What is a cult?

1. The wrong question

“Is it a cult?” sounds like the important question. It is not, for three reasons.

First, the word carries so much stigma that it stops thought rather than starting it. Almost nobody recognises their own community in a word that conjures robes and compounds, so the label lets a harmful group slip past — we’re nothing like that — while the actual mechanisms go unexamined.

Second, it invites the wrong argument. Once the question is whether a particular organisation deserves the label, the discussion becomes definitional, tribal and unwinnable. Members defend the group; critics escalate; the specific harm done to a specific person on a specific day disappears from view. High-control groups are extremely comfortable in that argument. They have rehearsed it.

Third, it implies a binary where there is a gradient. There is no bright line between a demanding community and a coercive one. There is a set of measurable pressures that can be present in small amounts or large ones, and can intensify slowly over years.

2. A structural definition

For the purposes of this site:

A cult, or high-control group, is a social system that maintains itself by systematically reducing its members’ capacity to evaluate it — concentrating authority, controlling information, penalising dissent, and relocating responsibility for harm from those with power to those without it.

Every clause in that sentence is about structure. None of it is about doctrine. A group with beliefs you find bizarre may be perfectly healthy; a group with beliefs you share may be doing enormous damage. Content is not the variable. The variable is the relationship between power and accountability.

Three features follow from the definition, and they are worth stating separately because they are what make these systems so hard to see from inside.

It is self-sealing

A well-formed high-control system contains a built-in answer to every objection, including the objection that it contains a built-in answer to every objection. Doubt is evidence of ego. Distress is evidence of resistance. Criticism from outside is evidence that outsiders cannot understand. Criticism from someone who left is evidence of what leaving does to a person. The system cannot be falsified from within, because every possible disconfirming signal has been pre-assigned a meaning that confirms it.

It attacks the instrument, not the argument

Ordinary disagreement addresses claims. Coercive systems address the person making them: your perception, your motives, your stage of development, your unresolved history. This is the deepest injury and the hardest to repair, because the faculty you would use to notice what is happening is precisely the faculty being disabled.

It is usually gentle

Popular imagination requires drama — fences, weapons, a siege. Most harm in most high-control groups is done with warmth: a concerned look, a gentle reframe, a teaching that arrives exactly when your doubt does, an invitation that quietly stops arriving. Nothing in the room ever looks like coercion. The coercion is in the accumulated shape of the year.

3. The six domains, in detail

The questionnaire measures six domains. They are presented in the order in which they typically consolidate, though real groups rarely proceed so tidily.

Domain 1 — Authority & leadership

What it looks like. Leaders are described as uniquely enlightened, chosen, awakened, or simply further along than anyone else could reasonably expect to get. Decisions are announced rather than explained. Questioning leadership is not forbidden — that would be crude — it is simply unbecoming, a sign of where the questioner still is. Loyalty is recoded from a practical judgement about a person’s trustworthiness into a moral or spiritual virtue.

Why it works. Once authority is framed as a property of a person’s state of being rather than of their role, it becomes unfalsifiable. You cannot dispute a decision without disputing their level, and you cannot dispute their level without revealing yours.

The healthy contrast. Authority attaches to a role, is bounded by the role, is exercised transparently, and can be lost. Leaders can be wrong out loud without the whole structure trembling.

Domain 2 — Control of information

What it looks like. Members are steered away from “negative” sources. Critical accounts are dismissed without examination — typically by classifying the source rather than answering the content: they are bitter, they were never really committed, they misunderstood, they have an agenda. Only sanctioned interpretations of the teaching are treated as valid, and the difference between a real question and a disloyal one is never written down.

Why it works. Modern groups rarely need to block information; the internet made that impossible. Instead they inoculate: they tell you in advance what you will find and what it will mean, so that when you encounter it, the encounter confirms the teaching. Pre-emptive framing is far more efficient than censorship, because the member does the censoring themselves and experiences it as discernment.

The healthy contrast. A group secure in its own value can tolerate you reading its critics. It may disagree with them specifically and in public.

Domain 3 — Suppression of dissent

What it looks like. Raising a concern produces social cost rather than argument: a subtle chill, a defensive response, a reputation for being difficult, an unexplained drop in inclusion. Critical feedback meets defensiveness rather than discussion. Harmony is prioritised over addressing harm — which sounds like a virtue and functions as a gag.

Why it works. Explicit punishment creates martyrs and evidence. Ambient cooling creates neither. Nobody can point to the moment they were punished, so nobody can appeal it, and the member is left to conclude that the coldness is a response to something defective in them.

The healthy contrast. Concerns are routed somewhere, answered, and sometimes acted on. People who raise them are not less included afterwards. There is a known process, and it has, at least once, produced a change.

Domain 4 — Reverse responsibility

What it looks like. Harm is returned to the harmed person as information about them. Your distress is the real problem. The leader’s mistake was a lesson arranged for your growth. The failure of a promised outcome reflects your insufficient commitment, openness or surrender. Accountability travels away from authority as reliably as water travels downhill.

