Manipulation defense, oriented

Philosophers usually set manipulation apart from rational persuasion, which works by offering reasons, but no definition is settled. One account says manipulation bypasses deliberation; another says it leaves deliberation worse off. Trickery accounts tie it to planting a false belief or an out-of-place emotion, pressure accounts to attaching costs to saying no, and real manipulators often mix both. Coercion is often set apart as interfering with a person's options, while manipulation leaves the options alone and corrupts how the choice gets made.

Dual-process models describe a careful route, where people weigh a message's substance, and a quick route, where cues about the source, such as credentials, carry the decision. Careful checking needs both motivation and ability. In one lab study, people told to judge emails within five minutes told phishing from real messages slightly less accurately. In field experiments, an abrupt swing from fear to relief raised agreement to requests.

The trap is believing only careless people get manipulated. Older adults have not been shown to be overrepresented among online fraud victims, and measured traits predicted phishing errors only modestly. Say Priya, who checks everything at her desk, gets a message while rushing between meetings: she is now in the condition that made lab participants miss more. The opposite error is suspecting manipulation behind every choice, since treating all influence other than argument as wrongful would leave nothing else legitimate.

Rule: ask whether influence works through reasons you can check or around them, and expect checking to weaken under rushing and sudden emotional swings.

Because the reciprocity norm says favors should be returned, a small, unexpected favor before a request can create a sense of obligation. Door-in-the-face opens with a large request and, after a refusal, retreats to a smaller one, a retreat presented as a favor to be returned. Some scams manufacture the favor: a tech support scam starts with a bogus virus warning, then charges to remove the malware it claims to find. US law draws one line plainly: you never have to pay for merchandise you did not order, and you need not send it back.

Foot-in-the-door wins a small yes first, and meta-analyses suggest it raises compliance on average, though the research also holds many failures to replicate. Low-balling wins a decision on attractive terms, then worsens them, counting on the buyer to stick with the choice. The sunk-cost effect lets money or time already spent pull on a decision that economic theory says should weigh only future gains and losses.

The trap is treating an earlier yes as a promise about the next one. Say Marcus decides on a bargain car, and at signing the price has risen. The pull he feels to finish is the commitment the tactic relies on, not a reason to finish. In one scenario study, people asked to justify their choice resisted the sunk-cost pull better.

Rule: treat a favor you did not request, and a yes you already gave, as pressure to notice rather than debts to pay, and judge each request on what it costs from here.

A 2024 Federal Trade Commission (FTC) rule targets manufactured agreement. It bars reviews by people who do not exist or never used the product, paying for reviews that must express a set sentiment, knowingly buying fake followers or views for commercial use, using intimidation to remove negative reviews, and implying the reviews shown are all or most of those submitted when low-rated ones were held back.

Affinity fraud works inside a group. Fraudsters are often members of the group or pretend to be, and many affinity schemes are Ponzi or pyramid schemes. Regulators describe fraudsters convincing respected community or religious leaders that an investment is legitimate and relying on them to spread the word, and many of those leaders become unwitting victims. Problems then tend to stay inside the group, and being told to keep an opportunity confidential is a listed warning sign.

The trap is treating a trusted messenger as a check on the message. Say an elder in a congregation tells members about a fund that has paid him well. He may be sincere and still have been fooled. Securities and Exchange Commission (SEC) guidance is to check out everything no matter how trustworthy the person bringing it seems, including a seller's registration and disciplinary history in its free search.

Rule: a crowd, a page of praise or a trusted messenger is not evidence that a claim is true, so check the claim the same way whoever brings it.

Impostors claim to be from the Internal Revenue Service, Social Security or a utility, then threaten arrest, lost benefits or shutoff unless payment comes at once. Each marker of authority they lean on can be imitated: any number can be spoofed onto caller ID, a phishing email can carry the real company's logo, and an impostor may already know a name, address or Social Security number.

A phishing message usually pairs a borrowed identity with an invented problem and one requested action: click a link, open an attachment or confirm details. Business email compromise uses a hacked or spoofed account, such as an executive's, to instruct finance to wire funds, or a supplier's to send invoices to a new account. The FTC is categorical that anyone who asks for your account verification code is a scammer. A short audio clip can let a cloned voice sound like a family member, so familiarity no longer proves identity.

The trap is checking a contact with the contact's own details. Say Leon gets a call showing his bank's number, and the caller offers a number to call back. That number, and the one on caller ID, verify nothing. The FTC advises hanging up and calling a number known to be correct, such as the one on a statement, and the Federal Bureau of Investigation (FBI) suggests families agree on a secret word in advance.

Rule: never verify a request with contact details the request supplied; end the contact and reach the organization through a channel you start.

Scam scripts start a clock so the target acts before thinking: pay today or face arrest, stay on the line so the story cannot be checked. Secrecy recurs across family-emergency, business-payment and romance scams, because talking to anyone else would expose the lie. Insistence on gift cards, wire transfers, cryptocurrency or payment apps completes the set. No real business or agency demands payment by gift card, and a card's number and personal identification number (PIN) let a scammer spend the balance while the target still holds the card.

