Reference

The frameworks behind the tools

35 named models, each with its originator, its elements, where the science is contested, and the citation. The encyclopedia, the detector, the Defense Playbook, and every AI prompt draw on this same list, so what a page says and what a tool applies never drift apart.

Definitions

Definitions

What Manipulation Is (Working Definition)

Robert Noggle; Cass Sunstein; Daniel Susser, Beate Roessler & Helen Nissenbaum · 2019

Persuasion Lab uses a convergent definition from moral philosophy: manipulation is influence that works by bypassing or subverting a person's capacity for rational deliberation — rather than by giving them reasons, evidence, or honest emotional appeals they can evaluate. Deception, exploitation of known biases or vulnerabilities, and concealment of the influence attempt are the three usual routes.

  1. Sunstein's test. An action is manipulative to the extent that it does not sufficiently engage or appeal to the person's capacity for reflection and deliberation (Sunstein, "Fifty Shades of Manipulation", 2016).
  2. Noggle's test. Manipulation is the attempt to lead someone astray from the ideals of belief, desire, and emotion that the manipulator would themselves acknowledge — e.g. inducing a belief you know to be false, or an emotion you know to be unwarranted (Noggle 1996).
  3. Susser, Roessler & Nissenbaum. Online manipulation is hidden influence that exploits decision-making vulnerabilities; its defining feature is covertness — the target cannot see the lever being pulled (2019).
  4. The publicity condition. A practical proxy: if the influence attempt would fail once the target understood exactly how it worked, it is probably manipulation. Legitimate persuasion survives disclosure.
  5. Not the same as: emotion. Emotional appeals are not manipulation when the emotion is warranted by the facts and the facts are available. Manipulation is the mismatch between the feeling induced and the reality.
  6. Not the same as: nudging. A nudge that is transparent, easy to opt out of, and in the person's own interest is not manipulation; the same architecture used covertly against the person's interest is "sludge" (Sunstein 2021).
Used forThe harm-class taxonomy (neutral / dual-use / manipulative / abuse) and every "where the line is" sentence in the encyclopedia are applications of this definition.
  1. Sunstein, C. R. (2016). Fifty Shades of Manipulation. Journal of Marketing Behavior, 1(3-4), 213-244
    The deliberation-bypass test for manipulation.
  2. Noggle, R. (1996). Manipulative Actions: A Conceptual and Moral Analysis. American Philosophical Quarterly, 33(1), 43-55
    Manipulation as leading someone astray from ideals the manipulator acknowledges.
  3. Susser, D., Roessler, B., & Nissenbaum, H. (2019). Online Manipulation: Hidden Influences in a Digital World. Georgetown Law Technology Review, 4(1), 1-45
    Covertness and vulnerability-exploitation as the core of online manipulation.
  4. Sunstein, C. R. (2021). Sludge: What Stops Us from Getting Things Done and What to Do about It. MIT Press
    Sludge as the manipulative inversion of nudging.

Ethics tests

Ethics tests

TARES Test for Ethical Persuasion

Sherry Baker & David Martinson · 2001

Five conditions a persuasive message must satisfy to count as ethical. Written for advertising and public relations, it generalizes to advocacy, political, and campaign communication. Persuasion Lab runs every campaign self-audit against it.

  1. Truthfulness of the message. Are the claims true — not merely technically true, but not designed to create a false impression (no paltering, no card stacking)?
  2. Authenticity of the persuader. Do you sincerely believe in what you are advocating, and are your motives and identity what they appear to be (no astroturf, no hidden sponsorship)?
  3. Respect for the persuadee. Does the message treat the audience as rational agents entitled to decide — or does it bypass their judgment through fear, urgency, or exploited vulnerability?
  4. Equity of the appeal. Is the appeal fair given the power and information asymmetry between you and the audience? Targeting children, the grieving, or the financially desperate fails this test.
  5. Social responsibility. Does the message serve or at least not harm the wider public interest — social trust, public health, democratic process?
Used forCampaign Integrity self-audit, Content Drafts, Marketing Funnel, Narrative Designer, and the ETHICS_SYSTEM_PROMPT rating logic (green/yellow/red maps to TARES pass/partial/fail).
  1. Baker, S., & Martinson, D. L. (2001). The TARES Test: Five Principles for Ethical Persuasion. Journal of Mass Media Ethics, 16(2-3), 148-175 · link
    The five principles and their original formulation.

Taxonomies

Taxonomies

Persuasion Lab Harm Classes

Filament: Persuasion Lab editorial standard · 2026

Every technique in the encyclopedia belongs to exactly one class. The class drives the badge on the page, the "where the line is" requirement, the refusal logic in the AI tools, and the severity weighting in the Manipulation Detector.

  1. Neutral. Legitimate communication craft; describing it does not describe a manipulation (tricolon, kairos, steelmanning, signposting).
  2. Dual-use. Legitimate when disclosed and grounded in real value; manipulative past a specifiable line that the entry must state (anchoring, scarcity, social proof, fear appeals with real threat and real efficacy).
  3. Manipulative. Deceptive by design; no honest deployment exists (astroturfing, paltering, dark patterns, fallacies deployed as tactics, scams). Described for recognition only.
  4. Abuse. Inherently harmful interpersonal control; no legitimate use (gaslighting, coercive control, grooming). Descriptive only; support resources render on the page; AI tools refuse to help deploy against a real person.
Used forBadges, detector severity, prompt refusals, campaign self-audit inventory.
  1. Filament: Persuasion Lab (2026). CONTENT_STANDARD.md, section 2: Harm classification. Repository documentation
    The four classes and the rules attached to each.
Taxonomies

The Seven Propaganda Devices

Institute for Propaganda Analysis (Clyde Miller and colleagues) · 1937

The first popular taxonomy of propaganda techniques, published to teach American students to recognize the devices used by domestic demagogues and foreign broadcasters alike. Still the most teachable starting set.

