EDUCATION SUBJECT ATLAS · LOGIC · Wintour House V1.0 · Rainbolt × CivDJ

What Is Logic?

Logic is the systematic study of good inference. It asks when a conclusion follows from premises, how arguments can be represented, which forms of reasoning preserve truth, how uncertainty changes inference, and how mistakes in reasoning can be identified without confusing disagreement with error.

Logic is often introduced through tidy syllogisms, symbols and truth tables. Those tools matter, but they are only part of the subject. Human reasoning also occurs in law, science, journalism, medicine, mathematics, public debate, everyday conversation and artificial intelligence. In those settings premises may be uncertain, language may be ambiguous and conclusions may be defeasible rather than certain. A complete education in logic must therefore connect formal proof with real-world reasoning.

Logic is not the art of always being right. It is the discipline of making the route from what we assume to what we conclude visible enough to test.

Arguments

An argument is a set of claims in which some claims—the premises—are offered in support of another claim—the conclusion. Arguments can be expressed in essays, conversations, diagrams, mathematical proofs, legal submissions, scientific papers and code.

The first skill in logic is not symbol manipulation. It is recognising what the argument actually is. Real language contains repetition, rhetoric, examples and unstated assumptions. Logical analysis often begins by extracting the inferential skeleton.

Premises and conclusions

Premises are propositions used as support. The conclusion is the proposition the reasoning is intended to establish. Indicator words such as “therefore,” “because” and “since” can help, but they are not mechanically reliable.

A paragraph can contain several nested arguments. One conclusion can become a premise for a later conclusion. Mapping these layers makes complex reasoning easier to evaluate.

Validity

A deductive argument is valid when it is impossible for all its premises to be true while its conclusion is false. Validity concerns structure, not whether the premises are actually true.

“All birds are mammals; all mammals are machines; therefore all birds are machines” is structurally valid even though its premises are false. Logic separates inferential quality from factual accuracy so that each can be tested independently.

Soundness

A sound deductive argument is valid and has true premises. Soundness therefore connects logical form with the world.

This distinction matters because perfect reasoning from bad information can still produce a false conclusion. Good thinking requires both logical control and reliable evidence.

Deduction

Deductive reasoning aims at necessity. If the premises are true and the form is valid, the conclusion must be true. Mathematics relies heavily on deduction because proofs derive consequences from definitions, axioms and previously established results.

Deduction is powerful because it preserves certainty, but it cannot generate reliable conclusions from unreliable premises.

Induction

Inductive reasoning moves from observations toward broader generalisations or predictions. Its conclusions are supported rather than guaranteed.

Scientific inference often depends on induction. Observing many instances can increase confidence in a pattern, but future cases may still differ. Statistical reasoning gives induction a more explicit account of uncertainty.

Abduction

Abductive reasoning infers the best available explanation from evidence. A doctor considers which diagnosis best explains symptoms. A mechanic asks which fault best explains a pattern of failures. A historian asks which reconstruction explains surviving evidence most coherently.

Abduction is defeasible. New evidence can make a once-plausible explanation inferior to another.

Formal logic

Formal logic represents reasoning using precisely defined languages and rules. Symbolisation strips away some ambiguity so inferential structure can be studied directly.

Formal systems contain syntax, semantics and proof rules. Syntax determines which expressions are well formed. Semantics assigns meanings or truth conditions. Proof systems specify which steps of derivation are permitted.

Propositional logic

Propositional logic treats complete statements as basic units connected by operators such as not, and, or, if-then and if-and-only-if.

Truth tables show how compound propositions depend on component truth values. They provide a mechanical way to test some forms of validity.

Predicate logic

Predicate logic adds variables, predicates and quantifiers so reasoning about properties and relationships can be represented. It can distinguish “every student read a book” from “there is one book every student read.”

Quantifier structure reveals ambiguities ordinary language can hide.

Necessary and sufficient conditions

A necessary condition must be present for something else to occur. A sufficient condition guarantees the relevant outcome under the stated rule. Confusing the two is a common reasoning error.

