Every manipulation mechanic, cognitive bias, and Undersignal term — named and explained with real examples. Tap any entry to expand.
FallacyAd Hominem
Attacks the person making an argument rather than the argument itself, using character or credentials as a reason to dismiss a position.
"We can't trust his economic plan — he's never run a business."
FallacyAppeal to Authority
Cites an expert or authority figure as evidence for a claim outside that expert's relevant domain, leveraging status as a substitute for applicable expertise.
A Nobel laureate in physics endorsing a specific nutrition protocol.
FallacyBandwagon
Treats popularity or widespread belief as evidence for truth — the majority accepting something is presented as a reason for accepting it, not as a separate phenomenon to evaluate.
"Millions of Americans can't be wrong."
FallacyCherry Picking
Selectively presents evidence that supports a conclusion while omitting comparable evidence that would undermine it, creating a false impression of consensus.
"Studies show coffee is healthy" — citing only the supportive studies from a contested and mixed literature.
FallacyCircular Reasoning
Uses the conclusion as a premise in the argument — the claim is assumed to be true in the course of proving it is true.
"The Bible is true because it says so in the Bible."
FallacyConflation
Merges two distinct concepts, claims, or events and treats them as a single unified thing — allowing attributes of one to transfer unexamined to the other.
The Cotton piece — a bioweapon origin theory conflated with a lab accident theory, treating both as interchangeable when they carry very different evidential and legal weight.
FallacyFalse Dilemma
Presents only two options as if they exhaust all possibilities, forcing a binary choice when other alternatives exist.
"You're either with us or against us."
FallacyFalse Equivalence
Treats two things as equivalent when they differ significantly in nature, scale, or context — creating a misleading sense of balance or symmetry.
"Both sides have extreme views" — conflating documented violence with rhetorical disagreement.
FallacyHasty Generalization
Draws a broad conclusion from an insufficient sample — treating isolated cases as representative of a universal pattern.
"Three people I know were harmed by the vaccine" — used to make claims about population-level safety.
FallacyLoaded Question
Embeds an unproven and often damaging assumption into a question, forcing the respondent to either accept the assumption or struggle to disentangle it.
"Have you stopped misleading your constituents?" — presupposes prior deception regardless of the answer.
FallacyMotte and Bailey
Defends an uncontroversial claim (the "motte") as a proxy for a more radical, indefensible one (the "bailey") — retreating when challenged, then re-advancing the stronger claim.
Defending "equity" as simple fairness, then advocating for specific controversial redistributive policies under the same label.
FallacyPost Hoc
Assumes causation from correlation — because event B followed event A, A must have caused B, ignoring the many other explanations for the sequence.
"Crime went up after the new mayor took office" — implying the mayor caused the increase without establishing a mechanism.
FallacyRed Herring
Introduces a consideration that is tangential to the actual argument. The digression is framed as relevant, but engaging with it means abandoning the original question without addressing it.
"Before we discuss the data breach, let's remember everything this company has done for the community" — pivots to reputation before accountability can be established.
FallacySlippery Slope
Claims that one event will inevitably lead to extreme or catastrophic consequences, without providing evidence for the chain of causation.
"If we allow this regulation, it will lead to total government control."
FallacyStraw Man
Misrepresents an opponent's actual position with a weaker or more extreme version, making it easier to refute while appearing to engage with the original argument.
"Critics of the climate bill claim opponents want to destroy the planet."
Cognitive BiasAffect Heuristic
Emotional state is substituted for rational evaluation — content that produces strong emotional responses (fear, outrage, disgust) makes the reader less likely to apply analytical scrutiny to the claims.
Fear-inducing language and imagery deployed early in a piece so that subsequent factual claims are accepted without independent evaluation.
Cognitive BiasAnchoring
The first information encountered disproportionately influences all subsequent judgment — when content opens with extreme framing, every fact that follows is filtered through that initial frame.
Opening with apocalyptic framing anchors all subsequent interpretation, making moderate evidence feel more alarming than it is.
Cognitive BiasAuthority Bias
Tendency to attribute greater accuracy and trustworthiness to the opinions of authority figures, regardless of whether their expertise is relevant to the claim being made.
A chemistry professor cited as expert on economic policy — the credential transfers perceived credibility across domains.
Cognitive BiasAvailability Heuristic
Vivid, easily recalled examples are judged as more common or more probable — detailed, emotionally loaded descriptions of specific incidents make risks feel more present than statistics show.