  1. Name-calling. Attaching a bad label to an idea or person so the audience rejects it without examining evidence.
  2. Glittering generalities. Attaching virtue words (freedom, justice, progress) so the audience accepts without examining evidence.
  3. Transfer. Carrying the authority or prestige of something respected (flag, church, science) over to something else to make it acceptable.
  4. Testimonial. Having a respected or hated person say that an idea or product is good or bad.
  5. Plain folks. Convincing the audience that the speaker and their ideas are "of the people".
  6. Card stacking. Selecting facts and falsehoods, illustrations and distractions, to give the best or worst possible case.
  7. Bandwagon. "Everybody — at least all of us — is doing it"; the appeal to follow the crowd.
Used forPropaganda Pattern Recognition tool, the Manipulation Detector's propaganda bucket, and the Political domain of the encyclopedia.
  1. Institute for Propaganda Analysis (1937). How to Detect Propaganda. Propaganda Analysis, 1(2)
    The seven devices as originally published.
  2. Sproule, J. M. (1997). Propaganda and Democracy: The American Experience of Media and Mass Persuasion. Cambridge University Press
    History and reception of the IPA program.
Taxonomies

The Firehose of Falsehood

Christopher Paul & Miriam Matthews (RAND Corporation) · 2016

A model of contemporary state propaganda distinguished by four features that together overwhelm rather than persuade. Useful for recognizing when the goal of a message stream is exhaustion and cynicism rather than belief.

  1. High-volume and multichannel. Many messages across many outlets, creating an impression of consensus and making rebuttal impossible to scale.
  2. Rapid, continuous, and repetitive. First-mover advantage on any story; repetition exploits the illusory-truth effect.
  3. No commitment to objective reality. Fabrications, partial truths, and contradictions coexist; the sender does not need to be believed, only to make truth feel unknowable.
  4. No commitment to consistency. Contradictory narratives are pushed simultaneously; the audience is invited to pick whichever fits, or to give up.
Used forPropaganda analysis and Counter-Campaign design; the model's own recommended counters (do not try to rebut every drop; raise the cost of lying; prebunk; redirect attention to the behaviour, not each claim) inform the counter prompt.
  1. Paul, C., & Matthews, M. (2016). The Russian "Firehose of Falsehood" Propaganda Model: Why It Might Work and Options to Counter It. RAND Corporation, PE-198-OSD · link
    The four features and the recommended counter-strategies.
Taxonomies

The 4 Ds of Disinformation

Ben Nimmo · 2015

A compact description of the moves an influence operation makes when confronted with an inconvenient fact. Recognizing the move is often faster than fact-checking the content.

  1. Dismiss. Attack the messenger or the evidence as fake, biased, or beneath notice.
  2. Distort. Twist the facts, invent a rival version, or reframe the story.
  3. Distract. Change the subject; raise a different accusation; whataboutism.
  4. Dismay. Threaten or frighten the audience or the critic into silence.
Used forPropaganda analysis, source-credibility profiling, and the detector's response-pattern heuristics.
  1. Nimmo, B. (2015). Anatomy of an Info-War: How Russia's Propaganda Machine Works, and How to Counter It. StopFake / Central European Policy Institute
    The original 4D formulation.
Taxonomies

ABC Framework for Disinformation (Actor, Behavior, Content)

Camille François · 2019

Separates three questions that get conflated in public debate: who is behind a campaign (manipulative actors), how it spreads (deceptive behavior such as fake accounts and coordination), and what it says (harmful content). Platforms act mostly on behavior; that is why an operation can be removed without anyone ruling on whether its claims were true.

  1. A — Manipulative actors. State or non-state operators who conceal their identity or intent (attribution question).
  2. B — Deceptive behavior. Sockpuppets, bots, coordinated posting, engagement farming, impersonation — detectable without judging the content.
  3. C — Harmful content. Falsehoods, incitement, targeted harassment — the part that is hardest to police without touching legitimate speech.
Used forExplains why "coordinated inauthentic behavior" is defined by behavior; used in the Propaganda Pattern tool and the Campaign Integrity audit (your campaign must pass on A and B, not only C).
  1. François, C. (2019). Actors, Behaviors, Content: A Disinformation ABC. Transatlantic Working Group on Content Moderation Online and Freedom of Expression
    The three-part framework and the argument for behavior-based enforcement.
Taxonomies

White, Grey, and Black Propaganda

Wartime information doctrine (Allied and Axis practice, codified in postwar scholarship) · 1945

A classification by source attribution rather than by truth. Persuasion Lab's ethical-influence rule is simple: only white influence — correctly attributed, truthful, and non-coercive — is permitted in any campaign the tools help build.

  1. White. Source correctly identified; information generally accurate. Public diplomacy, public health campaigns, disclosed advocacy.
  2. Grey. Source unattributed or ambiguous; truth status mixed. Front groups, unlabeled placements, "sources say" laundering.
  3. Black. Source falsely attributed, typically to the adversary; content usually fabricated. Forged documents, impersonation accounts, false-flag messaging.
Used forThe ethical gate in Campaign, Narrative, and PSYOP-framework prompts ("white only"); the Propaganda Pattern tool's source-analysis field.
  1. Jowett, G. S., & O'Donnell, V. (2018). Propaganda & Persuasion (7th ed.). SAGE
    Standard treatment of the white/grey/black distinction.
  2. Rid, T. (2020). Active Measures: The Secret History of Disinformation and Political Warfare. Farrar, Straus and Giroux
    Historical cases of grey and black operations.
Taxonomies

The Influence Continuum

Information-operations doctrine; popularized in strategic communication teaching · 2010

A spectrum from informing to coercing. The test at each step is how much of the audience's own judgment is engaged. Anything at the coercion end is flagged red in every Persuasion Lab tool.

