🔢 mathematics

Open-closed Deligne-Mumford field theories: construction

This paper constructs a unique, up to homotopy, open-closed Deligne-Mumford field theory associated with a relatively spin embedded Lagrangian, which extends the Fukaya $A_\infty$ algebra to higher genus and multiple boundary components as a foundational step toward proving Kontsevich's conjecture that the Fukaya category determines Gromov-Witten invariants.

Original authors: Amanda Hirschi, Kai Hugtenburg

Published 2026-05-06

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Amanda Hirschi, Kai Hugtenburg

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

The Big Picture: Mapping the Shape of Reality

Imagine you are trying to understand the shape of a complex, multi-dimensional universe (a symplectic manifold). Mathematicians have two main ways of looking at this universe:

The "Closed" View: Looking at the universe as a whole, like observing a planet from space. This involves counting specific types of loops or paths that exist inside the universe.
The "Open" View: Looking at the universe from the perspective of a specific surface floating inside it (a Lagrangian). This involves looking at paths that start and end on this surface, like a rubber band snapping back and forth.

For a long time, these two views were like two different languages that didn't quite translate into each other. This paper builds a massive, universal dictionary and a set of translation rules to connect them perfectly.

The Core Problem: The "Rubber Band" Problem

In the "Open" view, mathematicians study rubber bands (curves) that stretch across a surface.

The Ideal Scenario: If the rubber band is perfectly stable, it follows strict rules.
The Real Problem: In the messy reality of this universe, rubber bands can sometimes get unstable. They might shrink to a point, split into two, or bubble off into a tiny, unstable shape. When this happens, the mathematical rules break down. It's like trying to do calculus on a piece of paper that keeps tearing itself apart.

Previous attempts to fix this were like trying to patch a hole in a boat with duct tape. They worked for simple cases (flat rubber bands) but failed when the shapes got complicated (highly curved or multi-layered).

The Solution: A "Universal Construction Kit"

The authors, Amanda Hirschi and Kai Hugtenburg, have built a new, robust framework called an Open-Closed Deligne–Mumford Field Theory (DMFT).

Think of this framework as a Lego construction kit that can build any shape, no matter how complex.

The Bricks: Instead of standard Lego bricks, their "bricks" are mathematical descriptions of stable curves (shapes with holes and boundaries).
The Instructions: They created a set of instructions (an algebraic structure) that tells you how to snap these bricks together.
The Magic: This kit is designed so that even if a brick is unstable (a "bad" shape), the instructions automatically account for it. It's as if the kit has a self-correcting mechanism that says, "If this piece wobbles, here is exactly how to balance it with another piece so the whole structure stays standing."

Key Features of the Kit

1. The "Curved" vs. "Uncurved" Distinction
Sometimes, the universe is so curved that the rubber bands can't stay flat. The authors realized that to handle this, they first had to build a "Curved" version of their kit.

Analogy: Imagine trying to build a tower on a wobbly table. First, you build a "Curved" tower that leans with the table. Then, they show you how to add a specific "counter-weight" (called a bounding cochain) to straighten the tower out, turning it into a stable, "Uncurved" tower. This allows them to use the standard, clean rules of mathematics.

2. The "Chain-Level" Precision
Previous methods often looked at the "average" shape of these curves (like looking at a blurry photo). This paper works at the "chain level," which is like looking at every single pixel in a high-definition photo.

Why it matters: By looking at the pixels, they can see the tiny details that were previously hidden. This allows them to prove that the "Open" view (the rubber bands) contains all the information needed to reconstruct the "Closed" view (the whole universe).

3. The "Homotopy" Safety Net
In math, you often have to make arbitrary choices (like picking a specific angle to measure from). If your result changes based on that choice, the theory is broken.

The Fix: The authors proved that their construction is unique up to homotopy.
Analogy: Imagine you are drawing a map of a city. You can choose to draw the streets with a blue pen or a red pen, or you can draw them slightly wiggly or perfectly straight. As long as the connections between the buildings are correct, the map is valid. This paper proves that no matter which "pen" or "style" they use to build their theory, the final map of the universe is always the same.

The Main Achievement

The paper constructs a specific mathematical machine that takes a single surface (a Lagrangian) and generates a complete theory that describes:

How rubber bands behave on that surface.
How those rubber bands interact with the rest of the universe.
How to count the "loops" in the universe (Gromov–Witten invariants) purely by studying the rubber bands.

In simple terms: They have built a bridge. Before, you had to cross a river to get from the "Open" side to the "Closed" side, and the bridge was shaky. Now, they have built a sturdy, multi-lane highway that connects the two sides perfectly. This is the first major step in proving a famous conjecture (Kontsevich's Homological Mirror Symmetry) that says the "Open" and "Closed" views of the universe are actually two sides of the same coin.

Summary

This paper doesn't just fix a small math problem; it builds a new, universal language for describing the geometry of the universe. It solves the problem of "unstable shapes" by creating a flexible, self-correcting system that works for every possible curve, ensuring that the mathematical description of the universe remains consistent, no matter how you look at it.

Technical Summary: Open-Closed Deligne–Mumford Field Theories: Construction

Problem Statement
The paper addresses a fundamental gap in symplectic topology and algebraic geometry: the lack of a rigorous, chain-level algebraic framework that connects the Fukaya category of a Lagrangian submanifold to the Gromov–Witten invariants of the ambient symplectic manifold in arbitrary genus. While cohomological field theories (CohFTs) describe the algebraic structure of Gromov–Witten invariants, and Fukaya categories describe Lagrangian Floer theory, a direct construction of a field theory that operates at the chain level (rather than just on homology) and incorporates both open (Lagrangian boundary) and closed (interior) sectors has been elusive.

