{
  "id": "conformity",
  "title": "Conformity",
  "lang": "en",
  "source_lang": "en",
  "summary": "Conformity is the alignment of judgments, beliefs, or behavior with those of other people or with perceived group norms.",
  "category": "society",
  "subcategory": "psychology",
  "type": "concept",
  "aliases": [
    "Social conformity"
  ],
  "generated_at": "2026-11-02",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/conformity/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/conformity/",
      "json": "https://aiwiki.page/zh/conformity.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "Conformity is a change in judgment, belief, or behavior toward the positions expressed or implied by other people. It may involve publicly agreeing with a group despite private disagreement, accepting others’ judgments as evidence about reality, or adopting practices associated with a valued social identity. In social psychology, conformity is studied as a form of social influence; in sociology and anthropology, it is also examined as a process through which shared expectations and institutions are reproduced.\n\nConformity helps explain both ordinary coordination and consequential failures of collective judgment. Shared conventions make communication, cooperation, and social life more predictable, but pressure toward agreement can suppress information, sustain mistaken beliefs, or discourage dissent. Its consequences depend on what people conform to, why they do so, and whether the apparent consensus represents independent evidence. Research therefore treats conformity not simply as a personal weakness or a desirable social habit, but as a family of context-dependent processes connecting individual judgment with group life."
    },
    {
      "heading": "Conceptual boundaries and forms",
      "text": "Conformity concerns alignment with an actual or perceived group position. That position may be communicated directly, as when several people give the same answer, or inferred indirectly from clothing, voting patterns, professional customs, or online popularity indicators. An explicit request is not necessary. Someone entering a quiet reading room may lower their voice without being instructed, responding instead to a social norm inferred from the setting and the behavior of others.\n\nResearchers distinguish conformity from several neighboring concepts. Compliance usually means responding to a request, whereas obedience means following an order from a recognized authority. A person may comply with a colleague’s request without adopting the group’s preferences, or obey a supervisor while knowing that most coworkers disagree. Conversely, conformity can occur among equals and without an identifiable leader. Imitation is broader still: copying another person’s movement or technique need not involve agreement, social pressure, or a judgment about what is appropriate.\n\nA central distinction is between public conformity and private acceptance. Public conformity occurs when expressed behavior moves toward a group position, whether or not private belief changes. Private acceptance involves a corresponding change in what the person considers true, desirable, or appropriate. These outcomes can diverge. A committee member may vote with the majority while remaining unconvinced; a learner may adopt a group’s solution and continue using it after leaving the group. Neither visible agreement nor later persistence alone identifies the entire psychological process.\n\nHerbert Kelman’s 1958 account distinguished compliance, identification, and internalization as processes of attitude change. Compliance depends on obtaining a favorable reaction or avoiding an unfavorable one. Identification depends on maintaining a valued relationship or social identity. Internalization occurs when the adopted position fits the person’s own values or understanding of reality. These categories are not successive stages through which every conforming response must pass, and the same behavior can have more than one basis.\n\nThe relevant standard also varies. Descriptive norms concern what people commonly do; injunctive norms concern what people approve or disapprove of. A behavior may be widespread but condemned, or widely endorsed but rarely practiced. Conformity to one norm can therefore entail departure from another. Apparent nonconformity toward a national majority may simultaneously express strong conformity within a subculture, profession, religious community, or friendship network. Analysis must specify the reference group rather than assume that society contains a single, uniform consensus."
    },
    {
      "heading": "Classical experiments and the emergence of the field",
      "text": "Early experimental work showed that judgments could converge even when no authority demanded agreement. In studies reported in 1935, Muzafer Sherif used the autokinetic effect: a stationary point of light viewed in darkness appears to move because the observer lacks a stable visual frame of reference. Participants estimated how far it moved. Individual estimates initially varied, but judgments made in groups tended to converge toward a shared range. Group-established standards could also persist when participants later judged alone.\n\nSherif’s experiments were important because the task was genuinely ambiguous. There was no visible ruler or surrounding scene against which to check the apparent movement. Other people’s estimates could therefore function as plausible evidence. The findings demonstrated the formation of a group norm rather than merely submission to an obviously incorrect majority. They also raised a lasting question: does agreement reflect a changed perception, a changed interpretation, or a decision about which answer to report?\n\nSolomon Asch addressed a different situation in experiments reported from 1951 onward. Participants compared a standard line with three comparison lines, one of which clearly matched its length. A participant gave answers among several people who were actually cooperating with the experimenter. On designated trials, these confederates unanimously selected an incorrect line before the participant responded. The task was designed so that the answer was ordinarily evident without assistance.\n\nIn Asch’s principal series, involving 123 male college students, participants followed the erroneous majority on approximately 36.8 percent of critical judgments. About three quarters agreed with the incorrect majority at least once, while roughly one quarter remained independent throughout the critical trials. Errors were rare in the comparison condition without a misleading majority. These figures describe particular procedures and samples, not a universal rate at which people abandon their own judgment.\n\nThe contrast between Sherif and Asch became foundational. Sherif demonstrated convergence under uncertainty; Asch showed that unanimous social opposition could alter responses even when the stimulus appeared straightforward. Asch’s interviews also revealed varied experiences: some participants doubted their judgment, some suspected something unusual about the situation, and others knowingly gave an answer they considered wrong to avoid standing apart.\n\nNeither research program established that people are uniformly passive. Asch’s participants resisted on most critical trials, and substantial individual variation remained. The experiments instead provided methods for separating stimulus evidence from social information and for manipulating features such as unanimity, public responding, and group size. Later research extended these questions beyond perceptual judgments to attitudes, memory, consumer choices, and collective decisions."