  1. Inform. Accurate facts, presented so the audience can decide.
  2. Influence. Shape perception through emphasis, framing, and presentation — still truthful.
  3. Persuade. Argue for a change of attitude or behavior with reasons and appeals the audience can weigh.
  4. Manipulate. Bypass judgment through deception, concealed influence, or exploited vulnerability.
  5. Coerce. Compel through threat, force, or removal of real options.
Used forPSYOP Frameworks tool (maps each campaign element to a position), harm-class assignment, and the detector's coercion bucket.
  1. Paul, C. (2011). Strategic Communication: Origins, Concepts, and Current Debates. Praeger
    The inform-influence-persuade distinction in strategic communication.
Taxonomies

Ten Stages of Genocide

Gregory Stanton (Genocide Watch) · 1996

A process model in which dehumanizing rhetoric is a recognizable stage, not an aberration. Included because the language patterns (vermin, disease, infestation, "they are not like us") are detectable long before violence and are the strongest reason to treat dehumanization as a distinct technique.

  1. 1. Classification. "Us and them" categories.
  2. 2. Symbolization. Names and symbols assigned to the classified group.
  3. 3. Discrimination. Law and custom deny rights to the group.
  4. 4. Dehumanization. The group is equated with animals, vermin, insects, or disease — overcoming the normal human revulsion against murder.
  5. 5. Organization. Militias, special units, plans.
  6. 6. Polarization. Extremists drive the groups apart; moderates are silenced.
  7. 7. Preparation. Plans, euphemisms ("cleansing", "purification"), and training.
  8. 8. Persecution. Victims identified, separated, and expropriated.
  9. 9. Extermination. Mass killing, legally "extermination" because the victims are not considered fully human.
  10. 10. Denial. Evidence destroyed, witnesses intimidated, blame placed on the victims.
Used forSeverity weighting for dehumanizing and eliminationist language in the detector; the political domain's dehumanization and euphemism entries.
  1. Stanton, G. H. (2016). The Ten Stages of Genocide. Genocide Watch · link
    The ten-stage model (expanded from the 1996 eight-stage version).

Doctrine

Doctrine

Influence-Operation Kill Chain (DISARM-style phases)

DISARM Foundation (successor to AMITT, Credibility Coalition / SJ Terp & Pablo Breuer) · 2022

An open framework, modeled on cyber-security kill chains, that describes disinformation incidents as a sequence of phases with tactics and techniques under each. Persuasion Lab uses the phase structure descriptively so analysts can say where in a campaign's lifecycle a piece of content sits — and so campaign self-audits can check each phase for integrity.

  1. Plan. Set objectives, define the target audience, and decide the narratives to push.
  2. Prepare. Build assets: accounts, personas, pages, front groups, content libraries; establish legitimacy.
  3. Execute. Seed content, pump and amplify it, drive it across platforms, exploit real events, deliver to the target.
  4. Assess. Measure reach and effect; adapt; persist or conceal and withdraw.
Used forPropaganda pattern analysis (where in the chain is this?), Counter-Campaign timing (prebunk before Execute; expose during Prepare), and the ethical self-audit (an honest campaign has no concealed Prepare phase).
Caveats
  • DISARM is a descriptive vocabulary for defenders, not an operations manual; Persuasion Lab uses only the phase and behavior descriptions, never technique-level playbooks for deception.
  1. DISARM Foundation (2022). DISARM Framework (Disinformation Analysis and Risk Management). disarm.foundation · link
    Phase structure and the defender-oriented purpose of the framework.
Doctrine

Influence-Campaign Planning Stages (educational MISO/PSYOP structure)

US Army FM 3-53 / JP 3-13.2 (Military Information Support Operations) · 2013

The professional planning sequence that military and government influence practitioners use. Persuasion Lab teaches the structure because it is the same structure a public-health, advocacy, or marketing campaign needs — and because knowing the stages helps citizens recognize a planned operation. Each stage carries an ethics gate that an ethical (white) campaign must pass.

  1. 1. Planning and objectives. Define the behavior change sought and the measurable end state. Ethics gate: is the objective one you could state publicly?
  2. 2. Target audience analysis (TAA). Who can produce the behavior; their conditions, vulnerabilities, susceptibilities, and accessibility. Ethics gate: no targeting of vulnerabilities that bypass judgment (grief, addiction, cognitive impairment, minors).
  3. 3. Series and product development. Themes to stress, themes to avoid, lines of persuasion, products (posters, videos, posts, speeches). Ethics gate: TARES on every product; truthful and attributed.
  4. 4. Approval. Independent review before release. Ethics gate: someone not on the campaign signs off on truthfulness and harm.
  5. 5. Production and distribution. Channels, timing, cadence, messengers. Ethics gate: no sockpuppets, no undisclosed paid amplification.
  6. 6. Dissemination. Delivery to the audience. Ethics gate: the audience can tell it is a campaign message and who sent it.
  7. 7. Evaluation (MOE / MOP). Measures of effectiveness (did behavior change?) and measures of performance (did we execute the plan?). Ethics gate: report honestly, including unintended harms.
Used forPSYOP Frameworks tool, Campaign Planner, Campaign Integrity audit, and the "Influence Operations" learning path.
Caveats
  • Persuasion Lab teaches the stages and their ethics gates; it does not produce covert, unattributed, or deceptive products at any stage (AUP sections 1-2).
  • Doctrine documents are periodically revised; the seven-stage structure above is the enduring skeleton, not a citation of a specific edition's paragraph numbering.
  1. US Department of the Army (2013). FM 3-53: Military Information Support Operations. Headquarters, Department of the Army
    The seven-phase MISO process and TAA components.
  2. US Joint Chiefs of Staff (2011). JP 3-13.2: Military Information Support Operations. Joint Publication
    Joint doctrine definitions of MISO, target audience, and measures of effectiveness.

