Are geometrically trivial strongly minimal sets in differentially closed fields unimodular? (Nagloo's Conjecture 4.1)

By the Hrushovski–Sokolović trichotomy, every strongly minimal set definable in a differentially closed field of characteristic zero (a model of $DCF_0$) is either nonorthogonal to the field of constants, nonorthogonal to the Manin kernel of a simple abelian variety, or geometrically trivial. The first two classes have a developed structure theory; the trivial class remains poorly understood, with no general classification. Unimodularity — the requirement that interalgebraic tuples of equal Morley rank have symmetric algebraic multiplicities — is a strong structural property: Hrushovski proved that unimodular strongly minimal structures are locally modular, and inside $DCF_0$ unimodularity forces geometric triviality. Nagloo's Conjecture 4.1 asserts the converse: every geometrically trivial strongly minimal set in a differentially closed field is unimodular. Nagloo verified the conjecture for the solution set of the second Painlevé equation, where all finite-to-finite correspondences are automorphisms. The adjacent conjecture that motivated the problem — that all trivial strongly minimal sets in $DCF_0$ are $ω$-categorical — was refuted by Freitag and Scanlon, whose analysis of the differential equation of the modular $j$-function produced a strongly minimal, geometrically trivial, non-$ω$-categorical set. Since $ω$-categorical strongly minimal sets are unimodular (a general fact), the question is live exactly for the non-$ω$-categorical trivial sets: Casale–Freitag–Nagloo proved that in the genus-zero Fuchsian family $ω$-categoricity holds precisely for non-arithmetic groups, and no subsequent work determines whether the $j$-function solution set, or any other arithmetic case, fails to be unimodular. The conjecture remains open.

Background

A differential field of characteristic zero is a field $K$ equipped with a derivation $δ$, an additive map satisfying the Leibniz rule $δ(xy) = x δy + y δx$; the theory $DCF_0$ of differentially closed fields of characteristic zero is the model companion of the theory of such fields. $DCF_0$ is $ω$-stable — every definable set carries an ordinal-valued dimension, its Morley rank — and for tuples from a strongly minimal set this dimension refines differential-algebraic transcendence degree: $RM(y_1, …, y_n/K) = n$ means that $y_1, …, y_n$ together with all their iterated derivatives are algebraically independent over $K$. A strongly minimal set is an infinite definable set every definable subset of which is finite or cofinite; strongly minimal sets are the indecomposable building blocks of totally transcendental theories (those where every definable set has Morley rank), and their internal geometry is the subject of Zilber's trichotomy, the principle that such a geometry should be field-like, group-like, or combinatorially trivial.

Hrushovski and Sokolović established the trichotomy for differentially closed fields: every strongly minimal set definable in a model of $DCF_0$ is (i) nonorthogonal to the field $ℂ$ of constants (the elements with $δ(c) = 0$, a pure algebraically closed field) — two strongly minimal sets being nonorthogonal when some infinite definable relation between them has finite-to-one projections to both coordinates; (ii) nonorthogonal to the Manin kernel $A^♯$ of a simple abelian variety $A$ with trace zero (an abelian variety is a projective algebraic group, and its Manin kernel is its smallest Zariski-dense definable subgroup, itself strongly minimal); or (iii) geometrically trivial. Here a strongly minimal set $Y$ is geometrically trivial if whenever distinct $y_1, …, y_ℓ ∈ Y$ are such that no pair is algebraically dependent over the base field $K$, then $y_1, …, y_ℓ$ together with all their derivatives are algebraically independent over $K$: all differential-algebraic relations among elements of $Y$ decompose into pairwise ones. Classes (i) and (ii) have a developed structure theory; class (iii) is the least understood — there is no general classification of geometrically trivial strongly minimal sets of differential order two or more.