Previous attempts, such as Costello's open-closed topological conformal field theories (TCFTs), rely on modular properads of smooth framed surfaces. However, these do not directly utilize the compactified moduli spaces of stable curves with boundary that arise naturally in the study of pseudo-holomorphic curves. Furthermore, standard Lagrangian Floer theory often yields curved $A_\infty$ -algebras, and extending these to higher genus with arbitrary boundary components requires handling curvature and the specific geometric obstructions of moduli spaces with boundary.

Methodology
The authors construct "Open-Closed Deligne–Mumford Field Theories" (DMFTs), defined as symmetric monoidal $A_\infty$ -functors from a category of chains on moduli spaces of stable curves with boundary to the category of chain complexes. The construction proceeds through several technical stages:

Algebraic Framework: The authors define a category $\mathcal{OC}$ enriched over chains on moduli spaces of stable curves with boundary ( $M_{oc}$ ). To handle the geometric reality of Lagrangian boundary conditions, they introduce a "curved" version of this category, $\mathcal{OC}_r$ , which formally adjoins generators $D_0$ and $D_1$ to account for the instability of discs with few marked points. This allows the theory to accommodate curved $A_\infty$ -algebras.
Chain Models and Resolution of Singularities: A major technical hurdle is defining operations on chains where the moduli spaces are singular orbifolds with corners. The authors employ a specific chain model based on "analytic singular chains." They utilize the real analytic resolution of singularities algorithm (extending results from Bierstone and Milman) to handle fibre products of moduli spaces that are not transverse. This allows for well-defined pull-push operations (q-operations) even when the stabilization maps are not submersions.
Cubical Cobordisms and Kuranishi Structures: To achieve strict functoriality and handle higher homotopies, the construction relies on a system of "cubical cobordisms" (global Kuranishi charts) for moduli spaces of stable maps. These charts are constructed to be compatible with the boundary strata and the clutching operations. The authors define correspondences using these charts and Thom forms to define the operations of the field theory.
Deformation and Curvature Removal: The initial construction yields a curved open-closed DMFT. To obtain an uncurved theory (necessary for recovering standard Gromov–Witten invariants), the authors utilize the theory of bounding cochains. They show that if the Lagrangian admits a weak bounding cochain, the curved DMFT can be deformed into an uncurved one via an algebraic deformation process that eliminates the curvature term.
Homotopy Independence: The authors prove that the resulting field theory is independent of auxiliary choices (such as the almost complex structure and the specific Kuranishi charts) up to homotopy. This is achieved by constructing explicit homotopies using "pseudo-isotopies" of the underlying geometric data.

Key Contributions and Results

Definition of Open-Closed DMFT: The paper defines an open-closed Deligne–Mumford field theory as a symmetric monoidal $A_\infty$ -functor $F: \mathcal{OC} \to \text{Ch}_\Lambda$ . This structure extends the Fukaya $A_\infty$ -algebra to curves of arbitrarily high genus and with arbitrarily many boundary components.
Construction Theorem (Theorem A): For a closed symplectic manifold $(X, \omega)$ $(X, ω)$ and a relatively spin, embedded, weakly unobstructed Lagrangian $L$ $L$ , the authors construct a semi-strict, operadic curved open-closed DMFT.
- If $L$ is equipped with a weak bounding cochain, this deforms to an uncurved open-closed DMFT.
- The associated closed sector of this theory recovers the Gromov–Witten theory of $X$ .
Strictness and Operadic Properties: The construction is "semi-strict" (strict on stable chains) and "operadic" (compatible with forgetting marked points). This implies that the associated algebra on the Lagrangian is a cyclic $A_\infty$ -algebra.
Corollary 1.2: As a consequence, any compact symplectic manifold admits a strict chain-level action of the operad of stable genus-0 curves on the differential forms $\Omega^*(X; \Lambda)$ , lifting the quantum cup product.
Homotopy Independence (Theorem B): The constructed DMFTs are independent of all auxiliary choices (including the choice of almost complex structure and the gauge equivalence class of bounding cochains) up to homotopy of DMFTs.
Universal Curve-Counting: The theory is described as a "universal curve-counting theory" that contains the open-closed and closed-open maps as part of its data, bridging the gap between the Fukaya category and Gromov–Witten invariants.

Significance
The paper represents a critical first step toward proving Kontsevich's conjecture that the Fukaya category determines the Gromov–Witten invariants of a symplectic manifold. By following a strategy delineated by Costello, the authors provide the necessary geometric and algebraic machinery to lift the Fukaya category to a field theory that operates on the level of chains and moduli spaces of stable curves with boundary.

The construction removes the need for restrictive regularity assumptions (such as those required in previous works by Ganatra, Pomerleano, and Seidel) by utilizing virtual perturbations and global Kuranishi charts. The paper establishes that the algebraic structure of the Fukaya category, when enhanced by a bounding cochain, naturally encodes the enumerative invariants of the symplectic manifold. While the paper does not claim to fully prove the conjecture for all genera without further work (specifically regarding the generation of the Fukaya category and the trivialization of the $S^1$ -action), it successfully constructs the field-theoretic bridge required to make such a comparison. The work also generalizes previous results on genus-zero open Gromov–Witten invariants to arbitrary genus and boundary components.