Ancient / Platonic. The word aporia derives from Greek a-poros (“impassable,” “without a way through”). Plato’s early “aporetic” dialogues (Euthyphro, Charmides, Lysis, Laches, and the Meno) characteristically end without a positive definition, leaving interlocutors in perplexity. In the Meno (84), Socrates describes the purgative value of reducing someone to aporia: it shows a person who merely thought he knew that he does not, instilling a desire to inquire. The recent scholarly collection The Aporetic Tradition in Ancient Philosophy (eds. Karamanolis & Politis; surveyed in Bryn Mawr Classical Review 2018.11.08) distinguishes a positive use (Heraclitus, and in a roundabout way Protagoras and Gorgias) from a negative/eristic use (Zeno, Melissus). Aristotle, in the Metaphysics, made aporia methodological: he begins inquiry by surveying the aporiai that puzzled his predecessors. In the Pyrrhonist tradition systematized by Sextus Empiricus, unresolved difficulty yields suspension of judgment (epochē) and thence ataraxia.
Medieval — the insolubilia. Medieval logicians called liar-type paradoxes insolubilia (“insolubles”), and treated them as difficult but not truly insoluble — a striking contrast with the modern view that self-referential paradoxes are foundational and possibly unsolvable (per Wikipedia’s “Insolubilia” and the SEP entry by Paul Vincent Spade). Thomas Bradwardine composed his Insolubilia while a young master at Oxford in the 1320s (Stephen Read, ed., critical edition, Peeters, 2010), surveying and rejecting nine contemporary theories before setting out his own signification-closure solution — the best worked-out medieval treatment. John Buridan’s solution, developing Bradwardine, had the most lasting influence into the Renaissance. Notably, the insolubilia literature included more than the liar: it treated epistemic paradoxes (“You do not know this proposition”), pragmatic paradoxes (Buridan’s Bridge / promises and intentions in Sophismata ch. 8), and cases violating a semantic “principle of sufficient reason.” Buridan used self-reference to test his logic, whereas Ockham avoided paradox by denying that a term can refer to itself.
Kant — the Antinomies of Pure Reason. In the Transcendental Dialectic of the Critique of Pure Reason (1781), Kant structures reason’s illusions into paralogisms, antinomies, and the ideal of pure reason. The four antinomies are pairs of thesis/antithesis, each with an apparently valid proof, arising when reason seeks unconditioned totality in cosmology: - First (mathematical): the world has a beginning in time and is bounded in space vs. it is infinite in both (A426/B454). - Second (mathematical): every composite substance consists of simple parts vs. nothing simple exists. - Third (dynamical): there is causality through freedom vs. everything follows natural causal law. - Fourth (dynamical): there exists a necessary being vs. no necessary being exists. Kant classes the first two as “mathematical” (thesis and antithesis are contraries — both false) and the last two as “dynamical”/“cosmological” (they are subcontraries — both can be true) and resolves them via transcendental idealism. Left unresolved, the conflict yields the “euthanasia of pure reason” (A407/B434). Commentators (e.g., the Riksbankens Jubileumsfond project by Marcel Quarfood; the Logica Universalis 2020 paper) connect the resolution to the Aristotelian square of opposition.
Contemporary epistemology. The Münchhausen trilemma (also Agrippa’s trilemma) holds that every attempt to justify a knowledge claim collapses into one of three unsatisfying options — infinite regress, circular reasoning, or dogmatic (unproven) assumption. The term was coined by Hans Albert in Traktat über kritische Vernunft (1968); the ancient form is Agrippa’s, recorded in Sextus Empiricus’s Outlines of Pyrrhonism. In contemporary terms, foundationalism accepts the dogmatic horn, coherentism the circular horn, and infinitism the regress horn. Popper’s and Albert’s fallibilism accepts the trilemma as unsolvable and works by conjecture and criticism.
Gödel’s incompleteness (1931). The first theorem: any sufficiently powerful, consistent, recursively axiomatized formal system containing arithmetic is incomplete — there are true statements it cannot prove. The second: such a system cannot prove its own consistency. This ended Hilbert’s program of a complete, consistent axiomatization of all mathematics.
