Marathon groups

1. Introduction

Version followed: the weekend encounter-group marathon in its later, modified form, with sleep breaks, a no-coercion rule and three phases, as described by a practising UK marathon leader (Wibberley 1994, co-led with John Leary-Joyce from 1984). This guide uses the classic 1960s format of George R. Bach (1966, 1967a–c) and Frederick H. Stoller (1967, 1968a–b) only as history and as the source of the risks described in the Harm account. The primary texts by Bach, Stoller or Mintz are not drawn on here. Their titles and dates are confirmed from published reference lists. Their procedural content here comes from secondary and practitioner sources and is tagged [uncertain] where it goes beyond those. [source: Wibberley 1994; Rodríguez Zafra et al. 2019 reference list; Mintz 1969 record]

What it is. The marathon is a format, not a single therapy. An encounter or therapy group meets continuously for an extended block, usually 24–48 hours over a weekend. The leader may work from encounter, T-group, Gestalt, psychodrama or eclectic traditions. Stoller's version was grouped with the 'accelerated' treatments in Gazda's 1968 Innovations to Group Psychotherapy. Bach published 'The marathon group: intensive practice of intimate interaction' (Psychological Reports, 1966) and a 1967 series on marathon group dynamics. One paper in that series covers 'therapeutic aggression'. [source: Canadian Counsellor 1970 review of Gazda; Rodríguez Zafra et al. 2019 references]
How change is meant to happen. The fixed, continuous time limit creates urgency. That urgency is meant to push members further than ordinary weekly sessions would ('to reach further' than usual). The stated aims are leveling (frank mutual honesty), trust, honesty and the constructive expression of aggression. [source: Canadian Counsellor 1970] The group stands in for the world of emotionally significant others. Members show themselves 'as they are in the world outside', see how their patterns affect others, and face feelings now instead of postponing them. [source: UIA Encyclopedia entry] In the classic version, fatigue was also used on purpose: members would have 'less energy to resist'. Later practitioners dropped this, and this guide does not use it. [source: Wibberley 1994; see Harm account]

2. Framework

Format options (name the one you run):

Version	Time block	Sleep	Source
Classic 1970s UK marathon modelled on US marathons	Midnight Friday to midnight Sunday, in one room	About 5 h in total; none in the first 24 h	Wibberley 1994 (describing earlier groups)
Modified marathon (followed here)	7:30 pm Friday to 7:30 pm Sunday	Two sleep breaks, plus meal and tea breaks	Wibberley 1994
Research format (Stanford study)	30 hours of group time, as a single marathon or spread over weeks	Not stated	Smith c.1973, summarising Lieberman, Yalom & Miles 1973
Prison 'saturation marathon'	3–5 all-night sessions, inmates with their partners	Sleep not permitted	IJGP 1977, NCJRS abstract (not recommended; see Harm account)

Group size, leader–member ratio and fees are not stated in the sources used. A workable size is 8–14 members with two co-leaders [synthesis, uncertain].

Phases of the modified marathon [source: Wibberley 1994 unless tagged]

Pre-group screening and preparation (one or more individual meetings in the weeks before). Goal: exclude people at high risk, get informed consent and set personal goals. Move on when each member has been assessed, briefed on the format and rules, and has signed consent. [synthesis; Finkelstein et al. 1982 advised against borderline or psychotic people joining est, as reported in Wikipedia LGAT]
Beginning (Friday evening). Goal: lower anxiety and build connection through shared, structured exercises. Move on when everyone has spoken, the ground rules have been restated and accepted, and members start addressing each other directly instead of the leader [criteria: synthesis].
Middle (most of Saturday and Sunday morning). Goal: most of the exploration. Members respond honestly and own their feelings instead of projecting them. The work alternates between structured exercises, individual work, free-form group time and small groups of 4–5. Sleep, meal and tea breaks count as part of the process. Move on at the planned clock time, not according to how intense the group is. Leave several hours of the ending phase protected [synthesis].
Ending (Sunday afternoon). Goal: goodbyes, finishing loose ends, and planning how to carry the learning into ordinary life. Finish only when no one is left mid-process. Anyone still destabilised gets a check-in and follow-up arranged before leaving [synthesis].
Follow-up (about one month later). Goal: review what transferred, catch delayed adverse reactions and refer on. One study measured outcomes immediately after the group and again at one month [source: Putman 1973 abstract]. The follow-up session itself is [synthesis].