Graphic descriptions of violence against a demographic make crime risk feel dramatically higher than aggregate data supports.
Cognitive BiasConfirmation Bias
The tendency to seek, favor, and recall information that confirms prior beliefs — content structured to provide exactly what the reader already wants to believe exploits this directly.
An article structured so that every data point validates what the target audience already believes, with no friction or challenge presented.
Cognitive BiasFraming Effect
Identical information produces different conclusions depending on how it is presented — the same factual content can be interpreted as reassuring or alarming based purely on the linguistic frame used.
"90% survival rate" versus "10% fatality rate" — same number, measurably different emotional and decision responses.
Cognitive BiasIllusory Truth Effect
Repeated exposure to a claim increases its perceived truthfulness, independent of any supporting evidence — repetition alone shifts credibility.
Repeating "the corrupt mainstream media" across 20 instances regardless of evidence trains the reader to accept the characterization as fact.
Cognitive BiasIn-Group Bias
Preferential treatment and greater trust assigned to members of one's own group — content that constructs a clear in-group identity exploits this to make readers more receptive to claims from perceived in-group members.
"Real Americans know that..." — the framing pre-assigns group membership and signals what the in-group believes.
Cognitive BiasScope Insensitivity
Intuitive responses to large-scale problems remain roughly constant regardless of actual scale. Content that inflates or compresses numbers exploits the fact that ten thousand does not feel proportionally more alarming than one hundred.
An advocacy piece that describes a crisis affecting 10 million people and one affecting 100,000 in identical emotional terms — the scale difference disappears in the framing.
Cognitive BiasZero-Risk Bias
Complete elimination of a small risk is valued disproportionately over a large reduction of a bigger one. Content framing marginal safeguards as total solutions exploits this preference for zero over near-zero.
"This policy will eliminate the threat entirely" — preferred over a larger intervention that reduces a greater risk by 80%, because zero feels categorically safer than near-zero.
RhetoricAsymmetric Sourcing
Quoted or cited support drawn disproportionately from one side of an issue. (Media-analysis construct.)
Seven named critics of a policy quoted at length; one unnamed supporter offered a single sentence — the sourcing pattern implies consensus before the reader examines the evidence.
RhetoricEuphemism / Dysphemism
Wording that softens (euphemism) or sharpens (dysphemism) the same referent to shift its valence.
"Headcount reduction" softens; "mass layoffs" sharpens — the same action, re-labeled to control its weight before the reader evaluates it.
RhetoricGlittering Generality
Vague, positively-charged abstractions substituted for specific claims. (Institute for Propaganda Analysis, 1937.)
"This bill is about freedom and American values" — without specifying which freedoms or which values, the abstraction invites each reader to project their own meaning.
RhetoricHedged Attribution
A claim supported by vague or unnamed sources ("some say," "critics argue"); describes the sourcing pattern, not the inference drawn from it. (Distinct from the appeal-to-anonymous-authority fallacy.)
"Sources familiar with the matter say the deal is collapsing" — the claim gains insider plausibility without any source being accountable for its accuracy.
RhetoricInnuendo
A claim conveyed by implication rather than direct statement. (Classical relatives: paralipsis, insinuatio.)
"I'm not saying he lied — I'm just asking questions." The accusation is made without being made, and without the accountability of a direct claim.
RhetoricJuxtaposition
Two elements placed in adjacency so a connection is implied without being asserted. (Classical relative: syncrisis.)
A photo of the accused placed directly beside a crime scene image — no caption links them, but proximity does the associative work without making a claim that can be disputed.
RhetoricLoaded Language
Evaluative or emotionally charged wording used where neutral description was available.
Describing a policy as "reckless" or "common-sense" rather than describing what it does — the label does persuasive work before the reader evaluates the substance.
RhetoricNarrative Sequencing
Information ordered so that earlier items precede and frame later ones. (Classical relative: dispositio / arrangement. Distinct from Framing Effect — the angle, not the order.)
Publishing the accuser's account first so that all subsequent denials are read as damage control — the sequence creates a presumption the content alone does not establish.
RhetoricPresupposition
A contestable assumption presented as already-settled background within a declarative statement. (Pragmatics. Distinct from the Loaded Question fallacy, which embeds the assumption in a question.)
"Since the administration's failures on this issue are well documented..." — the failures are asserted as settled fact in the course of making a separate point.
