Noether's minimal-basis problem: characterizing rational intermediate fields of transcendence degree at least three

In her 1915 paper on fields and systems of rational functions, Emmy Noether proved that every intermediate field of transcendence degree $\varrho$ of a rational function field over a characteristic-$0$ constant field admits a rational basis, and asked when the minimal number $\varrho$ of algebraically independent generators suffices — a minimal basis, which exists exactly when the field is purely transcendental over the constants. For $\varrho = 1$ and $\varrho = 2$ the answer is classical (Lüroth's theorem; Castelnuovo's criterion over algebraically closed constants), and for $\varrho \geq 3$ minimal bases fail in general, rigorously since the 1971–72 counterexamples. The question Noether explicitly flagged as unattempted — characterize the exceptional fields that do possess a minimal basis — survives as the modern rationality problem for unirational function fields (equivalently, unirational varieties) of dimension at least $3$, also known as the open residue of the Lüroth problem. Despite five decades of obstruction theory (intermediate Jacobians, unramified cohomology, birational rigidity, decomposition of the diagonal), no characterization, complete obstruction theory, or decision procedure exists, and the flagship instance of the rationality of the very general cubic fourfold was settled in the negative only in 2025, by case-specific invariants rather than any general criterion.

Background

Let $\Omega$ be a field (the source works over an arbitrary field of characteristic $0$) and let $\Omega(x_1,\dots,x_n)$ denote the rational function field in $n$ algebraically independent indeterminates. An intermediate field is a field $\mathfrak{K}$ with $\Omega \subset \mathfrak{K} \subset \Omega(x_1,\dots,x_n)$; if its transcendence degree over $\Omega$ — the maximal number of algebraically independent elements — is $\varrho$, Noether writes $\mathfrak{K}_{n\varrho}$. A transfer principle proved in the same paper (§3) shows that it suffices to study the case $\varrho = n$, i.e. subfields of finite index. Noether proves (her Satz IV) that every such field admits a rational basis: finitely many rational functions that rationally generate it. A minimal basis (Definition IV, §6) is a rational basis consisting of exactly $\varrho$ functions, which are then necessarily algebraically independent; equivalently, $\mathfrak{K}$ has a minimal basis if and only if it is purely transcendental over $\Omega$, i.e. isomorphic to $\Omega(t_1,\dots,t_\varrho)$. Geometrically, $\mathfrak{K}_{n\varrho}$ is the function field of a unirational variety — a $\varrho$-dimensional variety $V$ admitting a dominant rational map $\mathbb{P}^{n} \dashrightarrow V$, or equivalently a function field embedding into a rational function field — and a minimal basis exists precisely when $V$ is rational, i.e. birational to $\mathbb{P}^{\varrho}$ (two varieties are birational when they are connected by a rational map with rational inverse). Classical terminology calls the fibration defined by the generators an involution, and rationality of the involution the existence of a minimal basis.

Noether records in §6 the state of the question: for $\varrho = 1$ a minimal basis always exists (Lüroth's theorem, in the formulation of Steinitz's field-theory monograph); for $\varrho = 2$ a minimal basis always exists after a finite algebraic extension of the constant field $\Omega$, and in general only then (Castelnuovo's numerical criterion for rationality of surfaces — a smooth projective complex surface is rational if and only if its arithmetic genus $p_a$, a cohomological Euler-characteristic invariant, and its second plurigenus $P_2$, the dimension of the space of sections of the square of the canonical line bundle, both vanish; modern proof by Zariski); for $\varrho \geq 3$ minimal bases do not exist in general — she cites Enriques' construction of a non-rational involution, i.e. a field $\mathfrak{K}_{33}$ without minimal basis — and she adds: 'Eine Charakterisierung der speziellen Körper $\mathfrak{K}_{n\varrho}$ mit Minimalbasis ist noch nicht versucht' ('a characterization of the special fields $\mathfrak{K}_{n\varrho}$ possessing a minimal basis has not yet been attempted').