Two finer invariants measure how rigid a trivial set is. Following Freitag–Jaoui–Marker–Nagloo, $Y$ is $ω$-categorical if $Y ∩ K⟨y⟩^{alg}$ is finite for every $y ∈ Y$, and strictly disintegrated if this intersection is always exactly $\{y\}$; here $K⟨y⟩$ denotes the differential field generated over $K$ by $y$ and all its iterated derivatives, and the superscript $alg$ its algebraic closure. Hrushovski proved (in a 1995 manuscript, as recorded by Freitag–Scanlon and by Aslanyan) that an order-one strongly minimal set in $DCF_0$ orthogonal to the constants is trivial with $ω$-categorical induced structure, which long supported the expectation that trivial strongly minimal sets in $DCF_0$ should be $ω$-categorical. Unimodularity, introduced by Hrushovski in his analysis of minimal structures, is a symmetry condition on multiplicities: a strongly minimal set $Y$ is unimodular if for every $n$ and every pair of tuples $y, y' ∈ Y^n$ that are interalgebraic over a differential field $K$ over which $Y$ is defined (meaning $K⟨y⟩^{alg} = K⟨y'⟩^{alg}$) with $RM(y/K) = RM(y'/K) = n$, the algebraic multiplicities agree: $mult(y/K⟨y'⟩) = mult(y'/K⟨y⟩)$, where $mult(y/K⟨y'⟩)$ is the degree of the finite algebraic field extension $K⟨y, y'⟩$ over $K⟨y'⟩$. Hrushovski proved that every unimodular strongly minimal structure is locally modular — its closure geometry behaves like linear span in a vector space, the linear case of Zilber's trichotomy — and Kestner and Pillay later showed that weak unimodularity is strictly weaker than unimodularity in general and that, for strongly minimal sets, unimodularity is equivalent to MS-measurability (Macpherson–Steinhorn measurability, an abstract dimension-and-measure structure on the definable sets). A basic example of non-unimodularity: in an algebraically closed field, the interalgebraic singleton tuples $y = (x)$ and $y' = (x^2)$ from the affine line have $mult(y/K⟨y'⟩) = 2$ and $mult(y'/K⟨y⟩) = 1$, which is one reason field-like structure is incompatible with unimodularity.

Inside $DCF_0$, Nagloo observed that unimodularity forces geometric triviality: by Hrushovski's theorem a unimodular strongly minimal set is locally modular, hence by the trichotomy nonorthogonal to the constants or to a Manin kernel, and neither is compatible with unimodularity (Manin kernels fail it, as shown in the same paper). His main theorem proves that for the second Painlevé equation $PII(α): y'' = 2y^3 + ty + α$ with $α ∉ 1/2 + ℤ$, the solution set is strongly minimal and geometrically trivial; since the finite-to-finite correspondences between its solutions are compositions of automorphisms (the classical Riccati/Bäcklund symmetry transformations of the Painlevé equations), the solution set is unimodular. He also records, from a private communication of Hrushovski, that for strongly minimal sets unimodularity is preserved under nonorthogonality. The paper then states Conjecture 4.1 — that conversely every geometrically trivial strongly minimal set in a differentially closed field is unimodular — presenting it as a restatement of Marker's conjecture that trivial sets in $DCF_0$ are infinite-dimensional, and its introduction frames it as a restatement of the conjecture that trivial strongly minimal sets in $DCF_0$ are $ω$-categorical.

