that the army closed after six weeks. The Tavistock Institute and the University of Leicester ran the first two-week Leicester Conference in 1957, which Margaret Rioch traces to Bion, Lewin and Bethel. The early conferences put small-group study first; A.K. Rice later moved the focus to authority and to the institution as a whole. Rice took the model to the USA (Mount Holyoke, 1965), where the A.K. Rice Institute still runs it. Rioch writes that the conferences "would scarcely have found their form without the ideas of Bion". [source: Rioch 1975, Group Relations: Rationale and Technique; Rioch 1983 via Whitwell; Gross 2015; Endnotes 2020; Tavistock Institute] Rice's Learning for Leadership (1965), the founding account of the conference design, has chapters on conference structure and culture, study groups, the large-group event, the intergroup exercise, application groups, the plenary sessions, the staff group, and the line between training and therapy. [source: Routledge, Learning for Leadership table of contents] Its detailed procedures are used here only through later accounts, and E.J. Miller's later accounts of the Leicester model are not used [uncertain].

Core idea of change. Every group works at two levels at once. One is the work group, the conscious and cooperative effort on the group's task, which Bion called "in however embryonic a form, scientific". The other is basic-assumption mentality: dependency, fight-flight and pairing. This is shared, unconscious and "instantaneous, inevitable and instinctive", and it pulls the group off its task as a defence against anxiety. In a conference the task, role, time and place are held constant. Consultants then name the covert processes that emerge as group-level interpretations or working hypotheses. Members learn by experience how they take up, give away, resist and are drawn into authority and roles, and they then apply this to their own work roles. [source: Endnotes 2020 quoting Bion 1961 p.143; Emery quoting Bion 1952 p.235; GREX 2012; Tavistock Institute LC19]

How a psychotherapist uses it. There are three routes: (a) attend as a member, the most common route and the only one open without further training; (b) train towards consulting or staffing conferences through AKRI, the Tavistock Institute or a sister organisation; (c) carry the consultant stance and the basic-assumption lens into therapy groups, staff teams, reflective-practice groups and organisational consultation. Route (c) is [synthesis] applied to clinical settings: the conference sources describe it only as "application" to members' home roles.

2. Framework

The framework is the conference itself, a temporary institution with a stated primary task. Leicester puts it as studying "the development and exercise of authority, role and organisation". [source: Tavistock Institute LC19]

Staff structure. A Director is responsible for the design and its delivery, oversees the consultants and holds the structure. Consultants ("consulting staff") observe and interpret and "do not lead in a traditional sense". Members are "the main actors", and their experience is the primary data. An administrator manages the practical boundaries. In institution-type events a Management group (Directorate plus event convenors, usually 6–8 people) works in public, apart from the consultants. [source: Tavistock Institute; De Loach 2014; AKRI LDRS 2024]

Phases (sequence as published; transitions are by the timetable, not by readiness):

Pre-conference. The applicant reads the brochure (primary task, events, staff, fees) and applies. Members are told to expect emotionally engaging, non-didactic work with "no formal lectures or predefined outcomes". No screening procedure is described in the sources [uncertain]. [source: Tavistock Institute; LC19]
Opening plenary. The whole conference meets to mark entry across the conference boundary. The Director states the task and the design. [source: LC19; GREX 2012]
Here-and-now study events (the bulk of the conference). These are SSG, LSG, Inter-Group and Institutional or Whole System events, plus the Embodying Role Event at Leicester and social dreaming at some conferences. They run in parallel strands across the days. In De Loach's design the institutional event starts after about a third of the conference events and takes about 26% of event hours. [source: LC19; Tavistock Institute; De Loach 2014]
Review and application. The RAG or RAAG groups meet throughout, usually daily. They begin by reviewing experience and, towards the end, move to planning how to apply the learning back home. [source: LC19; GREX 2012]
Closing plenary. This marks exit from the conference and its events. Some designs also have separate opening and closing plenaries for the institutional event. [source: LC19; GREX 2012]
Post-conference. GREX offered two follow-up review sessions at about two and four weeks, plus individual meetings, to help integrate the learning into work roles. The Tavistock Institute describes learning that continues after the event. [source: GREX 2012; LC19]

Typical sizes and durations:

