⚛️ high-energy theory

A local action for Kodaira--Spencer gravity

This paper presents a local, unconstrained action for Kodaira–Spencer gravity formulated as a cyclic $L_\infty$ algebra based on a Courant algebroid, which reproduces double field theory's gauge structure and linearized dynamics while introducing parity-odd non-linear interactions without requiring a level-matching constraint.

Original authors: Olaf Hohm

Published 2026-11-03

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Olaf Hohm

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Gravity is the force that holds the universe together, yet for physicists, it remains the most stubborn puzzle. While the other fundamental forces of nature are described by quantum mechanics with great precision, gravity resists being squeezed into the same framework. For decades, scientists have searched for a deeper theory that could unify these forces, often looking to string theory for answers. Within this vast landscape, there exists a peculiar, simplified version of gravity known as Kodaira–Spencer gravity. It is not the gravity we feel on Earth, but a topological version that lives on a special kind of six-dimensional shape called a Calabi–Yau manifold. Think of this manifold as a complex, multi-dimensional fabric where the very rules of geometry can be stretched and twisted. In this realm, gravity is not described by the bending of space and time, but by the deformation of the shape's internal structure. For years, physicists could describe the equations governing this deformation, but they could not find a local, step-by-step recipe—an action principle—that would generate those equations from first principles. Without such a recipe, it is difficult to understand how the theory behaves when things get complicated or to connect it to other areas of physics.

A researcher at Humboldt University in Berlin has now solved this long-standing problem by constructing a local action for Kodaira–Spencer gravity. The breakthrough came from a clever change of perspective. Previously, the theory relied on a mathematical constraint that made the equations non-local, meaning the state of the field at one point depended on the state of the field everywhere else in a way that was hard to calculate directly. The author realized that this constraint could be solved by introducing a new, simpler field, essentially a pre-existing blueprint that generates the complex gravitational field. By rewriting the theory in terms of this new field, the messy, non-local equations transformed into a clean, local action. This new formulation is local because it only involves interactions between fields at the same point in space, making the theory much more manageable and physically transparent.

The implications of this new formulation are surprisingly rich. When the author examined the symmetries of this new theory—the rules that dictate how the fields can be shifted without changing the physics—they discovered that the symmetry structure is identical to that of a theory called double field theory. Double field theory is a modern approach to string theory that treats space and its dual in a symmetric way, often requiring a special constraint to reduce the number of dimensions back to the familiar four. However, in this new Kodaira–Spencer formulation, the theory works perfectly without needing that extra constraint. It is an "unconstrained" version of double field theory, suggesting that the deep mathematical structures underlying string theory might be more flexible and robust than previously thought. The gauge algebra, which describes how these symmetries interact, forms a sophisticated mathematical structure known as an L-infinity algebra, built upon a framework called a Courant algebroid. This structure is not just a mathematical curiosity; it is the same type of structure that appears in other advanced theories of gravity and supergravity, linking this topological model to the broader quest for a unified theory of physics.

To test the physical relevance of this abstract construction, the author reduced the theory from its six-dimensional home down to a simple three-dimensional slice. In this lower-dimensional setting, the complex gravitational field splits into two distinct parts: a symmetric part that behaves exactly like a massless particle with spin two, and an antisymmetric part that behaves like a magnetic field known as a Kalb–Ramond field. The resulting action describes the free movement of these particles, matching the standard Fierz–Pauli action for gravity and the Kalb–Ramond action for the magnetic field. This reduction confirms that the theory contains the seeds of standard gravity. However, when the author looked at the interactions between these particles, the theory revealed something unusual. Unlike Einstein's theory of gravity, which involves smooth, two-derivative interactions, this theory produces interactions that involve three derivatives and violate parity, meaning the physics looks different if you view it in a mirror. These interactions are cubic, involving three fields at once, and they are the only interactions present in the theory.

The paper also explores what happens if one tries to remove the magnetic field component from this three-dimensional theory. The author found that integrating out this field leads to non-local effects, but these can be removed by redefining the gravitational field in a specific way. Up to a certain level of complexity, this process reproduces a known, exotic model of three-dimensional gravity that was discovered by other researchers. This connection suggests that the new local action is not just a mathematical reformulation but a genuine physical theory that encompasses known exotic gravity models. Furthermore, the author provided a "double copy" interpretation of the theory, showing that the complex mathematical structure of the gravity theory can be built from two simpler, independent mathematical objects. One object is related to the geometry of the manifold, and the other is related to the symmetries of the space. This double copy structure is a powerful tool in modern physics, often used to calculate complex gravitational interactions by squaring simpler gauge theory results. Here, it provides a deep insight into why the theory works and how its components fit together.

The work also addresses a subtle mathematical issue regarding the global shape of the manifold. In some cases, the manifold might have "holes" or topological features that prevent the simple local solution from working everywhere. The author showed how to extend the theory to include these global features, known as moduli, without breaking the local structure. These moduli act as background parameters that describe the different possible shapes the manifold can take, but they do not introduce new dynamic fields or change the fundamental nature of the theory. This ensures that the local action principle is robust and applicable to a wide range of geometric settings, including the specific shapes relevant to string theory and holography.

Ultimately, this paper provides a complete, local, and mathematically rigorous description of Kodaira–Spencer gravity. It bridges the gap between a non-local, topological theory and a local, dynamical one, revealing that the theory is a form of unconstrained double field theory with a unique set of interactions. By reducing the theory to three dimensions, the author demonstrated its connection to standard gravity and exotic gravity models, while the double copy interpretation offers a new window into the algebraic structures that underpin the universe. This work does not just solve a technical problem; it clarifies the nature of this topological gravity, showing that it is a genuine theory of gravity with a rich structure that could serve as a powerful model for understanding how gravity might be completed in higher dimensions. The author acknowledges that while the theory is mathematically consistent and locally defined, its full quantum behavior and its relation to non-topological gravity remain open questions for future investigation.