    },
    {
      "heading": "Informational, normative, and identity-based mechanisms",
      "text": "Morton Deutsch and Harold Gerard’s 1955 distinction between informational social influence and normative social influence remains central to explanations of conformity. Informational influence occurs when other people’s judgments are accepted as evidence about reality. Normative influence occurs when a person aligns with others to meet their expectations or secure social acceptance. These are analytical distinctions: both processes may operate during a single decision.\n\nInformational influence is especially relevant when evidence is ambiguous, unfamiliar, costly to obtain, or difficult to interpret. A newcomer may infer a procedure from experienced colleagues, while a traveler may choose an unfamiliar route after seeing local residents take it. Copying can be efficient when others possess relevant knowledge. Its reliability depends, however, on their competence and on whether their judgments are independent. Ten people repeating one unsupported claim do not provide the same evidential basis as ten independent observations.\n\nNormative influence arises from anticipated social consequences. Agreement may bring approval, trust, continued membership, or access to opportunities; disagreement may risk embarrassment, exclusion, or conflict. Explicit punishment is unnecessary. People can anticipate others’ reactions and regulate their behavior before anyone objects. Public responses are therefore often more susceptible to normative pressure than confidential responses, although confidentiality does not necessarily eliminate internalized expectations or concern about evaluation.\n\nIdentity-based explanations emphasize that people do not treat every group as equally relevant. Under social identity theory and related self-categorization approaches, individuals sometimes understand themselves primarily as members of a group rather than as separate persons. Agreement with an in-group can then seem like an expression of an appropriate shared perspective, not simply a concession made against personal preference. The same individual may adopt different standards when acting as a scientist, family member, sports supporter, or citizen.\n\nThese mechanisms can reinforce one another. A trusted in-group may be both a source of knowledge and a source of belonging. Agreement can initially be strategic, then become privately accepted through repeated exposure and participation. Conversely, someone may privately accept a group’s factual judgment while disliking its social demands. A simple division between rational information use and irrational social pressure therefore misses much of the phenomenon.\n\nConformity also interacts with cognitive dissonance and self-justification. After publicly endorsing a position, a person may become more favorable toward it, especially when the endorsement conflicts with an earlier attitude and lacks a sufficient external explanation. Such change is not inevitable, and dissonance is not required for conformity. It identifies one route through which an initially public adjustment can affect later belief, distinct from learning that the majority was correct."
