Source-linked AI summary
An Extension of Bochner's Problem: Exceptional Invariant Subspaces
David Gomez-Ullate, Niky Kamran, Robert Milson
TL;DR
The paper asks whether Bochner’s result can be extended when the polynomial sequence does not begin with a constant. It classifies codimension-one exceptional subspaces and constructs new complete orthogonal systems from second-order equations with rational coefficients. The resulting X1-Jacobi and X1-Laguerre families form self-adjoint Sturm–Liouville systems, while the classification identifies them alongside the classical systems.
Problem
The paper examines whether polynomial sequences solving Heine–Stieltjes equations with m > 0 can be defined when the sequence begins with positive degree rather than a constant.
Method
The authors classify codimension-one exceptional polynomial subspaces under projective transformations and use them to construct second-order differential operators preserving those subspaces.
Results
The X1-Jacobi and X1-Laguerre polynomials form self-adjoint Sturm–Liouville systems, and the classification also includes classical orthogonal polynomial systems up to affine transformation.
Takeaways & Limitations
Dropping the requirement that the sequence begin with a constant yields new complete orthogonal polynomial systems associated with exceptional codimension-one subspaces.
Takeaways & Limitations
The operator classification assumes a second-order differential operator mapping a codimension-one subspace M into P_n, with specified root multiplicities for q_M at infinity and zero.
Abstract
from arXiv · showhide
A classical result due to Bochner characterizes the classical orthogonal polynomial systems as solutions of a second-order eigenvalue equation. We extend Bochner's result by dropping the assumption that the first element of the orthogonal polynomial sequence be a constant. This approach gives rise to new families of complete orthogonal polynomial systems that arise as solutions of second-order eigenvalue equations with rational coefficients. The results are based on a classification of exceptional polynomial subspaces of codimension one under projective transformations.
1. INTRODUCTION AND STATEMENT OF RESULTS
The paper extends Bochner’s characterization by allowing polynomial sequences to begin at positive degree, yielding exceptional orthogonal systems governed by second-order equations with rational coefficients. It classifies codimension-one exceptional subspaces and identifies the resulting X1-Jacobi and X1-Laguerre families within self-adjoint Sturm–Liouville problems.
- Main construction: Complete polynomial sequences relative to positive-definite measures produce new families of orthogonal polynomial systems.These systems arise from exceptional codimension-one subspaces rather than the usual sequence beginning with degree zero.
- Extension of Bochner’s theorem: The extension theorem characterizes second-order operators whose eigenvalue equations have polynomial solutions in every degree n = 1, 2, 3, ... but not degree zero.Conversely, such operators have the stated form up to an additive constant and satisfy the listed conditions.
- Exceptional subspaces: For n ≥ 5, every codimension-one exceptional polynomial subspace is projectively equivalent to the classified X1 space.This restriction is the key ingredient in proving the converse direction of the extension theorem.
- Orthogonal systems: The X1-Jacobi and X1-Laguerre families are eigenfunctions of self-adjoint Sturm–Liouville problems with semi-bounded, pure-point spectra.The converse classification includes these two families, together with classical orthogonal polynomial systems, up to affine changes of variable.
- Orthogonal systems: Unlike general rational modifications of classical weights, the X1-Jacobi and X1-Laguerre polynomials satisfy a Sturm–Liouville problem rather than only degree-dependent second-order equations.The distinction concerns whether the differential equation’s coefficients depend explicitly on the polynomial degree.
2. THE EQUIVALENCE PROBLEM FOR CODIMENSION ONE SUBSPACES
The section classifies codimension-one polynomial subspaces through an SL(2,R)-equivariant covariant, reducing projective equivalence to the normalization of degree-n polynomial roots.
- Projective classification: The invariant bilinear form γ is symmetric for even n and skew-symmetric for odd n, and it is G-invariant.This invariance yields the equivariant correspondence used to represent a subspace M by Φ(M).
- Fundamental covariant: A codimension-one subspace M is represented by the polynomial qM, characterized up to scalar multiple by the annihilation condition γ(u,v)=0 for every v in M.Given a basis of M, Φ(M) can be computed by solving the resulting n linear equations.
- Projective classification: The SL(2,R) action on codimension-one subspaces is classified using a covariant that identifies each subspace with a degree-n polynomial up to projective equivalence.The construction uses invariant multilinear maps and an equivariant isomorphism between codimension-one subspaces and projective polynomial space.