Whole conference: a weekend up to 14 days. Leicester runs 14 days. Institutions of 50–70 members are typical in the Rioch era. [source: Tavistock Institute; LC19; Rioch 1975]
Small Study Group: up to 12 members at Leicester (maximum 10 at GREX) with one consultant. [source: LC19; GREX 2012]
Large Study Group: the whole membership, with 2–4 consultants (3 at GREX). [source: LC19; GREX 2012]
RAAG: 4–6 members and one consultant, 75 minutes, one per day at GREX. [source: GREX 2012]
Institutional System Event (weekend example): five sessions of about 60–75 minutes. [source: De Loach 2014]
Sessions begin and end promptly at the stated times. [source: GREX 2012]

Criteria for moving between phases. There are none apart from the clock: events start and end on time whatever is happening, and holding time as a fixed boundary is itself part of the method. [source: GREX 2012] Members are invited to take up their own authority to accept or reject the learning, so no competence criterion has to be met. [source: LC19]

Training and authorisation to staff. AKRI and the Global Relations Institute (GRI) publish a set of 26 foundational consultant competencies (2026), and AKRI runs a developmental learning programme (LDRS) for people with prior conference experience. The LDRS states that it does not fully prepare people to consult and that certification is still being developed. Formal certification steps and the required number of conferences or staff roles are not stated in the sources [uncertain]. In practice, directors invite staff. Treat that as [uncertain] general knowledge, not a sourced rule. [source: AKRI/GRI Consultant Competencies; AKRI LDRS 2024]

3. Guiding principles

Primary task. Every event has a stated task that defines what the work is. Practical: open each event by naming its task, and judge every intervention by whether it serves that task. [source: LC19; AKRI competencies; Rioch 1975]
Here and now. The data is what happens in the room as it happens, not reports of outside life. Practical: when members bring outside stories, ask what telling them here and now does for the group. [source: Tavistock Institute; GREX 2012; AKRI LDRS]
Group as a whole. The group, not the individual, is the unit of study. Practical: address interventions to "the group" or a subsystem, even when one person spoke. [source: Gross 2015; AKRI competencies ("systemic speech")]
Work group and basic assumptions co-exist. One dominates at a time, and they can oscillate fast. Practical: hold a hypothesis about which mentality is active, and expect it to shift within one session. [source: Endnotes 2020; French & Simpson summary]
The individual as carrier (valency). A member who acts something out, for example as scapegoat or rescuer, is voicing a group-wide experience. Bion's "valency" is a person's readiness to combine in basic-assumption activity. Practical: ask why the group needs this person to carry this now. [source: AKRI competencies; Emery citing Bion 1961 p.116]
Learning from experience; no lectures. Understanding comes from studying one's own experience, not from being taught. Practical: do not explain theory inside study events. Keep didactic input to review, application and plenary slots (GREX had about 8 hours there). [source: Tavistock Institute; GREX 2012]
Authority is taken up, given and resisted. AKRI defines authority as "permission to work on behalf of others". Practical: track who is authorised to speak or act, by whom, and how the group withdraws authority. [source: AKRI LDRS 2024; Tavistock Institute]
Role, not identity, is the unit of change. Roles are "behavioral commitments" taken up so that a group can pursue its purpose. Practical: in review groups, ask "what role were you in?" rather than "what kind of person are you?". [source: AKRI LDRS 2024]
Boundaries of task, time, territory and role are held constant. Holding them lets the other dynamics show. Practical: start and end exactly on time, stay in your allotted space and role, and treat crossings as data. [source: GREX 2012; Tavistock Institute; De Loach 2014]
Consultant as screen. The consultant stays relatively aloof so that projections onto authority become visible. Practical: minimise social engagement and self-disclosure in role, and do not reassure. [source: Gross 2015]
Use of one's own experience (self-containment). The consultant's feelings are data, but must be sorted from what belongs to them personally. Practical: before you speak, ask "is this mine, the group's, or the system's?". Voice your internal experience rather than act on it. [source: AKRI competencies]
Interpret what the group does to the consultant. Rice summarised Bion as "constantly trying to understand what the group is doing, particularly what it is doing to him". Practical: the group's treatment of you (ignoring, idealising, attacking) is a prime focus for interpretation. [source: Rice quoted in Rioch 1983]
Working hypotheses, not truths. Interpretations are offered as testable hypotheses, ideally with a "because" clause and supporting data. Practical: phrase interventions as "my hypothesis is… because…". Members are free to test or reject them. [source: De Loach 2014; AKRI LDRS]
Hold to the task under pressure. Staff keep to the stated purpose "no matter how difficult it may become", while also questioning their own activities. Bion's advice was to "think under fire". Practical: do not abandon the role to placate, and review staff conduct openly in staff meetings. [source: Rioch 1975; Endnotes 2020]
Rigorous honesty. The method values honesty about oneself and one's group, with no promise that it will resolve conflict. Practical: name the undiscussable even when it is uncomfortable. [source: Rioch 1975; AKRI LDRS]
Members define their own meanings. Staff do not prescribe what authority or leadership means. Practical: avoid normative feedback ("good leadership is…"). [source: Rioch 1975]
Identity, representation and socio-political context. Race, gender, age and wider social forces shape authority in the room, and the presence or absence of representation affects learning. Practical: include identity and power in your hypotheses, and stay aware of the projections your own identity attracts. [source: AKRI competencies; GREX 2012]
Management in public. In institutional events the staff do their managing openly and give their allegiance to the whole system. Practical: make no private deals with subgroups, and share information instead of withholding it. [source: De Loach 2014]
Regaining role. Losing the role is expected, and the work is to notice it, own it and take the role up again. Practical: if you slip into chatting, defending or rescuing, name it and return to the task. [source: AKRI competencies]
Dignity and restoration. Current AKRI material stresses dignity and supports restorative processes when offence is caused. [source: AKRI LDRS 2024]