RhetoricSelective Emphasis
Foregrounds supporting facts while de-emphasizing or omitting countervailing ones; describes what the text emphasizes, not the reasoning error of suppressing evidence. (Related: salience framing. Distinct from the suppressed-evidence fallacy.)
A financial report devoting three paragraphs to quarterly growth and one sentence to supply-chain risk — the proportion of coverage shapes what the reader takes as the dominant story.
UndersignalConstruction Assessment
The five-level classification of how deliberately content has been structured — derived from dimension score patterns after analysis, not independently assigned. Labels in order: Declarative (structurally non-rhetorical), Neutral (some mechanics present but not concentrated), Patterned (organic patterning, not deliberately constructed), Elevated (deliberate mechanics with visible intent), Structured (deliberately constructed framing with low transparency).
A report scoring Structured means framing, sourcing, and timing are deliberately constructed toward a conclusion — not just opinionated, but architecturally deliberate. The Disguise Factor is what separates Structured from Elevated.
UndersignalUndersignal Risk Score
The primary 1–10 numerical measure of detected persuasion mechanics — aggregated from 14 dimension scores, weighted by cluster and amplified by the Disguise Factor. Not a measure of factual inaccuracy. Not a verdict on the author or outlet.
A score of 8.2 means multiple high-severity influence mechanics are operating simultaneously — not that the underlying facts are wrong, but that the presentation is doing significant work on the reader's conclusions.
UndersignalInfluence Intent
The eight-label verdict system classifying what the content is structurally trying to accomplish. Two pieces can carry the same Undersignal Risk Score but different intent verdicts — one is transparent advocacy, the other is presenting advocacy as neutral reporting. The vast majority of analyzed content falls into verdicts 1–4. Verdicts 5–8 are uncommon by design.
- Legitimate Informational — Structured to inform. No dominant persuasion architecture.
- Legitimate Commercial — Explicit advertising or branded content. Disclosed commercial purpose.
- Editorial Framing — Declared opinion or analysis. Framing, not mobilization.
- Advocacy-Consistent — Declared advocacy or mobilization content. Transparent purpose.
- Undisclosed Agenda-Consistent — Concealed purpose. PR, hidden funding, or undisclosed interests driving the narrative.
- Ideological Recruitment Pattern — Identity-first architecture designed to build group commitment before introducing ideology.
- Authority Capture Pattern — Institutional or expert framing deployed to foreclose dissent and delegitimize alternatives.
- High-Risk Manipulation Pattern — Structured architecture with extreme mechanic intensity, active concealment, or both. The rarest verdict.
Two pieces can carry the same Undersignal Risk Score but different intent verdicts — one is transparent advocacy, the other is presenting advocacy as neutral reporting.
UndersignalContext
Context benchmarks this piece against other analyzed text of the same content type, so you can see where its Undersignal Risk Score falls within that group. The comparison line shows its position, for example how much of that content type scores lower. The score's color reflects absolute severity and stays the same regardless of this comparison, so a low-severity score keeps its color even when it ranks high within its group.
UndersignalAbsolute
The raw Undersignal Risk Score compared against the full corpus of analyzed pieces across all content types. The percentile reflects where the piece sits relative to everything in the corpus. The score itself is fixed, a structural measurement independent of content type.
Comparing an op-ed and a press release, the absolute score lets you evaluate which carries more persuasion mechanics on a direct, cross-format basis.
UndersignalRapid Compliance
The dimension measuring how easily an argument spreads without being read — designed for sharing via headline, pull quote, or image excerpt with no engagement with the underlying content required.
Content with high Rapid Compliance is engineered so the shareable summary carries the full persuasive payload — the actual article is beside the point.
UndersignalNarrative Trajectory
Where the piece is trying to take the reader — the implied next belief, action, or identity position that follows logically from accepting the content's framing and conclusions.
A piece with a high Authority Capture score and Ideological Recruitment verdict has a trajectory toward distrust of institutions and loyalty to an alternative authority — the score names the destination, not just the vehicle.
UndersignalSequential Analysis
Running multiple documents in a content series — an introductory email followed by a follow-up, an onboarding sequence, a series of press releases or investor communications — to surface escalation patterns that a single-document score will not reveal. Gateway content in manipulation sequences is often deliberately low-pressure, scoring in the MODERATE range. The Undersignal Risk Score on a single document is an accurate reading of that document's mechanics. The pattern across a sequence reveals the architecture of the full campaign.
An MLM homepage may score 3.2. The follow-up email scores 5.8. The onboarding call transcript scores 7.4. No single score tells the full story. The sequence does.