Clinical and high-control models

Clinical and high-control models

Lifton's Eight Criteria of Thought Reform

Robert Jay Lifton · 1961

Derived from interviews with people subjected to Chinese "thought reform" in the 1950s; still the reference checklist for evaluating whether a group environment is totalistic. No single criterion is decisive; the more that apply, the more totalistic the environment.

  1. Milieu control. Control of information and communication, both external and internal (what members may read, who they may talk to, even what they may think about).
  2. Mystical manipulation. Experiences are engineered to seem spontaneous or divine, proving the group's special mission.
  3. Demand for purity. The world is sharply divided into pure and impure; members must constantly strive for an unattainable perfection, producing guilt and shame that the group then manages.
  4. Cult of confession. Sins and doubts are confessed publicly; privacy is abolished and confessions become leverage.
  5. Sacred science. The group's doctrine is beyond question; questioning it is immoral and unscientific at once.
  6. Loading the language. Thought-terminating clichés and in-group jargon compress complex questions into slogans and mark outsiders.
  7. Doctrine over person. When experience contradicts doctrine, the person's experience is wrong.
  8. Dispensing of existence. The group decides who has the right to exist (socially, spiritually, sometimes literally); outsiders and ex-members are shunned or worse.
Used forThe high-control domain of the encyclopedia (one entry per criterion), the Intervention Planner, and the "High-Control Groups" learning path.
  1. Lifton, R. J. (1961). Thought Reform and the Psychology of Totalism: A Study of "Brainwashing" in China. W. W. Norton
    The eight criteria, chapter 22.
Clinical and high-control models

The BITE Model of Authoritarian Control

Steven Hassan · 1988

Organizes control tactics into four domains. Designed to be used by families and former members to assess a group or relationship; the 2018 revision aligns it with Lifton, Singer, and coercive-control research.

  1. Behavior control. Regulating where people live, what they eat and wear, sleep, finances, time; rewards and punishments; requiring permission for major decisions.
  2. Information control. Deception, restricting access to outside sources, compartmentalizing insider and outsider doctrine, encouraging spying and reporting.
  3. Thought control. Loaded language, black-and-white thinking, thought-stopping techniques (chanting, denial), rejecting critical questions, hypnotic and trance-induction practices.
  4. Emotional control. Manipulating and narrowing feelings; making members feel that problems are always their own fault; phobia indoctrination about leaving; love bombing; shunning.
Used forHigh-control entries (one per domain), Relationship Dynamics analysis, Intervention Planner.
  1. Hassan, S. (2018). Combating Cult Mind Control: The Guide to Protection, Rescue, and Recovery from Destructive Cults (30th anniversary ed.). Freedom of Mind Press
    The four BITE domains and their components.
Clinical and high-control models

Singer's Six Conditions for Thought Reform

Margaret Thaler Singer · 1995

A process description of how a person is moved without their awareness: the conditions that must be in place for coercive persuasion to work, which is also a checklist for spotting it.

  1. Keep the person unaware. They do not know there is an agenda to control or change them.
  2. Control time and environment. Contacts, information, and schedule are controlled.
  3. Create powerlessness, fear, and dependency. While supplying models of the new behavior the leader wants.
  4. Suppress old behavior and attitudes. Punish or ridicule the old self.
  5. Instill new behavior and attitudes. Reward the new self.
  6. Closed system of logic. No dissent; criticism is turned back on the critic.
Used forIntervention Planner readiness assessment; recruitment-at-transitions entry.
  1. Singer, M. T. (1995). Cults in Our Midst: The Hidden Menace in Our Everyday Lives. Jossey-Bass
    The six conditions.
Clinical and high-control models

Coercive Control (Stark)

Evan Stark · 2007

Reframes domestic abuse as an ongoing pattern of domination — isolation, micro-regulation of daily life, degradation, surveillance, and deprivation — rather than discrete incidents of violence. Now a criminal offence in England and Wales (Serious Crime Act 2015, s.76), Scotland, Ireland, and several other jurisdictions.

  1. Isolation. From friends, family, work, and independent sources of support and reality-testing.
  2. Micro-regulation. Rules about food, dress, movement, money, phone use; punishment for infractions.
  3. Degradation. Humiliation, name-calling, gaslighting; the erosion of self-respect.
  4. Surveillance. Monitoring of location, communications, and contacts — increasingly through devices (digital coercive control).
  5. Deprivation and exploitation. Financial control, withholding necessities, exploiting labor.
  6. Intimidation. Threats to the person, children, pets, immigration status, or reputation.
Used forAbuse-class entries, AbuseResources banner, the safety-first rules in every abuse-context countermeasure.
  1. Stark, E. (2007). Coercive Control: How Men Entrap Women in Personal Life. Oxford University Press
    The coercive-control model and its components.

Verification methods

Verification methods

SIFT — Four Moves for Evaluating Online Information

Mike Caulfield · 2019

A fast, teachable habit that replaces the slow and often useless "read closely and think hard" approach with the lateral-reading behavior professional fact-checkers actually use.

  1. Stop. Before sharing or reacting, notice the feeling the content produced and pause.
  2. Investigate the source. Leave the page; find out who is behind it and what others say about them (lateral reading).
  3. Find better coverage. Search for the claim itself and see whether trusted sources report it — and how.
  4. Trace claims, quotes, and media to the original context. Find the original study, full quote, or unedited image; check date and context.
Used forDefense Playbook, the Manipulation Detector's "what to do next" advice, source-credibility prompts, and the disinformation learning path.
  1. Caulfield, M. (2019). SIFT (The Four Moves). Hapgood (hapgood.us) · link
    The four moves as originally stated.
  2. Wineburg, S., & McGrew, S. (2019). Lateral Reading and the Nature of Expertise: Reading Less and Learning More When Evaluating Digital Information. Teachers College Record, 121(11), 1-40
    Fact-checkers read laterally and outperform historians and students who read vertically.
Verification methods

CRAAP Test

Sarah Blakeslee, Meriam Library, California State University, Chico · 2004

A library-instruction checklist for evaluating a source on its face. Useful as a vocabulary, but Persuasion Lab teaches SIFT/lateral reading first, because CRAAP-style close reading of a page is exactly what sophisticated fakes are designed to pass.