4. Intervention inventory

The inventory has three parts: A, the conference events, which are the method's "interventions" at the level of design; B, the consultant's in-role techniques; C, clinical adaptations for psychotherapists, which are all [synthesis]. Every example of therapist or consultant language is my own wording, labelled "(illustrative)". Sources do not give scripts.

A. Conference events

A1. Opening plenary

Purpose / when: At the start of the conference, and of the institutional event where it has its own plenary. It marks crossing the boundary into the temporary institution. [source: LC19; GREX 2012]
Steps:
1. The whole staff and membership meet.
2. The Director states the primary task, the events, the staff roles and the time and territory boundaries.
3. The Director takes questions about the conference and its parts (GREX).
4. In De Loach's institutional event opening, the Director is the only manager present. Two consultants arrange their own chairs. The Director explains the ways members can represent their groups when they cross into another subsystem: silent observer, delegate with a single function, or plenipotentiary. [source: De Loach 2014]
5. Start and finish on time.
Example language (illustrative): "The task of this conference is to study authority, role and organisation as they arise here, in the relationships we make during the conference. Staff will offer working hypotheses. It is up to each of you what you learn."
Watch for: Early dependency on the Director for rules and reassurance, and early fight about the design. Both are first data, not problems to fix.

A2. Small Study Group (SSG)

Purpose / when: The core event, held repeatedly throughout. Its task is to study the group's own behaviour, and how leadership and followership arise in it, in the here and now. [source: LC19; GREX 2012]
Steps:
1. Up to 12 members (10 at GREX) and one consultant meet in a fixed room for fixed sessions.
2. The consultant arrives on time, takes a seat and gives no agenda and no instructions beyond the task. [source for task and size: LC19; arriving and seating are [synthesis] from standard practice]
3. The consultant listens to the group as a whole and offers interpretations of group-level dynamics when they judge it will help learning: basic assumptions, the group's relation to the consultant, and what individuals carry for the group. [source: GREX 2012; Gross 2015; AKRI competencies]
4. The consultant ends exactly on time, even in mid-sentence.
Example language (illustrative): "The group seems to be waiting for me to tell it what to do, as if the work can only start once I take charge." "Perhaps the group needs Ana to stay the outsider, so that nobody else has to feel outside."
Watch for: Long silences, attempts to turn the session into introductions or a social chat, pressure on the consultant to lead, and pairs forming with hopeful excitement. Signs that it is working: members start to test hypotheses themselves and to link what happens here to their own pattern with authority.

A3. Large Study Group (LSG)

Purpose / when: The whole membership meets with 2–4 consultants to study what happens when face-to-face relating is no longer possible: cohesion and fragmentation, myth-making and reality-testing. [source: LC19; GREX 2012]
Steps:
1. Seating is arranged for the whole membership. Spiral or scattered layouts are common, but this is [uncertain]: the published materials do not specify it.
2. Consultants place themselves among members, work as a team and make group-level interventions about the whole.
3. The group meets on fixed times across the conference.
Example language (illustrative): "Voices from one corner are heard, and the rest of the room has gone silent. It may be that the large group is managing its fear of being swamped by leaving a few people to speak for everyone."
Watch for: Anonymity, loss of individual thinking, the rapid spread of rumour or myth, and splitting into factions. It is working when members notice how they lose and recover the ability to think in a crowd.