- Root normalization: Projective transformations normalize roots by sending the highest-multiplicity root to infinity, the next to zero, and the third to one.The resulting signature partition and remaining root locations fully solve the polynomial equivalence problem.
- Examples and normalized bases: Examples include the monomial subspace with qM=1, the exceptional monomial subspace with qM=x^(n−1), and single-gap spaces with qM=x^(n−j).The exceptional monomial examples are projectively equivalent and are identified as X1 exceptional subspaces.
- Examples and normalized bases: When qM has multiplicities λ at infinity and μ at zero, Proposition 2.5 supplies a basis of monomials and binomials for M.This gives a direct basis-level classification indexed by the root multiplicities of the covariant.
3. OPERATORS PRESERVING POLYNOMIAL SUBSPACES
The section studies second-order operators preserving polynomial subspaces, identifies exceptional codimension-one spaces, and connects their classification to the extension of Bochner’s theorem.
- Operators preserving Pn: Burnside’s theorem characterizes operators preserving Pn as quadratic elements of the enveloping algebra of the sl(2,R) generators.Such operators are often called Lie-algebraic operators.
- Rational coefficients: Assuming three linearly independent polynomial inputs and polynomial outputs forces the coefficients of a second-order operator to be rational functions.The coefficients are obtained by inverting a nonsingular linear system.
- Exceptional subspaces: For general polynomial subspaces, an operator preserving the subspace need not preserve Pn, so Burnside’s characterization does not apply directly.This distinction motivates the separate analysis of exceptional invariant subspaces.
- Exceptional subspaces: The monomial exceptional spaces and the broader Ea,b_n spaces are projectively equivalent, and the latter are shown to be X1 exceptional subspaces.The operator J5 preserves Ea,b_n but not Pn, establishing exceptionality; projective equivalence transfers the result to the related spaces.
- Classification and Bochner extension: Theorem 1.4 asserts that Ea,b_n are the only codimension-one exceptional subspaces, and this classification is then used to prove the extension of Bochner’s theorem.The argument proceeds from operator-preserving subspace classification to the polynomial eigenvalue result.
4. PROOF OF THEOREM 1.4
The proof shows that, under a root-multiplicity condition on a codimension-one subspace M, every second-order operator preserving M also preserves the full polynomial space Pn.
- Theorem 4.1 states that if n ≥ 5 and every root of qM has multiplicity at most n − 2, then D2(M) ⊂ D2(Pn).
- Operator decomposition: The proof expands rational-coefficient operators into homogeneous degree components Tk, where Tk maps xj to a scalar multiple of xj+k.
- Vanishing lemmas: An operator annihilating three distinct monomials must vanish, allowing the proof to eliminate high-degree components under the subspace constraints.
- Case analysis: Separate lemmas handle simple roots and the cases λ ≤ n − 3, λ = n − 2, λ ≥ 3, and λ = 2 with µ ≤ 2.
- Projective normalization: Projective transformations move a root of maximal multiplicity to ∞ and another root to 0, reducing the classification to multiplicities λ and µ.
- Conclusion: After showing Tk vanishes outside the permitted degree range, each remaining component preserves Pn, so their sum does as well.
5. PROOF OF THEOREM 1.2
The proof of Theorem 1.2 classifies rational-coefficient second-order operators with polynomial eigenfunctions in every positive degree but not degree zero, identifying their exceptional subspace structure and operator form.
- Assumptions: The converse assumes polynomial eigenfunctions Pn of degree n for every n ≥ 1, while the degree-zero eigenfunction is absent.
- Exceptional subspaces: For n ≥ 5, each codimension-one subspace Mn is forced into the exceptional form Ean,bn n rather than the alternative Ean n.
- Parameter constraints: The subspaces have the representation Mn = ⟨x − cn, (x − bn)2, . . . , (x − bn)n⟩, with parameters constrained across n.
- Parameter constraints: Compatibility of the eigenvalue equations for P1, P2, and P3 forces the parameters b and c to be independent of n.
- Operator classification: Using the classification of operators preserving the exceptional subspaces, T is reduced, up to an additive constant, to the form in (11).
- Conclusion: The resulting operator satisfies the required condition p(b) ≠ 0, completing the reverse implication of Theorem 1.2.