Tensorial Constraints for Commuting Endomorphisms of the Generalized Tangent Bundle

1. Problem Statement

The paper addresses the problem of identifying tensorial integrability conditions for families of mutually commuting endomorphisms acting on the generalized tangent bundle ($TM = TM \oplus T^*M$) of a smooth manifold.

• Context: Generalized geometry unifies complex and symplectic structures via the generalized tangent bundle, equipped with the Courant-Dorfman bracket. Previous work established integrability conditions for single endomorphisms (e.g., generalized complex structures) using the vanishing of the shifted Courant-Nijenhuis torsion.
• The Gap: While integrability for a single endomorphism is well-understood, the conditions required for a family of commuting endomorphisms (which may include both skew-symmetric structures like complex structures and symmetric structures like Riemannian metrics) to be compatible with the Courant-Dorfman bracket were not fully characterized.
• Core Question: What are the necessary and sufficient tensorial constraints (polynomial conditions) that ensure the compatibility of a commuting family of endomorphisms with the Courant-Dorfman bracket, such that the resulting tensors are $\Omega^0_M$-linear (i.e., tensorial)?

2. Methodology

The authors employ a novel synthesis of differential geometry and commutative algebra, specifically utilizing Gröbner basis techniques.

• Algebraic Formulation:

  • The authors define a polynomial ring $R = \mathbb{R}[x_1, \dots, x_N, y_1, \dots, y_N, z_1, \dots, z_N]$.
  • They introduce a polynomial action of $R$ on the space of trilinear forms $\Theta$ via the commuting family $\phi = (\phi_1, \dots, \phi_N)$.
  • A polynomial $P \in R$ is defined as universally tensorial if the action $P \bullet_\phi \tau_C$ (where $\tau_C$ is the Courant element) yields a tensor for any commuting family $\phi$ of a given signature $\epsilon$ (where $\epsilon_i = \pm 1$ denotes symmetry or skew-symmetry).
  • The set of all such polynomials forms an ideal $I_\epsilon \subset R$.

• Geometric Translation:

  • The condition of tensoriality is translated into the vanishing of the polynomial on a specific algebraic subvariety $V_\epsilon \subset \mathbb{C}^{3N}$.
  • By the Real Nullstellensatz, the ideal $I_\epsilon$ corresponds to the real radical of the intersection of three ideals: $I_x^\epsilon \cap I_y^\epsilon \cap I_z^\epsilon$, defined by linear relations between the variables based on the signature $\epsilon$.

• Computational Algebra:

  • To find the generators of $I_\epsilon$, the authors use Gröbner basis techniques to compute the intersection of ideals.
  • They utilize Frobenius splitting and the theory of Knutson ideals to prove that the intersection of the ideals is equal to their product (in specific cases), simplifying the computation.
  • A key technical innovation is the reduction of the general $N$-dimensional Gröbner basis computation to the $N=6$ case. They prove that any S-polynomial computation involving arbitrary indices can be mapped via an order-preserving bijection to a computation involving at most 6 indices, allowing for computer-assisted verification of the general case.

3. Key Contributions

A. Identification of Tensorial Ideals

The paper explicitly identifies the ideal $I_\epsilon$ of universally tensorial polynomials. The authors prove that this ideal is generated by two distinct types of polynomials:

1. Cubic Generators ($T_{\epsilon}^{ijk}$): These are direct generalizations of the shifted Courant-Nijenhuis torsion to multiple endomorphisms. They exist for any signature.
2. Quadratic Generators ($P_{\epsilon}^{ij}$): These appear only when the family contains at least two symmetric endomorphisms ($\epsilon_i = \epsilon_j = 1$). These have no analogue in the single-endomorphism case.

B. The Semiconcomitant

The authors introduce the semiconcomitant $K_\phi$, a bivariate version of the Courant-Nijenhuis torsion. Both the cubic and quadratic generators are constructed from this object.

• $T_{\phi}^{ijk}$ relates to the semiconcomitant of pairs of endomorphisms.
• $P_{\phi}^{ij}$ represents the antisymmetrization of the semiconcomitant for symmetric pairs.

C. Connection to Frobenius Splitting and Knutson Ideals

The paper establishes a surprising link between these geometric constraints and Frobenius splitting, a concept from algebraic geometry in positive characteristic. The authors show that the ideals defining the tensorial constraints are Knutson ideals, which implies they are radical and have specific geometric properties (e.g., rational singularities).

D. Uniform Integrability for Semisimple Families

For semisimple commuting families (where the generalized tangent bundle decomposes into common eigenbundles), the authors recast the vanishing of the tensorial ideal as involutivity constraints on the common eigenbundles.

• They derive conditions under which the Courant-Dorfman bracket of sections of eigenbundles $L_\lambda$ and $L_\mu$ must lie within specific sums of eigenbundles.
• This generalizes the known result that a generalized complex structure is integrable iff its $\pm i$ eigenbundles are involutive.

4. Key Results

1. Main Theorem (Theorem 26): The ideal $I_\epsilon$ is generated by the set:
   $$G_\epsilon = \{ T_{\epsilon}^{ijk} \mid 1 \le i,j,k \le N \} \cup \{ P_{\epsilon}^{ij} \mid \epsilon_i = \epsilon_j = 1 \}$$
   where $T_{\epsilon}^{ijk}$ are cubic polynomials and $P_{\epsilon}^{ij}$ are quadratic polynomials.

2. Reduction to $N=6$: The Gröbner basis computation for the intersection of ideals $I_x^\epsilon \cap I_y^\epsilon \cap I_z^\epsilon$ for arbitrary $N$ reduces to the case $N=6$. This was verified computationally for all signatures.

3. Uniform Integrability Conditions (Theorem 29): A semisimple commuting family is uniformly integrable if and only if:

   • The bracket of eigenbundles $L_\lambda, L_\mu$ is contained in $L_{\lambda + \mu}$ (scaled by eigenvalue differences).
   • For symmetric pairs, the bracket is contained in a specific Whitney sum of eigenbundles determined by a determinant condition involving the eigenvalues.

4. Special Cases:

   • Generalized Kähler Structures: Recovered as a commuting pair of generalized complex structures (or complex structure and metric) where the involutivity of all four eigenbundles is required.
   • Commuting Metrics: For two commuting generalized Riemannian metrics, uniform integrability implies the metrics are identical (if positive definite) or that the associated metric product structure is "bundle-like" with respect to the foliations defined by the eigenbundles.

5. Significance

• Unification: The paper provides a unified algebraic framework for understanding integrability in generalized geometry, extending beyond complex structures to include metrics and polynomial structures.
• New Constraints: It reveals that commuting symmetric endomorphisms (metrics) impose new, quadratic tensorial constraints that do not exist for single endomorphisms, fundamentally changing the landscape of integrability for metric structures.
• Methodological Innovation: The application of Gröbner bases and Knutson ideals to differential geometry problems is a significant methodological breakthrough. It allows for the rigorous derivation of tensorial constraints that would be extremely difficult to guess or prove using traditional differential geometric methods alone.
• Algebraic-Geometric Bridge: The connection to Frobenius splitting suggests that the moduli spaces of these generalized structures may possess deep algebraic properties (e.g., being reduced, having rational singularities), opening new avenues for research in the intersection of generalized geometry and algebraic geometry.

In summary, this paper rigorously characterizes the compatibility conditions for commuting generalized structures, identifying a complete set of tensorial generators using advanced commutative algebra, and applying these results to generalize classical integrability theorems.