Turing and Church (1936). Turing proved the halting problem undecidable — no algorithm decides, for every program/input pair, whether the program halts — and derived the unsolvability of Hilbert’s Entscheidungsproblem as a corollary. Church independently proved the Entscheidungsproblem unsolvable using the lambda calculus (Church, “An Unsolvable Problem of Elementary Number Theory,” American Journal of Mathematics 58(2):345–363, 1936). All these results share Cantor’s diagonalization structure: Gödel constructs a self-referential sentence; Turing diagonalizes over machines; Church over lambda-expressions.
Set-theoretic paradoxes. - Russell’s paradox (1901–1903): the set of all sets that are not members of themselves both is and is not a member of itself. Russell arrived at it by applying Cantor’s diagonal theorem to the universal set; he connected it explicitly to the liar. - Cantor’s paradox: there is no greatest cardinal (the power set always has strictly greater cardinality), so “the set of all cardinalities” cannot be a set. - Burali-Forti paradox (1897): the “set of all ordinals” would be an ordinal greater than every ordinal, a contradiction. Historically nuanced: Burali-Forti himself did not present it as a paradox (he misconstrued Cantor’s definition of well-ordering); Russell reformulated it as a contradiction and gave it its name in Principles of Mathematics (1903). Standard resolutions restrict set formation (Zermelo–Fraenkel; von Neumann–Bernays–Gödel “proper classes”) or stratify it (Russell’s simple and ramified theory of types, 1908). The Continuum Hypothesis is a further undecidability result: independent of ZFC (Gödel 1940, Cohen 1963).
The liar (“This sentence is false”) is the paradigmatic semantic paradox; the “strengthened liar” (“This sentence is not true”) resists truth-value-gap solutions. Major modern approaches: - Tarski’s hierarchy: no language contains its own truth predicate; truth for object language L₀ is defined only in a metalanguage L₁, generating an infinite hierarchy. - Kripke, “Outline of a Theory of Truth” (Journal of Philosophy 72:690–716, 1975): a three-valued (Strong Kleene) fixed-point construction in which sentences are evaluated in stages from “grounded” atomic facts. The liar is ungrounded and receives no truth value (falls into a gap); the truth-teller (“This sentence is true”) is ungrounded but not paradoxical. Kripke’s classification uses groundedness, paradoxicality, and intrinsicality — the last of which a 2024 arXiv result (Walsh) shows is not modally definable in the way the first two are. - Dialetheism / paraconsistency: Graham Priest holds that the liar is both true and false — a dialetheia — requiring a paraconsistent logic (his Logic of Paradox, LP, from “The Logic of Paradox,” Journal of Philosophical Logic 8:219–241, 1979) that rejects explosion (ex contradictione quodlibet: α, ¬α ⊢ β). Priest’s In Contradiction (1987; 2nd ed. Oxford, 2006) and Doubt Truth to Be a Liar (2006) develop the view; precursors include Newton da Costa’s C-systems (1963), F. G. Asenjo’s “A Calculus of Antinomies” (1966), and the Routley–Meyer relevance-logic tradition. Priest argues the naive-set-theory paradoxes (Russell) and the semantic paradoxes are dialetheia, and that dialetheism is rare and confined to limit cases, leaving ordinary discourse classical.
Wave–particle duality and complementarity. Quantum entities exhibit both particle and wave behavior; Bohr’s complementarity principle holds these descriptions are mutually exclusive but jointly necessary. The Heisenberg uncertainty relation (Δx·Δp ≥ ℏ/2) sets a joint limit on position and momentum.
The measurement problem. As von Neumann (1932) framed it, quantum theory has two evolutions: the continuous, deterministic Schrödinger evolution, and the discontinuous, indeterministic “collapse” on measurement. Since it is very hard to say what physically distinguishes a “measurement” from any other interaction, the boundary between quantum and classical is the core conceptual problem. Determinism-vs-indeterminism is genuinely contested: the Schrödinger evolution is deterministic; “indeterminism” strictly concerns measurement outcomes.