A4. Inter-Group Event (IGE)

Purpose / when: Members form groups that relate to one another through a common task. The aim is to learn how representative roles are taken up and how authority is negotiated and carried on behalf of others. At Leicester it includes action learning. [source: LC19; Tavistock Institute]
Steps:
1. A plenary opens the event.
2. Members form their own groups and choose territories.
3. Groups send representatives with defined levels of authority to other groups.
4. Consultants are available to groups.
5. A plenary closes the event. [The details of steps 2–5 are drawn from the GREX and De Loach descriptions of related events. Rice (1965) gives the intergroup exercise its own chapter, but the current Leicester IGE procedure is not published [uncertain].]
Example language (illustrative): "Your group sent a delegate with no authority to agree anything, then felt let down when nothing was agreed. What was the group protecting by tying her hands?"
Watch for: Competition, collaboration and conflict between groups, and the experience of exclusion when joining or failing to join a group. It is working when members can link how authority was given to their representative with the outcome they got.

A5. Institutional / Organisational / Whole System Event (incl. AKRI Institutional System Event)

Purpose / when: To study relations between the parts and the whole, how power and influence are distributed, and members' relatedness to the conference Management. It usually comes in the middle or later part of the conference. [source: LC19 (Whole System Event); GREX 2012 (Organizational System Event); De Loach 2014 (ISE)]
Steps (De Loach's ISE, version followed for detail):
1. In the opening plenary the Director explains the representation roles.
2. Management (6–8 staff) works in a public space with an open boundary. Member subsystems occupy numbered rooms. Consultants have a separate space and do not belong to Management.
3. Management writes working hypotheses about the whole system, reads them aloud and carries them to the subsystems. Each hypothesis has a "because" clause and supporting data and ends by inviting members to work with consultants.
4. Boundary managers greet visitors and record their time of arrival, role, room and purpose.
5. One member may join Management for a single session on a first-come basis, and that tenure cannot be extended or repeated.
6. Consultants go to member groups only when asked or in a crisis. They help members explore and extend hypotheses with pointed questions about feelings, fantasies and relatedness. They return to their own space 5–8 minutes before the session ends.
7. Management manages the closing plenary in the here and now, again through hypotheses.
Example language (illustrative, management hypothesis): "Management's working hypothesis is that the membership is treating Management as all-powerful, because three groups have sent observers but none has sent anyone authorised to negotiate, and the requests we receive ask us to decide for members."
Watch for: Refusal to hear hypotheses (De Loach reads this as rejecting the task), anonymity as a defence, occupation of unauthorised spaces, fantasies of an omnipotent or family-style Management, and staff rivalry or splitting between Management and consultants. In De Loach's example about two-thirds of members wanted to audit Management and about one-third refused to deal with it at all. [source: De Loach 2014]

A6. Embodying Role Event (ERE, Leicester)

Purpose / when: At Leicester, an event across all parts of the conference intended to build mindfulness and awareness of internal states while in role. [source: LC19]
Steps: Not described in published materials [uncertain]. Do not run it without the Leicester staff's own design.
Example language: none sourced.
Watch for: [uncertain].

A7. Social dreaming (in some conferences)

Purpose / when: Some conferences include it. Members use dreams to explore unconscious group processes. [source: Tavistock Institute]
Steps: The procedure (the social dreaming matrix, associated with W. Gordon Lawrence) is not described here [uncertain].
Watch for: [uncertain]. Rioch records that Bion responded to a dream told in a group by commenting on the form of the communication, not its content. This is a historical anecdote, not a protocol. [source: Rioch 1983]

A8. Review and Application Groups (RAG) / Role Analysis and Application Groups (RAAG)

Purpose / when: Small reflective groups that meet regularly. They review each day's experience, examine the roles members took up consciously or unconsciously, and later plan how to apply the learning in members' home work roles. [source: LC19; Tavistock Institute; GREX 2012]
Steps (GREX version):
1. Each group has 4–6 members and one consultant, and meets for 75 minutes once a day.
2. The consultant makes sure each member can state a clear learning intention.
3. Members review the roles they took in the events.
4. The consultant gives individual feedback on roles and links group-dynamics theory to each member's participation. This is one of the few places where didactic input is allowed.
5. The final session works on applying the learning to the home role.
Example language (illustrative): "What would you like to have understood by the end of the week?" "In the inter-group event you kept the minutes for your group. What did taking up that role allow you, and what did it stop you from doing?"
Watch for: Review turning into personal therapy, or into complaints about staff. It is working when members can describe roles, not traits, and can name one specific change to try at work.