  1. Currency. When was it published or updated? Does the topic require recency?
  2. Relevance. Does it answer the question at the right level and for the right audience?
  3. Authority. Who wrote it; what are their credentials and affiliations; who publishes it?
  4. Accuracy. Is it supported by evidence, reviewed, verifiable, and free of obvious errors?
  5. Purpose. To inform, sell, persuade, or entertain? Are biases and sponsorship disclosed?
Used forSource Credibility prompt vocabulary.
Caveats
  • Checklist evaluation of a single page is weaker than lateral reading; use CRAAP as vocabulary, SIFT as method (Wineburg & McGrew 2019).
  1. Blakeslee, S. (2004). The CRAAP Test. LOEX Quarterly, 31(3)
    Origin of the acronym and criteria.

Persuasion theories

Persuasion theories

Inoculation Theory and Prebunking

William McGuire (1961); Sander van der Linden & Jon Roozenbeek (2017-present) · 1961

People can be made resistant to persuasion the way bodies are made resistant to pathogens: forewarn them, expose them to a weakened form of the manipulation, and let them practice refuting it. Technique-based inoculation (teaching the tactic rather than the specific claim) generalizes across topics.

  1. Forewarning. Tell people they are about to be targeted; this alone raises resistance (Petty & Cacioppo 1977).
  2. Weakened dose. Show a diluted example of the manipulative technique.
  3. Refutational pre-emption. Give, or let people generate, the counter-arguments.
  4. Technique-based (logic-based) inoculation. Teach the manipulation technique itself — false dichotomy, scapegoating, emotional language, incoherence — so the resistance transfers to new claims (Roozenbeek et al. 2022).
  5. Booster. Effects decay over weeks; brief reminders restore them.
Used forThe academic banner over the defense side of the product; Counter-Campaign inoculation strategy; defensive roleplay mode; the learning paths.
  1. McGuire, W. J. (1961). Resistance to persuasion conferred by active and passive prior refutation of the same and alternative counterarguments. Journal of Abnormal and Social Psychology, 63(2), 326-332
    The original inoculation experiments.
  2. Roozenbeek, J., van der Linden, S., Goldberg, B., Rathje, S., & Lewandowsky, S. (2022). Psychological inoculation improves resilience against misinformation on social media. Science Advances, 8(34) · link
    Technique-based prebunking videos improved manipulation recognition in a YouTube field experiment.
  3. Roozenbeek, J., & van der Linden, S. (2019). Fake news game confers psychological resistance against online misinformation. Palgrave Communications, 5, 65 · link
    The Bad News game as active inoculation.
Persuasion theories

Persuasion Knowledge Model

Marian Friestad & Peter Wright · 1994

People develop theories about how persuaders try to influence them and use that knowledge to cope. When a persuasion attempt is recognized as such — "this is a tactic" — evaluation switches on and the tactic loses force. This is the reason the encyclopedia names techniques and the detector labels them.

  1. Persuasion knowledge. The target's beliefs about tactics, their effectiveness, and their appropriateness.
  2. Agent knowledge. Beliefs about the persuader's goals and competence.
  3. Topic knowledge. What the target knows about the subject itself.
  4. The change-of-meaning principle. Once an action is seen as a tactic, its meaning changes and its effect usually diminishes.
Used forJustifies "name the tactic" as the first countermeasure; the detector's labeling design.
  1. Friestad, M., & Wright, P. (1994). The Persuasion Knowledge Model: How People Cope with Persuasion Attempts. Journal of Consumer Research, 21(1), 1-31 · link
    The model and the change-of-meaning principle.
Persuasion theories

Elaboration Likelihood Model (ELM)

Richard Petty & John Cacioppo · 1986

Persuasion runs on two routes. The central route (careful evaluation of arguments) produces durable, resistant attitude change; the peripheral route (cues such as attractiveness, authority, number of arguments, mood) produces weaker change. Manipulation is largely the business of keeping people on the peripheral route.

  1. Motivation to elaborate. Personal relevance, need for cognition, accountability.
  2. Ability to elaborate. Time, distraction, prior knowledge, cognitive load — all of which manipulators reduce.
  3. Central route. Argument quality determines outcome; attitudes formed here persist and predict behavior.
  4. Peripheral route. Cues determine outcome; attitudes formed here are temporary and easily reversed.
Used forExplains why urgency, overload, and distraction are manipulation multipliers; used in the Analyzer's logic/emotion scoring rationale.
  1. Petty, R. E., & Cacioppo, J. T. (1986). The Elaboration Likelihood Model of Persuasion. Advances in Experimental Social Psychology, 19, 123-205
    The two-route model.
Persuasion theories

Extended Parallel Process Model (fear appeals)

Kim Witte · 1992

Fear appeals work only when perceived efficacy is high: the audience believes the threat is real AND that the recommended action works and they can do it. High threat with low efficacy produces fear control — denial, avoidance, reactance — not the intended behavior. This is the scientific basis for the "where the line is" on fear appeals.