3. Guiding principles

Here-and-now relationships in the room are the material. Work with what happens between the people present, not with case history. [source: Wibberley 1994; UIA]
Leveling. Members give direct, honest feedback on each other's impact, without the usual social politeness. [source: Canadian Counsellor 1970]
Own your feelings. Say 'When you do that I feel angry', not 'You are making everyone angry'. Prompt members to rephrase projections as I-statements. [source: Wibberley 1994]
Face feelings now, not later. Notice when a member is postponing or avoiding a feeling and invite it into the room, without forcing it. [source: UIA; the 'without forcing' qualifier is synthesis]
Free choice and self-set risk. No coercion. Each member decides how far to go, and declining a risk counts as a legitimate step. Say this aloud at the start and again whenever the group puts pressure on someone. [source: Wibberley 1994]
Name the group's hidden norms. Early marathons carried unspoken expectations that members be open, expressive and risk-taking. Notice and name any 'correct way to be' that the group is enforcing. [source: Wibberley 1994]
The leader is a facilitator, not the centre. Early marathons relied too much on the leader's charisma. Hand work back to the group and use the group's own creativity. [source: Wibberley 1994]
Warmth over stimulation. In the Stanford study there were no severe casualties in groups with warm, caring leaders. Severe casualties occurred under highly charismatic, intrusive 'Energiser' leaders, laissez-faire leaders and 'cool aggressive stimulator' leaders. Be warm, give meaning to what happens and set limits. Do not be provocative or absent. [source: Smith c.1973 on Lieberman et al. 1973]
Aggression is expressed, never acted out. Anger is welcome as words and feelings. There is no physical violence, and limits are set early. [source: aims per Canadian Counsellor 1970; the specific rule is synthesis, uncertain for Bach's original rules]
Breaks are part of the process. Protect sleep, meals and rest, and use the contact that happens during breaks. [source: Wibberley 1994]
Responsibility. Members are responsible for their own behaviour and its consequences. [source: IJGP 1977 abstract] Leaders are responsible for safety: monitoring, warmth and limits [synthesis].
No alcohol or drugs for the duration. [source: IJGP 1977 abstract; general rule, synthesis]
Transfer to life outside. Every insight is followed by a plan for using it elsewhere. [source: Wibberley 1994]
Do not screen in for vulnerability. People with psychotic or borderline-level fragility are not candidates for intensive, extended formats. [source: Finkelstein et al. 1982 via Wikipedia LGAT; extended to marathons as synthesis]

4. Intervention inventory

Scope note: the sources used name the components of a marathon (structured exercises, individual work, free-form group, small groups, nonverbal and contact work, honest feedback, ownership of feelings, breaks, closing) but give no scripts. The steps below are my operationalisation of those components [synthesis] unless tagged otherwise. The techniques attributed to Bach and Stoller (therapeutic aggression and fair fighting, focused videotape feedback) come from texts that are not drawn on here and are marked [uncertain]. All example language is illustrative.

4.1 Screening and informed consent

Purpose: keep out people the format could harm and make the format's demands clear in advance. [synthesis; exclusion of borderline and psychotic participants follows Finkelstein et al. 1982 via Wikipedia LGAT]
Steps: (1) Hold an individual interview. Ask about current psychiatric history (psychosis, bipolar disorder, recent suicidality or self-harm, active substance use, dissociation, recent trauma or bereavement), medical issues worsened by fatigue or strong emotion (epilepsy, cardiac conditions, pregnancy), and current therapy. (2) Exclude or defer people with active psychosis, mania or hypomania, borderline-level instability, active suicidality, or a current crisis. (3) Describe the schedule, sleep breaks, intensity, the possible physical-contact exercises (and that they are optional) and the right to pass or leave. (4) Get written consent, an emergency contact and the name of the member's own clinician. (5) Ask each member for one interpersonal goal.
Example language (illustrative): 'It's a long weekend with a lot of feeling. You choose how far you go, and saying no counts as part of the work.'
Watch for: people seeking the group as a rescue in a crisis, pressure from a partner or employer to attend, and minimising of psychiatric history. Working: members arrive with realistic expectations and their own goal.

