The generalized Ritt problem: uniform bounds on differentiations for detecting components of differential ideals, equivalently definability in families of irreducibility of differential varieties

The construction of differential Chow varieties — parameter spaces for differential algebraic cycles — by Freitag, Li, and Scanlon deliberately routed around, and thereby highlighted, a central unresolved obstruction in differential algebra: the Ritt problem. In its modern form the problem asks whether there is a uniform bound $r(d,n,m)$, depending only on the shape of the input system (number of variables and derivations, degrees, orders) and not on the base differential field, on the order of differential polynomials needed to test primality of a radical differential ideal — equivalently, on the number of differentiations needed to detect non-primality or to decide membership in the generic component of a differential variety. Equivalently, it asks whether irreducibility of differential varieties (and, closely related, the Kolchin closure) is definable in families in the theory $\mathrm{DCF}_0$ of differentially closed fields of characteristic zero. The problem traces from Ritt (1950) and Kolchin through Golubitsky–Kondratieva–Ovchinnikov, Harrison-Trainor–Klys–Moosa, and Freitag–León Sánchez–Li, and remains open as of 2026, with only partial results known: bounds for one quantifier of the primality test, effective bounds for radical membership (differential Nullstellensatz), computable bounds for the number of components, and definability in families of Kolchin polynomials and of weak irreducibility.

Background

Differential algebra studies systems of polynomial differential equations algebraically. A differential field is a field $k$ equipped with one or more commuting derivations $\delta_1,\dots,\delta_m$; the case $m=1$ is called ordinary and $m>1$ partial. Over $k$ one forms the differential polynomial ring $k\{x_1,\dots,x_n\}$, the polynomial ring in the indeterminates $x_i$ together with all their derivatives $\delta_1^{a_1}\cdots\delta_m^{a_m}x_i$. A differential ideal is an ideal closed under the derivations, and for a finite set $S$ of differential polynomials, $\{S\}$ denotes the radical differential ideal generated by $S$, the smallest differential ideal containing $S$ that is also a radical ideal. A differential ideal is prime exactly when the differential variety $V(S)$ it defines is irreducible in the sense of Ritt and Kolchin, meaning it is not the union of two proper differential subvarieties. A characteristic set, in the sense of Ritt's and Wu–Ritt's elimination method, is a canonical finite generating set for a prime differential ideal adapted to a fixed ordering of the derivative indeterminates; every prime differential ideal is generated, as a radical differential ideal, by any of its characteristic sets.

For algebraic (non-differential) polynomial systems the analogous boundedness question was settled classically: Hermann, Seidenberg, and later van den Dries and Schmidt gave uniform degree bounds for testing primality of radical ideals, and consequently, over an algebraically closed field, the irreducible members of any algebraic family of varieties form a definable (constructible) subfamily. The differential case fails at exactly this point: primality of $\{S\}$ requires quantifying over pairs of differential polynomials $f,g$ with $fg\in\{S\}$, and a witness to non-primality may appear only after many differentiations of the defining equations — for Ritt's example $f=(y^{(2)})^2-y$ one must differentiate three times before the reducibility of $V(f)$ becomes algebraically visible. No bound on the number of required differentiations in terms of the shape (number of variables and derivations, degrees, and orders) of the input system is known. Joseph F. Ritt posed the problem of computing the prime components of the radical differential ideal generated by a finite system in his 1950 monograph; Kolchin later listed the inclusion problem for prime differential ideals given by characteristic sets as Problem 3 of §IV.9 of his book, and R. M. Cohn studied the question through specializations of differential kernels and valuation theory in the 1970s and 1980s. Golubitsky, Kondratieva, and Ovchinnikov called the computational question — given characteristic sets of two prime differential ideals, decide whether one contains the other — the generalized Ritt problem, and showed it equivalent to several other existence-of-bounds questions, including the definability of zero-divisors modulo radical differential ideals. Harrison-Trainor, Klys, and Moosa recast the problem as the definability of primality conjecture: for each $d$ there should exist $r=r(d,n,m)$, independent of the base field, such that primality of $\{S\}$ for systems $S$ of degree and order at most $d$ is always detected by witnesses of degree and order at most $r$; they proved this conjecture equivalent to four other natural bounds conjectures and obtained a partial solution bounding one of the two quantifiers.