A9. Closing plenary

Purpose / when: At the end of the conference, and at the end of the institutional event. It marks the exit boundary. [source: LC19; GREX 2012]
Steps: The whole conference meets. Staff stay in role. In the institutional event closing, Management works with hypotheses and data, and members choose where to sit. [source: De Loach 2014]
Example language (illustrative): "We are in the last hour of this institution. It may be hard to leave something we made together, and hard to own what we did in it."
Watch for: Flight into celebration or premature farewells, and late attacks on the staff.

A10. Post-conference review sessions

Purpose / when: At about two and four weeks after the conference (GREX), for integrating the learning into work roles. [source: GREX 2012]
Steps: Group sessions for all participants and staff, plus individual meetings.
Watch for: Difficulty re-entering ordinary life. This is [synthesis]: the sources only say that the learning continues after the event.

B. Consultant techniques (in role)

B1. Group-level interpretation of basic-assumption activity

Purpose / when: To name a covert, shared mentality when it seems to be pulling the group off its task, so that it can be examined. [source: Endnotes 2020; Emery; Gross 2015; AKRI competencies]
Steps:
1. Notice the mentality from the group's behaviour as a whole. Dependency: the group acts as if it exists to be sustained by a leader. Members relate to the leader rather than to each other and show little wish to learn (Bion 1952, as quoted by Emery). Fight-flight: the group organises around an enemy, real or imagined, to fight or flee, is hostile to thinking, and is often busy or chaotic. Pairing: the group watches two members or subgroups with hopeful expectation, as if waiting for something or someone new to be born. [source: Bion 1952, as quoted by Emery; French & Simpson 2010, quoting Experiences in Groups]
2. Check your hypothesis against concrete data: who speaks, to whom, about what, and how the group treats you.
3. Look for the work group as well as the basic assumption. Bion held that work-group functions are always pervaded by basic-assumption phenomena, so both run at once. French and Simpson describe a work-group form of each assumption: dependence on a leader the group has authorised to help it think, rather than on an idealised knower; engagement with difference, rather than an us-and-them split; and a productive pair working for the task, rather than a hoped-for saviour. Where it is there, name the work-group capacity as well as the regression. [source: French & Simpson 2010, author manuscript, quoting Experiences in Groups]
4. Address the group as a whole, briefly, as a hypothesis.
5. Watch what follows, and treat the response as new data.
Example language (illustrative): Dependency: "The group seems to be acting as though only I hold the knowledge it needs, which may spare members from finding out what they know." Fight-flight: "The talk has turned to the hopeless administration. Perhaps attacking an enemy outside this room is easier than looking at what is happening in it." Pairing: "Everyone has gone quiet while Tom and Leila talk, as if the group is hoping the two of them will produce an answer that saves us all."
Watch for: Interpretations can become accusations if they are given too often or too cleverly. The individual being described can be shamed. A sign that it is working: the activity shifts, and members begin to name the mentality themselves. An intervention in the same form as the prevailing mentality (for example, fighting back against a fight-flight group) tends to reinforce it. [source: French & Simpson 2010]

B2. Interpreting the group's relation to the consultant (authority)

Purpose / when: The group's treatment of the consultant shows its relationship to authority. Rice's summary of Bion is the model: keep trying to understand what the group is doing, especially to you, and make that clear to the group. [source: Rice quoted in Rioch 1983; Gross 2015; Endnotes 2020 on Bion declining to lead]
Steps:
1. Refuse to take the leader or expert role the group offers. Do not supply agendas, introductions or reassurance. [source: Endnotes 2020; Gross 2015]
2. Track the projections onto you: idealisation, contempt, ignoring, seduction.
3. Interpret them as the group's current relationship to authority.
Example language (illustrative): "I've been spoken about for ten minutes as if I'm not in the room. It may be safer to dismiss authority than to be disappointed by it."
Watch for: Frustration and hostility towards the consultant are expected. Bion was also met with frustration, and his advice was to "think under fire". [source: Endnotes 2020] Do not retaliate and do not collapse.