    },
    {
      "heading": "Group size, unanimity, uncertainty, and commitment",
      "text": "The strength of conformity depends strongly on the structure of the immediate situation. Group size matters, but its effect is not usually proportional to the number of people present. In Asch’s line-judgment studies, increasing the opposing majority from one person to two and then three substantially increased conformity; further increases did not produce equally large increments. Later studies have found that the relationship varies with task, group organization, and whether additional members appear to contribute genuinely independent judgments.\n\nUnanimity is often more influential than sheer numbers. A dissenter can reduce conformity by demonstrating that disagreement is socially possible and that the majority is not the only available reference point. In variations of Asch’s procedure, support from another person markedly reduced yielding to the majority. Even an ally who gave a different wrong answer could weaken majority influence, indicating that disruption of unanimity can matter apart from confirmation of the participant’s own answer.\n\nTask characteristics alter the balance between personal and social evidence. Greater difficulty, ambiguity, or unfamiliarity can increase reliance on others, especially when they appear knowledgeable. A group with recognized expertise may influence judgments more than a larger group without relevant competence. Confidence also matters, but confidence is not equivalent to accuracy: a confident individual can resist sound advice, while an uncertain individual can appropriately revise an initial mistake.\n\nThe conditions of responding help distinguish mechanisms. Giving an answer publicly exposes a person to immediate evaluation, whereas an anonymous response reduces some interpersonal costs. Making an initial judgment before hearing others can create commitment and make later switching less likely. That commitment may protect an independent assessment, but it can also preserve an error. The timing of information therefore affects not only the amount of conformity but also the quality of the resulting decision.\n\nGroup cohesion, status, and interdependence can intensify pressure. Agreement may matter more when members value continued association, depend on one another for rewards, or believe that dissent will affect their standing. Yet closeness can also permit disagreement when relationships are secure and dissent is understood as helpful. A cohesive team is not necessarily a conformist team; the relevant question is which behaviors its shared expectations support.\n\nIndividual differences contribute, but research does not support treating conformity as one fixed trait that operates identically across situations. The same person may resist peers on a familiar technical question and defer to them on an unfamiliar social convention. Motivations such as accuracy, affiliation, and distinctiveness can change with circumstances. Predicting behavior therefore requires attention to the relationship between the person, the group, the task, and the consequences of disagreement."
    },
    {
      "heading": "Development, socialization, and cultural variation",
      "text": "Conformity develops within socialization, through which children learn language, conventions, roles, and expectations. Learning from others is indispensable because most cultural knowledge cannot be rediscovered individually. A child who copies a tool-use method or follows a classroom convention is not necessarily displaying the same process as an adult who publicly endorses an opinion under peer pressure. Developmental research distinguishes learning what works, learning what others prefer, and learning what members of a community regard as obligatory.\n\nStudies of young children show that majority information can influence judgments and behavior, but the effect depends on age, task, and how agreement is presented. Daniel Haun and Michael Tomasello’s 2011 research with preschool children found that public judgments could be influenced by a unanimous peer majority even when the children had access to the correct answer. Comparisons between public and private responses helped distinguish social accommodation from a straightforward loss of the original judgment. Such findings do not imply that children always prefer majorities over direct evidence.\n\nPeer influence becomes especially important during adolescence, when friendships, acceptance, and group membership acquire new significance. Yet influence is domain-specific rather than a uniform surrender of independence. Adolescents may converge with friends in clothing, leisure activities, or attitudes while remaining less influenced in other areas. Apparent similarity also reflects selection: people often choose friends who already resemble them. Separating selection from influence requires observations over time rather than a single comparison of friends’ characteristics.\n\nCultural research examines how conformity relates to shared expectations about autonomy, obligation, hierarchy, and interdependence. Rod Bond and Peter Smith’s 1996 meta-analysis examined 133 Asch-type studies from 17 countries. It reported cross-cultural variation associated with measures of individualism and collectivism, alongside historical variation within the United States. The findings helped establish that conformity rates depend partly on social context rather than constituting a single universal experimental constant.\n\nBroad cultural categories require careful interpretation. Countries contain substantial regional, class, institutional, and generational variation, while samples often consist of students rather than representative populations. Translated instructions, relations between participants and experimenters, and the meaning of public disagreement can also differ. A national average cannot establish how a particular individual will behave, and differences observed in a perceptual task need not generalize directly to political, religious, or family decisions.\n\nCulture also shapes the object of conformity. In some settings, displaying independence is itself a valued norm; members may conform by expressing originality or distancing themselves from a mainstream position. Countercultural communities can likewise maintain strong internal expectations about appearance, taste, or conduct. The contrast between conformity and individuality is therefore not absolute: culturally learned models help define what counts as an independent person and which forms of difference receive recognition."