In modern language this circle of questions is the Lüroth problem: whether every extension of an algebraically closed field $k$ of characteristic $0$ contained in $k(t_1,\dots,t_n)$ is purely transcendental. For $\varrho \geq 3$ the answer is negative: rigorous counterexamples were found in 1971–72 by Iskovskikh and Manin (smooth quartic threefolds, via birational rigidity — the phenomenon that a variety admits no birational self-maps beyond its biregular ones, while rational varieties have many), by Clemens and Griffiths (the cubic threefold, via its intermediate Jacobian — the complex torus built from the $H^{2,1}\oplus H^{1,2}$ part of cohomology, a birational invariant that for a rational threefold must be a product of Jacobians of curves (the abelian varieties parametrizing degree-zero divisor classes on curves) and for the cubic threefold is not), and by Artin and Mumford (unirational threefolds with nontrivial $2$-torsion in $H^3$, detected by the unramified Brauer group — the group of Brauer classes of the function field that are unramified along every divisor, itself a birational invariant). Hence the parent question is closed, and the surviving question is exactly the characterization: to decide, among unirational function fields of transcendence degree $\geq 3$, which are rational, by criteria playing the role that Lüroth's theorem and Castelnuovo's criterion play in lower dimension. The rational Fano threefolds of the tabulated families (Fano meaning that the anticanonical line bundle is ample) and projective bundles over rational bases are known rational classes, and Saltman's invariant fields are further known unirational non-rational fields.

Progress toward such criteria has come from birational invariants — quantities constant on birational equivalence classes whose nontriviality obstructs rationality: the intermediate Jacobian (dimension $3$ only), torsion in $H^3$ and more generally the unramified Brauer group and higher unramified cohomology, birational rigidity, and, since 2013, the decomposition of the diagonal — the property that in the Chow group of $V \times V$ (algebraic cycles modulo rational equivalence) the class of the diagonal equals the class of $V \times \{x\}$ for a point $x$ plus a cycle pulled back from $T \times V$ for some variety $T$ of dimension smaller than $\dim V$, a property equivalent to universal triviality of the Chow group $\mathrm{CH}_0$ of zero-cycles (every zero-cycle of degree zero on $V$ becomes rationally equivalent to zero after any extension of the constant field). The last-mentioned machinery governs stable rationality — rationality of $V \times \mathbb{P}^m$ for some $m$, a property implied by rationality but strictly weaker. Two adjacent questions frame the source paper: Hilbert's 14th problem (finite generation of rings of invariants, to which the paper contributes a positive class and which Nagata refuted in general) and Noether's problem (rationality of the fixed field of a finite group acting by permuting variables, posed in Noether's companion papers of 1915–1918); both originate in the same cluster of work and delimit what the 1915 paper did and did not settle.

Problem Statement

Let $\Omega$ be a field of characteristic $0$, let $\Omega(x_1,\dots,x_n)$ be the rational function field in $n$ algebraically independent indeterminates over $\Omega$, and let $\mathfrak{K}$ be an intermediate field, $\Omega \subset \mathfrak{K} \subset \Omega(x_1,\dots,x_n)$, of transcendence degree $\varrho \geq 3$ over $\Omega$ (the transcendence degree is the maximal number of algebraically independent elements over $\Omega$; Noether writes $\mathfrak{K}_{n\varrho}$, and by the transfer principle of the source paper one may take $n = \varrho$, i.e. $\Omega(x_1,\dots,x_n)$ a finite extension of $\mathfrak{K}$). Characterize those $\mathfrak{K}_{n\varrho}$ that possess a minimal basis in Noether's sense — a rational basis, i.e. a finite set of rational functions generating $\mathfrak{K}$ as a field over $\Omega$, consisting of exactly $\varrho$ functions, which are then necessarily algebraically independent; equivalently, those $\mathfrak{K}$ that are purely transcendental over $\Omega$, i.e. isomorphic to $\Omega(t_1,\dots,t_\varrho)$; equivalently, geometrically, determine when a unirational variety of dimension $\geq 3$ over $\Omega$ — one whose function field embeds into a rational function field over $\Omega$ — is rational, i.e. birational to projective space $\mathbb{P}^{\varrho}$ over $\Omega$ (two varieties are birational when they admit mutually inverse rational maps between dense open subsets). Over algebraically closed constants such as $\mathbb{C}$ this is the classical rationality problem for unirational varieties.