B3. Formulating working hypotheses (with a "because" clause)

Purpose / when: In any event, and above all in institutional events and RAAGs. The aim is to give members a testable statement rather than a pronouncement. [source: De Loach 2014; AKRI LDRS 2024]
Steps:
1. Collect data: behaviour, time, use of territory, who represents whom, and your own feelings.
2. Draft the hypothesis: "My or our working hypothesis is that X, because of data Y."
3. In the ISE, Management writes it down, reads it aloud, carries it to the subsystems and invites members to work on it with consultants.
4. Revise it in the light of the response.
Example language (illustrative): "A working hypothesis: the group is avoiding the question of who leads, because each time someone proposes a direction the topic changes within a minute."
Watch for: Hypotheses with no data behind them, which become the consultant's own projection.

B4. Using one's own experience: self-containment and voicing

Purpose / when: At all times. The consultant's feelings (boredom, fear, excitement, being deskilled) can be evidence of group processes. [source: AKRI competencies; De Loach 2014 on deskilling; Rioch 1983 on Bion valuing boredom in learning]
Steps:
1. Notice the feeling.
2. Stay still enough to sort what belongs to you, your history and identity from what belongs to the group or the system.
3. Voice the experience as a hypothesis rather than act on it.
Example language (illustrative): "I notice I've become very sleepy. I wonder if the group is putting something to sleep that it would rather not wake."
Watch for: Acting out your own needs, for example to be liked, to be clever, or to escape.

B5. Naming the individual as carrier for the group

Purpose / when: When one member is scapegoated, rescued, idealised or pushed into a fixed role. The aim is to return what the member carries to the group. [source: AKRI competencies ("individual as carrier"); Emery citing Bion on valency]
Steps:
1. Notice a pattern in which one person keeps carrying a function: anger, vulnerability, dissent, competence.
2. Interpret what the group gains from it, addressing the group and not the individual.
Example language (illustrative): "It seems Sam has become the group's spokesperson for anger. Perhaps that lets everyone else stay reasonable."
Watch for: Interpretations that leave the individual even more exposed. Protect the member by speaking to the group's use of them.

B6. Boundary management (time, territory, task, role)

Purpose / when: Constantly. Keeping boundaries fixed makes the dynamics visible, and boundary crossings are data. [source: GREX 2012; De Loach 2014; Tavistock Institute]
Steps:
1. Start and end every session exactly on time.
2. Stay in your designated territory. In the ISE, consultants do not enter member spaces unless asked or in a crisis.
3. Keep your role boundaries regardless of positive or negative responses.
4. Interpret late arrivals, the use of rooms, and moves out of role.
Example language (illustrative): "It's time." (said at the boundary, and nothing more). "Several people came in late today. Perhaps the group is testing whether its boundaries matter."
Watch for: Being drawn into negotiating the boundaries, and staff using breaks for private work, which creates confusion. [source: De Loach 2014]

B7. Managerial interventions

Purpose / when: Alongside interpretation, a consultant or Management may intervene to manage the event, for example by restating the task or the time, or by stating representation rules. [source: AKRI competencies ("group-level interpretation and managerial interventions"); De Loach 2014]
Steps: State the boundary or rule, briefly and publicly, then go back to the consultant role.
Example language (illustrative): "Observers are invited to remain silent in this space."
Watch for: Managerial statements turning into control or rescue.

B8. Distinguishing stated from actual task

Purpose / when: When the group's behaviour suggests it is pursuing a task other than the one it says it is pursuing. [source: GREX 2012 learning objective]
Steps: Contrast the stated task with what the group is actually doing, and ask what the substitute task protects.
Example language (illustrative): "The task is to study this group, yet for most of the session we have been planning a social evening. Perhaps building friendliness is easier than studying the tension in the room."
Watch for: Moralising. Rioch treats friendliness and hostility as secondary to the task, not as wrong. [source: Rioch 1975]

B9. Working with identity, representation and socio-political dynamics

Purpose / when: When race, gender, age, nationality or other identities shape who is given authority, who is silenced and what is projected. [source: GREX 2012; AKRI competencies]
Steps:
1. Include identity and power in your hypotheses.
2. Consider what your own identity draws in the way of projection.
3. Link dynamics in the room to the wider socio-political forces where the data supports it.
Example language (illustrative): "The two women of colour have been asked twice to explain the group's feelings about race. The group may be handing this work to them so that others don't have to own it."
Watch for: Re-enacting real harm. Current AKRI material supports restorative processes when offence occurs. [source: AKRI LDRS 2024]

B10. Consultation on request (institutional events)

Purpose / when: In the ISE. A member group asks a consultant for help. [source: De Loach 2014]
Steps:
1. Ask for a clear and specific request.
2. Go to the group's territory.
3. Help them explore and extend their hypotheses with pointed reflective questions about feelings, fantasies and relatedness to Management.
4. Leave 5–8 minutes before the session ends, so that you return to the consultants' space without being captured by members.
Example language (illustrative): "What do you imagine Management thinks of your group, and what makes you imagine that?"
Watch for: Being recruited as the group's advocate, or as Management's spy.