If oxygen is necessary for a certain combustion process, oxygen alone may not be sufficient. Logic helps unpack these causal and definitional relationships.

Contradiction

A contradiction arises when a set of claims cannot all be true together under the relevant logical system. Detecting contradiction is powerful because inconsistent premises can undermine an entire argument.

In everyday reasoning, apparent contradiction may instead come from changing definitions or contexts. The first task is to determine whether the claims really concern the same proposition.

Consistency

A set of claims is consistent when they can all be true together. Consistency is necessary for many rational systems but does not guarantee truth. A fictional world can be internally consistent while not being real.

Proof

A proof is a sequence of justified steps from accepted starting points to a conclusion. Mathematical proof aims at deductive certainty within a formal framework.

Proof differs from persuasion. A persuasive presentation can be logically weak, while a rigorous proof can be difficult to understand. Good education aims for both validity and intelligibility.

Direct proof

A direct proof begins from assumptions and derives the desired result through permitted inferences. It is often the clearest form when the route from premise to conclusion is visible.

Proof by contradiction

Proof by contradiction assumes the negation of the desired conclusion and shows that this assumption leads to inconsistency. The contradiction forces rejection of the assumption under classical logic.

Counterexample

A universal claim can be refuted by one genuine counterexample. If someone claims that all metals are attracted strongly to a magnet, one counterexample is enough to show the universal form is false.

Counterexample hunting is one of the most useful habits in logic because it stress-tests broad claims cheaply.

Informal logic

Informal logic studies arguments as they occur in natural language and real settings. It examines relevance, burden of proof, evidence quality, context, dialogue, rhetoric and fallacies.

Formal validity is not always the central standard. A legal argument, scientific inference or policy debate may involve uncertain evidence and defeasible conclusions. Informal logic asks whether the argument is cogent for its actual purpose.

Argument mapping

Argument mapping represents claims and support relationships visually. Independent premises, linked premises, objections and rebuttals can be separated into nodes and edges.

Mapping is especially useful when an argument feels confusing because several inferential layers are mixed together.

Hidden premises

Everyday arguments often omit premises because speakers assume shared knowledge. “She must be home; the lights are on” relies on an unstated link between lights and occupancy.

Making hidden premises explicit reveals where disagreement actually lies.

Burden of proof

The burden of proof concerns who must provide support for a claim and how much support is required. Burdens differ across contexts. Criminal law, scientific publication and casual conversation use different standards.

Logic becomes more realistic when inferential rules are connected to institutional contexts.

Fallacies

Fallacies are recurring patterns of poor reasoning. Labels can help, but merely naming a fallacy is not enough. The analyst should show exactly how the reasoning fails.

Ad hominem: attacking a person instead of addressing the relevant argument.
Straw man: replacing a position with a weaker version.
False dilemma: presenting too few alternatives.
Equivocation: shifting the meaning of a term.
Begging the question: assuming the conclusion in the premises.
Hasty generalisation: inferring too broadly from insufficient cases.
Appeal to popularity: treating widespread belief as evidence of truth.

Context matters. Criticising a person’s credibility is not automatically ad hominem when credibility is genuinely relevant to testimony.

Rhetoric and logic

Rhetoric studies persuasion; logic studies inferential support. They overlap but are not identical. A rhetorically powerful argument can be invalid, while a logically strong argument can fail to persuade because it ignores audience knowledge or language.

Responsible communication should aim to align persuasion with evidence rather than use rhetoric to conceal inferential weakness.

Dialectic

Dialectical reasoning develops understanding through objection and reply. Instead of treating disagreement as combat, dialectic treats an opponent as a source of stress tests.

The strongest version of an argument often emerges after it has survived serious criticism.

The principle of charity

The principle of charity asks us to interpret another person’s argument in its strongest reasonable form before evaluating it. This reduces false disagreement and prevents easy victories over weak caricatures.

Probability and logic

Classical logic represents certainty elegantly, but many real decisions involve degrees of belief. Probability gives a formal language for uncertain inference.

Logic and probability can therefore be complementary. A proposition may be logically possible yet extremely improbable given evidence, or logically valid conditional on premises whose truth is uncertain.