The characterization is sought in effective form: each criterion or invariant explicitly specified and computable on presented fields, and defined without reference to rationality itself. The stipulated constant field is, as in the source, an arbitrary field of characteristic $0$: the characterization is required for intermediate fields over any such $\Omega$, not only over algebraically closed constants (over nonclosed constant fields Galois-cohomological phenomena enter); the algebraically closed case (e.g. $\Omega = \mathbb{C}$) is the central instance of this same target, not a separate completion domain.

References

1. J. Lüroth. 'Beweis eines Satzes über rationale Curven.' Mathematische Annalen 9, 163–165 (1875). DOI: 10.1007/BF01443371. https://doi.org/10.1007/BF01443371
2. G. Castelnuovo. 'Sulla razionalità delle involuzioni piane.' Mathematische Annalen 44, 125–155 (1894). DOI: 10.1007/BF01446977. https://doi.org/10.1007/BF01446977
3. D. Hilbert. 'Mathematische Probleme.' Lecture at the International Congress of Mathematicians, Paris, 1900; reprinted in Gesammelte Abhandlungen, vol. III, pp. 290–329. Springer (1970). DOI: 10.1007/978-3-662-25726-5_19. https://doi.org/10.1007/978-3-662-25726-5_19
4. E. Steinitz. 'Algebraische Theorie der Körper.' Journal für die reine und angewandte Mathematik 137, 167–309 (1910). DOI: 10.1515/crll.1910.137.167. https://doi.org/10.1515/crll.1910.137.167
5. F. Enriques. 'Sopra una involuzione non razionale dello spazio.' Rendiconti dell'Accademia dei Lincei (5) 21, no. 1, 81–83 (1912). JFM 43.0702.01. https://zbmath.org/43.0702.01
6. E. Noether. 'Körper und Systeme rationaler Funktionen.' Mathematische Annalen 76, 161–196 (1915). DOI: 10.1007/BF01458137. https://doi.org/10.1007/BF01458137
7. E. Noether. 'Der Endlichkeitssatz der Invarianten endlicher Gruppen.' Mathematische Annalen 77, 89–92 (1915/16). DOI: 10.1007/BF01456821. https://doi.org/10.1007/BF01456821
8. E. Noether. 'Gleichungen mit vorgeschriebener Gruppe.' Mathematische Annalen 78, 221–229 (1917/18). DOI: 10.1007/BF01457099. https://doi.org/10.1007/BF01457099
9. L. Roth. 'Algebraic Threefolds (with Special Regard to Problems of Rationality).' Ergebnisse der Mathematik und ihrer Grenzgebiete 6. Springer, Berlin (1955). DOI: 10.1007/978-3-642-85531-3. https://doi.org/10.1007/978-3-642-85531-3
10. O. Zariski. 'On Castelnuovo's criterion of rationality $p_a = P_2 = 0$ of an algebraic surface.' Illinois Journal of Mathematics 2, no. 3, 303–315 (1958). DOI: 10.1215/ijm/1255454536. https://doi.org/10.1215/ijm/1255454536
11. J.-P. Serre. 'On the fundamental group of a unirational variety.' Journal of the London Mathematical Society s1-34, 481–484 (1959). DOI: 10.1112/jlms/s1-34.4.481. https://doi.org/10.1112/jlms/s1-34.4.481
12. M. Nagata. 'On the 14-th Problem of Hilbert.' American Journal of Mathematics 81, 766–772 (1959). DOI: 10.2307/2372927. https://doi.org/10.2307/2372927
13. V. A. Iskovskih (also transliterated Iskovskikh) and Yu. I. Manin. 'Three-dimensional quartics and counterexamples to the Lüroth problem.' Mathematics of the USSR-Sbornik 15, 141–166 (1971). DOI: 10.1070/SM1971v015n01ABEH001536. https://doi.org/10.1070/SM1971v015n01ABEH001536
14. C. H. Clemens and P. A. Griffiths. 'The intermediate Jacobian of the cubic threefold.' Annals of Mathematics (2) 95, 281–356 (1972). DOI: 10.2307/1970801. https://doi.org/10.2307/1970801
