THE PSYCHOLOGY OF CULTS: WHY PEOPLE FOLLOW BLINDLY

Imagine a room where everyone applauds a claim you find troubling. Nothing prevents you from remaining silent. Yet silence suddenly feels conspicuous. You glance at the people who welcomed you, helped you and made you feel understood. The question changes: not simply whether the claim is true, but what disagreeing might cost. This is an imagined scene, not testimony from a survivor. Its tension opens a real scientific problem.

People can conform, obey or defend a group while still thinking intensely. What narrows is sometimes the freedom to test conclusions, access alternatives or act on doubt. This investigation, with sources checked on 12 October 2026, examines how ordinary social needs can intersect with extraordinary control. Its title is a question about behaviour, not a diagnosis of everyone who joins an unconventional community.

Belonging is a need, not a defect

Friendship, faith, political action and shared interests can provide meaning and mutual protection. Roy Baumeister and Mark Leary's 1995 review argued that forming enduring interpersonal attachments is a fundamental human motivation. That helps explain why acceptance matters; it does not establish that loneliness automatically produces recruitment into abusive groups. People may join for service, knowledge, ambition, companionship or convictions they sincerely hold. Baumeister and Leary, 1995

The word cult is contested and frequently used as an insult. Unusual beliefs, intense commitment or a minority religion are not sufficient evidence of abuse. Here, high-control group refers to a setting where sustained pressure, restrictions and exploitation undermine meaningful choice. The decisive questions concern conduct: who can challenge decisions, what information is permitted, and whether people can refuse or leave without threats. Healthy belonging does not require surrendering those freedoms.

Asch: when the obvious becomes socially expensive

In Solomon Asch's line-judgment experiments, a participant heard other people give the same wrong answer before responding. The others were cooperating with the experimenter. In his 1955 account, involving 123 male college students, participants matched the misleading majority on 36.8% of critical judgments. Errors without that pressure were under 1%. These are proportions of answers in a particular experiment, not percentages of humanity susceptible to brainwashing. Asch’s original account, 1955

The experiment also revealed independence. Participants did not all yield, and breaking the majority's unanimity reduced conformity. Public agreement could reflect avoiding embarrassment rather than privately accepting the answer. The task was brief, artificial and unlike prolonged captivity; the sample cannot represent every gender, culture or historical period. Nevertheless, it isolates something consequential: social agreement can make a plainly visible fact harder to state aloud.

Normative influence means adjusting behaviour to gain acceptance or avoid rejection. Informational influence means treating others as evidence about reality, especially under uncertainty. They can overlap. Herd mentality is a loose popular label for these processes, not a single brain mechanism. Checking what everyone else thinks can be sensible when they possess expertise. It becomes dangerous when the apparent consensus is manufactured and dissent carries penalties.

Milgram: obedience, identification and resistance

Stanley Milgram's 1963 report described 40 men in a supposed learning experiment. They were instructed to administer increasing electric shocks to a learner. The shocks were simulated, but the participants were not initially told that. In this condition, 26 men, or 65%, reached the apparatus's highest marked level, 450 volts. Some displayed marked distress. The finding concerns this procedure and sample, not a measured universal willingness to kill. Milgram’s original report, 1963

Changing the situation changed the outcome across Milgram's programme. Distance, institutional context and other participants' behaviour mattered. Ethical problems include deception, distress and pressure to continue. Researchers also debate how fully participants believed the situation and what they understood the experimenter to be guaranteeing. Collapsing these questions into a claim that authority switches off conscience discards the very uncertainty the experiment needs us to examine.

