To discover truths is the task of all sciences; it falls to logic to discern the laws of truth."
"The laws of logic are not laws of nature; they are laws of the consequence of thought."
Gottlob Frege, Begriffsschrift (1879)
"The world is the totality of facts, not of things... Logic pervades the world: the limits of the world are also its limits." ~ Ludwig Wittgenstein, Tractatus.
Logic is the study of valid inference. It is the discipline that asks not what we believe, but whether we are entitled to believe it given our premises. The distinction is easy to state and remarkably hard to maintain. We routinely conflate the psychological force of an argument — how compelling it feels — with its logical validity — whether the conclusion in fact follows. Critical thinking, in large part, is the patient labor of keeping these two things apart.
Notice what logic does not do: It does not adjudicate the truth of premises. A perfectly valid argument can lead you straight to a false conclusion, provided the starting assumptions are themselves false. Logic is an instrument of consistency, not a guarantee of truth. Aristotle understood this with exemplary clarity: the syllogism shows that if all men are mortal and if Socrates is a man, then Socrates is mortal. Whether all men are in fact mortal is the business of the biologist, not the logician.
This makes logic both more limited and more powerful than is usually supposed. More limited, because it cannot by itself deliver knowledge of the world. More powerful, because its standards are universal: a valid argument in fourth-century Athens is valid in twenty-first-century San Francisco.
Informal logic treats reasoning in natural language — the arguments one finds in editorials, courtrooms, scientific papers, and ordinary conversation. Its central concern is fallacies: patterns of argument that mimic the shape of validity without inheriting its substance. The Socratic dialogues remain the founding documents of this branch. Their method is to expose, with embarrassing patience, that what seemed an unshakeable position in fact rests on a mistaken inference.
Formal logic abstracts entirely from content. An inference has formal validity when it holds for any substitution of terms: all A are B; all B are C; therefore all A are C is valid regardless of what A, B, and C designate. Aristotle inaugurated this project in the Prior Analytics (350BC) with one of the great conceptual innovations in intellectual history — the schematic variable. The decision to write a letter where ordinary language would have placed a noun marks the birth of logic as a science of pure form.
Symbolic logic extends formal analysis through a rigorous artificial language, treating the logical connectives — negation, conjunction, disjunction, implication — with algebraic precision. Its two principal branches are propositional logic, which operates on sentences treated as atomic units, and predicate logic, which analyzes their internal structure through quantifiers (∀, ∃) and predicates. The modern symbolic apparatus, developed by Frege, Peirce, Peano, Russell, and Whitehead between roughly 1879 and 1913, was animated by the ambition known as logicism: the demonstration that mathematics is nothing more than logic in extended form.
Mathematical logic turns formal systems into objects of mathematical study in their own right. Here one finds model theory, proof theory, set theory, computability theory, and the most disorienting result of twentieth-century thought: Gödel's incompleteness theorems, which established that any sufficiently powerful formal system contains true statements it cannot prove. The logicist ambition was not merely unrealized; it was demonstrated to be unrealizable: the gap between truth and proof is irreducible.
The concept of logical form is what makes logic formal in the relevant sense, and the word deserves deserves deep reflection. To say that logic is formal is not merely to say that it is rigorous or systematic. It is to say that the validity of an argument is determined entirely by its structure, independently of the particular content the structure happens to carry. The form is what remains when one subtracts everything specific.
The textbook examples of this principle are repeated endlessly: After all, Socrates really was a man, and he did not shy away from his own death. Logic becomes interesting when formal analysis begins to clash with human intuition. Let us look at some logical forms and how they operate in relation to human thinking.
Inference is the movement of thought from one claim to another — the step by which we take ourselves to be entitled to a conclusion on the basis of what we already accept. Not every transition between thoughts counts: drifting from one association to the next is not inferring, and neither is asserting a conclusion alongside a premise without any relation between them. An inference claims something stronger — that the conclusion follows, that accepting the premises commits one to accepting what comes after. The central task of logic is to make precise what "follows" means, and the history of the discipline is in large part a history of competing answers to that question.
Inference is also normative: it governs what one ought to conclude, not merely what one happens to think next. To recognize an inference as valid is to acknowledge a rational obligation — given these premises, this conclusion is what reason requires. One can of course fail to draw it; minds wander, motivated reasoning intervenes. But the failure is a failure against a standard, not simply a different route taken. This is what distinguishes logic from descriptive psychology. Psychology tells us how people in fact reason; logic tells us how reasoning must go if it is to count as reasoning at all. Frege put the point sharply: the laws of logic are not laws of thinking but laws of truth, and they bind anyone who aims at truth whether they recognize the binding or not.