4.2 Contracting ground rules

Purpose: create safety and a shared frame for intensity.
Steps: state the rules on Friday evening, check that each member agrees, and restate them at the start of each day. A suggested set: confidentiality; stay for the whole marathon if you can, and tell a leader before leaving (leaving is allowed); no alcohol or drugs; no physical violence or damage; physical contact only with consent; speak to people, not about them; the right to pass; sleep and meal breaks are kept. [Sources: no coercion and self-set risk from Wibberley 1994; no substances, no gossip or indirect comments, no private side-conversations from the prison marathon in IJGP 1977. That marathon also banned 'I don't know' and sleep, and both of those rules are rejected here. Composite list: synthesis]
Example language (illustrative): 'If you say something about someone in the room, say it to them.'
Watch for: members using the rules to police each other ('you're breaking the honesty rule'). Bring it back to choice.

4.3 Opening structured exercises

Purpose: lower initial anxiety and build connection. [source: Wibberley 1994]
Steps (illustrative structure; Wibberley does not specify exercises): (1) A go-round: name, what brought you, what you are nervous about. (2) Paired introductions, then each person introduces their partner to the group. (3) A round of 'one thing I want from this weekend'. Keep this to 60–90 minutes. [synthesis]
Example language (illustrative): 'Turn to someone you know least and tell them what you're most nervous about tonight.'
Watch for: members who say almost nothing; check in with them lightly. Working: the energy in the room rises, and members start addressing each other.

4.4 Here-and-now interpersonal feedback (leveling)

Purpose: help members see the impact of their behaviour on others, which is the core marathon aim. [source: Canadian Counsellor 1970; UIA]
Steps: (1) Notice an interaction in the room. (2) Invite the person affected to tell the other person directly how they are affected. (3) Ask for specific behaviour ('when you looked away'), not character judgements. (4) Ask the person receiving the feedback what they heard and how it lands. (5) Ask whether this feels familiar from outside the group. (6) Invite others' views, but stop it turning into a pile-on. [synthesis]
Example language (illustrative): 'Can you tell Sam directly what happened in you just then?' / 'Is this familiar? Do people outside here react to you like this?'
Watch for: several members converging on one person (scapegoating), and feedback used as an attack. Working: the receiver gets curious instead of defensive, and the pattern connects to outside life.

4.5 Owning feelings (I-statements)

Purpose: turn projection and blame into ownership. [source: Wibberley 1994]
Steps: when a member makes a 'you' or 'everyone' statement, invite a rephrase that starts with 'I', names the feeling, and names the behaviour that triggered it. Then ask what the feeling is like now.
Example language: Wibberley's model is 'when you do that I feel angry' (quoted). Illustrative prompt: 'Try starting with "I".'
Watch for: rote compliance in formula language. Working: the emotional tone becomes more alive and less accusing.

4.6 Individual work in the group

Purpose: focused exploration of one member's issue while the group witnesses it. [source: Wibberley 1994 names 'individual work' but not its method. The method is uncertain and depends on the leader's tradition: Gestalt, psychodrama and so on]
Steps: (1) The member asks for time; the leader never selects someone. (2) Agree a focus and time limit. (3) Work in the leader's own trained method. (4) Close with the member's own words. (5) Invite the group to share their own resonance with the material, not advice. [synthesis]
Example language (illustrative): 'What would you like from the next 20 minutes?' / 'Who was touched by that, and what in your own life does it connect to?'
Watch for: the leader escalating the emotion for its own sake, which is the 'Energiser' pattern linked to casualties [source: Smith c.1973]. Working: the member reports new understanding or relief and can re-join the group.

4.7 Free-form group time

Purpose: let interpersonal patterns emerge without structure. [source: Wibberley 1994]
Steps: announce an unstructured period. Intervene only to point out process ('I notice the group has gone quiet since Ana spoke'), to protect someone, or to enforce the rules. [synthesis]
Example language (illustrative): 'What's happening between us right now?'
Watch for: long silences that turn persecutory, and dominant members. Use the leader's limit-setting.

4.8 Small groups of four or five

Purpose: more airtime and a safer scale for less confident members. [source: Wibberley 1994]
Steps: set a clear task (for example 'what I'm not saying in the big group'), a time limit and a report back that is optional. Mix the membership across rounds. [synthesis]
Watch for: sub-groups turning into cliques. Working: quieter members bring something back to the large group.