In a 2012 PLOS Biology essay, S. Alexander Haslam and Stephen Reicher challenged the passive-obedience interpretation. They argued that identifying with a scientific mission helped explain cooperation: participants could understand persistence as serving something worthwhile. This engaged-followership account is an interpretation supported by a research programme, not the final explanation of every session. It changes the question from why people stop thinking to whose purpose they believe they are advancing. Haslam and Reicher, 2012

The moment I becomes we

Henri Tajfel and colleagues' 1971 minimal-group experiments showed that even arbitrary categorisation could favour in-group allocations. Later social identity theory, associated with Tajfel and John Turner, explored how membership becomes part of the self. These allocation tasks did not prove that all groups become hostile, or that a trivial label inevitably creates violence. They demonstrated a route by which a distinction between us and them can acquire psychological importance. Tajfel et al., 1971

Tribalism can therefore be a misleading shortcut if it suggests an ancient instinct explains every modern conflict. Identities are shaped by institutions, narratives and circumstances. In a coercive setting, criticism of a leader may be reframed as an attack on everyone's dignity. Defending the organisation becomes defending oneself. The same capacity for identification can support solidarity, whistleblowing and resistance. Group identity is not inherently the enemy of independent judgement.

Charisma and the architecture of control

A confident leader can offer an appealing account of confusing events and a role within it. Charisma alone is neither a crime nor a reliable test for abuse. The danger grows when personal authority becomes unaccountable: only the leader can interpret truth, mistakes become followers' failures, and ordinary boundaries are dismissed as obstacles to a greater mission. The relevant distinction is between persuasion that can be rejected and pressure that makes rejection costly.

Research on group psychological abuse makes these mechanisms more concrete. A 2015 taxonomy, reviewed by 31 experts through a Delphi procedure, addressed isolation, information control, control over personal life, emotional abuse, indoctrination and imposed authority. Expert agreement helps organise concepts; it is not a population survey or proof that every organisation using similar language is abusive. Repeated tactics, their context and their consequences matter more than a dramatic label. Rodríguez-Carballeira et al., 2015

Consider how restrictions can accumulate. Access to friends becomes conditional; private doubts must be confessed; money or work becomes difficult to separate from membership. A person may still make choices, but within an increasingly constrained field. Warmth and recognition can coexist with humiliation or threats. Indoctrination here means persistent attempts to make an ideology unquestionable, not simply teaching a belief. Criticism becomes evidence of disloyalty instead of information worth considering.

Why intelligence does not guarantee immunity

Being skilled at mathematics, law or medicine is not the same as having complete information about a new social setting. Neither professional accomplishment nor confidence establishes immunity to deception. This does not mean intelligence increases susceptibility, and there is no defensible single profile covering everyone affected. Asking why a person stayed requires looking at what they knew, what was concealed and which relationships or practical resources were at stake.

Leon Festinger and James Carlsmith's 1959 experiment illustrates one mechanism of self-justification. Students performed tedious tasks and some were paid $1 or $20 to describe them positively. Those given the smaller reward later rated the tasks more favourably, consistent with reducing dissonance when external justification was weaker. The study tested a narrow laboratory prediction, not prolonged group coercion; alternative explanations and subsequent research belong to the wider debate. Festinger and Carlsmith, 1959

In everyday terms, admitting that a trusted community caused harm may conflict with years of effort, public commitments and cherished relationships. That tension can invite rationalisation. It can also prompt reconsideration. Cognitive dissonance is not a trap with only one exit, and money already spent is not evidence that further sacrifice is worthwhile. Sometimes the obstacle is less philosophical: housing, employment, family ties or threats may make departure genuinely difficult.

Groupthink is a warning, not a universal law

Irving Janis popularised groupthink in 1972 to explain how striving for agreement might undermine careful decisions. Self-censorship, suppressed alternatives and an illusion of unanimity remain useful warning signs. But the model's historical case studies and causal claims have been disputed. A 1998 critique by Ramon Aldag and Sally Fuller emphasised limited empirical support for the phenomenon as commonly presented. Aldag and Fuller’s critique, 1998

That distinction matters in boardrooms, movements and laboratories alike. A disastrous decision does not retrospectively prove that everyone shared one mental state. Leaders can restrict information, incentives can discourage objections, and people can disagree privately while appearing united. A group that actively invites contrary evidence and protects dissent is structurally different from one that treats every challenge as betrayal. Agreement becomes more trustworthy when disagreement remains possible.