Why it works. It converts every piece of disconfirming evidence into fuel. A group that can do this reliably is, in the short term, unbreakable: the worse it performs, the more spiritual work its members are told they have to do. It also produces a devastating private experience — the injured person ends up apologising, and cannot say why that felt wrong.

The healthy contrast. Responsibility is proportional to power. When something goes wrong at the top, the people at the top say so, in words, to the people affected. See the full definition under Terminology.

Domain 5 — Social & emotional control

What it looks like. Outside relationships are devalued — not forbidden, just gently characterised as unconscious, draining, or unable to understand. Belonging is conditional and everyone can feel the condition. Fear, guilt and shame do the work that rules would otherwise have to do. And leaving is defined in advance as a moral failure, a regression, a loss of nerve.

Why it works. It makes exit maximally expensive by bundling it. Leaving the group is not one decision; it is simultaneously losing your friends, your meaning, your identity, your explanation of your own history, and your standing as a good person. Most people, facing that bill, stay — and then must find reasons why staying is right.

The healthy contrast. People leave, are wished well, and are still spoken to. This single indicator carries more weight than almost anything else on this page.

Domain 6 — Identity & dependency

What it looks like. Personal identity fuses with the group. Ordinary life decisions — work, relationships, where to live, whether to have children — start requiring approval, or the anticipation of approval. The prospect of leaving produces not sadness but disorientation, as if the floor were being removed. The group comes to feel like the only place where truth or safety exists.

Why it works. This is the end state the other five domains build toward. Once the group holds your framework for interpreting reality, evaluating the group requires standing outside a place you no longer have an outside of.

The healthy contrast. You can describe who you are without mentioning the group. You have relationships, opinions and plans that the group has no view on.

4. Frameworks that describe this

None of this is new. Six bodies of work underpin the material on this site; if you want to go deeper, these are the places to go, and the reading list gives full references.

Lifton’s eight criteria for thought reform (1961)

The psychiatrist Robert Jay Lifton, studying prisoners and students subjected to “thought reform” in 1950s China, identified eight themes present in totalist environments: milieu control (control of communication and environment); mystical manipulation (engineered experiences presented as spontaneous or providential); the demand for purity (an absolute moral binary the member always fails); the cult of confession (disclosure as surveillance); sacred science (a doctrine treated as both scientifically true and morally beyond question); loading the language (thought-terminating clichés that end enquiry); doctrine over person (the teaching outranks your actual experience); and the dispensing of existence (the group decides who counts as a real human being). Sixty-five years on, it remains the most precise description of the terrain.

Hassan’s BITE model

Steven Hassan, a former member of the Unification Church turned mental health counsellor, organises undue influence into four domains: control of Behaviour, Information, Thought and Emotion. The model’s value is practical — it converts a vague unease into an inventory you can actually check. Hassan also popularised the Influence Continuum, discussed below, and the distinction between an authentic self and an adopted group self that recovering members often describe as having spoken through them.

Singer’s conditions for thought reform

The clinical psychologist Margaret Singer, who interviewed thousands of former members, described the conditions a coercive programme needs: keep the person unaware of the agenda and of the changes being made; control their time and environment; create a sense of powerlessness; suppress old behaviours and attitudes; instil new ones; and maintain a closed system of logic in which criticism is proof of the critic’s defect. That last condition is this site’s central subject.

Lalich’s bounded choice (2004)

The sociologist Janja Lalich, herself a former member of a political cult, answers the question that puzzles outsiders most: why do intelligent people make choices that are obviously against their interests? Her answer is that they are making genuinely free choices — inside a frame that has been shaped so that all available options serve the group. Charismatic authority, a transcendent belief system, systems of control and systems of influence combine into a “self-sealing system” in which the member is neither a robot nor a free agent. This is the most humane account in the literature, and the one most former members recognise.

Stein’s attachment account (2016)

Alexandra Stein applies attachment theory to totalist groups. The leader or group is simultaneously the source of threat and the only permitted source of comfort — a structure that in infants produces disorganised attachment and in adults produces what she calls “fright without solution”: a state of chronic arousal in which clear thinking is neurologically difficult and the person clings harder to the source of the fear. It explains, better than any account of belief, why leaving is so physically hard.

Coercive control as a legal concept

Evan Stark’s work reframed intimate abuse as a “liberty crime” — a pattern of control and deprivation of autonomy rather than a series of assaults. Several jurisdictions have since criminalised controlling or coercive behaviour, though almost always within intimate or family relationships rather than group settings. The concept still matters here: it establishes in law the idea that a pattern of control can constitute serious harm even where no single act does. That is exactly the situation most people on this site are trying to describe.

5. A continuum, not a category

It is more accurate — and far more useful — to place a group on a continuum than to sort it into a box. At one end sit relationships and organisations characterised by ethical influence: informed consent, transparency about costs, freedom to question, freedom to leave, accountability that runs upward. At the other end sits undue influence: consent manufactured through deception and dependency, questioning penalised, exit ruinous, accountability running downward.