Subsequent literature transformed the landscape around, but not of, this conjecture. Freitag and Scanlon proved that the third-order algebraic differential equation satisfied by the modular $j$-function (the invariant classifying elliptic curves up to isomorphism) defines a strongly minimal set $J$ with trivial forking geometry that is not $ω$-categorical, refuting the $ω$-categoricity conjecture; Aslanyan recovered and generalized this via Ax–Schanuel-type predimension inequalities (transcendence inequalities in the style of Ax's differential-field theorem), and emphasized that $J$ was the first trivial strongly minimal set in $DCF_0$ known not to be $ω$-categorical. Casale, Freitag and Nagloo proved that for every Fuchsian group $Γ$ of genus zero, the differential equation of the uniformizing function $j_Γ$ defines a strongly minimal, geometrically trivial set that is $ω$-categorical exactly when $Γ$ is non-arithmetic — so within this family the non-arithmetic cases are $ω$-categorical and hence unimodular by the general fact recorded below. Meanwhile new classes of geometrically trivial strongly minimal sets were established: the generic Painlevé equations are strictly disintegrated (Nagloo–Pillay); the Liénard/Poizat equations $y''/y' = f(y)$ are strongly minimal and trivial with all algebraic relations classified, hence $ω$-categorical (Freitag–Jaoui–Marker–Nagloo); the degenerate third Painlevé family is strongly minimal, trivial and $ω$-categorical (Freitag–Nagloo); generic algebraic vector fields of degree at least two are strongly minimal and geometrically trivial (Jaoui); and the classical Lotka–Volterra system is strongly minimal and trivial (Duan–Nagloo, published 2026). Duan and Nagloo record in a footnote the general fact that $ω$-categorical strongly minimal sets are unimodular (citing Pillay's Geometric Stability Theory), and pose an order-two $ω$-categoricity conjecture of their own (their Conjecture 2.14, proven so far only for equations of Brestovski form); but whether the non-$ω$-categorical set $J$, or any other non-$ω$-categorical trivial set, is unimodular is not stated anywhere in the literature.

Problem Statement

Let $𝒰$ be a differentially closed field of characteristic zero (a model of $DCF_0$) with derivation $δ$, and let $Y ⊆ 𝒰^k$ be a strongly minimal set definable over a differential field inside $𝒰$: an infinite definable set every definable subset of which is finite or cofinite. For a differential field $K ⊆ 𝒰$ over which $Y$ is defined and a tuple $y$, write $K⟨y⟩$ for the differential field generated over $K$ by the components of $y$ together with all their iterated derivatives, and $K⟨y⟩^{alg}$ for its field-theoretic algebraic closure inside $𝒰$.

Say that $Y$ is geometrically trivial if for every countable differential field $K ⊆ 𝒰$ over which $Y$ is defined and all distinct $y_1, …, y_ℓ ∈ Y$: if $y_i ∉ K⟨y_j⟩^{alg}$ for all $i ≠ j$, then $y_1, …, y_ℓ$ together with all their iterated derivatives are algebraically independent over $K$. (This is the contrapositive form of Nagloo's Definition 2.4(i): any algebraic dependence among such elements and their derivatives must already occur inside some pair.)

Say that $Y$ is unimodular if for every differential field $K ⊆ 𝒰$ over which $Y$ is defined, every $n ≥ 1$, and all tuples $y = (y_1, …, y_n)$ and $y' = (y'_1, …, y'_n)$ in $Y^n$ that are interalgebraic over $K$ (i.e. $K⟨y⟩^{alg} = K⟨y'⟩^{alg}$) with $RM(y/K) = RM(y'/K) = n$ — where $RM(y/K) = n$ means that $y_1, …, y_n$ together with all their iterated derivatives are algebraically independent over $K$, and likewise for $y'$ — the algebraic multiplicities agree:

$$mult(y/K⟨y'⟩) = mult(y'/K⟨y⟩),$$

where $mult(y/K⟨y'⟩)$ denotes the degree of the finite algebraic field extension $K⟨y, y'⟩$ over $K⟨y'⟩$.

Question (Nagloo, Conjecture 4.1): Is every geometrically trivial strongly minimal set in a differentially closed field unimodular?

Since the converse implication — in $DCF_0$, every unimodular strongly minimal set is geometrically trivial — is already established, an affirmative answer would characterize the geometrically trivial class inside $DCF_0$ exactly as the unimodular class. A negative answer is a counterexample: a geometrically trivial strongly minimal set $Y$ definable in some differentially closed field of characteristic zero, a differential field $K$ over which $Y$ is defined, an integer $n$, and interalgebraic tuples $y, y' ∈ Y^n$ with $RM(y/K) = RM(y'/K) = n$ and $mult(y/K⟨y'⟩) ≠ mult(y'/K⟨y⟩)$. Note that strictly disintegrated sets are unimodular for trivial reasons and — by a general fact recorded by Duan and Nagloo, citing Pillay's Geometric Stability Theory — so is every $ω$-categorical strongly minimal set; hence any counterexample must be non-$ω$-categorical, with nontrivial algebraic relations among its elements.