No-go theorems as proven aporias. - EPR (1935): Einstein, Podolsky & Rosen, “Can Quantum-Mechanical Description of Physical Reality Be Considered Complete?”, Physical Review 47:777–780 (May 15, 1935). Their criterion of reality: “If, without in any way disturbing a system, we can predict with certainty … the value of a physical quantity, then there exists an element of physical reality corresponding to this physical quantity.” arxiv Applying this to entangled particles and non-commuting observables, they conclude the wave-function description “is not complete.” Einstein (1949) crystallized it as a forced choice between (1) completeness and (2) locality/separability. - Bell’s theorem (1964): J. S. Bell, “On the Einstein Podolsky Rosen Paradox,” Physics 1(3):195–200. No local hidden-variable theory can reproduce all QM statistics; the Bell inequality is satisfied by any local hidden-variable model but violated by QM. Bell: “If [the theory] is local it will not agree with quantum mechanics, and if it agrees with quantum mechanics it will not be local.” The experimentally testable form is the CHSH inequality (Clauser, Horne, Shimony, Holt, Physical Review Letters 23:880–884, 1969). - Kochen–Specker (1967): S. Kochen & E. P. Specker, “The Problem of Hidden Variables in Quantum Mechanics,” Journal of Mathematics and Mechanics 17:59–87. In Hilbert dimension ≥ 3 one cannot assign context-independent definite values to all observables consistent with QM; any hidden-variable theory must be contextual. The original proof used a set of 117 vectors (SEP, “The Kochen–Specker Theorem,” Held); later work simplified this dramatically — Cabello, Estebaranz & García-Alcaine, “Bell–Kochen–Specker theorem: A proof with 18 vectors,” Physics Letters A 212:183–187 (1996), present 18 four-dimensional vectors, “a record for this kind of proof” (note: the 18-vector record is in dimension 4, not 3). Contextuality was independently formalized by Bell (1966), hence “Bell–Kochen–Specker.”
Being vs. becoming; one vs. many (the Eleatics). Parmenides argued that only Being is, and is therefore one and motionless; Heraclitus that everything flows. Zeno of Elea (c. 490–430 BCE), Parmenides’ disciple, devised paradoxes by reductio ad absurdum to defend monism, showing that assuming plurality and motion leads to absurdity. The famous motion paradoxes — the Dichotomy, Achilles and the tortoise, the Arrow, and the Stadium — are reconstructed mainly from Aristotle (Physics VI) and Simplicius. The 5th-century CE commentator Proclus reports Zeno derived 40 arguments; the Stanford Encyclopedia of Philosophy (Palmer, “Zeno of Elea”) states “at least eight of Zeno’s paradoxes survive,” while the Internet Encyclopedia of Philosophy holds “Zeno probably created forty paradoxes, of which only … ten are known.” Plato’s Parmenides is the best source for his general intention. These strike at the heart of concepts of space, time, continuity, and the passage from one to many.
Subject/object, necessity/contingency. These are treated across Kant’s antinomies (the third and fourth concern freedom/causality and necessary being) and in the post-Kantian tradition. The one/many and being/becoming problems recur in Plato’s Parmenides through the “compresence of opposites” (the intelligible Being is both one and many, in motion and at rest).
Kant’s sublime. In the Critique of the Power of Judgment (1790), the sublime arises when imagination fails to comprehend the infinite/immense, while reason affirms we can think it — a painful-pleasurable encounter with the limits of presentation.
Lyotard — presenting the unpresentable. Jean-François Lyotard recast Kant’s sublime as the defining problem of avant-garde art: art that “presents” the fact that there is something unpresentable. Key texts: “The Sublime and the Avant-Garde”; “Representation, Presentation, Unpresentable” and “After the Sublime, the State of Aesthetics” (in The Inhuman, trans. Bennington & Bowlby, Polity, 1991); Lessons on the Analytic of the Sublime (Stanford, 1994). Lyotard reads Barnett Newman’s Onement I as targeting the impossibility of any pictorial representation of the oneness of God — the idea gestured at via negationis (Zuckert, “Presenting the Unpresentable,” Kantian Review 26(4), 2021).