4.9 Nonverbal and physical-contact techniques (flag: risk)

Purpose: the sources say nonverbal methods were used especially when anger or hostility was expressed openly, that physical contact was used to get past intellectual defences, and that comfort was given during emotional strain. [source: UIA]
Steps (safer form): (1) Offer the exercise and do not require it. (2) Describe exactly what contact is involved. (3) Get individual consent and offer a no-contact alternative. (4) Use non-sexual, low-risk forms: hand on shoulder for support; pushing palm against palm to physicalise resistance; cushions for anger. (5) Debrief. [synthesis]
Example language (illustrative): 'Would a hand on your back help, or would you rather I just sit here?'
Watch for (RISK): contact used to 'break down defences' against a member's will, sexualised contact, triggering of trauma survivors, and physical injury in force-based exercises. Never combine contact with exhaustion or group pressure. See the Harm account.

4.10 Giving space to a member in deep distress

Purpose: respect a member's deep distress instead of intruding on it. [source: UIA says the group usually gives space instead of intruding]
Steps: (1) Slow the group down. (2) One leader checks in quietly with the member: are they safe, grounded, do they want company or space? (3) Others stay present without crowding. (4) Before resuming, check orientation and safety. [synthesis]
Example language (illustrative): 'You don't have to say anything. I'm here. Can you feel your feet on the floor?'
Watch for: dissociation, disorientation, panic, unusual ideas, and statements about self-harm. These call for the safety procedure in section 5.

4.11 Constructive aggression and 'fair fight' exercises (Bach) (flag: risk; uncertain)

Purpose: Bach treated aggression as something that could be used therapeutically. His 1967 paper is titled 'Dimensions of helpfulness: therapeutic aggression'. He also taught couples rule-governed 'fair fighting' (Bach & Wyden 1969, The Intimate Enemy). The constructive expression of aggression is one of the stated aims of these groups. [source: titles from Rodríguez Zafra et al. 2019 references and Canadian Counsellor 1970. The texts are not drawn on here, so the procedures are [uncertain].] Bach's fair-fight vocabulary is widely reported in secondary literature. It includes a personal 'belt line' below which a partner must not hit, and 'gunnysacking' for storing up grievances. Those terms are not verified against the book [uncertain].
Steps (safer form; my operationalisation, not Bach's protocol) [synthesis]: (1) Use only with pairs who both consent, and never where there is intimate-partner violence, coercive control or a large power imbalance (screen for these separately). (2) Before starting, each person tells the other which topics are off-limits (their 'belt line'). (3) One complaint, about specific present behaviour, with one concrete request. (4) The partner repeats back what they heard before answering. (5) Set a time limit; the leader stops escalation, contempt, threats and references to the off-limits areas. (6) Debrief both partners, separately if needed.
Example language (illustrative): 'One complaint, about one recent thing, and what you'd like instead.' / 'Stop. That went below the line you both agreed.'
Watch for (RISK): escalation into humiliation, a dominant partner using the exercise to intimidate, physical posturing, and one member being 'fought' by the group. Do not use aggression exercises with exhausted members or in group-versus-individual form. See the Harm account. Working: both partners report feeling heard and the conflict becomes more specific and less global.

4.12 Focused videotape feedback (Stoller) (uncertain)

Purpose: members see their own interpersonal behaviour from outside. Stoller is associated with videotape feedback within accelerated groups. [uncertain: inferred from a 1978 thesis title on 'focused videotaped feedback' in groups and from general literature. Stoller's own texts are not drawn on]
Steps (safer form) [synthesis]: (1) Get separate written consent for recording, storage and deletion. (2) Record a segment. (3) Replay a short excerpt with a focus agreed in advance ('watch your hands when you talk about your father'). (4) The member comments first, then the group. (5) Delete the recording, or store it securely under data-protection rules.
Example language (illustrative): 'Before we watch, what do you think you'll see?'
Watch for: shame reactions to one's own image, and confidentiality breaches. Working: the member notices a pattern they had not seen before.