Bayesian reasoning

Bayesian reasoning updates probability when evidence arrives. Prior beliefs combine with likelihood to produce posterior beliefs.

This formalises a general reasoning principle: evidence should change belief in proportion to how expected that evidence would be under competing hypotheses.

Base rates

Base rates are background frequencies. Ignoring them can produce dramatic mistakes. A highly accurate screening test can still yield many false positives when the condition is extremely rare.

Logic for real decisions therefore needs statistical context, not only valid conditional statements.

Defeasible reasoning

Defeasible reasoning supports conclusions that may need to be withdrawn when new information appears. “Birds normally fly; this animal is a bird; so it probably flies” is reasonable until we learn it is a penguin.

Human reasoning relies heavily on defaults because complete information is rare. Non-monotonic logics model systems in which adding information can invalidate earlier conclusions.

Modal logic

Modal logic studies concepts such as necessity and possibility. Other modal systems reason about knowledge, obligation, time and belief.

These systems matter in philosophy, linguistics, computer science and AI because ordinary reasoning often includes statements about what must, may, should or could happen.

Temporal logic

Temporal logic represents relationships such as always, eventually, until and next. It is used in reasoning about systems that change over time.

Software verification can use temporal logic to express requirements such as “if a request is accepted, a response will eventually occur.”

Logic and mathematics

Logic provides foundations for mathematical proof and formal systems. Mathematical logic studies proof theory, model theory, set theory and computability among other areas.

The relationship is two-way: mathematics formalises logic, while logic clarifies what mathematical proof and consequence mean.

Logic and computer science

Computer science depends on logical structure. Boolean logic drives digital circuits; type systems constrain programs; formal verification proves properties; databases rely on logical query languages.

Programs are executable arguments in a loose but useful sense: given defined inputs and rules, they derive outputs through controlled transformations.

Computability and limits

Logic helped reveal that some problems cannot be solved by any general algorithm. The halting problem shows that there is no universal method that can determine for every possible program and input whether the program eventually stops.

This is an important intellectual boundary. Some failures are not caused by insufficient computing power but by limits built into formal reasoning itself.

Logic and science

Science combines deduction, induction and abduction. Theories generate predictions deductively; experiments provide observations; statistical inference evaluates patterns; explanatory reasoning compares mechanisms.

Logic helps keep these roles distinct so that evidence is not asked to do more than it can support.

Logic and law

Legal reasoning combines rules, precedent, analogy, evidence and institutional burdens of proof. A legal conclusion is not simply a syllogism because factual uncertainty and interpretation matter.

Logic helps separate whether a rule applies from whether the evidence establishes the facts required by the rule.

Logic and ethics

Ethical arguments contain both factual and normative premises. “This action will cause harm” is empirical; “avoidable harm should count against an action” is normative.

Logic cannot choose values for us, but it can show whether a moral conclusion follows from the values and facts we claim to accept.

Logic and language

Natural language contains ambiguity, implicature, metaphor and context. Formalisation can clarify structure but may also discard relevant meaning.

This is why logic and linguistics meet at semantics and pragmatics: what a sentence literally entails may differ from what a speaker communicates.

Logic and critical thinking

Critical thinking is broader than logic. It includes source evaluation, metacognition, statistical literacy and domain knowledge. Logic supplies the inferential spine.

A person can know fallacy names and still reason poorly if evidence is weak or domain assumptions are wrong.

Logic and AI

Artificial intelligence has long used formal logic for representation and inference, while modern machine learning relies heavily on statistical and learned representations. Contemporary AI combines several forms of reasoning rather than one universal logic engine.

Large language models can produce arguments that look coherent without guaranteeing validity or factual correctness. This makes explicit argument checking more important, not less.

Evaluating AI reasoning

Extract the actual premises.
Check whether cited facts are true.
Test whether the conclusion follows.
Look for missing conditions.
Ask whether uncertainty was represented honestly.
Construct a counterexample.
Compare an alternative explanation.

Fluent language should never be treated as a substitute for an inspectable inferential route.