Consider an inference called modus tollens: if P, then Q; not-Q; therefore not-P. The structure is unobjectionable. Now load it with an example: if the defendant was at the scene, his DNA would appear in the evidence; his DNA does not appear in the evidence; therefore he was not at the scene. The inference is valid — yet jurors routinely refuse it, because they don't trust the backward reasoning from absent effect to absent cause. The form compels what the psychology rejects. Logical training is precisely the closing of that gap.
There is no possible world in which P holds, the conditional holds, and Q fails. Along with modus ponens, it is the backbone of deductive inference, which lets us reason backwards from absent effect to absent cause. Popper's falsificationism is essentially modus tollens elevated to a methodology of science: a theory predicts an outcome, the outcome fails to appear, and the theory — in the ideal case — dies.
Negation is the simplest logical operation and the one that causes the most trouble. Formally, it is a one-place operator: given a proposition P, the negation ¬P is true exactly when P is false, and false exactly when P is true. Nothing could be more elementary. Yet negation in natural language almost never behaves this cleanly. Where the word "not" attaches — to a predicate, to a quantifier, to an entire proposition — changes the meaning decisively, and speakers routinely lose track of which scope they intend. "Not all swans are white" does not mean "no swans are white"; "I don't believe he is guilty" does not mean "I believe he is not guilty." The interesting questions about negation are therefore never really about the operator itself. They are about where it lands, what it takes in its scope, and how the ambiguities of ordinary language disguise distinctions that formal notation makes unmistakable.
Negation creates many difficulties. "Not all testimony is reliable" does not mean "no testimony is reliable" — it means "some testimony is not." Confusing these is no marginal error; it creates genuine disputes in epistemology and law. The contrapositive offers a related lesson: if a theory is scientific, it must be falsifiable is logically equivalent to: if a theory is not falsifiable, it is not scientific. The two formulations carry identical content but provoke very different reactions. Popper's celebrated criterion of demarcation is, formally speaking, simply the contrapositive — restated as a methodological policy.
Anselm's ontological argument illustrates what happens when formal analysis is brought to bear under maximum philosophical pressure. Necessarily, God exists; whatever exists necessarily exists in all possible worlds; therefore there is no possible world without God. The structure has survived seven centuries of scrutiny. What is contested is whether the first premise asserts a logical truth or smuggles existence into a definition. Formal analysis does not settle the matter. What it does — and what only it can do — is locate the dispute with surgical precision: the fight is about a premise, not about an inference.
The same principle exposes the deeper lesson: surface grammar and logical form frequently diverge. Bertrand Russell demonstrated this brilliantly in his theory of descriptions. "The present King of France is bald" looks, grammatically, like a simple subject-predicate sentence, parallel to "Kant is bald." But France has no king. If the sentence had the logical form it appears to have, it should be straightforwardly false — yet it also seems wrong to say "the present King of France is not bald," since that implies there is one and he has hair. Russell's solution was to show that the sentence's actual logical form is not P(a) but a complex quantified claim: there exists exactly one thing that is currently King of France, and that thing is bald. The apparent paradox dissolves once the form is correctly displayed. Natural language, in other words, is a treacherous guide to logical structure — a point Frege had already made central and one Wittgenstein would push to the breaking point in the Tractatus before reversing himself in the Philosophical Investigations.
The practical consequence is uncomfortable in a productive way. Once one has identified the form of an argument, one is obligated to evaluate it on those terms, regardless of whether the conclusion is welcome. An argument one finds repugnant may be formally valid — in which case the honest response is to challenge a premise, not the inference. An argument for a conclusion one already believes may be a fallacy — in which case believing it for that reason leaves one epistemically exposed. Logic does not care about preferences. That indifference is not a defect; it is the source of its authority.
Critical thinking turns on a distinction the formal apparatus makes especially crisp: the difference between deductive, inductive, and abductive reasoning.
Deductive reasoning is truth-preserving. If the premises are true and the argument valid, the conclusion is guaranteed. The price of this certainty is that deduction is non-ampliative — the conclusion contains nothing not already implicit in the premises. One cannot deduce new knowledge of the world; one can only unfold what one is already committed to.