  1. Perceived severity. How bad the threat is.
  2. Perceived susceptibility. How likely it is to affect me.
  3. Response efficacy. Whether the recommended action actually works.
  4. Self-efficacy. Whether I can actually do it.
  5. Danger control vs fear control. High threat + high efficacy → protective action. High threat + low efficacy → defensive avoidance. Fear without a path to safety is manipulation.
Used forAppeal-to-fear entry, public-health campaign guidance in the Campaign prompts, detector explanation text.
  1. Witte, K. (1992). Putting the fear back into fear appeals: The extended parallel process model. Communication Monographs, 59(4), 329-349 · link
    The model and the efficacy condition.
  2. Tannenbaum, M. B., et al. (2015). Appealing to fear: A meta-analysis of fear appeal effectiveness and theories. Psychological Bulletin, 141(6), 1178-1204 · link
    Fear appeals are effective on average, most so with efficacy statements and one-time behaviors.
Persuasion theories

Moral Foundations Theory

Jonathan Haidt, Jesse Graham, Craig Joseph · 2004

Moral judgment draws on several intuitive foundations that different political cultures weight differently. Its persuasion application — moral reframing — is one of the few cross-partisan techniques with experimental support: arguments framed in the audience's foundations move them more than arguments framed in the speaker's.

  1. Care / harm. Compassion for the vulnerable; sensitivity to cruelty.
  2. Fairness / cheating. Proportionality and reciprocity; resentment of free-riders (and, in some versions, equality).
  3. Loyalty / betrayal. Group solidarity; sacrifice for the group; treason.
  4. Authority / subversion. Respect for tradition and legitimate hierarchy.
  5. Sanctity / degradation. Purity, disgust, the sacred.
  6. Liberty / oppression. Resentment of domination and bullies (proposed later).
Used forPolicy Framing prompt, moral-reframing countermeasure, voter-segment literacy.
Caveats
  • The number and independence of foundations are debated; the framework is a useful map, not a settled theory of moral psychology.
  1. Haidt, J. (2012). The Righteous Mind: Why Good People Are Divided by Politics and Religion. Pantheon
    Accessible statement of the theory.
  2. Feinberg, M., & Willer, R. (2015). From Gulf to Bridge: When Do Moral Arguments Facilitate Political Influence?. Personality and Social Psychology Bulletin, 41(12), 1665-1681 · link
    Moral reframing experiments.
Persuasion theories

Cialdini's Principles of Influence

Robert Cialdini · 1984

Seven (originally six) shortcuts that reliably trigger compliance, each tied to a social function that makes it useful most of the time — and exploitable the rest of the time. Every entry in the psychological and social domains cross-references these.

  1. Reciprocity. We repay what we receive.
  2. Commitment and consistency. We act in line with prior public commitments.
  3. Social proof. Under uncertainty we do what similar others do.
  4. Authority. We defer to credible expertise and its symbols.
  5. Liking. We say yes to people we like, who are similar, who compliment us, who cooperate with us.
  6. Scarcity. We value what is rare or dwindling.
  7. Unity. We are influenced by those we share an identity with (added in Pre-Suasion, 2016).
Used forCross-referencing throughout the encyclopedia; the scam-triad signature in the detector (authority + scarcity/urgency + reciprocity or unity).
  1. Cialdini, R. B. (2007). Influence: The Psychology of Persuasion (Revised ed.). Harper Business
    The six principles.
  2. Cialdini, R. B. (2016). Pre-Suasion: A Revolutionary Way to Influence and Persuade. Simon & Schuster
    The unity principle and privileged moments.
Persuasion theories

The Story Model of Juror Decision Making

Nancy Pennington & Reid Hastie · 1986

Jurors do not weigh evidence item by item; they construct a story from the evidence, their world knowledge, and their expectations about what makes a complete story, then pick the verdict category that best fits. The story that is most coherent, complete, and consistent with the evidence wins — which is why trial persuasion is theme-and-narrative work and why evidence presented in story order is more persuasive than the same evidence presented witness by witness.

  1. Story construction. Coverage of the evidence, coherence (consistency, plausibility, completeness), and uniqueness.
  2. Verdict-category learning. Understanding the legal categories from instructions.
  3. Story classification. Matching the accepted story to the best-fitting verdict; confidence tracks goodness of fit.
Used forLegal Argument coach; theory-of-the-case and trial-theme entries; defense guidance for jurors (ask whether a story is compelling because of evidence or because of coherence alone).
  1. Pennington, N., & Hastie, R. (1992). Explaining the evidence: Tests of the Story Model for juror decision making. Journal of Personality and Social Psychology, 62(2), 189-206 · link
    Story order vs witness order experiments.
  2. Pennington, N., & Hastie, R. (1986). Evidence evaluation in complex decision making. Journal of Personality and Social Psychology, 51(2), 242-258
    The original story-model account.

Argumentation models

Argumentation models

Toulmin Model of Argument

Stephen Toulmin · 1958

A structure for laying out any practical argument so that its weak point is visible. Most manipulation in argument hides in the warrant (the unstated rule that gets you from evidence to claim) or in an absent qualifier.

  1. Claim. What you are asking the audience to accept.
  2. Grounds (data). The evidence offered.
  3. Warrant. The general principle that licenses the move from grounds to claim — usually unstated, usually where the argument fails.
  4. Backing. Support for the warrant itself.
  5. Qualifier. "Probably", "in most cases" — the strength claimed. Absence of any qualifier is a tell.
  6. Rebuttal. The conditions under which the claim would not hold — an honest arguer states them.
Used forDebate and Legal Argument coaches; argument-mapping and name-the-missing-premise countermeasures; the Analyzer's logic score.
  1. Toulmin, S. E. (1958). The Uses of Argument. Cambridge University Press
    The six-part model.
Argumentation models

Ten Rules for Critical Discussion (Pragma-Dialectics)

Frans van Eemeren & Rob Grootendorst · 1984

A normative model in which every classic fallacy is a violation of one of ten rules for resolving a difference of opinion reasonably. Useful because it turns "that's a fallacy" into "which rule did that break, and what does the discussion need to get back on track?"

