The Independence of the Continuum Hypothesis
Reference: Cohen, Paul J. (1963, 1964). “The Independence of the Continuum Hypothesis” (I & II). Proceedings of the National Academy of Sciences 50(6), 1143–1148 and 51(1), 105–110. DOI 10.1073/pnas.50.6.1143.
Summary
The papers that resolve Hilbert’s first problem and introduce forcing, the most powerful technique in modern set theory. Cantor’s continuum hypothesis (CH) asserts that there is no cardinality strictly between that of the integers and that of the reals — that |𝒫(ℕ)| is the least uncountable cardinal ℵ₁. Gödel had shown in 1938–40, via the constructible universe L, that CH and the axiom of choice cannot be disproved from the ZF axioms (they hold in L, an inner model of any model of ZF). Cohen supplies the other half: he constructs models of ZFC in which CH fails, and models in which the axiom of choice fails. Together these establish that CH is independent of ZFC — neither provable nor refutable — and likewise that AC is independent of ZF.
The method is forcing. Starting from a countable transitive model M of ZFC, Cohen adjoins a new “generic” object G (for ¬CH, enough generic subsets of ℕ to make the continuum larger than ℵ₁) to form an extension M[G]. The delicate point is that M cannot “see” G, yet one must guarantee M[G] still satisfies all the ZFC axioms and controls exactly which new statements become true. Cohen manages this with the forcing relation p ⊩ φ — a condition p (a finite fragment of the generic object) “forces” a sentence φ — defined entirely inside M, so that truth in the extension is decided by the ground model even though the generic object lies outside it. Genericity (the filter G meets every dense set of M) ensures the construction is consistent and that the intended statement holds.
Forcing turned independence from an exotic curiosity into a routine instrument: it is now the standard way to show a statement undecidable in ZFC, and it reframes the foundational picture. Where Gödel’s incompleteness theorems show that any sufficiently strong consistent theory leaves arithmetic sentences undecided, Cohen shows that one of the most basic questions about the size of the continuum is itself undecided by our standard foundation — not for want of cleverness but in principle. The combined Gödel–Cohen result is the canonical example of a natural mathematical statement that is independent of its axioms. Cohen received the Fields Medal (1966) for it.
Key Ideas
- Continuum Hypothesis: there is no cardinality strictly between |ℕ| and |ℝ|; equivalently |𝒫(ℕ)| = ℵ₁.
- Independence of CH from ZFC: Gödel (1940, via L) shows CH is consistent; Cohen shows ¬CH is consistent — so CH is undecidable in ZFC.
- Independence of AC from ZF, by the same model-building technique.
- Forcing: extend a countable transitive model M by a generic object G to get M[G] satisfying ZFC plus the desired (in)dependence.
- The forcing relation p ⊩ φ: definable inside the ground model, lets M control truth in M[G] without seeing G.
- Genericity: G meets every dense subset of the forcing poset that lies in M, guaranteeing consistency.
- Meta-mathematical limit: a basic question of set size is in-principle undecidable from the standard axioms.
Connections
Conceptual Contribution
- Claim: The continuum hypothesis (and the axiom of choice) is independent of the standard axioms of set theory — neither ZFC ⊢ CH nor ZFC ⊢ ¬CH.
- Mechanism: Forcing — adjoin a generic filter G to a countable transitive model M of ZFC, with a forcing relation defined inside M that fixes truth in M[G]; choose the forcing poset so the extension satisfies ¬CH (or ¬AC), complementing Gödel’s L model where CH holds.
- Concepts introduced/used: Forcing, Continuum Hypothesis, Constructible Universe, ZFC, Cardinality
- Stance: foundational independence-result paper (set theory / model theory)
- Relates to: Settles the cardinality question opened by Cantor 1891 as undecidable in Zermelo’s ZFC; the set-theoretic counterpart of the syntactic undecidability in Gödel 1931.