Inductive reasoning runs in the inverse direction: It moves from particulars to generalities, extending beyond the evidence to claims the evidence cannot strictly establish. Modern science is inductive at its core: no finite number of confirming observations leads to the final affirmation of a universal law. Hume saw this clearly and was disturbed by it — the inference from "the sun has risen every day so far" to "the sun will rise tomorrow" is psychologically irresistible and logically unjustifiable. Induction is ampliative: it generates genuine new knowledge. It is also defeasible: a single well-placed counterexample can topple a generalization built on thousands of confirming instances.
Abductive reasoning is inference to the best explanation. It is the working logic of clinicians, detectives, and scientists at the moment of hypothesis formation. It asks not what follows from the evidence but what account of the evidence is the simplest, most coherent and economical explanation. Abduction has neither the certainty of deduction nor the statistical weight of induction, and yet it is how we navigate nearly every decision made under incomplete information. Its formal status remains contested — which is precisely why it deserves more attention.
For most of the twentieth century, the working assumption among philosophers was logical monism: there is one correct logic — classical first-order logic, with its bivalent truth values and its laws of identity, non-contradiction, and excluded middle. That assumption has been steadily eroded. The proliferation of non-classical systems has made it impossible to ignore the question of whether logic is one or many. Intuitionistic logic abandons excluded middle in the name of constructive proof. Relevance logics restrict implication to block irrelevant entailments. Paraconsistent logics tolerate inconsistency without permitting it to detonate into triviality. Modal logics formalize necessity and possibility. Substructural logics weaken the structural rules of classical proof theory in the service of resource-sensitive reasoning. Each of these systems is mathematically respectable and philosophically motivated.
The most influential contemporary response is the logical pluralism of JC Beall and Greg Restall, who hold that there is more than one genuine relation of logical consequence — that classical, intuitionistic, and relevant logics each capture a legitimate way of making the informal notion of "follows from" precise. Their position has provoked vigorous defenses of monism and the more austere position of logical nihilism — the view that there are no correct logics at all, only formal systems with various technical properties. Whether logic is one or many remains a live question, and the answer one favors carries implications well beyond the seminar room: for the foundations of mathematics, for the semantics of natural language, and for the very meaning of rationality.
This is the contemporary horizon against which the historical material in the subpages should be read. The classical tradition is not a closed monument. It is a working inheritance, still under construction.
A Note on the Continental Connection
Logic is sometimes treated, in popular accounts, as the proprietary concern of the analytic tradition. This is a parochial mistake. Wittgenstein's Tractatus Logico-Philosophicus — the most ambitious attempt of the twentieth century to specify the relation between logical form and the structure of the world — sits at the boundary of both traditions and properly belongs to neither. Lacan's mathemes, Badiou's reading of set theory as ontology, the structuralist preoccupation with formal relation, Kimhi's recent return to the Parmenidean question of the unity of thought and being: these are all serious engagements with logical form undertaken from the Continental side. The strict opposition of "Anglo-American logic" and "Continental philosophy" is a pedagogical fiction — useful for organizing course catalogs and damaging when believed.
The Subpages
The pages that follow develop specific topics in greater depth:
History of Logic — from Aristotle through the Stoics, the medieval Scholastics, Leibniz's dream of a characteristica universalis, and the modern revolution that ran from Frege to Gödel.
The Three Laws of Thought — identity, non-contradiction, and excluded middle, together with the non-classical traditions that have systematically contested each of them.
Fallacies — a working taxonomy of the most common and consequential errors in argument, with attention to the psychological mechanisms that make them seductive.
Gödel's Incompleteness Theorems — what they show, what they do not show, and why popular accounts almost always overstate the philosophical conclusions.
Kurt Gödel — the life and singular mind behind the twentieth century's most disorienting mathematical result.
For deeper engagement with the primary literature, the Stanford Encyclopedia of Philosophy remains an indispensable resource, and selected primary texts — Frege's Begriffsschrift, Gödel's 1931 paper, Tarski on truth, Kripke on modal semantics — are linked in the bibliography.
Two final considerations as one approaches this material. Logic does not tell you what to think; it does not guarantee certainty. What it offers is something rarer and more difficult to dismiss — a public standard against which reasoning can be tested and, when it fails, honestly corrected. In an intellectual environment where the distinction between argument and assertion is increasingly blurred, the discipline of holding form distinct from content remains, in the strict sense, indispensable.