15. M. Artin and D. Mumford. 'Some elementary examples of unirational varieties which are not rational.' Proceedings of the London Mathematical Society (3) 25, 75–95 (1972). DOI: 10.1112/plms/s3-25.1.75. https://doi.org/10.1112/plms/s3-25.1.75
16. H. W. Lenstra Jr. 'Rational functions invariant under a finite abelian group.' Inventiones mathematicae 25, 299–325 (1974). DOI: 10.1007/BF01389732. https://doi.org/10.1007/BF01389732
17. T. Shioda. 'An example of unirational surfaces in characteristic $p$.' Mathematische Annalen 211, 233–236 (1974). DOI: 10.1007/BF01350715. https://doi.org/10.1007/BF01350715
18. R. G. Swan. 'Noether's Problem in Galois Theory.' In: Emmy Noether in Bryn Mawr (Bryn Mawr, 1982), pp. 21–40. Springer, New York (1983). DOI: 10.1007/978-1-4612-5547-5_2. https://doi.org/10.1007/978-1-4612-5547-5_2
19. D. J. Saltman. 'Noether's problem over an algebraically closed field.' Inventiones mathematicae 77, 71–84 (1984). DOI: 10.1007/BF01389135. https://doi.org/10.1007/BF01389135
20. E. Formanek. 'Rational function fields. Noether's problem and related questions.' Journal of Pure and Applied Algebra 31, 28–36 (1984). DOI: 10.1016/0022-4049(84)90073-2. https://doi.org/10.1016/0022-4049(84)90073-2
21. J.-L. Colliot-Thélène and M. Ojanguren. 'Variétés unirationnelles non rationnelles: au-delà de l'exemple d'Artin et Mumford.' Inventiones mathematicae 97, 141–158 (1989). DOI: 10.1007/BF01850658. https://doi.org/10.1007/BF01850658
22. J. Kollár. 'Nonrational hypersurfaces.' Journal of the American Mathematical Society 8, 241–249 (1995). DOI: 10.1090/S0894-0347-1995-1273416-8. https://doi.org/10.1090/S0894-0347-1995-1273416-8
23. J. Gutiérrez and D. Sevilla. 'Computation of unirational fields.' Journal of Symbolic Computation 41, 1222–1244 (2006). DOI: 10.1016/j.jsc.2005.05.009. https://doi.org/10.1016/j.jsc.2005.05.009
24. C. McLarty. 'Emmy Noether's first great mathematics and the culmination of first-phase logicism, formalism, and intuitionism.' Archive for History of Exact Sciences 65, 99–117 (2011). DOI: 10.1007/s00407-010-0073-y. https://doi.org/10.1007/s00407-010-0073-y
25. C. Voisin. 'Unirational threefolds with no universal codimension one cycle.' Inventiones mathematicae 201, 207–237 (2015). DOI: 10.1007/s00222-014-0551-y. https://doi.org/10.1007/s00222-014-0551-y
26. A. Beauville. 'The Lüroth problem.' Preprint, arXiv:1507.02476 (2015); published as a chapter of Rationality Problems in Algebraic Geometry (CIME, Levico Terme 2015), Lecture Notes in Mathematics 2172, pp. 1–27. Springer (2016). arXiv: https://arxiv.org/abs/1507.02476; DOI: 10.1007/978-3-319-46209-7_1. https://doi.org/10.1007/978-3-319-46209-7_1
27. B. Totaro. 'Hypersurfaces that are not stably rational.' Journal of the American Mathematical Society 29, 883–891 (2016; print issue of volume 29, published online 2015). DOI: 10.1090/jams/840. https://doi.org/10.1090/jams/840
28. J.-L. Colliot-Thélène and A. Pirutka. 'Hypersurfaces quartiques de dimension 3: non-rationalité stable.' Annales Scientifiques de l'École Normale Supérieure (4) 49, no. 2, 371–397 (2016). DOI: 10.24033/asens.2285. https://doi.org/10.24033/asens.2285
29. J. Guéré. 'On the irrationality of cubic fourfolds.' Preprint, arXiv:2603.04518 (2026). https://arxiv.org/abs/2603.04518