4.13 Extended continuous time (the defining feature) (flag: risk)

Purpose: continuity and the time limit build urgency and intensity, and leave no escape from the relationships in the room. [source: Canadian Counsellor 1970; Wibberley 1994] The classic format also used sleep loss and exhaustion to lower resistance. Later practitioners rejected this. [source: Wibberley 1994]
Steps (modified format): keep the continuous weekend but schedule two protected sleep periods plus meal and tea breaks [source: Wibberley 1994]. Members sleep in proper beds away from the group room [synthesis]. Run no emotionally intense work in the hour before a sleep break [synthesis].
Watch for (RISK): members skipping sleep to 'keep working', leaders extending sessions because the work feels 'hot', and signs of fatigue-related destabilisation (see section 5 and the Harm account).

4.14 Breaks as process

Purpose: contact during breaks carries the work forward and lets members recover. [source: Wibberley 1994]
Steps: keep the scheduled breaks. Invite members to bring anything significant from a break back into the group, and do not allow secret coalitions [synthesis; the ban on private side-discussions follows IJGP 1977].
Example language (illustrative): 'Did anything happen over lunch that belongs in here?'

4.15 Couples or partner marathon (variant)

Purpose: bring a partner in so that relationship problems are worked on directly. A prison programme had inmates attend all-night sessions with their partners to prepare for release. [source: IJGP 1977, NCJRS abstract; authors not shown on the abstract page]
Steps (safer adaptation) [synthesis]: screen each partner separately for intimate-partner violence and coercion. Keep the sleep breaks. Set individual consent and a right to pass for each partner. Give couples time to talk in private with a leader present when needed. Arrange follow-up couple sessions.
Watch for (RISK): that programme also banned sleep and the answer 'I don't know', and used fairly confrontive leaders. These are coercive features; do not copy them. See the Harm account.

4.16 Closing: goodbyes, loose ends, transfer planning

Purpose: close what has been opened and carry the learning into ordinary life. [source: Wibberley 1994]
Steps: (1) A round on unfinished business with any member. (2) Short pair or group exchanges to finish it. (3) Each member names one learning and one concrete step for the coming week. (4) Brief the group on possible after-effects: tiredness, emotional rawness, a 'high' followed by a dip. (5) Give contact routes, the follow-up date and the advice to delay major life decisions for a few weeks. (6) A formal goodbye round. [steps 4–6 are synthesis]
Example language (illustrative): 'What will you do differently on Tuesday, with whom?' / 'If you feel flat midweek, that's common. Here's how to reach us.'
Watch for: members leaving euphoric and planning drastic changes, and members still distressed. Hold an individual check-out before they leave.

4.17 Follow-up review

Purpose: check what has transferred and catch delayed adverse reactions. [synthesis; the one-month timing follows Putman 1973]
Steps: hold a group or individual review at about four weeks. Ask what changed, what got harder, and about sleep, mood and any intrusive memories. Refer to individual therapy where indicated. Send feedback to the referrer with the member's consent.
Watch for: the Stanford finding that intense groups gave peak experiences but lowered self-esteem afterwards [source: Smith c.1973]. Ask about both.

5. Handling problems

The sources used give little problem-specific guidance. They support three points: warm, meaning-giving leaders with clear limits had no severe casualties; dropping out can be self-protection; and group pressure and leader charisma are risks [source: Smith c.1973; Wibberley 1994]. Everything else below is [synthesis].

