STRATA
Skip to study
XLVII

Algebraic geometry and topology

Hodge conjecture

Which topological classes on a smooth projective complex variety come from algebraic subvarieties?

Clay: Unsolved

Status and source review: 2026-11-19

State the mathematical question precisely.

For a smooth projective complex variety X and any p, every class in H²ᵖ(X,Q) ∩ Hᵖ,ᵖ(X) should be a rational linear combination of cohomology classes of codimension-p algebraic cycles. Rational coefficients, projectivity and smoothness are part of the question.

Identify what a complete argument must establish.

  1. Start with an arbitrary rational Hodge class in the stated category and construct algebraic cycles with the required rational cohomology class.
  2. Distinguish a harmonic differential-form representative from an algebraic-cycle representative.
  3. Keep integral, rational, projective and general Kähler variants separate; a theorem or counterexample in one category may not transfer.

Trace the results that changed the problem.

  1. Hodge formulates the geometric question.

    The conjecture appears in his International Congress contribution on topological invariants of algebraic varieties, published in the proceedings in 1952.

  2. Divisor classes provide the accessible model.

    The line-bundle argument through the exponential sequence gives the degree-two case; extending its conclusion to arbitrary codimension requires new input.

  3. Atiyah and Hirzebruch disprove the integral strengthening.

    They find torsion cohomological obstructions to classes represented by analytic cycles, including examples in the projective category.

  4. The locus of Hodge classes is shown to be algebraic.

    Cattani, Deligne and Kaplan prove a precise result about variation in algebraic families without assuming the full conjecture.

  5. Voisin tests the Kähler extension.

    Her counterexample shows that even Chern classes of coherent analytic sheaves need not generate all rational Hodge classes on general compact Kähler manifolds.

Separate established results from the remaining claim.

Inspect the upstream dependencies.

Make each proposed approach answer for its obstacle.

These are research directions and diagnostic tests. Their inclusion does not establish novelty, feasibility at full scale, or a solution.

Construct cycles in a specified family and compare spans.

Choose a family with computable cohomology, list geometric cycles and calculate their rational span inside the Hodge subspace. State exactly which codimensions and fibers are covered.

Obstacle. A match in a special family does not exhaust all smooth projective varieties. Dimension counting may depend on an unproved claim that every Hodge class has been found.

Try the next discriminating test.

Independently certify the Hodge-space dimension and cycle-class matrix in a bounded example; retain any unresolved rank calculation as a gap.

Investigate deformation with the representation problem visible.

Transport a class along its Hodge locus and ask whether a known cycle can deform with it. Track the obstruction to extending the actual cycle, rather than only its cohomology class.

Obstacle. The algebraicity of the locus does not imply a global family of representing cycles. Monodromy, special fibers and extension obstructions can intervene.

Try the next discriminating test.

Produce a local deformation calculation in one family, then name the theorem that would justify passage to arbitrary fibers. If that theorem is the conjecture itself, the argument is circular.

Work through a restricted calculation.

Rational generation and integral generation differ.

Use the additive subgroup 2Z inside Z as a simple model for the coefficient distinction.

Its rational span contains 1 because (1/2)·2 = 1, but no integer multiple of 2 equals 1. An exact rational check confirms the first identity; reduction modulo 2 proves the second claim.

This is an elementary algebraic counterexample to conflating coefficient systems. It is not itself a geometric counterexample to the Hodge conjecture.

Compare this example with the interactive proof diagnostics

Follow the downstream consequences conditionally.

Locate the actual quantum connection.

Hodge theory and algebraic geometry occur in mathematical physics, but their shared use does not make this conjecture a quantum-mechanical prediction. No inspected source here establishes a quantum experiment that would decide the universal algebraic-cycle question.

Follow quantum mechanics through the historical map

Inspect the supporting sources.

  1. The Millennium Prize Problems

    Clay Mathematics Institute · Current official categories and original seven-prize fund.

  2. The Hodge Conjecture

    Pierre Deligne / Clay Mathematics Institute · Rational Hodge classes on smooth projective complex varieties; divisor case and restrictions.

  3. Analytic cycles on complex manifolds

    Michael F. Atiyah and Friedrich Hirzebruch / Topology · Integral cycle obstructions from torsion cohomology operations; published 1962.

  4. A Counterexample to the Hodge Conjecture Extended to Kähler Varieties

    Claire Voisin / IMRN 2002 · Failure of a generalized Kähler statement even using Chern classes of coherent sheaves.

  5. On the Locus of Hodge Classes

    Eduardo Cattani, Pierre Deligne and Aroldo Kaplan / arXiv; JAMS 1995 · Algebraicity of the parameter locus where a transported class retains Hodge type; distinct from algebraicity of a representing cycle.