Technical Summary: A Local Action for Kodaira–Spencer Gravity

Problem Statement
Kodaira–Spencer gravity, formulated as the closed string field theory of the topological B-model (BCOV theory), describes the deformation of complex structures on a Calabi–Yau threefold $X$ . The fundamental field is the Beltrami differential $A$ , a $(1,1)$ -form representing complex structure deformations. While the linearized equations of motion ( $\bar{\partial}A = 0$ ) are well-defined, the theory historically lacks a local action principle. Previous formulations, such as the BCOV action, rely on non-local operators (specifically $\partial^{-1}\bar{\partial}$ ) derived under the constraint that the associated $(2,1)$ -form $A'$ is $\partial$ -closed. This non-locality obscures the gauge structure and complicates the analysis of the theory's relation to gravity and double field theory (DFT).

Methodology
The paper constructs a local action principle by solving the constraint $\partial A' = 0$ explicitly. The author introduces a new prepotential field $e$ , a $(1,1)$ -form, defined via the relation $A' = \partial e$ . This substitution replaces the non-local operator in the kinetic term with a local two-derivative term.

The methodology proceeds through the following steps:

Local Reformulation: By substituting $A' = \partial e$ into the non-local BCOV action, the kinetic term becomes local: $S_{kin} = -\frac{1}{2} \int_X \partial e \wedge \bar{\partial} e$ . The cubic interaction term, originally local in $A$ , is rewritten in terms of $e$ , resulting in a cubic interaction with three derivatives.
Gauge Algebra Analysis: The introduction of $e$ introduces a gauge redundancy ( $e \to e + \partial \bar{\lambda}$ ). The author analyzes the full gauge symmetry, which includes volume-preserving holomorphic diffeomorphisms. They identify the underlying algebraic structure as a Leibniz–Loday algebra based on a Courant algebroid, which is then extended to a full $L_\infty$ algebra.
$L_\infty$ and BV Formulation: The theory is encoded in a cyclic $L_\infty$ algebra defined on a "folded" subspace of the Dolbeault bicomplex. This structure allows for the construction of a Batalin–Vilkovisky (BV) master action. The author explicitly constructs the $L_\infty$ brackets ( $\ell_1, \ell_2, \ell_3$ ) and verify the classical master equation.
Double Copy Interpretation: The $L_\infty$ algebra is shown to be constructed from two "single copies": the Dolbeault algebra of antiholomorphic forms and the Roytenberg–Weinstein $L_\infty$ algebra of holomorphic forms (mapped to divergence-free vector fields). The full theory is realized as a tensor product of these algebras, analogous to the double copy in scattering amplitudes.
Dimensional Reduction: The theory is reduced to a real three-dimensional slice ( $z = \bar{z}$ ) to compare with standard gravity theories.

Key Contributions and Results

Local Action Principle: The paper provides the first local action for Kodaira–Spencer gravity. The action is local in the prepotential $e$ and includes a kinetic term with two derivatives and a cubic interaction with three derivatives.
Unconstrained Double Field Theory: The linearized gauge transformations of the field $e$ coincide exactly with those of the generalized metric fluctuation in Double Field Theory (DFT). However, unlike standard DFT, this theory does not require the level-matching (or section) constraint to be consistent. Thus, Kodaira–Spencer gravity is identified as an "unconstrained double field theory."
Reduction to 3D Gravity: On a real 3D slice, the free action reduces to the sum of the Fierz–Pauli action for a massless spin-2 field ( $h_{ij}$ $h_{ij}$ ) and the Kalb–Ramond action for a $B$ $B$ -field.
- The non-linear interactions are parity-odd and carry three derivatives, distinguishing them from the parity-even, two-derivative interactions of Einstein–Hilbert gravity.
- Truncating the antisymmetric part of the field yields a theory matching the exotic 3D gravity coupling found by Boulanger and Gualtieri.
- Integrating out the $B$ -field generates non-localities that can be removed by non-local field redefinitions up to quartic order.
Cyclic $L_\infty$ Structure: The full non-linear theory is encoded in a cyclic $L_\infty$ algebra. The author explicitly derives the BV master action, which includes quartic couplings in the ghost sector arising from the higher gauge algebra.
Inclusion of Moduli: The framework is extended to include complex structure moduli (harmonic $(2,1)$ -forms) on compact Calabi–Yau manifolds. The moduli enter the action as background parameters without kinetic terms, consistent with the Tian–Todorov lemma which implies they are unobstructed.

Significance
The paper claims that this formulation clarifies the nature of Kodaira–Spencer theory as a genuine theory of gravity, albeit topological. By providing a local action and a clear $L_\infty$ structure, it bridges the gap between topological string theory and gravity-like formulations.

The author highlights that the theory serves as a potential toy model for studying the UV completion of gravity. While the 3D reduction is non-renormalizable by power counting, the 6D origin in Kodaira–Spencer gravity may possess a well-defined quantization (referencing the perturbative quantization of BCOV by Costello and Li). The work suggests that this "gravity-like" formulation could shed light on the quantization of non-topological gravity and offers a new perspective on twisted holography.

The paper explicitly notes that the double copy interpretation and the specific $L_\infty$ morphisms were largely developed through interaction with artificial intelligence, but the author takes full responsibility for the mathematical consistency and the final presentation.

More like this