B11. Regaining role

Purpose / when: After the consultant has lost the role, for example by chatting, defending themselves, giving advice or retaliating. [source: AKRI competencies]
Steps:
1. Notice that you have lost the role.
2. Own the mistake briefly.
3. Interpret what drew you out of role, if it is useful.
4. Take up the role again.
Example language (illustrative): "I've just been defending myself rather than doing my job. It may be worth asking what the group needed me to defend."
Watch for: Over-apologising, which turns the session into a session about you.

B12. Staff group work (staff meetings and management in public)

Purpose / when: Before, during and after the conference. Staff study their own dynamics, which mirror the membership's, support colleagues and share information. [source: De Loach 2014; AKRI competencies (collaboration, supporting colleagues)]
Steps:
1. Hold regular staff meetings.
2. Report events and voice differences in ways that support the Director's task.
3. Name staff rivalry, envy, narcissism and splits between Management and consultants.
4. Keep staff work out of the breaks.
Watch for: Consultants trying to join Management to escape discomfort, and "we and them" splits. De Loach reads these as ways of avoiding uncertainty. [source: De Loach 2014]

C. Clinical adaptations for psychotherapists ([synthesis] throughout: these are not part of the published conference method)

C1. Basic-assumption lens in therapy groups

Purpose / when: In psychodynamic or process therapy groups, when the group is stuck. Historical note: Bion's own Tavistock groups were therapy groups in which he declined to lead and encouraged members to examine the group's tensions, including their wish for him to take charge. [source: Endnotes 2020] The specific technique below is [synthesis].
Steps:
1. Hold a private hypothesis about which basic assumption is active.
2. Offer occasional group-level comments, mixed with individual work as your model of therapy allows.
3. Keep your duty of care for each patient. Unlike a conference, a therapy group has a clinical contract.
Example language (illustrative): "Today the group seems to be waiting for me to make it better. I wonder what we lose by waiting."
Watch for: Patients with fragile self-structure, psychosis or active risk, who may experience a withholding, group-only stance as abandonment. Use a more supportive and explicit style with them.

C2. Consulting to staff teams and reflective-practice groups

Purpose / when: When teams show repeated conflict, blaming of management, or idealised dependence on one clinician. [synthesis, extending the "application" aim of RAAGs]
Steps:
1. Contract a clear task, time and role with the commissioning manager.
2. Work in the here and now with working hypotheses about the team's relation to its primary task and its anxieties.
3. Review regularly against the contract.
Watch for: Dual roles. Do not consult to a team you clinically manage or are a member of.

C3. Role analysis in supervision

Purpose / when: To help a supervisee examine the role they take up in the system, not only their countertransference to individual patients. [synthesis from RAAG]
Steps: Ask what role the person is in, who authorised it, what task it serves, and what the system may be asking them to carry.
Example language (illustrative): "Whose anxiety are you holding when you take every crisis call yourself?"

5. Handling problems

The published sources address staff dynamics and member "resistance" to the task, but say little about distress, screening or ruptures. Guidance marked [synthesis] is the writer's own.