The scam threats the FTC describes come with the way out already chosen: pay, in the specific way the caller names. Guilt appeals had a small average persuasive effect in one meta-analysis, and romance scammers may imply that a partner who truly loved them would do what is asked. Sextortion draws force from shame, and reporting stays very low. Health and immigration scams trade on hope, promising a miracle cure or special access. The loss frame alone adds little, though: pooled studies found almost no difference in persuasiveness between gain- and loss-framed appeals.

The trap is assuming a real emergency would forbid checking. Say a caller tells Ruth her grandson is in trouble, begs her to tell no one, and asks for gift cards now. Each instruction exists to prevent the one call that would expose the story. FTC guidance is to resist the pressure, since honest businesses give time to decide, and to tell a friend or relative before doing anything.

Rule: when a request comes with a deadline, a demand for secrecy and a payment that is hard to reverse, read the three together as a script, and pause to tell someone before acting.

FTC pricing guides describe a seller briefly marking a product up, selling few or none, then advertising the usual price as a cut. Some sellers advertise part of a price and add required fees late in the buying process. Describing identical outcomes as lives saved or as deaths shifts people's risk preferences, and when a baseline risk is small, a relative reduction is perceived as larger than the same absolute one. Not every framing effect is large: across 3.6 million UK wine purchases, decoy options moved preference by only about 1%.

Deceptive design builds the same moves into an interface. A roach motel makes signing up easy and leaving hard. Confirmshaming words the decline as a bad decision, such as a line saying the shopper does not want to save money, and a pre-checked box on a free trial can authorize later charges.

At a timeshare pitch, the FTC calls an offer said to be gone tomorrow a warning sign that the seller may want to keep buyers from consulting advisers or taking time to think. A lower monthly payment can come from a longer loan and more total interest, and add-ons such as warranties are optional.

The trap is comparing offers in the terms each seller chose. Say Jun weighs two plans, one quoted monthly with fees added at checkout and one quoted as a total. Only figures in the same terms compare, one all-in price or one shared denominator, which is the idea behind the FTC's fee rule for live-event tickets and short-term lodging.

Rule: an offer you cannot restate as an all-in total, a trial whose cancellation terms you cannot find, and a rush to sign are all signals to stop and compare.

Romance scams follow a rough script. Scammers almost always try to move talk to a private app, then build trust through constant contact, compliments and quick declarations of love, sometimes for months. A partner who always has a reason not to meet in person is a recurring sign, and a video call does not settle it, since calls can be spoofed. Money requests often start small, such as help with a plane ticket, then grow with new medical, legal or travel crises, almost always with a promise of repayment. In relationship investment scams, the target is coached into deposits on a platform showing fake profits, and new taxes or fees appear at cash-out.

Elder financial exploitation draws on trust that already exists. The Consumer Financial Protection Bureau notes exploiters can be strangers but also family members or friends, and a financial power of attorney grants broad authority without regular oversight. Listed warning signs include unexplained withdrawals, usually paid bills going unpaid, changed beneficiaries, and a relative, caregiver or friend who blocks visitors or will not let the older person speak for themselves.

The trap is picturing a typical victim. A systematic review found no evidence that older adults make up a larger share of online fraud victims, and the National Institute on Aging says abuse can happen to any older adult. Say a neighbor notices that an older man's bills are going unpaid and a new friend now speaks for him. That matches the listed signs, and someone who reports a concern does not need to prove abuse, since professionals investigate.

Rule: a sweetheart never met in person who needs money, and a helper who controls an older person's contacts and accounts, both fit documented exploitation patterns, whatever story comes attached.

Every investment carries some risk, and investments that pay more typically carry more of it, so a pitch pairing high returns with safety contradicts that trade-off; the SEC lists guaranteed returns among its red flags. A Ponzi scheme pays existing investors with new investors' money, so it needs a constant inflow, and its tells include returns that stay positive whatever the market does and difficulty withdrawing. In a pyramid scheme, income depends on recruiting rather than retail sales.

Advance-fee fraud asks for money to release money, whether the payment is called a fee, tax or commission. The FTC says real prizes are free. Real lenders may charge an application fee, but none says a fee guarantees a loan. Fake job offers can end with the new hire paying for training, or depositing a check and sending part of it on; a fake check can take weeks to be discovered, leaving the target to repay their own bank. Government agencies never charge to help with a refund.

The trap is asking the seller to prove the seller. Say Omar is offered a fund with a polished registration certificate and glowing references. A fraudster gains nothing by correcting him, and the SEC does not issue registration certificates, so the document proves nothing. SEC guidance is independent research, including a free check of a broker's or adviser's disciplinary history, and it says there is never a reason to make a quick investment decision.

Rule: treat high return with low risk, and paying first to receive money, as red flags, and verify a seller through sources the seller does not control.

Researchers note gaslighting is defined inconsistently. A 2025 review of nearly 100 sources proposed defining it as manipulation of perception, reality and memory that undermines a person's ability to know their world. The tactics it found most often described were misrepresenting situations to create false narratives, denying or dismissing what happened, making the target feel or be called crazy, and recasting the target as the one at fault. Under that definition, disagreement alone is not gaslighting, since questioning someone's view counts only when it produces effects such as self-doubt.

