The conjectures of Kumbhakar, Roy, and Srinivasan
James Freitag, Omar León Sánchez, Wei Li, Joel Nagloo · 2026
Read The conjectures of Kumbhakar, Roy, and Srinivasan by James Freitag, Omar León Sánchez, Wei Li, Joel Nagloo (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-12387.txt.
Introduction
Freitag, León Sánchez, Li, and Nagloo prove two conjectures on how many independent solutions algebraic differential equations can hold. Both bounds come from recent work in differential Galois theory read through model theory. Stronger forms hold over the field of constants, where the equations are autonomous. Inverse Galois theory shows the bounds are optimal over fields finitely generated over their constants. Tools from Jaoui and Moosa on abelian reductions also recover parts of the earlier work in this line.
Catalog record
| CREATOR | James Freitag, Omar León Sánchez, Wei Li, Joel Nagloo |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 59987 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.12387
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The conjectures of Kumbhakar, Roy, and Srinivasan Thanks: JF is supported by NSF grant DMS-2452197 and CAREER award 1945251. OLS is partially supported by EPSRC grant EP/V03619X/1. WL is partially supported by National Key R&D Program of China 2023YFA1009401, NSFC Grants (No.12122118, 12288201), CAS Project for Young Scientists in Basic Research with Grant No. YSBR-008. JN is partially supported by NSF grant DMS-2348885. James Freitag, Omar León Sánchez, Wei Li, Joel Nagloo Abstract . We prove a conjecture of Kumbhakar, Roy, and Srinivasan (2024) on the classification of order one differential equations, and a conjecture of Kumbhakar and Srinivasan (2025) on higher order equations. Both conjectures are shown to be results of recent work in differential Galois theory. In both cases, stronger versions of the conjectures hold when working over the field of constants. We use inverse Galois theory to show the conjectures cannot be improved over any nonconstant differential field. We also show how certain recent results of Jaoui and Moosa (2024) on abelian reductions of differential equations can be recovered from the work of Kumbhakar and Srinivasan (2025) and vice versa. 2010 Mathematics Subject Classification 03C60, 03C98, 12H05 1. Introduction In [ 4 ] it is shown that given X an order one differential equation over a differential field (k,\delta) of characteristic zero, if any four solutions y_{1},y_{2},y_{3},y_{4} of X are algebraically independent over k , then all solutions are algebraically independent over k . When k equals its constant subfield C , four can be replaced by two. In the context of considering solutions in an arbitrary differential field extending k (or even arbitrary meromorphic solutions), there is no meaningful way to change the above universal quantification over solutions to an existential one - one might always build a differential field extension of k containing arbitrarily many independent solutions of X . When restrictions are placed upon the field extensions of k under consideration, such an existential version becomes possible. In [ 20 , Conjecture 1.2] , the authors consider only differential field extensions K of k adding no new constants, conjecturing: Conjecture 1.1 . [ 20 ] A first order differential equation over k (respectively, an autonomous differential equation over C ) is not of general type if and only if it has at most three (respectively, at most one) algebraically independent solutions in any given differential field extension of k having C as its field of constants. In the above conjecture, one can assume without loss of generality that the differential field k is algebraically closed and that the differential field extending k having C as its field of constants is k^{\operatorname{diff}} , the differential closure of k . We will see that the conjecture is closely related to an existential version of the above result of [ 4 ] : Let X be a first order differential equation over k . If there exist four algebraically independent solutions in k^{\operatorname{diff}} to X , then there exist infinitely many algebraically independent solutions of X in k^{\operatorname{diff}} . When X is autonomous, four can be replaced by two. Modulo understanding the model theoretic content of the definition of an order one equation of general type, the above conjecture of Kumbhakar, Roy, and Srinivasan is essentially equivalent to this existential version of the above result of [ 4 ] . We prove Conjecture 1.1 and some extensions in Section 3 . In Section 4 , we show that over any nonconstant differential field, the condition of having four algebraically independent solutions can not be improved to any smaller number of independent solutions. In a recent article [ 21 ] , Kumbhakar and Srinivasan independently give a proof of the conjecture. One advantage of the proof given here is that we obtain more information in the case that the equation is of general type. Our proof also applies in the case of partial differential equations whose solutions lie in a differential field with finitely many commuting derivations. In [ 21 ] , Kumbhakar and Srinivasan give a higher order analog of Conjecture 1.1 , proving the result for equations of order at most four. We next introduce the notation to understand their higher order analog of Conjecture 1.1 . Consider the equation (1.1) f(x,x^{\prime},\ldots x^{(n)})=0 where f is a polynomial over k . Let y be a nonalgebraic solution to equation ( 1.1 ). Let \mathfrak{S}_{k,f,y} be the set of solutions of ( 1.1 ) satisfying: (1) For any y_{1}\in\mathfrak{S}_{k,f,y} , there is a differential k -isomorphism between the differential fields k\langle y\rangle and k\langle y_{1}\rangle that maps y to y_{1} . (2) For any distinct y_{1},\ldots,y_{m}\in\mathfrak{S}_{k,f,y} , the field of constants of k\langle y_{1},\ldots,y_{m}\rangle is C and \text{tr.deg}_{k}k\langle y_{1},\ldots,y_{m}\rangle=\sum_{i=1}^{m}\text{tr.deg}_{k}k\langle y_{i}\rangle . Remark 1.2 . The above definition is given in [ 21 ] , and there may be some questions around the specific language defining the set \mathfrak{S}_{k,f,y} , since the implied uniqueness of this set can only be up to isomorphism of differential fields over k . We would have perhaps defined \mathfrak{S}_{k,f,y} simply as a maximal set of independent generic solutions of equation ( 1.1 ) for which the field of constants of the differential field over k generated by the solutions is equal to the field of constants of k . Concretely, the issue is that if one identifies specific functions in some F\supset k , there is not a unique maximal set of functions. For instance it would be problematic if y_{1} were to be a transcendental constant multiple of y_{2} , even though both y_{1} and y_{2} themselves could be members of such sets (though not together in the same set). In any case, it is easy to see that the set is well-defined up to k -differential isomorphism. The result of [ 21 ] : Theorem 1.3 . Let k be an algebraically closed differential field and y a nonalgebraic solution of ( 1.1 ). (1) \mathfrak{S}_{k,f,y} is a finite set if and only if there is k -irreducible differential subfield F of k\langle y\rangle that can be embedded in a strongly normal extension of k and that either F=k\langle w\rangle , where w is a nonalgebraic solution of some Riccati differential equation over k , or F is an abelian extension of k , in which case, \mathfrak{S}_{k,f,y}=\{y\}. (2) Suppose that k=C . Then, \mathfrak{S}_{k,f,y}=\{y\} if and only if there is an intermediate differential field C\subset M\subset C\langle y\rangle such that either M=C(x) , where x^{\prime}=1 or x^{\prime}=cx for some nonzero c\in C , or M is an abelian extension of C . (3) Suppose that n\leq 4 . Then, \mathfrak{S}_{k,f,y} is a finite set if and only if it has at most n+2 elements. In Section 5 , we prove part (3) of the above theorem for equations of any order. When the field k is a field of constants, we prove a stronger form with the bound of 1 in place of n+2 . Our proof again works for systems of partial differential equations in a suitable context. In section 6 , we show that over any nonconstant differential field, one can not improve the bound of n+2 in the above theorem. The proof of this optimality is