The everyday argument pipeline

Identify the conclusion.
List explicit premises.
Recover hidden premises.
Check definitions for ambiguity.
Ask whether premises are supported.
Classify the inference: deductive, inductive, abductive or defeasible.
Test the argument with a counterexample or competing explanation.
Calibrate confidence to evidence.
Separate disagreement over facts from disagreement over values.
Revise the conclusion when a premise fails.

Why fallacy spotting is not enough

Introductory logic often becomes a catalogue of fallacy names. This can create a new bad habit: using labels as conversation-ending weapons. Strong reasoning requires reconstruction before criticism.

Ask first whether the argument has been understood fairly. Then explain which premise, inference or assumption fails. Logic should improve the shared model, not merely score points.

Uncertainty and confidence

Not every conclusion deserves the same confidence. Logical education should develop a vocabulary of certainty: entailed, strongly supported, probable, plausible, speculative, contradicted and unknown.

This prevents a common failure in public reasoning where weak evidence is expressed with absolute language.

Common misconceptions

“Logic tells us what is true.” Logic tells us what follows from premises; premises still need evidence.
“If an argument is valid, its conclusion is true.” Validity alone does not guarantee true premises.
“Induction is bad logic because it is not certain.” Inductive reasoning is essential when conclusions must extend beyond observed cases.
“A fallacy label refutes an argument.” The specific inferential failure still needs explanation.
“Formal logic captures all reasoning.” Real-world arguments also involve uncertainty, context and defeasibility.
“AI reasoning can be trusted when it sounds coherent.” Coherence is not proof of validity or truth.

Mini case: the umbrella argument

“The pavement is wet, so it rained.” The conclusion is plausible but not deductively guaranteed. A street-cleaning vehicle, sprinkler or burst pipe could also explain the observation.

The correct response is not to reject the argument entirely. It is to classify it as abductive and compare alternative explanations.

Mini case: a viral claim

“Millions of people shared this claim, therefore it must be true.” The premise about popularity may be factual, but the inferential link fails. Popularity is evidence of transmission, not truth.

Logic separates the social fact from the epistemic conclusion.

A CivDJ model of logic

ENTITY: propositions, premises, conclusions, agents, evidence sources and formal systems.
STATE: truth assignment, belief level, proof status, uncertainty and context.
OCCURRENCE: inference, revision, contradiction detection, proof and update.
RELATIONSHIP: implication, support, contradiction, equivalence, relevance and dependence.
INTENT: prove, explain, predict, persuade, decide or challenge.
OBSERVATION: statements, data, testimony, examples and counterexamples.
ARTIFACT: arguments, proofs, truth tables, diagrams, formal languages and models.
CLAIM: conclusions about validity, probability, explanation or rational support.
VOID: hidden assumptions, unknown premises, ambiguous terms and unmodelled alternatives.

Rainbolt traversal makes logic inspect the edges of the obvious argument. CivDJ rotates premises against counterexamples, probability, context, alternate definitions and hidden assumptions before a conclusion is allowed to settle.

How to think like a logician

Find the conclusion.
Recover the premises.
Define ambiguous terms.
Classify the inference.
Check logical form.
Check factual support independently.
Search for counterexamples.
Test rival explanations.
Calibrate confidence.
Revise rather than defend a broken route.

Logic across the learning journey

Young learners can begin with if-then reasoning, categories, contradiction and giving reasons. Secondary learners can study arguments, validity, fallacies, probability and proof. Advanced study adds predicate logic, modal logic, proof theory, model theory, computability, informal logic and applications in computer science and philosophy.

The progression is from “give me a reason” to “show me the complete inferential architecture.”

Why logic belongs inside education

Logic teaches learners to make reasoning inspectable. It separates truth from inference, evidence from rhetoric, certainty from probability and disagreement from contradiction.

These habits support mathematics, science, law, writing, coding, citizenship and everyday decision-making. Logic is not one more subject beside the others. It is part of the wiring through which every subject justifies what it claims to know.

What Is Logic? | Arguments, Inference, Proof, Probability and the Architecture of Good Reasoning