Silent or stuck group (SSG). Bion valued boredom as part of learning, so treat silence and boredom as data before you treat them as a problem. [source: Rioch 1983] Interpret what the silence may be doing, for example waiting for the consultant (dependency) or avoiding a feared conflict, rather than breaking it with a question or an exercise. Do not introduce structured exercises: the design is non-didactic. [source: Tavistock Institute; synthesis for the response]
Demands that the consultant lead, explain or reassure. Refuse the leader role and interpret the demand as part of the group's relation to authority (B2). Expect frustration. [source: Endnotes 2020; Gross 2015] Restate the task once if a member is genuinely confused about the design, then return to role. [synthesis]
Hostility and attacks on staff. Hold to the task "no matter how difficult it may become" and "think under fire". [source: Rioch 1975; Endnotes 2020] Interpret the attack as data about authority. Do not retaliate, justify yourself or collapse. If you do lose the role, own it briefly and take it up again (B11). [source: AKRI competencies]
"Resistance": rejecting hypotheses, anonymity, occupying unauthorised spaces. De Loach reads these as acting out a rejection of the primary task. Management addresses them directly and in public, and avoids private dealing. [source: De Loach 2014] Address the resistance at the group level, not by singling out individuals. [synthesis]
Non-attendance, lateness and leaving early. Treat these as boundary data and interpret them in the group. [source: GREX 2012 on fixed times; synthesis for the response] If a member wants to leave the conference, the Director or a designated staff member should meet them out of the study event to clarify their authority to decide, and check their safety and support. [synthesis: no procedure is stated in the sources]
Overwhelm, acute distress or decompensation (psychological casualty). AKRI states that the work is not a substitute for mental health care, and no screening procedure is published [uncertain]. Rioch names the dangers of mobs, fanatical leaders and the "warm glow of togetherness", but gives no casualty procedure. [source: AKRI LDRS 2024; Rioch 1975] The writer's recommendations [synthesis]: (1) the application or brochure should state the emotional intensity plainly and ask members to name a current therapist or support; (2) name a staff member (not the member's SSG consultant) who can see a distressed member outside the events; (3) step out of the interpretive stance in that meeting and use ordinary clinical assessment, including risk assessment by current standard practice, a safety plan and links to local services; (4) let the member choose whether to return, and frame withdrawal as an exercise of their own authority, not as failure; (5) record what happened and review it in the staff group.
Scapegoating and attacks on an individual. Interpret the group's use of the person (B5) quickly, before the person is further exposed. Where identity-based harm has occurred, use the restorative processes AKRI describes. [source: AKRI competencies; AKRI LDRS 2024] In a serious case, the Director should meet the people involved outside the event. [synthesis]
Re-enactment of oppression (race, gender, other identities). Include identity in your hypotheses (B9), and watch what your own identity draws. [source: GREX 2012; AKRI competencies] Distinguish "the group is using this person to carry X" from real discriminatory acts. The second needs naming and repair, not only interpretation. [synthesis]
Staff splitting, rivalry and deskilling. Staff can feel competitive or envious, consultants can try to join Management to escape discomfort, and "we and them" splits can form. Name these in staff meetings. Keep staff work public and out of breaks. Use the experience of being deskilled as data about dependent or counter-dependent dynamics. [source: De Loach 2014]
Consultant over-interpreting or interpreting individuals. This shames members and turns the event into therapy without a contract. Keep interpretations short and addressed to the group, and only offer them when they will aid learning. [source: GREX 2012 ("where interventions will aid learning"); synthesis]
A member uses the RAAG as personal therapy. Acknowledge what they have said. Bring it back to the role and the system: what role were you in, and what did the system need from you. If needed, suggest outside therapeutic support. [synthesis based on GREX RAAG task]
Difficulty after the conference (re-entry). Use follow-up review sessions at about two and four weeks, as GREX did. [source: GREX 2012] Advise members not to make major work or life decisions immediately after an intense conference. [synthesis]
Carrying the method into therapy groups (C1). A withholding, group-only stance can be experienced as abandonment by patients with severe personality pathology, psychosis or active risk. Use a more explicit and supportive style with them, keep individual risk management and contract for group-level work. [synthesis] Note that Emery, a critic writing from a different tradition, argues that Bion's patients "remained patients" in a leader-led structure. [source: Emery]
Dual relationships. Do not staff a conference attended by your own patients, and do not consult to a team you manage or belong to. [synthesis]

Sources

Pages used:

Books and papers cited (book contents known through the publisher tables of contents and secondary sources above):

Bion, W.R. (1961). Experiences in Groups and Other Papers. London: Tavistock.
Bion, W.R. (1952). Group dynamics: a re-view. International Journal of Psycho-Analysis, 33.
Rice, A.K. (1965). Learning for Leadership: Interpersonal and Intergroup Relations. London: Tavistock.
Colman, A.D. & Bexton, W.H. (Eds.) (1975). Group Relations Reader 1. A.K. Rice Institute.
French, R. & Simpson, P. (2010). The ‘work group’: redressing the balance in Bion’s Experiences in Groups. Human Relations, 63(12), 1859–1878.
Armstrong, D. (2010). Working with groups: the consultant stance. British Journal of Psychotherapy, 26(2), 161–166.

← Previous · V-084Group analysis Next · V-088 →Group cognitive-behavioural therapy