Stuck group (circular talk, flat energy, intellectualising). Name the process ('We've talked about feelings for an hour without anyone showing one'). Switch to a structured exercise or small groups. Ask what the group is avoiding. Do not raise intensity with provocation, extended hours or forced contact. Those are the 'Energiser' and 'aggressive stimulator' patterns linked to casualties. [synthesis; casualty link from Smith c.1973]
Resistance in a member (refusing to engage, passing repeatedly). Treat it as a legitimate choice [source: Wibberley 1994]. Get curious about it instead of overcoming it ('What tells you it isn't safe yet?'). Stop the group from pushing. Never use fatigue or a ban on 'I don't know' to get past it (compare the prison marathon in the Harm account).
Scapegoating or group pile-on. Interrupt. Ask each speaker what in themselves the target stirs up. Turn the focus to the group process. Check in privately with the target before the next break.
Overwhelm, panic or dissociation. Slow down and offer space and grounding (4.10). Move to a quieter room with a co-leader if needed. Do not resume the member's piece of work that day. Check orientation and sleep.
Emerging psychotic symptoms, mania, suicidal statements or severe agitation. Stop the member's participation in the process work. One leader stays with the member. Do a standard risk assessment (current ideation, intent, plans, protective factors). Contact the member's emergency contact or own clinician and arrange urgent psychiatric assessment if needed. Do not let them drive or travel alone. Write a record. Lethal-means counselling covers safe storage and removal of means at home only. A transitory psychotic episode after an intensive group has been documented (see Harm account) [source: Lieberman 1987 via Wikipedia LGAT].
Exhaustion. Enforce the sleep breaks and stop sessions running over. Members who cannot sleep get a quiet room, not more group time.
Non-adherence (substance use, sexual contact between members, leaving early). Restate the rules privately first. Substance use during the marathon means stepping out of the process work. A member who wants to leave may leave. Agree a check-out and same-week follow-up, and never block the exit or shame the decision. Treat sexual contact between members during the event as a boundary issue to raise in the group with care. Sexual or romantic contact between leader and member is prohibited.
Rupture with the leader (a member attacks the leader or feels humiliated by them). Acknowledge your part without defensiveness ('I pushed too hard; I'm sorry'). Repair in the group if the member agrees, otherwise in private. Ask the co-leader to observe and give you feedback.
Leader charisma and dependence. Wibberley's critique of early marathons [source]. Co-lead with someone of equal standing, rotate who leads, turn requests for 'answers' back to the group, and arrange supervision after each marathon.
Post-marathon crash or euphoria. Warn members at closing (4.16). At follow-up, ask about self-esteem as well as peak experiences [source: Smith c.1973 for the self-esteem drop]. Refer anyone with persistent distress.
Modality-specific: escalating aggression in fight exercises. Stop the exercise and separate the partners. Check whether the conflict involves violence or coercion at home, and if so switch to individual safety assessment instead of conjoint work.
Modality-specific: pressure to disclose. If members or leaders demand disclosure ('you're holding back'), name the norm being enforced [source: Wibberley 1994 on hidden expectations]. Restate the right to privacy. Remember that the material shared will leave the room with other participants.

Harm account

This section covers the practices linked to harm in classic marathons, in coercive programmes that adopted the format, and in large-group awareness trainings (LGATs). It states only what the sources used support. Unverified claims are marked.

What was actually done

Classic exhaustion marathons (1960s–1970s). Participants often arrived straight after a full working day. The group ran from midnight Friday to midnight Sunday in one room, which members left only for the toilet. They got about five hours of sleep across the two days, none in the first 24 hours, so that they would have less energy to resist. The group put strong pressure on members to drop their 'image' and express feelings. Hidden norms required them to be open, expressive and risk-taking, and the group relied on the leader's charisma. [source: Wibberley 1994, describing the marathons he took part in, which were modelled on US groups] The US originators were Bach (1966 onward) and Stoller (1967–1968). Their exact sleep schedules and rules are not verified here [uncertain: primary texts].
Physical and nonverbal confrontation. Physical contact was used to get past intellectual defences, and nonverbal methods were used especially when hostility was openly expressed [source: UIA]. Bach framed aggression as therapeutic and taught structured fighting to couples [source: titles only; uncertain].
Aggressive leadership styles (Stanford study, early 1970s). Eighteen San Francisco group leaders each ran a 30-hour group for students, one of them in marathon style. 'Energiser' leaders were highly active, charismatic and controlling. 'Cool aggressive stimulator' leaders confronted members while staying emotionally distant. Some leaders were laissez-faire. [source: Smith c.1973, summarising Lieberman, Yalom & Miles 1973]
Prison 'saturation marathons' (reported 1977). Inmates attended three to five all-night sessions with their partners, led by 'fairly confrontive' therapists from outside the prison. Sleep was not permitted. Answering 'I don't know' was banned as a cop-out, and private conversations were prohibited. The stated aim was to break down the inmates' 'con code' and their manipulation. [source: International Journal of Group Psychotherapy 27(1), 1977, NCJRS abstract] This combines coercive features: sleep deprivation, a ban on not-knowing, no privacy and a captive population whose freedom to volunteer is questionable [synthesis].
Large-group awareness trainings (from the late 1960s on). These were commercial programmes descended from the encounter movement: Leadership Dynamics, Mind Dynamics, est, Lifespring, The Forum and Landmark, among others. Typical formats ran 3–5 days or consecutive weekends with 12–15-hour days, hundreds of participants and one central leader. They followed a written plan that did not vary between trainings. Reported methods include confinement and harassment in long sessions (Paglia); about half the time spent making participants feel bad, followed by effusive praise (McWilliams's 'pressure/release'); named processes such as 'lifeboat' and 'cradling'; confrontational sessions likened to 'attack' therapy (Singer); and, in a 1987 Washington Post report on Lifespring, participants crawling on their hands and knees and wailing. Cushman (1989) described the mass marathon training he studied as a 'pre-meditated attack on the self'. [source: Wikipedia, Large-group awareness training, citing these authors; the individual primary studies are not drawn on]
Synanon (US, from 1958). Published histories widely describe a confrontational encounter 'Game' and extended multi-day sessions at Synanon, which began as a drug-rehabilitation community. [uncertain: check before citing]