References

1. Joel Nagloo, Geometric triviality of the strongly minimal second Painlevé equations, Annals of Pure and Applied Logic 166(3) (2015) 358-368. DOI: 10.1016/j.apal.2014.11.008; arXiv:1302.4338. https://doi.org/10.1016/j.apal.2014.11.008
2. Ehud Hrushovski, Unimodular minimal structures, Journal of the London Mathematical Society (2) 46(3) (1992) 385-396. DOI: 10.1112/jlms/s2-46.3.385. https://doi.org/10.1112/jlms/s2-46.3.385
3. David Marker, The number of countable differentially closed fields, Notre Dame Journal of Formal Logic 48(1) (2007) 99-113. DOI: 10.1305/ndjfl/1172787548. https://doi.org/10.1305/ndjfl/1172787548
4. Charlotte Kestner and Anand Pillay, Remarks on unimodularity, Journal of Symbolic Logic 76(4) (2011) 1453-1458. DOI: 10.2178/jsl/1318338860; arXiv:1012.0982. https://doi.org/10.2178/jsl/1318338860
5. Joel Nagloo and Anand Pillay, On the algebraic independence of generic Painlevé transcendents, Compositio Mathematica 150(4) (2014) 668-678. DOI: 10.1112/s0010437x13007525. https://doi.org/10.1112/s0010437x13007525
6. Joel Nagloo, On transformations in the Painlevé family, Journal de Mathématiques Pures et Appliquées 107(6) (2017) 784-795. DOI: 10.1016/j.matpur.2016.10.006; arXiv:1502.01389. https://doi.org/10.1016/j.matpur.2016.10.006
7. James Freitag, Painlevé equations, vector fields, and ranks in differential fields, preprint (2016). arXiv:1608.04756. https://arxiv.org/abs/1608.04756
8. James Freitag and Thomas Scanlon, Strong minimality and the j-function, Journal of the European Mathematical Society 20(1) (2018) 119-136. DOI: 10.4171/jems/761; arXiv:1402.4588. https://doi.org/10.4171/jems/761
9. Vahagn Aslanyan, Ax-Schanuel and strong minimality for the j-function, Annals of Pure and Applied Logic 172(1) (2021) 102871. DOI: 10.1016/j.apal.2020.102871; arXiv:1805.03985. https://doi.org/10.1016/j.apal.2020.102871
10. Guy Casale, James Freitag, and Joel Nagloo, Ax-Lindemann-Weierstrass with derivatives and the genus 0 Fuchsian groups, Annals of Mathematics (2) 192(3) (2020) 721-765. DOI: 10.4007/annals.2020.192.3.2. https://doi.org/10.4007/annals.2020.192.3.2
11. James Freitag, Rémi Jaoui, David Marker, and Joel Nagloo, On the equations of Poizat and Liénard, International Mathematics Research Notices 2023(19) (2023) 16478-16539. DOI: 10.1093/imrn/rnac322; arXiv:2201.03838. https://doi.org/10.1093/imrn/rnac322
12. Rémi Jaoui, On the density of strongly minimal algebraic vector fields, preprint (2023, revised 2025). arXiv:2301.06362. https://arxiv.org/abs/2301.06362
13. James Freitag and Joel Nagloo, Algebraic relations between solutions of Painlevé equations, Journal of the London Mathematical Society 111(6) (2025) e70190. DOI: 10.1112/jlms.70190; arXiv:1710.03304. https://doi.org/10.1112/jlms.70190
14. Yutong Duan and Joel Nagloo, Algebraic independence of the solutions of the classical Lotka-Volterra system, Annals of Pure and Applied Logic 177(5) (2026) 103707. DOI: 10.1016/j.apal.2025.103707; arXiv:2502.17194. https://doi.org/10.1016/j.apal.2025.103707