The ineffable in music/language. Wittgenstein’s Tractatus distinction between what can be said and what can only be shown (and the closing injunction to silence about “the mystical”) frames the ineffable; his Philosophical Investigations, by problematizing private language and emphasizing rule-governed public use, reframes ineffability as a question of how putative ineffability actually gets expressed (Springer, Philosophy of Religion 2008).
Genuine moral dilemmas and “moral residue.” Bernard Williams (“Ethical Consistency,” Proc. Aristotelian Society Supp. 39, 1965; repr. in Problems of the Self, 1973) argued that in genuine conflicts, the obligation not acted on does not vanish but leaves a “moral remainder” (regret, guilt), which is evidence it was real. He argued deontic logic must drop the agglomeration principle while keeping “ought implies can,” and later developed “agent-regret” (“Moral Luck,” 1976; Moral Luck, Cambridge, 1981). Ruth Barcan Marcus (“Moral Dilemmas and Consistency,” Journal of Philosophy 77(3):121–136, 1980) argued genuine dilemmas exist and cannot be resolved “without residue,” that a consistent set of moral principles can still generate dilemmas, and proposed a second-order regulative principle to arrange life so as to minimize dilemmas.
Value incommensurability. Isaiah Berlin (“Two Concepts of Liberty,” 1958; Four Essays on Liberty, Oxford, 1969) held that fundamental values are objectively plural, conflicting, and incommensurable — no common measure ranks them, and no general rule resolves all conflicts. He wrote that human goals are “many, not all of them commensurable, and in perpetual rivalry with one another” PhilArchive (1969, p. 131).
The Other and interiority. Emmanuel Levinas (Totality and Infinity, French 1961, English 1969) argued that Western ontology reduces the Other to the Same, whereas ethics begins in the face-to-face encounter with the Other, whose alterity (“infinity”) exceeds every totalizing category and commands an asymmetrical, infinite responsibility (“the face … formulates the first word,” and its command is “Thou shalt not kill”). This produces a structural aporia: the singular ethical demand of the Other can never be fully met, and responsibility to one Other sacrifices “the other others.”
Sartre’s dilemma. In Existentialism Is a Humanism (1946), Sartre’s student is torn between staying to care for his mother and leaving to join the Free French Forces — a conflict between “the morality of sympathy, of personal devotion” and “a morality of wider scope but of more debatable validity,” Brandeis University which no doctrine (Kantian or Christian) can decide; the student must freely invent his solution.
Derrida’s ethical aporias. For Derrida the gift is “the impossible”: the conditions that make it possible (recognition, gratitude, return) also destroy it as a gift (Given Time, 1991). Likewise hospitality, forgiveness, mourning, and responsibility (The Gift of Death, on Kierkegaard’s Abraham: “tout autre est tout autre”) are structured as possible only as impossible. In Aporias (1993) he analyzes death as “the possibility of an impossibility” (deconstructing Heidegger’s being-toward-death), asking “Is my death possible?”
Wittgenstein’s rule-following paradox. Philosophical Investigations §201: “This was our paradox: no course of action could be determined by a rule, because every course of action can be brought into accord with the rule.” Saul Kripke (Wittgenstein on Rules and Private Language, Harvard, 1982) reads this as “the most radical and original skeptical problem that philosophy has seen to date” — the “Kripkenstein” paradox: there is no fact that determines whether I mean plus or quus, and the “skeptical solution” grounds meaning in community assertibility conditions, from which the impossibility of a private language follows as a corollary. (Most commentators accept the paradox but not the attribution of the skeptical solution to Wittgenstein.)
Reference and pragmatics. Related meaning-aporias include the semantic paradoxes of definability (Berry, Richard, König) and the intensional paradoxes that, per recent scholarship (Tucker & Thomason; a 2011 Review of Symbolic Logic paper), fit neither of Ramsey’s two groups.