Apps

🧪 This platform is in early beta. Features may change and you might encounter bugs. We appreciate your patience!

Dog Whistles

Also Known As: Coded Language Racial Coding Whistleblower Politics Implicit Messaging
Manipulation & Propaganda 🎯 Discrimination Detection☠️ Toxic Discourse ID: dog_whistles

Definition

Dog whistles are coded expressions that appear neutral to the general public but convey a specific, often discriminatory message to an intended audience. They allow speakers to signal discriminatory attitudes while maintaining plausible deniability. Identifying dog whistles requires understanding both the literal meaning and the historical or cultural context that gives the phrase its coded significance. Context is critical: the same phrase may be innocent in one setting and a dog whistle in another.

Examples

A politician campaigns on 'restoring law and order in our inner cities' — ostensibly about public safety, but historically a coded reference to racial minorities in urban areas.

A commentator repeatedly refers to 'globalist elites controlling the media,' using 'globalist' as a coded antisemitic reference while maintaining the surface meaning of internationalist economic policy.

A social media post about 'protecting our culture and traditions' uses language that, in context, signals opposition to immigration and multiculturalism rather than genuine cultural preservation.

Formal Logic Pattern
FOL Pattern
The First-Order Logic formula representing this reasoning pattern's logical structure.
FOL (First-Order Logic) uses quantifiers (∀ = for all, ∃ = there exists), connectives (∧ = and, ∨ = or, ⇒ = implies, ¬ = not), and predicates to capture the essential form of a reasoning pattern. For example, the Ad Hominem fallacy: Person(x) ∧ HasFlaw(x) ⇒ Invalid(Claim(x)). These patterns allow automated verification of logical validity.

∃m∃s(Message(m) ∧ Surface(m,s) ∧ Neutral(s) ∧ ∃h(Hidden(m,h) ∧ Discriminatory(h) ∧ ∃g(InGroup(g) ∧ Understands(g,h))))
Formal Verification:
Formal Verification
Checks whether a reasoning pattern is logically valid or invalid using an automated theorem prover.
Formal verification uses an SMT (Satisfiability Modulo Theories) solver — specifically Z3 — to mathematically check whether an argument's logical structure is valid. Each reasoning pattern is translated into First-Order Logic and tested: Can the premises be true while the conclusion is false? If yes, it's formally invalid. If no, it's formally valid. Many real-world patterns (analogies, heuristics) cannot be fully captured in formal logic — these are marked as not formally decidable, which doesn't mean they're wrong.
Not formally decidable

Verification Steps
Verification Steps
Binary yes/no questions that an AI must answer to detect a reasoning pattern in a text.
Each of the 452 aspects has verification steps — simple yes/no questions designed to systematically detect whether a pattern appears in a text. For ad hominem: "Does the argument attack a person rather than their claim?" For false dichotomy: "Are only two options presented when more exist?" This ensures consistent, reproducible analysis.

Binary (yes/no) questions an LLM must answer to identify this aspect:

  1. 1

    Does the language have a surface meaning that appears neutral or innocuous?

    Type: binary
  2. 2

    Could the same phrase carry a secondary meaning understood by a specific group?

    Type: binary
  3. 3

    Does the context suggest the secondary, discriminatory meaning is intended?

    Type: binary
  4. 4

    Does the speaker benefit from plausible deniability about the discriminatory meaning?

    Type: binary
Deep Dive
The expandable detail section on each aspect page with examples, psychology, and counter-strategies.
The Deep Dive section provides in-depth information about each aspect: a real-world example showing the pattern in action, an explanation of why it works psychologically, practical advice on how to counter it, alternative names, and links to related aspects.