The model-theoretic formulation uses $\mathrm{DCF}_0$, the theory of differentially closed fields of characteristic zero (with $m$ commuting derivations), which is the model companion of the theory of differential fields and the setting in which the geometry of differential varieties is tame. A definable family of differential varieties is a family $(V_a)_{a\in B}$ in which the parameter set $B$ and each fiber are given uniformly by formulas; a property of varieties is definable in families when the parameters of the fibers having the property always form a definable set. The Kolchin closure is the differential analogue of the Zariski closure, the smallest differential algebraic variety containing a given set. The Kolchin polynomial of a differential variety is a numerical polynomial measuring the transcendence growth of its generic points and provides the differential-algebraic notion of rank; a variety is weakly irreducible if it has a unique component of maximal Kolchin polynomial.

This is the obstruction left open by the construction of differential Chow varieties. The classical Chow variety is a parameter space for effective algebraic cycles of given dimension and degree; Gao, Li, and Yuan developed the differential Chow form and constructed differential Chow varieties only in special cases, leaving general existence open. Freitag, Li, and Scanlon proved that differential Chow varieties exist in general, but precisely because the Ritt problem is unsolved their proof works around it with a mixed strategy of prolongation spaces (the jet-type algebraic varieties $B_h(V)$ encoding a differential variety $V$ together with its first $h$ prolongations) and characteristic sets, rather than parameterizing directly by generators of differential ideals. They record explicitly that neither the Kolchin closure nor irreducibility of differential varieties is known to be definable in families in $\mathrm{DCF}_0$, and they remark that a proof of definable Kolchin closure would greatly simplify the construction, allowing the differential Chow coordinates of differential cycles to form a differentially constructible set directly.

Problem Statement

Let $k$ be a differential field of characteristic zero equipped with $m$ commuting derivations, and let $S$ be a finite system of differential polynomials in $k\{x_1,\dots,x_n\}$ of degree and order at most $d$; write $\{S\}$ for the radical differential ideal generated by $S$.

(Bound form, the definability of primality conjecture.) Do there exist, for each $d,n,m$, numbers $r=r(d,n,m)$, independent of the base field $k$, such that for every such $k$ and $S$: the ideal $\{S\}$ is prime if and only if $\{S\}$ is proper and every pair of differential polynomials $f,g$ of degree and order at most $r$ with $fg\in\{S\}$ satisfies $f\in\{S\}$ or $g\in\{S\}$? Equivalently: is there a uniform bound, depending only on the shape (number of variables, derivations, degrees, and orders) of the input system, on the number of differentiations of the input equations needed to detect non-primality of the differential ideal they generate, or to decide whether a given point — for instance a point at which the separants of a characteristic set vanish — belongs to the generic component rather than a singular component of the corresponding differential variety?

(Definability form, the question recorded in Freitag–Li–Scanlon.) Working over a differentially closed field modeling $\mathrm{DCF}_0$: is irreducibility of differential varieties definable in families, that is, for every definable family $(V_a)_{a\in B}$, is the set of parameters $a$ for which $V_a$ is irreducible itself definable? Is the Kolchin closure of differential varieties definable in families?

The bound form, the definability of irreducibility, and the Kolchin Problem 3 inclusion decision are equivalent to one another (Golubitsky–Kondratieva–Ovchinnikov, Theorem 1; Harrison-Trainor–Klys–Moosa, Theorem 5.7; Freitag–León Sánchez–Li). The definability of the Kolchin closure is a closely related further question, recorded by Freitag–Li–Scanlon alongside irreducibility; a strict equivalence between it and the irreducibility question is not pinned down in the cited literature. The problem is open already in the ordinary case of a single derivation ($m=1$). The question is whether such $r(d,n,m)$ exists; a refutation is an explicit construction assigning to each $r$ a system of uniformly bounded shape whose non-primality is not witnessed by any pair of differential polynomials of degree and order at most $r$.