  1. Freedom rule. Parties must not prevent each other from advancing or questioning standpoints (violations: threats, ad hominem, declaring topics sacred).
  2. Burden-of-proof rule. Whoever advances a standpoint must defend it when asked (violation: shifting the burden).
  3. Standpoint rule. Attacks must address the standpoint actually advanced (violation: straw man).
  4. Relevance rule. Defend with relevant argumentation (violations: red herring, appeals to pity or force).
  5. Unexpressed-premise rule. Do not deny your own implicit premises or invent your opponent's.
  6. Starting-point rule. Do not present a premise as agreed when it is not (violation: begging the question).
  7. Argument-scheme rule. Use appropriate, correctly applied schemes (violations: hasty generalization, false analogy, post hoc).
  8. Validity rule. Formal reasoning must be valid (violations: affirming the consequent, etc.).
  9. Closure rule. If a defense fails, withdraw the standpoint; if it succeeds, withdraw the doubt (violation: moving the goalposts).
  10. Usage rule. Be clear and interpret charitably (violations: equivocation, ambiguity).
Used forDebate coach fallacy audit and the "which rule did this break" explanation style in the detector's fallacy bucket.
  1. van Eemeren, F. H., & Grootendorst, R. (2004). A Systematic Theory of Argumentation: The Pragma-Dialectical Approach. Cambridge University Press
    The ten rules and the fallacies-as-violations account.

Checklists

Checklists

Vulnerability Windows

Synthesis of recruitment research (Singer, Lalich), scam-victimization data (FTC, AARP), and stress physiology · 2026

The states and moments in which people are measurably more susceptible to influence — and which manipulators, recruiters, and scammers deliberately target. Persuasion Lab uses the list two ways: as a self-check for the reader, and as a hard rule in the AI tools (campaigns may not target these states).

  1. Life transitions. New city, new job, first year of college, bereavement, divorce, retirement, illness — the ordinary support network is thin and identity is in flux (classic cult-recruitment windows).
  2. Loneliness and social isolation. Romance scams and high-control groups offer belonging first and demands later.
  3. Grief. Psychic and medium scams, "inheritance" fraud, and predatory funeral or financial pitches cluster around loss.
  4. Financial distress. Predatory lending, job scams, "guaranteed" investment pitches, MLM recruitment.
  5. Acute stress, fear, or time pressure. Narrowed cognition and reliance on heuristics; the reason every scam manufactures urgency.
  6. Sleep deprivation, hunger, intoxication. Reduced executive control; retreat formats and marathon sessions exploit this (HALT: hungry, angry, lonely, tired).
  7. Cognitive impairment and age-related change. Elder financial exploitation; undue influence in estate matters.
  8. Youth and inexperience. Grooming, loot-box mechanics, engagement design aimed at minors.
  9. Identity threat. When a group identity feels under attack, people accept in-group claims with less scrutiny and out-group claims with more (identity-protective cognition).
Used forAudience Profiler and Campaign prompts refuse targeting of these states; the Manipulation Detector notes when a message appears aimed at one; the Neuro & Human Nature learning path.
  1. Lalich, J., & Tobias, M. (2006). Take Back Your Life: Recovering from Cults and Abusive Relationships. Bay Tree Publishing
    Recruitment at life transitions.
  2. Federal Trade Commission (2024). Consumer Sentinel Network Data Book 2023. FTC · link
    Fraud-loss patterns by age and contact method.
  3. Arnsten, A. F. T. (2009). Stress signalling pathways that impair prefrontal cortex structure and function. Nature Reviews Neuroscience, 10(6), 410-422 · link
    Stress and the shift from deliberative to reflexive control.
Checklists

The Debunking Handbook 2020 — Core Rules

Stephan Lewandowsky, John Cook, Ullrich Ecker, and 19 co-authors · 2020

A consensus summary of what the evidence says about correcting misinformation. It replaced the 2011 edition's emphasis on the backfire effect (which later research found to be rare) with a practical structure for effective corrections.

  1. Prebunk when you can. Inoculation (forewarning + weakened dose + refutation) before exposure is more effective than debunking after.
  2. Debunk often and properly. Corrections generally work; the "backfire effect" is not a reason to stay silent.
  3. Fact. Lead with the true fact, stated plainly and memorably.
  4. Warn about the myth. Mention the myth once, clearly labeled as false — do not lead with it, do not repeat it in the headline.
  5. Explain the fallacy. Say how the myth misleads (which technique: cherry-picking, fake experts, impossible expectations, conspiracy, logical fallacy).
  6. Fact again. End with the truth, so it is the last thing remembered.
  7. Provide an alternative explanation. A correction that leaves a causal gap is weaker; fill the gap (continued influence effect).
Used forCounter-Campaign design, the Defense Playbook, and the ethical-rewrite feature in the Analyzer.
  1. Lewandowsky, S., Cook, J., Ecker, U. K. H., et al. (2020). The Debunking Handbook 2020. Databrary / Center for Climate Change Communication · link
    The fact-myth-fallacy-fact structure and the reassessment of the backfire effect.
  2. Wood, T., & Porter, E. (2019). The Elusive Backfire Effect: Mass Attitudes' Steadfast Factual Adherence. Political Behavior, 41, 135-163 · link
    Corrections generally move beliefs toward accuracy; backfire is rare.
Checklists

Rapoport's Rules (Steelmanning)

Anatol Rapoport, as formulated by Daniel Dennett · 2013

How to compose a successful critical commentary — and the fastest way to tell a good-faith argument from a bad-faith one: the good-faith arguer can pass step one.