How it caused harm

Casualty rates. In the Stanford study, 8% of participants met the criteria for persistent psychological decompensation attributed to the group, against 3% of non-participants. The comparison has been criticised as methodologically uneven. Casualties were 17% under Type A (Energiser) leaders against 3% in Type B groups. Severe casualties occurred only under Energiser, laissez-faire and cool aggressive stimulator leaders, and none in groups with warm leaders. Members of the intense groups had more peak experiences, but their self-esteem fell and they saw others less favourably. Dropout was high (17%), which the commentator reads as self-protection. [source: Smith c.1973 on Lieberman et al. 1973]
LGAT outcomes. Lieberman (1987) found that 5 of 289 participants had stress reactions, one of them a transitory psychotic episode. Lifespring part-funded the study, and Lieberman noted it is unknown whether these reactions would have happened anyway. Haaken & Adams (1983), observing Lifespring as participants, judged the reported well-being to be 'essentially pathological', with weakened ego functioning and impaired reality testing. Finkelstein, Wenegrat & Yalom (1982) reported anecdotal psychiatric casualties after est and noted that psychologists advised borderline and psychotic people not to take part. [source: Wikipedia LGAT summarising these papers] Glass, Kirsch & Parris (1977, American Journal of Psychiatry) is widely cited as a case series of psychiatric disturbances after est [uncertain].
Deaths. None of the sources used document deaths in marathon groups. Deaths and lawsuits involving LGAT participants have been reported in the press, but this is not verified here [uncertain]. Check court records or contemporary reporting before citing a specific case.
Mechanisms [synthesis, from general clinical knowledge, not the sources used]:
- Sleep loss weakens emotional regulation, judgement and impulse control, and increases suggestibility and conformity to the group. It can trigger mania in bipolar disorder, psychosis in vulnerable people and seizures in epilepsy.
- Coercive disclosure under group pressure leaves people exposed and ashamed, with private material in the hands of strangers. Without time to integrate it, disclosure of trauma can retraumatise.
- Aggression exercises can humiliate, rehearse dominance, escalate partner conflict and cause physical injury. In couples where there is violence they hand the abuser a script.
- Pressure-then-relief cycles and peak 'highs' create dependence on the group or leader, which commercial programmes then use to sell further courses and recruit.
- Captive or dependent members cannot freely refuse: prisoners, employees sent by their employer, and partners under pressure.
Most at risk: people with psychotic, bipolar or borderline-level vulnerability, trauma survivors, people in acute crisis, people with low self-esteem, and anyone with little freedom to refuse. Leaders with intrusive or charismatic, high-stimulation styles raise the risk. [source: Smith c.1973 on leader style; Finkelstein et al. 1982 via Wikipedia on diagnostic groups; the rest is synthesis]

Recognising it today

[synthesis, using the features above]

Format red flags: 12-hour or longer days, overnight sessions or 'sleep is optional' framing; no breaks, or restricted toilet breaks; rules against leaving; secrecy about content beforehand; a large audience with one star leader; a script that does not vary.
Process red flags: public 'hot seats' a member did not ask for; bans on saying 'I don't know' or 'pass'; the group told to confront a member; humiliation or regression exercises; touch without individual consent; staged anger 'fights' between strangers; alternating attack and praise; pressure to phone family or recruit others during the event; a hard sell for the next level at the end.
Rationales and phrases to listen for: 'breaking down defences', 'breakthrough', 'getting you out of your head', 'fatigue makes it real', 'your resistance is the problem', 'radical honesty', 'transformation in a weekend'. Rebrandings include intensive leadership retreats, 'immersion' or 'boot-camp' personal-growth weekends, residential 'transformational' courses and some intensive couples 'fight' weekends.
How legitimate versions differ: sleep is protected, risk is self-set, and declining is respected [source: Wibberley 1994]. Members are screened, leaders are warm and qualified, a co-leader is present, and follow-up and referral routes exist [synthesis]. Ordinary intensive or extended group therapy has a clinical contract and does not use exhaustion.