    },
    {
      "heading": "Dissent, minority influence, and collective decisions",
      "text": "Conformity research initially emphasized majorities, but groups also change because minorities maintain positions that others reject. Minority influence concerns the conditions under which a numerically or socially subordinate position affects the majority. Serge Moscovici and colleagues investigated this experimentally in the late 1960s. In a well-known color-judgment study, a consistent minority describing blue slides as green influenced a portion of majority responses more than an inconsistent minority did.\n\nConsistency can make a minority position salient and suggest commitment, encouraging others to reconsider the issue. Its effects depend on more than repetition, however. A minority perceived as inflexible, irrelevant, or acting only from self-interest may be dismissed. Flexibility, credibility, and the way a position connects with shared concerns can alter its influence. Claims that majorities produce only superficial compliance while minorities uniquely produce deep change are too categorical; both can affect public responses and private processing.\n\nDissent can improve a decision without persuading the group to accept the dissenter’s answer. Research associated with Charlan Nemeth has examined how minority disagreement can stimulate consideration of alternative perspectives and solutions. A group may benefit because members search more broadly, examine assumptions, or distinguish evidence from consensus. This effect is different from simply increasing the number of opinions expressed: the disagreement must enter the group’s reasoning rather than remain an ignored formality.\n\nGroupthink, a concept developed by Irving Janis in the 1970s, describes a proposed pattern in which pressure for concurrence undermines critical evaluation in cohesive decision-making groups. Suggested symptoms include self-censorship, an illusion of unanimity, and pressure on dissenters. The concept has been influential in interpreting organizational and governmental decisions, but its explanatory limits are important. Retrospective case studies can select evidence to fit a known failure, and cohesion by itself does not reliably establish that groupthink occurred.\n\nAnother group outcome is group polarization: discussion can move members toward a more extreme version of their initial average tendency. This differs from simple convergence on a moderate compromise. Explanations include exposure to arguments favoring the initially preferred side and adjustment toward what participants perceive as a valued group position. Whether polarization occurs, and in which direction, depends on the distribution of initial views, the information exchanged, and the social meaning of the issue.\n\nFormal procedures interact with these processes. A jury, scientific panel, or organizational committee may contain unequal status, uneven expertise, and different thresholds for agreement. Unanimity requirements can encourage fuller discussion but also increase pressure on holdouts. Majority voting can settle a decision without producing shared belief. The recorded outcome therefore cannot by itself reveal whether members independently agreed, learned from one another, compromised, or concealed continuing disagreement."
    },
    {
      "heading": "Social learning, information cascades, and networks",
      "text": "In economics, conformity-like patterns often arise in models of social learning. People observe previous actions and use them to infer information they do not directly possess. A restaurant with a queue may appear better than an empty competitor; an investment attracting earlier buyers may seem promising. Following others can be individually reasonable even when it generates a mistaken collective outcome, because each observer sees only part of the information behind earlier decisions.\n\nAn information cascade occurs when the public history of earlier actions becomes sufficiently influential that a decision maker follows it rather than acting on private information. Models developed in the early 1990s by Sushil Bikhchandani, David Hirshleifer, and Ivo Welch, alongside related work by Abhijit Banerjee, demonstrated how sequential choices can produce such patterns. An early sequence generated partly by chance can lead many later actors in the same direction.\n\nA cascade is not identical to normative pressure. Participants need not care about approval or even know one another. The mechanism can arise from inference alone: each person assumes that earlier actors had information favoring their choices. The difficulty is that observed actions may cease to reveal private signals once people begin copying. A long sequence of similar choices can therefore contain less independent evidence than its size suggests. Some cascades are consequently vulnerable to disruption by new public information.\n\nNetwork effects provide another reason for coordinated choices that should not automatically be labeled psychological conformity. A communication service becomes more useful when more contacts use it; a technical standard can become attractive because compatible equipment and expertise are available. People may adopt the same option because its payoff depends on adoption by others, not because they accept a majority judgment or fear rejection. Several mechanisms can nevertheless operate together in real markets.\n\nNetwork structure also matters. A person surrounded by a tightly connected cluster may encounter repeated versions of the same information, whereas someone linking different communities may encounter more independent perspectives. Damon Centola and Michael Macy’s work on complex contagion emphasized that some behaviors require reinforcement from multiple contacts rather than a single exposure. This differs from simple transmission in which one contact may suffice. Social reinforcement can be particularly relevant when adoption is costly, unfamiliar, or socially consequential.\n\nObserved similarity does not establish influence. Friends and neighbors may resemble one another because they selected similar contacts, share circumstances, or face the same incentives. This problem is often described in terms of homophily and common exposure. Distinguishing these explanations requires temporal evidence, experiments, or other credible identification strategies. Without that distinction, the spread of a behavior through a network can be mistaken for conformity when much of the pattern reflects pre-existing similarity."