  1. 1. Re-express. Restate your opponent's position so clearly, vividly, and fairly that they say "I wish I'd put it that way."
  2. 2. Agree. List every point of agreement, especially if they are not matters of general consensus.
  3. 3. Learn. Mention anything you have learned from your opponent.
  4. 4. Only then rebut. Now, and only now, say a word of criticism.
Used forDebate coach, steelman-before-refuting countermeasure, roleplay scoring for empathy and clarity.
  1. Dennett, D. C. (2013). Intuition Pumps and Other Tools for Thinking. W. W. Norton
    The four rules as stated in chapter 3.
Checklists

Ten Commandments of Cross-Examination

Irving Younger · 1975

The classic discipline for controlling a hostile witness. Included both as legitimate craft and as a recognition guide: several commandments are also how skilled interrogators and manipulators control a conversation outside court.

  1. Be brief. Short cross; three points the jury will remember.
  2. Short questions, plain words. No compound questions; no jargon.
  3. Leading questions only. Every question is a statement the witness can only confirm.
  4. Never ask a question you do not know the answer to. Cross is for confirmation, not discovery.
  5. Listen to the answer. Do not read the next question while the witness is speaking.
  6. Do not quarrel with the witness. Argument invites the witness to explain.
  7. Do not let the witness repeat direct testimony. Repetition reinforces.
  8. Do not let the witness explain. No "why" questions.
  9. Avoid the one question too many. Stop when the point is made; save the conclusion for closing.
  10. Save the ultimate point for summation. The jury should draw the inference themselves.
Used forLegal Argument coach; one-fact-per-question countermeasure; the note in the interpersonal domain that "leading questions only, never let them explain" is also a manipulation signature when used in ordinary conversation.
  1. Younger, I. (1975). The Art of Cross-Examination. American Bar Association, Section of Litigation (monograph)
    The ten commandments as commonly transmitted.
Checklists

PEACE Model of Investigative Interviewing

UK Home Office and police services (England and Wales) · 1992

The ethical, information-gathering alternative to accusatorial interrogation. Adopted after wrongful-conviction scandals; meta-analytic evidence finds it elicits as many true confessions with far fewer false ones. Included as the reference model for "getting the truth without manipulation", applicable to journalism, HR investigations, and hard conversations.

  1. Preparation and planning. Know the facts, define objectives, plan the interview.
  2. Engage and explain. Rapport; explain the process and the person's rights.
  3. Account, clarification, challenge. Free recall first; open questions; probe; only then put inconsistencies to the person.
  4. Closure. Summarize, confirm, explain next steps.
  5. Evaluation. Assess the information and the interviewer's own performance.
Used forThe ethical counterpoint to the Reid technique in the legal domain; the Conversation Architect prompt; a defense (recognize when you are being interviewed accusatorially).
  1. Meissner, C. A., Redlich, A. D., Michael, S. W., Evans, J. R., Camilletti, C. R., Bhatt, S., & Brandon, S. (2014). Accusatorial and information-gathering interrogation methods and their effects on true and false confessions: a meta-analytic review. Journal of Experimental Criminology, 10, 459-486 · link
    Information-gathering approaches reduce false confessions without reducing true ones.
  2. Kassin, S. M., et al. (2010). Police-Induced Confessions: Risk Factors and Recommendations. Law and Human Behavior, 34(1), 3-38 · link
    Risk factors for false confession under accusatorial interrogation.
Checklists

Baloney Detection Kit

Carl Sagan · 1995

A working scientist's habits for skeptical thinking, written for the public. Persuasion Lab keeps it because it is the most widely known checklist and because its items map directly onto encyclopedia entries.

  1. Independent confirmation. Seek independent confirmation of the facts wherever possible.
  2. Substantive debate. Encourage debate by knowledgeable proponents of all points of view.
  3. Arguments from authority carry little weight. Authorities have made mistakes; there are no authorities, only experts.
  4. Spin more than one hypothesis. Then test each.
  5. Do not get overly attached to your hypothesis. Ask why you like it; compare fairly.
  6. Quantify. If it can be measured, measure it; vague claims are hard to test.
  7. Every link in the chain must work. Including the premise.
  8. Occam's razor. Prefer the hypothesis that explains the data with fewer assumptions.
  9. Falsifiability. Ask whether the hypothesis can, at least in principle, be shown wrong.
Used forDefense Playbook cross-reference; roleplay debrief prompts.
  1. Sagan, C. (1995). The Demon-Haunted World: Science as a Candle in the Dark. Random House
    Chapter 12, "The Fine Art of Baloney Detection".
Checklists

HALT Self-Check

Recovery-community heuristic (Alcoholics Anonymous tradition); consistent with stress and sleep research · 1970

Before making a decision or responding to pressure, check whether you are Hungry, Angry, Lonely, or Tired. Each state reduces deliberative control and raises susceptibility; each is also a state manipulators try to create or exploit (marathon sales sessions, late-night calls, retreats with restricted sleep and food).

  1. Hungry. Low glucose and discomfort narrow attention; "let me eat first" is a legitimate delay.
  2. Angry. Anger increases certainty and risk-taking and is the emotion outrage bait targets.
  3. Lonely. Belonging offers land harder; the recruitment and romance-scam window.
  4. Tired. Sleep loss impairs judgment and increases suggestibility and false-memory formation.
Used forVulnerability windows; the Neuro & Human Nature learning path; detector advice when urgency is detected ("do not decide while flooded").
Caveats
  • HALT is a folk heuristic, not a validated instrument; the physiological claims behind each item are supported separately (see the neuro domain entries).
  1. Frenda, S. J., Berkowitz, S. R., Loftus, E. F., & Fenn, K. M. (2016). Sleep deprivation and false confessions. Proceedings of the National Academy of Sciences, 113(8), 2047-2050 · link
    Sleep-deprived participants were far more likely to sign a false confession.

See the frameworks applied: every encyclopedia entry names its harm class and cites its sources.

Browse the encyclopediaThe science