What to do instead and how to help those harmed

Safer alternatives [synthesis]: weekly interpersonal-process group therapy (Yalom tradition); time-limited intensive groups with protected sleep (the modified marathon above); and, for couples conflict, established couple therapies such as Emotionally Focused Couples Therapy or Integrative Behavioral Couple Therapy instead of fight training. Where partner violence is present, use individual safety-focused work, not conjoint conflict exercises.
Assessment of affected clients [synthesis]: take a detailed history of what happened: duration, sleep, coercion, disclosures, physical contact, any sexual exploitation, and whether the person is still involved or being recruited. Screen for PTSD and acute stress, dissociation, depression, shame and suicidality, mania or psychosis after the event, and loss of trust in their own judgement. Ask about money or relationships lost.
Treatment [synthesis]: validate the experience without minimising or sensationalising it. Give psychoeducation about how sleep loss and group pressure work, which reduces self-blame for having 'given in'. Use trauma-focused therapy where PTSD is present, and work on rebuilding autonomy and reality-testing. Avoid recreating the group's intensity in therapy. Refer for psychiatric assessment if there are psychotic or manic features.
Ethical, reporting and legal duties [synthesis; check local law and codes]: informed consent for group work with roles and the limits of confidentiality made clear (for example APA Ethics Code standards on informed consent and group therapy). Do not take part in or refer to programmes with the red flags above. Report a colleague's unsafe or exploitative practice to their licensing or registration body. Report sexual exploitation and follow mandatory-reporting rules where minors or vulnerable adults are involved. Keep safeguarding duties where a client is at risk. Clients may have consumer or civil remedies against commercial programmes; refer them for legal advice and do not give it yourself.

Sources

Sources

Further reading (not drawn on) [uncertain where used]

Bach, G. R. (1966). The marathon group: intensive practice of intimate interaction. Psychological Reports, 18, 995–1002.
Bach, G. R. (1967a–c). Marathon group dynamics I–III. Psychological Reports, 20.
Bach, G. R., & Wyden, P. (1969). The Intimate Enemy: How to Fight Fair in Love and Marriage. Morrow.
Stoller, F. H. (1967). The long weekend. Psychology Today, 1, 28–33.
Stoller, F. H. (1968). Accelerated interaction: a time-limited approach based on the brief, intensive group. International Journal of Group Psychotherapy, 18.
Stoller, F. H. (1968). Marathon group therapy. In G. Gazda (Ed.), Innovations to Group Psychotherapy. Charles C. Thomas.
Mintz, E. E. (1967). Time-extended marathon groups. Psychotherapy: Theory, Research & Practice, 4(2), 65–70; (1971). Marathon Groups: Reality and Symbol. Appleton-Century-Crofts.
Lieberman, M. A., Yalom, I. D., & Miles, M. B. (1973). Encounter Groups: First Facts. Basic Books.
Lieberman, M. A. (1987). Effects of large group awareness training on participants' psychiatric status. American Journal of Psychiatry, 144(4), 460–464.
Haaken, J., & Adams, R. (1983). Pathology as 'personal growth'. Psychiatry, 46(3), 270–280.
Finkelstein, P., Wenegrat, B., & Yalom, I. (1982). Large group awareness training. Annual Review of Psychology, 33, 515–539.
Cushman, P. (1989). Iron fists/velvet gloves: a study of a mass marathon psychology training. Psychotherapy, 26.
Weigel, R. G. (2002). The marathon encounter group: vision and reality. Consulting Psychology Journal, 54(3), 186–198.
Glass, L. L., Kirsch, M. A., & Parris, F. N. (1977). Psychiatric disturbances associated with Erhard Seminars Training. American Journal of Psychiatry [uncertain citation details].
Singer, M. T., & Lalich, J. (1995/2003). Cults in Our Midst. Jossey-Bass.

← Previous · III-101Encounter groups Next · III-103 →T-groups and sensitivity training