DARVO stands for Deny, Attack, and Reverse Victim and Offender: the accused denies wrongdoing, attacks the accuser's credibility and casts themselves as the real victim. In one experiment, readers who saw it rated a fictional perpetrator more believable and the victim less believable. Whataboutism and tone policing deflect in a similar way, answering a criticism with someone else's faults or with how it was voiced.

Workplace bullying research lists acts such as withholding information someone needs, changing their tasks so the work is hard to complete, taking credit for their work and setting impossible deadlines. The definition requires persistence over time against someone who finds it hard to push back, so one crunch is not the pattern.

The trap is letting the dispute move. Say Tessa raises a skipped safety check with a colleague, and within a minute the talk is about her tone and whether she can be trusted. Recognizing DARVO means noticing that a dispute about a specific act has become a dispute about the person who raised it, and earlier work reported that awareness of the pattern reduced its impact.

Rule: when a reply to a specific complaint turns to the complainer's tone, memory or credibility, or to someone else's faults, the original claim is still unanswered.

Evan Stark describes coercive control as a course of conduct rather than separate events, with physical violence often, not always, among its tactics; its harm lies in combination and repetition. Federal guidance names forcing isolation from family, friends, school or work, and threats to harm a partner's children, family or friends, as psychological abuse. Microregulation, control over ordinary daily life, can look mundane, and each act can seem minor. Federal guidance also defines economic abuse, restricting access to money, credit or financial information, and technological abuse, using devices, apps or trackers to monitor or control a partner.

Walker's cycle of tension building, an acute incident and reconciliation describes one pattern, but partner violence ranges from one episode to chronic abuse over years. Leaving is hard for practical reasons such as money, housing, children and animals, and abuse can continue or intensify after separation, a recognized risk period; professionals use structured risk assessment to inform safety planning.

High-control groups show parallel moves. Joining usually happens through small steps, often beginning with warmth and shared mission. Information control manages which sources reach members, stock answers close down doubts, and progressively larger demands make each next step feel cheaper than resisting. Leaving can cost relationships, identity, money and standing, though former members also report relief. Brainwashing is a poorly supported explanation: the methods involved are standard persuasion, and of about 3,500 American prisoners indoctrinated during the Korean War, probably under 1 percent kept the ideology.

The trap is judging each act alone. Say a friend notices that Ana has stopped making plans because defending each friendship became too tiring, and that her partner takes the car just before she needs it. No single item looks like violence; the pattern is the finding.

Rule: judge control by the pattern and its effect over time, not by whether any single act looks like violence.

Scammers push people to act before they can think, and in lab studies time pressure made people worse at spotting fraud, though one study found this only for scams framed as avoiding a loss, so the first refusal is a pause. One refusal protocol teaches a firm no with at most a short reason and no self-justification, then the broken record step: repeating the same refusal instead of adding explanations. The two programs that teach it did not test the technique on its own. In Asch's studies, a single dissenter cut conformity to a unanimous wrong majority from about a third of answers to about 5 percent, and consumer guidance makes a related point: telling a friend before acting can help someone see a scam.

After a scam, FTC guidance sorts next steps by payment method, from the bank or card issuer to the app, gift card issuer or crypto exchange. Federal law limits losses from unauthorized card use, while cryptocurrency lacks the same protections. The FTC and the FBI's Internet Crime Complaint Center use reports to build cases and track trends. Recovery scams then target people who already lost money, and the FBI's center says it never contacts people directly for money.

The trap is blaming the victim. Fraud victimization is associated with shame and self-blame, and fear of judgment is a commonly reported barrier to seeking help. Say Grace realizes the refund agent she paid was a second scammer. Shame pushes toward silence, yet a report with the dates, amounts and recipients the FBI's form asks for is what agencies use to build cases and warn others.

Rule: pause, say no once and repeat it rather than defend it, and treat any offer to recover lost money for a fee as another scam.

Results run both ways. Preregistered replications of ego depletion found effects near zero, and a 2012 replication of the elderly-priming study found no slower walking with automated timing; the effect appeared only when experimenters expected it. The 1957 movie-theater subliminal advertising claim was fabricated. Yet a 2023 replication of Asch's conformity study found an error rate of about a third, close to the original. Nudges shrink too: in 126 nudge-unit trials free of selective publication, the average effect was 1.4 percent, against 8.7 percent in published studies.

Lab studies usually attach no cost to a choice, and when that Asch replication paid for correct answers, errors fell from 33 to 25 percent. Short videos naming manipulation techniques improved recognition in experiments, but the effect was small in a field test. In a field experiment with over 10,000 ministry staff, 22 percent of the control group gave away a password, and one simulated phishing email cut that by about 8 points.

The trap is trusting a defense because it sounds scientific. Say a training vendor claims its course makes staff immune to phishing. Most message choices in persuasion research change persuasiveness only slightly, and traits predict scam susceptibility only modestly, so a claim that large needs its evidence shown.

Rule: before trusting an influence finding or a defense, ask which tactic was tested, on whom, against what comparison, measured how, and whether it replicated.