References

1. James Freitag, Wei Li, Thomas Scanlon, Differential Chow varieties exist, Journal of the London Mathematical Society 95 (2017), 128–156, with an appendix by William Johnson. DOI 10.1112/jlms.12002; arXiv:1504.03755. https://arxiv.org/abs/1504.03755
2. James Freitag, Omar León Sánchez, Wei Li, Effective definability of Kolchin polynomials, Proceedings of the American Mathematical Society 148 (2020), 1455–1466. DOI 10.1090/proc/14869; arXiv:1806.02060. https://arxiv.org/abs/1806.02060
3. Oleg Golubitsky, Marina Kondratieva, Alexey Ovchinnikov, On the generalized Ritt problem as a computational problem, Journal of Mathematical Sciences 163 (2009), no. 5, 515–522 (the arXiv version title spells "generalised"). DOI 10.1007/s10958-009-9689-3; arXiv:0809.1128. https://arxiv.org/abs/0809.1128
4. Matthew Harrison-Trainor, Jack Klys, Rahim Moosa, Nonstandard methods for bounds in differential polynomial rings, Journal of Algebra 360 (2012), 71–86. DOI 10.1016/j.jalgebra.2012.03.013; arXiv:1105.0600. https://arxiv.org/abs/1105.0600
5. Wei Li, Alexey Ovchinnikov, Gleb Pogudin, Thomas Scanlon, Algorithms yield upper bounds in differential algebra, Canadian Journal of Mathematics 75 (2023), no. 1, 29–51. DOI 10.4153/S0008414X21000560; arXiv:2005.01608. https://arxiv.org/abs/2005.01608
6. Joseph Fels Ritt, Differential Algebra, American Mathematical Society Colloquium Publications, vol. 33, New York, 1950. zbMATH Zbl 0037.18402 (original edition; the Dover reprint is reviewed at Zbl 0141.03801). https://zbmath.org/3056446
7. E. R. Kolchin, Differential Algebra and Algebraic Groups, Pure and Applied Mathematics, vol. 54, Academic Press, New York, 1973. zbMATH Zbl 0264.12102. https://zbmath.org/3415973
8. Xiao-Shan Gao, Wei Li, Chun-Ming Yuan, Intersection theory in differential algebraic geometry: generic intersections and the differential Chow form, Transactions of the American Mathematical Society 365 (2013), 4575–4632. DOI 10.1090/s0002-9947-2013-05633-4. https://doi.org/10.1090/s0002-9947-2013-05633-4
9. Ehud Hrushovski, Masanori Itai, On model complete differential fields, Transactions of the American Mathematical Society 355 (2003), 4267–4296. DOI 10.1090/s0002-9947-03-03264-1. https://doi.org/10.1090/s0002-9947-03-03264-1
10. Anand Pillay, Martin Ziegler, Jet spaces of varieties over differential and difference fields, Selecta Mathematica 9 (2003), 579–599. DOI 10.1007/s00029-003-0339-1. https://doi.org/10.1007/s00029-003-0339-1
11. Richard M. Cohn, Specializations of differential kernels and the Ritt problem, Journal of Algebra 61 (1979), 256–268. DOI 10.1016/0021-8693(79)90316-8. https://doi.org/10.1016/0021-8693(79)90316-8
12. Richard M. Cohn, Valuations and the Ritt problem, Journal of Algebra 101 (1986), no. 1, 1–15. DOI 10.1016/0021-8693(86)90093-1. https://doi.org/10.1016/0021-8693(86)90093-1
13. Oleg Golubitsky, Marina Kondratieva, Alexey Ovchinnikov, Agnes Szanto, A bound for orders in differential Nullstellensatz, arXiv:0803.0160 (2008). https://arxiv.org/abs/0803.0160
14. Omar León Sánchez, Alexey Ovchinnikov, On bounds for the effective differential Nullstellensatz, Journal of Algebra 449 (2016), 1–21. DOI 10.1016/j.jalgebra.2015.10.009; arXiv:1508.07508. https://arxiv.org/abs/1508.07508
15. Alexey Ovchinnikov, Gleb Pogudin, Thieu N. Vo, Bounds for elimination of unknowns in systems of differential-algebraic equations, International Mathematics Research Notices 2022 (2022), no. 16, 12342–12377. DOI 10.1093/imrn/rnaa302; arXiv:1610.04022. https://arxiv.org/abs/1610.04022
16. Wei Li, Partial differential Chow forms and a type of partial differential Chow varieties, arXiv:1709.02358 (2017). https://arxiv.org/abs/1709.02358
17. Wei Li, Yinghong Li, Computation of differential Chow forms for prime differential ideals, arXiv:1501.02755 (2015). https://arxiv.org/abs/1501.02755
18. William Simmons, Henry Towsner, Proof mining and effective bounds in differential polynomial rings, arXiv:1609.07509 (2016). https://arxiv.org/abs/1609.07509