Paradox (from Greek para, against, and doxa, opinion; literally “contrary to opinion”) — A statement, argument, or situation that, from apparently acceptable premises and by apparently valid reasoning, leads to a contradictory, absurd, or self-undermining conclusion. Paradoxes constitute a privileged field of logical and philosophical inquiry because they reveal hidden tensions in the most basic intuitions about language, truth, reference, and infinity.
W.V.O. Quine, in The Ways of Paradox (1966), proposes a tripartite typology. Veridical paradoxes are those whose conclusion seems absurd but is in fact true — such as the Banach-Tarski paradox or the birthday paradox in probability. Falsidical paradoxes involve reasoning that appears valid but conceals an error — the conclusion is in fact false. Antinomies are the most serious: paradoxes in which the reasoning appears equally valid in the direction of the conclusion and in the direction of its negation, generating genuine contradiction within the logical or semantic system.
Zeno’s Paradoxes (5th century BCE): Zeno of Elea formulated several arguments in defence of Parmenides’s position that motion is illusory. The most famous is that of Achilles and the Tortoise: Achilles grants the tortoise a head start, but each time he reaches the point where the tortoise was, it has moved a little further ahead — and so on ad infinitum. Zeno concludes that Achilles can never catch up. The modern resolution employs the mathematical concept of a convergent series (the sum of infinitely many terms can be finite), but the paradox continues to stimulate debates about the infinite divisibility of space and time.
The Liar Paradox (attributed to Epimenides of Crete, 6th century BCE; later classical formulation): “This proposition is false.” If the proposition is true, then it is false (for that is what it asserts); if it is false, then it is true (for what it asserts is precisely that it is false). The antinomy is a semantic antinomy: it involves self-reference and the predicate “true” or “false” applied to a statement that refers to itself. Alfred Tarski, in Der Wahrheitsbegriff in den formalisierten Sprachen (1933), showed that a rigorous solution requires the distinction between the object language (in which one speaks) and the metalanguage (in which one speaks about the object language), eliminating the possibility of unrestricted semantic self-reference.
Russell’s Paradox (1901): Bertrand Russell discovered an antinomy at the heart of naive set theory: let R be the set of all sets that do not contain themselves as members. Does R contain itself as a member? If yes, by definition it should not contain itself; if no, by definition it should. Russell communicated the paradox to Frege in a celebrated letter (June 1902), undermining Frege’s logicist programme at the moment of its completion (Grundgesetze der Arithmetik, vol. II). Russell’s solution was type theory (theory of types), which forbids sets containing themselves as members through a hierarchy of logical types. Zermelo independently resolved the problem through axiomatic set theory (restricted separation).
Grelling-Nelson Paradox (1908): Kurt Grelling and Leonard Nelson formulated a paradox of semantic self-reference: an adjective is autological if it applies to itself (for example, “short” is short; “polysyllabic” is polysyllabic) and heterological if it does not apply to itself (“long” is not long). The question: is “heterological” heterological? If yes, it does not apply to itself, so it is not heterological; if no, it does apply to itself, so it is heterological. The antinomy is structurally analogous to Russell’s paradox.
The Sorites Paradox (from Greek soros, heap): Attributed to Eubulides of Miletus (4th century BCE). A heap of sand with 1,000,000 grains is a heap; removing one grain does not turn it into a non-heap; by induction, 1 grain of sand is a heap. The paradox exposes the vagueness of gradable predicates: expressions such as “heap”, “bald”, and “tall” have no precise boundaries of application. Philosophical responses include supervaluationism and logical revisionism with trivalent logics.
← Glossary