Documented harm, without a single survivor story

The NXIVM case provides legal evidence of coercion rather than a speculative diagnosis. Keith Raniere was convicted in 2019 and sentenced in October 2020 to 120 years for racketeering and sex-trafficking offences. The US Department of Justice described how the secret DOS organisation collected compromising collateral used to enforce obedience. These documented crimes must not be attributed indiscriminately to everyone who attended NXIVM courses. US Justice Department: Raniere sentencing

Jonestown offers another caution about language. On 18 November 1978, Congressman Leo Ryan was murdered during a visit investigating Peoples Temple abuses in Guyana; mass deaths followed. The FBI describes murder and suicide, not a simple account of unanimous voluntary choice. Children and coercion make that distinction essential. Neither tragedy supplies a representative sample of religious communities. Historical investigation must preserve individual circumstances rather than turning victims into a parable of foolishness. FBI: Jonestown investigation

Online communities: similarity is not diagnosis

Digital belonging can deliver support across enormous distances. It can also make approval visible and criticism punishable through exclusion, harassment or threats. In a 2021 PNAS study, Matteo Cinelli and colleagues analysed interactions across Facebook, Twitter, Reddit and Gab, finding platform differences in homophily and information diffusion. Such observational patterns describe echo chambers; they do not demonstrate that algorithms caused coercive control or that every tightly connected community is a cult. Cinelli et al., PNAS 2021

The boundary is behavioural. Are members encouraged to check claims elsewhere, or told outsiders are inherently corrupt? Can moderators be questioned? Are money, sexual material or personal secrets demanded as proof of loyalty? A political forum, spiritual community or fandom can share strong preferences without practising abuse. The label becomes analytically useful only when it directs attention to evidence of domination, not when it merely condemns people we dislike.

What research can and cannot establish

A 2019 study developed an emotional-distress measure using 706 Spanish-speaking former group members, separated according to reported psychological abuse. It contributed tools for studying survivors' experiences. Retrospective accounts and selected samples cannot establish population prevalence, isolate every causal pathway or determine an individual's diagnosis. Former members' reports are important evidence, but different recruitment methods and comparison groups can change what conclusions are justified. Saldaña et al., 2019

Experiments, interviews, court records and survivor surveys answer different questions. Laboratory studies isolate influence; legal proceedings establish particular offences; clinical research investigates distress. No single test calculates how cult-like a community is with scientific certainty. Our investigation of hypnosis likewise distinguishes influence from a magical ability to possess a mind. Explaining social pressure is not excusing exploitation, and recognising resistance is not blaming people who could not safely resist.

Recognising manipulation and recovering room to think

Look for patterns rather than a charismatic voice or unfamiliar ritual. Warning signs include punitive responses to questions, enforced secrecy, interference with outside relationships, escalating demands and threats around leaving. Ask concrete questions: what happens after a refusal, who audits the leader, and can a claim be checked independently? One sign alone needs context. A consistent pattern that removes alternatives deserves serious attention, whatever the group's ideology.

Regaining autonomy may begin with small, practical freedoms: an unmonitored conversation, time away from meetings, access to documents or contact with someone outside the organisation. Compare claims with independent evidence and separate disagreement from personal failure. If threats, violence or blackmail are involved, prioritise safety and seek qualified local support before confrontation. These are general principles, not a validated self-treatment protocol or a demand to leave abruptly.

For relatives, ridicule may make discussion feel like another contest of loyalties. A more respectful approach leaves room for conversation, acknowledges what membership provided and asks about specific restrictions. Recovery can involve rebuilding relationships, finances and confidence, sometimes with a qualified mental-health professional. Distress varies; being a former member is not itself a psychiatric diagnosis. Replacing one unquestionable authority with another would miss the purpose of restoring choice.

A community strong enough to tolerate doubt

The disturbing lesson is not that humanity is destined to obey. It is that sincere ideals and meaningful relationships can be used to narrow the world in which decisions are made. The hopeful lesson is equally empirical: unanimity can break, participants can refuse, and collective identity can organise resistance. Our exploration of good and evil considers how moral behaviour depends on circumstances without being reducible to them.

We cannot live entirely outside influence, and we would lose much of what makes life valuable if we tried. The more demanding freedom is to belong without allowing belonging to become ownership. A healthy community need not fear a member who asks for evidence or takes time to think. Perhaps the deepest test of any promise of liberation is whether it leaves us free to say no.