Most real groups sit somewhere in between, and — this is the part that matters — they move. A community can drift toward the coercive end as a founder ages, as money arrives, as a succession crisis approaches, or as criticism mounts. A group that was genuinely good to you in 2019 can be dangerous in 2026 without anyone announcing the change.

Two practical implications follow. First, your memory of the group being good is probably accurate and is not a reason to doubt your present perception. Second, the question is never settled; it is worth asking again periodically about any structure you are inside.

6. How it actually happens

Nobody joins a cult. People join a meditation class, a training, a church, a startup, a therapy group, a movement, a relationship. What they join is usually genuinely good at first — that is not manipulation, it is the actual product, and it is why the memory is confusing later.

The mechanism is gradual commitment. Each step is small, and each is justified by the one before it. You attend a weekend. It helps. You attend a longer one. You take on a role. The role requires more time. Time with outsiders shrinks — not by instruction but by arithmetic. Your closest friends are now inside. You have made statements, donated money, brought people in. And each of those investments makes re-evaluating the group more expensive, because re-evaluating it means re-evaluating everything you already did.

Then the conditions tighten. Perhaps a scandal, perhaps a decline, perhaps just a leader with an appetite. When the group needs more, the members who have most is the ones it asks. The people who are hurt worst are almost always the most committed, most capable and most generous — the ones who gave the most and therefore had the most to protect by staying.

7. Six persistent misconceptions

“It only happens to vulnerable or gullible people.”

Recruitment favours the capable: people with social skills, income, energy and idealism are more useful. Studies of former members repeatedly find average or above-average education. What raises risk is not a defect of character but a moment of transition — a move, a bereavement, a breakup, a new city, an illness, a search for meaning. That is to say: ordinary life.

“They’re brainwashed — they can’t think.”

Members think constantly and often brilliantly. What has changed is not the capacity to reason but the premises reasoning starts from, and the cost attached to certain conclusions. Treating someone as though their mind is gone is both false and, in practice, the fastest way to lose their trust.

“If it were really that bad, they’d leave.”

Leaving means losing your entire social world, your framework of meaning, your account of your own past decisions, and often your income and housing, in a single week, while being told the distress you feel proves you were wrong to go. People stay in far less coercive situations for far less reason.

“It has to be religious.”

The mechanisms are indifferent to content. See below.

“There has to be a charismatic leader.”

Usually, but not always. Some systems are leaderless and hold their members through doctrine, group pressure and an internalised process alone. Leaderless does not mean accountability-free; the question remains where responsibility travels when someone is hurt.

“If I got something valuable from it, it can’t have been coercive.”

This is the hardest one, and it deserves a direct answer: both things are usually true. Most people did receive something real — skills, friendships, relief, meaning, a period of genuine flourishing. Coercive systems work precisely because they deliver. Holding the value and the harm at the same time, without cancelling either, is one of the central tasks of recovery.

8. Beyond religion

The same six domains appear, in recognisable form, well outside anything anyone would call a religion:

• Therapy and personal-growth organisations, where the framework that is supposed to help you evaluate your life is supplied and interpreted by the organisation itself, and where doubt is readily reinterpreted as resistance.
• Wellness, yoga and healing communities, where a teacher’s spiritual authority extends without limit into students’ bodies, finances and relationships, and where injury is reliably explained as the student’s process.
• Multi-level marketing and business “masterminds”, where the structure guarantees that failure is individual and success is systemic — a pure inversion of responsibility, financialised.
• Political and activist formations of any orientation, where moral urgency is used to license internal control and to make dissent a form of betrayal.
• Workplaces with mission language, extreme hours, an inspirational founder, aggressive loyalty norms and a habit of characterising departures as failures of commitment.
• Individual relationships. Every mechanism on this page can be run by one person against one other person. The literature calls this a cult of one; the law calls it coercive control.

9. An honest note on a contested field

Intellectual honesty requires saying this plainly: scholars disagree about parts of this subject, and a site that pretended otherwise would be doing the same thing it criticises.

The main dispute concerns “brainwashing.” A body of sociological work on new religious movements — Eileen Barker’s study of Unification Church recruitment is the best-known example — found that most people approached by such groups do not join, and that most of those who do join leave of their own accord within a couple of years. That research is a genuine corrective to a mid-century picture of irresistible mind control, and it is why the crude version of “brainwashing” has been largely abandoned. Critics of the field, in turn, note that studying only current members systematically undercounts those most harmed, and that voluntary entry tells you nothing about the conditions under which someone later stays.

This site takes the position that the argument about entry has been allowed to obscure the more important question about conditions. Whether a member walked in freely is largely irrelevant to whether the arrangement they are now inside permits them to question it, correct it, or leave it without ruin. Coercive control is a description of a relationship over time, not a claim about a single moment of consent. That framing — pattern rather than event, autonomy rather than belief — is the one used throughout this site, and it is why we measure behaviour rather than doctrine.