    },
    {
      "heading": "Institutions, everyday practices, and digital environments",
      "text": "Conformity helps maintain institutional routines. Schools, workplaces, professions, and voluntary associations communicate expectations through training, evaluation, promotion, and everyday interaction. Some expectations are explicit rules; others are tacit, such as when to question a senior colleague or how much uncertainty to acknowledge in a meeting. Participation in education and professional training therefore involves learning both substantive knowledge and the social conventions governing its use.\n\nInstitutional convergence is not always interpersonal imitation. Paul DiMaggio and Walter Powell’s 1983 account of institutional isomorphism distinguished coercive, mimetic, and normative processes through which organizations become more alike. Legal requirements and funding dependencies can impose similar structures; uncertainty can encourage imitation of apparently successful organizations; professional training can spread common methods. Organizational resemblance may thus reflect regulation, dependence, or shared expertise as well as pressure to appear legitimate.\n\nIn scientific work, agreement can result from converging evidence, shared methods, and successful replication. It is therefore not valid to infer conformity merely because specialists reach a consensus. At the same time, professional incentives and expectations can affect which questions receive attention or how readily contrary findings are heard. Peer review and methodological criticism involve social judgment, but their epistemic value depends on the quality of evidence and reasoning, not on agreement alone.\n\nDigital platforms make social information unusually visible. Ratings, likes, repost counts, follower totals, and “trending” labels can influence what users notice and how they evaluate it. A 2006 experiment by Matthew Salganik, Peter Dodds, and Duncan Watts created artificial online music markets. Showing participants others’ download choices increased inequality and unpredictability in song success relative to an independent-choice condition. Quality still mattered, but popularity was not simply a transparent ranking of intrinsic appeal.\n\nA 2013 randomized experiment by Lev Muchnik, Sinan Aral, and Sean Taylor manipulated initial ratings of comments on a social news site. An experimentally assigned positive rating increased subsequent positive evaluation, illustrating how small early signals can affect later collective judgments. Such findings demonstrate influence in specific platform settings; they do not establish that every popular item owes its success to conformity or that negative and positive signals always have symmetrical effects.\n\nOnline agreement is also shaped by ranking systems, moderation, coordinated activity, and selective participation. An apparently unanimous comment thread may reflect which messages an algorithm displays rather than what most users believe. People who expect hostility may remain silent, further distorting visible opinion. In these settings, data privacy, audience size, and persistence of records can alter the perceived costs of dissent. The relevant social group may be a small community rather than the platform’s entire population."
    },
    {
      "heading": "Measurement, research ethics, and unresolved questions",
      "text": "Conformity cannot be measured adequately by agreement alone. Two people may independently reach the same answer, or both may respond to the same evidence. An experimental design therefore usually compares judgments under different social-information conditions, ideally assigning participants randomly. Studies may measure switching from an initial answer, distance moved toward a group estimate, adoption of a majority choice, or persistence after the social source is removed. Each measure captures a different aspect of influence.\n\nSeparating public behavior from private belief remains difficult. Anonymous responses, delayed tests, confidence ratings, and transfer to new tasks can provide evidence about private acceptance, but none is an infallible indicator. A participant can infer the purpose of a study, continue to imagine an audience, or retain social information as part of the available evidence. Self-reports explain how participants interpret their behavior, yet they may not accurately identify all the processes producing it.\n\nClassical paradigms also raise questions about research ethics. Confederates and false majority feedback often require deception, while deliberate social isolation or ridicule can cause distress. Contemporary review procedures consider the necessity of deception, the proportionality of risks, informed-consent arrangements, confidentiality, and debriefing. The scientific value of a design depends partly on whether it can isolate social influence without imposing unnecessary interpersonal harm.\n\nGeneralizability is another continuing issue. Laboratory tasks offer control but often involve brief encounters with strangers, low stakes, and simple judgments. Real conformity may unfold over years in relationships where material dependence, identity, and institutional authority overlap. Field studies offer greater realism but usually less control over competing explanations. Cross-cultural research, longitudinal designs, and behavioral records can extend the evidence, provided that measurement differences and selection effects are addressed.\n\nNeuroscientific research has examined how disagreement with a group relates to conflict, reward, and subsequent adjustment. For example, a 2009 study by Vasily Klucharev and colleagues linked discrepancies between individual and group evaluations with brain responses associated with performance monitoring and later changes in ratings. Such findings connect conformity with more general learning processes. They do not identify a unique “conformity center,” and activity in a brain region cannot by itself establish whether a person sought accuracy, approval, or belonging.\n\nA central unresolved problem is distinguishing useful reliance on others from socially amplified error. The same outward act—changing an answer after seeing a majority—may represent sound learning, strategic accommodation, identity expression, or avoidance of exclusion. Resolving these alternatives requires evidence about information quality, source independence, incentives, and private judgment. Current research increasingly treats these features as interacting components rather than seeking a single explanation for all forms of conformity."
    }
  ],
  "links": [
    "social-psychology",
    "sociology",
    "anthropology",
    "social-norm",
    "subculture",
    "informational-social-influence",
    "cognitive-dissonance",
    "socialization",
    "counterculture",
    "jury",
    "economics",
    "network-effect",
    "education",
    "peer-review",
    "algorithm",
    "data-privacy"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}